ML23001A005
| ML23001A005 | |
| Person / Time | |
|---|---|
| Site: | 99902078, 05200050 |
| Issue date: | 12/31/2022 |
| From: | Fosaaen C NuScale |
| To: | Office of Nuclear Reactor Regulation, Document Control Desk |
| Shared Package | |
| ML23001A004 | List: |
| References | |
| LO-133414 | |
| Download: ML23001A005 (1) | |
Text
LO-133414 December 31, 2022 Docket No.52-050 U.S. Nuclear Regulatory Commission ATTN: Document Control Desk One White Flint North 11555 Rockville Pike Rockville, MD 20852-2738
SUBJECT:
NuScale Power, LLC Submittal, NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report, TR-121354, Revision 0
REFERENCES:
- 1. NuScale letter to NRC, NuScale Power, LLC Submittal of Planned Standard Design Approval Application Content, dated February 24, 2020 (ML20055E565)
- 2. NuScale letter to NRC, NuScale Power, LLC Requests the NRC staff to conduct a pre-application readiness assessment of the draft, NuScale Standard Design Approval Application (SDAA), dated May 25, 2022 (ML22145A460)
- 3. NRC letter to NuScale, Preapplication Readiness Assessment Report of the NuScale Power, LLC Standard Design Approval Draft Application, Office of Nuclear Reactor Regulation dated November 15, 2022 (ML22305A518)
- 4. NuScale letter to NRC, NuScale Power, LLC Staged Submittal of Planned Standard Design Approval Application, dated November 21, 2022 (ML22325A349)
NuScale Power, LLC (NuScale) is pleased to submit, NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report, TR-121354, Revision 0 (CVAP M&I). This report supports Chapter 3 of the Standard Design Approval Application, Design of Structures, Systems, Components and Equipment, Revision 0.
Chapter 3 supports Part 2, Final Safety Analysis Report, (FSAR) of the NuScale Standard Design Approval Application (SDAA), as described in Reference 1. NuScale submits the report in accordance with requirements of 10 CFR 52 Subpart E, Standard Design Approvals.
As described in Reference 4, the enclosure is part of a staged SDAA submittal.
NuScale requests NRC review, approval, and granting of standard design approval for the US460 standard plant design.
From July 25, 2022 to October 26, 2022, the NRC performed a pre-application readiness assessment of available portions of the draft NuScale FSAR to determine the FSARs readiness for submittal and for subsequent review by NRC staff (References 2 and 3). The CVAP M&I was not available for NRC readiness assessment review.
Enclosure 1 contains NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report, TR-121354-P, Revision 0, proprietary version. NuScale requests that the proprietary version (Enclosure 1) be withheld from public disclosure in accordance with the requirements of 10 CFR § 2.390. The enclosed affidavit (Enclosure 3) supports this request. Enclosure 1 has also been determined to contain Export NuScale Power, LLC 1100 NE Circle Blvd., Suite 200 Corvallis, Oregon 97330 Office 541.360-0500 Fax 541.207.3928 www.nuscalepower.com
LO-133414 Page 2 of 2 12/31/2022 Controlled Information. This information must be protected from disclosure per requirements of 10 CFR § 810. Enclosure 2 contains the nonproprietary version.
This letter makes no regulatory commitments and no revisions to any existing regulatory commitments.
If you have any questions, please contact Mark Shaver at 541-360-0630 or at mshaver@nuscalepower.com.
I declare under penalty of perjury that the foregoing is true and correct. Executed on December 31, 2022.
Sincerely, Carrie Fosaaen Senior Director, Regulatory Affairs NuScale Power, LLC Distribution: Brian Smith, NRC Michael Dudek, NRC Getachew Tesfaye, NRC Bruce Bavol, NRC David Drucker, NRC Enclosure 1: NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report, TR-121354-P, Revision 0 (proprietary)
Enclosure 2: NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report, TR-121354-NP, Revision 0 (nonproprietary)
Enclosure 3: Affidavit of Carrie Fosaaen AF-133415 NuScale Power, LLC 1100 NE Circle Blvd., Suite 200 Corvallis, Oregon 97330 Office 541.360-0500 Fax 541.207.3928 www.nuscalepower.com
LO-133414 : NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report, TR-121354-P, Revision 0 (proprietary)
NuScale Power, LLC 1100 NE Circle Blvd., Suite 200 Corvallis, Oregon 97330 Office 541.360-0500 Fax 541.207.3928 www.nuscalepower.com
LO-133414 : NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report, TR-121354-P, Revision 0 (nonproprietary)
NuScale Power, LLC 1100 NE Circle Blvd., Suite 200 Corvallis, Oregon 97330 Office 541.360-0500 Fax 541.207.3928 www.nuscalepower.com
NuScale Comprehensive Vibration Assessment rogram Measurement and Inspection Plan Technical Report ember 2022 ision 0 ket: 52-050 Scale Power, LLC NE Circle Blvd., Suite 200 allis, Oregon 97330
.nuscalepower.com pyright 2022 by NuScale Power, LLC pyright 2022 by NuScale Power, LLC i
COPYRIGHT NOTICE report has been prepared by NuScale Power, LLC and bears a NuScale Power, LLC, yright notice. No right to disclose, use, or copy any of the information in this report, other than he U.S. Nuclear Regulatory Commission (NRC), is authorized without the express, written mission of NuScale Power, LLC.
NRC is permitted to make the number of copies of the information contained in this report is necessary for its internal use in connection with generic and plant-specific reviews and rovals, as well as the issuance, denial, amendment, transfer, renewal, modification, pension, revocation, or violation of a license, permit, order, or regulation subject to the uirements of 10 CFR 2.390 regarding restrictions on public disclosure to the extent such rmation has been identified as proprietary by NuScale Power, LLC, copyright protection withstanding. Regarding nonproprietary versions of these reports, the NRC is permitted to ke the number of copies necessary for public viewing in appropriate docket files in public ument rooms in Washington, DC, and elsewhere as may be required by NRC regulations.
ies made by the NRC must include this copyright notice and contain the proprietary marking e original was identified as proprietary.
pyright 2022 by NuScale Power, LLC ii
Department of Energy Acknowledgement and Disclaimer material is based upon work supported by the Department of Energy under Award Number NE0008928.
report was prepared as an account of work sponsored by an agency of the United States ernment. Neither the United States Government nor any agency thereof, nor any of their ployees, makes any warranty, express or implied, or assumes any legal liability or ponsibility for the accuracy, completeness, or usefulness of any information, apparatus, duct, or process disclosed, or represents that its use would not infringe privately owned rights.
erence herein to any specific commercial product, process, or service by trade name, emark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, ommendation, or favoring by the United States Government or any agency thereof. The views opinions of authors expressed herein do not necessarily state or reflect those of the United tes Government or any agency thereof.
pyright 2022 by NuScale Power, LLC iii
Table of Contents tract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 cutive Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Purpose and Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Measurement Plan Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Benchmark Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 TF-1 and TF-2 Benchmark Testing for Turbulence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 3.1.1 TF-2 Modal Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 3.1.2 TF-1 and TF-2 Vibration Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 TF-2 Benchmark Testing for Fluid Elastic Instability . . . . . . . . . . . . . . . . . . . . . . . . . . 115 3.2.1 TF-2 Test Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116 3.2.2 TF-2 FEI Benchmarking Analysis Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128 TF-3 Build-out Modal Testing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148 3.3.1 Data Acquisition and Test Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150 3.3.2 Analytically Predicted Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152 3.3.3 Frequency Response Function Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152 3.3.4 Damping Estimation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168 3.3.5 Mode Shapes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175 3.3.6 Build-Out Modal Testing Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181 Validation Methodology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184 Use of Test Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184 Methodology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184 Validation Approach and Uncertainty Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186 4.3.1 Propagation of Uncertainties into a Result . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187 4.3.2 Calculation of Input Parameter Uncertainty . . . . . . . . . . . . . . . . . . . . . . . . . . . 188 4.3.3 Calculation of Mesh Numerical Uncertainty . . . . . . . . . . . . . . . . . . . . . . . . . . . 189 Evaluation Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190 Evaluation Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191 4.5.1 Turbulent Buffeting of Steam Generator Tube . . . . . . . . . . . . . . . . . . . . . . . . . 191 4.5.2 Vortex Shedding. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195 pyright 2022 by NuScale Power, LLC iv
Table of Contents 4.5.3 Fluid-Elastic Instability of SG Tubes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200 4.5.4 Acoustic Resonance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205 Experimental Bias Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208 4.6.1 Steam Generator Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208 Expected Results and Validation Range of Experimental Results. . . . . . . . . . . . . . . . 211 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212 Validation Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214 TF-3 Validation Test. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214 5.1.1 TF-3 Testing Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214 CNTS Main Steam Line Branch Connections Validation Testing . . . . . . . . . . . . . . . . 232 5.2.1 CNTS Main Steam Line Branch Connections Test Design . . . . . . . . . . . . . . . 232 Initial Startup Measurement Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241 Component Vibration Noise Estimation Methodology . . . . . . . . . . . . . . . . . . . . . . . . . 241 6.1.1 Tube Noise Estimation Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242 6.1.2 Background Turbulent Noise Fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248 Noise Estimation Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250 6.2.1 Sound Radiation due to Turbulent Buffeting of a Steam Generator Tube . . . . 254 6.2.2 Sound Radiation Due to Vortex Shedding of a SG Tube . . . . . . . . . . . . . . . . . 259 6.2.3 Sound Radiation Due to Fluid Elastic Instability of a Steam Generator Tube . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262 6.2.4 Sound Radiation Due to Unknown Reactor Vessel Internals Sources. . . . . . . 264 6.2.5 Sound Due to Component Impact . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268 Sensor Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269 Proposed Sensor Placement and Mounting. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270 Benchmarking to Confirm Detection Capabilities. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276 Results and Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276 Inspection Program . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278 Inspection Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278 Inspection Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279 Pre- and Post-Initial Startup Testing Inspection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279 Summary and Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283 pyright 2022 by NuScale Power, LLC v
Table of Contents Referenced Documents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283 pendix A TF-3 Instrumentation Plan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .A-1 pendix B TF-2 FEI Spectral Plots by Test Series (0 - 300 Hz). . . . . . . . . . . . . . . . . . . .B-1 pendix C TF-2 FEI Spectral Plots by Flow Rate (0 - 300 Hz) . . . . . . . . . . . . . . . . . . . . .C-1 pendix D TF-2 FEI Spectral Plots by Test Series (0 - 1000 Hz). . . . . . . . . . . . . . . . . . .D-1 pendix E TF-2 FEI Spectral Plots by Flow Rate (0 - 1000 Hz) . . . . . . . . . . . . . . . . . . . .E-1 pendix F TF-2 FEI Content within Frequency Ranges of Interest . . . . . . . . . . . . . . . . F-1 pendix G TF-2 FEI Frequency-Specific Amplitudes versus Flow Rate . . . . . . . . . . . G-1 pendix H TF-3 Build-out Testing Frequency Response Function Calculations . . . . .H-1 pyright 2022 by NuScale Power, LLC vi
List of Tables le 1-1 Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 le 1-2 Definitions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 le 2-1 Analysis Program Verification Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 le 3-1 Fluid Conditions for Each Test Case Analyzed . . . . . . . . . . . . . . . . . . . . . . . . . 12 le 3-2 Cross-Section Geometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 le 3-3 Adjusted Material Properties Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 le 3-4 Sensor Positions and Nearest Node Numbers. . . . . . . . . . . . . . . . . . . . . . . . . . 31 le 3-5 Full Bundle Modal Analysis Results for TF0001 Case . . . . . . . . . . . . . . . . . . . . 33 le 3-6 Full Bundle Nodal Analysis Results for TF0002 Case . . . . . . . . . . . . . . . . . . . . 34 le 3-7 Full Bundle Modal Analysis Results for TF0007 Case . . . . . . . . . . . . . . . . . . . . 35 le 3-8 Single Tube Modal Analysis Results for Sliding TF0001 Case . . . . . . . . . . . . . 40 le 3-9 Single Tube Modal Analysis Results for Sliding TF0002 Case . . . . . . . . . . . . . 41 le 3-10 Single Tube Modal Analysis Results for Sliding TF0007 Case . . . . . . . . . . . . . 42 le 3-11 Single Tube Modal Analysis Results Pinned TF0001 Case . . . . . . . . . . . . . . . . 43 le 3-12 Single Tube Modal Analysis Results Pinned TF0002 Case . . . . . . . . . . . . . . . . 44 le 3-13 Single Tube Modal Analysis Results Pinned TF0007 Case . . . . . . . . . . . . . . . . 45 le 3-14 X-Direction Modal Analysis Results for Mesh Sensitivity Study Using Single Tube Model, Sliding TF0001 Case. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 le 3-15 Y-Direction Modal Analysis Results for Mesh Sensitivity Study Using Single Tube Model, Sliding TF0001 Case. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 le 3-16 Z-Direction Modal Analysis Results for Mesh Sensitivity Study Using Single Tube Model, Sliding TF0001 Case. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 le 3-17 Major Mode Comparison Between Full Bundle and Single Tube Models . . . . . 56 le 3-18 Applicable TF-2 Test Cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 le 3-19 TF-1 Secondary Side PSDs Used . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 le 3-20 Giraudeau PSD Correlation Empirical Constants. . . . . . . . . . . . . . . . . . . . . . . . 72 le 3-21 Placement of Strain Gauge Instrumentation . . . . . . . . . . . . . . . . . . . . . . . . . . 121 le 3-22 Test conditions and File Names for Fluid Elastic Instability Data Acquisition . 122 le 3-23 Summary of Applied Processing Parameters. . . . . . . . . . . . . . . . . . . . . . . . . . 128 le 3-24 Dynamic Root Mean Square Strains Measured During TF-2 Fluid Elastic Instability Tests. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 le 3-25 Dynamic Peak Strains Measured During TF-2 Fluid Elastic Instability Tests. . 134 le 3-26 Dynamic Peak Strains by Test Series (Average and Maximum) . . . . . . . . . . . 135 le 3-27 Dynamic Peak Strains by Flow Rate (Average and Maximum) . . . . . . . . . . . . 136 pyright 2022 by NuScale Power, LLC vi
List of Tables le 3-28 Relative Comparison of Content within Frequency Ranges of Interest (0-10 Hz Inclusive), Sorted by Test Series. . . . . . . . . . . . . . . . . . . . . . . . . . . . 140 le 3-29 Relative Comparison of Content within Frequency Ranges of Interest (Excludes 0-10 Hz), Sorted by Test Series . . . . . . . . . . . . . . . . . . . . . . . . . . . 141 le 3-30 Relative Comparison of Content within Frequency Ranges of Interest (excludes 0-10 Hz), Sorted by Test Series. . . . . . . . . . . . . . . . . . . . . . . . . . . . 142 le 3-31 Acquisition Parameters for Time and Frequency Domain . . . . . . . . . . . . . . . . 150 le 3-32 Group B, Span C (5) Accelerometers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156 le 3-33 Group C, Multi-Span. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159 le 3-34 Energy Transfer across Spans. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161 le 3-35 End Span (A and AD) Frequency Response Function Summary . . . . . . . . . . 163 le 3-36 Support and Tube Frequency Response Function Summary . . . . . . . . . . . . . 167 le 3-37 Frequency Response Function Summary Per Group . . . . . . . . . . . . . . . . . . . 182 le 4-1 SG Test Model Inputs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192 le 4-2 SG Tube Inputs to TB Margin Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194 le 4-3 SG Tube TB Margin Uncertainty Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194 le 4-4 Steam Generator Test Model Inputs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196 le 4-5 Steam Generator Tube Inputs to Vortex Shedding Margin Calculation:
Method B . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197 le 4-6 Steam Generator Tube Vortex Shedding Margin Uncertainty Method:
Method B . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198 le 4-7 Steam Generator Tube Inputs to Vortex Shedding Margin Calculation:
Method A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199 le 4-8 Steam Generator Tube Vortex Shedding Margin Uncertainty Method:
Method A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199 le 4-9 Fluid Elastic Instability Analysis Input Parameter Types . . . . . . . . . . . . . . . . . 201 le 4-10 Steam Generator Tube Inputs to Fluid Elastic Instability Safety Margin Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203 le 4-11 Steam Generator Tube Fluid Elastic Instability Margin Uncertainty Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203 le 4-12 Decay Heat Removal System Steam Pipe Inputs to Acoustic Resonance Margin Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207 le 4-13 Decay Heat Removal System Steam Pipe Acoustic Resonance Margin Uncertainty Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208 le 4-14 Summary of Components and Flow Induced Vibration Analysis Validation Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213 pyright 2022 by NuScale Power, LLC vii
List of Tables le 5-1 Summary of Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214 le 5-2 Tube Array Design Geometry. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216 le 5-3 Helical Steam Generator Tube Array Details . . . . . . . . . . . . . . . . . . . . . . . . . . 217 le 5-4 TF-3 Test Facility Operating Conditions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217 le 5-5 Instrumented Tubes (Accelerometers and Strain Gauges) . . . . . . . . . . . . . . . 220 le 5-6 Instrumented Tubes (Accelerometers Only). . . . . . . . . . . . . . . . . . . . . . . . . . . 220 le 5-7 Instrumented Tubes (Strain Gauges Only). . . . . . . . . . . . . . . . . . . . . . . . . . . . 220 le 5-8 Accelerometers Placement for Instrumented Tube #1 . . . . . . . . . . . . . . . . . . . 221 le 5-9 Accelerometers Placement for Instrumented Tube #2 . . . . . . . . . . . . . . . . . . . 221 le 5-10 Accelerometers Placement for Instrumented Tube #3 . . . . . . . . . . . . . . . . . . . 222 le 5-11 Accelerometers Placement for Instrumented Tube #4 . . . . . . . . . . . . . . . . . . . 222 le 5-12 Accelerometers Placement for Instrumented Tube #5 . . . . . . . . . . . . . . . . . . . 222 le 5-13 Accelerometers Placement for Instrumented Tube #6 . . . . . . . . . . . . . . . . . . . 223 le 5-14 Tube Strain Gauge Placements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223 le 5-15 Steam Generator Tube Support Accelerometer. . . . . . . . . . . . . . . . . . . . . . . . 224 le 5-16 Other Pressure and Temperature Instrumentation. . . . . . . . . . . . . . . . . . . . . . 225 le 5-17 Steady-State Flow-induced Vibration Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . 229 le 5-18 Vortex Shedding Test Range1, 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230 le 5-19 Fluid Elastic Instability Flow Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231 le 5-20 Measurements and Sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234 le 6-1 Increase in SPL due to n Identical Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . 254 le 6-2 Noise at RPV Wall Due to Radial SG Tube Motion Caused by Turbulent Buffeting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256 le 6-3 Noise at RPV Wall Due to Axial SG Tube Motion Caused by Turbulent Buffeting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257 le 6-4 Tube Motion Due to Vortex Shedding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261 le 6-5 Noise at RPV Wall Due to Tube Motion Caused by Vortex Shedding . . . . . . . 261 le 6-6 Noise at RPV Wall Due to Tube Motion Caused by FEI . . . . . . . . . . . . . . . . . 263 le 6-7 Noise at RPV at Riser Exit Due to RVI Vibration . . . . . . . . . . . . . . . . . . . . . . . 267 le 6-8 Noise in Downcomer Due to RVI Vibration . . . . . . . . . . . . . . . . . . . . . . . . . . . 268 le 6-9 Candidate Dynamic Pressure Sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269 le 6-10 Proposed Dynamic Pressure Sensor Locations. . . . . . . . . . . . . . . . . . . . . . . . 271 le 7-1 Pre- and Post-Initial Startup Testing Inspection Locations. . . . . . . . . . . . . . . . 280 le F-1 Amplitude of Spectral Content between 0 - 10 Hz . . . . . . . . . . . . . . . . . . . . . . . F-2 pyright 2022 by NuScale Power, LLC viii
List of Tables le F-2 Amplitude of Spectral Content between 0-10 Hz (Excluding Test Series G) . . . F-3 le F-3 Amplitude of Spectral Content between 16-28 Hz . . . . . . . . . . . . . . . . . . . . . . . F-4 le F-4 Amplitude of Spectral Content between 35-55 Hz . . . . . . . . . . . . . . . . . . . . . . . F-5 le F-5 Amplitude of Spectral Content between 70-85 Hz . . . . . . . . . . . . . . . . . . . . . . . F-6 le F-6 Amplitude of Spectral Content between 140-160 Hz . . . . . . . . . . . . . . . . . . . . . F-7 le F-7 Amplitude of Spectral Content between 10-300 Hz . . . . . . . . . . . . . . . . . . . . . . F-8 le H-1 Testing Matrix. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .H-2 pyright 2022 by NuScale Power, LLC ix
List of Figures ure 3-1 TF-2 Test Specimen Assembly Drawing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 ure 3-2 TF-2 Full Bundle Geometry and Mesh . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 ure 3-3 TF-2 Full Bundle Geometry and Mesh Close-up . . . . . . . . . . . . . . . . . . . . . . . . 12 ure 3-4 Naming System for Various Tube Support Sections . . . . . . . . . . . . . . . . . . . . . 13 ure 3-5 Neglected Steam and Feedwater Piping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 ure 3-6 Neglected Shift of Header Pposition and Change in Tube Inclination Angle . . . 16 ure 3-7 Neglected Longitudinal Ribs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 ure 3-8 Neglected Internal Structures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 ure 3-9 Separate Plates That Are Modeled As Single Monolithic Plate . . . . . . . . . . . . . 18 ure 3-10 Coordinate Systems, Global Cartesian and Global Cylindrical Shown from Global Isometric View. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 ure 3-11 Radial Coupling between Barrels and Tube Supports . . . . . . . . . . . . . . . . . . . . 21 ure 3-12 Diagram of Screw Interface Between Barrels and Tube Supports. . . . . . . . . . . 22 ure 3-13 Tube to Tube Support Coupling (Column 5 Tubes Shown) . . . . . . . . . . . . . . . . 23 ure 3-14 Tube Supports to Thick Slab Coupling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 ure 3-15 Header to Header Attachment Plate Coupling . . . . . . . . . . . . . . . . . . . . . . . . . . 25 ure 3-16 Displacement Constraint on Radial Cantilever Plate . . . . . . . . . . . . . . . . . . . . . 26 ure 3-17 Primary Fluid Regions for Hydrodynamic Effects on Barrels. Front View on Left, Top View on Right . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 ure 3-18 Internal Barrel and External Barrel Remote Points for Hydrodynamic Coupling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 ure 3-19 External Barrel and Vessel Nodes for Hydrodynamic Coupling . . . . . . . . . . . . . 30 ure 3-20 Boundary Conditions for Single Tube Model . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 ure 3-21 Fundamental Mode for TF0001 Case (Full Assembly Rocking Mode). . . . . . . . 36 ure 3-22 Mode 3 for TF0001 Case (Tube Bundle Twisting Mode) . . . . . . . . . . . . . . . . . . 37 ure 3-23 Highest X-Participating Mode for TF0001 Case (Tube Bundle Shifting along X-Axis) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 ure 3-24 Highest Y-Participating Mode for TF0001 Case (Tube Beam Mode). . . . . . . . . 39 ure 3-25 Fundamental Mode for Single Tube Sliding TF0001 Case (Breathing Mode) . . 46 ure 3-26 Highest X-Participating Mode for Single Tube Sliding TF0001 Case (Tube Sliding through Supports with Beam Mode) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 ure 3-27 Highest Y-Participating Mode for Single Tube Sliding TF0001 cCase (Beam Mode) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 ure 3-28 Fundamental Mode and Highest Y-Participating Mode for Single Tube Pinned TF0001 Case (Beam Mode). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 pyright 2022 by NuScale Power, LLC ix
List of Figures ure 3-29 Highest X-Participating Mode for Single Tube Pinned TF0001 Case (High Order Beam Mode). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 ure 3-30 Mesh Size Comparison for Mesh Sensitivity Analysis . . . . . . . . . . . . . . . . . . . . 52 ure 3-31 Breathing/Twisting Mode Comparison between Single Tube and Full Bundle Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 ure 3-32 Highest Y-Participating Mode Comparison between Single Tube and Full Bundle Models (Sliding/Shifting and Beam Mode) . . . . . . . . . . . . . . . . . . . . . . . 58 ure 3-33 Highest X-Participating Mode Comparison between Single Tube and Full Bundle Models (Beam Mode). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 ure 3-34 Block Diagram of the Vibration Analysis Methodology. . . . . . . . . . . . . . . . . . . . 60 ure 3-35 Block Diagram of the ANSYS Solution Sequence . . . . . . . . . . . . . . . . . . . . . . . 61 ure 3-36 Strain Versus Temperature Ranges for the TD, TF, and TW Tests . . . . . . . . . . 76 ure 3-37 Strain Gauge Data for a Case with Secondary Side Boiling . . . . . . . . . . . . . . . 77 ure 3-38 TF0004_0769 Column 5 Lower Strain Sensor with Pinned Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 ure 3-39 TF0004_0769 Column 5 Upper Strain Sensor with Pinned Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80 ure 3-40 TF0004_0773 Column 5 Lower Strain Sensor with Pinned Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81 ure 3-41 TF0004_0773 Column 5 Upper Strain Sensor with Pinned Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 ure 3-42 TF0004_0773 Column 5 Lower Strain Sensor with Sliding Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 ure 3-43 TF0004_0773 Column 5 Upper Strain Sensor with Sliding Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84 ure 3-44 TF0003_0762 Column 5 Lower Strain Sensor with Pinned Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 ure 3-45 TF0003_0762 Column 5 Upper Strain Sensor with Pinned Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86 ure 3-46 TF0007_0777 Column 1 Lower Strain Sensor with Pinned Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87 ure 3-47 TF0007_0777 Column 3 Lower Strain Sensor with Pinned Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 ure 3-48 TF0007_0777 Column 3 Upper Strain Sensor with Pinned Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 ure 3-49 TF0007_0777 Column 5 Lower Strain Sensor with Pinned Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 pyright 2022 by NuScale Power, LLC x
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 List of Figures Figure 3-50 TF0007_0777 Column 5 Upper Strain Sensor with Pinned Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 Figure 3-51 TF0007_0781 Column 5 Lower Strain Sensor with Pinned Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 Figure 3-52 TF0007_0781 Column 5 Upper Strain Sensor with Pinned Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 Figure 3-53 TF0007_0781 Column 5 Lower Strain Sensor with Pinned Boundary Conditions and the TF-1 PS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 Figure 3-54 TF0007_0781 Column 5 Upper Strain Sensor with Pinned Boundary Conditions and the TF-1 PSD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 Figure 3-55 TF0007_0781 Column 5 Unfiltered Strain Sensor up to 600 Hz . . . . . . . . . . . . 97 Figure 3-56 TF0007_0781 Column 5 Lower Strain Sensor with Pinned Boundary Conditions and the Au-Yang/Jordan PSD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 Figure 3-57 TF0007_0781 Column 5 Lower Strain Sensor with Pinned Boundary Conditions and More Mode Combinations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 Figure 3-58 TF0007_0781 Column 5 Lower Strain Sensor with Pinned Boundary Conditions and Fine Mesh . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 Figure 3-59 TF0007_0781 Column 5 Lower Strain Sensor with Pinned Boundary Conditions and ((2(a),(c),ECI Damping . . . . . . . . . . . . . . . . . . . . . . . . . 101 Figure 3-60 Test Case Geometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 Figure 3-61 Displacement Spectrum Comparison for Test Case 1 . . . . . . . . . . . . . . . . . . . 105 Figure 3-62 Displacement Spectrum Comparison for Test Case 2 . . . . . . . . . . . . . . . . . . . 106 Figure 3-63 Displacement Spectrum Comparison for Test Case 3 . . . . . . . . . . . . . . . . . . . 107 Figure 3-64 Displacement Spectrum Comparison for Test Case 4 . . . . . . . . . . . . . . . . . . . 108 Figure 3-65 Bending Strain Spectrum Comparison for Test Case 1 . . . . . . . . . . . . . . . . . . 109 Figure 3-66 Bending Strain Spectrum Comparison for Test Case 2 . . . . . . . . . . . . . . . . . . 110 Figure 3-67 Bending Strain Spectrum Comparison for Test Case 3 . . . . . . . . . . . . . . . . . . 111 Figure 3-68 Bending Strain Spectrum Comparison for Test Case 4 . . . . . . . . . . . . . . . . . . 112 Figure 3-69 Tube Axisymmetric Model Mesh . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114 Figure 3-70 Tube Axisymmetric Model Mesh . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115 Figure 3-71 TF-2 Fluid-Heated Test Section Tubing Column Scheme . . . . . . . . . . . . . . . . 116 Figure 3-72 TF-2 Tube Support Detail. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117 Figure 3-73 Placement of Strain Gauges on Tube Coils (Typical) . . . . . . . . . . . . . . . . . . . 119 Figure 3-74 Placement of Strain Gauges on Tube Coils (S1101-1 and S1101-2 Shown). . 120 Figure 3-75 Example of Average Variations in Strain Gauge Signals . . . . . . . . . . . . . . . . . 127 © Copyright 2022 by NuScale Power, LLC xi
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 List of Figures Figure 3-76 Overall Power Spectral Density Content Comparison - Column-3, Side Strain Gauges (Top=S07, Bottom=S05) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132 Figure 3-77 Scatter Plot of Peak Dynamic Strain Versus Flow Rate. . . . . . . . . . . . . . . . . . 143 Figure 3-78 Spectral Comparison, 0-50 Hz, Datasets D5 Versus G5, Column-3 Sensors . 144 Figure 3-79 Spectral Comparison, 0-50 Hz, Datasets D6 Versus G6, Column-5 Sensors . 145 Figure 3-80 General Layout of Helical Coil Steam Generator Prototype for Vibration Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150 Figure 3-81 Typical Single, Long Span Vertical; Frequency Response Function (Blue) and Coherence (Orange) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154 Figure 3-82 Frequency Response Function Response for Roving Accelerometer Along Single Span (Mass Loading) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154 Figure 3-83 Lower Frequency Response (Below 25Hz); Frequency Response Function (Blue) and Coherence (Orange). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155 Figure 3-84 Coherence Only for Test 1C-1Z (Horizontal/1Z and Vertical/3Y Direction) . . . 157 Figure 3-85 Impact Location Effects on Span C (Left: Mid-Span; Right: Near End of Span) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157 Figure 3-86 Coherence Only for Test 2A-EfwZ (Vertical and Horizontal Direction). . . . . . . 160 Figure 3-87 A-EfwZ Energy Transfer across Spans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162 Figure 3-88 Coherence Only 5A-5Z (Horizontal and Vertical Direction) . . . . . . . . . . . . . . . 164 Figure 3-89 Support/Tube Testing Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166 Figure 3-90 Coherence for Test Sup-103 + Col-12, Span E Tub 1+5+9, Impact 1-X . . . . 168 Figure 3-91 Global Polynomial Curve Fit of Frequency Response Functions (Top Frequency Response Function Magnitude, Bottom Imaginary Magnitude on Log-Scale) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169 Figure 3-92 Half-Power Damping Estimations over the Frequency Response Function Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170 Figure 3-93 Peak Damping Values ((( }}2(a),(c),ECI:1C-1Y-Left and 2(a),(c),ECI:1A-3Z-Right). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170 (( }} Figure 3-94 Amplitude-Dependent Damping from Three Time Histories (Raw): Exponential Fit - Blue; Log Dec - Red . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172 Figure 3-95 Amplitude-Dependent Damping from Three Time Histories (Averaged): Exponential Fit - Blue; Log Dec - Red . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173 Figure 3-96 Amplitude-Dependent Damping From Pull Test: Exponential Fit - Blue; Log Dec - Red . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174 Figure 3-97 Amplitude-Dependent Damping: Pull Pest - Black and Grey; Impact Tests - Red and Blue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175 © Copyright 2022 by NuScale Power, LLC xii
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 List of Figures Figure 3-98 Span C, 5 Accelerometers (Frequency Response Function, Axial-Red, Vertical-Green, Horizontal-Blue) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177 Figure 3-99 Mode Shape for 1C-1Z at (( }}2(a),(c),ECI . . . . . . . . . . . . . . . . . . . . . . . 177 Figure 3-100 Mode Shape for 1C-1Z at (( }}2(a),(c),ECI . . . . . . . . . . . . . . . . . . . . . . . . 178 Figure 3-101 Mode Shape for 1C-1Z at (( }}2(a),(c),ECI . . . . . . . . . . . . . . . . . . . . . . . 178 Figure 3-102 Mode Shape for 1C-1Z at (( }}2(a),(c),ECI . . . . . . . . . . . . . . . . . . . . . . . 179 Figure 3-103 Comparison of Mode Shapes between 1C-1Z ((( }}2(a),(c),ECI) and 1A-5Z ((( }} 2(a),(c),ECI ) Using Modal Assurance Criteria . . . . . . . . . . . . . . . 180 Figure 6-1 Cylinder Vibration Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245 Figure 6-2 Sound Radiation Pattern for Radially-Vibration Tube. . . . . . . . . . . . . . . . . . . . 246 Figure 6-3 Sound Radiation Pattern for Axially-Vibrating Tube . . . . . . . . . . . . . . . . . . . . . 247 Figure 6-4 Estimated Spectrum of Noise Radiated by a Vibrating Tube . . . . . . . . . . . . . . 248 Figure 6-5 Noise at RPV Wall Due to Radial SG Tube Motion Caused by Turbulent Buffeting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256 Figure 6-6 Noise at RPV Wall Due to Axial SG Tube Motion Caused by Turbulent Buffeting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258 Figure 6-7 Noise at RPV Wall Due to Radial SG Tube Motion Caused by Turbulent Buffeting at Design Analysis Predicted Amplitudes . . . . . . . . . . . . . . . . . . . . . 259 Figure 6-8 Noise at RPV Wall Due to Tube Motion Caused by Vortex Shedding . . . . . . . 262 Figure 6-9 Noise at RPV Wall Due to Tube Motion Caused by FEI . . . . . . . . . . . . . . . . . 264 Figure 6-10 Noise at RPV Due to ICIGT Motion Caused by Turbulent Flow in Riser . . . . . 266 Figure 6-11 Noise at RPV Due to Component Motion (Assuming ICIGT Frequencies) Caused by Turbulent Flow in Downcomer . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267 Figure 6-12 Proposed Location of Dynamic Pressure Sensors - Option 1 . . . . . . . . . . . . . 273 Figure 6-13 Proposed Location of Dynamic Pressure Sensors - Option 2 . . . . . . . . . . . . . 274 Figure 6-14 Proposed Location of Dynamic Pressure Sensors - Option 3 . . . . . . . . . . . . . 275 Figure A-1 Instrumentation Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .A-2 Figure B-1 Test Condition A, Channel Set 1 - Side Sensors . . . . . . . . . . . . . . . . . . . . . . . .B-2 Figure B-2 Test Condition A, Channel Set 2 - Top Sensors . . . . . . . . . . . . . . . . . . . . . . . .B-3 Figure B-3 Test Condition B, Channel Set 1 - Side Sensors . . . . . . . . . . . . . . . . . . . . . . . .B-4 Figure B-4 Test Condition B, Channel Set 2 - Top Sensors . . . . . . . . . . . . . . . . . . . . . . . .B-5 Figure B-5 Test Condition C, Channel Set 1 - Side Sensors . . . . . . . . . . . . . . . . . . . . . . . .B-6 Figure B-6 Test Condition C, Channel Set 2 - Top Sensors . . . . . . . . . . . . . . . . . . . . . . . .B-7 Figure B-7 Test Condition D, Channel Set 1 - Side Sensors . . . . . . . . . . . . . . . . . . . . . . . .B-8 Figure B-8 Test Condition D, Channel Set 2 - Top Sensors . . . . . . . . . . . . . . . . . . . . . . . .B-9 © Copyright 2022 by NuScale Power, LLC xiii
List of Figures ure B-9 Test Condition G, Channel Set 1 - Side Sensors . . . . . . . . . . . . . . . . . . . . . . .B-10 ure B-10 Test Condition G, Channel Set 2 - Top Sensors . . . . . . . . . . . . . . . . . . . . . . .B-11 ure C-1 Nominal Primary-Side Flow Rate of 114 kg/s, Channel Set 1 - Side Sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .C-2 ure C-2 Nominal Primary-Side Flow Rate of 114 kg/s, Channel Set 2 - Top Sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .C-3 ure C-3 Nominal Primary-Side Flow Rate of 143 kg/s, Channel Set 1 - Side Sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .C-4 ure C-4 Nominal Primary-Side Flow Rate of 143 kg/s, Channel Set 2 - Top Sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .C-5 ure C-5 Nominal Primary-Side Flow Rate of 173 kg/s, Channel Set 1 - Side Sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .C-6 ure C-6 Nominal Primary-Side Flow Rate of 173 kg/s, Channel Set 2 - Top Sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .C-7 ure C-7 Nominal Primary-Side Flow Rate of 201 kg/s, Channel Set 1 - Side Sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .C-8 ure C-8 Nominal Primary-Side Flow Rate of 201 kg/s, Channel Set 2 - Top Sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .C-9 ure C-9 Nominal Primary-Side Flow Rate of 230 kg/s, Channel Set 1 - Side Sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .C-10 ure C-10 Nominal Primary-Side Flow Rate of 230 kg/s, Channel Set 2 - Top Sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .C-11 ure C-11 Nominal Primary-Side Flow Rate of 263 kg/s, Channel Set 1 - Side Sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .C-12 ure C-12 Nominal Primary-Side Flow Rate of 263 kg/s, Channel Set 2 - Top Sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .C-13 ure D-1 Test Condition A, Channel Set 1 - Side Sensors . . . . . . . . . . . . . . . . . . . . . . . .D-2 ure D-2 Test Condition A, Channel Set 2 - Top Sensors . . . . . . . . . . . . . . . . . . . . . . . .D-3 ure D-3 Test Condition B, Channel Set 1 - Side Sensors . . . . . . . . . . . . . . . . . . . . . . . .D-4 ure D-4 Test Condition B, Channel Set 2 - Top Sensors . . . . . . . . . . . . . . . . . . . . . . . .D-5 ure D-5 Test Condition C, Channel Set 1 - Side Sensors . . . . . . . . . . . . . . . . . . . . . . . .D-6 ure D-6 Test Condition C, Channel Set 2 - Top Sensors . . . . . . . . . . . . . . . . . . . . . . . .D-7 ure D-7 Test Condition D, Channel Set 1 - Side Sensors . . . . . . . . . . . . . . . . . . . . . . . .D-8 ure D-8 Test condition D, Channel Set 2 - Top Sensors . . . . . . . . . . . . . . . . . . . . . . . . .D-9 ure D-9 Test Condition G, Channel Set 1 - Side Sensors . . . . . . . . . . . . . . . . . . . . . . .D-10 ure D-10 Test Condition G, Channel Set 2 - Top Sensors . . . . . . . . . . . . . . . . . . . . . . .D-11 pyright 2022 by NuScale Power, LLC xiv
List of Figures ure E-1 Nominal Primary-Side Flow Rate of 114 kg/s, Channel Set 1 - Side Sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .E-2 ure E-2 Nominal Primary-Side Flow Rate of 114 kg/s, Channel Set 2 - Top Sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .E-3 ure E-3 Nominal Primary-Side Flow Rate of 143 kg/s, Channel Set 1 - Side Sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .E-4 ure E-4 Nominal Primary-Side Flow Rate of 143 kg/s, Channel Set 2 - Top Sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .E-5 ure E-5 Nominal Primary-Side Flow Rate of 173 kg/s, Channel Set 1 - Side Sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .E-6 ure E-6 Nominal Primary-Side Flow Rate of 173 kg/s, Channel Set 2 - Top Sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .E-7 ure E-7 Nominal Primary-Side Flow Rate of 201 kg/s, Channel Set 1 - Side Sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .E-8 ure E-8 Nominal Primary-Side Flow Rate of 201 kg/s, Channel Set 2 - Top Sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .E-9 ure E-9 Nominal Primary-Side Flow Rate of 230 kg/s, Channel Set 1 - Side Sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .E-10 ure E-10 Nominal Primary-Side Flow Rate of 230 kg/s, Channel Set 2 - Top Sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .E-11 ure E-11 Nominal Primary-Side Flow Rate of 260 kg/s, Channel Set 1 - Side Sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .E-12 ure E-12 Nominal Primary-Side Flow Rate of 260 kg/s, Channel Set 2 - Top Sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .E-13 ure G-1 Dynamic Strain, 0-10 Hz Versus Flow Rate . . . . . . . . . . . . . . . . . . . . . . . . . . . G-2 ure G-2 Dynamic Strain, 16-28 Hz Versus Flow Rate . . . . . . . . . . . . . . . . . . . . . . . . . . G-3 ure G-3 Dynamic Strain, 35-55 Hz Versus Flow Rate . . . . . . . . . . . . . . . . . . . . . . . . . . G-4 ure G-4 Dynamic Strain, 70-85 Hz Versus Flow Rate . . . . . . . . . . . . . . . . . . . . . . . . . . G-5 ure G-5 Dynamic Strain, 140-160 Hz Versus Flow Rate . . . . . . . . . . . . . . . . . . . . . . . . G-6 ure G-6 Dynamic Strain, 10-300 Hz Versus Flow Rate . . . . . . . . . . . . . . . . . . . . . . . . . G-7 ure H-1 315-12-1-C-5 (5y). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .H-3 ure H-2 zSgle Tube 1sec. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .H-4 ure H-3 1A-1Y . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .H-5 ure H-4 1A-1Z . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .H-6 ure H-5 1A-3Z . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .H-7 ure H-6 1A-5Z . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .H-8 pyright 2022 by NuScale Power, LLC xv
List of Figures ure H-7 1C-1Y . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .H-9 ure H-8 1C-1Z . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .H-10 ure H-9 1C-5Z . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .H-11 ure H-10 1E-1Y . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .H-12 ure H-11 1E-1Z . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .H-13 ure H-12 1E-5Z . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .H-14 ure H-13 2A-CmsZ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .H-15 ure H-14 2A-EfwZ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .H-16 ure H-15 2A-EmsZ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .H-17 ure H-16 2A-GfwZ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .H-18 ure H-17 2C-CmsZ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .H-19 ure H-18 2C-EfwZ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .H-20 ure H-19 2C-EmsZ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .H-21 ure H-20 2C-GfwZ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .H-22 ure H-21 3C-EmsZ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .H-23 ure H-22 4A-2Z . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .H-24 ure H-23 4A-3Z . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .H-25 ure H-24 5A-2Z . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .H-26 ure H-25 5A-4Y . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .H-27 ure H-26 5A-5Z . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .H-28 ure H-27 5A-Support Z . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .H-29 ure H-28 Sup-103+Col-12 Span E, Tube 1, 5, 9 Impact 1x . . . . . . . . . . . . . . . . . . . . . .H-30 ure H-29 Sup-103+Col-12 Span E, Tube 1, 5, 9 Impact 1Z . . . . . . . . . . . . . . . . . . . . . .H-31 ure H-30 Supp-to-Tube 103 Span-E FW-X . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .H-32 ure H-31 Supp-to-Tube 103 Span-E FW-X on Plate . . . . . . . . . . . . . . . . . . . . . . . . . . . .H-33 pyright 2022 by NuScale Power, LLC xvi
report describes the Comprehensive Vibration Assessment Program (CVAP) for the cale Power Module (NPM) that verifies the structural integrity of the internals for
-induced vibration (FIV). The CVAP conforms to the guidance of Regulatory de 1.20, Revision 4. The content of this licensing technical report provides additional rmation to substantiate the statements made in the NuScale Power Plant US460 standard ign, thereby facilitating a comprehensive review by the NRC of the NPM-20 design.
CVAP design analysis program is documented in the NuScale Comprehensive Vibration essment Program Analysis Technical Report, TR-121353-P (Reference 9.1.4). The results of analysis program serve as the basis for the choice of components and areas to be monitored e measurement and inspection programs, in order to validate the CVAP design analysis gram. This technical report provides the details associated with the CVAP measurement and ection program plans. pyright 2022 by NuScale Power, LLC 1
omprehensive Vibration Assessment Program (CVAP) for the NuScale Power Module (NPM) stablished in accordance with Regulatory Guide 1.20, Revision 4. The CVAP ensures that the ponents of the NPM exposed to fluid flow do not experience detrimental effects of
-induced vibration (FIV).
