ML20138K953
| ML20138K953 | |
| Person / Time | |
|---|---|
| Site: | Comanche Peak  |
| Issue date: | 11/13/1984 |
| From: | NRC COMMISSION (OCM) |
| To: | |
| References | |
| OL-2, NUDOCS 8512190121 | |
| Download: ML20138K953 (44) | |
Text
_..n..,
. y,y..
~
5,$$,%
2
': Q ;; Q; Q..i??:'
Twa q tw e
. y ( *M n
< w.
t,,, gs;<
.r "t
. 7
'~^
~.
r a. ~ #
ra C 5f '+
hbbSh ~n C
-.12 J 4
SSION pgwa
.m/
@Ykk 4'-
,y*
.~. y _,
. ~~
'W/St gas.
5; cL2,
q
' ~~
- frf..,w j, h ~f:
2# ' ' ;' g:GEt a
W, 3 M Y
_~
es_M,s. =
m.,
sc
'^"
M _..
JgA; ins:
'gp *,.
kM 12 -
up-
+1x
.~
y +,
~
_ _ _w re.< " $
).
A
~-
g-
.. ~ "
.* r..,
m y;g4_%
id..
. p- -
...'9Jh* g4,.
g b*
g'eg,~gp]{hgg7ht9, M
- g
.,H g
Q
.,w
~
e
- %yl9.~. y=rg4 n.
hAIUf.+M s,_.
.+
s _.
Wf" -
$D
^
~
- ^ {$;)(g p e W99m~.L.v..w..
w
~w.
e<.w:
4
^
~'
k
-[f?)hd
.;4,; ;
,v ammh..
w
= -
kh A wN
- r.
' a..g'$ Ty pr.}V
&fA?~'l
%f'N@g$$6'MQ
_~ %.
e
, weg g
ggA2qgan ggaaab 2d$%%%i
- ~. 4,
~2 Q'M;f.9_.
~,
^'
ll:k 4
$xw f
Q ~ ~79 yy Ax u
y. r.
..g, n.,m,
.,,55hhkAff?&Y b
~,,en..h.
J hnunc r
v.
m,si%.f*f.,,,4 u.,M ffY p.g$' up.,
?l d i.
m c.1
. g,Z w a 51,,. v. $.W ' $'$,4 %
f ~
,[h* >
l-
~
,x yfyys, <l;.
CR21069.0 1
BRT/Ojg 1
T-UNITED STATES OF AMERICA 2
NUCLEAR REGULATORY COMMISSION
- ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~*
3 In the Matter of:
4 TEXAS UTILITIES GENERATING Docket Nos. 50-445-OL2 5
COMPANY, et al.,
50-446-OL2 4
(Comanche Peak Steam Electric Station, Units 1 and 2) 7 8
Room 110 Phillips Building 9
7920 Norfolk Avenue B'ethesda, Maryland 10 Tuesday, November 13, 1984 11 The briefing in the above-entitled matter was convened I2 purnuant to notice at 10:00 a.m.
I 13 APPEARANCES:
14 JACK REDDING, Texas Utilities Electric Company 15 P. T. KUO, NRC FRANK RINALDI, NRC l
C. J.
COSTANTINO, BNL S. K.
SHARMA, BNL 18 MORRIS REICH, BNL 19 HSIANG CHANG 20 SPOTTSWOOD B. BURWELL 21 ROBERT IOTTI, Ebasco Services Inc.
22 ANIL M. KENKRE, Gibbs & Hill, Inc.
23 l
24 4.4 eses neswers, sac.
23
21069.0 2
CRT 1
PROCEEDINGS 2
MR. BURWELL:
My name is Spottswood Burwell, I 3
am the project manager for Comanche Peak, and I am i.
4 employed by the NRC.
5 The topic this morning relates to the applicant's 6
motion for summary disposition relative to the upper i
7 lateral restraint beam.
This motion was transmitted to us 8
by a letter dated May 20, 1984.
9 This particular motion, I believe on this particular 10 one case, is filed in a partial answer to the applicant's 11 statement -- to the applicant's motion on August 29.
12 MR. REICH:
You should have our answer.
13 MR. BURWELL:
Yes.
The applicant also filed a 14-response to that, dated October 26th.
15 With that, perhaps I could ask P.T. Kuo, or Franklin, t
16 to pick up on where we are.
17 MR. RINALDI:
I'm Frank Rinaldi, I'm the 18 structural reviewer for NRC.
19 Basically, in June '84 we had a meeting to address this 20 problem, and following that meeting, our staff and the i
21 staff from BNL, acting as our consultants, visited Gibbs 22, and Hill for an audit of the calculations in the middle of 23' August.
24 And following that audit, the applicant provided some 25 information on the NASTRAN code that was used in the
--- - _ j-- -
x--
21069.0 3
[RT 1
evaluation of this problem, and also some sample problems 2
related to this.
(
3 BNL has reviewed this, and also pirfo.aad somewha't of 4
an independent check on the specifiy NASTRAN code problem 5
and they have had discussions with the apr11 cant -- their 6
consultant -- to resolve differences in the code.
7 At this point the Staff feels that we should addr'ess 8
any problem so that we could come to a conclusion in the 9
evaluation of this upper lateral restraint beam.
10 I turn the floor to --
11 MR. BURWELL:
May I add a further statement on 12 that?
I believe our meeting this morning is specifically 13 to address the calculational methods used to predict 14 concrete structural failure, which is one part, only, of 15 our review of the upper lateral restraint beam.
16 This has more to do with a resolution -- an 17 understanding between the Staff's consultants and the 18 applicant's consultant, rel'ative to the method used to 19 derive some of their conclusions, rather than to the 20 conclusions themselves.
21 You, I believe, were going to turn the meeting over to --
- 22 MR. KUO:
If I may, I would like to add another 23 statement.
My name is P.T. Kuo.
/
2d During the audit, I believe, the Staff and their 25 consultant, BNL, found the calculations that we reviewed I
e
--n-
,,h
~
21069.0 4
ERT 1
were adequate, with the exception that the modified 2
version of the NASTRAN code was not verified.
And 3
discussion today is basically to discuss with the utility,
+
~
4 about the verification of this modified version of the 5
NASTRAN code.
6 MR. IOTTI:
I guess I would have to make a 7
statement to that, if I may be permitted?
I don't want to 8
imply that we would disagree with the conclusion that 9
today's meeting is to discuss additional verification of 10 the NASTRAN program, but it is my impression that the 11 initial question from the Commission was not whether the 12 code was or was not verified, because the code had been 13 verified against a certain set of' problems; but whether i
14 the verification was all-encompassing and whether it might 15 not be prudent to have additional verification.
This is 16 what we set out to do.
17 So, I don't want to leave the impression that the 18 Commission characterized the code as being used without 19 initial verification.
There were some questions in their 20 consultant's mind that the verification problems used to 21 date m y not have been sufficient to totally exercise the i
s code,ahdye.wereaskedtorunadditionalverification l
22 23 problemsb I
24 e
(
~, tiR. KUO:
You are right.
Thank you.
25 MR..BURWELL:
Thank you.
