ML20246M051
| ML20246M051 | |
| Person / Time | |
|---|---|
| Issue date: | 03/16/1989 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | |
| References | |
| ACRS-T-1722, NUDOCS 8903270031 | |
| Download: ML20246M051 (457) | |
Text
-
L .. . Acts 7~/7 4 OR G \A_
C UNITED STATES NUCLEAR REGULATORY COMMISSION
')
.. a 2-------------------===========----------------------g ADVISORY COMMITTEE ON REACTOR SAFEGUARDS In the Matter of: )
)
-)
SUBCOMMITTEE ON MATERIALS AND )
METALLURGY )
O 1
I Pages: 205 through 417 Place: Bethesda, Maryland ee n -
Date: March 16, 19 N r 6 cAOS v 7!!c'e u$o3yT } ram -
...-------...........-..-..s...,..-.- .:0 f 7 " g Q " F e 3d DLD.iH M
~ e---c HERITAGE REPORTING CORPORATION O on e 1220 L Street, N.W., Suke 600 '
8903270031 890316 Washington, D.C. 20005 h$72 pdc (202) 6M .,
J
/
J' 1 PUBLIC NOTICE BY THE j
) 2 UNITED STATES NUCLEAR REGULATORY COMMISSION'S 3 ADVISORY COMMITTEE ON REACTOR SAFEGUARDS 4 March 16, 1989 5
6 7 The contents of this stenographic transcript of the 8 proceedings of the United States Nuclear Regulatory 9 Commission's Advisory Committee on Reactor Safeguards (ACRS),
1 10 as reported herein, is an uncorrected record of the discussions 11 recorded at the meeting held on the above date.
12 No member of the ACRS Staff and no participant at 13 this meeting accepts any responsibility for errors or 14 inaccuracies of statements or data contained in this
/N 15 transcript.
L]
16 17 18 19 20 21 22 23 24 4C Heritage Reporting Corporation (202)628-4888 A
V
205 UNITED STATES NUCLEAR REGULATORY ~ COMMISSION ADVISORY COMMITTEE /A REACTOR' SAFEGUARDS
, In the Matter of: )
L )
)
SUBCOMMITTEE ON MATERIALS AND )
METALLURGY )
Thursday, March 16, 1989 Battelle Building II 505 King Avenue Conference Room G Columbus, Ohio The meeting convened, pursuant to notice, at 8:30 a.m.
BEFORE: PAUL G. SHEWMON Chairman, Metal Components Subcommittee Professor, Metallurgical Engineering Department Ohio State University Columbus, Ohio SUBCOMMITTEE MEMBER PRESENT:
MR. CARLYLE MICHELSON Retired Principal Nuclear Engineer Tennessee Valley Authority Knoxville, Tennessee, and Retired Director, Office for Analysis and Evaluation of Operational Data U.S. Nuclear Regulatory Commission Washington, D.C.
CONSULTANT:
J. HUTCHINSON ACRS COGNIZANT STAFF MEMBER:
AL IGNE Heritage Reporting Corporation (202) 628-4888 O
4 I
.206 NRC STAFF PRESENT: .
- MR. ARLOTTO MR. LEE l REPRESENTATIVES OF BATTELLE, INC.: d i
DR. SCHMIDT ,
1
.DR. WILKOWSKI MR. SCOTT l'
f' l
Heritage . Reporting Corporation (202) 628-4888 t
i _ - _ - - - - - _ - - - _ - - _ _ _ _ _ _ .
1 I- '
207.
1- E B Q C E E'R I E E a
() 2 DR. SHEWMON: Good morning. This is a 3 continuation of the Metal' Component Subcommittee of the
'4 ACRS. This morning we hear about the degraded piping
, 5 program which has been going on here for the last several 6 years and must be pretty close to completion since I have a l
7 draft of what is called a final report here.
8 . Gary, do you want to begin.
9 (Slides being shown.)
10 DR. WILKOWSKI: Right.
11 The degraded pipirg program was completed last 12 January and the final report is in the process of being 13 published'by the NRC. I have today slides as opposed to 14 overheads.
15 And what we will be showing you today -- I'll skip
(}
16 right on through there -- I will give a general introduction 17 into the program. Afterwards Charles Marschall, Dr.
18 Marschall will talk about the material characterization i 19 efforts. Paul Scott will describe the full scale pipe 20 fracture experimental work and some limit load analyses 21 associated with those expeftiments.
22 Somewhere around that time if you see fit then 23 maybe we can take a break there. Then we can get into the 24 elastic plastic analysis efforts by Jalees Ahmad. And then 25 afterwards I will talk about the significance of the results Heritage Reporting Corporation (202) 628-4888 ,
_ - _ __ A
f .
l l
208 6 1 and what it means to the NRC in leak-before-break and in-2 service flaw evaluation regulatory aspects.
(( )
3 So the objective of the degraded piping program i 4 :was to verify and improve fracture mechanics analysis for 5 nuclear piping. It was started almost five years ago. It 6 will be five years next March.
7 The scope of the program was to assess various 8 materials at reactor temperatures. Specifically looking at 9 circumferential cracks in piping.
10 Also it was limited to looking at quasi-static 11 bending loads or pressure loads or combined at bending and 12 pressurt loads on cracked pipe. Whereas the IPIRG program 13 then went ahead and extended it to dynamic loads.
14 The pipe sizes that we have evaluated have varied
/") 15 from four inches in diameter to 40 inch diameter. And as V
16 you saw yesterday on the tour there was a case where we had 17 wall thicknesses up to three and a quarter inches.
18 As far as the analysis work we verified existing 19 . analysis, modified them, or developed new analyses to 20 predict either loads, displacements, or crack stability.
21 Applications for this program are obviously 22 leak-before-break. Applications to leak-before-break. When 23 this program started the NRC was just starting to develop 24 leak-before-break concepts and it was in the preliminary 25 stage of it. In-service flaw ascessment analyses were just Heritage Reporting Corporation 73 (202) 628-4888 U
I i
1 209 l
1 being started within the ASME code at the start of this
<~s
() 2 program.
3 You can see the total funding level for the 4 program.
5 Some of the interactions that have involved. In 6 BWR pipe cracking aspects we've had contributions to 7 NUREG-1061, Volume 3 from scrue of our efforts. And also ;
8 looked at the NUREG-0313, Rev. 2 which, for instance, 9 includes criteria for design of weld overlay repairs. And 10 we have conducted ervr iments specifically on weld overlay 11 repairs where there were cracks in the pipe to evaluate 12 their criteria.
13 For pipe whip restraints and jet impingement 14 removal that work came within NUREG-1061, volume 3, and also 15 GDC-4 broad scope revisions that are ongoing.
(')T 16 Future NRC policy might include environmental 17 equipment qualification, but I think since I made the slide 18 there has been some aspects of that that there's other ways 19 that the NRC might circumvent some of those problems.
20 Mike, perhaps you can help me out here a little j 21 bit.
22 MR. MAYFIELD: That's still under consideration.
23 DR. SHEWMON: You ask different people you get a 24 different answer.
25 While you're stopped though I would be interested Heritage Reporting Corporation (202) 628-4888
l 210 1 in, has anything from this got into section 11 or what is ;
1
(')
Lj 2 the device for working that interface? You've talked about )
3 uses in section 11 some,this morning and a good deal 4 yesterday.
t 5 DR. WILKOWSKI: Yes, I will be showing you some 6 examples of where it fits into section 11 later on. And, 7 yes, I'm on the Section 11 Flaw Evaluation Committee and so 8 I go to the committee meetings and supply them with data 9 from the pipe experiments or material toughness data for use 10 in their assessment of the criteria. And also make 11 independent assessments of the criteria on occasions for the 12 NRC.
13 DR. SHEWMON: Okay.
14 DR. WILKOWSKI: Some industry aspects, cf course 15 there's the IWB-3640 which is austenitic piping. That
! (^)3
\-
16 criteria was essentially well underway and developed even 17 before we had experimental data. So our data was used j 18 mainly to verify their criteria. )
l l 19 They recently have developed a 3650 criteria for 20 carbon steel pipe flaw evaluations. That criteria, we have
! 21 developed data prior to them developing the criteria, so our 22 data had direct input into assessing their criteria. And 23 also supplied them with some lower bounds for determining 24 material toughness properties.
25 Some other aspects: we have made improvements in l
Heritage Reporting Corporation (202) 628-4888 O)
\-
1 211 i 1 the NRC's -- Ray Clecker had developed a procedure called
["')
\ /
2 NRC LBB. We have extended some of the linear elastic 3 formulations that could be used in that.
We have developed 4 a computer code called'NRCPIPE which is PC base code that 5 has various through-wall crack pipe estimation schemes in it 6 as well as a number of sample J-R curves from different 7 materials that we have tested for exercising that code.
8 Also, we made assessments of ASME Section 11 code 9 criteria for flaw evaluation.
10 Some of the interactions we've had directly with 11 industry. For instance, the pipe test data and the raaterial 12 property data that we have developed have been suppli.ed ,
13 directly to the Section 11 pipe crack task group. We have 14 procured pipes from canceled power plants, most of our pipes (j'
A 15 came from canceled power plants. And in one case we had a 16 cracked pipe that was removed from service and we 17 decontaminated that pipe and tested that particular pipe.
18 That was a stainless steel recirculation line pipe, it was 19 28 inch diameter, so it was a fairly good size pipe to clean 20 up.
21 We had stainless steel submerged-arc welds which 22 were supplied by EPRI to the program for verification of the 23 fracture toughness of stainless steel welds.
24 We had aged cast stainless steel pipe, 25 artificially aged that was donated by Framatome to this Heritage Reporting Corporation (202) 628-4888 1 \-)
l l
I 212- .
l 1 _ program; and then we also sent some of that pipe to Bill 2 Shack at Argonne for his efforts.
3 We've obtained ferritic welding procedures from !
l 4 Babcock & Wilcox. And'our weld overlay repair evaluations 5 have been developed to using weld overlay repairs 6 manufactured by NUTECH as per the way they would make the 7 weld _ overlay repair in plant. We would send them a pipe and 8 they do it exactly to the procedures. And we've also given 9 material to Babcock & Wilcox and several other industry 10 people for testing for confirmation of our data. ;
11 We've also worked a lot with other NRC programs.
12 The one we worked quite extensively is over at' David Taylor 13 Research Center. In their program they have done --
14 MR. ARLOTTO: Gary, that slide -- isn't there a 15 slide before that. That's continued.
16 DR. WILKOWSKI: Yes, there's the first one. I'm 17 sorry, I had those two out of order.
18 Some of the programs that we have dealt with here 19 are materials. engineering. Associates where they have 20 developed the material property data base called FIFRAC for 21 the NRC that has J-R curves and tensile test data within it.
22 They also have a PC version of that data base. And we 23 supplied them with quite a number.
24 Charlie, do you remember how much stuff you sent 25 them?
Heritage Reporting Corporation (202) 628-4888
l l
213 1 MR. MARSCHALL: You mean material?
() 2 DR. WILKOWSKI: Our J-R curve test. Probably well 3 over 100 J-R curves.
4 MR. MARSCHALL: Yes.
5 DR. WILKOWSKI: About 100, 150 tensile test 6 records. So that's quite a bit additional data that you 7 send in raw form.
8 We've also procured materials and sent some of 9 those materials that we have not tested in our program.
10 Some of the surplus materials that we bought, we also send 21 some of those to MEA for their testing in their program as 12 well.
13 And Argonne, we have supplied them with cast 14 stainless steel pipe for aging studies. PNL we supplied 15 them with cast stainless steel pipe and elbows for some of 16 their NDE work.
17 And at David Taylor's I was starting to say, we've 18 done an awful lot of work with them in pipe testing and 19 evaluations of some round-robin efforts on tensile testing, l 20 on electric potential testing used to document crack growth 21 and fracture mechanics specimens. In comparisons and ways 22 to calculate J-R curves from ASTM type of procedures to make 23 sure we are all calculating the same types of information 24 the same.
I 25 Some of the technical disciplines then that we l
l 1
Heritage Reporting Corporation ;
(202) 628-4888 l 1
214 1 have involved in our program are in the material
(')
v 2 characterization area which Charlie Marschall will talk 3 about, is we have done tensile test and fracture toughness 4 testing that was necessary on each of the pipes that we 5 tested in the program.. So we have data specifically on 6 those pipes not generic data for those materials.
7 There have been a number of other metallurgical 8 investigations or investigations that evaluated size effects 9 and how you can extrapolate J-R curves from small specimens 10 to bigger specimens.
11 The pipe fracture experimentation, Paul Scott will 12 talk about that. There has been a fair amount of 13 instrumentation development and data acquired in the 14 program. The instrumentation difficulties increase r' 15 considerably once you go above 300 degrees Fahrenheit. At b} 550 F it becomes difficult to do many things because typical 16 17 strain gauges, for instance, won't work above 300 F.
18 The analysis work that has been done, Jalees Ahmad 19 will talk about some of that.
l 20 In the limit load we evaluated the net section ]
21 collapse analysis. We have developed a method, a screening 22 method which would tell us when you can use limit load 23 safely, a simple method, and when you need elastic plastic 24 fracture mechanics analyses. l 25 We checked some of the ASME code and NRC l l
Heritage Reporting Corporation (202) 628-4888
. _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ a
l +
215 l 1 procedures. A lot of work has been done in the elastic- j
()
k l 2 plastic fracture mechanics estimation schemes and some i
3 limited amount of work on finite element analyses. !
4 I guess at this time, if you don't have any 5 questions for me I will let Charlie Marschall start getting 1
6 into the materials characterization part. f 7 There's a separate handout for Charlie going 8 around.
9 (Slides being shown.)
10 MR. MARSCHALL: I want to talk a little bit about 11 material characterization efforts and degraded piping 12 program.
13 This effort was an integral part of the program, 14 even though a relatively small part. It provided support 15 for the pipe fracture experiments at both in planning the
()
16 experiments and in analyzing the results of the experiments.
17 I provided an input to the pipe fracture data 18 record books that were issued for each of the piping 19 experiments. And I will say a little bit more about that in 20 a moment. It also, as Gary said, provided data for the NRC 21 pipe material data base.
22 We attempt to stay current in the testing 23 technology, primarily through interaction with ASTM and 24 specifically the ASTM-24 Fracture Committee and also to 25 interact with the fracture community at large, fracture Heritage Reporting Corporation gg (202) 628-4888 V
216 1 testing community.
() 2 And our testing, I'm talking about laboratory 3 specimen tests en specimens removed from the pipe.
4 The objective is to provide the following material 5 characterization data for pipes that are subjected to full i
6 scale fracture test: chemical composition; complete tensile 7 stress-strain curves; Charpy V-notch transition curves 8 including not only the energy but the fracture appearance 9 and the expansion; and then the crack growth resistance 10 curves, commonly called J.
11 I feel a little sheepish explaining what J is with 12 Professor Hutchinson here, but I will anyway for those of 13 you who aren't that familiar.
14 (Laughter) 15 MR. MARSCRALL: If there are any questions we will f]
a 16 refer to him.
17 The Charpy, of course, goes back a long, long way 18 and it is simply a machine notch three point bend specimen.
19 And the notch radius is fairly about ten thousands of an i
20 inch radius. And the specimen is supported at these ends j 21 and struck behind the notch and simply measure the energy 22 consumed in fracturing specimens.
23 So that energy term then includes the energy to 24 initiate the crack to produce plastic deformation to grow 25 the crack and everything. So it's not very well separated.
Heritage Reporting Corporation (202) 628-4888
h l i l
217-
- 1. Plus the' specimen is relatively small. It may not very well 2 predict what thick material is doing.
f J 3 A more recent test for measuring fracture 1
4 resistance is, one would be a compact tension specimen shown 5 here. And this as a machine notch but then the.tip of it is 6 finished off by fatigue precracking. So we have a;very.
i 7 sharp flaw and the specimen is loaded by pin loading this 8 way. You measure the force applied. We put a clip gauge, a 9 displacement gauge across these knife edges to measure the 10 opening displacement along the loading line. And then we 11 also measure the growth of the crack. And thero are several 12 ways to do that.
13 The way that we do'it is with direct current 14 electric potential method. That involves putting current 15 through the specimen top to bottom or on the top and I')T 16 measuring the potential across the notch mouth. And by the 17 change in that potential we can determine when the crack 18 begins'to grow and how much it has grown.
19 MR. MICHELSON: How large a precrack is normally 20 prescribed?
21 MR. MARSCHALL: How large? How long?
22 MR. MICHELSON: Yes.
23 MR. MARSCHALL: It is a minimum of fifty I 24 thousands, and I believe it's two and a half percent of the 25 remaining on crack ligament. So whichever is larger.
Heritage Reporting Corporation (202) 628-4888 L._--_------___.-___ . _ - _ - _ _ . - _ . . - _ - . _ - . - _ - _ - _ - - . - _ . - - . _ - - . . _ _ _ _ _ _
^,_
^
21E-1 A typical set of test data then would be the load y
.( ) 2- displacement curve shown here and a crack extension versus 3 displacement ' curve shown ' here. The load displacement curvii 4 rises elastically and then we begin to get plastic 5- deformation. Somewhere in here we get crack initiation.
6 And then sven though the crack is not growing the load 7 continues to rise for some time because of material strain.
8 But ultimately the crack extension prevails and the load I
9 starts to drop off. ;
\
10 So from these data, load and displacement crack 11 extension we can calculate a J resistance curve of a curve 12 of J versus crack extension.
13 This is actual data for one specimen. This is a 14 ferritic steel, an 8106-P steel. And I'm showing two curves-15 which I will explain in just a moment. But the point is,
()
16 the higher we go on the curve before we start to get crack 17 extension and the tougher the material is and also the 18 steeper the curve the more the material is resistant to 19 crack extension.
20 MR. MICHELSON: That is zero then at the time you 1
21 set up the test specimen.
22 MR. MARSCHALL: Yes. The J reflects the area with 23 the load displacement curve.
24 DR. HUTCHINSON: What would be the ligament -- the 25 original ligament length in that specimen?
Heritage Reporting Corporation (202) 628-4888 O
L
I 219 !
1 MR. MARSCHALL: In this specimen? Since I don't g 2 have a specimen number I'm not sure, but probably about one 3 inch. Yes, it is possible, we did grow these much harder i l
4 than ASTM describes.
This would be a conventional measure of J called J j 5
6 deformation and it has a tendency to gradually diminish in 7 slope as the crack grows. This is called the modified J and 8 it is an attempt to normalize data from specimens of 9 different size and also to aid in extrapolating data. And 10 it's kind of an involved issue but we have done some work on 11 it and there has been -- since we did work on the degraded 12 piping program there has been quite a bit of additional work
' 13 done elsewhere. And the feeling is that we are kind of 14 getting this in hand. Which of these we want to use and how 15 we want to use them.
16 DR. SHEWMON: Which would you feel safe in using 17 the tearing modules?
18 MR. MARSCHALL: Well, the current trend is to use 19 the JB.
20 DR. SHEWMON: Is that a best estimate or always 21 conservative estimate?
22 MR. MARSCHALL: Down here is where tearing modules 23 is usually determined. It doesn't really make any 24 difference. That's usually done in first 60 mills of crack, 25 so in the tearing modules it isn't terrible important which Heritage Reporting Corporation (202) 628-4888
220 I
1 we use.
g (j 2 The features of our laboratory tests is that we 3 take our specimens from,the same lot of pipe that is used in 1
4 the pipe fracture test. We don't mechanically flatten the j 5 pipe because we don't want to introduce the cold work 6 effect.
7 And in the fracture mechanics test the conditions 8 that we model are those in the pipe test. That is, we try 9 to use the came notch acuity. If a pipe test is done with a 10 sharp machine notch, for example, we would also use some 11 compact specimens with a sharp machine notch. If it's a 12 repeat crack we would use a y crack.
13 The orientation of the crack plane and the crack 14 in both direction would simulate that in the pipe test. The 15 test temperature would be the same and the testing speed
() l 16 would be above the same. And one thing I haven't sut here 17 is that the thickness of the specimen would be roughly the 18 same thickness as the pipe. Now, if the machine is a flat 19 specimen out of a curve pipe, we obviously can't get the 20 full thickness but we get about 80 percent of the thickness 21 and we think that's a pretty good approximation.
22 This is how we machine specimens from pipe. And I 23 have already talked about the compact specimen, the Charpy 24 pecimen, this is simply a tensile specimen in the 25 longitudinal direction. Then there are two other types of Heritage Reporting Corporation g-) (202) 628-4888
\)
l 221 1 fracture specimens that we use in some materials. This is a
(')
\-)
2 three point bend specimen.again showing the crack growing in I i
3 the circumferential direction it does in the pipe test. And f i
4 this would be essentially equivalent to this kind of test 5 except this uses less material.
6 The so-called single edge tension test was one 1 7 that we did some development work on in this piping program. >
8 I will say more about that, too. But in that one the crack 9 is oriented so that the crack grows through the wall 10 thickness to simulate the growth of the surface crack pipe 11 and that would be loaded in tension.
12 The results of our studies are put into pipe l
13 fracture data record books. There is a separate data record 14 book for each pipe fracture experiment. And in that data rg 15 record book then there will be a section on material kai 16 characterization which we give a detailed description of the 17 procedures and the way we analyze the data.
18 We also include tables of data: chemical 19 composition; tensile properties such as yield, elongation, 20 RA; tensile stress-strain data, that is tabulated numbers 1 21 from the test so that it makes an -- it provides an analyst 22 an easy way to analyze the data rather than picking numbers 23 off a curve; Charpy V-notch impact test results; the J at 24 initiation and the dj/da or slope of the J resistance 25 curves.
l i
Heritage Reporting Corporation ]
(202) 628-4888 l
('% l
\> l I
l I
222 1 In addition to give graphs of data: complete !
/~S
( j' 2 stress-strain curves; the Charpy test plotted to show
]
i 3 energy, shear area, and, lateral expansion versus 4 temperature; and the J-resistance curves.
5 Here is a list of materisis that we characterized 6 in the degraded piping programs. It includes wrought 7 ferritic pipes in different sizes. You see the diameters 8 over there. I didn't show the wall thicknesses but those 9 are varied as well, i i
10 Submerged-arc welds in wrought forritic pipe.
11 Austenitic pipe including two kinds of stainless and 12 inconel.
13 Cast austenitic pipe, CF8M stainless.
14 Submerged-arc welds in wrought austenitic pipe. And then 15 some gas pumps and are welds, one in wrought austenitic pipe I~)T u
16 and one is a weld overlay repair on wrought austenitic pipe.
17 In addition to the straight material 18 characterization we did a few other tasks in the program.
19 We did some work on predicting large crack growths, JR 20 curves from small specimen data. The reason for this --
21 there is a problem here in a cracked pipe. We may draw the 22 crack several inches but in a compact specimen machine from 23 the pipe we really only have the potential to grow it 24 perhaps a few tensile an inch or maybe even less than that 25 depending on the size of the pipe.
Heritage Reporting Corporation f y- (202) 628-4888 j t-l l
_ _ - - _ _ - - _ - - . - - )
223 H 1 So we aren't getting large crack growth data and o we did some experiments in an attempt to put that on better i ) 2 3 and firmer ground. ,
4 We did some work on the development of the single 5 edge tension test. We did a fairly cmall study of pipe 6 anisotrophy effects. And then we participated in eeveral 7 round-robins with other NRC contractors such as MEA and 8 David Taylor and Oak Ridge. These included tensile test; 9 round-robins. Round-robin to calculate JR curves on the 10 same data. And use of direct current electric potential to 11 monitor crack growth.
12 I already mentioned that we have studies to in an 13 attempt to extrapolate small specimen dealing with the large 14 crack load. And in that work we looked at specimens 15 machined from plate because we couldn't get very large
( ')
16 specimens on the pipe. And I did have a slide of these 17 specimen but somehow it got mislaid so I will have to 18 describe them.
19 We did several series but the main series we did 20 was one inch thick plate we made one T-specimens which are 21 about two and a half inches square. We made three T-22 specimens which are about seven and a half inches square.
23 And we made ten T-specimens which are about 25 inches 24 square. >
25 So in the ten T-specimen we had an uncracked Heritage Reporting Corporation (202) 628-4888 N]s
224-1 ligament of ten inches long, so here we could run the crack
() 2 depth' farther. We saw'some cize effects in that study _and 3 that'has received some additional attention since then.. But 4 especially.in weld metal tests, for example, if we have a 5' weld in a_one inch plate and the weld itself is perhaps one 6 inch wide, and then in a small specimen that's a large 7 fracture of the total volume'put in a specimen.- This large 8 it's a small fracture and that seems to have an effect on-9 the result which is not surprising.
10 DR. SHENMON: Do you do any stainless steel welds?
11 MR. MARSCHALL: Yes.
12 DR. SHEWMON: It seems to me I was told one time 13 that when you got into casting the aging properties didn't 14 really get much worse than any of the welds began, started
() 15 life with. That is that the welds were really the places 16 with the low ductility. Did you get into that kind of weld 17 or what :tind of welds did you look at?
18 MR. MARSCHALL: We looked at submerged-arc welds 19 and cast tungsten arc welds.
20 DR. SHEWMON: In stainless steel.
21 MR. MARSCHALL: In stainless steel, in 304 -- i 22 DR. SHEWMON: Which of these are used to repair ;
23 elbows or housings for valves?
24 MR. MARSCEALL: I'm not sure I can answer that.
25 Well, the overlay repairs are gas tungsten arc, is that i
Heritage Reporting Corporation
, ,q (202) 628-4888 l V
l 225 1 correct.
TN
() 2 3
DR. SHEWMON:
MR. MARSCHALL-Those are the nice kinds.
Those are the nice kind. Those 4 are very tough. The submerged-arc welds are considerably 5 less tough than the base metal or than gas tungsten arc I
6 welds.
7 DR. SHEWMON: I haven't heard submerged-arc welds 8 on big cast stainless steel, but if you don't mind I will 9 save it.
10 MR. MARSCHALL: So we did develop empirical 11 methods for extrapolating generic curves. But as I said 12 there has been considerable work go on in other programs, 13 NRC programs to look at this and there has been considerable 14 progress made in that area as there has been in this third f3 15 point, comparing the usefulness of J deformation and J 16 modified for extrapolating JR curves. So that is a complex 17 issue that I don't have time to talk about now but there has 18 been --
19 DR. HUTCHINSON: Will someone mention later what 20 the other programs are that the NRC is supporting in this 21 area. I would be interested to hear about that.
22 MR. MAYFIELD: In the general area of pipe 23 fracture?
24 DR. HUTCHINSON: Specifically of looking at the 25 issue of extending extrapolated --
Heritage Reporting Corporation (202) 628-4888
226.
1 MR. MAYFIELD: It is principally the work done at 2 David Taylor Research Center with Ed Hackett and Jim Jcyce.
3 They are sort of the driving force. We've got money going 4 to Hugo Mertz at Georgia Tech; John Landis at Tennessee. I 5 have to think who else is playing. Merckel and Hansted from 6 Oak Ridge. And there are a couple of industry people.
7 DR. HUTCHINSON: That's good. I didn't know.that 8 this was going on actually.
9 DR. SHEWMON: This comes in then to what kind of 10 analyses?
11 MR. MAYFIELD: We're driving right now principally 12 for the low upper shelf welds in vessels. That's really 13 where it has gotten to be at issue because the draft ASME 14 criteria are based on an instability calculation for those
- 15 pressure vessels and you need six, seven tenths of an inch 16 crack growth. But you're only going to get half T to one T, 17 CT specimens in the surveillance capsules. So you have to 18 extrapolate the arc.
19 It has also gotten to be at issue to a degree for 20 the piping work. But the primary emphasis for us right now 21 is on the low upper shelf weld.
22 DR. SHEWMON: And they are welds so you have to 23 face it there.
24 MR. MAYFIELD: Yes.
25 MR. SERPAN: Let me just pick up something on that Heritage Reporting Corporation (202) 628-4888 O
[ ,
227 l
1 JM, JD. Mike has had this task group of these people going rm l)
% 2 now for about a year or so and they have been working pretty 1
3 hard on this. There will be another meeting of that group 4 in Knoxville at the end of this month, if you're interested I
5 in it at all. i l
6 DR. HUTCHINSON: The end of March?
7 MR. MAYFIELD: March 29th in particular. I 8 DR. HUTCHINSON: I can't come. But actually I'm 9 going to be visiting Landis at the middle of next month.
10 HR. SERPAN: That's going on every couple months I 11 or so. So that's very intense.
12 DR. HUTCHINSON: I was unaware of that. Thanks.
13 MR. MARSCHALL: Now, I mentioned er.rlier the 14 single edge tension test that was to simulate the growth of
() 15 the surface flaw in the pipe to the wall of the pipe. And 16 this is just a schematic of how such a specimen is tested.
17 It's loaded with rigid grips and we simply measure load and '
18 displacement, overall displacement and displacement across 19 the notch and we also measure crack growth by electric 20 potential.
21 The objectives of those tests I have already !
22 mentioned. And the status of these tests now is that we 23 have submitted a topical report to NRC describing the 24 technique development, the development of an estimation 25 formula, calculated JR curve, verification of the formula by i
i Heritage Reporting Corporation rT (202) 628-4888 .
() !
I l
l
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ . _ _ _ _ __ ___ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ . .J
e 228 1 finite element analysis, and the application of the method 2 to evaluation of cracks in welds,
{~}
s- i 3 We are not currently using this test. It is a !
4 test that we see some promise for in this area of surface 5 cracks but it has not been developed sufficiently to be used 6 in the -- it's not being used in the IPIRG program. l 7 This shows some submerged-arc weld specimens that 8 were tested in using these single edge tensions. And it 9 shows -- this is stainless steel base metal. This is heat 10 affected zone specimen where we put the starting crack in 11 the heat affected zone of the weld and this is with the 12 crack right in the weld metal. And you see the crack {
13 growing through the thickness of the material.
i 14 The bending that you see is characteristic of the 3 15 way we are loading it. But an interesting feature that we u) 16 noted on this test was that in this particular heat affected 17 zone specimen the crack when it got to the fur.icn line 18 followed along the fusion line at a very small crack opening 19 angle. And that is indicative of a low -- of a very low 20 fracture. It's much lower even than you see in the weld 21 where you can see the reasonable crack opening angle here, 22 it's supposed to be a very small crack opening angle there.
23 It's an observation that we think needs some further study 24 but we have not done this.
25 DR. SHEWMON: Just for my edification. When Heritage Reporting Corporation (202) 628-4888 O
i l
j
229 1 somebody does a crack opening angle do they do it at this
() 2 magnification or do they try to go back at a lower mag and 3 closer to the crack? ,
4 MR. MARSCHALL: Well, you would probably do it at 5 a higher mag. Yes. Or you could do it from the crack 6 opening, that is the displacement.
7 DR. SHEWMON: If you do that then you assume that 8 it is sort of on that weld, just a triangle that goes from 9 the very end of it --
10 MR. MARSCHALL: Yes.
11 DR. SHEWMON: Fine.
12 MR. MARSCHALL: Now the crack, if you're talking 13 about a crack tip opening angle or crack tip opening 14 displacement that is something that is measured right near
,'~l 15 the crack tip. And there is a standard for doing'that which A/ !
16 was not invoked at the time this was done. But this is just 17 kind of an after the fact observation when we sectioned the !
18 specimens.
19 As I said, we did a very brief look at anisotrophy 20 effects in pipes. What we noted in the pipe test as well as 21 in our compact specimen test in ferritic steels we 22 frequently saw the crack growing at an angle from the plane 23 that we intended -- that we thought it was going to grow.
24 The stainless steel grew the kind of way an analyst imagines 25 that it would grow.
Heritage Reporting Corporation (202) 628-4888 O
230 )
.1 But in'the carbon steel they tended to veer off l
.( 2 almost at right angles, frequently at, at leastL60 degrees, j
3 And we looked at some seamless pipe, looked at
, .h i
4 nonmetallic inclusions and saw that in one pipe they were 1 5 oriented at about 20 to 30 degrees from the pipe axis.- This 6 probably came about in twisting during the forming of this 7 pipe. But the fracture-resistance we looked at in several 8 different directions and it was significantly lower in the 9 direction of the inclusions which is what you would expect.
10 And normally you would expect inclusions to just 11 be oriented longitudinally and that too may be the reason 12 why in ferritic steels the crack tends to go off almost at 13 right angle because you form this plastic zone and the crack 14 starts growing -- from the crack tip it starts growing this
'( ) 15 way and this way. And it has already got a component in 16 longitudinal or inclusion direction and it may simply be a 17 simple manner in obtaining instability.
18 DR. SEEWMON: Is there anything in the spec that 19 prescribes shape control on the sulfides here or were they 20 strung out or were they balled up?
21 MR. MARSCHALL: They're strung out.
22 DR. SHEWMON: Okay. There is a shape control.
1 23 And that way -- is there any particular -- and there is no i 24 sulfur spec either outside of 02 sulfur and something very 25 old-fashioned.
l Heritage Reporting Corporation (202) 628-4888
)
L l 231 1 1
1 MR. MARSCHALL: I don't'believe there is any l
) 2 sulfur spec on these.
3 DR. SHENMON: ,So they can give you anything they 4 want to in that regard.
5 MR. MAR 6CHALL- Yes. But it was something that l 6 the pipe testers observed frequently. And often we went 7 -- in our laboratory specimen test we went to side grooves.
8 We would test the specimen without side grooves and then if 9 the crack grew off at some large angle we simply would do 10 another specimen of the side groove to get some better 11 measure of the fracture.
12 DR. SHEWMON: What was the sulfur content of the 13 steel that showed this behavior, do you know?
14 MR. MARSCHALL: I know it -- I have it but I don't
(') 15 know what it is right now.
16 DR. SHEWMON: I would like to hear it later.
17 Thanks.
18 MR. SHACK: Just to comment on that, typically the 19 Japanese do have a much lower spec on the sulfur in their 20 carbon steel pipes. I just noticed that one Japanese pipe 21 they tested yesterday went straight.
22 MR. MARSCHALL: You mean it went the way --
23 MR. SHACK: It went the way it expected it to.
24 DR. SHEWMON: I have studied this some and 25 remember being appeased by some Japanese in saying, we can't i
Heritage Reporting Corporation '
(202) 628-4888 C
L-232 1 find steel with that much sulfur in it over here to do those in
(_) 2 studies anymore.
3 DR. WILKOWSKI: ,
Bill, we also did another.through-4 wall crack pipe test on-the Japanese and in that one it grew 5 off at a slight angle. The one went straight yesterday 6 because it was broken at low temperature and it was a 7 cleavage fracture.
8 MR. SHACK: But they do have a much stricter 9 specification on sulfur in their carbon. What would be an 10 SA-106 in Japan has a different designation, it has a much, 11' much more restrictive sulfur.
12 DR. SHEWMON: Now is SA-106 used in new plants? I 13 have the impression that is a pretty old --
14 MR. MAYFIELD: It is still widely used.
() 15 DR. SHEWMON: It would be nice if the ASTM or 16 something could get a more modern sulfur spec on it. The 17 toughness depends entirely on inclusions in this stuff and 18 if you've got these old-fashioned sulfur contents you've got 19 problems.
20 MR. MAYFIELD: It's an uphill battle with the 21 steel manufacturers, obviously.
22 DR. SHEWMON: Well, other customers ask for it and 23 pay for it and get it.
24 MR. MAYFIELD: They buy it from the Germans and 25 Japanese.
Heritage Reporting Corporation (202) 628-4888
233 1- DR. SHEWMON: No, they get it from the domestic O 2 outfits and you put it in your specs. You may have to pay
(/
3 something extra but they make it for customers that care.
4 The nuclear industry ought to care, too.
5 DR. HUTCHINSON: When you repeat a test with side 6 grooves, do you find that the toughness bends and the JR 7 curve drops?
8 MR. MARSCHALL: The JR curve is lower but at 9 initiation there is sometimes not much difference, sometimes 10 there is significance difference. But generally the affect 11 more is on the crack growth.
12 DR. HUTCHINSON: And it is lower.
13 MR. MARSCHALL: Yes.
14 'Some interesting findings of the material
() 15 characterization was that the flux welds are significantly 16 less tough than inert-gas welds.
17 I showed you an example of unusually low toughness 18 along fusion line, but that is a pretty isolated case and we 19 don't know a lot about that.
20 We also found that many carbon steel pipes are 21 susceptible to what's called dynamic strain aging. And 22 dynamic strain age doesn't necessary -- the dynamic does not 23 refer to the grade on the test, it simply means that the 24 aging is occurring while the specimen is being tested. As 25 it's at an elevated temperature the mechanism aging, that is Heritage Reporting Corporation (202) 628-4888 O
-l 1
234 1 the' movement of carbon and nitrogen can occur simultaneously-2 with the strain, so that's why it is called dynamic.
3 But there's also dynamic strain aging are that the 4 tensile strength at 550 F is greater than at room 5~ temperature which seems like a plus. But the negative.part 6 is at the J-1-C, the initiation J and slope of the dj/da 7 curve at 550 F is often less than at room temperature.
8 We also observed in carbon steel pipes and in the L 9 compact specimens that there were crack instabilities at'550 10_ F. Not cleavage fracture but sudden jumps.
11 This is some data from Babcock & Wilcox showing 12 J-1-C versus temperature for several carbon steels that are 13 used in the pipes. In these too we see a minimum of 14 toughness at around 400 F. And we think this is related to
, . 15 the dynamic strain aging, although the full explanation of 16 the mechanism is pretty lacking of what's going on.
17 This is illustrating the unstable crack growth in 18 compact specimen. This is a specimen tested at 550 F 19 showing the load displacement coming over the top and there 20 you get a rather large jump then stable crack growth and 21 another jump, stable growth. You see four different crack 22 jumps in this specimen. And at 300 F we have an uneven 23 curve partly due to duration in yielding which are typical 24 in strain aging. But there may be some small jumps here, 25 too, but they are certainly not significant.
Heritage Reporting Corporation (202) 628-4888 O
. . j
- )
235' 1 DR. SHEWMON: And each one of those is a yield 2 point then; is that right?
/^)
.a-3 MR. MARSCHALL: Well, it's a low drop associated l 4 with a rapid extension'of the crack. You can hear a report 5 on the specimen that does this.
6 DR. SHEWMON: You may not like the word, but if 7 you go back to the previous slide there was then subsequent 8 work hardening or if not that at least a yield in low 9 stress --
10 MR. MARSCHALL: You mean here?
11 DR. SHEWMON: Yes. No, immediately after the 12 drop.
13 MR. MARSCHALL: Oh, right here.
14 DR. SHEWMON: Yes. Now it rises.
- 15 MR. MARSCHALL: Yes, okay. Well, this is --
16 DR. SHEWMON: Is that all going in --
17 MR, MARSCHALL: -- the crack that the load drops 18 so far that the cross head couldn't keep up with it and so 19 now it's just reloaded. But that could be a yield point as 20 well.
21 DR. HUTCHINSON: Has anybody analyzed the specimen 22 to see -- under the circumstances to see if it's consistent 23 with the instability criteria?
24 MR. MARSCHALL: You mean from the instability due 25 to the compliance of the -- j l
Heritage Reporting Corporation (202) 628-4888 O
l l
_ _-______ _ ______ . _ _ _ _ _ _ _ 1
. j F 236 1 DR. HUTCHINSON: Yes. I mean, up until the peint-
, )' ~2- .of the first instability, say, you've got presumably a L i 3- ; fairly nice JR curve. ,
4 MR. MARSCHALL: Yes.
5 DR. HUTCHINSON:. And obviously you can analyze the 6 specimen in various ways. .I'm just-curious if'anybody took 7- the effort to'see if the instability'you are predicting 8 .there'is consistent with the theory as to where.the theory 9 would say instability should set in.
10 MR. MARSCHALL: You're speaking of instability 11 created by the compliance of the --
12 DR. HUTCHINSON: Correct.
13 MR. MARSCHALL: Yes. No. Well, I can't say that
- 14. we went into great depth on it, but a cursory look at it
() 15 says, no, it can't be that this is being driven by machine 16 compliance.