NPM-20 represents a first-of-a-kind design in its size, arrangement, and operating ditions. Accordingly, the first operational NPM-20 is classified as a prototype in accordance Regulatory Guide 1.20. The terms NPM-20 and NPM are used interchangeably throughout report. After the first NPM-20 is qualified as a valid prototype, subsequent NPMs classified as non-prototype. en its prototype classification, the CVAP addresses the applicable criteria of Regulatory de 1.20, Section C.2. NPM is a small modular reactor that is an integral, movable nuclear steam supply system. NPM design is passive, with primary coolant driven by natural circulation flow. Low natural ulation flow velocities decrease the propensity for detrimental FIV effects. The CVAP blishes the scope of analyses, testing, measurements, and inspections required to ensure components of the NPM are not subject to unacceptable vibratory degradation. When pleted, the CVAP provides the requisite assurance that the NPM components are not subject etrimental effects of FIV. report provides the details of the CVAP measurement and inspection programs. These grams consist of benchmark testing and analysis, validation testing and analysis, an rumentation plan, and inspection plan. The primary goal of the testing is to validate that imental effects do not occur during limiting operating conditions. owing the completion of each test, post-test analyses are performed to complete the dation effort. Assessments are performed based on the initial startup testing and inspection ervations. Combined with the benchmarking efforts, the measurement and inspection work pe validates the FIV screening and design analyses in the NuScale Comprehensive Vibration essment Program Analysis Technical Report, TR-121353-P (Reference 9.1.4). pyright 2022 by NuScale Power, LLC 2
Purpose and Scope This report describes the Comprehensive Vibration Assessment Program (CVAP) for the NuScale Power Module (NPM)-20 to verify the structural integrity of the components to flow-induced vibration (FIV). The terms NPM-20 and NPM are used interchangeably throughout this report. The CVAP conforms to the guidance of Regulatory Guide 1.20, Revision 4 and demonstrates a sufficient margin of safety and structural integrity through: design vibration and stress analysis testing to quantitatively validate analytical methods and results inspection prior to and following initial startup testing to qualitatively validate analytical methods and results The measurement and inspection work scope validates the FIV screening and design analyses in Reference 9.1.4. The measurement and inspection program explains the basis for the components assessed, the details associated with the planned testing and inspections, and how the results are used to validate the CVAP design analysis program and to confirm the adequacy of the design analysis, including the predicted safety margins. The measurement program consists of two components. The first is benchmarking the design analysis using test data that are not fully prototypic, but are applicable to susceptible components and aspects of the design analysis and overall validation approach. The second component of the measurement program is prototypic validation testing. This testing is informed from the design analysis and for components with the lowest predicted safety margins. Validation testing is performed either at specially designed test facilities, to permit higher quality and quantity collection of test data, and the ability to operate above licensing basis limits where the onset of strongly-coupled flow induced vibration (FIV) phenomena is predicted to occur, or on the first NPM during initial startup testing. The inspection program includes components that meet the screening criteria for any FIV mechanism. Details associated with the extent of inspections and inspection acceptance criteria are provided in this report. To finalize the CVAP, one additional technical report is developed. Upon completion of validation testing and inspection, this final report provides the post-test evaluation and inspection program results. In this final report, the differences between the expected and measured experimental results are either resolved or confirmed to be in the analytically predicted allowable ranges. pyright 2022 by NuScale Power, LLC 3
Table 1-1 Abbreviations m Definition acoustic resonance ME American Society of Mechanical Engineers TS containment system D control rod drive AP Comprehensive Vibration Assessment Program S data acquisition system Average RS decay heat removal system fluid elastic instability fast Fourier transform flow-induced vibration frequency response function feedwater SG helical coil steam generator T in-core instrument guide tubes Latin hypercube sampling C modal acceptance crtieria main steam V main steam isolation valve S main steam system M NuScale Power Module outer diameter Operations and Maintenance power spectral density S reactor coolant system S root mean square V reactor pressure vessel T Societ Informazioni Esperienze Termoidrauliche steam generator sound pressure level turbulent buffeting AR test equipment error and accuracy report vortex shedding visual test Table 1-2 Definitions m Definition ustic resonance A phenomenon where an acoustic wave is generated at a frequency that coincides with the natural frequency of a confining structure. In design analysis, bias is the difference between a best estimate and conservative parameter. In an experiment, bias is another term for a test distortion, i.e., a feature of the test that is different from the design analysis condition. fidence Interval The probability that the true value lies within the specified limits. pyright 2022 by NuScale Power, LLC 4
m Definition cal instrument An instrument whose proper function is required in order to accomplish the objectives of the test campaign. anded Uncertainty An estimate of the plus-or-minus limits of total error, with a defined level of confidence (usually 95%). gue Usage Factor Ratio of the number of vibration cycles anticipated during the lifetime of the component to the allowable cycles. d boundary condition For the fixed boundary condition, bonded contacts are used to simulate the situation that tubes are fixed to support channels. This boundary condition simulates high frictional forces between the tubes and NPM tube supports. or Pitch Velocity (Vgap) Local velocity to which tubes are subject, for flow in a closely pack tube array. This velocity is developed based on the overall flow area blocked by tubes (and supports as applicable). Calculated based on ASME N-1331.1 guidance of the approach flow velocity multiplied by the ratio of the tube pitch divided by the pitch minus the diameter. eral visual Method- This level of inspection is made with direct, assisted, or remote visual methods. A mirror may be used to enhance visual access to exposed surfaces in the inspection area. This level of inspection is made under normally available lighting conditions such as hangar lighting, flashlight, or drop-light, and may require removal or opening of access panels. Stands, ladders or platforms may be required to gain proximity to the area being checked. Criteria- A visual examination of an interior or exterior area, installation, or assembly to detect general mechanical and structural condition of components and their supports, and to detect surface discontinuities and imperfections. General mechanical and structural condition of components is verified by parameters such as clearances, settings, and physical displacements. Abnormally positioned components, such as misalignment of supports or a pipe outside a pipe hanger, are noted for closer examination. The inspection includes an observation of the condition of the material surfaces, including welds, within the inspection area to detect surface discontinuities and imperfections, such as a loss of integrity at bolted or welded connections, loose or missing parts, debris, cracks, corrosion, erosion, discoloration, and geometric discontinuities, such as gouges, chips, or dents. g span Based on arrangement of the SG tube supports in the NuScale design, lengths of helical tubing that span beneath each steam plenum are 64 degree arcs. These lengths of tubing are generically referred to as long spans. er riser section Reactor internal components from the upper core plate to the upper riser section. te Carlo Monte Carlo simulations provide statistical results to problems by performing repeated calculations with randomized input variables, and analyzing the trends in the output data. -Test Analysis The effort to model the test apparatus and assess input, numerical, and measurement uncertainties to inform the expected range of experimental results suitable for validation. pyright 2022 by NuScale Power, LLC 5
m Definition pagation of Uncertainty A test, for example, may report measured values for density ( ) and velocity ( V ), in addition to their respective uncertainties. A validation 1 2 analysis calculates dynamic pressure ( --- V ), so its uncertainty must be 2 estimated by propagating the uncertainty of density and velocity. ector block The reflector sits inside the core barrel and is made of sections referred to as reflector blocks. sitivity coefficient The instantaneous rate of change in the result due to a change in a parameter. rt span Based on arrangement of the SG tube supports in the design, lengths of helical tubing that span between each steam plenum are only 26-degree arcs. These lengths of tubing are generically referred to as short spans. plenum There are eight SG plenums (four steam and four feedwater (FW). The plenums contain the SG tube sheets. ing boundary condition For the sliding boundary condition, no separation contacts are used to simulate the situation that tubes can slide in the support channels. This boundary condition models no frictional forces between the tube and the NPM tube support. ndard uncertainty For a dispersion of values about a mean value, the standard uncertainty is the estimated standard deviation. uhal number A dimensionless frequency associated with vortex shedding (VS). 1 Test facility designed to study the effects of secondary side boiling in HCSG tubes. Dynamic pressure measurements collected during flow testing. 2 Test facility designed to study primary and secondary flows in HCSG tubes, and heat transfer. Strain gauge measurements were collected during flow testing. 3 Test facility designed to study fluid elastic instability, VS, and turbulence due to primary side flow in HCSG tubes. Testing consists of modal testing in air and in water, and primary side flow testing with extensive instrumentation to detect vibration. ertainty Relating to the presence of an unknown error in a measured quantity or a model calculation. er riser section Reactor vessel internals (RVI) components from the lower riser section to the top of the riser. dation The process of determining the degree to which a model is an accurate representation of the real world from the perspective of the intended uses of the model. pyright 2022 by NuScale Power, LLC 6
This report describes the scope of the CVAP measurement and inspection programs. The following sections provide an overview of the components that require testing to validate the design analysis per Reference 9.1.4. The methodology used to account for uncertainty and bias to demonstrate the test design meets validation objectives is described. Additionally, existing benchmark testing is used to supplement the measurement program validation of the design analysis. Measurement Plan Overview The analysis program includes a list of FIV phenomena and a list of components that could be subjected to these phenomena as described in Reference 9.1.4. The FIV analysis results demonstrate that many components have large margins of safety. The margin of safety is the means by which structural integrity against FIV degradation is assured. Therefore, when a margin of safety is sufficiently large, validation by testing is not necessary. As part of completing the CVAP, experimental testing is conducted to verify the vibration analysis conducted for the NPM. Analysis that shows less than 100 percent safety margin is judged to require verification in the measurement program. If a component screens for an FIV mechanism, but analysis is not performed, validation testing is required. Validation testing of novel design changes to preclude FIV, such as flow disrupters, also undergo confirmatory validation testing. To validate the FIV inputs, analytical results, and the margins of safety determined in the analysis program, a combination of separate effects and initial startup testing are performed. Separate effects testing is performed on a prototypic portion of the steam generator design. Initial startup testing is performed under full-power normal operating conditions. The testing results are used to validate the prototype NPM design. Table 2-1 summarizes the testing to be performed to verify the FIV analysis program for the prototype NPM. The results of the measurement program are used to validate FIV analysis inputs, results, and the margins of safety. pyright 2022 by NuScale Power, LLC 7
Table 2-1 Analysis Program Verification Testing Mechanisms with Prototype Testing Susceptible Component less than 100% Mechanisms Test Facility Initial Startup Safety Margin TS main steam CNTS steam piping branch AR - - testing nections helical tubing FEI, VS, TB FEI, VS TF-3 testing - Where, AR = acoustic resonance, CNTS = containment system, SG= steam generator, TB = turbulent buffering FEI = fluid elastic instability, and VS = vortex shedding pyright 2022 by NuScale Power, LLC 8
In the CVAP, benchmark testing is used to justify important aspects of the design analysis. Benchmarking supplements literature and industry accepted methods to provide confidence that the measurements taken during the validation program do not produce unanticipated results requiring modifications to the design analysis methods. TF-1 and TF-2 Benchmark Testing for Turbulence 1 TF-2 Modal Analysis This section analyzes the modal response of the TF-2 steam generator (SG) test specimen. The TF-2 test specimen is a full-scale representation of columns 1 through 5 of the NPM SG tubes. This calculation is based on the Societ Informazioni Esperienze Termoidrauliche (SIET) Helical Coil Steam Generator Test Program - Fluid Heated Test Facility Design. This section applies only to the TF-2 test specimen design and not the NPM design. A full bundle model as well as single tube models are used in the analysis. Both models are run for three test conditions, each with unique fluid temperatures and pressures that represent the test conditions. The single tube model is run for two sets of boundary conditions. The results generated in this section are: Modal frequencies, mass participations, and mode shapes for the full bundle and single tube models for significant modes. Mode shape text files for three tubes of interest for the full bundle and single tube models for modes (up to 160 Hz for full bundle and up to 600 Hz for single tube). A mesh sensitivity analysis to validate the mesh size used in the models. 1.1 Model Overview The TF-2 global assembly drawing is shown in Figure 3-1, with major components labeled. The ANSYS model explicitly models these major components except the vessel, which is modeled as rigid nodes. The TF-2 model is shown in Figure 3-2 and Figure 3-3. The model details are described in the following sections. pyright 2022 by NuScale Power, LLC 9
Figure 3-1 TF-2 Test Specimen Assembly Drawing
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 10
Figure 3-2 TF-2 Full Bundle Geometry and Mesh
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 11
Figure 3-3 TF-2 Full Bundle Geometry and Mesh Close-up
}}2(a),(c),ECI 1.2 Test Cases The model was run for three test cases. Each test case is for a different set of primary and secondary fluid conditions. These temperatures and pressures were measured directly from the sensors in the test specimen during the tests and are summarized in Table 3-1.
Table 3-1 Fluid Conditions for Each Test Case Analyzed Primary Side Secondary Side Case Notes Pavg (psi) Tavg (°F) Pavg (psi) Tavg (°F) (( F0001 drained secondary side F0002 filled secondary side (liquid) F0007 boiling secondary side (liquid and steam)
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 12
The geometry of the TF-2 full bundle model includes: Tube coil geometry Header geometry Tube support geometry Barrels and plates geometry Cross-section geometry Multiple cross-sections are used for the beam elements in the model. The geometry values and source references are listed in Table 3-2. The tube support regions that have cutouts use a representative rectangular cross-section that has equivalent bending stiffness as the detailed geometry. A diagram explaining the naming system for the tube support sections is shown in Figure 3-4. Figure 3-4 Naming System for Various Tube Support Sections
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 13
Table 3-2 Cross-Section Geometry Design Model Name Dimensions (inches) Dimensions (mm) (( der e_end e_transition e_middle ow_end ow_middle
}}2(a),(c),ECI Details of the TF-2 assembly were simplified to reduce model and mesh complexity. These modifications have negligible impact on mass, stiffness, and thus the overall results of the analysis. Minor simplifications include neglecting small features such as fillets, fasteners, and instrumentation. The larger simplifications are listed below.
Neglected steam and feedwater piping: Steam and feedwater piping extend from the headers and exit the vessel (Figure 3-5). The numerous pipe bends do not contribute to the stiffness of the assembly. pyright 2022 by NuScale Power, LLC 14
Figure 3-5 Neglected Steam and Feedwater Piping
}}2(a),(c),ECI Neglected shifted headers: In order to avoid interferences between the tubes and headers, the headers are shifted axially (Figure 3-6). This shift causes the first and last wrap of a given tube column to have an increased inclination angle. This design aspect has negligible effect on the overall behavior of the bundle. The sensors on the instrumented tubes are also located in the main helix of the bundle, not on these first and last wraps, meaning that slight changes to the end condition of the tubes do not affect the measured results.
Therefore, a constant nominal inclination angle is used for the tubes throughout the bundle. pyright 2022 by NuScale Power, LLC 15
igure 3-6 Neglected Shift of Header Pposition and Change in Tube Inclination Angle
}}2(a),(c),ECI Neglected longitudinal ribs: There are six longitudinal ribs that run along the bottom 1/3rd of the external barrel (indicated in Figure 3-7. These ribs are not included in the model.
Figure 3-7 Neglected Longitudinal Ribs
}}2(a),(c),ECI Neglected internal structures: Inside the bundle are a series of pipes and conduit as well as a ladder (Figure 3-8). These structures were neglected.
pyright 2022 by NuScale Power, LLC 16
Figure 3-8 Neglected Internal Structures
}}2(a),(c),ECI Merging of upper plate and lower plate: At the bottom of the tube bundle is a pair of plates called upper plate and lower plate that bolt together (Figure 3-9). These plates were merged into a single plate in the model.
pyright 2022 by NuScale Power, LLC 17
Figure 3-9 Separate Plates That Are Modeled As Single Monolithic Plate
}}2(a),(c),ECI During fabrication, the radial and azimuthal positions of the headers were re-positioned slightly to avoid interferences with the tubes. The radial positioning of the headers is a negligible change, but the azimuthal position change is incorporated into the model, as this affects the lengths of the end segments of the tubes.
1.4 Mesh The tubes were meshed with BEAM189 elements, which include midside nodes. The element length is set to 8.3 inches, which gives (( }}2(a),(c),ECI between two sets of supports that are 90 degrees apart. There is an additional element between the two immediately adjacent supports spaced 2 inches from each other. The tube mesh is shown in Figure 3-2 and Figure 3-3. The tube supports are meshed with BEAM188 elements that do not include midside nodes. Midside nodes are not necessary due to the small element size of less than 1 in. The tube support mesh is shown in Figure 3-2 and Figure 3-3. pyright 2022 by NuScale Power, LLC 18
midside nodes. This mesh is shown in Figure 3-2 and Figure 3-3. Primary fluid mass is modeled with MASS21 elements and FLUID38 elements. Remote points, which are used to couple the FLUID38 elements to the barrels, use TARGE170 and CONTA175 elements. 1.5 Materials Relevant components in the TF-2 test assembly are made of Type 304 stainless steel. The elastic modulus of barrels, headers, and tube supports is taken at the primary fluid temperature. The elastic modulus of the tubes is taken at the average of the primary and secondary fluid temperatures. Densities of the materials for the tubes and tube supports are adjusted to account for the hydrodynamic mass (displaced primary fluid mass) as well as the contained secondary fluid mass in the tubes. For the TF0007 case, which has boiling inside the tubes, the steam region of the tubes is assigned a different density than the liquid region. The material property values are summarized in Table 3-3. Table 3-3 Adjusted Material Properties Summary Barrels and Headers Tube Supports Tubes Case E E E (106 psi) (lbm/in3) (106 psi) (lbm/in3) (106 psi) (lbm/in3) (( TF0001 TF0002 TF0007
}}2(a),(c),ECI 1.6 Coordinate Systems and Circumferential Numbering Three different coordinate systems types were used in this evaluation. They are described below and in Figure 3-10. Note that the coordinate systems are located at global zero, but are shown at different locations in the figure for clarity. The pyright 2022 by NuScale Power, LLC 19
apply to the individual tube models as well. Global Cartesian coordinate system: X and Z are horizontal directions, and Y-points vertically upward. Global cylindrical coordinate system: X is the radial direction, Y is the circumferential direction, and Z points vertically upward. Local coordinate systems: Twenty unique local Cartesian coordinate systems are used for coupling tubes to tube supports. There are four coordinate systems for each column of tubes, one at each tube support group. From the perspective of the interface, the local X is along the tubes axis and Z points radially inward toward the center of the bundle. The local Y axis is perpendicular to these two, and is angled away from the global vertical direction by the tubes inclination angle (approximately 14°). igure 3-10 Coordinate Systems, Global Cartesian and Global Cylindrical Shown from Global Isometric View
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 20
This section explains boundary conditions and constraint equations used in the TF-2 full bundle model to couple the components. Nodes along the height of the external barrel are coupled to the outermost tube supports. Nodes along the height of the internal barrel are coupled to the innermost tube supports. The nodes are coupled in the radial direction only (UX) of global cylindrical coordinate system. A visualization of the constraint equation coupling is shown in Figure 3-11. Figure 3-11 Radial Coupling between Barrels and Tube Supports
}}2(a),(c),ECI This connection represents the interface between a series of screws and studs that penetrate the barrels and contact the tube supports to hold them in place. The studs are welded to the tube supports and pass through a large clearance hole in pyright 2022 by NuScale Power, LLC 21
supports to be pulled radially by the stud. The large clearance hole does not provide vertical or circumferential restraint on the stud. The screws are located beneath each stud and pass through a nut that is welded to the barrels. However, the screw only pushes on the flat part of the tube support and thus does not provide vertical or circumferential support. A diagram of this is shown in Figure 3-12. Figure 3-12 Diagram of Screw Interface Between Barrels and Tube Supports
}}2(a),(c),ECI The tube nodes are coupled to the tab nodes of the tube supports at each interface. The local UX, UY, and UZ of the local coordinate systems are coupled.
A visualization of the constraint equation coupling is shown in Figure 3-13. pyright 2022 by NuScale Power, LLC 22
Figure 3-13 Tube to Tube Support Coupling (Column 5 Tubes Shown)
}}2(a),(c),ECI This connection represents the tubes sitting in the cutouts of the tube supports.
The tube axial direction (local UX) is coupled to represent the high friction at the interface (no sliding). Since there are two supports closely spaced, it only takes a slight amount of axial misalignment of the supports to create high friction forces at the tube support. The fact that most supports carry tubes on both sides of the support also contributes to the high amount of interlocking in the assembly. The nodes at the ends of the tube supports are coupled to corresponding nodes on the upper and lower thick slabs. The six degrees of freedom are coupled using the global coordinate system. A visualization of the constraint equation coupling is shown in Figure 3-14. Note that the figure shows the upper region coupling, although the bottom region is identical. pyright 2022 by NuScale Power, LLC 23
Figure 3-14 Tube Supports to Thick Slab Coupling
}}2(a),(c),ECI Nodes on the header attachment plates are coupled to the nearest nodes on the headers. The six degrees of freedom are coupled using the global coordinate system. A visualization of the constraint equation coupling is shown in Figure 3-15.
pyright 2022 by NuScale Power, LLC 24
Figure 3-15 Header to Header Attachment Plate Coupling
}}2(a),(c),ECI The entire assembly is supported with a displacement boundary condition at the edge of the radial cantilever plate that extends outward from the external barrel, as shown in Figure 3-16. Displacements are constrained, but no rotations are constrained. This configuration simulates the cantilever plate resting on the ledge inside the test vessel.
pyright 2022 by NuScale Power, LLC 25
Figure 3-16 Displacement Constraint on Radial Cantilever Plate
}}2(a),(c),ECI The following connections are made through conformal meshes and therefore no constraint equations are required.
Header attachment plates to interior barrel Exterior barrel and interior barrel to thick slabs Exterior barrel and interior barrel to bottom plate 1.8 Primary Fluid Hydrodynamic Effect on Barrels The hydrodynamic effect on the tubes and tube supports is incorporated by adding the displaced primary fluid mass back onto those components. However, the effect on the barrels is more complicated. There are three fluid regions that impact the barrels: the fluid cylinder inside the internal barrel, the fluid annulus between the internal and external barrels (fluid annulus 1), and the fluid annulus between the external barrel and the vessel (fluid annulus 2). These regions are shown in Figure 3-17. The hydrodynamic effect on these regions occurs in the horizontal directions only, as the vertical direction is not contained, and primary fluid is free to flow. pyright 2022 by NuScale Power, LLC 26
gure 3-17 Primary Fluid Regions for Hydrodynamic Effects on Barrels. Front View on Left, Top View on Right
}}2(a),(c),ECI The contained fluid inside the internal barrel (fluid cylinder) is modeled by calculating the contained fluid mass and scoping it to the walls of the internal barrel in the horizontal directions only.
To ensure the mass is distributed evenly, the mass applied to each node is weighted by the area apportioned to the given node relative to the total apportioned area of all nodes. The mass is applied via MASS21 elements using the X and Z direction real constants only. The hydrodynamic coupling between the interior and exterior barrels is accomplished using FLUID38 elements. These elements have two nodes, each node representing the centerline of the concentric cylinders. The elements require three geometric inputs, inner radius, outer radius, and cylinder height. A material with the density of the fluid is also applied to the element. For the TF-2 ANSYS model, the barrels are sectioned along their heights into seven similarly sized regions. Each region of each barrel contains a remote point that is scoped to a circumferential edge at the center of the region. The pair of remote point pilot nodes for a given region serve as the two nodes for the FLUID38 element. For example, region 1 of the internal barrel has its pilot node connected to the pilot node of region 1 of the external barrel via the FLUID38 pyright 2022 by NuScale Power, LLC 27
Figure 3-18. Note that the remote point scoping skips the nodes on the barrels that are part of the barrels to tube supports coupling to avoid potential over-constraint. ure 3-18 Internal Barrel and External Barrel Remote Points for Hydrodynamic Coupling
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 28
the actual amount of fluid is less than the nominal volume of the annulus. To correct for this difference, the volume of the tubes and tube supports is subtracted from the nominal annulus volume, and new effective diameters are determined. Note that the volumes of the headers and thick slabs were deemed negligible. The hydrodynamic fluid coupling between the external barrel and the vessel is also modeled using FLUID38 elements. The thick vessel is considered rigid compared to the rest of the TF-2 test assembly and is not explicitly modeled. Instead, seven nodes are created at the same locations as the seven external barrel remote points. These seven vessel nodes have all degrees of freedom set to zero to simulate the vessel rigidity. The vessel nodes are connected to the external barrel pilot nodes though an additional set of FLUID38 elements. A visualization of the remote points, vessel nodes, and FLUID38 coupling is shown in Figure 3-19. pyright 2022 by NuScale Power, LLC 29
Figure 3-19 External Barrel and Vessel Nodes for Hydrodynamic Coupling
}}2(a),(c),ECI Because fluid annulus 2 does not contain large obstructions in the annulus, the nominal dimensions for the annular region are used.
pyright 2022 by NuScale Power, LLC 30
There are three tubes instrumented with strain gages. To aid in downstream calculations, the nodes on the tubes nearest to the strain gages are determined. A summary of the sensor positions used to determine the nearest node numbers are presented in Table 3-4. The node numbers of the nearest nodes are included in the table. Table 3-4 Sensor Positions and Nearest Node Numbers Sensor Elev Elev Radial Pos Radial Pos Azimuth ol Tube Node Num Name (m) (in) (mm) (in) (°) (( S1101 86192 1 20 S1102 78712 S3101 19559 3 21 S3102 19501 S5101 99520 5 11 S5102 106811
}}2(a),(c),ECI 1.10 Single Tube Model A single tube model is created from the full bundle model. This single tube model contains the three individual instrumented tubes listed in Table 3-4. Other components are suppressed. The model is shown in Figure 3-20.
Note that since the single tube model does not contain the barrels or vessel, the fluid-structure interaction and hydrodynamic mass on these components are not present to interact with the tubes. The purpose of this configuration is to benchmark the applicability of the single tube modal analysis results when trying to emulate the behavior of the full bundle. The tube supports are retained in the model, but for visualization purposes only, specifically when viewing results. The tube support nodes are fixed, and the bodies are assigned a zero density material to prevent them from affecting the model. Figure 3-25 is an example. The single tube model has boundary conditions applied directly to nodes on the tubes to simulate the interaction with the headers and tube supports. The header connection has degrees of freedom fixed to simulate the weld. The tube support connections are run for two cases: sliding and pinned. The sliding case has the local UY and UZ constrained. The pinned case has the local UX, UY, and UZ constrained. These boundary conditions are shown in Figure 3-20. Other model details are identical to the full bundle model. pyright 2022 by NuScale Power, LLC 31
Figure 3-20 Boundary Conditions for Single Tube Model
}}2(a),(c),ECI 1.11 Modal Analysis - Full Bundle Model Modal analysis up to 160 Hz is performed for the TF-2 full bundle model for the three fluid condition cases described in Section 3.1.1.2. The top 20 participating modes for each case are presented in Table 3-5 through Table 3-7. The highest participating mode in each direction is highlighted in yellow.
Visualizations of major modes are shown in Figure 3-21 through Figure 3-24. These modes are shown for the TF0001 case only for brevity. The other cases have similar modes but with slightly shifted frequencies. pyright 2022 by NuScale Power, LLC 32
Table 3-5 Full Bundle Modal Analysis Results for TF0001 Case
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 33
Table 3-6 Full Bundle Nodal Analysis Results for TF0002 Case
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 34
Table 3-7 Full Bundle Modal Analysis Results for TF0007 Case
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 35
Figure 3-21 Fundamental Mode for TF0001 Case (Full Assembly Rocking Mode)
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 36
Figure 3-22 Mode 3 for TF0001 Case (Tube Bundle Twisting Mode)
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 37
gure 3-23 Highest X-Participating Mode for TF0001 Case (Tube Bundle Shifting along X-Axis)
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 38
Figure 3-24 Highest Y-Participating Mode for TF0001 Case (Tube Beam Mode)
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 39
Modal analysis was performed for the TF-2 single tube model for the three fluid condition cases described in Section 3.1.1.2 and the two boundary condition cases described in Section 3.1.1.10. The top 20 participating modes up to 600 Hz for each case are presented in Table 3-8 through Table 3-13. The highest participating mode in each direction is highlighted in yellow. Visualizations of major modes are shown in Figure 3-25 through Figure 3-29. Only TF0001 case modes are shown. The other cases have similar mode shapes but with slightly shifted frequencies. Also, since the tubes do not interact with each other, only one tube participates for a given mode, and only the participating tube is shown in the figures. For all instances, the column 5 tube is shown since it has the largest mass and lowest frequencies. Table 3-8 Single Tube Modal Analysis Results for Sliding TF0001 Case
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 40
Table 3-9 Single Tube Modal Analysis Results for Sliding TF0002 Case
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 41
Table 3-10 Single Tube Modal Analysis Results for Sliding TF0007 Case
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 42
Table 3-11 Single Tube Modal Analysis Results Pinned TF0001 Case
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 43
Table 3-12 Single Tube Modal Analysis Results Pinned TF0002 Case
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 44
Table 3-13 Single Tube Modal Analysis Results Pinned TF0007 Case
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 45
igure 3-25 Fundamental Mode for Single Tube Sliding TF0001 Case (Breathing Mode)
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 46
igure 3-26 Highest X-Participating Mode for Single Tube Sliding TF0001 Case (Tube Sliding through Supports with Beam Mode)
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 47
gure 3-27 Highest Y-Participating Mode for Single Tube Sliding TF0001 cCase (Beam Mode)
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 48
ure 3-28 Fundamental Mode and Highest Y-Participating Mode for Single Tube Pinned TF0001 Case (Beam Mode)
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 49
Figure 3-29 Highest X-Participating Mode for Single Tube Pinned TF0001 Case (High Order Beam Mode)
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 50
A mesh sensitivity analysis is performed for the single tube model using the sliding TF0001 case. A fine mesh of 2.0 inches is compared to the nominal mesh size of 8.3 inches, which is shown in Figure 3-30. Both meshes use BEAM189 elements with midside nodes. The mesh sensitivity results for the top 20 participating modes in each direction are presented in Table 3-14 through Table 3-16. Agreement of frequency and mass participation is shown between the two mesh sizes, which validates the use of the nominal mesh. For modes below 160 Hz (the upper bound in the full bundle model), the percent error in frequency is less than (( }}2(a),(c),ECI, and the percent error in effective mass ratio is less than (( }}2(a),(c),ECI (most modes under (( }}2(a),(c),ECI). The percent error increases with higher frequencies, as the higher order modes are increasingly difficult to represent with the coarser mesh, although the coarse mesh is still acceptable. pyright 2022 by NuScale Power, LLC 51
Figure 3-30 Mesh Size Comparison for Mesh Sensitivity Analysis
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 52
able 3-14 X-Direction Modal Analysis Results for Mesh Sensitivity Study Using Single Tube Model, Sliding TF0001 Case
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 53
able 3-15 Y-Direction Modal Analysis Results for Mesh Sensitivity Study Using Single Tube Model, Sliding TF0001 Case
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 54
able 3-16 Z-Direction Modal Analysis Results for Mesh Sensitivity Study Using Single Tube Model, Sliding TF0001 Case
}}2(a),(c),ECI 1.14 Modal Analysis Results Discussion This discussion compares the full bundle modal results to the single tube results for the TF0001 case. The purpose of this comparison is to determine the applicability of the single tube results when trying to emulate the behavior of the full bundle. Table 3-17 summarizes the major mode comparison between the models. Figure 3-31 through Figure 3-33 compare the mode shapes of similar modes. In these figures, bodies are hidden except for the instrumented tube in column 5, in order to help identify similar modes between models.
pyright 2022 by NuScale Power, LLC 55
Table 3-17 Major Mode Comparison Between Full Bundle and Single Tube Models
}}2(a),(c),ECI The single tube models do not appropriately characterize the behavior of the full bundle model, as they cannot adequately account for the flexibility of the tube supports. Although the tubes are pinned to the tube supports in the full bundle model, the sliding case of the single tube model does a better job at emulating the full bundle behavior, as the sliding action of the tube somewhat simulates the flexibility in the tube supports. The pinned case of the single tube model cannot capture this behavior. For the major horizontal shifting/sliding mode, the single tube sliding model is about (( }}2(a),(c),ECI higher in frequency than the full bundle model.
The single tube pinned case can characterize the pure tube beam modes as well as the sliding case, as the flexibility of the tube supports is less important in these modes. The major vertical mode is (( }}2(a),(c),ECI higher in frequency in the single tube model (both cases) than the full bundle model. The twisting/breathing mode of the full bundle model is somewhat captured by the single tube sliding model. The single tube model breathing mode is (( }}2(a),(c),ECI higher in frequency than the lowest twisting mode of the full bundle model, but this full bundle mode has low mass participation, and is not in the top 20 participating modes for any of the directions. However, this breathing mode is only (( }}2(a),(c),ECI lower in frequency than another twisting mode of the full bundle model with a similar mode shape. This full bundle mode has higher mass participation, although it is in the horizontal direction, whereas the single tube mode has the mass participation in the vertical direction. pyright 2022 by NuScale Power, LLC 56
gure 3-31 Breathing/Twisting Mode Comparison between Single Tube and Full Bundle Models
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 57
igure 3-32 Highest Y-Participating Mode Comparison between Single Tube and Full Bundle Models (Sliding/Shifting and Beam Mode)
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 58
igure 3-33 Highest X-Participating Mode Comparison between Single Tube and Full Bundle Models (Beam Mode)
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 59
This section includes analysis of the turbulent buffering (TB) vibrations in the instrumented TF-2 tubes during the test set investigating fluid elastic instability. For the TF-2 tests, the secondary side is either quiescent fluid, single-phase moving liquid, or moving liquid boiling to vapor. For the cases of quiescent fluid, no secondary side PSD is applied as there is no turbulence to cause pressure fluctuations. For the cases with moving fluid, a secondary side PSD is applied based on either established literature or the measurements from the TF-1 test data. The analytical results are compared to the TF-2 strain gauge data to provide further basis for the validity of the analytical approach to TB vibrations. The methodology for determining the structural response spectra due to turbulent buffeting is based on the acceptance integral methodology. Figure 3-34 is an overview of the solution sequence. First, ANSYS is used to run a modal analysis on the TF2 structural model (Figure 3-35). This information provides the required structural properties. The loading on the structure is quantified by pressure power spectral density (PSDs). Separate PSDs are used for the primary and secondary sides. The spatial distribution of the PSDs are characterized by a coherence function. With the mode information, pressure PSDs, and coherence functions, the acceptance integrals are calculated. The acceptance integrals represent the contribution to the response from different mode combinations. Figure 3-34 Block Diagram of the Vibration Analysis Methodology pyright 2022 by NuScale Power, LLC 60
Figure 3-35 Block Diagram of the ANSYS Solution Sequence 2.1 Turbulent Buffeting Response The response spectra due to turbulent buffeting (Equation 3-1) is calculated using the acceptance integral methodology. Equation 3-1 is simplified to be a single summation instead of split into two summations. The acceptance integral, Equation 3-2, is related to how the energy added to a particular mode by the pressure PSDs is transferred between modes. Equation 3-3 is the modal transfer function which describes how vibrations are transferred between modes. Using the default option in ANSYS, the mode shapes are normalized based on the mass matrix. This normalization causes the generalized mass in Equation 3-3 to be unity. S y ( x , ) = AS p H ( )H ( ) J ( ) Equation 3-1 1 S p ( x' ,x , ) J ( ) = --- A -------------------------- S p ( x' , )
- dx'dx Equation 3-2 AA 1
H ( ) = ------------------------------------------------------------ Equation 3-3 2 2 m - + i pyright 2022 by NuScale Power, LLC 61
S y ( x , ) = Displacement response spectra (in2/Hz) x = Location on structure
= Frequency (rad/s)
A = Surface area (in2) or = Mode index or = Displacement mode shape (in/in) H or H = Modal transfer function shape (in/lbf) J ( ) = Acceptance integral (in2) S p ( x' ,x , ) = Cross spectral density of pressure between two locations (psi2/Hz) S p ( x' , ) = Power spectral density of pressure at a single location (psi2/Hz) m = Generalized mass (lbf-s2/in)
= Modal frequency (rad/s)
= Modal damping ratio (-)
Rearranging the equations above provides the following three equations for the displacement response spectra, acceptance integral, and modal transfer function. The frequencies are converted to be in units of Hz for convenience. This form of the acceptance integral becomes a factor for the mode shapes in Equation 3-4. This step allows the acceptance integral to be calculated once and then used multiple times to calculate the response due to turbulent buffeting for variables other than displacement, such as strain. S y ( x ,f ) = A ( f ) Equation 3-4 pyright 2022 by NuScale Power, LLC 62
A ( f ) = H ( f )H ( f ) Sp ( x' ,x ,f ) dx'dx Equation 3-5 1 H ( f ) = ------------------------------------------------------------ Equation 3-6 2 2 2 m 4 f - f + i f f where: A ( ) = Redefined acceptance integral (in2/Hz) f = Frequency (Hz) Equation 3-7 replaces the displacement mode shapes with strain mode shapes. Due to using beam elements in ANSYS, the strain mode shapes are bending strain in two directions and the axial strain. As the cross spectral density of the turbulent pressure is generally not available, the cross spectral density is replaced with known quantities using the definition of the coherence function in Equation 3-9. The form of the coherence function is further discussed in Section 3.1.2.1.5. S ( x ,f ) = A ( f ) Equation 3-7 A ( f ) = H ( f )H ( f ) ( x' ,x ,f ) S p ( x' ,f )S p ( x ,f ) dx'dx Equation 3-8 S p ( x' ,x ,f ) ( x' ,x ,f ) = ----------------------------------------- Equation 3-9 S p ( x' ,f )S p ( x ,f ) where: S ( x ,f ) = Strain response spectra (1/Hz) or = Strain mode shape (in/in2) ( x' ,x ,f ) = Coherence function defined in Section 3.1.2.1.5 (-) The acceptance integral shown in Equation 3-8 includes two area integrals. Those integrals are not calculated analytically. Instead, the integrals are evaluated numerically by two summations over the elements that make up the tube. pyright 2022 by NuScale Power, LLC 63
dependent. The displacement mode shape and PSDs are approximated as being constant over an element. This approximation is appropriate because the mode shapes vary over larger length scales than a single element. The displacement mode shapes and surface areas are vectors and are combined with the dot product operator to get the portion of the pressure acting on the surface area in the direction of the mode shape. A ( f ) =
- elems elems Equation 3-10 H ( f )H ( f )
( x' ,x ,f ) S p ( x' ,f )S p ( x ,f ) ( i , A i ) ( j , A j ) i j Acceptance integrals are not calculated for the possible combinations of alpha and beta mode indices. The mode combinations that do not significantly contribute to the summations in Equation 3-7 are eliminated. The mode combination significance is shown in Equation 3-11. The relative mode combination significance is calculated by normalizing by a hypothetical mode at the largest frequency of interest combined with itself. Acceptance integrals are calculated only for those mode combinations with a relative significance above 0.1. This ensures that all combinations of modes with themselves are included as well as large cross-modal combinations. A sensitivity on the threshold for significant mode combinations is provided in Section 3.1.2.2.5.
= max H ( f )H ( f ) Equation 3-11
,rel = m 8 f 2 2 Equation 3-12 max where:
= Mode combination significance (in2/lbf2)
,rel = Relative mode combination significance (-)
f max = Maximum frequency of interest (Hz) As beam elements are used in the tube model, the PSDs can be applied in the element y and z directions. The orientation node (node L) is on the element x-z plane. The ANSYS input file sets the location of the orientation node so that the element z direction is radially outward from the global vertical axis and the element y direction is offset from the global vertical axis by the tube inclination angle. pyright 2022 by NuScale Power, LLC 64
With the ANSYS structural model providing the structural information, the next input required for calculating the acceptance integrals is the pressure PSD. The pressure PSD is the input to the system that causes vibrations. As turbulence is a nondeterministic phenomenon, the forcing function cannot be represented exactly. Instead a PSD is used to provide a distribution of how the input pressure varies across the range of frequencies. 2.1.2 Primary Side PSD The pressure fluctuations on the outside of the tubes is modeled with a PSD for cross-flow in a tube bundle and repeated in Equation 3-15 below. The stored factor on the frequency converts a frequency to the reduced frequency. The stored factor on the pressure converts the normalized pressure PSD to the pressure PSD. The flow area along with the primary side density are used to calculate the velocity. The primary side density is calculated with the hot primary side temperature. F < 0.1 G p ( f ) = 0.01 0.1 < F 0.4 G p ( f ) = 0.2 Equation 3-13 3.4 F > 0.4 G p ( f ) = 5.3 x 10 F fD F = --------h- Equation 3-14 vf 1 2 3 G p = G p --- v D h Equation 3-15 4 f where: F = Reduced frequency (-) G p ( f ) = Normalized pressure PSD (-) G p ( f ) = Pressure PSD (psi2/Hz) f = Frequency (Hz) D h = Hydraulic diameter (in) pyright 2022 by NuScale Power, LLC 65
= Fluid density (lbf-s2/in4) v f = Free stream velocity(in/s) 2.1.3 Secondary Side PSDs For the cases listed in Table 3-18, the secondary side is either quiescent fluid, single phase moving liquid, or moving liquid boiling to vapor. For the cases of quiescent fluid, no secondary side PSD is applied as there is no turbulence to cause pressure fluctuations. For the cases with moving fluid, a secondary side PSD is applied based on either established literature or the measurements from the TF-1 test data.
The inlet quality, heat flux, and secondary side pressure are used to find the TF-1 test that most closely matches the conditions of each TF-2 test. Table 3-19 below lists each TF-2 case, the secondary side conditions for that case, and the TF-1 case that most closely matches the conditions. TF-2 cases with no secondary side flow are marked as N/A as no TF-1 PSD is applicable. pyright 2022 by NuScale Power, LLC 66
Table 3-18 Applicable TF-2 Test Cases Total Secondary Primary Flow Primary Primary Secondary se Name Model Case Side Temp. (kg/s) Temp. (°F) Press. (psia) Side Flow (°F) (kg/s) (( 001_0751 001_0752 001_0753 1 001_0754 001_0755 001_0757 002_0744 002_0745 002_0746 2 002_0747 002_0748 002_0750 003_0759 003_0761 003_0762 2 003_0763 003_0764 003_0766 004_0767 004_0768 004_0769 2 004_0770 004_0771 004_0773 005_0786 3 006_0784 3 006_0844 007_0777 007_0778 3 007_0779 007_0781
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 67
Table 3-19 TF-1 Secondary Side PSDs Used TF-2 Case Name TF-1 Case Used Secondary Side Flow (kg/s)1 (( TF0001_0751 TF0001_0752 TF0001_0753 TF0001_0754 TF0001_0755 TF0001_0757 N/A TF0002_0744 TF0002_0745 TF0002_0746 TF0002_0747 TF0002_0748 TF0002_0750 TF0003_0759 TF0003_0761 TF0003_0762 TA0070 TF0003_0763 TF0003_0764 TF0003_0766 TF0004_0767 TF0004_0768 TF0004_0769 N/A TF0004_0770 TF0004_0771 TF0004_0773 TF0005_0786 TD0023 TF0006_0784 TD0024 TF0006_0844 TF0007_0777 TF0007_0778 TD0048 TF0007_0779 TF0007_0781
}}2(a),(c),ECI e 1 The flow rates in the last column are for all five tube columns Each TF-1 case has five sensor measurements, and thus five PSDs, from the active coil (tube). The PSDs from the five sensors are averaged to get a representative PSD for the entire length of tube. This averaging is acceptable as the PSDs do not vary significantly with elevation.
The PSDs show a noise floor at (( }}2(a),(c),ECI. The noise is removed from the PSD by subtracting (( }}2(a),(c),ECI from the tube averaged PSD. After this subtraction, values less than 1.0e-12 are pyright 2022 by NuScale Power, LLC 68
These processed TF-1 PSDs are output for use in the strain response PSD calculations. The TF-1 PSDs contain a large amount of low frequency excitation in the (( }}2(a),(c),ECI range. There is also another peak in excitation between (( }}2(a),(c),ECI. The high frequency of this secondary peak is the reason why the turbulent buffeting analyses using the TF-1 PSDs are run out to 600 Hz. As an alternative to the TF-1 PSDs, PSDs can be implemented from open literature. Two PSDs are required to cover the single-phase and two-phase regions of the tube. The single-phase Chen PSD is shown in Equation 3-13 (from page 233 of Reference 9.1.11) and the single-phase Au-Yang/Jordan PSD in Equation 3-19 (from page 236 of Reference 9.1.10). The Chen PSD is used by default while the Au-Yang/Jordan PSD is used to evaluate the sensitivity of the results to the selected PSD.
