ML12243A447
| ML12243A447 | |
| Person / Time | |
|---|---|
| Site: | Palo Verde  |
| Issue date: | 08/30/2012 |
| From: | Mary Anderson, Moran T, Carol Nove, Pardini A Battelle Memorial Institute, Pacific Northwest National Laboratory, NRC/RES/DE |
| To: | Arizona Public Service Co |
| Gibson L | |
| References | |
| Download: ML12243A447 (30) | |
Text
1 UT in Lieu of RT for Nuclear Power Plant Applications Carol Nove - NRC Traci Moran, Tony Cinson, Michael Anderson, Al Pardini - PNNL NRC/APS Public Meeting August 30, 2012 NRC, Rockville, MD
Outline Project Objective Equipment Specimens Data Acquisition Data Analysis Summary to date Recent Modeling Work Upcoming PNNL Work Not yet considered Summary of Navy UT/RT Program 2
Project Objective Effectiveness and Reliability of UT & RT for NDE Activities (JCN-V6097)
- Objective of the project is to evaluate the capabilities and effectiveness of ultrasonic testing (UT) to replace radiographic testing (RT) for Section XI, repair/replace activities (RRA).
- Assess whether would also be applicable to Section III, new construction
- Related to ASME Code Cases N-659 and N-713.
3
Equipment
- 450 KV X-ray system
- Computed radiography w/ digital flat panel detectors
- ZETEC Dynaray PA System
- 0.2 - 20 MHz, 256 channels 4
Specimens 4 borrowed Carbon steel pipe-to-pipe welds
- Diameters - 14 and 16
- Thickness - 0.75 through 1.091
- Implanted Welding flaws
- Lack of fusion (LOF), Lack of penetration (LOP), Crack (CRK),
Slag (SLG), and Porosity (POR) 5 Flaw Type Length, mm (in.)
Min Max Mean Planar LOF 5.7 (0.23) 51.4 (2.02) 11.1 (0.44)
LOP 3.4 (0.14) 13.8 (0.54) 6.5 (0.26)
Crack 10.3 (0.41) 39.2 (1.55) 23.9 (0.94)
Volumetric Slag 6.4 (0.25) 51.1 (2.01) 14.6 (0.57)
Porosity 3.2 (0.13) 7.8 (0.31) 4.8 (0.19)
Data Acquisition RT
- Single-wall, film and CR
- True State - length sizing and flaw type UT
- 4.0 MHz TRS, linear PA (32 X 1 elements per probe)
- Insonification angles 45 - 75 degree
- Axial Raster Scanning, both sides of weld
- 1/2 V, full V, and 1-1/2 V sound path 6
Detection Reliability - UT and RT found both implanted (planned) flaws as well as bonus (unplanned) flaws on all four carbons steel specimens 7
Data Analysis Implanted Bonus indications Top View (C-Scan)
Data Analysis Detection Reliability - Flaw Maps were created for each pipe depicting length and location for both RT and UT (from each side)
- Green = Flaw manufacture provided implanted flaw location
- Red = RT indication
- Blue = UT indication 8
Data Analysis Detection Reliability - Implanted flaws in carbon steel piping UT and RT appear to have similar detection capability for volumetric flaws
- UT missed 1 small porosity (5.3 mm (0.2 in.) in length)
UT has a better detection capability for planar flaws
- RT missed 5 planar flaws
- Planar flaws are more likely to grow throughout the service lifetime of the plant and could be more detrimental 9
- Average UT Signal-to-Noise Ratio
- for planar flaws = 29.1dB
- for volumetric flaws = 22.6 dB
Data Analysis Detection Reliability - Bonus flaws in carbon steel piping UT missed 38 Flaws 35 volumetric flaws that were nominally less than 4 mm in length (smaller than focal spot size of PA probe) 1 porosity was large enough to be unacceptable RT missed 32 Flaws All assumed to be planar flaws (LOF)
- Missed due to relative orientation of x-ray beam and/or being masked by flaws in the same axial and circumferential positions 10
Detection Reliability -
Single sided vs. double sided exams - Some fabrication flaws may not be detectable from a single side UT exam Depth Positioning - Important advantage of UT is its ability to detect flaws stacked throughout thickness in same circumferential location Data Analysis
