ML20212F577
| ML20212F577 | |
| Person / Time | |
|---|---|
| Site: | Prairie Island  |
| Issue date: | 12/22/1986 |
| From: | NRC |
| To: | |
| Shared Package | |
| ML20212F113 | List: |
| References | |
| NUDOCS 8701120066 | |
| Download: ML20212F577 (6) | |
Text
- - =
Aooendix TT cvaluation of Vestinghouse Report WCAD 10031, Devision 1 "Touahness Criteria For Thermallv Aned Cast Stainless Steel" introduction Westinahnuse Deoort WCAD 10431, Devision 1, "Touchness Criteria For Thermall.y Aced Cast Stainless Steel," provides criteria for evaluatino the fracture resistance of thermally aged cast stainless steel oicino for Westinghouse Nuclear Steam Sunol.v Svstems.
The criteria in the report are divided into three categories. Rased on the predicted end-o' li#e KCll innact energv value #or the heat of material, the material is considered to be either cateanrv 1, ' or 3.
The catecnev 1 #racture toughness 'oronerties and the equations #or predicting end-of li#e KCll imnact enerav were previous'y documerted in a Vestinghouse Deport (Def. 11 The staf#'s review of this report is contained in Apoendix I.
As discussed in WCAD 10931 and reference 1, cast stainless steel is a two phase allo.y consisting o# austenite and #errite.
it has been #cund that the chrome enriched ferrite of the two phase allov becomes hardened and om5rittled when thermally aced at primarv loop water terneratures (550-600"ri.
Discussion WCAD 10931, Devision 1, describes the fracture preparties of materials which have fracture resistance belov that of the reference material discussed in Re f. 1.
These materials are either catecorv ' or 3 in addition, VCAD 10031, Devision 1, provides estimates and the bases for the uncertainty in the end-of-life VCil innact energv nrediction equations and the uncertaintv in the fracture properties.
fhlihu 5
2 P
-?-
The Catecory 2 and 3 fracture-toughness properties are based on the extrapola-tion of existing aged reference naterial for higher ferrite content and a fracture energy model described in Ref. (?). The aged reference material is the heat of material which was the basis for the bounding fracture properties (J-integral) reported in Ref. (11 However, Westinghouse has performed additional J-integral and Charpy impact tests on the aged reference material. The test temperatures for the J-integral tests were -320*F, -200 F, room temperature, 150"F, 700 F, 550"F and 600*F. The test temperatures for the Charpy impact tests were room temperature 200*F and 550 F.
At -320 F test temperature, the ferrite phase in cast stainless steel is completely embrittled and the J-integral test results would represent the toughness of the austenitic phase in the reference material. Hence, the J-integral data for the reference material at -390*F would conservatively represent the fully aged fracture toughness for this material. Although Charpy impact tests were rot performed at -3?0 F, the predicted fully aced value was estimated from the
.1 integral tests using correlations in references 3 and 4.
The J-integral test at 600*F was used to establish the fracture toughness for category 1 materials.
The J-integral test at 200*F indicate that the tearing modules (Tmi at 200*F could be less than the value at 600*F. The material's Tm describes its resistance to ductile fracture after a postulated crack has initiated growth.
In order for crack growth to initiate, the loads on a pipe must exceed the J
value for the material. The reduction in loads during transients in PWR's 1c I
occurring at 200 F as compared to transients occurring at 600*F has been evaluated. This evaluation indicates that piping loads at 200*F are below those required to initiate crack growth. Hence, it is likely that the loads at 200*F will not be large ennugh to exceed the naterial's J and the reduction 1c in Tm from 600 F to 200*F is not considered significant.
To extrapolate the fully aged fracture properties from the reference material to other materials, the authors utilized a fracture energy model designated as the tortuous beam model. This model is based on the fracture energy theory described in Ref. (2) which indicates that the total fracture energy is proportional to the volume of the plastica 11y deformed material.
Test data from Charpy V-notch impact and dynamic tear tests on ASTM A 533, Grade B, Class 1 material are reported in Ref. (2) which support this theory. The tortuous beam model considers the fracture path for the cast stainless steel as a series of cracks through ductile austenite and embrittled ferrite. The model proposes that the ratio of the total fracture energy of the reference material to that of another heat of material is dependent upon the volumetric fraction of austenite and ferrite in the two materials.
