LR-N11-0057, Official Exhibit - ENT000197-00-BD01 - Salem Nuclear Generating Station, Units 1 & 2 - Close-out of the NRC Audit Associated with Use of Westems Related to the Plant License Renewal Application
| ML12338A524 | |
| Person / Time | |
|---|---|
| Site: | Indian Point  |
| Issue date: | 02/24/2011 |
| From: | Davison P Public Service Enterprise Group |
| To: | Atomic Safety and Licensing Board Panel, Document Control Desk, Office of Nuclear Reactor Regulation |
| SECY RAS | |
| References | |
| RAS 22112, 50-247-LR, 50-286-LR, ASLBP 07-858-03-LR-BD01, LR-N11-0057 | |
| Download: ML12338A524 (20) | |
Text
United States Nuclear Regulatory Commission Official Hearing Exhibit Entergy Nuclear Operations, Inc.
In the Matter of:
(Indian Point Nuclear Generating Units 2 and 3)
" ....~~R REGU<.q" ASLBP #: 07-858-03-LR-BD01
!~~.~
Docket #: 05000247 l 05000286
" 03:
Exhibit #: ENT000197-00-BD01 Identified: 10/15/2012 I-
<1> . ~
Admitted: 10/15/2012 Withdrawn:
~
'I"'~
i Rejected:
0.... Stricken:
" ***** ... Other:
PSEG ENT000197 P.O. Box 236, Hancocks Bridge, NJ 08038-0236 Submitted: March 29, 2012 OPSEG Nuclear LLC 10 CFR 50 10 CFR 51 10 CFR 54 LR-N11-0057 u.s. Nuclear Regulatory Commission ATIN: Document Control Desk Washington, DC 20555-0001 Salem Nuclear Generating Station, Unit No.1 and Unit No.2 Facility Operating License Nos. DPR-70 and DPR-75 NRC Docket Nos. 50-272 and 50-311
Subject:
Close-out of the NRC Audit associated with use of WESTEMSTM related to the Salem Nuclear Generating Station, Units 1 and 2 License Renewal Application
Reference:
Letter from Mr. Robert C. Braun (PSEG Nuclear, LLC) to the NRC, "Follow-Up Responses to Questions Raised during January 18-19, 2011 NRC Audit of WESTEMSTM Program Benchmarking Activities, Related to the Salem Nuclear Generating Station, Units 1 and 2 License Renewal Application," dated January 31,2011 On January 18 and 19, 2011, the NRC Staff audited various activities related to the use of WESTEMSTM associated with the Salem Nuclear Generating Station, Units 1 and 2 License Renewal Application. In the referenced letter, PSEG Nuclear provided the Staff with information responding to five specific questions raised during the Audit.
Audit activities were resumed on February 8, 2011, at which time PSEG Nuclear, Westinghouse, and NRC staff reviewed the results of calculation updates that documented the basis for stress analyst activities related to the license renewal environmentally-assisted fatigue (EAF) calculations. After this review, the NRC determined that follow-up actions would be needed from PSEG Nuclear in order to satisfactorily complete the Audit.
Enclosure A of this letter is an update to the information provided in Enclosure A of the January 31, 2011 Reference letter, and replaces that response package in its entirety, with changes indicated. Similarly, Enclosure B of this letter updates and replaces the License Renewal Commitment List changes that had been provided as Enclosure B in the January 31,2011 Reference letter. Although unchanged, Enclosure C from the January 31,2011 referenced submittal is again provided as Enclosure C to this letter, for completeness. Together, these Enclosures provide the information needed to close out NRC Staff questions associated with the WESTEMSTM Audit.
Document Control Desk LR-N11-0057 Page 2 of 2 There are no other new or revised regulatory commitments associated with this submittal.
If you have any questions, please contact Mr. Ali Fakhar, PSEG Manager - License Renewal, at 856-339-1646.
I declare under penalty of perjury that the foregoing is true and correct.
Executed on Sincerely,
~J~.\)~
Paul J. Davison Vice President, Operations Support PSEG Nuclear LLC
Enclosures:
A. Updated Responses to NRC Questions associated with Use of WESTEMSTM associated with the Salem Nuclear Generating Station, Units 1 and 2 License Renewal Application B Update to License Renewal Commitment List C. "Method for Selecting Stress States for Use in an NB-3200 Fatigue Analysis,"
Proceedings of the ASME 2010 Pressure Vessels & Piping Division /
K-PVP Conference, July 18-22, 2010, Bellevue, Washington, USA cc: William M. Dean, Regional Administrator - USNRC Region I B. Brady, Project Manager, License Renewal- USNRC R. Ennis, Project Manager - USNRC NRC Senior Resident Inspector - Salem P. Mulligan, Manager IV, NJBNE L. Marabella, Corporate Commitment Tracking Coordinator Howard Berrick, Salem Commitment Tracking Coordinator
Enclosure A LR-N11-0057 Page 1 of 9 Enclosure A Updated Responses to NRC Questions associated with Use of WESTEMSTM associated with the Salem Nuclear Generating Station, Units 1 and 2 License Renewal Application
Enclosure A LR-N11-0057 Page 2 of 9 In January and February, 2011, the NRC conducted an audit of PSEG Nuclear's use of WESTEMSTM associated with the Salem Nuclear Generating Station, Units 1 and 2 License Renewal Application (LRA). During the audit, the NRC requested additional information. PSEG Nuclear provided responses to these requests in its letter LR-N 11-0042 dated January 31, 2011. As a result of follow-up audit discussions held on February 8, 2011, these responses are being updated (and new items added) to clarify or supplement the earlier responses. Changes to the earlier responses and affected LRA Sections are shown with bold, italics font for additions and strikethrough text for deletions. As in the January 31,2011 letter, the specific request for additional information is stated below, followed by the PSEG Nuclear response.
Audit Question No.1:
In order to close-out the Salem WESTEMS audit, for the WESTEMS "Design CUF" module analysis of the BIT and surge nozzles, provide written explanation and justification of any user intervention in the process including the user intervention applied to the peak and valley selection process.
PSEG Response:
Westinghouse has revised their environmentally-assisted fatigue (EAF) calculations that supported the Salem License Renewal Application (LRA) for the Unit 2 Pressurizer Surge Nozzle Safe End to Pipe Weld and the Unit 2 Safety Injection Boron Injection Tank (BIT) Nozzle Coupling to Cold Leg Weld. The revision specifically added a new section to an existing Appendix to document the following:
- 1. Description of the WESTEMSTM stress peak and valley selection algorithm.
