ML13333B186
| ML13333B186 | |
| Person / Time | |
|---|---|
| Site: | San Onofre  |
| Issue date: | 06/11/1984 |
| From: | Hamid A, Harris H DREXEL UNIV., PHILADELPHIA, PA |
| To: | |
| Shared Package | |
| ML13333B185 | List: |
| References | |
| NUDOCS 8412060202 | |
| Download: ML13333B186 (50) | |
Text
~~ATTAC4AIM~12 EVALUATION'OF THE SAN ONOFRE NUCLEAR CENURATING STATION, UNIT I MASONRY WALL TEST PROGRAM RESULTS AND CORRELATIONS for The Franklin Research Ceater Philadelphia, Pennsylvania Prepared by Dr. Ea::ry G. Harris and Dr. Ahmad A. Hamid Yepartment )f Civil Engineering Drexil University Thiladelpnia, Pennsylvania June 11, 1984 8412060202 841106 PDR ADCPDR P
INTRODUCTION A review of the various References 1-5 was carried out in support of the FRC seismic evaluation effort of the reinforced concrete masonry walls of the San Onofre Nuclear Generating Station, Unit 1. A very extensive analytical effort coupled wit. a confirmatory test program was undertaken by the licensee in an attempt to meet the NRC requirements.
Although this has been a very Com mendable undertaking, breaking new ground in the analysis and testing of rein forced masonry structures under earthquake loading, it is felt by the writers that several concern. must be resolved before the NRC requirements are satisfied.
The specific items of concern are discussed in detail below under the several sub headings.
CORRELATION WITH ANALYTICAL MODEL
- 1. Effect of Vertical Reinforcement Eccentricities An interesti.tg question israised in the study (Ref.
- 1) by the fact that the eccentricities in the rebars for specimen-lB were measured to be 0.8 in. off center. The fact that the analysis does not take these variations into account in the actual walls is of concern. In the actual wails these variations can be expected to be at least as large as the ones found in the "carefully constructed" lab test spec imens but most probably they would be expected to be larger. Thus the discrepancy found in correlating the detlections of Wall 1B points to a weakness in the analytical model since the eccentricities in real walls will cause larger deflections than those predicted by the model.
- 2.
Masonry Strains The reported values of masonry strain (Table 3.1 of Ref. 2) do not agree with reported analytical face-shell stresses for the normally 1*
expected ialues of masonry modulus of elasticity.
This discre pancy and lack of stress-strain data is inconceivable in a compre hensive program of this type.
The cost of making instrumented prisms and obtaining a stress-strain curve for the material is small compared to the overall effort undertaken.
Only if such information is obtained can a realistic comparison of the test data and the analytical.predic tions be made.
- y
- 3. Steel Yield Strength The comparisons of analytical and test results cannot be made as indicated in Ref. 1 since very different values of steel yield strength were.. sed in.the test-specimens lA, 13 and IC and the analytical model. A reanalys-s of these walls should. be made to check the.correlation.
The analytical model should also include the yield length as a parameter to check the.sensitivity of thcse predictions.
\\4.
Length of Yielding Rebars Large differences were found between the length of yielding rebars used
-in the analysis and that measured in the tests.
It is argued that the reason could be partially due to the higher steel strength. Other factors could contribute in increasing the length of the plastic hinge.
The use of shorter yielding length is conservative in determining steel strain ratio but may be unconservative in predicting response and de flection oi the test walls. A senstivity analysis should be conducted to study the impact of the length of yielding rebars on the wall re sponse and overall behavior under prescribed loads.
- 5. Permanent Set in Wall Defections The permanent offset of the walls of type 1 is a prime concern and it cannot be atributed to solely observed rebar eccentricities or the 2
higher dynamic input of testing.. The real concern is that it did occur, even when a much stronger steel was used and thus most pro bably the actual walls will have permanent set if subjected to such an earthquake as that of the test program.
The fact that the analytic l model did not predict any permanent set is a concern and it indicates that the model needs refinement.
TEST SPECIMENTS
- 1.
Top and bottom 3 courses of the speciments were fMlled solid.
Is this a similar detail to the actual walls? How was partial grouting modeled in the analysis?
- 2.
It is stated that 6 #5 were used horizontally.
The drawings show only 5 #5 (pq.
15 Ref. 1)
- 3.
Dowels at footing are unsymmetrically placed with only need for the vertical steel.
Why are the other thzee bars included?
(Fig. 4.1, Ref. 1).
- 4.