17 DR. WILKOWSKI: We looked at'it a little bit, 18 John, and for one sample in particular and found out that 19 the machine stiffness, it was stiff enough that that 20 shouldn't have been the reason why there was an instability.
21 There's something else other than machine compliance.
22 MR. MARSCHALL: And also we get in some cases 23 instabilities on the rising part of th, curve.
24- DR. WILKOWSKI: Right. Which has nothing to do
- 25. with the compliance of the machine.
Heritage Reporting Corporation c (202) 628-4888
J L
si 237 l 1- MR. MARSCHALL: In fact, I.think that's the;next.
([ 2 one I'm. going to show.
3' , DR. WILKOWSKI- You'can see a few little'ones. 1 4- MR. MARSCHALL: This'one is in a pipe. test'on the 5 same' materials that I just showed or the previous. slide i 6 showing numerous small instabilities. There must.be about 7 10 ' <nc .15 instabilities on . that. These are deliberate 8 unloadings'and reloadings.
9 DR. WILKOWSKI: Those look like small 10 instabilities but that was a 28 inch diameter pipe and.so 11- the crack jumps were arranged from maybe a quarter ot an 12 inch to four inches.
13 MR. MARSCHALL: Well, just a few words-in summary 14 then. The' material characterization studies have been an important part of the degraded piping program. And the data
( ) '15 16 obttined have been used in planning the pipe experiments and 17 analyzing the results of the experiments. And the data are 18 available to others for further analysis of pipe experiments 19 vis-a-vis pipe fracture data record , books and vis-a-vis the
-20 NRC pipe fracture data base.
21 If there are no questions I'll go to Paul Scott.
22 DR. SHEWMON: Will you get me a sulfur content on 23 that --
24 MR. MARSCHALL: Yes.
25 (Slides being shown.)
Heritage Reporting Corporation (202) 628-4888
238 l' MR. SCOTT: I will be discussing task 4,000 of the r'$ ' 2. degraded piping program ~which is,the full scale pipe
.V 3 experiments. ,
'4 , The objectives of-task 4,000 was obtained: fracture 5 data'on circumferential1y cracked piping. We would use --
a c V '
T '
6- these full scale pipe experiments we would use 7 representative piping materials,. crack, and' loading ls. 8 . conditions.
9: .The area. alluded to earlier, most of the materials
,~.
10 we evaluated were pipe samples that we obtained from.
11 cancaled nuclear power plants.
12 The majority of experiments we conducted were 13 conducted at temperatures 550 Fahre'nheit. Part of these 14 experiments we determined crack initiation loads; maximum-15 loads; and cracked growth data for the comparison to the JR 16
~
analysis.
17 We evaluated the limit load analysis using the 18 net-section-collapse in which part of this would develop the 19 screening criteria to show when net-section-collapse 20 analysis appropriate would be appropriate and when some more j 21 detailed elastic plastic fracture analysis would be 22 required.
23 And we also established and maintained I think 24 extensive pipe fracture database.
25 As far as my presentation it will begin with a Heritage Reporting Corporation (202) 628-4888
()
L _
239-1 review of the' test matrix. We will look at some examples of
() 2 some'of the pipe fracture experiments. .As part.of~this we 3 -will look at a short video tape of some of the experiments.
'4 Review the pipe fracture database. We will look at some.
5 comparison between the net-section-collapse analysis and the 6 experimental 1'results and also the screening criteria which 7 was developed.
8 As part of this program we evaluated three 9 different crack geometries under four different loading 10 conditions. The loading conditions were: pure four point 11 bending; pure pressure loads; combined pressure and bending; 12 and compliant bending. In the compliant test we either used.
13 long lengths of pipe.or a series of bell springs in series 14 of the. load-train to simulate the excess compliance of.a
~( ) 15 'long piping rod.
16 The flaw geometries we looked at were idealized 17 through-wall crack. Typically 37 percent of the pipes come 18 from some length. A part through surface crack which was 50 19 percent of the pipes comes from some length and 20 approximately 6 percent of the wall thickness and depth.
21 And the third was a complex crack which was a long internal 22 surface crack that had penetrated the wall in a localized 23 area similar to the crack found in Arnold.
24 This is a break down of the experiments by pipe 25 diameter. As you can see the majority of the experiments we Heritage Reporting Corporation (202) 628-4888
240 1 conducted were on the 10 to!6 inch diameter range.
4
'('N~ 2 Typically the larger diameter experiments, you will see in a y.
'3 subsequent / viewgraph, were all through-wall crack 4 ~ experim nts.
5 '?
!This.is a bheak down of experiments by. crack 6 geometryL YN lot of six inch' diameter surface crack 7 experiments as I alluded to.- All the large diameter 8 experiments 28 inch in diameter were through-wall-cracks.
9_ The one large diameter. surface crack experiments we.
10 conducted to.date.was conducted as far as the IPIRG program, 11 that was a Japanese pipe test that you saw yesterday-set'up 12 in the facility.
13 This is a break down of experiments by material.
714 The 37 inch diameter experiments were the two cold-leg
.15- experiments we conducted. One was conducted on a section of
.O 16 cold-leg or through-wall crack in.the base metal. The other 17 was in a through-wall crack and a shop fabricated growth 18 weld.
19 The 28 inch diameter, the stainless steel weld
'20 that Gary alluded to was a section o? recirculation line in 21 the Nine Mile Point plant that was cut out, we cleaned up 22 and tested.
23 The 12 inch diameter test were two sections of 24 line that were artificially aged at Argonne. One had a 25 crack in the weld and the other was a crack in the base i
l Heritage Reporting Corporation I (202) 628-4888
()
l o >
241 1
1 metal. We also in the age cast stainless there was one 16 1
/^%
(_/ 2' inch diameter test specimen that was from a section of age 1
3 cast stainless steel provided by Framatome.-
4 As far as some of the stainless steel welds we 5 looked at two different conditions. In one condition the
, 6 welds, we looked at the average C weld and the other we 7 looked at a weld that had beer solution annealed.
8 This next series of viewgraphs will just show some 9 of the pipe fracture experiments and the load frame that we 10 used in some of the experiments.
11 This is the typical loading frame'that we used.
12 This was the large strong facility that you saw at West j 13 Jefferson and it was used for all the experiments 16 inches 14 in diameter and greater.
() 15 The ends were restrained by some wire ropes. Two 16 large actuators with about three quarters of a million 17 pounds capacity each to put the pipe section into four point 18 bending. The roller saddle assembly will allow the pipe to 19 translate and also rotate so that we didn't build up an 20 axial membrane stresses during the bending.
21 The spring assembly shown here is a typical, we 22 just conducted one experiment with the spring assemblies in 23 place, tno simulator compliance, excess compliance with all 24 piping systems.
25 This is a post test photograph of the cold-leg Heritage Reporting Corporation (202) 628-4888
---_ _- i
$i f 242
[ 1: weld' experiment. .Here is the wire ropes that reacted the I) 2 resistance bending moments. Here is the rams. As"you.can
_3 see, the pipe -- all the experiments were conducted at 550.
4 degrees Fahrenheit. The pipes were wrapped-in a very high 5 . temperature heater' tapes and insulation.
6 This'is the post' test photograph. . This.was the
.7 pipe sample'that you saw out there yesterday sitting outside 8 the building.
9 This.ic post test photograph of one of.the weld 10 . overlay-experiments we conducted. This is a 16 inch 11 diameter specimen. The crack was a fatigue sharpened 1:2 through-wall crack.approximately 40 percent of the pipe
- 13 circumference in" length. It was sent to NUTECH Engineering 14 in San Jose,' California. They fabricated a weld overlay
- ().15 over top the cracks. We tested the repaired section under 16 combined pressure and four-point bending.
17- This is a photograph right before the-test.
18 , specimen failed. You can see that the pipe went through a 19 tremendous amount of deformation. The excess thickness of 20 the overlay supported the cracked section such that all the i
- 21. deformation occurred outside the crack section. It would l-22 take a tremendous amount of deformation, you can bend these 23 things almost like a pretzel.
24 This is a photograph or two photographs of the 12 5 spring assembly in the uncompressed condition and the L
Heritage Reporting Corporation (202) 628-4888 l-O
_m - __..m .. _ _ _ _ . _ _ _ __ _ _ . _ _ _ _ _
I 243 1 compressed condition, like the excess compliance in the 2' large piping system.
3 DR. SHEWMON: ,
Keith, what.is NRR's position or 11 what are they being asked to do these days on leaving 5 overlays in for two, three, and four runs?
6 MR. WICHNEN: Fortunately, that's not my 7 responsibility.
8 DR. SHEWMON: Well, by. association here.
9 (Laughter).
10 MR. WICHMAN: My. understanding that originally 11 this was, quote / unquote, "a temporary repair." 'I'm not sure 12' what the official position is. NUREG-0313, Rev. 2.I believe 13 it gives a limited time that it even. stays in because you 14 still have to inspect on a more frequent basis. There is 15 and at more and more weld overlays it does affect the stress
[}
16 analysis. You have to reanalyze the piping system.
17 They have gotten away with two or three or four 18 cycles perhaps in some cases.
19 DR. SHEWMON: At some point that's -- at least in 20 the stress corrosion crack is pretty strong.
21 MR. WICHMAN: In fact on Brunswick recently it was 22 well over the safe end to the nozzle.
23 DR. SHEWMON: Brunswick was the sister plant that 24 Dwayne Arnold was --
25 MR. WICHMAN: That's right.
Heritage Reporting Corporation (202) 628-4888
"244' 1 DR. ' SHEWMON: They had to keep, going in and
'O 1,j l2 inspecting the thing but never find any cracking.
3 MR. WICHMAN: ,That's right. You had_a thermal 4 sleeve of the same configuration as Dwayne Arnold, cracking 5 of the same nature you had with crevice. 'And also the 6 cracks grew into the - from the 182 --
7- DR. SHEWMON: That's at Dwayne Arnold or they have 8 seen this at Brunswick?
9 HR. WICHMAN: At Brunswick and at Dwayne Arnold.
10 The official position I don't know. At least there is no 11 push on that I'm aware in NRR.
12 DR. SHEWMON: They try and keep a certain level of 13- pain, but they will accept a technical argument for leaving:
14 it in if they want to inspect it and pay the price.
() 15 MR. WICHNaN: It is virtually -- a decision made 16 virtually on a cycle, on a separate basis.
17 DR. SHEWMON: Okay, fine. Thank you.
18 MR. SCOTT: This is a short video of some of the 19 experiments conducted as part of this program.
20 (Video tape being played.)
21 MR. SCOTT: This is one of the weldfoverlay 22 experiments. This is -- I'm not sure if it's a six inch or 23 a 16 inch. I can tell in a second. But it was pressurized 24 with subcooled water and bent to failure. At six inch the 25 wall thickness of the pipe -- it's schedule 120 pipe, wall Heritage Reporting Corporation (202) 628-4888
l 1
245 1 thickness of about .56 inches and the overlay thickness is about .315.
(-)
V 2
3 You can see the -
amount of deformation that this 1 1
4 little six inch pipe also withstood. This is a different 5 loading fixture than what you saw. This is just what we ]
I 6 refer to as our small strong back. We did the six --
7 intended six experiments on.
8 DR. SHEWMON: That's insulation that is raining 9 down there.
10 MR. SCOTT: That's insulation that's raining down.
11 Now this is a low toughness submerged-arc weld. I 12 believe it is also a six inch test. The thing to notice in 13 this particular video is the difference in the deformation 14 between the previous segment and this segment of the video.
15 This piece of pipe still almost looks straight. This is a 0 16 very low toughness submerged-arc weld in a six inch 17 stainless steel.
18 This was a section of pipe. The weld was donated 19 by EPRI.
20 DR. SHEWMON: How do they get the low toughness, 21 put a lot of swag in it or do you know?
22 MR. SCOTT: It's just typically the welding 23 procedures. You can see almost still straight.
24 The crack opening here. This is some of the 25 instrumentation that was around the crack.
Heritage Reporting Corporation j (202) 628-4888 '
G O
l i
I
. .. .. 1
t 246
.1 DR. SHEWMON: And that was an inside crack?
~ l
()s 2 MR. SCOTT: That was internal surface crack. I 3 Probably~50 percent of circumference in length and about 66 i
4 . percent circumference in depth.
.5 This is a load toughness submerged-arc weld except 6 -this is a 16 inch diameter pipe. This particular weld was 7 solution annealed. I believe this is an end view of the
- 8. same experiment.
9 This is a section of an aged cast stainless steel.
10 I believe'I have to' check, I'm not sure if this is 16' or '12 11 inch. This is a 16 inch.
12 DR. WILKOWSKI: That last test I believe, John, 13 when you.went out to West Jefferson on your last visit here 14 you were looking down at the end of the pipe.
() 15 DR. HUTCHINSON: Yes. That's it.
16 DR. WILKOWSKI: That's the pipe that was tested.
17 DR. HUTCHINSON: Yes.
18 MR. SCOTT: Most of this was instrumentation or 19 insulation. ,
i 20 (Slides being shown.)
21 MR. SCOTT: Typically as part of these experiments 22 the measured applied loads and displacements at the low
.23 points. From the applied loads and the crack rotations the 24 applied loads we could calculate the applied moments. And 25 from the crack rotations to the applied loads we could come Heritage Reporting Corporation (202) 628-4888 O
247 !
\
1 up with a moment rotation relationship for the particular 2 experiment.
3 From the d-c electric potential data and the crack 4 open displacement data we could come up with the crack 5 growth data for the experiment. And from the crack growth 6 data and the crack open and displacement data we calculate 7 the crack opening area data that might be used in a leak-8 rate type calculation.
9 In addition we measure the pipe and fluid 10 temperature and the internal fluid pressure as far as these 11 experiments.
12 Next I would like to discuss briefly the pipe 13 fracture database that we have developed as far as the 14 degraded pipe program.
15 The objectives of this particular effort was to O 16 establish and maintain a comprehensive database for the pipe 17 fracture experiments. These include both a raw and the 18 reduced data from the actual experiment as well as detailed 19 material of property data for each of the experiments.
20 For each experiment that we conducted Charlie 21 conducted a corresponding series of material properties just 22 for that same material and same heat that are included in 23 these data record books.
24 We would establish as part of this effort the 25 format for addition of future data record books. And we Heritage Reporting Corporation j (202) 628-4888 l r~% l
(_/ J 1
248 1 have used these data record books to evaluate present and
(,) 2 future elastic plastic fracture mechanic analysis.
3 The current status is that we have five volumes of 4 data record books. It includes 61 experiments from the 5 degraded piping program as well as six experiments from 6 other programs. The former Battelle/EPRI' program as well as 7 some of the David Taylor data.
8 I would like to shift gears a little bit and start 9 looking at some of the results from these pipe fracture 10 experiments as they compare to the net-section-collapse 11 analysis and the screening criteria that we have developed.
12 The summary of the net-section-collapse results we 13 found that the initiation and the maximum load the margins 14 were larger for the through-wall cracks than ~~-f were for n
(,) 15 the surface cracks.
16 As I have mentioned we developed a separate 17 screening criteria for the through-wall cracks and the 18 surface cracks. )md they show the limits of when the 19 net-section-collapse analysis is appropriate and when more 20 detailed elastic plastic fracture mechanics analysis is 21 required. We will discuss the screening criteria in the 22 subsequent viewgraph.
23 We also have developed a unified simple screening
~
24 criteria that combines both the through-wall crack data and 25 the surface crack data that could possibly be used in the 1
l l Heritage Reporting Corporation (202) 628-4888 (r]>
249 1 code application. It's a criteria that gives a lower bound th
's_) 2 on.the failure loads. It gives a 95 percent confidence 3 level for the failure loads. Basically the 95 percent 4 confidence levels are two standard deviations below the best 5 fit to the data.
6 This viewgraph shows some of the results from the 7 two cold-leg experiments. The weld' metal experiment and the 8 base metal experiment as they compare to the net-section-9 collapse analysis.
10 In the blue a flow stress for the net-section-11 collapse analysis has been defined as the average of the 12 yield in the ultimate strength, whereas in the -- I guess 13 that's red -- has been defined as 2.4 times the value of 14 ASME from the code, 15 The one discouraging point looks like for the weld t( )
16 metal data -- for the metal test, when you use the actual 17 yield and ultimate strength from the weld material the point 18 to note there is that almost immediately after the crack 19 initiated it grew into the base metal. The base metal 20 having a significantly lower yield strength and ultimate 21 strength than the weld metal. So the fact that it's lower 22 that it did not reach net-section-collapse condition based I
23 on the weld metal data is probably not surprising. !
24 This is the screening criteria that I have
\
25 referred to. Basically the screening criteria is based on, I
Heritage Reporting Corporation r"y (202) 628-4888
(/ l l
1 250 H 1 this is a-simplified version, but basically it's based on
(). 2 the assumption-that when the remaining ligament from the 3 crack tip to the neutral axis or the plastic zone that hao I
4 the crack tip is as large as the remaining ligament between 5 the crack tip and the neutral axis then fuller plastic .
l 6 conditions exist and net-section-collapse analysis should be 7 appropriate. That occurs when this dimensional plastic zone 8 parameter is greater than one. When that's the case we can 9 see that typically the net-section-collapse analysis 10 provides a conservative assessment of the maximum stress for 11 the experiment.
12 When this parameter is less than one that means 13 that you have contained plasticity of the crack tip that the 14 plastic zone is less -- the size of'the plastic zone is less 15 than the remaining ligament and as it goes smaller and
{}
16 smaller we see that the maximum stress from the experiment 17 is less than predicted by the net-section-collapse analysis.
18 DR. SHEWMON: Is Sigma F a flow stress?
19 MR. SCOTT: Sigma F is a flow stress. In this 20 particular case it was defined as the average of the yield 21 in the ultimate strength of material.
22 DR. HUTCHINSON: And D is the diameter of the 23 pipe?
i 24 MR. SCOTT: D is the diameter of the pipe. j 25 And this is the 95 percent confidence level curve for the I
Heritage Reporting Corporation (2C2) 628-4888 l
1
1 l
251 1 data.
2 I would like to conclude the presentation with one 3 summary viewgraph. As part of this effort we obtained 4 materials from canceled plants with the predominate 5 materials we used in our full scale experiments. ,
I 6 We had a large test matrix that was conducted -- J 7 typically conducted at prototypical temperatures and 8 . pressures of 550 degrees F.
9 We evaluated three different crack geometries: the i
10 idealized through-wall crack; the part through surface H 11 crack; and the complex cracks similar to one might find at 12 Dwayne Arnold Plant. And we generated an extensive database 13 both material property data as well as pipe fracture data.
14 And as a result of this effort the limits of the net-15 section-collapse analysis is better defined.
16 Any questions?
17 (No response) 18 MR. SCOTT: Thank you.
19 DR. HUTCHINSON: Is the reason that the screening 20 criteria doesn't have any crack length in it, I assume, is 21 because in fact you're assuming a very long crack.
22 MR. SCOTT: Basically we're just assuming that 23 from that crack tip -- this doesn't have the figure on it.
24 But basically assume from the crack tip, the bottom part of ,
1 25 the fracture -- well, this is a simplified curve and it j Heritage Reporting Corporation j (202) 628-4888 ;
1
() 1 l
l E _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ___ __________
252
.1 'doesn't in the actual entire parameter, that we kind of I
-O(,) 2 eliminated some of the terms for a simplified parameter. It 3 does have the crack length tied up in it.
4 But it assumes that from that crack tip to the 1
5 neutral axis that distance is the -- when that plastic zone 1 6 encompasses that entire distance than you have fuller 7 plastic conditions. But this is a simplified thing that we 8 have thrown in there for code application.
9 DR. WILKOWSKI: We have a separate technical paper 10 on that if you want to look at that.
11 DR. HUTCHINSON: Yes.
12 DR. SHEWMON: Hopefully the code isn't just 13 ' evaluating cracks that are two-thirds of the way through the 14 wall and 50 percent around the circumference. So you say
() 15 this is the simply one you use for code applications. What 16 arc you doing for the code for realistic crack or is that --
17 DR. WILKOWSKI: This was a procedure that was 18 developed actually before the code had dc7 eloped their own 19 screening criteria. Now in the ASME code they essentially 20 developed techniqyes similar to the R-6 method which 21 determines when you may have -- John, are you familiar with 22 the R-67 23 DR. HUTCHINSON: No.
24 DR. WILKOWSKI: Well, in the R-6 method it 25 essentially gives bounds between linear elastic behavior, Heritage Reporting Corporation (202) 628-4888
i 253 i elastic-plastic, and bounds between elastic plastic and 2 fully plastic. And so they have developed a much more 3 detailed screening criteria, whereas here we have a one line 4 equation and they have a page full of equations that you 5 have to.go through to determine whether you are linear 6 elastic or more elastic-plastic.
7 DR. HUTCHINSON: But I guess it is true for these 8 pipes, in a sense, those experimental points have been 9 established using very big cracks. So that if indeed the 10 cracks are smaller and you check your criteria and you find 11 that you' re satisfying your -- you' re in the net-section-12 collapse region, then you certainly are because the cracks 13 are even smaller. I mean, I think that's the way to 14 interpret that, isn't it.
15 See, normally if I don't see a crack in a criteria 0 16 I wonder what's going on. But I think the answer to it 17 basically is that the experiments and the establishment here 18 has been done with very long crack and so any cracks, in 19 answer to Paul's question, any cracks that are shorter 20 lengths if you establish from this diagram that you meet the 21 ret-section criteria, then fine.
22 MR. SCOTT: That's right.
23 DR. WILKOWSKI: These are probably -- the crack 24 lengths that were selected here are probably the most 25 sensitive crack lengths to whether you're going to be in Heritage Reporting Corporation (202) 628-4888 O
1
\
i 254
-1 net-section-collapse or limit load.
2 DR. HUTCHINSON: Yes.
)
3 DR. WILKOWSKI- Crack lengths that would be' longer 4 or shorteriwould tend to go more.towards: limit. load. And 5 perhapaLthe greatest objection to using this sort of' thing 3 6 is maybe it's.too conservative for come other case.
7 1 DR. HUTCHINSON: It could be too conservative, 8 that's correct.
9 DR. WILKOWSKI: I know the Japanese first were 10 very interested in it. However, they believe that'you needed 11 to have also crack growth resistance incorporated into'it
- 12. somehow. We do indirectly in that there's crack growth in 13 the' pipe experiments, in the empirical correlation.
14 DR. HUTCHINSON: But this is -- the purpose of 15 this is to give one guidance as to whether one should do a 16 more sophisticated analysis or not; right?
17 DR. WILKOWSKI: Right.
18 DR. HUTCHINSON: And so if it's conservative and 19 one goes about a more sophisticated analysis we will find 20 that out.
21 DR. WILKOWSKI: Yes. Frequently we'll do this 22 sort of thing and, you know, it's just a one line equation.
23 So you plug it out real quickly. And if you find out that 24 you're in a region of concern you might go back and use the 25 ASME screening criteria which is a page full of equations Heritage Reporting Corporation (202) 628-4888 O
E
255 1 'and then calculate it out over more detail.
(,) 2 MR. SCOTT: I will turn it over to Jalees now and 3 talk about the analysis.
4 DR. SHEWMON: Why don't we take a break.
5 (Whereupon, a 10 minute break was taken.)
6 DR. WILKOWSKI: We will continue with Jalees 7 Ahmad.
8 (Slides being shown.)
9 MR. AHMAD: This is a continuation of the degraded 10 piping program presentation, a less glamorous part because I 11 don't have any video tapes.
12 I'm going to talk about fracture mechanics 13 analyses and the degraded piping program. A list of people 14 who worked on it: Bud Brust is here who worked on some the
(( ) 15 pipe estimation schemes.
16 The emphasis in this work, in the degraded piping 17 program was a pretty well integrated program combining 18 experiments and analytical work. However, I want to 19 emphasize that the objective of the analyses work here was 20 basically to come up with simple methods for analyzing and 21 predicting the behavior of flawed piping, piping in nuclear 22 power plants for applications such as leak-before-break and 23 in-service flaw evaluation.
24 And as stated here to provide NRC with simple and 25 generic models. In other words, providing a single finite l
Heritage Reporting Corporation (202) 628-4888 1
l 1
l 1
_ _ _ _ _ _ _ . _ l
256 1 element calculation in detail of a given situation was not
) 2 an objective,.but rather to come up with some simple as in 3 -- or at worse something that could be done on a PC.
4 The majority of the work that I'm going to present 5 is in the regime of elastic plastic fracture mechanics. I'm
,6 not going to talk about the limit load type analysis that 7 Paul Scott presented earlier.
8 Just as a reminder, looking at simple methods to 9 predict failures present in pipes, on a failure stress 10 toughness space we are looking for nuclear materiale 11 basically in the regime where the boundary of the elastic-12 plastic fracture mechanics and limit load applications, 13 because generally the materials are very tough.
14 The approach in the overall program was really to
() 15 mesh the analyses very closely with experiments in the hope 16 of coming up with practical solutions.
17 The flow of information shown here is just some 18 small specimen tests, developed data such as Charlie 19 Marschall showed earlier. Given to some kind of an analysis 20 procedure, called estimation analyses here. These are 21 simplified methods again in the form of equations. For 22 example, the ASTM in finalizing crack growth data.
23 You go through an estimation type analysis and 24 come up with a J-R curve for the material. Take that J-R 25 curve now and feed it into what I have called here the I
Heritage Reporting Corporation rm (202) 628-4888 v
257 1 predicted estimation analysis which will be a J-estimation
([ 2 analysis now for a pipe containing a flaw. So you feed that 1
3 information along with the stress strain material behavior 4 and make predictions about pipe behavior.
5 Now in this particular case we have experimental 6 data on pipes. So you check your predictions against what-7 you found in the experiments. In'some cases we also'did 8 elastic-plastic finite element analysis, modeling the 9 experiments and compare the results with what you would get 10 with a simple equation that you used.
11 And based on that, and come up with recommended 12 methods for analyzing each class of problems.
13 You see a little loop over there with a small 14 specimen analysis and that also contains some finite element The purpose there was that when we used specimens
() 15 analysis.
16 that were non-standard, for example, very large CT specimen 17 or the single edge notch tension specimens for which we 18 didn't have much confidence in the simple analysis methods 19 or we developed new methods. In that case we did detailed 20 analysis again to verify that what we were doing with simple 21 methods was reasonably accurate.
22 So there were accuracy checks and validation tests l
23 at all stages. )
1 24 With that as the overall approach what I'm going 25 to have time to go through is basically list some of the l
Heritage Reporting Corporation p- (202) 628-4888 L) l 1
258 1 important topics that we covered during the program and then
() 2' go ahead and show you some specific examples, perhaps not 3 under each one of those topics but some of the more 4 important ones.
5 Key topics. First we go look at the laboretory C specimens. Later on pipe, flawed pipe geometries. In the 7 laboratory specimens there are several studies, and these 8 numbers over here are the codes published on each one of 9 those topics.
10 One is large extension -- large crack extension 11 J-R curves. The kind of thing that Charlie Marschall talked 12 about earlier on the experimental side. We did some 13 calculations and show the results later looking at using J-D 14 versus J-M for large amounts of crack extensions. And 15 again, I want to point out that the purpose of doing that
()
16 was that in a typical small compact specimen you get a J-R 17 curve for very small amount of crack growth. In actual pipe 18 analysis, for example, a through-wall crack analysis you get 19 several inches of crack growth. If you want to make 20 predictions in the pipe you need J-R curve for large amount 21 of crack or how do you get it. There's a need for 22 developing or there was a need for developing a method that 23 you could take a small J-R curve and extrapolate it.
l 24 And this study was aimed at really checking out 25 the different extrapolation methods. What is the reasonable Heritage Reporting Corporation I (202) 628-4888 J
\
)
1 259 j 1
1 way-to extrapolate.
/~\ 2 Surface crack growth resistance characterization. ,
\-) \
3 And the surface crack growth we are talking about crack' 4 growth in the radial or the through thickness action of the 5 _ pipe. Most test specimens tested are for the v circumferential or axial direction of the crack growth in 7 the other two directions of the pipe.
8 We did this study to look at a specimen design 9 which was a single edge notch specimen design in which we 10 put the. initial notch so that the crack will grow'in the 11 through thickness direction. And there is a difference 12 between toughness in the two directions. The hope was to be 13 able to capture that.
14 The third topic, J-R curves for cracks in welds, rw 15 That poses a problem because, for example, the simple ASTM
(-) 16 analysis method for getting J-R curves or any other simple 17 analysis methods assumes homogeneous material.
18 If we are going to look at cracks in weld then we 19 need J-R curves in welds or how do you get it? What is the 20 right way to analyze? We looked into those considerations.
21 To basica,lly check our analysis results with those f 1
22 of others even in the case of finite element analyses we 23 conducted three international round-robin type efforts in 24 which we gave the same problem to several organizations who 25 did it with finite element analyses and estimation efforts ,
l i
Reritage Reporting Corporation (202) 628-4888 O
I l .
260 1 and compared theJresults. As another means of checking our
() 2 analysis results for accuracy.
3 Moving on to the topic on the pipes with cracks we 4 looked at both circumferential through-wall cracks and the 5 main application there being leak-before-break j 6 considerations. Several methods -- several topics covered 7 here. I will go through very briefly here.
8 We had in the case of through-wall crack certain 9 analysis, simple analysis, methods in the literature when we 10 started working on this program. The problem was that there 11 were proposed methods only in the sense that there was no 12 experimental verification with regards to their accuracy.
13 There was some equations available to analyze through-wall 14 crack pipes but not a real good feel how good those methods
() 15 were.
16 Part of the work was to evaluate those methods.
17 While doing so if a given method showed or was found to be 18 less than accurate then we made some improvements on those 19 methods. Sometimes based on experimental results that we 20 were getting or modifying existing methods.
21 So topic B here is the method that existed and was 22 developed at NRR basically at the beginning of this program.
23 We made some improvements in that method and other analysis 24 methods and those two reports show those developments.
25 We did experimental validation with cracks in Heritage Reporting Corporation (202) 628-4888
1 i
261 J
1 welds, both in TIG welds and flux welds. And again as I j
(') 2 mentioned, the consideration there was how do you analyze :j 3 these things using simple analysis methods.
i I
4 Do you have a question?
-5 DR. SHEWMON: When you say you did experimental 6 validation, what was it you validated or will you get to 7 that later?
8 MR. AHMAD: I will show you some examples. We 9 made predictions using analysis methods. For example, the 10 maximum load and check it using experimental' maximum loads.
11 In some cases we also compared the predicted load 12 displacement response from these equations and compared it 13 directly with experiments. Those kinds.
14 DR. SHEWMON: All right.
() 15 MR. AHMAD: Again, some round-robin type analysis 16 efforts to check our calculations against others.
17 We did somo limited amount of work. Most of this 19 work was circumferential through-wall pipes on the pure 19 bending. Some work where you have combined pressure and 20 bending, so you get into the business of non proportional 21 loading, for example. The pressure is applied first and 22 then bending. There's very limited amount of work in that 23 area. ,
24 Again, I want to emphasize that these are not 25 really basic research studies; these are studies saying, Heritage Reporting Corporation (202) 628-4888 O l
262 1 okay, here are all the experimental data, what kind of.
- ( ) 2 simple analysis methods you can come up with to utilize this 3 data in order to predict what goes on in the degraded 4 piping.
5 In addition to through-wall cracks we did some 6 work on circumferential surface cracks and this will be, for 7 example, in-service flaw assessment applications. When we 8 started this program we didn't really have a simple analysis 9 method to analyze surface cracks in pipes, internal surface 10 cracks in pipes and the bending. So we did some development 11 work in putting together an estimation method.
12 And some work done on the complex crack problem 13 where you have a long circumferential crack which is broken 14 through part of the way to look at that problem.
15 Some examples of analysis of laboratory specimens.
{~)
16 Here is as an example a C(T) specimen with a plane of 17 dimension of 25 inch square. A large C(T) specimen used to 18 develop data for large amount of crack extension.
1 19 What we did here was both estimation and scheme 20 analysis and finite element method analysis. This just 21 shows the finite element model of the specimen.
22 We looked at -- our main interest was to look at 23 far-field J as a function of crack growth in these analyses, 24 but we also -- and that larger shaded area shows the J 25 contour crack initially being about here. And a small Heritage Reporting Corporation (202) 628-4888 l
R 263 1 . contour which in this-analysis was moved as the crack grew q
/~% !
(,) 2 maintaining the same shape. And the idea was to get some ;
3 kind of a measure of near crack J because -- for large crack 4 growth' amounts we will expect the J to be in these
'5 calculations.
6 That was, as Charlie had mentioned earlier, was a-7 'part of a. study where we did a series of tests. And this 8 was the largest of the three specimens which was a 10-T and L
9 we also did a 3-T and a 1-T specimen.
10 This viewgraph shows the results of all the three 11 tests analyzed by finite element analysis and the J-R curves 12' developed thereby. You have a J versus crack extension 13 corresponding to the 1-T specimen which goes up to about 14 here and corresponding to a 3-T specimen which goes up to 15 here_and corresponding to a 10-T specimen that goes all the
(~)h u
16 way up to about six and a half inch of crack growth.
17 And these are the far-field values of J.
18 DR. HUTCHINSON: Could I ask a quick question.
19 MR. AHMAD: Sure.
20 DR. HUTCHINSON: In doing the finite element 21 analysis did you input the relationship between load and 22 crack or load point displacement and crack extension?
23 MR. AHMAD: That's right. That was -- well, we 24 modeled the experiment using the crack extension 25 displacement data from the experiment. By modeling that we Heritage Reporting Corporation (202) 628-4888
J
.I l
264 J l calculate J. J
() 2 DR. WILKOWSKI Just one other point of. i 3 clarification. In these series of tests of different size 4 specimens we kept the thickness the same which is a little 5 bit different than what is being done in the reactor 6 pressure vessel low upper shelf concern where they're 7 looking at standard size 1-T specimens and the standard size 8 3-T and the standard size 10-T. In our case the pipe 9 thicknesses or the specimens from the machine to the pipe we 10 can get about 80 percent of the thickness but we can't get 11 it to be any bigger size in the plant form direction because 12 of the curvaturs of the pipe.
13 So we have slightly different problems. We are 14 dealing more towards the plane stress condition here where 15 the pressure vessel concern is more toward the plane stream.
(}
16 DR. HUTCHINSON: So the thickness here was one 17 inch.
18 MR. AHMAD: Here the comparison now of the finite 19 element results and what you would calculate in using an 20 estimation formula given in the ASTM standards. Keep in 21 mind that if you go by the standard doesn't really apply for 22 such large amount of crack growth. But that is what is 23 called JD over here and shown in the dark symbola. The open 24 symbols are the finite element results.
25 And what you see that after -- it's hard to see Heritage Reporting Corporation (202) 628-4888
()
265 1 over here -- but after a certain amount of crack growth you tD,) 2 see a deviation between the JD and the finite element 3- results and you see a big difference and you come out to a 4 large. amount of crack extension. And you would expect that 5 to happen using the deformation J definition.
6 ,
Going back to the finite element results we did 7 input -- we did use incremental theory plasticity 8 formulation, so it was -- we took into account of the 9 elastic loading property. So because of that then you would 10 expect that' kind of a difference between the two. methods.
11 Surprisingly or maybe not so surprisingly, when 12 you compare the finite element results with the modified J 13 formulation of curves we found a good correlation between 14 the'far-field J values and JM even for a fairly large amount of crack extension.
( ) .15 16 And if you consider that JM in an approximate way 17 takes account of the incremental theory type effects on J, 18 then this makes sense. And this.would suggest that GM or at 19 least these re sults suggest that JM may be a good parameter 20 to use if you want to extrapolate J-R curves to a larger 21 amount of crack extension. At least it's consistent --
22 seems to be consistent with the finite element calculations 23 that we adopt.
24 DR. WILKOWSKI: John, one point of concern there 25 is, though, if you look at that last figure. Could we go ,
Heritage Reporting Corporation (202) 628-4888
y 4 ..
\
266 aa,
- f. , 11 backJto your last slide, please. Note that'the 1-T specimen
. ().- 2 does have a'much' steeper' resistance --
3 DR.xHUTCHINSON: Yes.
4 DR. WILKOWSKI: -- than~the 10-T specimen.- So you 5 got to1have -- be a little bit careful in taking small 6 specimen data.and extrapolating it with JM.
.7 DR.-HUTCHINSON: I have another_ question.I was 8 -going to hold off but maybe this is the time to ask it.
9 When you use JM data or JM, when you turn around and analyze 10 a pipe, do you use the analogous. definition of JM in 11 ascertaining, you know, in analyzing the pipe?. I don't knowL 12 if I made myself clear.
13 MR. AHMAD: Yes, you did. Yes.and no. We started-14 out doing that. In simple terms JM is basically you take~
15 the crack growth correction term from the right hand' aide
()
16 and put it on'the left hand side. When you'do application 17 analysis you think'that's what you've.got to do as well and 18 that's what we.did initially. .
19 However, it was pointed out to us mainly by Ernest-20 that that's not really the intention of JM. The intent is 21 basically to collapse small specimen data. When you apply 22 it then you always use the regular J interval, okay, that's 23 one thing.
24 The second thing is --
25 DR. HUTCHINSON: Is that a conservative procedure?
Heritage Reporting Corporation ,
(202) 629-4888
.O
S 267
-i. 1 MR. AHMAD: It is.-- well, I'would expect"-- no, 2" n o', it wouldn't be on the. conservative side.
3 DR. HUTCHINSON: Well, in the. application side JD 4 'is lower than JM, right.-
5 MR. AHMAD: Correct.
6 DR. HUTCHINSON: So it'actually should be L 7- considered.
8 MR. AHMAD: Yes, but when you use an application 9 you apply J --
10 DR. HUTCHINSON: As large.'
11 MR. AHMAD: -- as larger and it will go the other 12 way.
13 DR. HUTCHINSON: It will go the other way.
14 MR. AHMAD: But the procedure right now, whoever
/ 15 is using JM including us is to use just a. regular definition 16 of J when you. apply it. But you use the resistance curve
, 17 with a JD or'a JM.
18' DR. HUTCHINSON: That's inconsistent.
19 MR. AHMAD: Yes, it's kind of inconsistent.
20 However, I think the argument of Ernest was that this is 21 only an approximate way of saying that, okay, take the crack 22 growth term. Actually JM is somewhat different than that. l 23 It's approximate.
- 24 Now there is no equivalent definition of apply J 25 really for a pipe that allows you to do that.
Heritage Reporting Corporation (202) 628-4888
a 268 y 1
1 RDR. HUTCHINSON: I guesssit's also true in the l
() 2 application, the difference between JM if you could 3 calculate it or estimate it and JD is not as great, right, l 4 as in the specimen; is that a correct statement?
5 MR. AHMAD: Well, for small amount of crack 6 extension there is no difference. The difference comes 7 about in very large amount of crack extension. Now for very 8 large amount of crack extension even to apply J -- I mean, 9 it's violating some of the rules using J.
10 DR. HUTCHINSON: Sure.
11 MR. AHMAD: Another difficulty comes about if you 12 want to use something like a JM say in application and 13 supposing you want to do it very accurately, the finite 14 element, you can't really definite JM in a finite element.
15 It's not a contour inte;tal, so that's another difficulty.
(
16 But that's the way these things are done.
17 These data, by the way, have been used and there's 18 a code on that, we don't present results in this fashion for 19 the NRC. We use a lot of th(Je data and analyses and came 20 up with an equation recently. We say how to extrapolate J-R 21 curves from small amount of crack extension to large in a 22 simple equation rather than giving finite element.
23 DR. WILKOWSKI: We also have a draft of a 24 technical paper on that.
25 MR. AHMAD: That's a practical concern. I mean, Heritage Reporting Corporation '
(202) 628-4888
)
269 1- if you're' going to analyze a pipe-on the crack-extension 2 what J-R curve you use. Here is a proposed method.
3 Moving:on to.the next topic of the single edge 4- ~ notch specimen. Again, you have seen this picture before.
5 The crack.-- the important thing is tho crack growth l i
6 direction in this-specimen, this is a thickness area. And' 7 that's one of the things that'you want to look at. It's a
'8 fixed -- excuse me, loading condition.