- 2. A new WESTEMSTM program run with no analyst intervention (i.e. no manual removal by the analyst).
- 3. Numerical comparison of the number of stress peaks and valleys selected by the analyst in the original revision of the calculation to the number of stress peaks and valleys selected during the new WESTEMSTM program run.
- 4. Justification for analyst manual removal of the stress peaks and valleys on a transient-by-transient basis only when WESTEMSTM selected more stress peaks and valleys than the analyst. In the cases where the analyst selected more stress peaks and valleys than WESTEMSTM, justification is not required since this method is more conservative for the fatigue evaluation.
The justification is illustrated in the new section of the Appendix by use of plots generated by a spreadsheet containing downloaded data of the Total Stress Intensity and Primary plus Secondary Stress Intensity values for each transient. The plots depict the stress peaks selected by WESTEMSTM and by the analyst. Documentation is provided in the new section of the Appendix justifying removal of redundant stress peaks and valleys for each transient.
Enclosure A LR-N11-0057 Page 3 of 9
- 5. For the Unit 2 Safety Injection BIT Nozzle Coupling to Cold Leg Weld location, two new tables are added to list the fatigue pairs and corresponding fatigue usage for the original revision of the calculation (analyst intervention) and where no analyst intervention was involved for comparison of the total cumulative usage factor (CUF). For the Unit 2 Pressurizer Surge Nozzle Safe End to Pipe Weld location, the fatigue usage values were compared and were the same value in both the original revision of the calculation, and the revised calculation where no analyst intervention was involved.
Although the 60-Year Design CUF value for the Unit 2 Safety Injection BIT Nozzle Coupling to Cold Leg Weld location was higher in the case of no analyst intervention during the stress peak and valley process, justification is provided for removal of redundant stress peaks and valleys. The 60-Year Design CUF listed in LRA Table 4.3.7-2, "Salem Unit 2 60-Year Environmentally-Assisted Fatigue Results," reflects justified analyst intervention during the stress peak and valley process.
The revised proprietary calculations have been approved by Salem, and will-ge were made available for NRC review during the Re*t February 8, 2011 phase of the WESTEMSTM audit.
During the February 8, 2011 review of the BIT Nozzle Coupling to Cold Leg Weld location calculation update, the basis for analyst removal of two of the peak and valley times from the data was unclear and not sufficiently documented in the calculation. The following discussion provides the detailed basis for the analyst removal of two of the peak and valley times from the data.
Specifically, for Transient 2 (Primary Side Leak Test), WESTEMSTM selected five (5) peaks. The analyst selected two (2) peaks for use in the fatigue evaluation.
The first of three (3) peaks was removed because it represented the same Total Stress as a peak approximately seventy-two minutes prior to this peak, and since the Primary plus Secondary stress in this evaluation does not result in any Ke (simplified elastic plastic penalty factor applied to alternating stress when the Primary plus Secondary Stress Intensity Range limit is exceeded), values greater than 1.0, it is redundant with the previous peak, and not required. The last (2) peaks in the transient are redundant peaks of the initial state captured by the first p~ak time, since the transient returns to the same stress state as it started. This stress state is redundant because it is the same stress state as the initial stress state of another Primary Side Leak Test transient cycle, or any other transient that begins at a similar plant no-load condition (e.g., Unit Load, etc.). For Transient 11 (High Head Safety Injection Boron Injection [Inadvertent Safety Injection]
transient), WESTEMSTM selected nine (9) peaks. The analyst selected six (6) peaks for use in the fatigue evaluation. There were five (5) peaks selected by WESTEMSTM early in the transient. The analyst removed two of the five peaks because they were selected based on peaks in the Primary plus Secondary Stress.
Since the Primary plus Secondary Stress Intensity Range is less than the allowable of 3Sm, a Ke penalty factor on alternating stress is not required, and therefore, two peaks were considered redundant and able to be removed. Later in this transient, two peaks were selected by WESTEMSTM within several seconds of each other. WESTEMSTM selected the first of these two peaks because of a stress
Enclosure A LR-N11-0057 Page 4 of 9 intensity inflection point created by a stress component sign reversal and an accompanying change in the controlling principal stress. A revision to the fatigue evaluation re-established the basis for removal of one of these two peaks by reviewing a plot of total stress component history. Only one true peak had occurred and was selected by the analyst for use in the fatigue evaluation. The stress intensity inflection point of the other peak does not represent the final extreme stress range caused by the transient excursion, and was considered unnecessary for inclusion in the fatigue evaluation. Towards the end of the transient, there were two (2) peaks within 65 seconds of each other. Since the stress states were the same, the analyst selected one of these two peaks for use in the fatigue evaluation. The analyst added one (1) peak that was not selected by WESTEMSTM at the initial time of the transient for additional conservatism in the fatigue evaluation.
The associated calculation has now been updated to properly capture the basis for this user intervention activity.
Audit Question No.2:
For any WESTEMS "Design CUF" module analyses performed for the remaining monitored locations at Salem (Le., other than the BIT and surge nozzles), provide written explanation and justification of any user intervention applied in the process including the user intervention applied to the peak and valley selection process prior to two years before entering the period of extended operation.
PSEG Response:
Salem will revise the fatigue GalGulations for all IOGations monitored at Salem Units 1 and 2 to inGlude a written explanation and justifiGation of any user intervention applied for any
'I'lESTEMSTM "Design CUF' module analyses, inGluding the user intervention applied to the stress peak and valley seleGtion, at least hyo years prior to entering the period of extended operation.
PSEG Nuclear has reviewed the justification for the stress peak and valley editing provided in revisions to the Unit 2 Pressurizer Surge Nozzle Safe End to Pipe Weld and the Unit 2 Safety Injection Boron Injection Tank (BIT) Nozzle Coupling to Cold Leg Weld EAF calculations. Discussions with Westinghouse concluded that stress peak and valley editing during the fatigue calculation process for the remaining locations monitored by WESTEMSTM at Salem Units 1 and 2 is consistent with the two locations that were the subject of the WESTEMSTM benchmarking audit, and that the two calculations that were revised are considered to be the most limiting with respect to cumulative usage. Therefore, PSEG Nuclear has deemed it unnecessary to revise the existing EAF calculations performed for the remaining WESTEMSTM monitored locations to include a written explanation and justification of any user intervention applied for any WESTEMSTM "Design CUF" (NB-3200 module) analyses, including the user intervention applied to the stress peak and valley selection.