Data Analysis Sizing Reliability - limited data set on carbon steel piping Length Sizing
- Many factors influence sizing results such as size of flaw, surface connectedness, flaw type, inspection angle, use of 2nd and 3rd leg of sound
- POR, SLG, and LOF sized better with -6 dB sizing method
- LOP and CRK sized better with -12 dB sizing method RMSE within Section XI, Appendix VIII acceptance criteria of 19.05 mm
- Typically applied to cracks and may not necessarily translate to fabrication flaws Over-and undersizing
- Potential to make a call to accept a rejectable flaw or reject an acceptable flaw 12 UT Length (4.0 MHz)
Flaw Type Near Side Far Side 6 dB RMSE 12 dB RMSE 6 dB RMSE 12 dB RMSE Porosity 3.18 (-)
7.08 (+)
3.45 (+)
6.89 (+)
Slag 5.41 (-)
8.96 (+)
8.70 (-)
8.99 (+)
LOP 8.46 (-)
5.44 (-)
6.40 (-)
2.30 (-)
LOF 8.24 (+)
10.45 (+)
6.03 (+)
8.26 (+)
Crack 11.08 (-)
7.31 (+)
9.46 (-)
7.45 (+)
Summary of detection reliability result to date Detection Reliability - carbon steel piping
- UT has a better detection capability for planar flaws
- UTs most significant detection limitation is detecting small volumetric flaws that have less of an impact on the structural integrity of the specimen
- One of the main advantages that UT has over RT in detection is its ability to assess through-wall depth information as well as its ability to discriminate flaws stacked throughout the thickness in the same circumferential location
- UT was able to detect flaws using the 2nd and 3rd leg of sound when the 1st leg of sound could not be used due to a weld crown
- UT did not always detect flaws from both sides which would impact single-sided examinations 13
Summary of length sizing results to date Sizing Reliability (limited data set on carbon steel piping)
- RMSE within Section XI, Appendix VIII acceptance criteria of 19.05 mm for all fabrication flaws evaluated
- Typically applied to cracks and may not necessarily translate to fabrication flaws
- UT is more likely to oversize SLG, POR, and LOF, which may cause the flaws to be rejected when they should have been accepted
- UT tended to undersize CRK and LOP, which may cause an unacceptable flaw to be accepted 14
Recent modeling work PAUT Probe Design 4 MHz TRS, linear PA 2x(32x1) elements per probe used to date Probe sufficient for 1 thick CS Modeling performed to determine extent of probes capabilities Idealized probe designed for use in thicker CS samples (approx. 2)
Resulted in 5 MHz probe with 2x(32x4) elements Modeling Beam directivity acoustic pressure calculation using Huygens principle existing 4.0-MHz probe produces an unwanted grating lobe when sweeping (directing) from the part center angle Beam simulations Ultravisions modeling tools 5.0-MHz probe produces a better focal zone throughout the thickness of the material and has better focusing capabilities, thereby improving signal-to-noise values for flaw detection and characterization for thicker carbon steel materials (i.e., Navy plates) 15
Recent modeling work Simulations in 1 pipe specimen TD Focus 3/4T, 45 Deg. (Bounce Metal Path) 16 4.0 MHz 5.0 MHz
Examination of several Navy UT/RT test plates Collect and evaluate data on several of the remaining Navy UT/RT test plates New PAUT probe designed for thicker material will be used Flaw maps are available to provide true state (location, size, and classification) of discontinuities in the plates.
Sectioning and metallography cost prohibitive to perform.
Cannot section mockups scanned to date as they belong to industry.