-5 0.272 x 10 S < 5.0 G p ( f ) = ------------------------------
0.25 S Equation 3-16
-5 S 5.0 22.75 x 10 -
G p ( f ) = ----------------------------- 3 S 2 fD S = ---------------h- Equation 3-17 vf 2 3 Gp = Gp v Dh Equation 3-18 f where: S = Reduced frequency (-) G p ( f ) = Normalized pressure PSD (-) G p ( f ) = Pressure PSD (psi2/Hz) f = Frequency (Hz) D h = Hydraulic diameter (in)
= Fluid density (lbf-s2/in4) pyright 2022 by NuScale Power, LLC 69
v f = Free stream velocity(in/s)
-3.0F F < 1.0 G p ( f ) = 0.155e Equation 3-19
-1.26F F 1.0 G p ( f ) = 0.027e fR F = min 5.0 ,-------h- Equation 3-20 vf 2 3 Gp = Gp v Rh Equation 3-21 f
where: F = Reduced frequency (-) G p ( f ) = Normalized pressure PSD (-) G p ( f ) = Pressure PSD (psi2/Hz) f = Frequency (Hz) R h = Hydraulic radius (in)
= Fluid density (lbf-s2/in4) v f = Free stream velocity(in/s)
For the two-phase region, the correlations by Giraudeau from Reference 9.1.12 are implemented. These correlations are force PSDs of two-phase fluids around 90° elbows. Giraudeau found that the force PSD on an elbow is caused by the changes in the mass flux through the tube due to turbulent two-phase flow. The helical tube experiences similar behavior in the two phase region. The Giraudeau correlations are based on Equations 17, 19, 20, and 21 from Reference 9.1.12, repeated below in Equation 3-22, Equation 3-23, Equation 3-24, and Equation 3-25. The five empirical constants that go into the Giraudeau PSD are defined in Table 1 of the reference and Table 3-20 below. The constants are defined for four void fractions. When the actual void fraction in the tube is between the tabulated values, linear interpolation is pyright 2022 by NuScale Power, LLC 70
void fraction is between 0.05 and 0.99. The Giraudeau PSD is a force PSD on a 90-degree elbow. To convert the force PSD to a pressure PSD, it is divided by the area of the element squared. Also, each element only sweeps out a portion of a 90-degree bend. Therefore, the PSD is also multiplied by the square root of one minus the cosine of the angle swept out by the element. This factor ensures that the sum of all force vectors on elements that make up a 90-degree bend sum to the appropriate value. The Giraudeau PSD is for flow around an elbow and is therefore only applied radially on the tube. The single-phase PSD is also applied in the two-phase region. The single-phase PSD is insignificant compared to the Giraudeau PSD in the radial direction, but provides excitation in the vertical direction. f <f0 = k 1 f m1 Equation 3-22 m2 f f0 = k2 f j 0.8
= -------------------- ------- We Equation 3-23 2 2 D (j D ) h fD f = ---------h- Equation 3-24 j
2 l j Dh We = ---------------- Equation 3-25 where: f = Reduced frequency (-)
= Normalized force PSD (-)
= Force PSD (lbf2/Hz) f = Frequency (Hz)
D h = Hydraulic diameter (in) l = Liquid density (lbf-s2/in4) pyright 2022 by NuScale Power, LLC 71
j = Mixture velocity (in/s) We = Weber number (-)
= Surface tension (lbf/in)
Table 3-20 Giraudeau PSD Correlation Empirical Constants id Fraction f0 k1 k2 m1 m2
}}2(a),(c),ECI 2.1.4 Fluid Conditions for PSDs The fluid conditions on the secondary side change significantly over the height of tube due to the heating and boiling of the secondary side coolant. To more accurately calculate the local PSD at each element in the tube, elevation dependent fluid properties are required. The local fluid conditions are generated using the results of an NRELAP5 model of the TF-2 test facility.
The NRELAP5 results are steady state and do not show the temperature oscillations discussed in Section 3.1.2.2.1. 2.1.5 Coherence Function Coherence is a measure of the degree of relationship between two signals. In turbulence, coherence is used to measure how pressure fluctuations are related between two different points in a flow. The response of the structure is different if the pressure fluctuations occur in phase and with the full magnitude compared to pressure fluctuations that occur with phase offsets and different magnitudes. In Equation 3-26, the first exponential is based on the correlation length, which decreases the coherence as the two points are separated by more distance. Distances far apart tend to have pressure fluctuations that are not well correlated as turbulent eddies interact and dissipate. The second exponential is based on the convective velocity of the flow. This term adds a phase offset to the coherence based on the time it takes a turbulent eddy to travel the distance between the two points. This calculation uses the free stream velocity for the convective velocity. pyright 2022 by NuScale Power, LLC 72
the ANSYS model may not be fine enough to capture the rapid changes in coherence. To overcome this issue, additional points are added within each element when calculating the element coherence. The coherence at these additional points is calculated and an average value is used for the overall element coherence. The number of additional points is selected to give an average spacing of 0.125 in, as smaller spacing is shown to have a negligible effect on the results (Section 3.1.2.2.6). x1 - x2 2 if ( x 1 ,y - x 2 ,y ) C ( f , x 1 , x 2 ) = exp - ------------------- exp - ---------------------------------------
- Equation 3-26 v
where: C ( f , x 1 , x 2 ) = Coherence function (-) f = Frequency (Hz) x 1 = Point location 1 (in) x 2 = Point location 2 (in)
= Correlation length (in) v = Convective velocity (in/s) 2.1.6 Response Power Spectral Densities With the response PSD calculated using Equation 3-7, the mean square response can be calculated as the integral of the response PSD. The RMS response is the square root of the mean square response. The crossing frequency is a measure of the average frequency that the strain switches direction (compression/tension) and can be calculated using Equation 3-27.
2 fc = 0 f S ( f )df
-------------------------------- Equation 3-27 0 S ( f )df where:
f c = Crossing frequency (Hz) f = Frequency (Hz) pyright 2022 by NuScale Power, LLC 73
S ( f ) = Strain response PSD (in/in/Hz) These equations require many levels of summations which are implemented in a series of scripts. 2.2 Calculation Body 2.2.1 Strain Gauge Data Limitations The strain gauge data are available as described in Section 3.1. The data extend up to 300 Hz. Extended data up to 1000 Hz are also available, but do not have the same filtering of electrical noise as the 300 Hz data. The data from the top sensors for column 1 are not presented, because those sensors were not functional. Additionally, the data show a noise floor around (( }}2(a),(c),ECI µ2/Hz. This noise floor creates a threshold below which direct comparisons to the analytical approach are not possible. However, it is possible to infer that analytical results above the noise floor must bound the actual vibration response in the test facility. The TF-2 test facility has flow restrictor orifices at the inlet of the tubes. These restrictor orifices are not prototypic and have a lower loss coefficient compared to the NuScale design. There are differential pressure sensors located on select tubes to measure the pressure drop across the inlet flow restrictors which is related to the flow rate through the orifice. The differential pressure instruments indicate that for the fluid elastic instability (FEI) tests with boiling secondary side flow, there are significant oscillations in flow. Additional investigation of the available outer diameter temperature sensors shows that there are also significant temperature fluctuations caused by the boiling region inside the tubes moving up and down. The temperature fluctuations occur every (( }}2(a),(c),ECI seconds depending on the test conditions. The temperature data are sampled for at least 50 seconds at approximately 1 Hz while the strain gauge data are sampled for (( }}2(a),(c),ECI at 5000 Hz. Figure 3-36 below correlates the maximum range of strain for a set of sensors (column 3 lower top and side sensors for example) to the maximum temperature range for the nearest outer diameter temperature sensors (pair of intrados and extrados sensors). The ranges are calculated as the difference between the minimum and maximum values from the signal time histories. The results in Figure 3-36 show a somewhat linear trend indicating that larger recorded temperature ranges correspond to larger recorded strain ranges. The comparison is limited by two factors. First, the temperature sensors and strain gauges are not in the same location. The temperature sensors are nearby, within about four feet of tubing, but may not be representative of the temperature changes at the strain gauges. In some cases, the temperature pyright 2022 by NuScale Power, LLC 74
the same column is used. The second limitation is the duration and frequencies of measurements. The temperature data are collected for a long duration at low frequency and the strain data are collected for a short duration at a high frequency. The two data streams are collected through different systems and may not start at exactly the same time. Therefore, it is not possible to determine what temperature oscillations were occurring during the time the strain data are collected. Due to the (( }}2(a),(c),ECI period of temperature oscillations, the five second strain data may not coincide with the full range of temperature oscillations. This limitation provides justification for why the outlier points are lower than the trend. Section 3.1.2.5 describes a simplified ANSYS model of a tube that shows how temperature fluctuations correspond to strains in the tube. The results indicate that thermal strain is generated at a rate of (( }}2(a),(c),ECI µ/°C, which corresponds well to the trend in Figure 3-36 with a slight under prediction. The under prediction could be due to the assumption of unrestrained thermal expansion in the ANSYS model or differences in material properties compared to the test. pyright 2022 by NuScale Power, LLC 75
Figure 3-36 Strain Versus Temperature Ranges for the TD, TF, and TW Tests
}}2(a),(c),ECI Figure 3-37 is an example of the strain gauge data for a test case with boiling on the secondary side flow. The largest strains occur at frequencies below
(( }}2(a),(c),ECI. As shown in Figure 3-36, the low frequency temperature fluctuations correspond to the low frequency strain fluctuations. Due to the limited frequency of the temperature data, no direct conclusions can be made about the thermal strain in the (( }}2(a),(c),ECI range. While no direct comparisons are possible, the conclusion that the (( }}2(a),(c),ECI strains are due to thermal strains fits the trends in the TF-2 data. The low frequency strains occur only in cases with secondary side flow and are largest in cases with two-phase secondary side flow. The low frequency strains also only occur in sensors that are near sensors that show temperature oscillations. In Figure 3-37, the lower sensors have low frequency oscillations while the upper sensors show no low frequency strains. pyright 2022 by NuScale Power, LLC 76
Figure 3-37 Strain Gauge Data for a Case with Secondary Side Boiling
}}2(a),(c),ECI The thermal strains detected in TF-2 are not prototypic of the design. The design of the SG inlet flow restrictors is such that the inlet pressure loss stabilizes the flow at full power. A stable flow would not have such large thermal strains. Therefore, no attempt is made to match the analytical results to the strain data less than (( }}2(a),(c),ECI.
Without the strain data attributed to thermal oscillations, the TF-2 strain gauge data are predominately made up of noise at around (( }}2(a),(c),ECI. There are a few peaks that may be distinguishable from the noise but there does not seem to be a repeatable trend in those peaks. While the noise in the strain data prevents quantitative comparison for analytical results below the noise level, analytical results that are above the noise level are bounding with respect to the test data. pyright 2022 by NuScale Power, LLC 77
The following sections compare the analytical results using the acceptance integral method to the strain test data from TF-2. As described in Section 3.1.2.2.1, the strain data have low frequency content caused by thermal strains. No comparison is attempted for the strain data below (( }}2(a),(c),ECI. Cases without Secondary Side Flow Three TF-2 FEI test series had no secondary side flow, TF0001, TF0002, and TF0004. Of those test series, TF0004 is selected for detailed discussion because it represents the most bounding cases. TF0004 has stagnant liquid on the secondary side and the primary side is at slightly elevated temperatures. Both of these factors contribute to lower modal frequencies and increasing responses. Figure 3-38 and Figure 3-39 show the strain response for the lower and upper sensors of column 5 for TF0004_0769 with the pinned boundary condition. The analytical results are mostly below the noise floor of the strain sensors. The peak just below (( }}2(a),(c),ECI is due to the first modes of the tubes. The increased response between (( }}2(a),(c),ECI is due to the increased pressure input from the primary flow PSD in Equation 3-13. The top of tube sensors show higher response than the side sensors due to mode shapes being predominately in the vertical direction compared to the radial direction. pyright 2022 by NuScale Power, LLC 78
Figure 3-38 TF0004_0769 Column 5 Lower Strain Sensor with Pinned Boundary Conditions
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 79
Figure 3-39 TF0004_0769 Column 5 Upper Strain Sensor with Pinned Boundary Conditions
}}2(a),(c),ECI Figure 3-40 and Figure 3-41 show the strain response for the lower and upper sensors of column 5 for TF0004_0773 with the pinned boundary condition.
The responses are generally larger due to the much larger primary flow rate. The higher flow rate also extends the region of higher pressure PSD content to coincide with the first mode frequencies, which amplifies the first mode. pyright 2022 by NuScale Power, LLC 80
Figure 3-40 TF0004_0773 Column 5 Lower Strain Sensor with Pinned Boundary Conditions
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 81
Figure 3-41 TF0004_0773 Column 5 Upper Strain Sensor with Pinned Boundary Conditions
}}2(a),(c),ECI Figure 3-42 and Figure 3-43 show the strain response for the lower and upper sensors of column 5 for TF0004_0773 with the sliding boundary condition.
The responses are significantly larger than the pinned boundary condition and the response is shifted to lower frequencies. The lower frequency peaks are the sliding modes. The large low frequency oscillations occurring with the sliding boundary condition are orders of magnitude higher than the strain gauge noise floor and orders of magnitude higher than the pinned boundary condition. The pinned boundary condition already predicts larger responses than the strain gauge data and is therefore conservative. The sliding boundary condition responses are overly conservative. The remainder of the comparison is focused on the pinned boundary condition as it is found to be a better match to the test data. pyright 2022 by NuScale Power, LLC 82
Figure 3-42 TF0004_0773 Column 5 Lower Strain Sensor with Sliding Boundary Conditions
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 83
Figure 3-43 TF0004_0773 Column 5 Upper Strain Sensor with Sliding Boundary Conditions
}}2(a),(c),ECI Cases with Single Phase Secondary Flow The TF0003 test series have single-phase secondary side flow. The
(( }}2(a),(c),ECI kg/s of flow through the 252 tubes in TF-2 is comparable to the feedwater flow rate for the design at 100 percent power. However, the design would have boiling in the tubes while these tests are all single phase liquid. Figure 3-44 and Figure 3-45 are from TF0003_0762, which has a comparable primary side flow rate to TF0004_0769. Therefore, the main difference in test conditions between the plots below and Figure 3-38 and Figure 3-39 is the moving secondary side liquid. The results are nearly identical, which indicates that the secondary side single-phase PSD is insignificant compared to the primary side tube bundle PSD. pyright 2022 by NuScale Power, LLC 84
that create some detectable strain above the noise for less than (( }}2(a),(c),ECI Hz. Figure 3-44 TF0003_0762 Column 5 Lower Strain Sensor with Pinned Boundary Conditions
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 85
Figure 3-45 TF0003_0762 Column 5 Upper Strain Sensor with Pinned Boundary Conditions
}}2(a),(c),ECI Cases with Two Phase Secondary Flow Of the tests with two-phase secondary side flow, TF0007_0777 is most representative of the design at full power and maximum design flow. However, due to thermal power limitations in the TF-2 test facility, the secondary side flow is much lower than the actual design.
The results for all three instrumented columns are shown in Figure 3-46 through Figure 3-50. The upper sensors for column 1 are not shown as they were not functional during the test. In general, the analytical results overpredict the test data. The resonance peaks for the first tube modes are an order of magnitude higher than the noise floor of the test data. The lower sensors of columns 1 and 3 show some low frequency response that is due to the two-phase pressure PSD. The upper sensors do not show that same pyright 2022 by NuScale Power, LLC 86
of the two-phase region tends to excite modes near the bottom of the tube. Figure 3-46 TF0007_0777 Column 1 Lower Strain Sensor with Pinned Boundary Conditions
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 87
Figure 3-47 TF0007_0777 Column 3 Lower Strain Sensor with Pinned Boundary Conditions
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 88
Figure 3-48 TF0007_0777 Column 3 Upper Strain Sensor with Pinned Boundary Conditions
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 89
Figure 3-49 TF0007_0777 Column 5 Lower Strain Sensor with Pinned Boundary Conditions
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 90
Figure 3-50 TF0007_0777 Column 5 Upper Strain Sensor with Pinned Boundary Conditions
}}2(a),(c),ECI For comparison, the column 5 results for TF0007_0781 are shown in Figure 3-51. The higher primary flow rate in this test causes overall higher responses. Besides the thermal strain content, the test data does show clear peaks above the noise floor. The analytical results are orders of magnitude higher than the test data.
pyright 2022 by NuScale Power, LLC 91
Figure 3-51 TF0007_0781 Column 5 Lower Strain Sensor with Pinned Boundary Conditions
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 92
Figure 3-52 TF0007_0781 Column 5 Upper Strain Sensor with Pinned Boundary Conditions
}}2(a),(c),ECI 2.2.3 TF-1 Secondary Side PSD Sensitivity Additional cases are run to assess how well the results using the TF-1 pressure PSD data match with the TF-2 strain data. The TF-1 data also have some different features compared to the literature PSDs that require explanation.
Figure 3-53 and Figure 3-54 show the strain results for a typical case upper and lower sensors. The TF-1 pressure PSD has a large amount of low frequency content ((( }}2(a),(c),ECI) which increases the analytical response greatly. While this result matches the TF-2 test data better for the lower sensor (Figure 3-53), the comparison is worse for the upper sensor (Figure 3-54). Section 3.1.2.2.1 discusses that the low frequency strains in the TF-2 test data are due to thermal strains caused by unsteady fluid pyright 2022 by NuScale Power, LLC 93
results significantly overpredict the low frequency strain response. The TF-1 facility used piston pumps to supply feedwater to the flow. The TF-1 pressure data less than about (( }}2(a),(c),ECI is attributed to the pumping frequency. This interpretation is corroborated by the available pressure data from the feedwater inlet. While the feedwater inlet pressure is sampled at a much lower frequency, the data show oscillations that are consistent with the low frequency portion of the pressure PSDs measured inside the tube. This information indicates that the low frequency pressure data are caused a phenomenon that affects the whole test loop, like the piston pumps, and not just the tubes. The other important feature of the TF-1 PSDs is the broad spectral peak around (( }}2(a),(c),ECI. One analysis of the test data describes this peak as being due to bubble formation in the flow. While that explanation is reasonable, it does not entirely fit the data. When present, the spectral peak appears in all pressure sensors of the active tube. The peak also appears in the adiabatic cases that have two-phase flow at approximately constant quality. The (( }}2(a),(c),ECI peak is greatly reduced in magnitude for the diabatic cases with low electrical power and with high subcooling. Explanations other than bubble formation are possible, but without a clear explanation for the (( }}2(a),(c),ECI peak, it is conservative to assume that it is a real phenomenon. The TF-1 cases are run for frequencies less than 600 Hz. Looking at the response results for frequencies greater than 300 Hz, the strain content is on the order of (( }}2(a),(c),ECI and the displacement content is on the order of (( }} 2(a),(c),ECI . Both of these results are not large enough to be significant. Even if the (( }}2(a),(c),ECI peak is a phenomenon applicable to the design, it does not have a significant impact. The strain gauge data up to 300 Hz shown in Figure 3-53 and Figure 3-54 remove electrical interference. Figure 3-55 shows the strain gauge data out to 600 Hz without the same processing. The response in the data around (( }}2(a),(c),ECI is consistent with the noise floor and does not show indication that the (( }}2(a),(c),ECI pressure data measured in TF-1 is manifested in the strain response for TF-2. pyright 2022 by NuScale Power, LLC 94
Figure 3-53 TF0007_0781 Column 5 Lower Strain Sensor with Pinned Boundary Conditions and the TF-1 PS
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 95
Figure 3-54 TF0007_0781 Column 5 Upper Strain Sensor with Pinned Boundary Conditions and the TF-1 PSD
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 96
Figure 3-55 TF0007_0781 Column 5 Unfiltered Strain Sensor up to 600 Hz
}}2(a),(c),ECI 2.2.4 Au-Yang Secondary Side PSD Sensitivity In order to assess the sensitivity of the results to the single-phase PSD on the secondary side, additional cases are run using the Au-Yang/Jordan PSD described in Equation 3-19. The Au-Yang/Jordan PSD is developed to bound highly turbulent single phase flows. Figure 3-56 shows that the responses are significantly higher with the Au-Yang/Jordan PSD compared to the Chen PSD in Figure 3-49. The differences are particularly large around the modes near
(( }}2(a),(c),ECI. The Au-Yang/Jordan PSD adds response that results in strains higher than the noise floor of the test data. The test data do not show the same peaks and therefore, the Au-Yang/Jordan PSD is only adding conservatism with respect to the test data. The Chen PSD is a better match to the test data. pyright 2022 by NuScale Power, LLC 97
Figure 3-56 TF0007_0781 Column 5 Lower Strain Sensor with Pinned Boundary Conditions and the Au-Yang/Jordan PSD
}}2(a),(c),ECI 2.2.5 Mode Combination Significance Sensitivity To aid in the efficient computation of the response spectra, only the most significant of the possible mode combinations are used. The significance criteria are described in Equation 3-11 and Equation 3-12. The selected significance threshold is 0.1, but a sensitivity is run using a value of 0.01.
Using TF0007_0781 column five with the pinned boundary condition as an example, the threshold of 0.1 uses the 1683 most significant mode combinations out of 4753 possible combinations. The threshold of 0.01 uses the 3682 most significant combinations. Figure 3-57 shows the results using the threshold of 0.01. The results are largely the same as those in Figure 3-51. The RMS strain amplitudes are within a fraction of a percent for the x and y directions. The z direction results are within approximately one percent, which is acceptable. pyright 2022 by NuScale Power, LLC 98
Figure 3-57 TF0007_0781 Column 5 Lower Strain Sensor with Pinned Boundary Conditions and More Mode Combinations
}}2(a),(c),ECI 2.2.6 Coherence Integral Mesh Sensitivity The numerical solution of the acceptance integrals includes a term for the coherence which is a combination of a decreasing exponential and a sinusoid as shown in Equation 3-26. Since this function can change on length scales smaller than the finite elements, a mesh sensitivity is performed to show that the selected size of 0.125 inch is appropriate for the acceptance integral calculation.
To assess the mesh sensitivity, cases are run with the mesh size reduced by a factor of two. Figure 3-58 shows fine mesh results not noticeably different from Figure 3-51 with the larger mesh size. The calculated RMS strain amplitudes from the two mesh sizes are less than (( }}2(a),(c),ECI different which is acceptably small. pyright 2022 by NuScale Power, LLC 99
Figure 3-58 TF0007_0781 Column 5 Lower Strain Sensor with Pinned Boundary Conditions and Fine Mesh
}}2(a),(c),ECI 2.2.7 Damping Sensitivity To investigate the effect of damping on the results, additional cases are included with a damping of (( }}2(a),(c),ECI. Figure 3-59 shows the same overall trend as Figure 3-51, but with overall smaller response near resonance. This damping sensitivity study is inconclusive with respect to what damping provides a better match to the TF-2 test data. This sensitivity study does show that damping would have needed to be significantly larger to move the resonance peaks below the noise level in the strain gauge data.
pyright 2022 by NuScale Power, LLC 100
Figure 3-59 TF0007_0781 Column 5 Lower Strain Sensor with Pinned Boundary Conditions and (( }}2(a),(c),ECI Damping
}}2(a),(c),ECI 2.3 Results and Conclusions 2.3.1 TF-2 Low Frequency Strains Analysis of the TF-2 test data shows that the low frequency strains are mainly due to temperature oscillations caused by secondary side flow oscillations.
The limitations of the duration and frequency of the TF-2 test data prevent a direct comparison, but the trends indicate that low frequency strains occur when nearby temperature sensors also show low frequency temperature oscillations. Simple ANSYS model described in Section 3.1.2.5 provides further indication that the magnitude of the low frequency strains corresponds well to the magnitude of the temperature changes. The temperature oscillations in the TF-2 test facility are characteristic of secondary side flow oscillations that may occur at low secondary flow rates. pyright 2022 by NuScale Power, LLC 101
strains in the data are categorized as noise. This observation indicates that strains due to vibrations in the tests were lower than the noise floor for the higher frequencies. 2.3.2 TF-1 PSDs Cases run using the TF-1 measured pressure PSDs show that the TF-1 PSD is overly bounding. Noise in the TF-1 PSD and low frequency pressure oscillations, potentially caused by the piston pumping frequency of the facility, create challenges in using the TF-1 PSDs. The combination of the Chen single-phase PSD and the Giraudeau two-phase PSD provides a better match to the TF-2 test data. The TF-1 pressure PSDs show a peak near (( }}2(a),(c),ECI. There is currently no conclusive physical explanation for this peak. Therefore, the data are assumed to be realistic and included in an analysis. The responses in strain and displacement due to the (( }}2(a),(c),ECI peak are not significant. The reduced impact of these high frequency pressure oscillations is due to the natural decrease in the modal transfer function at higher frequency inputs. 2.3.3 Sensitivity Studies Sensitivity studies on the mode significance threshold and mesh size indicate that the values used in this analysis are appropriate. A lower threshold or a smaller mesh size did not significantly affect the conclusions in this calculation. A sensitivity on the secondary side single-phase PSD shows that using the bounding Au-Yang/Jordan PSD is overly bounding and compares worse to the test data compared to the Chen PSD. A damping sensitivity showed that the resonance peaks are sensitive to damping as expected. Based on the noise in the TF-2 strain data, no conclusive comparison is possible about what damping is present in the test. 2.3.4 Vibration Analysis Discounting the thermal strains in the TF-2 test data, there are no clear resonance peaks in the data. The data are dominated by a noise floor around (( }}2(a),(c),ECI. This noise floor corresponds to small strains, on the order of (( }}2(a),(c),ECI RMS. The actual strains in the TF-2 facility are washed out by the noise floor and therefore are less than or equal to the noise floor. As the noise floor corresponds to small responses, it can also be concluded that the TB vibrations in the test facility are small even for scaled primary flow rates that are much higher than in the design. pyright 2022 by NuScale Power, LLC 102
above the noise floor in the data. The cases where the highest response is expected, highest primary and secondary flow rates, show analytical peak responses orders of magnitude higher than the test data. The shape of the analytical results matches typical responses to turbulent excitations with the majority of the response clustered around the lowest modal frequencies. Assuming that the actual vibration signal in the TF-2 facility is of similar shape, the comparison of the peaks in the analytical results to the lack of peaks above the noise floor in the test data indicates that the analytical approach is bounding compared to the test data. The analytical results indicate that the primary PSD and the two-phase secondary PSD are the most significant inputs. The primary PSD drives the largest resonance peaks at the low frequency tube modes. The two-phase PSD is larger in magnitude, but acts at frequencies well below the tube fundamental frequency. The two-phase PSD is also local to the lower portion of the tube which only excites modes in the same area. 2.4 Acceptance Integral Implementation Validation To validate the implementation of the acceptance integral method, several test cases are used to compare the methodology to the Random Vibration module in ANSYS Mechanical. ANSYS has the capability to calculate the response spectra due to PSD excitation, but has limitations on what types of coherence can be used. ANSYS can do fully correlated PSDs, PSDs correlated by a phase offset, and spatially correlated PSDs. The spatial correlation is limited to a linear ramp between fully correlated and zero correlation at two distances. Figure 3-60 shows the geometry used for the test cases. The beam is straight and aligned in the positive x direction starting from the origin. The beam is 60 in, long and meshed with 1-in. elements. All three translational degrees of freedom are set to zero on the left end and the y and z direction translations are set to zero on the right end. Figure 3-60 Test Case Geometry An arbitrary pressure PSD is applied to the beam in the y and z directions. The first test case uses a fully correlated PSD. In order to use the phase offset or spatial correlation options in ANSYS, the pressure PSD is converted to a force PSD. The equivalent force PSD yields also the same results as the original pressure PSD. The four test cases are as follows: pyright 2022 by NuScale Power, LLC 103
- 2. Wave - Force PSD correlated by a phase offset due to an x direction velocity of 600 in/s
- 3. Spat - Force PSD with spatial correlation (fully correlated for points within 5 in. and uncorrelated for points greater than 10 in away)
- 4. Half - Force PSD on the left half of the beam with spatial correlation (fully correlated for points within 1 in and uncorrelated for points greater than 3 in.
away) The same four test cases are executed using the acceptance integral methodology. The results are compared for a point at x = 25 inches. The resulting RMS values are within (( }}2(a),(c),ECI for displacement and within (( }} 2(a),(c),ECI for bending strain. Spectra for the four cases for displacement and bending strain are shown in Figure 3-61 through Figure 3-68. pyright 2022 by NuScale Power, LLC 104
Figure 3-61 Displacement Spectrum Comparison for Test Case 1
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 105
Figure 3-62 Displacement Spectrum Comparison for Test Case 2
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 106
Figure 3-63 Displacement Spectrum Comparison for Test Case 3
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 107
Figure 3-64 Displacement Spectrum Comparison for Test Case 4
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 108
Figure 3-65 Bending Strain Spectrum Comparison for Test Case 1
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 109
Figure 3-66 Bending Strain Spectrum Comparison for Test Case 2
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 110
Figure 3-67 Bending Strain Spectrum Comparison for Test Case 3
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 111
Figure 3-68 Bending Strain Spectrum Comparison for Test Case 4
}}2(a),(c),ECI 2.5 Tube Thermal Strain Model To get an approximate measure of how temperature fluctuations would induce strain in the tubes, a simplified axisymmetric model of the tube is developed in ANSYS. Figure 3-69 shows the model and mesh. The thermal boundary conditions are selected based on estimates from the NRELAP5 cases. The exact boundary conditions are not critical as the purpose is to understand how much strain is generated for a given change in outer diameter temperature. An axisymmetric model is used for its simplicity and efficiency. Modeling the curvature of the tube is not critical to get an estimate of the effect of temperature on strain.
The inner and outer wall heat transfer coefficients are 15 kW/m2-K and 12 kW/m2-K respectively. The inner wall fluid temperature is 420 degrees F and oscillates as a square wave with an amplitude of 40 degrees F zero to peak and a frequency of 0.1 Hz. The square wave is selected to represent the quick fluid pyright 2022 by NuScale Power, LLC 112
conditions. The precise frequency is not important as the temperatures reach an approximate steady state after only a few seconds. The structural boundary conditions include a fixed boundary condition in the vertical direction at the lower edge and a constraint that the nodes on the upper edge have the same y coordinate. These boundary conditions emulate a situation where the tube is allowed to grow unrestrained in the vertical direction. This modeling choice is an approximation as there is some resistance to lengthening of the tube which may result in additional strains. Figure 3-70 shows the resulting strain time history due to the temperature changes. The strain change occurs over about a one second period out of a 10 second cycle which would be captured by only some of the five second sets of strain data. The strain plot in Figure 3-70 occurs for a 28-degree F temperature change on the outer diameter of the tube. Therefore, a strain range of (( }}2(a),(c),ECI on the outer diameter of the tube corresponds to a temperature change of (( }}2(a),(c),ECI. This rate is important because it relates two measured quantities from the TF-2 test, temperature and strain, on the outer diameter of the tube. pyright 2022 by NuScale Power, LLC 113
Figure 3-69 Tube Axisymmetric Model Mesh
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 114
Figure 3-70 Tube Axisymmetric Model Mesh
}}2(a),(c),ECI TF-2 Benchmark Testing for Fluid Elastic Instability In the process of benchmarking the thermal-hydraulic performance of the prototypic helical soil steam generator (HCSG), NuScale commissioned SIET to construct and test two separate test specimens. Test Facility #1 (TF-1) utilized electric direct heating of several individual helical coil tubes to investigate thermal hydraulic behavior within the secondary-side (HCSG tubes). Test Facility #2 (TF-2) was a more-complex specimen, consisting of a 5-column, partially-prototypic HCSG assembly inserted into a test vessel and operated at a range of conditions, with variations in both primary- and secondary-side parameters. Given that it represents a smaller version of the ultimate fully-prototypic design, TF-2 was used to obtain a limited sample of data to characterize the response of individual HCSG tubes to various flow conditions. These measurements were accomplished by affixing axially-oriented strain gauges to the outside of specific tubes. During testing, TF-2 was subjected to both normal and above-normal operating conditions, including a series of cases wherein primary flow velocities were increased to approximately (( }}2(a),(c),ECI above normal operating conditions to assess whether FEI excitation of the HCSG tubing is feasible.
This section provides a detailed analysis of the strain gauge measurements obtained from TF-2 during FEI testing conditions, including the methodology, results, relevant observations, and the ultimate conclusion that the onset of FEI was not observed in the tests. While the TF-2 test was not fully-prototypic, the measurements and subsequent results presented herein are used in conjunction with other testing program results to validate the HCSG CVAP design analysis. pyright 2022 by NuScale Power, LLC 115
The primary purpose for fabrication and testing of TF-2 is qualification of the NuScale thermal hydraulic models, and confirmation of fluid property specifications and assumptions. SIET possesses an existing test facility (GEST) with the pressure vessel and associated pumps and equipment necessary to perform high-pressure, high-temperature testing of vessel internal components. A prototypical assembly was created to mimic the as-designed HCSG, with representative tubes of identical OD and pitch. Modifications were made to the end portal designs to simplify manufacturing. The size of the specimen was limited by the size of the GEST vessel, which necessitated using Columns 1-5 (of 21 total) as shown in Figure 3-71. This limitation is less desirable for the FEI assessment because the inner-most tubes have the lowest reduced velocities and require the highest driving flow to reach the onset point. Figure 3-71 TF-2 Fluid-Heated Test Section Tubing Column Scheme
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 116
during initial scoping activities for the test program. The opportunity to obtain additional, value-added data in support of CVAP benchmarking efforts was acted on once recognized; however, at that stage in testing, it was not possible to implement a comprehensive instrumentation configuration for FIV. There were physical and environmental challenges associated with the sensors themselves. In addition, the design and construction of TF-2 did not facilitate use of a fully-prototypic tube support configuration. The TF-2 SG tubes and tube supports are shown in Figure 3-72. The support columns are arranged in a consistent pattern at azimuths of ((
}}2(a),(c),ECI, and so on, such that the longest unsupported span of the HCSG tubing is (( }}2(a),(c),ECI.
Figure 3-72 TF-2 Tube Support Detail
}}2(a),(c),ECI The NPM-20 SG has eight tube support columns, whereas TF-2 was fabricated with four support columns (at (( }}2(a),(c),ECI), for an effective span length of (( }}2(a),(c),ECI. This configuration is beneficial for the FEI assessment because, compared to the fully-prototypic support configuration, the longer spans result in lower frequencies. The TF-2 supports restrain motion in a pyright 2022 by NuScale Power, LLC 117
configuration, the interface at each tab/guide approximates a fixed (pinned) connection. The presence of multiple tabs in close proximity serves to reduce rotation about the support column and constrains the significant modes to individual bending of the unsupported spans. Minor differences in thermal expansion between the tubing, tube supports, reactor pressure vessel (RPV) riser, and RPV wall cumulatively result in the tubing being pinned tightly against one side of the supports, further supporting the above determination on modal constraints. Variations in support interface rigidity may introduce variations in localized static strains within individual tubing spans. Given the preceding discussion, the natural frequencies and mode shapes of the TF-2 tubing are not expected to match those from NuScales fully-prototypic design models, including the FEI analysis. Modal predictions for several single tube configurations within TF-2 facilitate comparison and evaluation of the resulting data. Because the corresponding modal analysis has not been finalized as of issuance of this report, the corresponding mode shapes and frequencies presented herein are treated as an assumption. Given the geometry and support arrangement of the helical array, there are a number of modes, many with similar, closely-spaced frequencies and response profiles. The frequencies that are of interest based on mass participation are provided in Section 3.1.1.14. Section 3.1.1.14 and Table 3-17 discuss the frequencies that are expected to provide significant contribution to potential tubing movements based on ANSYS modal analysis of the TF-2 test specimen. The TF-2 models were generated in a global/cartesian coordinate system, wherein the Y-direction is vertical and the X- and Z-directions are either axial to or perpendicular with the tubing, depending on azimuth. The in-vessel instruments were exposed to a fluid environment over the full range of operating parameters, which restricts the applicable sensors to a small range. Also, there were physical challenges in terms of limited space to install or affix sensors in the space between the inner (annual) and outer walls. Given these challenges, SIET chose to employ Kyowa strain gauges (Model KHC-20-120-G9-16), which are affixed to the tubing by capacitive discharge welding. These sensors have been deployed for in-vessel measurements of boiling water reactor steam dryers. The noise floor is quite high compared with 350 strain gauges, which have a higher gauge resistance and improved signal-to-noise ratio. The indicated strain gauges were installed in pairs at six locations (12 total sensors). Columns 1, 3, and 5 were chosen to represent the range of available geometries within TF-2. For each column, strain gauges were installed on one select tube, at a Lower and an Upper location. These locations were generally 1.5 helical turns from the respective FW or steam plenums. At each location, one strain gauge was installed on the tube extrados (side) and another installed on the top. Both strain gauges were oriented axially. Figure 3-73 is a schematic illustrating the placement of strain gauges at each location and Figure 3-74 is a photograph of the sensors installed at the lower pyright 2022 by NuScale Power, LLC 118
installed strain gauges. Because the actual damage mechanism associated with FEI is excessive tube motion leading to adverse wear and impact, it is important to understand how measured strains relate to actual tube displacements. This relationship varies depending on sensor placement and support configuration. Figure 3-73 Placement of Strain Gauges on Tube Coils (Typical)
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 119
igure 3-74 Placement of Strain Gauges on Tube Coils (S1101-1 and S1101-2 Shown)
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 120
Table 3-21 Placement of Strain Gauge Instrumentation sor Description Sensor Placement (General) Sensor Placement (Specific) (/ to Elevation Linear Azimuth ame ID Column Tube Plenum Position Nearest (m) (m) (°) Support (( 101-1 S01 Side FW 101-2 S02 Top 1 20 102-1 S03 Side MS 102-2 S04 Top 101-1 S05 Side FW 101-2 S06 Top 3 21 102-1 S07 Side MS 102-2 S08 Top 101-1 S09 Side FW 101-2 S10 Top 5 11 102-1 S11 Side MS 102-2 S12 Top
}}2(a),(c),ECI es: 1. The Plenum column defines sensor location based on the nearest inlet/outlet - -FW or MS.
- 2. Global elevations are measured from a datum several meters below the steam generator.
- 3. Linear positions are measured from the inlet (FW) orifice of an individual tube, along its centerline.
- 4. Global azimuths are measured from a 0° reference at the inlet (FW) plenum. In this arrangement, tube supports are at global azimuths of (( }}2(a),(c),ECI and so on.
- 5. Sensors S03 and S04 (Column-1, Upper/MS Location) were determined to be non-functional upon installation, and the data from these channels is excluded from subsequent analysis hereinafter.
Table 3-22 summarizes the test series selected for further processing, including the applicable steady-state fluid parameters at the time of acquisition. It also presents a shortened alpha-numeric naming convention used to identify or discuss specific datasets throughout the remainder of this section. Test series TF0005 and TF0006 contains datasets at ramped flow conditions, but only at maximum flow conditions. It is assumed that these datasets are outliers wherein SIET was attempting to establish steady-state flow conditions at elevated temperature and pressure. pyright 2022 by NuScale Power, LLC 121
ble 3-22 Test conditions and File Names for Fluid Elastic Instability Data Acquisition Secondary st Series Primary Side Test ID Side Flow Temp. Pres. Temp. Conditions (kg/s) (°F) (psi) (°F) (( 114.4 A1 141.7 A2 172.5 A3 TF0001 Ambient 199.2 A4 230.5 A5 263.3 A6 113.7 B1 143.3 B2 174.6 B3 TF0002 Ambient 205.0 B4 229.5 B5 262.5 B6 113.2 C1 142.2 C2 175.5 C3 TF0003 Ambient 200.7 C4 234.5 C5 263.6 C6 113.5 D1 145.7 D2 170.6 D3 TF0004 249.9 202.0 D4 232.6 D5 266.8 D6 TF0005 252.7 580.0 E6 260.2 F6a TF0006 499.9 262.3 F6b 168.4 G3 199.0 G4 TF0007 449.8 227.0 G5 256.0 G6
}}2(a),(c),ECI es: 1. The indicated temperatures represent inlet values. The secondary-side parameters represent average values across tube columns. Where present, singular values represent an average or nominal value for datasets within a particular test series.
- 2. The indicated test IDs (right-most column) were assigned to simplify reference to individual datasets or conditions. The first character (alpha-indicator) refers to a specific test series with consistent primary and secondary side parameters (note: primary flow varies); the second character (numeric) is a sequential index increasing with flow rate.
pyright 2022 by NuScale Power, LLC 122
spectral (frequency) content. A consistent processing approach was applied to each of the data files in Table 3-22, to enable like-for-like comparisons. This processing consists of eliminating spurious signal noise and computing statistics or estimates to inform the FEI evaluation. The general steps and characteristics of the processing script are described below for reference. The raw waveform files exhibit variation between datasets and individual channels in terms of average (DC) value(s). Figure 3-75 exemplifies these DC variations by plotting the raw waveform data for sensors S01, S05, and S09 (one in each column) at the maximum primary flow for each test series (i.e., datasets A6, B6, C6, D6, and G4). These variations were anticipated given the sensor type and configuration, as well as the uncertain boundary conditions on individual tubes when the HCSG assembly is thermally loaded. Given the randomness of the variations, it is not possible to infer meaningful results by comparing the static values (DC/average) between datasets or individual channels. Removal of the DC content (by subtracting the average) results in reasonable dynamic waveforms, which are comparable between datasets or channels. Thus, the DC content of the waveforms is removed during processing. Digital filtering was used to restrict the signal content to the frequency range of interest for the HCSG and to eliminate spurious electrical noise. Filters are based on ((
}}2(a),(c),ECI. The filters were applied in the forward and reverse direction to avoid phase distortion, using the filtfilt function in Matlab.
If present, FEI excitation within the HCSG is expected to amplify the first-mode bending frequencies of the individual tube columns. These frequencies range from approximately (( }}2(a),(c),ECI (radial direction) to (( }}2(a),(c),ECI (vertical direction), depending on the assumed boundary conditions. Given the support configuration and general flexibility of the thin-walled tubes, higher-order modes may also be visible within the data, up to 100 Hz per the modal analysis. For the FEI analysis herein, a digital bandpass filter was used to de-emphasize content outside of the frequency range of interest. Specifically, the filter negates the impact of low-frequency drift and slow oscillations that are visible in many of the unadulterated waveforms. ((
}}2(a),(c),ECI.
The TF-1 dynamic pressure data exhibits a notable response peak between (( }}2(a),(c),ECI within the dynamic pressure data obtained from the secondary-side sensors. Thus, a second or alternate processing run was performed with a bandpass filter range of 5 to 1000 Hz to evaluate whether the data contains higher-frequency peaks of interest that align with the TF-1 content. A separate set of processing parameters are applicable to the higher-frequency bandpass, which are referred to as Run-2 hereinafter. The 3-300 Hz bandpass noted in the previous item is referred to as Run-1. Table 3-23 provides a summary of the exact processing/filtering parameters applied during Run-1 and Run-2. pyright 2022 by NuScale Power, LLC 123
The strain gauge signals exhibit some electrical line noise ((( }}2(a),(c),ECI) or its multiples ((( }}2(a),(c),ECI , and so on). In order to delineate content at apparent structural frequencies and normalize overall results between datasets, these electrical frequencies (where present) were removed using bandstop (notch) filters. For each peak, a 6th-order filter was applied, with a bandwidth of (( }}2(a),(c),ECI depending on the amplitude and width of the target frequency. For Run-2, filtering of electrical multiples (i.e., (( }}2(a),(c),ECI and higher) introduced additional spectral noise. Thus, the bandstop filters for Run-2 were applied in an identical fashion as Run-1, such that some amount of electrical noise remained present at (( }}2(a),(c),ECI and above. For certain channels, the digital filtering operations produced a ringing effect at the beginning and end of the resulting waveform where the digital filters struggle to track the target raw signal from unknown initial conditions. Therefore, before calculating the overall values and frequency transforms for each channel, one-half second of data are truncated from the start and end of each channels waveform to remove potential ringing effects. The net effect of this operation is that the overall length of each channel waveform is reduced from ((
}}2(a),(c),ECI.
