ML20147D960
| ML20147D960 | |
| Person / Time | |
|---|---|
| Site: | Clinch River |
| Issue date: | 10/31/1978 |
| From: | Christie A ENERGY, DEPT. OF, CLINCH RIVER BREEDER REACTOR PLANT |
| To: | |
| Shared Package | |
| ML20147D956 | List: |
| References | |
| WARD-D-0218, WARD-D-218, NUDOCS 7812200247 | |
| Download: ML20147D960 (23) | |
Text
{{#Wiki_filter:. . . WARD-D-0218 CATEGORY 2
, 3;.f 1
.,Z Clinch River Breeder Reactor Plant STRUCTURAL RESPONSE OF CRBRP SCALE MODELS TO A SIMULATED HYPOTHETICAL CORE DISRUPTIVE 4
ACCIDENT I O OCTOBER 1978 Prepared for the United States Department of Energy under contracts EY-76-C-15-2395 and EY-78-C-15-0003. Any Further Distribution by any Holder of this Document or of the Data Therein to Third Parties Representing Foreign interest, Foreign Govern-ments, Foreign Companies and Foreign Subsidi-aries or Foreign Divisions of U.S. Companies Should be Coordinated with the Director, Division of Reactor Research and Technology, United States Department of Energy. ( W Westinghouse Electric Corporation w ADVANCE 0 REACTORS DIVISION
- 427tL2. BOX 158 M ADISON, PE NNSY LV ANI A 15663 I
78122002'[f
l l O WARD-D-0218 STRUCTURA' ,dSPONSE OF CRBRP SCALE MODELS TO A SIMULATED HYP0THETICAL CORE DISRUPTIVE ACCIDENT AUTHOR ALAN M. CHRISTIE PRINCIPAL C0clTRIBUTORS GENERAL ELECTRIC FAST BREEDER REACTOR DEPARTMENT (m~) N. W. Brown B. W. Joe SRI INTERNATIONAL C. M. Romander D. J. Cagliostro WESTINGHOUSE ADVANCED REACTORS DIVISION M. A. Todd S. Ranatza Approved: A " ll L. E. Strawbridge Manager Margin Analysis and Design O
l l O l l lNFORMATION CONCERNING USE OF THIS DOCUMENT ! PRELIMINARY DOCUMENT l This document contains Information of a preliminary nature prepared in the course of work ' for the U.S. Department of Energy. This information is subject to correction or modification upon the collection and evaluation of additional data. NOTICE This document was prepared as an account of work spon :ored by the United States Government. Neither the U.S. Department of Energy, nor any of their employees, nor any of their contractors, subcontractors, or the employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, comr,leteness or l Os usefulness of any information, apparatus, product or process disclosed, or represents that its use would not infringe privately owned rights. l WESTINGHOUSE ELECTRIC CORPORATION I ADVANCED REACTORS DIVISION BOX 158 MADISON, PENNSYLVANIA 15663 O V
TABLE OF CONTENTS O O Page 1.0 Introduction 1 2.0 Sunmary and Conclusions 8 2.1 Summary and Conclusions from Experimental Program 8 2.2 Sunmary and Conclusions from Analytic Program 9 3.0 The Experimental Program 15 3.1 Phase 0 - Evaluation of Material Properties 15 3.2 Phase 1 - Energy Source Calibration 23 3.3 Phase 2 - The Scale Model Tests 27 3.3.1 The SM-1 Test 27 3.3.1.1 Objectives of the SM-1 Test 27
- 3. 3.1. 2 The SM-1 Model and Instrumentation 28 3.3.1.3 The SM-1 Test Results 29 3.3.2 The SM-2 and SM-3 Tests 30 3.3.2.1 Objectives of the SM-2 and SM-3 Tests 30 p
3.3.2.2 The SM-2 and SM-3 Models and Instrumentation 31 3.3.2.3 The SM-2 and SM-3 Test Results 32 3.3.3 The SM-4 and SM-5 Tests 34 3.3.3.1 Objectives of the SM-4 and SM-5 Tests 34 3.3.3.2 The SM-4 and SM-5 Models and Instrumentation 35 3.3.3.3 The SM-4 and SM-5 Experimental Results 37 4.0 Analytic Simulations of the SM Series Tests 83 4.1 Pre-Test Predictions 83 4.1.1 Comparison of REXC0-HEP Pre-Test Predictions with SM 83 Series Test Results 4.1.1.1 REXC0-HEP Models 84 4.1.1.2 Pre-Test Analysis Results 85 4.1.1.3 Conclusions from Pre-Test Analysis 93 4.1.2 Comparison of ANSYS Vessel Head Predictions with the 94 Test Results 4.1.2.1 The ANSYS Models 95 4.1.2.2 The Pre-Test Analysis Results 96 4.1.2.3 Conclusions from Pre-Test ANSYS Analysis 100 iv
TABLE OF CONTENTS (Continued) Page 4.1.3 Assessment of Differences Between the 100 Predictions and Test Results 4.2 Assessment of Deviations in Response Associated with 161 Scale Model Non-Prototypicalities 4.2.1 Review of Scaling Laws 1 61 4.2.2 Application of Scaling Laws to Scale Model Tests 164 4.2.3 Scale Model Simulation Using Prototypic Properties 170 4.3 Cover Gas Response and Its Leakage Implications 1 91 Appendix A Detail'Comprison of Experimental and Calculational 200 Data on SM-2, SM-3 and SM-4 Scale Model Tests Appendix 8 Scale Model Drawings for SM-2 Through SM-5 220 0 O
LIST OF TABLES Table Page Number 2.1-1 Summary and Conclusions of Scale Model Tests 13 3.3-1 Head Deflections for Static Test SM-1 43 Energy Partitioning in Models SM-2 and SM-3 44 3.3-2 3.3-3 Energy Partitioning in Models SM-4 and SM-5 45 Summary of Data from SM-2, SM-3, SM-4 and SM-5 46 3.3-4 Material Properties Used in REXC0 Pre-Test Analyses 107 4.1-1 of the SM Tests 4.1-2 Comparison of SM-4 Analyses Using Expected and 108 Calibrated P-AV Curves 4.1-3 Comparison of Predicted and Experimental Energies 109 and Slug Impact Velocities SM-2 Head Element Connectivities and Real Constants 1p 4.1-4 4.1-5 Nodal Coupling and Component Definition for SM-4/SM-5 111 Head Model Element Real Constants for SM-4/SM-5 Head Model 112 4.1-6 Summary of SM-2 Vessel Head Analytical and 113 d 4.1-7 Experimental Results Summary of SM-4/SM-5 Vessel Head Analytical and 114 4.1-8 Experimental Results Comparison of Pre-Test with Post-Test ANSYS Vessel 115 4.1-9 Head Responses Significant Dimensional Quantities 173 4.2-1 Operations on Dimensional Matrix 174 4.2-2 Similitude Relationships 175 4.2-3 Comparison of Prototypic and Scale Model Physical 176 4.2-4 Parameters Comparison of REXC0-HEP SM-4/SM-5 Prototypic and 177 4.2-5 l Scale Model Results Comparison of Energy Partitioning in SM-4/SM-5 178 4.2-6 prototypic and Scale Model Calculations (At , Slug Impact) J ya
LIST OF TABLE (Continued) Table Number Page 4.2-7 Comparison of Energy Partitioning in SM-4/SM-5 179 Prototypic and Scale Model Calculations (At Dynamic Equilibrium) 4.2-8 Comparison of Core Barrel and Vessel Wall Strains 180 from REXC0-HEP SM-4/SM-5 Prototypic and Scale Model Calculations A.2-1 Summary of Appendix A Figures 202 A.2-? Comparison of Pressures and Strains - REXC0 vs. 203 Experiments O vii.
LIST OF FIGURES U Figure Page Number 1.0-1 Pressure-Volume Relationship for SMBDB Loadings 5 14 2.1-1 1/20 Scale Models of the CRBR 3.1-1 Stress Strain Properties of Ni-200 and 304 Stainless 18 & 19 Steel Stress Strain Properties of the SM-3 and SM-5 UIS 20 3.1-2 Columns 21 3.1-3 Stress Strain Curves for 533B Class 1 Carbon Steel Strain Rate Effects in 5338 Class 1 Carbon Steel 22 3.1-4 Energy Source Calibration Apparatus 25 3.2-1 Pressure-Volume Change and Gas Work - Volume 26 3.2-2 Change Curves for Energy Source 48 3.3-1 Schematic of SM-1 Vessel Head SM-1 Test Apparatus 49 3.3-2 Instrumentation Layout for Static Test SM-1 50 3.3-3
' Profile of Head at Six Pressures; Static Test SM-1 51 3.3-4 Pressure-Volume Change Relation for SM-1 Hydrostatic 52 3.3-5 Loading 3.3-6 Final Deformed Shape of Head: Static Test SM-1 53 SM-2 with Instrumentation 54 3.3-7 SM-3 with Instrumentation 55 3.3-8 Comparison of Loading Pressures: SM-2 and SM-3 56 & 57 3.3-9 Comparison of Accelerations: SM-2 and SM-3 58 & 59 3.3-10 Comparison of Strain Response: SM-2 and SM-3 60 & 61 3.3-11 Comparison of Deformed Shape Profiles: SM-2 and 62 3.3-12 SM-3 Comparison of Water Surface Displacements:SM-2 63 3.3-13 and SM-3 3.3-14 Upper Internals Structure and Head Assembly from 64 SM-3 65 3.3-15 Deformation Profiles for UIS Columns in SM-3 66 3.3-16 SM-4 with Instrumentation O
V viii
/
LISTOFFIGUPES(Continued) Figure Number _ Page 3.3-17 SM-5 with Instrumentation 67 3.3-18 SM-5 Vessel Head Sealing Arrangement 68 3.3-19 Comparison of Loading Pressures: SM-4 and SM-5 69 & 70 3.3-20 SM-5 Cover Gas, Inlet Plenum and VIS Pressures 71 & 72 3.3-21 Comparison of Strain Response : SM-4 and SM-5 73 & 74 3.3-22 Loadings on SM-5 UIS Columns 75 3.3-23 Deformation Profiles for VIS Columns in SM-5 76 3.3-24 Comparison of Deformed Shape Profiles: SM-4 and 77 SM-5 3.3-25 Comparison of Accelerations: SM-4 and SM-5 78 & 79 3.3-26 SM-5 Vessel Head Accelerations and Displacement 80 3.3-27 Comparison of Water Surface Displacements: SM-4 81 and SM-5 4.1-1 aEXC0-HEP Model for SM-2 ~ Pre-Test Analysis 116 4.1-2 REXCO-HEP Model for SM-3 . Pre-Test Analysis l' 4.1-3 REXCO-HEP Model for SM-4/SM-5 Pre-Test Analysis 118 4.1-4 Core Barrel and Peactor Vessel Stress Strain Curves 119 Used in Analysis 4.1-5 Pressure-Volume Change and Gas Work-Volume Change 120 Relationships for Pre-Test Analysis 4.1-6 SM-2 Predicted and Experimental Deformation Profiles 121 ! 4.1. 7 SM-3 Predicted and Experimental Deformation Profiles 122 4.1-8 SM-4 Predicted and Experimental Deformation Profiles 123 4.1-9 SM-2 Predicted and Experimental Core Pressure 124 at Gauge P) 4.1-10 SM-3 Predicted and Experimental Vessel Wall Pressure 125 at Gauge P 5 4.1-11 SM-3 Predicted and Experimental Vessel Wall Pressure 126 at Gauge P 5 4.1-12 SM-4 Predicted and Experimental Vessel Wall Pressure 127 l at Gauge P 4 4.1-13 SM-2 Predicted and Experimental Vessel Wall Pressure 128 l at Gauge P l 7 4.1-14 SM-3 Predicted and Experimental Vessel Wall Pressure 129 at Gauge P l 7 ix '
LIST OF FIGURES _(Continued) Page Number SM-4 Predicted and Experimental Vessel Wall , 130 4.1-15 Pressure at Pressure Gauge P 5 131 4.1-16 SM-2 Predicted and Experimental Head Pressure at Pressure Gauge P 8 132 4.1-17 SM-3 Predicted and Experimental Head Pressure at Pressure Gauge P 8 133 4.1-18 SM-4 Predicted and Experimental Head Pressure at Pressure Gauge P6 134 4.1-19 SM-5 Predicted and Experimental Vessel Wali Pressures at Pressure Gauge P 5 135 4.1-20 ANSYS Model for SM-2 Pre-Test Head Analysis 136 4.1-21 ANSYS Model for the SM-4/SM-5 Pre-Test Head Analysis 137 4.1-22 Comparison of Scaled SMBDB Head Load with Predicted SM-2 Head Load 138 4.1-23 Comparison of Scaled SMBDB CSP Load with Predicted p V SM-2 CSP Load 139 4.1-24 Predicted Force on SM-2 Margin Ring 140 4.1-25 Predicteo Displacement of SM-2 Vessel Head 141 4.1-26 Comparison of Predicted and Experimental SM-2 Head Plug Accelerations 142 4.1-27 Comparison of Predicted and Experimental SM-2 Vessel Flange Accelerations 143 4.1-28 SM-4/SM-5 Predicted Force on Vessel Head 144 4.1-29 SM-4/SM-5 Predicted Force on CSP 145 4.1-30 SM-4/SM-5 Predicted Pressure Under VIS 146 4.1-31 Comparison of Predicted and Experimental SM-5 LRP Acceleration 147 4.1-32 Comparison of Predicted and Experimental SM-5 IRP Acceleration 148 4.1-33 Comparison of Predicted and Experimental SM-5 SRP Acceleration /~T U x
LIST OF FIGURES (Continued) Figure Number Page 4.1-34 Comparison of Predicted and Experimental SM-5 IRP 149 Displacements 4.1-35 Comparison of Predicted and Experimental SM-5 LRP 150 Frequency Spectra 4.1-36 Comparison of Predicted and Experimental SM-5 IRP 151 Frequency Spectra 4.1-37 Comparison of Predicted and Experimental SM-5 SRP 152 Frequency Spectra 4.1-38 Predicted SM-4 LRP Margin Ring Load 153 4.1-39 Predicted SM-4 IRP Margin Ring Load 154 4.1-40 Predicted SM-4 SRP Margin Ring Load 155 4.1-41 Comparison of Pre- and Post-Test Analysis Vessel 156 Wall Stress-Strain Curves 4.1-42 Comparison of SM-2 Pre- and Post-Test REXCO-HEP 157 Deformation Profiles 4.1-43 Comparison of SM-4/SM-5 Pre- and Post-Test REXC0-HEP 158 Defonnation Profiles 4.1-44 Comparison of Pre- and Post-Test REXC0-HEP Head Loads 159 As Applied to ANSYS Head Model 4.1-45 Comparison of Pre- and Post-Test ANSYS Simulations 160 of SRP Resonse 4.2-1 Vessel Stress-Strain Input Data for REXCO-HEP 181 Post-Test Analysis 4.2-2 Core Barrel Stress-Strain Input Data for REXC0-HEP 182 Post-Test Analysis 4.2-3 Comparison of Core Barrel Pressures from 183 Calculations Using Scale Model and Prototypic Properties 4.2-4 Comparison of Average Slug Velocities from 184 Calculations Using Scale Model and Prctotypic Properties l
i LIST OF FIGURES (Continued)
--) Figure Page Number 185 4.2-5 Comparison of Slug Forces from Calculations Using Scale Model and Prototypic Properties 186 4.2-6 Comparison of Net Force on CSS from Calculations Using Scale Model and Prototypic Properties 187 4.2-7 Comparison of Pressure Under UIS From Calculations Using Scale Model and Prototypic Properties 188 4.2-8 Comparison of Core Barrel Strain Energies from Calculations Using Scale Model and Prototypic Properties 189 4.2-9 Comparison of Vessel Wall Strain Energies from Calculations Using Scale Model and Prototypic Properties 190 4.2-10 Comparison of Peak Vessel Wall Radial Displacements from Calculations Using Scale Model and Prototypic Properties 193 4.3-1 Region Modelled in MAXPRES Analysis 194 4.3-2 Driving Pressure at Lower Shield Plate Elevation Calculated Pressure Response in Dip Seal Region 195 4.3-3 (No Orifice losses) 196 4.3-4 Comparison of Calculated and Experimental Pressure Response in Dip Seal Region (Orifice Losses Accounted For) 197 4.3-5 Comparison of Calculated and Experimental Pressure Response Above Marain Rina (No Orifice Losses Accounted For Across Margin Ring)
A-1 Pressure Versus Time SM-2 Pre-Test Analyses 204 & 205 A-2 Strain Versus Time SM-2 Pre-Test Analyses 206 Pressure Versus Time SM-3 Pre-Test Analyses 207 & 208 A-3 Strain Versus Time SM-3 Pre-Test Analyses 209 A-4 A-5 Pressure Versus Time SM-4 Pre-Test Analyses 210 & 211 Strain Versus Time SM-4 Pre-Test Analyses 212 O A-6 ( xii
LIST OF FIGURES (Continued) Figure Number Page A-7 Pressure Versus Time SM-2 Post-Test Analyses 213 & 214 A-8 Strain Versus Time SM-2 Post-Test Analyses 215 A-9 Pressure Versus Time SM-4/SM-5 Post-Test Analyses 216 & 217 A-10 Strain Versus Time SM-4/SM-5 Post-Test Analyses 218 B-1 Detailed Cross Section of SM-3 221 B-2 Detailed Cross Section of SM-4 222 .B-3 Detailed Cross Section of SM-5 . 223 O xiii
Acknowledgments O The scale model tests and supporting analyses discussed in this report l involved the technical cooperation of personnel in the General Electric Fast Breeder Reactor Department. SRI International, the Westinghouse i Advanced Reactors Division and the U.S. Department of Energy. In ) addition, personnel of the Terra Tek company were involved in obtaining materials data for the tests. N. W. Brown and B. W. Joe of GE were responsible for performing the REXC0-HEP simulations of the tests, reducing and inte.rpreting the results and organizing those aspects of the report '. elated to these analyses. C. W. Romander and D. J. Cagliostro were the principal experimental investigators at SRI International. D. J. Cagliostro was responsible for the overall test program. In addition to being involved in the testing, C. M. Romander performed most of the post-test analysis of the experimental results and provided significant input to this report. C. M. Romander also provided the Fast Fourier Transforms of the vessel head plug responses. At Westinghouse Advanced Reactors Division, M. A. Todd provided support primarily in the area of pre- and post-test dynamic simulation of the responses of the scale model vessel heads. In this effort, he was supported by T. M. Frick. The simulations of cover-gas response were performed by S. Ranatza. Additional support effort in the early planning stages of the program was provided by R. G. Jones. l In addition to the above, personnel from the U.S. Department of Energy were involved in the planning and implementation of the program. The administration of the program was under S. Berk of the Reactor Research and Technology Branch while M. McKeown of the CRBRP Project Office was responsible for ensuring that the tests met the needs of the CRBRP Project. i 1 O . x1y l l
1.0 INTRODUCTION
A series of scale model tests has been performed to simulate the dynamic loadings that could arise from a hypothetical core disruptive accident The primary (HC]A) in the Clinch River Breeder Reactor Plant (CRBRP). objectives of these tests was to verify that scale models of this reactor can withstand the loads resulting from the simulation of an HCDA. Scale model tests sia:ulating fast reactor response to HCDAs have been extensively used in the last two decades both in the U.S. and abroad. For example, in the late 1950s and early 1960s, the Naval Ordnance Laboratory used idealized scale models with bare explosive charges to investigate the structurai .esponse of the Fermi fast reactor to Hr.nA l loadings (Reference 1-1). In the early 1970s, Stanford Research Institute (now SRI International) I conducted a broad testing program to examine the response of the Fast Flux Test Facility (FFTF) to an HCDA (Reference 1-2). In that program a series of simplified and detailed models of FFTF was tested. An explosive source l was used whose characteristics closely matched those postulated to result from an energetic fuel-coolant interaction. In addition to tests of the reactor vessel and head, the response of both the FFTF Primary Heat Transport System and individual pipes was examined in this program. As a result of the test program and subsequent analyses by the Hanford Engineering Development Laboratory (Reference 1-3), it was concluded that FFTF has significant structural margin against failure for the postulated accident energetics. The above programs have also generated data against which to compare the results of computer codes such as REXC0-HEP (Reference 1-4) used to simulate the hydrodynamic-structural response of fast reactors. Such comparisons (References 1-5,1-6,1-7) have allowed improvements to be made, and confidence to be gained in the use of these codes. A secondary objective of the CRBRP scale model HCDA test program is to provide additional information for the continued qualification and improvement of such codes. In particular, the experimental data can be used to provide assurance that b 1
reactnrs having CRBRP-like features (e.g., upper internals structures and three plug rotating head) can be adequately simulated. Three computer codes were used to simulate different aspects of the reactor response. REXC0-HEP was used to simulate the system hydrodynamic-structural response. ANSYS (Reference 1-8) was used to simulate the response of the individual plugs of the vessel head and MAXPRES (Reference 1-9) was used to simulate cover-gas pressurization. Summarizing, the objectives of this program were: e Assess the ability of CRBRP scale models to withstand the loads resulting from a simulated hypothetical core disruptive accident. e Obtain a systematic understanding of the response and interaction of reactor components such as rotating head plugs and upper internals structure to the simulated HCDA loadings. e Verify that the above computer codes provide adequate represent-ations of the response of reactors whose configuration is similar to that of CRBRP. To meet these objectives, a combined experimental and analytic program was established. Early in the program, it was agreed that the HCDA to be simulated would be that resulting in the Structural Margin Beyond the Design Base (SMBDB) loads being used in the design of CRBRP. This case is based upon an HCDA whose working fluid is a high-pressure two-phase mixture of fuel, the thermodynamic characteristics of which define the pressure-volume relationship shown in Figure 1.0-1. If the volume of this working fluid is expanded until a pressure of one atmosphere is reached, the resulting work energy release, assuming an isentropic expansion, is 661 MJ. The experimental program (Section 3) is divided into three phcses: materials testing (Phase 0), explosive source calibration (Phase 1) and scale model testing (Phase 2). In Phase 0, the stress-strain properties of the simulant materials at various strain rates were assessed. In Phase 1, the explosive source characteristics (i.e., the pressure-volume relationship) 2
were made to match those of the postulated fuel-vapor expansion. In Phase 2, a series of five 1/20th scale models was tested. For the first of these tests (SM-1) a model of the three-plug vessel head was hydro-statically loaded to the point of plug disengagement. This allowed an estimate to be made of the head deformation profile and ultimate static ; load carrying capability. In the second and third tests (SM-2 and SM-3) l l idealized reactor vessels were tested dynamically using the source developed in Phase 1. These tests provided a basis for comparison with analytic predictions and for the systematic understanding of basic response l cha racterstics. Tests SM-4 and SM-5 were essentially identical dynamic ter Y (although SM-5 was more extensively instrurnented) to assure reproductibility of response. These two tests were also the most detailed, having all reactor internal components necessary to produce a prototypic system response. In none of these tests was the primary heat transport system modelled. The analytic program (Section 4) supporting the tests consisted of pre-and post-test phases. In the pre-test phase, coupled hydrodynamic-structural calculations were performed using the REXC0-HEP code to predict the reactor system response. In addition more detailed structural cal- , 1 culations using the ANSYS code were performed to assess the vessel head ; O response to the slug load predicted by REXC0-HEP. These analyses, performed i for tests SM-2 through SM-5, predicted that the test components were capable of withstanding the dynamic loads and provided computer simulations of the tests unprejudiced by the experimental results. The post-test analyses served two main functions. First, they were used to resolve differences between the pre-test analyses and the test results. Secondly, an assessment was made of the deviations from non-prototypicality in tests SM-4 and SM-5 resulting from the scale model materials (e.g. Ni-200 for the vessel, water for the coolant) having properties slightly different from those of the actual reactor. This allowed adjustments to the experimental results to be made so that a prototypic response could be determined. Also covered in Section 4 are the experimental and supporting analytical efforts used to assess the potential for CRBRP vessel head leakage. l 3 l
l This report discusses the tests and supporting analysis from the perspective of structural boundary integrity. However, SRI International l personnel have also prepared a detailed report on these tests l (Reference 1-10). The SRI report provides more detail on the exper' mental l procedures used and further explanation of the experimental results. ! l l l 1 l 9 4 9
O O O 8 v
'P
- Initial pressure = 273 bars Initial volume = 2.558 meters3 250 -
200 - . E i ac I E a 150 - N
<n E
a. 100 - i 50 - J I I I t i l l i I I I 0 30 40 50 60 70 80 90 100 110 0 10 , 20 i VOLUME (METERS3 ) i Figure 1.0-1 Pressure-Volume Relationship for SMBDB Loadings i
1 References for Section I l-1 W. R. Wise, Jr. , et al . , " Response of Enrico Fermi Reactor to TNT Simulated Nuclear Accidents", N0LTR 62-207, November 1964. , 1-2 A. L. Florence and G. R. Abrahamson, " Simulation of a Hypothetical Core Disruptive Accident in a Fast Flux Test Facility," HEDL-SRI-1, May 1973. (Availability: U.S. DOE Technical Information Center) 1-3 G. L. Fox, et al., " Scale-up of Test Results from Simulation Experiments of a Hypothetical Core Disruptive Accident in the Fast Flux Test Facility," HEDL-TME-74-54, 1974. (Availability: U.S. DOE Technical Information Center) 1-4 Y. W. Chang and J. Gvildys, "REXC0-HEP: A Two-Dimensional Computer Code for Calculating the Primary System Response in Fast Reactors," ANL-75-19, June 1975. (Availability: U.S. DOE Technical Infonnation Center) 1-5 T. J. Marciniak, et al., " Analysis of FFTF Primary-Containment Complex Model Experiments," ANL-8062, January 1974. (Availability: U.S. DOE Technical Information Center) 1-6 Y. W. Chang and J. Gvildys, "Arialysis of the FFTF Primary-System Response to an HCDA," ANL-8066, January 1974. (Availability: U.S. DOE Technical Infonnation Center) 1-7 G. Nagumo and C. Fiala, " Comparison of FFTF Simple-Model Tests with REXC0 Predictions," ANL 8071, January 1974. (Availability: U.S. DOE Technical Information Center) 1-8 G. J. DeSalvo and J. A. Swanson, ANSYS Engineering Analysis System User's Manual, Swanson Analysis Associates, Inc., Elizabeth, Pa., 1975. 6
1-9 S. Ranatza, "MAXPRES-2 Computer Code Verification," CRBRP-ARD-0175, July 1977. (Availability: U.S. DOE Technical Information Center) 1-10 C. Romander and D. J. Cagliostro, " Experimental Simulation of a Hypothetical Core Disruptive Accident in 1/20-Scale Models of the Clinch River Breeder Reactor," Technical Report 4, to be published. (Availability: SRI International, Menlo Park, Calif.) l O 7
2.0
SUMMARY
AND CONCLUSIONS The experimental and supporting analytic programs, together with major conclusions, are summarized below. 2.1 Summary and Conclusions from Experimental Program _ The experimental program was divided into three phases: Phase 0 (materials testing), Phase 1 (onergy source development) and Phase 2 (scale model dynamic testing) . In Phase 0, the structural properties of the simulant materials (Ni-200 was substituted for 304 stainless steel while SA-533B carbon steel was substituted for SA-508 carbon steel) were tested under low and high strain rate conditions. In Phase 1, a low-density explosive source, used to simulate the expanding fuel vapor, was developed. Calibration experiments were carried out in a simplified 1/20 scale rigid vessel and core barrel which was filled with water. The explosive was detonated inside a vented steel cannister placed in the core region and surrounded by a volume of air. The resulting bubole pressure and volume changes were measured to give the source pressure-volume relationship. l In Phase 2, five models, one hydrostatic and four dynamic, were tested. All the tests (SM-1 through SM-5) were conducted at 1/20 scale and were sequenced with increasing complexity. A schematic of the dynamic models is shown in Figure 2.1-1. The test objectives and conclusions are summarized in Table 2.1-1. In test SM-1 a three plug model of the CRBR vessel head was loaded undar hydrostatic pressure until the structural instability point was reached. Test SM-2 was the simplest of the dynamic tests of the reactor system. It included a simplified vessel, core barrel and one-piece vessel head. 8
Test SM-3 was identical to SM-2 except that an upper internals structure (UIS) was added. By comparing the response of the SM-2 test to that of SM-3, the effect of the UIS was determined. Test SM-4 incorporated the essential features of the reactor which are necessary to assess system structural response. As such it included the UIS, vessel thermal liner, detailed core support structure with inlet plenum and a vessel head which incorporated three rotating plugs. The SM-5 model tested was identical
- with SM-4 to check for reproducibility. i However, test SM-5 was more extensively instrumented to obtain a broader range of experimental information.
l The tests confirmed the conservatism of the methods used to generate the SMBDB loads. In addition, they have resulted in an improved l understanding of both the vessel head and reactor system response I to these loads. Test SM-1 showed that the most likely head failure mode is neither through local distortion at the margin-ring juncture nor margin ring rollout. (These were considered possible failure mechanisms prior to the test.) Failure was found to result from margin-ring disengagement I due to doming of the large rotating plug. However, the disengagement was found to occur at a very high static pressure (ml160 psi). From the SM-2 and SM-3 tests, the magnitude ad profile of vessel and core barrel deformations and the effect of the UIS on these deformations was obtained. Tests SM-4 and SM-5, the most prototypir of the tests, further confirm that the reactor vessel and the three plug head 6esigns had significant margin against failure with respect to the SMBDB loads. The inclusion of the UIS and thermal liner significantly reduced the peak strains in the reactor vessel wall. In addition, these tests have confirmed the conclusion from the SM-1 test that the margin-ring junctures have adequate strength to accommodate the SMBDB loads. 2.2 Summary and Conclusions from Analytic Program The supporting analytic program was aimed at veri'fying that computer models based on the REXC0-HEP and ANSYS codes provide adequate representation of p the structural response of CRBR under HCDA loads. An additional aim of the b
*dxcept fo nnealing differences in the UIS columns as indicated on Figure 3.1-2.
9
analyses was to assess the changes in reactor response which would result in going from the simulant materials to the prototypic materials. In the first of these two aims, pre-test predictions and post-test analyses were performed with the structural / hydrodynamic code REXCO-HEP and the structural code ANSYS. Pre-test analyses were performed to support tests SM-2 through SM-5. These analyses provided guidance, for example, in the placement and specification of instrumentation and allowed an assessment to 'oe made of the predictive power of such computer models, unbiased by the results from the tests they are simulating. From these analyses, the following conclusions can be drawn:
- 1) The predicted pressure loads within the vessel were, with a few exceptions, comparable to or greater than the experimental loads.
Where deviations did exist these were explained in terms of the modelling assumptions inherent in the code Bee Section 4.1.3). 1
- 2) The peak vessel wall strains predicted by REXCO-HEP conservatively enveloped the experimental strains. In the SM-2 and SM-3 calculations, REXCO-HEP under-predicted mid-vessel wall strains. However, this l
underprediction was not observed in the SM-4/SM-5 calculation which included a thermal liner. l
- 3) The ANSYS simulations of the SM-2 and SM-4/SM-5 vessel head responses predicted elastic response for all plug junctures except at the SM-4/
SM-5 large rotating plug (LRP) juncture where a small amount of plastic deformation was predicted. These results are in close agreement with the tests where virtually no plastic deformation was observed at any of the plug junctures. Post test analyses with REXC0-HEP were performed in an attempt to reduce the differences between the pre-test predictions and the test results. The major change was in the use of material properties reflecting high strain rate characteristics. From these post-test analyses, the following conclusions can be drawn: 10 O
- 1) The use of the high strain rate material properties brought the calculated vessel peak strain into closer agreement with the tests.
I
- 2) The improved modelling in the post-test analyses did not result in substantial changes in the predicted pressures.
The continuing differences result from modelling assumptions (for example related to the core barrel shielding and the UIS) inherent in the version of the REXC0-HEP code that was used (see Section 4.1.3 for further discussion). 3
- 3) The post-test SM-4/SM-5 vessel head analysis using the ANSYS code showed some reduction in margin ring forces compared to the pre-test analysis. This reduction resulted from the use of the REXC0 post-test analysis head load. No plastic deformation occurred at any of the plug junctures. This is more consistent with the test results than were the pre-test predictions where some limited deformation occurred at the LRP juncture.
Post-test analyses were also performed to assess the sensitivity of reactor response as a result of using simulant materials rather than prototypic materials. This was addressed first through analytical application of the scaling relationships resulting from the Buckingham Pi Theorem and second through direct numerical simulation using the prototypic material properties in the REXCO-HEP code. Conclusions from these analyses are:
- 1) The prototypic slug load on the head would be less severe and delayed in time compared to that predicted in the test. However, those differences are very minor.
- 2) The core barrel and vessel wall deformation may be marginally greater in the prototypic case (the peak vessel strain changes from 2.66 percent to 2.96 percent in the REXC0-HEP calculations).
O 11
- 3) Dimensional analysis predicted deviations which were very consistent with those calculated by direct numerical simulation.
Also in the post-test analysis cover-gas pressure response through and above the large margin ring was simulated using the MAXPRES-2 code. This analysis showed that, as a result of neglecting energy dissipating mechanisms, predictions of cover-gas pressurization in the riser annuli are likely to be very conservative. O-12 1
'm Table 2.1-1 Sumary and Conclusions of Scale Model Tests Ocjectiva of Test conclusica s M Description of Model Obtain understanding of margin ri and Pargin rings do net roll out of head SM-1 Three-plug model of head loaded recess.
hyd-ostatically. plug .;eformations. A%ist in defining failure modes. High limiting pressure (s1160 psi). Obtain estimate of limiting pressure. l'ltimate failure through large plug doming. not through lxal margin ring Obtain quantitative estimate of head force / distortion. deflection characteristics. Simpliffed vessel with one-plug head Verify that simplified model of CRBR can Peak strain in vessel wall and core SM-2 accomodate SM3DB loads (reactor model barrel much lower than plas*ic in-and no upper internals. stability strain (4.4 and 1.3 percent closely matches that used to calculate REXCO-HEP SFBDB loads), respectively). l!se as base case against which to assess 1.ccal deforwtion in head-to-vessel results of SM-3 (which has upper internals), margin ring regMn virtually clastic. Simplified vessel with one-plug head and Assess effect of upper internaisst-ucture SM-3 on reactor response. Vessel peak strain reduced by 36 per-including upper ir.ternals structure. cent over SM-2. Core barrel strain increased by 15 d percent over SM-2. Deak head pressure 34 percent less thae that in SM-2. Slight vertical rise of upper internals and bending of columns. Verify adequacy of response model, whic5 Vessel peak strain reduced by 61 Most detailed reactor model; includes percent over SM-2. SM-4 three-plug head with margin rings, includes the essential design features upper internals, horizontal baf fle, affecting shock wave and mechanical load transmi ston. Core barrel strain 38 percent less and thermal liner, than SM-2. Yerify structural adequacy of three-plug Local defomation at all margin ring head derign to SM308 slug load. junctures virtually elastic. Peak head pressure 21 percent less than SM-2. , Verify reproducibility by comparing with Coolant boundary response very similar 5M-5 A duplicate test of SM-4 with increased to that of SM-4 instrumentation. SM-4 responses. Obtain broader range of experimental infor-mation than in SM-4.
ACCE LE ROMETERS 1 COMPLEX HEAD M I SIMPLE HE AD 4
*T f-t I t1 / ST A D\
^ -
= 7 -;-
.}.- ----- g - =g 133,: ;;- i . ;, __ _
PRESSUR E :; t - ?l L L 2 M' i I, j d ri__ f TRANSOUCERS { -- _T- j l jf STRAIN GAGES l{ _- _ - . 2 d l sa' l k i a If i - t g .
'i 12.15 IN. 60119 IN VESSEL WALL d THERMAL LINER =
STR AIN .j-GAGES $ $ UPPER INTERNAL :
$ (10 PLACES). STR UCTUR E t. .g F
- 3. 72 r $ 1 !!
' t.
\a C IN.
k DIAPHR AGM WITH I k l r PRE SSURE l l u =q CORE BARREL 5 ] A-w Q g TRANSDUCERS g g f.Cg-s E (0.100 IN. Th.Ck Ni2001 a
]e r! -; u=.
