last modified: 16-JAN-1991 | catalog | categories | new | search |

NESC1070 PRAISE-C.

PRAISE-C, Double-Ended Guillotine Break (DEGB) Breaks from Weld Cracks in Light-Water Reactor Piping System

top ]
1. NAME OR DESIGNATION OF PROGRAM:  PRAISE-C.
top ]
2. COMPUTERS
To submit a request, click below on the link of the version you wish to order. Only liaison officers are authorised to submit online requests. Rules for requesters are available here.
Program name Package id Status Status date
PRAISE-C NESC1070/01 Tested 16-JAN-1991

Machines used:

Package ID Orig. computer Test computer
NESC1070/01 CDC 7600 CDC CYBER 830
top ]
3. DESCRIPTION OF PROGRAM OR FUNCTION

PRAISE-C is a probabilistic fracture mechanics code used to estimate the probability of double-ended guillotine break (DEGB) in light water reactor piping due to the growth of cracks at welded joints. Pipe failures are considered to occur as the result of crack-like defects either introduced during fabrication, or that initiate after plant operation has begun, and that escape detection during inspections. PRAISE was developed to estimate the influence of earthquakes on the probability of failure at a weld joint in the primary coolant system of a pressurized water reactor. An initial hydrostatic proof test, pre-service non-destructive inspection, and periodic in-service inspection can be simulated. PRAISE treats the inter-arrival times of operating transients, such as system heatup and cooldown, either  as constant or exponentially distributed according to observed or postulated rates. Leak rate and leak detection models are also included. Earthquakes of varying intensity and arbitrary occurence times can be modeled.

PRAISE-C extends the capabilities of PRAISE-B to include a tearing instability failure criterion for carbon steels (supplementing the original net section stress criterion used for austenitic materials), and an advanced probabilistic model of stress corrosion  cracking in stainless steels (Type 304, Type 316NG "nuclear grade")  used for BWR reactor coolant piping. The stress corrosion model is semi-empirical in nature, and is based on experimental and field data. The model considers crack initiation, including the number, time, and location of initiated cracks, in addition to the effect of stress corrosion on crack growth rates. Various phenomena are considered, including environment (i.e., coolant temperature, dissolved oxygen content, level of impurities), applied loads, residual stresses, material type, and degree of sensitization. By allowing cracks to initiate after reactor operationhas begun, the simulation is not restricted to the original "single crack" assumption. Consequently, crack linking (i.e., multiple small cracks joining to form a single large crack) is considered. Separate descriptions of welding residual stresses are included for small-, intermediate-, and large-diameter piping.

PRAISE-C can be readily modified by the user to accommodate longitudinal welds, as well as a wide range of crack growth relationships, failure criteria, and stress intensity factor formulations. The model could also be modified to include measures now being used on actual reactor piping to mitigate stress corrosion cracking, such as inductive heating stress improvement and weld overlay.
top ]
4. METHOD OF SOLUTION

Failure probabilities are estimated by applying Monte Carlo methods to simulate the life history of the selected weld joint. During each replication of the Monte Carlo simulation, PRAISE assumes that failure, defined as either a through-wall defect (leak) or a complete pipe severance (LOCA), results from the fatigue-induced growth of an as-fabricated interior surface circumferential defect (crack) assumed to be two-dimensional and semi-elliptical in shape. The initial crack size is selected by stratified sampling from probabilility distributions of crack depth  and aspect ratio. Crack propagation rates are governed by a Paris-type relationship with separate cyclic RMS stress intensity factors for the depth and length. Both uniform through-the-wall and  radial gradient thermal stresses are included in the calculation of  the stress intensity factors.
top ]
5. RESTRICTIONS ON THE COMPLEXITY OF THE PROBLEM:  Maximum of -, 10 seismic intensity classes.
top ]
6. TYPICAL RUNNING TIME

The four sample problems require 3, 0.9, 0.9, and 1.6 CP hours, respectively, on a CDC7600. NESC executed the second sample problem in 2 CP hours on a CDC CYBER170/875.
NESC1070/01
NEA-DB executed the test cases included in this package (with a reduced number of replications) on a CYBER 830 computer. The following CPU times were required - case 1 (500 repl.): 5308 secs; case 2 (with 1000 repl.): 1992 secs; case 3 (1000 repl.): 1992 secs; case 4 (10000 repl.): 15032 secs.
top ]
7. UNUSUAL FEATURES OF THE PROGRAM

When stress corrosion is considered, execution time is significantly longer than for PRAISE-B.
top ]
8. RELATED AND AUXILIARY PROGRAMS

PRAISE-C is an updated release of the stratified Monte Carlo simulation code, PRAISE (Piping Reliability Analysis Including Seismic Events) and supersedes the PRAISE-B code previously distributed. PRAISE assumed that exactly one initial defect existed in the weld and that the earthquake of interest was the first earthquake experienced at the reactor. While  retaining the "single defect" assumption, PRAISE-B extended the capabilities of PRAISE to include: the effect of stress corrosion on crack growth rates (no initiation), the contribution of welding residual stresses to the stress intensity factors, and the influence of high-cycle low-amplitude vibratory stresses.
top ]
9. STATUS
Package ID Status date Status
NESC1070/01 16-JAN-1991 Tested at NEADB
top ]
10. REFERENCES