Determine if Navy study conclusions that UT is a good alternative to RT (for submarine hull weld inspections) may translate to NPP CS materials/piping systems as well as the PAUT method 17 Upcoming PNNL Work
18 Not yet considered
- Consideration of whether ASME Code,Section XI, Appendix VIII UT performance demonstration requirements are applicable to repair/replacement activities
- Flaw types and inspection volumes very different from pre-/in-service inspection
- Acceptance criteria for fabrication type flaws
- Applicability of RT acceptance criteria? Differing physics of methods
- Applying Section XI acceptance criteria may result in accepting welds with poor workmanship
- Applying Section III may reject acceptable flaws causing unnecessary repairs
Supplemental Information Summary of U.S. Navy program Ultrasonics as an Alternative to Radiography (UT/RT) for Submarine Hull Weld Inspection 19
Summary of Navy UT/RT Program
- U.S. Navy program Ultrasonics as an Alternative to Radiography (UT/RT) for Submarine Hull Weld Inspection
- Mid 1980s - early 1990s
- The Navy believed that UT for weld inspection would result in reduced cost of inspection, increased productivity, immediate inspection results including knowledge of depth of discontinuity, and potentially more accurate sizing of the discontinuity.
- Objectives of program:
- to determine if structural welds could be ultrasonically inspected with repeatability and reliability comparable to that obtained with radiographic inspection
Publications summarizing this program:
Lebowitz, Carol A. and DeNale, Robert, Evaluation of a Computer-Assisted Ultrasonic Inspection System (P-Scan) for Structural Weld Inspection, Proceedings for the 4th Canadian Forces/CRAD Meeting on Research in Fabrication and Inspection of Submarine Pressure Hull, 4-6 June, 1991, Halifax, Nova Scotia, Canada.
DeNale, Robert and Lebowitz, Carol, "A Comparison of Ultrasonics and Radiography for Weld Inspection," Review of Progress in Quantitative Nondestructive Evaluation, Vol. 8, Eds. D.O. Thompson and D.E. Chimenti, Plenum Press, New York (1989).
DeNale, Robert and Lebowitz, Carol, "Detection and Disposition Reliability of Ultrasonics and Radiography for Weld Inspection," Review of Progress in Quantitative Nondestructive Evaluation, Vol. 9, Eds. D.O. Thompson and D.E. Chimenti, Plenum Press, New York (1990).
Lebowitz, Carol A., "Evaluation of an Automated Ultrasonic Scanner," Review of Progress in Quantitative Nondestructive Evaluation, Vol. 10, Eds. D.O. Thompson and D.E. Chimenti, Plenum Press, New York (1991).
DTRC-SME-90/30 February 1990, Ultrasonics as an Alternative to Radiography for Submarine Hull Weld Inspection, by Robert DeNale and Carol A. Lebowitz (not open literature; however, copies may be requested from NAVSEA 05P24) 21 Summary of Navy UT/RT Program
Techniques evaluated in UT/RT program Standard practice manual ultrasonics (MUT)
- Krautkramer-Branson USL-48 flaw detector with 3/4 x 1, 2.25 MHz, 60-degree shear wave transducer
- Eight shipyard inspectors (equivalent to ASNT Level II)
Standard practice radiography (RT)
- X-ray, cobalt-60 and iridium-192 sources with Kodak types AA and M film (Ir/AA most commonly used in shipyards)
- Eight shipyard film interpreters (equivalent to ASNT Level II)
Computer-assisted ultrasonic (CAUT)
- Danish Welding Institute P-scan System (Model PSP-3) with 3/4 x 1, 2.25 MHz, 60-degree shear wave transducer
- Five to eight shipyard inspectors using a manual weld scanner (CAUT) -
each completed 40 hours4.62963e-4 days <br />0.0111 hours <br />6.613757e-5 weeks <br />1.522e-5 months <br /> of formal training in operating the P-scan and evaluating data prior to use
- One experienced operator using a manual weld scanner (CAUT*)
- One experienced operator using an automated scanner (CAUT**)
22 Summary of Navy UT/RT Program
Evaluation of standard practice UT and RT:
Test plates:
18 x 2-ft.welds fabricated for study with purposely induced discontinuities SMAW and GMAW (automated and semi-automated) welding processes 24 x 24 x 1-1/2 thick HY-80 steel Slag, lack of fusion, incomplete penetration, cracks, slugs, clustered porosity, and scattered porosity 18 x 1-ft. welds cut from decommissioned submarine Consensus discontinuities identified by reviewing all inspection results from the exams of the 36 test plates, characterizing as to discontinuity type, and for several welds, verification by sectioning and metallography.