9 Finite element.model after the specimen. Again, 10 large far-field contour, the near-field contour.
11 'We look at J -- maybe this didn't move. Getting 12 down to the results. We did two analyses of' detailed finite 13 element analyses of two specimens. What.you see is J.versus 14 crack extension for the two different analyses. The circles r- 15 are showing the far-field J. The triangle is showing the V) 16- near-field J.
17 And the purpose in showing this picture is, the 18 point onJone of the things that we found is that you see a 19 deviation between the two contours almost right after crack
- 20. initiation. And that in an indirect way is pointing to the 21 fact that there is very little J dominant growth in these 22 experiment. In other words, if you want to talk in terms of 23 Omega limits then it's a very small Omega limit. In 24 retrospect that's what you would expect because it's a 25 predominately tension type specimen rather than bending.
Heritage Reporting Corporation !
(202) 628-4888
270 !
/
1 The purpose was, again, not to-do just the. finite i I
2 element results analyses but to come up with an estimation !
3 And so we derived an estimation formula for model here.
4 that specimen analogous to the one used in the ASTM 'for C (T) 5 specimens using basically the old method done by guys in 6 Paris. And you come up with a simple equation to calculate-
'7 J as'a function of crack extension.
8 Again, the finite element results are those 9 circles over here and the estimation formula, the simple 10 equation that we came up with give a result which is 11 represented by the straight line here using the same 12 experimental data in both cases.
13 The difference between the two is roughly in the 14 same order that you get by using the ASTM formula and 15 compare-it with finite element results. So the formula is O 16 quite accurate.
17 For those who are not familiar with this method it 18 is basically one of -- J comes out to be basically 19 proportional to the area in the load displacement curve.
20 And so use the experimental load displacement curve, get the 21 area under that, and use that to calculate J.
22 Just for comparison purposes we said, okay, don't 23 use the experimental J load displacement curve but rather 24 use the load displacement curve that you get out of the 25 finite element analysis and then what you get, well, this is l
Heritage Reporting Corporation (202) 628-4888 O
271 1 the kind of result you get then.
I
(~~') ~2 And estimation formula is somewhat in between.
x-3 And the accuracy is again in line with 1ost of the 4 estimation methods.
5 Getting to the round-robin exercise that we did 6 and I show the result. Eight participants in this analysis i
7 round-robin. The only interesting part I guess is here you 8 see -- or nine participants -- four out of the nine are from 9 Japan and only one from U.S. Other from Europe, different 10 places.
11 One of the analysis problems that we posed was 12 again a specimen with large amount of crack extension 13 because we didn't really find much analysis that were done 14 in the literature for very large amount of crack extension.
f^ 15 We thought it's a good round-robin exercise to see how
(_)z 16 others would analyze this.
17 Calculated J-R curves by different participants 18 here. This line of the JD line, that's not a finite element 19 result. This is the JM resistance curve and the rest are 20 finite element results.
21 We later found out -- this is a pretty old result 22 -- that the organization would produce -- these results were 23 actually off by a factor of two, we calculated J over two 24 rather than J. So if you multiply those by two they're in 25 line with the rest of them.
Heritage Reporting Corporation (202) 628-4888 b
272 1 A detail that explains some of the discrepancies
('N
-(_j. 2: as'to how you model crack growth in a finite element 3 calculation. ,
4 DR. SHEWMON: Now is the correct answer to this 5' quiz JM or which of the two J curves is --
6 MR. AHMAD: Well, our analyses show that the far-7 field J.comparee well with the JM. And that's what all 8 these people 5 :,2ying as well.
9 There is one outlier over here and that's what I 10 was going to talk about. In this case it turns out that in 11 the finite element calculation they use a deformation 12 essentially. So it compares with the J.
13 DR. WILKOWSKI: What they did'is have several 14 meshes.
15 MR. AHMAD: Several meshes to solve the problem.
()
16 DR. WILKOWSKI: So they didn't get any history 17 dependents.
18 MR. AHMAD: I think given that, I don't know how 19 many, six countrit 4 involved, I think the results are not l
20 too bad in terms of a variety of codes used, different net 21 sizes. In some cases different models, model the material.
22 Because all we provided was a specimen design and the final 23 result and load displacement data.
24 DR. WILKOWSKI: And stress-strain curve.
25 MR. AHMAD: And the stress-strain curve. But how l Heritage Reporting Corporation (202) 628-4888
9 273 1 they used that stress-strain data in their models could be
,e] 2 different. So that was pretty consistent.
V 3 We also posed a pipe problem, a three-dimensional 4 analysis problem, finite element analysis, through-wall i
5 crack pipe under four point bending. That was subjected 6 just to find out our finite element model for that and ;
7 showing some stress contour.
8 If you look at the calculated J now as plotted in 9 the function of load line displacement, again, most of the h 10 finite element results are pretty much in line. This is the 11 kind of thing, John, that we used to do. We had a j 1
12 definition of JM for a pipe where it says a simple analysis 13 J for the pipe, and this is the kind of difference you would 14 get between the two.
-3 15 So this, if you like, is a simple estimation U 16 result. This is the detailed finite element analysis i
17 result. And what you see is some discrepancy between the 18 two.
19 I want to emphasize that all these simple analysis 20 methods, you do make a lot of assumptions. That's how you 21 simplify the problem down regarding the strains here, 22 applied geometry and so on. So this -- those kinds of 23 results are why in some cases may not be acceptable. j 24 They're expected. It turns out that when you look at the 25 real applications these differences are not really that bad.
)
Heritage Reporting Corporation ]
(202) 628-4888 1 r^g i
%-)
1 l
__,]
%f ,
, 274.:
1; ~DR. HUTCHINSON: So the solid'line.J is from W
V 2 ~ where?-
3 -- 'MR. AHMAD: .T1,ie' solid'line J is'a simple method, 4 information-method of calculating J. This is modified JM.
'5- MR.' SHACK: Using JM specimen data?
~
6 MR. AHMAD:' These are using - -
7 HDR . WILKOWSKI: The pipe load displacementLdata 8 are used'in calculating these J values.
9- .MR. AHMAD: The pipe -- the small specimen.
10 LIf you look at the predicted' load displacement --
1:L ' I think'you were shown this viewgraph in an earlier 12 presentation -- what we found is consistently being shown.is-13 that this.being the experimental load displacement behavior,
- 14 this.one the predicted.
15 Now in the analysis you are inputting displacement O 16 and calculating load. You go ahead and apply the load 17 displacement. That you consistently under-predict. And all 18' of us did that or participating in that round-robin. Since 19' ;this work was done roughly two, two and a half years ago 20 others have analyzed this. For example, Rickstead in 12 1 Sweden, and he is right in with the rest of them.
22 We are looking or starting to look at that 23 discrepancy right now. We have looked at large deformation 24 effects and a possible organization type effects and nothing 25 seems to explain it. What we are focusing right now is that Heritage Reporting Corporation (202) 628-4888
/~
i
275 1 in most'of these calculations,-what we use for plasticity
.)
( 2 models wasLthe --'with the associated floor room type of' 3 ' simple equation. The suspicion now is that perhaps the 4 model that we are using for the model at'that temperature 5' may not be the right one. Particularly we are;looking at 6 .possible discrepancies between yield surfaces and 7 compression. .Because you do have a' bending situation, a 8 large compression reason and tension reason and that.could-9 be a possibility.' So tests are going on right now to-
-10 develop.that. data and then try to explain what's going on 11 here.
12 MR. MAYFIELD: Jalees, may I interrupt. There is 13 a French pipe test, at least one that we have been' told 14 about and apparently we now have those;results, that suggest (J'15- -- for that pipe test I don't see the same thing, that the i 116 finite element results agree fairly well with the 17 experimental results.
18 We also just a month or so ago got word that Fred 19 Neilson took just a solid bar with a circumferential flaw in 20 it and put it in one of his test machines, bent it, and he 21 finds this result: that his finite element results don't !
i 22 agree with the experiment load reflection behavior. So we
-23 are left scratching a little bit. But this little problem i 24 has a lot of attention worldwide and it's leaving a lot of I 1
25 people scratching their heads not quite sure why their best Heritage Reporting Corporation (202) 628-4888
.O 1
i
1 276 1 1~ cut ~ finite element analyses don't do a better job.
2 It's an interesting twist.
i
- 3. DR. WILKOWSKI: We have taken this a step _further .;
4 into:trying to develop this tension and compression data and 5 supply that for a round-robin in the IPIRG program for next ;
-l 6 May to help us shed some light on why this behavior is 7 occurring.
8 So a number of people will be reanalyzing that ,
9 pipe -- the French pipe test. We'11 be analyzing the same
~
10 test that_Bricksted has analyzed. And perhaps we'11 have 11- compression data on this particular material and reassess 12 this particular pipe test as well.
13' MR. AHMAD: Another aspect that we looked at was, 14 maybe the way the model crack growth in finite element and 15 the non-linear release and cracks don't grow by non-linear 16 release. Maybe that's where the problem -- then go back and j 17 look at it, well even before initiation and start seeing a i
18 difference.
19 DR. HUTCHINSON: Where is initiation?
20 MR. AHMAD: Probably about here. Would you say, ,
21 Gary? l 22 LR. WILKOWSKI: Yes.
j 23 MR. AHMAD: You start seeing differences there.
24 DR. HUTCHINSON: So you're seeing differences 25 earlier anyway.
I Heritage Reporting Corporation (202) 628-4888
.O l
i 277 1 MR. AHMAD: Yes.
- .2 DR. WILKOWSKI
- Even before crack growth.
- 3. 'DR. HUTCHINSON: Yes.
4 DR..WILKOWSKI: Charlie Marschall made aui 5 interesting comment, it was that the French material just 6 .even in a tension specimen if you looked to the cross 7 section it'was quite oval in shape indicating that it has 8 some anisotrophy in the material properties that are 9 different than the stainless steel we have been testing.
10 MR. AHMAD: The French got good correlation l 11 between their experiments, their analysis, but for somewhat 12 different material. That maybe again points to the material 13 modeling part, I don't know.
14 DR. WILKOWSKI: Interestingly, one of the I 15 participants up there is the same French organization.
()
16 DR. SHEWMON: Can you say what participant number 17 one did, number two did that the others didn't. They seem I 18 to fall off in the right direction.
19 MR. AHMAD: Well, we do have all those details. I 20 don't recall seeing anything drastically different in their 21 approach other than maybe a finer model, you know, finite 22 element model, something like that. But we couldn't really 23 point to a trend at all of what's going on.
24 DR. WILKOWSKI: When we wrote up the report on j l
25 this, for instance, participant number one results !
Heritage Reporting Corporation g (202) 628-4888 g\-) i 1
l u )
278
.1 1 originally.were almost right on the experimental losd d (f 2: displacement curve.
i 3 .MR. AHMAD: Yes, these were right up here.
'4 DR. WILKOWSKI: They were right up there.
5 MR. AHMAD: .Then they found an error.
6 DR. WILKOWSKI: And everybody was saying, how come ;
7- your agreement is so good. And what they did is, they did 8- not do iterations to check for conversions for the stress-9 strain curve. When they iterated they came down to 10 everybody else's'results.
11 MR. AHMAD: This is being looked at.
12 Moving back to the simple analysis methods which 13 was our main charter by simple analysis methods. But those 14 provided some benchmark solutions that we could judge our
(~T 15 estimation analysis results against.
V:
16 Some of the analysis methods that were used and/or 17 developed, the GE/EPRI estimation method of Kumar and Shea 18 used the method as originally was developed. And the 19 original method -- when doing it you know there's the -- is 20 that J is written as J-elastic and J-plastic, some of the 21 two. But the J-elastic also contains some plasticity 22 correction. And we said, well, you counted plasticity 23 twice, let's take that out and use that as a method. i
)
24 EPRI -- new GE/EPRI handbook edition on the 25 estimation methods and several other modifications. A Heritage Reporting Corporation (202) 628-4888
f m
o 279-4 il method proposed by Paris, this is the method-also known as' q
( 2 the Clocker method as being in NRC' analysis.
3-- .Two. methods here proposed by Bud Brust. Basically .)
4 improving on the LBB NRC method which assumes'a perfectly.
5 plastic material but try to. incorporate straight and 6- hardening, that-kind of an approach. And he also used the-l7I R-6b method that's popular in Europe, mainly'in Great 8 Britain.
9 For circumferential surface cracked pipes, we o 10 didn't really have a simple method so we made one up. I 11 think I'm missing _a viewgraph somewhere. It may be in your 12 handout but not here.
13 DR. WILKOWSKI: No.
'14 MR..AHMAD: It's a picture. No. -I guess the way 15 this is arranged is, I could show the results'of the 16 through-wall crack simple analysis methods and then get..on 17 to"the surface cracks.
18 If-you look at the prediction of say maximum load 19 and compare it with the. prediction load,. now with simple 20- estimation analysis methods this is plotted as a ratio. So 21 if you are larger than one and the measured load is larger 22 your quote on the conservative side for this kind of an 23 application. Plotted against pipe diameter and for
^
24 stainless steel.
25 And you see that the predictions using different Heritage Reporting Corporation (202) 628-4888
h 280' 1 methods.- 'Three listed here: the'LBB NRC; Paris method; GE 2- ' method. Again,;the; solid dots now.are using'the J-D-R
{ ,
3 curve,.the open ones using:the-J-M-R curve and you pretty 4- much expect-the way they would move. Using the higher JM 5 resistance curve we will get results'which gives..you a
'6 higher. predicted load.
7 So in general for most experiments you will see 8 this kind of results. In some cases it's quite.large, but 9 in general on the conservative side.
10 This'is for carbon steel, the same kind of 11 results. There are some outliers over here using the JM L12 resistance curve which is the open circle.
13 So depending on the method that you use and the 14 resistance curve, the JD or the JM you've got to come up fs 15 with a combination that most of the time will give you
' R16 conservative-result. The one thing to point out here is 17 that, don't look at it as a direct comparison within the use 18 of JD and JM in terms of final result because there is also.
19 an additional approximation involved through the simple 20 estimation analysis J, okay, that kind of messes up the 21 picture.
22 But here is, okay, given JD or JM, given this 23 method what do you get as a result and compared with the 24 experiment.
25 So in general -- also I must point out that as Heritage Reporting Corporation (202) 628-4888 bb l I
l u_____________ _ . _ _ _ _
281-
'l soon as this work was done and the results are reported in
() 2 .the final report there's some statistical analysis done on 3 all the various predictions and different methods and 4 different JD and JM curve would say, okay, what kind of 5 result that you would expect in general on the average. And 6 those results are reported in there.
7 For example, one of the finding is that if you use.
8 the GE/EPRI method with a JD resistance curve you're always 9 conservative. However, you may beloff the mark in terms of 4
10 accuracy quite a bit.
11 Comparison between predicted load displacement 12 curves, now using the simple method versus the experiment.
13 What you saw earlier was the finite element result. What 14 you see is that, maybe it's not a big difference here 15 between -- if you look at the error between experiment and
()
16 analysis they're about the same order as in the finite 17 element analysis. So estimation scheme is maybe not too 18 bad. And these are different estimation procedures used.
19 Another example of that, perhaps a better one, the 20 previous one where you saw maybe a typical result. If you 21 are lucky you get a better comparison with experiments.
22 Moving on to the surface crack estimation model.
23 In order to come up with an estimations scheme which wasn't l
24 simple. Our approach was to -- well, first of all, we l
l 25 looked at the crack -- the surface crack in the pipe and the 1
Heritage Reporting Corporation p ,q (202) 628-4888 l V f
A
--_--___________o
L I 7-I 282 1 pure bending with.a flaw integration like this. In general ;
(~N 3
(,) 2 the method, it turns out that we develop doesn't restrict 3' you to this flaw shape. You could have an electrical or any j i
l' But most of the experiments were done 4 of the flew shape.
1.
I 5 using.this kind of a flawed geometry. So that's what is 6 shown here.
7 The approach was to look at the center, cross-8 section across the center here and that would look like 9 this. The pipe isn't bending which means the stress 10 distribution on that segment would be somewhat like this. A 11 predominately tension -- non-uniform tension but tension.
l 12 This is a two dimensional problem versus a three 13 dimensional problem here. You can find simple estimation.
14 solutions for this problem in, for example, the GE/EPRI Use that solution and try to integrate it while
() 15 handbook.
16 satisfying the capability and equilibrium, that sort of 17 thing, come up with the solution for the surface crack.
18 The nice thing that you get out of this is that 19 you use existing two dimensional solutions in order to 20 generate a three dimensional solution for the pipe, without 21 going into detail.
22 The final result again, compared versus experiment 23 here. Plotted out over a two day show of the different pipe 24 sizes used. Now the thing is reversed here and it's 25 predicted divided by measured, so everything below this line Heritage Reporting Corporation (202) 628-4888
iT
- 1 1
283 1
1 one is on the conservative side.
( ) 2f Using that simple estimation formula you make i
3 predictions. In this case initiation load for the surface 4 crack problem. And the events you see over here represent 5 different ways you could use the small specimen property 3 1
6 data, JM versus JD, or if you use -- the curve fit, how do -1 7 you use it. There's some different ways you could interpret 8 this small specimen data, would give you this kind of a 9 scatter in the prediction. That's what those bands -- and 10 you see that in general that you're making conservative 11 predictions. Again, sometimes off by quite a bit compared-
-12 to the experiment result but on the conservative side.
13 The same thing you see if you look at the maximum 14 load predictions rather than the initiation load prediction for the surface crack pipes. And in this case most results
( ). 15 16 are' conservative but maybe as much ac 35 percent or so lower 17 than the measured values.
18 I want to point out that this -- the report on 19 this has been written but the evaluation of the method that 20 we came up with is limited to looking at our experimental 21 data which if you remember are pretty much fixed flaw sides, 22 67 percent along the circumference, fixed depth. The method 23 is not being evaluated against data because both data don't 24 generally exist.
25 For different flaw sizes, different aspect ratios Beritage Reporting Corporation (202) 628-4888 O
__-_____________ - _ _ - i
284 1 of the flaws and so on. Also there has not been yet a
() 2 thorough evaluation of the prediction versus finite element 3 calculations. The reason being that the finite element 4 calculators can't agree among themselves yet. So you don't 5 really have a good baseline. So maybe taken as a good 6 preliminary method that you use because there's nothing else 7 available but not totally verified yet.
8 Also I must point out, all this thing was done for 9 pure bending. Now on the surface crack case you may 10 typically want to look at combining pressure and bending.
11 That extension of this method to combine pressure and 12 bending is pretty straightforward but is not being done yet.
13 What is the bottom line after all these 14 calculations, what do you come up with as a deliverable?
(~T 15 What we come up with is a PC base computer code called
%-)
16 NRCEIPE in which you basically input the material property 17 data from small specimen testing. You input the flaw 18 geometry and height sizes that you want to look at. And the 19 code is set up basically for LBB assessments, so you're 20 looking only right now for through-wall crack assessments.
21 This code contains most of the estimation schemes 22 that we validated. But the document that goes with this 23 code also provides you with an assessment of which kind of 24 guides you in a sense. Which analysis method to use in a 25 given situation rather than giving just a handbook to Heritage Reporting Corporation g (202) 628-4888
'%-) .
1 l
l l
285 1 solutions. And outcomes the prediction of pipe fracture and
() 2 deformation behavior. That's deliverable of the code.
3 And this, as I understand now, has been delivered 4 as NRCPIPE computer code user's guide and a floppy disk. We 5 missed the commercial for that program.
6 DR. SHEWMON: Does that run on an old-fashioned 7 8088 or do you prefer a more advanced chip-or what kind of a 8 PC is necessary?
9 MR. AHMAD: Bud Brust.
10 MR. MAYFIELD: I think it can rely on anything, 11 but it will be slower on earlier machines. From personal 12 experience, on the 8088 machines it depends on how many 13 years you want to wait on it. Yes, it will run.
14 DR. SHEWMON: Okay.
()'15 DR. WILKOWSKI: In the documentation of the code 16 we supplied a sample calculation times for the different 17 methods with the different types of computers just so you 18 can get a feel for how much time you might spend on making 19 the calculation, one method or another method or whether 20 you use a 386 or an old PC.
21 DR. SHEWMON: Is it set up to use an 8087 or math 22 co processo ?
23 DR. WILKOWSKI: Yes.
24 DR. SHEWMON: Did you have one of those in your 25 machine. Okay.
Heritage Reporting Corporation (202) 628-4888 O
hw-,---
' l 1
286 1 MR. MAYFIELD: Well, we've also run a version
(_j 2 without one and it makes a difference. )
3 DR. SHEWMON: -
Then it's decades.
4 MR. MAYFIELD: Then it's decades.
5 MR. AHMAD: Well, to summarize I don't want to 6 just give you -- there are a couple of things that I think 7 need a little bit more looking into as we didn't get into it 8 too much. But the fact that we found in looking at the 9 single edge notch specimen that you lose J dominates very 10 quickly saying that right after initiation this question-11 over there, you can really use J for that kind of a problem.
12 The fact that that specimen was meant to mimic a 13 surface crack in a pipe also raises some questions, well, 14 what kind of a J dominates you have in a surface crack in a e'
() 15 pipe. Can you really use with confidence a J-R type 16 analysis for those problems.
17 From a practical viewpoint the difference between 18 initiation and max load in surface crack problems is not 19 much, okay, In other words, maximum load is reached only 20 after a small amount of crack extension and small addition 21 and load.
22 However, if you're looking at applications where 23 you want to look at the breakthrough of a surface crack and 24 then subsequent growth as a through-wall crack, in those 25 class of problems it may be important. There is some I I
Heritage Reporting Corporation f S (202) 628-4888 (u .
287 1 Japanese work coming out recently also raising same kinds of
() 2 concerns abotc the appropriateness of assuming an HRR type 3 f2el in surface type problems. That's one thing.
4 The second thing is, maybe some more analytical 5 work in the area'of cracks in welds. What we have done L
6 right now is more or less empirical correction to existing 7 methods, but not a really detailed study of cracks in welds.
8 And third and maybe a very important one is the 9 detailed analytical assessment of what Charlie Marschall j 10 initially called as the straining -- the dynamic straining. -h 11 It looks like analysis modeling point of view. It's a
- 12 combination of grade and temperature, strain grade and 13 temperature and that perhaps needs to be looked at in some 14 detail. With that I'll shut up and let Gary present his.
15 DR. SHENMON:
() "7es the strain aging behavior tend 16 to make the sample absot* more work, more energy raise J 17 or --
18 MR. AHMAD: Well, the answer it depends. From the 19 data I have seen and because there seems to be a crossover 20 at some temperature strain aged combination where the yield 21 is going up and then starts going down, straight down. So 22 it depends where you are in that space. 'That would depend 23 on the material I guess.
24 So you can't really make a common -- a general t,
25 comment like that, that's my understanding. But it looks Heritage Reporting Corporation (202) 628-4888 O
l l
288 1 like as the strain ~ grade really grows to the crack, in n
( ).
2 combination with where you are in the temperature side of )
3 things, okay, which will determine what happens. I don't 4 ~know if it is the dynamic strain aging type phenomena.
5 Typically hard the strain. rate, you know, I would 6 expect less plastic deformation of the crack test. The data 7 I have seen that can change depending where you are with 8 temperature. I 9 DR. SHEWMON: Okay.
10 (Slides being shown.)
11 DR. WILKOWSKI: I will continue on with the 12 closure of the discussion of-the degrade piping program and 13 .in doing'so -- I've got the wrong viewgraph in there. But I 14 want to talk about the state of the art developments. Some
() 15 of the applications and where they apply, where it's 16 regulatory policies. Program deliverables. And the last 17 area that I had to discuss was just responding back to some 18' questions that you had raised from the last meeting.
19 As far as the : bate of the art developments this 20 was a viewgraph that I think was in the first package of my 21 handout but I thought it was perhaps more appropriate to 22 discuss it here at the end.
23 At the very start of the degraded piping program 24 where were we versus where are we now. For instance, at the 25 start of the program the net-section-collapse or limit load Heritage Reporting Corporation (202) 628-4888
(
289 1 analysis was developed. It was verified on small diameter r~s 2 pipe in the program that we had done for EPRI on stainless b At that time about four, five years ago GE/EPRI 3 steel pipe.
4 estimation schemes were'under initial development. There 5 were one'or two solutions for pipes but nobody really went 6 in and did a detailed evaluation of that estimation scheme.
7 The NRC's LBB or NRC LBB analysis method.was just 8 being developed by Ray Clecker. At that time there was very 9 limited material property data that existed about four years 10 ago and the amount of pipe fracture data that existed was a 11 very small amount. Generally a small diameter pipe, and so 12 we didn't have much of a database to really check a lot of 13 the applications.
14 Now where are we at the end of the program. The s 15 material property database has been greatly expanded. We 16 have developed, I think, in the neighborhood of 150 J-R 17 curves on a variety of different materials and a comparable 18 w.aount of tensile test curves. All of that data of load 19 displacement crack growth and the stress-straining curves 20 have been incorporated into the NRC's PIFRAC database, so we 21 just filled up a portable hard disk drive and sent it over 22 to MEA and had them incorporate it into that database. j 23 A detailed database on circumferential1y cracked 24 pipe under slow monotonic loading has been developed. In l
25 doing -- developing that database we collected loads. We "
Heritage Reporting Corporation (202) 628-4888
('
d
, 290 1 collected low-line displacement, crack growth, but'also 2 crack opening in case'anybody wanted to use this for
~
3' verification of leak-rate calculations in the future.
4 We tried to collect as much detail as we could or 5 as much data as we could so that these experiments would be 6 heavily documented so that any new analysis procedure could l
7 be assessed. They wouldn't have to go back and redo any of 8 these experiments.
9 We verified and developed new analyses for leak-10 before-break and through-wall stability. We verified the 11 GE/EPRI analysis. You have seen that. We did some work 12 with a method that Ray C1ecker had initially developed at 13 .NRC. We have developed two of our own methods in addition 14 to that.
15 One of the bottom lines in our final report, which 16 Jalees didn't discuss as 'such and one of the questions that 17 you had was, what do you use JD or JM in doing these type of 18 calculations? Well, it depends on the estimation scheme 19 that you use. For instance, in general the GE/EPRI 20 estimation scheme is the most conservative technique 21 compared to all.of the others. If you use JM with a GE/EPRI 22 estimation scheme you get pretty good predictions compared 23 to the pipe fracture results.
24 However, if you were to use JM with any of the 25 other methods you might over-predict the experimental loads.
{
Heritage Reporting Corporation j (202) 628-4888 i
( )-
1 l
a=--_--_---_____-_
1 1 L 291 1 If you were.to use just JD with the power lock extrapolation rx ;
l_) 2 type of technique then we found that the methods that Bud- i 3 Brust has developed in this program, LBB, ENG, and GE 4 method, give us reasonably good results. That is they 5 predict about 93 percent of the maximum loads. And the 6 variation in the predictions is only about 10 percent. So 7 those are pretty good results.
8 So you could use either JD or JM but you've got to 9 be careful what estimation scheme you use for the structural 10 analysis.
11 We've also, as Jalees.was talking about, had done 12 an initial development of a finite length surface crack 13 elastic-plastic engineering estimation scheme. This was 14 really just the first step in that. Nothing existed before
() 15 the program when we started to look at finite length surface 16 cracks and came up with a nice idea, Carl Poplar first 17 originally was a consultant to us and came up with a 18 procedure to use in the thin wall procedure and then Jalees 19 extended that to a thick wall shelf solution. There is 20 still improvements that can be done to that.
21 We have assessed the accuracy of finite element 22 analysis. Jalees showed you some of the results on C(T) 23 specimens and through-wall cracks in pipes. We also do the 24 similar thing for a surface crack in a pipe. And for a I
l 25 surface crack in a pipe we see that the opposite trend is to Heritage Reporting Corporation (202) 628-4888 1
i
L i l 292 l
1 tend to over predict the experimental loads rather than 2 under-predict the experimental loads.
{
,}
3 The third bullet there briefly discusses the fact 4 that we've got another method that we didn't have time to 5 get into. Like so many things on this program, when you l 6 have been working on something for five years and try to 7 discuss it in a morning time frame it's difficult to do.
8 But we have developed an energy balanced method 9 that would help us to predict not only when an instability 10 starts but-to approximate if that crack is going to arrest.
11 So for instance, if you were interested in determining 12 what's the maximum credible leakage area for a surface crack 13 in a pipe or a through-wall crack in a' pipe after a loading 14 event you could approximate that from this method.
- 15 I use the term " approximate" because it's not 16 highly accurata because there are strain rate effects that 17 we're neglecting in such an approximation.
18 And this really is based on the fact that you need 19 from the estimation schemes predictions of the moment 20 rotation behavior of the surface crack. You have to predict 21 the moment rotation behavior of the through-wall crack and 22 know the compliance of the piping system. If you know those 23 three key things then you can start using an energy balance !
24 to give you an upper bound of the crack jump. Or you could l
25 make a lower bound estimate by not doing an energy balance Heritage Reporting Corporation (202) 628-4888
() )
293 1 approach.- So you can bound it.
X
(,) 2 Using that same moment rotation type of ;
J 3 calculation you could input that moment rotation curve as a 4 non-linear in your spring element into a' piping code. And f 5 by that then you could start making assessments of cracked i
6 piping systems under inertial loading like in the IPIRG 7 program. And that's essentially what we did in the IPIRG l 8 program for designing the test system. It was used in that 9 non-linear spring from either experimental data or from 10 estimation schemes to go into a simple piping dynamic code.
11 And so now we're doing types of calculations there in that 12 program that were unthinkable five years ago where you're 13 trying to estimate what happens to a surface crack in a pipe 14 under elastic plastic loading conditions and now.we can do 15 it within a couple of hours, ten hours of computer time on
(}
16 just a simple work station computer; whereas, we thought gee 17 you need a couple craze to do this five years ago, 18 Let's look a little bit at some of the application 19 of the results, the in-service flaw acceptance criteria 20 specifically for the stainless steel criteria. What we 21 found is that the base metal criteria and the weld metal 22 criteria for low toughness welds gives a better than average 23 prediction. That is if you took their criteria and took 24 away the safety factors they predict better than the average 25 of all the experimental data. There's a few points below Heritage Reporting Corporation (202) 628-4888 !
()-
h, - _ . - _ _ _ ._ _ _ . . _ _ _ _ _
i l 294 l
12 what.they would predict-without their safety factors, but ;
2" their safety factors more than accommodate those 3 differences. ,
l4 We did find some inconsistencies in the ASME flux 5 weld criteria. -For-instance, they use too high of a 6 ' toughness for shield metal arc welds and submerged-arc welds 7 compared'to. data that we had.
8 They tended -- in those calculations they also 9- used.the weld metal strength in doing their calculations 10 rather than the base metal strength, which from our results 1,
11 show that should be on the'non-conservative side. However, 12 they used conservative estimation scheme for'the pipe that-13 they could use which seemed'to counteract the other non-14 conservative aspects. And that more than compensated for it 15 so that overall their safety factors were quite good and 16 reasonable.
17 Charlie Marschall showed you the results of an 18 experiment that we did with the side edge tension specimen 19 where the crack was growing along in the heat affect zone 20 .and in one specimen, on part of that specimen it went into 21- the fusion line. We see that that fusion line toughness was 22 about half.the toughness in terms of crack tip opening angle 23 of the submerged-arc weld, whereas before we thought the 24 submerged-arc weld was the lowest toughness that we could 25 have. So perhaps we need to look at that a little bit Beritage Reporting Corporation (202) 628-4888 O
i
4 .
295 1 closer in the future.
fl' t,,j 2 A final point was that we have done a number of 3 experiments on artificially aged cast stainless steel, some 4 pipe that Bill Shack had aged at Argonne. Some pipe that 5 Framatome had aged and sent to us. And what we found from 6 those results is that they reached the predicted loads by 7 the ASME criteria. Now the ASME criteria for flaws in 8 austenitic piping says it can be applied to cast stainless 9 steels if the ferrite number is 20 or less.
10 And for the steels that we looked at, the cast 11 steels that we looked at they were right around that 20 12 number. And we found out that their toughness was 13 sufficiently high, that they essentially reached the limit 14 load criteria that was in the ASME code for stainless
() 15 steels. And perhaps the worse case would really be the L 16 welds which would be much lower toughness than the aged 17 material that we examined.
18 Bill, do you want to comment on that any further?
19 MR. SHACK: I think that's true. That wasn't a 20 worse case aged cast material, but that's basically l 21 consistent. Most of the aged cast stainless are not going l
l 22 to be ae bad as the welds. There's only going to be a few 23 and even those few are probably not going to be a whole lot 24 worse than the levels.
25 DR. WILKOWSKI: The one that we saw that was far l
Heritage Reporting Corporation (202) 628-4888 O
]
o i
I 296 )I 1 worse in a French paper had a very high ferrite number I c's l i
(_) 2 compared to what's applicable in the ASME code for cast 3 pipe. ,
4 DR. SHEWMON: Was it a worse case weld?
5 MR. SRACK: We don't know.
6 DR. WILKOWSKI: Some application from NUREG-0313 7 for weld overlay repairs, we found that the criteria that 8 they suggested to use within 0313 for weld overlay repairs 9 were quite well, quite satisfactory. Also our weld overlay 10 repair pipe test, as you saw, there was tremendous amount of 11 deformations that occurred adjacent to the weld overlay 12 repairs in the end cracked pipe section.
13 We also did a mini round-robin where for that 16 14 inch diameter pipe test that had such large deformation, we 15 sent out the test conditions to NUTECH and GE and a bunch of
{
(])
l 16 other people on the ASME code and said, predict what the 1
17 loads are by code procedures and you can also use elastic-18 plastic fracture mechanics if you want. And we found out i
19 that when they used the 0313 suggested procedures that they l 20 were all within a couple of percent of each other, which 21 also e. greed well with experimental data. So that gave us 22 pretty good confidence with that procedure and those 23 results. If they used elastic-plastic fracture mechanics 24 the scatter increased. But if you just use the limit load 25 it goes right out.
Heritage Reporting Corporation (202) 628-4888 O
297 1 Currently the ASME code is getting approved the r
(_yj 2 article IWB-3650 for analysis of flaws in carbon steels. In 3 this particular case we,had developed a large amount of pipe 4 fracture data that was used in the initial development and 5 verification of their procedure.
6 We found in our work that some submerged-arc weld 7 procedures can have a very low toughness.- Not all of them.
8 Some of them can be very high at toughness. And the low 9 toughness submerged-arc weld that we had found was used as a 10 lower bound in the ASME code criteria. It had a J-1-C of 11 about 350 inch pounds per inch square.
12 A lot of our Charpy and J-1-C data were used in 13 the screening criteria that.they had developed. The 14 question about anisotrophy we talked about that a little. bit 15 in our work on anisotrophy. We also found that the
()
16 toughness for the axial crack direction was significantly 17 lower than it was for the circumferential crack direction in 18 the carbon steel pipes.
19 Originally in one of the drafts of this procedure 20 they assumed that the toughness in the circumferential 21 direction and axial direction was the same. However, ue 22 pointed out some data from pipes we had tested and pipes we 23 sent to MEA and was in the PIFRAC database and they 24 subsequently revised the criteria for axial flaws in pipes.
25 And we had conducted a couple low toughness pipe 1
1 Heritage Reporting Corporation i
! (202) 628-4888 !
l ($) I i
298 1 experiments, this is underway. This slide was made at the
-( f '2 Water Reactor Safety meeting, we conducted those experiments 3 since then and found that for this very low toughness weld 4 procedure that the failure stresses of the pipe were quite 5 well above what was predicted by the ASME criteria.
6 For leak-before-break applications we had some 7 earlier contributions to 1061, volume 3 methodology report.
8 We have made some minor improvements to the LBB NRC method 9- by having more linear elastic solutions which were really 10 Sander solutions that we put in terms of K instead of his 11 energy release rate parameter.
12 Then we developed some additional through-wall 13 crack fracture estimation schemes that gave us some better 14 accuracy.
15 We have developed this NRCPIPE code, that can be
( })
16 used in licensing applications. And we have developed a 17 pretty large database that can be used to assess generic 18 lower bounds or look at trends in data for different 19 materials. And developed the pipe experimental data that 20 has been used to make leakage area predictions in other NRC 21 program such as the IPIRG program.
22 Our final program deliverables, we've had eight 23 program summary reports one of which is the final report l 24 that you have in draft form. Had 11 topical reports. We j i
25 had an extensive amount of data reported on all of our i Heritage Reporting Corporation
-) (202) 628-4888 J
l l
l
. _ _ _ . _________________A
299 1 experiments as well'as some experiments from past EPRI 2 programs or from David Taylor ResearchLCenter. The PC-
- 3. computer code is mainly for circumferential'through-wall ,
4 cracks which would be-used in'LBB fracture analyses. And
.5 -also contributed a lot of material property data to the 6 NRC's PIFRAC database.
7 Switching _now very briefly, just a couple quick 8 viewgraphs. The last time we met'there was a couple of 9 ' questions that were raised. One question was the need to 10 further_ evaluate why crack grows in a helical direction in 11 some of the ferritic piping. Some:of the things that we
- L2 have done is we have documented the toughness as a function 13 of crack orientation inLone of our reports. As a matter of 14 fact, a lot of' people come back to that report only for that 15 appendix, just to see what the effects are.
.O 16 We found that in-seamless pipe the use -- where 17 the management process is used to fabricating pipe that the 18 pipe is actually twisted so that the working direction is in 19 the helical direction and that's how the inclusions are 20 oriented which is the low toughness orientation. )
i 21 Also when they fabricate the pipe by this process 22 they use a cross rolling process with rollers that will cold 23 work the material in a helical path at the same time, so you
.i 24 'have two effects.
25 DR. SHEWMON: If you assume the crack will run i
Heritage Reporting Corporation ]
(202) 628-4888 j (2) i
__E_-_____________.____ _ _ _ _ _ _ _ __ ___ _ _ _ _ _ _ _ _ _
,' l 300 1- circumferential1y with your analysis does this alwaysE--
j 2 usually.-in a; tensile and that you always end up with'
< 3- conservative estimates. That is if you-don't take account I l
4 'of the' fact that1this crack runs hither and beyond. . )
[ f5 DR. WILKOWSKI: 'For the case of pure bending we 6: have reasonable predictions. There could be I think --
7 DR. SHEWMON: By reasonable you-mean it fits it 8 well or --
'9 DR. WILKOWSKI: Fits it well. Slightly 10 conservative.- Okay. But not overly conservative. -Again, 11 'you get into the issue as to what estimation scheme of JD' L
12 and JM. And I think we can handle that and get something 13 that is slightly conservative but gives reasonable accuracy.
14 at the same time.
() 15 A possible question might be, well, what happens
{- 16 if you have combined loading such that now the principal 17 stress direction is normal to that helical direction. And 18 that's something that we haven't really investigated at this 19 time.
20 DR. SHEWMON: This would have to be a substantial i 21 torsjonal load. !
22 DR. WILKOWSKI: It could be a torsional load. It 23- could be just the combination of hoop stress for pressure.
24 with longitudinal stress giving you a principal stress 25 direction. Perpendicular to the helical end.
l l
Heritage Reporting Corporation l (202) 628-4888
.O
'I
i
.301 s
1 MR. MAYFIELD: Gary, can I interject something.
2 We are just now putting together a statement work and'
.{
3 starting to circulate it through the NRC, management, and to 4 our colleagues in licensing, looking at some of the things 5 .we found in the degraded piping program that appeared to 6 need more work. 'This is one of the areas. .And it's 71 exactly, what --.is there some realistic combination that 8 could cause a crack to run off in this helical direction.
9 DR. WILKOWSKI: And'we've had some other' examples 10 where cracks have turned because of anisotrophy effects.- As 11' I explained a little bit yesterday on the. tour, we had a 12 cold-leg pipe test where in this particular case the pipe 13 had an axial seam weld so the~ low toughness direction was in 14 the' axial direction.
15 We had a circumferential crack in the pipe under 16' pure bending, no hoop stress but the crack grew in the axial 17 direction, four or five inches first and then it went in the 18 circumferential direction.