!i 3_
y l
i
.y.'
-, - -,. - ~ _ - - -, _ -,,,, - - -, - -
'5 21069.0 5
CRT 1
Okay.
How would you like to start?
2 MR. REICH:
It's up to you.
(
3 MR. KUO:
I guess at this time we can turn to
~ 4 Morris Reich and ask them what concerns they might have 5
with regard to the few example problems that were 6
submitted to them.
7 MR. REICH:
Okay.
Let me go over the sample 8
problems and how we looked at them and I would ask my 9
colleagues here to come in whenever they feel they should 10 add on or subtract from whatever words I say.
11 First, four check problems were submitted to BNL.
Of 12 these four, we only were able to take one check problem 13 and ' thoroughly go through it.
t 14 This problem I think is classified as CA-29, and it 15 involves a beam which is held rigidly between two walls, 16 and it is loaded in the middle.
And it was sent to us, we 17 got the problem sometime the beginning of September.
18 A finite element grid was constructed in the same 19 manner as tha one submitted to us in the problem, with the 20 same elements, number of elements -- we tried to put the 21 steel in the same place and everything.
22 Similarly, the same material properties were used.
And 23 we ran this particular problem.
I 24 Now, for low applied roads, that is, let's say we were 25 coming up to 10 or 12 kips; the two runs yielded similar
21069.0 6
ERT 1
results.
However, when we started going higher, the BNL 2
calculation showed that the besm would fail at between 22
(
3 and 24 kips.
And we found that the failure would be due 4
to diagonal tension cracks near the fixed ends.
5 And this value, indeed, was close to a value which was 6
predicted in a paper by S. Sushil -- that's a name.
This 7
is an equation which this man brings in a paper, and it is 8
based essentially on a statistical evaluation of many 9
experimental data points.
10 It is also in the ACI september-October 1984, pages 81 11
-- I think it's 81-83, by Bazant and Kim.
We gave a, copy 12 of this report to EBASCO -- so, the predicted failure is 13 verified by tests, physical tests.
I 14 Now, the calculations that EBASCO did, in that first 15 run that we had, did not show failure even at 32 kips.
It 16 showed cracking b'ut it did not show failure.
17 Furthermore, the results that we saw did not show any r
18 diagonal tension cracking.
And at this point we had a 19 meeting, att=M=4 -- at BNI. -- att=Med by Bob Iotti, and 20 also Harry Chang.
And we presented this differences -- we 2
21 had these differences between our answer and their answer, 22 and at that point we mentioned that, you know, that 23 perhaps it is possible that there should be more than one
(
24 crack there, because we didn't see in any of the answers 25 that we had, that there was more than one crack.
And we a
wy y
p.w-9
--wy---
9
-r--.-y
21069.0 7
CRT 1
talked about that a little bit.
And it was agreed that 2
EBASCO should go back and look at the problem again;
(
3 perhaps there was something missing in the input or 4
something in the code -- and they needed some time for 5
that.
And I also recommended at the time to NRC that we 6
should allow for EBASCO,'two to three weeks, at least.
7 This is, indeed, what happened.
EBASCO returned on 8
November 5th with new runs, and they also gave us a little 9
write-up of that, sort of summarizing the runs.
10 Now, in this summary -- and let me sort of quote from 11 it -- there are tables.
And I will sort of direct you to 12 table 1, where it says that -- first of all, in this new 13 run there were more than single cracks.
In other words, g
14 there were two cracks -- multiple cracks, let's say.
Okay?
15 However, this table shows that, for instance, for shear 16 interlock -- and that's shear retention; in other words, 17 after you get a crack there's still shear retention.
And 18 they changed it.
Harry, or whoever ran the code, changed 19 the shear retention factor from.2 to
.4.
There was 20 virtually no difference in the maximum reenforcement 21 stress, even though that was changed.
22 Now, this part we do not understand, because it's not 23 supported from other studies, including BNL's work, which i
24 we have done in concrete analysis -- it shows it does --
25 shows that it does have an effect.
It is not negligible.
L
21069.0 8
CRT 1
So we didn't understand that.
2 Also, according to this -- the discussions we had with
(
3 Harry -- said that they were only able to obtain shear
' 4 failure when the concrete is assumed to have a low tensile 5
strength of about 120 psi, and then it gets a shear 6
failure close to 24 kips.
That's about a quarter of what 7
the strength normally is.
If it goes higher, it doesn't 8
get shear failure at all.
We can't understand that too 9
clearly.
10 Another thing which this report says is that -- it 11 states that they calculated large vertical displacements 12 before the diagonal tension failure occurred, and that's 13 another thing we don't understand, because our analysis
(
14 does not indicate this to be the case.
We find it to 15 happen suddenly and the displacements are not that 16 abnormal.
17 Another thing this table also indicates is that 18 regardless of whether you use the angle of limitation 19 criterion or not,. tha. difference in the answer is very 20 little.
Virtually nothing.
And this seems to indicate to 21 us that the angle criterion doesn't even come into play 22 ' when you are talking about the answer that they were 23 getting for this beam.
(
24 So, that brings us to the final point that we have on 25 this.
The figure in EBASCO's report which shows the crack i
l
_ ~, _.., _-,.. _._.. _ _,. _ _ _ _ _ _ _ _
21069.0 9
BRT 1
pattern -- and again the cracks are shown to be virtually 2
all orthogonal, and the reason for this is not clear to us,
.(
3 too.
f 4
These are the things we don't understand and our code i
1 5
gives us different answers.
From our experience, from the 6
data that that we have, we feel that we are correct.
We 7
don't know why EBASCO is getting this.
8 MR. BURWELL:
That's quite a laundry list, 9
however you want to take them.
10 MR. IOTTI:
Well, I don't want to go over every 11 point of the laundry list because these points are now 12 well known to us.
We have discussed them with Brookhaven.
13 The first thing we looked at is whether this could be a 14 problem with the code, in the terms of, given the theory 15.
is correct, there may be some coding input errors that we 16 would have to look at.
We found none.
And of cource that 17 puzzles us quite a bit, because if the code is doing the 18 thing right and the theory is corrcet, then why aren't we 19 getting the same answers?
20
-All we can conclude so far is that there is a mechanism 21 of possible failure which the EBASCO NASTRAN version does 22 not address.
23 In-plane, shear-type failures are simply not considered 24 in our code.
25 A lot of the conclusions presented in the last set of l
..._-..,me.,,..w----r, e rm-
-- --* +-~~ ~ ++
21069.0 10 CRT 1
example problems, for example, the relative lack of effect 2
of the shear interlocking character, were presented.for a 3
tensile strength of 546 psi.
Those conclusions may or may
. 4 not be the same if the tensile strength were lower.
5 Likewise, the angle limitation, that conclusion was 6
valid for 546 psi.
It may not be the same conclusion 7
where the tensile strength is lower.
8 We kind of looked at what the code limitation may be.
9 We ran a simply supported beam.
This is something that we 10 had promised'Brookhaven we would do, and we didn't have a 11 chance to have it completed before our latest submittal.