Evaluation of Computer Assisted UT Up to 8 inspectors used the CAUT equipment to inspect 15 fabricated test plates and 17 of the welds cut from the decommissioned submarine Discontinuities in several test welds were verified by sectioning and metallography 23 Summary of Navy UT/RT Program
Initial Results From 3024 inspection reports on 36 test plates, 212 consensus discontinuities identified.
UT had highest detection rate, followed by RT with X-ray, Ir-192 and Co-60
- For RT, as energy of source increases, number of calls decreases. Also, the higher energy sources preferentially missed planar discontinuities.
MUT detected more discontinuities than RT, and in particular, MUT detected more planar discontinuities. However, MUT detected approximately the same percentage of volumetric discontinuities as RT.
Of rejectable discontinuities (those that are detected and rejected by consensus), UT rejected more than RT.
Specifically, UT rejected more planar discontinuities than RT.
24 Summary of Navy UT/RT Program
Metallography Five plates containing 34 consensus discontinuities were sectioned for verification purposes.
The classification assigned to 78% of the discontinuities verified by metallography agreed with the metallographic interpretation. This result indicated that the discontinuity identification procedure was a reliable method for classifying discontinuities.
UT detected more metallographically rejectable discontinuities than Ir/AA radiography, regardless of discontinuity type.
- Data indicated that UT performed by experienced operators correctly dispositions 2x as many metallographically rejectable discontinuities as did Ir/AA RT.
- UT had fewer incorrectly dispositioned discontinuities than did Ir/AA RT.
25 Summary of Navy UT/RT Program
26 MET MUT CAUT CAUT*
Ir/AA Detection 34 25 16 29 17 Correct disposition 34 20 10 23 12 Summary of Navy UT/RT Program
27 MET MUT CAUT CAUT*
Ir/AA Total 34 20 10 23 12 Planar 22 14 6
14 7
Volumetric 12 6
4 9
4 Summary of Navy UT/RT Program
Assessment of inspection repeatability MUT and CAUT are more consistent than RT when discontinuity is planar RT is more consistent when discontinuity is volumetric Overall, UT and RT repeatability comparable when examining purposely introduced flaws UT will be most consistent method for rejecting flaws in automated welding processes where flaw types are expected to be planar RT will be most consistent method for rejecting flaws in manual welding processes where flaws are expected to be volumetric 28 Summary of Navy UT/RT Program
Error Analysis Error analysis was performed on 34 discontinuities that were found during metallographic evaluation of 5 test plates Null hypothesis is that a discontinuity is acceptable Type 1 errors occur when an acceptable discontinuity is rejected - this results in increasing production costs Type 2 errors occur when a rejectable discontinuity is accepted - this results in not meeting the design criteria Results indicated that a similar number of errors occurred with each method evaluated; however, when looking at the data in terms of errors made per total number of discontinuities detected, UT performed better than RT as it detected a greater number of discontinuities resulting in a smaller percentage of errors.
29 Summary of Navy UT/RT Program CAUT*
MUT Ir/AA CAUT Total Detected 29 25 17 16 Type II Error 5
4 5
6 Type I Error 1
1 0
0
Navys Conclusion (from DTRC-SME-90/30):
- Based on the results of the work presented herein, ultrasonic weld inspection, if implemented under carefully controlled conditions, will provide an acceptable alternative to radiography as an inspection method for submarine hull welds.
30 Summary of Navy UT/RT Program