19 HPA has seen similar anisotrophy effects in 20 Germany in some of their tests where they had an axial crack 21 in the pipe. However, the low toughness direction was in 22 the circumferential orientation. What happened, as soon as 23 t. hat axial surface crack broke through it got to the end of 24 the surface crack and it turned into circumferential j 25 direction.
Heritage Reporting Corporation (202) 628-4888 O
302 1 So material anisotrophy in an elastic-plastic
(~) 2 fracture can have a significant effect on the failure mode L
3 that occurs. .
4 '
And as I mentioned before we've tried to account 5 for some'anisotrophy effects in the ASME ferritic pipe 6 ' criteria for axial cracks.
7 Another point that was briefly mentioned before is 8 we had done a very small -- made a very small comment in-one 9 of our reports where we've noticed that the amount of 10 nitrogen that was seen in some of the submerged-arc welds 11 seem to correlate with the lower toughness of the weld. .j 12 That was also an observation from some work that was done in 13 the Section 11 Task Group where they found that the lowest' 14 toughness specimen that was ever tested for a weld from
. 15 David Taylor: had been subjected to metallographic 16 evaluations and chemistry evaluations by TVA. The only 17 thing they found out that was unusual about that weld is it 1
18 happened to have a high amount of nitrogen in its chemical 19 composition.
20 And we found in a few of ours, we said, gee that's 21 interesting, why don't we just take a more detailed chem l
l 22 analysis and we found that when the nitrogen was higher that 1
23 we tended to have a lower toughness weld. That was more of 24 an observation. It wasn't the intent at that time for us to 25 do a detailed evaluation of why this has occurred. And that l -
Heritage Reporting Corporation (202) 628-4888 L _ _ _ _ _ __ _
(- e ,
l 1
303 f
1 work'was documented in our; report on stainless steel flux f 2L welds.
3 Another point,was the need to assess why finite-4 element analysis of through-wall crack pipes under predict 5 experiment loads.- Jalees discussed this a little bit and I 6 don't.think I have to belabor any of these bullets because 7- we have gone through all of these points'already. ;About the l
8 fact that the Swedish:results can under predict-the load 9 displacement curves and the IPIRG program-may evaluate 10' constitutive modeling behavior explaining this. And with 11 that I'm'done.
12- DR. SHEWMON: A couple other questions. One, I 13 was going to call Everett Odebaugh on, some of you guys made 14 almost as well. There's a line in here in the executive
( f 15 summary of your piping report that says, the grade six pipe 16 can be either seamless or electric resistance weld of the 17 toughness of the electric resistance weld it can typically-18 be one-tenth of the lower bound or one-third of the lower 19 bound toughness used in IWB-3650 procedure for ferritic 20 steels.
21 So you know how much electric resistance welded !
a '
22 grade six pipe is out there and what sort of things would 23 come in, that sort of systeras? l
-24 DR. WILKOWSKI: Do yov want to enswer? Yes, I can 25 do it a little bit.
Heritaga Reporting Corporation (202) 628-4888 i
304 1 MR. MAYFIELD: Okay.
() 2 DR. WILKOWSKI: That's specifically for A333 grade 3 six pipe. That the specification for that particular pipe l 4 material allows one to either fabricate the pipe by. seamless 5 procedures or by an electric resistance seam welding 6- ' procedure.
7 DR. SHEWMON: .I'm familiar with how you make 8 electric resistance seams, that's not my. question.
9 DR. WILKOWSKI: I was just providing a little 10 background. And we had found from some work that we had 11- - done for the American Gas Association that these electric )
12 -remistance seam welds have extremely low toughness compared 13- to the base metals.
14 - Within the ASME Section 11 code I raised -- at the
() 15 - committee meeting I raised this concern and the response 16 that I got back from some of the members who are cognizant ;
17_ of what they do when you fabricate a plant was, gee, when we 18 order 333 grade six pipe we really don't know if it's ,
19 . seamless or electric resistance seam weld.
20 I think one thing that perhaps should be done is 21 just doing a survey of manufacturers to see if they really l
'l 22 do make 333 with electric resistance seam welding and what j l
23 size of pipe ranges might they use it in. I l
24 DR. SEEWMON: Is it viable for safety systems? )
I 23 MR. MAYFIELDr ASME 333 grade six and there is no l Heritage Reporting Corporation (202) 628-4888
.O I
1
-._.________._._______mm__ ___.._. _m __
i h .
L 305
' s 1 distinction.
2 DR. .SHEWMON: I' don't know A333 grade six frcm a 3 tin;can. ,
L 4 'MR. MAYFIELD: It's a cleaned up version of A106
'5 grade B.
6 MR. SHACK: It's widely used in GE BWRs, for l- 7 example.
8- MR. MAYFIELD: You will find a lot of it in L 9 boilers. Substantially less in PWR.
10 DR. SHEWMON: There must be a limit on how thick a 11 wall'you can weld, isn't-there?-
12 DR. WILKOWSKI: Yes. Also in pipe diameter _too.
13 MR. MAYFIELD: By and ihrge, what you're talking
.14 .about apparently are' going to be 10 inch.or less. Probably 15 more in the four and six inch diameter where you will find (16 this. And then you're talking schedule 40, schedule 80.
17 DR. SHEWMON: This is then the potential axial 18 flaw which could run from here to the next elbow.
19 MR. MAYFIELD: In principal, yes. One of the-20 things that Gary said is that the ERW welds are lower in 21 toughness They can be very good welds. They can also be !
22 real bad welde. From the line pipe er.perience, we don't 23 know. It can be a real problem. It can also provide a very 24 good piece of pipe. That's something we're trying to sort 25 out. It's not being ignored; it's just not solved yet.
Heritage Reporting Corporation (202) 628-4888
306 l
1 DR. WILKOWSKI: The newer procedures of l
() 2 fabricating ERW pipe uses a high frequency welding procedure 3 which has been used I think for about the last 15 years.
4 They give a relatively good pipe quality welds as far as 1 5 free from defects. Their toughness is still lower than what 6 we would like for the base metals. But at least the quality 7 of the weld is better.
8 DR. SHEWMON: One other question. Back earlier, ;
9 results from complex crack pipe experiments show that even a 10 shallow surface crack adjacent to a through-wall crack can 11 significantly lower the apparent fracture resistance of the 12 pipe. What comes to mind is that the pipes that get l 13 evaluated, the flaws that get evaluated under these things 14 are -- I'm tempted to say -- never continuous; they usually 15 get called continuous because there's a series of these
(( )
16 things and they may overlap. Can that lead to any lack of 17 conservatism then or just how big is this a factor? What 18 does the secondary crack have to be before it makes this 19 difference you're talking about?
20 DR. WILKOWSKI What we saw from our results -- I 22 don' t have a viewgraph of it -- but essentially we found 22 that the effective or apparent material toughness that we 23 calculate from pipe tests with thesa surface cracks going 24 around the inside in a through-wall crack can lower the l whole resistance curve by either -- it's kind of a I 25 l
1 Heritage Reporting Corporation I (202) 628-4888
307 1 constraint condition. That is a function of how deep that n 2 crack is.
V 3 It's interesting that the whole resistance curve 4 . changes in proportion to the depth of that crack.
5 Suggesting it's just a geometric term.
6 DR. SHEWMON: Now both of these cracks are coming 7 in from the same side.
=
8' DR.'WILKOWSKI: I see a couple blank ones here.
9 A complex crack, for instance, would be something 10 that looks like -- that's supposed to be round pipe. You 11 may have a flaw that's all the way through and some surface 12- crack around the inside. So this is flaw area.
13 And if you put the pipe in bending what you might 14 see is that the crack wants to grow like this. It will
,, 15 follow along that internal flaw or discontinuity. Quite a 16 bit more you get a constraint of large amount of plasticity 17 that occurs.
18 What we found was that if you look at the J from 19 the complex crack versus the J from a simple through-wall 20 crack which is typically what we analyze a leak-before-21 break. That's a function of the depth to thickness of this 22 internal surface crack. And it goes something like this.
23 Nhereas about 10 percent of the wall thickness, okay, you i
24 could see a reduction from the simple through-wall crack j 25 geometry of in the neighborhood from 25 to 50 percent. We 1
Heritage Reporting Corporation ]
(202) 628-4888 O ;
u__u_______________________.________
308 1 just have a scatter of the data in that region..
r 2 DR. SHEWMON: Oh, okay. I.didn't understand what 3 you said then. You're saying that if you have something 4 that you call a complex crack that can crack much -- ;
-i 5 DR. WILKOWSKI: Much quicker, much earier because 6 you have an apparent reduction in the materials fracture 7 resistance from constraint of that surface crack.
8 DR. SHEWMON: When you said adjacent to I. thought 9 you meant they ran parallel to.
10 DR. WILKOWSKI: No , it's in the same plane.
11 DR. SHEWMON: Fine. Okay, that's all the 12 questions I have.
13 DR. HUTCHINSON: One question. I know that when 14 you talked about the single edge notch specimen'you said you
() 15 didn't do too much with it. Did you.do any test to -- on 16 the same material to compare the J resistance curve you got 17 from that specimen with the J resistance curve you were 18 getting from compact tension specimens?
19 DR. WILKOWSKI: Yes, we did. We did a number of 20 them that way. And we saw that they were fairly comparable.
21 DR. HUTCHINSON: They were fairly comparable.
22 DR. WILKOWSKI: Yes. There wasn't that much 23- difference when we used the far-field J values from the 24 finite element or the estimation scheme, and then compare !
i 25 that to the C(T) specimen using the ASTM procedure. Not too I
i Heritage Reporting Corporation f- (202) 628-4888 V)
O t
309 L
l' much different there.
() 2' DR. HUTCHINSON: What about ~just the standard l- 3 reduction of the data using, you know, the area under the 4 curve type formulas.. Again,~do you find that those two 5 classes of resistance curves were more or less? Actually, 6 I'm not_-- wanting to even hear'about the large amounts of 7 crack growth but just moderate amounts of crack growth.
8 DR. WILKOWSKI: That's all we can see in that 9 specimen, just small' amounts.
10 DR. HUTCHINSON: So you found that they were more-11- or less comparable.
- 12. DR. WILKOWSKI: More or less comparable.- There 13 was.a couple that were a little bit lower and a couple that 14 are higher. And on the average they're within the,.I think,
() 15 three materials. Not terribly different.
16 Of course you know one of the theoretical-17 objections that you have is that the near-field and the far-18 field deviate credit initiation which doesn't-give you a 19 warm feeling about analyzing either of that type of specimen 20 or surface crack in a pipe. But we have a practical 21 problem, we have to analyze a surface crack in a pipe. So 22 perhaps using the far-field is a good engineering v
23 compromise.
24 DR. SEEWMON: Okay, fine. Thank you.
May I mak'e one note about the 25 MR. MAYFIELD:
Heritage Reporting Corporation (202) 628-4888
1 310 1 report you have. It's a draft. That report has been 2 submitted for publication and the copy you have is what will
(~}
s 3 be published. ,
4 DR. SHEWMON: I'm sorry, it's finalized.
5 Can you do anything with half an hour or will it 6 just break up your whole presentation continuity.
7 Why don't we take a few minuto break. Pardon me, 8 Joe.
9 MR. MUSCARA: I was going to suggest that Steve 10 take.one of the shorter topics towards the end of the 11 program and do it now because I think a number of people 12 would like to leave a little bit earlier than what the 13 agenda shows.
14 (Slides being shown.)
/^ 15 MR. DOCTOR: I'm going to discuss and describe a G}'
16 program entitled Steam Generator Group Projects. The only 17 emphasis in this particular presentation will deal with the 18 NDE aspects and code activities. This was a rather large 19 program. It extended over six years. It was an NRC project 20 and there was participation by the EPRI, France, Italy, and 2J Japan. It basically studied the Westinghouse Surry steam I
22 generacor type from 2-A.
1 23 The regulatory issues that were going to be 24 addressed in this are the adequacy of the NRC regulations 25 governing the frequency, extent, cnd procedure for ISI. And Heritage Reporting Corporation (202) 628-4888 '
O
311 1 then the adequacy of ASME requirements.
l
()T 2 The project objective shown here with regard to 1 3 NDE.was to determine the reliability and the effectiveness 4 of conventional and what they classified as the near-term l l
5 field practice eddy current techniques to both detect and to )
6 size degradation occurring in the tubes.
7 And furthermore, as an impact to develop input for j l
8 the revision of the ISI Reg. Guide 1.83; and two, plugging ]
l 9 Reg. Guide 1.121. '
10 The process I've -- just for one second. This is 11 a Westinghouse model 51 steam generator. It was the one 12 that was employed in this study. Effectively the only 13 portion that was involved was this lower portion here 14 containing all the tubes.
()
15 The baseline inspections were conducted. There 16 were basically two teams that were involved in that: one 17 from Zetec and one from Intercontrole. They made 18 predominately the equipment employed to perform eddy current 19 examinations.
20 And the idea of conducting these baseline studies 21 was to look at a number of aspects of selection of tubes to 22 be involved in the round-robin studies that were being
'23 discussed and going to follow on in the program.
24 And when they conducted these they were expectino 1
25 to see a large amount of overlap. They found that the best s Heritage Reporting Corporation
.S (202) 628-4888 i V
l
312 1 case detection cgreement was at about 70 percent. And the gg 2 reported size of indications was significantly different 3 between the teams. But,this information did serve as a 4 basis for the tube selection for the round-robin studies.
5 Now there are basically four round-rotins that l 6 were conducted. However, in this discussion we're going to 7 focus on the first one, the data acquisition and analysis 8 round-robin that dealt with the reliability of current 9 multifrequency eddy current bobbin-coil type of round-robin.
10 And then looking at the advanced alternate type of the 11 round-robin where you looked at not only improvement in 12 terms of eddy current equipment but also there were some 13 ultrasonic probes that were also used in the inspection of 14 the tubes.
15 The data acquisition and analysis round-robin is 16 shown here. There were basically five very experienced 17 teams that had done many field examinations that were 18 involved in this round-robin. They used the same identical 19 equipment. The same set of tubes were inspected in all 20 cases. And the same inspection procedure was used by all 21 five of the teams.
22 There are a number of copies of the handout that 23 are going around. I brought 20 so there should be plenty 24 for everyone.
25 The NDE validation objectivos are shown here.
Heritage Reporting Corporation (202) 628-4888 9
c 313 l1 What they chose'to do was to select 550 specimens to be
)
~
2 removed from the steam generator. They wanted to use these
.3- to confirm the' reported indications; to measure the defect 4 severity; to identify if there are defects that were missed' 5 -by the eddy current examinations; to take some of the 6 segments and do actual burst tests on those to validate to 7 ' integrity equations; and to document and validate all of the 8 ' findings from the study.
'9 DR..SHEWMON: How hot were these tubes-and how 10 much could you clean them up without changing defects?
'11 MR. DOCTOR: I think the decontamination procedure 12 was quite effective after removing much of the -- because
- 13. most of this is surface contamination' problems that they had 14 to deal with. So we were able to remove that and there was
.()15 :not that much effect since most of the kinds of defects that 16' were in the generator were pitting and wastage. Although 17 there are other defects. I suspect based on other things
-18 they do not have much of an effect.
19 This is a s'ummary of the different types of 20 defects that were found. As you can see there is a variety l21 of different kinds of defects. The row one and row two, for 22 example, could not be inspected because simply there was too 23 much bending and so there's a separate round-robin that was 24 conducted on that. We will not discuss that today. !
25 The results are really going to pertain primarily l
l 1
Heritage Reporting Corporation (202) 628-4888 O
- - - _ _ _ _ _ _ _ _ _ _ _ _ _ _ - - )
314 1 to the pitting and wastage since these were the predominate 2 kind of defect, although in the study all of these defects
{~
3 stress corrosion, et cetera, the cepper rich deposits, all 4 of these were in the set of tubes that were involved in the 5 study but the bulk of the indications were the pitting and 6 wastage.
7 What we mean by that is shown here. This is an 8 example of a cross section of one of the tubes. And you can 9 see here the amount of pitting and wastage on this 10 particular tube. And of course when you would go through 11 and evaluate a tube like this and come up with an estimate 12 you would look for the smallest remaining ligament as being 1 13 the greatest amount of degradation in the tube.
14 Now what I wanted to discuss for a moment is the 15 detection probability here versus the metallographic wall 16 loss. And what we're trying to do is to come up with a 17 lower bound performance. What they chose to call this, this 18 90/90 lower tolerance limit. What that means is that 19 they're 90 percent confident that 90 percent of the teams 20 are actually above this particular line. ,
1 21 If you lock at this you can see there are a total
'22 of seven teams that actually were involved in this data 23 acquisition and analysis round-robin. The results are shown 24 here. You can ses a scattering if you follow any particular 25 team as X here. You can see sometimes they're high and ]
l Heritage Reporting Corporation (202) 628-4888 )
O
l 315 1 sometimes they're low relative to the other people. So ,
lh 2 there's no trend there.
3 However, you can see here that this lower bound 4 starts out around 25, goes up, and then plateaus up here, 5 and then they put it around 65 percent through-wall. By 6 beyond this there were not that many available specimens in 7 the -- specimene that actually were up here. This is 8 somewhat of an extrapolation of the data because just the 9 limitations of the specimens that they did have.
10 This shows the same kind of a plot by selecting 11 one of the teams. I happened to be this team five. This 12 particular team was the overall best performer in terms of 13 the inspection both in terms of detection and sizing. This 14 is the results that this particular team actually was able lll 15 to achieve. And you can see here that their performance got 16 up in the neighborhood of around 50 percent through-wall.
17 They had 100 percent detectability.
18 MR. MUSCARA: The same problem exist, there is 19 very few data points with the large flaw sizes. This is the 20 actual data not the 99, low tolerance level.
21 MR. DOCTOR: So I suspect the air bars on this go 22 somewhat upward like that.
23 Now this is a plot of performance in terms of --
24 it's typical of most of the teams where they came up with an 25 estimate of the actual wall loss versus the actual Heritage Reporting Corporation (202) 628-4888
316 1 metallographic and evaluation of the two.
rx
(,) 2 You can see here that there are a number along 3 these two axis. As you,can see these are indications that 4 would result in defects that one simply would classify as a 5 closa call. And down here are ones that are simply missed, 6 classified as actually being zero in size.
7 If you look at this you can see there are quite a 8 few Xs that are at that particular axis. There are a number 9 of points, systematic trend though in this data is actually ]
10 to do an undersizing. So one has to conclude that that is a 11 generic trend. And if you look in fact at the best team you 12 can see that even the best team had some false calls. They 3 13 missed a number. But for the ones that they did actually 14 detect you can see they had a much better performance. A
<s
() 15 much tighter clustering of the data around the ideal 16 performance line right here.
17 If one looks at kind of generically all of the 18 data for both the baseline and the data analysis and 19 acquisition round-robin you can see the data is somewhat 20 almost scattered around.
21 There is a general trend, however, again when you 22 look at this for an undersizing, a substantial number of 23 indications that are close calls and that turn out to be 24 actual defects that are missed.
25 This is the same kind of a curve showing the Heritage Reporting Corporation e- (202) 628-4888 (j)
L 317 1 results for the advanced teams. You can see a somewhat
/~'s
() 2 better in terms of an improvement There's not nearly as 3 many false calls nor as many missed defects.
And you can 4 also see that this data does tend to cluster somewhat better 5 towards'the idea of curve line.
6 What I would like to do now is move on and 7 basically go through a series of conclusions that have come 8 out of this validation exercise on the round-robin.
9 One is that the copper-rich deposits affected the 10 eddy current inspection by creating false calls and in 11 possibly masking tube degradation. For example, there was 12 some wear due to the vibration bar that was plated over by 13 the copper-rich deposit and this was clearly missed. So
'14 that copper-rich deposit very clearly masked that kind of
() 15 degradation.
16 The high probability detection for pitting and-17 wastage was found for defects that were greater than 40 18 percent through-wall.
19 For stress corrosion occurring there was a low 20 probability detection.
21 The automated data screening seemed to improve the 22 detection for pitting and wastage. Team number five thLr I 23 showed you, this was one of the things that they did. They 24 employed an automatic system and screened all the data to 25 identify the areas that they would then do analysis on. So Heritage Reporting Corporation (202) 628-4888 i
318 1 that appeared to be a significant thing that provided
/~g 2 improvement with regard to the section of indications or
()
3 degradation. ,
4 Eddy current generally underestimated the pitting i
I 5 and wastage, especially for the severe degradation that you 6 saw from those scattered plots that I put in the last few 7 viewfoils.
8 There was a wide variation in the eddy current 9 response both in terms of team to team variability, but also 10 in terms of tube to tube. Two different tubes that had the 11 same kind of wastage. There was a very large dispersion of
)
12 the data with regard to performance on those even though 13 they had very similar kinds of defects.
14 There was a complex defect morphology. For
<~s 15 example, dents and deposits that made the interpretation N_]
16 quite difficult. There was a position in many cases where 17 you had several different things occurring simultaneously 18 and it complicated the analysis, it did not look like the f 19 general types of eddy current signals that most analysts are 20 accustomed to evaluating. 1 21 The analyst interpretation of these complex eddy 22 current signals was decided as being the largest source of 23 variation in the reported depth for the specimens. They 24 would look at these and pull out different features and use !
25 those and this resulted in the variability that was found or i
Heritage Reporting Corporation !
(202) 628-4888 l
<-)
- t. ,
cm- - - -
~319
- 1. 1 _ helped to be the cause of'that.
.rm .
(_) 2 The improved sizing accuracy was noted for the one 3 team. They.used a special frequency mix to improve the 4 signal-to-noise ratio. They also used ultrasonic'and 5 rotating probes to augment their conventional bobbin-coil 6 type of inspection. So they did a number of things. The 7 more data that they had the better able they were to apply 8 that and come out with an accurate answer.
9 One of the things that one would like to use the j 10 information from the study for is to give you guidance in 11 how do you find degradation as occurring in a steam 12 generator. So that was one of the things that was set out.
13 And the way'that they approached this was to use the eddy 14 current Surry reliability data to model the eddy current
( ) 15 sizing there. And what they called the probability of 16 exceeding the eddy current plugging limits, so it's P 17 basically E and L. So it's probability of exceeding the 18 eddy current plugging limit as a function of the true defect 19 depth.
20 They also used the same information up here to 21 model probability of detection as a function of the true 22 flaw size. And then you estimated the ranges of both of 23 these to evaluate and compare various kinds of sampling 24 plans, sampling intervals, the amount of sampling that 25 actually occurs on how effective defective tubes were Heritage Reporting Corporation (202) 628-4888
320 1 detected and plugged. And what we're going to do in the r% 2 next few viewfoils is step through some of that analysis
.V 3 with regard to applying what was actually gained from the 4 Surry-program and how that impacts sampling plans.
5 What is showing right here is a plot of the true 6 depth of a defect versus the, if you will, the eddy current 7 response in terms of estimating the through-wall extent of a 8 defect by eddy current information.
9 This here is the 40 percent plugging limit. And 10 what's shown here is, kfyouwill, kind of the mean value 11 for defects at various depth. If one would go out here to a 12 true depth, for example, right here.of around 75 percent and 13 you go and take all the data for that type of defect'from 14 the Surry and plot what kind of through-wall extent each of
- q. 15 the different inspectors pr)vided you would come out with
\J 16 some kind of a distribution function that is shown here.
17 The mean value is located right here.
18 And you define then this PEL as being the. total 19 area of that curve that lies above then the plugging limits.
20 So it's the number of people who actually looked at this and 21 actually classified it as being greater than 40 percent.
22 In the case of the Surry program this turned out 23 to be for the 75 percent defect a value of around .73.
l 24 Ideal would be of course 1.
L 25 MR. MUSCARA: May I interject a comment. Recently l
Heritage Reporting Corporation (202) 628-4888
<~T l ks/
321 1 we chose 75 percent,-we are identifying that as being a
(~ ,
(_,h/ 2 defect of 2. 75 percent is a flaw size that if it grows 10 i 3 percent in the inspection periods it would fail under loss 4 of secondary pressure. And 85 percent tube will burst. 85 5 percent defected tube will burst under 2,000 psi. And then 6 we also put into it 10 percent for growth during inspection j 7 periods. So that size --
8 DR. SHEWMON: Are these usually pits or are they 9 extended wastage?
10 MR. MUSCARA: The 75 percent through-wall depth 11 because of bursting. The wastage and cracking and pitting, 12 they're all about the same kind of relationships, so it 13 could be any of those.
14 DR. SHEWMON: Well, you said earlier most of them
/-~~
{) 15 were for pitting and pits if it leaks you don't care so 16 much, it sort of puts out a small stream of waste that gets 17 bigger.
18 MR. MUSCARA: That's right.
19 DR. SHEWMON: If it's a longer defect --
20 MR. MUSCARA: We found it once -- once you get a l l
21 flaw that's above tube diameter, regardless if it's more --
1 22 the burst fractures are the same, fractures are depth. With !
i 23 the pitting there is some strengthening factor and we have I 24 also those in the equation. And the pits do in fact, 25 because they're short, have a higher burst pressure. So the l
Heritage Reporting Corporation rm, (202) 628-4888 V
i
322
'7'5 percent is based on an infinitely long flaw.
1
'T 2 DR. SHEWMON: Yes. Okay, go ahead.
i(d..
3 MR. DOCTOR: That's all I have to say on that 4 viewfoil.
'5 The strategies that one can use for the 6 inspection, of course, is set out and accomplish this goal; i 7 namely, you want to be able to identify and plug all the 8 -defective tubes. .l 9 one can do 100 percent inspection and look at 10 every one of them or one can go through some kind of a l
11 sampling scheme and look at a smaller subset of that 100.
12 . percent. There is obviously a tradeoff there in terms of
, 13 the amount of time to perform an inspection, the cost 14 associated with it, et cetera.
15 If one looks at 100 percent one can potentially 16 you would say find all of the defected tubes. But let's j 17 look at what in fact one is doing. When you're actually 18 detecting you are actually looking at the POD which I just 19 showed you earlier. For the detection performance you found 20 this lower limit value that most of the teams were above.
21 However, even looking at those values you were p 22 looking at numbers like 65 percent. You're going to j 23 multiply that by the -- not only the fact that you detected i 24 it, but then you have to actually size it correctly so that 25 you in fact will be able to make the correct decision that Heritage Reporting Corporation (202) 628-4888 O
\J l
I
_ _ . - - _ _ _ _ _ _ - _ = _ _
g" .. .,
s
-323
- 1. it exceeds that threshold value and go" ahead and plug'it.
h '2' ,
.So this is really what you wanted;to; achieve, You
'3 want to detect everything ideally, have a'.value of one in' 4- there, and you also want this.PEL value'to be, as I said, a 5- value of one. However, if you'look at.what I showed you on 6 the previous viewgraph the number from the Surry data showed l 7 that that value was about .3.
8- If you go to a sampling approach things change'in 9 a minor. amount and that is the first term here. You not l 10 only have to detect it, but that particular tube has to.be 11 in the sample set that'you're actually inspecting because if 12 it isn't there'is no way that you can, in essence, 13 accomplish what you're trying to do. .In other words, find 14 that defected tube l'E you're not. going to look_at it.
15 ' o the goal then is to look at this performance -
16 and we would'like to have that performance come as close as 17 practical as 100 percent inspection. And so what.you would 18 like to do is to look at some various ways'of comparing 19 these numbers, finding out what 100 percent inspection 4' 20 provides you and what some various sampling schemes might 21 provide' yoi2.
22 If you just take a value of .9 for the POD and the 23 value of .7 and plug those into the pre-setting equation, if 24 you're doing a three percent random inspection you would 25 then have .9 times .7 times .03 which would give you a value Heritage Reporting Corporation (202) 628-4888
1 324 1 here of .02 which is quite low.
() 2 50 percent random sample .9 times .7 times .5, 3 .s32. If you did 100 percent of the inspection it would be 4 .9 times .7 or a maximum value of about .63.
5 So even looking at this kind of an approach doing l
6 a 50 percent random sample is only half as good as doing 100
- 7. percent inspection. And 100 percent inspection is not l
8 perfect, it only finds 63 percent of all the cases that one 9 would like to be able to find.
10 The way that it was decided to approach this was 11 to look actually at the Surry team in terms of whether some 12 particular aspects of the way the defects that were 13 occurring that could allow you to give some guidance in 14 terms of how you selected your actual sampling plan.
15 And one of the things that came out of this was Iw )'
16 that they did not have isolated single defects. The defects 17 appeared in some form of a cluster. So the approach that 18 was taken was to say, well, I've got some kind of a 19 degradation mechanism going on. I will have, for exaruple, a 20 defective tube, one that has a through-wall defect in it 21 that exceeds this 75 percent, but surrounding around that t 22 will actually be degraded tubes, tubes that do have 23 degradation from the same mechanism. So that if I go 24 through and do a sampling and I'm using this kind of 25 cluster, if I in fact find one of these and detect it then Heritage Reporting Corporation fs (202) 628-4888 U
-_______--_----- l
325 g
1 'what I would then put into my sampling plan would be then
~T' 2' that I would have to start increasing the sampling around (V.
3 where that particular defective tube is located until I 4 found tubes that in fact were not defected.
5 The idea of taking this cluster approach is that 6 it's a conservative approach or felt to be conservative 7 because of the fact that the degradation is occurring in 8 clusters, at least in Surry. And it was felt that this can j 9 be extended to other steam generators based on data that has 10 been obtained primarily on terms of looking at the kinds of 11 defect signals that have occurred in clusters of tubes. So 12 it's not something that is not without a technical basis.
13 One of the things that also that this cluster 14 assumption and the motivational factor that was employed is
. 15 that one didn't want to just jump out and do 100 percent t
16 because traditionally people have only been doing about a 17 three percent inspection. A three percent inspection and 18 . jumping to.100 percent is something that would be hard to 19 get industry I believe to do. I think there would be a lot 20 of resistance. And as a consequence one would like to find 21 something that would be equally effective.
22 If you assume this kind of a cluster and one then
- 23. would go in and do 20 percent sampling one should include at 24 least one of these tubes in your sample set.
25 When we actually did some additional work looking l
Heritage Reporting Corporation ,
(202) 628-4888 I
g 4 I f j
'326 l' at this the actual systematic sampling intervals-that were
~2-; selected was a 20 percent one and~a 40 percent one.
3' .When.the eddy; current indication:due to.the !
- 4. degradation *is observed, and I.said this. earlier, we're 5 . going to expand the inspection. We would like to build-
'6 actually a two tube buffer zone around that defective area.
7 What that means is, if I find -- as in this previous case --
8- this defect, I would' expand out until there were two tubes 9 in this direction, in this direction, in this. direction, in 11 0 this direction,'in which I had no indications from. Then 11' -they would consider that I had actually. included all of the 12- defective' tubes in that particular sampling.
13 So whenever one is found the sampling then 14 dictates that you expand until you have that two tube-
' buffer.
'( ) 15 .
- 16. Each tube with an. eddy current indication 17 exceeding the plugging limit is assumed to be.pluggedfor 18 repaired. That is part of the analysis that-has been 19 assumed.
20 Now this shows a plot of 20, 40, and 100 percent 21 inspection sample. And in this particular case where i
And 22 plotting the probability of plugging a defective tube.
23 what we're saying here is that this is at PI and D. In i l
24 other words, if we're doing a sampling we have to include. 1 1
25 the defective tube in our sample. And furthermore, we have j i
,. Heritage Reporting Corporation (202) 628-4888
.O L
i - - - --- - - - - - - _ ---
327 1 to be able to detect it.
(~) 2 This is the probability exceeding the eddy current v'
3 threshold. And you can,see if you look at this that we've 4 actually got three pairs here. But what you will notice is 5 that 40 percent and the 100 percent lie basically one on top 1 i
6 of the other. What this says is that there is no net 7 benefit as long as this assumption about clustering is in 8 fact valid between 40 percent and 100 percent inspections.
9 But there is a substantial difference between the 20 percent 10 and these other two cases, t
11 Conclusions from this work are that if one 12 actually goes to 100 percent inspection based on the data of l l
13 Surry that you would end up plugging 65 percent of all of 14 the defective tubes. So even 100 percent doesn't get you to l
,~)
, 15 where you would like to be in terms of this performance. l
~
I 16 And 40 percent systematic sampling, what we're showing here, i
17 is that the same is effective as 100 percent as long as you !
18 are applying this assumption about clustering and expanding j 19 the sample plan around any defective tube that you find.
20 I have a couple more viewfoils. One of the things
( 21 that's also being looked at is doing some Monte Carlo 22 simulations. These computer simulations are being performed 23 to supplement the work that has juct been described and 24 evaluate the performance of sampling scheme when clustering 25 does not hold, if in fact, that assumption of clustering is Heritage Reporting Corporation (202) 628-4888
(~
V)
- - - - _ - _ _ - _ _ _ _ _ _ _ _ _ _ _ _ I
328 1 not valid..
p.
'(_) 2 What they want to look at is what the effect is on ,
\
3 POD,. sizing, plugging limits, and of course, false calls.
4 The conclusions that have been reached to date, 5 there has been a variety of all of those parameters that 6 have just been described, have gone into looking at a whole 7 series of different Monte Carlo simulations. The 8 conclusions from that work are the simulation result 9 support, the analytical work when clustering assumption is 10 approximately valid. !
j 11 'All sampling inspection schemes are equally j 12 effective when the defective tubes are surrounded by a large J
13 number of degraded tubes. In other words, if in fact that 14 is valid. No matter what you do as long as one of those
() 15 tubca gets involved in your sampling then you're going to be 16 able to find it. I 17 The 40 percent sampling is not as close to 100 18 percent inspection when defective tubes are not surrounded 19 by degraded tubes. However, 40 percent is still better than j i
20 20 percent. And the reason for stating that is that -- my 21 understanding is that industry is pushing for a 20 percent And what we're saying from our data is that )
22 sampling plan.
23 40 percent is systematically better in all the cases we have 24 looked at.
25 DR. SHEWMON: There is some rule in place now that Heritage Reporting Corporation (202) 628-4888
7______ _
~
a
- {
'3 329 1~ ~ says that.you find:with a smallfsample'a certain number of
- ( ) 21 defectsEyou have to expand the. size of your sample. Is
- 3. there.anything in the' regulations now that says how you 4 ' choose those?? -Is it a' cluster? Are'you supposed to choose 5 them nearby.or completely across'the board?
- 6. MR. MUSCARA:. There is nothing in the regulation.'
7 But generally it's done near the area where we've located
?- 8 the defects.
9 MR. DOCTOR: The last.viewfoil then is one that 10 describes and summarizes the work in~ regard to Reg. Guides 11 and code activities.
12' The'first one up there on revised Reg. Guides 1.83 13 and 1.121 have been prepared and have been reviewed by NRC L14 ' staff.
j{ f 15 The draft value impact study to support these Reg.
16 _ Guides has been completed and currently is undergoing NRC 17- review.
18 Under ASNE code there has been participation in 19 the very special ASME working group on eddy current p 20 examination for revision of appendix IV on the development 21- of performance demonstration criteria.
22 This was originally following the one on 23 reliability where I was going into detail about the 24 performance demonstration. And the bottom line is that' 25 section 11 is going away from prescriptive requirements for 1
Beritage Reporting Corporation
. (202) 628-4888 t
j
)
1 r
330
~1 a11Lof the inspection. They're going to performance
) -2 demonstration, that people have to be able to work with' 3 their equipment' and procedures and demonstrate the'ir '
4 .capabilit'y to'do that. And this is also being. applied'in
'5 'the area of eddy current analysis.
6' There has been input into this-appendix IV from 7 the' steam generator. There's work that has been funded by-8 the NRC downfat Oak Ridge impacting things'such as the rate 9 that one uses.for recording information as well as linking 10 the frequency. response of your systems that you probe travel 11~ speed.. The specific requirements.that have been developed 12 out of that work that's just going into that.
13 This special working group right now on 14' performance demonstration,.there's an agreement that'that.I'
() 15 believe should be done. 'The disagreement is the fact-that 16 some people feel that needs to be done on a plant-specific 17 type of basis versus a generic basis because the feeling is ').
1 18 that in some plants we only have.this kind of a' defect and 19 that's the one that the requirement demonstration.needs to 1
- 20 be done on. And we don't care if you can find other kinds !
21 of defects. So that's an issue right now that's being 22 debated.
23 And another issue is in terms of do you do the 24 entire. system or is the analyst the only one that has to go- .
25 through the performance demonstration. And that's another ]
Heritage Reporting Corporation ;
(202) 628-4888
331 1 issue that's being debated.
(~T
-(). 2 So although there is some consensus that 3 performance demonstration is the direction to go., the way 4 that's going to actually come about and how that's going to 5 be implemented at this time isn't clear. That's still being 6' under a large amount of discussion with industry.
7 DR. SHEWMON: How do they reactor to your cluster 8 assumption in trying to cluster the sample or haven't you ,
9 gotten into that yet?
10 MR. MUSCARA: Steve wasn't involved in the work, a 11 he's just reporting on it. But the reason for going to the 12 cluster, of course, was --
13 DR. SHEWMON: I understand that, that wasn't my 14 question.
O 1s .ua ausc^a^= rue reeoeio= 1e eaet e=1te ortea eae 16 assumption holds true, sometimes it does not. Now in the 17 Monte Carlo simulation we evaluated a number, 12, 13 l
)
18 different flaw maps where the cluster assumption is not 19 holding true. So we've had tubes that are defective and 20 nothing else wrong in there. And what we find is that ;
21 although the result is not quite as good the 40 percent 22 sampling still is almost as effective as 100 percent.
23 So you only need very few. You don't need to have 24 the four surrounding, even with one or two surrounding it 25 becomes quite effective, the 40 percent. No, 20 percent is Heritage Reporting Corporation f's (202) 628-4888 tj l _
l l
1 332
.1 very much poorer than-the 40 percent.
2 DR. SHEWMON: Okay. Fine. Is that the end of I 3 this study? ,
j 4 MR .~ DOCTOR: 'les, it is. l t
1 5 DR. SHEWMON: Why don't we.take a break and try to j l
6 start'about 1 o' clock. I l
7 (Whereupon, at 12:20 p.m. a lunch break was taken 8 to reconvene at 1:00 p.m. this same day, Thursday, March 16, 9 1989.)- ,
, 1 10 11 12 13 14 h 15 16 17 18 19-20 21 22 23 24 l 25 !
Heritage Reporting Corporation (202) 628-4888 i
--_m_._mm___-____________.m__ _ _ . _
I b 333 1 AEIEENQQH EE'EEIQH L( )-) 2l (1:00Lp.m.) ;
3 DR. .SHEWMON: ,Where do we go from here?
'4 (Slides'being<shown.) 1 5 MR.-DOCTOR: What I will describe now is a program H
6' entitled Evaluation and Improvement of NDE' Reliability.for i
'7 In-Service Inspection of Light Water Reactors and we have 8 shortened this to NDE Reliability Program.
-9 . The program objective is shown here. 'Trying.to '5
. l 10: ~ quantify the reliability of in-service' inspection technology 11 and techniques for light water reactors. Specifically the 12 primary system components through independent research-and 13 establish means for obtaining. improvements in the H14 reliability of those in-service' inspections.
The kind of regulatory issues that we're
( ) 15 16 addressing andLI'll talk quite a bit about the first one in 17 regard.to our code activities-as qualification criteria 18 ~ requirements for in-service ultrasonic examination.
19 Dealing'with the effectiveness and adequacy of 20 current inspection requirements.
21 And then they're technique evaluation and 22 improvement for a variety of different components and a- 23 degradation mechanisms as delineated in this list.
24 Now the kind of things that we do and how this i i
25 program facilitates and'holps nuclear regulation. It Heritage Reporting Corporation (202) 628-4888
334 1 provides on-call expert consultants in the area of NDE.
p).
(, 2 Provides a resource'for' fast response to
- 3. regulatory. requests. ,
4 Provides third party evaluation of industry 5' practices.
6 Provides assistance in moving NRC research results
~, into national codes and standards.
8 And develops needed technical data basis for
.9 supporting positions. Those are the kind of things that i .