12 And on the simplest supported beam, we would suspect, as 13 the load increases at a certain point, at the location of f
14 the two supports, you would begin to see cracking under 15 essentially pure shear, and you would expect that crack to 16 be 45 degrees.
But the way the code predicts -- and that 17 code, by the way, the paper of -- what's the name --
18 Bazant and Kim -- would predict to start when your tensile 19 strength or your shear exceeds 120 psi.
20 Well, the way that our code handles that, the shear is 21 equal to the tensile stress.
And with a tensile stress of 22
.546, we simply aren't getting a crack.
And the reason 23 that we aren't getting the same result as Brookhaven as we i
24 lower the tensile strength is not, perhaps, because the 25 mechanism is the same, but because when you set the
T 21069.0 11 ERT 1
tensile strength equal to 120, you, in fact, are 2
predicting the diagonal shear failure at the ends.
Then 1
3 we would get the corresponding results.
4 We are puzzled at Brookhaven's results in the sense 5
that the explanation given to us is that their only 6
mechanism for failure is also the check against the 7
tensile strength of the concrete, which is what we are 8
doing.
9 This is not a question of computer code -- you can 10 simply run this calculation by hand for a simply supported 11 beam.
What is the shear at the end and what is the 12 tensile strength and they are going to be equal.
You can 13 just run the calculation by hand and you find it's 120 for 4
14 this particular behm.
So how are you going to fail if 15 your figure is 480?
16 MR. CHANG:
On the last item -- let's forget 17 about the code.
Whatever the code is saying.
Let's talk 18 about a simple' support beam.
19
'A simple support beata, a load to the center with 22 to 20 24 kips, the average stress on the section -- in the i
21 section is approximately, average stress is approximately 22 150 psi.
23 Now, taking the distribution of the shear -- it's a 24 parabolic shape, on the end of the beam, so it would go i
25 about 200 psi, 220 psi, something in that area: 200 to 220 J
.-.w-y y,, - -
,,,-,w.--,
i 21069.0 12 ERT l
1 psi.
2 Under the shear condition, the diagonal tension will be 3
equal to 220 psi.
If we put in the concrete strength of 500 psi, there is 4
5 just no way to crack this beam.
The coding is dding it 6
correctly.
And we compute -- we also show that at the end 7
beam section we have about 200 psi.
And I also show that 8
coming out, the tensile strength of 500 psi; it will not 9
crack.
10 From the ACI code or from the paper by Bazant and Jin-Keun 11 Kim, the cracking of the diagonal' tension to the shear 12 should initial at approximately 2 square root of FC pi.
13 This -- for this case it is approximately at 120 psi, 14 the tensile strength.
15 So, if we input 120 psi in tensile strength in our code, 16 we will get a failure ~ mode of from 22 to 24, as shown by 17 this experiment.
18 However, if we input 500 psi, tensile strength, we will 19 not get any crack due to shear, because the shear crack --
20 the shear stress around that area is only about 200, and 21 with the diagonal tension is only about 200, only.
22 And that, therefore, what we are really saying is this 23 way:
if BNL's code, using the same criteria we are using --
24 because they review our theory correctly, they don't find 25 anything wrong with our theory -- then I expect ENL also l
21069.0 13 CRT i
1 will show up a crack with a lower tensile strength as a 2
higher tensile strength, 500 psi.
(
3 Usually it's due to fracture failure of the beam -- the 4
fracture is bending structure of the beam it's usually 5
1/10 of the compression.
6 The initial cracking of the shear from all the test 7
results and from ACI code, initiation of the cracking, is 8
around 120 psi.
9 Therefore, what I said:
If we input 120 psi, 20 psi to 10 the code, the code would correctly predict the failure, 11 Therefore the difference --
12 MR. IOTTI:
Let me see if I can summarize this.
13 We are puzzled at this point.
We-know there's what we 14 perceive to be a limitation in the code that we are using 15 because we are only checking tensile strength caused by 16 flexure.
Maybe we should have a dual criterion and also 17 check for diagonal tension.
So that we see a potential 18 weakness in the code and we can address whether that has 19 any effect on it - that has been done.
20 What is puzzling is, if in fact BNL also has a failure 21 caused by flexure, how they can get results different from 22 ours.
That's what is puzzling.
i 23 MR. COSTANTINO:
When we first got together in 24 August, everyone agreed that failure prediction in 25 concrete is a difficult task.
The BNL code is really not l
i i
I
21069.0 14 ERT 1
the question.
I don't know if we should be --
2 MR. IOTTI:
I don't want to discuss it.
We are 3
puzzled why we are not getting the same answer.
. 4 MR. COSTANTINO:
I'm puzzled, listening to this 5
talk.
Everyone agrees that the checking does not agree 6
with the data on shear failure of the beams -- for 7
whatever reason.
Aside from the BNL calculation.
8 The conclusion from that would be that for the kinds of 9
problems to which NASTRAN -- the NASTRAN code was applied 10 in the main -- the steam generator compartment, you would 11 estimate, then, that the response predicted for combined 12 moment, shear, and axial force would probably also be 13 incorrect for predicting the amount and extent of crack.
14 MR. IOTTI:
It would underpredict the extent of 15 cracking and therefore overpredict --
16 MR. COSTANTINO:
Whatever.
It seems to me the 17 primary conclusion then is you would question the results 18 of the cracking predictions from the NASTRAN.
19 MR. IOTTI:
Of course.
We went back and looked, 20 because the next question is obviously how much difference 21 would it make?~ And then you have to flag the areas where 22 the differences can arise, how extensively --
~
23 MR. OOSTANTINO:
Then everyone would agree that 24 the results predicted by the NASTRAN for the cracking of 25 the compartment --
l
21069.0 15 BRT 1
MR. IOTTI:
For the high tensile strength, yes.
2 MR. COSTANTINO:
I want to make sure we all
(
3 agree with this.
~ 4 MR. CHANG:
I would --
5 MR. COSTANTINO:
Let me finish my statement 6
first.
I want to be sure we are all saying the same thing.
7 MR. IOTTI:
Absolutely.
The data conflicts with 8
what is predicted by our code at high tensile strength.
9 We have already gone back and found out how much of a 10 difference.
I think where we started straying is, because 11 we couldn't quite reconcile the differences between the 12 two codes, we were asking ourselves how can we go wrong?
13 We know what our limitations are.
We think we have found 14 out where our limitations are.
15 We are puzzled where why, if in fact the theory is 16 identical with Brookhaven, how could we come up with a 17 different answer?
So the conclusion is maybe the theory 18 isn't the same.
Okay?
19 MR. QQSTANTINO:
That.'s probably the case.
20 MR. SHARMA:
The only thing we mentioned in that 21 meeting is we agreed with the theory EBASCO is using.
We 22 didn't say we are using the same theory.
In fact --
23 MR. IOTTI:
I wish you had told me that, because i
24 I am going crazy.
If you have a slightly different 25 criterion it's not surprising.
7 21069.0 16 BRT 1
MR. SHARMA:
We have a dual failure criteria a
2 model --
3 MR. IOTTI:
Oh, you have a dual failure in; and g
4 that's why we differ.
That's the answer.
5 MR. SHARMA:
But you implied one is in tension 6
and one is in shear.