(=; y . q: z. p= i 7 F_ j . fb ' CHARGE CANISTER f 7M,2 " jy i V p/ fn- j i 91 Mi 1 i ( r(ig1m m /g
\ACCE LEROMETERS
(_ SM 3 SM 4,SM 5 SM 2 M A-3929-160 Figure 2.1-1 1/20-SCALE MODELS OF THE CRBR
# # 9
3.0 The Experimental Program In Reference 3-1, a detailed account of the experiments and test results is given. This section summarizes that information with emphasis on results which are of particular relevence to structural boundary integrity. 3.1 Phase 0 - Evaluation of Material Properties Stress-strain tests were performed on samples of the materials used to fabricate the 1/20-scale models for two reasons. First, tests were performed to demonstrate that the materials in the 1/20-scale models of the CRBRP reasonably simulate, at room temperature and at strain rates 20 times higher than those in the prototype, the average stress-strain properties of the prototypic materials at reactor operating temperatures. In the scale models, 533-8 Class 1 carbon steel simulated the 508 Class 2 carbon steel vessel head material, and Ni-200 simulated the 304 stainless steel vessel wall, core barrel, core support structure and the 316 stainless steel upper internals column materials. The second reason was to provide reliable material property data that could be used in pre- and post-test analysis. For this reason, stress-strain tests were performed at both low and high strain rates. Low strain rate tests were also performed as part of a quality assurance program to ensure that proper materials were being used and that proper annealing procedures were being followed on completed models. High strain rate tests were performed to provide more accurate data for use in computer analysis of the models and to demonstrate the strain rate sensitivity of the materials. To maximize reproducibility between the test models all similar components (e.g. , all core barrels) were fabricated from the same billet of material. The test specimens were then exposed to the same heat treatment of the corresponding component, thus assuring consistency of material properties. (The only except to this was the UIS columns as discussed below.) O 15
Figure 3.1-1 shows the resulting stress-strain properties for the vessel walls, the core barrels and the core support plates, both at low and high strain rates. The batches of Ni-200 used for the core barrels and vessel walls were annealed so that their stress-strain properties at a strain rate of 100 per second (the anticipated peak scale model strain rate), simulated the low strain rate stress-strain properties of the prototypic stainless steel at operating temperature. This approach was taken because the test components experience strain rates 20 times those of the full scale components. Figure 3.1-2 then shows the stress-strain curves for the upper internals columns for tests SM-3 and SM-5 where they can be compared to the stress-strain curve for the prototypic 316 stainless steel columns. Unlike all other components, the SM-3, SM-4 and SM-5 UIS columns were inadvertently annealed under varying conditions. The conditions for the SM-3 and SM-5 columns are defined beside the curves in Figure 3.1-2. The columns for SM-4 were unannealed. This produced columns which were too strong and they responded elastically during testing. The stress-strain properties for the SM-4 columns were therefore not detennined. Low and high strain rate tests were performed on 533B carbon steel, the material used to fabricate the heads of the CRBRP models. Figure 3.1-3 shows the low strain rate (c = 0.5 sec-l) stress-strain curves for 5338 carbon steel. Two sets of stress-strain measurements of the 533B carbon steel were made as shown on the figure. The differences between these sets of measurements probably result from the fact that the specimens were removed from different parts of the plate. Figure 3.1-4 demonstrates the strain rate insensitivity of 533B carbon steel up to a strain rate of 100 sec -I . At higher strain rates (greater than 100 sec-l) 5338 steel is strain-rate sensitive, but at strain rates less than 50 sec -l , 5338 is a reasonable simulant for 508 Class 2 carbon steel, the prototypic head material. 16
i ! = l X' 50 L l 40 -
, , - l f . ::
- s#
/
lis
- w:
5 s 'a -
#'W. ,/
/-
g 30 - p 7, - s ,, ' p ' ', i f - "
,/
vs g i
@ / p m /f',, -
z 20 - ' 5 ~'
~
C 5 .* 304 STAINLESSSTEEL (LOWER TEMPERATURE OF VESSEL WALL-750 F) 8
/
r' - - - 304 STAINLESS STEEL (UPPER TEMPERATURE OF VESSEL WALL-840 F) 8 10 .-- ----- Ni-200 (R00M TEMPERATURE AT STRAIN RATE OF 100 lN/lN/SEC) N -200 (ROOM TEMPERATURE AT STRAIN RATE OF 0.001 IN/IN/SEC) I I ' I I 0 O 1 2 3 4 5 6 AXl AL STR AIN (PERCENT) i I'igure 3.1-1. Stress-Strain Properties of Ni-200 and Stainless Steel I As for the Vessel Wall 1
=
X' 50 L ts 40 - ' s /, g f ,. h n 5
; 30 - ,x-
" ~'
z ,-
' ~',
/
a /:: - , /- m E l# -
/
o 20 - / ' 5 / 's '
/
" (f/ I s l 304 STAINLESS STEEL O_0t".'ER TEMPERATURE OF CORE BARREL-750 F)
- /
334 STAINLESS STEEL '. UPPER TEMPER ATURE OF CORE BARREL-1015oF) ) 8 10 - - - - - Ni-200 (TWO TESTS AT ROOM TEMPERATURE AT STRAIN RATE OF 100 IN/IN!SEC) 3 I - Ni-200(R00M TEMPERATURE AT STRAIN RATE OF 0.001 IN/IN/SEC) i I l f I I I I I 0 O 1 2 3 4 5 6 AXIAL STRAIN (PERCENT) Figure 3.1-1. Stress-Strain Properties of Ni-200 and Stainless Steel till for the Core llarrel O --- O - - - - -- - O
T 3 50 ,_ g - L ' a' /
,/
40 - - 3 ,-
,/ '
Y
,/
en /
/ ,
cc 30 - f
/ ,/
/ <
z E / I III
- z i -
3 z 20 -I o ua 8 304 STAINLESS STEEL (TEMPERATURE OF CORE SUPPORT PLATE -7500F) e I
---- Ni-200(R00M TEMPERATURE AT STRAIN RATE OF 100lN/IN/SEC)
-- Ni-200 (ROOM TEMPERATURE AT STRAIN RATE OF 0.001 IN/IN/SEC) 10 --
i I I I I I 0 1 2 3 4 5 6 O 4 AXI AL STRAIN (PERCENT) i Figure 3.1-1. Stress-Strain Properties of Ni 200 and 304 Stainless Steel <C i for the Core Support Plate. j i
0 100 l l l l l l l l ! 90 - _ UlS COLUMNS FOR SM 5 80 - (At Room Temperature)\ _ l l 70 - ,,' _.
/
g 60 -
,/ _
p' k' UlS / COLUMNS FOR SM 3 h 50 G e' (At koom Ternpnrature)
$ s' l ? 40 -
p _ i t j s' 316 Stainlens Steel (At 995 F) 30 ' - -
/
20 I- .- SM-3 columns were annealed at 1300"F for 60 min. l SM-4 columns were not annealed. The stress-j 30 - strain curve was not measured but would be ._ l above the SM-5 curve. SM-5 columns were annealed at 1300 F for 30 min. I I I I I I I l l o 0 0 02 0.04 0.06 0.08 0.10 0.12 0.14 0,16 0.18 0.20 TRUE STRAIN M A-39 29 -269 Figure 3.1-2 STRESS STRAIN PROPERTIES OF THE SM-3 AND SM-5 UIS COLUMNS l l 20 9ll I i
O O O
'2 1 I t l l 1 100 -
b
*m-
~
I E e
"' so - / _
8 a N y
@ SA 533n mnterini as rostr ' b- Terret Tek (70*F) h 49 _. h SA 533B material as tested by TJ-ARD (70*F) -
m
@ Nomirtal properties for SA 508 Class 2 (400 F'i 20 -
4 1 I I i i 1 o 0 1 2 3 4 5 6 7 AXtAL STRAIN - percent M A-3929-251 Figure 3.1-3 STRESS-STRAIN CURVES FOR C33 B CLASS 1 CARBON STEEL , 1
O
'dU i l I I i 130 -
A e - 2v. - 3 ot = 4% 120 - v> 0 110 go - H 100 - @# A - 2 # 0 f
@ 90 -
c.O k0 , N U 80 - # AA )
- i b 'a 70 -
O l l I I I Y 10-3 10-2 10 1 1 to 102 103 STFI AIN FI ATE, i hn/in/sec) MA 3929-162 l Figure 3.1-4 STRAIN RATE EFFECTS IN 533 8 CLASS 1 CAHBON STEEL 9 22
3.2 Phase 1 - Energy Source Calibration _ O To simulate the energy released by the expanding fuel vapor, a low density explosive source was used. This source consisted of a 90:10 mixture by weight of PETN* powder and Microspheres (hollow plastic spheres) contained in a cannister consisting of stacked rings, separated by gaps, and bolted between two end plates. The explosive gas is discharged through the gaps into the surrounding air volume. The gaps and air volume cause attenuation of nonprototypic shock waves. To match the experimental charge characteristics to those of the scaled SMBDB energy source, a thick walled vessel and core barrel scale model of CRBRP, shown in Figure 3.2-1, was used. The explosive source was placed inside the core barrel, sealed off above with a mylar diaphagrm, and the vessel filled with water. The charge was then detonated and the resulting change in gas volume AV(t) and core barrel pressure P(t) were measured as functions of time. Two piezoelectric transducers P) and P2 measured the pressure change while the wooden float resting on the surface of the water measured the volume change associated with water displacement. This displacement together with a knowledge of water compression and vessel elastic expansion were used to determine the actual change in bubble volume AV(t). The effects of water compression and vessel expansion were accounted for using an analytic model. The time variable t was then eliminated to give core pressure as a function of the gas volume i.e. P = P( AV). Twenty calibration tests were run to match the explosive source P(AV) to the scaled SMBDB P(AV). Calibration was performed by varying the mass of the charge and the initial gas volume in the core. The final four tests, which had identical specifications to assure reproducibility, had characteristics close to those of the scaled SMBDB source. These four tests had 19.7 gms of charge and an initial gas volume of 962 cm3. The peak pressures varied from 260 to 265 bars while the gas work released, at a time equivalent to that of slug impact with the vessel head, varied from 12.8 to 14.0 KJ. The
- Pentaerythritol Tetranitrate (C H 058124) N 23
peak pressure is thus slightly lower than that predicted for the SMBDB case (273 bars) while the energy released at slug impact is slightly higher than that for the scaled SMBDB case (12.6 KJ). Figure 3.2-2 shows~the resulting experi-mental curve in comparison with the scaled SMBDB curve. Also shown is a comparison of the experimental and scaled SMBDB gas work curves. l l l l 1 l l 24
I LIGHT SOURCE !
/ - LIGHT PICKUP TUBE \
PHOTO DIODE \ SENSOR
/
/ -
LIGHT
/
LADDER l
/
~
/ /. WOODEN FLOAT /
WATER SURF ACE f
,/
f / f v
/ /
,' ! J
[ l
/
l / 7 l
/ / l
/ // 12.86 in.
" PRESSURE 3.0 in. -
,- 12.15 in. -
~
l G AGE LOCATION ] MYLAR / / - 4 plam) DIAPHRAGM \ / p CHARGE CANISTER
/N ~ 2 in.
~ /
AIR / ~ 3.72 in,* [. CHARGE
/ (PETN-Microspheres p' \ /
~
/- 90/10 by weight) i
( 4
/ r4 pq,e/ "' ' / ORE BARREL P,
/ P2 ---- vs
[-
'/
}
[, JJ / ,1 y
'/. l CORE ,,
'/ { ~T~ T / 'h INST
"./g[ RUMENT ATION FEEDTHROUGH SU PPOR T -
PLATE {f p-f A
' t. /
1
' Q .y a. TV x g x '
l'V\\ 7
.4 il' N y' lxh x l
; l .
.(
'+ s
/K
' I 2, , r
;~ g N 'N, ' g ~ h\\\ \'N g \
~
~
s K g'g g g 'N \ N M A-39 29-144 Figure 3.2-1 ENERGY SOURCE CAllBRATION APPARATUS v 25
i l l l O l I I I I I l l l l l I I l 3(N) } } [ ] l l l
*- PE A KS (2f32 tiars) -
25 2 '.,0 - Average and Spread of Four Cahtuahon E xperiments 19 79 90!10 PET N-M riosphere Mix 3 20 902 cm Initial Volum" - 200 [- {
^ E l s a
l .- I t20-Scale SMBDB ; i 3 ' l
! COVE H G AS VOLUME 42560 cm ) "',,- ' ~ r- 15g )
i ,s y 150 }-. C ~
\ f ,o# ~ g l Y
- m w \ ,
4 1
\ 's ~ O g
100 -\ '
,#, j l
\ ,/
\ ,s N /
~
50 /
/
/
/
/ ~~~~~~~~~~~______~~~~ ~~ .
l 1 f ' ' 0 I I I 1 i l 1 1 l I I O 3000 4000 0 1000 2000 VO L UMt. CH ANGE t m 'I Fit;ure 3.2-2 PRESSURL VOLUME CHANGE AND GAS WORK VOLUME CHANGE CURVES FOR ENERGY SOURCE 9 26
+
_ __ --y-m.w y
3.3 Phase 2 - The Scale Model Tests The five 1/20th-scale models, SM-1 through Sfi-5, are described below along with the experimental results. No primary heat transport loops were modelled in any of the tests. This approach was taken since loads outside the reactor vessel have been shown both experimentally and analytically to be rapidly attenuated through plastic yielding of thin walled pipes such as are used in CRBRP (References 3-2, 3-3 and 3-4). 3.3.1 The SM-1 Test In the SM-1 test a scale model of the three plug vessel head as shown in Figure 3.3-1 was hydrostatically loaded to the point of structural instability. By loading the head quasistatically much information on head deformation shapes can be obtained that could not be obtained reliably in the dynamic mode. 3.3.1.1 Objectives of the SM-1 Test The SM-1 test was run to satisfy the following four objectives:
- 1) Obtain an understanding of margin ring juncture and plug deformations.
Analytic simulations of the combined effects of local plastic response at the plug junctures and the large displacements associated with plug bending and doming are very difficult. Thus SM-1 was run to obtain direct information both on local deformations at the margin ring junctures and on plug bending. The deformation profile resulting from this static test will be similar to that experienced in the subsequent dynamic tests. This is true for two reasons. First, upon slug impact, the vessel head will respond initially in its fundamental mode whose shape is very similar to that of the static test. Second, the dynamic force-deflection characteristics at the margin ring juncture can be approximated quite accurately by the static characteristics for typical impact velocities experienced by the plugs. This results because typical margin ring impact velocities in the dynamic tests are about one percent of the speed of sound in 27
the structure. Thus insight into the dynamic head response will be gained through the static test. O
- 2) Assist in defining failure mode.
While neither analysis nor experiment predict head response close to the failure point under the SMBDB loads, an understanding of the ! mechanisms producing such failure are of interest as they orovide insight into structural margin at lower loads. Since analytic prediction of the failure mode is uncertain, assessment through testing is appropriate. Static rather than dynamic testing is ) possible since as discussed in 1) above, the structural response l characteristics of the two cases will be very similar. 1
- 3) Obtain an estimate of limiting pressure. l A knowledge of the limiting static pressure is of use as it will provide an order of magnitude estimate of the load carrying capacity of the head.
l
- 4) Obtain a quantitative estimate of the head force / deflection char.pc teris tics .
Knowledge of the head force / deflection characteristics allows a determination of the effective elastic limit of the head. These characteristics also are used in modelling the head stiffness in the SM-4/SM-5 REXC0-HEP simulation. They are also used, in conjunction with SM-5 head deflection data, to assess the degree of defonnation in the SM-5 head plugs (see Section 3.3.3) and to determine how close the SM-5 head comes to its effective elastic limit. 3.3.1.2 The Test Model and Instrumentation 1 The SM-1 model, shown in Figure 3.3-1 consists of the three head plugs, margin rings and vessel flange. Recesses were drilled into the intennediate rotating plug (IRP) to simulate the holes required for control rod drives and l upper internals structure (UIS) columns, and into the small rotating plug (SRP) for the in-vessel transfer machine (IVTM) column. Beneath the head assembly, rubber and aluminum discs prevented leakage of the pressurizing fluid (water) from between the head plugs. i 28
Figure 3.3-2 shows a schematic of the apparatus used to pressurize the test model and measure the under-head pressure, while Figure 3.3-3 shows the instrumentation layout. The above-head instrumentation was composed of linear potentiometers (linipots) to measure displacement and were mounted primarily along the axis of symetry. These allowed a determination of both absolute vertical displacement and relative displacement across the margin rings. Other linipots were added off the symmetric axis to check symmetry of response (linipots 13 through 16) and to measure hold down bolt extension (linipot 17). The whole apparatus was loaded continuously at 150 psi / min and continuous linipot displacement readings were taken. 3.3.1.3 The SM-1 Test Results Upon loading the underside of the head as described above, the deformation profiles shown in Figure 3.3-4 were obtained. The length of the vertical tick marks at each linipot location represents the bounds on displacement measurement based upon a 95 percent measurement accuracy. Tha corresponding volume change under the head remained linear with pressure up to about 700 psi after which significant plasticity became evident. This can be seen in Figure 3.3-5 which also shows that the peak pressure reached was 1160 psi. Once in the plastic regime, the most significant deformation took place in the bo'dy of the large plug. Neither the IRP, the SRP nor the margin ring junctures suffered significant distortion. In the photograph of the final head deformation profile, shown in Figure 3.3-6, perman at plastic strain in the LRP is quite apparent. Beyond the elastic limit, the l margin ring at the LRP-IRP juncture began slipping laterally away from the adjacent plug lip. Final ' failure' of the head thus occurred above 1160 psi through margin ring disengagement between linipots 9 and 10. If the under-head shielding had been accounted for in the SM-1 model it is likely that an even higher pressure would have been reached before failure and the jamming of the shield plates would have reduced head doming. For detailed information on the response of each linipot, Table 3.3-1 should be examined, As reflected in the table, no readings were obtained from linipot 13. O - 29
3.3.2 The SM-2 and SM-3 Tests The models on which the SM-2 and SM-3 tests were based are identical except that SM-3 incorporated a UIS. Because these two tests are very similar, comparisons between them are appropriate and they are discussed together. Figures 3.3-7 and 3.3-8 respectively show ae layouts of these models. 3.3.2.1 Objectives of the SM-2 and SM-3 Tests The objectives of the SM-2 test were: J
- 1) To verify that a simplified model of CRBRP can accommodate the SMBDB loads.
SM-2 has very similar features to the analytic REXC0-HEP model used to generate the SMBDB loads. Its response can be used to experimentally verify the adequacy of this idealized model by con.peing it to the response of the more complex models SM-4 and SM-5. In addition, by , simulating this scale model directly with REXCO-HEP (see Section 4.1.1), f the ability of this computer code to model such cunfigurations is assessed. 1 I
- 2) To experimentally assess the effect of the upper internals structure.
Because of the difficulty in analytically modelling the flow restriction and redirection caused by the UIS, an experimental assessment of the effect of this component is important. To see this effect it is first l necessary to understand the fluid and structural response of a ! system in which the UIS is absent. l The objective of the SM-3 test was:
- 1) To assess in conjunctior with SM-2, the ef fect of the upper internals structure on the reactor structural sys tem response. The upper internals structure was expected to significantly change the energy partitioning within the vessel. This was determined by comparing l SM-2 and SM-3 results. l l
l 1 30 l l l l
3.3.2.2 The SM-2 and SM-3 Models and Instrumentation Figures 3.3-7 and 3.3-8 respectively show the layouts of the SM-2 and SM-3 models. Detailed information on model dimensions can be obtained from Appendix B. In both cases, the vessel heads were modelled as solid discs whose thickness simulated that of CRBRP. Weights on the top of each head simulated the masses of the head mounted components, while stacked discs under the head simulated the under-head shielding. The margin ring and the surrounding structure was accurately simulated at the juncture of the plug and vessel flange. The vessel walls had unifonn thickness equivalent to that of the vessel below the elevation of the overflow and make-up nozzies. In each model the core support structure and bottom head were replaced by a rigid plate whose mass was the scaled equivalent to that of the lower structures and fluid. Inside the core barrel, the core surrounding structures were simulated by a set of stacked rings, each one being cut radially to remove hoop strength. The rings were also lubricated to facilitate differential radial movement. Between the rings and the core barrel a thin aluminum cylinder was inserted to prevent O detonation products from directly loading the core barrel. The charge cannister was mounted on a tripod structure rather than being bolted on to the centerline of the core plate so that no direct mechanical load would be transmitted to the inner part of the plate. This procedure is not significant in SM-2 and SM-3 but was desirable in SM-4 and SM-5 where the response of the core support plate was to be assessed. Lead shot was placed on top of the itylar diaphragm to simulate the upper pin structures that might get ejected from the core during the initial bubble expansion phase. The total mass of these core internal structures was such that it simulated the actual in-core mass. For SM-3, the UIS was modelled from a solid block of aluminum with vertical holes drilled to represent the chimneys. The exterior profile and total mass were kept consistent with the prototype.. The columns were modelled from Ni-200 tubes with their inside and outside diameters appropriately scaled. O 31
The instrumentation for SM-2 and SM-3 was identical except that SM-2 had an additional strain gauge SG10 mounted on the vessel wall close to the point of peak circumferential strain. In addition to pressure trans-ducers, strain gauges and accelerometers, two water surface gauges (WS) and WS 2
) were instailed in the head. These allowed for a determination of water surface displacement during slug movement. In conjunction with the core mounted pressure transducers, these pennitted a check on core energy release.
3.3.2.3 The SM-2 and SM-3 Test Results Figures 3.3-9, 3.3-10 and 3.3-11 compare selected pressure, acceleration and strain information respectively for SM-2 and SM-3. It can be seen that the in-core pressure responses (gauges P), P 2 and P3.) are quite similar be-tween tests SM-2 and SM-3. The in-core pressure in SM-3 is slightly more sustained as a result of the UIS. These results are consistent with the good agreement seen between the SM-2 and SM-3 core barrel strain profiles (see Figure 3.3-12). Thus the response of the core barrel appears to be insensitive to the addition of the UIS. Outside the core barrel the addition of the UIS does have a significant effect in reducing loadings. At the outlet nozzle elevation, the SM-3 pressure is reduced to about 54 percent of the SM-2 value (see gauge P5 on both SM-2 and SM-3). The peak SM-3 under-head loading (gauge P8) is reduced to about 66 percent of the SM-2 value. This is very consistent with the ratio of the slug impact velocities shown in Figure 3.3-13 where the reduction is to about 68 percent. These reductions are also reflected in slug kinetic energy at impact (see Table 3.3-2). I l The head accelerations for SM-2 and SM-3, which are shown in Figure 3.3-10 are difficult to compare because the exact plug movement gap closure times are highly sensitive to loading and design differences. However, it can be seen that, before slug impact, the SM-3 results show more response because of the upward movement of the UIS. After slug impact, peak response is comparable in the two cases. The SM-3 core plate acceleration (gauge A4 ) is initially greater (before 1.0 msec) as a result of the UIS. However, upon slug impact (at about 3.0 msec) SM-2 shows greater core plate acceleration l as a result of the more severe slug impact with the head. 32 O'l
The SM-2 and SM-3 vessel residual strain profiles can be compared in Figure 3.3-12, while details on the transient responses can be compared in Figure 3.3-11. The latter figure shows the upper vessel wall strain responses (SG5 on both SM-2 and SM-3). The SM-2 response occurs at a strain rate of about 93 per second, which is close to the anticipated value of 100 per second as discussed in Section 3.1. As a general conclusion, the effect of the UIS is to reduce the 1 deformations on the vessel wall and hold down bolts. No residual deformation was found in the vessel hold-down bolts of either SM-2 or SM-3. The final deformation shapes of the SM-3 columns are shown in l Figures 3.3-14 and 3.3-15. The hollow sections of the columns in the test compressed vertically by 0.55 inches (about 7 per cent). Since l I the engagement length of the UIS keys was 0.37 inches, the keys disengage by 0.18 inches vertically. This is a very short distance l i relative to the distance the UIS would have to move (2 inches) to challenge coolant boundary integrity. l From these two tests it is evident that the UIS has a major attenuating effect on the coolant slug motion. The coherence in upward slug movement ' was lost to a considerable exter,t in SM-3 with the result that energy was much less efficiently transfurred to the head. This can be seen in Table 3.3-2 where the estimated kinetic energy in entrained and locally acc91erated water increased from 4 percent of the total released gas work in SM-2 to 48 percent in'SM-3. This is consistent with the fact that the final strain energy absorbed in the structures (as a percent of total energy) reduces from 47.3 per cent in SM-2 to 31.5 per cent in SM-3. Table 3.3-2 also shows that the energy directly absorbed by the columns is small and is unlikely to have a major impact on the overall energy distribution. A summary of the SM-2 and SM-3 results is given in Table 3.3-4. O 33
3.3.3 The SM-4 and SM-5 Tests The SM-4 and SM-5 tests were the most prototypic of CRBRP in the sense that they included more detail than any of the previous tests. Their response would thus be closest to that of CRBRP were it subjected to the SMBDB loads. The only significant difference between the tests was the degree of instrumentation; SM-5 was much more heavily instrumented than SM-4. This additional instrumentation has minimal effect on the model's response. 3.3.3.1 Objectives of the 532 and SM-5 Tests The objectives of the SM-4 and SM-5 tests were:
- 1) To determine ability of models, which contain all the necessary features affecting reactor systems response, to withstand the simulated 661MJ HCDA. While these models do not have vessel nozzles, it has been shown (Reference 3-5) that such local geometry changes do not affect the local or system loads in any signi ficant way.
- 2) To verify reproducibility of response by comparing experimental O
information from SM-4 and SM-5.
- 3) To obtain a broader range of experimental results through the SM-5.
SM-5 was most heavily instrumented for two reasons. First, since it was the last test in the series, experience had been gained in setting up and instrumenting the earlier models. Second, being one of the two most detailed models, this test was capable of giving the greatest amount of detailed information. In particular SM-5 was instrumented to obtain the pressure response of the cover gas. 34 9
3.3.3.2 The SM-4 and SM-5 Models and Instrumentation Figures 3.3-16 and 3.3-17 show the SM-4 and SM-5 models respectively with their instrumentation. While these models are identical, SM-5 has over twice the instrumentation of SM-4. In addition to the SM-2 and SM-3 features, each of these models has a three-rotating-plug head, a more prototypic core support structure, a vessel hottom head, a thermal i I liner, a horizontal baffle, an in-vessel transfer machine .(IVTM) column and a vessel wall whose thickness varies in a more prototypic manner. The three rotating plug head is perhaps the most important additional feature since it provides direct experimental confirmation of the ability of the three-plug head to withstand the scaled SMBDB loads. The under-head shielding was modelled accurately enough so that the effect of shielding flexibility on the load transmitted to the head was taken into account. The core support plate thickness was accurately modelled and blind holes were drilled to j account for the inlet module penetrations. The variation in core barrel wall ! thickness was also accurately represented. A vessel bottom head was added since this structure and the liquid above it will affect the dynamic response of the core support plate. The core plate response in turn can affect the kinetic energy imparted to the liquid slug. The thermal liner is important for prototypic response since it has the effect of shielding the vessel wall from the direct pressure loads in the upper plenum. It will also affect the load felt on the head. The IVTM column was included since it could potentially produce an asymmetric response of the UIS. Finally the vessel wall thickness was made more prototypic, being 119 mils at elevations below the overflow and make-up nozzles and 140 mils at higher elevations. The thickness of the vessel flange transition section was also modelled more prototypically than in SM-2 and SM-3. Figure 3.3-18 shows the sealing arrangement used in SM-4 and SM-5 to prevent fluid escaping from between the head plugs. Similar, though less complex, arrangements were used in SM-2 and SM-3. A gas space above the margin ring, O 35
whicn had a volume equivalent to that of the large riser annulus, was incorporated so that an estimate of the pressure in this region could be made in the SM-5 test. The 0-ring seal adjacent to the large margin ring was added for two reasons. First, it guaranteed that the ring would be in its correct position immediately prior to slug impact. Second, it allowed for a specified constant gap connecting the regions below and above the margin ring. The 0-ring acts to block direct communication across the circumferential gap 'as shown in the I diagram opposite. However, it does ,. radial gap not block communication through the *fn f . r es radial gaps between the ends of the 4 v .4i
,UMc r nf t margin rings. These latter gaps //
have a constant (and known) orifice l fJ margin ring gap l area throughout the transient. This lj . k
-margin ring gap allows a well defined geometry to be margin /
used in the MAXPRES calculation # "9 ! simulating the cover gas response
----]
(see Section 4.3). The cross-sectional area of the radial orifice gaps for gas communication is equivalent to about 10 percent of the fully I open margin ring circumferential gap. l The amount of instrumentation on SM-4 was slightly less than on SM-2 and l SM-3. There were two pressure transducers (P), 2P ) in the core, three on the vessel wall (P3 , P4 and P5 with a penetration through the thermal liner for P 4), and one on the head (P ). Seven strain gauges were mounted on 6 the vessel wall (SG) through SG6 nd SG 10) and three on the vessel hold-down bol ts (SG , SG 7 8 nd SGg ). Three accelerometers (A), A2 and A3 ) were mounted on the head, one on each plug. These are shown 6n Figure 3.3-15. Two water surface gauges (WS) and WS 2 ) were mounted under the head shielding to measure the upward movement of the water slug. Thus SM-4 had a total of 21 instrument channels. These are listed in Table 3.3-4 where they can be compared to the instrumentation of the other tests. In SM-5, two pressure gauges were mounted in the core (P),2P ), five along the vessel wall (P5 through gP ), four in the vicinity of the head (P through P13), three in 10 the gas gap between the large rotating plug and the vessel flange (P)4 through 36
we're used P16) and two around the UIS (P3 and P4 ). Gauges P)4 through P 16 to assess gas pressurization up the riser annulus. Although the precise annular shape was not modelled, the equivalent gas volume was modelled so that gas pressurization in this region will be correctly characterized. Gauges P3 and P 4were mounted at the bottom and top of the UIS chimneys respectively i thus allowing a determination to be made of the pressure differential l across the UIS. Strain gauges were mounted on the core barrel (SG), SG 2 ), the vessel wall (SG3 through SG8and SG21) the LRP (SGg and SGl0),theUIS columns (SG j ) through SG13)theUIScaps(SG)4 through SG)7) and three of the holddown bolts (SG 18 through SG20). The UIS caps are located at the position of the jacking mechanisms but are not intended to be prototypic of these components, i Most of the accelerometers wert eted on the head plugs (A 2 through A8 )' two were located on the vessel flange (A) and Ag) and one, (A10)which failed to function,was placed on the core support plate. 4 1 A linear potentiometer (LP1) to measure vertical displacement was placed at the point of expected highest rise of the IRP, Three water surface gauges (WS), WS2 and WS 3
)'were included. Thus, in total, SM-5 had 51 channels of instrumentation. These are listed in Table 3.3-4 where they can be compared to the instrumentation of the other tests. l 3.3.3.3 The SM-4 and SM-5 Experimental Results Figures 3.3-19 through 3.3-27 present the pressures,. strains, accelerations, I
deformation profiles and water surface displacements from SM-4 and SM-5. From these results it can be seen that excellent reproducibility of response was obtained between the two tests.
\
l 37
First 'onsider c the pressure response. The core pressure readings for SM-4 and SM-5 are in very close agreement (see gauges P) and P 2 in Figure 3.' 3- 19 ) . In addition, Figure 3.3-27 shows equally good agreement between the measurements of SM-4 and SM-5 water surface displacements. These results indicate that the source pressure-volume characteristics are behaving in a quite reproducible manner. The slight variations in the vessel wall pressures below and above the horizontal baffle (gauges P and P in SM-4 3 4 and P6 and P7 in SM-5 respectively) are probably due to that fact that, inadvertently, the SM-4 UIS columns were not annealed. Thus the SM-4 columns did not bend as was observed in SM-5, and the area for radial flow between the core barrel and VIS was less. As a result slightly lower pressures are seen at the SM-4 mid-vessel eleva tions. This effect is a local one as can be observed in the comparison of upper vessel loadings (P in SM-4 5 and Pg inSM-5). The comparison of head loadings (P in SM-4 and P g in 6 SM-5) is not quite as good but this is because P)) is not centrally located on the head as is P6 (the central gauge, P10, broke at about 3.1 msec). Table 3.2 4 shows that the SM-5 peak head pressures are higher than the other tests. However, other indicators such as slug impact velocity suggest that these readings are anomalously high. In comparing the SM-4 and SM-5 pressures to those of SM-2 and SM-3, it can be seen that the core pressures are quite similar. This is also true with the vessel wall loads at the mid-core elevation. However, at the outlet nozzle elevation, greater vnriations msulting from geometry changes can be observed. While the SM-3 pressure at the outlet nozzle was significantly reduced in comparison to the SM-2 value as a result of the UIS, the SM-4 and SM-5 values increase again approximately to the level of the SM-2 pressure. This results from the reinforcing effect of thermal liner. However, as discussed later, this reinforcing also results in a major reduction in vessel wall hoop strain. Comparison of the central head loads of SM-2 through SM-5 shows that as a result of increasing the geometric complexity and associated fluid turbulence the secondary pressure spike is rapidly suppressed. 38
figure 3.3-20 shows the cover-gas pressure transients f rom the SM4 test. The under-head pressure (P13) reaches a peak of only 30 psi. This pressure may be somewhat low since in the model more limited communication can take O place between the main body of the cover gas and the gas trapped between the head and the upper shield plate. Figure 3.3-20 also shows the pressure response in the flange dip seal region (P)4) and in the cavity representing the annular volume in the large riser (P 15 nd P16). The pressure in the dip seal region is considerably lower and more smoothed out than the loading under the head shielding, although slug impact is reflected by the small peaks at 3.0 msec. The effect of the narrow annulus and margin ring constriction is to further reduce the pressure seen in gauges P15 and P16. The high frequency overtones seen in these pressure profiles reflect structural ringing in a low signal-to-noise environment and not true pressure response. However, it can still be concluded that pressures above the margin ring are in the range of 10 to 30 psig. With the additional restrictions of a prototypic dip seal configuration and dip seal sodium, it is very likely that the pressures above the margin ring would be even lower than found from the SM-5 test. Figure 3.3-20 also shows the pressure at the bottom and top of the UIS ( chimneys (P 3 and P 4 respectively). The cable to the upper gauge was severed just after 1 msec and pressures beyond this tine are thus uncertain. However, the response beyond 1 msec is likely to be between the responses of gauges P7 and P8 , also shown in Fiqure 3.3-20. Finally, gauge P5 shows the pressure response in the inlet plenum as a result of core support plate movement. The peak pressure (480 psi) represents about 1.5 percent reduction of inlet plenum volume. Figure 3.3-21 compares the strain records of SM-4 and SM-5. While the individual vessel wall records show local differences in the final vessel wall deformation, profiles are very consistent. Notice, in particular, that the hoop strain at the outlet nozzle elevation is very low especially for SM-5 where it is almost elastic. The strain gauges on the hold down bolts (SG 7, SGg and SGg in SM-4 and SG18, SG)g, and SG20 in SM-5) indicate virtually elastic bolt response. O 39
,-,v, .-
~ , - , , , . . , . ,, -w-, , . - - - , . - - , -~
The upward ledge load resulting from the stretching (including the 0.2 percent preload strain) of the 72 bolts is 95,600 lbs for SM-4 and 84,300 lbs for SM-5, these being equivalent to 38.2 million and ?3.7 million 1bs respectively for CRBRP. These loads are well below the CRBRP ledge load limit of 50 million 1bs. Consider now the response of the upper internals columns. Since the SM-4 columns were neither instrumented nor did they undergo plastic deformation (as a result of their inadvertantly not being annealed), the loads to these columns could not easily be determt u. However. from the SM-5 columns significant data were obtained. Figure 3.3-22 presents the strains in the four SM-5 columns. SG 13 shows the strain in one of the hollow columns while SG j4 through SG)7 show the strains in the upper solid sections of the columns. SG)4 on column 1 responded significantly more than gauges 15 through 17 on the other columns. This may be due to the fact that column 1 experienced greater initial column compression (see Figure 3.3-23). Thus, on slug impact with the head, column 1 was preferentially pulled upward. SG 13 shows that the hollow nickel part of column 1 compressed by 2.1 percent. Using Figure 3.1-2, it can be determined that the associated stress is 48 ksi. This load is very consistent with that of SM-3 where, as seen in Figure 3.3-15, the residual strain is 7 percent. Using Figure 3.1-2, 7 percent strain results in a column load of 47 ksi. It is thus likely that the upward compressive load on the columns.is in the range of 4800 dnd 4900 lbs, per column. In CRBRP the corresponding upward liad on one 2 column would therefore be around 1.9 million 1bs (i.e. , 20 x4800lbs). Upon slug impact, the rapid upward movement of the head generates tensile loads in the columns. For column 1, the peak tensile strain is about 0.22 percent (SGj4). This results in a peak column load of 22,900 lbs for the model or 9.1 million lbs in the prototypic case. The other column loads are considerably less, the average being about half that on column 1. These loads may be non-prototypically high since the upper solid sections of the columns are more rigid than in the actual design. With reaard to the UIS columns, two main conclusions can be 40
drawn. First, the most severe load imposed on the columns (and the jacking mechanism) results from upward head doming and not from the core press loading. Second, the tests showed that non-axisymmetric effects may p.my a significant role in determining peak loads both on the UIS columns and the jacking mechanisms. While no detailed measurements are available on the SM-4 and SM-5 core support plates or horizontal baffles, post-test inspection was made. This revealed that some doming of the core plates occurred. The residual downward displacement of the core plate centerline relative to the : periphery was very close to 0.1 inch in both cases. Inspection of the horizontal baffles revealed virtually no residual deformation. Figure 3.3-25 compares the SM-4 and SM-5 head plug accelerations. Considering the complex nature of plug-to-plug interaction, these acceleration records show close correlation. This is consistent with the good agreement between the SM-4 and SM-5 slug impact velocities (Figure 3.3-27) and also with the fact that-the responses of all the margin ring junctures were vitually elastic. Figure 3.3-26 shows some additional acceleration records from the SM-5 head. Also shown is the response of the one linear potentiometer (LP1) used on the SM-5 head. This was placed at the point of anticipated maximum By head displacement and shows a peak displacement of 0.06 inches. examining the readings of linear potentiometer 9 from the SM-1 test (see Table 3.3-1), it can be seen that a displacement of 0.06 inches corresponds to an equivalent static under head pressure of about 540 psi. Using this pressure in Figure 3.3-5, it can be concluded that the head is still within its elastic range (plastic deformation starts around 700 psi). This result is consistent with the fact that no residual deformation in the head plugs was observed. 41
No nozzles were modelled on any of the SM series vessels. This approach was used since it has been shown that vessel wall pressure is insensitive to the existence af an adjacent nozzle (Reference 3-5). Thus, while direct experimental data are not available from the SM series of tests, some conclusions regarding the ability of nozzles to withstand the loadings can still be made. Consider the pressure loadings and resulting hoop strains on the SM-2 and SM-3 vessels at the outlet nozzle elevation. Table 3.3-4 shows that the SM-3 pressure loading is reduced to 70 per cent of the SM-2 value (gauges PS ) while the corresponding strain is reduced to 55 per cent (gauges SG2 )* These reductions are the direct result of adding the UIS. By examining the same infonnation for the SM-4 and SM-5 tests, it can be seen that the pressure at the outlet nozzle elevation increases to a level slightly above that of SM-2. This is due to increased rigidity resulting from the addition of the thermal liner. However, the liner also has the effect of strengthening the coolant boundary and this results in a reduction of vessel hoop strain in SM-4 to about 3.0 percent of the SM-2 value. The effect of such a reduction is to significantly enhance the ability of an outlet nozzle to survive the HCDA pressure load. By examining the vessel data at the elevation of the overflow and makeup nozzles, it can be seen that the thermal liner would have a similar strengthening effect. A still noticable but lesser reduction in strain is also observed at the cover gas nozzle elevation. Since the liner does not extend up to the elevation of the cover gas nozzle, the strain reduction is more likely to be due to the slightly thicker, more prototypic vessel cross-section used at this elevation in SM-4 and SM-5 (0.119 inches vs. 0.104 inches). Table 3.3-3 compares the energy partitioning for SM-4 and SM-5 and shows that the energy was similarly distributed in the two tests. The differences in the energies absorbed by the SM-4 and SM-5 UIS columns result from the annealing differences as indicated on Figure 3.1-2. The other differences are not considered major (e.g. , see Figure 3.3-24 for comparison of reactor vessel wall and core barrel deformation profiles). 42
1 1 1 1 1
]
V I 4
*3
@=
J asanassanaassassomosssammassamosas21 3..e32 2 233 33322 a24o22 I 8 h ,
-s - - ~ ~ e, N ~ N ~ N m m m m m m m . . . . . . o o o r e c o . . . . . - ~ ~ ~ ~ . 4 . e e . . - m ~ m e - > s em m=mmmmmmmm v I
cTMNOk omeo..*N=-4& N, n,
- s
- - - - - - - - ve , NE9 S-, #.w~vvw s ? *hN1 m @unmm*,mm @ ?.
m . . M. .
. f.
.N.