10 come out of the program and were involved.
11 The kind of regulatory products that are affected 12 by this research are shown here. One area is the mandatory 13 Appendix VII on NDE personnel training and qualification 14 that officially passed the entire code process in the latter 15
.() part of '88.
! 16 This mandatory Appendir VIII cin UT/ISI performance 17 demonstration. The status on it 1s that it passed main 18 committee effective February of '89.
19 We're working on developing UT equipment operating 20 tolerance parameters. I'll describe them in a little more 21 detail in the presentation in a few moments.
22 We're also looking at developing requirements for 23 surface roughness preparation for ultrasonic in-service 24 inspection. I do not plan to describe that one in this 25 particular presentation because it's something that is Heritage Reporting Corporation (202) 628-4888 g-)g L l l
h 335 I
1 started and underway but there's not a lot of progress to
(,A) 2 report on, so I decided not to include it in this 3- presentation. ,
4 The last one'I will be discussing is new criteria-5 for UT/ISI inspection programs.
6 The program is laid out in this manner. And I 7 will be talking about really activities involved in the code
.8 activity piping inspection, new criteria, and also the 9 pressure vessel inspection.
10 For this particular presentation I have chosen to ;
I 11 highlight and discuss in detail our cast stainless steel 12 that we're doing. Talk about the code activities, in 13 particular qualification. Talk about UT equipment operating 14 tolerances. Our involvement in the PISC program. And then 15
() the new criteria.
16 So let's move on and talk about the first one here 17 on cast stainless steel. The problem with cast stainless 1 l
18 steel with regard to.the inspection of it is that it's a 10 very anisotropic material. It has very large coarse grains i 20 that are not homogenous. As a result when you try to 21 propagate sound through it, it brings the sound up. It does 22 things like it steers it to places where you don't expect 23 the sound to go but you don't know where it has gone; and 24 that's the difficulty.
25 There's a large variations in terms of the Heritage Reporting Corporation (202) 628-4888
336 1 velocity, the. attenuation, field skewing, as well as what we 2 mean by_ partitioning. It might take the sound field and 4( )
3 break it up in two beams effectively'and send two beams 4 through and if there is a defect there you might get two 5 responses back that you think might be associated with 6 particular properties of the defect you're looking at, but 7 really~is-a property of the media that the sound is going 8 through.
9 Highlights is that PNL has documented 10 microstructure and sound field maps from a variety of cast 11 stainless steel specimens. We have conducted extensive 12 studies on these cast stainless steel three round-robin 13 studies and looking at them through technologies such as 14 SAFT.
15 We're currently putting together a topical report.
(
16 That should be drafted and available for internal review the 17 first of April.
18 Now the way that we actually make the measurements 19 are shown here. I'll just briefly put this up. What we do 20 is we place the specimen in a supporting jig and we then put 21 a small little indentation on the surface. We then spring 22 load a very small microprobe into that small indentation.
23 We then scan the transducer -- a transducer is typically of 24 what one might use, for example, for in-service inspection 25 -- across the surface. And then we measure the sound field Heritage Reporting Corporation (202) 628-4888
i 337 1 1
'1' as insonified by the transmitting transducer that's shown
<- j i ). 2 there.
3 The kind of materials that we look at then are l
4 shown here. You've seen some of these before, but this is a 5 columnar type of microstructure. .The dimensions on this.is I
6 about an eighth of an inch. You can see those dendrites go 7 nearly all the way through-wall. ;
8 Here is an equiaxed type of. structure. This is a 9 more interesting structure in terms of how they dendrite 10 structurepredominatelhhereontheOD. And the equiax
.11 structure predominately here on the ID. This view down here
- 12. is an axial view from the edge of the part that's shown 13 there.
14 This is another example where we've got a rather complicated structure. You can see the andritic structure
( J 15 16 that's in here and more of an equiax structure in this zone.
17 You get over here and you have a very small amount of 18 dendritic structure versus here. It's a very complicated 19 structure. You can see that that microstructure is clearly 20 not homogeneous and it varies completely around the part.
21 A representation of the kind of information that 22 we get is shown here. And I will get into the kind of 23 results that we obtained.
24 We normally scan the carbon steel block because it 25 is a very nice isotropic type of material. It has service Heritage Reporting Corporation l (202) 628-4888 I l
(~)%
t i i l l
l j c 1 l
4 338 :)
i p- 1- curvature and everything else that removes that effect. So ]
2 when we look at this -- this is our standard.for comparison.
'3' If Lwe now look at what happens when' it goes to an equiax l 4 structure you-can see here, this beam is centered right in-5'- 'the center of the plot and well as it is there at 45 6 degrees. .When you go over here and you look at this equiax 7 structure you can see there's been a skimming down to a.-
(:
8 ' lower; inspection angle. And if you'actually plotted this 9 from the center you would see that it has been shifted off-10'- center.
- 11. DR. SHEWMON: With that experimental setup you.run 12 the beam over, rotate the sample, and run the beam over 13 again?
14 MR. DOCTOR: -No. What we do is we position it and 15: then we do. a raster. scan like this. So this represents the
'16 raster scanning. It makes a small stepover. So this is 17~ actually a description of the sound field as measured by
'18 this point source. The point source is about ten thousands 19 of an inch in diameter. You're looking at a wave length 20 here that is in the neighborhood of probably 200 mills in 21 diameter -- 200 mill wave length.
22 So you can see it's very large. It's effectively 23 a point source and does a really good job of measuring 24 actually what the sound field is. The colors, by the way, 25 show a DB range here. You can see we've got a 20 decibel Heritage Reporting Corporation (202) 628-4888 l
____-_- . i
9 5
-339 c.
L , 1- range for displace.
- - -2 .DR. HUTCHINSON
- So when you say the.-carbon ~ steel-3 -is. isotropic what you really mean is that the grain size is 4 sufficiently.small such that anisotrophy at'that level is 5 'not-affecting.
6L MR. DOCTOR: That's correct. That's correct, yes.
7- If you go to a columnar microstructure youtcan see 8 a system that has happened. You-can see that this angle has 9- been-shifted up and you can see a compression that's 10 occurring in this direction. That's a focusing effect that 11' that type of columnar type. microstructure' consistently and 12 systematically does.
- 13. If'we-go now to those two complicated 14 . microstructure that I have.shown you, you can.see here now.
15' syou have reduced the semidry in both cases versus these
{
16' three cases up here. You now see that you no longer have an 17 iscmetric sound field. And you can see that this one is 18 spread out over from about-35 degrees for this red,..this.one 19 DB range up here to about 46 degrees. And you can see a 20 severe distortion in the field.
21 We've gone through and have made measurements, 22 . quantified them for one megahertz which is the predominate 23 frequency. The 45 degree is the dominate inspection angle 24 and longitudinal mode is typically what people use to 25 perform inspections.
Heritage Reporting Corporation (202) 628-4888
340 1 What we've shcwn here is we went in and selected
(^g 2 four places that were discreetly different on the specimens L) 3 that I've showed you, measured what the refracted angle is.
4 For carbon steel you can see a fairly type cluster. In this 5 particular case it says that for the velocity and that which
<6 is slightly different in the austenitic steels. We've got 7 an inspection and average of about, oh, 43 and a half 8 degrees.-
9 You can also see a small variability in that; 10 that's due to the repeatability of the measurement.
'11 Now we go to the equi &x structure and you can see 12 here that we now have a fairly large range. We're going 13 over from about 42 to 45 degrees for the center of'our sound 14 field plus and minus one and a half degrees.
15 The columnar microstructure as I showed you in 7-V 16 that previous viewfoil it tends to shift it to higher 17 angles. You can see a very tight cluster here centered 18 around about 46 degrees.
19 Then when you go to these other two mix and layer 20 type of microstructure where you have both dendritic and 21 equiax you can see a much larger variability here in terms 22 of the inspection angles as a function of spacial position.
23 What this tell us is that we're going to try and 24 make an effective examination on this material. And you're 25 going to have to understand that as you scan around your Heritage Reporting Corporation (202) 628-4888 O
l l a
341 1, sound field is going to be wavering all over the place and i 1
) 2 you won't be getting the sound where you think you are.
o l3 If you really want to do an effective examination 4 you're going'to have, on the fly basically, determine what 5 that microstructure is so that you can compensate for it in 6 terms of the inspection that you're performing.
7 Another property that we looked at in terms of the 8 sound field maps was, what happens to the size of the sound 9 field. And we measured this both in terms of a 10 circumferential variation and in terms of an axial 11 variation.
12 So what we did was to go to the four maps. We 13 went in and we picked off this, in this particular case, 14 3-D-B levels and we measured the width of that sound field
/ ) 15 in both the axial and circumferential direction. We looked 16 for whichever one of those was the minimum and then we 17 looked at what was the max variation that we found in terms 18 of the various measurements on those four different field 19 maps.
20 We then applied this as a function of this ratio 21 which is shown here. When you look at the carbon steel you 22 can see the variability is down about 10 percent. That's 23 the kind of repeatability that we see in terms of that kind 24 of measurement.
25 When we went to the equiax structure you can see Heritage Reporting Corporation (202) 628-4888
c-I 342 1 that for the circumferential and axial variability it was )
!) 2 about the same, but it is systematically skewed up here to j
3 approximate 1y'.3 or about 30 percent.
4 For the columnar microstructure you saw there was 5 a systematic shift but there was also a focusing that 6 occurred. And that focusing resulted in a very small 7 circumferential variability where you found a quite large 8 variability. The same order of magnitude'is for the equiax 9 structure in that the axial direction.
10 When we went to the two complicated 11 microstructure you can see that the sound field size, in 12 essence, changed by approximately 50 percent.. What that 13 means is that if you're using the sound field properties to 14 try and make measurements about the indication that you're looking at you're going to tend to have a fairly large error
( ) 15
- 16. in terms of what the indication is in terms of its size.
17 And that's a very important fact.
18 We're looking at this in terms of the implication 19 for sizing and the implication of what this means in terms 20 of changing the shape of the sound field because what we see 21 as one of the elected consequences of this le that one may 22 have to have a smaller stepover amount between subsequent 23 scans in order to compensate for this effect.
24 DR. SHEWMON: Sometimes they have trouble getting 25 a beam through the thing at all, where that seems a first Heritage Reporting Corporation (202) 628-4888
343
.1 . test they do. Have they overcome that with different
. l()_ 2 frequency or more powerful something or other?
- 3. MR.. DOCTOR: There is improvements being made in 4 that direction with proper selection of transducers and 5 that. You're starting to get sound fields through. I just 6 saw last week a piece of equipment on a calibration standard 7 that we've used in our round-robin test, a piece of 8 equipment that had 1,000 volt pulser, add a single 9 frequency, namely one megahertz, strung out over about six 10- cycles. It had the best signale that I've ever seen looking 11 at the 10 percent end mill slots that are the calibration i
12 reflectors.
13 As a matter of fact, one of our round-robin -- we 14 had a bet as to whether or not anybody.could actually find those and nobody did. They couldn't even' set up and
( } 15 16 calibrate on them. And this had a beautiful 12 DB signal 17 and noise ratio looking at those.
18 So I think improvements are being made in that 19 direction. We still have the problem though that the sound i
20 field properties are varying and we don't know how to 21 compensate for that at this time. That's one of the 22 problems.
23 The other problem is that you can get in if you're 24 down, you can turn the grain up and get a response coming 25 back. The difficulty la that because you've got these very l Heritage Reporting Corporation (202) 628-4888 1
4 m__________-_._ _ _ . _ _ _ _ _ _ . . _ _ _ _
f
- 344 1- large grains that are in there you get coherent scattering
() 2' coming from them and you will look at that and it has all 3' the properties of a flaw. It walks in time, it does. 5:
4 everything, exactly what you would expect from a machine
!' 5 reflector. There's an acoustic impedance mismatch'between 6 those grains. The scatter is just a beautiful signal. And
-7 you have a real discrimination between knowing whether 8 you're looking at that grain structure or actually looking 9 at a flaw and that's a tough one. That's going to be really 10 the challenge. I think we're going to be able to get energy 11 back. But the real bottom line in the challenge I think in 12 the future is going to be whether or not we're going to be 13 able to discriminate effectively.
14 Well,.our conclusions from the work are these cause significant more sound
( f 15 complicated microstructure 16 ' field distortion than the pure microstructure do. And we 17 feel that you're going to have to either on-the-fly 18 determine what the microstructure is; you're going to have 19 to assume that you've got the worse case. And you're going 20 to have to make some corresponding assumptions with regard 21 to that, with regard to perhaps the sensitivity exam that 22 you' re using in terms of the amount of overlap between
'23 subsequent passes as an example.
24 Four is indicated, the standover lap we feel is 25 going to have to be increased. What we're trying to do is Heritage Reporting Corporation ,
(202) 628-4888 I
'345 1 to pull this information together and go to the ASME code.
(} 2 and present the case for what we feel should be done.
.3 That's being fully drafted in a topical report. So we 4 should have that prepared about the end of April. That was 5 May 1, wasn't it.
6 What I wanted to do was to move on and talk about 7 the code activities. We're involved in upgrading ASME code 8 and NRC requirements usirag the programmatic research results 9 from this program. And we're specifically directing this 10 effort to the Section 11 inspection requirements.
11 Some of the highlights are that this code' case 12 N-409-1 was just recently published. It includes 13 statistically designed performance demonstration to qualify 14 personnel, the equipment, and procedures for the ultrasonic 15 testing in-service inspection of piping welds. That was in 16 place in 7/88. And that's a significant step forward.
17 The earlier requirements were based around 409 in 18 which we simply did a test looking at maximum flaw links. In 19 this case we're introduced the concept of grading units and 20 that's the basis of this statistical design.
21 We're preparing this NUREG on 4882 dealing with 22 documenting our qualification activities and hopefully we 23 will have that out in the next couple of months.
24 The next item here I've already mentioned both 25 Appendix VII and Appendix VIII. I want to talk about them Heritage Reporti.ng Corporation (202) 628-4888 O
i l
346 J 1 in a little more detail _in a few moments and'present some fs
'(~)' 2 viewfoils that gives the strength and the weaknesses of how 3 we see these two particular appendix with regard to their 4 impact.
5 So why don't we just move on and talk about it.
6 The strengths are very significant. They're important. The 7 things we've been -- I think you're aware, Paul, back in '83 8 you were at one of our meetings at Seattle dealing with thic 9 issue of qualification. It has come a long ways and there 10 has been a large amount of industry participation. They 11 fully support it.
12 What we're trying to'do here is just kind of 13- ' compare these two appendices to what is currently in place.
14 And some of the things that these two appendices provide is
( ') 15 ' guidance and rationale for conducting the performance 16 demonstration, whereas what currently is in existence here 17 does not provide such information.
18 Furthermore, you have referenceable qualification 19 criteria, both to the flaw detection and flaw sizing.
20 Whereas now there's a plan but that plan doesn't have any 21 detail for referenceable criteria.
22 It includes specific requirements for personnel, 23 the equipment, and procedures versus what currently is going 24 on. It deals only with the personnel.
25 It provides explicit criteria for evaluating Heritage Reporting Corporation j (202) 628-4888
>O i
l i
347 l 1 . performance demonstration results'both in terms of the r~p (mj; 2 grading and in terms of'the criteria, the pass / fail criteria 3
'31 that one uses to judge whether or not someone-has 4 successfully passed the test. j 5 It contains explicit criteria regarding the test 6 specimens. It tells you what kind of specimens you need'to.
7 'use. What kind of defects. The number of these. flaws. The' 8 proximity of flaws. How close can you have these things one 9 to another.
10 Industry has been involved in this whole process 11 pre'ty t much from the start and they are supporting it.
12 They've endorsed.it. As a matter of fact, when we 13 originally started this work it was dealing only with Lg 14' ultrasonics. When we went to the ASME code with-it they
() 15 immediately endorsed this approach, but they wanted to 16 extend it in a plant to all NDE for in-service inspection..
17 We are hoping down the road that one day that 18 would actually happened. But it. happened almost-19 instantaneously.
20 The responsibility and qualification criteria for 21 NDE instructors are involved and clearly defined in these 22 two appendices. Currently no requirements or criteria 23 exists for them. They require statistically designed 24 performance demonstrations, that's important. This code 25 case that we talked about 409-2 does have that in there but Heritage Reporting Corporation (202) 628-4888
.O
1 348 i
1 of course that's not mandatory. So it is really the first !
I 2 time it has been put in in a mandatory manner.
{a~}
3 It requires periodic training. 10 hours1.157407e-4 days <br />0.00278 hours <br />1.653439e-5 weeks <br />3.805e-6 months <br /> per year 4 is a compromised position, as I'll show in the weakness that 5 we feel it should be more than that. But we simply had a 6 large amount of resistance from industry to try and push for 7 more. Generally the people that are doing the work agree >
8 that they would like to have more, but the people that are 9 paying the bills who have to provide that training expense 10 c.re the ones that are reluctant to support it.
11 Requires nuclear application experience plus 12 classroom training for the level IIIs. And that's 13 important. In the past that has not been the requirement.
14 Qualification examinations both written and 7s 15 practical must emphasize Section 11 applications.
b 16 DR. SHENMON: What does the one before that mean, 17 what is required nuclear application experience? It just 18 means they must have worked in a nuclear power plant?
19 MR. DOCTOR: Yes. When you're going through in 20 terms of like being a level II and then they pass level III 21 experience they must have nuclear experience to have l
22 successfully gone through this entire process.
23 The training syllabus was upgraded and expanded to 24 cover level III topics. This refers to SNT C-1-A basically 25 took that and expanded that for level IIIs.
Heritage Reporting Corporation (202) 628-4888
l 349 1 . Appendix VIII involves or includes specific
( 2 supplements for.the' clad / base metal interface, nozzle inner
- 3. radius, pressure vessel,shell welds, nozzle-to-shell welds, 4 bolting and studs,.as well as for austenitic and ferritic.
5 piping welds. So it addresses every one of those in terms 6 of the supplements that are in there.
, 7 There are, however, two supplements that.are
- 8. currently being developed: one is on the dissimilar metal' 9 welds; and the other is on cast materials. Those are in
! 10 development and are not part of that Appendix VIII at this 11 time.
12 DR. SHEWMON: In your earlier talk you mentioned 13 the problem in 'the steam generators in defining what kind of 14 flaw it is you wanted to certify people on or the spectrum
( ) 15 of flaws. That used to be a problem here. I can remember.
16 the NRC -- industry people telling that the NRC wouldn't say 17- what kind of flaws they wanted to look for, so what the 18 hell. But now through PISC or whatever it is and other 19 things is it pretty well agreed what kind of defects you 20 need to certify people on?
21 MR. DOCTOR: I think that there is a consensus on 22 that. When you look at, for example, what's in Appendix 23 VIII there's some options, there's some flexibility built 24 in. When you look at it you find that there are 25 conservative approach taken. There are different kinds of Heritage Reporting Corporation f (202) 628-4888
(
p 350 1 cracks as an example that one can put into their specimen,
.(^s
(,) 2 stress corrosion cracks,. general fatigue cracks, mechanic 3 fatigue cracks, those are predominately the' kind of 4 mechanisms that you find that's in piping.
5 What you find is that thermal fatigue cracks are 6 quite conservative and those are specified as one of the 7 things that one can use. And there's a number of those that 8 are specified.
i 9 DR. SHEWMON: When you say conservative, do you 10 mean they're easier to see --
11 MR. DOCTOR: No, they're harder to see than, for 12 example, a traditional mechanical fatigue crack. A 13 traditional mechanical fatigue crack it's just like putting 14 saw cut'in. From an ultrasonic standpoint it's very easy to
() 15 find that particular condition. j 16 A thermal fatigue crack has a residual compressive 17 stress that makes it tight and it has a lower reflective 18 coefficient, as a consequence it's much more difficult to.
19 detect that.
20 Let's talk about the weaknesses with regard to 21 these two appendices. One area that has not been addressed 22 at all is implementation. The code has steered clear from 23 that particular issue and it's something that needs to be 24 addressed.
i
-25 There are a number of things about that issue that 1
Heritage Reporting Corporation l l
(202) 628-4888
() {
)
1 L __ __ - _ - - .
351 1- has to be addressed in terms'of, for example, over what kind
) 2 of time frame are you going to require this to be done? Are 3 they going to wait for ten years versus putting in place 4 maybe a three or five year time period?
5 There are some other areas of implementation that 6 also have to be addressed. For example, what we put into 7 the code spells everything out. It spells so much out that 8 if one sits down to actually -- and would go through and 9 apply that in a simplistic sense you could almost not know 10 anything and go in and have a chance of passing the test.
11 For example, if you were told you were going to ;
12 have a true/ false test which is basically what you're doing 13 in crack detection and you were told that you were going to 14 be taking, let's say, a three out of three type of test. In O 15 other re ' v ='re so1=e te aeve three tree co ettio= i=
16 there and you were told that there were going to be twice as 17 many false conditions and if the number of falses that were 18 in there that they would be dispersed such that no two trues 19 could be adjacent to one another you would have a tremendous 20 amount of information without even knowing what the answer 21 is with regard to just successfully passing them.
22 So in this area of ireplementation has to look at 23 this testing that's going to actually apply because you 24 spelled out so many conditions that if you don't disguise 25 the test you've have given people so much information that l
Heritage Reporting Corporation (202) 628-4888
352 1 they could just use that information and have testmanship 2 apply and be able to have a chance of passing it. So that's (V~N 3 an area of implementation that has to be addressed.
4 The ad hoc task group concept of criteria for 5 agencies. Agencies that would deal in, for example, 6 training, performing the qualification, or performance 7 demonstration was eliminated from the code committee 8 consideration, so it's not part of Appendix VII or VIII. I 9 think that, to a certain extent, falls in this area up here 10 of implementation.
11 If you address this you're also addressing I ;
12 believe these implementation issues. So that's another area 13 that does have to be addressed.
14 As I mentioned earlier those 10 hours1.157407e-4 days <br />0.00278 hours <br />1.653439e-5 weeks <br />3.805e-6 months <br /> of per year f, 15 of periodic training we feel is insufficient with regard to Y.]
16 the complexity of the kinds of problems that 10 hours1.157407e-4 days <br />0.00278 hours <br />1.653439e-5 weeks <br />3.805e-6 months <br /> of 17 simply trying to learn how to do IGSCC as an example simply 18 isn't adequate. I mean, they have expanded that test or 19 that training session from, you know, a week to basically a 20 two week time period because it is difficult. There are 21 skills that are involved and 10 hours1.157407e-4 days <br />0.00278 hours <br />1.653439e-5 weeks <br />3.805e-6 months <br /> per year is probably 22 not adequate.
23 Requirements and criteria for requalification are 24 not specifically addressed. They're only covered by l
25 implication. And that's an area of concern that we have.
Heritage Reporting Corporation (202) 628-4888 r~T
\_)
_ _ _ _ - - - - -- --.--____ ____________ _ _ _ _ a
l 353 0 1 This, of course, relates back to this
) 2 implementation issue because anybody can get involved in 1
3 this process. There is the possibility that you're going to 4 have a.large amount of variability with regard to the actuali 5- implementation of these two appendices. So that has to be 6 addressed and I think the NRC needs to look at that and 7 establish their position with regard to the implementation.
8 of both of these appendices.
9- DR. HUTCHINSON: How many inspectors are there --- l
.10 how many people does this apply to?
11 MR. DOCTOR: I'm trying to think in terms of like 12 the people that have gone through down at EPRI. Your 13 talking like 3, 4, 500 people for doing ultrasonic in-14 service inspection.
MR. MUSCARA: Can I comment briefly on the 10 hour1.157407e-4 days <br />0.00278 hours <br />1.653439e-5 weeks <br />3.805e-6 months <br />
({) 15 16 maintenance training. The 10 hours1.157407e-4 days <br />0.00278 hours <br />1.653439e-5 weeks <br />3.805e-6 months <br /> per year is training 17 that everybody should take on new topics. But Steve 18 mentioned that it was not enough to learn how to detect 19 IGSCC, that's not the intent for those 10 hours1.157407e-4 days <br />0.00278 hours <br />1.653439e-5 weeks <br />3.805e-6 months <br />. If you 20 want to detect IGSCC you must go to the performance 21 demonstration test.
22 So this training is just training for everybody to 23 take just as they current. We start out at 40 hours4.62963e-4 days <br />0.0111 hours <br />6.613757e-5 weeks <br />1.522e-5 months <br /> was our l
24 recommendation and came down to 10 hours1.157407e-4 days <br />0.00278 hours <br />1.653439e-5 weeks <br />3.805e-6 months <br />. I really don't J 25 think that's too much of a drawback since they have to prove i
Heritage Reporting Corporation (202) 628-4888 O
1 1
W 354-1 :themselves in.allLthe types of flaws they must. detect
'2 : anyway.
~
'3- MR. DOCTOR: But as new mechanisms or degraded is 4 found outlthere that's'when they obviously go into that 5 training. I don't feel.that 10 hours1.157407e-4 days <br />0.00278 hours <br />1.653439e-5 weeks <br />3.805e-6 months <br /> is enough, I think. i 6 that's-the bottom line. But eventually they're going to 7 work inLthe performance demonstration in that training 8 that's associated with that.
9 Well, I would like to move on and discuss now is 10 the equipment interaction matrix study. What we're trying 11 to do here is that --
12 DR. SHEWMON: I. agree with your last statement 13 that once you get the performance certification through then-14- the argument over how much time they have to' pay for to .;
bring them up to date passes away some. When'is that likely 1
( ) . 15 16 to be? That is that there will be a performance 17 certification and since on the IGSCC there is one now, isn't 18 there?
19 MR. DOCTOR: There is one now except the weakness j i
20 of it is the fact that there are no referenceable, you know, 21 requirements and criteria for that. And what you want to be
- 22. able to do is just open it up so that, for example -- I 23 think one of the issues of implementation is to have some 24 regional centers as one of the options so utilities would go 25- together and have a location where they could have a series Heritage Reporting Corporation
. es . (202) 628-4888 dj
i 355 1 of specimens, have their training, and do the performance 2 demonstration.
3 If you go with that concept then you have to have 4 enough criteria and requirements that you can get uniformity 5 among these different places.
6 DR. SHEWMON: How many utilities have their own l
7 inspectors? I thought most of these people would be '
l 8 Southwest Research and a few others that would come in on 9 contract.
10 MR. DOCTOR: Generally you find, you know, l 11 different utilities from one extreme having almost nothing 12 to having a partial support crew that might be involved in 13~ doing supervision while they actually have outside 14 contractors in there. There's a spectrum. But to my 15 knowledge I don't know of any that have an adequate staff to
-16 deal with the entire ISI, for example, of a nuclear plant.
17 Because that's a very large load in terms of standpoint of 18 the amount of radiation exposure.
19 A lot of the trend is towards job shopping and 20 with regard to, I hire you and you use you until you're 21 burned out and then you can go off and farm or whatever i 1 22 until the next quarter.
23 The equipment interaction matrix study was set up 24 and designed to look at trying to control one of the things 25 that we have engineered; mainly the equipment. How do you i l
l Heritage Reporting Corporation ]
(202) 628-4888 3
r g,/
\_
356
[ -1' -set limits on how'much the. equipment parameters can vary so
() '2 that you limit the influence and the effect.of that in terms i
I3 -of:the' reliability of the inspection.
4 The only existing requirement-are currently what 5 were put.into effect.in N-409-1. They're-alsocinc409-2.
6- And they basically were engineering judgments based on a-7' database that we had assembled.
6 And we came up to the conclusion that if you 9 limited such things as pulse amplitude, pulse rise time, 10- . pulse-duration to parameters such as 10 percent, you found 11 that we werellooking at a bunch of the data. It looked'like 12 we.could get_ reproduce the ability of the results.
13.' However, what we wanted to do was'to put that on a 14 scientific basis so that we could look at that and say, have
'15 :we covered all the cases? Are there. things that are 16 slipping through that we aren't covering?
.17 So what we did basically was to build a model of
'18 this-entire process: the pulser; the transducer; 19 acoustically transmitting energy down off from.a plot back 20 :to the transducer and to your receiver and display 21 associated with that. We wanted to model that.
22 And our justification was, when we went to a study 23 from an empirical standpoint looking at several different i
24 pulsers, receivers with different amounts of tuning I 25 associated with them, traditional transducers that we found Heritage Reporting Corporation (202) 628-4888 1
.357 .
1 l' with people used.out in the field and a series of rather k l
j )I 2 standard type reflectors, side drilled holes, standsrd ASME
]
'3 ' calibration notches, angled notches, fatigue cracks and 4L integrated stress. corrosion cracks. We built up this matrix 5 -- and if I just show you some results from one particular 6 case. )
7 In this particular case we're looking at a series 8 of saw cuts, kept the pulse receiver combination the same 9 and we just changed the transducers, you.can see here.
10 'Looking at a couple of different of these saw cuts you can 11- see a very large variability. The worse we found when you~
. l 12 just switch the. transducer it went from a plus 52 percent to 13 a minus almost 29 percent. That gives you about an ADB 14 roughly amount of variability with regard to this equipment.
'r) p
\m-15 Now when I give these numbers what we did was we ]
]
16 looked at all six of the combinations. For example, change {
j 17 in transducer, computed the average and what we have shown 18 here are the largest' deviations from that average.
19 Now, specifically what really happens in an 1
20 instrument as shown here, we've got an instrument with a 21 transducer and it has some kind of a transfer function.
22 Transfer function means its responds as a function of l l
23 frequency. ]
24 If in fact I had a nice broad band defect then 25 there would be no important elements associated with my Heritage Reporting Corporation (202) 628-4888
-,,.__._,.,__n_ _ . - - . _ - - . , _ - . , , - _ _ - - ._
358 1 equipment: parameter. -But in fact.what happens to the j 2- ' defects is that they have dimensional sizes and as a
-3 consequence the energy. that comes off of them is associated 4 with that dimensionality. And in fact.you;will find that 5 'you can get this kind of response. .I'll show you some-6 actual.ones that we computed and done some' comparison with
, 7 the experimental work and show this very nicely.
8 But what happen is when I multiply this times
~
9 that, this is what I'get for a response for.one system; this 10 is what I get for a' response from another system. Both of 11 these wereLset up_and calibrated on the same side drill-hole i
12 astan example as required by ASME~ code; and yet what I-did-13- here is a difference shown here in terms of the response 14 from'that same reflector. j J15 I want to put this up and only say one' thing about
-16 it. -What we're doing. basically is using-elasticdynamic 17 physical optics model which means it's just the same one 18 that's used in typical optical lens designs. We do ray 19 tracing and. accommodate -- what happens to the-transducer by.
20 simulating it by a series of point sources that we're going 21 to sum together to simulate an entire transducer. And of 22 course we do the same thing with the receiver.
23 The kind of performance that we find from this 24 model is exemplified here. This is just taking a 90 degree 25 corner, in other words the corner such as this and what I i
)
Heritage Reporting Corporation (202) 628-4888 I) i i
. _ - .:- _ . - - = . . _ - . . _ - - _ - _ . . _ - _ _ _ _ - _ - - - - - - - _ _ _ . _ _ _ - . _ _ - . _ _ _ _ _ _ _ _ - _ _ - - - - - - .
l
359 ,
l 1 'want to do is to see what the response is as a function of A
A) s 2 distance from the corner. When I've done that you can see 3 here our model predicts the solid line or dash is what the 4 experiment shows. An extremely good agreement.
5 Down here is another example where in this case 6' what I've done is I've put a flaw in this over here, a 45 7 degree angle, so I'm coming down and striking it in a normal ,
l 8 fashion and you can see that you get very good performance 9 from that. And then we.also did a through transmission. A 10 variety of different ways.
i' 11 Here is another case where we have done a 12 comparison - where we' re showing in this particular case 13 some results on tandem. A benefit of being involved in the 14 PISC program these were some data that were generated by a 15 nuclear laboratory and provided them to us before the report
.( )
16 was out so that we~could actually test the effectiveness'of 17 our model for these different cases. This was a 10 18 millimeter strip buried into 193 millimeter thick steel slab 19 at 2.25 megahertz. This is a 25 millimeter flat bottom 20 hole. And the same thing for 25 millimeter strip. So we 21 got extremely good agreement.
22 The kind of results that we generated are shown 23 here. Let me draw your attention to this upper left hand 24 corner. What we're modeling then is transducer on a wedge, 25 sending energy down. We've got a flaw or a crack, if you 1
1 \
Heritage Reporting Corporation )
(202) 628-4888 G,f g t
J.j . r 360'
- 1 will, located here. Tie finest link by this dimension. -And l '~ '2 this angle measured . frc m this surface.
3 Now what'I'm howing here in this particular case 4 is that.I've got a vertically oriented flaw, so it's 90 5' degree to the back surface. For'E flaw of length or depth,.
6 if you will, from a. half millimeter up to 25 millimeters you 7 can'see that this is the uniformity that-I.get. There:is 8' .no, if you will, low pass nature associated with it. This 9- would be kind of'a' typical pasaband for many equipment 10 systems. .So you can see over this range they're almost all-11 basically flat.
12 However, if I go over, if you will, to this corner 13 and now-if I have-that flaw instead of being. vertical I-14 rotated over to 85 degrees,. Now you can see that in fact
-() 15 I'm starting to develop gnaws that are associated .with the
-16 flaws due to that angular property and due to the dimension 17 of the flaw. So you can see these gnaws.
~
18 Now what we concluded was, a worse case condition 19 would be that if I took this gnaw and I placed it right in 20 the center here of the passband so that this thing would 21 come down like this, and that would be a worse case 22- condition.
'23 And so what we did then was to basically put I 24 together a series of about 10 different flaws that were 25 worse case. So we said if you could vary this parameter and Heritage Reporting Corporation (202) 628-4888 ,
t
l 361-l' the result from going through our analysis showed that you
-c 2 can strain the output to no variation larger than what you
-3 can strain;that parameter to being that that was acceptable. -
4 And what we found, in essence, was that this could 5 be met for most of the conditions. There is one where it 6 doesn't meet it and I will talk about that in'just a moment. l
'7 Basically I just wanted to give you a status. We 8 validated our~model within sections. We determined whether 9 this worse. case flaw condition, those that give us maximum 10 DB change in the middle of our equipment passband.
11_ One of the things we get was, the real key to this 12 whole approach that we've taken is this worse case type of
- 13. flaw. So we put together a paper and we've submitted it to 14 materials evaluation to get a peer review on that because
() 15 .that's critical when we go to ASME code. We want to make 16 sure that there is nothing that we haven't covered. And so 17 we've taken'this approach.
L 18 The modeling analysis and experimental testing we 19 have completed to date show that it took the changes 20 affecting both gain and bandwidth are complete and they show 21 very good consistency. In other words, if I can strain a l
22 parameter such as gain to 10 percent or can strain the 1
23 bandwidth to a ship of no more than 10 percent the output is 24- always less than 10 percent, okay.
25 However, when we get to this one on center Heritage Reporting Corporation (202) 628-4888 g-)g
\-
l 1
-1______ _ . _ _ . _ _ _ _ _ _ _ . _ _ _ _ _ _ __
.j
362 1 _ frequency we find that for very narrow band systems, if you -l 7_
i,s_) 2 change the center frequency by 10 percent for these worse 3 case flaw conditions we, find that you can get a 40 or 50 4 percent change in the response from a given flaw. l 5 If you go to a broad band system than this center ;
6 frequency requirement gets totally relaxed and you can 7 strain it 10 percent. It's always less than 10 percent in 8 the output.
9 So the bottom line is what we're going to be doing 10 in terms of going to code is that there's going to have to 11 be different requirements. If you've got a broad band 12 system 10 percent is something that's easily achievable.
13 It's something that you can measure relatively easily and it 14 gives you a constrain within that 10 percent.
() 15 But for narrow band systems you're going to have 16 to control frequency rather tigL ly. And that's the bottom 17 line.
18 I kind of jumped ahead of myself here and said 19 that these defects do act as low pass filters. And I've 20 already discussed what we mean by the impact on narrow band 21 and broad band system. But it's nice because we're 22 qualifying all this. It has got a good firm foundation and 23 it's something that now can solidify those numbers that were l
24 actually put into the code.
25 We're going to expand this work in terms of Heritage Reporting Corporation
,, (202) 628-4888 k-)
363 1 looking at thick section. We don't really anticipate any f) 2 surprises when we go to thick section. We feel that pretty 3 much the analyses we have done will be further verified for 4 the exception.
l 5.
^
When we go to things such as double' compound.
6 radius of curvatures that are associated, for example, of 7 nozzles we think there's going to be perhaps some 8 implications on that and we're going to have to look at that 9 problem in detail.
10 DR. SHEWMON: Are there any, what you referred to 11 as thin materials in nuclear systems?
12 MR. . DOCTOR: Oh, yes. Thin materials are for 13 piping. Things that are rigid up to several inches in.
14 thickness. When we did --
15
() DR. SHEWMON: Somehow I thought of thin as 16 somewhat less than a couple of inches.
17 MR. DOCTOR: No. We' re talking -- we arranged up 18 to actually three inches in thickness. We looked at some of 19 the cast stainless steel thickness and those go up to about 20 80 millimeters.
21- MR. MUSCARA: That piping in vessels.
22 MR. DOCTOR: That's right.
o 23 DR. SHEWMON: Tell me this again, thin materials 24 you said 80 millimeters.
25 MR. DOCTOR: 80 millimeters.
Heritage Reporting Corporation
. f-) (202) 628-4888 LJ
o f'
364 1 DR. SHEWMON: Not 80 mill, okay.
f~T g ,j- 2 MR. DOCTOR: 80 millimeters, yes.
3- DR. SHEWMON: ,
Why is that thin? It's much-thicker 4 than your wave length.
5 MR. DOCTOR: Yes. It's just in contrast with the 6 vessel which is typically in the range of six to 10 inches 7 in thickness. So it is a relative just thicker.
8 I would like to describe the FISC activities, if I 9 could next.
10 The kind of things that we're involved with, for 11 example, under the PISC II data re-analysis. The database 12 has been generated by that round-robin test that 50 teams 13 participated in. And we have that data available at our 14 laboratory on our computer. We are looking at.that data to 15 analyze it in regard to what the implication of that is and
()
16 regard to conditions in ISI practice here in the United 17 States.
18 Some of the things we're addressing is, for .
19 example, the analysis methodology that's employed is not 20 typically of what we use here in the United States. We 21 would like to evaluate it in light of, for example, the 22 requirements that are going in place, Code Case N 409-1 23 which is basically the same thing that's in Appendix VIII.
24 So we'd like to see how in fact if one constructed 25 a series of flaws how many of those teams would have, for I
Heritage Reporting Corporation (202) 628-4888 -
()
365
- 1. example, passed.
() 2 Exam the PISC results in light of the ASME code j 3 for reactor pressure vesssi performance demonstration, I 4 just said that.
5 The materials and surface finishes that'are used 6 on those blocks are typical of what is representative of 7 conditions in Europe but not typical of conditions here in 8 the United States. And we're going to look at it in terms 9 of what is the implication of that.
10 The PISC results basically address.the upper bound 11 measure of capability. They sent the specimens around, 12 people used them in their laboratory. Even though it might 13 be a field type of system putting it into a laboratory and 14 using standards that are not typical of, for example, what
( ) 15 one might use out in the field and things like that, provide 16 you with an upper bound in terms of the performance with 17 regards to the techniques.
18 Maybe just go on and show you some of the results 19 from PISC and the kind of things that are important from 20 that and what we' re looking at.
21 Here, for example, are some results from PISC I j 22 and from PISC II and basically what they did was they looked 23 at it and they broke up the kinds of defects into smooth 24 sharp planar defects into volumetric defects and then these 25 rough cracks or combinations of smoother planar defects and I
i Heritage Reporting Corporation ;
(202) 628-4888 '
O
366 1 volumetric combinations complicated things.
2 So you have these three different clusters that 3 are shown here in all cases. And what we're looking at 4 really is what happens'with regard to the ability to detect 5 something. You can see here is a function of defect size.
6 There's a fairly strong comparison between using ASME 50 l 7 percent DAC and performance to being able to detect these 8 different classifications of defects.