We have two criteria in the sense 7
that we check the strengths --
8 MR. REICH:
But this is irrelevant.
The code is 9
not under r'eview.
10 MR. SHARMA:
This is irrelevant.
Given that 11 beam, we know what its failure to shear is and we are not 12 using the ACI equation to get the failure because as Harry 13 Chang said right, that only gives the initiation.
We are i
14 using the other mode as stated before.
15 That is a question which gives the failure load -- not 16 the initiation of cracking, but the failure due to 17 diagonal tension.
And according to that equation, this 18 beam should fail in shear around 22 kips, assuming a i
19 compress strength of 4800 PSI.
20 MR. IOTTI:
What we are saying is that the 21 failure for diagonal tension would occur at the lower 22 value of diagonal tension than it does if you assume a 23 diagonal tension occurs at 480 or 500.
And we don't i
2d change that fall.
We still assume it's 500 and that's why 25 we are not predicting that failure.
i l
i
. _, - - - _. - - - -....,, -,,, ~, - -
21069.0 17 BRT 1
MR. SHARMA:
Now, if we assume a concrete of 2
4800 psi compress strength, then the most reasonable value 3
would be 1/10; not 1/40.
We disagree here.
We would 4
assume the tensile strength of this concrete to be around 5
480 psi.
6 MR. IOTTI:
For diagonal?
7 MR. SHARMA:
For anything.
Now Harry is saying 8
he has to assume 120 psi in value to predict that failure, 9
which in our opinion is too low.
10 MR. REICH:
One more thing, while we are still 11 answering you, Harry -- you mixed in the moment and shear 12 diagrams.
That only applies if there's no crack in 13 starting at all.
This beam will start cracking in 14 beforehand.
You can't tell us the moment of shear diagram 15 theory will stay the same once you have cracking going on.
16 The elastic theory isn't there anymore.
We told you this 17 before.
18 MR. SHARMA:
Another point, if I may, is I also 19 gave Harry some saferences in which people have predicted 20 shear strength of the beam using similar kinds of 21 formulations, and in those papers the tensile strength 22 might assume even more than 1/10, like 12 percent of the 23 compressive strength, and they were able to predict the
(
24 shear failure at the experimental load level, so it's not 25 that you have to assume 1/40 of the tensile strength.
The i
21069.0 18 ERT 1
value will be 10 percent and one should be able to predict --
2 MR. REICH:
And the other thing, Harry, why do 3
you get large displacement?
4 MR. CHANG:
The large displacement is because we 5
want to show you that observation is after it has failed.
6 We have shown that it is small if you just take starting 7
of the failure.
It is reasonable --
8 MR. COSTANTINO:
You get a major increase in 9
load capacity before you call it failure?
10 MR. CHANG:
No.
We don't.
At 24 kips, when you 11 increase the 24 kips, first you start a diversion solution.
12 A diverging solution in that spacing is small.
If there's 13 something is wrong -- you would t,ake the next shot to try I
14 to converge again.
The next shot -- the stiffness is 15 almost failed already, it is very small.
Then you are 16 putting reiteration issue of the large displacement.
17 MR. COSTANTINO:
That's the wrong answer.
The 18 shear failure of a beam is a brittle structure-type 19 failure.
20 MR. CHANG:
Yes.
21 MR. COSTANTINO:
So collapse of the beam starts 22. at its normal displacements; displacements of the order of 23 less than 1/10.
24 MR. CHANG:
It has collapsed.
25 MR. COSTANTINO:
If the code requires a large
e-
-, ~, - - - - -,.
21069.0 19 CRT 1
displacement to enforce collapse --
2 MR. CHANG:
No, it does not.
(
3 MR. BURWELL:
One at a time, please.
4 MR..COSTANTINO:
I mean, Harry, what you are 5
really saying is under the initial assumed stiffness 6
coefficient you get small displacement which don't satisfy 7
equilibrium, so you need a zero stiffness correction which 8
gives you large displacements.
So the conclusion is the 9
code is not predicting experimental behavior.
I think you 10 agree with that.
11 I don't think there's any point in discussing too much 12 else beyond that, except that everyone agrees that the 13 code is not predicting concrete failure behavior properly.
14 That's the overall assessment.
That's all everybody is 15 trying to say.
16 The implication of that on the -- on the generic stress 17 analysis is something else.
18 MR. CHANG:
Well, one more thing I would like to 19 say isk this way:
The code is doing correct what we want 20 to do, according to the theory.
We may not have a second 21 theory -- second -- I don't know that area.
But if we, 22 for this particular application is concerned, if we lower 23 the tensile strength to zero -- okay?
-- that would be i
2d predicted at much lower failure criteria, failure loads.
25 For this particular upper stress concern, we actually did
21069.0 20 ERT 1
run certain runs with zero tensile strengths, certain runs 2
with very high tensile strengths --
3 MR. COSTANTINO:
We are talking about detail.
g
- 4 MR. IOTTI:
We are talking about a different 5
issue.
6 We concur that there is a limitation in the code --
7
.MR. COSTANTINO:
When it comes to cracking in 8
concrete --
9 MR. IOTTI:
Right.
At the higher loads.
Right.
10 And there's a point at which the two solutions obviously 11 depart.
I don't know why that point is; we tried to 12 assess it.
13 MR. COSTANTINO:
The cause of that is something 14 else.
15 MR. IOTTI:
We can't find it.
16 MR. COSTANTINO:
If there's no programming error 17 there's an error in formulation.
If not, then there's --
18 but obviously there's a problem someplace.
19 MR. IOTTI:
We are trying to learn also.
Given 20 that that's the case, first of all does any of the 21 analysis that has been done for the upper lateral 22 restraints involve the type of loads that cause this vast 23 change in the behavior?
That's the question we asked 24
'ourselves and we went back and reviewed it and we can 25 assess it later.
1 k
r--
21069.0 21 ERT 1
Give me one minute, Frank, and then I'll turn it over 2
to you.
(
3 The other part is that, since we don't -- well, we 4
think we've found what the limitation of the code is.
So, 5
obviously if we ever have to go back and do anything 6
further, we will address it in that fashion.
And at this 7
point, I guess, Frank, if we want to get into the 8
discussion of the upper lateral and the analysis and the 9
consequences of the limitation of the code we can do so.
10 If not -- that's up to you.
11 MR. RINALDI:
Basically, the way I see it, you 12 use this code to predict the value at this upper lateral 13 restraint beam.
Now we have a question about this code, 14 so apparently we should use another approach to evaluate 15 it.
16 MR. IOTTI:
We used the code with zero tensile 17 strength to assess the values on the concrete.
Okay?
We 18 used the higher strength to assess how much of a load we 1%
will get on the upper lateral.
There is a dual approach 20 to this.
For the concrete we use the zero tensile 21 strength, if you look at the affidavit.
And when you look 22 at the upper lateral in terms of maximizing the effect on 23 the upper lateral, the bounding on the upper lateral so
(
24 you maximize the stresses on the steel itself, we use this 25 high tensile strength concrete.
Let's not forget that.
I l
)
-c..
-e.