Enclosure A LR-N11-0057 Page 5 of 9 Therefore, PSEG Nuclear is retracting what had been proposed as commitment
- 53 as described in PSEG letter LR-N11-0042, "Follow-Up Responses to Questions Raised during January 18-19, 2011 NRC Audit of WESTEMSTM Program Benchmarking Activities, Related to the Salem Nuclear Generating Station, Units 1 and 2 License Renewal Application", dated January 31,2011. The License Renewal Commitment List is revised accordingly, as shown in Enclosure B.
As a result of this response, sommitment #63 is added to LRA Table A.6, Lisense Renewal Commitment List, as shown in Enslosure B of this letter.
Audit Question No.3:
For any use of the WESTEMS "Design CUF" module in the future at Salem, include written explanation and justification of any user intervention in the process.
PSEG Response:
Salem will include written explanation and justification of any user intervention in future evaluations using the WESTEMS ' "Design CUF" (NB-3200 module).
As a result of this response, commitment #64 53 is added to LRA Table A.5, License Renewal Commitment List, as shown in Enclosure B of this letter. The seventh paragraph of LRA Section A.4.3.7 is also revised, as shown following the response to Audit Question No.4, below.
Audit Question No.4:
Provide a commitment that the NB-3600 option of the WESTEMS "Design CUF" module will not be implemented or used in the future at Salem.
PSEG Response:
Salem will commit to not use or implement the NB-3600 option (module) of the WESTEMSTM program in future online fatigue monitoring and design CUF calculations.
As a result of this response, commitment #ee 54 is added to LRA Table A.5, License Renewal Commitment List, as shown in Enclosure B of this letter. The seventh paragraph of LRA Section A.4.3.7 is also revised, as shown below.
A.4.3.7 Environmentally-Assisted Fatigue Analyses The evaluations showed that no cumulative usage factors with environmental penalties exceeded 1.0 for 60 years of service for the identified plant-specific locations. Future fatigue evaluations using WESTEMSTM "Design CUF" (NB-3200 module) will include written explanation and justification of any user intervention. Future fatigue design calculations will not use or implement the NB-3600 option (module) of the WESTEMSTM program. The Metal Fatigue of Reactor Coolant Pressure Boundary aging management program (B.3.1.1) will be
Enclosure A LR-N11-0057 Page 6 of 9 used to manage the aging effects of environmentally assisted fatigue for the components in Salem LRA Tables 4.3.7-1 and 4.3.7-2.
Audit Question NO.5:
Provide a description of the peak and valley selection process used by WESTEMS and how that process aligns with ASME Code NB-3216 methodology.
PSEG Response:
WESTEMSTM isa software program developed by Westinghouse Electric Company LLC (Westinghouse). It is used to perform fatigue evaluations for components using NB-3200 stress models, referred to as Analysis Section Number (ASN) models by Westinghouse, according to the ASME Boiler and Pressure Vessel Code,Section III, Subsection NB-3222.4, 1986 edition. As part of the ASME Code fatigue evaluation, the person performing the fatigue evaluation (analyst) is required to select the extremes of the stress cycles (stress peaks and valleys) imposed by the component's transient loads (thermal, mechanical, etc.). WESTEMSTM uses an automated approach to assist the analyst in selecting the stress peak and valley times in each transient. The approach is described in general, with respect to the associated ASME Code fatigue evaluation requirements, in Westinghouse's publication, "Method for Selecting Stress States for Use in an NB-3200 Fatigue Analysis," PVP201 0-25891, Proceedings of the ASME 2010 Pressure Vessels & Piping Division / K-PVP Conference, July 18-22, 2010, Bellevue, Washington, USA. This paper is attached to this letter as Enclosure C.
The WESTEMSTM stress peak and valley algorithm and associated analyst options for its use are described in the WESTEMSTM User's Manual, Volume 2. The alignment between the WESTEMSTM automated approach and ASME Code NB-3216 methodology (NB-3216) is illustrated below with the aid of excerpts from the previously referenced Westinghouse publication and sections of the WESTEMSTM User's Manual.
The following paragraphs are based on, or taken directly from, Westinghouse publication PVP201 0-25891 :
Performinga fatigue evaluation per ASME Code Section III, Subsection NB-3222.4 (NB-3222.4) requires calculating the stress differences for each type of stress cycle in accordance with NB-3216. In determining the number of cycles for each type of stress cycle, consideration is given to the superposition of cycles from various stress cycles that produce a total stress range greater than that of each individual stress cycle alone.
This procedure is outlined in NB-3222.4(e} (5), Step 1. The resulting cycles and alternating stress intensities from this procedure are then applied in a cumulative manner using the appropriate design fatigue curve, in NB-3222.4(e} (5), Steps 3-6, to calculate fatigue usage factors. In traditional ASME Section III Code fatigue evaluations, the extremes of the stress cycle have been selected by the analyst based on experience and review of the stress component and/or stress intensity histories produced by the various transients.
Enclosure A LR-N11-0057 Page 7 of 9 The stress state selection method incorporated in WESTEMSTM fatigue evaluations employs a stress intensity based approach that is a practical method used to interpret and apply NB-3216.2. It can capture Primary plus Secondary stress and Total stress ranges for complex transients, allowing for the proper application of NB-3222.4. The approach emulates considerations employed by analysts for decades in applying various calculation methods to NB-3200 requirements.
The method used by WESTEMSTM to select the stress peak and valley times utilizes a straightforward mathematical process to select times where the stress states are at a relative minimum or maximum. Additionally, the method employs controlled options that provide the ability to control the treatment of initial condition stress states in the selection process.
The basic algorithm used by WESTEMSTM is as follows. For each transient cycle in the component fatigue evaluation, the six stress components of Primary plus Secondary stress and of Total stress are calculated for the entire transient time history. Then the stress intensities for the Primary plus Secondary stress and the Total stress time histories are calculated. Relative maxima and minima within the Primary plus Secondary stress and Total stress time histories for each transient are identified using the second derivative test (comparing the slopes of the stress history around a time point).