9 You can see that this class A, the smooth sharp 10 cleaner type of defects are more difficult to detect. And 11 the reason for that is that unless you specifically do 12 something to detect that tip signal or you have a receiver 13 where that specular energy is reflected to, you're simply 14 not going to see that.
15 And if you look at traditional 50 percent DAC, 16 I'll show you in a few moments, you just simply don't have a 17 sensitivity to determine those defracted signals. So as a 18 consequence you don't find them.
19 You can see that when you go to volumetric defects 20 things improve. When you go to rough, these complicated 21 things the energy gets scattered everywhere so that you have 22 a much higher probability of being able to see them at any I 23 given location corresponding where you can see the shift and i
24 the change.
25 One of the outputs from this is shown here where Heritage Reporting Corporation (202) 628-4888
() ;
j
367 1 you're plotting the mean defect detection frequency as a
() 2 function of DAC level for three of the different pipes that 3 were employed. You can,see there's a very strong systematic 4 trend up here to 20 percent DAC and virtually no change 5 between that and going to 10 percent DAC.
6 So as a result of this one of the changes that did 7 occur in ASME code was to change from 50 percent DAC to 20 8 percent DAC.
9 I mentioned the fact that for a smooth planar type 10 defects one of the things that you have to be able to 11 detect --
12 DR. SHEWMON: One of the reason that people went 13 with 50 percent earlier is that there was enough noise in 14 other problems with implementing 20 percent. That was
() 15 crappy steel, it's not going to get any better if the i
16 equipment depended it may.
17 MR. DOCTOR: Yes. One of the things that they've 18 done in PISC, if you look at plate number one which isn't 19 one of the things that I've showed you in the previous 20 viewfoil.
21 Plate number one was one that was made in which 22 there was a very large number of unintended defects that 23 were put into that when it was fabricated. What they found 24 is that teams that went in and carefully characterized those 25 unintended defects, they subtracted them out and have j l
l Heritage Reporting Corporation (202) 628-4888 i
1 l
368 1 something -- no influence on the performance. They did as 2 good a job of using that as if they had done it in actual 3 perfectly clean steel.
4 So the conclusion from that is that if in fact you 5 do have dirty steel you have to be more careful, for 6 example, normal inspection to determine what those 7 unintended type of conditions are and then subtract those 8 out and then you can characterize what remains.
9 .Now it does require obviously more time. All the 10 vessel inspections are done with remote tools. There is no 11 radiation exposures. It's a matter of simply collecting 12 data and it takes time on the analysis side.
13 This particular viewfoil I think demonstrates the 14 difficulty that you have. This is showing two different 15 things in terms of the two axis. What's over here is the DB O 16 response level. What you find plotted over here is the 17 response level but as a function of the DAC. So five, ten, 18 20, 35 -- here's 50 percent DAC up here.
19 And what we're looking at actually is this lower 20 crack tip response. The lower crack tip when you eliminate 21 a vertical planar type defect you get a response of the 22 fraction off the top and the one off the bottom. The one 23 off the bottom is always substantially larger than the one 24 off the top, okay.
25 And so if you're going to detect it that's the one Heritage Reporting Corporation (202) 628-4888 m
y
i 4
369 1 you're going to.see. And so what this is is a series of
() 2 measurements for vertical planar defects in the 10 to 25 3 millimeter through-wall, depth size as a function of location l l
4 going from very near the surface to 200 millimeters below 5 the surface.
6 And what we have plotted here than as result of 7 detecting that lower tip at 70 degrees and 60 degrees and 8 with a 45 degree inspection angle. And what you see here is 9 that, if you're going to use a 45 degree which is 10 traditional that the basis for most of the work within ASME 11 code, people felt that this was the most reliable. They did 12 this but they felt that 60 degrees did not provide much new 13 information.
14 You can see in order to effectively use that 45
() 15 degrees to detect that lower tip you had to be working down l 16 here with something like the 8 percent DAC level. And 17 that's extremely difficult to work down at that level.
18 When you go to 60 degrees you can see that working 19 at 20 percent DAC you will get a response above that level 20 down to about 75 millimscers or, you know, basically the top 21 six inches. If you go with a very effective probe such as 22 the 70 degree and work in the 20 percent DAC you can see all 23 the way down through it with that and get a good response 24 off of that lower tip.
25 So this is the reason that relying on 45 degrees l
Heritage Reporting Corporation (202) 628-4888 o_- - _-- -_ _ - - - - _ _ - - -
370 1- you are dropping down to 20 percent DAC is simply not going
(:,)-
2 to pick up that defracted tip signal.
3 This shows taking some of the results from plate 4 number three and plotting them as a function of probability 5 detection and currently sentencing them in term:s of, you 6 know, their size and what they are. So this has really got 7 a kind of a sizing element in it, if you will, because of 8 this correct sentencing. As a function of the defect 9 through-wall size.
10 -And what we have plotted here is the 50 percent 11 DAC level and an air bar around there showing that it
.12 applied that 50 percent DAC. They had just kind of a 13 variability associated with the -- but you can see very, 14 very poor performance in the neighborhood of about 15 O 15 verce=t-16 We now go to a 20 percent DAC which is currently 17 in the code analyses and much improved performance level, 18 getting up here in the 75, 80 percent range. Although there 19 still is a fairly sizeable amount of variability associated 20 with it. J i
21 Some of the European teams use special procedures )
i 22 and those were separated out. And you can see that one can ]
1 J
23 get a substantial amount of improvement if you apply some of 24 these special procedures.
25 DR. SHEWMON: Are these usually toss and catch or 1 i
Heritage Reporting Corporation q (202) 628-4888 )
L/
j 371 1 snake or whatever sney call it?
2 MR. DOCTOR: They're a variety of things. They're 3 using SAFT, holography,,using a tandem, that. type of thing.
-4 So it was a variety of things that were actually' lumped 1
5 together into this category. And this occurred at the ,
6 Marshall committee said that one had to be able to achieve I 7 with regard to finding defects.
8 DR. SHEWMON: Now none of this is through 9 cladding, is that --
10 MR. DOCTOR: This is all through cladding.
11 DR. SHEWMON: Okay. These tests --
12 MR. DOCTOR: All of these. All these tests were 13 made through cladding. But again, the cladding war a very
-14 high, yo. know, surface finish. A very good coupling in all
(,) 15 four of those plates.
16 DR. SHEWMON: And if somebody hired you away from 17 the NRC and said they wanted to do a test to look for half 18 inch flaws underneath cladding and convinced the NRC that 19 none of them were there so they could run their PTS limit on 20 up, could they do it over the next five years and convince i
21 you or -- j i
22 MR. DOCTOR: Convince me that defects of that size )
p 23 were not there by -- l 24 DR. SHEWMON: Knowing that you're looking just ]
25' beneath the cladding.
l Heritage Reporting Corporation !
(202) 628-4888
l I
372 l I
1 MR. DOCTOR: I think that other clad region ~-- if
.fm
(_) 2 tne surface finish is adequate I believe.you can find those 3 with high reliability. ,
The real question and the unknown '
4 with regards to many of the reactors is what is the quality 5 of that surface finish. We found in terms of the~ work that L- 6 we had done on surface finish that'we looked at the 7 requirements that existed for surface finish, the only ones 8 we could find in code was based on radiography and based on 9 that we saw that you could easily have like a 60 mill step
- 10 which is not very much of a step.
11 We set up a calibration reflector and made a 12 measurement and then milled in a 60 mill step. And the DB 13 response dropped by 12 DB off of a standard 10 percent slot.
14 So if the surface finish is poor you're going to 15' have severe degradation in terms of being able to find
$()
16 things. And I see that as being the real key issue. I 17 mean, you've got good surface conditions you can penetrate 18 it. There's no question in my mind. I've been working for 19 the laat 10 days down at the youth center inspecting a 20 series of pressure vessel dropouts that they've' implanted a 21 number of defects in and I'm fairly convinced that I'm doing 22 a good job of penetrating that.
23 DR. SHEWMON: I don't know if we're going to live 24 long -- if I'm going to live long enough to see one of these 25 things around, but that's certainly a defense out of the PTS 1
Heritage Reporting Corporation (202) 628-4888 l
L i 'o 373' i if .ruleLwhich'is getting.more stringent.
?( ) ~2 MR. DOCTOR: I'think the key there is going to .o
't 3 the-surface' finish. I think'it's going' to be one of the e.
L 4 dictating things-in terms of being able to build a case that 5 you can inspect reliably through it.- And I think in order
.6 for you to do it, in'some cases, there's going to have to be 7 1some machining done.
- 8. I mean, if you leave -- it has been manually 9- welded and you leave it in that condition that's.really-10 tough to penetrate that effect. It just breaks up the sound 11 field so much.
,12 DR. SHEWMON: Okay, thank you.
13 MR. DOCTOR: I think -- I want to talk about PISC
- 14' III. I' m going to jump ahead a series of my handout. I'm going to go to this one and perhaps just jump to -- explain
{ } 15
- 16. what kind of activities are involved and then go to a status
-17 viewfoil'on that in light of the time since I've chewed up 18 most of the hour already.
_19 But in PISC III we're' working in a cooperative 20 international participation in quantifying and understanding 21 the limits of ISI on light water reactor components, both 22 conventional and advanced NDE techniques.
23 Some of the things that we have done in terms of 24 analysis, we run a screening study published, this is 25 NUREG/CR-4970. Had 18 teams from around the world to look Beritage Roporting Corporation (202) 628-4888
I b-374 1 at a series of cast stainless steel specimens.
() 2 We're involved in the design and analysis of 3 austenitic stainless steel round-robin tests to give us the 4 kind of data we feel we need to build cases on both 5 capability and reliability of being able to inspect defects 6 in cast type structures.
7 Provide technical input for the human reliability 8 studies that are being conducted in conjunction with the 9 PISC III program, We're looking at it in terms of ensuring 10 that all of the results that come out can be applied in 11 terms of conditions and practices here in the United States 12 in terms of the ISI that is there.
13 One of the thing;s that comes out of this kind of 14 participation is that we get early feedback of PISC results
(~') 15 so that you can use these with regard to getting things set w/
16 up for making recommendations to code and regulatory bodies.
17 Work in terms of coordinating U.S. input and 18 participation into the PISC III program. And we have 19 provided some samples. For example, we had part of the 20 shell course from, it was Hope Creek, contained two BWR 1
21 recirc inlet nozzles. That was provided as part of one of 1 22 the actions. We provided austenitic steel pipes that 23 contained integrated stress code and cracks and some blank 24 material. And we've also provided thermal fatigue cracks.
25 Action four has three different activities Heritage Reporting Corporation
,, (202) 628-4888
()
375 1 associated with it. One-is a capability study looking at n
(,) 2 what is the capability of procedures to detect in size and 3 discriminate between flawed and unflawed materials in both 4 the wrought and austenitic cast structures.
5 Pai'ametric studies are designed specifically to 6 look at important parameters such as what is the importance 7 of microstructure. What is the importance of, for example, 8 the weld geometry. If, for example, one has a sharp and.
9 very close counterbore transition zone. What happens if you 10 get a flaw coming out of that, at what depth does it get to 11 before you're able to start detecting it. So those are the 12 kind of things that are being addressed here in the 13 parametric studies.
14 And then, of course, the reliability studies are
( ) 15 addressing the issue from the reliability simulating actual 16 in-service inspection type conditions with realistic type of 17 cracks. And we have the people do the same thing as what 18 they would do out in the field to contrast with this where 19 the specimens are actually shipped to'the various 20 laboratories that are participating in the study.
21 And if you'll allow me I'll just give kind of a 22 highlight of the status of some of the activities going on.
I 23 Currently right new we have one nozzle plate that 24 is going to be here in the U.S. shortly. It's going to two i 25 teams here in the U.S. spending about five weeks. We have Heritage Reporting Corporation (202) 628-4888 l
376 1 two other nozzles that are being shipped to the U.S. this
'~
2 summer, one in June and one in July. Right now there are D) 3 two teams that are scheduled to receive and inspect both of L 4 'those.
5 In terms of the action four here in austenitic 6 steel test, the capability test the wrought stainless steel 7 test is scheduled to start this summer. The. wrought to the 8 cast is something that we haven't finalized and exempt 9 materials that are going to be used in that. So that one is 10 open at this point.
11 The casts right now we're looking at, the fall of 12 '89, to initiate that. Microstructural studies are in this 13 coming summer. Defect type will begin this summer. Weld 14' geometry effects will begin in the spring of this year.
15 Reliability tests for wrought material will begin 16 in the fall. And the cast material, because of the number 17 of defects that have to be introduced we won't have the 18 material until towards the end of the summer, so it will be 19 early 1990 before we'll actually be able to initiate 20 reliability tests on the cast material.
21 That's the kind of starting time frame. These are 22 probably going to be about two years in duration in terms of 23 the number of teams that have proposed to look at the !
24 various tests in wrought stainless steel materials.
j 25 And the reliability test down here, we've got Heritage Reporting Corporation (202) 628-4888
377 1 quite a few people that are interested in this. A number of f $
(_/ 2 people interested in this but it's contingent upon their 3 performance in the capability study.
4 DR. SHEWMON: Those will be in the U.S., a couple 5 of places in Europe, and once in Japan or where?
6 MR. DOCTOR: Well, in terms of the capability 7 those specimens will be shipped anywhere to any laboratory.
8 In terms of the reliability test, you know, that's 9 going to be restricted. The number of sites or wherever 10 those tests will be conducted hasn't been turned up yet.
11 Probably one in -- probably Italy. One here in the U.S.,
12 the site of which I don't know where that would be. And in i l
13 terms of the Japanese I'm not sure whether there's enough 14 interest from the Japanese at this point to warrant actually
( ) 15 having it set up for them to do the inspections there.
16 DR. SHEWMON: Well, is there any interest in the 17 capability part in the Japanese?
18 MR. DOCTOR: Oh, yes. Yes. )
19 DR. SHEWMCN: Okay.
20 MR. DOCTOR: If we could move on and talk about 21 the new inspection criteria tasks. The objective of this 22 was to develop methodologies which will provide technical 23 bases for new inspection criteria to meet safety goals and 24 simply low component failure probabilities.
25 The scope of this pertains to basically all the Heritage Reporting Corporation J
-) (202) 628-4888
e:
378 1 light water-reactor materials.
y ' ( )' 2 I would like to just kind of highlight a couple of 3 things and talk about these last two in a little more 4- detail.
5 Two documents that we have been working on: this 6- conceptual approach document was written and is at the NRC 7 undergoing review. And we've done that review of available 8 past experience from the NPRDS database. We drafted a 9 report and it's currently undergoing review within PNL. At 10 the end it will be sent off to the NRC I would say at the 11 end of April.
12 We've done some work looking at -- to the use of 13 PRAs for establishing ISI priorities. We have looked at 14 actually four different plants: Oconee 3; Surry 1; Crystal River 3; and Calvert Cliffs 1.
[ } 15 16 We've done the first three.and we're currently 17 working on the fourth one. And I'll describe the kind of 18 work that we're doing in regard to that by using the pilot 19 study that we did over at Oconee 3.
20 And then there's also a special ASME task force on 21 risk based inspection guidelines that we're participating 22 in.
23 Now the kind of thing that we're doing with regard 24 to the pilot study that has an expanded now to four plants 25 was, we first went through this and we wanted to demonstrate Heritage Reporting Corporation (202) 628-4888
379 1 . whether,or not the PRA type of approach could be-used as a 1(O_j. 2 . basis for establishing ISI priorities.
3 In this particular case there was a good PRA that 4 was in existence that we could employ. We had good 5 information here on operating experience from this NUREG 6 document. And we used what we called a weld inspection 7 importance parameter that I'll describe in the next 8 viewgraph. But it was applied to establish the risk-based I
9~ ISI priorities for ten systems or components.
10 And then we did a detailed analysis of the 11 emergency feedwater system. And I may just highlight just 12 one little part of that. I want to get through this and 13 since I've got to cover those other topics I can't -- this 14 is the last part of this NDE program. .But I can't get into 15
() too much detail because of time constraints.
16 Well, basically there are two parameters here.
17 The birnbaum importance parameter that's shown over here.
18 And what this is it says, if in fact we lose this system 19 what.is the probability that we're going to uncover the 20 core.
21 And if you look at this you're going to see that 22 the reactor pressure vessel, obviously if you lose that 23 you're going to uncover the core. So that's rated number 24 one and then you rate everything else accordingly. And 25 these were the kind of numbers that we came up with.
Heritage Reporting Corporation (202) 628-4888
- _ _ _ _ = _ _ _ _ _
380 l
1 Then on that NUREG document we look at that in !
2 terms'of a failure history and we said, what's the 1
3 probability that we' re actually going to have some kind of a - j l
4 failure in this kind of a system that might lead to taking I
5 that system out of this normal mode of operation. And so 6 then we got some -- a new series of numbers.
7 And what we're trying to look at here is, this is l
0 really the safety importance of the system, if you will, and .]
4 9 this is based on operating experience. And this is where we 10 had problems. What we're trying to do is identify whether 11 or not there are certain systems. For example, number five 12 here high pressure injection, it was rated number five in 13 terms of safety importance. Yet in terms of, if you look at 14 these numbers, in terms of where they have had problems you
() 15 can see that they've had a lot of problems with this 16 particular system.
17 So what this rating does, it tell us which ones of 18 these we should move up. Based on this we'll never move a 19 system down. But it says, these are the ones important to 20 safety and which other ones are they having a lot of 21 problems with in terms of system that might elevate the 22 importance of it.
23 DR. SHEWMON: Did you keep the number fixed, if 24 you move emergency feedwater up steam generators is going to 25 come down, isn't it?
Heritage Reporting Corporation (202) 628-4888 l i
. __________._m_____._m__
381 1 .MR. DOCTOR: Well, no, but what we're trying'to do-O
(_) 2 .is we're'trying to find class of systems that,'let's say --
3' you know, over here let',s maybe take all-the ones that are 4 either.this material or higher. And you have this system,.
l 5 this system, and'this system. So those are all -- they're 6 really important systems from safety.
.7 DR. SHEWMON: Okay, fine.
8 MR. DOCTOR: And then you look at these and'you 9 would say, well,'all of the ones that are 10 to the minus 10 six are the more important ones. Then you say which ones 11 are newLthat weren't included in that list; steam generator 12- becomes important.
13 So this is the kind of approach that we've taken.
-14 And we've gone on and looked, for example, a system --
( ) 15 emergency feedwater that was shown earlier and'we've listed 16 the various-elements within that. Over here we put what 17 kind of inspection requirements, ASME code currently 18 considers for those.
-19 You can see here that in general there's a fairly 20 good correlation here between our ranking that we came up 21 with this approach and in terms of how the ASME code deals 22 with it. The only one that is somewhat of an anomaly is 23 this supply lines from the upper surge tank. And those are 24 basically a gravity feed type lines. At this time they're 25 in no class until they can do basically anything they want Heritage Reporting Corporation (202) 628-4888 l:
382 l 1 to with them. There is no requirements from the ASME code 2 standpoint that one does anything with those. But it's 3 unlikely with just a gravity type of a flaw system that, you 4 know, they're going to' fail. But perhaps one should do or 5 have some kind of a program where you at least do some kind i
6 of a visual inspection to make sure that the line isn't 7 severed let's say.
8 So there are a few anomalies that this thing is 9 pointing out. But in general what we're seeing here is a ;
i 10 fairly good consistency among the various elements or 11 components within the system and what ASME code has said.
l
^
12 Our selection plans for the PRA is to cover all 13 manufacturers, all the major models of the manufacturers.
14 The next four PRAs that we're going to address are shown 15 here at Peach Bottom 2; Sequoyah 1; Zion 1; and Grand Gulf.
O 16 In those we haven't put in the second combustion engineering 17 one where we have two or three of the others'and the reason 18 is that-there are not very many good PRAs available right 19 now in combustion engineering plants other-than the one that 20 we have. In about a year there will be several good ones 21 available, but currently there are not.
22 Our conclusions in terms of looking at those first 23 plants are that in general we have seen that there is good 24 consistency from the three plants that we have already 25 compared of looking at the different systems of those three Heritage Reporting Corporation (202) 628-4888 )
(2) 1 f
l
383 1 plants from plant plant. And that's extremely 1
/"N - 2 encouraging.
V 3- As I mentioned, we are finding a few exceptions 4 and when we get into the detail I suspect that there will be 5 probably be more because of perhaps some plant-specific 6 considerations. But right now the only one that I've 7 highlighted here is the upper surge tank, is there something 8 that needs to be done with regard to that, for example, to 9 do a visual exam or so ne other requirement with regard to 10 that system.
11 Well, what we're doing, if I can conclude on this 12 particular viewgraph, is working with this ASME task group 13 for tash force on risk based inspection guidelines. The 14 first meeting on that was held in last November and 15 basically set goals and initial assignments.
gq
%/
16 We plan on meetings four times a year. And we 17 pull together a cross section of people, not only from 18 nuclear but civil, industrial, insurance, aircraft, and
.19 academic organizations.
20 The phase one of this effort is focusing on the 21 nuclear concerns of Section 11. However, they plan to 22 expand that to generically across the board.
23 The next meeting is planned in April.
24 DR. SHEWMON: What does across the board mean, how 25 wide is the board? Is this all nuclear or are you going to Heritage Reporting Corporation (202) 628-4888
,G v
384
- 1 get'into aircraft or what?
2 MR. DOCTOR: My. understanding, what I mean by that' 3 is across the_ board as far as what ASME code applies to.
4 That is the intent anyway'and why they've gone to this cross 5 section of people.
6- The output from this is going to be a document 7 containing recommendations for-changes of code and that 8 highlights one thing that once they generate that document 9 there's going to have to be work done in terms of taking 10 what is the series of recommendations and' putting that into 11 ' code language and then trying to move that through the code 12 process.
13 ,
So even once it's generated there still'is, I'm !
1 14 sure, quite a bit of work involved with moving it' through f
15 the code. However, initial feedback that has been received 16 has been promising. So I think that if this is'done 17 properly and the code is informed of what's happening as 18 this' work evolves, then there will be hopefully a' fairly 19 straightforward path and a follow-up to move that in the 20 code. ]
21 DR. SHEWMON: Sounds very good.
1 22 MR. DOCTOR: And that's all I wanted to say on the 1 l
23 NDE reliability program. And I took longer than I should 24 have so I'll have to reduce some of the material in the 25 other presentations.
, l Heritage Reporting Corporation (202) 628-4888 O
385
~
.1- -DR. SHEWMON: Do you want'to'get'a drink before h( ) 1F :you gozon'to the.next.part.
3 MR. . DOCTOR: I appreciate that. .These people wanti I
'{
- ) 4 -to be out of here'by 3
- 00 so.I'better get started.
~
5 What I wanted to move on to-next and discuss is'a 6 program dealing with field validation, acceptance, and 7 training for nondestructive' examination technology.: Andi 8 this is focused on'the. technologies that the NRC has'been 9 ' developing.such as the SAFT technology, acoustic emission 10 ' technology, and the eddy current technologies.down at Oak 11 Ridge.
12- The-objective is to develop field' procedures, I
13 performance of field-validation testing,-and providing
'14- training-for.NRC headquarters and regional staff, and then j 15- working'with ASME code for the'use of these advanced
- 16L technologies.
17 The kind of things that we' re doing is conducting 18 workshops with NRC headquarters and regional staff in these
'19 technologies. Pursuing the acceptance of these in the ASME
.20 code. Working with regional staff trying to identify places 21 where this technology can help them to solve problems.that 22 they're having in ISI. Pursuing the validation testing of 23 both the acoustic emission and the SAFT technologies through 24 utilities, international agreements, and things such as PISC 25' program.
Heritage Reporting Corporation (202) 628-4888 i
9 r
kr ~f '
386 1 And of. course.doing the field validation testing:
j } 2' of the. technology at sites where we can get the technology
!- '3_ in to help resolve and clarify defects or indications.
4 Regulatory issues are improved defect 1 detection 5 reliability. . Improved defect characterization accuracy.
6 Alternative methods for surveillance of flaws. Improve leak ,
7 ' detection. And of course ALARA as you like to reduce the
.)
8 exposure so that you can use remote sensing, for example, ;
9 with AE, you can put that on you don't have to spend time 10 scrubbing a pipe as an example. . Or you can use an automated 11 system versus sending one in to do a manual inspection, you 12 can reduce ALARA.
13 I'd like to first start out talking about the 14 SAFT-UT technology. How specifically does this help nuclear
'15 regulation? It provides a mechanism for the technology to 16' flow from NRC contractors at PNL to the NRC staff.
17 Provides a mechanism to allow the code activities 18- to continue to support the long process of moving the 19 technology into the code.
20 Furthermore, it provides technologies that can 21 solve inspection problems and thus can aid NRC staff in 22 assuring the structural integrity of components.
23 I would like to just talk briefly about some !
24 highlights, some things that are being closed up on terms of 25 the original SAFT program, l u
l' j Heritage Reporting Corporation j
. (202) 628-4888 1
()
4 387 1 We completed the system documentation, thisi
.2. .NUREG/CR-5075.- With that you can duplicate our system-3' .completelyLand it tells you all of.the operating features
~
4~ that the system has.
5 We've developed a final report. The one element- ]
1 6- that;we.need to put in~that is' current 1'y what's going on and- !
7 it is in progress right:now at the EPRI NDE center. We're 8 looking at a. series of thick session, in other words J9 pressure vessel dropouts. 'And.that contains a variety.of.
10 Edifferent induced defects, cracking, lack of fusion, slag, 11- velocity, et cetera.
12 We're'also will be looking at a series-of.under?
113 ' clad current blocks that they have. And.then we will be 14- going-through the integrated stress corrosion detection and sizing' performance demonstration.- So we've.got a lot of
( } l 15'
.16 work in progress; we're partly through.that right now'.. I' JL7 interrupted that to come up here for this meeting.
^* -
18 Some of the things that we've done. We have 19 conducted a joint inspection on Indian Point unit two.
20- .There was a vessel indication in this, in-the bulk line 21 region on the OD and we worked in conjunction with Dynacon 22 and Westinghouse in which they collected the data for us 23 using a probe that we supplied and we then processed it.
24 .And so we worked in conjunction with other commercial
- 25. equipment to actually employ SAFT technology.
Heritage Reporting Corporation i
, (202) 628-4888
388 1 DR. SHEWMON: It probably was on the outside.but I)) 2 the sensor was on the'inside.
3 MR. DOCTOR: That's correct. That's correct. So 4 you look through the cladding and you had to look, you know, 5 it's about 10 inch vessel wall in that location with a metal 6 path and 45 degrees. You're looking like 15 inches away.
7 So it's a long metal path.
L 8 We have done some other things, we worked in 9 conjunction with Amdata, Combustion Engineering to look at a 10 series of indications in the cast elbow down at Trojan 11 Nuclear Plant.
12 We've conducted a training session with NRC Region 13 I personnel. They came out and. looked at a series of our 14 IGSCC that we had that were field removed. And they got an exposure to the SAFT technology and saw it in operation.
(~ f 15
,. 16. And we took the system -- in October this last 17 year back there and put on a demonstration of the technology 18 and also presented a seminar on SAFT fundamentals to them.
19 There was also participation of headquarters as well as 20 other regions at that particular meeting.
21 Currently, as I said, SAFT is undergoing this 22 evaluation on this series of blocks. We've also had 23 discussion with MPA on the use of the SAFT technology on the 24 full scale vessel test and we're still trying to work out 25 how that in fact can become a reality.
Heritage Reporting Corporation (202) 628-4888
389
'l I put this in here just a mention. One of the 2 things that the Japanese will be doing, applying line SAFT 3 in the program of PISC phase III on several of the different 4 actions. So SAFT is going to be included, although it will 5 be only the two dimensional form of it and we want to be 6 involved in applying the three dimensional.
7 DR. SHEWMON: How many hours would it take the 8 level III on UT to pick that up and actually use it; one 9 week?
10 MR. DOCTOR: Yes. It's not really that 11 complicated because it's all based on the same acoustic 12 fundamentals. -It's just a matter of getting accustomed.
13 Most people, you know, if you put a very diverging radiation 14 field into a part, they know they're going to get a lot of g- 15 noise. Well, if you don't get processing you obviously are V 16 going to have a very noisy signal. And so when you look at 17 the scope you see a lot of noise, but when you look at the 18 processed image you've got a very high signal noise ratio.
19 And the thing that you have to overcome is the 20 fact that they look at something that doesn't look very good 21 and they get a feel of the picture and they don't have a 22 good warm understanding of where that comes from. So that's 23 the main hurdle.
24 Let's see, I've got a few pictures in the handout 25 that describes some pictures of results from Indian Point Heritage Reporting Corporation (202) 628-4888 m
U i
~
l-p 390
-,- 1. and' Trojan _. I mean, .you can look'at those'and they're
([ 2 pretty self-explanatory. I have slides of them, I don't 3 have viewfoils. But they show z what' kind of results we 4 obtained, you know, looking at'a calibration-notch at' 5 Wallace Mills. And then it shows the following one -- well, 6 if I'm going to say'that I might as well turn on the --
7 wrong way.
8 Maybefjust point this out. If'you're not' familiar 9 at looking at this kind of picture. This happens to'be --
10 there's a notch located right here and this is'a corner' 11 trap. And.the' point that I want you to see here, as this 12 signal is smeared out across, where surface'is located, q 13 basically uniformly distributed both above and behind that'
.14 line.
,Fi ' 15 This kind of pattern that'you see in here like
%)
l
.16 this is an effect due to the cladding.. Okay.- If you look
(
17 now, this is the response that we obtained from indication 18 at Indian Point. had you can see,.here is thel indication.-
19 Here is the back surface. You can see the signal is 20 predominately above the back surface. And that was the 21 basis for.why we concluded and just did not look at all like. 1 22 the signal that we obtained when we had a surface connected i
23 response. i i
24 And then.I'll.just put this in. Here shows some 25 of the kinds of response looking at the cast elbow from the l b 1 i Heritage Reporting Corporation (202) 628-4888 ]
O l
1 i
391 1 Trojan plant. You can see this is a C-scan looking down on (f 2 the~part. We scanned across the surface in a pattern'like 3 .this and we're looking down on that information. And you 4 can see that there is a substantial number of indication 5 located in these three. areas.
6 I just wanted to briefly show that and really not 7 spend anymore time.
8 What I wanted to do was really go on to this.
9 We've been pushing, as I mentioned, in the Section 11 work 10 to go to performance demonstration. And we wanted to try 11 and get the SAFT technology involved in the code. What we 12 found in Section 11 is that they don't want to go to any 13 kind of prescriptive procedures. Simply we're having large 14 difficulty with that.
15 So we went to Section 5 and discussed it with them 1[)i 16 and they're very interested in having this kind of
( l'7 technology and procedures into Section 5. The reason is 18 that currently the only technologies that they have area 19 conventional technologies. They have no procedures for 20 automated; for data processing; or imaging system currently 21 in Section 5. So they wanted to create a new Article 4 22 entitled " Computer Processed Imaging" in which we will draft 23 input for the SAFT technology and other people will be 24 encouraged to provide input. People, for example, that have l- 25' P-scan or the M-data interspect 98 system. If they want to Heritage Reporting Corporation (202) 628-4888
392 1 go to the trouble of developing the procedure and putting it
,-).(, . 2 in there, the code will be excited about getting'it and it 3 would go into this article.
4 DR. SHENMON: How what is Section 57 5 MR. NUSCARA: Section 5 essentially is the section 6 of'the code that writes all the NDE procedures. And then 7 the other sections of the code like 11 references those.
8 So, 11 doesn't write the specific procedural types of the 9 code but Section 5 does.
10 DR. SHEWMON: Okay.
11 MR. DOCTOR: Well, our FY ' 89 plans are of course 12 to finish the blind testing that's going on at the EPRI NDE 13 center. Publish our final report. And get this draft 14 version of the write on Appendix IV for the SAFT_ technology
(
.]) 15 developed and distributed so that we can start getting 16 feedback from it.
17 Now the acoustic emission technology is the next 18 item. This is part of the same program so I will just 19 continue on and describe it.
20 The AE technology, how does that help nuclear 21' regulations? Well, it performs an alternative method for 22 surveillance of identified flaws to detect growth which is 23 extremely important because you can do this continuously on 24 line and not at some periodic interval when you don't know 25 how rapidly, for example, something may be going.
Heritage Reporting Corporation
_ (202) 628-4888 v.)
_ . _ _ _ _ . _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __m
p' V r -
I 393
'l- - So it'enhancesisafety by continuously surveilling 2 and.-it increases the regulatory flexibility in dealing with' g, 3 such. cases. ,
4 It provides a very sensitive leak detection
! ~
.5 methods'that~can provide you with.very timely information to 6 because you can.do it' continuously.. It's very sensitive and 7 so'it increases operational safety.
8 It reduces the need for manual inspection'of 9 service sensitive piping and nozzles and thus reduces 10~ personal exposure.
.11 - -And it provides continuous monitoring for early 12- detection.of cracking in the primary pressure boundary. It i 13 increases operational safety by timely detection and
- 14. location of' active flaws sites.
15 So those are the kind:of problems that it-2{
16- addresses and deals with. Some of the -- since this hasn't 17 been presented for a while I went back and recapped things 18- in?FY.'88 and included things in FY '89.
19- One of the things that has been done is that it 20- has performed -- the AE is performed in a source location 21 . calibration work with' installed AE system at Watts Bar unit' 22- 1. This is a case where they started out very early with 23 Watts Bar and saw the system and went through cold hydro and 24 hot functional test. And since it's going to be several 25 years before they start up one of the things they wanted to Heritage Reporting Corporation (202) 628-4888 O
p 394
$ IL do.wasito go back in there and actually be able to. inject 2' . signals at various locations on:the vessel and be able to f 3 evaluate the performance ~of their-system with regard to how 4 accurately they.could actually detect it and locate those
'5 . signals.
'6 So'they were able to do that and got the. level of 7 performance.that they had expected to=have and were very.-
8 happy with'the results.
9 A non-mandatory appendix on AE monitoring was 10 prepared. Itwassubmb.ttedtoSection11. It's part way-11 through~the acceptance. process. And.it was actually tabled.
12 then at the subcommittee level.- The decision at the .
- 13. subcommittee was that they needed to have'some kind of 14 feedback from actual applying it out:in the field before- j 151 they put in place a non-mandatory. appendix. They said, 16 submit a code case, we will get through the process, let's 17 'get some data assembled'on actual field application,'and 18 then we will then move this through.
19 So at the current. time they set up through the 20 main committee this code case. There were three negatives 21 that were received and there were several good technical 22 things. Some other were just simply misunderstandings.
23 There's no problem and they feel at the next main committee 24 meeting in May that it should go through and then on to the 25 MCS for their approval.
Heritage Reporting Corporation (202) 628-4888
-0
395 1 So-that looks very good. It's moving along.
(b 2- There really-is no problem with that and:I think it's' fairly
.I 3 standard to have a few things pop up like what happened with l 4 .that code case. .;
5 There are some discussions that have occurred with 6 AE monitoring with Davis-Besse, Toledo Edison plant. It's-
- 7. back-on information that has been supplied to them. They j i
- 8. requested a cost estimate for applying the AE tschnology to 9 a high pressure injection nozzle which contains some fatigue 10 cracks. So they're in the process of evaluating that .
l 11 information. We're hopeful that that will work out and 12 there will be an opportunity to apply.the technology in this 13 particular case.
14 Another case that has been pursued is with Taiwan
()15 Power. They express interest in the AE monitoring and 16 actually paid to have Phil Hutton go over there and discuss e 17 the inspection problem with them and requested a proposal be 18 submitted and it was. For some reason it was rejected and 19 the technical reasons for why it was rejected I really don't
[ 20 have a sound foundation. We are not sure as to yet, you 21 know, the reason why it was rejected. But that was 22 something that was looking very hopeful and it's unfortunate 23 that to date nothing has panned out on that.
24 DR. HUTCHINSON: Would you characterize this work 25 as an attempt to develop a method? Is it proven technology l
Heritage Reporting Corporation (202) 628-4888
396 1 in any sense?
h2 MR. DOCTOR: -It has been proven in a number of 3 cases. As a matter of fact, ASME code allows and accepts 4 the'use of AE for some particular special applications. So, 5 I mean, it's not a technology that doesn't have a firm 6 basis.
7 The technology unfortunately was sold early and 8 conceptually it's an easy thing to understand. But there's 9 many noise sources that also provide information and the 10 problem is it's discriminated against those. That's one of 11 the major things in this program, it has been the 12- development of the ability to reject noises and sources that 13 don't pertain to the indications coming from flaw growth.
14 DR. HUTCHINSON: Is it imagined that the flaw has 15 been identified and that you're focusing on a given form?
(
16 MR. DOCTOR: Well, either one. I mean, you can go 17 in and instrument part of the system. Let's say, maybe a l
18 particular loop, put a number of sensors on that to simply 19 listen, to use that as a way of checking as to whether or 20 not in fact soma kind of degradation is going on. It's 21 omitting acoustic type event. So it could be used in 22 replacement for potentially down the road of going in and 23 doing, let's say, ultrasonics. It hasn't been done and it's 24 going to take quite a bit of data before people are going to 25 buy that.
Heritage Reporting Corporation (202) 628-4888 1
. 1 I
397 1 But potentially that's one of the things that l
()
/ ~, ,
2 could happen. Right now we're at a place where if you do i 3 have e, crack growing there has been a substantial amount of 4 data compiled, not only by us but other researchers, that 5 show that you can go in and you can monitor that. And you 6 can understand if, for example, the flaw is growing in 7 length dimension. It's very clear that that is happening.
8 You can monitor, you know, the depth, the thing is growing 9 deeper.
10 The difficulty it has is trying to quantify the 11 events with how much actual depth growth has occurred. But 12 you can tell that depth growth has occurred. It's just 13 quantifying that with high accuracy that has been the 14 struggle.
( 15 MR. MUSCARA: We have been trying to characterize 16 the flaws and we have correlations that relate the acoustic 17 emission to the crack growth rate. So if you have some i 18 indication of the stresses you can calculate. And that's 19 done with -- there's some scatter in the data but some data 20 has been developed and we have a conservative line to
~
21 describe that. And all of this is part of the monitoring 22 equipment, discriminate the flaw from other noise source.
23 Once it's discriminated it goes through a model we are now 24 characterizing.
25 All of this has been proven in a large scale test, Heritage Reporting Corporation 3 (202) 628-4888
(_)
l l
l f\.
= _ _ - _ _ _ _ _ _ _ _ - _ _ _ _ - _ _ -
398
,1 butinot.'on reactor.
2 MR.. DOCTOR: Some of the activities. involved in
'} 3 technology. dissemination are highlighted-here.- The AE 4 . program participated in the AE technology seminars.that-
-5 . occurred in October at. King of Prussia, Region 1.
6 Supplement to the ASME Section 11 code case and non-l 7 mandatory appendix includes details of the technology so 1
8 that the user can understand what is being recommended with 1'
9 regard to actually applying the technology for the 10 applications that are highlighted in both code case and the-11 non-mandatory appendix.
12- The technology has been applied to monitoring the 13- high' flux isotope reactor vessel at Oak Ridge during 14 overpressurization testing for checking for crack growth.
15 This was done, I believe it was in '87.
LO 16 Technologies have also been applied by the 17 laboratories for monitoring vitrified waste canisters to 18 detect cracking of the glass during the cooling cycle. In 19 other words, you mix this up with radioactive waste and you 20 pore it into the canister and one of the critical things you l
21 want to know, you would like to have that thing be a solid 22 cake in there. So that if something should happen to the 23 outer containment you simply have a solid cake. If you get 24 a lot of fracturing the concern was that you could get 25 leaching and et cetera.
Heritage Reporting Corporation !
(202) 628-4888 0>
)
399 1 So one of the things we did was instrument and k 2 show that you could actually determine if you had cracking 3 occurring during the cooling cycle.
4 The AE program results have been published in a 5 number of different forms. They've been published in 6 Materials Evaluation; Journal of Acoustic Emission; and 7 there's a volume 5 of the ASNT Nondestructive Testing 8 Handbook on Acoustic Emission. And Phil was and the 9 organization were involved in developing a number of 10 chapters in that, editing and getting that published. So 11 there's a heavy element of this work that is involved in 12 that particular handbook.
