Analysis of JNES Seismic Tests on Degraded Piping (NUREG/CR-7015)
On this page:
Download complete document
Manuscript Completed: May 2010
Date Published: July 2010
1 T. Zhang, 1F.W. Brust, 1D.J. Shim, and 1G. Wilkowski
2 J. Nie, 2C. Hofmayer
1Engineering Mechanics Corporation of Columbus
3518 Riverside Drive
Columbus, OH 43221
2Brookhaven National Laboratory
P.O. Box 5000
Upton, NY 11973-5000
S. Ali, NRC Project Manager
NRC Job Code N6076
Office of Nuclear Regulatory Research
U.S. Nuclear Regulatory Commission
Washington DC 20555-0001
This report describes efforts conducted by Engineering Mechanics Corporation of Columbus (Emc2) under subcontract to Brookhaven National Laboratory (BNL) for the U.S. Nuclear Regulatory Commission to analyze and better understand a series of degraded pipe tests under seismic loading that were conducted in Japan. These efforts were undertaken as part of collaborative efforts between the United States Nuclear Regulatory Commission (NRC) and the Japan Nuclear Energy Safety Organization (JNES), who conducted a multi-year test program for the Ministry of Economy, Trade and Industry (METI) of Japan to investigate the behavior of typical Nuclear Power Plant piping systems under large seismic loads. JNES provided the test results for the evaluations given in this report. Emc2 worked with BNL on the degraded piping system analyses. This report describes Emc2’s post-test analyses of the large-scale piping system tests and discusses insights gained from this program. The analyses in these efforts are of value to understanding margins that degraded piping has under seismic loading for: (a) Transition Break Size rule for CFR 50.46a applications, (b) Leak-Before-Break analyses in NRC’s SRP 3.6.3, and (c) pipe flaw tolerance by ASME code evaluations.
The JNES tests analyzed were of both the simple degraded component tests, e.g., a nozzle with a short section of straight pipe and a circumferential surface crack in the joining girth weld; as well as combined component tests from a 1/3-scale section of the main recirculation system of a BWR. The combined component tests had circumferential cracks at a nozzle girth weld and elbow girth weld of the main pipe run, and at a tee girth weld in the branch pipe. The combined component tests were conducted at room temperature with pressure and simulated seismic loading in one direction.
Emc2’s analyses used a “cracked-pipe element” (CPE) concept, where the element represented the global moment-rotation response due to the crack. This approach was developed to simplify the dynamic finite element analysis. The failure modes automatically captured by this method are ductile tearing during the final failure event, and very low cyclic fatigue damage with large-scale plasticity. Crack growth due to high-cycle fatigue or low-cycle fatigue with small-scale plasticity requires a separate analysis than the current “cracked-pipe element” approach. This “cracked-pipe element” methodology was developed in the International Piping Integrity Research Group (IPIRG) programs that the USNRC and Japan were members of in the early 1990’s. In this program, the validation efforts increased in complexity by comparisons with pipe tests with circumferential through-wall and surface cracks under quasi-static bending, fully reverse cyclic loading, and inertial loading to examine dynamic loading aspects with cyclic hysteresis loops of a circumferential through-wall crack. Those pipe tests data were developed in prior USNRC or IPIRG programs. The JNES simple component test was used to fine-tune the “cracked-pipe element” model since the toughness of the weld metal in the JNES tests was not available.
In the JNES combined-component experiment where leakage occurred (Test FTP-4), it was determined that the loading at the crack location was in the small-scale plasticity range of the “cracked-pipe elements.” Hence, the leakage was caused by low-cycle fatigue with very small-scale yielding, rather than by ductile tearing or low-cycle fatigue with larger plasticity that would occur if the loading were near the ultimate load capacity of the cracked pipe. The procedure used to make predictions of low-cycle fatigue crack growth were based on the Dowling ΔJ procedure, which is an extension of linear elastic based fatigue crack growth methodology extended into the nonlinear plasticity regime. The predicted moments from the CPE approach were accounted for using a cycle-by-cycle crack growth procedure. Only moments large enough to produce significant crack growth per cycle need be included, and the U.S. Air Force fatigue crack growth code, AFGROW, was used to make these predictions. The predictions using this approach compare quite well with the experimental measurements where leakage occurred during the sixth load block.
As part of the analyses for the FTP-4 experimental case, the amplitude of the seismic loading was increased until failure of the surface crack occurred by cyclic tearing in the ABAQUS model with the “cracked-pipe element” with just one loading block. The margins on the applied accelerations in the FTP-4 test were about a factor of 5.0. The material property parameters were then changed for the cracked-pipe sections as well as the rest of the model to estimate the margins if the JNES test was conducted at the BWR operating temperature at 288°C (550°F). The margin at this temperature was conservatively estimated to be 3 on the applied acceleration in the FTP-4 test. This was a slightly conservative analysis since the viscous damping of the uncracked piping system at this higher level of excitation was kept at 2% (the “cracked-pipe element” and the elbows behave plastically and contributed to the increased overall damping effect). The minor conservatism from the damping at the higher excitation is also true for the room temperature case as well.
To better assess how a full-scale pipe system might behave with similar loading and crack size, an elastic-plastic correction factor for larger-diameter pipe was applied to the load capacity of the FTP-4 pipe analysis at 288°C. An evaluation of one of the component tests showed that the JSME Z-factor for GTAW/SMAW welds was quite close to the experimental results, whereas the ASME code would have over predicted the test failure load by 30-percent. Using the JSME GTAW/SMAW Z-factor equation, the full-scale pipe system margins on failure load and acceleration that were applied in this test for a similar non-dimensional crack size were conservatively estimated to be 1.06 and about 2, respectively. Results here suggest that larger cracks could be tolerated in this system test.
These margins were still above the assumed nonlinear stress correction used in the NRC’s work on “Seismic Considerations for the Transition Break Size” in NUREG-1903, although analyses as conducted in this report would have to be performed for the crack sizes and typical pipe system used in the Transition Break Size seismic analysis to better quantify the margins that might exist. Such margins would also affect LBB analyses, ASME code criteria, and on-going probabilistic pipe fracture analyses such as the NRC’s xLPR efforts that are just starting up.