After these operations were completed, overall values were obtained from the processed waveforms for each channel of each dataset. Values were obtained in terms of RMS, zero-to-peak (0-pk), and peak-to-peak (pk-pk) measures, although only the RMS and 0-pk results are used in the remainder of this report. The processed time domain data are converted to the frequency domain using a fast Fourier transform (FFT) algorithm; specifically, the pwelch function within Matlab, which calculates the broadband-normalized power spectral density (PSD) of a signal. A frequency resolution (bin width or f) of (( }}2(a),(c),ECI is applied for all channels, resulting in multiple waveform data blocks of (( }}2(a),(c),ECI duration. For the large frequency range in Run-2, a f=(( }}2(a),(c),ECI is applied to obtain 1-second data blocks). An overlap of 50 percent is used to increase the number of data blocks in the computation. A Hanning window is used in conjunction with the FFT algorithm to avoid spectral bin leakage due to varying start and end conditions of each block, thereby improving spectral bin resolution. Use of the Hanning window in this manner has a global scaling effect on each block, which in turn requires multiplication by an adjustment factor (1.5) in order for frequency domain amplitudes to match time domain content at specific frequencies. The windowed, scaled FFT results for each block are then linearly averaged to compute the overall PSD. This averaging process improves spectral resolution and normalizes the effects of frequency-specific amplitude perturbations. The entire process of resolving the waveform into overlapped blocks, performing the FFT conversion, windowing, scaling, and averaging is self-contained and controlled by the pwelch function and its parameter inputs. pyright 2022 by NuScale Power, LLC 124
spectral plots are represented in RMS units to simplify identification and characterization of dominant peaks. Thus, the overall PSDs computed as described in the preceding item were converted to RMS spectra, using the expression in Equation 3-28. S rms ( f ) = S psd ( f ) f F window Equation 3-28 Where, S rms ( f ) = Frequency domain spectral output in terms of RMS value, S psd ( f ) = Frequency domain spectral output in terms of PSD value (i.e., from pwelch), f = Spectral resolution or bin width, and F window = Window adjustment factor (1.5 for Hanning window). For any signal, Parsevals theorem holds that the total energy content in the time domain must equal the energy content within the resulting frequency domain. Thus, for the spectra obtained from the frequency domain conversions described above, it is possible to compute an overall RMS value over a specific frequency range using a square-root-of-the-sum-of-the-squares approach. This relationship is expressed in generic form (i.e., for an FFT algorithm without windowing, scaling, and averaging) in Equation 3-29. The integral limits for the spectral term (right-hand side) of Equation 3-29 were modified to reflect a frequency range of interest (i.e., flow to Fhigh instead of 0 to F). The expression in Equation 3-29 is applicable to the normalized (single-sided), raw (complex) output of generic FFT algorithms. For this application, use of the pwelch function in the manner described above automatically returns a single-sided, magnitude-only (non-complex) spectrum, which is properly scaled for the applied windowing and averaging parameters. Consequently, the expression used to compute overall RMS values from the pwelch spectral results is modified for span-specific calculations as shown in Equation 3-30. 1 T 2 2 F 2 W rms = T o
--- [ W ( t ) ] dt = S rms =
F o
--- [ S ( f ) ] df Equation 3-29 Where, W rms = Time domain RMS measurement from signal waveform, W ( t ) = Discretized time domain waveform, containing "N" number of points, pyright 2022 by NuScale Power, LLC 125
T = Period/length of time domain signal (s), dt = Difference in time between measurement points (i.e., inverse of sampling rate), S rms = Frequency domain RMS measurement from normalized (single-sided) Fourier transform, S ( f ) = Normalized Fourier transform raw output (in complex form), F = Maximum discretized frequency of the continuous Fourier transform. For N waveform points, F = N f 2 , and df = Difference in frequency between bins (i.e., f). f high 2 1 S rms_f = [ Spsd ( f ) ] f -------------------- F window Equation 3-30 f low Where, S rms_f = Frequency-band-specific RMS measurement computed from pwelch spectrum output, f low = Lower bound of the frequency span of interest, and f high = Upper bound of the frequency span of interest. pyright 2022 by NuScale Power, LLC 126
Figure 3-75 Example of Average Variations in Strain Gauge Signals
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 127
Table 3-23 Summary of Applied Processing Parameters Parameter- Run-1 Run-2 (( DC Removal Bandpass Filter Notch Filters (Electrical Noise) Waveform Truncation Frequency Spectra/PSD
}}2(a),(c),ECI es: 1. A nominal value of (( }}2(a),(c),ECI was used for notch filter bandwidth in most cases; the (( }}2(a),(c),ECI setting was applied in select cases where the width of the
(( }}2(a),(c),ECI primary electrical noise peak warranted additional reduction of the adjacent frequency bins.
- 2. The indicated notch filters at (( }}2(a),(c),ECI were only applied when review of the spectra indicated content present at those frequencies.
2 TF-2 FEI Benchmarking Analysis Results The datasets listed in Table 3-22 were processed and the results collected in intermediate output files, containing summarized overall or statistical values, as well as digitized spectral tables. The intermediate results were then formatted into representative tables and plots for further assessment and interpretation. The list below summarizes the results obtained in this fashion. Unless otherwise explicitly specified, the tables, plots, and summary values specified below were obtained from datasets processed according to the standard parameters specified in Run-1 in Table 3-23. The waveform duration ((( }}2(a),(c),ECI) is short and does not facilitate accurate characterization of frequencies when acquiring random vibration data as is typical of FIV excitation. Specifically, for turbulence or FEI, one can observe variations in the amplitude of response peaks over the course of (( }}2(a),(c),ECI due to minor differences or local effects in the fluid excitation function. The concern is particularly relevant at low frequencies ((( }}2(a),(c),ECI), as there are fewer cycles for the FFT algorithm to attempt to characterize and, if acquisition is triggered during a period of relative calm, the response may not be visible. The overall noise floor of the data, on average ((
}}2(a),(c),ECI (PSD units), is higher than desirable for such measurements. This are evident when viewing combined spectral or PSD plots in which the resultant data for the datasets is overlaid. A reference example is provided in Figure 3-76 for several of the Column-3 sensors. As noted previously, this elevated noise is not unexpected given the type of strain gauges used. The increased noise floor does not impact the ability to characterize observable peaks, but obscures minor peaks (in particular for modes that are not excited by the flow conditions). This issue is compounded by the aforementioned short waveform duration.
pyright 2022 by NuScale Power, LLC 128
tables are fitted with data bars to facilitate quick visualization of datasets/channels that fall outside of the norm. Table 3-26 and Table 3-27 provide summarized overall 0-pk values by test series and primary flow rate, respectively. Average and maximum values are reported for each category and the test series or flow rate with the overall largest value for each channel is summarized at the bottom of the table. These tables help to illustrate whether specific parameters have a marked effect on overall dynamic content. The FFT outputs were used to create comparison spectral plots for the datasets and channels. The plots were grouped in two different fashions: by test condition and by primary flow rate. Additional notes are as follows: Datasets E6, F6a and F6b were excluded from these spectral plots. The primary- and secondary-side flow conditions for these datasets are similar to dataset G6, but the spectral data are not a direct match. These datasets appear to be outliers, because flow was not ramped in a similar fashion as test series A-D and G. Appendix B contains plots from 0-300 Hz arranged by test series, consistent with filters applied according to Run-1. Appendix C plots contain the same 0-300 Hz data arranged by primary flow rate. Data are also plotted from 0-1000 Hz, consistent with the bandpass filter applied according to the alternate parameters specified in Run-2 in Table 3-23. Appendix D contains the plots arranged by test series and Appendix E contains the plots arranged by primary flow rate. Datasets within test series G exhibited more content at electrical noise multiples (i.e., (( }}2(a),(c),ECI, and so on) as compared to prior recordings. It was not possible to filter out the electrical peaks without introducing additional spectral noise, so bandstop filters were limited to multiples between (( }}2(a),(c),ECI. Thus, the plots in Appendices D and E exhibit electrical noise peaks at higher multiples ((( }}2(a),(c),ECI and above), specifically for test series G. The spectral plots were reviewed for frequency content. Most datasets or channels exhibit content at frequencies not predicted within the assumed modal analysis results; in many cases, the amplitude of these peaks is larger than any within the predicted frequency bands. The most-common additional peaks were: 1) a low-frequency response (<10 Hz), 2) a series/group of peaks in the (( }}2(a),(c),ECI range, and 3) another group or set of fairly-sharp peaks in the (( }}2(a),(c),ECI range. Accordingly, the method shown in Equation 3-30 and its accompanying discussion are used to compute frequency-range-specific RMS values. The results of those computations are provided in tabular form in Appendix F. The frequency ranges that were evaluated are as follows: pyright 2022 by NuScale Power, LLC 129
(( }}2(a),(c),ECI: Tube breathing mode, plus large responses observed on several channels. Application of the bandpass filter ((( }}2(a),(c),ECI per Table 3-23) served to minimize content at frequencies below (( }}2(a),(c),ECI. (( }}2(a),(c),ECI: First-mode bending in horizontal plane, visible responses within data. (( }}2(a),(c),ECI: First-mode bending in vertical plane, visible responses within data. (( }}2(a),(c),ECI: visible responses within data, predominantly at (( }}2(a),(c),ECI (frequency and amplitude varies by channel and dataset). (( }}2(a),(c),ECI: visible responses within data. The observed significant frequencies (above) were compared with potential sources of known excitation within the system. Outside of electrical content, which is removed by filtering, the only other potential source of forced excitation would be vane passing frequency pulsations from the driving pump(s). There were no consistent peaks within the data at (( }}2(a),(c),ECI or multiples thereof. The data from Table 3-24 and Appendix F are combined in Table 3-28, to create a consolidated view of the various frequencies and amplitudes of interest for each dataset and channel. The table was generated by comparing the frequency-bin-specific RMS values from the tables in Appendix F to the overall RMS values from Table 3-24; most of the numbers in the table thus represent a percentage of the total signal energy. There is no technical basis for this calculation, as it compares values calculated from a single-sided frequency domain representation to those obtained directly from the raw time domain. As such, the numbers in the table should be treated as representative trends, not an exact measure of fractional energy. Conditional formatting is used to further delineate peaks. Table 3-29 and Table 3-30 mimic the consolidated view within Table 3-28, but with low-frequency content (0-10 Hz) excluded to highlight changes in response at the presumed tubing response modes. Table 3-29 is sorted by test series and dataset, while Table 3-30 presents the same data sorted by flow rate. The overall peak dynamic strains from Table 3-25 were plotted versus flow rate to illustrate escalation of content with increased primary flow. Figure 3-77 provides the resulting plot. Data labels are included for points with overall values greater than (( }}2(a),(c),ECI to help identify datasets where large responses were observed. The plot provides a graphical representation of non-linear escalation in tubing vibration with flow rate. If one or more resonant responses were present within the range of the TF-2 test series, whether driven by FEI, VS, or TB, it would be evident within Figure 3-77. pyright 2022 by NuScale Power, LLC 130
frequency-range-specific RMS computations in Appendix F. The resulting plots are provided in Appendix G. The low-frequency responses observed primarily within test series G (primary conditions at (( }}2(a),(c),ECI, secondary side boiling at (( }} 2(a),(c),ECI) were further evaluated by plotting those spectra against their corresponding datasets from test series D (((
}}2(a),(c),ECI). Figure 3-78 provides a comparison for Column-3 (Sensors S05-S08) for Datasets D5 and G5. Figure 3-79 provides a comparison for Column-5 (Sensors S09-S12) for Datasets D6 and G6.
From the data, it is evident that the peak frequency of interest is at or near the lower cutoff for the bandpass filter ((( }}2(a),(c),ECI), with the exact response varying somewhat by channel and dataset. The corresponding raw waveforms were superimposed on the spectral plots to further clarify the difference in signals between channels with the low-frequency response component. pyright 2022 by NuScale Power, LLC 131
ure 3-76 Overall Power Spectral Density Content Comparison - Column-3, Side Strain Gauges (Top=S07, Bottom=S05)
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 132
Table 3-24 Dynamic Root Mean Square Strains Measured During TF-2 Fluid Elastic Instability Tests
}}2(a),(c),ECI e: 1. The blue data bars reflect a common scale across datasets/channels, from 0 to the maximum value listed in the table ((( }}2(a),(c),ECI from dataset G4, channel S10).
pyright 2022 by NuScale Power, LLC 133
able 3-25 Dynamic Peak Strains Measured During TF-2 Fluid Elastic Instability Tests
}}2(a),(c),ECI e: 1. The blue data bars reflect a common scale across datasets/channels, from 0 to the maximum value listed in the table ((( }}2(a),(c),ECI from dataset G3, channel S06).
pyright 2022 by NuScale Power, LLC 134
Table 3-26 Dynamic Peak Strains by Test Series (Average and Maximum)
}}2(a),(c),ECI es: 1. Values in the table represent the average or maximum across datasets within the indicated test series.
- 2. The blue data bars represent a common scale across the average values, from 0 to the maximum value listed in the table ((( }}2(a),(c),ECI).
- 3. The yellow data bars represent a common scale across the maximum values, from 0 to the overall maximum listed in the table ((( }}2(a),(c),ECI).
- 4. The green data bars represent a common scale across the maximum values, from 0 to the overall maximum listed in the table ((( }}2(a),(c),ECI).
- 5. The summary rows at the bottom of the table indicate the test series during which the highest average, maximum, and standard deviation values were observed within individual datasets.
pyright 2022 by NuScale Power, LLC 135
Table 3-27 Dynamic Peak Strains by Flow Rate (Average and Maximum)
}}2(a),(c),ECI es: 1. Flow rates listed above represent an average of the values listed in Table 3-22 for applicable datasets. For example, the first value ((( }}2(a),(c),ECI is the average flow rate for Datasets A1, B1, C1 and D1; the last value ((( }}2(a),(c),ECI is the average for Datasets A6, B6, C6, D6 and G4.
- 2. Values in the table represent the average or maximum across datasets within the indicated test series.
- 3. The blue, yellow and green data bars are as-described in the footnotes to Table 3-26.
- 4. The summary rows at the bottom of the table indicate the flow rates at which the highest average, maximum, and standard deviation values were observed within individual datasets.
pyright 2022 by NuScale Power, LLC 136
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Table 3-28 Relative Comparison of Content within Frequency Ranges of Interest (0-10 Hz Inclusive), Sorted by Test Series ((
}}2(a),(c),ECI Notes: 1. Values listed in the "Overall" columns were obtained from Table 3-24 (i.e., in units of -RMS).
- 2. The values in the colored columns were obtained by dividing the frequency-bin-specific RMS energy by the overall value, and thus represent an effective percentage of total signal energy within the given spectral band(s). These values should not be treated as absolutes; they are for trending/comparison purposes only.
- 3. The conditional formatting/highlights are frequency-bin-specific across the channels. For example, within the (( }}2(a),(c),ECI bin, formatting is applied over the range from the minimum value ((( }}2(a),(c),ECI from dataset G6, channels S09 and S10) to the maximum value ((( }}2(a),(c),ECI from dataset C1, channel S11), which includes 310 values in total (31 datasets x 10 working channels).
- 4. The rule(s) for the applied conditional formatting were arrived at through several iterations of parameters, with final selection qualitative in nature. Conditional criteria are as follows:
Green = Mean - StdDev Yellow = Mean + StdDev Red = Mean + StdDev x 3 © Copyright 2022 by NuScale Power, LLC 140
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Table 3-29 Relative Comparison of Content within Frequency Ranges of Interest (Excludes 0-10 Hz), Sorted by Test Series ((
}}2(a),(c),ECI Notes: 1. Values listed in the "Overall ((( }}2(a),(c),ECI)" columns were obtained from Table F-7 (i.e., (( }}2(a),(c),ECI inclusive, in units of -RMS).
- 2. Other footnotes from Table 3-28 are applicable herein.
© Copyright 2022 by NuScale Power, LLC 141
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Table 3-30 Relative Comparison of Content within Frequency Ranges of Interest (excludes 0-10 Hz), Sorted by Test Series ((
}}2(a),(c),ECI Notes: 1. Values listed in the "Overall ((( }}2(a),(c),ECI)" columns were obtained from Table F-7 (i.e., (( }}2(a),(c),ECI inclusive, in units of -RMS).
- 2. Flow rates listed above represent an average of the values listed in Table 3-22 for applicable datasets, similar to the representation in Table 3-27.
- 3. Other footnotes from Table 3-28 are applicable herein.
© Copyright 2022 by NuScale Power, LLC 142
Figure 3-77 Scatter Plot of Peak Dynamic Strain Versus Flow Rate
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 143
igure 3-78 Spectral Comparison, 0-50 Hz, Datasets D5 Versus G5, Column-3 Sensors
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 144
igure 3-79 Spectral Comparison, 0-50 Hz, Datasets D6 Versus G6, Column-5 Sensors
}}2(a),(c),ECI Observations on the TF-2 data evaluation are summarized below:
Data is acquired at various primary and secondary-side conditions, as documented in Table 3-32. The length of each waveform recording ((( }}2(a),(c),ECI seconds) was insufficient to effectively characterize the frequencies of interest given the random variations present in typical FIV data. This concern is most-pronounced at low frequencies. Channels S03 and S04 (Column-1, Upper Location) were inoperable for the evaluated test data. The sensor locations with respect to the nearest supports (Table 3-21) are not ideal from the standpoint of measuring maximum bending strains; many of the sensors are near the neutral-strain point for a fixed-fixed or pinned-pinned beam. However, any significant response that is present (such as FEI) should still be plainly visible. A mid-span measurement of (( }}2(a),(c),ECI would represent approximately (( }}2(a),(c),ECI of displacement for a representative tube. This static deflection correlation may be extended to vibration at a primary driving frequency, such as that which would be expected from significant FEI excitation. As shown in Table 3-25, the peak value observed during FEI testing was pyright 2022 by NuScale Power, LLC 145
(( }}2(a),(c),ECI, and the average value was below (( }}2(a),(c),ECI. Furthermore, the reported values represent a composite overall and values at individual frequencies are much lower.The maximum observed single-frequency peak was (( }}2(a),(c),ECI and on average, the maximum across the datasets and channels was (( }}2(a),(c),ECI. Reported value applies for frequencies greater than (( }}2(a),(c),ECI. Responses at low frequencies exhibited significant variation between datasets and it is not possible to establish bounding amplitudes considering the short duration of the recorded waveforms. The low amplitude of individual frequency peaks versus composite overalls suggests that the primary excitation on the TF-2 tubes was broad-band in nature, presumably due to general flow turbulence or buffeting.
- The resultant value of (( }}2(a),(c),ECI within the example calculation corresponds to a Displacement/Diameter ratio of approximately
(( }}2(a),(c),ECI, or displacement of (( }}2(a),(c),ECI in metric units. Comparing these values to several of the aforementioned literature sources, a measured strain of (( }}2(a),(c),ECI peak falls at the extreme lower end of most published test data. RMS frequency spectra were generated to compare the responses on each individual channel with respect to test series or primary flow. The plots contained in Appendices B through E were reviewed for significant content and the following frequencies were consistently observed:
- { }}2(a),(c),ECI: low-frequency content that appeared to be present only during test series with secondary-side flow (i.e., C and G). Figure 3-78 and Figure 3-79 contain additional spectral comparisons for select datasets and sensors that exhibited responses in this realm.
- (( }}2(a),(c),ECI: presumed to be horizontal-plane, first-mode bending.
- (( }}2(a),(c),ECI: presumed to be first-mode bending about the vertical axis.
- (( }}2(a),(c),ECI: prominent group of peak(s) present in most datasets; does not specifically match an assumed modal response; may be an upper-order mode.
- (( }}2(a),(c),ECI: prominent peak present in many datasets; does not specifically match an assumed modal response; may be an upper-order mode.
None of the observed frequencies align with potential sources of mechanical excitation. Specifically, the tubing does not appear to contain content at the driving pump(s) vane passing frequencies. Therefore, the observed peaks were determined to be structural responses, indicative of first-mode bending, as well as upper-order complex modes. Appendices D and E provide spectra plotted to 1000 Hz. These plots were reviewed and compared against the PSD results from TF-1 to determine whether the notable secondary-side response between (( }}2(a),(c),ECI is also evident in the TF-2 data. Although several peaks are evident at specific pyright 2022 by NuScale Power, LLC 146
response such as that observed in the TF-1 data. It is concluded that the (( }}2(a),(c),ECI secondary-side response was either not present within TF-2, or present but not strong enough to induce a global structural response of the HCSG tubing. Frequency-bin-specific RMS values were computed (Appendix F) and collated (Table 3-28 and Table 3-29) to illustrate where variation exists in the data. Based on review of those tables:
- The spectral plots do contain evidence of the first-mode bending peaks for the tubing; namely, responses at approximately (( }}2(a),(c),ECI are visible in select plots. The amplitude of these responses is small, typically on the order of (( }}2(a),(c),ECI, indicating that large motions of the tubing center-spans is not occurring.
- The variation responses within the FEI frequency ranges of interest are somewhat random. In general, the data appear to illustrate a slightly-increasing trend with primary flow, which is expected. In certain cases, the maximum values occurred during datasets with less-than-maximum primary flow. This behavior can be observed from Table 3-30 in the response within the (( }}2(a),(c),ECI band for sensor S07.
- The low-frequency responses ((( }}2(a),(c),ECI) are only visibly significant for datasets with secondary-side flow, in particular for test series G (boiling conditions). This behavior is readily-apparent from Table 3-28, but also evident within the overalls (Table 3-24 through Table 3-27) and spectral comparisons (Appendices B and C). Figure 3-78 and Figure 3-79 further illustrate this phenomenon by comparing select datasets and sensors within test series G to corresponding data from test series D (stagnant secondary flow).
Review of the raw waveforms (the inlays in Figure 3-78 and Figure 3-79 contain examples) suggests that the observed response is a physical effect measured by the strain gauge sensors, not a signal anomaly or sensor malfunction. The effect is consistently apparent within the lower sensor locations (those closer to the FW plenum), which suggests that it may be a function of secondary-side excitation. The responses in question were most pronounced within test series G (secondary-side boiling), even compared to test series C (single-phase flow, roughly six times the mass rate as series G), suggesting that the effect is exacerbated by elevated temperatures or fluid phase changes. The truncated (( }}2(a),(c),ECI recordings are insufficient to allow for consistent characterization of the low-frequency responses. For example, it is unclear whether the large spike observed in S05/S06/S09/S10 near the start of both the G5 and G6 datasets is a one-time or repeating event. Based on the limited data available, it appears that the observed response was generally between ((
}}2(a),(c),ECI.
pyright 2022 by NuScale Power, LLC 147
(( }}2(a),(c),ECI: the fundamental breathing mode of the helix is (( }}2(a),(c),ECI, which is not expected to be excited by primary-fluid flow. Thus, the observed low-frequency response is likely a combined structural mode of the TF-2 tubing and support configuration. Based on the data evaluation herein, including the observations above, FEI excitation of the HCSG tubes did not occur throughout the entirety of TF-2 FEI testing. This conclusion is largely based on the lack of amplification of responses within the assumed primary bending mode frequency ranges and is further reinforced by the low levels measured across the entire frequency span of interest. If FEI were present a significant increase or emergence of sharp peaks in the first-mode bending frequency range ((( }}2(a),(c),ECI) is expected. Table 3-29 shows that the actual FEI data contain no such amplification; values generally trend upward with flow, but in some cases even decrease at the maximum primary flow rates. Figure 3-77 plots the data from TF-2 and it is clear that an exponential increase in vibration levels at higher flow rates was not observed, as would be the case if the critical velocity were reached. Furthermore, the points at which elevated vibration levels were observed at elevated flow rates can be correlated to the low-frequency, secondary-side effect noted above. Even after accounting for a sub-optimal test setup (TF-2 size limited to Columns 1-5, i.e., shortest spans, (( }}2(a),(c),ECI recording length, and sensor locations), the lack of first-mode response escalation indicates the HCSG design is not susceptible to FEI excitation within its normal operating parameters. TF-3 Build-out Modal Testing For CVAP validation testing, there is a partially prototypic test facility (TF-3) with five tube columns of identical geometry as Columns 9 through 13 of the functionally prototypic design. At the time of the build-out modal testing, TF-3 was partially constructed, with two tube columns complete: Column-13 (outermost) and Column-12. Each column is divided into four plenums, each of which contains sixteen 304/316 stainless steel tubes (64 total tubes per column). The spacing and overlap of the tubes are such that access to the Column-13 tubes is extremely limited; therefore, the in-situ testing was focused on the accessible Column-12 tubes. The HCSG was positioned in a horizontal orientation that could be manipulated to adjust the relative position and azimuth of the tubes, plenums, and supports. Each tube has a (( }}2(a),(c),ECI OD and (( }}2(a),(c),ECI wall thickness. Tube spans between supports for Column-12 are approximately (( }}2(a),(c),ECI (long span) and (( }}2(a),(c),ECI (short span), with an approximate tubing weight of (( }}2(a),(c),ECI. Testing was targeted at the accessible Column-12 tubes, ranging from the bottom of the HCSG (denoted Span A, FW inlet) to the top (Span AD, MS outlet). The in-situ testing was primarily a discovery task, intended to inform approach and parameters for future regimented testing. As such, emphasis was placed on testing a multitude of span, support, and sensor variations, with less emphasis on repeatability and like-for-like pyright 2022 by NuScale Power, LLC 148
accurate determination of the following variables:
- 1. Modal parameter estimations including
- a. Natural frequency
- b. Damping
- c. Mode shape(s)
- 2. Excitation methods
- 3. Effects of boundary conditions
- 4. Measurement and signal fidelity A general overview of the HCSG test setup is shown in Figure 3-80, with annotations detailing terminology used throughout the remainder of this report. The sensing chain used for acquisition and analysis is documented in Table H-1.
The general nomenclature used for the test IDs includes a group to identify the sensor(s) azimuth and plenum (accounting for rotation of the pressure vessel), spans, identified as segments between supports (starting with Span A/FW plenum and progressing through Span AD/MS plenum), and the general impact location or direction. These references were modified slightly throughout testing to capture varying conditions and lessons learned, and are specifically denoted in the notes below the tables and figures herein. The individual tests were categorized into six groups by the unique test configuration. These configurations represent unique tubes, spans, or components that were tested and are summarized by the following: A) Single Spans - Single Accelerometer B) Span C - (5) Accelerometers C) Multi-span (C through G) - Single Accelerometer per Span D) Plenum transitions Span A (FW-side) and Span AD (MS-side) - Multiple Accelerometers E) Support and Tube Testing F) Support Only Note: the testing conducted as part of Group E included characterization of responses from both support(s) and tubing. Upon further analysis, these responses offered sufficient characterization of lower-frequency modes impacting the supports, such that the Group F tests could be regarded as redundant. Therefore, the Group F data are not analyzed further. pyright 2022 by NuScale Power, LLC 149
Figure 3-80 General Layout of Helical Coil Steam Generator Prototype for Vibration Testing
}}2(a),(c),ECI 1 Data Acquisition and Test Methods Modal testing to characterize the HCSG tubing and support parameters was conducted on impulse excitation from an instrumented hammer with various accelerometer configurations to measure the ensuing response(s). The vibration data are recorded primarily in the frequency domain and for select tests, in the time domain. The following acquisition parameters were used:
Table 3-31 Acquisition Parameters for Time and Frequency Domain Domain No. Samples Resolution Ending Time Linear Averages (( Time (sec) requency (Hz)
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 150
input was applied. The force window value was assigned unity over the leading five percent of the time record following the initial impulse trigger, followed by a cosine taper to zero for the remainder of the time record (eliminating noise). An exponential window was also applied and set to unity (identical to that applied to the response accelerometers) for the frequency domain calculations. Because the exponential window was set to 1.0 at the end of the acquisition (i.e., no amplitude effect), no corrections for damping are needed as the force/input signal was unaltered for the period of application. Frequency response functions (FRFs) were calculated by the Fourier spectrum of the acceleration response divided by the Fourier spectrum of the excitation force (impact hammer). Four FRF averages from four separate impacts were used to calculate these results; this number of averages was determined experimentally, and represents the best compromise for maintaining consistency of response peaks while reducing noise at non- responsive frequencies. The acquisition duration (typically 4 seconds) and tube damping resulted in a reduction of amplitude to (( }}2(a),(c),ECI of the maximum imparted response. This amplitude reduction is sufficient to reduce window leakage in FFT computations. Leakage is a signal processing bias error due to the limited definition of a periodic waveform or transient over the sampling period. The error is leaked across spectral lines and over the entire frequency range representing noise and reducing signal-to-noise ratio of the measurement. Given the limited time available, windowing was not completely optimized during this initial testing. A soft hammer tip with a mass extender was used for excitation. The soft tip is used to spread the force energy (pulse width) over a longer time and excite lower frequencies. The combination of hammer tip and impulse energy (mass extender, impact velocity) describe the force input used for modal parameter estimation. Any changes to these variables affects modal parameter estimation. Spectral coherence was calculated for each FRF, estimating the relation between two signals (input/hammer vs. the response). These values range from zero to one over the frequency range of interest, where one represents a perfect correlation. Values less than one can be attributed to several factors: Anti-resonance locations or locations where the FRF value is close to zero Resonance locations (FRF peaks) where the effect of leakage is pronounced. Note windowing reduces leakage and does not eliminate it. The measured response contains contributions caused by extraneous noise, non-linearity, or other forces not contained in the measurement input. Impact locations/directions were slightly altered between averages. The coherence metric was used to define various useable frequency ranges for the analysis groups; generally, values greater than (( }}2(a),(c),ECI were considered acceptable. pyright 2022 by NuScale Power, LLC 151
structure properties. However, the tube and HCSG support structure has a combination of local and global modes. The local modes or those specific to a span have difficulty transferring across a sufficiently stiff boundary. These modes typically require local excitation and measurement to characterize, similar to the single span testing. For multi-span tube and component testing, the stiffness of this boundary or tightness of supports change between each span and each support from fixed to relatively free. This change results in varying levels of energy transfer across these boundaries and the measurement of both local and global modes between supports, tubes, and spans. Mass loading of the structure can affect modal parameter estimation. The weight of an individual accelerometer used for this testing is (( }}2(a),(c),ECI ounces and the cable is (( }} 2(a),(c),ECI (not including conductor). This compares to about (( }}2(a),(c),ECI percent mass loading across a long span for a single accelerometer (most tests herein) and between (( }}2(a),(c),ECI percent for five accelerometers (depending on the cable length left suspended). This effect is discussed further when using multiple accelerometers to characterize a mode. 2 Analytically Predicted Modes A summary of modal predictions for Columns 9 and 13 (innermost and outermost of the TF) is provided below for comparison to the results presented herein. The lower end of the range represents the frequency for Column 9 and the upper end of the range is the frequency for Column 13. Highest mass participation in the vertical direction
- (( }}2(a),(c),ECI (sliding boundary conditions)
- (( }}2(a),(c),ECI (fixed boundary conditions)
Highest mass participation in the horizontal direction
- (( }}2(a),(c),ECI (sliding boundary conditions)
- (( }}2(a),(c),ECI (fixed boundary conditions)
For the in-situ testing that was conducted, there is good agreement for the first predominant mode in the vertical direction, which is more closely approximated by a fixed boundary condition (generally reported herein in the (( }}2(a),(c),ECI range). Similarly, the predominant horizontal mode from testing was most often present in the (( }}2(a),(c),ECI range. 3 Frequency Response Function Analysis Table H-1 summarizes testing parameters, sensor locations, and configurations used during the acquisition. FRF calculations from 0-400 Hz for each group of tests is provided in Figure H-1 through Figure H-31 of Appendix H. The following sub-sections discuss the FRF results for each of the groups. pyright 2022 by NuScale Power, LLC 152
Several initial tests were conducted on individual tubes, altering various acquisition parameters to identify an acceptable set for repeat testing. A typical acquisition is as demonstrated in Figure 3-81 for these tests. A subset of tests was selected for this group to illustrate the analysis over a simplified number of responses. A single long span (Span C, first long span after FW plenum) was tested, with the resulting FRF shown in Figure 3-81. The FRF and coherence is overlaid to demonstrate a spurious response (highlighted in yellow) along with the evident response peaks ((( }}2(a),(c),ECI, highlighted in gray). The FRF response is also slightly skewed, especially at the lower frequency peak. Both effects are indications of non-linearity. This skewing is also likely an effect of multiple closely-spaced modes, as detailed in Group B test responses (discussed more in the following section). Another test was conducted to enhance the resolution of mode shapes using a roving response test and Maxwells reciprocity. Twelve roving FRF measurements along a single span are shown in Figure 3-82 using a single impact location. The use of reciprocity and roving to define a mode shape assumes the structure is dynamically symmetric. The combination of mass loading (albeit to a lower extent) and non-linearity negated the principals assumed for reciprocity, with the peaks shifting between measurements. For this reason, roving response tests were not used for further analysis, and are not recommended for subsequent testing. Use of non-tubing impact excitation locations (e.g., supports) and reduced acquisition periods (i.e., one second) allowed for further characterization of lower frequency modes. The reduction in the acquisition period (from ((
}}2(a),(c),ECI) reduces the averaged noise (if present) in the signals and computed FRFs, but also adversely affects the frequency resolution (increased bin sizes from (( }}2(a),(c),ECI), which, in turn, affects the accuracy of damping calculations in the frequency domain such as half-power. As shown in Figure 3-83, a (( }}2(a),(c),ECI mode (normally where low coherence prevents characterization) was captured using a smaller time window to reduce the noise following the transient. It is not clear if the removal of the ring supports for this test also affected the improved coherence at lower frequencies. The discussion of Group E also describes the lower modes excited from support locations.
The variation of responses along spans is evident throughout the testing where various modes became difficult to excite repeatedly. Variation was mainly attributed to the boundary conditions (HCSG support tightness ranged from fixed to some variety of sliding), impact energy (direction/magnitude), and non-linearity. Not all modes were excited and in general the lower frequency modes (below (( }}2(a),(c),ECI) were the most difficult to excite consistently. pyright 2022 by NuScale Power, LLC 153
ure 3-81 Typical Single, Long Span Vertical; Frequency Response Function (Blue) and Coherence (Orange)
}}2(a),(c),ECI gure 3-82 Frequency Response Function Response for Roving Accelerometer Along Single Span (Mass Loading)
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 154
igure 3-83 Lower Frequency Response (Below 25Hz); Frequency Response Function (Blue) and Coherence (Orange)
}}2(a),(c),ECI Group B: Span C - Five Accelerometers This group of tests entailed instrumentation of one span (Span C, first long span after FW plenum) with five equally-spaced accelerometers. The span was rotated during the tests from a bottom-azimuth orientation (4:00-5:00), to the side (2:30-3:30), and the top (11:30 to 12:30). The primary purpose of this test group was to determine the effects of varying boundary conditions based on vessel rotation/azimuth (i.e.,
compression applied by support rings), and allow for improved mode shape characterization by multiple accelerometer locations. Generally, there were three frequency response ranges of interest: ((
}}2(a),(c),ECI. Rotation of the RPV or HCSG caused a slight shift among the responses. Shifts within the FRF peaks were mostly bounded within (( }}2(a),(c),ECI (depending on the mode). The lower frequency modes (between (( }}2(a),(c),ECI) were predominantly in the vertical direction, while the largest-amplitude FRF responses were in the horizontal direction (between
(( }}2(a),(c),ECI). The location of the impact (near each end and mid-span A summary of the FRF results during this testing are shown in Table 3-32 and the FRFs are plotted for each test in Appendix H, Figure H-3 through Figure H-13. Impact locations mid-span produced improved characterization for the first predominant mode and impact locations near the ends of the span offered the best characterization across the upper modes. The latter is demonstrated by a mid-span and end-impact for tests 1A-3Z and 1A-5Z. Additionally, the direction of the impact improves the characterization for predominant horizontal and vertical modes, pyright 2022 by NuScale Power, LLC 155
characterization is a result of a majority of the overall energy imparted into the tube directed along the impact axis (e.g., y-vertical, z-horizontal), resulting in a higher signal-to-noise ratio of the responses for modes predominant in the impact direction. The coherence between the input and output were generally calculated above 0.9 over the frequency range of (( }}2(a),(c),ECI as shown in Figure 3-84 (below). Table 3-32 Group B, Span C (5) Accelerometers
}}2(a),(c),ECI s: 1. Test ID references azimuth of Span C, 1A 6:00, 1C-3:00, 1E-12:00-sensor (1, 2, 3, 4, 5) impacted and direction (y-vertical, z-horizontal).
pyright 2022 by NuScale Power, LLC 156
Figure 3-84 Coherence Only for Test 1C-1Z (Horizontal/1Z and Vertical/3Y Direction)
}}2(a),(c),ECI ure 3-85 Impact Location Effects on Span C (Left: Mid-Span; Right: Near End of Span)
}}2(a),(c),ECI Note for Figure 3-85: Both impacts measured all five locations and the impact was imparted in the horizontal direction at two locations, midspan (location 3) and near the support (sensor location 5).
Group C: Multi-span (C through G) - Single Accelerometer per Span pyright 2022 by NuScale Power, LLC 157
E, F, and G), each with a single accelerometer mounted mid-span. Spans D and F represent a short span ((( }}2(a),(c),ECI) whereas Spans C, E, and G are long spans ((( }}2(a),(c),ECI). The set of spans was also rotated from an azimuth where Span C was orientated at 06:00 (BDC) and Span G was at 12:00 (TDC) to where Span C was oriented at 09:00 and Span G was at 03:00. The primary purpose of this group of testing was to determine the effects of changing boundary conditions through rotation and the variations in local modal parameters between like spans. Generally, two modes were present in the first orientation ((
}}2(a),(c),ECI. Excitation was limited to the longer tubes with no consistent presence or dominating frequency between the long spans and short spans. These frequencies increased (( }}2(a),(c),ECI during rotation elevating to ((
}}2(a),(c),ECI . Both lower modes were clearly responding for each long span with the lower mode mostly clearly represented from Span E impacts. It is unclear if the two modes were influenced by boundary conditions as the end conditions were quite similar among the spans. The lower modes ((( }}2(a),(c),ECI) were predominant in the vertical direction for this testing.
The next set of predominant modes was at (( }}2(a),(c),ECI; these also shifted by 2 Hz after the rotation. The last set of modes near (( }}2(a),(c),ECI (predominant within a group of closely spaced modes) was shifted by approximately (( }}2(a),(c),ECI upon rotation. The mid-span location demonstrated primarily vertical predominant modes over these upper frequency ranges. A separate test was also conducted within Span E (3C-EmsZ), where tubes 2, 6, 10, and 14 were tested with single mid-span accelerometers, impacting tube 6. The first mode frequencies ranged from (( }}2(a),(c),ECI. A single tube excitation was effective in exciting the first mode in adjacent tubes, albeit with elevated noise and elevated modes were no longer discernable. A summary of the FRF results during this testing is shown in Table 3-33 and the FRFs are plotted for each test in Appendix H, Figure H-13 through Figure H-21. pyright 2022 by NuScale Power, LLC 158
Table 3-33 Group C, Multi-Span
}}2(a),(c),ECI s: 1. Test ID references "azimuth of Spans, 2A (C-12:00, G-6:00),
2C (C-9:00, G-3:00), 3C (E-12:00)"-"Impacted Span on FW or MS side of span in the z-horizontal direction)." pyright 2022 by NuScale Power, LLC 159
(( }}2(a),(c),ECI over the frequency range of (( }}2(a),(c),ECI, as exemplified by Figure 3-86. Figure 3-86 Coherence Only for Test 2A-EfwZ (Vertical and Horizontal Direction)
}}2(a),(c),ECI The majority of FRF peaks are present not only in one FRF or span, but across each of the individual spans, even though each span exhibits localized modes (long and short spans). The energy transfer across spans (cross-communication) was investigated using both the FRF magnitude and coherence of the input (hammer) and response signals on each span. From a single test impacting the center span of Span E, two modes were excited ((( }}2(a),(c),ECI). From this test alone, it is unclear which mode is local to Span E or if both are local. As shown in Table 3-34, the first predominant mode at (( }}2(a),(c),ECI is dominant in Span E (impact location) at more than five times the response (g/lbf) of the other responses, but the peak is present in three out of four of the adjacent spans at a reduced magnitude.
This FRF peak occurs on similar span lengths (i.e., Spans C, E, G) at ((
}}2(a),(c),ECI of the impacted spans FRF response magnitude, which suggests a similar local mode is present along the spans although not properly excited. For the (( }}2(a),(c),ECI peak there is clearly amplification on Span G (FRF magnitude six times that of Span E) and very slight amplification on Span C (albeit with low coherence) indicating a clear resonance near this frequency for Span G. The presence of local modes specific to a span and the identification of modes cross-communicating from adjacent spans is only clear when analyzing the FRFs across multiple spans in a single test. The ambiguity in response (i.e., multiple closely spaced peak responses at each sensor) is more clearly identified by local exaction to each span while measuring adjacent spans.
pyright 2022 by NuScale Power, LLC 160
locations (Span E, top left; Span C, top right; Span G, bottom right). All five span responses are plotted for each impact (C, D, E, F, and G) varying by color. Coherence is plotted for the Span E impact test (lower left). From these tests it becomes apparent there are multiple closely-spaced modes between (( }}2(a),(c),ECI which are not easily distinguished in these tests. For Span E, there is a predominant mode at (( }}2(a),(c),ECI. For Span G, the predominant mode is at ((
}}2(a),(c),ECI. A comparison of boundary conditions between spans does not explain the differences in the predominant mode responses. The HCSG supports are relatively the same tightness for Span C and Span E (cumulative relative tightness of 5), while Span G was slightly looser (cumulative tightness of 6. Each end of the span was tested by hand to determine the relative tightness of the support on a scale of 0-5, where 0 was nearly a fixed boundary and 5 was very loose. The algebraic sum from each end of the span is reported here as the cumulative tightness.). Although clear peaks in the FRF response can be seen from Spans E and G (whether they were impacted directly or not), the Span C response was much less clear or pronounced and even when impacted directly (upper right plot, Figure 3-87), poor responses were measured. This span would require additional testing to fully characterize the FRF response and it appears the span was not properly excited with cross-communication from modes on Span E and Span G dominating the response.
In addition, modes below 140 Hz were not excited on short spans (D and F), which may be because of lack of excitation local to each span (i.e., impacts were only completed on long spans). Table 3-34 Energy Transfer across Spans
}}2(a),(c),ECI es: 1. Impact location was on FW side of Span E with responses measured across five spans (C, D, E, F, G) by single mid-span accelerometer. FRF magnitude and coherence are provided for this test across two peak FRF responses at
(( }}2(a),(c),ECI. The input/output ratio is the ratio of FRF amplitude from Span E (closest to excitation source) to adjacent span FRF responses to quantify amplification and attenuation. pyright 2022 by NuScale Power, LLC 161
Figure 3-87 A-EfwZ Energy Transfer across Spans
}}2(a),(c),ECI Group D: End Spans (A and AD) - Multiple Accelerometers This group of tests characterized the response of the end spans (transition bends) on both the top and bottom of the HCSG (FW and steam plenums), for the shortest and longest tubes within each of these spans. These spans are fixed on one end (plenum orifice) and supported by the HCSG supports on the other end. The tubes were excited by a hammer impulse on each tube and on the HCSG support. Predominant measured frequencies are summarized in Table 3-35.
pyright 2022 by NuScale Power, LLC 162
Table 3-35 End Span (A and AD) Frequency Response Function Summary
}}2(a),(c),ECI s: 1. Test ID reference indicates "Span (4A-spanA and 5A-span AD)-"impact sensor and direction." Two sensors were placed on the shorter tube (1) and three sensors were placed on the longer tube span (15 or 16).