13 And the main emphasis for the remainder of this 14 year are to continue efforts to try and locate an 15 appropriate place for installation of the acoustic emission h) 16 system because you want to be able to monitor a couple 17 different areas. We would like to monitor an area where 18 you've got a defect in there that you know it's going to 19 grow. And you would like te monitor another area as a 20 comparison, if you can, so that you've got a comparison to 21 show that, yes, it is working. You are monitoring and 22 seeing a chain. And another case showing that it isn't just 23 noise. That in fact -- or you have maybe applied or 24 repaired or mitigation that you can verify that nothing has 25 happened.
Heritage Reporting Corporation (202) 628-4888
400 1 MR. MUSCARA: May I update a little bit on this l) 2 activity.
v 3 DR. SHEWMON: ,
Briefly.
4 MR. MUSCARA: Very briefly. Just last week we 5 have been talking with people at PEPCO for the Limerick 6 plant, they have a crack in the safe end and they are very 7 interested in doing the acoustic monitoring. As of last 8 Monday this week we were working with them to try to 9 accomplish instrumentation of that component by mid April.
10 DR. SHEWMON: Maybe that will keep Keith off their 11 back.
12 MR. DOCTOR: That's all news to me, Joe.
13 Of course the other item here is to complete the 14 ASME approval of the AE code case. So that brings you up to g-) 15 speed in terms of activities. This is really winding down
%)
16 of both the SAFT and the AE technology. Doing some final 17 validation and getting them in the code and getting them 18 accepted by industry.
19 DR. SHEWMON: What happened to CE, was it, for a 20 while was talking about trying to get a SAFT/UT piece of 21 equipment on the street. )
22 MR. DOCTOR: If you look at their activities they 23 have been doing relatively little nuclear work. 90 percent 24 of the things that they have been involved with are 25 associated with Aerospace NASA and military inspection of Heritage Reporting Corporation (202) 628-4888
,/
i l
1 i
^
- l' m
401 1* -~ 1 missiles. So they have'been doing relatively little nuclear-
-h ?2 work and that's the reason'-- they've actually applied the-
,y 3. SAFT technology.in'some of-these other. application. areas, 4 but they haven't -- they haven't been able to do it in terms 5 of nuclear application.
6- That's-just kind of unfortunate that we haven't-7 been able to apply it there, but they're still learning and
, 8 gaining knowledge'of the technology.
9- DR.'SHEWMON: That's not all bad.
R10 MR. DOCTOR: That's not all bad, that's right.
11- MR. MUSCARA: Well, granted the nuclear 12 technology, this past fall thsy have used the SAFT 13: technology on an ir.spection in Japan on'a shipping. cast-
'14 that's' cast iron. And they had very good'results with SAFT and not so good results with the techniques. So they're
() 15-l 16 very pleased with using that technology. We' re using two 17 dimensional SAFT.
18 DR. SHEWMON: Maybe the Japanese will peddle 19 American -- go ahead.
20- MR. DOCTOR: I know three other companies that are 21 doing two dimensional line SAFT over there in terms of the 22' use of V-mass and in terms of ultrasonic growths for 23 inspection, developing the technology for nuclear 24 application.
25 As I noted earlier, they are going to use it in Heritage Reporting Corporation (202) 628-4888
}
,1 402
'l- the test three program,.several of the companies.
2_ What I wanted to wrap up with and it looks-like
)
3- I'm pretty much1on. schedule and should be done in the-4- - remaining 20 minutes, is to describe the work we have been 5 doing'on aismall little program on'looking at the 6 development of a program plan for non-destructive 7 measurement methods for the measurement of material 8 properties and property changes that are due to aging.
L 9 What.I would like to do is put up first what the 10 . objective is and what we're trying to accomplish. We wanted 11 to set out and review what's'known about this particular 12 topic through looking at the literature, talking with 13 experts,-looking at any related activities in this area, and 14 really compare which we call a state.of the art document 1 15 that continues a plan for the development of engineering 16 databases and validation of prototypic systems for non-17 destructive measuring material properties and the 18 degradation results in those due to aging.
19 Now how does this program help nuclear regulation?
20 Well, light water reactor materials age in grade as 21 exemplified by pressure vessel embrittlement, toughness loss 22 in cast austenitic steels, and of course fatigue damage.
23 And the NRC needs assurance that there is not an 24 undetecte'd reduction in this defense-in-depth principal -
25 that's codified in 10 CFR Part 50. This concept that's Heritage Reporting Corporation (202) 628-4888 O
403 l' . codified there includes multiple barriers to be established
) 2 to prevent the potential release of fission products that L
3 :couldfaffect public' safety.
l
- 4. .We also have here that aging should and could not 5- ' set the stage for common mode failures leading to' accident
~
l 6 initiation or reducing the capacity to mitigate accidents.
7 So that's part of the reason why we would:like-to 8 have something like this in place. In addition, there's.
9 been discussions in' industry where they have entailed 10 looking at. coming up and trying to compute what kind of 11 remaining lives.one might have. And making some assumptions.
22~ based.on amount of embrittlement, the loading, the cycles, 13 .and the means of combining these.
- 14. We have a lot of controversy in regard to what those values are and how do you actually combine them. NRC
_ ( } .15 16 really needs some kind of confirmatory tool if this is the 17 approach that industry is going to be taking to verify the 18 assumptions and the accuracy of the. extrapolations with 19 regard to how they come up with those numbers and how 20 'they're trying to extrapolate those out in the future.
21' And the bottom line here is then, there's also a 22 need that the NRC have technical expertise available to 23 provide input on aged-related NDE issues as they arise. So 24 that was the reason for putting this-together and looking at 25 it with regard to what the NRC we see needs and how a Heritage Reporting Corporation (202) 628-4888
404 4
1 program of this nature might provide that.
2 We structured this around, again, writing this llh 3 state of the art assessment with a real key element that ;
l 4 went into this was a workshop we had in which we pulled 5 together 31 people; about half experts and the other half 6 experts in material properties, and in particular in aging 7 of material properties to discuss what is the status of what 8 we know and where things can go.
9 Now the state of the art assessment, if you look 10 at the four bullets that are noted here, outlining the 11 important things that one has to address.
12 What are the technologies that potentially could 13 be used for making the measurements.
14 What kind of aging induced degradation mechanisms ggg 15 exist and how can you develop accelerated means for doing 16 this acceptable procedures that will allow to age material 17 quickly. You can't allow yourself, you know, 40 years to 18 age material, if you need that information for this in the 19 next, you know, five or ten years.
20 Based on assessments things such as the workshop, 21 technical experts input, write a state of the art assessment 22 of the NDM methods for material property measurements. And 23 of course put in a plan for how do you actually do the 24 assessment and what kind of information that will come out 25 of that plan.
Heritage Reporting Corporation (202) 628-4888
i 405 1 The status is that we have completed the research
. ;,,) _ . 2 work in terms of paper review. And if we pull that 1 3' together, we have already conducted the workshop here.
4 Currently the document is -- outline has been developed and 5 review and we have begun writing on it. Right now we' re 6 planning to have a draft report of that into the NRC by the 7 30th of May of this year.
8 Now the kind of things that came out of the 9 literature review are shown here. For example, looking at 10 various kinds of degradation mechanisms, the ones that were 11 identified in the radiation embrittlement and fatigue where 12 you can have mechanic, t14ermal, or corrosion fatigue.
13 And we looked at the'various kinds of material 14 components and it very quickly breaks down and highlighted 15 the most important of these three groups. You can see that
.[ }
16 the reactor pressure vessel is dominate in two of these: in 17 cast stainless stain, and the work that's already going on 18 in that area I think is -- I'm supporting that work.
19 Now the literature search showed that there are a 20 number of different techniques that have been shown to 21 correlate with various material properties. In this 22 particular viewgraph we selected four of the most important 23 from the standpoint of reactor applications. Of course, 24 fracture toughness and fatigue.
25 You can see here that there are a number of Heritage Reporting Corporation
_. (202) 628-4888 .
r~)5
\_
406 1 different techniques that are sensitive. And I think it lll 2 points out that taking a look in this area that there is a 3 high probability that you can develop these kind no non-4 destructive measurement approaches.
5 The real key here is that embrittlement work has 6 been done in particular with regard to the kinds of 7 materials that are applied in nuclear reactore.
8 What's shown up here is work that has been applied 9 in a broad spectrum of different applications. But 10 unfortunately most of the work has not had anything to do 11 with nuclear related type degradation and activities.
12 DR. HUTCHINSON: Could you explain to me what 13 acoustics or acoustic emission has to do with the fracture?
14 How that could possibly make a fracture toughness?
ggg 15 MR. DOCTOR: For example, if you're listening to 16 an acoustic emission event, depending upon taking, you know, 17 acoustic emission event means that, for example, you might 18 have some micro cracking or something like that, that's in 19 there. That under stress you're going to cause micro cracks 20 to grow. And if you have different fracture toughness the 21 nature of the --
22 DR. HUTCHINSON: That's awfully soft. I would 23 like to know what you mean, what you could possibly mean by 24 acoustic or acoustic emission, say, for degradation due to 25 aging?
Heritage Reporting Corporation (202) 628-4888
407 1 MR. DOCTOR: As you change the properties, for lll 2 example, and you embrittle something and you change the 3 acoustic properties at loss of propagation, for example, it 4 will shift as a function of --
5 DR. HUTCHINSON: I don't believe it.
6 MR. DOCTOR: You don't. Well, that's not what the 7 literature supports.
8 DR. SHEWMON: What you've got there is the same 9 problem you've had in acoustic emissions in stage. You've 10 got sensitive things where the signal bears no one to one 11 correspondence with what you're trying to measure.
12 And I think -- well, I won't quote you the words I 13 have written here. But I think it's a very hard and dubious 14 path to try to back anything firm out of those problems ggg 15 which will give you signals up to kazoo and relate them. I 16 suspect if you have a micro harden test you would do a 17 better job of getting yield stress and hardness which are 18 important than anything you've got up there and would be 19 less ambiguous.
20 MR. DOCTOR: I understand what you're saying. The 21 only thing is you would like to be able to go in and make 22 the kind of measurements in a zone, you know, of interest.
23 For example, you might be interested in making a measurement 24 on reactor pressure vessel, the cladding, the base metal 25 interface as an example. And you would like to have some Heritage Reporting Corporation (202) 628-4888
i 408-11; means of being'able to exam that area. Trying to do a micro
- t2 indication approach tells you what it's like on the OD or
!3 it's in on the. stainless steel. But it doesn't tell'you the 4' kind of'information that you would really like to have.
5 DR.'SHEWMON: At least a hard number. You know, 6 you've got:this -- you're pushing-it up hill with a wet 7 piece of spaghetti -just to try to get a connection between
'8 'here and there.
9 MR. DOCTOR: 'I understand. All I'm.saying here is l: 10 thatinreviewingtheb.iteraturewefoundthatpeoplehave
~
11- found correlations between these and these particular 12 . conference,.okay.
13 Now whether or not those are going to' work in
.14' actual' light water-applications, I don't'know the answer to 15 that.
' 16 DR.~SHEWMON: Yes, but there's a difference 17 between, say, if I change the yield' stress I can find some 18 change in what I get for a transmittance or internal.
19 fr'etion.. But your problem is, you want to go the other
-20 way. You want to say, I've got a change in internal 21 friction, what does.it mean; not-I changed the property and 22 'by gum it changes this in some way. And. going back is where 23 I'm very dubious and I expect my friend is here, too.
24 MR. DOCTOR: I understand it's a tough problem.
25 But there's one other things that came out in, for example, Heritage Reporting Corporation (202) 628-4888
.Q.
1 1
409 l 1 the magnetic measurements. Dave Jiles at Iowa State 1 1
() '2 University has been making measurements on bolts, looking at )
3 aging in those and'has been able to do a really excellent 4 job of characterizing those. They follow them up with 5 metallography in that they have been verified that he'can 6 actually detect very'small subtle changes in those bolts 7 that have been verified by the actual metallographic 8 analysis. DR.
9 SHEWMON: But he's coming.the easy-way. You're going to go 10 back in. You've got this hunk of steel that has been there 11 for 40 years and you're trying to go the other way.
12 MR. DOCTOR: But these were aged ones that they 13 aged, so they weren't radioactive. But he is developing 14 this, working with Westinghouse to actually be able to take 15 this out in the field so that they can make these
( })
16 measurements without pulling the bolts out. Okay.
17 I realize it's not a simple thing. But there is 18 work out there that shows that you can do this.
19 DR. HUTCHINSON: Ca7 you describe, is this 20 considered -- what you're describing now -- is this 21 considered a research program or is it consider --
22 MR. DOCTOR: At this point what we're trying to do 23 is to simply look at it and determine a state of the art.
24 Where does the technology stand? What has it shown? And 25 what are the limitations that we know?
Heritage Reporting Corporation L (202) 628-4888 l l
l
410 1 And if one was going to actually develop this 2 technology how would you develop a plan to actually assess 3 the effectiveness of this technology? Not to develop the 4 technology but to assess the technology in fact can make 5 measurements. What kind of weight and reliability can you 6 put into those measurements? What do they mean?
7 MR. SERPAN: There was a small amount of 8 exploratory -- there was a small amount of money that was 9 available in Fiscal ' 88 to look into this and get started.
10 The major product of that was to hold this workshop to try 11 to expose what had been developed. What was available. What 12 was known. Was there any hope at all.
13 What Steve is trying to do then, today, is to 14 expose that and say, yes, there have been some things done.
15 They have found some things. There are some preliminary 9 16 correlations.
17 And at this point in Fiscal '89 we don't have any 18 money in this program for it and there just isn't any in the 19 outyears yet. He's simply reporting on what we found. We 20 would love to be able to do it. It would be great stuff for 21 aging. But there's no question that it has problems. But 22 that's what we're trying to expose today.
23 MR. MUSCARA: What was most evident from the 24 workshop was the magnetic techniques and ones who have the 25 best chance of working. .And the way they' re working there Heritage Reporting Corporation (202) 628-4888 O
ga 411
~
1 as:you degrade'the material through fatigue,'for example,'in I 2 ; developing these locations. It's more difficult to move the
-3 magnetic' domains around. Because these techniques givela 4 measure of how difficult it is to move those domains.-
51 DR. SHENMON: But carbon influences.it.- Nitrogen l
6 influences it.. Infrastructure --
7 MR. MUSCARA: In the idea of developing that
'8- . technology.we.should look at a number of heats.-
Look at a L: 9 number of different degradation steps in performing-the 10 measurements and develop the correlation.- That would
~ 11 be the --
12 MR.. DOCTOR: Well, there were a number of things
-13 that came.out of the workshop ~in terms of things that had 14- ,been'donerin.a lot of different, you know, other areas.
j ).15- Measurements that'had been made on other kinds of steel.
- 16. The difficulty with most of these is that they 17 haven't been broad in terms of looking at a large spectrum.
18 Let's say if you buy 8535-B and getting it from a number of 19 manufacturers over a number of years, you know,'what kind of 20 -variability.
21- What they did on most of the studies is they have 22 taken a set of samples from a given block and proceeded to 23 do a series of different kinds of -- let's say introducing 24 dislocation densities into a, you know, tensile loading or 25 whatever and then proceeding to make the measurements.
l Heritage Reporting Corporation ,
(202) 628-4888 !
L O
4 i
C____...____._._____ __
412 1 What's reported here then are where people have n:
2 been-' successful. ,There are other. cases whereLother people have tried some of these things on other different alloys 3
4 and have not been successful.
1 5 The difficulty that you have in assessing what's 6 reported in'the paper is getting the sufficient amount of. l 7 detail of the actual experiments and the material properties 8 in a global sense. Not just only the ones that.they happen 3 J
9 to measure to really put to all the perspective.
-10 Now the workshop was the -- the really good. thing 11 was that we got-a good' chance to interact specifically with
- 12. people who had been doing this kind of work. The series.of' 13 measurements,Lfor example, that Jules has done. There's a 14 . series of 1 18 different measurements that he made on material 15 properties for these magnetic measurements'on the bolts.
' 16 And what he is doing right now is focusing on what are the 17 important ones that he needs to actually measure. He can 18' measure all 18, many of which he does not need to put a 19 system set up to do that. So he's been doing kind of an-1 20 overkill.
1 21 The difficulty with that then is of course I
22 app?ying it out in the field where you have access to only 23 one end of the bolt. They would like to do it without 24 actually removing the bolt from the component that it's
- 25. located in. And that.of course is a problem.
Heritage Reporting Corporat lon (202) 628-4888
,A
-_ = - - - - - - - - -
h 413 1 But again, there have been some very good detailed 2 experiments that were presented at this workshop where
.{ }
3 people found correlations to exist and have done extensive l
4 amount of work to show that it was not a result of many 5 other things. And this was a group that was there trying to 6 understand. And there were a lot of in-depth questions that 7 was asked with regard to the experiments that were actually 8 performed.
9 The conclusion from that series of presentations 10 was that these in fact do exist. That they can be measured.
11 It's not an easy measurement, but it can be measured. Some 12 of these techniques will never make it in the field.
13 Because they're simply not applicable. But others will.
14 I would like to respond on the terms of the 15 acoustics. We made a series of measurements on a set of --
16 not part of this program but on another program I had with 17 some internal -- we made a series of measurements on some 18 wrought stainless steel specimens that we had heat treated 19 to various levels to looking at forming different 20 precipitation levels in these. This was backup through 21 metallographic analysis.
22 And what we did to in fact make the measurements 23 was to make a viry fringence measurement. I don't know if 24 you know what that means or not. But the wrought stainless 25 steel is inherently slightly anisotropic. And what you do Heritage Reporting Corporation (202) 628-4888 O
414 1 is you put a sheer wave as far as only polarized stirrup.
ggg 2 And as you rotate that thing of polarization around you find l 3 that it will have maximum and minimum values. And you can 4 actually set up -- you can take two maximum -- readings and 5 what we found was that we could use this approach and make 6 these measurements, although going in and trying to measure 7 something like that velocity was extremely difficult.
8 But when we went in and actually made this kind of 9 a measurement we found it very reproducible as we aged in 10 terms of a thermal aging of material and changing the 11 precipitation levels. We go a very retrievable pattern that 12 occurred over several different set -- we had three 13 different sets for three different heat of wrought stainless 14 steel and we could actually measure this change in a 15 consistent manner.
16 DR. HUTCHINSON: The thing that was changing was 17 the --
18 MR. DOCTOR: Precipitation level.
19 DR. HUTCHINSON: But what acoustical property is 20 changing?
21 MR. DOCTOR: The viry fringence velocity was 22 changing.
23 DR. HUTCHINSON: So you're saying the speed of 24 weight propagation --
25 MR. DOCTOR: The speed of weight propagation for a Heritage Reporting Corporation (202) 628-4888 9
I I
415 l 1 particle motion -- you're going down in the part, the
('~ T ' 2 particle motion'like this. When you went with particle
%J 3 motion like this you could not see any change. It took a 4 particle motion in this plane in order to be sensitive to 5 it. I think that has to do potentially with perhaps some 6 longer range ordering that's occurring in terms of the 7 interaction of that particle motion along boundaries.
8 So that's the kind of thing -- I mean, I've done 9 that. I've done that on three different piece. It's 10 reproducible. I can send you the information if you're 11 interested. We published a paper at the last Material .
I 12 Research Society meeting last October. So it can be done.
13 Well, I've only got a couple of minutes. What 14 I've got in the next four or five viewgraphs was really just s 15 trying to show you the kind of objectives that we had and 7
\(^)
l 16 maybe only just put up the first one.
17 The general objective really was to identify and 18 assess non-destructive measurement techniques to quantify 19 the material microstructure and changes in those specific 20 properties related to fatigue life and fracture toughness 21 which controls structure reliability of light water reactor
, 22 components.
l 23 What's in the next five viewgraphs really just 24 goes into more detail. It shows the kind of discussions, l
25 the techniques that we looked at. We had people that came Heritage Reporting Corporation (202),628-4888 g}
L
416 'l 1 to it already' prepared with regard to the various topics and
'2 we had some'very timely discussions in terms of each one=of
([ }
3 .these so we could look at them one related to the other.
~
4 I'll put up maybe the last one which is the 5 conclusion one since it's about one minute till 3:00 and
- 6. just say.what the conclusions were.
7 The consensus was that techniques can be developed .
'i
'8~ to-non-destructive measure material properties for light.
9 water reactors -- in situ.
'10
+
Sufficient samples currently exist.and could be 11 -used to start screening techniqyes for effectiveness ~at the 12 .present' time.
13- Some work such as that on the magnetic' techniques 14 by Dave Jules at Iowa State University. And it's one of the 15 few that has really focused on materials in use in light 16 water reactors and he has shown extremely good correlation 17 'and he shows very promise reaction being able to make those 18 measurements on bolts in site.
19 Other technologies have the potential to be 20 successful but have'not been evaluated on LWR materials or 21 the appropriate degradation mechanisms. They have been 22' applied in other cases in extrapolation to this. I don't 23' believe it's justifiable but there is potential, you know, 12 4 they would need to be looked at.
25 DR. SHEWMON: Thank you. It's a tour of the Beritage Reporting Corporation (202) 628-4888
~417
' l' forest and you're doing some very interesting things. 'Thank
~
,m-- 2 you. .'Is that it?
3 MR. SERPAN: We have nothing more to present.
'4. DR. SHEWMON: Fine. The meeting is adjourned.
5 Thank you.
s 6 (Whereupon, at 3:00 p.m. the meeting was f
7 adjourned.) l
.8 9
10 11 12 13
.14 15 r
O 16 17 18 19 20 21 1
22 l 23 l-l 24 25 Heritage Reporting Corporation (202) 628-4888 0
1
- _ _ _ _ _ _ _ _ _ _ _ - . _ - _ _ 1
I 1 CERTIFICATE 2
3 This is to certify that the attached proceedings before the l
4 United States Nuclear Regulatory Commission in the matter 5 of: ADVISORY COMMITTEE ON REACTOR SAFEGUARDS SUBCOMMITTEE ON MATERIALS AND METALLURGY 6 , Name:
7 8 Docket Number:
9 Place: Bethesda, Maryland 10 Date: March 16, 1989 11 were held as herein appears, and that this is the original 12 transcript thereof for the file of the United States Nuclear 13 Regulatory Commission taken stenographically by me and, 14 bhereafter reduced to typewriting by me or under the direction of the court reporting company, and that the llh 15 16 transcript is a true and accurate record of the foregoing 17 proceedings.
18 /s/ Oh R 19 (Signature typed): JOAN ROSE 20 Official Reporter 21 Heritage Reporting Corporation 22 23 24 25 4 Heritage Reporting (202) 628-4888 Corporation
!l 1
. 1 l
l e i h
W G-l l
e t a
p 8
8
/
0 a E ,
u s
1 B r t s anW R .
HM e u
P o O
TA g r B wVi s i
a _
MR n F o
,d sD
,i v
OG a s no al s u RO M eL n uob FR t r aMopm SP
- c B , r u i
e .k yl TG k s
j o Ccp sC o
N N i
r ,v EI w P da wpla e MIP o mh o ae l
PP y k C hl u Pt t A KVBa l
. b i R .
OD W N J L
E E i
i
- y Gk k b r a gws V D M d s mk ea EA . l o
G e l t
ua al k DR i f s r Ni L G .W y eK r
E a .
M .
A f
oG ,G RD M l
,l d
EC . ni oeh r a n E i i r c a NR .
t ar st EN er t l
M i
puao G mGMc o . .
S CDCP O.e l1ll l l
1 l
l e W G-H e 9 8
/
3-a S
_ B R C
w O. l F
a A d
n a
S s B N s t
r B O t f o L I
T n f n
M A s e E o
A T N
t r i m
r i
s s s R E f o e p
y t
l u
G l
f S E x a s E n e O R n E e A R R P o r e t i
u r n P F n t
a z t u e O o c t c
r r
G i
i r a r a u E t c e N N u t
c F
F r C I I d a e c f
o P L T o r a
p i I
U r
i P
t s e s P O t
I n
h C e l a c n n o D l l l
a P- ai a a c c ct a
E r e
i r
e S- i t
i f
iu D n e
t a
l l
s a
nl ga A G M F u l E
iv SE R
G e e e e e E
D
O 3 l
l e ,
W G-e t s 8
) 8 l
l r s e n /
0 a a e h m 1
B l e c e R c W 34 u n
i nc 2l p a
r 4 s o ( i f
s md e m s, s d y 7 a s 6 0,lo l )
a F M n a 0 d a1 c t
A s 5
5 ooit d R
l E c (
e G V I
i n E C e r )s u e r
p T a P 8 O C h c O 8 sh n s c a O R E J
e C 2 t
e n c r i P B m S a p4 ( a t O e s d h G r u
l a nmt N
I t
c i
r k e c g a
mess a
P f r
t a an2 r i y
0 l y I
P e mcd n 1 at nl i
v sl emab a
i D o r uibt oya E p i o
r ei n cfr f t s i D i m
ve a esdok c a r t A d n s s
f mis t
s z/r i ma R a .
e u s f yc G f y g n
s sircu apreied n E
i ACQPV a i i r p ei D V p e e e e e O
6 l
l e W G-l l
e 9 8
/
3-a S B R C
O A i
s 9 s 8 y 9 l
1 a ,
n N y M a O r
a s
A S N
i s
t n
e I
T u n 0 R y A a 0 O l m R G G I T
a n s s U J
n N 3
6 O A a e s
D i d
I D 8, R C I
k a s M e N 5 7
P L e a A h U 7, P r R s F 5
b- w i G P A e a G i n $
N r l
f O F I o e R -
P f e c P s i
r I
b v a P k
- r e e y
a s D
- e n 4
- E L i D
A
. e R
- G E
_ D
O 5
l e l W
G-l a
l l
e v o
8 8
/
S 0 a E m 1 B I e R C R W O I L
O t
n e n P m o L e i t
A g a N n c i
p s i f
O s n i l
I n m o s a M HT o I i n u TA i t e i s l aQ A I WR N t a
u 1 J 3 e v P yn t
R S E l
a e d eR c e G N NT E v m2 na me u vt .
u p i
l op m O OI / y gl oen o co l
Pi u O
I R TL r n oiVRiaV y q CA r S r
P AR I t k , ,
t oE R T l a c1 3 s1 u ra6 1 e 6 a d t l a ta G ES gC1 00 R3 0 or l u n N T U e - -
p 1
- B ge I
P N D I
R eGGiG p 4- R em I N I CiEEh R PRRWRC E Co r n
P /
Y N RUU eUD Ri v D R t WNN pNG N n E O nB i
P eE T e r D A L
r r
u t u
A U C F
u -
R G G E e .
R E
D O
- 2 O
l l
e W G-H e S 9
8
/
3-a
- E I
s S
_ B C I
i s
y R
C O
- L l A
O a P a n i
r A L
at e B A i i r r B N
O t eC L.
I i
r g C HT Cn R M TA gp i
N I
N .
A WR)D iP ni pl R n t
o i a
r R S E E i e Ri t e
G N NT U P e N de i
r OI /IN iS ct C O I TL T t in Cta I
eo t b N p n R TC s r h r o u a oet i
G ES(
T U AC t
i w ncdo cSe N ND 0 0 r i I
P I
N 4 5 6 6 f t e s C E I
P I
/
Y 3- 3 BB
- sn eP nC M S D
R O WW I I amEA r
T eP T vI ss E A EE y oP D L MM gr op Ce s A U G t ySS r
s AA l
omRNA s R E u h nI G R d - - c e - - -
E i n T D e .
O.
i ! 1 i
o 7 l
e W G-l e
H e s a .
8 8
/
a b C .
0 1
A s -
B t a R i e R W
a
- J S
M d
y F
l P
d t
u A t o s r t g R e n n G p i i
o d g O r e a R p t r
P l a r o a o f M C i r p e -
R e p A N t a
r o i p
R R m c n l G E e i a
e e
O H p t t s
T i a -
oR P p .
O r d s .
o y s E H f t r
e D T l n N G I d e
e p i a r N W r u o t o -
i N r s f P c p t e O o r
l s p I I T a a _
P i
p i c p r
C s)Cte d d D A R
l a
iAa e e _
E E r eR m E i
l p
i l
p D T t F r aI u N p p -
A N N u u R
l M(P O O S S -
A G L G E e . R .
N .
E M A P D ._
o
,lll l _
8 l
l e W G-l l
e )
8 8
/
0 a D 1 B
T d R N e W s
O t
O s n e o C
(
t s
s i r
S d l an M e pi b
A R d w mo or -
G n cd O a d n M R d g
n ou A P t e i t
s ht or R C a e e gn .
G R c
i t at r a r mni ib O N b d o nnts o b o e r i
R a f R E f t l sa ot ta d i -
P H e
pt eir r l P n u
T i
pet n c ed uE o G O pa u a c rt di pmocd l
N H e r s I
P T I
de de et l
ps eee vr rvt e I
P W i wyzpte c u ul i N Al a nt e c s D O T neipRR n e - -
E I
T GASPJ J T D C A e e e e e e .
A R C R
R E T T
G N D E I D
, l '
m O 9 l
e W G-l e E 8 M e e 8 H S p /
0 a A i
p m o
1 B o d t a
R W
$ e J t d k c
mW a&
e a r i
l p
r c I F B p R ym Y u d P b o R s n Edf r T a e S t a s yt d U a t b ae M d p n d nn D a g A N y op ul i e oi a d t n t
I r r l
b i R pe o t G
H T eGe p
d pp uis s
e I
okl r se l
spe t O W pac seu l r r o
O R S l a
T n a . d ted Hf P N O
ik r c c c e es cEeCW l
G I T
t e amiwsoT & v a r o r AsrepU B N C m ef Crs eSlng N ot
^
I A l P R d pd nie meinyn etaib ev I
E aP r ot sl d P T I ur sf t ei ed g N tsX c od sswal D I s
e nr e sa eccmi a E t i opvl ondiR etr i
r D et e p c pmie a ge r Ota A i PSP r ei etSAFWM R
G . + . . . = e _
E D
.O 1 lt lll '
0 0
e H d 1
W G-e H
n a 8 8
a d e
/
0 _
B p 1
R o _
O l l
3 l
e v
d e
p t
n e
W P e o .
D d l
e m
) v p F d e o O a d l
e T o s v R
l t
i s e a e A
i d
M T mS l
y r
t a
l b _
A S i
l
( S a n d e d l a
R E s 4 0
a n t y i a
v G H i
s3 M u r e a T yr F d p O T l
ae P o o t a
R A nt E h r a 0 a e t e
p c P T e m n o
l a e R si a m r G
i i
r u A- .
pd t a s e t N
i E al l m.y s t a c l
a o a l
I H i r _
P cm m l
t T- s a f I
F n s
- e- n e e P O- o no J a i p
i p
i B p p D
I E t R B T cd d d E A e si e P E
L. t e t e
D T t
- f i / C i i A S ee N v r
E G
R N
i L
m i L
m R
G .
. e . e E
D 0
O 1 l
e 1 W
l G-H e 8 8
a sn
/
0 B eo 1
R r
O W i
ut a d
g r e em ci i
n na t os t
yt g
r t s
es i t
ow r p e s
S u s o n ids e Cc i s
E l
l i
u z R n i N a si ea t
qr i N a s M I t
eng c u d c h l y
A L P Mhgsna i t
as s n a m e a n
R I C dn ueot r e e
a G S ovia p d er d ain a tt d o u e O
I n d h i l D nelr t Ct c i
a O R L i o u c ae aw ni E ar t
e P A t t ac gi nF m i
Mf d C z ar r o0 S c t G
i I
N r f r uei pt 5 n e l a5 Ai ts e N
l H t d a xt na I
C c a a nte E nt e a da ol ps e m P E aemr r emg osk c mel e I
T h r u u n P
l l i
t ct i
e Cs nh et rctt s i i
e ah s et i D l a et a r n s ei s i mh l ci n E iTOFI r T sL CE y sF D e e l t
a p - a - - -
A - -
i n -
A M P R
G . . .
E
- D
.O
1 O M C-8 l
l e 8
/
0 l
e 1 R
l a S W B T R
,w O F
F E
N O
I M T n A A l o
aw l
Z i
R I hcdod r
e i s
v G R s l
n en c l ei D
E r l OO R T M a LaHi.M R
t t
e s u
C Bab P .P .N G
A W
.M
. P P
. m u
R C l
o N
I A C P H I
P C L
D A E I D R A E R T G A E M D
lj
- g
' l
2-M C-8 l
l e 8
/
0 l
e 1 R
l W a
B O s e
p i
p s r e o v r
f s u a e c J N t as v n e
O d s t r u i o c I n e n c t i
n T ot o n s a i
e i n
A t
E t ar i t i a a s V z u s r i Z
I I
T i t r c o p
t r
s t s e
h r R C t e a r m s c
E s t h E cf a o o r e t
J c e n- w T B a a l
t r
o C V l
O h c a s r g
c s c A l -
i m
l e y p k R
l i al u e s r a c i
r n a A e of h e h r
- H t at C T C C C mde et L d ce . . . e A
I i
vj ob R r u P s E
- T A
M e
3-O M C -
8 8
l e /
0 1
l e n R l h e W l
a tcm B / / oc i N- ep aw / >
Vyt S pc r
a p a
h CI m J'S l
(i T
E EE n e
CT m NT i c
/ / S e
p n
I o
SM i s
EI n RH e T
HT C > t c
a T p WO N m o
O ( C RD GN KS CT A
/ O/Od
\
k e
c a
r c
e AS P r
RE CT O
l
! l11 1 I 1
li!
4-u c
m 8
l l
e n o
/
8 0
1 l
e i s n l na w a k cteA B
ax r E Q C V-a V ,
A t n
e D
m e
9 c
R l a
O p s _
C i
D E V- ,
R P T
S E
T L P A d C
I a
o P L Y
T e.
5-O M -
M l
l e W l
l e M a .
,. mF B 0 0
0 0
0 0
0 0
0 2
1 1 8 6 aw 5 7
1 i ~ - - - - ~
f.
0 0 ' 6 0
5 1
i 0
- 0 x 0
- o x o o ' 5 5
2 i x o 0 o
N m m
1 x x g n
,I RO n x
9 o
o '
4 n
.o io01 i OIS F
s n
x g g
0i s
n O N t e
X t
e x
EE E x
- 5. i x
o 3E VT kc 7 x o 0k c
a RX a r
C x r C
UE
- o g ' 2 CK 5i m d 0
EC J J
= -
x o
CA 5. , x o x I 1
NR 0
2 AC 5 6
TE 0
- - - - - - - - F ><
SL I
I 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
6 0
0 0
5 0
0 0
4 0
0 0
3 0
0 0
2 0
0 0
1 0
ST 0 9 8 7 1
EC
" 5yc Eh5 RU-JD
. O l
6-M O C-8
- l l
e 8
/
0 l
e d 1
R l
e W a e l e
B i p
d p o aw i n m d s e n s e i o
u p i t
e i
p d n p n o i e o i t
p h c a t t
- f ,
n Y o f o t s e er R t o g s i r u l
n e ot d O e i n
t s
y er a e n ep ep t
T m e c i
u a A a s
t t
a i
n apm gs cl O R m l
f at h s h k e n O o l
a c e c ct t
i t
et oras es e t
B r f t i c mep AT L S s s n te n
a eip r
n ct t S e e h u n mru c t F E i e c a
i e
OT ct e c m r s o
pa o f SN Sf r N nh t EE I
li
7, O
l l
e l
l e ,
a B
^#**
/ _
/ / n o
/ / i
'~g ///
t aRCC aC t - -
nLLL xL l
i-
'/ i e
r A
/ O
~ /
D N
x -- )
S \ -
T ONE - (
x [ E )
\ S
)
- T B
(
E
\ A )
(
C S .
M I f n oV N -
x i C )
C I B o
ns) onn-eo d
n -
E // N i t t ih b e
P // o n a( sc nt
(
S N i geeoT e gt n g
/l x sd( - - d GS ee V e .
- D t n-E // \ e s
l ecyoe apil NP I
I
/
\
\
T p
y i
gprs g nnohann ei NP I
/\ n e
SCCTS m ABCDE HM i c
e CO S p
AR MF -
O l l 1 l(
8-M C-8 l
l e 8
/
0 l
e 1 R
l W a s B a l t
l a
t u r a e aw d s e
r s t are d el u n t s v r a a
,t a ae uer s
e t t at c r ap e r n d c n r u o npa i
aa m d
i e
tr et s t i e
c i
ss oet am r
sh s,u s e v
E o pt rsh i
i s
r R r s- c s ysu p m e opstoa ege r c
U r r s en nvenc f
T o cor er nd /
t n l a psVJ t
eeo CS o c eeyd t e yst i a AK s i
i t l p miipd l l s s r n pr pni a s R
F O i
r c a e nn a a t amape r h eeh a oh x O
t s -
daCTTCJ i
e d cCeJ E
P-B d f f
o d s o P D I
eis s - -
s - - -
R l iy al l e h p
NO I
t e n a b a
a r
C D a T G SE T
L R e e e UA -
O 9-M C-8 l
l e s 8
/
0 l
e h e
4 2
4 1
R l
c W a n 6 2 37 6 d 6
1 6 B i n
1d nd 1 n a d 1 p,w i
da n d n d r n ,a 7n 6 ,66 a n4 6 t
e a08 3a 1 1
2 a e 612 ,
8 6
1 6 m 4 ,
i a 4 D e M p A i P
DR i c
EG i t
n ZO I
e p
ee pt RR e i is Pu OEP T
i P
p P c i cA CG i
t t t ih c
i n n g AIN e e u i
i t s t ss s u ssos t
0r0 RP 67r7
.e se e nel surs AIP r r F r. pl sAleAWle n H e GGtG ini st nt CE D pB , ,h i 6 3 6 g6 6 c at
,B Pia e ehaih ni t s 0 p n gt goa l
t P 031 u10isL0it iu suys SD c135o51 n466 i
Pa o4oa t
r0 rl r0 4
L iAAArAAe01l t
t 33ei csW3We3 AA rMMM W i
r n mms ee o t
nim n n en v
I R eTTTiTTu FSSS sSS A yy pp c e F8 si ydpid Oy pe RG tAAAdAAtTTI n t sCu dTleeT l EE h g
l e
h g A l
e WW TD A
u o W u o t s W AA MIN r
A r a A TT W S W C S GG
. O l
l 1lll1 0
1 M
- O C-l l
e 8 8
/
0 l
l e 1 R
a h t W B s w e o p,w v r
g r
u k c c R- r a
J c
r h
t s o t
t si _
w c nn o
r f e s oo f n i g a m t
t s e i b
ut o l
k e y o c t p r c P a
r
) o l c- T
(
r dnt sc aE E
t o us c O e
g r
S f
i s
n o
r e ervd t
S l a o a nl e iic uf o K t e s nR - es f
o n p n S n e i p ieo m t t J U A i o p f o
a p
T t c l o y i
c LD e i
d v d i t
e u r AE r D
e t a NM P S P _
OR I
) ) ) )
2 3 4 TO I
F 1
( ( ( (
~
ll lll l -.
_ O 2 1
_ M C-l l
e 8 8
/
0 t
e s 1
R t
a e W
, v B s e
r u
v c w r u R-c J r R- g f o
S J i n M N S n t a J E E o l o d I
s p n M D t ct s a a I
U es r D
RC T S ff e e t t
x e
J J E o s l
f P E e ta r e L z e f o s v s r O THS L sm i
E s e u L T e d- d o n c WL T A
l bl e h l
uR -
OA i f B ss w t e eJ s g RM on pi m u n GS dl y d e d t ea i
l p
KM e a r l l
ac i o a o pp CO evp e l
e v mr a AR e s e ot x RF R e D Ce CS ) ) )
EE 1
(
2
(
3
(
. O
1 M
C-l l
e 8 8
/
0 H
e 1 R
a W B
O
_ =-
EE TP g
AIP L w
, e UD
/, 7 a
- 5
- = c ME 2
K .