@,#oonnoeoo--~rc,remsm e3-+
@m m--m--- vv 1
m,% s 3 m m - ~ , ~ m m m e e , , , o n e e o o - ~ ~ ~ o m e > , . . mmmm-m-m-m . , m - ~ N m m m _, m-----sssssn e o - > . v , o , ~, e, m, ,, ) I l
- \
l m#
-J N- ]
)
~ o,=---m---evsyssN, ,, ,, , ,,,,,,,,,,,,n.. ooeo-~~ro+.. -m. e--
@m I i
=
- 1 1
M a 1 mm a 3 3-m- v s , v n m m m e e , e n .n o c a c ~ ~ ~ ~ m s a m a---m----m-mm--m-mNNv% t a.3 a - m S N n n . , n e a s t r 3 N ,o svmmm N e r m, W b* H $3 m m s 3%, oS 2-me~2mmnes.w.c-s m - - - m m N N N ~ N a m m m m . ~, m, c, ,3~ ~, , c c, oor-m m e t . e 4m e vet-
~~~ .c r N. .o . .--~meo~
o u t e o s m > . ~, #e.u ~, W C- m--=-m-m-NvN99 b r- U
- m ,V ,~ ,7n .-, s, #, k,~~~~DN..,.h.
~@ m,4 a .N@~ @.*, h E4 om,4 ,
e H .J W m-_m-~N~N~m mmm,F,( 4,,@
, ne .C .m .C .9 # G e w & 4 ( 4 @. - m ~ N m ..n . . N. ~4 .N &. ~. m. i H *a - --*e---- g vw wv1 , j M.
mH
< 1
- p. .
m c7 e s s o -- -w.- -, mnma-N,e* j I
> 0 ram wr4t sm
- - N N s % N N a m m e m m m m,, , 4, 4,core c y #44 m e s s N .e r u m N s xessee.
4 4-~~~ - , - - 4 c .c N szuwe-<em
--Nm,o4.m+
4 r- M ms- m ---mmmmNNw , V no o i M% \
- p. ~<
,+~e-~ ,~~
i U @ J Ne~>-m
- e m - N e%,<1e~ N % % % ,cs m m m m,m.m..,e..>.mm m DJ'neppD&D&@k k~k~ ,ee-me~ ,mmsem 13.su ttDL@ .mNNmcom,eG%- e w e >-n ~, ~r m.e ein m as -- - m - - - e e .g ,% %
f e.e o i 5 . 2 U es- ~,mm~m,.-.-N,m....,..,N,,.,.,.,.,.,o.m~m,....-~,.~.mm.....~.,..,.nm.-
--wNy~~~~smmm,mmmm , , nnnnooooeaooco-~~-oeo>#=-~ , o~
---mm-4=
W +} ,es-12
-aNmm,sents.-Nmm,eenm.--Nen~se
- - - m - - - - m ~ .s v ~ ~ N N ~ N N m m m m m m m m , ,o , N, ,, ,+ ~, ,xoo . ,m ,s ,s ,-..--+ 1 ~ ee ,u ,s r e o ~ s, n os-I
+o-o.
mmm
" b -
N wa- -m-----m---NNNN~NN~NNmmmmmmmmmm ...b. . ceroco~~mer Nm m ,m vNO , $[Mhb MIdb OM~mm b MMb MM OM ~N, Mb Sk Mmnb M , bD g es
----mm---------NNNNNNNNNNNF"mmmmme<,#d# 4kK76 8 %
1 N , , @M vP4MMM.,@@@~M@ &$.--NM,@&~@ .N,4}N nJ
- ..M mmm---mmm-mmmmm--m~~NNNNN%NNmmmmm ,e.4mE
@peQ
~
A 4 $$ hS hommme , ,, ,$ h,hhhhh&hWh b h f h O ,$ = & esm e-m--- 3 W-r$
.d -
. ~. ,. ...~
. . .m---mNNNNNmMMMM,~,,
. ., .. . . .~
. . ., . .. .~ . .,.... . . . .,.,...~,..
. . . .~. . . . . ,....~,..=v....~
*AddA@edede-~~~>3Seese@@@@~ . . . . . . . . . . . . . . ., ... . ., . . ..~. .,. .
.3 m m-m
--mmm -m-a-
e
\
43
l TABLE 3.3-2 ENERGY PARTITIONING IN MODELS SM-2 AND SM-3
- Energy Energy Distribution (KJ) (%)
EM-2 SM-3 SM-2 SM-3 Gas Work at Slug Impact 14.40 13.90 100.0 100.0 Slug Kinetic Energy at Slug Impact 9.49 4.43 65.9 31.9 l Strain Energies at Impact **: Core Barrel 0.46 0.42 3.2 3.0 Vessel Wall (Radial) 3.86 1.60 26.8 11.5 Vessel Wall (Axial) 0.08 0.01 0.6 0.1 UIS Columns - 0.83 - 6.0 Core Support Plate 0.00 0.00 0.0 0.0 Total Strain Energy at Impact 4.40 2.86 30.6 20.6 Estimated Kinetic Energy in 0.51 6.61 3.5 47.6 Entrained and Locally Accelerated Water Total Gas Work After Impact 14.60 '14.20 100.0 100.0 Strain Energies After Impact *: Core Barrel 0.46 0.42 3.2 3.0 Vessel Wall (Radial) 6.19 3.18 42.4 22.4 Vessel Wall (Axial) 0.25 0.04 1.7 0.3 UIS Columns - 0.83 - 5.8 Core Support Plate 0.00 0.00 0.0 0.0 Total Strain Energy Af ter Impact 6.90 4.47 47.3 31.5
*SM-2 does not include a UIS; SM-3 includes a UIS.
- Strain energies computed on the basis of the high strain rate stress-strain curves 44
TABLE 3.3-3 ENERGY PARTITIONING IN MODELS SM-4 AND SM-5
- Energy Energy Distribution (KJ) (%)
SM-4 SM-5 SM-4 SM-5 Gas Work at Slug Impact 13.50 13.60 100.0 100.0 Slug Kinetic Energy at slug Impact 4.15 4.12 30.7 30.3 Strain Energies at Impact **: Core Barrel 0.24 0.37 1.8 2.7 Vessel Wall (Radial) 0.20 0.33 1.5 2.4 Vessel Wall (Axial) 0.01 0.01 0.1 0.1 Thermal Liner 0.02 0.02 0.1 0.1 UIS Columns 0.00 0.27 0.0 2.0 Core Support Plate 0.02 0.02 0.1 0.1 Total Strain Energy at Impact 0.49 1.02 3.6 7.5 Estimated Kinetic Energy in 8.86 8.46 65.6 62.2 Entrained and Locally Accelerated O Water Total Gas Work After Impact 13.80 13.90 100.0 100.0 Strain Energies After Impact *: Core Barrel 0.24 0.37 1.7 2.7 Vessel Wall (Radial) 0.83 1.19 6.0 8.6 Vessel Wall (Axial) 0.02 0.02 0.1 0.1 Thermal Liner 0.21 0.24 1.5 1.7 UlS Columns 0.00 0.27 0.0 1.9 Core Support Plate 0.02- 0.02 0.1 0.1 Total Strain Energy After Impact 1.32 2.11 9.6 15.2
*The SM-4 and SM-5 models were identical except for annenling differences in the UIS columns as indicated on Figure 3.1-2.
** Strain energies computed on the basis of the high strain rate stress-strain curves i
b) ( 45 l - .
Table 3 . 13 4 SUWW WY Ot DAT4 FRCN SM 2, SM 3 SM 4. AND SW 5 Gage No. Peaks SM 2 SM 3 SM 4 EM S Location SM 2 SM 3 SM 4 SM S P P P P Core 3300 pst 4600 pst 4200 pst 4000 pst 1 1 1 1 P P P P Core 3750 ;s1 4100 pst 3700 pet 4000 psi 2 2 2 2 P U P per core 2900 psi 2600 pet 3 '3 P Lower U!S 1620 pst P !!pper UIS 690 put P Lower vennel 480 put P P P P Vessel es11 at c re 540 pst 480 psi 575 pst 530 pst 4 3 6 P P, P P Vessel esti et U18 $70 pst 400 pst 610 pst 590 pst 3 7 P 6 6 8
"' "" E*I P' ' P' P P P P Upper vessel es!! 1650 pst 1000 psi 2700 pst 2850 pst 7 7 S g P
g P P 6 P g Head (center)' 5300 pst 3500 psi 4200 psi 7750 pst P Head (LRP)' 5350 pst l
'P Head (SRP) 8900 pst P Head air gap (center) 33 pst P Flange air gap 500 pst P Flange ring gap (180 ) 25 pst P Flange ring gap (0 ) 40 psi SG1 Core barret (C) 1. 7 5*.
SG2 Core barrel (A) --- SG1 SG3 Core support ring (C) -0.06% 0.04% 1 SG2 SG4 Core support ring (A) 0.06% 0.05% SGI SGI Vessel wall et core (C) 1.62% 0.87% SG2 SG2 SG3 SGS Vessel wall at UlS (C) 2.65% 1.45% 0.08% 0.32% SGJ SG3 SG4 SG6 Vessel well at U1S (A) 0.32% 0.53% 0.07% --- l SG4 SG4 Vessel well 2.30% 1.60% SGIO SGIO SG21 Vessel at peak ( (C) 4.53% ( ) 1.38% 1.95% SGS SC5 SG5 SG7 Upper vessel wall (C) 2.90% 2.15% 1.53% 1.87% SG6 SG6 SG6 6G8 Upper vessel wall (A) 1.03% 0.54% 0.38% 0.40% SG9 1.RP (6) ... SG10 LRP (R) --- SG11 U18 column (N1200) ~2.0% l l SCl2 018 column (NL200) l SG13 UIS column (NL200) -2.1% l l SG14 U1B column 1 (cap) -0,04% i
+0.14% l SGIS U1S column 2 (esp) -0.08%
+0.11%
SG16 U1S column 3 (cap) -0.04%
+0.04%
SG17 U18 column 4 (cap) -0.05%
+ 0. 05%
(Continued) O 46
_. - . . _ . _ _. ..-. ~ .~ ._ _-. - - _ - . _ - ~ __ -- .-
' Table S,14 Peams Gage No.
SM 2 SM 3 SM 4 SM S SM 2 SM 3 SM 4 SM S Locatton 0.16% 0.12% 0.13% 0 08% SG7 SG7 SG7 SG18 Stud 60 0.20% 0 , 08". 0.14% 0.08% SGS SG8 SG8 SGl9 Stud 180 0.12% 0.13% 0,09% SG9 SG9 SG9 SG20 Stud 300 0.36% 2100 g A Flange ring (0") 1200 g 1200 g 3400 g 3600 g A A A A, IRP at 0 g 2600 g A SRP near IRP (0 ) 4800 g 4600 g 3300 g 4500 g A A A A BRP near IRP (180 ) 3800 g A, IRP near center 1850 g 1300 g 3200 g 3500 g A A A 1RP near LRP (180 ) A) 4800 g A LRP near IRP (0 ) 3600 g A [JLP near edge (180 ) 3100 g A Flange ring (180 ) A A Platform 950 g 1230 g A 4 4 10 62 mits
!#1 Next to A IV IV IV IV 91.5 ft/sec 62.5 ft/sec 62.4 ft/sec 62.2 ft/sec Impact time 2.5 as 2.9 as 2.9 as 2.9 ma Peak value of spikes; recommend using impulse or slug impact velocity to calculate impact pressure, b
gage b?oke b G f 47
. +
Q i p -
~
/
d ; \ [ e* e _e -3 [\ _ _ g *h j s/. 1'<'l(0JVj r,J i
, %gy, /- ,
o,
,/+,
q
/ / \ \
BASE PLATE OUTER RING
/
LARGE PLUG
/ INTERMEDIATE PLUG
\ TOP CLAMPS SMALL PLUG INTERMEDIATE PLUG LARGE PLUG MARGIN RINGS O-RING FLUID OUTER RING SMALL PLUG INPUT an\ t \h w/
gwam, me_mmeg$$l "N.yk'NN khN'N. BASE v FLUID RESERVOIR
/ 1100-0 ALUMINUM )
PLATE SEALING PLATE M A-3929-143 Figure 3.3-1 SCHEMATIC OF SM-1 VESSEL HEAD 48
1 l O O O MECHANICAL PRESSURE l G AG ES
/ PRESSUR E RE LEASE: ON-O F F ON-OFF X X M
' ' METERING METERING O
')
M - 6000 psi PRESSURE TRANSDUCER P2 PRESSURE NITROGEN RELEASE: 3 i im... T lu m , it ir- .i ON-OFF
* (
V [- T> ,
.%Qp(:3%${4Q3 ^ ]@
VESSEL HEAD MODEL PROTECTIVE F LUID \ PRESSURE TRANSDUCER P WALL 7 0 3
- MANOSTAT N / GAS O
M A-3929-130 Figure 3.3-2 SM-1 TEST APPARATUS i i 4
e LINEAR POTENTIOMETERS s ,L,- f x N, e x/
. + .
g 4 ,
\ '
' ,/
STRAIN GAGE -3 QW IL
; . 3' 4 M F '
'5 " W 6- A
/ .
~ g <^)
/ -
6.572 l "4 d5 26.125 '6 l3 o . I g ! Fg - 5.897 16 i- ',
}. 5.4 ~i2
+ 7 1 g
- u Z:T- ' ' ,
g 5.342 l' - '/ 7 l \_ 4.978 ' 5
^F"('?
i ' ' y G 15i ' l' / ,
/' ' um 3~Q ,
-+-
.) 3.500 ? 8 2.028 " 154 e ' //
. f,13 --
- - e8*[**
m ' o -_
- . + -e-e -
14 i " " *14 ' o a { 1.665 1.360 - 7 '/ / . 3' hf 9 / 9
'N 6.572 ,
4h 91 l 12 l: l2'
- g ;
,17 r_4 x
, Nff,
- /
y Ql }II ^ Mk\ /' 7 . MA4929-113 Figure 3.3-3 lNSTRUMENTATION LAYOUT FOR STATIC TEST SM-1 1 I 9 9 - - 8
l l l l ( i i i i I i i i y / ' o u e i z w O s .
/ 9 3*a p ,'j l E a
hO[$ O u. v I . N x
/ / ,'
j
-- 6 572 ' \*! '-
//////Z" !)h_ d - $
%, ~ ' ' '
n n i t =: :;'d'
-/
I ' s / ,
-O o 4.9 78 ,/ /
/ >
t.n l- h -- ? -3.500 l",/h. u
/
w i', T _w i',' 3
- p. p -n 2078 h /r=r / , $
D W
-J i .-m 1 665 - -4" -
/ T
,/ ,
N xl / s
-X
- l$\' ,
,/ , ,'[
- _- 1 -- . _ .
o &-/f/-. 9
't
/
/ -__. 't 't i ^t ~R
- ~;' ; / ,'
/ 7, o
Q g g g g cn go n g .-
/ / / X g- j ' 7l .
w
. JN - / / o
; .-m 2.752 -[_' - ,' / w N \ -9 3.177 -
c ,. ,/ 7 ,/ - _J
\ "m ,! l O N / / C
,I ll L K l / '
/ / }
; _I / m.
i -C 6 125 M Wn 6572 S' *E. s .
- - - . , /
/
to y'~ N,
%rn , /
u g g'/ '/ t.0
]
.s
/ % w C N f.- ,' /!M. 4 N
t 8 t N t 8 I i 3 t t - O - 8 375 C -,. 5-'"
s ,/
- -/
/
' k ff Q
o o o o o o o o ,' , 584h - NOID37 3 30 g O O O
t l l l l i I l 1200 I I I I ')_ 1000 - 1 800 - I m w N 600 cc U$ 0 E 400 - 200 -
' ' I I 0 -
O 5 10 15 20 25 VOLUME CHANGE - in.3 MA-3929-138 Figure 3.3-5 PRESSURE-VOLUME CHANGE RELATION FOR SM-1 HYDROSTATIC LOADING e - -- - - O -- --- - -- -- - O - - -
- O O O i
OUTER RING LARGE PLUG g 4- ' R
~~ ~ a 7_;_ M f g O lf ,,
M _
.MN, 1M
\
^
~
1 - _. g _. ~((39 '
... s. ._ ..,.- E. -
' ~ ~ ' ~ ~ ~ ^ l
?. g ; ..
S j ' $x, { _i l g V
, < .t _
i LARGE PLUG INTERMEDIATE PLUG OUTER RING SMALL PLUG l 1 g
.* d # ..
g- g . .
. ,. _ j2 ;
* 'r 's
^
.- gr , .k; .s 4 .
. :e az ,,7.
e i~;; , ' . s g.. ,
, , p
#^* ',_ * * -' ' ' '
'6sy 4;1-l3. ,,
--- li * '
l 4 MP-3929-11,4 l Figure 3.3-6 Final Deformed Shape of Head: Static Test SM-1 l i i
l l l l l 1 HOLDDOWN ' STUDS (72 PLACES) SIMPLE HEAD A2 607,8. 9 COVER GAS REGION i A' UDS [ WATER SURFACEN -
' Pg ,
j N ; [ 3.45 in, Tf[ I 1.35 in. ) WS, l yf -L I W5 2 lP 7
, SGS,6 _- 7g;
'Y-l SG10 "
Trr DI APHR AGM lp f N SG4I s 12.86 14.10 in,
)
12.15 in. 0.119 in. f
~ 7.775 in.
(Ni 200)
---+- 3.72 in. -
CHARGE % SG2,3lN ; lp 0 LEAD SHOT CANISTER l ,#
; p 4
j 21.85 in. 3N' ig- 4 I ALUMINUM ll i , iiii
' ' Y AhP 2 SUPPORT gg SG11 3 , l p, STAND AIR CORE-~ .
BARREL f f
?"b ,
Y$$$j[y ~ SEGMENTED
' S C ANISTE R '
SUPPORT STANO
+
M A-3929-139 8 Figure 3.3-7 SM 2 WITH INSTRUMENTATION 54 1
p SIMPLE HEAD ~ ~~ HOLDDOWN
'd STUDS
' ' ~
COVER GAS REGION m n' a_sw I ' ' i WATER SURF ACE I
,.h s v p. -i .-- -
g- 3 A 5 in. p;_, 7- g 7-, 1.35 in. lf( ll [ 1Pg 7.80 in. l WS 2 ' WS 1 l SG 7,8,9 l
~
Y] -lf
' SGS,6
- (THREE 7 i
. STUOS) t 4 i
' lP 6 SG41 j 14.10 in.
UPPER 12.86 in. 12.15 inj l - 0.119 in. l VESSEL lNTERNAL ~% STRUCTURE N- '({r / - j, f-l'
?. D, UPPER CORE / '
j
/
/
R G l SG2,3 -'[-'~ ~ M 3.72 in. l ALUMINUM DI APHR AGM -
*~~
; D
/
CHARGE CANISTER ,
,ja.f n ff AIR Hy W CORE 1]P 2 -
iP
" 4 BARREL -
,, SEGMENTED 26.75 in.
(0.100 in. II , ;% , STEEL CANISTER THICK [ - -! RINGS S ORT I p O Ni 2001 STAND ( l SUPPORT < STAND A4 t l kh% hM+ bhkk UA "W =:v-sbfM'e%N fd'F MA-3929-140 A Figure 3.3-8 SM-3 WITH INSTRUMENTATION 55 l
q[- - A._ .. Ilp.a 4 ;t==P 4.C ht/Ai:[f]4 4
-,...-----P , .1 l g
p _.._.A--___ p j [____,8 1. y = _ _"'="_TN'~ aP P T P l - pa;-3^^ 7 ~ - ~
"f 7 P
% 6 g l *P 6
j g yP g I(,; l-) .;, P 5 3 w, :p M 8p IlM p !p 4 .!P I p:hp l 1
] == 27 1 t P 1:: ~--: ! 2) 4 1 3
)
.. $- 5 -. ; h l
..g :.--_._ _ __ .f SM 2 SM 3 5000 l I i i 5000 l i i i ;
SM 2 -- SM 2 4000 -4 SM 3 - --- -- _ 4000 . SM 3 -- --- E I I
't
, E l 3000 l 3000 Y w W w b.y c: ( : cr
@ 2000 e
-g.- D 2000 g ,,..
m [ 1000 N. w g 1000 Q ,, O - 0
! l !
0 1 2 3 4 5 6 0 2 5 1 3 4 6 TIME -- msec TIME -- msec Py CORE p CORE 2 5000 i i i i 700 g i ; i ; l SM 2 -- SM 2 600 !
; 4NO SM 3 - - - -- ; SM 3 - - -- -
- c. a 500 -
g . ,3 l 3000 l 400 gM. 3 2000 00 i kg n w V y 2oo ;J \\ l r;
- \
c 1000 l N' m m \
/
/ <\
f( I'T
'\.' '
F I \ J \ h. . . . /' ?\ 0 ~- 0 Y Ny^I f
.l~l I
-100 I j 0 1 2 3 4 5 6 0 1 2 3 4 5 6 TIME - msec TIME - msec P3UPPER CORE P4VESSEL WALL AT CORE M A-3929-167 Figure 3.3-9 COMPARISON OF LOADING PRESSURES: SM 2 AND SM 3 O
56
r p J~X: :"1}.
- g["3 i- [l 5 ,. ..
m*r pt= P . :r2
% l'1 ___g g--
-p ,
r.- ".p a r T y;g=- -- -- }l P 1 1 4 l- T y j 1 ll q rP 6 LP g l p ' 1 -'- l i i i ! l iP 5 . J. { k--) iP 5 lpg-~- .~ p %__ I l M - 3L-j y[ j P
- g g P2 +P 4 P
- g. , g y P2pP 4 4
=-
t s : i
. Ls =d __. .Shk.-4\ _ i L_ , _ _. y SM 2 SM 3 700 , i i i 700 , i , i i 600 SM 2 600 -SM 2 a
o 500 yO SM 3 - - -- - - 500 SM 3 - ---- - - l,
~
-E l -ih. i i j. ,,,fi' i 400 j - ,
i 400 ijy' 300 -f', 300 h", T\ h 200 1rM 1 dA . $ 200 _je g ;y ( i C
- a. 100 j M. If il i /\ f\ f\ . A ..m e a 100 f .up ,1 vV h p
. .y y. y.g. .y..y
, i ,
f t y 3 O 'l V ' 'I k h ^ D I V v \/ "T \ / W
' I
- 100 -100 O 1 2 3 4 5 6 0 1 2 3 4 5 6 TIME - msec TIME - msec Pg SOOM OMET EME P 6
^
2000 - ' 6000 SM 2' .,_ SM 2 SM 3 -- -- -- - 5000 S M 3 --- -- - a 1500 j (* 4000 I l
, j000 .
w 3000 a -
\.V , .
g a SM HA a 2000
? 1000 -[
__f.i MV. . .. n
.) G.Mw
~'"
0 *$ -
-t000 0 1 2 3 4 5 6 0 1 2 3 4 5 6 TIME - msec TIME - msec P UPPEH VESSEL WALL PgHEAD 7
M A -3929- 168 Figure 3.3-9 COMPARISON OF LOADING PRESSURES: SM 2 AND SM 3 (Concluded) 57
1 i A2 A3 A2 A3
+ l? -
D D
~
____..A.;_ J ' [ . I'[
. g - tm %r I :
% i j
[] i l i p,..l ,
! i hI I
c ;;gc F i
; g r1 ! '
-Il f '
)
s . e . t
.9 5
EE F._ F l I 1 {f
,== .
l! f
. .L A. _ -
L lA _ J , SM 2 SM 3
. 1500 , 1500 l 1000 l 1000 l
z r A -k z j,i O 500 O 500 G I- ' G '
%F"v^'(pakhv g 0 r+d Q,-3@f 4, <
l g o d -500 J -
! j -500 ,
o f u , S -1000 - -
$ -1000
-1500
-f , l -1500 0 1 2 3 4 5 6 0 1 2 3 4 5 6 TIM E .- msec TIME - msec A gSM 2 HE AD EDGE A g SM 3 HEAD EDGE 6000 0000 4000 4000 l l I...._
z 2000 z 2000 0 o i j O h o - { fr h!*^ u
^
0 O! f*",P Y ' n%d " w g w .
'l d 2000 -- -- ---
d -2000 8 8 4 -4000 at -4000
-0000 -6000 0 1 2 3 4 5 0 0 1 2 3 4 5 6 TIM E -. mice TIM E -- msec A 2 SM 2 HEAD CENTER A 2 SM 3 HEAD CENTER M A-3929-171 Figure 3.1-10 COMPARISON OF ACCELERATIONS: SM 2 AND SM 3 O
58
, A A3 m j 4%3_'Zflp --m ,-L==- L L l"1
** w n:= .-- - - .
{ . .. q=,.=_ _= ,
. -1 .
j ;r j t ( l I ! j l l I ml}-;j l i-l 1 I I .r - -
~;; - ,
L1 1- :.
- } l i .H 55 r.. + , '4 ES-E F.
= .:
=
s j s= I I
!L h._ Jqd k._),
i j s-A 4 T A4 SM 2 SM 3 l 1 1500 '
, 1500!
a a l 1000 f j, l 1000 6 5* 5 J . . i iti 5 k t - , P d' i14 1 iTA P A A M 1 ltLi >.h
..t / .n 0 o
[ jy ( j,- g,;,ii g -y g-- - y- v v // Lvri Mc' 's" n {-j$
"i -50 0 "i -500- '
i [ U v .1000 O I o -1000
-1500 , -1500 l 0 1 2 3 4 5 6 0 1 2 3 4 5 6 TIME - msec TIME -- msec A 3 SM 2 HEAD EDGE A 3 SM 3 HE AD EDGE o 1000 1000 g ,
! j h l
*
- l 5 .g 'O l bl o 5 _th 0
_fL b ' h
' A Mi ,
v' b 0
' ime' w"'
lsn < ~~ e E- e 3 r
-500 - b -
-500 0 fl Y I O l' l
-1000 _{ 1000 q j l !
O 1 2 3 4 5 6 0 1 2 3 4 5 6 TIME - msec TIME - msec A 4SM 2 PLATFORM A 4SM 3 PLATFORM M A-3929-172 Figure 3-3-10 COMPARISON OF ACCELERATIONS: SM 2 AND SM 3 (Concluded) 59
p r,3
. 1"_11.~.fl . I
[.- :IlteTO- - - -
,,..y . .
j i. q A y ____.A_,.____ SG5% ~TW ! SG5 "l{' _- lr 1 :> SG4 ' (1 SG4 ! !
! I r i lI l'
l l l
> l T -- >
SG2,SG3 ' (j 1_ , ,- SG2,SG3 1 j l N
.i u g
EEE
- r. +
l t j'=[ iu 1-p
.4 h . ._'. :a .. :A __j L v ' W SM 2 SM 3 3.0 0.6
^
2.5 f
~ - -- -
06 .....
$ 2.0 [ $ 04 l
^
1.5 0.3 rW r' u 1 : A l l 2 1.0 f
^
z 0.2
-f I y
0.5 [ SM 2 -- g 0.1 - SM 2 7 0 S M 3 , -- ,- - - -- 0 , SM 3 -- -- -
-0.5 I I I I -0.1 I I ' I I O 1 2 3 4 5 6 0 1 2 3 4 5 6 TIME - msec TIME - msec SG2 VESSEL WALL AT UlS (C) SG3 VESSEL WALL AT UlS (A)
(SODIUM OUTLET NOZZLE) 3 .0 3.0 l 2.5 L5 [u 2.0 j^- -r-~ j u 2.0 f [ i .s / l 5
" i.5 I
i l z 1.0
/ z 1.0 !
a E 0.5 l a E 0.5
$ / SM 2 --
$ SM --
0 -' ~" a I SM-3-------- 0 S M 3 - ,-- - --
-0.5 I I I I ' I I I I
-05 O t 2 3 4 5 6 0 1 2 3 4 5 6 TIME - msec TIM E - msec SG4 VESSEL WALL (C) SG5 UPPER VESSEL WALL (C)
MA 3920-109 Figure 3.3-11 COMPARISON OF STRAIN RESPONSE: SM-2 AND SM-3 60
A i / STUD ST UD l .-, 9C
,. 4 L" 4 ~ ^1 4
=hrst Cl y . _ _ . - - _ , _ _ _ . .
-_)
's --, A SG6 - -_ , SG6 L_ -- ---j
}
q l=J 1 - - l, I i i 1 j p+[r-r l
+ !, -l ag_. - 5 _ _ ._i l 4
sEs
=_: e > , <
1MI : > l = . 1=
; = jr - t.
l _ -} _ h .- - i A . . ..a' SM -2 SM 3 1.2 I I ! l 0.20 l j jj 1.0 SM 2 d SM-2 a { S M 3 -- -- - - fg g 015 S M --- -- / 4y e O.P lg e E 0.10 A E 0.6 i ( ~ -%~ / ( g l 2 0 '4 Z 0.05 L ,/ a a / *i 1 M 0.2 .
?N , N m
b 0 g .. 7 i, h : !!
'h,g.'
w
#' i
-0.05 O 1 2 3 4 5 6 0 1 2 3 4 5 6 TIME - msec TIME - msec SGG UPPER VESSEL WALL (A) SG7 STUD AT 60'
.20 0.20 j j l g l j j ; !
- SM 2 O E 0.15 . SM 2 !
\
[ 0 15 SM3-------- l g S M 3 - - - ---- pl ji
- l A E -I N %
a 0.10 0.10 - l ( / T l . l m \ E o 05 E 0.05 I g l - i- '\ - g l '. \-
$ 0 *f f% : 1 0 - -- M " f i h X F % @lA
-005 -0.05 0 1 2 3 4 5 6 0 1 2 3 4 5 6 TIME - msee TIME - msec SG8 STUD AT 180* SG9 STUD AT 300
M A-3929-170 Figure 3.3-11 COMPARISON OF STR AIN RESPONSE: SM 2 AND SM-3 (Concluded) ,O l 4
\
61
O STR AIN - percent : r 5 4 3 2 1 0 1 in, I I I I I j 1 in. AVERAGE DATA (6 MERIDI ANS)
,/ - 2 a*' SM 2
,6 -
3
/ ~~~~~~ SM-3 4 '4% I
~
4 8 .- 2.8%
- SM 2 Strain Gatfs 5 O
% SM 3 Strain Gages
~
0 O
\
s - 7
\
\- 8
-- 0.119 in
\_ g
+ :
i I '
/ - f essel V Wall
/
/ - ii
/ -
12 A - 13 3.5% 1.8% - 14 STR AIN - percent
\
\
15 3 2 1 0
\ _ ig i i i Toe
\
17
/
\
\
r 8
.f
- Core Barrel f
\ -
18 l- 7
) /
l - 19 I - 6 f -: ,- 0.100 in g - 20 - 5 21 ~ 4 j g 1.5%-j 22 1.3% -
~
3
\
g- 23 - 2 y 24 - 1 l -.- 1il l 1l f II I I I I I I 25 L.LL l i n 0 300 240 180 120 60 0 ! 120 60 0 6 R -- mils < , 6 R - rnils s i M A-3929-163 Figure 3.3-12 COMPARISON OF DEFORMED SHAPE PROFILES: SM-2 AND SM 3 62 0
l l O l l l I l l I I I l 1.2 - SM 2 f\ Om] su 3 ----- b 1.0 -
/ _
c 1 / I 0.8 _ [ IMPACT VELOCITY g / 62.5 f t/sec g 9
/
[ IMPACT VELOCITY 01 5 '"5"c 04
/
/
0.2 - O I I I I I l l 0 1.0 20 30 40 ' TIME - msec MP-3929-157 Figure 3.3-13 COMPARISON OF WATER SURF ACE DISPLACEMENTS: SM 2 AND SM 3 l O 63
I O d I lc UlS CAPS
.. . A NI .
HEAD SHIE LDING : 2 1 3 4 l l UlSCOLUMNS : l UPPER INT ER NA L I STRUCTURE 6
. .. . t . a . , a_ _ , -y.s.......
MP-3929166 Figure 3.3-14 Upper Internals Structure and Head Assembly From SM-3 O 64
a 400 m Column 1 _ 200 - O l I I I I I I I I I 400 m . Column 2 _ E i 200 - 0 I I I I 1 I L J
~
400 m Column 3 _ E i 200 - 0 I I I I I I I i i I \ 4 ._ Column 4 _ E ' i 200 - 0 1 l I I I l l I I I O I 2 3 4 5 6 7 8 9 10 11 ' 12 - CHANGE
--=- 1.0" ~ 0.55" ~ ~ IN
(~7%) LENGTH z-
-- l si m -
l 0 050 in. Thick Ni 200 NO AXIAL COMPRESSION AXI AL COMPRESSION UlS COLUMN l M A -3929-16 5 Figure 3.3-15 Deformation Profiles for UIS Columns in SM-3 l l 65
l man HOLD 00WN COMPLEX HEAD STUDS (72 PLACESI COVER GAS REGION ! A 3 A 2, i L ' [ WATER SURFACE k, i i l a i ' /\ 6_ d "
\j g a o; _s s=v=,n -
3.4"5 in. 1.35 in. y - _ .=., _ rl , i ll JP' S ' SG 7,8.9
, WS 2 31 (THREE
-=
l F'~ STUDS) ySG5,6l p! i "' f #
- 0.119 in.
(, i VESSEL UPPER 30.75 in. 12.86 in. ' INTERNAL .j ,ir- j
'p WALL g/ !'l >
STRUCTURE ' (Ni 2001 F l [
/ / '
d d l UPPER CORE SG3,4 {) [/ fl ) !
) -
f )P 4 0.078 in. ' FOR R 4 THERMAL _.j _ - 3.72 in. LIN E R G 21.85 in. (Ni 200)
...-i DIAPHRAGM La rpOz( _._
l (5 P, MMP 2 & I l CORE 8ARREL l m 2d9 E-j PdlNifl#,F-'Q.24[f" i I ALUMINUM l 27.78 in, 10.100 in. THICK CHARGE g g -.__.- gi ! ll SEUMENTED i CANISTER SG l ,2 ~
) STEEL RINGS Ni 200) % J[3-- W,
-+ ,
)p==$=-1 -',
CORE f a b SUPPOR T r STRUCTUR E l C ANIS TER l SUPPORT ' f STANO SUPPORT --*[ STAND wy
, ,, h &
fh h \ h M A-3929-1420 Fic,o e 3.3-16 SM 4 WITH INSTRUMENTATION O 66
i /N l SG16,17 i COMPLEX SG 14,15 HEAD A8 A4 NS ig ,
,j A 3 WMM A1 i
COV E R G AS p "" A 9 A7 A 3.45 in. REGION \ %! 6 S FA N L_gi ' W/5 27 --$ P 16 - WATER SURFACE i SG9,3 p' ~ u -1 M p '10 2P 12 7 SG18. - f ' '= g il ' ' * -j {o "" g ""~=< Pg 19,20 7,go 0.119 in. 1.35 in. 'g ] _ (3 -- in. VESSEL g3 WS 3
' WS WS 1 #
WALL P i 2l
,8 SG 11,12,13 p e
- n. 0.75 INTERN L STRUCTURE N P
, T E1 AL 21 E5 34.83 I
12.86 in. LIN E R UPPER CORE #l h l p l, 4 (N. 2001 FORMER ~
- f / r RING P P G 5,6 3 7 OIAPHRAGM ' 3 725"-
.%_ gx ---~ h w= q y M35 CANIS R ( , r,- ALUMINUM OA RL ,
(n. lP 6 [9 g T I k
,S,G
{_ _ [=r= 3 ---] F l SG3,4 G ---- f l \\~ * .I SUPP 1 STRUCTURE CANISTER i SUPPORT - lP S
-- l l SUPPORT STAND ST AN D .
Y h7 l
~*MS%%WMSRQi;WKWWW4WW M A-3929-141 A A
() Figure 3.3-17 SM 5 WITH INSTRUMENTATION 67
-m. ,
_a Rubber
- t.
e "0" Ring g ,
,-s, Seals '
- v. x L
;51 - , , x
_.,g (
/ ^ f *
6 '
- - '\ ) i T. - . _ . . _ . .
,,f- g ) _
,, s y' l --
N L' ; !. I< ; !- ; / j 1 'i i i
\
.[.' !
)..
' i c- ,% '
i
-Q ,
i
- I' -d ' f s 1 x. ., l 3
. ' ' J.? 6 %-
i t m e,+ .- C3 ' II s
-4 i,
,._ ll Detail "C" Detail "A" Detail "B" Pressure Gauges P15 - -- --
and P16 (At 0 and 180 ). Gas Space Simulates Volume of Riser Annulus Detail"B"/- Detail "A" m
'J !' I 'i ,
! Detail "C" f ;4--l!' f:;
l' ' Iy.
',C 1
- I, g,;-
- i)i ej p i
q ,q.- .~j i Pressure ; ,. ._ , :p? ? '.?'t Li
, 32 -, _.
Gauge P14 - , 'a ' ' ,3 f -1, - i Q_ .1
.<i
' - -P i' _s.
! i i
1! j' i
- , ,i j+ i
*"0" Rings Adjacent to Margin Rings Serve to '
Assure Correct Positioning .! of Margin Ring During Test Figure 3.3-18 SM-5 Vessel Head Sealing Arrangement e G G
. . _ . . . --- . . . - .~~ ~ .-- . .. . _ - _ _ . . . - ._ - - - .. - - . . . . -.
-(% (W f- y f.gl g .. .)
T - s " 4 I--. $$ 7F ~77 '- ' Mh,rh [ T7P
< I - .P . _.__
P S .I . Pg i i
< q
.P[ __
{ v
% l4 N{
l l '
- l l
_ i I , > P, i L P, o
- r. . ..
r] z.m l u r ._y.c r] 1
- - .T .--- ,
T _ I 5 5 bp
=
>P 3
i 5 [ I P
+P 6
=
P, P,
+h
.i
. 2 -- 7 I _._%
..- I .
\.kSii9il/Nl
.anusuu % a$'- Rrunu.a/
l 4 SM 4 A5 5000 i 5000 -- i
... j , ,
i 4 i i , i 1
! r_ ,
1--+4- - - - , e- , g 4000 ii p ;SM-4~~~ 4 g 000 .- -- ----* L SM 4 - - - .
--.- ' - SM S
-~ ' ~ ~~ SM S -
I l 4 l i
' -+.
)
w 3000 r-- ----- I 3000 - - - - - - - g i +-. A- i_... F--} I _8 1 i w '
--,-------+--a-----------j----"---------
, I i i
i e
"- . __f- j l l J _... !--' 24. kt---+f h2000 - ---' --
h2000 j r - " -f"i w k j
. - _ L. _
w _.
. _ a__.4_
_.7-._, . - r ._.- . . . 1000 l_- b Ou _ -.___; _. . .
