ML19212A535

From kanterella
Jump to navigation Jump to search
Research Info Ltr 13:transmits Verified Model for Predicting Residual Stresses Resulting from girth-butt Welding of Pipes & Estimation of Residual Stresses Resulting from Weld Repairs of Vessels
ML19212A535
Person / Time
Issue date: 11/11/1977
From: Levine S
NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES)
To: Case E, Minogue R
Office of Nuclear Reactor Regulation, NRC OFFICE OF STANDARDS DEVELOPMENT
Shared Package
ML19211A049 List:
References
RIL-013, RIL-13, NUDOCS 7912140493
Download: ML19212A535 (3)


Text

N i'a

've..:

33piW p

7

.s

..s NOV 11577 r~ 3 q

Wk't (. - : -

h(k[ ~

~~

U. C.

MEMORANDUM FOR:

E. G. Case, Acting Director Office of Nuclear ~ Reactor Regulation Robert B. Minogue. Director Office of Standards Developcient FROM:

Saul Levine Director Office of Nuclear Regulatory Research

SUBJECT:

RESEARCH INFORMATION LETTER - #13 - RESIDUAL STRESSES IN WELDS j

Introduction The purpose of this Research Infonration Letter is to transmit to NRR a verified model for predicting residual stresses resulting from the girth-butt welding of pipes and the estiration of residual stresses resulting from weld repairs of vessels. This model was developed by Battelle Colurbus Laboratory under the sponsorship of RES and can be used as needed in licensing reviews and standards developrent.

Backgrounc Residual stresses due to welding have been of concern in safety analyses for LWR prinary syste:n cor.ponents. Residual stresses.arise from the differences in temperatures due to the cooling of the deposited weld metal and the heating and coollnt of the neighboring base metal. Sub-sequent weld passes also contribute to residuel stre'sses by reheating and remelting metal. The heat input, weld speed, weld rod size, weld 4

metal and base metal properties, and the rechanical restraint of the structure all have an influence on the residual stresses. Ordinarily these stresses 'are relieved by a postweld stress relief treatment, but for most welds rade in the field, this treatrent is not possible. Residual stresses can be quite high, reaching and even exceeding yield stress levels. The subsequent addition of operating stresses produces localized stress levels that could be excessively high. Such conditions of localized high stress increase the tendency for cract initiation to occur. Thus, it is important that some r.ethod be available to reasure or estimate the residual stresses in welded components for LWRs.

"2 "2" 1568 132 l

e.t. >

i 2

NOV 111977 Desults Analytical methods have been developed and validated at Battelle Coliadms Laboratories to detannine residuar stresses in girth-butt type welds due to the weld-joining process. A computer code AXISOL has been developed to predict residual weld stressas in girth-butt welds in pressure vessels and pipes. Validation of the procedure has been accocplished by com-parison of predicted results with residual stresses obtained through measurements rede on four different pipes with three different pipe diameters and in a weld repair of the 6-inch-thick, 39-inch-diaceter HSST intermediate test vessel. The computer model predicts the actual residual stress levels well and, sore frportantly, the predictions accurately describe the compression-to-tansion changes in residual stress in going from the base metal on one side of the weld, through the weld metal, and to the base metal on the other side of the weld.

Comparisons between predictions and measured vale, agreed well for both the inside and outside of the pipes. Details of'the computer code and results of the validation comparisons are presented in the enclosure.

Evaluation The results of this program have been reviewed by the NRC Vessel Integrity Resean:h Review Grtup, and by the staff of the HSST Program at Oak Ridge.

It is the consensus of these experts that the calculations and measurements have been most carefully done and, thus, represent the best method for prediction of residual weld stresses, Furthennore, the code has been shown to be capable of predicting residual weld stresses in small diameter (4.5 inches) pipes as well as thick section pressure vessels which required the modeling of sor.= 1000 weld. passes in a 6-inch-deep pressure vessel weld repair. The procedure is versatile enough that both NRC and EPRI plan to use it as an aid in directing research in improving weld practtces for reduction of weld stresses thereby enhancing the quality of fabricated products.

4 This model could be used in the licensing process to aid in the evaluating of cracking that has occurred in girth-butt welds in.niping.

It should also prove to be useful in any safety evaluation of proposed repairs by weld buildup in nozzle corner regions after cracks have been removed, and in vessel weld repairs.

Attachcents The results of the study in which the model was developed and verified are described in the final rtport, " Residual Stresses in Girth-Butt Welds in Pipes and Pressure Vessels." prepared for the U.S. Nuclear Regulatory Cocr:ission Office of Nuclear Regulatory Research under Contract No. AT(49-24)-0293, Task 1 August 1977, currently in

.,,,u swoneas, W

, _ ~

$&TS D 1568 133

( ',

Q

^:

(

(

3

, NOV 1 1 1977 publication. This report is sumartzed in the enclosure, " Evaluation of Method for Predicting Residual Stresses in Girth-Butt Welds." A paper describing the model and comparisons with laboratory measurements for residual stresses has been accepted fo presentation at the American Society of Mechanical Cngineers Winter Annual Meeting held November 27 to December 2,1977, in Atlanta, Georgia. The paper is entitled, "A Finite Elesent Ndel for Residual Stresses in Girth-Butt Welded Pipes,"

by E. F. Rybicki, D. W. Schmueser, R. B. Stonesifer, J. J. Groom, and H.

W. Mishler.

Cris!nal SW kr Saul Levine Saul Levine, Director Office of Nuclear Regulatory Research Enclosum:

as stated Distribution SUBJ CIRC CHRON CZS rf CZS cy Branch rf CEJ cy LST cy TEM cy CJ cy PGN cy SL cy EXL cy y

. JM cy 1

l E pmdiga RS

'R h#

CZ mt RSR:W

, RSR:W RSR:DIR RES:TA:D..

RES:SA:D

_ E :D R

CEJohnson LSTong TEMurley

..C.

er.

.._PGNorr

...SL n e _.

11/ /77.

11/ /77 11/ /77

..ll/f/77

.11/ 7.177

.11/ /77.

NBC PJ EM 388 (9 76) NRCN 0,240 W un s. oovsamassart raini,no carecsa sere -h 1568 134