ML19291C041
| ML19291C041 | |
| Person / Time | |
|---|---|
| Issue date: | 01/09/1980 |
| From: | Mattu R Office of Nuclear Reactor Regulation |
| To: | |
| References | |
| REF-GTECI-A-07, REF-GTECI-B-06, REF-GTECI-CO, REF-GTECI-PI, REF-GTECI-RV, TASK-B-06, TASK-B-6, TASK-OR NUREG-0484, NUREG-484, NUDOCS 8001110145 | |
| Download: ML19291C041 (10) | |
Text
.
NUREG-0484 SUPPLEMENT 1 METHODOLOGY FOR COMBINING DYNAMIC RESPONSES R.K. Mattu Assistant Director for Engineering Division of Systems Safety Office of Nuclear Reactor Regulation U. S. Nuclear Regulatory Commission Washington, D.C.
20S55 1734 106 8001110/Y
CONTENTS 1.
INTRODUCTION 2.
SSE AND LOCA DYNAMIC RESPONSES - NUREG 0484 3.
CRITERIA FOR COMBINATION OF DYNAMIC RESPONSES OTHER THAN TH0:i. OF SSE AND LOCA 4.
BROOKHAVEN NATIONAL LABORATORY GENERIC STUDIES ON RESPONSE COMBINATION METHODOLOGY STATUS 5.
CONCLUSION 1734 107
di i
a NUREG 0484 SUPPLEMENT 1 INTRODUCTION In the analysis of structures, systems, and components important to safety, the NRC has required the combination of structural / mechanical responses due to various accident loads and loads caused by natural phenomena, particularly earthquakes. This requirement flows from 10 CFR Part 50, Appendix A, General Design Criterion 2 which was issued in 1971 and calls for an appropriate combination of the above loads to be reflected in the design bases of safety equipment. The requirement has been implemented in various ways both within the NRC and tha Nuclear Industry.
The loads due to postulated accidents and natural seismic phenomena often yield dynamic responses of short duration and rapidly varying amplitude in the structures and components exposed to the loads. These loads usually have no physical time phased relationship in the accident analysis either because the loads are random in nature or because the loads have simply been postulated to occur together without a known or defined coupling. An example of this would be an earthquake (SSE) which can be postulated to occur together with a large pipe break to give the classic LOCA plus SSE combination of response. Lacking a physical basis for relating some of the loads in question, they have been required to be combined for design purposes in the licensing process according to the absolute or linear sum methodology, i.e., summation of 1734 108
\\
\\
b
3.
'I 2_
peak structural responses due to each of the individual loads. This approach may lead to overly conservative design requirements for certain plant systems and may result in more rigid systems which is not beneficial when designing for thermal stresses which are present in normal day to day operation.
SSE AND LOCA DYNAMIC RESPONSES - NUREG 0484 The NRC staff Working Group established to develop recommendations for combining dynamic responses in the analyses of structures, systems and components important to safety published in September 1978, NUREG 0404 (Methodology for Combining Dynamic Responses) which allow the use of SRSS technique for combining short duration, rapidly varying dynamic responses of LOCA and SSE within the Reactor Coolant Pressure Boundary and its supports.
NRC has adopted the working groups recom-mendations which states:
"For the combination of the dynamic responses within the Reactor Coolant Pressure Boundary (RCPB) and its supports, which result from the coincidence of an SSE and LOCA, the Square Root of the Sums of the Squares (SRSS) technique is acceptable contingent upon performance of a linear elastic dynamic analysis to meet the appropriate ASME Code,Section III, Service Limits."
The study described in NUREG 0484 was limited to the primary reactor coolant system because of the limited amount of the time that was made available to the working group, and because of the amount of information 1734 109
. that was available regarding the design and responses of various primary system components to dynamic loads.
Since the publication of NUREG 0484, the staff has had time to do additional work and has had additional substantial information from the Brookhaven National Lab., General Electric Company, the Mark II Owner's Group and from Dr. R. P. Kennedy and Dr. Nathan Newmark.
It is now the staff's opinion that the recommendations and conclusions of NUREG 0484 regarding the use of SRSS methodology for combining the responses due to LOCA and SSE may be extended to any other ASME Class 1, 2 and 3 affected system, component or support. The extension to structures has not been accepted at this time. Caution should be used in combining dependent functions unless time phase relationship of the functions can be established.
In those cases where it cannot be established that the functions being considered for combination are independent functions (e.g., Annulus pressurization and jet impingement or reaction loading in case of a LOCA), absolute summation of responses is required. The staff position for the use of SRSS for combining dynamic responses of SSE and LOCA is that the response to the loads satisfy each of the following:
(i)
The dynamic time function is rapidly varying.
(ii)
Duration of the strong motion portion of the function is short.
(iii) Function consists of a few distinct high peaks in random appearance.
(iv)
Response is not associated with a normal plant operation.
1734 110
\\
w
k (v)
Phasing relationship among functions to be combined are random.