Comparing the short sections at each end of the HCSG, the FW side (4A-2Z) shows the predominant responses in the (( }}2(a),(c),ECI range, whereas the MS side (5A-2Z) shows the first set of predominant frequencies lower ((( }}2(a),(c),ECI range). This result is likely a cause of the boundary conditions where one end is fixed and the other simply-supported end is tighter on the FW side when compared to the MS side. Additionally, the unsupported length of each tube at the transition to the plenums is different, leading to differences in the measured frequencies. A predominant response from the support was also measured at (( }}2(a),(c),ECI and present on the shorter tube. Using a separate sensor affixed to the support (#6, Test 5A-Support6Z) and impacting that support, the (( }}2(a),(c),ECI response was clearly evident on the support sensor and through cross-communication was also observed on the shorter tube response (sensors 1 and 2). Excitation of the support and tube resulted in similar responses in the lower frequency range ((( }}2(a),(c),ECI); responses were muted above those frequencies when not directly excited at the tube. The fixed-end on these spans offered very clean responses and high signal-to-noise ratios. All modes were excited in the horizontal direction, which also corresponded to the largest energy response. Coherence pyright 2022 by NuScale Power, LLC 163
between the signals was generally above (( }}2(a),(c),ECI over the frequencies (( }} 2(a),(c),ECI as shown in Figure 3-88 (below). This coherence plot represents the best input/output relationship among the tests conducted, with a value of nearly 1.0 across the entire frequency range. Also, the plot exhibits little evidence of noise and cross-communication at FRF peaks, and a much more deterministic response below (( }2(a),(c). Figure 3-88 Coherence Only 5A-5Z (Horizontal and Vertical Direction)
}}2(a),(c),ECI Notes:1.Vertical coherence (4Y, midspan of tube 16) compared with the impact hammer (999Z) and horizontal coherence (5Z, near the first support) compared with impact hammer (999Z).
Group E: Support and Tube Testing Group E entailed testing of the HCSG support and tubing as one structure. Seven equally-spaced accelerometers were placed along the vertical axis of a support and mid-span accelerometers were placed on adjacent tubes. This group of tests is unique from other groups in that the support was impacted and used to excite modes within the tubes. In addition, the boundary conditions were unique in that the seven support rings were removed for these tests. The testing configuration is shown in Figure 3-89. Four tests were conducted, first measuring three adjacent tubes (1, 5, 9) impacting the support in the axial (tube direction) and the horizontal (z-direction). The next two tests instrumented tubes 1, 2, 5, 6, 9, 10, 13, and 14 representing the majority of the sixteen tubes across Span E. In addition, a larger hammer (1 mV/lbf compared to the pyright 2022 by NuScale Power, LLC 164
FRF results is presented in Table 3-36 for the four tests in this group. The signal-to-noise ratio decreased for this testing as the energy transferred through the support to the tubes was as much as two orders of magnitude less than other FRF peaks (FRF peaks on the order of (( }}2(a),(c),ECI). A relatively weak axial tube mode (lowest FRF magnitude of Group E) was detected at (( }}2(a),(c),ECI Hz for the initial axial impact direction (Sup-103 + Col-12, Span E Tub 1+5+9, Impact 1-X) and became less prominent in the second test in the same impact direction (Supp-to-Tube 103 Span E FW-X ) due to elevated noise in the lower frequencies. The FRF magnitude was similar between the support and tube response indicating a global mode. When impacting in the horizontal direction a (( }}2(a),(c),ECI mode was excited again in the axial direction (Sup-103 + Col-12, Span E Tub 1+5+9, Impact 1-Z) and present in nearly all measurement directions (likely a local mode from the support). In general, axial and horizontal modes were detected at (( }}2(a),(c),ECI across the tests most likely originating as local modes from the support because there was little to no variance for each tube and the magnitude of the FRF response were largest on the support (except for (( }}2(a),(c),ECI). Most of the tubes tested exhibited predominant vertical modes between (( }}2(a),(c),ECI with some outliers detected up to (( }} 2(a),(c),ECI . Coherence was above (( }}2(a),(c),ECI over the lower frequencies mostly due to the different method of excitation, as shown in Figure 3-90. pyright 2022 by NuScale Power, LLC 165
Figure 3-89 Support/Tube Testing Configuration
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 166
Table 3-36 Support and Tube Frequency Response Function Summary
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 167
Figure 3-90 Coherence for Test Sup-103 + Col-12, Span E Tub 1+5+9, Impact 1-X
}}2(a),(c),ECI e:1. Provided in axial (901x-direction), vertical (901y-direction) and horizontal (901z-direction) 4 Damping Estimation 4.1 Half-power Method for Damping Estimations A global polynomial curve fit was used to approximate damping using the half-power method. The curve fit was applied over multiple FRFs (all valid sensors per impact test) to calculate a global fit per measurement direction. The fit was applied over discrete frequencies to increase the accuracy of the fit function as shown in Figure 3-91. Figure 3-91 represents the FRF magnitude (g/lbf) in the top plot and the bottom is the imaginary part (log-scale) of the FRF used to identify peaks for fitting. Due to noise within the FRFs, a filter was applied to report damping values greater than (( }}2(a),(c),ECI. The removal of those peaks shifted the average value from (( }}2(a),(c),ECI.
Damping estimations were calculated for each group and represent various FRF peaks. These are plotted in Figure 3-92. The range of damping values is (( }}2(a),(c),ECI. Multiple closely-spaced peaks coupled with non-linear FRF responses resulted in broadening FRF peaks used to calculate damping with the half-power method. This effect is well demonstrated for peak damping values of (( }}2(a),(c),ECI in Figure 3-93. Based upon the 250 damping values, a 95 percent confidence interval describing the average damping over the reported frequency range using the half-power method is expected to be between (( }}2(a),(c),ECI. pyright 2022 by NuScale Power, LLC 168
average) from the tube would be expected to fall between ((
}}2(a),(c),ECI. Because more impact hammer tests were performed than pull tests, these statistics are skewed towards representing damping due to small amplitude vibration (TB) rather than the larger amplitude vibration expected from strongly-coupled phenomena such as VS and FEI.
Figure 3-91 Global Polynomial Curve Fit of Frequency Response Functions (Top equency Response Function Magnitude, Bottom Imaginary Magnitude on Log-Scale)
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 169
igure 3-92 Half-Power Damping Estimations over the Frequency Response Function Groups
}}2(a),(c),ECI Figure 3-93 Peak Damping Values ((( }}2(a),(c),ECI:1C-1Y-Left and
(( }}2(a),(c),ECI :1A-3Z-Right)
}}2(a),(c),ECI 4.2 Exponential and Log-decrement Methods In addition of the half-power methods, the exponential and log-decrement methods are also assessed.
Exponential Fit An exponential curve is fitted to subsets of the 45 positive peaks. The exponential
- ne t fit assumes a decay of the positive peaks in the form of e where is the pyright 2022 by NuScale Power, LLC 170
damping ratio, ne is the estimated natural frequency of the considered data, and t is time. Subsets of the positive peaks are used to calculate the exponents coefficient. The subsets include positive peaks in groups of 5, 9, 13, 17, and 21 positive peaks per data block. These are referred to as "runs" in the following text. For each of these runs, an exponential fit is performed starting with the first peak (i.e., 1 of 45) and the exponential coefficient (- ne ) recorded. The run is then shifted to the second positive peak (i.e., 2 of 45) to calculate a new exponential coefficient, then the third peak, and so on. The calculated exponential coefficient is divided by the runs corresponding natural frequency estimate to determine a damping ratio estimate. Because the natural frequency may change with amplitude the natural frequency estimate for each run is calculated as 2 ( N Cycles ) ne = ------------------------------------------------------------------------------
- Equation 3-1 Time Peak(Final) - Time Peak(Initial)
By developing a damping ratio estimate at various initial peaks, an amplitude dependent damping ratio estimate can be obtained. Logarithmic Decrement The logarithmic decrement is calculated as: 1 x i
= ---- ln ------------ - Equation 3-2 N x i + N where is the logarithmic decrement, x i is a maxima free vibration displacement amplitude (i.e., one of the 45 positive peaks), and x i + N represents the maxima free vibration displacement amplitude measured N periods from x i . The damping ratio is then written as:
= ------------------------ Equation 3-3 2 2 4 +
The damping ratio is calculated for an array of initial peaks and subsequent peaks (i.e., the Ns). The exponential fit and logarithmic decrement damping estimates from three processed displacement time histories are overlaid and shown in Figure 3-94. The three displacement time histories come from the 315-12-1-C-6 (5y) time ((( }}2(a),(c),ECI filter), 2A-EfwZt_8Y ((( }}2(a),(c),ECI filter), and 2A-GfwZt_10Y ((( }} 2(a),(c),ECI filter) data sets. Note the frequency range of interest for the three data sets centered around (( }}2(a),(c),ECI. pyright 2022 by NuScale Power, LLC 171
(( }}2(a),(c),ECI. The majority of oscillations used for damping calculations are less than (( }}2(a),(c),ECI. These amplitude ranges are expected to be most representative of the TB mechanism. The data shown in Figure 3-94 are the raw data from both the exponential fit and logarithmic decrement. As such, the data contain various N values because each N value for a time history is associated with an amplitude value. The damping values in Figure 3-94 that are significantly above or below the mean (Figure 3-95) are generally due to low N values. The negative damping values in both the exponential fit and logarithmic decrement methods (Figure 3-94, Figure 3-96, and Figure 3-97) are attributed to filtering, which resulted in positive peaks having lower displacement amplitudes than the subsequent peaks. The damping ratios greater than (( }}2(a),(c),ECI correspond to the amplitudes associated with N=1 and N=2. ure 3-94 Amplitude-Dependent Damping from Three Time Histories (Raw): Exponential Fit - Blue; Log Dec - Red
}}2(a),(c),ECI 4.3 Amplitude Dependency The three vibration mechanisms that are prone to the HCSG tube design are buffeting, VS, and FEI. These mechanisms are expected to manifest at differing pyright 2022 by NuScale Power, LLC 172
expected smallest amplitude to the largest: TB, VS, FEI. Amplitude dependency was not a specific focus of the initial testing. The various tests were post-processed into displacement units and a best effort was applied to evaluate the effect of amplitude dependency. Figure 3-95 shows a processed version of the data from Figure 3-94 where the damping estimates at each amplitude are averaged and then plotted. For example, the 20 logarithmic decrement damping estimates for a given displacement amplitude are averaged together and the average is plotted. There is a clear trend of damping increasing with response amplitude for both methods. Figure 3-95 Amplitude-Dependent Damping from Three Time Histories (Averaged): Exponential Fit - Blue; Log Dec - Red
}}2(a),(c),ECI While most of the data are collected with an impact hammer one pull test was performed, and the output recorded. The tube was physically displaced by hand, released, and allowed to undergo free vibration.
A plot of displacement amplitude versus damping using the exponential fit and logarithmic decrement methods are shown in Figure 3-96. Figure 3-97 overlays the damping data from the pull test to the damping data to the three previously discussed time histories. Peak displacement responses used for damping are roughly (( }}2(a),(c),ECI mils and represent the largest damping values pyright 2022 by NuScale Power, LLC 173
calculated, up to (( }}2(a),(c),ECI. Figure 3-97 shows that even neglecting the first (( }}2(a),(c),ECI positive peak data points from the pull test (starting near (( }}2(a),(c),ECI), the pull test indicates more damping than the three tests noted above. gure 3-96 Amplitude-Dependent Damping From Pull Test: Exponential Fit - Blue; Log Dec - Red
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 174
gure 3-97 Amplitude-Dependent Damping: Pull Pest - Black and Grey; Impact Tests - Red and Blue
}}2(a),(c),ECI 5 Mode Shapes A geometry file of a single tube and HCSG supports was used to evaluate specific mode shapes for Group B (Span C). Group B was chosen because it was unique in providing five measurement locations to fully define mode shapes, whereas other tests specifically focused on FRF responses with a single mid-span location. An overview of the five locations and FRFs used is provided in Figure 3-98. For the mode shapes, points between the measurements were interpolated and the points on the end ((( }}2(a),(c),ECI) were fixed. There are three frequency regimes of interest near (( }}2(a),(c),ECI. The first mode analyzed is predominant in the vertical direction. The second and third modes are most predominant in the horizontal directions while still contributing slightly in the vertical direction.
The first predominant mode is represented as multiple, closely-spaced peaks within in the FRF. Two modes were chosen to evaluate the mode shapes: 1) (( }}2(a),(c),ECI representing a broad smooth peak, and 2) (( }} 2(a),(c),ECI representing a sharp FRF peak. The mode shapes (shown both deflected and non-deflected) are provided in Figure 3-99 and Figure 3-100. These figures show very similar mode shapes between the closely-spaced modes predominant motion in the vertical directions and much less pronounced motion pyright 2022 by NuScale Power, LLC 175
modes most closely represent the first bending mode of a simply supported or fixed-beam. The second predominant mode near (( }}2(a),(c),ECI represents the largest FRF response. Again, multiple peaks are present near this mode as closely-spaced modes. A predominant horizontal mode is shown in Figure 3-101. This mode is best characterized as a second bending mode of a beam with the mid-point relatively motionless. The third predominant mode near (( }}2(a),(c),ECI is one of many peaks between (( }}2(a),(c),ECI. This peak was chosen consistent with its relative response in all directions. The predominant horizontal mode shape is shown in Figure 3-102. The mode is best characterized as a third bending mode of a beam with two inflection points. Because a single test (1C-1Z) was used to develop the mode shapes, a comparison to another test (1A-5Z) was used to quantify variability. Figure 3-103 shows two mode shapes for qualitative comparison. These shapes are very similar even at two frequencies and two different tests ((( }}2(a),(c),ECI). A quantitative evaluation using a modal assurance criterion (MAC) value demonstrates very consistent mode shapes (MAC value (( }}2(a),(c),ECI). A MAC value is a statistical indicator for correlating the complex vectors for nodes pairs between two tests (amplitude and phase). The indicator is most sensitive to large differences and relatively insensitive to small differences in mode shapes making it ideal for use in empirical testing. The numeric correlation is bounded between 0 and 1, with 1 indicating fully consistent mode shapes. A value near 0 indicates the modes are not consistent. pyright 2022 by NuScale Power, LLC 176
Figure 3-98 Span C, 5 Accelerometers (Frequency Response Function, Axial-Red, Vertical-Green, Horizontal-Blue)
}}2(a),(c),ECI Figure 3-99 Mode Shape for 1C-1Z at (( }}2(a),(c),ECI
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 177
Figure 3-100 Mode Shape for 1C-1Z at (( }}2(a),(c),ECI
}}2(a),(c),ECI Figure 3-101 Mode Shape for 1C-1Z at (( }}2(a),(c),ECI
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 178
Figure 3-102 Mode Shape for 1C-1Z at (( }}2(a),(c),ECI
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 179
ure 3-103 Comparison of Mode Shapes between 1C-1Z ((( }}2(a),(c),ECI) and 1A-5Z ((( }}2(a),(c),ECI) Using Modal Assurance Criteria
}}2(a),(c),ECI 6 Build-Out Modal Testing Conclusions Five groups of tests were completed to characterize frequency response, damping, and mode shapes. The individual test configurations for the tests in this report are summarized in the test matrix in Table H-1.
Local modes for various span lengths and boundary conditions resulted in three predominant modes over a range of frequencies for the majority of tests. The first predominant mode ranged from (( }}2(a),(c),ECI with one exception in Group A and those in Group E. Excitation of the HCSG support showed measurements on the adjacent tubes to have multiple low-frequency FRF peaks between (( }}2(a),(c),ECI (these tests had support rings removed). The FRF peaks specific to supports also introduced lower modes between (( }}2(a),(c),ECI for Group E. Lower modes below (( }}2(a),(c),ECI were difficult to excite and not present in responses for Groups B, C, and D. A summary of frequency ranges over predominant modes for the testing groups is provided in Table 3-37. pyright 2022 by NuScale Power, LLC 180
Table 3-37 Frequency Response Function Summary Per Group Predominant
}}2(a),(c),ECI Modes were excited more clearly within the measured FRFs when excited mid-span for the first predominant mode and near the ends for the second and third predominant modes. The influence of local modes unique to each span and boundary condition in the HCSG creates additional complexity in evaluation and multiple tests may be required to discern specific modes within each span. With impact excitation, the data suggest modes can be excited up to one adjacent span of similar length from the excitation source. Fixed-end conditions, or nearly fixed-end, offer the best location for impact testing and elevated signal-to-noise ratios.
Multiple methods were used to calculate damping with the following trends: Half-power method exhibited the largest variance Logarithmic decrement can improve variance using the third peak and subsequent peaks Half-power generally produced higher estimates when compared to logarithmic decrement Logarithmic decrement has issues resolving damping for upper modes greater than ((( }}2(a),(c),ECI). It is noteworthy that FEI is expected to be predominant for the first-mode bending, or less than (( }}2(a),(c),ECI based on the current testing data and TB and VS are also expected at frequencies below (( }}2(a),(c),ECI. Exponential fit produced generally lower damping values than logarithmic decrement with less variance (evident at larger initial tube displacements). Amplitude dependency on damping was observed for both logarithmic decrement and exponential damping estimations and can be summarized as follows:
- Impact testing was limited to (( }}2(a),(c),ECI of tube displacement with the majority below (( }}2(a),(c),ECI. Damping values generally increased with amplitude ranging from ((
}}2(a),(c),ECI with a few points between ((
}}2(a),(c),ECI.
- Pull testing for excitation was limited to (( }}2(a),(c),ECI of initial displacement (single test). Damping values were between
(( }}2(a),(c),ECI over the (( }}2(a),(c),ECI range and up to the pyright 2022 by NuScale Power, LLC 181
(( }}2(a),(c),ECI displacement with an exponential fit. The trend and scatter increased for elevated displacements with a maximum damping (logarithmic decrement) of (( }}2(a),(c),ECI. Mode shapes of Span C characterized the three predominant frequency responses as:
- first-mode bending in the vertical direction
- second-mode bending in the predominantly horizontal direction
- third-mode bending in the predominantly horizontal direction Impact testing provided good repeatability of FRFs across multiple tests. The mode shapes were also repeatable from various locations (i.e., different span locations) of excitation.
pyright 2022 by NuScale Power, LLC 182
The validation methodology provides a framework for selecting the aspects of the design analysis program to be validated and establishes or confirms that the experimental design provides sufficient data to validate the necessary aspects of the design analysis program. Pre-test prediction calculations implement the validation methodology. These calculations confirm the adequacy of the experimental scope, including identifying optimal test conditions and locations for sensors, and determining a range of expected and allowable experimental results, considering uncertainties and biases, that validate the design analysis. Use of Test Data Every measurement has some minor error that results in a difference between the measured value and the true value. This difference between the measured and true value is the total error that is comprised of two components: random error and systematic error (Reference 9.1.2). Random error varies randomly in repeated measurements throughout the conduct of a test, whereas the systematic error remains constant, for example due to imperfect calibration or data reduction techniques. Accurate measurement requires minimizing both random and systematic errors. The test data are accompanied by a test equipment error and accuracy report (TEEAR), which provides the estimated value of expanded uncertainty for each recorded quantity (direct and derived). When each test is complete and the TEEAR uncertainty values are available, they are substituted for experimental uncertainty (uD) as discussed in Section 4.2. Methodology Pre-test predictions develop and apply a series of calculations and finite-element models to generate best estimate and allowable responses for the components in the test. The output of the analyses is compared to prototype test results from the vibration and stress measurement program to validate the analytical approach in the design analyses and the margin of safety. The implementation of this methodology is discussed in detail for individual NPM components in the context of the corresponding FIV analysis method in Section 4.3, Section 4.4, Section 4.5, and Section 4.6. This validation process accounts for the fact that model predictions and analytical methods rely on engineering simplifications, and that test results are inevitably affected by practical differences (Section 4.6) and instrument uncertainty that introduces errors into the validation process. Equation 4-1 to Equation 4-6 are presented for derivation of the validation approach. Following the logic presented in Section 1-5 of Reference 9.1.2, a predicted value S is compared to an experiment data value D for purposes of validation. The comparison error or discrepancy is: E = S-D Equation 4-1 pyright 2022 by NuScale Power, LLC 183
test, as shown below. The true value of the variable of interest is denoted as T, so the error in the predicted value, S, is the difference between S and T: S = S - T Equation 4-2 Similarly, the error in the measured value, D, is the difference between D and T: D = D - T Equation 4-3 Using Equation 4-1 through Equation 4-3, the validation comparison error E is expressed as: E = ( S + T ) - ( D + T ) = S - D Equation 4-4 Knowledge of the true value T is not known with certainty, so the application of Equation 4-4 is continued through the definition of additional terms. The errors in the predicted value S are assigned to one of three categories included in the following summation: S = model + num + input Equation 4-5 i) The error model due to assumptions and approximations in design analysis ii) The error num due to the numerical solution of the equations (relevant to complex computer codes, e.g., finite element analysis) iii) The error input due to variability in the input parameters to the pre-test analysis. The objective of a validation exercise is to ascertain the magnitude of model to within an uncertainty range. However, this error can be obscured by errors from the analysis solution scheme or the error embedded in the pre-test prediction input parameters, as well as the test result error. Re-arranging Equation 4-5 and using Equation 4-4: model = E + D - num - input Equation 4-6 Equation 4-6 is not solved directly. Once a test is complete and measurement data are collected, the sign and magnitude of E are known from Equation 4-1, but the remaining terms on the right-hand side of Equation 4-6 are not known with certainty. In the Reference 9.1.2 approach, each error term is viewed as a single realization from a parent probability distribution. The standard deviation of each parent population (also called standard uncertainty) corresponding to these errors is taken as u D , u num , and u input . pyright 2022 by NuScale Power, LLC 184
Thus, the standard uncertainty associated with the estimate of model is expressed as a validation uncertainty as follows: 2 2 2 u val = u +u +u Equation 4-7 D num input Consequently, Equation 4-6 and the range defined in Equation 4-8, characterize an interval within which model falls. Therefore, the determination of the validation comparison error E provides a direct assessment of the prediction error with u val as the validation standard uncertainty. E +/- u val Equation 4-8 Validation Approach and Uncertainty Analysis Based on the validation methodology described in Section 4.2, the uncertainty analysis applies the following guidelines, with Equation 4-8 as the basis for the validation exercise in the post-test analysis: If the absolute value of the validation comparison error E is much greater than the validation standard uncertainty u val , then model is approximately equal to E (i.e., it accounts for most of the observed difference between analysis result and test data). In this case, there is an incentive to enhance the accuracy of the analysis to reduce model error. If the absolute value of the validation comparison error E is less than or equal to u val , then model is within the noise level created by uncertainties in the solution scheme, inputs, and test data used to perform the validation. In this case, there is small benefit to pursuing model improvements to achieve better accuracy. If u val is identified with a particular family of probability distributions, then a confidence interval can be defined. For instance, assuming ( input + num - D ) is from a Gaussian distribution, the expanded uncertainty with 95 percent confidence is U 95 = 2u val . Thus, E +/- U 95 provides an interval in which model resides about 95 times out of 100. The following sections provide the detailed methodology for calculating the validation metrics E and u val for each of the FIV phenomena examined in the measurement program. u val is calculated in the pre-test prediction and E cannot be finalized until the post-test analysis, when errors in the measured values can be assessed. u val is also used in the pre-test prediction to inform the expected and allowable range of experimental results that validate the design analyses. pyright 2022 by NuScale Power, LLC 185
Engineering variables of interest are often functions of other variables based on a well-known mathematical relationship. In this validation methodology, the engineering variables of interest are the safety margins associated with the onset of an FIV phenomenon or other FIV results such as the fatigue margin for a SG tube. The effect of random standard uncertainty in the constituent variables, denoted as u x , is i approximated by the Taylor series method (Reference 9.1.2). Consider a result R expressed in terms of the average or assigned values of the independent parameters X i that enter into the result. That is, R = f ( X 1 ,X 2 , ... ,X I ) Equation 4-9 Where I signifies the total number of parameters involved in R . When there is a known mathematical relationship between the result and its parameters, sensitivity coefficients i are calculated by partial differentiation (Equation 3-2-2 of Reference 9.1.2): i = ------- R-Equation 4-10 Xi The absolute standard uncertainty of the result is calculated based on square root sum of squares of each uncertainty term, as follows: 1--- I 2 2 uR = ( i ux ) i Equation 4-11 i=1 An alternative to partial differentiation by analytical derivation is to use central finite differences. This implies that the sensitivity coefficients are calculated as shown below: f .. ,X + X ,.. - f .. ,X - X ,.. 0 0 0 0 i i i i i = --------------------------------------------------------------------------------------- Equation 4-12 0 2X i It is implicit that the sensitivity coefficient in Equation 4-10 is evaluated at the nominal value of the parametric vector. Note that many design analysis input parameters are biased to be conservative (versus best-estimate or nominal). Therefore, care is taken in the pre-test prediction to use best-estimate inputs in the absolute standard uncertainty calculation to ensure an appropriate range for the expected measurement results is obtained. If using Equation 4-12, a choice is made to set the value of the pyright 2022 by NuScale Power, LLC 186
perturbation size X i . If X i is too large then truncation error is large, so a practically small value is recommended. It is often useful to set X i equal to the ratio of the variables standard error to its nominal (mean) value. 2 Calculation of Input Parameter Uncertainty Based on the guidance in Reference 9.1.2, there are two different approaches for estimating u input . The approach depends on whether a local or global view of uncertainty estimation process is followed. In this report, both are presented as valid options because of the diversity in the FIV phenomena and their analytical methods, and to provide flexibility for the analyst when performing pre-test analysis. 2.1 Local Method An analysis prediction S with n p uncorrelated input parameters is effectively a result developed from the arithmetic construction or manipulation of the underlying input parameters. Hence, the same uncertainty propagation method discussed in Section 4.3.1 is used to evaluate u input , namely: 1--- np 2 S 2 u input = X
-------- u x i i Equation 4-13 i=1 where u x is the standard uncertainty in input parameter X i . Ideally, u x should i i come from prior experiments, although engineering judgment may be accepted instead to estimate it or may require validation based on the measurement program results.
2.2 Global Method The sensitivity coefficient method presented in the preceding section is termed local sensitivity and uncertainty propagation because the function evaluations are in a narrow (local) neighborhood of the mean parameter value. This approach does not capture highly nonlinear behavior in the input parameter space. For this reason, sampling-based methods using a Monte Carlo technique is used to mitigate the limitation in the local method. Reference 9.1.2 specifies using the Latin Hypercube Sampling (LHS) method to achieve a reasonable number of samples, equivalent to n LHS n p + 1 , where n p is the number of variables in the function being evaluated. Once the analysis is performed using n LHS parameter vectors, whose constituent values are paired at random, the mean value and standard deviation from the pyright 2022 by NuScale Power, LLC 187
respectively (Equations 3-3-1 and 3-3-2 of Reference 9.1.2). n LHS 1 S = ------------ n LHS Si Equation 4-14 i=1 n LHS 12 1 2 u input = --------------------- n LHS - 1 ( Si - S ) Equation 4-15 i=1 The validation exercise that employs the LHS method demonstrates statistical convergence by performing sensitivity runs with an increasing number of samples. In addition, if the distribution function of the input variables is assumed, then sensitivity of u input to this assumption is explored. 3 Calculation of Mesh Numerical Uncertainty In some FIV evaluations, the analysis relies on a computer model that is developed using finite-element methods. In this case, the solution process introduces uncertainty in the overall model result due to the fact that discretized equations are used or iterative matrix solvers are executed (the latter only for nonlinear systems). Finite element models are used to determine the vibration mode shape and natural frequency in many FIV evaluations. Estimation of numerical uncertainty is not required for hand calculations in which the input parameters are obtained from mathematical or empirical correlations. The Grid Convergence Index method is a means to estimate the numerical uncertainty that arises from the use of computational grids with different resolution capabilities that aim to output a result . Considering three numerical meshes (fine; medium; coarse) to have characteristic cell sizes h 1 < h 2 < h 3 , and refinement factors r 21 = h 2 h 1 and r 32 = h 3 h 2 , the order of convergence, p , is calculated as: ln ( 32 21 ) + q ( p ) p = ----------------------------------------------
- Equation 4-16 ln ( r 21 )
where p r - s q ( p ) = ln ----------------- 21 { 0 if r 21 = r 32 } Equation 4-17 p r - s 32 pyright 2022 by NuScale Power, LLC 188
32 s = 1 sign ------- Equation 4-18 21 32 = 3 - 2 (difference in result between course and medium mesh) Equation 4-19 21 = 2 - 1 (difference in result between medium and fine mesh) Equation 4-20 A uniform and integer mesh refinement factor is used (for example r = 2 ) to implement this method. Numerical uncertainty is then estimated from the Grid Convergence Index, given below, with a factor of safety F S = 3 (recommended for unstructured grid refinement, and three grid solutions are sufficiently conservative): FS 1 - 2 u num = GCI = ---------------------------
- Equation 4-21 p
r -1 21 Note that if, for example, the order of the discretization scheme is known to be p = 2 (second order), and the medium mesh cell size (h2) is uniformly decreased by half in all directions (r = 2 ), then u num from Equation 4-21 becomes the absolute value of the difference between the result on the fine and medium grid ( 1 - 2 ). For linear modal analysis, this value is small because there is a weak dependence of calculated frequency on grid density. Evaluation Process The process for the pre-test prediction and post-test assessments is summarized below. Some values may require assumptions at the time of the pre-test prediction depending on the status of the test design. In the post-test assessment, the uncertainties are confirmed or adjusted as necessary based on the final test design and results. Pre-Test Prediction
- 1. Using best-estimate input and accounting for experimental biases, calculate safety margin and critical parameters for the test (frequencies, critical or lock-in velocities, and so on).
- 2. Calculate input, measurement and numerical uncertainties.
- 3. Using the parameters determined above, quantify the range of allowable test results that adequately validate the design analysis.
These steps provide confidence that considering the experimental biases, and input, measurement and numerical uncertainties, the test design is adequate to validate the design analysis. pyright 2022 by NuScale Power, LLC 189
- 1. Determine if changes to the expected or allowable range in the pre-test prediction are required based on considerations such as finalization of the test design, confirmation of measurement uncertainties, or the as-tested conditions.
- 2. Confirm test results match predictions and are within the allowable range for validation.
- 3. Quantify the validation comparison error and model error to confirm they are acceptable.
These steps complete the validation of the design analysis using the test results. If the modeling error is greater than the validation comparison error, the design analysis is to be updated to decrease modeling error. Additionally, if the test results do not match predictions, the design analysis is to be revised based on the conclusions of the testing. Evaluation Procedures 1 Turbulent Buffeting of Steam Generator Tube The method for validating the TB design analysis process by testing measurements is described below. 1.1 Overview Consider the results of the test used for validating the TB design analysis to provide the natural frequencies in water (f n_T ), mode shapes ( T ), and associated uncertainty ( u D ). An ANSYS model to simulate the test geometry is developed. The model prediction for the vibration modes and natural frequencies are M , and f n_M , respectively. The model results are examined on a column-by-column basis and an appropriate comparison to a corresponding test result is to be made. 1.1.1 Calculate Model Error in Modal Parameters Using Equation 4-1, the modeling error is determined as the maximum difference in frequency or mode shape comparison: E SG = max ( f n_M - f n_T , M - T ) The comparison may involve more than just the fundamental beam mode. 1.1.2 Calculate Model Uncertainty The uncertainty associated with the result from modeling of the SG test is referred to as u val , and consists of input, numerical, and measurement pyright 2022 by NuScale Power, LLC 190
sections. The analysis ensures that the resultant u val E SG , otherwise the modeling approach should be modified to reduce the error. Engineering judgment may be used to weigh the relative significance of the errors obtained from the comparison of different modes, and among the different SG columns. This step is dependent on the level of detail in the test data, and whether measured frequencies can be distinguished and matched to their counterparts in the model. 1.1.2.1 Input Uncertainty Significant inputs to the analysis of the test configuration are listed in Table 4-1. The sampling technique discussed in Section 4.3.2.2 provides a method to generate a series of model predictions from which an input uncertainty can be estimated (Equation 4-15). Note that sampling of different boundary conditions is not necessary unless the modeling error is much larger than the overall uncertainty ( u val << E SG ). When evaluating input uncertainty, nominal (best-estimate) values are used, as discussed in Section 4.3.1. Table 4-1 SG Test Model Inputs ut Parameter Basis for Variability metric Dimensions Manufacturing tolerances affect the nominal values specified for the model s The effect of hydrodynamic mass in actual tube bundle may be different than that estimated by formula which is based on a correlation for single flexible tube surrounded by an array of rigid tubes (note: the correlation relies on pitch and diameter) ndary Conditions SG boundary conditions are subject to variability based on considerations such as fit-up with the supports, manufacturing tolerances, compression and thermal expansion. Boundary conditions are expected to be fixed, sliding, or a combination. Boundary conditions may vary throughout the tube bundle. 1.1.2.2 Numerical Uncertainty The numerical uncertainty in the modal analysis (u num ) is estimated based on the sensitivity study described in Section 3.1.2.2.3 for the approximate solution of the acceptance integrals. The approximation method is described in Section 3.1.2.1. 1.1.2.3 Measurement Uncertainty In the post-test analysis, the measurement data uncertainty is used directly as provided by the test if it is reported for the frequency and mode shape and no other action is required. In the pre-test analysis, a pyright 2022 by NuScale Power, LLC 191
propagation calculation is performed to obtain u D similar to the approach for estimating input uncertainty. 1.1.3 Estimate Vibration Amplitude Uncertainty The turbulence-induced RMS displacement is determined analytically using the approach described in Section 3.1.2. The test program shall measure the displacement (y test ), and also quantify its uncertainty as u Dy . Using Equation 4-1, the prediction error is: E y = y rx - y test . The parameter of interest is the maximum RMS value of displacement in the two adjacent spans of a support. The input uncertainty in the RMS displacement is determined using the global method described in Section 4.3.2.2. The parameters sampled are the damping ratio, modal frequencies, and PSD magnitude. The sample standard deviation of the results provides the uncertainty in the RMS displacement. 1.2 Estimate Uncertainty in Safety Margin for SG Tube in NPM Margin to the TB acceptance criterion is determined via Equation 4-22 which allows determination of the allowable number of cycles and the fatigue usage due to impact stress. Equation 4-22 is rewritten as Equation 4-23. 4 2 2 15 E Me f y S rms = c --------------------------------- n rms 3 Equation 4-22 D 2 15 X 1 X 2 X 3 F ( X 1 , X 2 , X 1 , X 4 ) = ------------------------ Equation 4-23 3 X4 1 45 where = --- cE . The input parameters in Equation 4-23 are discussed in 2 Table 4-2, and assigned generic labels. The process in Section 4.3.1 is used to evaluate the effect of uncertainty in the different parameters, as given in Table 4-3. pyright 2022 by NuScale Power, LLC 192
Table 4-2 SG Tube Inputs to TB Margin Calculation porary Label Input Parameter Basis Effective mass of tube, Total mass includes the mass of the tube metal, secondary usually taken as 2/3 the fluid, and hydrodynamic mass (virtual mass on primary side). total mass of the two Uncertainty in those contributing factors propagate to the spans { M e } total. A nominal value for effective tube mass is used. Natural frequency of the This is obtained using an ANSYS modal analysis. tube { f n } Maximum mean square This is calculated using a PSD approach, which is verified to vibration amplitude of be bounding as part of the SG testing. A nominal vibration the tube in the adjacent amplitude is used. 2 spans { y } rms Outer diameter of the Manufacturing tolerances affect the nominal value specified tube { D } for the calculation. A nominal outer diameter is used. Table 4-3 SG Tube TB Margin Uncertainty Method p Description Procedure Analytic derivative of -4/5 2/5 1/5 X X X Equation 4-23 with respect to - --------------------------------
-- 1 2 3 X1 5 3/5 X
4 Uncertainty in X 1 Uncertainty in total tube mass is determined from the uncertainty in secondary fluid density Analytic derivative of -1/5 2/5 4/5 X X X Equation 4-23 with respect to 2 1 2 3 X2 5 3/5 X 4 Uncertainty in X 2 Standard deviation from the mean frequency of the set of vibration modes included in the average Analytic derivative of 1/5 2/5 -4/5 X X X Equation 4-23 with respect to 1 2 3 - X3 5 3/5 X 4 Uncertainty in X 3 u y from Section 4.5.1.1.3 Analytic derivative of -1/5 2/5 1/5 X X X Equation 4-23 with respect to -3 1 2 3 X4 5 8/5 X 4 Uncertainty in X 4 Obtain from standard deviation of SG outer diameters as generated by manufacturing tolerances distribution pyright 2022 by NuScale Power, LLC 193
p Description Procedure Equation 4-11. The result is 2 2 2 2 2 2 2 2 added to / subtracted from the u + u + u + u alternating stress in 1 x1 2 x2 3 x3 4 x4 Equation 4-23 evaluated at nominal values 2 Vortex Shedding Consider the results of the test to provide the natural frequency in water (f n_T ), mode shape ( T ), and associated uncertainty ( u D ). An ANSYS model to simulate the test geometry is developed using the same configuration, as well as environment and boundary conditions as the experiment. The model predictions for the fundamental mode natural frequency and mode shape are f n_M and M , respectively. 2.1 Calculate Model Error Using Equation 4-1, calculate the modeling error based on maximum of different modes: E SG_f = f n_M - f n_T E SG_ = M - T E SG = max ( E SG_f ,E SG_ ) The above comparison errors are examined for at least the first mode results. If higher modes are reviewed the same procedure is followed, but the first mode is more limiting to the margin assessment. Also, because is not a single value but a function of SG height, the maximum difference is used on the basis of a unity normalized mode shape. 2.2 Calculate Model Uncertainty The uncertainty associated with the result from modeling of the SG test is referred to as u val , and consists of input, numerical, and measurement uncertainties, combined per Equation 4-7, and discussed in the following sections. The analysis ensures that the resultant u val E SG , otherwise the modeling approach is to be modified to reduce the error. pyright 2022 by NuScale Power, LLC 194
The inputs to the modal analysis of the test configuration are listed in Table 4-4. The sampling method discussed in Section 4.3.2.2 is used to generate a series of model predictions from which an input uncertainty is estimated (Equation 4-15). Note that sampling of different boundary conditions is not necessary unless the modeling error is much larger than the overall uncertainty (u val << E SG ). Table 4-4 Steam Generator Test Model Inputs Input Parameter Basis for Variability Geometric Dimensions Manufacturing tolerances affect the nominal values specified for the model Mass Effect of hydrodynamic mass in actual tube bundle is likely different than that estimated by formula, which is based on a correlation for single flexible tube surrounded by an array of rigid tubes (note: the correlation relies on pitch and diameter_ Boundary Conditions SG boundary conditions are subject to variability based on considerations such as fit-up with the supports, manufacturing tolerances, compression and thermal expansion. Boundary conditions are expected to be fixed, sliding, or a combination. Boundary conditions may vary throughout the tube bundle. 2.2.2 Numerical Uncertainty The numerical uncertainty in the modal analysis (u num ) is estimated using the approach discussed in Section 4.3.3 using nominal model inputs. 2.2.3 Measurement Uncertainty In the post-test analysis, the measurement data uncertainty is used directly as provided by the test if it is reported for the frequency and mode shape and no other action is required. In the pre-test analysis, a propagation calculation is performed to obtain u D (Section 4.1) similar to the approach for estimating input uncertainty. 2.3 Estimate Uncertainty in Safety Margin for Steam Generator Tube in NuScale Power Module There are four methods to show acceptable margin to VS lock-in. The three methods applicable to tube arrays are described below for completeness; however, they may not be necessary to execute the pre-test prediction. Method B is described first. Method B requires that the reduced damping (C RD ) is greater pyright 2022 by NuScale Power, LLC 195
represented using Equation 4-24. l 2 C RD 1 4 m tot tube ( x )dx 0 1 SM VSB = ----------- - 1 = ------ ---------------------------------------------------- - 1 Equation 4-24 64 64 2 le 2 D 0 ( x )dx 1 The input parameters in Equation 4-24 are discussed in Table 4-5, and assigned generic labels. The process in Section 4.3.1 is used to determine the analytically predicted allowable range for the safety margin in Equation 4-24 and evaluate the effect of uncertainty in the different parameters. Equation 4-24 is re-written as: SX 1 X 2 g ( X 5 ,X 7 ) SM VSB = -------------------------------------- --1 Equation 4-25 2 SX 3 X 4 h ( X 6 ,X 7 ) where S = 16 , g , and h are the definite integrals. Table 4-6 shows the details of performing the calculation on Equation 4-25. The SG mode shape relative magnitude is zero at the beginning and end of the tube, so 1 ( 0 ) = 1 ( l tube ) = 0 Table 4-5 Steam Generator Tube Inputs to Vortex Shedding Margin Calculation: Method B porary Label Input Parameter Basis Damping ratio in air A nominal value of damping is used. This parameter is measured { } in a dedicated test, so its uncertainty is propagated. Total mass of tube Total mass includes the mass of the tube metal, secondary fluid, { m tot } and hydrodynamic mass (virtual mass on primary side). Uncertainty in those contributing factors propagates to the total. Fluid density { } The primary fluid temperature variation along the SG radius and height is a source of uncertainty in the calculation, which assumes a constant value for RCS (cold or hot region). Tube outer Manufacturing tolerances affect the nominal value specified for diameter { D } the calculation. Overall length of Manufacturing tolerances affect the nominal value specified for tube { l } the calculation. tube Length of tube In the design analysis, this is the length of the tube where the subject to cross mode shape is considered. Its uncertainty propagates to the flow { l e } margin calculation. pyright 2022 by NuScale Power, LLC 196
Method B (Continued) porary Label Input Parameter Basis Fundamental The free-vibration mode shape of the SG tube is obtained using modal shape { 1 } an ANSYS modal analysis. This parameter includes uncertainty in the modal analysis modeling, considering both the range of possible mode shapes based on boundary conditions and the components of the mode shape that could be excited by cross flow. Table 4-6 Steam Generator Tube Vortex Shedding Margin Uncertainty Method: Method B p Description Procedure Analytic derivative of SX 2 g ( X 5 ,X 7 ) Equation 4-25 with respect to -----------------------------------
- , where g and h are definite integrals 2
X1 X 3 X 4 h ( X 6 ,X 7 ) evaluated at the nominal values Uncertainty in X 1 Obtain from measurement data of damping ratio in air Analytic derivative of SX 1 g ( X 5 ,X 7 ) Equation 4-25 with respect to -----------------------------------
- , evaluated at the nominal values 2
X2 X 3 X 4 h ( X 6 ,X 7 ) Uncertainty in X 2 Uncertainty in total tube mass is determined from the uncertainty in secondary fluid density, tube material density, 2 2 2 0.5 and RCS density (i.e., u X u +u +u ) 2 S 690 rcs Analytic derivative of -SX 1 X 2 g ( X 5 ,X 7 ) Equation 4-25 with respect to ----------------------------------------- , evaluated at the nominal values 3 X3 X 3 X 4 h ( X 6 ,X 7 ) Uncertainty in X 3 Obtain from RCS density range across SG (uncertainty due to fluid temperature variation) Analytic derivative of -2SX 1 X 2 g ( X 5 ,X 7 ) Equation 4-25 with respect to -------------------------------------------- , evaluated at the nominal values 3 X4 X 3 X 4 h ( X 6 ,X 7 ) Uncertainty in X 4 Obtain from standard deviation of SG outer diameters as generated by manufacturing tolerances distribution Analytic derivative of 2 Equation 4-25 with respect to SX 1 X 2 ( l tube ) 1
- = 0 X5 2 X 3 X 4 h ( X 6 ,X 7 )
Uncertainty in X 5 Not propagated pyright 2022 by NuScale Power, LLC 197
Method B (Continued) p Description Procedure Analytic derivative of SX 1 X 2 g ( X 5 ,X 7 ) Equation 4-25 with respect to --------------------------------------- , evaluated at the nominal values 2 2 X6 X3 X 4 1 ( le ) Uncertainty in X 6 Obtain from range of tube length exposed to flow based on tolerances in the configuration of supports and their assumed effectiveness Analytic derivative of Equate to 1.0 (justified because model uncertainty is Equation 4-25 with respect to expected to be small) X7 Uncertainty in X 7 Same as u val in Section 4.5.2.2. Equation 4-11 2 2 2 2 2 2 2 2 2 2 2 2 2 2 ux + ux + ux + ux + ux + ux + ux 1 1 2 2 3 3 4 4 5 5 6 6 7 7 Uncertainty estimates are also provided for Method A and C. fn D X 9 X 10 SM VSA = -----------------------
- - 1 = -------------------------------
--1 Equation 4-26 v cos St X 8 cos X 11 Table 4-7 Steam Generator Tube Inputs to Vortex Shedding Margin Calculation:
Method A porary Label Input Parameter Basis velocity { v } This parameter includes uncertainties in the assumed primary coolant flow velocity. Natural Frequency This parameter includes uncertainties in the SG tube natural { fn } frequency. Tube outer This parameter includes uncertainties associated with diameter { D } manufacturing tolerances. Table 4-8 Steam Generator Tube Vortex Shedding Margin Uncertainty Method: Method A p Description Procedure Analytic derivative of -X 9 X 10 Equation 4-26 with respect to ------------------------------- 2 X8 X cos X 11 8 Uncertainty in X 8 Uncertainty in velocity. pyright 2022 by NuScale Power, LLC 198
Method A (Continued) p Description Procedure Analytic derivative of X 10 Equation 4-26 with respect to ------------------------------- X 8 cos X 11 X9 Uncertainty in X 9 Uncertainty in the SG tube frequency. Analytic derivative of X9 Equation 4-26 with respect to -------------------------------- X 8 X 11 cos X 10 Uncertainty in X 10 Uncertainty in tube outer diameter. Analytic derivative of -X 9 X 10 Equation 4-26 with respect to -------------------------------- 2 X 11 X 8 cos X 11 Uncertainty in X 11 Uncertainty in the Strouhal number. Equation 4-11 2 2 2 2 2 2 2 2 ux + ux + ux + ux 8 8 9 9 10 10 11 11 Method C uses the same variables as Methods A and B, but with different acceptance criteria limits. Specifically, the reduced damping needs to be greater than 1.2 instead of 64 and the reduced velocity needs to be less than 3.3 instead of 1.0. The reduced velocity limit does not affect the analytic derivatives. The safety margin value can be calculated using Equation 4-26 with 3.3 substituted for the value of 1 from Method A. For the reduced damping, Equation 4-25 can be used with the value of S modified to 0.3 instead of 16 . When using Method C, the reduced damping and reduced velocity both are evaluated to determine expected and allowable ranges. 3 Fluid-Elastic Instability of SG Tubes The method for validating the design analysis process by testing measurements is described below. 3.1 Calculate Model Error Model error is quantified at the completion of the post-test analysis per Equation 4-1. For the FEI assessment, model error is calculated for each measured or derived parameter that is a calculation result, such as the frequencies, mode shapes, and pitch velocity. The model error is then compared to the validation standard uncertainty to ensure the validation metric are consistent with the inequality in Equation 4-8. No estimation of model error is required during the pre-test analysis. pyright 2022 by NuScale Power, LLC 199
Model uncertainty is comprised of input uncertainty, numerical uncertainty, and measurement uncertainty. These uncertainties are evaluated in the pre-test prediction. Some values require assumptions at the time of the pre-test prediction depending on the status of the test design. In the post-test assessment, the uncertainties are confirmed or adjusted as necessary based on the final test design. In the pre-test analysis, the predicted uncertainties are used to determine the range of allowable test results that validate the design analysis. Section 4.7 contains additional details regarding this aspect of the pre-test analysis. In the post-test analysis, the range is confirmed using final uncertainty values, and the validation uncertainty is compared to the model error as described in Section 4.5.3.1. 3.2.1 Input Uncertainty The effect of input parameter uncertainty on simulation uncertainty is calculated using the local or global methods that are discussed in Section 4.3.2.1 and Section 4.3.2.2, respectively. The local method is sufficient for input parameters that are relevant to FEI. If the effects of large differences in the input parameter need to be investigated the global method is to be used. The categories of inputs relevant for the FEI analysis of the test configuration are listed in Table 4-9. The sampling methods discussed in Section 4.3.2.2 are used to generate a series of predictions from which an input uncertainty can be estimated (Equation 4-15). In the pre-test analysis, the significant inputs within each category in Table 4-9 are to be considered. Examples of the types of input parameter variability for each category are also identified in the table. The input parameters that are considered in the input uncertainty assessment are to be justified in the pre-test analysis. Table 4-9 Fluid Elastic Instability Analysis Input Parameter Types ut Parameter Example Input Parameter Variability egory al The effects of as-built conditions and allowed variability in test conditions on the pitch velocity are considered. rmal hydraulic The effect of the following on the pitch velocity are considered- allowed variation in test conditions, calculation methods for determining local velocities, and the effect of as-built conditions on local velocities. ign The effect of as-built conditions on the critical velocity are considered. pirical Uncertainties in the empirical correlations used to determine the critical velocity are considered. pyright 2022 by NuScale Power, LLC 200
Numerical uncertainty is assessed in the pre-test prediction. Completion of testing does not affect the predicted numerical uncertainty. Numerical uncertainty applies to the modal analysis of the SG tube and the integration of the mode shapes over the length of the tube, which is performed as a summation approximation. The numerical uncertainty in the modal analysis (u num ) is estimated using the approach discussed in Section 4.3.3 using nominal inputs. 3.2.3 Measurement Uncertainty Measurement uncertainty is estimated in the pre-test prediction and a final evaluation is performed in the post-test assessment. Depending on the maturity of the test design at the time of the pre-test prediction, engineering judgment is used to estimate measurement uncertainty values. In this event, appropriate assumptions are made that are reasonable both in terms of the test design and the ability to validate the design analysis. The results of the pre-test prediction are used to identify changes that need to be incorporated into the detailed test design to accommodate reasonable measurement uncertainty. Open items are established to track the assumption until it is incorporated. In the post-test analysis, the measurement data uncertainty is used directly as provided by the test and no other action is required. In the pre-test analysis, a propagation calculation is performed to obtain u D (Section 4.1) similar to the approach for estimating input uncertainty. 3.3 Estimate Uncertainty in Safety Margin The margin to the onset of FEI is based on the difference between the reduced mode shape weighted mean pitch velocity and the critical velocity, as shown in the following equation. The uncertainties in the modal, thermal hydraulic, design and empirical inputs are evaluated to quantify the effect of each parameter on the design analysis safety margin. 0.5 v cos ( )M i SF FEI = 1 - ------------------------------------ Equation 4-27 m'2 a f i DC --------------- 2 D The input parameters in Equation 4-27 are discussed in Table 4-10, and assigned generic labels in Equation 4-28. pyright 2022 by NuScale Power, LLC 201
0.5 X 1 cos ( )X 2,i SF FEI = 1 - ----------------------------------------------------- Equation 4-28 X 2X X8 X 9,i X 6 X 7 -------------------- 3 4 2 X5 X 6 The process in Section 4.3.1 is used to determine the effect of input parameter uncertainty on the safety margin value as outlined in Table 4-11. pyright 2022 by NuScale Power, LLC 202
Table 4-10 Steam Generator Tube Inputs to Fluid Elastic Instability Safety Margin Calculation porary Label Input Parameter Basis velocity This parameter includes uncertainties in the assumed primary coolant flow velocity. modal multiplier This parameter includes uncertainty in the modal analysis modeling uncertainty, considering both the range of possible mode shapes based on boundary conditions and the components of the mode shape that could be excited by cross flow. mass This parameter includes uncertainty in the mass of the tube and the surrounding fluid. damping This parameter includes uncertainty, which is finalized when the damping value is measured. density This parameter includes uncertainty in the operating temperature of the primary coolant. tube outer diameter This parameter includes minor uncertainties in the manufacturing tolerances of the tube. X 8 Connors constants These parameters include uncertainties associated with characterizing the onset of FEI for helical tubes. Tube frequency This parameter includes minor uncertainties in the frequency of the tube. Table 4-11 Steam Generator Tube Fluid Elastic Instability Margin Uncertainty Method p Description Procedure Analytic derivative of 0.5 Equation 4-28 with respect to cos ( )X 2,i X1 X 2 X X8 X 9, i X 7 X 6 -------------------- 3 4 2 X5 X 6 Uncertainty in X 1 Uncertainty in velocity based on differences in primary coolant flow rate, due to considerations such as calculation method and temperature. Analytic derivative of Equation 4-28 with respect to cos ( )X 1 X2 - ------------------------------------------------------------------- X 2 X X8 2X 2 X 9, i X 7 X 6 -------------------- 0.5 3 4 2 X5 X 6 Uncertainty in X 2 Uncertainty in modal multiplier considering differences in tube and fluid mass, material properties and boundary conditions. pyright 2022 by NuScale Power, LLC 203
Method (Continued) p Description Procedure Analytic derivative of 0.5 Equation 4-28 with respect to X 8 X 1 cos ( )X 2,i X3 X 2 X X8 X 3 X 7 X 6 X 9, i -------------------- 3 4 2 X5 X 6 Uncertainty in X 3 Uncertainty in linear mass density considering differences in tube and hydrodynamic mass. Analytic derivative of 0.5 Equation 4-28 with respect to X 8 X 1 cos ( )X 2,i X4 X 2 X X8 X 4 X 7 X 6 X 9, i -------------------- 3 4 2 X5 X 6 Uncertainty in X 4 Uncertainty in damping considering the potential range of expected damping at the design analysis condition. Analytic derivative of 0.5 Equation 4-28 with respect to - X 8 X 1 cos ( )X 2,i X5 X 2 X X8 X 5 X 7 X 6 X 9, i -------------------- 3 4 2 X5 X 6 Uncertainty in X 5 Uncertainty in density considering the range of potential primary fluid temperatures at the design analysis condition. Analytic derivative of 0.5 Equation 4-28 with respect to ( 1 - 2X 8 )X 1 cos ( )X 2,i X6 X 2 X X8 2 X X 7 X 9, i -------------------- 3 4 6 2 X5 X 6 Uncertainty in X 6 Deviation of SG outer diameters as generated by manufacturing. Analytic derivative of 0.5 Equation 4-28 with respect to X 1 cos ( )X 2,i X7 X 2 X X8 2 3 X 9, i X 6 X -------------------- 4 7 2 X5 X 6 Uncertainty in X 7 Consider a range of empirical constants appropriate to bound FEI for the helical SG tubes. pyright 2022 by NuScale Power, LLC 204
Method (Continued) p Description Procedure Analytic derivative of X 2X Equation 4-28 with respect to X cos ( )X 0.5 ln ------------------- 3 4 X8 1 2,i 2 X 5 6 X X 2 X X8 X 9, i X 6 X 7 -------------------- 3 4 2 X5 X 6 Uncertainty in X 8 Consider a range of empirical constants appropriate to bound FEI for the helical SG tubes. Analytic derivative of 0.5 Equation 4-28 with respect to X 1 cos ( )X 2,i X9 2 X X8 2 X X 6 X 7 X -------------------- 3 4 9,i 2 X5 X 6 Uncertainty in X 9 Uncertainties in the SG tube frequency. Equation 4-11 2 2 2 2 2 2 2 2 2 2 2 2 ux + ux + ux + ux + ux + ux 1 1 2 2 3 3 4 4 5 5 6 6 2 2 2 2 2 2
+ ux + ux + ux 7 7 8 8 9 9 4 Acoustic Resonance The method for validating the design analysis process by testing measurements is described below.