It is important to note that in following an NB-3216.2 procedure, the analyst is to pick a time point where stress conditions are known to be extreme and then find the maximum stress component range relative to this extreme. Using the stress intensity based approach, the time points where stress conditions are extreme are picked at the relative stress peaks and valleys, or maximum and minimum stress states along the stress intensity time history. Effectively, NB-3216.2 calculates stress component ranges from chosen extreme stress component states, where the stress intensity based approach picks extreme stress states based on stress intensity, which is a good indicator of stress component and related principal stress difference extremities. The stress intensity based approach identifies the time points of these extremes, then calculates stress component ranges, the principal stress ranges, and finally the resulting stress intensity range between two selected stress states using the corresponding component stresses at those time points (not the values of stress intensity used to select those points in time as extremes). This is consistent with the procedure used in NB-3216.2. In summary, the stress intensity time histories for each transient are used to select relative extremes, and the component stresses at those extremes are used in calculating stress ranges with other stress states that were selected in the same manner. This procedure is performed for both the Primary plus Secondary and Total stress time histories for all transients considered in the evaluation. Specific examples illustrating the approach are provided in Westinghouse publication PVP2010-25891.
The following discussion is based on and provides specific references to sections of the WESTEMSTM User's Manual. The proprietary WESTEMSTM User's Manual wi-U-ge was made available for NRC review during the Ae*t-February 8, 2011 phase of the WESTEMSTM audit.
Enclosure A LR-N11-0057 Page 8 of 9 The WESTEMSTM stress peak and valley selection algorithm that follows the approach described in the Westinghouse publication PVP2010-25891 is described in section 14.0 of the WESTEMSTM User's Manual, Volume 2. The process used by WESTEMSTM is designed to find all of the inflection points (also known as maxima and minima) in the controlling stress intensity histories. It considers the ASME Section III Code Total stress and Primary plus Secondary stress time histories, since both may influence the fatigue usage calculation. It should be noted that, while the WESTEMSTM algorithm uses the stress intensity history to find the times of the stress peaks and valleys, the individual components of stress available for the stress model are retained at the time points selected, so that the stress range pairs in the fatigue evaluation may be calculated based on the ASME Code Section III methodology. In the WESTEMSTM User's Manual, Volume 2, the general term "peaks" is used to refer to the set of inflection points that include the stress peak and valley stress states in a transient.
The WESTEMSTM algorithm is generally designed to ensure that no valid stress peaks are missed,and so it may, in many cases, conservatively select the number of stress peaks. Therefore, as discussed in section 8.1.3 of the WESTEMSTM User's Manual, Volume 2, the software program permits the analyst to provide inputs, such as stress filter and time constant merge parameters, to attempt to eliminate redundant or unnecessary stress peaks. In the WESTEMSTM design analysis mode, the results are dependent on the analyst time-history input. Therefore, the analyst has ultimate control over the stress peaks used in the fatigue evaluation, to the extent of final editing of the stress peaks selected by the program using the WESTEMSTM stress peak editing tool, which is described in section 8.11 of the User's Manual. The fatigue evaluation is independently verified by another analyst and approved by a manager.
Since the WESTEMSTM algorithm selects stress peaks and valleys consistent with the criteria in ASME Code Section III, Subsections NB-3216 and NB-3222.4, as described above, and the user controls are used to reduce WESTEMSTM program conservatism and donot change the overall basis for the stress peak selection, the final WESTEMSTM fatigue evaluations are performed consistent with the criteria in ASME Code Section III, Subsections NB-3216 and NB-3222.4.
Enclosure A LR-N11-00S7 Page 9 of 9 Audit Question No.6:
The response to "Bullet # 5" on Page 14 of Enclosure A of December 21,2010 PSEG Letter LR-N10-0445 stated that the stress models used in the governing EAF analyses are the same as the stress models employed in the WESTEMSTM online monitoring tool. Based on the discussions during the February 8, 2011 audit activities, it is understood that for the Salem Pressurizer Surge Nozzle Safe End to Pipe Weld location, a different version of the WESTEMSTM stress model was used for the fatigue analysis than what will be used for online fatigue monitoring.
Please explain.
PSEG Response:
As a result of performing the benchmark calculation for the Unit 2 Pressurizer Surge Nozzle Safe End to Pipe Weld, it was discovered that the WESTEMSTM online model was different than that used for calculating the 60 Year cumulative usage fatigue (CUF) value for this location, which is listed in LRA Table 4.3.7-2, "Salem Unit 260- Year Environmentally-Assisted Fatigue Results". Therefore, we are revising our RAI response to Bullet # 5 on page 14 in Enclosure A of PSEG Letter LR-N10-0445, dated December 21,2010, as follows, starting with the last sentence on page 14:
The stress models used in these EAF analyses are the same as the stress models employed in the WESTEMSTM online monitoring tool, with the exception of the Salem Pressurizer Surge Nozzle Safe End to Pipe Weld location and the Surge Line Hot Leg Nozzle to Pipe Weld location. The stress models used in the EAF analyses for these two locations are specific to each Salem Unit due to slight physical differences that were explained under Bullet # 1 [in PSEG Letter LR-N10-0445}. However, for online fatigue monitoring, Salem used a stress model common to both Units that was determined to be conservative and bounding for each of these two locations.
It was verified that the original statement made inPSEG Letter LR-N10-0445 relative to the stress models being the same is correct for all other monitored locations at Salem.
Enclosure B LR-N11-0057 Page 1 of 1 Enclosure 8 Update to License Renewal Commitment List As a result of this response, the commitments discussed above are added to LRA Table A.5, License Renewal Commitment List as commitment numbers 53 and 54 as shown below. Note that the item that had been designated as commitment 53 in PSEG letter LR-N11-0042 has been deleted and the other two commitments have been renumbered. Deletions from text in letter LR-N11-0042 are shown with strikethrol:Jgh format and additions are shown in balded italics. Any other actions described in this letter are not regulatory commitments and are described for the NRC staff's information:
A.5 License Renewal Commitment List UFSAR Supplement Enhancement or No. Program or Topic Commitment Location Implementation Source (LRA App. A) Schedule a.3 Salam ~atill:ls Galsl:llatiaAS Salam IIIill Fauisa lRa fali91:1a salsl:llaliaAs far all WA Allaast a yaafs filFiaf la Salam bettef far weSTeMSTM ~4aRilaras lasatiaRs maRitersS at Salam bJAits ~ aRS 2 le lRa filaries af axtaRsas b~ fll~ ~ QQ42 basatiaAs filreuisa FittaA a*fillaRatieA aRs jl:lstifisalieR ef efilaFalieA.
aAY I:ISaF iRlaFYaAlieA afilfillias faF aRY
\l\IeSTeMSIM "QaSilR G'J~" mesl:lla aRalysas iRsll:lsiRI tRa I:Isar iAteR'aAtieR afilfillias te tRa
,,.I ,~
9453 Salem Fatigue Calculations Salem will include written explanation and WAAA.3.7 Within 60 days of Salem Letter using WESTEMSTM program justification of any user intervention in future issuance of the renewed LR-N11-0042 evaluations using the WESTEMS "Design CUF" operating license.