Does the stel percentage represent all the walls?
p
= 0.25 %
ei =0.08% Discuss range of P in different walls at San Onofre Station, Unir 1.
- 5.
Test cylinders of mortar used are not standard.
- 6.
It is felt that the attachments which were not included in wall tests of the type 1 could be significant in terms of the stability problem posed and the tendency to increase the compressive stresses on the faceshell.
TEST SET-UP AND TESTING PROCEDURE
- 1. How is the top boundary rotation of the wall accounted for in the model? The vertical span of the wall is 22 ft rather than the
actual heigit of 24 ft.
2 hat is the effect of filtering low level frequencies of the input motion on the wall stability?
- 3. How were the joint opening measurements used in the correlation?
- 4. Wire pots (WP04) 'were plotted at 32 in. above the centerline. What about the results of WP05, WPO6 and WPO7 at the centerline for walls lB and 1C?
- 5. Was spalling of faceshell observed at mid-height region of the test walls?
- 6. No attempt was made to test the walls to failure to be able to predict the margins of safety.
PERFORMANCE AND SAFE!T OF THE WALLS It is concluded that the masonry walls were able to withstand without any significant damage the DBE load. The test wall ' perienced excessive deforma tions as indicated frum large dispfacements, thi length of yielding rebar, spalling of faceshellsand permenant set of deflecti.ons.
Because the walls are reinforced and securely anchored at top and bottomthe instability is highly unlikely. It is not clear what is the definition of wall failure and
- how the margin of safety against collapse can be determined for the masonry walls at San Onofre Station.
Also it is not clear whether or not permenant set and large displacements are acceptable from an operability standpoint.
CONCLUSIONS AND RECOMMENDATIONS It is concluded that direct correlation between the test data ahd the analytical model can not be made because of the discrepancies in material properties,detailing and input motions.
Assessment of the conservatism of the analytical model can not be made at the present time.
Reanalysis of the test walls reflecting actual material properties, length of yielding rebars, 4
eccentricity of the rebars and input motion should be done. Evaluation of the
-margins of safety against collapse of the masonry walls at San Onofre Nuclear Generating Station Unit 1 should be performed.
REFERENCES
- 1. Computech Engineering Services, Inc., Report No. R557.09,"San Onofre Nuclear Generating Station, Unit 1, Seismic Evaluation of Reinforced Concrete Masonry Walls - Masonry Wall Test Program Results from Testing Walls 1A, lB and IC and Appendice A, B and C," Berkeley, CA, February 1984.
- 2. Computech Engineering Services, Inc., Report No. R557.10,"San Onofre Nuclear Generating Station, Unit 1, Seismic Evaluation of keinforced Concrete Masonry Walls, Masonry Wall Test Program Correlation With Analysis Results," Berkeley, CA, February 1984.
- 3. Computech Engineering Services, Inc., Report No. R557.07,"Masonry Wall Test Program -
Test Results Summary -
wall No. 2A," Berkeley, CA, February 1984.
- 4. Computech Engineering Services, Inc., Report No. R557.0b,"Masonry Wall Test Program -
Test Results Summary -
wall No. 3A," Berkeley, CA, February 1984.
- 5. Computech Engineering Services, Inc., Report Nc. R557.11, "Masonry Wall Test Program -
Test Results Summary -
Wall No. 3B," Berkeley, CA, February 1984.
6
ATrA~mcG.NjT3*LaJc&
- pacJ 1
f BASHC PREMS 1F"T f~jh
- Reinforced am u a z. 0 %
cosierable nElast*-
':~ ins2~ Ed' ela:
re
BASIC AIMS OF TEST PROGRAM
- Demfonstrate validity of basic assumiptions of mfethodology I
- Demfonstrate conservatismf of analytica.1 methodoli0g
OVERVIEW OF TEST PROGRAM Three wall types tested:
-.Each specimen 8 feet wide.
- Each specimen 8 inch reinforced concrete masonry.
- Horizontal reinforcing #5 _
48" o.c.
- Wall t pe 1 representing Fuel Storage Building and Ventilation Building:
-2 committed, 3 tested.
- Height 24 feet.
- Vertically #7 rebars @ 32" o.c.
- Amplified inputs at top and bottom.
- Wall type 2 representing Reactor Auxiliary Building:
- 1 cointtedi I tested.
- Height 16' - 8".
- Vertically #4 rebars t 32" 0.c.
- Ground type motion applied both at top and bottom.
OVERVIEW OF TEST PROGRAM (Continued)
- Wall type 3 representing Turbine Building:
1 committeA, 1 tested.
- Height 21" - 4".