4.1 Calculate Model Error Model error is quantified at the completion of the post-test analysis per Equation 4-1. For the acoustic resonance (AR) assessment, model error is calculated for each measured or derived parameter that is a calculation result, such as the acoustic frequencies and velocity. The model error is compared to the validation standard uncertainty to ensure the validation metric are consistent with the inequality in Equation 4-8. No estimation of model error is required during the pre-test analysis. 4.2 Calculate Model Uncertainty Model uncertainty is comprised of input uncertainty, numerical uncertainty, and measurement uncertainty. These uncertainties are evaluated in the pre-test pyright 2022 by NuScale Power, LLC 205
prediction depending on the status of the test design. In the post-test assessment, the uncertainties are confirmed or adjusted as necessary based on the final test design. In the pre-test analysis, the predicted uncertainties are used to determine the range of allowable test results that would validate the design analysis. Section 4.7 contains additional details regarding this aspect of the pre-test analysis. In the post-test analysis, the range is confirmed using final uncertainty values, and the validation uncertainty is compared to the model error as described in Section 4.5.4.1. 4.2.1 Input Uncertainty The effect of input parameter uncertainty on simulation uncertainty is calculated using the local or global methods that are discussed in Section 4.3.2.1 and Section 4.3.2.2, respectively. The sampling methods discussed in Section 4.3.2.2 are used to generate a series of predictions from which an input uncertainty is estimated (Equation 4-15). The input parameters that are considered in the input uncertainty assessment are to be justified in the pre-test analysis. 4.2.2 Numerical Uncertainty There is no numerical uncertainty in the AR calculation method, so this term is zero for the purpose of the pre-test and post-test analyses. 4.2.3 Measurement Uncertainty Measurement uncertainty is estimated in the pre-test prediction and a final evaluation is performed in the post-test assessment. Depending on the maturity of the test design at the time of the pre-test prediction, engineering judgment is used to estimate measurement uncertainty values. In this event, appropriate assumptions are made that are reasonable both in terms of the test design and the ability to validate the design analysis. The results of the pre-test prediction are used to identify changes that need to be incorporated into the detailed test design to accommodate reasonable measurement uncertainty. Open items are established to track the assumption until it is incorporated. In the post-test analysis, the measurement data uncertainty is used directly as provided by the test and no other action is required. In the pre-test analysis, a propagation calculation is performed to obtain u D (Section 4.1) similar to the approach for estimating input uncertainty. pyright 2022 by NuScale Power, LLC 206
Margin to the AR acceptance criterion of Strouhal number is calculated as follows: cd i SF AR = ----------------------------------------- --1 Equation 4-29 4St ( L v + 0.3d i )V The uncertainty in Equation 4-29 is primarily from input parameters and variables that are measured during the startup test at a single location susceptible to AR. Table 4-12 discusses the influential parameters that contribute to the uncertainty in safety margin. The process in Section 4.3.1 is used to determine the analytically predicted allowable range for the safety margin. Hence, Equation 4-29 is re-written as: X1 X2 SF AR = -----------------------------------------------
--1 Equation 4-30 4 ( X 3 + 0.3X 2 )X 4 X 5 Table 4-13 shows the details of performing the calculation on Equation 4-30.
Table 4-12 Decay Heat Removal System Steam Pipe Inputs to Acoustic Resonance Margin Calculation ut Parameter Basis ed of sound In analysis space, the speed of sound is evaluated as a function of steam pressure and temperature in the rigid piping. During NPM steady-state operation and startup testing, the thermodynamic properties of the steam fluctuate about a mean. Accordingly, changes in pressure or temperature are accounted for in the speed of sound. al flow velocity at This input parameter is evaluated from the total secondary mass flow rate, steam in pipe density, and flow area of pipe to obtain an average value. During NPM steady-state operation and startup testing it is probable that the local velocity is slightly different due to thermodynamic fluctuations, turbulence, or geometry effects (e.g., effect of upstream elbow). gth of closed branch This parameter is measured during construction or in the test setup based on the as-built configuration. de diameter This parameter is measured during construction or in the test setup based on the nected to tee as-built configuration. tion uhal Number Specific geometric considerations are used to determine the best-estimate value and uncertainty range for the Strouhal number. pyright 2022 by NuScale Power, LLC 207
Table 4-13 Decay Heat Removal System Steam Pipe Acoustic Resonance Margin Uncertainty Method p Description Procedure Analytic derivative of 0.25X 2 Equation 4-30 with respect to -------------------------------------------- ( X 3 + 0.3X 2 )X 4 X 5 X1 Uncertainty in X 1 Uncertainty in the speed of sound is estimated based on the range of possible pressures, temperature and flow rates in the test. Analytic derivative of 0.25X 1 X 3 Equation 4-30 with respect to ----------------------------------------------- 2 X2 ( X 3 + 0.3X 2 ) X 4 X 5 Uncertainty in X 2 Uncertainty in the inner diameter of the cavity piping. Analytic derivative of -0.25X 1 X 2 Equation 4-30 with respect to ----------------------------------------------- 2 X3 ( X 3 + 0.3X 2 ) X 4 X 5 Uncertainty in X 3 Uncertainty in the length of the cavity. Analytic derivative of 0.25X 1 X 2 Equation 4-30 with respect to -------------------------------------------- 2 X4 ( X 3 + 0.3X 2 )X 4 X 5 Uncertainty in X 4 Uncertainty in the steam flow velocity. Analytic derivative of 0.25X 1 X 2 Equation 4-30 with respect to -------------------------------------------- 2 X5 ( X 3 + 0.3X 2 )X 4 X 5 Uncertainty in X 5 Uncertainty in the Strouhal number. Equation 4-11 2 2 2 2 2 ( 1 ux ) + ( 2 ux ) + ( 3 ux ) + ( 4 ux ) + ( 5 ux ) 1 2 3 4 5 Experimental Bias Considerations Some of the NuScale CVAP tests are performed on functionally prototypic mockups of the NPM components, as opposed to in-situ with the plant at normal operation. The following tests have reasonable differences from the plant configuration: Steam generator FIV testing is performed on a full-size subset of the total number of tubes corresponding to the 9th to 13th columns of the actual steam generator. Unheated water flow about the tubes (on the outside) provides the excitation forces, and no fluid is present within the tubes (on the inside). 1 Steam Generator Testing The geometry of the helical coil tubes used in the FIV testing is identical to the components used in the NPM. The boundary conditions in terms of supports and pyright 2022 by NuScale Power, LLC 208
water flow in the reactor primary side is simulated by water entering the lower part of the inner vessel, rising upwards and turning at the top into the annulus to cross the tube bundle in a downward sense. Nonetheless, there are practical considerations that make the test specimen differ from the NPM so those differences are considered. The following features introduce bias into the design analysis that relies on test measurements: The circulating water on the exterior of the tubes is at room temperature The helical coil tubes are not filled with boiling water, but contain instruments and cables The test specimen contains five tube columns instead of twenty-one in the full assembly. The objective of the testing is to collect data and confirm that VS and FEI are not active, that inputs to the analysis of the different FIV mechanisms (e.g., natural frequency, damping ratio, and so on) are justified, and that vibration amplitude predictions for TB are bounding. 1.1 Approach for Testing and Confirmation of Turbulent Buffeting Characteristics The natural frequency in water is a parameter in the calculation of safety margin against fatigue from alternating stress induced by random vibrations on the SG tubes, and its uncertainty is accounted for by comparison of simulations and measurements. The simulation accurately models the empty tube and specify the appropriate water temperature as the test to make the comparison valid. The structural or operating features of the test that deviate from the NPM design result in a different average frequency as determined in the design analysis. Therefore, the modal analysis of the test apparatus is compared to the modal analysis of the NPM to understand any differences. The effective mass of tube is another parameter that is affected by the test vs. NPM differences. Its influence on vibration appears in the measured frequency and mean square displacement amplitude. As discussed earlier, the pre-test prediction determines the modal characteristics and compares to the design analysis. Lastly, the maximum RMS displacement is a parameter in the safety margin calculation for the design analysis. The methodology uses an upper bound PSD approach from literature, so the expectation is that measured RMS responses are less than in the pre-test prediction (i.e., positive bias). To comprehensively consider the uncertainty in RMS displacement, the PSD analysis is validated against the recorded dynamic pressure data in the test to confirm that the literature approach is bounding for the HCSG under prototypical NPM velocities. Numerical evaluation of the differences between the literature pyright 2022 by NuScale Power, LLC 209
uncertainty and incorporate bias. 1.2 Approach for Testing and Confirmation of Vortex Shedding Characteristics The signature of VS is a periodic oscillation in the pressure field that translates to a distinct frequency in the PSD. If the frequency of the periodic driving force matches the natural frequency of a tube, then resonance occurs and the amplitude of the vibration becomes significant. Vortex shedding characteristics in terms of the coherent structure of the eddies are strongly dependent on Reynolds number. The SG testing, although only a subset of the total number of tubes, allows detection of this phenomenon if it occurs for a range of velocities that are considered based on flow rate changes with power in the NPM. The differences relative to the NPM design in terms of circulating water temperature and empty tubes, and their potential impact on VS are considered as follows: The water temperature affects the fluid density and viscosity, which influence the Reynolds number (the tube diameter and flow speed are similar to NPM conditions). Therefore, the pre-test prediction evaluates those changes in Reynolds number, and ensures that the test specifies sufficient flow rate changes to cover the possible range of Reynolds number at plant conditions. The vibration mode shape and natural frequency of the tube are important parameters that affect the tubes response to periodic flow excitations and potential for phase synchronization. As discussed earlier, the pre-test prediction determines the modal characteristics of the test apparatus and compares to the design analysis. 1.3 Approach for Testing and Confirmation of Fluid Elastic Instability Characteristics Fluid elastic instability is an intense vibration regime that causes significant motion and tube wear much greater than vibration caused by turbulence. Analytically, the onset of FEI is determined when the pitch velocity is greater than an empirically established critical velocity (evaluated using Connors coefficient). In order to comprehensively consider other parameters that may be affected by experimental biases, the following are evaluated: The effective pitch velocity is dependent on the vibration mode and natural frequency of the tube. The pre-test prediction determines the modal characteristics of the test apparatus and compares to the design analysis. The comparison ensures that the test specifies sufficient flow rate changes to compensate for differences that result in a decrease of the reduced pitch velocity in the test relative to the NPM design. The critical velocity is dependent on the tubes total mass per unit length (including hydrodynamic mass), damping ratio, and density, which may deviate from the normal operating state. The damping ratio is a measured pyright 2022 by NuScale Power, LLC 210
of safety margin calculation. The empty tubes provide a positive bias because that reduces the critical velocity (mass is in the numerator), and hydrodynamic mass is approximately 1/10 the metal mass. On the other hand, the test temperature provides a negative bias (density is in the denominator) because the test water density is greater than in the NPM. Therefore, the pre-test prediction compares values of the critical velocity calculated for the test apparatus and the NPM design. Expected Results and Validation Range of Experimental Results The purpose of the pre-test prediction is to calculate the expected experimental results and determine a range of results that are acceptable to validate the design analysis. If the experimental design exactly matches the conditions of the design analysis, and if there were no errors in the experimental measurements or modeling, the experimental result would exactly match the design analysis result. Distortions exist for most experiments, and the effect of the positive and negative distortion values on the range of validation safety margins and experimental results is determined. For example, testing at a lower pressure than the pressure associated with the limiting design condition is considered a distortion, if pressure affects the results of the analysis. To quantify the effect of the distortion, the design analysis calculation is re-performed, using the standard design analysis methods, at the test pressure and a new safety margin is determined. The difference between the best-estimate safety margin and the safety margin determined at the test pressure is the distortion adjustment. If the testing condition is more limiting than the design analysis condition, the value of the adjustment is positive and increases the validation range. Each distortion is individually calculated and then the safety margins are added together in Equation 4-31. The input, numerical and measurement uncertainties are calculated based on the guidelines in Section 4.5.1. These uncertainties are calculated in terms of the safety margin and are combined in Equation 4-31. The effect of the design analysis safety margin, distortions and uncertainties on the allowable range of the safety margin is described by Equation 4-31. The design analysis safety margin is provided as an upper bound to the range to provide a reasonable cutoff for the validation. For the lower bound of the range, the design analysis safety margin is not considered to ensure that even accounting for uncertainties there is positive margin to the onset of the phenomena at normal operating conditions. The expected result for each measured parameter are documented and an allowable validation range using the result of Equation 4-31 for that parameter specified. Range Upper = SM BE - SM d + SM DA + SM u Equation 4-31 Range Lower = SM BE - SM d - SM u pyright 2022 by NuScale Power, LLC 211
Range = Range of safety margins for validation (%), SM BE = Best-estimate safety margin (%), SM DA = Safety margin from the design analysis (%), SM d = Safety margin adjustment from distortions (%), and SM u = Safety margin adjustment from total uncertainty (%). Note that it could be possible to justify a higher upper bound for the validation range, since an experimental result that falls above the upper bound indicates that there are un-realized conservatisms in the design analysis. Particularly for design analysis results with small safety margins or known, highly conservative inputs, this approach should be considered to determine an alternate upper bound instead of the approach provided in Equation 4-31. Summary The NuScale Power CVAP Measurement Program relies on diverse testing campaigns to verify the structural integrity of NPM components that are evaluated in the Analysis Program to have a margin of safety less than 100 percent or contain novel design changes to preclude FIV that warrant validation testing. This section provides an approach to validate the analysis methods against the experimental results, and methodologies to quantify bias and uncertainty embodied with the results of the design analyses and the measurement program. Table 4-14 summarizes the components discussed in the above sections, and their analysis validation approach. When testing is complete, post-test analysis is performed to assess the experimental results and finalize the validation effort. The pre-test prediction provides a level of confidence in the test design and its ability to validate the design analysis. Differences between the pre-test predictions and experimental results are adjudicated in the post-test analysis. pyright 2022 by NuScale Power, LLC 212
able 4-14 Summary of Components and Flow Induced Vibration Analysis Validation Methods Margin Relative Variables with Uncertainty to omponent FIV Mechanism Validation Test To be Measured Speed of sound TS Steam Critical Strouhal AR Initial Startup Flow velocity Tees Number Cavity diameter and length Tube diameter and mass Reduced Mode shapes and natural SG VS Prototypic Damping frequencies Damping ratio in air Tube diameter and mass Mode shapes and natural Impact Fatigue frequencies SG TB Prototypic Usage Mean square response PSD Damping ratio in water Flow velocity Mode shapes and natural frequencies SG FEI Prototypic Stability Ratio Tube diameter and mass Damping Primary fluid density pyright 2022 by NuScale Power, LLC 213
The following sections provide a summary of the validation testing planned for the steam generator, and CNTS steam tees. The test design, as specified in the testing needs document, is discussed for each test. TF-3 Validation Test 1 TF-3 Testing Overview The general objective of the TF-3 test is to obtain vibration test data for a partially prototypic NuScale HCSG for validation of FEI, VS, and TB design analyses. The specific test program objectives are:
- 1. Determine in-air natural frequencies and mode shapes of the HCSG tubes; this includes ability to characterize modes of the tube bundle assembly (synchronized motion of full tube bundle and supports).
- 2. Determine in-water natural frequencies and mode shapes of the HCSG tubes and supports; this includes ability to characterize modes of the tube bundle assembly (synchronized motion of full tube bundle and supports).
- 3. Determine in-air and in-water damping values for a range of representative mode frequencies and vibration amplitudes.
- 4. Obtain data to characterize primary flow dynamic pressure fluctuations, SG tube and tube support vibration amplitudes for a range of primary flow conditions.
- 5. Obtain high flow rate vibration amplitudes to demonstrate margins to FEI and VS.
The TF-3 test program consists of modal testing during the fabrication process, modal testing on the completed test assembly, and flow testing as shown in Table 5-1. Table 5-1 Summary of Tests Test Type ir natural frequencies and mode Partial-tube array pe testing (during fabrication) ir damping tests (during fabrication) Partial-tube Array ir natural frequencies and mode Full-tube Array pe testing ir damping tests Full-tube array ater natural frequencies and mode Full-tube array pe testing ater damping tests Full-tube array ady-state flow testing Full-tube array, flow flow testing Full-tube array, flow flow testing Full-tube array, flow pyright 2022 by NuScale Power, LLC 214
performed following installation of a complete helical column, but before installation of the next column. Temporary support rings are used during this testing to simulate the restraint provided by the riser in a fully-assembled tube bundle. This testing includes measurements using both permanently-installed and temporarily-mounted accelerometers. Individual tube testing during assembly allows exploration of the various boundary conditions that may exist. Access limitations exist in a fully-assembled tube bundle array test, both for instrumentation placement and for means to perform excitation of tubes. Single-tube tests allow greater ability to examine variability in boundary conditions between tubes, based on the flexibility to use temporary, removable accelerometers that can be iteratively employed. Permanent sensor arrangement in the tube array cannot be altered once the array is constructed. Performing modal measurements during fabrication allows confirmation that sensor function is as expected. Tube array testing is performed using a representative tube bundle array, constructed using functionally prototypic supports. Array testing includes in-air modal and damping tests, performed as part of individual tube testing during fabrication of the test specimen. The array testing also includes in-water modal and damping tests and flow testing over the full range of nominal design operating conditions, as well as at FEI and VS on-set conditions. This testing is a comprehensive performance demonstration of the FIV design of the NuScale SG. A full-array test provides the most prototypic platform to characterize modal frequencies, shapes and damping for the NuScale design. In order to directly evaluate FEI, VS or confirm primary side flow PSDs for the NuScale design, flow testing of a prototypic tube array is necessary. 1.1 TF-3 Test Specimen The scope of the TF-3 test is the SG (including tubes and tube support structures) both as a fully-assembled tube array and during the fabrication process at points where full-helical columns are installed. This section identifies the design aspects that must be prototypic. Temporary Supports for Individual Tube Testing during Fabrication The in-process tube array is used for testing during fabrication; therefore, requirements for the full tube array (described in the following section) apply. During this phase of the testing, temporary support structure is provided to simulate the constraint provided by the riser in a fully assembled tube bundle. This structure is designed to interface with the innermost installed column of tube supports. This structure is adjustable to allow variation in the applied compression and to allow use at various intermediate steps in the specimen fabrication (i.e., be able to be deployed as each column is installed, and so on). pyright 2022 by NuScale Power, LLC 215
Confinement of flow through the SG tube bundle is represented, with a prototypic upper riser (including riser-to-SG-tube-support clearance) and reactor vessel (including vessel-to-SG-tube-support clearances). The full test specimen includes a mechanism to provide adjustable compression of the tube supports between the riser and the vessel wall. This mechanism is functional (including ability to adjust compression) during testing conditions, including flow testing. Prototypic geometry upstream of the SG is maintained by representing fluid confinement from the outlet of the riser to the inlet of the SG tube bundle. Flow downstream of the SG is maintained as balanced annular flow using a pressure drop plate or other feature to remove the effect of constrictions (transition of flow from an annulus to an exit pipe or plenum) at the outlet of the test fixture. The design of the SG tube array itself, including tube geometry and length, design of tube supports and physical interfaces between the tubes and tube supports and the riser and vessel (interfaces), is functionally prototypic unless otherwise noted below. The main departure from prototypic design geometry is that only helical Columns 9 through 13 are included. This implicitly requires the diameter of the riser and vessel to be scaled accordingly so that prototypic gaps between the riser and vessel, and between the inner and outer tube columns, are maintained. Other minor allowed deviations are described below. The design of tube array test fixture is in accordance with Table 5-2. Table 5-2 Tube Array Design Geometry Drawing Design Elements am generator
- SG tube geometry (including OD and wall thickness)
- SG tube arrangement (e.g., pitch) d plenum access port
- Feed plenum tube layout
- Geometry of vessel and tubesheet am generator tube supports
- SG tube supports er RPV section
- Steam plenum tube sheet layout
- Upper and lower SG supports
- Upstream SG flow geometry (pressurizer baffle plate)
- Downstream SG flow geometry (riser)
- Interface between outer column tube supports and the vessel; this implicitly includes maintaining prototypic separation between outermost tube column and vessel ctor vessel internals - upper riser
- Interface between inner column tube supports and riser; this implicitly includes maintaining prototypic separation between innermost tube column and riser.
- Riser geometry and supports (upstream SG flow geometry)
- Downstream SG flow geometry (riser)
The tubes in the test fixture include the entire helix, transition, and straight lengths. A general summary of tube requirements for this test are provided in Table 5-3. Tube lengths upstream of the shell-side face of the feed plenum and downstream of the shell-side face of the steam plenum are reduced because a pyright 2022 by NuScale Power, LLC 216
plenums. Table 5-3 Helical Steam Generator Tube Array Details Helical # of Helical Radius Length (in.) Column Tubes (in.) (( 9 10 11 12 13
}}2(a),(c),ECI The SG tubes for this testing use 304 or 316 stainless steel or other materials as proposed by the supplier and approved by NuScale. The SG tubes are 0.625-inch OD with a 0.050-inch wall thickness.
Test facility operating conditions are provided in Table 5-4. Table 5-4 TF-3 Test Facility Operating Conditions Parameter Maximum Minimum Nominal (( ary-side temperature(1) ary-side pressure(2) ary-side flow ondary-side temperature ondary-side pressure ondary-side flow
}}2(a),(c),ECI es: (1) Maximum primary-side temperature is based on accommodating heatup due to flow resistance without the need for dedicated cooling. Maximum design temperature may be reduced provided required flow rates are accommodated.
(2) Maximum primary-side pressure is based on providing margin to accommodate required flow. Maximum design pressure for test fixture or system may be reduced, provided required flow rates are accommodated (3) This volumetric flow corresponds to a gap velocity of (( }}2(a),(c),ECI. (4) This volumetric flow corresponds to a gap velocity of (( }}2(a),(c),ECI. Lower minimum flow capacity is acceptable. The minimum flow requirement is based on allowing flexibility in selecting pump and pump controls. Testing at flows less than minimum flow are not planned pyright 2022 by NuScale Power, LLC 217
The purpose of the TF-3 test is to obtain data on the vibration characteristics of the SG tubes; therefore, most test instrumentation is applied to the SG tubes, with a secondary emphasis on the tube supports and the vessel. Both strain gauges and accelerometers are used to characterize tube vibration amplitudes and associated frequencies: PCB Piezotronics (Model number W356A03, 10mV/g sensitivity, 2-5000Hz frequency range) tri-axial accelerometers are used for tube accelerometer instrumentation. Strain gauges (HBM 1-LY65-3/350) are used for tube strain gauge instrumentation. A total of eight tubes selected from Columns 9, 11 and 12 are instrumented. Two tubes are permanently instrumented with strain gauges and accelerometers, four tubes are instrumented with accelerometers only, and two tubes are instrumented with strain gauges only. Based on the potential for varying boundary conditions (fixed, sliding, hybrid of sliding and fixed) for any given tube or span, the general approach for instrumentation is to provide some instrumentation on as many spans as possible to maximize the capability to capture measurements of the full possible range of tube vibration modes. The total number of instrumented tubes and the distribution of instrumentation among the selected tubes provides a balance between obtaining intensive data for a representative number of tubes (tubes with both strain gauges and accelerometers), having the capability to assess influence of the types of instrumentation on data obtained (mix of tubes instrumented with only one or both types of instruments), and obtaining data on a broader sample of tubes (spreading total available instrumentation among a greater total number of tubes). There was also a preference to place instrumentation in spans that have an exciter coupling (or are proximate to the coupled spans) and a need to place multiple accelerometers on some spans to better characterize some higher order modes (anti-node(s) occur within the span, not only at supports) that have high relative mass participation factors. Each of the six instrumented tubes have a maximum of 10 accelerometers based on instrument cable constriction limitations. Accelerometers are placed in locations expected to have the largest vibration magnitude for the most dominant (highest mass participation factor) modes. For three accelerometer-instrumented tubes (one in each instrumented column), two accelerometers are designated for placement at a tube support, to ensure capability to detect rigid body motion. There are accelerometers placed on the tube supports that are paired with the tube accelerometers placed at a support. Strain gauges are placed in pairs at a given location, with one strain gauge on the tube extrados (oriented along major axis of the tube, on outside of major curvature) and one strain gauge on the top of the tube (oriented along major axis pyright 2022 by NuScale Power, LLC 218
placements characterize the maximum strains associated with tube bending deflections perpendicular to the axis of the tube at a given location. Specific placement of strain gauges is based on locations of maximum predicted strain associated with most dominant tube frequencies (maximum mass participation factor). Detailed descriptions of the specific instrument locations on each tube are provided in the following sections. The general basis for the selection of the tubes for instrumentation is as follows:
- Column 12. Because the outermost tube column has the highest susceptibility to FEI, two tubes from Column 12 are to be instrumented to provide the capability to characterize FEI.
- Column 11. The center column of the tube bundle is the most removed from the tube bundle interfaces (vessel and riser wall) and should represent the most nominal flow conditions. Design of the test specimen includes access ports that allow coupling up to four tubes to harmonic exciters. Therefore, four tubes are selected for instrumentation in this column to provide the greatest amount of instrumentation in the most representative column and to allow maximum utilization of the harmonic exciter. The harmonic exciter is the only means of performing modal measurements on tubes in the fully-assembled test specimen.
- Column 9. This is the innermost tube column in the specimen. In order to fully characterize the vibration response of the tube bundle, potential variations in vibration response due to flow differences along the perimeter (i.e., riser surface) are assessed. Two tubes are selected for instrumentation in this column. Tubes at the bottom periphery of the tube bundle (including the tubes in the innermost column) are potentially susceptible to VS.