,,.., _. _ ~,., < -
21069.0 22 CRT 1
think that's the point that Harry wanted to make.
2 So, clearly, yes.
We may be overestimating the
(
3 resistance of that wall.
The point is, by overestimating
' 4 all you have done is created a worse situation for 5
yourself in terms of stresses on the steel beam.
6 Obviously nothing worse can happen to the concrete than 7
the concrete having zero tensile strength.
So, if you go 8
back to the affidavit -- I'll read it to you, because 9
there is a tendency after you do all of this work to 10 forget why some of the work was done.
11 On page 11, first we address the impact on the concrete 12 structure; okay?
And the test for concrete structure was 13 conductsd for zero tensile concrete strength and then 14 later on it was for 450 psi.
15 MR. COSTANTINO:
Where are you reading, Bob?
16 MR. IOTTI:
Page 11 The actual affidavit 17 itself.
18 MR. COSTANTINO:
Oh, the May 20th document?
19 MR. IDTE :
Yes.
But I,guass ul.timately we 20 would concur, there is a question about the validity of 21 the 450 psi, or not.
22 MR. COSTANTINO:
Are you saying if you assume 23 zero you get a very conservative or very soft response to
?
24 the concrete and 450, a very hard response?
25 MR. IOTTI:
The concrete at 450 will certainly
21069.0 23 ERT 4
1 provide a much larger resistance to expansion --
2 MR. COSTANTINO:
As far as the evaluation of the
(
3 upper lateral restraint beam, I'm not sure that that's the 4
proper question.
It seems to me the question is:
Let's 5
assume the NASTRAN output is wrong.
In all cases.
Can 6
you conclude that the 450 calculation is c'onservative?
If 7
all you know is the NASTRAN calculation is wrong?
As far 8
as had the upper lateral --
9 MR. IOTTI:
From the standpoint of the upper 10 lateral itself.
11 MR. COSTANTINO:
The steel beam.
12 MR. IOTTI:
Yes.
13 MR. COSTANTINO:-
It's an opinion.
It seems to 14 me that some justification in terms of simplified I
calculations is needed.
16 MR. ICTTI:
We looked back to see what elements i
17 would have load,"where the particular application of the 18 code would affect it -- it turns out to be very few of the L9 alamanta naam where the upper lateral impediment is.
If 20 they crack as opposed to not crack as they are, what we 21 would get is additional relief.
Okay?
So it's an 1
22 educated opinion that there is lower cracking; okay?
(,
23 Now the next question is whether the zero tensile 24 strength represents the worst case for the concrete.
It 25 should be close.
It may not be the worst case absolutely i
I
\\
21069.0 24 ERT 1
because you also get expansion of the upper beam, but what 2
we found just by comparison, the extent of cracking for 3
zero tensile strength was practically everywhere when we 4
went up to 450 -- we expect the zero tensile strength 5
represents the worst conditions for concrete for the 6
cracking.
7 Now, we did go back and check everywhere in the 8
elements where the diagonal tensile strength could have 9
been a factor in calculating the initiation of cracking.
10 And we found it was six or seven elements in that region 11 which --
12 MR. COSTANTINO:
Should have been predicted --
13
. MR. IOTTI:
Should have been put as cracking in 14 accordance with the test data.
Okay?
Now we don't know 15 what cracking those elements would have done with the 16 surrounding elements without running it, but we would 17 expect you'd have, if anything, more cracking; not less.
18 To that extent, we believe the results, however 1.9 incorrect in a sanse o.f predicting the failuza or the 20 cracking in the concre.te, are conservative for the 21 analysis.
22 Understand that the off-shear cracking, that was 23 checked manually.
So that part was addressed properly, 24 per ACI.
25 MR. COSTANTINO:
Are you going to submit a piece 9
. =,.
~, -. _
.. ~.. ~.
_~
l 21069.0 25 CRT i
1 of paper on that or have you already done that?
2 MR. IOTTI:
We have already done that.
We gave
(
3 you the manual calculation, off-shear.
~
4 What we don't know is what else to do at this point to 5
correct Whatever deficiencies there are in the code.
In 6
other words, we are a little puzzled about how we are 7
going to go about correcting it.
I would like to separate, 8
perhaps if it's possible, the upper lateral restraint 9
analysis from Whether the code has limitations and Whether 10 the code should be corrected.
Okay?
11 As far as we can determine, the limitation of the code 12 actually imposed, if anything, more of a problem for the 13 upper lateral restraint than otherwise.
So Whatever i
14 results we have would have been conservative, as opposed 15 to not conservative.
16 Call it an opinion, if you will, but it is an educated 17 opinion.
j 18 MR. KUO:
See if I can summarize what I L
1.9 understand from Mr. Iotti.
I believe everyone here agrees 20 that the code itself may or may not correctly predict the 21 cracks in the concrete.
22 MR. IOTTI:
I believe we agree it does not, at 23 high levels.
What'is a fact is a fact; okay?
l l
24 MR. KUO:
Okay.
And that you tried two ways to 25 approach the problem:
One, by using zero tension in l
.r
21069.0 26 CRT 1
concrete to look at the concrete; and then another, assume 2
450 psi in the concrete, look at the upper lateral y
3 restraint.
And supposedly this approach is conservative.
~
4 And later on you, I guess you were asked whether this 5
is an educated opinion or you have something to support it.
6 MR. IOTTI:
We think we have something to t
7 support it.
J 8
MR. KUO:
Can you make that available to us?
9 MR. IOTTI:
Yes, we can basically.
The list of 10 all of the elements and what the diagonal -- what the 11 shear stress is in all those elements -- the in-plane i
12 shear stress, which is, after all, what is in question.
13 And wherever that shear stress is in excess of 120 or 150, 14 we' flag that element as being a possible candidate for 15 crack initiation.
Whereas right now there's no crack 16 prediction.
17 You can find the location of the zone and you kind of 18 mentally assess for yourself what would have happened if 19 those elements were also cracked.
20 Again, it's not a question --
)
21 MR. COSTANTINO:
What do you mean by "in-plain 22 ' shear?"
I thought we were talking about normal shear, 23 bending and normal shear?
24 MR. IOTTI:
This is a two-dimensional element.
25 The shear through the thickness is assessed manually.
21069.0 27 CRT 1
This is a shear through the plane.
2 MR. COSTANTINO:
Okay.
(
3 MR. IOTTI:
And we looked at 5 elements, really,
' 4 is all we saw so far.
There might be 6.
It is not that 5
many.
It's not that pervasive, in terms of the analysis.
6 And this is the worst case.
7 MR. CHANG:
Yes.
That's the worst.
8 MR. KUO:
Now I can continue?
By that approach, 9
if you were able to show that the lateral -- upper lateral 10 support beam is okay, then we have resolved the problem 11 for this lateral support beam.
But then, by using the 12 other approach, zero tension, you probably would show a 13 lot of cracking in the concrete.
14 My question is that:
Is there any impact of extensive 15 cracking in the concrete?
16 MR. IOTTI:
I don't know.
I guess I'm a little 17 at a loss to answer your question directly, as what is the 18 effect of having extensive cracking in the concrete.