(NB-3200 module). Salem Letter LR-N11-0057 5ft54 Salem Fatigue Calculations Salem will not use or implement the NB-3600 WAA.4.3.7 Within 60 days of Salem Letter using WESTEMSTM program option (module) of the WESTEMSTM program in issuance of the renewed LR-N11-0042 future online fatigue monitoring and design operating license.
calculations. Salem Letter LR-N11-0057
Enclosure C LR-N11-0057 Page 1 of 8 Enclosure C "Method for Selecting Stress States for Use in an NB-3200 Fatigue Analysis,"
Proceedings of the ASME 2010 Pressure Vessels & Piping Division I K-PVP Conference, July 18-22,2010, Bellevue, Washington, USA (7 pages)
Proceedings of the ASME 2010 Pressure Vessels & Piping Division / K-PVP Conference PVP 2010 July 18-22, 2010, Bellevue, Washington, USA PVP2010-25891 METHOD FOR SELECTING STRESS STATES FOR USE IN AN NB-3200 FATIGUE ANALYSIS Thomas L. Meikle V E. Lyles Cranford III Westinghouse Electric Company LLC Westinghouse Electric Company LLC Monroeville, PA, USA Monroeville, PA, USA MarkA. Gray Westinghouse Electric Company LLC Monroeville, PA, USA ABSTRACT Now, as analytical capabilities have improved, we are able In ASME Code Section III NB-3222.4 fatigue evaluations, to more precisely model postulated design transient conditions.
selecting stress states to detennine the stress cycles according In the past, when analytical capabilities were more expensive, to Section NB-3216.2, Varying Principal Stress Direction, can simplifications were employed to reduce costs and shorten become a challenging and complex task if the transient stress design times. These simplifications would use methods like conditions are the result of multiple independent time varying lumping external loads into minimum and maximum states and stressors. This paper will describe an automated method that creating groups of enveloped transient conditions to achieve the identifies the relative minimum and maximum stress states in a desired effects. In general, this approach is acceptable for component's transient stress time history and fulfills the criteria situations with a small number of loading conditions and little of NB-3216.2 and NB-3222.4. Utilization of the method or no differences in the response characteristics to the loads.
described ensures that all meaningful stress states are identified However, when the loading conditions are defined in a very in each transient's stress time history. The method is very complex manner, as is the case with some components in a effective in identifying the maximum total stress range that can PWR, this approach is not necessarily easy to apply to satisfy occur between any real or postulated transient stress time the analytical requirements in NB-3216.2 and NB-3222.4.
histories. In addition, the method ensures that the maximum Some reasons why these simplifications can fail are as follows:
primary plus secondary stress range is also identified, even if it is out of phase with the total stress maxima and minima. The a) Complex transient time histories typically have many method includes a process to detennine if a primary plus more significant stress states per global transient cycle secondary stress relative minimum or maximum should be than the simplified transients.
considered in addition to those stress states identified in the b) Complex transients may have significantly varying total stress time history. The method is suitable for use in local thennal conditions within one global cycle of the design analysis applications as well as in on-line stress and transient.
fatigue monitoring c) Complex transients produce primary plus secondary stresses that are at times out of phase with the total INTRODUCTION/BACKGROUND stress.
Traditionally, the task of selecting stress states for consideration in fatigue analysis was considered a relatively Subsection NB-3216.2 explains the method for calculating simple and straightforward process. This was true in large part alternating stress intensity for cases where the directions of the because of: principal stresses can change during the stress cycle at the a) the simplicity of the design transients, and location being considered. This method for computing stress b) simplifying assumptions that were introduced during the differences is described as pennitting the principal stresses to qualification process. change direction while still maintaining their identity as they rotate.
Copyright © 2010 by ASME
Perfonning a fatigue evaluation per NB-3222.4 requires effect, the lesser alternating stress range can be neglected, since calculating the stress differences for each type of stress cycle in both stress ranges originated from the same event.
accordance to NB-3216. In detennining the number of cycles for each type of stress cycle, consideration is given to the superposition of cycles from various stress cycles that produce NOMENCLATURE a total stress range greater than that of each individual stress Transient Cycle: Any stress cycle experienced by a component cycle alone. This procedure is outlined in NB-3222.4(e) (5), due to a plant loading event, regardless of whether it was Step I. The resulting cycles and alternating stress intensities caused by thennal or mechanical load conditions.
from this procedure are then applied in a cumulative manner Sp: Total Stress using the appropriate design fatigue curve, in NB-3222.4(e) Sn: Primary plus Secondary Stress (5), Steps 3-6, to calculate fatigue usage factors. Salt: Alternating Stress Intensity To meet the requirements of subsection NB-3216.2, one Sm: Material Allowable Stress would first compute the total stress component time histories Emod: Elastic Modulus Correction Factor for the complete transient cycle, considering all loading Ke: NB-3228.5 Elastic-plastic Penalty Factor conditions and structural discontinuities. Then, a point in time during the transient cycle must be chosen when stress TRANSIENTS VS. STRESS STATES conditions are known to be extreme (either maximum or Depending on the location of the component analyzed, the minimum). The stress components associated with this time component geometry, the location of the component within the point are then subtracted from the corresponding stress system, and loading conditions, the stress responses from components for every time in the transient cycle. The result is various transient cycles can be simple or complex. The a stress component range history relative to one extreme of the superposition of various loadings, whether thennal or transient cycle. From these component stress ranges, principal mechanical in origin, can make selecting an extreme time point stresses and the resulting stress differences are calculated for during the transient cycle per NB-3216.2 a challenge. It is also each point in time for the transient cycle. The alternating stress possible that there are multiple relative maximum and intensity is half of the largest absolute magnitude of the stress minimum stress ranges that should be considered within the intensity range time history. This procedure must be repeated component fatigue evaluation. Finally, each loading condition for each transient cycle to be considered for the component. must be evaluated per NB-3200 rules to detennine if the NB-To calculate the correct alternating stress intensity to use 3222.2 Primary plus Secondary stress range limit is met, and with the design fatigue curve from a computed total stress the possible Ke penalty that may result. Two relatively simple range, an additional factor that must be considered is the transients are shown in Figure I and Figure 2. These two satisfaction of NB-3222.2 for that specific total stress range. If transients were arbitrarily defined to reinforce the premise of the primary plus secondary stress range corresponding to the the stress intensity based method to relative maxima and total stress range fails to meet the 3Srn criteria set by NB- minima selection, as described in the next section.