A summary of instrumented tubes is provided in the following tables. pyright 2022 by NuScale Power, LLC 219
Table 5-5 Instrumented Tubes (Accelerometers and Strain Gauges) st Tube # Location Description Notes Column 11, tube 1 at steam plenum 4 (305.4°) This corresponds to the position of 1 Instrumented tube N.1. Column 11, tube 16 at steam plenum 2 (144.6°) This corresponds to the position of 2 Instrumented tube N.16. Table 5-6 Instrumented Tubes (Accelerometers Only) st Tube # Location Description Notes 3 Column 9 tube 15 at steam plenum 1 (~45°) To evaluate VS 4 Column 9 tube 16 at steam plenum 1 (~45°) To evaluate VS Column 12, tube 1 at steam plenum 3 (~225°) To evaluation limiting location 5 for FEI Column 12, tube 2 at steam plenum 3 (~225°) To evaluation limiting location 6 for FEI Table 5-7 Instrumented Tubes (Strain Gauges Only) st Tube # Location Description Notes Column 11, tube 2 at steam plenum 4 (307.1°) This corresponds to the position of 7 Instrumented tube N.2. Column 11, tube 15 at steam plenum 2 (144.3°) This corresponds to the position of 8 Instrumented tube N.15. At least ten tri-axial accelerometers, including mounts and adapters suitable to support removable placement at various locations on individual tubes, are available for modal testing to be performed during test specimen fabrication. Tri-axial accelerometers are used for permanent instrumentation of the tubes. Higher sensitivity (100 mV/g) is desired for the testing using removable accelerometers. Instrumented tubes (Columns 9, 11 and 12) contain 3.75 helical turns, which result in a total of 31 tube spans (length of tubing between two adjacent support) for each tube. Based on the non-symmetric placement of the tube supports, spans are alternatively long (64-degree arc) and short (26-degree arc). The first and subsequent odd-numbered spans are long and even-numbered spans are short. For the purpose of identifying instrument locations for the test specimen, instrument locations are identified to a specific span, based on numbering that originates at the steam plenum. The steam plenum numbering is based on FW plenum 1 being oriented at 45 degrees with the steam plenum numbering proceeding in a clockwise direction (steam plenum 2 is located at 135 degrees, and so on). For example, Span 1 is a long span and represents the tube span from the FW plenum steam tubesheet to the first support (e.g., steam transition bend). Span 31 is also a long span and represents the span of tubing from the last support to the FW tubesheet. pyright 2022 by NuScale Power, LLC 220
corresponding to the tube support located immediately counter-clockwise from steam plenum 1 and SG tube support numbering proceeds in a clockwise manner, such that positions #1 and #2 are on each side of steam plenum 1. Figure A-1 provides a figure illustrating this numbering scheme. Accelerometer placement is in accordance with the following tables and Figure A-1. Table 5-8 Accelerometers Placement for Instrumented Tube #1 elerometer Location Notes Span 1 (steam transition), A1 Fundamental mode (fixed) mid span A2 Span 11: mid span To detect prevalent fixed and sliding lower order modes To monitor for rigid body motion of the tube, e.g., movement of the tube due to FIV of the tube support (to A4 Span 14/15 at support which the tube is coupled), so that this can be differentiated from FIV of the tube itself. A3 Span 13: mid span Span 15 has harmonic exciter coupling A5 Span 15: mid span (nozzle N3) A6 Span 25: mid span To detect prevalent fixed and sliding lower order modes A8 Span 26/27 at support To monitor for rigid body motion of the tube A7 Span 27: mid span To detect prevalent fixed and sliding lower order modes A9 Span 27: 1/2 span to exciter location Span 29 has harmonic exciter coupling (nozzle N1), VS A10 Span 29: mid span susceptible location Table 5-9 Accelerometers Placement for Instrumented Tube #2 elerometer Location Notes Span 1 (steam transition): A11 To detect prevalent fixed and sliding lower order modes mid:span A12 Span 3: mid span Span 3 has harmonic exciter coupling (nozzle N4) A13 Span 5: 1/2 span A14 Span 11: mid span To detect prevalent fixed and sliding higher order modes, A15 Span 13: mid span proximate to exciter A16 Span 15: mid span Span 15 has harmonic exciter coupling (nozzle N2) To detect prevalent fixed and sliding lower order modes, A17 Span 17: mid span proximate to exciter To detect prevalent fixed and sliding lower order modes, A18 Span 25: mid span proximate to exciter A19 Span 26: mid span To detect response of short spans To detect prevalent fixed and sliding lower order modes, A20 Span 27: mid span proximate to exciter pyright 2022 by NuScale Power, LLC 221
Table 5-10 Accelerometers Placement for Instrumented Tube #3 elerometer Location Notes A21 Span 3: mid span To detect prevalent fixed and sliding lower order modes A22 Span 5: mid span To monitor for prevalent higher order modes A23 Span 7: mid span A24 Span 11: mid span To monitor for prevalent first order modes A25 Span 13: mid span To monitor for prevalent first order modes A26 Span 21: mid span To monitor for prevalent first order modes A27 Span 23: mid span A28 Span 27: mid span To monitor for prevalent first order modes A29 Span 30: mid span First short span from feed transition, susceptible to VS Span 31 (feed transition): VS shedding susceptible location, fundamental mode A30 mid span response (fixed) Table 5-11 Accelerometers Placement for Instrumented Tube #4 elerometer Location Notes A31 Span 7: mid span To detect prevalent fixed and sliding lower order modes A32 Span 11: mid span To monitor for prevalent higher order modes A33 Span 13: mid span A34 Span 13/14 at support To monitor for rigid body motion of the tube A35 Span 15: mid span To monitor for prevalent first order modes A36 Span 21: mid span To monitor for prevalent first order modes A37 Span 22/23 at support To monitor for rigid body motion of the tube A38 Span 23: mid span To monitor for prevalent first order modes A39 Span 29: mid span First long span from feed transition, susceptible to VS Span 31 (feed transition): VS shedding susceptible location, fundamental mode A40 mid span response (fixed) Table 5-12 Accelerometers Placement for Instrumented Tube #5 elerometer Location Notes Span 1 (steam A41 Fundamental mode (sliding) transition): mid span A42 Span 5: mid span To monitor for prevalent first order modes A43 Span 7: 1/3 span To monitor for higher order modes A44 Span 7: mid span A45 Span 9: mid span To monitor for prevalent first order modes A46 Span 21: mid span To monitor for prevalent first order modes A47 Span 22: mid span To monitor for first order response mode of short span A48 Span 23: mid span To monitor for prevalent first order modes A49 Span 29: mid span To monitor for prevalent first order modes Span 31 (feed transition): A50 VS shedding susceptible location mid-span pyright 2022 by NuScale Power, LLC 222
Table 5-13 Accelerometers Placement for Instrumented Tube #6 elerometer Location Notes Span 1 (steam transition): A51 Fundamental mode (sliding) mid span A52 Span 13: mid span To monitor for prevalent first order modes A53 Span 13/14 at support To monitor for rigid body motion of the tube A54 Span 15: 1/3 span To monitor for higher order modes and for prevalent first A55 Span 15: mid span order modes A56 Span 21: mid span To monitor for prevalent first order modes A57 Span 23: 1/3 span To monitor for higher order modes and for prevalent first A58 Span 23: mid span order modes A59 Span 23/24 at support To monitor for rigid body motion of the tube A60 Span 25: mid span To monitor for prevalent first order modes A total of 32 uni-axial strain gauges are mounted on SG tubes as shown in Table 5-14. Figure A-1 provides a composite view of strain gauge placements. Strain gauge placements are based on measuring strains associated with most prevalent vertical modes. Strain gauges are generally placed in spans that do not include accelerometers. This is both to increase the extent of specimen that contains at least some instrumentation and to minimize potential impact of instrumentation on measurements (heavily concentrating instrumentation in a single span further alters vibration response as compared to a span with no added mass due to instrumentation). Table 5-14 Tube Strain Gauge Placements Strain Gauge trumented Tube # Location Number S1, S2 Span 1, at steam plenum tubesheet face column 11, tube 1 S3, S4 S5, S6 Span 15, each end, as close as practical to tube support S7, S8 Span 31, as close as practical to tube support Span 15, as close as practical to tube support between S9, S10 span 15/16 column 11, tube 16 Span 17, as close as practical to tube support between S11, S12 span 17/18 S13, S14 S15, S16 Span 27, each end, as close as practical to tube support Span 11, as close as practical to tube support between S17, S18 span 11/12 Span 13, each end, as close as practical to tube support S19, S20 between span 12/13 column 11, tube 2 Span 15, as close as practical to tube support between S21, S22 span 15/16 Span 29, as close as practical to tube support between S23, S24 span 29/30 S25, S26 Span 31, as close as practical to tube support pyright 2022 by NuScale Power, LLC 223
Strain Gauge trumented Tube # Location Number Span 3, as close as practical to tube support between S27, S28 span 3/4 column 11, tube 15 Span 15, as close as practical to tube support between S29, S30 span 14/15 S31, S32 Span 31, as close as practical to tube support There are six SG tube accelerometers (as shown in Table 5-15) that are coincident with a SG tube support. An accelerometer is placed on the corresponding tube support at a location coincident (to the extent practical) with each of these SG tube accelerometers. Table 5-15 Steam Generator Tube Support Accelerometer Corresponding SG Tube Radial Tube Support SG Tube Instrumented Column Support Position Position celerometer Tube # (Circumferential) (Radial) SG tube support A3 1 11 Circumferential position #6 between Col 11/12 SG tube support A8 1 11 Circumferential position #2 between Col 11/12 SG tube support A34 4 9 Circumferential position #7 between Col 9/10 SG tube support A37 4 9 Circumferential position #8 between Col 9/10 SG tube support A53 6 12 Circumferential position #2 between Col 12/13 SG tube support A59 6 12 Circumferential position #8 between Col 12/13 Based on Table 5-15, there are a total of four upper and lower SG supports that have corresponding tube accelerometers to monitor for rigid body motions of the SG tube supports. Accelerometers are placed on both the upper and lower SG supports at each of these locations (eight accelerometers). Likewise, a paired accelerometer is placed on the SG tube support along with each of these accelerometers (eight accelerometers). Sixteen accelerometers are placed on the vessels to evaluate potential modal responses of the entire test specimen. Based on pre-test modal analysis, depending on the boundary conditions that exist, the vessel shells respond in beam modes (( }}2(a),(c),ECI, shell modes (( }}2(a),(c),ECI or torsional modes (( }} 2(a),(c),ECI . Eight inside of the inner vessel (i.e., riser), two sets of four accelerometers (at approximately 1/10, 2/5, 3/5 and 9/10 heights, vertically in-line), offset circumferentially by 90 degrees. Eight outside of primary vessel, two sets of four accelerometers (at approximately 1/10, 2/5. 3/5 and 9/10 height vertically in-line), offset circumferentially by 90 degrees. pyright 2022 by NuScale Power, LLC 224
to characterize a turbulent pressure force PSD for steady flow conditions. Other sensors included in this test are documented in Table 5-16. Table 5-16 Other Pressure and Temperature Instrumentation Total easurement Description Number Temperature 2 Primary-side inlet and outlet temperature Pressure 5 Primary-side static pressure spaced over height Primary-side pressure drop spaced over height (pressure drop Differential 4 measurements are developed based on differentials between five pressure pressure instruments) Flow 2 Primary-side flow For the purposes of this testing, frequencies up to 500 Hz are of interest; therefore, DAS sampling rates for modal and damping tests are at least 2000 Hz. Lower DAS operating frequencies are permitted for FEI and VS tests (these phenomena are associated with frequencies (( }}2(a),(c),ECI. Damping tests require DAS capabilities to provide near real-time calculated damping measurements. This includes the DAS being pre-programmed to determine transfer function(s) for fixed exciter damping tests. The DAS is capable of measuring both large-amplitude, slow-changing strain (i.e., static strain) and low-amplitude, rapid-cycling strain (i.e., dynamic strain). Evaluation of the FIV phenomena described herein requires only dynamic strains; however, knowledge of static strain changes and variations throughout testing may be of benefit in understanding boundary conditions at the support interface points. Before each dynamic strain measurement (e.g., damping, modal- or flow-test dataset), the static and average strains are measured to establish a baseline. In order to facilitate accurate measurement of both static and dynamic strains, the DAS includes one of the following features or equivalent measures to address accurate measurement: Capability to measure with high-range (low-gain) settings for static strains, output the analog (raw) signals, and use separate hardware to measure with low-range (high-gain) settings for dynamic strains (high-gain measurements is zero-centered for ease of data post-processing). Capability to auto-balance (null-calibrate) each channel, with procedures or software features to re-zero strain signals before each test series dataset (to avoid signal saturation). The DAS must record all such adjustments such that an accurate static strain representation can be reproduced. pyright 2022 by NuScale Power, LLC 225
1.3.1 Partial-Tube Array Testing During Fabrication In-air modal testing is performed on individual tubes in fully assembled columns during the fabrication process. Testing is performed after Columns 11 and 12 are fully installed, including associated tube supports. Testing is performed with up to three variable levels of compression applied by temporary tube bundle supports. Testing with variable levels of support compression evaluates potential variations in frequency that could result from differences in the boundary conditions between the SG tubes and supports. Tube excitation using impulse hammer strikes or other means such as a harmonic exciter, is utilized for partial-tube array testing performed during test specimen fabrication. Access to perform impulse tests is precluded once test specimen is fully assembled with the riser installed. Fixed point excitation capabilities using a harmonic exciter is provided for instrumented tubes #1, #2, #7 and #8 (column 11) through access nozzles N1, N2, N3, and N4. Fixed point excitation is only used for full-tube array tests (in-air and in-water modal and damping tests). The harmonic exciter is capable of providing a range of displacements and operating over a frequency range of at least 5 to 500 Hz. There are three technical objectives of the modal testing during tube bundle fabrication: Reproducibly measure mode frequencies and mode shapes (goal of characterization of up to four mode shapes and frequencies per tube) using impulse excitation. Evaluate range of tube modal responses. Based on preliminary modal measurements, a range of frequencies (representing sliding and fixed boundary conditions) are expected. Obtain consistent modal measurements for temporary and permanent instrumentation. Testing is conducted by placing temporary accelerometers, taking measurements and then repeating as necessary to obtain required data. The specific placement of temporary accelerometers to perform this portion of the testing is determined by the test performer as necessary to meet the test objectives. Necessary modal measurements are achieved if at least two modes can be measured for a given tube. Three-dimensional mode shape functions are developed for as many modes as possible (minimum of the first two local modes), including at least one pyright 2022 by NuScale Power, LLC 226
and frequency that can be characterized for each variable level of temporary support compression, at least two sets of duplicate measurements are obtained to ensure consistent results. Accelerometers are not required to be re-positioned to obtain duplicate data sets. Mode shapes are defined in a local coordinate system. Locations of removable accelerometers are recorded to within an accuracy of (( }}2(a),(c),ECI inches. Rotational position (e.g., bottom dead center, top dead center, and so on) of each span tested is also recorded to evaluate the impact of the horizontal orientation of the test fixture. In addition to data from removable accelerometers, data are recorded using permanently installed strain gauges and accelerometers in the tested tubes. 1.3.2 In-Air Damping Tests Damping tests are performed on the same individual tubes that are characterized as part of partial tube array fabrication modal testing. This includes testing performed with up to three variable levels of compression applied by temporary tube bundle supports. Tests are performed on tubes in Columns 11 and 12. The extent of damping measurements in each column is based on the number of individual modes that are successfully characterized as part of the partial tube array fabrication modal testing. Damping values for each tube and mode are developed using both logarithmic decrement and half-power bandwidth methods. Damping testing includes evaluation of amplitude specific damping. The specific placement of temporary accelerometers to perform this portion of the testing is determined based on the test objectives. In addition to data from removable accelerometers, data are recorded using permanently installed strain gauges and accelerometers in the tested tubes. There are two primary objectives of this testing: Reproducibly measure damping values for modal frequencies characterized during partial tube array fabrication modal testing. Damping values are determined using at least two methods (logarithmic decrement and half-power bandwidth methods) Determine variability in damping values between different tubes and as a function of modal frequency. Minimum duration of data files is agreed upon by NuScale and the testing services supplier. Logarithmic decrement and half-power bandwidth methods produce different ranges of damping values (results from each method are not expected to show strong agreement). Therefore, testing objectives are satisfied when reproducible results are produced for each method, even if the results obtained for each method are not in agreement. pyright 2022 by NuScale Power, LLC 227
This testing obtains frequency and mode shape measurements comparable to those observed in the tests during tube bundle fabrication. With the exception of possible limited access for tubes #3 and #4 (Column 9), impulse excitation is not possible for these tests. Therefore, a fixed point exciter is used to measure mode frequencies and mode shapes of only tubes in Column 11 (Instrumented tube #1, #2, #7 and #8). Initial testing is performed by frequency sweep testing for frequencies between 5 and 500 Hz. Following this initial sweep testing, harmonic sweep testing around frequencies characterized during individual tube testing of Column 11 tubes is performed if broad spectrum sweep testing did not adequately characterize these modes. Measurements are obtained for each tube using excitation at each of the fixed points (two to three points per tube, based on access through N1, N2, N3 and N4). At least two duplicate sets of measurements are obtained for each tube at each discernable modal frequency with excitation from each exciter coupling location. In addition, excitation in frequency ranges corresponding to any predicted predominant frequencies not characterized during the individual tube testing is performed. This testing is to confirm the boundary conditions (as indicated by observed modal frequencies) in the fully-assembled tube array are consistent with those during the testing during the fabrication process. During this testing, it is only necessary to record data for the individual tubes to which excitation is applied. Repetition of some or all tests using altered tube bundle compression forces is performed. 1.3.4 Tube Array Testing: In-Water Natural Frequencies and Mode Shapes The objectives, conduct, and necessary data for this testing are identical to the full array in-air natural frequency and mode shape testing, with the exception that the test specimen be maintained full of stagnant water. 1.4 TF-3 Steady-State Flow Testing These tests are devoted to fully characterizing the vibration amplitudes of the HCSG as function of normal primary flow velocities. Measurements are also performed with no flow conditions to determine background noise levels. When increasing flow velocity, displacement amplitudes are monitored so that the test can be suspended if the onset of instability is observed. Because flows in this phase of testing are limited to 100 percent normal design flow, no unusual vibration is expected. The general requirements of this testing are steady flow runs of a minimum of five minutes at each test run, with recording of a dedicated test point data file for a pyright 2022 by NuScale Power, LLC 228
observed for at least three minutes (based on concurrence with the NuScale test engineer, as supported by relevant instrument readings), a minimum of two minutes of data are recorded to a dedicated file for the test point. Minimum testing points are in accordance with Table 5-17. Table 5-17 Steady-State Flow-induced Vibration Tests Test Run Flow (gpm) Description (( 1 2 3 4 1136 5 1514 6 1893 7 2271 8 2650 9 3028
}}2(a),(c),ECI The PSD functions cover a range of non-dimensional reduced frequencies (frequency*tube diameter/Vgap) up to 10. The flow test points in Table 5-19 provide data to validate the design PSD over the prescribed range of reduced frequencies.
Each steady-state flow test run is repeated on a different day from the first test run. If divergent results are observed for the duplicate tests, additional test runs may be requested until consistent results are observed. 1.5 TF-3 Vortex Shedding Flow Testing This test is to demonstrate that the SG design is not susceptible to VS phenomena. The VS can cause excessive vibration when alternating vortices are induced at a frequency that is at or near a modal frequency of the structure. The purpose of these tests is to evaluate flow rates which could induce VS at frequencies coincident with modal frequencies of individual tubes at the bottom periphery of the tube bundle by measuring vibration amplitudes that result during these flowrates. Flow rates necessary to cover the range of VS frequencies are provided in Table 5-18. Vortex shedding may occur when the Vgap is in proximity to a structural frequency (perfect coincidence is not necessary); therefore, testing at only specific points is not sufficient to characterize this phenomena. As there are many closely-spaced modal frequencies of the tubes within part of the range where tube natural frequencies and VS frequencies coincide, the testing approach requires taking sufficient data at flowrates throughout this range to fully pyright 2022 by NuScale Power, LLC 229
with respect to fine adjustment of flow velocities using high-capacity pumps; therefore, the proposed test sequence may be adjusted to facilitate available control capability. Table 5-18 Vortex Shedding Test Range1, 2 inimum Validation Flow Rate Expected Flow Rate for Onset Maximum Validation Flow Rate
}}2(a),(c),ECI es: (1) There is some overlap of the vortex shedding test conditions and steady-state and FEI test conditions. Common flow conditions that satisfy evaluation of all phenomena can be consolidated in the text matrix.
(2) Test points may be adjusted based on results of modal testing to ensure test runs most closely correspond with actual measured structural frequencies of the test array. The VS tests include a minimum of five minutes of steady-state flow with recording of a dedicated test point data file for a minimum of two minutes at each test point. Once steady flow conditions are observed for at least three minutes, a minimum of two minutes of data are recorded to a dedicated file for the test point. If large tube displacements (characteristic of VS) are observed, duration of data files may be reduced to prevent possible damage to the test specimen. The VS is associated with the fundamental or other lower frequency modes (less than (( }}2(a),(c),ECI), therefore data acquisition rate may be reduced for this testing. Flowrate changes between test points are controlled such that changes in flowrate do not exceed 10 percent per minute. Each VS flow test run is repeated on a different day from the first test run. If divergent results are observed for the duplicate tests, additional test runs may be requested until consistent results are observed. Likewise, if indication of VS occurs during transitions, additional test points are added to examine flow rates intermediate to the test points specified in Table 5-18. 1.6 TF-3 Fluid Elastic Instability Flow Testing This test is to demonstrate the onset (or lack of) of FEI in the NuScale HCSG to demonstrate design margin. However, the ability to demonstrate the onset of FEI in this test may be impossible for two reasons, 1) it is possible that a helical tube design is immune to FEI phenomena based on unique design features (de-tuning based on array including tubes of various frequencies, cannot be synchronously excited in a single mode), and 2) if margin to FEI is significantly larger than predicted, test facility flow rates are inadequate to reach FEI. Based on achieving maximum possible test flow rates without onset of FEI, adequate design margin is demonstrated. The general requirements of this testing are that steady-state flow runs of five minutes with recording of a dedicated test point data file for a minimum of two pyright 2022 by NuScale Power, LLC 230
three minutes, a minimum of two minutes of data are recorded to a dedicated file for the test point. If large tube displacements are observed, duration of data files may be reduced to prevent possible damage to the test specimen. Fluid elastic instability is associated with the fundamental or other lower frequency modes (less than (( }}2(a),(c),ECI for the NuScale design); therefore, data acquisition rate may be reduced for this testing. Flowrate changes between test points are controlled such changes in flowrate do not exceed 10 percent per minute. Fluid elastic instability is characterized by synchronized vibration of the tube array at a common frequency (predicted to be between (( }}2(a),(c),ECI for this test specimen). Continuous monitoring of vibration amplitude frequency of available strain gauges and accelerometers is necessary during FEI testing to accurately identify the onset of FEI, if it occurs. If the onset of FEI is observed, flow rates are immediately reduced and concurrence from NuScale obtained before subsequent increases in flow rates. Table 5-19 Fluid Elastic Instability Flow Tests nimum Validation Flow Rate Expected Flow Rate for Onset Maximum Validation Flow Rate
}}2(a),(c),ECI For FEI flow tests, each test run is repeated on a different day from the first test run. If divergent results are observed for the duplicate tests, additional test runs may be requested until consistent results are observed.
1.7 TF-3 Results Scope The Summary Test Report includes the following: test description list and discussion of test anomalies and test procedure changes assessment of test data acceptability relevant data plots or tables The Final Test Report includes the following: raw data calculations, including methodology and results summaries checksums for the non-encrypted reduced and calculated data sets reduced and calculated data in an encrypted file instrument calibration certificates pyright 2022 by NuScale Power, LLC 231
completed, filled out, and signed test procedures test logs CNTS Main Steam Line Branch Connections Validation Testing Initial startup testing is performed on the first NPM after the first fuel load. Due to the natural circulation design of the NPM, it is not possible to obtain the limiting thermal hydraulic conditions that are necessary to verify the FIV inputs and results until the NPM is operating near full power conditions. Initial startup testing is performed for a sufficient duration to ensure one million vibration cycles for the component with the lowest structural natural frequency. It takes less than one day of operation to obtain one million cycles of vibration. This is a conservative estimate because the lowest natural frequency of any component evaluated in the CVAP is (( }}2(a),(c),ECI (outer column SG tube). The initial startup test is performed with online vibration monitoring of the CNTS decay heat removal system (DHRS) steam piping and the CNTS main steam (MS) drain valve branch piping. In the event that an unacceptable vibration response develops any time during initial startup testing, the test conditions are adjusted to stop the vibration and the reason for the vibration anomaly investigated before continuing with the planned testing. Vibration amplitudes in the steam lines are measured to confirm that flow disrupter components effectively mitigate acoustic resonance (AR). 1 CNTS Main Steam Line Branch Connections Test Design The CNTS MS branch connection to the DHRS piping and the CNTS MS drain valve branch haveflow disrupters installed at the leading edge of the cavity entrances to preclude the coherent shedding of vortices. Testing is performed to ensure this design effectively mitigates acoustic resonance at these locations. These locations are branch lines where there is normally no flow during operation. As a group the DHRS steam piping tees and the MS drain valve branch are referred to as the CNTS main steam line branch connections. This section develops an approach and testing requirements to perform in-situ measurements in the piping outside the containment vessel head. Flow-excited ARs, where instabilities in the fluid flow excite acoustic modes within valves, stand pipes, or branch lines can play a significant role in producing mid- to high-frequency pressure amplification and vibration. Flow separation and generation of unstable shear layers at closed branch lines can sometimes lead to AR. To determine if there is a concern for AR in the design, the piping locations where this source of flow excitation is possible are identified and the Strouhal number is calculated for each location. To determine the margin to AR, the mitigation of the flow disrupter is not credited and the calculated Strouhal number is compared to the critical Strouhal numbers based on geometry and flow parameters that could lead to the onset of AR. This analysis is applied to the CNTS main steam line branch connections, as described below, for full-power normal operating conditions when the pyright 2022 by NuScale Power, LLC 232
operating conditions produce the Strouhal numbers closest to the critical Strouhal number. At lower reactor power levels, velocities are reduced. This results in Strouhal numbers that are higher and therefore further from the critical limit. The main objective of the test is to verify that AR is not active or causing detrimental vibration in the CNTS main steam lines during tests that represent the full range of operating conditions. This includes testing for the detection of any acoustic excitation by a higher order shear layer mode during partial power operation. Unlike monitoring mechanical vibration of a particular component, ARs are also detectable by monitoring the magnitude and frequencies of dynamic pressure pulsations in the fluid, which help identify the presence and excitation of a standing wave in a flow occluded region. Such measurements serve as additional evidence beyond the vibration data that are collected on the exterior of the pressure boundary. 1.1 Vibration Testing Guidelines The ASME Standard for Operation and Maintenance (OM) of Nuclear Power Plants (Reference 9.1.1) Part 3: Vibration Testing of Piping Systems, provides test methods and acceptance criteria for assessing the severity of piping vibration. Steady-state and transient vibration testing are addressed along with applicable instrumentation and measurement techniques, recommendations for corrective action, and discussions of potential vibration sources. The test specification developed for the AR initial startup testing shall comply with the ASME Operations and Maintenance code requirements as follows: Reference 9.1.1 under General Requirements, stipulates that a test specification be prepared to ensure that the objectives of the tests are satisfied and that results obtained are accurate or conservative. The test specification shall include the minimum list of items in Reference 9.1.1 Section 3. Namely, (a) test objectives, (b) systems to be tested (including boundaries), (c) pretest requirements or conditions, (d) governing documents and drawings, (e) precautions, (f) quality control and assurance, (g) acceptance criteria, (h) test conditions and hold points, (i) measurements to be made and acceptable limits (including visual observations), (j) instrumentation to be used, (k) data handling and storage, and (l) system restoration. Classification of the piping is in accordance with the requirements and guidance of Reference 9.1.1, Section 3.1.1 (steady-state vibration) because transient operations such as pump actuation or rapid valve motion are not relevant for acoustic resonance. Due to operating experience and design analysis, the NuScale CNTS main steam line branch connections are classified as Vibration Monitoring Group 2. pyright 2022 by NuScale Power, LLC 233
steam line branch connections as described in Reference 9.1.1, Section 5.1.1.4. The determination of an allowable deflection limit is to be developed based on the methodology defined in Reference 9.1.1, Section 5.1.1.5 and NuScale design inputs. The measurement technique and deflection limits are specific to the validation of the NuScale CNTS main steam line branch connections and AR phenomena. Instrumentation and the Vibration Monitoring System specifications comply with the requirements discussed in Section 7 of Reference 9.1.1. 1.2 Instrumentation and Data Acquisition Requirements A general description of the minimum measurements and sensors proposed to perform the testing is provided in Table 5-20. Sensor specifications, signal conditioning equipment, data conversion and storage procedures, and calibration procedures are to be accepted by NuScale. Table 5-20 Measurements and Sensors Minimum Number easurement Description of Sensors The accelerometers are mounted on the two MS piping lines, two DHRS branch piping legs (ten sensors total), and two MS drain ration amplitude 18 valve branches, and two MSIVs in order to record their vibration response. The high frequency(b) dynamic pressure sensors probe the two branch lines to the DHRS actuation valves, and two MS drain namic pressure 6 valves, and MS lines. Strain gauges may be substituted if dynamic pressure sensors cannot be accommodated(c). These sensors are used to determine the superheated steam Temperature 2 temperature inside the pipes. The main steam system (MSS) has these as part of its control and monitoring function. pyright 2022 by NuScale Power, LLC 234
Minimum Number easurement Description of Sensors The mass flow rate sensor in the MSS is used, along with pressure and temperature sensors in the MS line, to calculate the free stream flow velocity in the MS pipes at the DHRS branch Flow rate 2(a) lines. The uncertainties involved in this calculation are considered in the pre-test prediction. A velocity reading at the DHRS branch line location is not required and a free stream velocity is sufficient. es: Quantity indicates a final datum of average velocity in each MS line. The velocity is calculated from the mass flow rate measurement in the downstream MSS line The frequency should be greater than the reciprocal of transit time for a fluid particle across the branch avity, and the transit time of a sound wave traversing the length of the cavity and back to the MS pipe. f strain gauges are used, they would be installed in the DHRS branch and MS drain valve branch to ransform strain oscillation readings into pressure amplitudes. Two sets of four symmetrically ircumferential strain gauges placed at two axial locations in each branch pipe cancel out the shell modes of vibration such that the frequency range is well below the breathing shell mode of the pipe. Any deformation of the pipe as measured by strain gauges is caused by an acoustic pressure wave nside the pipe. The axial distance between the two measuring locations is less than half the wavelength of the upper limit frequency. Acceleration, velocity, and displacement are measured with the use of accelerometers. The accelerometer also provides the frequency signature of a vibration such that the vibration response can be correlated to a vibration source, (i.e., from turbulence or from AR). Velocity and displacement readings are obtained through single and double integration, respectively. The advantage of accelerometers is they measure absolute acceleration and do not need to be referenced to a structure position. Each of the main steam isolation valves (MSIVs) are to be monitored with accelerometers mounted to the valve body to ensure that vibration amplitudes are acceptable. If the instabilities in the stub piping resonate with the structural natural frequencies of the MS piping and MSIVs, dynamic loads could result in high cycle fatigue. Thus, it is important to also instrument nearby components such as the MSIVs. The MS lines are not fully symmetric and a slight disparity in steam velocity between each line results from minor differences in flow resistance through the FW and MS piping. Although unlikely, with one MS line at a slightly higher velocity, resonance could occur in one MS line and not the other. Also, if resonance in a cavity were to propagate into the MS piping, there may be small differences in vibration response due to geometry, support locations, and damping, thus accelerometers are placed on both MS lines for this testing. The branch piping exterior walls are also instrumented with accelerometers to ensure deflection is below the limits for the piping, as calculated using ASME OM Part 3 Section 5.1.1.5 (Reference 9.1.1). The amplitude measured on the branch pipe is expected to be larger than the process piping. As discussed in Section H-3.1.3 of Reference 9.1.1, measurement of true peak-to-peak displacement is pyright 2022 by NuScale Power, LLC 235
pipe mode shape and vibrational stress. RMS measurements cannot be readily converted to peak-to-peak measurements except for pure sinusoidal signals, so RMS displacements can only indicate averaged stress. The count of two MSIVs, two MS line piping legs, four DHRS actuation valves, two MS drain valve and two DHRS branch piping legs requires 18 sensors, which is the minimum for accelerometers if tri-axial accelerometers are used. If biaxial accelerometers are used, then two accelerometers offset by 90 degrees are needed at each location. Consideration of additional sensors for redundancy in case some fail is assessed in the test plan based on review of vendor data, instrument specifications, and ease of installation. Pressure data are best obtained through the use of dynamic pressure transducers directly in contact with the fluid, which requires tapping into the piping. This is feasible in the DHRS branches as indicated by the small-bore taps. These particular penetrations are used for sensors related to the nuclear steam supply system control, so additional ports are needed for the first prototype to accommodate using dynamic pressure transducers for this test. Each of the two DHRS closed-end legs and two MS drain valve branches has dynamic pressure measurement in order to detect the presence of a standing wave in the branch piping and allow for the investigation of resonance. Because the acoustic transmission of pressure waves propagates in the MS lines, the remaining pressure sensors are installed in the flow path of the MS piping in order to compare to the branch measurements, and determine the strength of the reflected pulsation if it is active. Alternatively, two sets of four strain gauges may be placed at two axial locations along a run of straight pipe to measure hoop strain in the pipe. This non-intrusive technique may be used to measure dynamic changes in pressure inside the pipe. At a given axial location, the four strain gauges need to be placed around the circumference of the pipe every 90 degrees. The axial distance between the measuring locations is designed to avoid half-wavelengths of the acoustic pressure waves. The half-wavelength would be calculated as L = c/2f. In order to detect an excitation of a higher order acoustic mode by a higher order shear layer mode, the spacing should be less than the half wavelength based on the first and second mode acoustic frequencies. A third or higher acoustic mode is not expected to be excited because this requires higher flow velocities and the velocity is limited at full power operating flow rates during the initial startup testing. Strain gauges are sensitive to changes in hoop stress due to local pressure changes. A final method is selected in the test specification based on the accuracy and reliability of measurement (from field experience) with consideration of installation options. Thermocouples are needed to measure the fluid temperature exiting the steam generators and flowing through the MS lines. Measurement uncertainties are quantified and provided. The plant has permanent instrumentation in the MS line pyright 2022 by NuScale Power, LLC 236
dual transmitters with dual temperature transmitters to support the flow measurement. These temperature sensors are also used to allow for an assessment of the heat losses through the MS line, and the impact to the calculated velocity at the CNTS main steam line branch connections. The test plan specifies monitoring steam temperature at these locations and save data during the test interval. The pre-test prediction evaluates the uncertainty in the MS flow measurement, but the need to perform velocity measurement for this test is optional because the plant has permanent instrumentation to measure the mass flow rate in the downstream MSS piping outside of the NPM. The free stream velocity in the vicinity of the DHRS steam line tees can be back-calculated from the downstream steam flow rate measurement. Although the velocity profile at the DHRS steam line entrance may exhibit turbulence and swirl from the upstream pipe bends and tee junction, this does not need to be measured explicitly because the AR analysis method is based on the free-stream velocity upstream of the cavity, and not a local velocity at the cavity entrance. Design of the DAS is documented in test facility design documents submitted to NuScale for acceptance. The highest fundamental acoustic frequency of the pipe cavities is less than (( }}2(a),(c),ECI. A time signal with a sampling rate at least five times the highest frequency of interest is sufficient to accurately record the expected pressure pulsations. Therefore, a DAS with a sampling rate of 3000Hz or higher is used. 1.3 Description of Required Tests This test includes gathering vibration, flow, and acoustic measurements at various power levels during the initial startup testing. The test should gradually increase the FW pump flow rate such that the CNTS main steam line branch connections are exposed to a range of partial-power steam flow rates to detect any acoustic excitation by a higher order shear layer mode. When the flow velocity is ramped up from a low value a given resonance mode can be excited by a higher-order shear mode before it is excited by the first order shear mode. The shear-layer excitation is the strongest at the first shear layer mode, where the most severe pressure pulsations are developed. The test procedure specifies a range for the partial power flow rates and a time to hold at each flow rate for sufficient data collection. The partial power testing exercises the test procedures and ensures that vibration levels are acceptable before increasing to the full 100 percent power conditions. The test procedure also specifies a time to hold at full power for sufficient vibration and flow/acoustic measurement data collection. The plant performance parameters in terms of steam pressure, steam temperature, and flow rate are documented in the final test plan. The test procedure specifies a range for those values in order for data collection to be performed. pyright 2022 by NuScale Power, LLC 237
branch lines, but the reactor power level must remain at or below the maximum licensed power level during initial startup testing. The accelerometers measure the acceleration on the valves, cavities, and piping while the plant is operating at each tested condition, which can be converted to a RMS velocity, or a peak displacement. Typically, vibration levels above a certain RMS velocity or displacement limit warrant corrective actions to reduce the vibrations. The test plan needs to establish these thresholds based on ASME OM Part 3 guidelines, and reviewing the vendor design limits for the valves. Stress limits on the piping/welds due to fatigue may also inform the point at which to stop testing and take corrective actions. Vibration responses are manifested differently depending in their source, i.e., responses due to turbulence are typically random, low level spectral amplitudes along the same order of magnitude, whereas an AR source vibration is distinguished as large amplitude responses at distinct frequencies. Vibration signals usually consist of very many frequencies occurring simultaneously, such that a resonant frequency cannot immediately be seen by looking at the time history response. Therefore, a frequency analysis must be performed to break down the vibration time signals into individual frequency components such that a spectrogram is generated and analyzed to verify the presence or absence of AR. Online and offline frequency analysis are performed to determine the amplitude and frequency content of the vibration signals. The dynamic pressure transducers are used for continuous monitoring of the pressure fluctuations in the DHRS branch lines, MS drain valves branch lines, and the MS lines. These measurements can be used to determine if resonance exists at acoustic source frequencies, and to understand the characteristics of the excitations if the piping or valve vibration amplitudes are excessive. Having pressure sensors in each MS line and each DHRS branch line helps determine if local velocity differences have an effect on the pressure fluctuations. The pressure measurements, along with temperature data, determine the uncertainty and bias in the predicted speed of sound and velocity. The as-built dimensions of the CNTS and DHRS piping are key inputs to the analysis. The test procedure shall specify that as-built measurements of these variables are recorded to ensure that uncertainty in design inputs are adequately accounted for in the analysis. Measurements need to be obtained for the inside diameter of MS lines at the flow sensor location and at the CNTS main steam line branch connections inside diameters. These measurements can be performed in the factory upon receipt inspection of the fitting or when it is welded as a piping assembly. pyright 2022 by NuScale Power, LLC 238
The testing to be performed follows at a minimum the sequence of activities specified in this section. Additional steps are included by the test supplier as needed to effectively and safely conduct the test. For each test period, the vibration monitoring from the accelerometers and dynamic pressure transducers are recorded simultaneously to allow for data interpretation and decision making about test continuation. In the event that an unacceptable vibration response develops any time during initial startup testing, the test conditions are adjusted to stop the vibration and the reason for the vibration anomaly investigated before continuing with the planned testing. Online and offline spectral analysis and time-history analysis are performed to determine the amplitude and frequency content of the vibration signals. Spectral analysis is used to detect large amplitude responses at distinct frequencies, which is indication of the acoustic frequency modes of the CNTS main steam line branch connections resonating with the MS line VS frequencies. The resulting vibration of the piping systems and MS drain valve, or MSIV valve bodies could cause dynamic loads and fatigue on the locations where maximum stress is expected. The measurements are compared to the established acceptance criteria for each location. The acceptance criteria are based on the allowable vibratory stress limits for the instrumented components. Depending on the practical limitations of sensors placement, cable routing and DAS channels, the in-situ flow measurements (pressure, temperature, and flow rate) do not have to be performed at the same time as the accelerometer measurements, although the impact to overall test duration of sequential measurements should be considered in the test plan. The dynamic pressure transducers or accelerometers can independently indicate if there is an AR condition during operation. The duration for data collection needs to be determined in the test specification and should be sufficiently long enough to allow for stabilized statistical averages in the data. If vibration levels are below the acceptance limit and the test is not stopped due to detection of resonance, a reasonable period for data collection is (( }}2(a),(c),ECI hours for the full power test in order to achieve 1 million cycles of the CNTS main steam line branch connections natural frequencies, which is less than (( }}2(a),(c),ECI for the DHRS piping tees which have the lowest margin in the CNTS main steam line branch connections. For the second order shear layer test, it is not required to achieve 1 million cycles at each flow rate hold point, but the total test duration depends on the heatup and power ascension rate limits. A final test report is developed to summarize the testing activities and results. The content includes the following, at a minimum. description of the testing infrastructure, such as pyright 2022 by NuScale Power, LLC 239
instrumentation mounting
- instrumentation diagrams
- details of the DAS and instrumentation
- dates of the testing
- identification of tester or data recorder discussion of testing methodology actions taken as a result of any deviations types of observations collected, for example
- vibration amplitude
- dynamic pressure or strain
- temperature
- mass flow rate/calculated velocity test results and results evaluations
- data files
- data post processing (extent to be determined in the test specification)
- identification of personnel evaluating test results
- documentation of critical instrument channel total uncertainty in an official TEEAR
- instrument calibration certificates
- test readiness inspection report
- completed and signed test procedures pyright 2022 by NuScale Power, LLC 240
This section recommends startup instrumentation plans for monitoring flow-induced vibrations in the NuScale Power Module (NPM)-20 design during initial startup testing. The CVAP consists of three phases: analysis, measurement and inspection. Validation in the measurement program is required for all components with an analysis result of less than 100 percent safety margin for a flow-induced vibration (FIV) mechanism, or if a novel design feature is implemented to preclude the onset of FIV. Based on the high safety margins for most components and mechanisms analyzed, this results in a very limited scope of required validation testing compared to traditional pressurized water reactor designs, which often have components that experience high alternating stresses due to turbulent flow and thus necessitate a significant CVAP measurement program validation scope. The results of the analysis program show that the only component inside of the containment vessel that requires validation in the measurement program is the steam generator (SG). Validation testing is pursued in a separate effects test facility to provide sufficient quantity and quality test data to validate the design analysis results. This testing is completed prior to core loading to confirm the adequacy of the steam generator design for FIV. Although startup instrumentation is not required for specific validation testing needs, it is desirable to provide some instrumentation in the prototype NPM for assurance that there are not destructive levels of component vibration once the plant is operating. This report recommends instrumentation plans for CVAP initial startup test sensors. It provides a methodology and assessment of using dynamic pressure sensors to detect destructive levels of flow-induced vibration in the NPM, including flow-induced vibration from known and unknown sources that may be activated during testing. Sensor models, locations, and mountings are proposed. Use of dynamic pressure sensors for the CVAP initial startup testing program is recommended and assessed in this report because these sensors are expected to better meet the intent of detecting unknown and unanalyzed vibration sources compared to the use of locally-mounted accelerometers, strain gages or optical sensors. Use of dynamic pressure sensors also minimizes the risk of introducing additional components in the flow stream that could contribute to FIV or become a source of loose parts. Dynamic pressure sensors are commonly used in other CVAP measurement programs, as well as scale model testing. Component Vibration Noise Estimation Methodology A proof-of-concept calculation is performed to estimate the detection capability of vibrations in the NPM using dynamic pressure sensors. The pressure signals produced by FIV mechanisms are estimated by calculation and compared to conservative estimates of the background noise levels to ensure that the estimated pressure signals can be sensed by a dynamic pressure sensor. Section 6.5 discusses how testing data is used to confirm this assessment, and to develop acceptance criteria for initial startup testing. pyright 2022 by NuScale Power, LLC 241
simplifications. One simplification is that tube noise is generated by a single vibrating tube. The methodology does not consider the effect of incoming acoustic waves, nor reflection of generated waves. It would be unusual for some locations in the NPM, such as in the SG tube bundle, to only have one tube span vibrate. Assessment of multiple acoustic waves and wave reflection are outside of the scope of this proof-of-concept evaluation. Sound radiation equations are two-dimensional. In order to apply them to components in the NPM, the vibrating component is treated as infinitely long and vibrating with uniform amplitude. In practice, a component that is much longer than the distance to the RPV wall (L >> r) satisfies this assumption. This is the case for tube column 21 but is satisfied less well for tubes in the inner columns or for detecting an unknown vibration source. A more detailed calculation could use mode shapes to compute three-dimensional sound radiation patterns, and integrate sound from all sources at the sensor location. However, such a procedure is not necessary for this proof-of-concept evaluation. 0.5 For the SG tube assessments, the bundle speed of sound c bundle = c ( 1 + ) is used, where is the fraction of the bundle volume taken up by the tubes. This approximates the isotropic scattering of sound within a bundle. No attempt is made to quantify the local effect of sound diffraction by nearby components, or sound attenuation due to turbulent flow. This is judged to be reasonable for this feasibility study. 1 Tube Noise Estimation Methodology The pressure generated by a vibrating tube is computed based on the dipole sound equations for a two-dimensional vibrating cylinder using the coordinate system shown in Figure 6-1. The cylinder is assumed to be executing one-dimensional harmonic motion in the x-direction, with velocity as: (( Equation 6-1
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 242
((
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 243
(ka << 1), which allows a simplified solution for the sound intensity in the form given as: ((
}}2(a),(c),ECI The sound intensity is interpreted in terms of the root mean square (rms) pressure given as:
((
}}2(a),(c),ECI The sound pressure level (SPL) is given as:
((
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 244
Figure 6-1 Cylinder Vibration Diagram
}}2(a),(c),ECI Equation 6-4 through Equation 6-6 are used to compute the noise radiated from tubes vibrating in the radial and axial directions, as shown in Figure 6-2 and Figure 6-3, respectively. In each case, the tube radiates in the two-lobed intensity pattern characteristic of dipole sound and generated by the cos2 term in Equation 6-4.
Figure 6-2 indicates that the optimal location for a dynamic pressure sensor to detect noise due to vibration of a single tube is at the same axial position as the tube. Since the SG tubes are helical, this location would be the axial and azimuthal position of greatest tube motion for the first bending modes of the long helical spans (i.e., the center of the long spans). In the case of axial vibration of a tube, the location of maximum signal at the RPV wall is at an axial distance equal to (( }}2(a),(c),ECI the distance of the tube center from the RPV wall, either above or below the tube, as shown in Figure 6-3. Both figures include a general location on the RPV wall given by Laxial, R, and . This position is found through calculation as the limiting location where a pressure sensor can be expected to distinguish noise generated by a vibrating tube from the background noise due to turbulence. pyright 2022 by NuScale Power, LLC 245
Figure 6-2 Sound Radiation Pattern for Radially-Vibration Tube
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 246
Figure 6-3 Sound Radiation Pattern for Axially-Vibrating Tube
}}2(a),(c),ECI Equation 6-4 provides the total sound intensity generated by a single tube and radiated to a location given by coordinates r, . In order to compare the noise signature of the tube with that of the background turbulence, a noise spectrum is required. This spectrum is obtained by spreading the tube noise into the spectrum of a damped one-dimensional oscillator following, as shown in Figure 6-4. This figure shows the amplitude multiplier A1dof as a function of the frequency ratio fr = f/fn, where pyright 2022 by NuScale Power, LLC 247
fn is the natural frequency of the tube. The power spectral density (PSD) for the tube noise radiation is then given by: ((
}}2(a),(c),ECI Figure 6-4 Estimated Spectrum of Noise Radiated by a Vibrating Tube
}}2(a),(c),ECI 2 Background Turbulent Noise Fields The background turbulent noise field depends on whether the sensor is placed in the SG tube bundle region or in a less obstructed region.
Within the tube bundle, the turbulent spectrum is taken as the suggested envelope given by Equation 6-8. This PSD provides a bounding representation of the normalized random pressure PSD for single phase cross flow in a tube bundle. For a pyright 2022 by NuScale Power, LLC 248
that the use of the cross flow PSD to represent the background noise that would be seen at the sensor is conservative. Depending on the location of the sensor relative to the tube bundle, it would likely measure a response less than that of the cross flow spectrum but greater than that of the annular flow spectrum. Note that the velocity scaling used in this equation Vp, the pitch velocity, is equal to the velocity in the gap between the tubes. G p ( f ) = 0.01 F < 0.1 G p ( f ) = 0.2 0.1 F 0.4 Equation 6-8
-4 5.3 x 10 G p ( f ) = ------------------------- F > 0.4 3.5 F
Df F = ------ Equation 6-9 vp Where: G p ( f ) = PSD of the turbulent pressure for cross flow in a tube bundle (psi2/Hz) f = Modal frequency (Hz) F = Reduced Frequency (-), per Equation 6-9 D = SG Tube outer diameter (in) v p = SG pitch velocity (in/s) In regions outside of the tube bundle, the turbulence spectrum is assumed to be the envelope shown in Equation 6-10. This PSD represents a boundary layer turbulent energy spectra for straight flow channels. This spectrum is used to evaluate turbulent noise in the upper riser and the downcomer. In this correlation, the velocity vf is the free steam velocity in the flow channel and the length scaling Dh is the hydraulic diameter. pyright 2022 by NuScale Power, LLC 249
-5 2 3 0.272 x 10 v D h f
G p ( f ) = ------------------------------------------------
- S<5 0.25 S
Equation 6-10
-5 2 3 22.75 x 10 v D h f
G p ( f ) = ------------------------------------------------
- S>5 3
S 2 Dh f S = ---------------- Equation 6-11 vf Where: G p ( f ) = PSD for general annular flow (psi2/Hz)
= Fluid density (lbf-s2/in4) v f = Free stream velocity (in/s)
D h = Hydraulic Diameter (in) v f = Reduced Frequency factor (-), per Equation 6-11 Noise Estimation Results Estimates for SG tube noise generated by TB, VS, FEI, and unknown reactor vessel internals (RVI) sources are presented in Section 6.2.1 through Section 6.2.4. For each vibration mechanism and component evaluated, the tube noise received at the RPV or riser shell is compared with an estimate of the background turbulent noise, which is approximated as the turbulent forcing function per Section 6.1.2. A detection goal for this ratio is set and the maximum axial distance at which this goal is met is computed (this distance is denoted as Laxial in Figure 6-2 and Figure 6-3). The comparison is done by first computing the frequency range over which half of the radiate power exists. Referring to Figure 6-4 it can be seen that for a lightly damped structure the half-power range is quite narrow, (( }}2(a),(c),ECI of the natural frequency in the case where the structural damping ratio is (( }}2(a),(c),ECI. The power of the turbulent noise over pyright 2022 by NuScale Power, LLC 250
of rms pressures this criterion is ((
}}2(a),(c),ECI The goal is set equal to 10 to ensure that the radiated noise is not obscured by the background turbulent noise. The distance from the equilibrium position of the vibrating component to the furthest wall location that meets this goal is the maximum value of R in Figure 6-2 for radial vibration and Figure 6-3 for axial vibration.