-_E 1000 --t I- r e - 1 2- -
-h_ a l-__.
_ _ _ . _ , 4__ ;_.l_ __L.a.__ _L- ' O ----+ q ;
,- - - - 0 q
i Ml ~~ I l 0 1 2' 3 4 5 6 0 1 2 3 4 5 6 TIME - msec TIME - msec P3 CORE P2CORE 700 700
-p.d-"-l"p_-t., --*, SM t -- i i . -- !
600 _ f ---- _.y 4p 3 __.._ - H- 4-H+H-~r-' SM
- 44 P 4h ~-- --*-- t--
8 500 600il_-('_4- j L .--H SM 5 P ,_& ---
-4 J{ -. ~,7p , -.-. - 6 -
500 l- lH6~---
. SM 5 P 7
't - A-l !
400 f l 4004 -d+ ' Lr_. _+ _4H
. ,H.- r_. . .. - --4 g y
? _{ - e--+- w _n.y' H
. _1 -.
3 300 .7
-7 _".g.p.- y> ;-- -+7 g 300l 4t- ! , ! J l g t 3.,
4 ;&q . .t , __;_ _, ._ _7._ b w 200 g: . ._. _.--- P., 4.. ..,..._-_--___f,s 200
-1_+_
h-pM --_ g , ._. . . . _ _ . g _ 7 .;q . r__ g. a 100 -, L J. r m.. ~h. a. 100 -.-------k e _4 % -,--.--.--f '
.*%,c}. q-- ' -
+- - ---
L_. 0 -..p.;r- _ .
- ? -4_ .-
_ -a!-f '- - - - - - - 0 Ma i ! ! ! ! t i--- i._..-
-100 ' r--t '
100
- M -- J ' - 2
' ' ' L--
O 1 2 3 4 5 6 0 1 2 3 4 5 6 TIME - msec TIM E -- msec VESSEL WALL AT CORE VESSEL WALL AT UlS (SODIUM OUTLET NOZZLE) l M A-3929-210 Figure 3.3-19 COMPARISON OF LOADING PRESSURES: SM 4 AND SM 5 69
O m n gla M b2&&& q'" "T"? grg Ibn
'7"7<
P' n i
~
wrr < ' -i W 3 a:Et:[:p'
; 6 i : ,5 33 9
l p- 3 s l i , l ( LP 4 LP 7
- y. r. :Q f r~
--Q 1
- ~_. .
=
b P,Q;ih p-- Pj # 6
- A P
+P 3 'h i l
P p . j ; g 2l TN ._ .FF=d J lQ&Lnk l usuno ; Q1:nndS./, ~m - I i SM 4 SM-5 O
. . _ ._L_... . . _ _ . . . . _ . . _ _
5000 3 2500 -'- SM 4 P 5'**-* g ..._.1_.' ---.. ,, 4000 - - -- - Si4P 6 ---- 2000 - ~ _ .
, . SM 59 P - E - - - -
SM 5 P3 - l - - i-7 l 3000 - - - - - --
+- -
W 1500 - - ,- - - - - - - - -- w . _ , _ . _ , . _ _ _ _ j , 3 -._.14 . ._ . _ . . . (r -) $ 1000 - . L- - +- . J -- - - . . -._;.. 3 2000 w - .;. , . , 24 g _-._7______ ._. .. - - - a 0: 1000 - - - - - - - .
- _ . _ _ , ----+.3.g - , .
1__t E _ yg]J- ; n. 4 0
,- _.. ;_... - -e -
, _ . . _ _ , 0 - --c \ _. - . . . . . . .-~
-500 0 1 2 3 4 5 6 0 1 2 3 4 5 6 TIME - msec TIME -- msoc UPPER VESSEL WALL HEAD M A-3929 - 211 Figure 3.3-19 COMPARISON OF LOADING PRESSURES: SM-4 AND SM 5 (Concluded) 70 O
_ _ _ _ _ - .__ _ _ _ _ _ .- _ _ _ = ___ __ P 15
']
Pg 'll v% 3 T~ .' -A b i77"~ r <
]
'i !
I k > car &
&5 :
k bu,Ld A sua m. /' SM-5 60 50 l 40 40 d g 30 l30
. A ( 20 1
I! g - 4 u-I EI f- \ N 'I r 'o
-h 2 t t 10 r u_
p g 0 . h ir h i 0 ~I
.a. i On A 6 1 I N Ofi 1
-10 -10 V1 I O 2 4 6 8 10 0 2 4 6 8 10 TIME - msec TIME - mste P15 FLANGE RING Ani' G AP (180) P16FLANGE RING AIR GAP (O')
700 50 { 600 40
._E _E 500 g 30 (g l 400 / \ $ 20 300 / \
h'* W (-N l jf .Ama _ h' I v '! '
~ ' ~
0 -# W 0 - 10 I -100 6 8 10 0 2 4 6 8 10 0 2 4 TIME - msec TIME - msec P13HEAD-AIR GAP AT CENTER Pg4RANGE-Am GAP M A-3929 241 Figure 3.3-20 SM-5 COVER GAS, INLET PLENUM AND UIS PRESSURES O 71
,,--,a --~w ty- y - - a- ~ * -'-7--TP- N'-
w------,,---,c 9--r3 w-
JTTa MI ho 3 91_ [A E 77 m n iETI-Y..N._
^t l
777-I s Q'. ' ~ ~ ~ ~ ' p 700 , , i , l
- j. 0 '
, p 600 -
- I d
j d 500 - i ~ ' ' ' k_L\ 400l-f f
.T3 l Q Ff l
+
g 300 -- , 1, 1
\% ,
l
$ 200 - *
$a. ._ y e. ;, ,
I . y ,<,n . ,
; , i a. 100 - -
wa ' (%usu! dt uuw val <W 6, , O k~ hP -100 T '-! ' I ' ' I ' i ' 0 *1 2 3 4 5 6
* #) TIME - msee sg.5 P7VESSEL WALL OPPOSITE UlS 1400 -
3500 j , 1200 - 3000
. t
{ i .
.& 1000 - i E 2500 -
l 800 - I I 2000 w - ,W g 600 . g 1500 fg e M l m l V 400 l sp ^ -
$ 1000 -
3 m . . , ,- - c i
- 200- - ' ^g' a 500 g o m: __ ._-
,'N
^
f~ ;' , ' 4
.$00 1 2 3 5 6 0 2 4 6 8 10 TIME - msec TIME - msec PgVESSEL WALL P3LOWER UlS
. .a 700 600 - -
600 g 500 -l g 500 l 400 l 400 - g E 300 I
$ 300 --
h 200 V
~
f 200 - 100 --
, 100 l ,
%k2 0 f
} _ 0 l 100 100 0 2 4 6 8 10 0 2 4 6 8 10 TIME - msec TIME - msee P4UPPER UlS P 5
M A-3929-239 Figure 3.3-20 SM-5 COVER GAS, INLET PLENUM AND UIS PRESSURES (Concluded) 72
av. T v. f M ri 7 m - l
*jL.,ld.ip-[n t n :
y! lA - ]" .
' 7"P'""' y T~l"'r'" fQ _g..l.f._.{ g i 7T*r" SG S, G l, $..___f&_ __ . SG 7,8 6 __
, k .
s N: lc q;
~ ~
~*
((Ilf[ , , SG 3t :
> SG 5 t L
- y. L b ,
yElk I
% h:-
[: i, , j {
=
p , j
-l .
I
.h/
l h FG'y uuun : h, %ulun,d rh m , 1
\
! ! i
! l
'%. SM 4 SM 5 0.40 _..
..._.g_,__,._....___o.. 25 , . , ,
i _: + a _. " - - -
._ . ' ~
E 0.30
~
2.0 - SM-4 SG 5 - --- ._ . ; _ 7 _ .. _ y W - --& - . g g -- - - ' "-
- - ' ~ ' - '
.- SM S SG 7
& i
& 1 '5 - - - - - - -+ d-- "
l 0.20 $
._,_...._'....._t..,_-.!
~ SM 4 SG 3 -- ' - ~ i 2 ._ p .. ._.L...SM-5SG5 _. _ . z I
1.0
~~'*---I~'--~'
*-~t'-~~'
4 . .-, . . , , - _ .
._ g . 4 --+-+-+- - * ' - - - - -
e 0.10 _ .i _ - _,; 1 _ _ ._..4_ , _ _ H_ _+
.. g H ._ 4 _ . . . ----k-- 1 -- + -
M
-: e . . 4 .._i ..J_ _ p. .
H m 05
.i t. . - ;. s.w
, -, _. ci,.,;?+^.c : c, . - - - - ~ ~ ~ - - - - .
-,--7-,-N'-.-.._--.
0 .a i- - o
. . . , ';__.._.. l
-0.05 ~~T" ~"~ ~'
(j] [ O 1 2 TIME - msec 3 4 5 6 0 1 2 TIME - msec 3 4 5 6 VESSEL WALL AT UlS (C) UPPER VESSEL WALL (C) (SODIUM NOZZLE OUTLET) 06
.. ._.-_.__._4._._-._...._ ...
0.5 SM -4 SG 6 . .. . . .. _. _ _ ' . _ . _ . _ .
$ ~ SM 5 SG 8 -i j M~ E
? O4 g --*-*--~T."_y_.p*--__~___*_-.___
_ _ . _ _ t__ . l 03 - - - I __ ., _ -+-41 i _.-I -_-_- +..t_
~~-
z
.: ~~+- -
q 0.2 g - . - _ - . F- 4 -.
* - t- -"
, 4 y 0.1 '
-. ; . ~ .- - y} - . z- y- s -. -
-.7.-- ,
- .v - -- ..;;e 0 [t 1 --- c -- -h-+ .; , s-..< '~ . s.. .
0 1 2 3 4 5 6 TIME - msec UPPER VESSEL WALL (A) M A-3929-213 Figure 3.3-21 COMPARISON OF STRAIN RESPONSE: SM 4 AND SM 5 m t f] 73
r, en
'" SG 18.19. 20
(" f SG 7, 8. 9 h p 1 l'
- A_.q p .. If _.h, ,.j p ,p
'"Y'7~- k _ hy$$ . q7"P '77
" A.
' ,f----- ~ A---l%Ii. s MT
~
.y]__ _
;_.,)<
1
.it 0
_W ; - Mt ; l lP T l T l ' i .'
! , l
{ , _. - ..g-b r t > t > r-- - -" '
-Le i r ' ~ ' ~
5l - t, A'O. ; Gl
'.p: I p- -~~
.1 ::= + '
.4 F ..' +
a-- - i _. ! O W
!-b
. l _.
~
A :. k-N'; u./nk. mm
;A,%. WL' ~Cruna /
i l 1 SM 4 SM 5 0.20 ; , 0.20 , i
+ #-
SM 4 SG 7 - - * " - - - - - - - SM 4 SG 8
-'-- ~ ^- ^-- ~
g 0.15
~~~
0 15 ' ~ ~ - ~ ~ SM 5 SG 18 S M- 5 S G 19 - 7 y - . . _ . . . - _ . .
--', ,; - .. 8 _..__..._J_..._._ _. ~ - . - . . . . - .
k 010 + --
- 2 d + - -- - -- - $ o,jo - :- L - H - - --
l . . _ . _ . _ _ _ . _ , _ . _ . . _ l _._,._.2 l l . . . . . . __ _ E O 05 - - - - * - - * - - - - - - 3 0.05 -- - "-+1---- < !_, < < 1 1 g _ _.___....# ._.. _ _. g > r i H m . . _ . _ .A~.e.,L." .? - ~i m 0 R~9._ # -~-'--q. ~~^ -- - * - --. *:r:. 3. \,Q c
~*
o -
+,,, . _..+. ~-
.- . . - . _ . . v :._-_.. .
p 0.05 -0.05 0 1 2 3 4 5 6 0 1 2 3 4 5 6 TIME - msec TIME - msec FLANGE STUD AT 60* FLANGE STUD AT 180* 0 20 . :
-- H - l
- +--
- SM 4 SG 9 I g 015 u
sM 5 sc 20 l - - I t -* '
; - 3_.a ._ . _7 -
8 0.10 - --- 4 -J r- t L -- - l _ 4_._. iI ' z t > i i__I' _ 4 _4 ' i:, i .. 5
- 0.05 ' r 6- - r-+--t+- -
c ._. _ L.M_!_; _k 4. 1_ _. -- 0 ~~$c.m.Y:-- i - -i,- x-5-n-i ,
. _ _ _ _ . . . . - _ . __. 1
.__ . . 4
-0.05 >
0 1 2 3 4 5 6 TIME - msec FLANGE STUD AT 300' M A-3929-214 Figu're 3.3-21 COMPARISON OF STRAIN RESPONSE: SM 4 AND SM-5 (Concluded) 74
/% SG 15. SG 16 SG 14 SG 17 m h -
. . A .1,Af.
~~T .Eh g. . .
{2*7 7 --- 0.5 {~ '~i O 1 l \ l' - h r_.___ SG 13
-0.5
\
I l -1.0 __) t > 7 E -1.5 u- r -" l" e - \ jhm _ _i __
-2.0 g E]:- ,
C I l
- 2.5
[ Y. 0 2 4 6 8 10 [iIO TIME - msec I l [ ' SG 13 UlS COLUMN 1 (A)
/
M7 M5 0 15 0 25 j 02 1 - 0.10
$ b
- 0.15 -- -
i p __.._ .__._ $ 0 05 - l 010 - - - l H {- E y - ' -- 3 0 + f~% ; o os -- - . - t. y h [
^
')
( , & - ---
&._ n j
'%d
- _' * -0 05 - - -- -
O g;---, . -
}-
I
~M i
-0 05 1 _o.10 0 2 4 6 8 10 0 2 4 6 8 10 T IME -- msec TIME - msec SG 14 UlS COLUMN 1 (TOP) SG 15 UlS COLUMN 2 (TOP) 0 15 0 15 0.10 g 0 10 3 b $
$ 0 05 $ 0 05 L3SA-0 0 -
L /ad % _f a T //f I F m -0 05 -- - I e m -0 05 -( b ---
-0 10 -0.10 4 6 8 10 0 7 4 6 8 10 2
TIME - msec TIME - msec SG 16 UlS COLUMN 3 (TOP) SG 17 UlS COLUMN 4 (TOP) M A-3929-245 O
\ Figure 3.3-22 LOADINGS ON SM-5 UIS COLUMNS
(
'vl 75
1 l l l Oll l l l 300 i f - COLUMN 1 l 1 l200 l
> 100 - -
g i i l 1 -? I I I I I I l l 300
? COLUMN 2 E 200 - -
I i i l I I I l ' 1 300
? COLUMN 3 E 200 - -
I l I l l l l l o 300
~ -
E COLUMN 4 E 200 - I
>100 -
0 I I I I I I I I O 1 2 P 4 """" 5 6 ' 7 8 9 10 11 12 1.0 in 0.17 in - - C(ANGE (2.1 %) IN LENGTH n
-.r-I ~ I * ~ ~
i - i SThEt l l O 0.050-in.-Thick Ni 200 NO AXIAL COMPRESSION AXi AL COMPRESSION UlS COLUMN M A-3929-218 Figure 3.3-23 DEFORMATION PROFILES FOR UlS COLUMNS IN SM S O 76
_ . _ . . _ _ _ . _ . . __ - . . _ _ _ _ . ~ . . _ . _ . _ . _ . . _ _ _ . _ _ . . . _ _ ._ -. _, O STR AIN - percent 2 1 0 l I In- STRAIN F 98 in. 3 f percent
/- 2 2 1 0 AVERAGE DATA (6 MERIDIANS)
/ -
3 I V- 13 SM 4 9 1.4% ./ --- SM 5 1.8%--D # - 4 - 12
\ _ :.._.0.140 in.
- SM 4 Strain Gages
\ -
5 g 11 o SM 5 Strain Gages
\- 6 -
10
-: ~0.125 in.
7 9
) 8 -
8+ -0.078 in. VESSEL WALL THERMAL ~ - 9.6 10 LINER - 6 11 - 5 CORE BARREL 12 - 4 6 R - mils 9 120 60 0 13 - 3 IIIiIiI 14 -
-2 STR AIN - percent
-l 15 -
'I i 2 1 0 16 -
0-f I I I I ' 17 lllll[ - *1 j 8 18 120 60 0 f- 7 19 6R - mits l 6 f I- 20 - 5 1.1% / 0.7% d h 21 0.119 in. -:- f' 4 0.4%- b- 22 - 3 k 0.100 in. --- - 23 - 2 24 s g( (- 1
; 1 ,n. i g
2 % 0.2% ~ 27 1 28 29 llll11 120 60 0 6 R - mils s _.~ ' . i{ MA-g?9-219 Figure 3.3-24 COMPARISON OF DEFORMED SHAPE PROFILES: SM 4 AND SM 5 77
A3 ^2 Aj Ay k J A 4
. . i kgy.}. $ fiq~?"Y
'YT """'"[ - ~--. 9"~" 'W"'~~-
,' .g fi g mlll.iH; 1, l y-,7
'O I
o, :F r r + {" F t I 1 > l,1 L[ 1 - l up_ LlLlL' kg-j , s m. - , .: u . m , .m. ,
- l L- d i '
l l E.id i '
%f l. '#t 19 .3':t
.j==t A]==,)K
'tGnsau Nj.Ci au aw i
)-
)
N # gya# SM 4 57 SM 5 4000 4000 e 3000 m 3000
! 2000 f 2000 0
[ b 1000 b 1000 0 N & y? " - - 0 . jg.: = y ~~ w -1000 w .1000 V at 1.
.R.
_J d -2000 __l - _,,_ a d -2000 N-3000 N-3000 4000 4000 0 1 2 3 4 5 6 0 1 2 3 4 5 6 T IM E - msec TIME -- msec SM 4 Ag SRP NE AR 1RP AT 180* SM 5A4 SRP NE AR 1RP AT 180* Ol 4000 4000 e 3000 e 3000
! 2000 l 2000 2 > .--
', 2 . .
g 1000 O 1000 l 0 If4 A'.^ VO 2_ gad .) - Apdr uj -1000 ri j
}g w -1000 O
fg d-2000 y.2000 I N-3000 N.3000
-4000 4000 0 1 2 3 4 5 6 0 1 2 3 4 5 6 TIM E - - msec TIM E - msec SM 4 Ap IRP NE AR l RP AT 180* SM5A6 lRP NE AR LRP AT 180*
M A-3929-216 Figure 3.3-25 COMPARISON OF ACCELERATIONS: SM 4 AND SM S i l 1 l l O' 78 l l
2 1 ^6 ^4 A l A7 3 1 i) ,} N- . 'l; [: ngl _r= ! j . ^yLN r,
,i 73 vr~ j@_555%4@Gx7n- r _ .)t-
,f- -
- .f52l%-
% hm - .
'O 'i ,
{ I L I E q C. b l y_ ' l C~ .
.L- :
l -4 h p-I! p I y -i. p 1
- - 3.
= E l ._.}
% bhDbLU : h%$$fibNb
! 1 l SM 4 SM S 5000 5000 i 4000 , 4000 l 3000 l 3000 2 2000 o 2 2000 p 1000 9 - 1 I )
_' F
< 1000 A ;
4 _ _ . e . h &N h%n* $4- [ d -1000 0 " 1 di n
';ff y 5 a
w -1000 0 "
&f5_ +-
2 p 1 O -2000 O
< -2000
< _p_
-3000 t-t j -3000 1 I i
-4000 -4000 ;
O 1 2 3 4 5 6 0 1 2 3 4 5 6 i TIME - msec TIM E --- msec SM 4 A3 LRP NEAR 1RP SM5Ay LAP NE AR 1RP M A-3929-217 F.lgure 3.3-25 COMPARISON OF ACCELEi ATIONS: ' SM 4 AND SM 5 (Concluded) O 79
i Ag 4, y, A' 6 A v __! Aa* 7 2 -- 97T
' lt 1 9
4000 l
+ e 3000
[< ' t > ! 2000 uf 0;[h b 1000 - 3= i. a -1000 dtd $"$AZ l-b . b-2000 Ni i -3000 i I 1 -4000 0 2 4 6 8 10 t,
\
) TIME - msec t/
SM-5 A 6
^" "
6000 0 07 ; i e rA00 - -
--- ~
[o 0 06 l-[-f p l 4000 . _ _ _ . _ _ . _ E O 05 _ _'
, 3000 1 0.04 -.
~
$ 2000 _
.d---- .- 0.03 -
4
"^
2
---- -4 d
-- - +
Q 1000 b 0 (pw Aqs- ' [ U
.2 7 ,;
2-~ 1. 22
--2 11 i
~
d-1000 l ,~ O-2000 Z-- -
~
-Z-
~~
- h 0
4 Z-
~
Zf 9 -0 01 -
-3000 -
- -002
-4000 --- -
6 8 10 0 2 4 6 8 10 0 2 4 T I M E -- m sec TIME - msec A 7LRP NEAR 1RP LPI NEXT TO A 6 4000 4000 --. 3000 7;;- - --- 3000 l 2000 7_7t--- 7 O 2000 i g 1000 j g -1000 0 3 . _ _ _ 4 t) 8A [ -
$ 1000 uj
]
( h ,
. { y,, u_
"i.2000 ---- -
0 "_.' i._ k@ 1
-3000
~~~
k-1000 - ]'T " ( 4 -4000 4 5000 ~~.,,. O 2 4 6 10 0 2 4 6 8 10 TIME - 1sec TIME - msec A LRP NEAR EDGE AgFLANGE RING 180" 8 M A-3929-248 Figure 3.3-26 SM-5 VESSEL HEAD ACCELERATIONS AND DISPLACEMENT O 80
O l4 ~ I l l l l i 1.2 - ~ ~ ~ ~ SM 4 - SM 5 % IMPACT VELOCITY 10 - SM 4 < 62 4 f t/sec - j SM 5 a 62 2 f t/sec l o8 - w 3 o 06 -
<t Y /
V 5 04 -
/
/ -
/
/
/
0.2 -
/ -
o l l l I I I o 1.0 20 3o 4o TIME - msec M A -3929 -212 Figure 3.3-27 COMPARISON OF WATER SURFACE DISPLACEMENTS: SM 4 AND SM 5 O 81 i l
References for Section 3 3-1 C. M. Romander and D. J. Cagliostro, " Experimental Simulation of a Hypothetical Core Disruptive Accident in 1/20-Scale Models of the Clinch River Breeder Reactor", Technical Report 4, to be published. (Availability: SRI International, Menlo Park, Calif. ) 3-2 G. L. Fox, Jr. , and D. D. Stepnewski, " Pressure Wave Transmission in a Fluid Contained in a Plastically Deforming Pipe," Trans. Am. Soc. Mech. Engrs_. 96, Series J, pp. 258-262 (1974). 3-3 C. M. Romander et al. , "The Response of Water-Filled Pipes to Pressure Pulses", in Transactions of the 4th International Conference on Structural Mechanics in Reactor Technology, Vol. E, Paper No. E417, Commisson of the European Communities, Brussels, 1977. 3-4 M. J. Moneim and Y. W. Chang, " Comparison of ICEPEL Code Predictions with Straight Flexible Pipe Experiments," to be published in Nuclear Engineering and Design,1978. D. J. Cagliostro and C. M. Romander, " Experiments on the Response O 3-5 of Rigid and Flexible Reactor Vessel Models to a Simulated Hypothetical Core Disruptive Accident," Fifth Interim Report,1976 (Availability: SIR International, Menlo Park, Calif.) 82
l 4.0 . Analytic Simulations of the SM Series Tests O Extensive analysis was undertaken in parallel with the tests. This supporting effort was aimed primarily at verifying computer models and increasing confidence in the predictive powers of the analytic tools used to generate the SMBDB loads. Pre-test analysis was performed with the REXCO-HEP and ANSYS codes. This allowed for predictions of reactor system and vessel head responses respectively, unbiased by the test responses the codes are simulating. In addition, this analysis aided in test planning ar.d design. A , post test evaluation was then made to reconcile differences which were observed between the experimental results and the pre-test predictions. Post test analysis was also performed to assess the effect of using simulant materials (i.e., nickel and water) rather than prototypic materials (i.e., stainless steel and liquid sodium). Finally, the transient response of the cover gas was simulated using the MAXPRES code. This provided an assessment of the degree of conservatism which is usually inherent in the use of that code. O 4.1 Pre-Test Predictions 4.1.1 Comparison of REXC0-HEP Pre-Test Predictions with SM Series Test Results This section describes three pre-test analyses performed for the SM-2, SM-3 and SM-4 scale model tests. These analyses used the REXC0-HEP (Release 2} computer code (Ref. 4-1) and the input information available at the time the analyses were initiated. The SM-2 and SM-3 analyses used similar REXCO-HEP models while SM-4 analyses used a REXCO-HEP computer model approximating the more detailed SM-4 scale model test. 83
4.1.1.1 REXCO-HEP Models Figure 4.1-1 shows the REXCO-HEP model used for the SM-2 analvsis and Figure 4.1-2 shows the same for SM-3. Also shown on these figures are the approximate locations of strain gauges and pressure sensors used in the experiments. There are minor differences in the head stiffness and vessel length in the two models that are a result of the evolutionary improvement in the pre-test information. The SM-3 values are the more exact values. The figures also show a difference in the initial grid structure for grids between the core barrel and the reactor vessel . The slanted grid structure was found to be more efficient and required less rezoning but did not affect the numerical results. The major difference in these two test analyses was the inclusion of the upper internals structure (UIS) in SM-3. Fluid-structure interaction associated with this component cannot be directly modeled in the REXCO-HEP computer code. Modelling of the UIS as a hydrodynamic mass was thus used in the SM-3 pre-test analysis. This consisted of simply replacing the material in the 20 zones (5 zones high by 4 zones wide) immediately above the top of the core barrel with a material of uniform density that made this region equal in mass to the upper internals structure plus the water mass in the region. The other properties, such as the bulk modulus, were modified appropriately to model a mixture of steel and water. Figure 4.1-3 shows the model used in the SM-4 pre-test analysis. The dimensions and inlet plenum detail corresponded to the SM-4 information available at the time the analysis was performed. A grid structure similar to the SM-3 model was incorporated along with the scaled core support structure. The properties data used in the models were derived from standard sources for materials data (Refs. 4-2, 4-3, 4-4 and 4-5). The nickel and the pressure source term are the only two exceptions. Table 4.1-1 summarizes the properties input to REXCO-HEP. The stress-strain curves for the nickel material used to model the stainless steel in the full scale system 84 O
_ _ . _ ~ _ _ _ . _ -- _ _ ___ __ __ _ _ . - _ _ - _.__ - were obtained from low strain rate (0.001 in/in/sec) tests on samples from the material used in the test models. The stress-strain curves for these data are shown in Figure 4.1-4 along with the points input to ,
, REXC0-HEP to represent the material. The calculation uses the piecewise linear lines shown in Figure 4.1-4. These curves are engineering stress vs. strain whereas REXC0-HEP uses true stress and true strain. The curves shown were used on the basis that for the anticipated strains the l difference between the two types of stress versus strain would not be j significant.
l The pressure-volume relationship for the PETN explosive charge was modeled on the basis of SRI pre-test development work. This curve is l identified as the " expected P-AV" curve. Figure 4.1-5 shows the pressure as a function of volume change and compares it to the ideal 1/20 scale SMBDB curve. The SM-2 and SM-3 pre-test analyses were performed using I the expected P-AV curve for PETN because it was the best information available at the time the calculations were initiated. The tests were actually run with a 19.7 g charge closely matching the " calibrated P-aV" curve. The SM-4/SM-5 pre-test analysis was performed with both the expected l P-aV (20.7 g) and the calibrated P-AV (19.7 g). As shown later the differences in the REXC041EP analyses resulting from the expected and cal-ibrated P-AV curves are minor. The SM-2 pre-test prediction described in the next section is thus based on the expected curve while the SM-4/SM-5 pre-test prediction is based on the slightly more accurate calibrated curve. 4.1.1.2 Pre-Test Analysis Results The pre-test analysis results will now be compared on a selective basis with the test measurements. The comparison is selective because considerably more information is computed than is measured. Appendix A provides a more comprehensive set of experimental versus computational data. curves than are discussed in this and in the following sections. Curves from Appendix A that are useful to the discussion are duplicated in this and other sections. Reference to Appendix A is only necessary if greater data detail is desired. O 85
- - _ _ . . . ~ . . . _ _ . , _ _ . _ . _ - . , _ _ _ _ . -_ .. _ _ _ . _ -
The most meaningful and direct comparison of the experiments and calculations can be made using the reactor vessel and core barrel final hoop strains. I In these comparisons the calculated strains are those that exist at the completion of the REXC0-HEP run (s6 msec). In the case of the core barrel, strains computed earlier in time were somewhat larger than those present at the completion of the run. Those computed at the end of the run are deemed the more appropriate analog of the post-test profile l 1 measurement. In the case of the reactor vessel the pressure is still abov'e : 100 psi at the completion of the calculation whereas in the corresponding experimental situation, the residual pressure is muc', lower with the result that more elastic spring-back will have occurred. Figures 4.1-6, 4.1-7 and 4.1-8 compare the calculated and measured SM-2, SM-3 and SM-4 vessel and core barrel strain profiles. As a result of conservative assumptions in the REXCO-HEP code (for example, lack of energy dissipating mechanisms), the predicted strain profiles generally envelope the corresponding experimental profiles. The differences between the predicted and experimental profiles can be explained by examining the l loadings within the reactor vessel. In the paragraphs below, the pressure loads causing the strains in the core barrel and in the different sections 1 of the vessel wall are discussed, j l From Figures 4.1-6, 4.1-7 and 4.1-8 it is evident that REXC0-HEP tends to overpredict core barrel strains. This results from four known , e f fec ts . First, REXC0-HEP models the core shielding as a hydrodynamic medium having no structural resistance. Thus the loads are transmitted to the core barrel more readily than in the experiment and greater core barrel expansion resul ts. The more rapid initial expansion of the core barrel leads to more rapid decay of the core region pressure. This can be seen in Figure 4.1-9 which compares the SM-2 predicted and experimental core region pressure histories. Comparisons for SM-3 and SM-4 are very similar. 86
Second, because of the particular Lagrangian mesh formulation used in the Release 2 version of REXC0-HEP, there is a coupling between the shielding V and core region zones which resists upward movement of the bubble-liquid interface. This resistance to upward fluid movement tends to retard bubble expansion and thus adds to the overprediction of core barrel 1 deformation. Overall bubble expansion is however dominated by the first of the two effects above and hence the predicted core pressure still decays more rapidly than in the test. Third, while no experimental strain histories of the core barrel are available, REXC0-HEP pre-test calculations indicate core barrel strain rates of up to 95 per second. Thus the strain rate effect is likely to strain harden the core barrel material as a result of its rapid dynamic response. This effect was not accounted for in the pre-test analysis. Fourth, the computed core b.crel deformations are also influenced by the wave reflected after slug impact. Because of the lack of energy dissipating effects in REXC0-HEP, the reflected wave, which travels into the annulus between the vessel wall and core barrel, is more intense in the calculations than in the experiments. The differences between the analytic and experimental post-impact pressures (i.e. , beyond three milliseconds) can be seen in Figures 4.1-10, 4.1-11 and 4.1-12. These differences result in the analysis predicting some reverse defonnation of the core barrel while the tests do not. The magnitude of this effect for SM-4 is discussed in the post-test analysis section. Consider now the reactor ves',el mid-wall response (i.e., from the core support cone to the elevat9n midway between the upper core barrel and the vessel head) and the associated loadings. From Figure 4.1-6 it can be seen that the SM-2 pre-test prediction significantly underestimated the vessel strain around the elevation of the outlet nozzle, but this discrepancy is much less noticeable for SM-3 and SM-4 (Figures 4.1-7 and 4.1-8). The relative magnitudes of these deformations are seen to be consistent with the relative magnitudes of the corresponding vessel wall pressure loads before 2 milliseconds, as shown in Figures 4.1-10 through 4.1-12.
.r 87
The early loading on the SM-2 vessel is considerably more severe than in the pre-test prediction. As discussed in the following sentences this results from the same considerations affecting the core barrel response. Because of the more rapid expansion of the core barrel in the pre-test analyses, less liquid is initially ejected upwards than in the test. Also the non-prototypic coupling between the shielding and the upward moving fluid tends to retard the fluid movement. A similar coupling is also present between the inner and outer fluid zones above the core barrel and helps retard the rapid upward movement of the liquid ejected from the core barrel. Figures 4.1-11 and 4.1-12 compare the predicted and experimental mid-vessel pressures for SM-3 and SM-4 respectively. Early in these transients, the predicted and experimental results are in much closer agreement than the predicted and experimental SM-2 results. This agreement results in the favorable vessel wall strain comparison at the mid-vessel elevation. Since REXCO-HEP does not model the. energy dissipating mechanisms in the liquid, the longer term calculated pressures (i.e., from 3 to 6 milliseconds) tend to be higher than the experimental values. These later loadings produce little additional vessel strain, since the early loadings have a strain hardening effect on the structures. In the SM-3 test, the UIS plays a major role in dissipating energy through the creation of fluid turbulence and produces the significant reduction l in the early vessel wall loadings. The addition of the hydrodynamic UIS in the SM-3 model, however, does little to change the calculated fluid ! response. The final effect of the above competing processes is that remarkably good agreement between test and calculated mid-vessel loadings occurs when the UIS structure is included in the model. Agreement is also excellent between the SM-4 pre-test prediction and experimental results early in the transient. The higher load at 0.5 milliseconds on the SM-4 vessel wall results from the stiffening effect of the added thermal liner. 88
1 l The predicted upper vessel wall hoop strains are all greater than the corresponding test values. Specifically, the experimental peak vessel hoop strains for SM-2, SM-3 and SM-4 are 4.4, 2.8 and 1.7 percent respectively 4 l while the corresponding predicted peak vessel hoop strains are 6.4, 6.8 and i 4.0 percent respectively. The overprediction are partly due, in SM-2 and SM-3, to the higher predicted upper vessel wall loads as shown in Figures 4.1-13 and 4.1-14, and the fact that the REXCO-HEP model did not account for material strain rate effects. However, the SM-4 predicted and experimental upper vessel wall loads are comparable (see Figure 4.1-15) and it is unlikely J that strain rate effects are responsible for the fact that the predicted hoop strain is about 2.5 times the measured strain. In addition, on approaching the vessel flange, the experimental strain profile flattens off gradually to a zero gradient while the predicted strain profile makes i an abrupt change in gradient at the flange. These observations tend to indicate that REXC0-HEP is not fully accounting for the bending stiffness of vesse.1 shells adjacent to structural discontinuities. This conclusion is supported by similar observations of the predicted and experimental strains in the lower vessel section of SM-4. This is discussed later in the section. By comparing the predicted and experimental responses of SM-2 and SM-3 at the upper vessel wall, it is evident that the approach used to model the UIS in REXC0-HEP does not fully account for energy dissipating mechanisms introduced by the UIS. While the SM-3 vessel wall hoop strain is about 36 percent less than the SM-2 value, the predicted SM-3 strain increased over I that of SM-2 by about 8 percent. This resulted because the hydrodynamically modelled VIS in SM-3 produced a fluid slug having more impulse than in the case of SM-2 where no VIS was modelled. In addition, dissipative fluid turbulence created by the actual VIS was not generated by the simulated VIS. Thus, in the REXC0-HEP model used to generate the SMB08 loads, considerable conservatism in the system response is built in as a result j of not including the dissipative effects associated with the UIS. O 89
1 l Figures 4.1-16, 4.1-17 and 4.1-18 compare the SM-2, SM-3 and SM-4 experimental ._nd predicted central head loads. The SM-2 comparison shows good agreement in loads although the experimental peaks occur earlier. This results because of the limitation in modelling in-core fluid slippage,which as discussed earlier, is inherent in the version of REXCO-HEP used. The SM-3 predicted peak load is significantly higher than that for SM-2 l because of the increased slug impulse associated with the hydrodynamic UIS. From these upper vessel and vessel head predicted responses, it is apparent that the approach used here to analytically model the UIS results in greater loads to the head and upper vessel. The test results indicate that a prototypic UIS is likely to have the reverse effect reducing head and upper vessel loads. While the hydrodynamically i modelled UIS increases the slug kinetic energy through the addition of higher density fluid above the core, the prototypic UIS would have the effect of reducing slug axial kinetic energy through generation of fluid turbulence. Where comparisons can be made, the SM-4/SM-5 pre-test analysis shows ! similar trends to those of the SM-2 and SM-3 pre-test analyses. However, unlike the SM-2 and SM-3 test models, the SM-4 and SM-5 test models included scaled core support plates and bottom heads. For these components, the SM-4/SM-5 pre-test analysis predicted considerably more strain than measured in the experiments. In both the SM-4 and SM-5 tests, residual downward displacement of the centers of the core support plates relative to the edges was approximately 0.1 inch. However, in the pre-test prediction, the corresponding displacement was 0.31 inch. In addition, as shown in Figure 4.1-8, the predicted peak residual hoop strain in the vessel wall was about 1.8 percent while the measured strain was only about 0.2 percent. 90
These over-predictions may result from two effects. First, the core support structure was modelled as a solid elastic-plastic nickel plate supported by a cone-shaped shell attached to the reactor vessel. The density and yield stress of the plate were adjusted in an attempt a to model the flexibility of the actual core support plate, but the resulting stiffness of the analytic model may have been too low. Second, the residual hoop strain of the lower vessel wall (0.2 percent) is much less than would be expected from the applied inlet plenum pressure loading of about 400 psi (see Figures 4.1-8 and 4.1-19). This conclusion results from comparing similar pressure loadings and correspondin9 mid-vessel wall hoop strains measured in SM-2 and SM-3. (With the thermal liners of SM-4 and SM-5, mid-wall loading and resulting hoop strains are not representative of the lower vessel wall which has no thermal liner.) For example, with a less severe loading of about 350 psi on the SM-3 mid-vessel wall (pressure gauge 5P in Figure 3.3-9), the vessel wall hoop strain is about 1.5 percent, which is much higher than the measured hoop strain in the SM-4 and SM-5 lower vessel walls. These results indicate that not only the applied pressure but also the boundary constraints are important in determining the final vessel wall hoop O v strain. It is thus likely that the added hoop stiffness of the cone-to-vessel junction and the vessel bottom head are partly responsible for the reduced vessel wall strain measured in the SM-4 and SM-5 lower vessel walls. This reduction is not very evident in the REXC0 SM-4/SM-5 pre-test prediction and it is possible that the code under-estimates the vessel bending stiffness resulting from the cone-to-vessel junction and the vessel bottom head attachments. This is consistent with the earlier discussion on upper vessel wall response. The responses of the lower vessel wall and core plate are likely to be strongly coupled through the inlet plenum fluid and their relative importance in assessing fluid pressurization has not been determined. /O 91 V
As m'e ntioned previously the SM-4 pre-test analysis was performed with both the expected P-aV and the calibrated P-AV curves. Table 4.1-2 compares several significant parameters for both of these cases. It can be seen that there is not a major difference in the consequence from these two pressure sources. The higher peak pressure in the expected P-av curve (290 bars) contributes higher axial kinetic energy and slug impact energy. This in turn causes higher impact loads and slightly higher (15 percent) reactor vessel peak strain. The fact that the pressure drops below the calibrated P-AV curve (peak pressure 265 bars) with only a slight change in volume and has less energy (/PdV) in the l pressure range above 50 bars causes a reduction in core barrel strain. The peak core barrel strain is slightly (2.5 percent) greater for the calibrated curve than for the expected curve. The pre-test analyses for i SM-2 and SM-3 would be affected in a similar manner by the use of the l expected P-AV. There are several integral parameters that can be derived from the j experiments, such as kinetic energy of the water slug above the core barrel, and strain energy for the core barrel. These parameters are also computed or can be derived from the REXCO-HEP calculation. Table 4.1-3 l summarizes se.veral of these parameters and compares those derived from the experiment to the REXCO-HEP calculated value. The differences are consistent with the more detailed results. The experimental values in this table are derived from measured experimental data and Reference 4-7 should be consulted to understand the assumptions made in the derivations. In general the assumptions are reasonabic and do not introduce any major error; however, it should be understood that the values identified as experimental are not measured. The experimental values are those present after slug impact and thus include the strain energy in the reactor vessel that occurs on slug impact. The core barrel strain energy is the same before and after slug impact in the case of the experiment while this is not the case for the REXCO-HEP calculation. The value of strain energy reported for the core barrel is taken at a time following slug impact but prior to the distortion that occurs from the impact of the reflected wave. This value is judged to be a better value for the purpose of this comparative table. 92 O
4.1.1.3 Conclusions from Pre-Test Analysis From a comparison of the pre-test predictions with the experimental () results (see Figures 4.1-6 through 4.1-19) the following conclusions can be made: e Because of the general lack of energy dissipating mechanisms in REXCO-HEP, the predicted pressure loads and resuli.ing strains are
.predominantly conservative, as would be expected. The pressure load with its 'as-sociated hoop strain at the vessel mid-elevation for SM-2 is the major-exception to this , as discussai in thn le t paragraph in this section.