- Vertically #5 rebars I 32".c.
- Weights added to test specimen.
- Ground input applied at bottom of specimen Amplified motion applied at top of specimen.
- Measured data includes:
Displacement profile.
- Acceleration profile (including input).
- Joint gap openings.
- Steel strains.
- Tests performed were:
- Material tests for determining material properties.
--Low intensity tests for determining damping.
- Full intensity tests.
SUMMARY
OF RESULTS WALL lA WALL 18 WALL I WAU. 2A WALL 381 Center Displacement (inches)
Maximum 9.24 7.86 11.28 0.21 2.13 Minimum
-10.47
-12.38
-10.75
-0.19
-2.87 Maximum Masonry Compressive Strain 0.0029 0.0038 0.0029 0.0003 0.0012 Steel Strain Ratios Center 4.2 4.8 4.8 1.0 End 10.5 (1) 10.0 (1) 10.0 (1) 3.2 Length of Yielding Rebar 881 96' 96" NOTE (1.)
Strain values at the end of the wall are minimum values for wail type 1 as the strain gage readings exceeded their maximum scale values.
FUEL STORAGE BUILDING
_AVERAGE ANALYSIS Center Displacement (Inches)
Maximum 9.46 9.03 Minimum
-11.20
-9.93 Masonry Compressive Strain 0.0032
()
Steel Strain Ratios Center 4.6 18.7 End 10.2 (2) 20.7 Length of Yielding Rebar 93*
18" NOTES (0.)
Analytical values did not provide a maximum masonry compressive strain. The evaluation criteria required a check of the average face shell comresslve stress.
(2.).
The test value is a minimum value as the strain gage readings exceeded the maximum values for the instrument.
&?4eWAo.'
-TSEN-iE GM 7dUS PAGE 13 PROPMT, 5OUTHERN4 C=-OCFMA EDW*OM4 COMPANY Figure Withheld from Publ ic Disclosure
MW"Ma"=N PRESEMTED ON THS PAGE 53 PFCPFIEETA:I TO ZOUY CALLFO4MA EDMOM 00APANY Figure Withheld from Publ ic Disclosure
.AFomwATION P4MEENT.D C'M 7~
- ^G a
,prTJw Tro
.3 Figure Withheld from Public Disclosure
WFORATION PRESENTED ON THU PAGE I PRPRETARV TO SOUTHEN CAUFORNA EDISON COMPANY Figure Withheld from Public Disclosure 12
e0 CORRELATION WITH ANALYTICAL MODEL
- Effect of reinforcing eccentricities a
Masonry Strains a
Steel Yield Strength
- Length of Yielding Rebars a
Permanent Set in Walls
EFFECT OF REINFORCING ECCENTRICITIES "The analysis does not account for eccentricity of the rebar. Variations in actual walls would be expected to be at least as large as those in the test specimens.
I "Perfect" construction assumed in analysis
- 0.8" offset noted in one wall specimen a
Such offsets not noted in SONGS-I walls.
- SONGS-i walls constructed using nuclear quality control standards.
- SONGS-1 walls 40'-O" between control joints, therefore more likely to be uniformly distributed.
The effects noted in the wall tests are therefore likely to form upper bounds on those to be expected in the actual walls.
MASONRY STRAINS "Stress-strain values do not agree with normally expected modulus of elasticity values. Stress-strain data is needed to compare test data and analytical predictions".
- Strains greater than strain at which maximum stress occurs.
- Tests would require falling branch - difficult to obtain on most test machines.
- Parameter studies previously performed examined possible shapes of stress-strain curve.
- Pre-test conclusion was that strains would not cause compression fai-lure and this was confirmed.
Parameter study examined variations in masonry strengthi steel overstrength, elastic modulus) falling slope and steel yield length.
Steel yield length was by far most important parameter.
Stress-strain relationship is not an integral part of the analysis methodology. -Therefore this information was not deemed neccessary.
- r.
J 0V S40 44 0/
Cbsbo-O
/.,'
STEEL YIELD STRENGTH "Comparisori of test and analytical results cannot be mane,,
because of lifferent values of gield strength.."
- Steel with a minimum specified gield strength of 40 K 1 was purchased. However) sample testing of mater: 31 showed steel yield strength was 507.
highei than used in analysis.
- Actua. SONGS-1 yield strength beyond specified minista of 40 ksi is unknown - usual experience is
+101 o +254.
- Incre sed steel strength would:
nerease face sh-ell stresses and strains nd thus potential for compression ailure mode.
roduce higher bond stresses) therefore ncreasing yield length.
elay onset of yielding) and therefore
,oftening.
crease steel strain ratios.