3222.2, then the Sa used with the design fatigue curve must be increased by a penalty factor, Ke, as part of the simplified STRESS INTENSITY BASED APPROACH elastic-plastic analysis requirements of NB-3228.5. This Although with today's current computational power it mandates that primary plus secondary stress range history and might be possible to program an algorithm that would be a total stress range history must be computed for each transient literal interpretation of NB-3216.2, there would still be some cycle using the procedure described in NB-3216.2. issues that would have to be resolved in selecting the correct Also, because primary plus secondary stress is influenced stress ranges. Those issues include capturing the effect of the by different properties of the component location, the time possible Ke penalty induced from a failure of NB-3222 .2, and when its magnitude is at an extreme may not be the same time applying NB-3216.2 to a complex transient with multiple stress point as the extreme for the total stress history of the single ranges. Westinghouse engineers have developed a repeatable point where fatigue usage is being calculated for the method to identify all significant stress states in a given stress component. The magnitude of the primary plus secondary time history. This method does not rely on the engineer's stress at this maximum, and the associated total stress at the experience but rather utilizes a straightforward mathematical same time, may produce a more conservative alternating stress process to select times where the stress states are at a relative in the forthcoming fatigue analysis. In this case, at least two minImum or maximum. Additionally, the method employs separate total stress ranges resulting from the same event must controlled options that provide the ability to control the be considered in computing the applicable alternating stress. treatment of initial condition stress states in the selection These include at least one based on the primary plus secondary process.
extreme time points and one based on the total stress extreme The stress state selection method has been incorporated time points. After it is established which combination of total into Westinghouse's internally developed stress and fatigue stress range and primary plus secondary stress range will analysis program called WESTEMSTM and is described below.
produce the highest alternating stress, considering the Ke The method employs a stress intensity based approach that is a 2 Copyright © 20 I 0 by ASME
practical method used to interpret and apply NB-32l6.2. It can Transient 1 capture primary plus secondary stress and total stress ranges for complex transients, allowing for the proper application of NB- 600 3222.4. The approach emulates considerations employed by 500 engineers for decades in applying various calculation methods iL:"
to NB-3200 requirements. ~400 The basic algorithm is as follows. For each transient cycle e.
in the component fatigue evaluation, the six stress components f!
- J 300 of primary plus secondary stress and of total stress are ~ I--
calculated for the entire transient time history. Then the stress '~200
"
{!!.
intensities for the primary plus secondary stress and the total 100 stress time histories are calculated. Relative maxima and I -Temperature I minima within the primary plus secondary stress and total stress time histories for each transient are identified using the 20 40 60 80 100 120 140 160 180 200 second derivative test. Special considerations are used to Time(s) address relative flat spots that are plateaus in the stress intensity Figure 1 Loading Conditions for Transient 1 time histories.
It is important to note here that in following an NB-3216.2 Transient 2 procedure, the analyst is to pick a time point where stress 600 , - - - - - - - - - - ' - - - - - - - - - - - - 0 300 conditions are known to be extreme and then find the maximum stress component range relative to this extreme.
500* "r-_____- - - - - - - - - - - - - - t 250 Using the stress intensity based approach, the points where iL:"
stress conditions are extreme are picked at the relative peaks g> 400 I--l---I----~___- - - - - - - - - - _ _ + 200 "[
and valleys, or maximum and minimum stress states along the e. :;:
~ C stress intensity time history. Effectively, NB-3216.2 calculates ~ 300 150 :::.
stress component ranges from chosen extreme total stress ~ g
~ 'e
{!.
component states, where the stress intensity based approach Q. 200 100 picks extreme stress states based on stress intensity, which is a ! 100 +--f------------,-----~___-__+ 50 good indicator of stress component and related principal stress difference extremities. The stress intensity based approach identifies the times of these extremes, then calculates stress 20 40 60 80 100 120 140 160 180 Time(s) component ranges, the principal stress ranges, and finally the resulting stress intensity range between two selected stress Figure 2 Loading Conditions for Transient 2 states using the corresponding component stresses at those times (not the values of stress intensity used to select those points in time as extremes). This is consistent with the After the total stress component histories for each transient procedure used in NB-3216.2. In summary, the stress intensity are calculated, the process of selecting the largest total stress time histories for each transient are used to select relative range was conducted using both methodologies. A rigorous extremes, and the component stresses at those extremes are analysis using the procedure described in NB-3216.2 was used in calculating stress ranges with other stress states that performed to calculate the maximum local component stress were selected in the same manner. This procedure is performed ranges relative to each time step within the transient and for both the primary plus secondary and total stress time between the two transients. For Transient 1, the maximum histories for all transients considered in the evaluation. stress intensity range occurred at 13 seconds. For Transient 2, Several tests were performed to prove that the stress the maximum stress intensity ranges occurred at 13 seconds intensity approach satisfies NB-3216.2, One example problem and 40 seconds. Refer to Figure 3 and Figure 4 for total stress is provided here that uses both the literal interpretation ofNB- component time histories of Transient 1 and Transient 2, 3216.2 and the stress intensity based approach to show that the respectively.
resulting maximum stress range calculated between two transients would be the same. First, two arbitrary transients are defined that include a simple thermal transient (Transient 1) and a second thermal transient that also includes a varying torsion load (Transient 2). The torsion load was included in the second transient to introduce a varying principal stress direction response. Figure I and Figure 2 illustrate Transient 1 and Transient 2 loading conditions, respectively.
3 Copyright © 2010 by AS ME
Transient 1 Stress Response Trans. 1* 2 Stress Difference Relative to Trans. 1 at 13 sec.