Using Equation 6-4 it can be shown that for any location on the wall, the sound intensity due to radial vibration of a tube is given by ((
}}2(a),(c),ECI The geometry shown in Figure 6-2 leads to (( }}2(a),(c),ECI. Therefore, for radial vibration
((
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 251
Equation 6-12 then gives ((
}}2(a),(c),ECI Combining Equation 6-14, Equation 6-15, and Equation 6-16 and solving for Rgoal gives
((
}}2(a),(c),ECI For axial vibration the geometry in Figure 6-3 is such that
((
}}2(a),(c),ECI and
((
}}2(a),(c),ECI Equation 6-4 then becomes
((
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 252
Where (( ) }}2(a),(c),ECI is enforced at the wall. Maximizing Equation 6-20 gives (( }}2(a),(c),ECI, where R max denotes the value of R at the location labeled Optimal Location for Sensor in Figure 6-3 for axial vibration. Combining Equation 6-20 with Equation 6-15 and Equation 6-16 gives ((
}}2(a),(c),ECI This is a cubic equation with three real roots. The roots of interest are solved for i=0 and i=1, which correspond to the axial range over which the power ratio goal is satisfied. The maximum power ratio is located at an axial location between these two roots.
((
}}2(a),(c),ECI The specific values of Laxial in Figure 6-3 at which the goal are satisfied is labeled Lgoal and is found as
((
}}2(a),(c),ECI As noted previously, the sound pressure level (SPL) computed for each FIV mechanism assumes that only one component is vibrating. If multiple components, e.g., SG tubes, pyright 2022 by NuScale Power, LLC 253
logarithmically to obtain the total SPLTotal as ((
}}2(a),(c),ECI This equation is derived from the fact that sources combine by adding their squares.
Equation 6-6 is then used to turn this into a relationship between sound pressure levels, as given by Equation 6-24. If all n sources have the same SPL at the receiver then ((
}}2(a),(c),ECI The 10log10(n) term gives the increase in sound pressure level over that produced by one of the n sources. This term is computed in Table 6-1.
Table 6-1 Increase in SPL due to n Identical Sources n 10log10(n) (dB)
}}2(a),(c),ECI 1 Sound Radiation due to Turbulent Buffeting of a Steam Generator Tube The noise generated by tube motion caused by turbulent buffeting loads is assessed for a tube motion of y/D = 0.02, where y is the maximum tube displacement and D is the tube OD. This limit is motivated by the Tubular Exchanger Manufacturers Association (TEMA) standard (Reference 9.1.15) which sets y/D = 0.02 as the maximum tube displacement that should be allowed under turbulent buffeting loads.
The estimated turbulent buffeting vibration amplitude from Reference 9.1.4 is also pyright 2022 by NuScale Power, LLC 254
calculations for turbulent buffeting of the SG tubes. Once the displacement is known, the velocity v0 in Equation 6-4 is then computed as ((
}}2(a),(c),ECI Turbulent buffeting can excite tube motion in any direction. Therefore, both the radial and axial cases shown in Figure 6-2 and Figure 6-3 are evaluated.
The results of turbulent buffeting-induced radial tube vibration are presented in Table 6-2 and Figure 6-5. The SPLs recorded in Table 6-2 and the spectra shown in Figure 6-5 are taken at the RPV wall immediately opposite the center of the span of a single, radially-vibrating tube. Both the spectral peaks shown in Figure 6-5 and the overall SPLs in Table 6-2 rank as Column 21, Column 11, Column 1. This ranking is due to the (( }}2(a),(c),ECI term in the sound intensity equation (Equation 6-4), which implies a (( }}2(a),(c),ECI dependence on tube natural frequency and distance to the sensor. Tube column 11 with the highest natural frequency provides the strongest signal. However, as can be seen in Figure 6-5, the background turbulent noise is highest at the natural frequency of tube column 1, resulting in a low power ratio of ((
}}2(a),(c),ECI This phenomenon can be understood qualitatively by noting how distinct the tube column 11 signal is from the background turbulence in Figure 6-5 as compared with how indistinct the signals from tube columns 1 and 21 are from the background turbulence. If neighboring tubes vibrate simultaneously at the same frequency and in-phase, the noise level may be high enough to meet the 10X sensing goal for all tube columns.
pyright 2022 by NuScale Power, LLC 255
Table 6-2 Noise at RPV Wall Due to Radial SG Tube Motion Caused by Turbulent Buffeting SPL at Optimal Maximum Axial Tube Maximum Power Distance for Sensing Sensor Location Column Ratio (Laxial- inches) (dB) ((
}}2(a),(c),ECI Figure 6-5 Noise at RPV Wall Due to Radial SG Tube Motion Caused by Turbulent Buffeting
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 256
Table 6-3 and Figure 6-6. The SPLs recorded in Table 6-3 and the spectra shown in Figure 6-6 are taken at the RPV wall in the location marked Optimal location for sensor in Figure 6-3. Because of the reduced signal reaching the wall when the tube is vibrating axially, ((
}}2(a),(c),ECI ble 6-3 Noise at RPV Wall Due to Axial SG Tube Motion Caused by Turbulent Buffeting Minimum Axial Optimal Maximum Axial SPL at Optimal Maximum Distance for Distance for Distance for ube Sensor Location Power Sensing Sensing Sensing olumn (dB) Ratio (Laxial- inches) (Laxial- inches) (Laxial- inches)
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 257
ure 6-6 Noise at RPV Wall Due to Axial SG Tube Motion Caused by Turbulent Buffeting
}}2(a),(c),ECI The case of radial tube motion due to turbulent buffeting at the design analysis predicted vibration amplitude is evaluated. For this case, the vibration amplitude is approximately an order of magnitude less than at the TEMA value of 2 percent of the tube outer diameter. This lower amplitude results in a lower detection capability,
(( }}2(a),(c),ECI, as shown in Figure 6-7. In this figure, ((
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 258
Figure 6-7 Noise at RPV Wall Due to Radial SG Tube Motion Caused by Turbulent Buffeting at Design Analysis Predicted Amplitudes
}}2(a),(c),ECI 2 Sound Radiation Due to Vortex Shedding of a SG Tube In vortex shedding, alternating pairs of vortices are shedding due to flow over a cylinder or other bluff body. The combination of fluid forces on the cylinder and the motion of the cylinder itself radiates noise in a double dipole pattern. The motion in the drag direction is an order of magnitude smaller than that in the lift direction and is generally ignored. The sound radiated in the lift direction reaches the RPV wall radially, as shown in Figure 6-2.
The fact that there are two noise sources present in vortex shedding presents a detection challenge. The noise due to the vortex shedding forces, known as Aeolian tones, occurs whether or not the tube is in motion. Significant tube motion only occurs near lock-in, when the vortex-shedding frequency approaches the natural frequency of the tube n. This situation is shown in Reference 9.1.17, Figure 5.28, in which the relative contribution of the two sources is plotted. In this figure, it is only in pyright 2022 by NuScale Power, LLC 259
Equal Contributions that the noise due to tube motion can be sensed over the Aeolian tone caused by the vortex shedding forces. Blevins shows analytically that equal contribution occurs when y/D = 0.07 and states that experimentally this departure typically occurs in the range y/D = 0.05 to 0.10. Therefore, the y/D ratio must be checked first, before noise levels are computed. This check is accomplished as follows: (( Equation 6-27
}}2(a),(c),ECI As described in Section N-1324.2 of Reference 9.1.13, this equation provides a conservative estimate for the amplitude of periodic vortex-induced vibration assuming the VS is fully correlated along the cylinder span and the lift coefficient is one.
Furthermore, the maximum displacement along the span is used, so = 1. n The results are presented in Table 6-4. ((
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 260
Table 6-4 Tube Motion Due to Vortex Shedding Tube Tube motion at Sound due to Tube Motion Column lock-in (y/D) Exceeds Vortex Sound?
}}2(a),(c),ECI The second check is to determine whether noise due to tube motion can be sensed over the turbulent background noise. This check is identical to the calculation described in Section 6.1.1 except that the vibration amplitudes in Table 6-4 are used.
The results are presented in Table 6-5 and Figure 6-8. ((
}}2(a),(c),ECI
((
}}2(a),(c),ECI Table 6-5 Noise at RPV Wall Due to Tube Motion Caused by Vortex Shedding SPL at Optimal Maximum Axial Tube Maximum Power Distance for Sensing Sensor Location Column Ratio (Laxial- inches)
(dB) ((
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 261
Figure 6-8 Noise at RPV Wall Due to Tube Motion Caused by Vortex Shedding
}}2(a),(c),ECI 3 Sound Radiation Due to Fluid Elastic Instability of a Steam Generator Tube FEI is characterized by self-sustained motion of tubes drawing energy from the flow and executing closed-loop motion. Typically, a number of tubes in a tube bundle execute coordinated motion.
In the case of a helical tube steam generator, Chen (Reference 9.1.18) found that the tube motion is primarily in the flow direction. Therefore, in estimating the directionality of the sound radiated by the tube motion the axial arrangement shown in Figure 6-3 is used. The onset of FEI occurs suddenly as the flow rate over the tube bundle is increased. Figure N-1331-1 of Reference 9.1.13 shows typical test data. In this figure the y axis shows tube motion as y/D x 100. Therefore, the maximum tube motion shown in this figure is 6% of the tube diameter. Because of this sudden onset, FEI is normally avoided rather than quantified. Critical velocities are computed but tube motion is not. pyright 2022 by NuScale Power, LLC 262
}}2(a),(c),ECI Using this value yields results shown in Table 6-6 and Figure 6-9. ((
}}2(a),(c),ECI It is possible that an acoustic signal strong enough to be sensed above the background turbulence could occur just before the critical velocity is reached. In this circumstance there would be a strong possibility that the critical velocity would be reached as primary coolant flow rate is increased. If FEI occurs, this would be accompanied by large-scale tube motion and possible tube-to-tube contact, which could result in impact noise (Section 6.2.5).
Table 6-6 Noise at RPV Wall Due to Tube Motion Caused by FEI Minimum Axial Optimal Maximum Axial SPL at Optimal Maximum Distance for 10x Distance for 10x ube Distance for Sensor Location Power Sensing Sensing olumn Sensing (dB) Ratio (Laxial- inches) (Laxial- inches) (inches)
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 263
Figure 6-9 Noise at RPV Wall Due to Tube Motion Caused by FEI
}}2(a),(c),ECI 4 Sound Radiation Due to Unknown Reactor Vessel Internals Sources
((
}}2(a),(c),ECI
((
}}2(a),(c),ECI This signal is compared to the boundary layer background noise in the upper riser and the downcomer, to provide a sensitivity for measuring at these two locations (different fluid velocity, density, hydraulic diameter or distance to sensor). The distance over which the radiated sound intensity at the pyright 2022 by NuScale Power, LLC 264
computed. ((
}}2(a),(c),ECI The results of this calculation are shown in Table 6-7 and Table 6-8. ((
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 265
Figure 6-10 Noise at RPV Due to ICIGT Motion Caused by Turbulent Flow in Riser
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 266
Figure 6-11 Noise at RPV Due to Component Motion (Assuming ICIGT Frequencies) Caused by Turbulent Flow in Downcomer
}}2(a),(c),ECI Table 6-7 Noise at RPV at Riser Exit Due to RVI Vibration SPL at Optimal Maximum Axial Maximum Power Distance for Sensing Frequency (Hz) Sensor Location Ratio (Laxial- inches)
(dB) ((
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 267
Table 6-8 Noise in Downcomer Due to RVI Vibration SPL at Optimal Maximum Axial Maximum Power Distance for Sensing Frequency (Hz) Sensor Location Ratio (Laxial- inches) (dB) ((
}}2(a),(c),ECI 5 Sound Due to Component Impact Within the NPM, vibration of a component may occur without impact, in the case of existing contact at a support or prior to the onset of a vibration that generates sufficient force to displace the component at the support location. If gaps exist between the support and component, impact may occur at the onset and/or during the duration of the component vibrating condition. For extreme FIV mechanisms, such as FEI, the component response may be so strong such that impact occurs between tubes.
Part 11 of Reference 9.1.14 discusses the use of indirect measurement devices for the detection of impacting signals due to component vibration. Section 7.2.2.(b) of Reference 9.1.14 recommends band-pass filtering of 2 kHz to 7 kHz for structure-borne sound due to impacting, which indicates that impacting tends to occur in the kHz signal range. During initial startup testing, it is desirable to detect impacting sound that is transmitted in the primary coolant. Reference 9.1.14 discusses the use of a microphone for detection of tube vibration via transmission of the tube impact sound via the air column of the tube. Since water is a better conductor of sound than air, the principals discussed in Reference 9.1.14 are applicable to dynamic pressure sensor use in the primary coolant, with the exception that the designs proposed herein recommend fixed-location sensors. To detect possible impacts during startup testing, the dynamic pressure sensor should have a maximum frequency measurement range of at least 10kHz. As demonstrated in the previous sections, this frequency is well above the range at which the turbulent power spectra is elevated and detection of impact is expected to be achievable for the distance scales of interest in the NPM. In the event that impact occurs, the signals from multiple sensors can be compared to support identifying the impact location. pyright 2022 by NuScale Power, LLC 268
To support measuring noise generated due to FIV during NPM-20 initial startup testing, dynamic pressure sensors must meet the following requirements: ((
}}2(a),(c),ECI
(( }}2(a),(c),ECI This requirement drives the selection to a ((
}}2(a),(c),ECI. Three such instruments are presented in Table 6-9.
Table 6-9 Candidate Dynamic Pressure Sensors Pressure Max. Frequency Radiation anufacturer Model Max./Range Temperature Response Capable?
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 269
}}2(a),(c),ECI This information, in combination with sensors information provided in other CVAPs, demonstrates that there are no expected challenges in selecting and qualifying dynamic pressure sensors for use during initial startup testing.
Proposed Sensor Placement and Mounting Based on the sensor selection goal in Section 6.3 and the assessments in Section 6.2, (( }}2(a),(c),ECI dynamic pressure sensors are proposed to provide detection capabilities throughout the NPM. Three different sensor arrangement plans are also proposed. These arrangement plans provide similar detection capabilities, but have tradeoffs in terms of design and manufacturing. Multiple plans are proposed to provide the customer a range of options to meet initial startup testing vibration and impact detection goals. The three arrangement plans are shown in Table 6-10 and Figure 6-12, Figure 6-13 and Figure 6-14. ((
}}2(a),(c),ECI At each location, ((
}}2(a),(c),ECI A high coherence between sensors for a spectral peak implies that the dynamic pressure contains a large component of propagating acoustic pressure fluctuations. A low coherence may imply the dynamic pressure measurement is dominated by local hydrodynamic fluctuations or a non-propagating acoustic fluctuation. Sensors are specified to be flush-mounted to prevent significant increases in the measured dynamic pressure due to local flow disturbances.
((
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 270
}}2(a),(c),ECI All portions of the sensors and cables exposed to primary flow are designed and analyzed in accordance with ASME Section III. Cables for all sensors can be routed through and out of the containment vessel (CNV) at the CNV electrical penetration assembly.
Table 6-10 Proposed Dynamic Pressure Sensor Locations Installation Number of Sensors at Sensor Location Basis for Location rangement Plan Location
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 271
Installation Number of Sensors at Sensor Location Basis for Location rangement Plan Location
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 272
Figure 6-12 Proposed Location of Dynamic Pressure Sensors - Option 1
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 273
Figure 6-13 Proposed Location of Dynamic Pressure Sensors - Option 2
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 274
Figure 6-14 Proposed Location of Dynamic Pressure Sensors - Option 3
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 275
sealing. It is desirable that the tip of the sensor be nearly flush with the inside diameter of the mounting surface. A custom adaptor is designed for each installation elevation, to be compatible with the RPV or RVI surfaces, and provide adequate mounting and sealing properties. Benchmarking to Confirm Detection Capabilities Section 5.1.1 provides the instrumentation outline for the SG validation testing program for FIV, TF-3. TF-3 testing is used to demonstrate the adequacy of the vibration and impact detection method prior to initial startup testing. The TF-3 test specimen is extensively instrumented with accelerometers and strain gages (Section 5.1.1.2). Dynamic pressure sensors are provided at various elevations in the tube bundle. The diverse set of sensors as well as the ability to expose the SG to beyond-design-basis flow velocities allows for characterization of the turbulent power spectra and the vibration response of the SG tubes. If onset of a strongly-coupled FIV phenomena like VS or FEI occurs, the dynamic pressure signals can be compared to the locally-mounted strain gage and accelerometer readings to confirm the detection capability of the dynamic pressure sensors. The test specimen has ports to allow for in-water modal testing, where tubes are mechanically excited under no-flow conditions. This configuration test also provides the opportunity to assess the detection capabilities of the dynamic pressure sensor to a known excitation frequency and location with no turbulent flow noise. The TF-3 results help to inform the development of acceptance criteria for the dynamic pressure sensor output during the initial startup testing program. As the NPM is a natural circulation design, initial startup testing is the first opportunity to collect vibration data at full power primary side flow conditions. However, during non-critical testing the plant is heated and some primary coolant flow are developed using the module heatup system. Operation of the module heatup system is designed to provide sufficient primary side flow for heatup and mixing. Measurements taken during non-critical testing aid in characterization of noise transmitted to the NPM from the building and via the reactor pool, and provides the opportunity to diagnose and filter non-structural noise from the sensor outputs. Results and Conclusions Calculations have been performed to estimate the sound produced by single vibrating tubes experiencing FIV due to TB, VS, and FEI. The calculations show that noise generated by the motion of the SG tube columns is expected to be detectible over the background turbulence in many cases, although some signals are not ten times greater than the background noise, which served as the detection goal in this assessment. These assessments show that the acoustic signals generated from lower frequency vibrations are more challenging to measure. ((
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC 276
successive row of tubes, the flow in the upper riser and the downcomer is much less constricted and the background turbulent noise is significantly lower than in the SG region. ((
}}2(a),(c),ECI While the potential sources of FIV in these regions other than turbulence have been analyzed and show to be inactive in the design analysis program, it is proposed that the ability of acoustic sensors to detect unknown sources of vibration be provided by ((
}}2(a),(c),ECI This proof-of-concept assessment demonstrates the feasibility to monitor for component vibrations and impact in the NPM during the initial startup testing program using dynamic pressure sensors. Three equivalent installation arrangement plans are developed based on the assessment of acoustic signals generated from SG and RVI vibrations compared to the background noise levels. Considerations for installation and testing are provided.
Future design work will benchmark the sensor performance against locally mounted strain gages and accelerometers during the TF-3 validation testing. The TF-3 results help to inform the development of acceptance criteria for the dynamic pressure sensor output during the initial startup testing program. pyright 2022 by NuScale Power, LLC 277
This section identifies the prototype components that require inspection before and after the CVAP measurement program. because the measurement program focuses on the limiting components, inspection is used to confirm the assumptions regarding which components are limiting. For the components that are instrumented to support the measurement program, inspection provides a secondary confirmation of the FIV performance and integrity of these structures. Before and after initial startup testing, components are inspected for mechanical wear and signs of vibration induced damage. Initial startup testing provides a sufficient duration for the limiting NPM component to experience a minimum of one million cycles of vibration. Components that are evaluated in the analysis program undergo inspection. For the components validated in the measurement program via testing, the inspection provides a secondary confirmation of the FIV integrity of the NPM components. For components that do not require testing due to large safety margins, the inspection confirms that the testing performed on more limiting components sufficiently bounds the performance of the non-tested components. Inspection Methodology The components most susceptible to FIV are examined in limiting and representative locations to demonstrate acceptable performance. Inspection areas include:
- a. Major load-bearing elements of the reactor vessel internals that position the core support structure
- b. Lateral, vertical, and torsional restraints inside the RPV
- c. Locking and bolting components whose failure could impact reactor vessel internals integrity
- d. Contact surfaces and potential contact surfaces
- e. Critical locations identified by the analysis program
- f. RPV interior for loose parts or foreign material Components may be removed and inspected outside of the pressure vessel, but many NPM components cannot be removed from their installed locations. For those components, or when practical, an in-situ inspection is performed. Components like the SG tubes and SG tube supports that are too long to examine their entire surface and have inaccessible areas are inspected at least at the accessible ends of their length.
Initial startup testing provides for 1 million cycles of the most limiting (lowest fundamental frequency) component. This testing duration provides a reasonable number of cycles so that if rapid degradation is occurring due to FIV, there is evidence detectable by inspection. pyright 2022 by NuScale Power, LLC 278
The NPM components are inspected following the guidelines and requirements provided in ASME Section III (Reference 9.1.5), Paragraph NG-5111, Paragraph NB-5111 and using the methods defined in the ASME Section V (Reference 9.1.6), Article 9. The visual inspections are performed using VT-1 and VT-3, as defined by ASME SectionXI, Subarticle IWB-2500, Tables IWB-2500-1 B-N-1, B-N-2, and B-N-3 (Reference 9.1.7). These nondestructive surface examinations are used to inspect the surfaces and welds of the components identified for inspection. Visual examinations are performed on the NPM components to satisfy the following objectives: On critical surfaces to see if there are any cracks, defects, or abnormal distortion On welds to see evidence of cracks On interface surfaces to see evidence of wear, distress, or abnormal corrosion On fittings to see if they are tight At reasonable locations to see if loose parts or debris have collected The inspection results are documented in the CVAP Measurement and Inspection Program Results report. Any inspection findings and repairs/modifications are documented along with a complete record of the pre- and post- initial startup testing inspections including notes, photographs, and video. Pre- and Post-Initial Startup Testing Inspection The inspection and documentation for the NPM components are completed in two stages. The baseline inspection stage (pre-initial startup testing) takes place as the NPM components are assembled. The post-initial startup testing inspection stage takes place after the completion of the initial startup testing. During post-initial startup testing inspection, the core support structure and lower riser assembly are examined in the pool while other internals are examined in the dry dock. The post-initial startup testing inspection results are compared with the baseline inspection data. The comparison provides an independent method of corroborating the conclusions of the CVAP analysis program, that no severe FIV related degradation is occurring. The CVAP inspection locations are listed in Table 7-1. These locations include all inspection elements required to cover the six inspection areas listed in Section 7.1. The inspection examination methods defined in Table 7-11 are based on and consistent with the methods specified for in-service inspections of the NPM to meet ASME Code Section XI requirements. VT-1 inspections are specified for welded core support structures, attachments to the RPV, or identified areas of low margin. VT-3 inspections are specified for the majority of the remaining features except for locations that are mainly inspected for loose parts using a general visual exam. pyright 2022 by NuScale Power, LLC 279
Table 7-1 Pre- and Post-Initial Startup Testing Inspection Locations Inspection Req. ation Exam. Method Feature to be Inspected Category ID with Notes 4 (Section 7.1) 8 Core Supports and Flow Diverter Core Support Assembly Mounting Bracket to RPV Bottom VT-1 1 1 a, b Head Welds Core Support Assembly Mounting Bracket Exterior VT-3 1, 2 2 a, b, c Surfaces 3 Lower Core Plate Surfaces VT-3 a, b, c, d 4 Lower Core Plate to Core Barrel Weld VT-1 1 a, b 5 Reflector Blocks General Visual 5 f 6 Core Barrel Exterior Surface VT-3 6 a 7 Upper Support Block Weld and Fittings VT-1 1, 2 a, b, c, d 8 Core Barrel to Upper Core Plate Interface VT-3 a, b, d 9 Shared Fuel Pins VT-3 3 d 10 Flow Diverter General Visual f Lower Riser 11 Upper Core Plate to Lower Riser Section Weld VT-3 1 a, b 12 Upper Core Plate Surfaces VT-3 2, 5 a, b, c, d, f 13 Lower Riser Section Surfaces VT-3 5, 6 a, b 14 ICIGT Assembly to Upper Core Plate Welds VT-1 1 f 15 Fuel Pin VT-1 3 d 16 CRA Lower Flange Surfaces VT-3 b, c 17 CRAGT Interior and Exterior Surfaces VT-3 b 18 CRA Card to CRA Guide Tube Interface VT-3 3 d 19 CRD Shaft Alignment Cone to CRD Shaft Interface VT-3 3 d CRD Shaft Alignment Cone to CRAGT Support Plate VT-3 20 b, d Interface Upper Riser and Pressurizer Spray Nozzle 21 Upper Riser Bellows VT-3 b 22 Upper Riser Section Surfaces VT-3 6 b 23 RCS Injection Line Pipe VT-3 - 24 CRD Shaft Support Surfaces VT-3 b 25 CRD Shaft Supports to CRD Shaft Interface VT-3 3 d CRD Shaft Supports to ICIGT/ Riser Level Sensor GT VT-3 3 26 d Interface 27 Bellows threaded limit rod surfaces VT-3 3 b 28 CRD Shaft Sleeve Surfaces VT-3 2 e 29 Upper Riser Hanger Plate Surfaces VT-3 2 b, c 30 Hot Temperature Thermowell External Surfaces VT-3 - 31 Pressurizer Spray Nozzle Surfaces VT-3 1 - 32 Set Screw Assemblies VT-3 2, 7 b, c SG and Downcomer 33 SG Inlet Flow Restrictors VT-3 2, 3, 7 c, f 34 SG Inlet Flow Restrictor to SG Tube Interface VT-3 3, 7 d pyright 2022 by NuScale Power, LLC 280
Inspection Req. ation Exam. Method Feature to be Inspected Category ID with Notes 4 (Section 7.1) 8 35 SG Lower Support Surfaces VT-3 b, d 36 SG Tube to Tube Support Interface VT-3 3, 6 b, d, e 37 Tube Support to Upper Riser Section Interface VT-3 6 b, d 38 RCS Injection Line Pipe Surfaces VT-3 - 39 Cold Temperature Thermowells External Surfaces VT-3 - 40 RRV Nozzles VT-3 - Secondary Side Components 41 Steam Plenum VT-3 - 42 Steam Plenum Nozzle VT-3 - 43 SGS Piping VT-3 6 - 44 Steam Temperature Thermowells, External Surface VT-3 - 45 CNTS MS Tees VT-1 3 e 46 DHRS Condensate Piping Inside Containment VT-1 - es: isually examine welds erify that fittings are tight isually examine for evidence of vibration wear ll visual examinations should include checking for loose parts spection limited to accessible surfaces exposed while assembled spection limited to the accessible ends of the feature due to the large surface area spection limited to a sampling of components due to the large quantity ows with a - indicate that the feature to be inspected does not fall within one of the six categories entified in Section 7.1 pyright 2022 by NuScale Power, LLC 281
This report provides the details of the NuScale CVAP measurement and inspection program. This program consists of benchmark testing and analysis, validation analysis and testing, an instrumentation plan to detect large amplitude vibration during initial startup testing, and inspection of components screened as susceptible to FIV before and after initial startup testing. Following the completion of each test, post-test analysis is performed to complete the validation effort. Assessments are also performed based on the initial startup testing and inspection observations. Combined with the benchmarking efforts, the measurement and inspection work scope validate the FIV screening and design analyses in Reference 9.1.4. pyright 2022 by NuScale Power, LLC 282
Referenced Documents 9.1.1 American Society of Mechanical Engineers, Operation and Maintenance of Nuclear Power Plants, Division 2: OM Standards, ASME OM-2017, Part 3, Vibration Testing of Piping Systems, New York, NY. 9.1.2 American Society of Mechanical Engineers, Standard for Verification and Validation in Computational Fluid Dynamics and Heat Transfer, ASME V&V 20-2017, New York, NY. 9.1.3 Au-Yang, M.K., Flow-Induced Vibration of Power and Process Plant Components, APractical Workbook, ASME Press, New York, NY, 2001. 9.1.4 NuScale Power, LLC, NuScale Comprehensive Vibration Assessment Program Analysis Technical Report, TR-121353-P. 9.1.5 American Society of Mechanical Engineers, Boiler and Pressure Vessel Code, 2017 Edition, Section III, Rules for Construction of Nuclear Facility Components, New York, NY. 9.1.6 American Society of Mechanical Engineers, Boiler and Pressure Vessel Code, 2017 Edition, Section V, Nondestructive Examination, New York, NY. 9.1.7 American Society of Mechanical Engineers, Boiler and Pressure Vessel Code, 2017 Edition, Rules for Inservice Inspection of Nuclear Power Plant Components, New York, NY. 9.1.8 Ziada, S. and Shine, S., Shrouhal Number of Flow-Excited Acoustic Resonance of Closed Side Branches, Journal of Fluids and Structures, Volume 13 (1999), pg. 127-142. 9.1.9 Schardt, J.F., Flow-Induced Vibration Characteristics of BWR/6-238 Jet Pumps, GEAP-22201, UC-78, General Electric Company, September 1982. 9.1.10 Au-Yang, M.K. and Jordan, K.B., Dynamic Pressure Inside a PWR - A Study Based on Laboratory and Field Test Data. Nuclear Engineering and Design 58, pg 113-125, 1980. 9.1.11 Chen, Shoei-Sheng. Flow-Induced Vibration of Circular Cylindrical Structures. ANL-85-51, June 1985. 9.1.12 Giraudeau, M., et al. Two-Phase Flow-Induced Forces on Piping in Vertical Upward Flow: Excitation Mechanisms and Correlation Models. Journal of Pressure Vessel Technology, Vol. 135, p. 030907, 2013. pyright 2022 by NuScale Power, LLC 283
Dynamic Analysis Methods, 2017 Edition, New York, NY. 9.1.14 American Society of Mechanical Engineers, Operation and Maintenance of Nuclear Power Plants, Division 3: OM Guides, ASME OM-2017, Part 11, Vibration Testing and Assessment of Heat Exchangers, New York, NY. 9.1.15 TEMA, Standards of the Tubular Exchanger Manufacturers Association, 9th ed., 2007. 9.1.16 Blevins, R.D., Flow-Induced Vibration, Second Edition, Krieger Publishing Company, Malabar, FL, 2001. 9.1.17 Blake, W., K., Mechanics of Flow-Induced Sound and Vibration, Vol. 1, Academic Press, London, Second edition, 2017. 9.1.18 Chen, S. S., Jendrzejczk, J. A., and Wambsganss, M. W., "Tube Vibration in a Half-Scale Sector Model of a Helical Tube Steam Generator, J. Sound and Vib., 91(4), pp. 539-569, 1983. pyright 2022 by NuScale Power, LLC 284
Appendix A TF-3 Instrumentation Plan pyright 2022 by NuScale Power, LLC A-1
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0
}}2(a),(c),ECI
Appendix B TF-2 FEI Spectral Plots by Test Series (0 - 300 Hz) pyright 2022 by NuScale Power, LLC B-1
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure B-1 Test Condition A, Channel Set 1 - Side Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC B-2
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure B-2 Test Condition A, Channel Set 2 - Top Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC B-3
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure B-3 Test Condition B, Channel Set 1 - Side Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC B-4
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure B-4 Test Condition B, Channel Set 2 - Top Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC B-5
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure B-5 Test Condition C, Channel Set 1 - Side Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC B-6
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure B-6 Test Condition C, Channel Set 2 - Top Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC B-7
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure B-7 Test Condition D, Channel Set 1 - Side Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC B-8
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure B-8 Test Condition D, Channel Set 2 - Top Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC B-9
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure B-9 Test Condition G, Channel Set 1 - Side Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC B-10
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure B-10 Test Condition G, Channel Set 2 - Top Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC B-11
Appendix C TF-2 FEI Spectral Plots by Flow Rate (0 - 300 Hz) pyright 2022 by NuScale Power, LLC C-1
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure C-1 Nominal Primary-Side Flow Rate of 114 kg/s, Channel Set 1 - Side Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC C-2
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure C-2 Nominal Primary-Side Flow Rate of 114 kg/s, Channel Set 2 - Top Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC C-3
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure C-3 Nominal Primary-Side Flow Rate of 143 kg/s, Channel Set 1 - Side Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC C-4
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure C-4 Nominal Primary-Side Flow Rate of 143 kg/s, Channel Set 2 - Top Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC C-5
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure C-5 Nominal Primary-Side Flow Rate of 173 kg/s, Channel Set 1 - Side Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC C-6
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure C-6 Nominal Primary-Side Flow Rate of 173 kg/s, Channel Set 2 - Top Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC C-7
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure C-7 Nominal Primary-Side Flow Rate of 201 kg/s, Channel Set 1 - Side Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC C-8
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure C-8 Nominal Primary-Side Flow Rate of 201 kg/s, Channel Set 2 - Top Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC C-9
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure C-9 Nominal Primary-Side Flow Rate of 230 kg/s, Channel Set 1 - Side Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC C-10
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure C-10 Nominal Primary-Side Flow Rate of 230 kg/s, Channel Set 2 - Top Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC C-11
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure C-11 Nominal Primary-Side Flow Rate of 263 kg/s, Channel Set 1 - Side Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC C-12
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure C-12 Nominal Primary-Side Flow Rate of 263 kg/s, Channel Set 2 - Top Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC C-13
Appendix D TF-2 FEI Spectral Plots by Test Series (0 - 1000 Hz) pyright 2022 by NuScale Power, LLC D-1
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure D-1 Test Condition A, Channel Set 1 - Side Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC D-2
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure D-2 Test Condition A, Channel Set 2 - Top Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC D-3
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure D-3 Test Condition B, Channel Set 1 - Side Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC D-4
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure D-4 Test Condition B, Channel Set 2 - Top Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC D-5
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure D-5 Test Condition C, Channel Set 1 - Side Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC D-6
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure D-6 Test Condition C, Channel Set 2 - Top Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC D-7
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure D-7 Test Condition D, Channel Set 1 - Side Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC D-8
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure D-8 Test condition D, Channel Set 2 - Top Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC D-9
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure D-9 Test Condition G, Channel Set 1 - Side Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC D-10
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure D-10 Test Condition G, Channel Set 2 - Top Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC D-11
Appendix E TF-2 FEI Spectral Plots by Flow Rate (0 - 1000 Hz) pyright 2022 by NuScale Power, LLC E-1
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure E-1 Nominal Primary-Side Flow Rate of 114 kg/s, Channel Set 1 - Side Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC E-2
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure E-2 Nominal Primary-Side Flow Rate of 114 kg/s, Channel Set 2 - Top Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC E-3
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure E-3 Nominal Primary-Side Flow Rate of 143 kg/s, Channel Set 1 - Side Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC E-4
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure E-4 Nominal Primary-Side Flow Rate of 143 kg/s, Channel Set 2 - Top Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC E-5
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure E-5 Nominal Primary-Side Flow Rate of 173 kg/s, Channel Set 1 - Side Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC E-6
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure E-6 Nominal Primary-Side Flow Rate of 173 kg/s, Channel Set 2 - Top Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC E-7
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure E-7 Nominal Primary-Side Flow Rate of 201 kg/s, Channel Set 1 - Side Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC E-8
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure E-8 Nominal Primary-Side Flow Rate of 201 kg/s, Channel Set 2 - Top Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC E-9
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure E-9 Nominal Primary-Side Flow Rate of 230 kg/s, Channel Set 1 - Side Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC E-10
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure E-10 Nominal Primary-Side Flow Rate of 230 kg/s, Channel Set 2 - Top Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC E-11
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure E-11 Nominal Primary-Side Flow Rate of 260 kg/s, Channel Set 1 - Side Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC E-12
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure E-12 Nominal Primary-Side Flow Rate of 260 kg/s, Channel Set 2 - Top Sensors ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC E-13
Appendix F TF-2 FEI Content within Frequency Ranges of Interest pyright 2022 by NuScale Power, LLC F-1
Table F-1 Amplitude of Spectral Content between 0 - 10 Hz
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC F-2
Table F-2 Amplitude of Spectral Content between 0-10 Hz (Excluding Test Series G)
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC F-3
Table F-3 Amplitude of Spectral Content between 16-28 Hz
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC F-4
Table F-4 Amplitude of Spectral Content between 35-55 Hz
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC F-5
Table F-5 Amplitude of Spectral Content between 70-85 Hz
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC F-6
Table F-6 Amplitude of Spectral Content between 140-160 Hz
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC F-7
Table F-7 Amplitude of Spectral Content between 10-300 Hz
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC F-8
Appendix G TF-2 FEI Frequency-Specific Amplitudes versus Flow Rate pyright 2022 by NuScale Power, LLC G-1
Figure G-1 Dynamic Strain, 0-10 Hz Versus Flow Rate
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC G-2
Figure G-2 Dynamic Strain, 16-28 Hz Versus Flow Rate
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC G-3
Figure G-3 Dynamic Strain, 35-55 Hz Versus Flow Rate
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC G-4
Figure G-4 Dynamic Strain, 70-85 Hz Versus Flow Rate
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC G-5
Figure G-5 Dynamic Strain, 140-160 Hz Versus Flow Rate
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC G-6
Figure G-6 Dynamic Strain, 10-300 Hz Versus Flow Rate
}}2(a),(c),ECI pyright 2022 by NuScale Power, LLC G-7
pendix H TF-3 Build-out Testing Frequency Response Function Calculations pyright 2022 by NuScale Power, LLC H-1
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Table H-1 Testing Matrix ((
}}2(a),(c)ECI
© Copyright 2022 by NuScale Power, LLC H-2
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure H-1 315-12-1-C-5 (5y) ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC H-3
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure H-2 zSgle Tube 1sec ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC H-4
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure H-3 1A-1Y ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC H-5
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure H-4 1A-1Z ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC H-6
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure H-5 1A-3Z ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC H-7
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure H-6 1A-5Z ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC H-8
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure H-7 1C-1Y ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC H-9
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure H-8 1C-1Z ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC H-10
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure H-9 1C-5Z ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC H-11
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure H-10 1E-1Y ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC H-12
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure H-11 1E-1Z ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC H-13
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure H-12 1E-5Z ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC H-14
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure H-13 2A-CmsZ ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC H-15
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure H-14 2A-EfwZ ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC H-16
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure H-15 2A-EmsZ ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC H-17
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure H-16 2A-GfwZ ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC H-18
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure H-17 2C-CmsZ ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC H-19
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure H-18 2C-EfwZ ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC H-20
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure H-19 2C-EmsZ ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC H-21
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure H-20 2C-GfwZ ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC H-22
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure H-21 3C-EmsZ ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC H-23
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure H-22 4A-2Z ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC H-24
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure H-23 4A-3Z ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC H-25
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure H-24 5A-2Z ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC H-26
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure H-25 5A-4Y ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC H-27
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure H-26 5A-5Z ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC H-28
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure H-27 5A-Support Z ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC H-29
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure H-28 Sup-103+Col-12 Span E, Tube 1, 5, 9 Impact 1x ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC H-30
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure H-29 Sup-103+Col-12 Span E, Tube 1, 5, 9 Impact 1Z ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC H-31
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure H-30 Supp-to-Tube 103 Span-E FW-X ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC H-32
NuScale Comprehensive Vibration Assessment Program Measurement and Inspection Plan Technical Report TR-121354-NP Revision 0 Figure H-31 Supp-to-Tube 103 Span-E FW-X on Plate ((
}}2(a),(c),ECI
© Copyright 2022 by NuScale Power, LLC H-33
LO-133414 : Affidavit of Carrie Fosaaen, AF-133415 NuScale Power, LLC 1100 NE Circle Blvd., Suite 200 Corvallis, Oregon 97330 Office 541.360-0500 Fax 541.207.3928 www.nuscalepower.com
NuScale Power, LLC AFFIDAVIT of Carrie Fosaaen I, Carrie Fosaaen, state as follows: (1) I am the Senior Director of Regulatory Affairs of NuScale Power, LLC (NuScale), and as such, I have been specifically delegated the function of reviewing the information described in this Affidavit that NuScale seeks to have withheld from public disclosure, and am authorized to apply for its withholding on behalf of NuScale (2) I am knowledgeable of the criteria and procedures used by NuScale in designating information as a trade secret, privileged, or as confidential commercial or financial information. This request to withhold information from public disclosure is driven by one or more of the following: (a) The information requested to be withheld reveals distinguishing aspects of a process (or component, structure, tool, method, etc.) whose use by NuScale competitors, without a license from NuScale, would constitute a competitive economic disadvantage to NuScale. (b) The information requested to be withheld consists of supporting data, including test data, relative to a process (or component, structure, tool, method, etc.), and the application of the data secures a competitive economic advantage, as described more fully in paragraph 3 of this Affidavit. (c) Use by a competitor of the information requested to be withheld would reduce the competitors expenditure of resources, or improve its competitive position, in the design, manufacture, shipment, installation, assurance of quality, or licensing of a similar product. (d) The information requested to be withheld reveals cost or price information, production capabilities, budget levels, or commercial strategies of NuScale. (e) The information requested to be withheld consists of patentable ideas. (3) Public disclosure of the information sought to be withheld is likely to cause substantial harm to NuScales competitive position and foreclose or reduce the availability of profit-making opportunities. The accompanying report reveals distinguishing aspects about the process by which NuScale develops its CVAP Measurement and Inspection Plan. NuScale has performed significant research and evaluation to develop a basis for this process and has invested significant resources, including the expenditure of a considerable sum of money. The precise financial value of the information is difficult to quantify, but it is a key element of the design basis for a NuScale plant and, therefore, has substantial value to NuScale. If the information were disclosed to the public, NuScale's competitors would have access to the information without purchasing the right to use it or having been required to undertake a similar expenditure of resources. Such disclosure would constitute a misappropriation of NuScale's intellectual property, and would deprive NuScale of the opportunity to exercise its competitive advantage to seek an adequate return on its investment. (4) The information sought to be withheld is in the enclosed report entitled NuScale CVAP Measurement and Inspection Plan. The enclosure contains the designation Proprietary" at the top of each page containing proprietary information. The information considered by NuScale to be proprietary is identified within double braces, "(( }}" in the document. (5) The basis for proposing that the information be withheld is that NuScale treats the information as a trade secret, privileged, or as confidential commercial or financial information. NuScale relies upon the exemption from disclosure set forth in the Freedom of Information Act ("FOIA"), 5 USC § AF-133415 Page 1 of 2
552(b)(4), as well as exemptions applicable to the NRC under 10 CFR §§ 2.390(a)(4) and 9.17(a)(4). (6) Pursuant to the provisions set forth in 10 CFR § 2.390(b)(4), the following is provided for consideration by the Commission in determining whether the information sought to be withheld from public disclosure should be withheld: (a) The information sought to be withheld is owned and has been held in confidence by NuScale. (b) The information is of a sort customarily held in confidence by NuScale and, to the best of my knowledge and belief, consistently has been held in confidence by NuScale. The procedure for approval of external release of such information typically requires review by the staff manager, project manager, chief technology officer or other equivalent authority, or the manager of the cognizant marketing function (or his delegate), for technical content, competitive effect, and determination of the accuracy of the proprietary designation. Disclosures outside NuScale are limited to regulatory bodies, customers and potential customers and their agents, suppliers, licensees, and others with a legitimate need for the information, and then only in accordance with appropriate regulatory provisions or contractual agreements to maintain confidentiality. (c) The information is being transmitted to and received by the NRC in confidence. (d) No public disclosure of the information has been made, and it is not available in public sources. All disclosures to third parties, including any required transmittals to NRC, have been made, or must be made, pursuant to regulatory provisions or contractual agreements that provide for maintenance of the information in confidence. (e) Public disclosure of the information is likely to cause substantial harm to the competitive position of NuScale, taking into account the value of the information to NuScale, the amount of effort and money expended by NuScale in developing the information, and the difficulty others would have in acquiring or duplicating the information. The information sought to be withheld is part of NuScale's technology that provides NuScale with a competitive advantage over other firms in the industry. NuScale has invested significant human and financial capital in developing this technology and NuScale believes it would be difficult for others to duplicate the technology without access to the information sought to be withheld. I declare under penalty of perjury that the foregoing is true and correct. Executed on 12/31/2022. Carrie Fosaaen AF-133415 Page 2 of 2}}