19 Carta. inly, when we ran the case of the tensile strength 20 equal to zero, we found that everywhere the rebar stresses 21 were low as compared to the allowable maximum was 22 kai 22 ' (sic) versus 24; the compressive strength in the concrete 23 would also be well within the allowable, like a third of a
24 the allowable.
The strain in the rebar was something in 25 the order of 20 percent of the allowable use strain -- not
21069.0 28 BRT 1
the allowable use strain -- the use strain.
So, yes, we 2
have cracking in the concrete.
But there are no adverse
(
3 consequences from that cracking.
4 MR. KUO:
That's basically what I'm asking.
5 MR. IOTTI:
Right.
6 MR. KUO:
If you can make an assessment --
7 MR. IOTTI:
That is in table 1 of the affidavit.
8 There is an assessment of the rebar stresses, compress 9
stresses in the concrete, straining the rebar as a function'of the tensile strength assumed; for zero psi you 10 11 have all the values.
We don't think there is a problem.
12 I don't know whether you remember this or not -- the first 13 three columns of table 1 give you the results for the f
14 tensile strength equal to zero pai.
15 MR. KUO:
Right.
These are all numbers.
I 16 think what I'm asking is that, if you can provide us an 17 assessment, in words, what the effect of this concrete 18 cracking would be, in terms of the structure.
19 MR. IOTTI:
Oh.
None.
None.
I don't know what 20 words I can give you other than "there is none."
21 MR. KUO:
But why?
I think we have to have a 22 little basis.
23 MR. IOTTI:
Okay.
There are two things that 24 thosa walls have to fulfill.
25 They are not -- these are interior structures of the r
21069.0 29 ERT 1
containment.
There is no requirement for them to be leak 2
tight.
They have to be able to carry load.
Okay?
And
(
3 they have to be able to distribute the load due to post --
4 call it post-event, if you will, conditions initially the 5
codes that these structures will seal, which will cause 6
the extensive cracking, will be the mechanical loads.
At 7
the time the mechanical loads are imposed on the 8
structures, the concrete has not really cracked very much.
9 Even if it did it would still carry the mechanical load.
10 Bear in mind the temperature load occurs later.
11 After the temperature load has occurred there is 1
12 virtually no loading left for these structures to accept 13 other t,han possibly a seismic event.
14 They have assessed the capabilities of the structures 15 to accept the coincident seismic event together with the 16 mechanical and thermal load.
17 If we were not to input the seismic event 18 simultaneously but' input it later; number one, we would 19 find there would, be lass cracking to start with and; 20 number 2, certainly if we accept all three of them 21 simultaneously we can accept one of them separately by 22 itself.
23 So, as far as I'm concerned, the structures fulfill 24 every single functional requirement expected of them.
And 25 there is none other.
Those are the words that I would use i
l
4 21069.0 30 i
CRT j
l 1
to justify there is no adverse consequence of the cracking 2
in the concrete.
I 3
It's not a containment.
It's not that I have to
' 4 maintain leak tightness in these structures.
All I have 5
to do is maintain structural integrity and the ability to 6
accept loads coming in from environmental and other places.
7 MR. COSTANTINO:
The case really has to be more 8
that the seismic is small?
9 M.. IOTTI:
I don't recall any.
Is the seismic 10 really that small?
11 MR. COSTANTINO:
Do you remember -- that was the 12 August -- the numbers were very small.
13 MR. IOTTI:
I just didn't want to say anything 14 because I don't remember.
15 MR. COSTANTINO:
Okay.
If you are going to 16 write something down about the seismic cases, they are 17 small rather than --
18 MR. IOTTI:
Right.
I recollect that the case L9 was made where the seismic loads per so were samall.
20 MR. COSTANTINO:
Because the seismic analysis 21 was so incorrect, because of improper boundary condition.
22 MR.. IOTTI:
Was it boundary or asymmetry?
i.
23 MR. COSTANTINO:
Due to the asymmetry, that was 24 assumed.
That last thing should be because the seismic 25 analysis was incorrectly performed.
k
,,m-
21069.0 31 ERT 1
MR. BURWELL:
Let's go off the reccrd a minute.
2 MR. IOTTI:
I just want to qualify if we can
(
3 stay on the record for a minute.
The seismic analysis
~ 4 that Professor Costantino is referring to wau the seismic 5
analysis of the structures themselves because there was 6
another of the primary system, whose loads come back into 7
the upper lateral, that one is not in question.
Or at 8
least it wasn't discussed.
It was assumed that it was 9
done right.
~
10 MR. BURWELL:
Let's go off the record a minute.
11 (Discussion off the record.)
12 MR. BURWELL:
On the record.
13 MR. KUO As we said before, we al!/ r'ecognize 14 there is some limitation in your program,and you have 15 provided us some verbal justification a while ago, why you 16 think the approaches that you have taken are conservative >.
17 We basically agree with your concluhion, based'on our 18 own independent check.
We still are not quite sure what
~
19 conservatism is inherent in your calculation.
20 We would still have a few questions to clarify with you.
21 But before we do that, I would like to make a request that
'you provide a written justification for your approach --
22 23 for the approach you have taken, both for the upper 24 lateral restraint beam and for the concrete support.
We s
25 don't expect any new calculations.
\\
^
-(
5 n
i i
4
-y
,-c.
,-#_r,
,s
~
21069.0 32 BRT 1
MR. IOTTI:
Just words.
2 MR. KUO:
Just words.
Put together what you I
3 have.
With that, I would like to turn, more or less to 4
clarify a few more details in your calculations.
5 MR. BURWELL:
Do you understand what he's asking 6
for when he asks for a verbal justification?
7 MR. IOTTI:
My understanding is that applicants 8
have performed a set of calculations, some of which can be 9
judged to be suspect because of the limitation in the 10 computer codes.
So he needs words that places the 11 limitations of the code in terms of result predicted by 12 the code in a proper perspective as to why, even though 13 the prediction may not be correct, it would be 14 conservative.
And he wants those words to reflect 15 conservatism for the prediction of the upper lateral 16 restraint itself, stresses; and he would like similar 17 words insofar as the concrete.
So I would have to discuss 18 both the calculation performed for zero tensile strength 19 and the 450, without doing any additional cariculations.
20 Just place it in perspective.
21 Is that a fair statement?
22 MR. KUO:
Yes, I think so.
23 MR. REICH:
Let me take you to a couple of s
24 questions on your write-up.
This pertains to your 25 write-up dated August 21st.
i i
-~
21069.0 33 BRT 1
MR. IOTTI:
This letter here, Gibbs and Hill, 2
August 21, '84; submittal from Gibbs and Hill to --
(
3 actually it's really a --
4 MR. KUO:
It's really a letter from Mr. Ballard 5
of Gibbs and Hill to Mr. Joe B. George, of Texas Utilities, 6
with copies to the Commission.
7 MR. BURWELL:
The letter is identified by GTN-69363.
8 MR. REICH:
I take you to sheet number 55.
9 There is where you take a very conservative case a'nd you 10 say that the stresses in the beam assumed fully 11 constrained at each end.
This is for the upper lateral 12 restraint; okay?