- E.Y. Lim:
  Probability of Pipe Fracture in a Primary Coolant Loop of a PWR
  Plant, Volume 9: PRAISE Computer Code User's Manual, Load
  Combination Program Project I Final Report,
  NUREG/CR-2189 (UCID-18967), August 1981.
- T. Loo and R.W. Mensing:
  Probability of Pipe Failure in the Reactor Coolant Loops of
Combustion Engineering PWR Plants, Vol. 2: Pipe Failure Induced by    Crack Growth,
  NUREG/CR-3663 (UCRL-53500), September 1984.
- A. Brueckner-Foit, Th. Schmidt and J. Theodoropoulos:
A Comparison of the PRAISE Code and the PARIS Code for the Evalua-    tion of the Failure Probability of Crack-containing Components
  Preprint Nuclear Engineering & Design 110 (1989) 395-411
  (Available from Elsever Science Publishers B.V. (North-Holland
  Physics Publishing Division), Amsterdam.)
NESC1070/01, included references:
- G.S. Holman and C.K. Chou:
  Probability of Failure in BWR Reactor Coolant Piping
  Vol. 1: Summary Report
  NUREG/CR-4792 (UCID-20914) (March 1989).
- T.Lo, S.E. Bumpus, D.J. Chinn, R.W. Mensing, G.S. Holman:
  Probability of Failure in BWR Reactor Coolant Piping
  Vol. 2: Pipe Failure Induced by Crack Growth and Failure of
  Intermediate Supports
  NUREG/CR-4792 (UCID-20914) (March 1989).
- D.O. Harris, D.D. Dedhia, E.D. Eason, and S.D. Patterson:
  Probability of Failure in BWR Reactor Coolant Piping
  Vol. 3: Probabilistic Treatment of Stress Corrosion Cracking in
  304 and 316NG BWR Piping Weldments
  NUREG/CR-4792 (UCID-20914).
- D.O. Harris, E.Y. Lim, D.D. Dedhia, H.H. Woo, and C.K. Chou:
  Fracture Mechanics Models Developed for Piping Reliability
  Assessment in Light Water Reactors
  Piping Reliability Project
  NUREG/CR-2301 (UCID-15490) (June 1982).
- N. Storch et al.:
  TV80LIB Graphics Library
  LCSD-436 Rev. 0 (February 3, 1981).
- L. Eyberger:
  PRAISE-C Tape Description and Implementation Information
  NESC Note 89-46 (March 31, 1989).
top ]
11. MACHINE REQUIREMENTS

The sample problem required 155,000 (octal) words of memory on a CDC CYBER170/875.
NESC1070/01
To run the test cases on a CYBER 830 computer, 132712 words of main storage were required with an additional 21610 words of LCM.
top ]
12. PROGRAMMING LANGUAGE(S) USED
Package ID Computer language
NESC1070/01 FORTRAN-IV
top ]
13. OPERATING SYSTEM UNDER WHICH PROGRAM IS EXECUTED:
NOS 2.4 (CDC CYBER170).
NESC1070/01
NOS2.5.1 (CYBER 830).
top ]
14. OTHER PROGRAMMING OR OPERATING INFORMATION OR RESTRICTIONS

Memory
must be preset to zero prior to execution.
PRAISE-C calls the LLNL TV80LIB graphics routines FRAME, MAPARAM, MAPBOX, MAPS, PLOTE, SETCH, SETCHM, TRACE, and UX80ID; these routines are not included. Suitable alternatives must be provided to obtain graphical output.
top ]
15. NAME AND ESTABLISHMENT OF AUTHORS

            G. S. Holman,
            Lawrence Livermore National Laboratory.
top ]
16. MATERIAL AVAILABLE
NESC1070/01
File name File description Records
NESC1070_01.001 Information file 71
NESC1070_01.002 JCL and control information 17
NESC1070_01.003 PRAISE-C FORTRAN source 8073
NESC1070_01.004 Sample problem 1 17
NESC1070_01.005 Sample problem 2 17
NESC1070_01.006 Sample problem 3 17
NESC1070_01.007 Sample problem 4 17
NESC1070_01.008 Sample problem 1 output (by author) 812
NESC1070_01.009 Sample problem 2 output (by author) 767
NESC1070_01.010 Sample problem 3 output (by author) 767
NESC1070_01.011 Sample problem 4 output (by author) 668
NESC1070_01.012 Sample problem 1 output (NEA-DB) 813
NESC1070_01.013 Sample problem 2 output (NEA-DB) 768
NESC1070_01.014 Sample problem 3 output (NEA-DB) 768
NESC1070_01.015 Sample problem 4 output (NEA-DB) 669
top ]
17. CATEGORIES
  • I. Deformation and Stress Distributions, Structural Analysis and Engineering Design Studies

Keywords: Monte Carlo method, fracture properties, light-water reactors, loss-of-coolant accident, pipes, probabilistic sys assessment, reliability, seismic effects, stress analysis.