ML19343A176
| ML19343A176 | |
| Person / Time | |
|---|---|
| Site: | Sequoyah  |
| Issue date: | 09/10/1980 |
| From: | Greimann L AMES LABORATORY, ENERGY & MINERAL RESOURCES RESEARCH |
| To: | Schauer F Office of Nuclear Reactor Regulation |
| References | |
| CON-FIN-A-4141 NUDOCS 8009160335 | |
| Download: ML19343A176 (5) | |
Text
'
lowa State University Ames, Iowa 5001i HIIIeS lahornton-Energy & Mineral ResourceyResearch Institure s
September 10, 1980 Dr. F. P. Schauer Chief Structural Engineering Branch Division of Systems Safety Office of Nuclear Reactor Regulation Nuclear Regulatory Connission Washington, DC 20555
SUBJECT:
AMES LABORATORY TECHNICAL ASSISTANCE TO THE DIVISION OF SYSTEM SAFETY, NUCLEAR REACTOR REGULATION
" REVIEW 0F NUCLEAR PLANTS STRUCTURAL DESIGN" (FIN N0. A-4141).
Dear Dr. Schauer:
During our meeting of September 3,1980, you requested that I perform certain tasks concarning this and next year's work:
(1) Sumary and evaluation of the various analyses of the Sequoyah containment presented at the ACRS meeting of September 2:
see enclosure.
(2)
Interactive computer program in FORTRAN for approximate analysis of steel containments: This will be incorporated into next year's program - see forthcoming Form 189 (current terminology is WPAS),
Task 3(a).
(3) Limitation to above approximate analysis: This will be incorporated into next year's program:
see forthcoming Form 189, Task 3(c).
(4) Usefulness of instrumentation of full-scale containment during over-pressure test:
Instrumentation of an actual containment vessel during the overpressure test would provide useful data for comparison with analytical results.
By placing strain gages at a number of locations on the shell and stiffeners, strains could be measured in the linear behavior range of the containment vessel. The data could be used to verify and/or improve the linear analytical results.
For example, the effect of stiffeners on shell behavior could be examined.
However, these experimental results should not be expected to verify a nonlinear, limit pressure analysis of the containment.
Only pressure tests of models to leakage would provide such information.
Development and implementation of the instrumentation and data
.5 acquisition scheme for such a test would require careful study.
If you decide that such an experimental program would be useful, j
s p
we would be very interested in submitting a proposal to perform f
the work.
eovoien335 under Contract With The United States Departenent Of Energy
Dr. F. P. Schauer September 10, 1980 (5) Extension of previous dynamic analysis te cover other size compart-ments and vent areas: Work is beginning on this item.
Before this work can be completed, we request that you furnish us with realistic information regarding the pressure pulse for the dynamic loading.
In particular, the pulse magnitude and duration are requested along with tte relationship of these parameters to compartment size and venting area.
(6) Form 189:
This will be forthcoming when we receive the Statement of Wor'< and NRC Form 173.
(7) Comments concerning proposal by Morris Reich for interactive struc-tural hardware and software system:
I believe Mr. Reich's proposal would be a worthwhile effort, and I would look fonvard to participating in such an effort. For the immediate future (FY 81, before the inter-activa capability is available) we would intend to use his program, NFAP, for a significant portion of our finite element analyses.
Sincerely,
}
A_---
h.
Lowell Greimann Project Engineer Enclosure cc w/ encl:
Director, Division of Systems Safety Attention:
B. L. Grenier Harold Polk Program Manager Delwyn D. Bluhm, Head Project Engineering
L. Greimann SUPeiARY & EVALUATION OF SEQUOYAH CONTAINMENT ANALYSES Presented at ACRS MEETING, SEPTEMBER 2, 1980 APPR0XIMATE METHODS Common Assumptions
- penetrations do not control
- uniform static internal pressure
- von Mises yield criteria
- local bending effects neglected
- axially symmetric Ames Lab L. Greimann (January 1980)
-Fy = 32 ksi
- uniform stress in stiffeners & shells at limit pressure
- limit pressure = 36 psi (+30%, -10%)
R & D Associates, F. Parry
- Fv = 32 ksi
- l'inear elastic analysis of stringers
- rings neglected
- yield pressure = 27 psi TVA
-F
= 45.7 ksi
- n$alect all stiffeners
- yield pressure = 38.2 psi NRC Research, G. Bagchi
-Fy = 32 ksi
- strincers carry pressure load in bending and transmit it to the rinas
- limit pressure = 34 psi Ames Lab, L. Greimann (September 1980)
-F
= 35.2 ksi
- c$mplete vielding of rings and strincers (mechanism)
- limit pressure = 44 psi Offshore Power Systems, R. Orr
- F, = 45 isi
-sdearedeings
- stringers neglected
- limit pressure = 50.5 psi (9'6" rina soacing) 56.8 psi (6'6" ring spacing)
Page Two L. Greimann FINITE ELEMENT ANALYSES Common Assumptions
- penetrations do not control
- uniform static internal pressure
- von Mises yield criterion (except as noted)
Ames Lab, L. Greimann
-F
= 35.2 ksi
-gEometricandmaterialnonlinearitiesincluded
- axisyncetric analysis of complete shell including rings and stringers (neglect circumferential variation of displacement)
- ANSYS program
- limit pressures by ASME b linear slope method = 47 psi Offshore Power Systems, R. Orr
-F
= 45 ksi
-gkometricandmaterialnonlinearitiesincluded
- analysis of typical curved 5" panel bounded by a ring and stringer
- ANSYS program
- no limit pressure predicted but loa '-displacement curve becomes quite flat at 50 psi pressure Franklin Research Center, Z. Zudans I
-F 32 ksi
- 15n=arly elastic analysis e
- analysis of typical curved 5/8" panel with ring and stringer which verified smearing of stringers
- axisymmetric analysis of a shell length including rings and smeared stringers (neglects circumferential variation of displacement)
- pressure with average hoop stress at yield = 30.5 psi (max. shear stress criterion)
COMPARISON For a valid comparison, multiply pressures by 45 ksi/F since 45 ksi is the approximate actual yield strength.y If the maximum shear stress criterion was used, convert to the von Mises criterion by multiplying by 1.15.
Yield Pressure R & D Associates (Approximate) 38 psi TVA (Approximate) 38 psi Franklin Research (Finite Element) 49 psi i
a Page Three L. Greimann Limit Pressure Ames Lab - Jan. 1980 (Approximate) 51 psi NRC Research (Approximate) 48 psi Ames Lab - Sept. 1980 (Approximate) 56 psi Offshore Power Systems (Approximate) 50 & 57 psi Ames Lab (Finite Element) 60 psi Offshore Power Systens (Finite Element)
Not Specified OBSERVATIONS
- The yield pressure of 38 psi predicted by R & D Associates and TVA is very conservative because all stiffening is neglected.
- The yield pressure predicted by Franklin Institute (49 psi) is unconservative because it involves complete through thickness yielding.
On the other hand, 49 psi is a conserva-tive limit pressure because a full mechanism has not formed in the analytical model.
- The limit pressures predicted by Ames Lab - Jan.1980, and NRC Research, are conservative because a complete collapse mechanism has not been formed in the analytical model.
- The finite element analyses of a typical panel by Offshore Power Systems and Franklin Institute show that the circum-ferential variation of displacement is insignificant.
Therefore, an axisymmetric finite element model, as used by Ames Lab and Franklin Research, is adequate.
CONCLUSIONS Within engineerino accuracy (T.+10%), the liuit pressure is probably between 55 and 60 ps A limit pressure can be achieved if the structure has adequate ductility.
L