Time SXX SYV SZZ SXY SYZ SZX SI 50 0.09 -79.63 -79.75 0.00 32.61 -0.02 112.39
-10 r::.. 100 150 20 60 0.08 -77.57 -77.69 0.00 30.44 -0.02 108.15
-20 I 70 0.08 -76.48 -76.59 0.00 28.26 -0.02 104.87
! ~30 / -s)()(
-SVY 80 0.08 -75.89 -76.00 0.00 26.09 -0.02 102.11 e
III
-40 f -sZZ
. SXY 90 110 0.08 0.08
-75.52
-75.30
-75.62
-75.40 0.00 0.00 23.91 19.57
-0.02
-0.02 99.56 94.99 U; -50 I -SYZ
-SZX 130 0.08 -75.25 -75.35 0.00 15.22 -0.02 90.60
-60 I 150 0.08 -75.20 -75.30 0.00 10.87 -0.02 86.20
-70 II 190 0.08 -75.21 -75.30 0.00 2.17 -0.02 77.51 200 0.08 -75.21 -75.30 0.00 0.00 -0.02 75.38
-80 Time(s)
The process of locating the local relative maximum stress Figure 3 Total Stress Component Time History - Transient 1 ranges was then conducted using the stress intensity based approach. After the total stress intensity histories were Transient 2 Stress Response calculated, the program algorithm was run to select the local 60 stress intensity peaks and valleys. Relative maxima and 50 minima within the total stress intensity time histories for each 40 ~ transient were identified using the second derivative test.
30 \ Figure 5 and Figure 6 illustrate the stress intensity time
\ histories for Transient 1 and Transient 2. The peak and valley
~ 20 -S)()(
times selected using the stress intensity based approach are
~
= 10
\. -SVY
-SZZ summarized in Table 2.
'-
-~SXY e
iii 0
-SYZ
-SZX
_ _ _ _ _ 20
-10
\ 50 100 150 Transient 1 Stress Intensity Response
-20
-30
-40
\
\ '---""
~ ----- 80 70 l Time(s) i--
/\
Figure 4 Total Stress Component Time History -Transient 2
\
\
When the relative stress component ranges, principal stress I-
\
ranges, and the resulting stress intensity ranges were computed \
between transients, the two relative maximum stress intensity ranges occurred between Transient I at 13 seconds and Transient 2 at 13 seconds, and also between Transient 1 at 13 seconds and Transient 2 at 40 seconds.
10 o
o 20
""---
40 60 80 100 Time(s) 120 140 I- Stress Intensity 160 180 I
200 Figure 5 Total Stress Intensity Time History for Transient 1 Table 1 NB-3216.2 - Relative Stress Intensity Range Results Trans. 1* 2 Stress Difference Relative to Trans. 1 at 13 sec.
Time SXX SYV SZZ SXy SYZ SZX SI 0 0.08 -75.21 -75.30 0.00 0.00 -0.02 75.38 10 0.08 -75.21 -75.30 0.00 0.00 -0.02 75.38 11 0.08 -106.29 -106.39 0.00 0.00 -0.03 106.47 1,,\'OTi'3' 1:1125\35::.: 11i::'£j'25:51fl 110:00 "yx,?';SC¢
- 'n:mA,"".' Ii 0,00:: 125:64:):
15 0,14 -118.74 -118.90 0.00 0.00 -0.03 119.04 25 0.11 -97.62 ~97.75 0.00 13.04 -0.03 110.84 35 0.10 -86.92 -87.05 0.00 26.09 -0.02 113.16 I>*::',',*:**,c(*,r
- ~(IiI"(ji' ICii-.U!091\ I;>
- ~sl~ ~G:;;" I'l,;;'?:::;;;:! "LiO'
- j *..* ,.;~
7)' I'Y?H6t38{!
4 Copyright © 20 I 0 by ASME
Transient 2 Stress Intensity Response evaluation, the cycles for each stress intensity range would be assigned based on the approach described in NB-3222.4.
70
"" ~
60 PRIMARY PLUS SECONDARY STRESS 50
/ CONSIDERATION As described in the introduction, a circumstance that can f\V / ~ arise is the possibility of the primary plus secondary stress time
~ history lagging the total stress time history. This is because primary plus secondary stress is influenced more by properties 20
~ like section thickness and geometric and material 10 ~ discontinuities, and can have a slower stress response that the total stress at a point. The WESTEMS' program I -Stress Intensity I~ implementing the method stores the primary plus secondary 20 40 60 80 100 120 140 160 180 200 stress components and the total stress components for each Tlme(.) stress state time selected, regardless of whether it was selected based on primary plus secondary or total stress. This is Figure 6 Total Stress Intensity Time History for Transient 2 because when two stress states pair in NB-3222.4, the primary plus secondary stress range of that pair must be measured Table 2 Selected Maxima and Minima* Stress Intensity against NB-3222.2 limits, and any resulting Ke penalty must be Method applied to the total stress intensity range of that pair to calculate the altemating stress. The example problem below illustrates Time Transient 51 how WESTEMS ' incorporates primary plus secondary stress 13 1 75.38700 into its algorithm to calculate alternating stress.
15 1 65.48500 If two stress states were selected because the primary plus 195 1 0.00100 secondary was out of phase but was caused by the same event, then one of the stress states can be disregarded. This can only 0 2 0.00500 be done once it is determined which stress state is contributing 13 2 50.25200 to the pair resulting in the higher alternating stress. Then, the 25 2 35.49900 stress state associated with the lesser altemating stress is 40 2 65.22001 disregarded.
50 2 65.22000 If, for a specific component, it is known that the total stress 200 2 0.00600 state is always controlling, the program also has the option to automatically disregard the primary plus secondary peak/valley based on a time constant. If there is a primary plus secondary Once the peaks were selected using the stress intensity peak/valley within a specified time relative to the total stress based approach, all possible pairings of maxima and minima peak/valley, then the primary plus secondary stress peak/valley were calculated to reveal those that would yield the highest will be disregarded. The time constant for a location can be total stress intensity range. This was done by calculating the approximated by evaluating a transient composed of only component stress ranges between the selected stress state times, thermal stress and reviewing the primary plus secondary stress computing the principal stress ranges, and the resulting stress and total stress responses. The period between these two intensity ranges. The two highest* stress intensity ranges are peaks/valleys would be a good initial estimate of the time shown in Table 3. constant.