LENGTH OF YIFLDING REBARS "Large differences were found between the length of yielding rebar in the analysis and the tests. This may be unconservative in predicting deflections."
t Average of 93" in tests, equivalent to maximum strain levels over approximately one-half this length or 46". Length of 18" was used in analysis.
Parameter study previously reported studied effects of varying yjield length. Reducing length from 18" to 10" decreased deflections by 2%.
Increasing length from 18" to 30" increased deflections by 5%.
e:
Shorter lengths were conservative for steel strain ratios. Increasing yield length from 18" to 30" decreased steel strain ratios by 30%.
- Shorter lengths may underestimate deflections by a small amount.
PERMANENT SET IN WALLS "The permanent offset in wall type 1 is a prime concern. The actual walls will probably have a permanent set which was not predicted by the analysis."
I Agreed that permanent-set likely in actual SONGS-I walls.
- Likely causes are self-weight and rebar eccentricity - effects not included in analysis.
- -Did not impair wall stability.
- Wall type 1 walls are located in Fuel Storage Building above fuel deck level with no safety related equipment likely to be effected.
- No permanent set in Auxiliary Building walls and only 0.4" in Turbine Building walls.
- From rigid block deformed shape geometry, a gap opening of 1/4" at the center could cause a permanent set of 8" either way.
TEST SPECIMENS
- Grouting detail. How modelled'?
- Number of horizontal rebars (Wall type 1)?
- Number and placement of dowels?
- Steel percentage. Representative?
- Mortar cylinders?
- Attachments (Wall type 1)?
GROUTING DETAIL
- Grouting detail for test specimens is similar to SONGS-i walls.
- Weight of all grouting included in model.
Only cracking and yielding stages of wall response modelled.
Effect of grouting on stiffness and strength minor after cracking occurrs.
/*
NUMBER OF HORIZONTAL REBARS
- For wall type 1 representing the Fuel Storage Building 6 #5 rebars were used horizontally.
C 0 MPUTEH
"'"~~~
3W'-s.'
.41 0
f I.
no"u 4.1 Pma Sa mug Bu TGO SP~ckywns AWPO.Ue OusSf 13
r
~
NUMBER AND PLACEMENT OF DOWELS I
Vertical steel #7 @ 32" a. c. at SONGS-1.
- Dowels #5 @ 16" o.c. at SONGS-1.
- Dowel steeltarea same as vertical steel area.
- Non-symmetry required in test specimens by construction layout.
- All vertical rebars were required to be lapped by a dowel.
STEEL PERCENTAGE All SONGS-i walls have #5 rebars 48" o.
- c.
horizontally (0.08%).
Vertical reinforcing ratio of 0.25% in Fuel Storage and Ventilation Buildings.
- Vertical reinforcing ratio of 0.08% in Reactor Auxiliary Building.
- Vertical reinforcing ratio of 0. 12X in Turbine Building.
- Steel ratios in test specimens representative of all relevant SONGS-i buildings.
MORTAR CYLINDERS M
Mortar cylinders not an ASTM standard.
- Cylinders have been standard practice in California for the past 20 years.
- Cylinders have been used in all Berkeley masonry tests since 1972.
- No correlation between mortar cube strength and mortar cylinder strength exists.
- Mortar strength is not used directly in either correlation with analytical results or safety evaluation of SONGS-1 wallS.
ATTACHMENTS - WALL TYPE 1
- Wall type 1 represents upper levels of Fuel Storage Building.
a Existing equipment is lightweight electrical conduit etc.
- Total equipment weight is less that 2%
of wall weight.
- Attachments mounted near top and bottom where effects on response are minimized.
TEST SETUP AND TESTING PROCEDURE
- Top boundary rotation in model?
Vertical span of wall type I?
- Effect of filtering on response?
- Joint opening measurements in correlation?
- Wirepot plots in report?
- Faceshell spalling at midheight?
Failure testing - Margin of safety?
TOP BOUNDARY ROTATION IN MODEL
- Model top boundary condition rotation is similar to that in test specimen and actual SONGS-I walls.
- No rotational restraint.
- No vertical restraint.
- Lateral (out-of-plane) fitity.
VERTICAL SPAN - WALL TYPE 1
- Course by course correspondance of test walls to actual SONGS-1 walls.
Walls have a support I foot below top elevation for an effective span of 23 feet.
EFFECTS OF FILTERING e Input motions filtered to remove all frequencies below.0. 35 Hi.