13 Now, in this case here you did this calculation.with a 14 delta T 282 minus 70 -- 212; right?
15.
MR. IOTTI:
Right.
16 MR. REICH:
Now, there is yet another case.
If 17 you look at your letter, page 3,. the attachment to the 18 letter, " minutes of the meeting."
And there is page 3 on 19 top, you have the main temperature -
20 MR. IOTTI:
It is higher.
21 MR. REICH:
It says over there 370.
I want to 22 ask you two questions about that, actually.
23 I have seen different numbers.
In some places you had 24 340, which you said you calculated.
Some places I have 25 seen a number like 352, being used.
21069.0 34 BRT 1
MR. KENKRE:
355?
2 MR. REICH:
355.
I'm sorry.
355.
(
3 What is the real number for that main steam case r okay?
4 And why did you cnly do this very conservative case?
Why 5
didn't you show the other case, too?
That's one question.
i 6
MR. KENKRE:
This calculation was done only 7
because it was requested.
Okay?
8 MR. RINALDI:
We were trying to get to the 9
limiting case, I guess.
10 MR. IOTTI:
There may be a misunderstanding.
11 The reason this calculation was done is because, if you go 12 back to the records of the hearings, you will find that 13 this is the calculation that is always appearing because 14 the issue was always the local case.
We issued the main 15 steam later.
16 You have to understand, if you assume this you can have
~ 17 problems.
Wie have always maintained that's a bad thing to 18 assume, the wall is infinitely quick, to begin with.
19 We essentially reproduced what had been discussed at 20 hearings; okay?
The main steam line does give you higher 21 temperature.
22 MR. REICH:
And would give you a higher force?
23 MR. IOTTI:
Certainly.
24 MR. REICH:
What should the number be for the i
25 main steam line, 255 --
I l
.. ~
21069.0 35 ERT 1
MR. lOTTI:
355.
2 MR. REICH:
355?
34'0?
In terms of temperature,
(
3 or should it be 3707 4
MR. IOTTI:
370 is not the correct number, I can 5
tell you that much.
It was a conservative -- 355, I think, 6
was the calculated --
7 MR. KENKRE:
71, 340 for the upper and 350 for 8
the lower --
9 MR. KUO:
355 was referred to.
What is the real 10 one?
11 MR. IOTTI:
That's the one reported on the 12 affidavit, I believe.
13 MR. REICH:
Is it 370 you have there?
14 MR. IOTTI:
The temperature on page 10 of the 15 affidavit would state the temperature in both upper and 16 lower lateral supports rises for approximately 355 17 fahrenheit at about --
18 MR. REICH:
It should be 355.
19 MR. IOTTI:
Approximately 355; right.
20 MR. COSTANTINO:
Then that letter should be 2
21 changed; right?
Bob?
Or whoever?
On page 3 of the 22 letter from Gibbs and Hill that we are talking about, it 23 implies that the beam is designed for 370.
At the top of 24 the page.
25 MP.. IOTTI:
I'm not so-sure that it is implied.
l
21069.0 36 BRT 1
I think what needs to be changed is the word "as desi gned."
2 It's really "as analyzed."
Because we used 370 in our
(
3 analysis.
That beam was never designed for any 4
temperature in terms of the beam itself.
5 So we need to change is the word there to "as designed."
6 MR. COSTANTINO:
What do you mean "as analyzed"?
7 In the NASTRAN computer run?
8 MR. IOTTI:
Yes.
9 MR. KUO:
What's the real number we should use?
10 MR. REICH:
He said 355.
11 MR. KUO:. I didn't hear it.
12 MR. IOTTI:
My best recollection is when I look 13 at the output of the heating 5 code and the rehab that 14 - generated the input, it was about 355.
15 MR. KUO:
So we could go ahead and use that 16 number?
17 MR. IOTTI:
Right.
I don't know why this is 340 18 and 350.
19 MIL. REICE:.
Let ma ask you this, then:
When you 20 later on -- let's say, for instance, on page 60, okay?
21 When you deal with the main steam break case, 1-9, and you 22 get these values, let's say F sub BX; and F sub BY, in the 23 beam itself you are getting some sort of bending; am I 24 right?
l 25 MR. KENKRE:
Yes.
w w
- ,--w
,m' r
mp
-v w
"es---
I-->--+v-w f-'--
~*e
- " +
-mi e-~~~*-
-.=+ -
21069.0 37 ERT 1
MR. REICH:
Now, those values there come because 2
of the computer code.
Those forces, actually -- forces
(
3 that go into bringing you that BX and BY, result because 4
you are using the computer code with zero?
Is that with 5
zero shear?
Zero tension; I'm sorry.
6 I'm just asking if the two values of F sub BX and F sub 7
BY; need number 607 8
MR. KENKRE:
Those are really coming from the 9
fixity of the beam on the wall.
10 MR. REICH:
Yes, but they are coming -- at what 11 condition did you get this?
The fixity is due to some 12 sort of force interaction that you are getting between the 13 beam and the wall and it comes from the computer output.
14-Now, in that computer output you assume zero tensile 15 strength?
or am I wrong?
I'm just asking you.
16 MR. RINALDI:
I guess the question is:
What do 17 you assume?
18 MR. REICH:
Yes.
What do you assume?
Let me 19 not give you the words.
20 MR. KENKRE:
Run 9.
I'm not sure whether --
21 MR. COSTANTINO:
We are not asking for those 22 answers now.
We are just bringing up --
23 MR. REICH:
No.
Maybe just add in the letter "what 24 do you use"; that's all.
25 MR. IOTTI:
I can speculate and we can give you 1
l l
21069.0 38 BRT 1
the correct response, but I think it will be 450.
2 MR. BURWELL:
Particularly in case 2.
(
3 MR. IOTTI:
That's correct, because we are 4
talking about maximum compression and that occurs at 5
maximum compression at the wall.
6 MR. RINALDI:
So that would be the worst case in 7
predicting --
8 MR. IOTTI:
For the upper lateral.
9 MR. KENKRE:
For the maximum compression.
10 MR. COSTANTINO:
I have some more questions.
4 11 MR. IOTTI:
Wait a minute.
I have to make a 12 note.
So you need an explanation as to what case sheet 60 13 corresponds to; and confirm max temperature of main steam
. 14 line break; B, is what case page 60 concerns.
And you 15 want all of this as part of a written reply.
16 MR. COSTANTINO:
I have some more little items.
17 MR. IOTTI:
Sure.
I just want to make sure I 18 don't miss any.
19 MR. COSTANTINO:
On page 4 of the letter, if you 20 look at the middle of the page it says, "to support the 21 position that the assumed direction boundary conditions 22 'are adequate - " that sentence it seems to me to need 23 change, the s direction boundary conditions are incorrect.
I 24 The boundary conditions are not adequate.
25 MR. IOTTI:
I don't want to quibble over p._
,..cy,,
.,w..--
.,y
.cs
-9
,__-,.,.e
21069.0 39 CRT 1
semantics.
2 MR. COSTANTINO:
I don't know if that's so much
(
3 semantics.
Boundary conditions are either right or wrong.