(vi)
Response is calculated on linear elastic basis.
CRITERIA FOR COMBINATION OF DYNAMIC RESPONSFS OTHFR THAN THn9F OF 99F AND LOCA Although the staff recognizes that in several respects the generic Kennedy-Newmark (K-N) Criteria (NEDO 24010-2 Supplement 2) as prepared in cooperation with the Mark II Owner's Group may be conservative, we have not completed the review of the criteria for the generic use of SRSS dynamic response combination methodology. The entire spectrum of which accider' events and natural phenomena require combination, how responses in components, equipment, and structures should be combined and the specification of suitable design margin for such combinations is under active review in general task action plan B-6, " Loads, Load Combination and Stress Limits." It is expected that the staff will complete its review of the generic K-N criteria soon and at that time the conclusions of this review will be published.
Brookhaven National Laboratory under technical assistance contract from DSS is performing sensitivity studies for investigating methodologies of combining dynamic responses, is evaluating the acceptability of the Kennedy Newmark Criteria for application to various dynamic response combinations, and is evaluating whether specific responses resulting from postulated dynamic events satisfy either of the Kennedy Newmark Criteria for SRSS combination method in the case of Mark II plants.
1734 11i
=
- s 8,
BROOKHAVEN NATIONAL LABORATORY GENFDIC STUDIES ON RESPONSE COMBINATION METHODOLOGY STATUS BNL studi.s consists of the following three phases:
(a)
Identify parameters governing response characteristics.
(b) Conduct Sensitivity Studies.
(c) Develop recommendations on combinations criteria and quantify conditions under which ABS cr SRSS or CDF should be used.
Phases A and B of tha Study are nearly complete ar.d tiie ccikpleticr cf the entire study and final report is due at the end of September,1979.
Preliminary indications on sensitivity studies are:
1.
For the combination of two steady state sinusoids:
(a) Two equal frequency sinusoids have non-exceedence probability (NEP) of SRSS near 50 percent.
(b) Two different frequencies cause decrease in NEP of SRSS below 50 percent.
(c) Amplitude ratio effect of NEP of SRSS is not significant.
(d) Different duration times will decrease NEP of SRSS below 50 percent.
(e) The density function for random phase angle has the following effects:
1734 ll2
1
. (i) The primary effects on combining two equal-frequency sinusoids, i.e. fo'r normal distribution is that the smaller the standard deviat#an, the lower the NEP-of-SRSS (lower than 50 percent);
(ii) it has an insignificant effect on combining two differer.t frequency sinusoids (i.e. NEP of SRSS in smaller than 50 parcent because of their difference in frequency).
2.
For combining two sinusoidal time varying (damped) transients, NEP-of-SRSS may exceed 50 percent if the adverse effects of 1 (b) and 1 (d),
and 1 (e) are overridden by the effect of rapid amplitude variation (i.e. high damping).
3.
For combining more than two sinusoids:
(a) NEP-of-SRSS is greater than 50 percent if all sinusoids have equal frequencies, durations and random time lags; (b) NEP-of-SRSS may be greater than 50 percent if the adverse effects of difference in frequency and duration are overridden by the effects of rapid time varying in amplitude.
4.
For combining two steady state multi-frequency responses:
(a) Two equal predominant-frequency responses will decrease NEP-of-SRSS with increasing frequency band width.
The amount of decrease in NEP-of-SRSS is less for lower predominant frequencies.
1734 113 u
3, o
. (b) Two different predominant fiequency responses are a primary factor causing decrease in NEP-of-SRSS.
(c) Shorten duration time in one response will decrease NEP-of-SRS5.
(d) Density function for random time lag is not a significant factor to effect the NEP-of-SRSS.
(e)
In general, NEP-of-SRSS is greater than 50 percent. Less than 50 percent tray occur under jointed effects of 4(b) and 4(c).
5.
For combining two multi-frequency transient responses with sparsity of peaks, NEP-of-SRSS is generally far greater than 50 percent.
6.
For combining more than two multi-frequency transients, a high percentage (much greater than 50 percent) in NEP-of-SRSS is generally expected.
CONCLUSIGN The staff considers the use of SRSS appropriate for combination of SSE and LOCA dynamic responses for all ASME Class 1,
$nd 3 systems, component or e.upport as discussed in the section of this NUREG supplement entitled"SSE and LOCA Dynamic Responses - NUREG 0484."
The criteria for comti.a'.isn of Dynamic Responses other than those of SSE and LOCA are still awaiting BNL study results and subsequent staff review and conclusions as part of Task Action Plan B-6.
1734 ll4 s
6 One of the problems facing the staff regarding Kennedy /Newmark Criteria is that the K-N Criteria uncouples the problems of how loads are combined from which loads are to be combined. Nature fundamentally couples these problems, and no technique which uncouples them is capable of addressing the overall load combination problem. Task Action Plan B-6 is making an attempt to resolve these problems by addressing the directions of work suggested in NUREG 0484.
1734 115
.