e The hydrodynamic modelling used to simulate the UIS increased upper vessel and vessel head loads. In the tests, fluid turbulence generated by the UIS reduced these loads. Although hydrodynamic modelling of the UIS is thus not recommended it does result in conservative upper vessel and head loads. e REXCO-HEP overpredicted hoop strains in the vicinity of structural discontinuities (e.g., at the upper vessel wall next to the (n) flange and at the lower vessel wall between the bottom head and the core support ring). This may result from insufficient shell bending stiffness as modelled in REXCO-HEP, e The 1/20 scale model core barrel and upper vessel wall respond sufficiently rapidly that strain rate effects should be included I in the definition of material properties. For the prototypic i case, vessel wall and core barrel strain rates are likely to be too low to significantly alter material properties. I e The materials properties and modelling used in the analysis to characterize the core support plate and cone produced too flexible a response in these components (SM-4/SM-5 only). This will lead to higher inlet plenum pressures but will make little difference to upper vessel response. N (d 93
e Hydrodynamic modelling of the in-core shielding does not account for the structural resistance of the shield rings and leads to too rapid an expansion of both the shield region and the core barrel. This results first in too rapid decay of in-core pressure and second, in reduced energy being imparted to the liquid slug, o In REXCO-HEP, the Lagrangian zones associated with the high-pressure vapor source are attached to the hydrodynamically modelled shielding zones. The resulting constraint on the vapor bubble expansion retards the liquid slug, reducing the loading on the mid-vessel wall, and delaying slug impact with the vessel head. The effects discussed in the last two conclusions above both tend to reduce the loading on the mid-vessel wall, and in this sense are non-conservative. However, other modelling assumptions, such as those associated with the UIS more than compensate in the conservative direction. The considerations discussed in the last two conclusions are examined more fully in Reference 4-6. 4.1.2 Comparison of ANSYS Vessel Head Predictions with the Test Results In conjunction with the REXCO-HEP analyses of the reactor system responses ANSYS (Reference 4-10) analyses were performed to assess the structural response of the SM-2 and SM-4/SM-5 vessel heads. The SM-3 head was not analyzed since its response was considered to be similar to, but somewhat less severe than, that of SM-2. As with the REXC0-HEP analyses, these vessel head analyses served both to support test planning efforts and to provide confirmation of this approach in assessing dynamic head response. 94
m 4.1.2.1 The ANSYS Models t The ANSYS code is based on the finite element method. This allows the l construction of a composite vessel head model consisting of a set of l linked elements each of which has one or more of the physical properties l necessary to simdlate the dynamic response of the head. The SM-2 and l SM-4/SM-5 head analytic models are constructed entirely of dynamic elements l which include spring, slider, mass, damping and gap components in their l formulation. l
. \
It should be emphasized that these simplified lumped mass models do not I have the degree of detail (e.g. in head plug bending) necessary for exact l simulation of the tests. The objectives of usir; 5ese simplified piodels are (1) to obtain pre-test information on the likely response of the experimental models particularly in the margin ring region, and (2) to assess in an unbiased manner the ability of such simplified models to simulate the dynamic response of single and multi-plug vessel heads. m U 0 The SM-2 Vessel Head Model 1 1 Figure 4.1-20 shows the lumped mass model used for the SM-2 head analysis. ' This model incorporates a simplified vessel wall representation to include the effect of the downward core support structure load. In the group of elements which, for example, represent the region around the margin ring, the spring and l slider elements result in an elastic-perfectly plastic response. The gap element is used to simulate the margin ring gap closure response while , the dashpot accounts for local damping. Table 4.1-4 defines the connectivity between the head components, and in addition, defines the inass and physical l l i l
)
l r l 95 l 1 i l
constants used in the model. The stiffness of the head in the vicinity of the margin ring was derived from a set of static experiments performed to assess this and other head properties in the margin ring region. The damping coefficients reflect a 2 percent of critical structural damping and a 20 percent of critical impact damping. 6 The SM-4/SM-5 Vessel Head Model The SM-4/SM-5 vessel head model is constructed in a manner very similar to that of SM-2, although more elements are included to account for the three head plugs, in addition, it accounts for the upper internals structure and the flexibility of the large and intermediate rotating plugs (the small plug is very stiff). The UIS columns were modelled in a simplified elastic perfectly plastic manner, buckling taking place at the effective yield point. The force required to yield of the four columns was 8560 lbs. Figure 4.1-21 shows the resulting ANSYS model while Tables 4.1-5 and 4.1-6 show the element connectivities and physical constants respectively. For SM-4/SM-5, 2 percent of critical damping was used for structural response and approximately 10 percent of critical impact damping was used. 4.1.2.2 The Pre-Test Analysis Results 0 The SM-2 Vessel Head Response To assess the SM-2 vessel head response, the SMBDB head force, reduced to 1/20th scale, was used since at the time the head analysis was performed, ) the SM-2 REXC0 calculation had not been completed. Figure 4.1-22 compares ) the scaled SMBDB with the SM-2 REXCO-HEP head load and it can be seen that l in terms of impulse, the histories are not significantly different. For the l purpose of the comparison, the SMBDB pressure spike has been aligned in time to a position coincident with that of the SM-2 pressure spike.
)
TheSMBDBcoresupportplate(CSP) load,reducedto1/20 scale,was l applied to node 4 of the model. A comparison of this load ith the 1 96
l l l REXCO-HEP calculatea SM-2 CSP load is shown in Figure 4.1-23. While the comparison is not as good as with the head load, the CSP load is of ! p secondary importance as far as head response is concerned and as such the f U scaled SMBDB load adequately represents the predicted load. l Figures 4.1-24 through 4.1-27 show results of the analysis. In Figure 4.1-24 it can be seen that the peak total force on the margin ring is 1.77x105 lbs. From the static force deflection experiments discussed I earlier, the elastic limit for the full-scale large rotating' plug is predicted to be 99.2 x 10 l6 bs . As a result, the elastic limit on the 1/20th scale SM-2 model is: l 6 99.2x10 /(20)2 = 2.48x10 lbs Thus no residual plastic deformation was expected in the region of the SM-2 margin ring. Results of the SM-2 test confirmed this conclusion. This is a particularly severe test of the LRP margin ring since the stiff one-piece head will transmit a more severe load to this juncture than would occur with the more prototypic head configuration. Figure 4.1-25 shows the head upward displacement close to the margin ring while Figure 4.1-26 shows the associated head acceleration and compares it to the experimental value. In this comparison, the calculated slug impact peak has been shifted back in time by 1.3 msec in order to align it with the experimental peak. As can be seen the respone.es show quite similar trends. The initial experimental peak is less than the calculated peak (1250 gs against 2200 gs). However, because of r' - flexibili ty, the experimental peak at the center of the head is o-(4800 gs) than at the edge. Since the model simulates the head a ed mass, it cannot predict this central acceleration. In addition, s seen that the experimental result exhibits more high frequency res, than does the predicted response, possibly due to effects such as pl, l bending or non-asymmetric response which is not accounted for in the analytic model. 1 l
Figures 4.1-27 compares the analytic and experimental acceleration histories , at the vessel flange. As with the head, the time base for the calculated flange response has been shifted back to align slug impact times. It can be seen that the flange accelerations match quite well except that the analytic response does not exhibit the higher overtones seen in the experimental response. Table 4.1-7 summarizes the analytic data and compares them to the experimental data where available. While the peak accelerations do not appear to correlate too well, it should be remembered that the extremely narrow high frequency spikes in the test results are beyond the resolution of the analytic model. This is not critical since the energy associated with the peaks of these spikes is generally small, o The SM-4/SM-5 Vessel Head Response 1 1 To assess the response of the SM-4/SM-5 vessel heads, five b d histories were applied to the SM-4/SM-5 model (F1 through F4 and F6 shown in Figure 4.1-21). ' ese loads, obtained from the SM-4/SM-5 pre-test analysis discussed in Sect .n 4.1.1, were (1) the downward force applied to the core support plate, F6 (Figure 4.1-29); (2) the upward force applied to the underside of the UIS, F4 (Figure 4.1-30) and (3) the slug load applied to the head (Figure 4.1 78). Th#s latter load was divided between the three head plugs on an i basis to give forces F1 on the SRP, etively. F2 on the IRP and F3 on Figures 4.1-31, 4.1-32 ant , m. . mpare the LRP, IRP and SRP experimental and predicted acceleration responses for SM-5. Unlike the corresponding plot for the SM-2 head response, the slug impact peaks for the calculated response were not aligned (in time) to the experimental peaks. From these figures, it can be seen that the best agreement occurs between the SRP experimental and predicted responses while the worst agreement occurs between the corresponding LRP responses. This results because the rigid plate analytic model of the SRP is a good representation of the actual plug. However, the simple analytic model of the LRP cannot represent the complex flexural response modes of the LRP. An explicit 98
three-dimensional model of the head may thus be necessary to pick up the more complex LRP response. The accuracy of the predicted IRP acceleratfon response lies somewhere between those of the LRP and SRP. However, if the analytic IRP response is integrated to give the plug displacement history, the resulting profile is found to follow, although at lower amplitude, the exper,imental SM-5, IRP response. This can be seen in Figure 4.1-34 which compares these results. The experimental curve is taken from that location on the plug which experiences highest deflection. Thus use of such an analytic model would appear to result in a conservative prediction of head doming. Figures 4.1-35, 4.1-36 and 4.1-37 compare the experimental and analytic acceleration frequency spectra for the three plugs. These spectra were derived using a Fast Fourier Transform of the acceleration histories discussed above, and from them, three conclusions can be drawn. First, both the analytic and experimental data indicate all plugs have fundamental excitation below 500 Hz. Second, the IRP and SRP responses indicate primarily rigid body motion. This conclusion is drawn since the experimental frequency spectra for these plugs follow quite closely the predicted spectra in which the flexural modes of excitation are not modelled. Third, the experimental LRP response shows significantly greater higher-frequency content than does the experimental response. This indicates a quite complex plug response which is not found in the analytic model. The SM-1 test indicated that the LRP flexibility was significantly greater than the flexibility of either the IRP or SRP. This is consistent with the conclusion drawn from these spectral responses. Figures 4.1-38, 4.1-39 and 4.1-40 show the predicted total forces on the 5 4 margin rings. The peak forces are 2.25 x 106 , 1.25 x 10 and 4.72 x 10 lbs across the LRP, IRP and SRP margin rings, respectively, while the local yield forces at the corresponding margin ring junctures are 2.0 x 5 5 4 10 , 1.75 x 10 and 5.25 x 10 lbs*, Hence the analysis predicted some slight plastic deformation at the LRP juncture while none was predicted at the IRP and LRP junctures. In the actual SM-4/SM-5 tests no plastic deformation.was observed at any of the margin ring junctures.
*These effective yield forces were taken from finite-element analyses to determine the static force-deflection characteristics of the margin ring O, junctures. The results of this analysis are consistent with the results of the experimentally derived force-deflection characteristics discussed with respect to the SM-2 test.
99
Table 4.1-8 summarizes the information above. As in the SM-2 comparison, the pr'edicted peak head accelerations do not correlate too well with the experimental values, since the experimental peaks are extremely narrow and reflect very complex plug interactions. However, as discussed, the predicted IRP displacement history was similar in shape, though somewhat greater in amplitude than the experimental values from SM-5. In addition the pre-dicted frequency response spectra of the pit.g accelerations showed similar characteristics to the experimental spectra. 1 4.1.2.3 Conclusions from Pre-Test ANSYS Analysis From a comparison of the pre-test predictions of head response with the experimental results, the following conclusions can be drawn: e The response characteristics of the SM-2 head resulting from the l pre-test analysis were in good agreement with the experiment. The l predicted peak acceleration resulting from slug impact was, bewever, significantly higher than the experimental value (see Figure 4.1-26). e No residual deformation was predicted around the SM-2 margin-ring and none was found in the experiment. e The predicted responses of the SRP and IRP of the SM-4 and SM-5 tests matched the experimental histories in their general characteristics. However, because the analytic model did not account for the flexibility of the LRP, the simulation was unable to generate the higher frequency response of this plug (see Figures 4.1-31 through 4.1-37). e While the SM-4/SM-5 pre-test analysis predicted some alight plastic deformation'at the LRP juncture, the tests showed no plastic deformation. 4.1.3 Assessment of Differences Between the Predictions and Test Results A post-test quantitative and qualitative assessment of the importance , of some of the input factors for the REXC0-HEP analysis is provided in this section. The predicted load on the vessel head changes slightly as a result of this re-analysis. Hence an updated vessel head analysis, using the ANSYS head model, was performed and is also reviewed in this section. 100
8 The REXC0-HEP Re-Analysis Section 4.1.1 compared the SM-2, SM-3 and SM-4 pre-test predictions , with the test results. Several factors were identified as having contributed to the differences between the analysis and experiments. These factors, as associated with REXC0-HEP modelling, were: (1) neglect of strain rate effects, (2) hydrodynamic modelling of the UIS, (3) hydrodynamic modelling of the core barrel shielding, (4) artifically strong coupling between hydrodynamic core barrel shielding and the pressurizing gas, (5) insufficient accounting of shell bending stiffness at geometric discontinuities and (6) limitations in core support plate modelling. The first factor involves an input data assumption while the last five primary involve code limitations. Later in this section, the importance of strain rate effects on the core barrel and vessel wall responses of SM-2 and SM-4/SM-5 are discussed. With respect to the second factor, the Lagrangian formulation of REXCO-HEP code makes modelling of orifice flow through the UIS virtually impossible, and it has thus not been possible to address this assumption through the use of REXC0-HEP. Code modifications are required to address factors O 3 and 4. Assessing these effects is beyond the scope of this report since the analyses are being used to validate the mathematical model in the version of REXC0-HEP which was used to generate the SMBDB loads. However, the effect of these factors is addressed through the use of a modified version of REXC0-HEP as discussed in Reference 4-6. This report showed, for example, that with improved assumptions with respect to factors 3 and 4, the mid-vessel wall loading and corresponding hoop strain in SM-2 matched the experiment much more closely. Factors 5 and 6 have not as yet been addressed quantitatively in any detail. In the post-test SM-2 and SM-4 analyses performed to assess the first factor,_ some additional minor input changes were also made. These minor changes were made with the objective of making the analysis more consistent with the as-tested geometry and materials properties of 1 01
the scale models. Their total effect is not believed to be very significant to the analysis. The changes that were made for the post-test SM-2 analysis can be summarized as follows:
- 1. Stress-strain data at a strain rate of 100 in/in/sec were used for the core barrel, core support cone and reactor vessel rather than the data obtained at 0.001 in/in/sec.
- 2. The calibrated P-AV curve was used rather than the expected curve (see Figure 4.1-5).
- 3. Water level was changed from 23.6625 in to 24.9151 in.
- 4. Air space between the water surface and head was changed from 1.3675 in to 1.35 in.
- 5. The bulk modulus for the region above the core containing 2
lead shot was changed from 2.3488 x5 10 lb/in to 5 2.839 x 10 lb/in . The partial derivative of the bulk modulus with respect to pressure was changed from I 6.749 to 8.51. l The first two changes are the most significant. The stress-strain properties for the high and low strain rate will bound those properties actually experienced by the Ni - 200 (50 to 100 in/in/sec). With the limited strain rate data available on this material, this approach was felt to be most appropriate. Figure 4.1-41 shows the stress-strain curve for nickel 200 used in the reactor vessel at a strain rate of 100 in/in/sec and compares it to the low strain rate curve and the piecewise linear curve used in the REXC0-HEP calculation. A similar adjustment to the stress-strain properties for the core barrel material was made in the post-test analysis. The higher yield stress associated with the high strain rate reduces the maximum deformation in both the core barrel and the reactor vessel . Figure 4.1-42 shows the deformation profile for the reactor vessel and core barrel for the SM-2 pre-test and post-test analysis. The core barrel comparison is shown at two times. One is prior to the 102 9
return of the reflected pressure pulse and the second is after this c pulse has acted on the' core barrel and corresponds to the time of final reactor vessel deformations. The higher strength resulting from the high strain rate properties is seen to reduce both the maximum core barrel strain early in time and the degree of reverse deformation caused by the computed reflected pressure pulse. As a result, the final strain of the core barrel is greater than in the low strain rate analysis. The significance of the change in the P-AV curve was discussed in the pre-test analysis of SM-4 in Section 4.1.1. As mentioned in that section, this , change would increase the core barrel strain (by about 2.5 percent) and decrease the vessel strain (by about 15 percent). The change would be expected to have the same impact in SM-2. Changes 3, 4 and 5 have not been evaluated separately but they are expected to have an even smaller effect than the P-AV change. The correction for materials properties does improve the correlation between the SM-2 analysis and test results, especially in the upper reactor vessel region. However, there are still some differences that result from the other factors discussed, particularly those addressed in Reference 4-6. Further details of the SM-2 post-test results can be j I found in Appendix A while details of the REXC0-HEP analysis can be found in Reference 4-8. The model for the post-test analysis of SM-4/SM-5 was the same as for the pre-test analysis with the calibrated source term except that the item 2 and item 5 changes made in the SM-2 post-test analysis were also made in SM-4/SM-5 post-test analysis. Figure 4.1-43 compares the SM-4/SM-5 pre- and post-test deformation profiles. The degree of improvement in the analysis is similar to that in the SM-2 comparison. The other modeling assumptions discussed earlier will mainly be responsible for the continued differences. In particular, the lack of proper . 103
~
characterization of the UIS is likely to be most significant and will contribute to a large share of this continuing difference. Over-prediction of the deformation of the upper reactor vessel is still present in SM-4 whereas in SM-2 the change in material properties brought the experiment and analysis into close agreement. This difference is attributed to the limited ability of REXCO-HEP to model the UIS structure and its effects on fluid flow. Thus, changing the material properties in the reactor vessel would not be expected - to completely resolve the discrepancy in upper vessel deformation in the SM-4 analysis as it did in the SM-2 analysis. Over-prediction of strain in the core barrel and the reactor vessel at the core elevation is still substantial. The failure to properly model the shield region is apparently a major source of this discrepancy and changing the core vessel material properties only produced a small improvement. The changes in material properties for the thin-shell sections of the core support structure did not affect the core support structure response, and the core plate deflection is still excessive. The deflection is almost identical with the pre-test analysis , which is to be expected since the core plate modeling is unchanged. In suninary, the changes in the stress-strain properties for shell structures and correction of other minor discrepancies in the pre-test analysis improve the comparisons of the analytical results with the test results. The apparent causes of the remaining discrepancies have been identified and partially addressed through Reference 4-6. It is unlikely that the Lagrangian method used in REXC0-HEP could be modified to properly account for the effect of the UIS. 104
Further details of the SM-4/SM-5 post-test results can be found in' O Appendix A while details of the post-test REXCO-HEP calculation can be O 'found in Reference 4-9. O ANSYS Vessel Head Re-Analysis The SM-4/SM-5 vessel head re-analysis used the finite element ANSYS model discussed in Section 4.1.2. The model was not modified since the material properties used previously were still appropriate. The only changes made were to the loads applied to the under-head shielding, the core support structure and the upper internals structure, these being derived from the post-test REXCO-HEP analysis. Figure 4.1-44 compares the pre-test SM-4 head load with the post-test load used in the re-analysis. The post-test loads resulted from a REXCO-HEP simulation of the SM-4 test similar to that discussed at,ove but with liquid sodium properties substituted for water, and stainless steel properties substituted for nickel . As shown in Section 4.2.3. where these loads are discussed, (see Figures 4.2-5, 4.2-6 and 4.2-7), the substitutions make little difference on the pressure profiles. U Table 4.1-9 summarizes the changes in response from the pre-test and post-test analysis. From this table several conclusions can be d rawn . First, the minor amount of plastic defonnation at the juncture of the large plug-to-vessel flange disappeared, making the result more consistent with the SM-4/SM-5 tests. In general, head response (e.g., displacements, forces and accelerations) was slightly less severe in the post-test analyses while vessel wall response was slightly more severe. The more severe response of the vessel wall and also the vessel flange and flange bolts is thought to be due to the change in frequency content of the head load. After slug impact, the post-test head load in Figure 4.1-44 shows a pronounced oscillation around 1000 Hz. This is not present in the pre-test loading. A 105 ()
Since the natural frequency of the vessel flange on the flange bolts is atrout 1050 Hz, it appears that the flange and bolts resonated with the slug force. As a result, the amplitude of the flange motion ' increased, increasing the strain in the bolts and lifting the flange off the support ledge. When the flange was arrested by the ' edge on its return, the inertia of the vessel induced greater axial deformation of the vessel wall. This distinct 1000 Hz resonance does not appear in tne experimental head load since the underhead pressure decays too rapidly for sustained pressure oscillations. As shown in Figure 4~.1-45, head response, typified by the SRP acceleration, is not unduly affected by the change in applied load. In conclusion, the use of strain-rate dependent material properties produced structural deformations in the vessel wall and core barrel which were more consistent with the experimental values. In addition, use of the post-test REXCO-HEP loading on the vessel head removed the slight plastic deformation previously predicted at the LRP juncture, thus making the results of the head analyses more consistent with the experiment. O 106
t 1 Table 4.1-1 lO Properties of Materials Used in REXC0 Pre-Test Analyses of the SM Tests Density Bulk Modulus Pressure Derivative Material. (lb/in3) (Psi) of Bulk Modulus 5 Water .036 2.89 x 10 6.9 7 Nickel .321 3.57 x 10 4.87 Steel .284 1.86 x 10 4,0 6 lead .410 4.76 x 10 6.0 6 Stainless Steel .284 8.19 x 10 1.0 5 Sodium .030 6.24 x 10 7.67 107 0
1 1
. 1 l
Table 4.1-2 Comparison of SM-4 Analysis Using Expected And Calibrated P-AV Curves l l l SM-4 with SM-4 with l COMPUTED PARAMETER Expected P-aV Calibrated P-oV Peak' Radial Kinetic Energy, Joules 3 2.3x10 3 2.4x10 (time, msec) (.5) (.5) l 3 3 l Peak Positive Axial Kinetic Energy, Joules 8.0x10 7.3x10 (time, msec) (3.2) (3.3) l Peak Slug Velocity, f t/sec 62. 56. (time, msec) (3.1) (3.2) Peak Slug Force, lbs 5 5 4.0x10 3.1x10 (time, msec) (3.2) (3.3) l Peak Head Total Energy, Joules 2 3.5x10 2.7x102 (time, msec) (5.3) (5.7) l l Peak Vessel Strain, % 4.7 4.0 (time, msec) (4.7) (4.8) (radial displacement, in) (.28) (.24) Peak Core Barrel Strain, % 8.0 8.2 (time, msec) (1.2) (1.2) (radial displacement, in) (.30) (.31) cA larger value of 2.5x103 occurs at 3.4 msec but it is associated with the reflected pressure pulse and reverse loading of the core barrel and therefore is not correlated with the peak core barrel strain shown. O 108
,q Table 4.1-3 Comparison of Predicted (2) and Experimental (I) Energies and Slug Impact Velocities SM-2 SM-3 SM-4/SM-5 Exp. Comp. Exp. Comp. Exp. Comp {3) Slug Impact Velocity, ft/sec 91.5 65.6 62.5 57.4 62.4 57.4 Slug Kinetic Energy, KJ 9.49 3.67 4.43 3.05 4.85 2.56 Core Barrel Strain Energy, KJ .46 3.83 .42 4.14 .24 3.52 p Reactor Vessel Strain Energy, KJ 6.19 2.73 3.18 3.07 .83 2.76 V (1) These parameters are computed from measured parameters; refer to Reference 4-7. (2) REXCO-HEP Pre-Test analyses. (3) Calit' rated pressure source REXCO-HEP case. 109
Table 4.1-4 SM-2 HEAD ELEMENT CONNECTIVITY AND REPRESENTATION ELEMENT N0DAL NUMBER CONNECTIVITY COMPONENTS SIMULATED 1 1-2 Plug Weight, Margin Ring Force-Deflection 1 2 2-1 . Bearing Ring Force-Deflection l 3 4-2 Vessel Weight & Force-Deflection i 4 2-3 Flange Weight, Bolt Force-Deflection 5 3-2 Ledge Force-Deflection 9 SM-2 HEAD ELEMENT REAL CONSTANTS Stiffness Damp. Coef. Mass Margin Ring Yield Force 2 Element k 106 C /LBjEC M LhSEC GAP (IN) Fy (10 3LB) Number / 1 21.07 1099 0.3583 0.0061 231.3 2 34.95 1415 0.0 -0.0001 78.34 3 5.481 76.91 0.6745 0.0 99.99 4 4.300 30.91 0.1389 -0.000005 104.5 5 14.52 56.81 0.0 -0.01359 0.0 0 110
I O O O t TABLE 4.1-5 NODAL COUPLING AND COMPONENT DEFINITION FOR SM-4/SM-5 HEAD MODEL Element Nodal Number Connectivity Components and Properties Simulated I l-2 SRP plug weight, margin ring force deflection 2 2-1 Small riser force deflection 3 4-2 UIS columns, UIS mass, column force deflection 4 9-2 IRP plug weight, plug flexibility 5 9-3 IRP margin ring, margin ring force deflection s 6 3-9 Intermediate riser force deflection j _ 7 8-3 LRP plug weight, plug flexibility
~~
8 8-5 LRP margin ring, margin ring force deflection > 9 5-8 Vessel flange and LRP riser, flange and riser force deflection 10 6-5 Vessel wall, wall mass, force deflection 11 5-7 Vessel hold-down bolts, bolt force deflection (force upward) 12 7-5 Vessel flange, ledge force deflection (force downward) l h l i i -
TABLE 4.1-6 ELEMENT REAL CONSTANTS FOR SM-4/SM-5 HEAD MODEL Stiffness Damp.Coef. Mass Margin Ring Yield Force Element K C M Gap Fy Mass At Number (lb/in) (1b-sec/in) (lb-sec /in) (in) (lb) Node 1 1.70 x 10 7 181 .0239 .006 52500 1
-5 2 1.64 x 10 11.6 .1405 -5.63 x 10 -
2 3 1.34 x 10 6 7.5 .0324 0.0 8560 4 4 9.80 x 10 6 47.1 .0001 0.0 - 9 5 3.80 x 10 7 3 01 .2539 .006 175000 3 6 2.50 x 10 4 9.5 .0000 .003 - - 7 4.60 x 10 6 43.2 .0001 0.0 - 8 7 8 3.28 x 10 298 .0924 .006 225000 5 9 1.10 x 10 4 10.5 .0000 .01 6 - - 10 5.35 x 10 6 79 .8001 0.0 100000 6 6 11 4.03 x 10 284 .0000 -7.5 x 10~ 120000 - 12 1.50 x 10 31 7 .0000 -7.46 x 10- - - O O O
O O O TABLE 4.1-7 Summary of SM-2 Vessel Head Analytical and Experimental Results Component Peak Force Residual Plastic Peak Displacement Peak Acceleration (lbs) Deformation (ins) (ins) (gs) Analysis Analysis Experiment . Analysis Analysis Experiment 4 Head / Flange at Margin Ring 178,000 0.0 0.0 - - Vessel Head 412,500 - 0.0 0.036 2200 1250 i ' j Core Support Plate 101,160 - 0.0 0.056 260 1000
)
Lower Vessel Wall 100,000 0.0372 0.003* 0.056 260 1000 i 3 Vessel Flange 137,000 - 0.0 0.023 1800 1040 l Flarge Bolts 100,000 0 0.009 0.0069 - -
*The .003 inches represents an upward movement of the lower vessel wall / core support plate.
This results because the radial bulging of the vessel wall produces more upward movement than the loading on the core plate produces downward movement. Because of less vessel radial strain, this effect is not se'n in the subsequent tests. I
TABLE 4.1-8 Sunriary of SM-4/SM-5 Vessel Head Analytical and Experimental Results Peak Displacement Component peak Force Residual Plastic - (lbs) Defonnation (ins) (ins) l Peak (absolute) (gs) Acceleration Analysis Analysis Experiment Analysis nalysis Experiment I SM-4 SM-5 SM-4 SM-5 Small Rotating Plug 47,000 5) 0.0 0.0 0.0 0.120 4383 3300 i 4000 Inter. Rotating Plug 125,000(5) 0.0 0.0 0.0 0.114 , 1897 3200 3500 Large Rotating Plug 225,000(5) 0.001 0.0 0.0 0.091 I 1679 3300 3700 U15 Columns 8,600 1.53 0.0(I} 0.17 1.566 4340 - - Core Support Plate 104,000 - 0.l(2) 4.1 - I 323 - - k Lower Vessel Wall 100,000 0.035(4) 0.014 0.032 0.058 - - - Vessel Flange 108,000 0.0 0.0 0.0 0.033 734 - s2600 Flange Bolts 120,000(3) 0.01 0.0 0.0 0.033(.007 test) 359 - - (1) Columns not annealed (2) Relative displacment of center to preiphery of CSP (3) Experimental value is 97,600 lbs. (4) Axial defonnation (5) Force across margin ring O O - - - - - O
O O O TABLE 4.1-9 Comparison of Pre-Test with Post-Test ANSYS Vessel Head Responses Peak Force (lb) Peak Maximum Residual Plastic Acceleration (g's) Displacement (in) Deformation (in) Pre-Test Post-Test Pre-Test Post-Test P re-Test Post-Test Pre-Test Post-Test 0.0 47,176 39,000 4,383 3,520 SRP Margin Ring 0.120 0.117 0.0 0.0 124,980 104.110 1,897 2,210 IRP Margin Ring 0.114 0.110 0.0 0.0 225,000 206,870 1,679 1,580 LRP Margin Ring 0.092 0.092 0.001 0.0 0.0 107,960 0 734 1.780 Vessel Flange 0.033 0.037 0.037 0.010 0.015 120,000 120,000 - - Flange Bolts 0.033 ,
, g 4
1 i e 1 f 'f
l HEAD , o075IN. + r=3.79 IN >l' l AIR
\ U
" SG5 , Pp *P8 1.3675 IN.
VESSEL h (0.119 IN) 20 19 18 o SG4, P6 WATER 17 11.607 IN. N
\ 16 15 1
14 l t E AD SHOT AND " SG ,2PS l W ATE R N 13 l CORE BARREL 12 (0.1 IN) N \ h SHIELD P3 oh CORE x g 9 10.055 IN. ALUMINUM l p4
\ \ 6 STEEL 5 (E L ASTIC- U 4
PLASTIC) Ng 3 2.0IN h 2 34 5 6 7 8 Figure 4.1-1 RExco.IIEP Model f or SM 2 Pre-Test Analpis 0526-1 0 116
HEAD 6.075 lN. *j 4 3.79
"'" X x
'"~ '
[, __t_ VESSEL SG ,PS7 *P 8 1.35 IN. (0.119 IN) 22 j , WATER 21 ' 20 UPPER INTERN ALS <
,P6 12.86 IN.
x - -
;e WATER \ / 14 LEAD SHOT AND /
WATER j CORE BARREL 11
~~
g 10 i~~ h ll CORE 8 ALUMINUM sg /p 7 10.0551N. g\g,hg pt, / P2
\ /
STEEL N s'A*2 \ w l 3 _3__ 2.0IN. I 2 - - - - - 2 3 4 5 6 7 8 d Izigure 4.1-2 RiiXCO-illiP %1odel for SN13 l're-Test Anal) sis 0526-2 O V 117
+ 0.075 lN. H HEAD f l p3.79 IN >g l AIR m
\ \ U o
SG7 P10 28 c VESSEL + LINER m 25 (0.197 IN) g o Pg g WATER m
\ 22 12.86 IN UPPER INTERN ALSm STRUCTURE - -
/ 18 P SGS, Py p 17 WATER H /
LEAD SHOT AND WATERm
\ // 'O 15 CORE ARREL m \ N f 14 h
\
CORE BARREL (0.2525 IN) m
\ // 12 CORE m jj ALUMINUh ,Pj,P2 /[ 10 9.52 IN SHIELD m \ /[ 8 CORE SUPPORT (E L ASTIC PL ASTIC) m i g \ // 7 p
CORE SUPPORT m CONE (0.235 IN) N / 6 1.2IN. WATER m
-/ / 5 h
_~ yj 4 o p5 0'0 _s / 3 y / 2 p
- 7 2 34 5 6 liigure 4.13 illiXCO-lit.I' Model for SM-4 Pre-Test Analy sis 0526-3 118
O O 3
,] Se i g i i i i
$ 5 5 40 -
CORE BARREL E NICKEL 200 E 30 _ a e z g 20 - LEGEND: U MEASURED AT 0.001IN/IN/SEC E 10 M INPUT TO REXCO-HEP w PRE-TEST ANALYSIS 0 50
$ RE ACTOR VESSEL E
~
40 - NicKEt 200 o m {0 g 30 _ m - C E ,# 20 _
"~ -
5 ' ,, LEGEND: E 10 ME ASURED AT 0.001 IN/IN/SEC 5 O---O INPUT TO REXCO-HEP 2 . PRE-TEST AN ALYSIS m l l 1 1 I t 0 i i 0 1 2 3 4 'S 6 7 AXI AL STR AIN (PERCENT) Figure 4.1-4. Core Barrel and Reactor Vessel Stress-Strain Cunes Used in Analysis i
O 350 LEGEND: EXPECJED P.AV,20.7 gm 90/10 PETN 96'i t. .J INITI AL VOL UME 300 - 1/2* 3CALE SMBDB
--"- C A lBRAT10N P dV,19.7 gm 90/10 PETN 962 cm3 INITIAL VOL UME (G AS WORK ABOUT THE SAME AS EXPECTED P dV) 250 g
g 200 - 20
; \ \
cc h k y 150 .( a 15
\ ,e 6 100
\ ~ / g
__( f', 10 g _x \ s
- m a
50 - 4 -
/ ' - . 5
/ _ _ _ _ _ _ _
/
/ -
O l l I O O 40 80 120 160 VOLUME CHANGE -(IN3 ) l'igure 4.1-5 Piewure-Volume Change.md Gas Wmk Volume Change lielationships For Pre-Test Analysis 05265 O 120
STR AIN (PERCENT) , 7 6 5 4 3 2 1 0 s I l l l l l V i LEGEND:
,/ 2 TEST D ATA
. . ... SM 2 PRE TEST
/ ,, /* ' -
3 4 ANALYSIS
/
/ _ 5
\ '
\ -
6
\ -
7 g - 8 : N g . + VESSEL WALL s s - 10 , N 11 l
\ \
- 12 13 1 -
STR AIN (PERCENT) l
- 14 ,
1 - 15 ; 4 3 2 1 0 f TOP g 16 . I l' I ; I - 17 - l 8 I #
- 18 - 7 CORE 1 - 19 / - 6 fRREL V l / - r
- 20 j
- 5
' /
- 21 - - 4 ;
f
- 22 l j - 3
- 23- ( - 2 N 24 , N- 1 1iiliiiiiIiiliiIiilii 25 lIiil 0 420 360 300 240 180 b R (mils) 120 60 0
] N 180 120 6 R (mils) 60 0 T Figure 41-6 SM-2 Predicted and lixperimental 1)eformation Profiles 0526 6 121
STR AIN (%) 7 6 5 4 3 2 1 0 1 I I I s
,l/i l I
- 2 LEGEND:
' TEST D ATA
,- - 3
/ -------- SM 3 PR E T E ST 4
/
ANALYSIS l - 5 i
\ - 6 1 1
\\ -
8 g N g + VESSEL WALL s s i-s - 10
's 11 f
's
\
N - 12 l
~
\
g STR AIN (%) j g 14 0 l l - 15 TOP l I I I l/l 16
/ ,
I - 17 / 8 18 7 l l - 19 's - 6
- 20
/
)- S CORE BARREL l -
21 g - 4 l g l - 22 g -- 3 .