~
4 MR. IOTTI:
I understand.
But when you do 5
modeling, you know very well on occasion you make 6
assumptions that may seem to be incorrect and you know 7
that they are not, you know, the proper conditions, but 8
you think they are adequate for the purpose that you want.
9 MR. COSTANTINO:
Well, I agree with you.
All 10 I'm saying is the reason why we look at calculations is 11 because everybody agreed the boundary conditions were 12 incorrect for the seismic problem.
Okay?
So it just 13 seemed that should be -- I guess we agree to that.
.. 14
-And on page 5 of the letter, now --
15 MR. IOTTI:
Do you want that in writing?
16 MR. COSTANTINO:
Something has to modify that 17 statement, it seems to me.
Okay?
18 MR. IOTTI:
Okay.
19 MR. COSTANTINO:
On page 5, subparagraph C, 20
" calculations were developed and reviewed based on the 21 assumptions that the beam - " page 5, subparagraph C, 22 which applies to the calculation of upper bound estimates 23 of thermal stresses in the beam.
24 MR. IOTTI:
Right.
We don't look at the high 25 case.
21069.0 40 BRT 1
MR. COSTANTINO:
So that paragraph has to be 2
modified, giving an upper bound.
(
3 MR. IOTTI:
C on page 5, right?
4 MR. COSTANTINO:
C on page 5.
Right.
5 MR. SHARMA:
I have a question.
When the 6
concrete -- these walls are fixed at the bottom?
7 MR. IOTTI:
I believe that's correct.
Yes.
8 MR. SHARMA:
Now, what of the shear if you have 9
zero strength concrete and you have shear developing at J
10 the bottom?
What will happen to the shear force?
11 MR. IOTTI:
We'll have to go back to the 12 calculation.
There is a shear, obviously.
I 13 MR. CHANG:
The concrete stress is assumed to be
-- -zero but not necessarily assumed the concrete and steel 15 together still can hold up the structure.
In the design 4
16 of concrete structure we always assume the tensile 17 strength of the concrete structure is completely zero.
18 That's the usual assumption.
19 MR. SHARMA:
That's right.
20 MR. CHANG:
So, when we design with zero tensile 21 strength, that steel can hold up the structure because the i
22 steel, combined action with the concrete and the portion 23 of the shear Which is not cracked -- portion of the shear 24 resistance which is not cracked.
25 MR.'IOTTI:
Let me make sure I understand your
,,,~-,.,.,~ene
---,,,---e----
,,,,-,,,,e---v----
-,,---wm--
rn--w-e-~r-s,m.-n--,--,,o--y--~~
~
~v
, - -, -- - -, ~ -, - -
21069.0 4
41 BRT e
l 1
question.
You want to know what the shear is at the l
]
2 bottom?
(
3 MR. SHAnMA:
The steel will then provide the
' 4 membrane resistance.
Now, if you have shear at the bottom 5
and yon assume the strength of the concrete to be zero --
6 MR. KENKRE:
The tensile strength of the l
7 concrete is zero.
But there's shear strength because of 8
interlocking of everything.
9 MR. SHARMA:
If you put tensile strength to be 10 zero then I assume --
11 MR. IOTTI:
That 's the off-plane.
That shear 12 will be exhibited as an off-plane shear through the i
13 thickness.
14 MR. SHARMA:
What shear-do you keep?
Do you 15 have any idea?
r 16 MR. IOTTI:
.2, I think.
17 MR. CHANG:
That shear -- we took a bend -- we i
18 talk of bending, that shear would be consistent to the
{
19 bending moment.
j 20 MR. SHARMA:
Okay.
In my opinion, shear 21 strength and tensile strength are related.
All right?
If 22 you assume shear strength to be in one plane, then 23 obviously if you take a component, it will have a tensile t
24 strength at 45 degrees.
So what does it mean to have a 25 zero tonsile strength concrete?
[
i l
i
[
l
21069.0 42 BRT 1
MR. CHANG:
Zero tensile strength concrete does 2
not mean the concrete cannot have resistance, okay?
i 3
It doesn't even address the shear in that direction.
^
4 The shear is computed inconsistent of the bending moment 5
of the wall.
There's an assumption there --
6 MR. IOTTI:
I think you lost them.
7 MR. COSTANTINO:
Am I allowed to use the board?
8 I think all Sushil is saying, Harry, if you look at the 9
failure it's something like this with some kind of a cap 10 someplace.
And if this is the tensile capacity, then you 11 have some kind of a shear capacity.
If you are saying you 12 have zero, at zero you have zero shear strength.
Is that 13 right, Harry?
(
14-- -
MR. CHANG:
Zero strength.
15 MR. COSTANTINO:
That's all Sushil is saying, he
~
16 questions that part.
17 MR. SHARMA:
I need some clarification.
18 MR. IOTTI:
What you want to know is what is the L9 shaar st,the. bottom of the model.
We'11 provide you thet.
20 There is a number.
21 MR. RINALDI:
One more item on this response.
22 I'm assuming --
23 MR. IOTTI:
Excuse me a second, Frank.
Let me 24 write this down.
Go ahead.
25 MR. RINALDI:
We are assuming that this modified
21069.0 43 ERT 1
NASTRAN code was only used for the upper lateral beam 2
restraint evaluation.
t 3
MR. IOTTI:
That's correct.
4 MR. RINALDI:
If, in the reply you could restate 5
that?
Maybe you have done it before, but I think we would 6
like to have it restated.
7 MR. IOTTI:
Okay.
8 MR. COSTANTINO:
One point of clarification.
9 The linear seismic run, the SSE run that we looked at, was 10 that a SAP or a STRUDL?
Was that SAP?-
11 MR. IOTTI:
I think it was STARDYNE.
12 MR. KUO:
What kind of schedule are we talking 13 about?
4 14---
MR.- IOTTI:
What is today, Tuesday?
I could try 15 and have something in the mail Monday.
16 MR. BURWELL:
Fine.
In the meantime, I hope you 17 gentlemen know what you will say, whether or not to 18 proceed t.o put pencil to paper and give us some results 19 soon -- get this work in reasonably soon.
20 Okay.
I thank you gentlemen.
I think that was a 21 useful meeting and I think we have come together.
22 (Whereupon, at 11:55 a.m., the conference was 23 concluded.)
24 25
- e., ---,,
- ~., - - -, -, - - -
CERTIFICATE OF OFFICIAL REPORTER This is to certify that the attached proceedings before the UNITED STATES NUCLEAR REGULATORY COMMISSION in the
(
matter of:
NAME OF PROCEEDING:
TEXAS UTILITIES GENERATING COMPANY, et al.
(Comanche Peak Steam Electric Station, Units 1 and 2)
DOCKET NO.:
50-445-OL2; 50-446-OL2 PLACE:
BETHESDA, MARYLAND DATE:
TUESDAY, NOVEMBER 13, 1984 were held as herein appears, and that this is the original transcript thereof for the file of the United States Nuclear Regulatory Commission.
(sigt (TYPED)O
/
JOEL BREITNER Official Reporter ACE-FEDERAL REPORTERS, INC.
Reporter's Affiliation
[
.