Example Problem Table 3 Maximum Stress Intensity Range Results* Stress This example problem illustrates a simple example of Intensity Based Approach when a peak/valley time chosen from primary plus secondary Trans B Time B Trans A Time A 51 stress can actually be the controlling stress state over the total 2 13 1 13 125.64 stress peak/valley for an event. Figure 7 and Figure 8 below 2 40 1 13 116.38 illustrate the loading conditions within two arbitrary transients.
A simple comparison between Table I and Table 3 reveals that the stress intensity based approach yields the same results as the approach used in NB-3216.2. In the ensuing fatigue 5 Copyright © 20 I 0 by AS ME
Transient A Sp and Sn Component Stress Histories Transient A 600 3000 500 2500 I Sp: SXX Sp: SYV 2000 :::-
C/I ~ -20j--ft------------=--'/-----_;
.e.. II)
II)
Sp: SZZ Sn: SXX 1500 :; ~ ~o~~--~~~~~--------_; Sn: SYV C/I <II
~ Sn: szz 1000 "-
100 500 I: Temperature I -Pressure Time(s) 500 1000 1500 2000 2500 3000 3500 Time(s)
Figure 9 Transient A Sp and Sn Component Stress Histories Figure 7 Loading Conditions for Transient A Transient B Transient B Sp and Sn Component Stress Histories 600 3000 f\.
500 2500 I
u:-
5l'400 e...
~ 300 2000 :::-
C/I
.e e
...
=-
50 40
""""
~~
I
~ -
-
-
sp: SXX Sp: SYV
~~
Sp: SZZ e" I/
1500
~~ k~ I --
- l
.
1!
~200 1000 "-
C/I C/I e
C/I II) iii
-y~y Sn: sxx Sn: SYV Sn: SZZ
~ --.::J 100 500 10
-Temperature I I -Pressure 500 '000 '500 ,"'" '500 500 1000 1500 2000 2500 3000 3500 Time(s) Time (5)
Figure 8 Loading Conditions for Transient B Figure 10 Transient B Sp and Sn Component Stress Histories A stress model of a thick-walled component was used to Table 4 Fatigue Significant Peaks and ValJeys*
evaluate these transients. A thick-walled component was chosen because it will increase the primary plus secondary Trans. A Sp Sn Peak Time (ksi Sn Trans. B Peak Sp (ksi stress lag behind total stress for a thermal excursion. The (5) ) (ksi) Time (5) (ksi) )
primary plus secondary and total stress component histories are 0 1 1 0 3 3 shown in Figure 9 and Figure 10 for transient A and transient B, respectively. The stress intensity responses for these two 167 78 28 158 76 24 transients are shown in Figure 11 and Figure 12. The 531 65 48 496 62 43 peak/valley times were selected using the stress intensity based 3500 32 34 2500 18 15 approach for primary plus secondary stress and total stress histories. A summary of the peaks/valleys selected can be seen 2700 63 60 in Table 4.
3500 57 55
- See Figure 11 and Figure 12 6 Copyright © 2010 by ASME
Transient A Total Stress ys. Primary Plus Secondary Stress Table 5 Alternating Stress Intensity 9o.---------------------------------------~
Sp Peak@ 167 seconds
, / SI:Sp = 78 ks; SI:Sn =28 ksi Trans. Time Trans Time 3Sm Sn Ke Sp Emod Salt A 531 B 2700 60 107 3.3 129 1.05 226
~60+4+---~~------------------------------~
~ A 167 B 3500 60 83 2.3 135 1.04 160
'iii Sn Peak @ 3500 seconds c
.! SI:Sp = 32 ks;
.E SI:Sn = 34 ksi B 0 B 158 53 27 1 79 1.1 55
\
ill
~30~+-~~~~--~~~~~_=----_.~~~~~ B 496 A 0 58 45 1 63 1.06 37 iii Sn Peak @ 531 seconds SI:Sp = 65 ksi SI:Sn = 48 ksi A 3500 B 2500 57 24 1 24 1.07 14 j-SI:SP! Note: All recorded hme IS III seconds and stress IS III kSI
-SI:Sn O~.~--~----~----~----~----~----~==~
500 1000 1500 2000 2500 3000 3500 From Table 5, it can be seen that a pair including a peak Tlme(s) selected from a primary plus secondary stress intensity peak Figure 11 Sp and Sn Stress Intensity Responses for resulted in the highest alternating stress intensity. This is Transient A because of the primary plus secondary stress range and the resulting Ke penalty, which significantly increased the total alternating stress intensity calculated for that pair. This Transient B Total Stress YS. Primary Plus Secondary Stress illustrates the importance of considering both stress quantifies 90 ........................................................................................................................................... in the selection of peak/valley times to consider in the fatigue Sp Peak @ 158 seconds
, / SI:Sp = 76 ks; Sp Peak @ 3500 seconds analysis.
SI:Sn = 24 kSI Sp Peak @ 2700 seconds SI:Sp = 57 ksi SJ:Sp = 63 ksi SI:Sn = 55 ksi "w SI:Sn= 60 ksi ~ CONCLUSIONS
~60ttt---~~--------------------~~~==~~
~
'iii I:
The stress state selection method described here is notable S because it is a repeatable process that can be taught and applied
.E II) using automated methods. The method improves the overall e30+4~~~~~~~~~~----~--f---------~
Cij quality of the work performed because the method ensures that all significant states are identified in the stress time histories.
I-SISpl While it is possible to apply the method manually using
-51:5n graphing techniques, the method is best implemented using an 500 1000 1500 2000 2500 3000 3500 Time(s) automated process. Because the method is easily automated, it is ideal for use in both design calculations as well as in an Figure 12 Sp and Sn Stress Intensity Responses for online monitoring role.
Transient B From Figure II and Figure 12, it is not clear which stress ACKNOWLEDGMENTS state pairs would result in the highest alternating stress range. The authors wish to acknowledge the significant The algorithm employed by WESTEMS' automatically contributions made by Dr. C. Y. Yang and A. L. Thurman, who calculates the actual primary plus secondary stress and total helped to develop and refine the method described in this stress ranges, based on the stress components, for each possible paper.
pair to reveal which stress state pairs would result in the highest alternating stress intensity. Table 5 summarizes the REFERENCES stress state pairs that result in the highest alternating stress intensity, Sa. [I] ASME Boiler and Pressure Vessel Code,Section III, Division 1, Subsection NB, 2007, "Rules for Construction of Nuclear Facility Components, Class I Components," American Society of Mechanical Engineers, New York.
7 Copyright © 20 I 0 by ASME