\ \
g - 23 - - 2
\
s 24 l',, - 1 Iiiliiliiliiliilii) ih- 25 I ' 'IUh 0 420 360 300 240 180 120 60 0 ] ' 180 120 60 0 6 R (mils) b R (mils) liigure 4.1-7 SM 3 l'redicted :md thperimental 1)eformation l'rofiles 0526-7 9 122
LEGEND: STR AIN (%) SM-4 TEST DATA 5 4 3 2 1 0
----- REXCO HEP PRE TEST ANAL.YSIS O' I I L#7 _ j STR AIN (%)
)
2 1 0 1 '
\ -
3 I I - 13 s N - 4 - 12
\ \
- 5 - 11 :
6 10 s
\ 7 9
\
\ 8 - 8
\[ 0 THERMAL ~
LINER 6 6 R (mils) 120 60 0
'u 11 - 5 12 4 ,,
k _ J 13 ~ 3 i STR AIN (%) VESSEL WALL _,2
+ 14 L 2 1 0 15 _ ; ,
/- 16 _ . m, TOP 0
17
/ Iiili r ,,
,L
-' f 7
/ 18 120 60 0
/ . _ Ig
[ .- 6 ( 20 6 R (mits) t
\
L 5 CORE N, - 21 g _ 4 BARREL N, 22 % 3 Q 1 23 / - 2
, /- 24 's -
1 25 s f { 26 , QQ g i s - 27
'%g 28 liiliiliilii 29 240 180 120 60 0 ,
b R (mils) I l'igure 4.1-8 SM.4 Predicted and L'sperimental 1)efi>rmation Protiles 0526-8 ( 123
6 r CD I Z b c I , 5 l,i H O hij p l m m m W W c1 W e & E E m L d l $ s I e 5 l s I l7 l= ll ' e I 8 l ? l .' t s d E !! l.' ' .m LLJ C E b P 'C I,
- 7 Ib ",
I a I<^
" ?
f E m
/ - -
1
/
/ 0
/ ?
~
/
s/ ar 1-r **" ___b_____ I g o e o o o o O C3 O c o C O O O O 6t) C M N e (ISd) 380SS3Hd o s 2 c>.<> 124 O I
a_. _a 4. - a ..- m.a -
.2 .-2 .Aa _ . - . _ - ._.%sr_w.- + __ mu_ ,a. m _ - . ,
a.av a. - 7e-d -6m . t e * @ E i 3 LA
>=
(J g "W * ' ' * " " * " L [ p 5 W %> g ct; llll M f ,d U Fs - e e ['*. D tlJ . N
.m*""
.. y M ~ & ~~ p y A e
Q - - M =__m .mo a e m o a m m'.* w 5 - =--- 9 n g g 1 -
- w.m.
E a zene-l _,, r_- - _- - - ~~~ e
,-e,y a-4-- y
- 9
$pe* U i
es"~ $ S q S 4*. 24 m :""*~"~ 1 1 Lu 2 ,$ m w r> d W y
===[., ~a C
d.#. , , - N M
.!:f
% a
,m/ v ,
# g \
- e> e4 a
8"" ;
# ens e
$ Z i 7 -
y O i ~
\
4 4 W- C M, .$ u,
- -w e"-
8 8 (isa) aunSS3Hd ' os2t -10 125
=
$ 700 L
~
LEGEND: l , SM-3 TEST 600 --
------ PRE TEST PREDICTION 500 - '
! l .kI 400 -
.i
, .. i
^
- .... i g ' '
- ::) . t .ha, ,, ;j',,:::: ),
_ ,, ,i, ,i , ,,, , .- .. .
= d ., i, ;;,",
!::;;:,? : ', lip 'a -
g ':: ; : 'l,. .s: 'i ro m 200
- N,, ..- : ,,::
s i is
, s .* '
. 4
8
s a ,
' ' i' e
I
*' \, ,'
s
- i. i .o' 100
-l, '#
,, y','
}
- ^
0
-100 O 1 2 3 4 5 6 TIME (msec)
Figure 4.1-1 i S\i-3 Predicted and Experimental Vessel Wall Pressure at Pressure Gauge PS e 9 - - --
e 4 , u ' c
< ~p, a.="
- - m C O
e-.===* "'"" 3 g" ,%w $
- act - - - - e en* M JE .E " -
em EW - e O
".==e- o m
=
Y .== """ ""* w / r o N*==~-~,,.,,,,,, i.=
~
4 y "m e'** er" e
- - - - . " ama.
" " -= " '--"--" ", g j N EEEEE. N M g g g u s J y
- "E m w -
a g - c ~' - r e
*C
<~h o C
"ll> ev 2 e- ;,;,,)
> m u 8"" C p
g ='.======m- --
='
~ ,
M ( f < E da w w ~ ~ ~- ~ > so
~ n P
~
Y LU -- lE cc to c. 7
=
x ( W c'% _
- m
- l D 4
_- , - ~" 4
" m- e
~~" ~~ '
N enWW W e.,,, ,,,,, l l l l l "I ~ ' N = c Q O C3 O C3 O O O C3 O c O O O Q C N t.0 m W M N w w I USd) 3HIISS3Ud 0526 12 L 127 t 1
e 0
\
4 7 f' - m
/ P k
/ h
.) =
aut C ~~
~
1_ , *~ 2
?
% ^
=cM~ m/ e ~ q E
o n A- -~~ ~
._.= 2."_ _ *- g 8
- =
~% sn ~~
~
." T ~--.~- I
-. ==ames.arsm-8. * . - _- y,
, 2- a t
1 W Q
~
1
/ :
o 3
.E
._ c
- N O
^
1 3 / O
~=
u e o d , c 0 W l> - m , b h b 5 M m I b
= m I ,
, =W
>= p N y I i
~ w b c
g a-A i m D P4 e
- q O i Z
Z e s m W L3 - e uJ I ~. l 4 I I I I I
= E r,,
e w o C3 e O o C o o o o o O == O W C kn O kn O .
'i m N N e e sn I
(ISd) 3HIISS3Hd 0526-13 128 0
l ov k
)
b C% f% (, , s air" , ' ~ W Y
' r3
=
N :> b l 7 Y fw#. e 3 ed I
= = e
< - - , e M E~ E
*- ~
e m b $ t e 4, _.~== -
" ~ ~
== s ,
jr" t [ N E S' m w .C E # 5 ~S
=
0 E
'C O
1 ? t 9
~ W
-s 3 C
- ) >g -
M c 1 > 4 , bha E i .O
& 4 ' T==
>= g y O i
6 M m i
- e e 1(
a= , U, m n a at l 4 f "" W l 1 l 9 a
- O I y n l l l ! '
o C o o = = =
~
~
S E-E i (ISd) 380SS38d 052(wl 4 r 129
c 3000
'e LEGEND:
Y SM4 TEST 3 _______ PRE-TEST PREDICTION 2500 __ h t it 2000 (- si li 8 G 1500 li S in 31 I
= li !
E m l ', P at g i s E 1000 8 tile : lgl', ,,", f n g g i iiI i 4 /g i 18 g a e i, 8s le I ig 500 - e
' ' g/ V s',#'
l) s s l ii ir .'s/ s o ____-s- s,. - _. .-__-_ h _j $ s
- 500 I ) -
I I i 0 1 2 3 4 5 6 TIME (msec) Figure 4.1-15 SM-4 Predicted and Experimental Vessel Wall Pressure at Pressure Gauge PS 9 O O
1 4 l i o
,1 1
\
%,b l
,, s 1 -
m d J%
-_ x
- - _ _ _ __ _ _____ ._~_ _ _._
0
~ 7IJ
=
-ns_ _.;;._ ~_ _
1_ - . L, I O _. . =
-_-_ ~ . ,. =.=______
__:_._.__t .. . $ i I e-ii
~
I n l , y x_____________._____.__q- _____________ a m g _a w - l E 5 n l H- G i : I e z l E O O A E eo
~ e .
l
- w. m e_
M / Q. W i m >m =
= m =
H d
~
z
$a = .
_0
.a a l < -
- m <
Z l . I h W ' o l < A m l di l r4 I .i g e. I d'
- c 1 I I I I I l *
= o w
e o o o o o e o o = 8 o o a o o o o o e o o o o o a a o -Of# ! oo c o m = m m - - L l (ISd) 380SS38d 0526-16 C \ 131 1.
4 r $ o
,e I
%~~
y a0 *
""e.
44 g p e *
^~==~
== *=
( ,' $ d
. N
- A
===ms'E. - - -.
I g 9 q old
~ 3
.-- ~ ,;". ,M
== # .
---== m .W W e M=-"".e"eM MM W p ,
l O s g / O l 2" l
-_- __,_3.____________------
M I w I m l u O, 8 6 m E w I - m l W 5 i E M o l E i -
.M.
P= 2 o u 9 1
.5-O i -
u k h
./
"cc bid
> H T
$" M c N
m 1"J
>=
M da T 2 mE ,. a b o M s-g ~ d l 5
- Z 0 w
LD a W G ' e
/.
d 1 M i 4 4 1 b M e l I I I I I I l o e O O O C3 Q O O O O C3 O O C3 O c3 o o 3, os O Q O C3 O Q Q
- en O O Q o L s m m e m u - -
1 (ISd) 380SS38d 0526 17 O 132
i
/
% i
%<l
! / ' I
<i ,
~<.
o - -i 3 m se a g' , e w a u
.m . c <
u,
- =
@ g- . ?:* ,: 3 vw - - . . _ u e.,
-s___ _ m 3 'm =
.r.=_,- -
n so s= I = _ , . , _ _ _ _ _ _ . _ ".
. _ :t w l i c1 LD l I w i Q
" i l
=
_' - e
- y =
\ - m E 5 1
o w i l
- s
=_n
* =
I l c
=_
c _k I - N ' W
-e
=
i n
-s
- h
, b
,g e
O A > r 7 m
< x i
4 l 1 , -r e l I I I I I I i = 53 LC o C3 o C o C o cm cm cp e o e CD C C C c c C O O O O C Q c 00 N (JD kn att m N , c 1 (ISd) 38 ASS 38d 0526 18 A U 133
_ _ _ _ _ _ _ _ - _ - - _ _ - - - - - - - r-4 l s** h
~,%
/
o'
'\
a.
~
.r x y
,=' a s
- go $8A m* _
=
es 6 Y
- t~' ~
~
J u
=
4
* == A 6
=ll
% r3
%w# $
~
\ h l 9 $
' '- N >
E
$ I -
m
- o / w =
u x '/ 3 '# W p E 5 ( *C m k w s 7 Q cC p **
**- U W / %
=
b /
/ n cc - N O o- /
w . 3 p 4 g
.v-a Y cx: I h
O ~a f a w 4 cc a: %. T p w % / M M ) Z
< E w a.,e .
E ) ~ " ~ l \ E
.. . ss
- Q 1 % "
$ 8 i 4) =
a .e 4 N L
~ **
- ene O O C C O O C C O Q Q O O O O LD m W M N w w I
(ISd) 3B ASS 3Ud 0926-19 134 9
1 O " BEARING i : RING LEDGE STIFFNESS
.L. L.
l T T
/ ///////
PL UG l
@ FLANGE l l//////,
h LEDGE I
] BOLTS
.,L. - L. ,
T T
,, VESSEL WALL
,, STIFFNESS hN VESSEL
' M ASS ,
NOMENCL ATURE: O =^ss f o^aeca SPRING GAP SLIDER Figure 4120 ANSYS Model for the SM 2 l>re Test fle:ul Aiul) sis 052420 O 135
I6% /
/
/
__ 'G / k A
/
/
A t, f L
/
/
O
/ 9D I f C
1 t
/ E
/ I
/
D t L
- /
/
I i "' V /
/
* /
' /
/
/
/
1 t
/
/
O L Ea E L f l V "* SS S' EV' 0 h/@i* ; 1 t 1 l h i s s D A ly L
. E a E
l y G 1 T CE
%GG n
F l" R G k AR LA n A
' A l
G LRi d
- a l
et l l s
' i' t o s b e m T- y eS r
Pf 5- o e C c , l n
.i i o I Si t i 1
t A L E< I i M E 4 I f N e SD e r in O D I L 0 h o D E 4 t f L r i f V
=c 1 T f o 8 1
' aO -
p A a R 1 t t =I le !. d4 l o e
\. r u
E S ig 1 L M YF I C u S e B N e 5 h AS t 5 w 0 4 c , p S u L 0
's 1 2
o , S O - UC n 1
, I
< 4
,G e r
,t 1
,i sR u t
ig F o t i .G
< i V
~
',.c , . A a 5
D" .
& L PS"Ef A
S I H =$? O w'"
l l LEGEND.
- REXCD SM 2 LOAD
--- SCALED SM8DB LOAD l
l l 40 .- l l l l'
; l.
Il
^
v e x i 'I E 30 - I d ;l
@ I I
l 5 \l u ;I 5 4 I I
- u. 20 .-
i I
$ ll p
a i-i l i l I I MI gI L ) 10 I
- lf; 1
I \ h 'l
\ $
' 't
\, l ? s l
I 4i I I I I 0 0 1 2 3 4 5 TIM E (msec) Figure 4.122 Comparison of Scaled SS1111)ll llead lead willi Predicted S%12 Ilead lead 0526 22 O 137
i l O l 12 LEGEND: i REXCD SM 2 LOAD SCALED SMBOB LOAD I 10 - . ! Co y V~ ~ 'g 5 ' ( 5 j n C $
$ I \
I I E 6- I I I i # h u l l I h s"ii s 1 i g i c I g Il [
$ l g /
g 111 s
- j 4 J. ',b 1{, l'
@ l ll f',3i e 1 ii il i /\ s Il I
,/ g
% s,' 8 2e ,,
I l I I I I I O O 1 2 3 4 - 5 TIME (msec) l'igure 4.1-23 Comparison of Scaled SMllill! CSP Load willi Predicted S\l 2 CSP Load 0526 23 O 138
l e O' 200,000 l l l l l 160,000 - l 120,000 - J k G U
=
E 80,000 - 0 1 40,000
^ I I " I O
O 2 4 6 8 TIME (msec) Figure 4.1-24 Pre (liclet! Force on SM 2 Margin Ring 0526-24 O V l 139 l b- , . - - .~_ _ _ . . - , _ _ . . _ , _ , . _ __._ _
J
/
0.048 O 0.040 - G u E 0.032 - W u 5 85 0.024 - O 0.016 -
- 7. >;
0.008 I I I O 2 4 6 8 TIME (msec) Figure 4 l-25 Predicted I)isplacement of SM 2 Veswl Head 0526-25 9 140
(w V i U k 5l" m b 6 - s
- 34
- m
.5
/ - & ;
n
%wi-~
b
~
n Y
/ _n g- =
r., g-.N 4
* - L
~~~~ _ , sW W
_ s E NW-N ,_
- g
= %__,,,,,,,,,
.e g 4 2w asg m
3 N b f E*m e %> m >- -4 E N5 m y N .2 4Y '2 l c 6 - - m f .
,, C o C '
4 w I , es
.C b I '
, !a o :s
.5 o e 8
e o o 8 8 e u. m " 7 (6) NolLVy33333y 05 2b-26 O O 141
O w w c
%> .o
~
=
9 * - t U o
- s L O M <
5g W - m o E s. ? xm f e t; , 5 UE / j' b i:i i 8 WL
< /
o x l w I e ~
- m 0 >
8 l s=
=,
e. Z
-w -
-- e g n -
, ___ w a z -
3
's, 2 p%s .g
~
O a
- p -
?
W s - t.
,f S
c t'.
~ ~
2 s C
.f u
,, e I 's -
, = a ,
o 8 8 g ) R 7 '7 .9 L
@) N011VU31333y 0526-27 142
O 4x105 3x105 _ E
$ l 8 2x105 _
8 u. l < 5 l Ix10 _ lt
+
0 I I I I I I O 1 2 3 4 5 6 TIME (msec) l l Figure 4 l-28 SM-4/SM-5 Predicted l'oice on Vessel llead 0526-28 iO 143
l e 120000 i 90000 W
- 1 w
fiD000 - 4 f [,3 e s J I l l l l 0 I 2 3 4 5 6 TIME (msec) q 1%ure 41 29 Sgt.4/SSI-5 Prealieteil I ,,ree ,,,, csp 0526 29 144 9
O 2500 2000 . g 1500 .. 5 N E 1000 - N J j \ l SC O - [ i v 0 - I I I I I O 1 2 3 4 5 6 TIME (msec) Figure 4.130 SN1-4/SNI.5 l'redicted Pressure Under UIS 0526-30 0 145
' JI i I 1 I. 1 0
1 O l 9 n i o t ar l e 8 e I c c N A O P L I T C R A L T ID N E
' 7 5-E R M M P IR S S E Y l a
P S . t fW X N n E A e I 6 m
- h i r
e
) p
- .W 5
( c e s m E d x
- E n M a I
T d J t e i c d 4 er P O f o 4
\g n o
s i 3 r a p m o C
'e ' 2 1 3
1 4 e r
' 1 u
g i e F
=.
0 0 0 0 0 O 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 5 3 2' 1 1 2- 3 4-75os$*m # 3 )' _ O
O O O 3 x 4000 EXPERIMENTAL OM -
---- ANSYS PREDICTION 2000 - ~
3 Z 1000 - i g 8 t
< r f,,g ; -
-1000 -
d II Y
-2000 -
-3000 I I I ' '
O 1 2 3 4 5 6 7 8 9 10 TIME (msec) Hgure 4.1-32. Comparison of Predicted and Experimentzt SM-5 IRP Acceleration
e 0 1 I 9 n i o t a r
' 8 l e
N e c O c L I T A A C P T N ID E R E R ' 7 S M P 5-IR S E Y M P S S X N l E A a t
~ I 6 n e
- ~ i m
r
)
c e
- e p
- d j I 5 (
s m E x E d
- n IM a T d l l t
e i c I I 4 d bl I l l e r f ii 1 P f I
\g\
l I 3 f i o n o s r a o p m o C I 2 . 3 l 3-1 4 1 er I ug i F
- - - - - - O 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 3 2 1 1 2 3-4- 5 - 3 !@Eu4 8g0 e me
, 11il
,----.- ,- ---,, y a .3, , , _ u . - ,
i l e Q Z e U W " $ E -
- 5 NE cc m E
W
- a. M
> G X 2 y
w a ,y l - l I 5 a !
,n E
2 W l s w i - N m w _ a
=
0 C
\ _
m j
\ 7
% _ w
/ s M m s s' * %
- D w
a E
=
n p# s - e,- . H u-o
/ :
j N ' * == % ' q
*%. e O ~~~%w, %
\
.{
h
\ - -
m E o V e 9
~
/ N v
N z I .D
\ 'a.
s - [ s > E I I l l l j
=
N N
" " O c h =
O
= =
= = s e s c o o l
(S3H3NI)IN3W33VldSl0 0526-34 4 149
i j l
- l 8 '
8
)
J r } o
~
.j g
\ m I 8 c.
a w 3 T #
>i m \ =
o a [y T E I o w: e e s.l a a e h C
=
E S z m j 5 E 3 S ' I > E l - 8 E E g w c l - )- = c E. I
~
/ m
- a I ( s -
?a
)- 3
-f ,N l k
A- ( s C
- / o $
c
~
~ E
*# u
,# 1 $
_ ~~ V m __ __w - 2 . ? a
-r p* *#
y g
~ } ~ ~-= ~~%-
e c O C . . . o g g 8 h m c
=
C g n ~ ~ - -
=
(6) 3anindWV 0582-4 O 150
8
=
o L, -
)
'1 1 e
4 8e e 8
) =
m
\ c >.
8
'\ 8 m o 3
=-
) k u ,
8 m l
= m -
R 5 ) E# E M m l
.:. )
E <. Z x W a:
/_ =g w 8 - E l E
-- E
>. c u o E Z @
W l r 5 $ e , e l (/ o a w x w m n o e n .t O '5 = Ic g - 8 g e% 'c
-j 2 7 , _ g o x m
< C'~, ,
<;n . 4 u 6
- g
-- , - e
-- ~ o
# , Cl)
F y e.* LC M ~ ~ ~~ _ _ _ y - .--
, -~~ ,
o e e o E 8 m 8 m 8 0582 5 151
0 0 0
' 0 1
O 0 q 0 0 9 t a r c e p h^ t 0 9 0 8 S y c n e u q
' e r
L F A T N D f 0 0 0 P R E M I E T IC RL 7 S 5 R g - E P D g M E X R / 0 S E P 0 l i 0 a t 6 l ) z n H e
- l'
( Y i m r
- \ i 0
0 0 C N e p x E
- 5 U E i Q d
- ,I E n -
i R a k \ 0 F d
\ 0 e f[ i 0
4 i t c d e
'y r O
P g 0 f f 0 o Jr , 0 3 n o s i f l r a
^\ 0 p
0 m
) ,
0 o i 2 C 7. V!, 1-1 0 N 0 i 0 4
\
\} 1 e j
r - i u g l\liI l l1 ll i F lIil I I j -l l l l _ O 0 0 0 0 0 , 0 0 0 0 0 0 0 0 0 0 5 4 3 2 2 35E 3h&
"u~
i O
400,000 O V 320,000 - 240,000 - Gi 160,000 _ 80,000 - 0 ! l l , 0.0 1.6 3.2 4.8 6.4 9 2 TIME (msec) I l Fipne 4.1-38 l'redicted S%1-4 LRP Ntargin Ring Load 0526-15 O 153
1 l i I 15000 12000 - l 9000 - G a m l _ 4 6000 - l i i 3000 -
,1 O l li I ' I O 1.6 32 4.8 6.4 TIME (msec) l'igure 4.1-39 Predicted SM 4 IRP Margin Ring Load 0526 36 O
154
1 O 5000 4000 - 3000 _ l l G 1 2000 _ 1
\
l O 1000 - ( 0 l l l 0.0 1.6 3.2 4.8 6.4 TIME (msec) l igure 4.1-40 l'n clicted S%14 SitP %1argin f(ing 1.o nl 0926-37 155
l! l 8 e
. i 7
~ ,
~
~
~
I 6 e n
- u C
-. I 5 )
T i n a N tr _ E S-C s
- R s E
e r
- P t
( S
- I 4
N I A 0 0 2
. R l T e S k A
T _p ~ L A 1 X i N l c la
- A _
i 3 A D - w
- T -
A S le e T d, - . A w D -
- e L
A T C E C T S O P C
~ V 1
S E E ~ 2 N I S P S l 4-E N I
/ E /
1 N H N M
/
I N I
/
O / I 4
- R I N
D E 1 IN C I
- r e
N P 0 X E X 0 - 0 0 E 0 0 u
' ig G E 0 1 1 E
L i F 1 _ - - - - 0 0 0 0 0 0 0 0 2 0 8 6 4 2 1 1 _Er EwC$ eEcww c , m c .J s f &i* e
._M
STR AIN (PE RCENT) 5 4 3 2 1 0 STRAIN (%-3 msec) O 1 I I I V/ l 5 4 3 2 1 0
, - 3 I I l l/l f( _ 4 , / -
8
~
/ - 7 g\ -
6 ' VESSEL / - WALL / Nh - 7
/ [
5
\ - 8 l \
\
~
4
\ l
_ g
- 3
- 10
- 2
' - 1 11
)I- '
I,ili I"T' i 4 ' 0 l1 - 12 l t
~
240 180 120 60 0 6 R (mils) 14 1 STRAIN (% 6 msec) l - 15 ; 4 32 1 0 Il - 16 ; I I l jP TOP ' f- 17 ; / 8 . l l i
- 18
[ - 7 , V - 19 6 CORE [f - 20 ;
]
5 : BARREL
- 21 _ 4
'll1
- 22 -
3
\
23 ,
- 2 N 24
- 1 Lil,ili,1,il 25 II' 0 l 300 240 180 120 60 0 s
, 180 120 60 0 6 R (mils) b R (mils)
LEGEND:
--- REXCO POST TEST ------ REXCO PRE TEST EXPERIMENT Figure 4.1-42 Comparison of SM 2 Pre and Post-Test RiiXCO-HliP 1)eformation Profiles l
0526-39 O 157
S T R AIN (PE RCEN T) 4 3 2 1 0 AVERAGE DATA I 1 STR AIN (%) SM-4
/ / /( 22 1 0 ----- SM-5
's p_ 33 ----- REXCO POST TEST p _ 3 l p s, f _ 4 _ g --------- REXCO PRE TEST 5 g
- 11 i s, -
6 10 J
's 7 -9 CORE BARREL
\
\
6 R (mib) 6 R (mils) 4 10 - 6 180 120 60 0 300 240 180 120 60 0 4 11 - 5 , lllI lllllI 4 12 - 4 , STR AIN (%-6 msec) STR AIN (% 3 msec) 4 13 - 3 q' 14 -,2 4 3 2 1 0 7 6 5 4 3 2 1 0 J 15 -- 1 i ; i ; , i j i l l ;l ll 16 -- I 9 j
' 8 / 8 8
/
II
- 1. , l , 7 li s' -- 7 / l 7
O f'fff- I-- ' 18 120 60 0 5 R (mils) h
'se I -
['s( - 6 5 , ,'
,/ '
l - 6 5 f( l-- 20 t l s _ 4 123 24
,_('* l
/
ff'u /
} 25
-0 0 f
1 f 26 h s -- 27 i
'gj 28
- t. 1, , i i , I i d) 29 240 180 120 60 0
$ R (mils)
I l'igure 4.14.1 Comparisim of S\l-4/S%15 l're- amt Post /lesi RI.X( () lil.I' i)eformation l'rofiles 0526 40 0 158
i l O 1 I 1 400.000 LEGEND.
------- PRE-T E S T SL U G L D AD POST TEST SLUG LOAD I
300.000 - g B il li
$ I
- f g 200.000 -
lIl 8 II lI l1 h I", a l's I i' 100.000 l 'l gig
- h ,l \i~f -
,,',7 m -
l 'I,l 1 0 I I l ' ' f I l 0 1 2 3 4 5 6 TIME (msec) Figiire 4.1 -14 Coinp:irisoii of I're- :iint Post -l'e.st 1(' Xm-Ill l' lle:ul Lo:uls :is Applicil to ANSYS Ile:nl \1otlel 0526 41 0 159
O 2000 LEGEND:
------- PRE TEST SRP RESPONSE POST TEST SRP RESPONSE D
t* l' I
*\
8 i 1000 _
,I 3 :
a l 0 A ilI
$ i'F l w I i 1%
!$ 8 I 's _ s 0 O A / I J g ,1
$',his l! s O
iii, l I I
-1000 I I O 1 2 3 4 5 6 TIME (msec) 1~igme 4 l-45 Compasison ol' Pre- aml Post-lest ANSYS Simulations of SRP Response 0526-42 O
160 t - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
4.2 Assessment of Deviations in Response Associated with Scale Model Non-Prototypicalities
~
In applying the results of the scale model tests to the prototypic reactor, the effects of scaling must be assessed. The following sections thus identify those physical parameters whose similitude cannot be maintained exactly and assess the resulting deviations between the scale model and prototypic responses. 4.2.1 Review of Scaling Laws Provided certain conditions are met, the dynamic response of a scale model can be made to simulate exactly the prototvpic response. This has been shown theoretically (Reference 4-11), has been Anonstrated experi-mentally (Reference 4-12), and has been used to great advantage for both destructive and non-destructive tasting of scale models. Such techniques have not only been used in the reactor field but have been used extensively in, for example, the automobile and aircraft industries. As a specific application in the fast reactor field this technique was used in evaluating deviations from non-prototypicality in the FFTF scale model tests (Reference 4-13). For exact similitude in the scaled dynamic responses of a coupled fluid-structural system, only two requirements must be met. These are: o All linear dimensions must be reduced by the geometric scale factor and e All structural and fluid properties must remain the same as the prototype. If these constraints are met, then exact similitude in the response of the scale model will be maintained even when non-linear characteristics in material properties (e.g. strain harder.ing) and geometry (e.g. gaps) are present. b a 161
In practice the first of the above requiremerits can be met quite accurately for a reactor system. However, more significant deviations from simili-tude arise from changes in material and fluid properties. In particular, the vessel wall, major in-vessel components and coolant are simulated using room temperature nickel-200 and water, respectively, rather than stain-less steel and liquid sodium at CRBRP operating temperature. It is pri-marily for these material substitutions that scaling effects must be examined. The coupled fluid-structure system being analyzedis subjected to time varying loads originating from the energy release in the core region. For such a system the important dimensional units are force (F), length (L) and time (T). The dimensional quantities of interest can all be related to these three units as shown in Table 4.2-1. Quantities which are inherent to the model (e.g. lengths) are termed " characteristic" while quantities which are related to the response of the model (e.g. stresses) are termed " typical". The variables identified in Table 4.2-1 thus deter-mine the system and its response and may be described by some complex function 1 f (D,E, p s ,g, f
,k,W,o,c 6, v, a, f, p, t) = 0 The Buckingham Pi Theorem (Reference 4-11) then states that an equation of the form above can be reduced to an equivalent equation for a complete set of dimensionless products, i.e.
F (ni , n2 >"3 *"n) = 0 where each n$ is a dimensionless combinacion of the variables in question. To determine the nj s, a matrix as shown in Table 4.2-2 (a) is generated. The numbers in the matrix correspond to the exponents of the dimensional quantities given in Table 4.2-1. Dimensionless products are formed by first reducing to normal form the first three columns of the matrix, com-posed of important physical quantities. D, p and p have been chosen as 162
~
can be seen in Table 4.2-2(b). The next operation consists of sweeping out the remaining' columns using the first three. The final . form of the matrix is shown in Table 4.2-2 (c). The dimensionless ratios, which are , equivalent to the ng , may then be taken directly from the normalized s matrix. Strict similitude requires that the corresponding dimensionless ratios be equal for model and prototype. Using "m" and "p" as subscripts on the model and procotype dimensionless ratios, the final similitude re-lations are. then given in Table 4.2-3. It should be noted that the explicit form of the modulus E in equation (11) of Table 4.2-3 is not defined. This would be taken as the elastic modulus were the problem in the elastic range. However, it could equally well be defined for example as the secant modulus thus allowing differences in plasticity between model and prototype to be accounted for in the scaling evaluation. As a result the ratio of the moduli can be more generally interpreted as the ratio of the amplitudes (i.e., the strains at a particular value of stress) of the respective stress-strain curves. The additional constraint that the model and prototype stress strain curves are geometrically similar must be imposed for this more general case. O O ,e3
l 4.2.2 Application of Scaling Laws to Scale Model Tests The equations derived in the last section allow an assessment of the degree of non-prototypicality which may be expected in scale model simulation of the reactor response. These equations can be used to gain a quantitative estimate of the deviations resulting from such simulation. Table 4.2-3 showed that if exact geometric similitude and identical material properties are maintained between model and prototype, then similitude of response will be maintained. However, in practice, it is difficult to obtain model materials which will precisely fit these requirements. Table 4.2-4 compares the properties of the materials chosen (water and nickel-200) to those of the prototypic materials (liquid sodium and 304 stainless steel). While there is some variation in the properties, adjustments can be made to account for these. This can be done in two ways. First, using the scaling laws, adjustments resulting from material property changes can be made both to the time scale and to the slug impact load experienced by the vessel head. Second, a computer simulation of the test, in wh1ch the model material properties are replaced with prototypic properties, can be made. Using this latter approach the changes in response can then be directly related to the property changes. In both cases, only material scaling and not geometric scaling is considered. This approach is taken since effects due to geometric scaling can be assessed relatively simply through application of the geometric scale factor A D. For the discussion below, it is assumed that i = (1/20)/(1/20) n = 1. The dimensional analysis approach is discussed below while direct numerical simulation is discussed in Section 4.2.3. Consider first the change in fluid response between the two models, one being sodium filled and one being water filled. Of particular interest are the ratio of slug impact times, defined byt A = pt /tm, and the ratio of the slug impact pressure, defined by pA = pp /pm ' First, to assess the relative impact times, Table 4.2-3, equation 12 states: t Ap/A t or A t= t p/tm m p * )Ap / 164 O
where A p is the ratio of the pressures before impact and X I if the ratio of the fluid densities. Since the system pressures will be identical at the start of bubble expansion and very similar just before slug impact (the bubbles l will both have expanded by the cover-gas volume), A p
% 1 over tl.is time period. Hence
?f l At " \/ A p In the REXC0 calculations used to simulate the two cases above (see Section 4.2.3) the upper internals structure was modelled as a hydrodynamic medium of density 121 lbs/ft3. Hence the effective upper plenum density is taken as volume weighted average of the UIS and coolant densities. For the sodium filled model this weighted density is p f = 64.9 lbs/ft3 while for the water filled model this is p f = 73.6 lbs/ft 3. Hence the time i m i scale ratio to slug impact is given by: 1 At = tp /t m =,
=
64.9/73.6 = 0.939 From the REXC0 results (see Table 4.P.-5 or Figure 4.2-5) the corresponding value is: l Os REXC0 A = t p/t = m 3.02/3.19 = 0.947 REXC0 Hence A is less than one percent different from A
- t Consider now the change in slug impact pressure between the water filled and sodium filled models. The impact pressure is given by p = p fvc where v is the fluid flow velocity at slug impact and c is the fluid sonic velocity. Hence f
P I A p(atslugimpact)'= = =A AA yc m pmmV e, O ,ee
Consider each of the tenns A f ,A y and A c in turn: , e The density ratio A p is given above. e The velocity ratio A is y defined through Table 4.2-3, equation 8: Ay = vp/vm *I\Ap /A p' " ' II A p In the above A p =A p (before impact) % 1, as discussed above. e The sonic velocity c is defined by c = ,/ Keff /p , where K eff is the effective fluid compressibility as defined 'in Table 4.2-4. Hence (..-. (. Ac = cp/cm " ( \/ eff /P )pI ( \'Keff/P )m r
" \.Ak /py Note that this is consistent with Table 4.2-3, equations 1 and 8.
Combining the above terms: O 1 A p (at slug impact) = Af A yc A
=Af / 'A/A k
" \' Ak
= 1.129 From the REXC0 results (see Table 4.2-5 or Figure 4.2-5) the corresponding value is:
166 O
l R A"EXC0(at slug impact)' = 3.48 x 510 /3.04 x 510 i O = 1.145 Hence A REXC0 is about 1.5 percent different from A . p Following slug impact, it is more difficult to define scaling ratios accurately since, for example, A p or Ac will n t clearly be dominated by a single physical process. Since post impact pressures do have an effect in de-termining the response of the coolant boundary, it is more . difficult to obtain accurate structural scaling ratios. Consider for example the difference in peak vessel strain resulting from fluid impact and its after pressure. Ideally the deviation in strain between the models should be zero (see Table 4.2-3, equation 6) but becuase of material property deviations, it will not be. Consider the strain ratio Ag = c p/cm s s
=(c/E)p/(c/E)m Es is the secant modulus
= Ay/AE Since the stress is physically more related to the applied equivalent static pressure than to structural properties for a simple hoop loading, and stress and pressure are related through Table 4.2-3, equation 5, Ag = A p/AE Note that for the ideal case Ag = 1 implies that A =A This is consistent p E.
with Table 4.2-3 equation 11. From Figure 4.2-1, A 20.9 in the plastic regime and from the above E discussion, A p will be in the range 1.0 to 1.125. Hence A will be in the 167
range of 1.1 to 1.25. From Table 4.2-5, the peak REXC0-HEP vessel wall REXC0 strain give A REXC0 = 3.0/2.7 = 1.1. Thus A is consistent with A g and seems to agree more closely with the result derived from the assumption that Ap = 1. Sectio ' T discussed the creation of fluid turbulence which resulted from adding the upper internals structures to the models. This effect was shown, at least in the scale models, to have a significant energy dissipating effect and led to a reduction of loads imposed on structures. Since geometric scaling and material property differences alter the fluid response, an estimate of the non-prototypic deviations associated with turbulent effects is important. Because REXC0-HEP cannot currently simulate turbulent dissipation, it is not possible to assess this effect through use of this code. The approach taken below is thus to compare the Reynolds numbers (R) for the scale model and prototype. Such a comparison is too limited to quantify the change in energy dissipating effects. However it will indicate whether energy dissipation in the model under or over estimates prototypic energy dissipation. For the scale model Rm = (pvD/p)m O For the prototype, Rp = (pvD/p)p wnere p is viscosity. The other parameters are defined in Table 4.2-1. Hence AR = R p/Rn1 b b b "E P Y D U m m m m "A Ay AD A p f 1 1/2 x A p[ p U O 168
f 1/2 "
=
A, Ag /A, f Table 4.2-4 defines A , A and A p D A R = (0.827)1/2 x 20./0.224
= 81. l l
Thus turbulent effects in the prototype are likely to be considerably greater than in the scale model, and will further reduce structural loadings. l In summary, this section has applied the scaling equations derived in Section 4.2.1 to assess deviations in system response (i.e. in time base, pressure and strain) when simulant material properties are substituted for prototypic properties. In assessing the adequacy of this approach, whose basis is the Buckingham Pi Theorem, the calculated deviations were compared to the deviations associated with explicit REXC0-HEP simulations of the transient response. This comparison shows that the dimensional analysis approach to predicting deviations in slug impact response is excellent. In addition, while assessment of post impact response is more difficult due, for example, to stronger fluid-structure coupling and greater material deviations in the plastic regime, it is still possible to obtain meaningful results if insight into the dominant physical processes is applied. Finally it was shown that fluid energy dissipation is likely to be considerably greater in the prototype than in the scale model . 169 l
4.2.3 Scale Model simulation Using Prototypic Properties This section presents results of two REXC0-HEP analyses related to the SM-4/SM-5 tests. The objective of these analyses is to assess by direct simulation the significance of deviations from prototypic properties. The analyses successively change the water and nickel used in the SM-4/ SM-5 post-test analysis to elevated temperature sodium and stainless steel properties. Except for these material property changes the analyses are the same as the post-test SM-4/SM-5 analysis discussed in Section 4.1.3 Comparisons are made between the earlier analysis and the two additional analyses discussed herein. Correction factors are computed l for use in scaling the experimental results to the full scale prototypic conditions. l Table 4.2-4 compares the water properties in the SM-4/SM-5 post-test analysis to the operating sodium temperature properties used in the REXCO-HEP analyses discussed in this section (References 4-2, 4-4 and 4-5). ' Water at room temperature is a denser more compressible fluid than the sodium at operating temperature. Figures 4.2-1 and 4.2-2 compare the REXCO-HEP input stress-strain data for the Nickel 200 at a strain rate of l 100 in/in/sec and at room temperature to 304 stainless steel at low strain rate and elevated temperature. The stainless steel in both the core barrel and the reactor vessel have higher yield points than the nickel, but nickel has a somewhat greater capacity to absorb energy in the strain range of zero to 10 percent. High strain rate nickel properties are used here since strain rate scales by the scale factor (i.e., strain rates in the model will be 20 times those in the prototype.) The nickel in the water-nickel analysis was changed to stainless steel , first to produce a water-stainless steel analysis. The water was then changed to sodium to produce the sodium-stainless steel analysis. The intermediate water-stainless steel case allows for separate evaluations of the non-prototypic effects of water and nickel. Table 4.2-5 summarizes the results for these two cases and compares them to the SM-4/SM-5 post-test analysis from Section 4.1.3. 170
i There are no major differences in the results. Changing nickei to stainless steel results in slightly greater deformations of the core barrel and vessel wall. This result is consistent with the greater capability of the nickel to absorb energy. However, as shown in Table 4.2-6, the radial kinetic energy is less with stainless steel than with nickel because of the higher yield strength of the stainless steel. More energy is contributed to the water slug prior to yielding in the case of the stainless steel. This leaves less energy to contribute to radial motion and radial kinetic energy. Changing water to sodium causes a slightly higher slug velocity and earlier impact due to the lower density of sodium. The lower density of sodium causes a lower slug kinetic energy at impact even though the slug velocity is higher. Figures 4.2-3 through 4.'2-10 compare selected variables from the two analyses. Figure 4.2-3 shows the pressure on the core barrel. The sodium-stainless steel system has a lower core barrel pressure. The total impulse to the sodium slug is less but it achieves a slightly
' higher velocity and impacts the head slightly earlier in time-as shown in Figure 4.2-4. The time difference between the two systems is consistent with the relationship t
sodium
=t water.(densityofsodium/densityofwater)1/2 as discussed in Section 4.2.2. A similar time separation is shown in Figure 4.2-5 for the force history on the reactor head. The pressure pulse is higher for the sodium-stainless steel analysis, again consistent with the dimensional analysis. Figure 4.2-6 shows that the net force on the core support is not much different in the two cases.