Using the proposed model the authors define the limiting fracture properties (J-integral and Charpy impact) for fully aged cast stainless steel. These properties have been verified by evaluating the fracture toughness data from a heat of aged cast stainless steel supplied by a Westinghouse Licensee. The Charpy impact data on this experi-mental heat indicates that the cast stainless steel has been fully aged. The J-integral test data from this experimental heat are higher than the values predicted by the Westinghouse model. Hence, the model appears to provide conservative estimates of the J integral properties for fully aged material.
I Westinghouse has used equations proposed in Ref. (1) to predict the end-of-life KCU impact erergy values for any heat of cast stainless steel. There are two equations for predicting the end-of-life KCU impact energy value of aged cast stainless steel. One equation is for CF8M material and the other is for CF8 and CF8A materials. The authors of Westinghouse Report WCAP 10931 have refined the the ecuations to determine the 95% lower confidence level for the aged material.
The 95% confidence level for the calculated KCU value were determined by a procedure given in Reference 5.
These confidence limits have been incorporated into the predicted end-of-life KCU impact energy equations to ensure that the
- predicted value for the aged cast stainless steel at the end-of-life of the nuclear plant is greater than that of the reference material.
The uncertainty in fracture properties for aged cast stainless steel material was determined by evaluating the uncertainty in the ferrite centent. The ro-efficient of variation of the ferrite was deternined using procedures ir. r fer-e ence (6) and data in Reference (7). The coefficient of variation of the ferrite was used to establish the 95% lower confidence limit for category ? ard 3 materials, which are defined below.
The category 1 material is defined as material in which the calculated kcl!
impact energy exceeds the uocer 95% confidence KCU value for the aged reference material. The fracture properties for category 1 materials are those o' the aged reference material reported in Re'erence 1.
The category 2 materials is defined as material in which the calculated KCU impact energy is less than the upper 95% confidence KCU value for the aced reference material and greater than the fully aged predicted mean KCU value for the reference material. The fracture properties for the categorv P material are predicted by a linear interpolation between the fracture properties of the aged reference material and the fully aced 95% lower confidence limit reference material.
The category 3 material is defined as material in which the calculated KCU impact energy is less than the fully aged predicted mean value for the reference material. The fracture properties calculated for the category 3 materials are those of the fully aged reference materiel at the 95% lower confidence limit.
Conclusions The equations and methodology that are documented in VCAD 10931 Devision 1 may be utilized for establishing the fracture criteria for thermally aced cast stainless piping apolicable for the leak-before-break analyses.
The end-of-life KCU impact energy for CF8M cast stainless may be calculated using
..eouation I?-17) in VCAD 10031 Revision 1 The end-of-life KCU impact energy for CF8 and CFRA material mav he calculated usina eouation 5 2 in WCAD 10456 (Ref. II.
Rased on the calculated KCU impact energy the fracture touchness #or a heat of therrally aged cast stainless steel may be calculated usino the procedure in Table 9-8 of WCAD 10031 Devision 1, except that for category 1 material, th6 maximum.1 applied should be limited to 3000 in lbs/in.2 The maximum.1 applied limit was established hv the staff in its evaluation of Reference 1, which is contained in Aonendix T.
. Re#erences (11 WCAP - 10456, " Effects of Thermal Aging on the Structural Integrity of Cast Stainless Steel Piping for Westinghouse Nuclear Stean Supply Systems," Westinghouse Proprietary Class ?, November 1983.
(71 F.J. Witt and R.G. Berggren, " Size Effects and Energy Disposition in Impact - speciman Testing of ASTM A 533 Grade 8 Steel, Experimental Mechanics, Volume 11, No. 5, pp 193-701 (May 1971).
(3)
F.J. Witt, " Relationships Between Charpy Impact Energies and IJpper Shelf K Values for Reactor Pressure Vessel Steels," international IC
.10urnal of Pressure Vessels and Pipino, Volume 11, op. 47-63 (1983).
(41 S.T. Rolfe and J.M. Rarsom, Fracture and Fatique Control in Structures-Applications to Fracture Mechanics, Prentice-Hull, Chapter 6 (19771 (5)
N.R. Draper and H. Smith, Applied Regression Analysis, John Wiley and Sons, Inc., New York, 1967, pp. 21-74.
(6)
E.B. Haugen, Probabilistic Approaches to Design,.lohn Wiley and Sons, New York, 1968, pp 105-128.
(7)
L.S. Aubey, et al. " Ferrite Measurement and Control in Cast Duplex Stainless Steel," Stainless Steel Castings, ASTM STP 756, American Society for Testino and Materials, 1982, pp 126-164