- Filtering, required by 'velocity and' displacement limitations of test actuators.
Wall type I (most flexible) had after cracking frequency of1.0 1.2 Hz.)
reducing to 0.5 0.6 Hft after yielding.
- Wall1 frequency al-ways higher than cut-off frequency-of the high pass filter.
S Thereforet effect of lowremo4ency friltering considered negligible.
JOINT OPENING MEASUREMENTS IN CORRELATION e Measurements not used directly in correlation.
SMeasurements used to
-calculate neutral axis locAtion and section curvature of test specimens.
- Neutral axis location and section curvature then used to calculate masonry faceshell strain.
Direct measurements not used because of eccentra~.
location of instruments requiring the use of, simultaneous readings from both wall faces.
I
WIREPOT PLOTS IN REPORT
- Body of test report includes displacement traces at which maximum displacement occurred.
- For wall lA maximum disph ement occurred at mid height.
- For walls-1B and 10 maximum displacement occurred 32" above mid-height.
- Time histories for all measured points included in Appendices A, B and C for walls lA, lB and 10 respectively.
FACESHELL SPALLING AT MID-HEIGHT s No spalling whatsoever of faceshell was observed during or after the tests at mid-height region of the test walls.
FAILURE TESTING - MARGIN OF SAFETY
- Objective of test program was to demonstrate conservatism of evaluation procedures, a
Failure defined by wall collapse.
- rebar fracture
- anchorage fracture
- not masonry compression a
Loading used caused rebar strains less than one quarter of fracture strain.
- Input motion sufficient to cause wall failure beyond capability of test equipment.
- Level of input motions mutually agreed upon with NRC staff prior t.o commencement of test program.
- Repeated tests on reinforced masonry walls and obtaining meaningful results not possible because of cumulative, permanent extension of yielding rebar.
POGMATON PRES~Ifl! ON 1TU8 PIAOC 0 P~FAM7 TO SOTNP CM436 WcoHw" Figure Withheld from Public Disclosure
WOOAOTMON PWESETEO CM flS P*0e 1 PR~pEW1V" To SOUTtMi CALWFOMMA EDMM CMPY Figure Withheld from Public Disclosure
PERFORMANCE AND SAFETY OF WALLS
- Criteria for steel fracture, masonry compression failure and wall instability were met in both tests and analyses.
- Margins of safety were included in the input spectra and enveloping time histories. Tests were "proof test" in nature rather than "fragility" type.
- Displacements would not impair operability of SONGS-1.
CONCLUSIONS
- Test results and analytical predictions were extremely close for non-linear behavior.
- With.in practical bounds, the assumptions of the methodology have been validated.
a Seismic safety of the SONGS-1 masonry walls has ben denonstrated.
ATTACHMENT 4 ACTION ITEMS SAN ONOFRE NUCLEAR POWER PLANT
SUBJECT:
MASONRY WALL DESIGN (IE BULLETIN 80-11)
In accordance with the agreement reached during the meeting on September 5 and 6, 1984 with the representatives of Southern California Edison Company, the following list of action items is to be completed by the licensee:
- 1. Information regarding the connection details of the walls in the fuel storage building (details at the intersection of the intermediate concrete support slab and wall) will be provided.
Verify whether the connection details in this case were reflected in the analysis and test panels. Provide drawings nos., 568138, 568141, 568135.
- 2. Assess the impact of the vertical control joint spaced at a 4-ft interval in the walls (while the test panels were 8 ft wide) on the overall testing and analysis program. Also, provide information regarding the connections between the bond beam and these vertical joints.
- 3. Verify-whether the load of the roof in several buildlngs was included in the analysis. If not, please provide justification.
- 4. Please provide information regarding the QA/QC program for masonry wall construction at the plant. From the available data relative to the QA/QC subject, please provide necessary assessments (by
-2 some statistical means) about the existence of the reinforcement and boundary connections (i.e., dowels, fully grouted at the bottom three courses, etc.). It is understood that some portions of walls were removed and photographs taken to confirm the details of wall construction. Please provide these photographs for review.
Resolution of this item is necessary prior to return to service.
A7TAC&C-JT K.-&4ASONZ0 tojl PNOTOWLAPHI
T.
1op vc.k ~j
.4
4b Me-J
.C---... I ft.7$-
s'reu..-I,
'1.
1~
oh' £~-.79
.ir.hd..
S a
- a.
w a
~*
I.A&t~
o~ wr'.A Mckts ~o t~.scdt Sbtas ~ W0J44