The sodium-stainless steel analysis causes a less severe downward force in the first 1 msec. Figure 4.2-7 compares the pressures under the UIS region. Figures 4.2-8 and 4.2-9 show the buildup of strain energy in the core barrel and reactor vessel respectively. The slight increase in the 171
3 core barrel strain energy at the 5 to 6 msec interval is due to deformation caused by reflected pressure waves. These curves are consistent with the higher yield stress and lower energy absorption capability for stainless steel discussed previously. Figure 4.2-10 shows the radial displacement of the reactor vessel at the point of maximum strain near the top of the reactor vessel. The curves are similar showing time phasing and maximum reactor vessel straining consistent with the dimensional analysis. The larger strain associated with the stainless steel occurs even though the slug energy is less. This appears to be because the energy absorption capability of the stainless steel is less. Tables 4.2-6 and 4.2-7 provide the energy partitioning in the three cases for just prior to slug impact and at 6 msec. The latter time represents dynamic equilibrium (i.e., when all plastic deformation of the coolant l boundary has ceased). l l Table 4.2-8 sunmarizes the strain produced in the core barrel and reactor vessel for the water-nickel computation and the sodium-stainless steel computation. It should be noted in reviewing the table that the stainless steel vessel wall properties as defined in Figure 4.2-1 reflect a temperature of 1000 F rather than normal operating temperature (nominal 304 stainless steel properties for normal operating temperature are shown in Figure 3.1-1). If this discrepancy were removed from the data in Table 4.2-8, the strain difference shown in that table would be reduced to even smaller values. In conclusion, it can be seen that the use of the simulant materials selected for the scale model experiments results in very small deviations from prototypicality. In addition, deviations, as derived by direct namerical simulation, are very consistent with those derived through the use of the dimensional scaling laws (as shown for selected variables). Deviations resulting from fluid turbulence cannot currently be assessed through numerical simulation. However, application of the scaling laws indicated that energy dissipating effects resulting from prototypic fluid turbulence would be greater than in the scale models. 172 O
Table 4.2-1 Significant Dimensional Quantities O Quantity Dimensions Symbol Characteristic length L D 2 Characteristic elastic modulus F/L E (structure) Characteristic density of structure 2 4 s FT /L p 2 Characteristic bulk modulus (fluid) F/L K Characteristic density of fluid 2 4 I FT /L p Characteristic stiffness F/L k Characteristic work energy FL W 2 Typical stress F/L , Typical strain - c Typical displacement L 6 Typical velocity L/T v 2 Typical acceleration L/T a Typical force F f 2 Typical pressure F/L p Elapse time T t 173 1
Table 4.2-2 1 Operations on Dimensional Matrix l l l f s D p o K p k W o e 6 y a f E t F 0 1 1 1 1 1 1 1 0 0 0 0 1 1 0 L 1 -2 -4 -2 -4 -1 1 -2 0 1 1 1 0 -2 0 T 0 0 2 0 2 0 0 0 0 0 -1 -2 0 0 1 (a) Original Matrix of Dimensional Exponents 2 D p0 D l K s k W o e 6 y a f t p p E F 1 0 0 1 1 1 1 1 0 0 0 0 1 1 0 L 0 1 0 -2 -4 -1 1 -2 0 1 1 1 0 -2 0 T 0 0 1 0 2 0 0 0 0 0 -1 -2 0 0 1 (b) Diagonalization of First Three Rows of Matrix W WI U2 "3 US W6 38 "9 "10 W4 U7 *11 812 p f2 D f f E / p' 2 K k o 6 V fp_ ap D 2 pD D p p p f % pD p c' D sp p pD p t f,f D2 F 1 0 0, 0 0 0 0 0 0 0 0 0 0 0 0 L 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 T 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 (c) Final Dimensionless Form of Matrix 174 0
Table 4.2-3 Similitude Relationships
- A)
(K/p)m " (E/P)p "" Em " (I/A p)Kp (I) s (pf,f)m"(Plo )p "" (o')m " (1/^p I(o$)p (2) (k/pD)m = (k/pD) => k m (I/A pAD)k p (3) 3 3
=> W,= (1/Apx39)wp (4)
(w/pD ), = (w/pD )p (o/p), = (o/p)p => (5) m " (1/Ap ) p cm " 'p "" (6)
'm * 'p 1
=> 6 (7)
(6/D)m = (6/D) m "_(l/AD)6 l f f (v /p /p ), = (v /p /p), => v,=fx f /x p v p (8) f f f
=>
(ap o/p), = (ap o/p)p a,= (xg Ag/xp)ap (9) 2 (f/D p)m * (f/0 P)p "> fm * (II A)p p (10)
=>
(E/p)m = (E/p)p Em * (I/^o) E p (11) r2 (t / J u >m = (1 lg,p D >, => t m = /x,,x,< Dx 2' t, (,2) l where: A = p /p ' A = D /D " (P )p / (P )m p p m D p m' A p A U 775
TABLE 4.2-4 l l Comparison of Prototypic and Scale Model Physical Parameters j Parameter Prototypic Value Scale Model Value Ratio (A) Upper Plenum Fluid Density 0.030 0.036 0.827 (lbsf/in3) 3 Density of Structure (lbs /in ) f 0.283 0.321 0.882 6 Structural Elastic. Modulus (psi) 22.6 x 10 30.0 x 10 6 0.753 5 Fluid Bulk Modulus (psi) 6.235 x 10 2.885 x 10 2.16 0 Pressure Derivative of Bulk Modulus 6.915 7.67 0.901 5 5 Effective Bulk Modulus (psi)* 2.308 x 10 1.811 x 10 1.274 2 3.264 x 10- -7 Viscosity (1bs -sec/in ) f 1.458 x 10 0.224
*The effective bulk modulus takes into account the elastic response of the vessel wall i.e.
K t1eff vessel
= K/[l+(K/E)(D/t)] where and liner (see D is4.1-3).
Figure the vessel diameter and t is the effective thickness of O O O
Table 4.2-5 O Comparison of REXC0-HEP SM-4/SM-5 Prototypic and Scale Model Results SM-4/SM-5 Nickel and Water Stainless Steel Stainless Steel Parameter System and Water System and Sodium System Maximum Radial 1.82 1,73 1.70 Kinetic Energy, Kw-sec (time, msec) (0.42) (0.45) (0.44) Maximum Positive 5.81 5.89 5.76 Axial Kinetic Energy, Kw-sec (time, msec) (3.13) (3.16) (2.98) l l Maximum Slug 56.1 57.1 58.7 Velocity, f t/sec (time, msec) (3.13) (3.14) (2.95) 5 Maximum Slug 3.04 x 10 5 3.57 x 10 5 3.48 x 10 Force, lbs. (time, msec) (3.19) (3.22) (3.02) Maximum Head 1,48 1,44 1.28 Total Energy, KJ (time, msec) (4.73) (4.77) (4.53) Maximum Vessel 2.7 2.8 3.0 Barrel Strain , %' (time, msec) (5.21) (5.14) / ( 3. 85 ). (radial displacement, in) 0.162 0.172 0.180 Maximum Core 6.6 7.3 7.3 Barrel Strain, % (time, msec) (1.14) (1.21) (1.24) (radial displacement, in) 0.250 0.277 0.278 177
1 l l l Table 4.2-6 l Comparison of Energy Partitioning in SM-4/SM-5 Prototypic And Scale Model Calculations (At Slug Impact) i Encrqics in Kilo. joules Nickel and Water Stainless Steel Stainless Steel l Energy Component System and Water System and Sodium System l Head Total Energy 0. O. O. Radial Kinetic 0.741 0.721 0.72' Energy Upward Kinetic 5.73 5.79 5.68 Energy Downward Kinetic 0.0279 0.0386 0.0193 Energy l Core Barrel Strain 2.99 3.12 3.16 l Energy l l Vessel Strain 2.37 2.37 2.21 Energy Fluid Internal 0.151 0.198 0.150 Ene rgy Core Internal 69.0 68.8 69.0 Energy Total 81.010 81.038 80.940 Percent Change * -1.95% -1.92% -2.04%
- Total energy = 82.625 Kilojoules and the percent change indicates that ,
percent not accounted for at this time in the analysis. 178
9 Table 4.2-7 t Comparison of Energy Partitioning in SM-4/SM-5 Prototypic And Scale Model Calculations (At Dynamic Equilibrium) _ Energies in Kilo.ioules Nickel and Water Stainless Steel Stainless Steel Energy Component System and Water System and Sodium System Head Total Energy 1.12 1.06 .852 Radial Kinetic .248 .251 .179 Energy l
)
Upward Kinetic .497 .469 .482 I Energy i Downward Kinetic .370 .471 .409 Energy O Core Barrel Strain Energy 3.38 3.38 3.43 Vessel Strain 5.21 5.30 5.05 Energy Fluid Internal .725 .737 .749 i Energy l l l Core Internal 68.3 68.1 68.2 ; Energy Total 79.9 79.8 79.4 l l Percent Change * -3.36 -3.46 -3.96 l
- Total energy = 82.625 Kilojoules and the percent change indicates that
, percent not accounted for at this time in the analysis.
179
Table 4.2-8 Comparison of Core Barrel and Vessel Wall Strains From REXC0-HEP SM-4/SM-5 Prototypic and Scale Model Calculations Core Barrel l SM-4/SM-5 Axial Nickel and Water Stainless Steel l Posi tior , in* Analysis and Sodium Analysis Difference l 8.97 4.12 % 4.20 % + 0.08 % 11.16 6.60 % 7.35 % + 0.75 % l Reactor Vessel l l SM-4/SM-5 l Axial Nickel and Water Stainless Steel Analysis l Posi tion, in* and Sodium Analysis Difference ' 4.59 1.28 % 1.31 % + 0.03 % 9.15 0.0115 % 0.0119 % + 0.004 % 12.17 2.28 % 2.38 % + 0.10 % 19.87 0.158 % 0.106 % - 0.052 % 24.03 1.25 % 1.26 % + 0.01 % 26.59 2.66 % 2.96 % + 0.30 % 29.16 1.67 % 1.58 % - 0.09 %
- From bottom of reactor vessel O
180
O O O e g 60 -- L LEGEND:
'- NICKEL AT ROOM TEMPERATURE 000 IN/IN/SEL)
- --- STAINLESS STEEL AT ELEVATED TEMPERATURE (LOW STRAIN RATE)
/
s**
- 40
=
# , s E ., =* #
M **" w .aa cc # M n* # ! 5; ' _. ='*- j 20 _ - I l O O 2 4 6 8 10 STRAIN (PERCENT) Figure 4 2-1 Vessel Stress-Strain input Data for REXCO-ilEP Post-Test An:ilysis
0 1
===.
s O i s y l a n
# A t
t 8 s e
- T t
s o P P E
# H -
)
E O T p C A R X E N I R - A R 6 r o T f S W s ) T a t a O N L ( 8 E D 8, s C t
)
C R E a R u E p E U n S T
/
P I ( N A n I
/
N EP R # N I i a I A r 0 M R t 0 E 1 ( T - T S S-s s E G I 4 e R N r t U T I S T A A R E P M T P R E O
==.
l B er r a O E T A e r M LE # o O E C O T R S - 2-T S ' 2 A SE 4 L L E N I 2 e r K I C AT u I ig N S F 0 N E G - - E L
- r7i 0 0 0 0 O 6 4 2 552 m$5=
Ob; O g~
water-nickel
^ sodiun. stainless steel
~
1500 l
?
1000 I 1 R 500 ( 1
, f I d' g :'
I , l 0 4, o I 4 0 ; a A 1 & A 0 .1 .2 .3 .4 .5 6 l l l TIME (msec) 1 Comparison of Core Barrel Pressures from Calculations using Scale Model and Prototypic Properties Figure 4.2-3 183
800
,y y(.
600
//
I
/
./ k 400 ,/ '\
.,i
\
/ 5 b
/ \\\ ,
, 200 l
M \ m D C s\\' % v. . O o 3
- \
l wa ter-nickel
-200 <
-- sodium-stainless steel
-400 - - - -
0 1 2 3 4 5 6 TIME (msec) Comparison of Average Slug Velocities from Calculations Using Scale Model and Prototypic Properties Figure 4.2-4 184
O 400 -- l 300 - l 1 l 1 200
- o
'~
$ 100 ()
'o b + ,
8 h s-0 '
~0 O
' ~
-100 water-nickel l
--4-- sodium-stainless steel
-200 -
0 1 2 3 4 5 6-TIME (msec) Comparison of Slug Forces from Calculations Using Scale Model and Prototypic Properties Figure 4.2-5 O 185 1
100 - 50 - f s 0 7 / O " l ng C 1 water-nickel ; i
-100 -
sodium-s tainless steel n n a E 0 1 2 3 4 5 6 I TIME (msec) Comparison of Net Force on CSS from Calculations using Scale Model and Prototypic Properties fi<;are 4. 2-6 186 1
1 (D v - 7000 -I e i 1 1000 1 l 5 .& . 4am- . 3 --- - - - .. Oi El l c' 1
-1000 ~
0 1 2 3 4 5 6 TIME (msec) Comparison of Pressure Under VIS from Calculations Using Scale Model and Prototypic Properties N Figure 4.2-7 187
4 O 3
/
2 1 l l l 1 , n l
;a
$1 l
b Ji 0' ; water-nickel
-1 sodium-stainless steel '
-2 '
0 1 2 3 4 5 6 TIME (msec) Comparison of Core Barrel Strain Energies from Calculations Using Scale Model and Prototypic Properties Figure 4.2-8 188
s U,r 8 6
,c
_w ": : --e
/ ;
4 l
^
^ -
-____I 2
G I 21 b (~jh a w 0< l
--- water-nickel
-2
-C- sodium-stainless steel
-4 - -
0 1 2 3 4 5 6 TIME (msec) Comparison of Vessel Wall Strain Energies from Calculations Using Scale Model and Prototypic Properties Figure 4.2-9 V(% 189
0.20 , wa ter-ni ckel
-C- sodium-stainless steel 0.10 -
2 C. U n as M 0.05 2. a g 9<hn ^ =-2 Y 0 1 2 3 4 5 6 TIME (msec) Comparison of Peak Vessel Wall Radial Displacements from Calculations Using Scale Model and Prototypic Properties O Fi gure 4.2-10 190
i 4.3 Cover Gas Response and Its Leakage Implications O v In Section 3.3.3.3 the cover gas pressure response from the SM-5 test was discussed. This section compares these results with results from simulations of gas response below and above the large margin ring. The SM-5 test had pressure gauges at the lower elevation of the head shielding (P), g in the dip seal region (P)4) and in the gas space above the margin ring (P 15 and P16). The first of these pressure histories is used in conjunctxn with the MAXPRES-2 code (Reference 4-14) to detennine the pressures at These are then compared with the corresponding the locations of P)4 and P 15 experimental loads. The MAXPRES-2 code models a series of annular regions of differing cross-section. The liquid and gas remain, throughout the transient, as two distinct f' ids separated by one planar interface. As the liquid moves up the channel, the gas above it is assumed to compress adiabatically. While throttling effects are simulated, shock wave effects are neglected. Figure 4.3-1 shows a schematic of the region, the corresponding MAXPRES-2 geometry and its dimensions. The gap across the margin ring was assumed to be constant throughout the transient for the reasons discussed in Section 3.3.3.3. An overview of the region together with the locations of the relevant pressure transducers is given in the SM-5 drawing in Appendix B , while details of the margin ring region can be seen in Figure 3.3-18. The pressure loading from the Pg gauge (see Figure 4.3-2) was used to drive the column of water up the annular space. The first 2.6 milliseconds of Pg data were removed from this figure since it takes that time for the transient loading to be transmitted to the lower shielding elevation. Two calculations were performed. In the first no form losses resulting in energy dissipation.or reflection were assumed so as to provide a bounding calculation. As a result, the pressure calculated in the dip seal region, as shown in Figure 4.3-3 was severe and quite non-prototypic. In the second calculation, realistic form losses were introduced between the lower shielding elevation and the dip-seal region, cnd as a result, 191
there is a drastic reduction in peak pressure. The response with form losses is shown in Figure 4.3-4 where it can be compared with the experimental pressure detennined from the dip seal region (P)4). It should be emphasized that the magnitude of the fonn loss used was arbitraily chosen to give an experimental result whose peak pressure approximates that of the calculated value and to show that with such a loss coefficient the calculated response characteristics are similar to the experiment. The most significant conclusion from this analysis is that, by not including energy dissipating fonn losses, predicted pressure can overestimate the actual pressurc. by at least an order of magnitude. This is of considerable importance in leakage assessments, l Figure 4.3-5 compares the predicted and experimental (P15) pressures above the margin ring in the gas space simulating the large riser annulus. The analytical curve resulted from the second of the two calculations above, in which form losses were included below the margin ring. The j observed overprediction may be the result of two effects. First, energy dissipating form losses were not assumed to exist through the complex geometry of the margin ring gas space. As discussed above this is an extremely conservative assumption. Second, the gas gap across the margin l ring region was modelled with uniform cross section which may not have simulated the actual gas gap closely enough. For discussion of the actual gas gap geometry across the margin ring, see Section 3.3.3.3, while for the dimensions of the analytical model, Figure 4.3-1 should be consulted. Even with these conservatisms, attenuation is significant enough to reduce this pressure to less than 100 psi. Current design analysis techniques do not generally include energy dissipating form losses because these are difficult to determine. However, by excluding them, gas pressures associated with short term dynamic loads may be significantly over-predicted in the convoluted annular regions close to and above the vessel head. O 192
O O O a Y N 7
' v 7 l
HEAD 6 VESSEL FLANGE 5 3 /l l 4 4 1 _ l
$ l3 l J3 [
o 2 2 w - - - JL l UNDER HEAD REGION HEIGHT DIAMETER WIDTH SHIELDING 1 (H) (D) (W) Hj j MAXPRES 1 1.69 12.10 0.05 MODEL 2 0.19 12.13 0.075 ,, 3 0.18 12.14 0.056 0; ,, 1 f 4 0.53 12.53 0.4175 L 5 0.19 12.91 0.038
+ + Wj 6 0.17 12.91 0.001 i l _
7 0.87 12.83 0.10 V EXPERIMENTAL PRESSURE
- ALL DIMENSIONS IN INCHES HISTORY APPLIED HERE Figure 4.3-1 Region Edelled in St.\XPRES Anal) sis
3000 i
=
y 2000 - w CC m M w E ti 2 E e h 2 E 1000 - 1I o T C l
/
I I O - 0 1 2 3 4 5 6 TIME (msec) l'iyure 4J 2 I) rising Preouie at Lower Sliichi I' late I:lesation 0926 46 O 194
6000 5000 - 4000 - i
=_2 y 3000 -
i 1
- s <
M m
)
f , I l 2000 - O 1000 I I I i i 0 0 1 2 3 4 5 6 TIME (msec) i a j Figuie 4 3 3 Calculated Pressure liesguinse in l>ip Seal Region (No ()rince 1 osses) l O 195
500 LEGEND: EXPERIMENT
------- C AL CU L ATE D
*\
I, 400 - t , ', Ii s I r
\
I \ l )
* \
I \ I I 300 - 8 4
= * \
? ' \
I \ g i \ m 8
\
g I 8 \
- u. \
E \ 200 - \
\
\
\
\
m N x s e 100 _f
/
~s,'~
~,*%
I I I I I I I I O O 1 2 3 4 5 6 7 8 9 TIME (msec) l'iyuie 4.3 4 Comp.irison of Calt ut.ited anti 1.speiimental l'iewure Itesponse in I)ip Seal itegion (Orifice 1. owes At counteil l or) 0526 4% 9 196
O O O C 100
'j LEGEND:
P
? 15 EXPERIMENT --- m N
2-C / w P16 EXPERIMENT y N CALCULATED 80 ^
/
/
/
60
/
/
/
_5 /
= 40
/
5 i' ) c, n .u ., J., m . .
,. ., e i .,
e m . , ~ m ** , , a s p ' l'- ,sA
'* ' ' 5 'I ,'
',,/, ,\.,
/ 1 '
's l'. ,8 s j ', , ' .
e ,, s. i l', 20 - s . , s ,s , , , - >
- ,s . ;,, i, , s i . ,, '
s,e e
- s a s , , s
... . \l. :, , 'I .
f ., '
- s' -
* ' s j' * se s
,\
\r N', ,' - ' e l
's g ',
0 vv -v yx ' t
,-,'*,-- y s , ','
,8
. y I I I I I I l l
-20 2 3 4 5 6 7 8 9 0 1 TIME (msec)
Figure 4.3-5 Comparison of Calculated and Experimental Pressure Response Abme Margin Ring (No Orifice Losses Accounted for Across \1argin Ring)
References for Section 4 4-1 Y. W. Chang and J. Gvildys, "REXCO-HEP: A Two-Dimensional 0 Computer Code for Calculating the Primary System Response in Fast Reactors," ANL-75-19, June 1975. (Availability: U.S. DOE Technical l Information Center). 4-2 " Nuclear Systems Materials Handbook," TID 26666, Hanford Engineering Development Laboratory, Richland, Washington,1975. 4-3 Metals Handbook, Vol. 2, American Society for Metals, Metal Park, Ohio, 1964. 4-4 ASME Steam Tables, 3rd Ed., American Sociaty of Mechanical Engineers, New York, 1977. 4-5 G. H. Golden, "Thermophysical Properties of Sodium", ANL-7323, August 1967. 4-6 Y. W. Chang and J. Gvildys, " Comparison of REXCO Code Predictions with SRI SM-2 Experimental Results", ANL-78-18 (to be published). (Availability: U.S. DOE Technical Information Center). 4-7 C. M. Romander and D. J. Cagliostro, " Experimental Simulation of a Hypothetical Core Disruptive Accident in 1/20-Scale Models of the Clinch River Breeder Reactor" Technical Report 4, to be published. ( Availabili ty: SRI International, Menlo Park, Calif.) 4-8 B. W. Joe, "SM-2 Post-Test Analysis of the Effect of Material Strain Rates", CRBRP-GEFR-00287, October 1977. (Availability: U.S. DOE Technical Information Center). 4-9 B. W. Joe, " Assessment of Deviations from Prototypic Properties in SRI SM-4 1/20-Scale Model", CRBRP-GEFR-00311, November 1977. (Availability U.S. DOE Technical Information Center). O 198
4-10 G. I. DeSalvo and J. A. Swanson, ANSYS Engineering Analysis Systems i User's Manual, Swansen Analysis Associates, Inc, Elizabeth, Pa, 1975. 4-11 W. E. Baker, et al., Similarily Methods in Engineering Dynamics: Theory and Practice of Scale Modelling, Hayden Book Company, Inc., Rochelle Park, New Jersey, 1973. 4-12 G. Murphy, et al., " Similitude of Dynamically Loaded Burf ed Structures", WL TR-64-142, March 1965. 4-13 G. L. Fox et al., " Scale-Up of Test Results from Simulation j Experiments of a Hypothetical Core Disruptive Accident in the Fast Flux Test Facility," HEDL-TME-74-54,1974, (Availability: U.S. DOE Technical Information Center). 4-14 S. Ranatza, "MAXPRES-2 Computer Code Verification", CRBRP-ARD-0175, July 1977. (Availability: U.S. DOE Technical Information Center). O O 199
O Appendix A Detail Comparison of Experimental and l Calculational Data on SM-2, SM-3 and SM-4 Scale Model Tests O 200
A.1 Introduction This appendix provides a comprehensive set of comparisons of calculated and measured data on the SM-2 through SM-5 series of tests. The calculated data are from the REXC0-HEP (Release 2.) computer code (Ref. A-1). SM-4 and SM-5 were essentially duplicate tests and the test results are very similar. Thus in the post-test comparisons the SM-4/SM-5 analysis results are compared only with SM-5 test recults. A.2 Sumary Description of Data Table A.2-1 provides a breakdown of the data (pressure and strain histories) presented in this Appendix. The pre-test analyses, which were all run with the low strain rate data, are shown in Figures A-1 through A-6 (designated "REXC0-HEP" or " LOW c"). These results came from the analyses discussed in Section 4.1.1.2. The post-test l analyses were run with both low strain rate data (designated " LOW &") and high strain rate data (designated "HIGH L'.') and are shown in Figures A-7 through A-10. These results were derived from References A-2 and A-3 and were discussed in Section 4.1.3. Note that no post-test analysis was performed on the SM-3 test. Table A.2-2 summarizes the data shown in the Figures. l l l l l l l lO l 201
O TABLE A.2-1
SUMMARY
OF APPENDIX A FIGURES Fiqure Nes, Description Test Analysis
- A-1 Pressures SM-2 Pre-Test SM-2 A-2 Strains SM-2 Pre-Test SM-2 A-3 Pressures SM-3 Pre-Test SM-2 A-4 Strains SM-3 Pre-Test SM-2 A-5 Pressures SM-5 Pre-Test SM-4/SM-5 A-6 Strains SM-5 Pre-Test SM-4/SM-5 A-7 Pressures SM-2 Post-Test SM-2 A-8 Strains SM-2 Post-Test SM-2 A-9 Pressures SM-5 Post-Test SM-4/SM-5 A-10 Strains SM-5 Post-Test SM-4/SM-5
- Pre-test analysis designated as "REXCO-HEP" or " LOW 6" Post-test analysis designated as " LOW t" or "HIGH 6" 202 6
~
O O O TABLE A.2-2 COMPARISON OF PRESSURES AND STRAINS REXCO* VS. EXPERIMENTS Gage No. SM-2 SM-3 SM-5 SM-2 SM-3 SM-5 Location REXC0 Expt. REXC0 Expt. REXC0 Expt. P j Core 4000 psi 3300 psi 4400 psi 4600 psi 4000 psi 4000 psi P) P) P P P Core 4000 psi 3750 psi 4000 psi 4100 psi 4000 psi 4000 psi 2 2 2 P P Upper Core 2500 psi 2900 psi 2800 psi 2600 psi 3 3 P L wer Vessel 540 psi 480 psi 5 P P P Vessel Wall at Core 525 psi 540 psi 410 psi 480 psi 575 psi 530 psi 4 4 6 P P Vessel Wall at UIS 460 psi 570 psi 350 psi 400 psi 655 psi 590 psi 5 "5 7 P P P Vessel Wall 750tpsi 530 psi 750tpsi 520 psi 700tpsi 595 psi 6 6 8 P P P g Upper Vessel Wall 2900 psi 165C psi 1750 psi 1000 psi 2800 psi 2850 psi 7 7 P P Head 5500 psi 5300 psi 7500 psi 3500 psi 4900 psi 5350 psi 8 8 P)) SG) SG) Vessel Wall at Core 1.40% 1.62% 1.90% 0.87% SG 2 b6 2 b6 5 Vessel Wall at UIS 0.80% 2.65% 1.25% 1.45% 0.15% 0.32% SG SG Vessel Wall 4.15% 2.30% 6.20% 1.60% 4 4 SG 5 SG 5 b6 7 Upper Vessel Wall 1.70% 2.90% 4.10% 2.15% 1.65% 1.87%
- General Electric calculations using high strain rate properties for SM-2 and SM-5, low strain rate properties for SM-3 1 Average value
1 Pg HEAD w r. _______1 P . SG S (- - ---g WATER ' -- 7 ,--- %
.i
_c=- d_ p P , eSG 4 i
,,,,, i.E LEAD SHOT D[ [, , ,9 ,
STEEL o
, --d +---
y CORE _77 p, , p2
.__d
- CORE BARREL ~
--;,6- y4 pP , 4SG 1 i Py b p.- ) N 1((.'
y,77;pj/gg,gf , f ' "' N'J- % i % p -s . . 3. SM 2 REXCO MODEL SM 2 5000 5000 . _ 4000 . - - ~ ~ - - - -
.. 4000 E . - _ . . _ _ . . _ - - . . . - . . . . - E g...--.- .. - - - - -
l 3000 - - - - - - - - - - - - - - - - - - - l 3000 -- E X P E R I M E N T --- - - ---- - - -- ~ EXPERIMENT,_ ._.
$ 2000 $ 2000 '.'. - RE XCO-HEP .
g R E X CO-H E P ~ ~~ ~ ~ "_- - -- - - - g _ . _ . - _ . - . _ . p 1000 ' - - - - ~ ~ - - - - - p 1000 -h - - - - - - - -- o- ,c.__._._..__...-, c- _. . : , .,c_..__._._._. . . _ . __. -
~
0 - - 0 - " - - - - 0 1 2 3 4 5 6 0 1 2 3 4 5 6 TIME - msec TIME - msec PjCORE P2CORE 600 700 500 -
- - - EXPERIMENT - - 600 g . - . - _ . . _ . - .
g [i [~,EXPEP yN -
' MENT _ , - ~ -- - - -
400 . - - M/ - R E XCO-H EP R E X CO-H EP h l ~ l 400 - '( -- [- - - - - '" r I ~ w l _ [_5 y 300 4' I '-- --h-i l 3 200 - 3 - 6 J 'I - . , , y o _ _ _ , _.
-100 -100 O 1 2 3 4 5 6 0 1 2 3 4 5 6 TIME -- msec TIME - msec P4VESSEL WALL OPPOSITE CORE PSVESSEL WALL ABOVE CORE M A-3929-2 70 FIGURE A-1 PRESSURE VERSUS TIME SM-2 PRETEST ANALYSIS O
204
Pg HEAD P AIR 8 a n A I kA 1 F ~7 P 7, SG S
.m,1 pl=--rh~~ ~p A -A]
s , WATER
-- >/ n i w=~ __ __._,/"6,P SG 4 I />P 6 $..I i ,
i
- I WATER .. .. ~___.~yP _. ' 5SO 2 LEAD SHOT .... ---.- ..
P STEEL u i *
# CORE . _. p,, p
. _ , ~
p su gP2 f CORE BARREL ALUMINUM N F i]p_J, _c_. _ [ pP , 4SG 1
]
1 . m $. : . STEEL
~-* -~~
I
} 3 / N g 'Z111 2.~
a d ~ NEA$$N//dlfS~ , . . . . , . SM 2 REXCO MODEL SM 2
, i i - , 3000 ' ' , , ,
1000 h '
.._ . L_ ! d _ _ , __ _j ---r-
~
2500 A 800 _ ..._..._._._.;_ _R'E XCO-HEP g-f ; 1 R E XCO-H E,P 2000
...j._....- ; - ,
r- i , '~ w 1500 EXPERIMENT \:. 600 . - EXPERIMENTg _ n. g.;;._.g p w -
.2,. - +-- ;
u ...
+
. .t , ;. <; . _ . 'f_ - w 1000 t; j
-H
. -~._ i .e 4 l
200 ,:- p. h.., ye - ,* 7 ' - l ; e 500 - - ' 2 3- h- m- .I- - 0
!i - t. .
, f-l-~.r 0
, g'5-d: r- !-q -
riri--
')
0 1 2 3 4 5 6 0 1 2 3 4 5 6 TIME - msec TIME - msec P6VESSEL WALL P7UPPER VESSEL WALL 8000 , . 7000 "
~ ~ ~
.a a
6000 [ REXCO-HEP_~h. .
. _..__~
5000 1 -t - - - -- w 4000 EXPERIMENTQ QU _i . J 4, g
> 3000 _. 4 _! ,- _ -i __ , _ + ' , . , _ _ . . ._....-
m _ . w r M 200Q j . _ g _ _. k_ g _} _ - ~_g MM . 3 . ._ ._, g.,! y l _ . . _
- o. 1000 ; ; j i 1
-r v itJ + --- -
0 A&rW "~
._r._.__.. ,
-1000 0 1 2 3 4 5 6 TIME - msec P
8 HEAD M A-3929-2 71 n FIGURE A-1 PRESSURE VERSUS TIME SM-2 PRETEST ANALYSIS (Concluded) 205
P HEAD O 8 p
- AIR 8 a_ _...._._n \
I- -T ,p P , 7SG 5
,m - _ _ - . _ -
~-
P WATER 3
.$. 7 77 VESSEL N ~"
q
/rP 6 Nt- fi "
WATER LEAD SHOT [_i_ d +- 5. P5' SO 2 CORE DJ P,P i 2 la. v' CORE BARREL % --r. +. +- p P4, SG 1 q.p . r.y,. 7 [P 2 ALUMINUM N O STEEL N g[ _,M.E ..+ g . _{- Y , g w . ..
?!!'h$$$$$- _., ,.h. .I. '. $
SM 2 REXCO MODEL SM-2 2.0 3.0 ,
. EXPERIMENT . - - - " - - -
yEXPERIMENT - A 2.5 - -- e-E 1.5 . g y
$ 10
-Q R E XCO-H E P 8
g 2.0 - -
-. - - - . R E XCO-H E P - .J 1.5 -
l __ _ ../.......-.- ---_--- l __, 2 0.5 -: - z 1.0 -- T ,/ x 0.5 * --- -~ - - - - --- $ 0 0
$ - - ~ ~ - - - " - - '
-- ~- -- - '
0 - - r
-0 5 ' -0.5 0 1 2 3 4 5 6 0 1 2 3 4 5 6 TIME -- msec TIME - msec SG 1 VESSEL WALL OPPOSITE CORE SG 2 VESSEL WALL ABOVE CORE 8 6
. 5 -----~--~y.' ' s E 6 - - -
y ~ b'~~~~~'-"~ g - - - - . 4 - ;_. . _ - _ , 5 '
'~- _R E..X C_ - -O_ -i .H_E_ P._\_/ - & - ' C_. l _ . R E X CO-H E P .. - - - - - - ~ - -
4 -' 3 l _ l E X_PE R lM E NT_ ,_,'- -- g 3 - -- 2 2 - - ---
- F -- - --
$ 2 . EXPERIMENT y -----------:----.. -. g . .._ g j _ . _ _ . . _ . . . _ _ . _ _ _ __,;.-.. M 1
~ * * -_'---.----:_.l_._._.._..._
0 --f------ 0l 0 1 2 3 4 5 6 0 1 2 3 4 5 6 TIME - msec TIME - msec SG 4 VESSEL WALL SG 5 UPPER VESSEL WALL M A-39 29-2 72 FIGURE A-2 STRAIN VERSUS TIME SM-2 PRETEST ANALYSIS 206
p P HEAD p* 'd en U e T ^'" WATER P7,SG 5 m :.-- -- : =- - ._ ( i . _ VESSEL _._ _ STRUCTURE ,
# WATER w --._.4 ps, SG 2 L NU m q 9 r ej
_ :. &. sg ""' Ny - SM 3 REXCO MODEL SM 3 5000 5000 4000 . 4000
] - *- -- -- -- -
j l 3000 r -- - -EX? T'j--- l 3000 h / / E X PE R IM UN_T _ _, _ _ . _ m ;. k-[pII._ENE'NT m p I L_._ ....
$ 2000 -h
_ _ b WI.. 2000 A - - -- - g _i,f _ . ._ . _ . . p 4 [ _. _ _ ._ .._J-. _ ..
$ 1000
*,- E 1000 *;-
._.'..._m-A Q . _ _ _ j a .
k..,,. _ 0 -- - ----
' j - -t - }[ 0 -
7j j 0 1 2 3 4 5 6 0 1 2 3 4 5 6 TIM E --- msec TIME - msec PgCORE P 2 CORE 3000 600 ;;i;; 2500 $ ------ - - - - 500 b l {1.4
.M E E I l 2000 i-y E XPERIM ENT -- ----
l 400 L, . LOWi - %{.. ..
.g..,
g g 1500
- f. ,--
--t---
/
g ;- - y g 300 f- - b,
- 1. .
._ -- E 2,rTr [ '
3_. 3 1000 , i,n.J-__ c4 3 200 g i ---- 1 _ b,,_y ,- /
.y . .
w 500 l- - N's -- -- - -- p 100 1 jm -- .- -
-. . - -% 4--.-y ._h r -.
s w g: . 0 - - -- -- -- 0 - - .--... . - -
-500 100 0 1 2 3 4 5 6 0 1 2 3 4 5 6 TIME - miec TIME - msec P3 UPPER CORE P4 VESSEL WALL OPPOSITE CORE M A-3929-260 FIGURE A-3 PRESSURE VERSUS TIME SM-3 PRETEST ANALYSIS O :
207
P HEAD p AIR k.WA a
' l 3=p 3 , nn VESSE L V'**
% 9 '
6* N 4 UPPER , P INTERNAL ------ '* 6 " STF1UCTURE _
's- WATER % --.-< - P5 , SG 2 P
CORE BAR CO DQ N P,P 2 3 u cg 4,PSG 1 1.i __. p y EP 2[ g ALUMINUM N g "-'r . b
~
STEEL I q
= s s l
), - , ;
- A . . -
i 8
* ,,...,.' l l
SM 3 REXCO MODEL SM 3 l l 700 . 600 - _(_ _; , 1000 - -
-h u -
b b LOW i f 5 500 - - - -- -
-LOWili- E 800 -
I-h '3
+
l 400 - t- 1 l [ ! 9- 4 .e 600 --- - - Ott * $ 300 w:I h _. u ... -@ -dNh --
-s $ -EXPERIMENT - - ' I lib \
3 -
'tr tv /t9 1 if/U D 400 f.
$ 200 - w t - - H,
.. r , a .V ,,l y, 4' t.- o g .
.. j j hg,.
m ,
&f
-+
k IOO G 200 l ,. t I O
;EXPERI ENT o ' . .. . . .
9 j { { 0 1 2 3 4 5 6 0 1 2 3 4 5 6 TIME - msec TIME -- msec P5VESSEL WALL OPPOSITE UlS P6VESSEL WALL 3000 8000 2500 - 7000 - ' g 6000 __-;_7- g
. 2000 -- - -- - - - - - ~
l ._ _ . ... . . _ ._
' 5000 l ._-i, --+: -
4 1500 - - - - - - -
-- LOW i -
w 4000 - - ) -'~f -.. .l 3000 -
- --+- --+ -s
$ 1000 --- - - r- -t- - -i -
$ --EXPERIMENT i -d w
a 500
-EXPERIMENT -- i 6- kfj y'f g 2000 ----- 4
-- 1 -
t a 1000 b()N h r h.g ..
