Validation of analysis methods for assessing flawed piping subjected to dynamic loading

Technical Report · Mon Aug 01 04:00:00 EDT 1994

DOI:https://doi.org/10.2172/10176738· OSTI ID:10176738

Olson, R J; Wolterman, R L; Wilkowski, G M ^[1]; Kot, C A ^[2]

---

1. Battelle, Columbus, OH (United States)
2. Argonne National Lab., IL (United States)

Argonne National Laboratory and Battelle have jointly conducted a research program for the USNRC to evaluate the ability of current engineering analysis methods and one state-of-the-art analysis method to predict the behavior of circumferentially surface-cracked pipe system water-hammer experiment. The experimental data used in the evaluation were from the HDR Test Group E31 series conducted by the Kernforschungszentrum Karlsruhe (KfK) in Germany. The incentive for this evaluation was that simplified engineering methods, as well as newer ``state-of-the-art`` fracture analysis methods, have been typically validated only with static experimental data. Hence, these dynamic experiments were of high interest. High-rate dynamic loading can be classified as either repeating, e.g., seismic, or nonrepeating, e.g., water hammer. Development of experimental data and validation of cracked pipe analyses under seismic loading (repeating dynamic loads) are being pursued separately within the NRC`s International Piping Integrity Research Group (IPIRG) program. This report describes developmental and validation efforts to predict crack stability under water hammer loading, as well as comparisons using currently used analysis procedures. Current fracture analysis methods use the elastic stress analysis loads decoupled from the fracture mechanics analysis, while state-of-the-art methods employ nonlinear cracked-pipe time-history finite element analyses. The results showed that the current decoupled methods were conservative in their predictions, whereas the cracked pipe finite element analyses were more accurate, yet slightly conservative. The nonlinear time-history cracked-pipe finite element analyses conducted in this program were also attractive in that they were done on a small Apollo DN5500 workstation, whereas other cracked-pipe dynamic analyses conducted in Europe on the same experiments required the use of a CRAY2 supercomputer, and were less accurate.

• Research Organization: Nuclear Regulatory Commission, Washington, DC (United States). Div. of Engineering; Battelle, Columbus, OH (United States)
• Sponsoring Organization: Nuclear Regulatory Commission, Washington, DC (United States)
• DOE Contract Number: W-7405-ENG-92; W-31109-ENG-38
• OSTI ID: 10176738
• Report Number(s): NUREG/CR--6234; ANL--94/22; BMI--2178; ON: TI94017485
• Country of Publication: United States
• Language: English

Related Subjects

22 GENERAL STUDIES OF NUCLEAR REACTORS
220200
36 MATERIALS SCIENCE
360103
COMPONENTS AND ACCESSORIES
COMPUTER CALCULATIONS
CRACK PROPAGATION
CRACKS
DYNAMIC LOADS
FINITE ELEMENT METHOD
FRACTURE MECHANICS
MECHANICAL PROPERTIES
NUCLEAR POWER PLANTS
PIPES
STRESS ANALYSIS
TESTING
WATER HAMMER