~,,
0 q-
~j --q! 'T [; k' 0
._. _'p _ q _.{ ..[ 5 2--- f l - - l j j {
0 1 2 3 4 5 6 0 1 2 3 4 5 6 TIME - m:ec TIME - msec P7UPPER VESSEL WALL P8 HEAD M A-39 29 -261 FIGURE A-3 PRESSURE VERSUS TIME SM-3 PRETEST ANALYSIS (Concluded) 0 208
f O P HEAD AIR % P8 SG S 7
- ' - ~ ~ -
p s s" m . p P p SG , _ M[ VESSEL ,_ __ ,
- _d ,
_~Il pP , 5SG 2 LEAD SHO , 3 BAR 4 G1 l
-- ~'~
ST EL SG l ' _
,,,,,, .k e _ ',
J ..... . l SM 3 REXCO MODEL SM 3 2.0 3.0 I g "
. _ _ .y%. _ _ -. _ . .._. LOW I ~
" " ~
.. j 2'5 7 1.5 - - - -[- -O" E
8 . _ . _ .r_._._. . _ _ _._. - . -- 32.0 k 1.0 -: ---- -
-E XPERIMENT' 1.5 E X P E R IM E NT_._ .. _ _ _ 4 .'~ ~
4__. pea
-- F- , . ,; . ...p l l j z 0.5 M - --- - -
z 1.0 ,f LOW {7 , y 1,l, . -.- -. . _ . .. - ng _ _ __ y 0 "' - - -- -- y 0 j -
~
-O 5 -0.5 0 1 2 3 4 5 6 0 1 2 3 4 5 6 TIME _ msee TIME - msee SG 1 VESSEL WALL OPPOSITE CORE SG 2 VESSEL WALL OPPOSITE UlS 0
.. .I _ { .
7 ,,
] ._, 5 . _ __
l c g _._ . , _ ._ _ . . _ , g
- t. ._ .- . . . _
. .f y 4 _ _ ,,, sg g 5 -
g y )
-._ LOW i_ . .,f __..
LOWi
,f 1 3 -
l 4 . - ---- - -
..- .. .' -- l _ _ . .-
7 3 .-._.---. --- .. .. . _. ..
. ; y _
g f
'~
f2 g Z _ .._ _ .T..
- -y.
~1.((~
F m 1 --EXPER IM ENT _- _ _ _h,/ . I ER g jf __ _ ._ 0 Z.Z[_. , , m . . .i .... v. 0 1 2 3 4 5 6 0 1 2 3 4 5 6 TIME - msec TIME - msec SG 4 VESSEL WALL SG S UPPER VESSEL WALL M A-3929-262 HDURE A-4 STRAIN VERSUS TIME SM-3 PRETEST ANALYSIS
)
209
P " 6 HEAD P { AIR 6 37, mm, WATER I _ _ , P5, SG S b, i
% v "::::: INTERNAL x :n i
l STR UCTUR E
-I b' V M " M^""" %:q ---@Z LE D SHO
- STEEL % -
y P, j' =. 1 j,Qhg, :x=% q
- n 5?W, \
CORE SUPPOHT CONE N ,pSG 1 I Y k%---.--{ WATER , _-m . N --.--[
/
l SM 4 REXCO MODEL ; SM 4 l 5000 I 5000 l 4000 $- -- 6 - - - - - - - - "
. 4000 .
" ^-~~~-~~ ----~-'
l*3000 E X P E R I M E N. T. -. . .- . l 3000
- - EXPERIMENT R E X CO-H E P .
R E XCO-H E P l w w : l- . g 1000 a 1000 -k.----------_------, _ _ _ l l 0 -- - O - - - 0 I 2 3 4 5 6 0 2 3 4 1 5 6 TIME - msee TIME - msec Pj CORE p 2 CORE 70 700 1 ._ 600 - - -- 600 -- -
- l -- -
E 500 EXPERIMENT
$ 500 - - - -
I 400 h , , +-- R E XCO-H EP 400 4' !) _ .
~
' _ l; l
w - - -
.__.._o 300 y 300 _ . . ._,;,_. _ __ ',
] 200 5 200 f..' < 5 '00 t \ 100 -
,f
~
0 1 2 3 4 5 6 0 1 2 3 4 5 6 TIME -- msec TIME - msec P3VESSEL WALL OPPOSITE CORE P4VESSEL WALL OPPOSITE UlS (SODIUM OUTLET NOZZLE) M A-3929 -2 73 FIGURE A-5 PRESSURE VERSUS TIME SM-4 PRETEST ANALYSIS 210
k P6 " HEAD A R
- P5, SG 5
*TrT-" "I h , ,
^ * * % ------ **
1 UPPER STRUCTUR E
- - ..# WATER w
_ ___ -<- p..
, P 4, SG 3 LEAD SHOT ----'-" .
U"' ^RL y j], CORE
$k[h p bb CORE SUPPORT CONE k
N s h E2 h'/ h SG 1 kW
~~~
i / WATER k __--d ,; l ----y
/ m
.b.b. . . . .
SM 4 REXCO MODEL SM 4 l 3000 8000 L- - - , 7000 , _7 _ .7_ 2500 -
._g -)4 - .
g 800o ....___... .. ._. l2000 EXPERIMENT __
~
REXCO-HEP l 5000
\!'
-- - -- L .-
4000 EX3ERdENi f g 1500 _
- D 3 D 3000' t.
S 1000 -. . f,000
., ~~ -
0 1 soo
. _ . - - .hj _
.p.)lil; y,,
- . _ . .- - . "c 1000 - -
l [ !;U I , s 0 0 1 2 3 4 5 6 0 1 2 3 4 5 6 TIME - mise TIME - msec PSUPPER VESSEL WALL p6 HEAD M A-3929 -274 FIGURE A-5 PRESSURE VERSUS TIME SM-4 PRETEST ANALYSIS (Concluded) 211 O
I P 6 HEAD f h AIR P 6
% _" 13. _.n*q E '
P VES E
# ~* ' (
UPPER N d j STRUCTURE _
- ,# WATER w -
S .. 9 pP , 4SG 3 LEAD SHOT ORE BARREL N ---[Q[-* ' , '. bfWh!5~'U
)
j .- v
.W
-I CORE SUPPORT P CORE SUPPORT CONE
%N g
q kh4
----'/ , e
'/ ','/ SG 1 (h / WATER % W-2222 .
N
----Q
--y SM 4 REXCO MODEL SM 4 r 0.25 0 *~~% l 0.20 /R E XCO-H E P - l p -0.2 EXPERIMENT p j .p s l
-0.4 \ -~
EXPERIMENT I -06 s# REXCO-HEP } 0.10 l . l
-08 z 0.05 [sY,VN, g
~4 !
c -1.0 2 0 t' .
- 1. 2 -0.05 l
- 1.4 -0.10 0 1 2 3 4 5 6 0 1 2 3 4 5 6 TIME - msec TIME - msec SG 1 CORE SUPPORT RING (C) SG 3 VESSEL WALL OPPOSITE UlS (C) l (SODIUM OUTLET NOZZLE)
G i _ s . E - R E) CO- A E P - - -- {4 - - -
\.
l3 I 2 EXPERIMENT t; 'l i O 1 2 3 4 5 6 Tif-1C - msec SG 5 UPPER VESSEL WALL (C) M A-3929-2 75 FIGURE A-6 STRAIN VERSUS TIME SM-4 PRETEST ANALYSIS 212
I Pg HEAD
\
% AIR 8 f
r- -7 'f pP SG 5 M - -- 7 d -- =---] p WATER ' n
! _ g_ - ! P6, SG 4 WATER -1 y 2 'e"L %
e p SG 2 EEEEMF"I '
- CORE % --U y G[Gf "',% "" %Q
. p, , p
- h. _#
Em&f~ ' '
~
EdWlRlM2%- .....,.l I SM 2 REXCO MODEL SM 2 4000 4000 3500 3500 '
'l 3000 ll I j 'i 3000
% EXPERIMENT l 2500 l 2500 ,t w 2000 { w 2000 \\ -
lw 1000 _n(EXPERIMENT
's*
'. \ HIGH i lw 15* 1000
\\
'h
. ,HIGH i g E 500 E 500
~
0 " ---~~~~"" 0 i i ii
-500 -500
O 1 2 3 4 5 6 0 2 4 1 3 5 6 TIME - msec TIME - msec PjCORE P 2 CORE 3000 , i 600 . . . . 2500 ' ' ' EXPERIMENT' n'
._ 500 ,;i g- \ l 'l 2000 "
EXPERIMENT E 400 '! N '/ I
! 1500 _
HibH i ! j , [ glow' 1 7%
$ 1000 '.t I
LO'W i $ N-E - N '- ' ' ' ' '1 Y 500
~
h \ l \ f i^\ y 0 %c
.r w 100
# " I I *
& lI \
a - E I v . GJ M' i-Ifl\ .. / \ J. OM,#
-500 ,
0 ,l '-%F.'LWFut,i. -,- y
-1000 -100 0 1 2 3 4 5 6 0 1 2 3 4 5 6 TIME - mnec TIME - msec P
3 UPPE R CORE P4VESSEL WALL OPPOSITE CORE M A-3929-2 56 F!GURE A-7 PRESSURE VERSUS TIME SM-2 POSTTEST ANALYSIS l l 213
m O vv w c
;= - - - g _ _ . -- ;k , , , PT*. '
g - -' n P7, SG 5 WATER 7 -
,hP , 6SG 4 7,p VESSE L N WATER ---* 5 P ' 5SO 2 LEAD SHO 2222 ;
P 3 P 5 STEEL --- [ CORE ^T P,P j 2 ALU NUM 22 2 kY " p.
'I ._ fh STEEL m h22222
;@@4Sda:97,
- v 2 J 4 6 4 F S SM-2 REXCO MODEL SM 2 700 . . .
1200 ii i . EXPERIMENT- tow ; 600 -- yl t j 1000 8 500 H Hi -E
. HIGH i s- Q
- h--z LOWi l l 400 I' - ,
l 800 gq 4;l'e~ W ( C 300 ( l' 600 - - I. U- 2 O'ih.' 11
- f. . ' 3 -EXPERIMENT p l W"i '
200 1 t, y y 'fU:( }}_ f 400 q+5N? $ 100 0 i
-7
$ .D nh k -
E 200 g.ey.. ,S i, Il 1i i _a 0
!/ i O<<4 ,10 4
1 A-A
) y y _A 0 1 2 3 4 5 6 0 2 3 4 1 5 6 TIME - msec TIME - msec PSVESSEL WALL ABOVE CORE P6VESSEL WALL 3000 8000
.----f _-( !l ll i i LOW i--
, 2500 HIGH is LOW i- -
ylc'H 'i g -- g ,s g 6000 -g ll l 2000 l 5000 w 1500 !: w 4000 -; ' $ $ 3000 -EXPERIMENT h 1000 -EXPERIMENT -
,7g ;
@ SOO f 1000 ,1 hj 0 '
$5 N. . O ? ^
~
0 1 2 3 4 5 6 0 2 3 4 5 1 6 TIME - msec TIME - msec P7UPPER VESSEL WALL P8 HEAD M A-3929-257 FIGURE A-7 PRESSURE VERSUS TIME SM-2 POSTTEST ANALYSIS (Concluded) O 214
P 8 HEAD Pg
\ AIR
'Q . A P~\ 2 9. P 7, vv ww I.- .dA ~ ,-;ip _ bh,, % . ,M SG 5 SG 51. Ir --- _ WATER n y
.1, '_a--- I 7
/ 6*SC4 P SG 41 frP 6 ,,
SG 21 > ORE P 3
- CORE BAR % ' ! P4, SG 1
^
SG 1 J P g m E. - e ,,. .,. I SM 2 REXCO MODEL SM 2 2.5 3.0 iiii
! I EXPERIMENT i 2.5
--y*.. LOW ;
. 2.0 [
l 2.0 h t '5
--^ f 8 / t / 'T, 8 r LOWe.
1.5
- y. . . ....*-g...
, j ,
i n ', l 1.0 , 7 N HIGH e, ; l g 1.0 y f .- HIGH E . ,,,,,e, j ,, ,,, z f . ' [ 0.5 0.5 f,#
\ $ 0 Il $ 0 -
-0.5 -0.5 0 1 2 3 4 5 6 0 1 2 3 4 5 6 TIME - msec TIME - msec SG 1 VESSEL WALL OPPOSITE CORE SG 2 VESSEL WALL ABOVE CORE O 3.0 f%y t l l LdWE-, ! LOV[i 5 2.5 I g g EXPERIMENT f '* fp 4 g 2.0 l HIGH i g 4 - .;
HIGH I/ 8 f , -- .. . ~~J j ir - l 3 /, ! I / ' g 2
/ M -
2 l 'O i [i
< r/l, 4 '
cc EXPERIMENT 0.5 ' b 1
/
/ f *.'
/ 1: m )
0
-0.5 3 4 5 6 0 1 2 3 4 5 6 0 1 2
' TIME - msec TIME - msec SG 4 VESSEL WALL SG S UPPER VESSEL WALL M A-3929-258 FIGURE A-8 STRAIN VERSUS TIME SM-2 POSTTEST ANALYSIS
/^ ( s 215
P 11 HEAD W' E P 7E g
~~ * # '
N 24
, UPPER 8
. .#i p
h STRUCTURE WATER %
~l ,
< yPp n i O_ -.-,_4
_ SG 5 LEAD SHOT -
-;A .
" ^
CORE l p- . ((l. M CORE SUP RT P -~~" CORE SUPPORT CONE N / ( g p< WATER kh q.-
-r -
j, fp
'i P r p 5 __._ 5
~ ---
%g, Q ,
- 5. . . . . . ,
SM S REXCO MODEL SM S 6000 6000
} i : i I.. 7 7._+.-. . 7 i
i i
}--}
_ 4000 .._.p. . -7 g .___._ ._ 4000 q_ ,7-8 . . . . . --
-._f..- L. E --
73-l 3000 t- .--(d r-F -...- l 3000
, ,;_ ,__._._L___._._ m p p_ . __ ._ __.. _ . _e. t. _-
$ 2000 h EXPERIMENT--
$ 2000 M EXPERIMENT b-- - - --
g _., __t_.. __. _ _ -- .__ m 2_ i j ! ! 4 _ _ -. d-
$ 1000
-b.-
y 4 - p--f. ..- 1000
-Kb m T - -
E--~~~ - l
-.7-- 3 0 0
-- - j -- 7-j-q3- j -j j j 0 1 2 3 4 5 6 0 1 2 3 4 5 6 TIME - miec TIME - msec Pj CORE P2CORE 700 i i i i i i 700 j i
_ _ _ ; } I I EXPERIMENT ! 600 M ' fi - E XPERih1ENT ! 600 5 500 ~' ' kc LOW l' d 500 LOW) I
! L .
i l 400 ._: ;4# 1 +.h,. 7 I HIGH I ! l 400 ,' ;
!*DI l ' HIGH I l4- f.,M vb .
j
.4 . .pm
> o,s E 300 .-p- ' l'h j i/.a
;] y 300 ,, 3hj;r[
200 y 4 -+ h 200 h g, g thN- Uk g 100 J- 4,%,/, -+qt,.h E 100 'f.[,.;% f,' Y l t- y M 0 # .-' O q b( ~
-100 ' i i I i i 17 100
! ' ' i I 0 1 2 3 4 5 6 0 1 2 3 4 5 6 TIME --- msec TIME - msec P5LOWER VESSEL P6VESSEL WALL OPPOSITE CORE M A -3929 -264 FIGURE A-9 PRESSURE VERSUS TIME SM-4/SM-5 POSTTEST ANALYSIS O
216
l l l l I ' HEAD P AIR 33 A.y _)Lf" I - Ik 4 . .ss
- 1^ _ _ _ .
Pg SG 7 WATER
~ 77 g& Jew & 77-~ ,
f
;k V-
~ ~~ # 9 VESSE L f N
__ _ la
'+
flP ' UPPER n INTERNAL a
---t'-, j h STRUCTURE _
,# WATER w -
5 _~t._ - y y,SG S
- P l
LEAD SHOT -..-.%. ubh , CORE STEEL % D P l P/ h b2g
]
i f ALUMINUM -g % _,_gr ,pP6 f 3 i i *h CORE SUPPORT PLATE N. f -
-k,. - a N
CORE SUPPORT CONE %r ---' /,<> I r w ', l'J(b L Mw p
" WATER kW fp
' P g
__._ - a# S
~'
SM S REXCO MODEL SM 5 700 - 1400 1---y. ,LOWl ! ! ! GOO w%--HIG H I i 7 d 1200 HIGH J i 500 .i j ,, g; ;i & juo l' . , l Til i -"'J, "T Nii.e. i f; w 400 g,g 9 (is-
,.y
;7 7,
~;
w I 800 %a7
"#Q" l ,-
d-
, [. ,
1 N? h-W
$ 200 p,p khh 1M t-n --%j.- $ 600 -E Xc P E R IM,E NT- hl'f' yk[lj.
g 2
; m 4 i j .
y f' , 100 , W , , f
}
EXPE NT 0 l ) - [ 100 0 1 2 3 4 5 6 0 1 2 3 4 5 6 TIME - msee TIME - msec P7VESSEL WALL OPPOSITE UlS P8VESSEL WALL 3000 5000 . - .--
! ! I -h , ! -d- h h ---
ll l 2500 -E XPE RIM E NT tJLOW l 4000 -EXPERIMENT -
-l LOW i -
._E E I -
-t p l 3000
'MU+- -~~~- --
l 2000 ._ w 1500 d w e-E Utt
$ ; 3 'fg $ 2000 -
g 1000 - ' r, g c g . . ,, ! 4 y 500
;p M%
-n
" ,h, y 1000 a j!
l i b:Tj[g, a , . p, j ,,:,7, ,, , A I'I O -^ 0 r IN l E[ -If I 6 I il l ~ I [ 0 1 2 3 4 5 6 0 1 2 3 4 5 6 TIME - msec TIME - msee P6UPPER VESSEL WALL PjjHEAD M A-3929-265 FIGURE A-9 PRESSURE VERSUS TIME SM-4/SM-5 POSTTEST ANALYSIS (Concluded) s 217
- i. - -.
7 O I " HEAD AIR 11 f" I]t . (-}.
.y_tp-p' xeny ,",^,',5" m - .
N
.f"a "
[ [ UPPER ,", STRUCTURE
- 7{: #j '
L WATER % 7777;;
----' p P7 , SG 5 l
' t.
a4Mo"EAD SHO;;^,"""%:q STEEL % jp!yP j ,
; ;] g l P[Ja jP2 gI ALUMINUM %
[{7 ,MP 6 _; CORE SUPPORT PLATE % ----'/
. f f CORE SUPPORT CONE N q 'l
{\
-. : v WATER k%----j p .
.__--/ - *yP 5
% --- -p
(
- f;;/ .
SM 5 REXCO MOOEL SM-5 0.40 s 0 0.35 5 ~ E 0.30 & e
.es-m w EXPERIMENT u m _,~ _%
e W 0.25 y4 !.OW e, & ._! ' , tow'; g 0.20 j lh - 3 l _ 2 0.15 ..q"M_t.n -* "'A -
- ' ,s z 2 2 0.10 'H;gc;47bA/ ! .. .J s.
EXPERIMENT H 0.05 - p 1
/ ,1 HIGH i M -
L.---. m , 0 'l I n ~~" l l I
-0.05 I 0 1 2 3 4 5 6 0 1 2 3 4 5 6 TIME - msee TIME - msec SG S VESSEL WALL OPPOSITE UlS SG 7 UPPER VESSEL WALL M A-3929-266 FIGURE A-10 STRAIN VERSUS TIME SM-4/SM-5 POSTTEST ANALYSIS O
218
1 l l References A-1 Y. W. Chang and J. Gvildys, "REXC0-HEP: A Two-Dimensional Computer Code for Calculating the Primary System Response in Fast Reactors", ! ANL-75-19, June 1975 (Availability: U.S. DOE Technical Information Center). i A-2 B. W. Joe, "SM-2 Post-Test Analysis of the Effect of Material i Strain Rates", CRBRP-GEFR-00287, October 1977. (Availability: U.S. DOE Technical Information Center). A-3 B. W. Joe, " Assessment of Deviations from Prototypic Properties in SRI SM-41/20 Scale Model", CRBRP-GEFR-00311, November 1977. (Availability: U.S. 00E Technical Information Center). l O
. \
)
O 21 9 : I l l
i l i l I t i O; l i 3 i I 1 j t t .: i 4 e i l l 4 1 i APPENDIX B SCALE MODEL DRAWINGS FOR SM-2 THROUGH SM-5 i U 4 I i a N I j O 220
K b
\
f 4 e hip -- ' - - - --- i b .
+G '#
h '- 1y f_ *
' 4' 'N
, ]-- .au . zic{, L,Lsl ,
w mN - m = ='
. p.,q
/ ;e r ,, .m
=-m m y
f / -TT[* , . - o.. J-9 ~
- ._--_.9 x s s
.__..___g t-M..ch </ j/-
3 4.u
' Js/
m I, _ ..
'*T' "
/
-, f h
/
/ .) %y i
w svu e, -n. - "g-Et 2 1/ y , f// j'y e,, p - A :. , .
' -'_ 3 --N --_. _. y 1--
- - -H-Q, ]4f,5&
~. ,- ;j. .; _
i7
+ . _,. -
s 5 3 g t v.,
--' . 41 i
r - t 4 et ~~ y'v" _..-.,h , l l \.
\.
i
,,,- J. . ( ' ~~ . 1 ,
- ,a 4- - lT -r/ - dl :. [b. ,p(/
. l .. N1
*.1.4
> l' _ ..~llil
.\
\
\
ry
~ I j!
l t
\' V ,f i/' i -
"R' j'~4, ml q
1.-
\_n =om? Y, 4"e ' ' /p h.,q.- . .
s u /- / :-._ L e
. L, , . c yp I jg 1,p.
i rr -- x .,. m_ n. '
\ e,.....,
q \ ' f* ~. _ j L. ,. g.: _..,...,y/. ..,,.
. m.,-- -/ % .r,g.a
~ .,e., ,- l; I l
s\ - - s dJ'" _.'_
, N '., . ;,/ D %f, j
~
g
'A j/ / k J[ -
.1 ~
( - m y..wwww1 - - mf
<y -
/ ,D n . ,w,,,
.f':4< u ,.4, s1(/c, o_ ,
x. g, c- - ; .p;4' :,'7.', c_, jf_/
.L. o
*. g %-,y;e;-(
.. --, . -tk*** s G.\g y s N, ,; . , . ,/ w. M *2'A~~' y.- ~.D $s.
~. 9
.N
~N s'
%,.u..sa.uuu l
(a-w \+
... my ,. ,,, u ~
l )' A* u-. . . ,, . . , , % *.~.l N y[4-iw . . , < . . ........' .a e-. e 1 l l l l l 1 1 1 I I :( l \ l l h h l 1N ., i l 1 l l 1 i .m < = u.u.
- ca,, a. wu
- n~ m >~M r i) l iu ,, , i
, /*- /* e. .: /,* j' l% y . i / l / . ... .ur. . .,.m , [., ,, / . ,/ t i ! l w/ .i.,*~ ..... a k / = ,/ , ,/ \. 'n a j u.j. . . J. ,j *****!!! M ( j y s s 3 s. & .S t. a2s r' s 1 - W => r:1 r r --' -r- s-h:' t ~r- rr -r" M ". 2y' l / p ,, .y _ + . _ _ _ _ , g- s N, gl,_ ; , , , , . - . - - . . . - . / \ . .... .,. .. y. e. j i ;; ,r,r.- ,- / ; . ., f . . . f -...mam.._1\j 7_ ' gr o ' '~;/ _ - ,., y s j[ y!,n<:r ~,7: r -~: , ~ L, ,,, ,! l'J:H ! $ py,w:[; u s W e[ ' t!l , ..,,y,-., A*~-- - -- -- , 'l --J i; , iy ;q u UI : g,l::i: Li.4 q4, i H jj 4 . > ~ . - - . - j ' i , .. j -- - l ~ ' }. . 'e l i x . lr~. <. ,- . ,m m
- f. ~. ~.p l = , ,
t 4 4 , is ~ -. :,,;. *~: ,* " _ ~- _ . $. 4 J.,._nl,,_7,,4.' j 3
- _i!4 ,k_;t
_. I . _.._,.. .r . , 1 llq1 .c i r4 j ,~~ y1 a ,;. - t j , ii N \x __ , .pt[l'_L.1'pLL:7. _J ; c x. .s. }) ',.i ., .r; eN'\ g'Ab ;,;j': \ \y'&.. ' :~'::."::=:':::::'*' ], ~ g- - l l Lp ., l 3kg;j 7"~~ 1 A '"[ .M , d k y*p AMN4 ,.' . .~ ~.:,'.w:c., , ' t'r.:: = ~.:-*~ Yt :r*:~a *:', ; ~ l ',".,.:.l\. i . :.,~:. . . \ s 'N s J \ 6L \ w] l'x l .. 7.: _ n 4 , , ,, , a.u,-,=.- y< = = )y=v=< == (q 3 I ' - . . . . . . . . . v n,- s x a~ %, ::. : ;~ .;- . . , .. , \ t.,,,,t...... - -*~~-- \ A / j ~ p~ j ; - ~ ~ ~ ~ - . \'s <}d ,.,_,,,,.,y 'Nw'l s .'l+ . . , \ <.,. .,.,. iM. . U,' ' .,'.~' ~ ~ ,i ~ ~ > 3. -u
- l. . . sr. ..
,-. ,,,.m.,,,,,,, . ~ - . .. ~~~ D w..va,... W. 4 ,, tQ, J). s W. #c" ~~~ '~ ~" > . ~. .. .a .~c- c..,; <-> \.,.'...,,.-. \ ,' I i 1 y6, , g ..sa:ruu [ f**f ,o ,,, L.** e., W y4%. eJ M B-3929-221 FIGURE B-1 DETAILED CROSS SECTION OF SM 3 (SM-2 dimensions identical. In SM-2, UIS mass located on top of vessel head.) I l 4 s 221 1 3 4 1 l [ 1 ' 6 ./ 1, () AJ.- ', i.": ,*-**~, ._, _ t.,,-..., 4 _ g ,[ . t ny - - ^ , , , ,.. _ i., i 7,,,,,,,, ,.., l.A' %u..p r-- e *_p,. % ~ '.',-m**- .- 7 , ,,og ~ .a. ,. g.i j , 4 . . .-.a.... . , ,-> _ w , ,e . ,, /, - i ' ~- ~ < / e ._ % ., -a.,, a n,! ~ ,.tr . .
- s. .., w e
/ ..,7 ( , f, , , ......h.s,-, - , Q, N. g a rt tu4 (~ ~ .e ~~~~s -- m - - 4 ,/ 4 3 3
- 49. + -m -
, j -,r -tg,. , i ,. ,- />.; . .e..,.,m. a, - 4 , s , ., . ___. m .1 , 9 i l \_ > - \ ' ' , --' *) _'~ \ . rkeJ ,7 . ~J . I .. + / , J .x.. . . . _ . . c ;- {'s s-. m M v ,, q .{vPg-r, _' f ( , %.n_dQQgg 4s .x a ._ ._ Q pq p c ,._ _ . . . _ . .g - .,,,,_.. T d 4.- s w' %OO.w.q :;:u.c - . . _ - - , ~ '~ 4 s- 0 0 e r / j_ i g ' .' %[.lC 1 ' \ A' ' " i t D. 5, c - j ,/ , J ..,, ,e . -. .,.,...,.. , ,,._.%q~..- 1 am. . s. Ix ?%.q- g-t .5 -,-. ,f i ..ib-s _,~, .w f! .. , ;,< 'y L_.._.," 'k a (.. h,_; ,, i m % m g'. 1-ld.) [y, ". .L , w.e ~.._- [q_['.,'.ww/, [.x-N E..h f!l} *g m ,,- _y _ ... p}, W' ' g ...,,.y ^ - 2.- _U %.y p.., ->< n , aa....,,..,/, - wp ""** . M," .. 17t 7 ,.,,o.2,, -' / ""o _ '. / =-, i .- - + u+ , ...a....,,,. . . , .
- s. t . . . , -
~ ' ' * ~ ' <; - . . . , _ . . . . ~ , . . . .7.. ;;,g ..-.,-.n.. , , . . , 1 l l .D 1 i 1 ( / .. 4 '),.. .g - 4C ,- 1 N f 1 1 1 1 1 1 1 i 1 1 I I I 1 l l l \ l l ! l t se a ' ' * *
- 9. M *
- w * * * *
- in X\ NN , , , , _I
! ' "IN s's N t 1 ', . .p b '\m. A 1. ,* a p ~~'.<. r \' m .r. . , . m ,.. _. , , a - - .m. -,- , =3,,nA a :=m ; ~m ' ' o, ', ' ff *T',K',' ",,',"h. .,, ,, . \ j ~.a,o.a..,e,..nr.i.,.,< ~ Y s.jw l ~ ~ , , l La , , ,, , - . - - y , j k * '..::D . = ' ( _.. ' 't 6 'd. ^ .fp- J a $9 i; V i.f,4J,/yl, 1. - < , ew gm y . . {T; w,olGk.Y,. ; j d , g75, i . .w- P- . ! ;1,'
- ! y ,, !
q ; ,gd , . ene, I -----. >1-.' > .; ;!. Jii! i
- e. ,,.A. . . 4. - ,it. g. l;i l
, - - - . . . n,7 e . 1 e o ** . * *s ore 9 ...e . .se .4,; .y, ,,,. ; ; ;~ ; - . - . . ' . - . - - - . - -h;,. , -.7 ; ; y, ,l ;3 . - .. .. .. .. .. ,. ,. ,. ~ , ~ , , . l qt ] y. - - - s .]- , . , , , 4 . . . . ~ n, e ., , . . - , . , . 4..._, , ,,=.,j ..g_ -.gl . p . 134- . [ gtl. j, .,ggl ;, lL[._, j _ {l _-.Mq . -.-+ ' j -__. ~ f~ a, i -3, _q. . , t._._ .. i , pJq mum fj~ j[21 % L n ._ sa ,
- m. _J, .
1r **d f ( af g
- w. y p y'MN'lMv j . .
,e . f.(;,:A.. ,x n! Ji/fLW 1- 1 l JW ~*- --
- s. . . .. - . - /
~~ ~~* ~'s /4 1. . . . i .;d,. ,w *'.
- J~ , s%, ,J wsGw n, a<%%.& .us g/]e jq g '.U14
&i . . . . . , ~ ~ ~ ~ - ' - " ~ ' ' g// . .......~,o .- y, <'. n , , , k, or f t g y . 'e,).. ,! yi -~Y )* *"* .'.a - _ , , , - .u., ,
- n. ., p.' ,
An t / i.y 5 L.: , .y. , . /, q n .x a : .. y. L x = 4 =- - - J' sm % ' ....,.~.n. .-.....,,,...o , . . ~ . . .*-*"' . . . . ~ . , .-- -v y'KvW .'\qN . o osu a o. ;, g NN1%. e o ne s eu en e**. <**wo*.m= r ' ,a - A f \ ee g .4 e n'o.w 'st 'E _ _ : . , , . , ' , , ** ' ' . .= * . , , f,n . ses., e er .a oe ,4 m A. .#d ar.yd .e' e d, 8. Og,Fg3 y f ee. -e e e < e'., o o. . ' * *
- ne "%" , . '*
. ? LF.L " . .e., M B -3929-23 2 FIGURE B-2 DETAILED CROSS SECTION OF SM 4 l l l \ l l l l l l 222 l 1 l 1 I i i 1 l I f \ ( $ 1 1 ee b , se 4% * / , 'a{ l .e.e*a.er,.,e,, A \ \ ~ 1* I ' .*\' g * ,e r f N ,N
- il"Y, * * **,*'Ya
\o [ p r : y v' . . :; * \. R , . aa s g s'J ;.: ; 4 *;,';" y '. ' /..qb N A'q & ,n/s '- js ll{' ,' ' % . , sfin g, , l*.:,: ',::",,',, ~YM IG ,f'b ' ,-ig Q , 3
- td e.
~ ,,. ' ~ .n g~f( y'. .&1 - f &' -; / ^; w . j 'o l m x , _, k.~t -,ug%m;,4 N q, t ~,y. & ' 4 - c : -~*y. f i'. ', . . ,fQ j . % n; \~ .
- g. ..,
7 ; 9 . u.. .a-, , %e 4 . I ' - y-r.-,%'N$mi,T.~TS NG c..,,, k,, [ [ f cf s 'O 'N 5\ U~ i,- ; ,; s.___.._. $+ j'{9: ,- , ~. s s ,3 , , . _ . I s. ' ef I ~ ~ %f f - b --4 ..._1 - 4 4 /j Q 'sf 4 y. *'-k,. f '\ . ,.k ~ ' ~ ' + a .? . ,. ./ .' .- - , t. i.' h. , .t$' .. -- ]l _h gA < 4~^~ + .E--- . 0 s w e g): - ^-^ s ~ T' j l O 1 p; ^^.can..3 N:n}t ,, g H e . '. - .o. ~~< - ~ y a. ,2 <~ ~ a - A .. - tc g' o b ,,; g.. ,d ea pf3 wg f
- \./..h i,q
-t , .1 n <^~ .n n i I (47' # t> \ , , . ,(/J ' *y - ./ - Ol u U. 4,'[:.[n)(&,&,, W. ~ ' ~- 4-.--Q)?tyht ",J' u &'\ %jI b b },l',s.'g,' W .,'. < -jl,~, ;l
- ,9p k>:w , .n hlt qf ' , ::. ,
. L t< v' N-c y'_- w s " ~~ .L S o- . <Nt...7,- .J- 1- , , . .y$ 3 .; ,-t.P,(k~ >"ts. p p- d,/,, . .r.,- .' (, p ' , yo'. -h,-- 3./ c > s s ole'e s,- : aan1 i y e:q . w ),j t , ,_ , , *~ \,','Q<0 0m.%- Oy,Of &-va, l':
- v. ... . , ,
. ~.',o . {4 ,. en&nJ. L f y.4.. - * /,, ,, nel 46 7 - ( p ~C, ,' x- Q f*b :* t ~Q .. y , . .~- r'.. '}'v!'a,. ,Q..g ,, .,
- s t..p.ql, r e .&- ., c m
, g-d: )1-
- s. .
_h-il o I' , s'_ ; .j,9,; q. .,, ,, eV 4] 'ji ., R ,'A k;u... +, . - a p E '.*,,+'",~:: I O.. v,y e.3 " ~ ,'7,,3 .Qy .k<=~ 7 :,,;;~g",* - 4 , u , ,\y &. lt , g *\ ,,- ~ ,1 5 . ~
- w ,
g , e - , , ' ' ,f,,, . , ';, ;', *7 ' .,..,a I= ,e* . e ss I 8 v-**_ ,i-4 A. es 4.m-e : 6g ..a,e, , , se Je F I,, t.
- p I,
8 ~-va. n o'e, *e
- j l', '7 }'.. . 1 b w.' .).
.us.. ; n -a . ~ _4 ';, ,' . m-i[jM. p ...( c - , , , ec c. e-<eg s ,et rda-e J, 4 J , 8e ,,, p g *,*s *p 'M ' "t' ,j_'*_.,f _,; t.4- ,L<-
- y. ~~ -
.y.y i 1 1 1 1 1 1 /f l ). t! l 1 -t t m g 4 Vq' . - ~ ~ . , . .. . -a N s, . ..g.,.... 5, - * '
- k. .. 3 a,
M ai a
- 2. . .5.
~~ ;*' ,:~:..=. ddLL. . .t w' ? .'dW1.' . J, v;,A--h- M <- a > ', n (, _- , . ..-c. ~...<... . ' - om .i,. .' \_ , ,f T.'.**. ., s, ',y s \" ,l, <, . <-- M r !cw , , . .,..~.,.,..u., q s_ ,, \ ' N ' q 4 l(,N l ,'W~s . ^_(' l. , . m' d ;t,, //p s ; nj~.,.; m !Wh ' (' ', . \ i i' II . M, p.
- 3 3
.t # + i 4-3, e- -{ ; ' '(*/ j.p ! j I l dli .,l.l )! ,l ! , M, ti,f i ' d c,!- j H; A, , t I . -~ a r c t < .r. / . , ,s .\, ) , e - - - -+y .: ;,z : ::::,- ,\ \ %. ., .m,4 ,.,c 3- ' ~' , i kp; - r ! ;~ ;;--!; s 4 a nn 8i q L t. r / \, ' 4g93y j l l !. i u,, j , g .f.L n31 ~" " .p k , pre' ,, 2 f'* ji . _ \, i . *L' \, l ?3 }' , l l pf, v fr d -*jr "l4- - ~~r ar', 1,~ p t 1 ..\ * (!(0 1ll - ~ - i ..\;y .y I- {, jj * * ~ * - r v l-l .' i,l ,ji . ,;I ,f,,k:1 '. ' f;5 ',1,1,;.",.,', r .' ? V '4 .r .'//'I," F/f ** 't.._., << e < t i/.! v' t (s"'s.;_ li,N 6 , I! r fi g;:ll/ v
- n. ~ ,
e j .J s ' E 1 { . . . , , . , */ / !~. , _,;, u =7.:~~' ~~ ; , ,y;,a';.h,,j..,a ,.., as:.1+Ll'L,./'3~-~~~~~~.' . ;/ u '^ ., . L ,, ,,'.? * \, ~ ~ ,y-~ / ,r U i .#, .ep. e \ i / \ . ., ad , ./ r a w ;. a: ,. s - . ,4' ^ s s
- r . z :o . . - r- 1 :x- x O-,
, . a /z., 2. z.. r-.2 ...s y : , . w. . r .a : ,s --.z-.-- 32 *< % /'.,' \ x 4 . . , . , i-e n-n o e,,e 8 a * ' y & , . ~ . * * * * - * * - ~r. ,,,. _ , , , , , , . s I . , , , , , . , . . . . ,a ,, e ,..,.,,.....u,mes . ~ . . n m in.no , m . - .----- -. u a, i;. ~ J s pa W* i . , , . . . . , t _ e .q, w .+ , , . . . . l>.st_.s , ,, I 1 eende 5 g ges f . L . p , /. '~ I ' M 8-3929 -238 , (t FIGURE B-3 DETAILED CROSS SECTION OF SM 5 223 .,}}