/j+
I SC OA -
- SE EC TAl F
TR I
- EU SS
4 O
1 M
C-
_ l l
e 8 8
/
0
_ l l
e e 1
R a i p t e W
_ B p i an f a t otaf i
o
_ p,w n nt d y n
_ i ea t b o k gml su a i c n pmtel u t a
a bl ro i
u r
c e
i rc evfo em h r t o l
a v
c s en e a ed omf ni o
f r d,uet arof o t u i t d s Cqm ea o S a Ri nt vm i h
E s r V
f o S No ch e cuts s e i t oRf eims I
T n U t ef O C o T A d t
l t s yf ol ol d
E i s T e a nJ a e J n g S i t t t
n e a oao n n nw B me m te n ei n i
O xd i
b mptoaa nt t a t i e n ui r l l
c e c s e e e uf mlpc k se pvci i i h b r t
t r x e a el pr e l
a t
et o oED cV eA c p
S ad e r
ut e l
T l S i mce a c
. e e e E sj b
i p
T ou o l
T s T T'
f' E
S
. O l
O m3 <
l0( l6, l9 4 li t 19 1 II ' I ' ' '
..,3 '
7 ae 2 - \
e < t 1
{
y
. k F ;IIr. ;
- f5hg !
II,\{%{?
O ;
t u gM; 3 m 3 E mt g:r t o
3;$ " ;
- a %
0 E 1
1 O.
6 O 1 M
C-l l
e 8 8
/
0 l
l e i n
o 1
R a t c W B e r
g i m ow 0 d o r
d 3 n f m l e o i r d o at e e r e0 t f v2 w l u
e et o s l
g r a l y e n ed pe l
t r
y a i pn t n l n a b a s e i c a g si r f b n a no i
t l
e o n r oi m
msi s g pr wi i nf o ec t
a n x s 0S r e gr e oai s s e e ot T yd i
a u p c
n i
t o ah C l t
n a l
f on l
c p i
t a
s t
n g E i e n et m s s i i F u n n c o n i
r r i
F qe r i oi l r e o r ou E e r e t l a a s f
es i
nd f f n e et r u og Y kc mu i mmr e t ul c i
s n ut i
P a an g c n s r rc l
x oe ai ci O Cc i
E nd r f F o it nw RS TE OP e . e e SIP I
NN AI
. 0
1 M
C-l l
e 8 8
/
0 e s n t 1 H d i a R l
e a s a t r
r e W B w_
s s e n a i i
t c t O
l a i a h i b
g m e t a
r s a t
r g s t
s e n e n n e y )
S i n d C l
)
i i
k G n l o 8 C c N a n t 8 a N I h o e (
2 8 8
r c
D t i s
l b 2 O N I
h g f u i t
F 0
( d e
I F u p F T G o g e 5 5 0 i
b t
A N t n o
c s t 5 i h
Z I s
s l
a u a 5 x I
T e s h t a e R S E
l s s s t g a s E R l y e e p n d e p
T E t
n n e /
h i
p r TJ i p
C T N
a c g l t
sRd l A I i
f u
o e
e et an d e e
R i l n t t s
i s n aT t s
A i g
w n n a c R n)
H s l o o eh t t J a l t oC s y br b C d l
l l
a r
a8 8
L e a c -
c (2 w u yg -
A I x s u nn yF n
u n ai g a0 R l F U Ma M 55 E
T A e e e e M
1ll I1i i 8
O 1 M
w C-8
,mu 8
/
h 0 H
e 1 R
a W F B .
aw lf ;'
0 n 0 0
2 1
e g
0 4 0 &
0
- - 3 AB f .
au t ! ,
0 0
5 i
ee HH
.0
/
/ ,-
cC7 r r r /
oGG s66 0o1 ,
y 1 i5 0
0 i AAA SSS ,Noe/, q O 0 0
4 OeA s 2 F ,N _ *
- - C m
O e
r u e.
r A t a0 r
e0 p3 i -
t u
a r
e S -
T m
e N
p m
e D -
, de O O T
- N O 0 0
2 I
' 1 T ,
eN
\
SE a L L 0 0
1 i
g N
EB EI -
O O
M TT 0
0 0 0 0 5 0 SP 2 1 1 33 j F E OC S I cU JS .
. O .
. I
9 1
M
- 0 C-l l
e 8 8
/
0 e 1 R
W M
Z E3 aw 0 0
0 0
0 0
0 0
0 0
0 0 0 0 0 0 0 0 0 0 0 5 5 0 0 0 4 3 2 2 1 O5 4 0
l e
e 4 t
0 0 1 i ) s 5
5 0 1 7 A
) )
5 F n '3 H
i M r 0 0 T G 5 4 S
T 8i(
A 6 1
5 5 6
(
6 3
W m 6 A 2
F M C 0(
8
'0 n OE N m - T 2 S P A 8
2 m m
5 .
_I 9
R i 2. tn t t 9 D a n m GIM 6 0 e e6i e o m d 1
~
m p r e e i f e c P t = c s
KC l a
d e e T D u )
F
)
n i 2. l 0 p a
p a i CP E i D
s m j 0 0 5 3 0 5 D i
s AS 4 , (
C 0 6 i 1 0
R 8
(
CTC 4 m 1 m t 4 a .
i .
1 0
E A 2, d 1 5
L e 5 B P t - i 0 s
e n 0 T B AM r TO - ' -
- - - - - ~
O 0
0 0 n 0 0 0 0 0 0 0 SC 0 0
0 0
C 0
0 0
0 0
6 0
0 5
0 0
4 0
0 2
0 0
1 NN 8 0 9 7 1
d UI =
O.
O 2-M C-l l
e 8 8
/
0 f
l e 1 R
a W B
aw d5J 1 t P 0 0 0 0 0 g 0 0 0 0 0 g 5 5 4 3 2 ,
5 1
O H 2 1 s e
T m m
h c
n WT T
, i O OS N ,
E T M N RE E C
9 E
M GT R L
P n o
E C
R KE S I
D i
t a
L P
S CR i
t I E i n 6 D N
- a I r
D L C A. C.
C R -
O D L ' R O
ER 3 L L F B
AE TP I
- - - - i O
SP .
0 0
0 0
0 0
0 0
0 0
0 0
NN 0 0
5 1
0 5
2 1
0 0
0 0
5 7
0 0
5 0
5 2
UI 1
d5g b t l R
. O
O 1 l
l e K G-H e 8 8
/
0 a 1 B M R W
O A ,
E DG R ,
R i
r n EO i e oi UTR r sk r l s
TCP C U eOw u
.o n
o .
G J. k isA.
A RD G . l i F NN A. R ,Wi i DU v S.
I Dd E CIOP t l
, ef M.
s o ui P P t
n I y o c eG bh mO O P S T D b S sd on u s ,
EN E
.R a l ou LE D Cb M.
P J.s e emu AMR A r ,Mn l
,ar l
CI e el o SRG mh ae B t
t t
aC L EE r i K V R.
B L PD .
UE XE S. H. C GP F H T
N I
3
2 l
l e K G-H e s n
8 8
/
o a 0 a g i
t a
1 R
B n t
i d d e W s
O n i
p h p i
p o t w a c l d g o r
l o
e n g c- e k i d k n s c a c o a a
r o a i b c l r t c a y d c e t
a l
l n d s- d a a ) n t e i
t s C a e r n k d n u e c 8 a g t r 8 c e a r 2
( l o i n a r
f c s f m , F m u e u s 0 u s p c
r l
a 5 m e i p
i c
i r 5 i x s e t a y a n t a a l a n o m s m n i a
a t n n a t t
a e e o d n d
v i a
i a
i t m t a o S e r t a i r
e i
t l
m E u n e p i
n t
i d n
V t c s x e
i e i m a I
a e l T f r r p
t c
n e h s
C i
- t n e u m a i l
E i
a r d r e u b J t e n t l
a a b s o e v t
s
- B O U C D E E O
. e e e e .
O. 3 K
m G-i a 8 8
r / _
e 0 E i C
t r
g 1
R W
AW n i
n e
e r
s c _
t S
n e d m n i
r a e
p C x S E N E e t s
N r u n OI L x t
c e
s i
a g
T a a A
i r r b U t a F a d O M e p
t a e r
t i D a N s e
P e r
p _
O f m _
I T o t u o -
T f
o s c C y e a r A w l
p F r
t s a T i e m e l
u m N v e a x p s e
m u
E R E i
P R S S
E R
P
- O lll
ll ,
i !l K D 4 O
C C ND K H A E N G-e R B BE 8 8
H C & T /
0 a X EEN 1
B E G RRA R W
L N UUI L O P M
C BPPC K D C N A D E R NB Y C EBN ET R E x GR A O A C N SSU &I A P L
M I
F e
DSS M R U
NEE M ERRO U S BPPC S
X I K R C A
T R _
A C D N
M L L
E _
T A E B GR S W- NU &
E I SS -
H a
T DSS -
G NEE U ERR O BPP _
R O -
H T
l 1: llll l
l lll l O
e 2 l
l 4
t t
e _
a - _
B -
7 3
O 8 _
2 .
R E -
T 6 s e
E 1
- h c
n M
I r,
e A
t e _
I -
2 m a
D _
- 1 i _
D e
O E -
i P
p .
P I - 0 1
P -
Y - _
B 6 S _
T _
N _
E M
I 2 o 8 6 4 2 0 8 6 4 2 0 8 6 4 0
4 -
R 3 s 2 2 2 2 2 1 1 1 1 1 E
P X 3 U $
E O
!Il !
W G-9 e l
l e 8
/
3-t e S R
t C a A B
n2 aw w4 Y
n3 7 R
T E
A 0
E K
C A
R m8 2 C
6
- 1 m
m2 ..
- 1 m O
- x0 1 Y
YR BTE 46 S
T M NO _ - - - - - - - - -
- 4 EE 2 6 4 2 0 8 6 4 2 0 -
3 2 2 2 2 1 1 1 1 1 MG
- RK I
3E {U f$
EC PA XR EC 0
- W G-
- O 9 e
H 8
/
3-e S R
n C
a 2 A
4 B
O 7
w la 3
- = t M
e
- e s
- e==
- a E B P - -
Y - - 8 T m. 2 i w le te W L s s S S A
R == n E
T A
M s s ae oa **
- n t
C C 6
1 s
e h
2D l c
Z3 n ZE ME I
S L I E r, e
t e
A I
2 1
iD m
o R lo P
e OTA E
E0 1 M
Y
[ h 6 B
S _ y , - - - -
_4 T 2 3
2 33o 2 8 6 4 1 1 1 No E 3l U M
I .
R E .
P X
E O
l
3 K
l e G-l e i a 8 8
r /
H e 0 a t i
1 R
B C r
W O i g
n n
e e
r s c t
n S e d m n i
r a e C p S x N E
E e t s
N r u n I
L t c
e i a
x s g T i r
a r b a A U t a F a d O M e p
t a e r
t i D a N s e
P e r
p O T f
o u m I t o T
f o s c C y e a r A w l
p F r
t s a m
T i e m e l
u m
N v a p s e x e u E
i R E P R S S
E R
P
- l
! , l 1 i
.s O R er e
g
_ 9 l
l e W g, g
i .. ..!:
K G-e
_ 8 8
n -
/
0 a
_ } 1
= -
B R W
O
_ i..*'l!:
_ lem
_ er t
S A
_ t
_ nn ;... .. !:
_ oe br m "s ao T '3 .s CM m 1 o 2 j
N i..*'l!:
R
/ _
E / ,
N 2 (.. .
M l e
e
] . W '-
- -. f
m oE j
i R E
t
=-
S" r_:... . .!:
=
s s -
G l ia e
n
- "6
's N
l t
S le ly ly l e
- .i*':!l
>d b b i
gm C A j/ m / d e/
/o e e
/ Si s r m d Rs A n
i e r s ps d
a - ( - _
r o R SA n
O R L
.iii.!l A - .
_ x!.
....!l
=
m
= ; R
] . U. ::
o(l G. -
Y _
r_ . l.....!!
N I
D lem er A
l..ii. !:
A t
S t nn oe O bm r
a o CM l.. e'i!
L f L / l...ii! :
A C
j .,...!:
I P i. ,. ..! :
Y T
O
O l k
~
,a.. . .
n .j [~;.
I -
>j
, -p
= , .
i$
'{ .
1 S 24 , *e ,
! , Jt t.
.m i .. . .t 1
- lP x4 ,
f j,
,- 4 L
-. .. :m L .
O
O l
I a
l n.
g., , . ,-
w-6
,.n , .,
- . ..r, 1. .% , .n 9
,.xs ia ;
. 9 m . Ae < s
' ?- -
N Ef 'fY ?.,
~
- . , , l '~ ^ ,
gw%w ,.,. >:. . , .
o .
n s .1. l,.
g M> M
,j* .
ag '( - e' *' a
- n *- a" a i. .. N " **
1 s 7 ' * '
l e
=,./
- g , *.
, A, s
J.I,]
r / . -
i
{ # ', , ".f,--* , '
l
'h - 5:j ,
>: 9 s .,
)
I
.O
l l
e f
l e
a a e
B s A r
- t n g
- 4 e h n m t i n
D o w e E M o r p T d G O A e k k R i l
c c E p a a p r r N C E A C G
D s E s t T t n n e e A i o m r
u R P e c
t a
E d a
r e a e r N L o
l a
l p p u s
i s m AE t t
n i e
s e
TG /
s a
s s
n t e
o D
g T P
r AD d t
n o P n d i
u d DE i i a e t a c n i u
o l
e F LR D L m t o
i r p l
F e t d
AU E R d e
c a
R c e O n l a
CS U i l
l p
k c
l E
k c
a e
n r
P A I
S p a a e p s c p A r r t Y E E A i
D C d C
i P n TM I
M
O 3 l
l e K G-l l
e i a
r 8
8
/
0 a t e 1 B
i r R C W g
Ce n i
n e
e r
t s c n S e d m n i
r a e
p C S
x N E
E e t s
- N r n OI L t u
c e s
i a
g -
T x a a i
r r b A U t a F a d O M e p
t a e r .
t i D a N s e
P e p O
I T
f o t r
u m o
T f
o s c C y e a r A
w l
p r t s a _
F _
T i e m e l
u m N v e a x p s e
m u
E R E i
P R S S
E R
P
- O (l =
8 l
e 1 K
l G-l l
e s e 8 8
a s y
/
0 B l 1
a R n W 7#
$ a e s M s t n F a e P b E a m t
a i
r e r
d e u p t e x u i
v a e f s t a e d E S n e d r
u n
a S E V h t a y t u t A I T
e r d a t r
e f n e
B C p
d p f
o s A E m o
e c
o r n o
e r
T J B c u p i p
A O a d e
l a
i d
t t
e D n r i r d a u
i a d e a E
l t n t a r a R i n a m f o v e
a U m w a d t
a o T d r l e m t
e C n h i
a r o s a
A t
a o t
e f h b b R s e d
e h
s t a
F i l
d d i
l a b u u b d E t a l c
l c t a e P s n n s s I E i i E U P . e e . .
l
O 9 l
l e s 1
K G-e e 8 l g 8 l a /
a p 0 1
B 0 R 0 W O 6 3
y l
t e
a m d i
x d e e d o
r d l u
S p u c U p l c n E T a n i
s A ; i S T s
e m m A S m a r
a r
B T u g g O A N l o o r
o r
T E v p p R e A R v i s r e
D U i f h t
h t
o E C f o
- m o m R t e r o U s f r i
s s f T n t n s C
t
- o e n c e A e m m
R i
r s e i r
F a p e b x p E t a e x P a 1 e
I D 6 6 P . . .
'O
, l'
1\1j 3 O l
l e K G-e i a
r 8
8
/
0 t
e 1 M i C
r R W
O i g
n n
e e
r s c t
n S e d m n i
r a
- e p C _
x S E N E e t s
N r u n I
L t c e i a O x s g T i r
a r b a A -
U t a F a d O M e p
t a e r
t i D a N s e
P e.
r p
O T f
o u m I t o T
f o s c C y e a r A w l
p F r t s a T i e m e l
u m N v e a x
p s e
m u
E R E i
P R S S
E R '
P
- O l l ll l
O 1 l
e 2 K
G-e l
8 f
l C 8
/
a S C
0 1
B n a S R 3$ h t d n
h t
o W
C a b W C r o
T W f ;
r s o T d d f r e a r o p o e f o l g
r d l
e e r
S a e v u l p e T
l s o d i
a L n l
e n f U v o i
g d r e C i n
S a d S r
e u O E m n N t i
r o
R o f c b d
a i
r o r e s g e C l o t i t n w o e i
r S
i m c m n l N u g i e s c l
e e n m n w r ve F i x
i n o c s id O a e e
h s e gi n n f
m r l Y o c a pC oo mS ic i
R t s r r e e i e%
A n o t a
i t
r sd n
t i
r 5 M i t r c de a c9 a a iC p f M 9 1
e -
iWn -
U i n S UT S e . .
O
1l ' '
1l l 2 _
2 W
G-O 9
,u 8
/
,i l
3-l e S R .
l C a A B .
aw lai l
a te M ~.
w s _
a _
a z
e ._
S l
ta 2
T e s
o v.
N B .
E a
+ O a
u M
I
/
R /
v E
P X 1 1
. . u 7,
- 1 -
1 FE 1 C O
G 1;3;t i SE TL L -
O._
O 3 2
l l
e K G-l l
e 8 8
/
a 0.
0 1
B @O O 4
R W
3$ O I
5 3
4 3
= o D O C A
t a J e g o d 0E a o
t I k 3 2 2 I a ct o w d aa R l la l r
cd k e r
E f r
o n
lo a wr
- c c te a
m T
I is o
s hgce ua of I 5 2
a P
r e
e r r R p e
r m
h u TS Z n
o C
c e OO t e
s 0 s I
2 la n P G
I r s o s O N I o
f tw le le n
o s
n s
e 5 e N
c i m
n D i
e E f di n
o d
f e
E O */.
c 0
h p
m R
! a O 5 1 S 9 9 C
S !
5 0
D O O E
I g
- - O @ @1 -
F I
2 1
0.
i 8
0 6
0 4
0 2
0 O
O L iG g I2E M 5 P
M I
S O
- , 1 l (rl l iIl lll
3 l
l e K G-H e i a
r 8
8
/
0
_ e a .
i t 1 B C r R W
O i g
n n
_ e e
r t
s c n S e d
- m n i
r a e C p S x N E
E e t s
N I
r u n L t e i a
x c s g T i r
a r
b a A U t a F a d
- O M e p
t a e r
t i D a N s e
P e p O
I T
f o t r
u m o
T f
o s c C y e a r
A w
l p r t
s a F
T i e m e l
u m N v e a x p s e
m u
E i
R E P R S S
E R
P
O e
9 2
l K
l G-H e d )
a 8
8 e /
a s e
t a
t a 0 1
B r u
u d R 4
t l
a e W a v p e
i r p e
p s d m n n
~ e i o a t
t i a
d d t e n a t
a o d v c l a
e g i r
l e i n t e
d t d a e a a m r
u d l o (
c e d O o r
t c
d n t e d p u a a e d s r n s n e e i l
o n f i
a c i
r e e r t g d e s e r t
a t
s m e e e o s a t t
m t e e e f g b b v o k t a
C Y
i t x c a S a i r a d R t n t a
r c e N A e s m s i v
s f
o M e e u n s r
p g o e i t
M i
r r e a a t
x m U
i R L V E L S . . . . .
.O -
- 1 l
l e 1 A
J-8 H
e 1 E
8
/
0 a S 1
R B
S A W gw o E H S P Y -
L A M N A A R M G s o
A S O l R C I R t ul oa i
d s pi G N P a ru o d a
O ANG mB pG h
r yh k HI N AWsaD P C I .
E 'P J F p .
G PaN I
N M P .
I E D V P
I R E P U D .
D T A E C R Dl l
A G AE R E F D RS.
GA E EH H DP T
2 l
l e A J-8 l
l e 8
/
0 a 1 R
B r . W o s e
aw f l
s t nq u eai d p l
n o h mer c e
gwt E i n oM pF M V r I
T er e aP A C n e Eg R E i
gc l
G J n u n i B e nsu O O eo s f
R S l
p yk P E mict S i sgr r a G Y ec L h N A i t
t ng I
N win P
I A Ch e i i
n P Rt t a
N gn D no ei s c E d ve s s i
D1 1
os e p AE r si P ap RS GA EH DP llll ' i
3 l
l e A J-8 l
l e 8
/
0 a 1 R
B W C
d a
- l o
t i
M i m
A L R =
G y
r M s bP t e
m o
e F
P E
e h
g u
s n
g o G e T
N I i p M p F P
I d E
L P n a
D s e w e a r s E m l f u es D1 1 i l g n i r at FS AE R e e v
RS e i
g GA r u a EH l i
a r
o DP F F
. O.
l l
4 l
l e A J-8 l
l e 8
/
0 a 1 R
B W aw g
n l is t a
U ne s n a is ly io imtDa r
, lp a ,iI t
a e d p 1
n i x
_ A - la E
- _ E - V
_ F - -
M _
A _ -
R H _ - s isy s G C 4 d e
d o
A *c- n o s
)
e n e v
e v la n ) d n h t n o O O P a a
t r a yis it
, mla i
m c
f 1 C r
u I l
i t
c A E id n P io PI f
I e it ma M miimss e
R R P
sD i
i
=
it Es A n p S
e R-J P e
r t a
(
o s c
e E t
ly P P s
=
(
l i
E t
m s R A
a n
G A ' ,'
N I
P I
s isy n o
P it la a An d E ila D F V E l Dl AE RS GA EH DP
O 5
0 ) )
A l
) ) J-l 5 3 6 3 8 H
e 7 5
0 1 8 0 7 5
8
/
0 4- 5- 4- 4-a R R R R 1
R B C C C C W O /
G E
/
G E
/
G E
/
G E
R R R R _
U U U U N
(
N
(
N
(
N
(
M s A S e
v e
c R C r
u n a d s
G I P
c t s
l e
O O R- i s w
~
eR T J n
e r i n
s P Y o h s e E i t k s G K s w c y n o a l N s t e r gn r c a n
I x o r a P n e e ki t o n I
k c a f P m c a r i z s i
b i
c a cr e e o D e r c et v r
r E p e cc aa u d l S c n Dl y g
r f r r a R- u AE r a uh o o L Sc J R RS t a
GA r o
EH b a ) ) )
)
DP L 1
(
2
(
3
(
4
(
O l1ll!
6 9 -
e A -
l
) ) ) ) ) ) ) J-l 3 2 4 6 8 32 8 H
e 5 8
77 0 5 5 8 7
8 78 5 0 8
/
0 4- 4 - 4 4- 4- 4- 4 a RRRR R RR 1
R B CCCC C CC W g
,w ////
GGGG
/
G
//
GG EEEE E EE RRRR R RR UUUU U UU NNNN
( ( ( (
N
(
NN
( (
s M k e c s ru -
A a r
c d d o e R l h c n ee l l e d G t e n s a l
a eoot r t -
O w- mptisa s e s s se yr R h g
ssd i i i s s l u 9 P u s yy sl e al l e as ns o l l vn n aes G r a al i a
i a
h nnat t n r d ipa N
I l
t aat s n s s n s b d o P a) sdCen o r e l
I i
t e
i n s v y o eR midid N.
l rik s wk e s l
e d
- ng i n P s ec wn bi D
k c
r feapB l r paGrax c umdn a maInmB xr LECTCf RCb I uoo e _
E C r l Dl l uB c
h r B)
AE t i
i C( L ab ( ( ( (
) ) )
cd
)
(
e )
f
( (
g
) _
RS W GA s e
EH p )
DP i P (
1 9
ll _
_ O 7 l
l e )
2
)
7 A
J-8 l
l e 7 8
8 6
8
/
0 4- 4-a R R 1
R B C C W
/ /
aw G E
G E
R R U U N
(
N
(
M A s k s
R k c G ct a a
r n r c
O c e m x mR )
d e
c s l e
P e a s e
p u f r s m G i n u s a s o c
N t n
o l l I
a w a P C i t a i t
I ( nf l
n P k s e r e e r
c e c e D a f
i f
E r mrv )a ei r m
I C u s e u DI h c
r - t c
r AE t i
i CI (
ni r c
i C
RS W GA s e
EH p ) )
DP i P (
2 (
3 O
7 l
, !Il1 9 O i
l e A J-8 l
l c . 8
/
0 a 1 R
B W -
O n o
i s n n u n
e 4
t 5 x 2 E =
h k c n
c i a
r 1
C :
e g
r t
N e
o m
- e r ,
io a *d L S
~ .
5 2 O
' O h -
t i
W n
hJ =
J%~/(C e
i m "
5 c : --
S e %
p l~ ;. i I
I S ' - 8 S
I
) ,
Y T
(
L C A
N A
E F
O
O 0 i
l c 1 A
J-8
- e 8
/
a 0 1
B d l
e r R
W O
iu _
Fo t _
rn ao FC -
h t
w .
l or t
a G .
o k T c ar C
p _
O c o
r C
s r
u .
t o -
n o
C iN N d l
i e _
F r
S I r N a
e S
Y L
A N
A E
F O
ill l _
1 l
l e A 1
J-l l
e .
/
8 8
a 0 1
B R e mN r4 W
$ 0 2
5 1
0 1 5 ,
5 -
i 7 7j 1
^
0 . 6 5 i 1
a s
5 . 5 e 2 I h
m A s m 1
c n
e ,
a I v
r n
o ,
u i 0 A . 4 n s 0I o C
1 n i e s t a n R- E x
t e
g x J k 5
7
. 3 E
c E a A k c
F C r
a A r 0 . 2 C
5 S ^
E o, S 5 2 i TTB I 3I T
,o ao g
. I Y ona L o A 8 - - - - - O N
I g 0 8 0 8 0 0 8 6 4 2 A
1 mC - ' 3 g C
, l
1 I' o 2 i
l e 1 A
J-l c 8 8
l n /
a o 0 1
B i s ,e 2 % R n W O t e
x s s a
2 1
e e *
- a E
i s 7 r I
- - - _ - ~ ~ -
i k
c a
r -
e 6 C s t
i -
e o
. e aoo g . 4 r s . s a z i
I o 4*.* s e
m . TTT h L m o . I 3U c n
r n
, I f
o o e i e I -
, 4 n
s s o
n s e
, i o o s e t x , e n
v r
E s o - 3 t
x k i 1, E u c a o . k C C r c a
r o
- R- e s I C
D o J
S h o"o>
E t i
s a I . 1 S W Y n - - -
L o e e
a
- e
,i e
A i s
r I
s
- 2 N a p ."
A m u
r j 7 5 'm~ '
) o T
( C -
C
.o IlI
m 3
- e 1 A
l l
J-e 8 8
H n /
0 a o =42 h 1 B
i s R n W O e e 5 '
r 1 ' ,
t x .
- - ~
6 E -
a ^
e k 4 1
c ^,
a r , . 5 C e 2 i 1
a s
egA e ,a yOaa g s s
r a Tri e
a m e
e,
- I 3e l . 4 h
c L e s
, n r n A
i f
o i o a n s e, o s n s e
' . 3 i
s e t x
n e
v r E ' t x
E u k c s e,
p C C a
r
. 2 k
c a
r R- .
a C
. e, M 4 n Ao S J ,
a E h t e *., 4
. 1 S ". 0 i i 2
W 0 Y n 0
L o a - - - - ,
a A s 1
i 9
2 s
s m e s
e 4
e 2
r N
1 i a
p A m ,c E T5 mG ~ ,%
o Q
C C
O 4 l
l e 1 A
J-t t
e 8 8
/
a 9
0 1
B R W
Q
-A ~
O H
g p
U-i t
e O S C r ;
q y s : . e
- ~
C t
s / - C
/
e C C
~
J T U e,
E o o 8
E o o N
E M
I C
E P
S E
S O
f i !
i
5
_ e H
1 A
J-e 8 8
H /
0 a 1 B R W
O g
- t r
u o
n
._ S
, ., sC o u
i t,, , e l
a .
f k ml d i
e e % :: -
r aF l
d .e ,E e l i l a e
o d 1 l r N eu l
l
- o u\
M F
- n r o \ E aC E F . i I ! l ' ,gI J ;
F , r.,
a .
3, ;
g' s
n
- e E
L s
i N
E M
I C
E P
S
)
T i E
S
.l
O 6 l
l e 0 5
o o
8 0
4 1
A J-e 7 s 8 ,
' 7 , , 0 1 i 5
O 8
t 2 8
/
t 0 a 1 B o i wrw r . -
a 0
R W
- r r
o
- me Oa h c
3 in 3 1
m, e n. 6-n o f o a is -
is a n 2 n e te P te x O a E n D E 3 2 k )
k 0 c b c a a r
(
o r
Ca a C O a a
o o 0 1
2 o a e a O 3 O ^
0 o
s o o o <' 0 0 t
o o
o c
a o
o s 4 l w u h4 s
O e ks4 R e c
o s
o o
, o o a 3 s 3
, w c, S 00 E 3 0
0 F r o
a S 2
0 s w 0 w
N e O
r c
a r..
E 5 e
r Oa 2
o 9h b o nc 5 i 2 M
I m
m, ns o o a n, 9 o 2 S is f c n -
t C
is n O wieP 2 te O 3 E D s
E k
E s. c )
ko o a c
o mr oC
(
P C r
2 O a 0 S c 5
0 0
- e O ^ 0 o
Q 0
o 0 E s o 0 0 0 o 0 0 c
s 2
0 0
2 0
S o
so
- S h4 O
_ l l: il i' ,l i
11l!l 7 O l
l e 1 A
J-l l
e 8 8
/
a 0 1
B R W
O fy 4 0
0 0
4 0
5 3
0 0
3 m2 0 0 0 1
0 0
1 0
5 O 8 0
t s 5 1
i s
0 _
l e u glag u
s n iVi s
n s 5 e 5 2 UoU e -
es 0
0 R
t m-ame e eu hPh cla d c l Sa SVo o mt ioe t
n nn o-4
- 0. h h
t t
mia aia -
t 0 c n
m imremP -
e n
, t pt E sx sE i
eI,n o FEEEF o -
M i s
n5 3
0 is n
e e7 .
O-n t 0 x
- 0t x
o E O
E O k k i
c s c t o O o a C5 r0
/
2 r
- 0. C i
m 0 s
, O 0
t - -
s 'O' N E 5 2
0 f
'O, 1
0 0
E d n i M
a - -
f
'[( -
0 0
I E 0 (
0O C
0 0 0 0 0 0 0 0 0 0 F 5 2
0 2
5 1
0 1
5 E h4 P
S
)
T
(
E S
O
O A l
l e e 8
1 A
H e c n
J-8 a 8 a r /
0 B F y n 1 R
,t n a aw e s a p W h e a '
c m aW r ,
r e
J r
n d a .
G e n a nd aa H. t s
wapA opaS l
nC,m b. e nP aJ,U i
F,uie t Wa pi c
,a r J y
,t n r
eqe h t r J t si e s oi S i a cir er ei s T
t nmre tg o,E l t sivv N
Cteoh c t uk y
u nD i
A Ai t f
oodUn P S s I
h e s r T eI l u
O C ci r go l t
a ou yn f t ab I
T R
ar t ee ka t t s tyi t aiu v om s
e n e ni s et al o A a s P R 'E Rr l
s r nH N C I
i ge h k a l
a a ct f u
ni v c aml l e
at u nr as N a ae tt i
f nif r SIB h a h a eU u
e s whoa I c s c r e ak B es s p R K o SO. Ti mle l
l a Y Y
L R ms i
r e
l a
r AD oe ya t
t n
FR G O
B 8
l l
e .
1 A
J-t t
e 8 8
/
0 a 1 B .
- I l l
- :.
R W
O l .
e l d .
p o .
i T
M
- k
- c i
s .
- - L a
r s
y
' : $ l:'
r , i C
( l a
l a i t
n i n
A '
I n
e i
m c
e N
I B S
)
p )
b N
(
O T
(
R C 7 D.
N .
U O
R E
F
O 9 l
l e 1 A
J-l l
e e ,
8 8
/
E a 0 1
B R 4 W C 2 0
04 4
2 0
2 6
1 2
1 9
4 g
8 s
n o
i) t2 5 a ,7 7 1 u4 q
1 edn ga n 1 i
, a is 4 o ',
l 0 '
( , 6 5 i ,
1 y D g
',', d J e ',',
x ~ -
v r 5 $ ,
2 i
{
', - , 5 u 1 ,
C ',
S m O
E
',X 0
0 s
e OTL h
R- m0
,0i O '
, 4 i c
n J G 1 ',', U ,
A , o U )
T oW
[',
- O a
S (
C 5 0
' -,x A
U , 3 E
i 7
W ,
O R b
.y O
W
',x
', D s i 2345 o N
I 0
5 i
t lu s
e t t on on a op a p p p i p t
n n n t t 0
f D
, 2 R iciiiiiicc c ic y '", A u B t n
e mPPPPP t
r r t
o a oo o t
r t
r t
r o '
M ',, U O 5 2
i le E xMa oa o ',
y's g
R i F
te in o O
/,
D D 0 o
p' O
)
' 0 N g g
0 2
1 g
g g
0 8
0 6
0 4
g U.
O wf \f h- nm~
.n R.
. O
m m
m 4 6 "5 17 0 l
l e (
6 6 6 0 (2 3 4 "3 0
2 A
J-f l
e "6 1
0 1
=
D r
8 8
/
a Dt
/
o o
0 1
B Z R W
O t
-s D
7 Ok O c j 2r a
! C m :
l e
b o
r .,
P ;
'7
- 7' ,
e p
i
,6- -
P C 7
- 3 N W
- n I T ** ew $" .
B -
T - T,o.
O l
?
m ds/ 0 .~
l
$==
R i
=
t s
e '
)
Z se
- ii N m
D 0 4
S
~
2*
5, N
=
p=
U -
q O '
R .
E F i.
N lt UJQ I","
o . '
O 1 l
l e 2 A
J-l l
e 8 8
/
a 0 1
B R W.
3$ t t
e u
4 l
e d
o M
i s
s y
O l a
n A
e N i p
I P
B O
R D
N U
O ,
R E
F O
Of
.a.. 2 E
a v=
w/NM 't d
~
I I i, I i
! N N -
R
\ 4 o n N , e l N * -
Ni 5
.3 _ N E.
- N % , .
N O C g < * =W N
@ O e g $
h 2 $g s o d $ .
-.J O -
ss o -
es g W u) ' % ~- - a d E J
5 d! }kkkk jswww j a
!'! b l -
Ekkkb g j o o <! +
- j
- a
. -- 2 3 O , , , , , ,
u g .
O l: l
~ l l
l l-l
~
Z ul/sql 't
)
O E
9.
.l1 11 3
2 W
O G-9 l
l e 8
/
3-l e S R
l C a za d a N d$e A B
0 M
Ce 5 0 2 1 5 O 5
3 o-n 0 0 I i
o 8 o ' 0 3
t _
i c 0 0 I 7 0 s d e e
r m
m M A
' 5 2
h c
0 n P T
, 0 6
I a
i t N o
- ts n E M lu s
T N
e E 0 C 0 1 o
o y^
la e
r ' 0 2
E M
E R 5 m tn C S e L
P S
y a e n
e A
L OTL
- r c I D 0 04 U e p
x b y P S
l a E i
h d e
e s
lt u
1 2345 t t t t
' 5 1
I D
p N s 4 n n n n U
t I [ c e R e eaa vap ao cp E s L 0 g*
r r
o ; wect t t ticq N
I S
- . t i
D 0 I C . r r r r t r L
D- R 3
- k
. boooo I PPPP -
E d O
L D M
4 E it e OOa+M
' a D l
R O
R a o
0 0
2 i
og+ i F
n L N L 4+ '
5 .
I e 0 0 4 B i p 1 O P 0
O R 0 0
0 0
0 0
0 0
0 0
8 0
0 M D 5 4 3 2 g N " - g ady1 yr$
U O
R
= O pt(<!
4 l
l e A 2
J-f l
e 8
/
8 a e 0
1 B J R n W O s e
o n 3
)
s y i o 8 3 5
t 8 l
a c 4- 8 n e 4- s r P e A r N R s e o C y p c o /
l a
i e t G n P n n n o E A d o zitoi R e e - t c U p N
k c i t
cac er 2 N i P
a i sf r 7 (2 O
I C r
) l adioc )
5 4 L d
e 1 po 4 -
C T l l
3 M 53 6R k c
AD S W a 9 1 o3 8 t
u 4 3/BC B. a r .
MO h P th 88 4R nR GE LB )B C -
m a
Niw4 CioC/R Gc e I
r TH g u ( t d
g o
ST o l aaN l
O/aGU GcEN nE aCr a f r
p EE h r
nn i EfiR( ( u
- h e
K i ii EM T I ggRRdUCL 3S*
i i r r aRPOOENM( PU o 1
N R C v. taI NC t
g VS l l I
i n
r I
i t
s N. B. R eiHH u TI nE e/ iBB n e T. T. d CY S r E eG r
aBB6 r CC d e
I PLLR eSS f p DL f o
- m. m l e
EA u c
e e e e u c
e . v e
RN r r D PA i i C C
- O 5 l
l e 2 A
J-e 8 8
/
0 1
M R W
O l t
e e
W 0 5
S s
s " A0 , O a l
e 0 0 I I 0
4 0
i n
t a
S r
f o 0 5 l I 0 7 3 s -
- n .
r.
e
=
i ot e.
e O t m i
d cw e
r 0
0 5
l C
R I
R I
0 2
w a P P N. E d . O , rA t
B rs/
BoE LPG t
a J .OAD S o 4eAE L
E E
I L 0 T
S O
S 5 l I 2
S m S Y u t s
e o
E L
N L m 5 6 p
A O
I A
T A i x
a g -
S N M 0
- - ~ - -
O A
- 3 1
2 t
w 3yk 0.
1 w3 E)j 9
0 8
0 C
W T
. O
,L
_ 6 l
l e A 2
J-l l
e /
8 8
0 a 1 B .
R 4 W
_ C l e
e 0
5 t
S n
o A I 0
br 4 a
C r
f o
s s n m A . O 0 h e
i o m r,
e cs r
t te e i
c m e
te m
a D
d i
D e
e r P o e o
mP P
d a C I o R R P
S I
L BaE LPG E
N. s /
B ir L E
E S m OaO T
S u
Y m GA5 N O
B L i x
R A
A a C
N M - - - - -
A
- o 4 3 2 i i
0 1
9 8 1 1 1 0 0
]J 53I1 31 C -
W T
O 7 l
l e 2 A
J-t t
e 8 8
/
0 a e p
1 B i P
Z .E3 OhE 3,$O ,
R W
O l e
e e ei N
r t
e s
e t
e s
e e
e s
0 0
4 e
e 2
e 2
t - - - - - -
s S
n l a
o t n
e - C R LL I 2 br e ir m
e s N.C B.CB. i 6 eI p iB B B 1 a 4 x E aBB EGPLL L r
B s C m m
h e
c a T
, O -- -
i n
r -
o N
E 0
- -- ' 2 T
M 0I 1 N f E 3 o E C M s A L o E
C n P A S o S L OE I , P i D S I
t c E e O , D N
S ei ' 8 i I 2 ' E N
d L o g\
Y
- I e
r D
R o '-. V L
L P O
L o
D R
A t n s k( .
O L
N e e3 l
o
'4 A
m '
e '
/
J N c a - -
Ci ; i l '
i- c O
5 l e e I
p e e
0 5
e 0
0 5
e e
0 5
e T i s 3 2 2 1 l A D- py E3 0 E gh8 d
M a I
o T L S
- E
. O
8 O l
i e 2 A
J-8 8
/
m f a c8" $f gg*
0 1
R W
O e s
e s
8 '
9 2 s
- 3 e
7 n l t
a 2 o
I n
- e m
CLL R '
5 i
t c e s i o ' i e
r p
xE aBB r N.C CB. B.
isBBB B
2 i EGPLLL s o e d m -
- h e
r e
,e
~
O - ' 2 I c
n P T s i N
E e
- - - T t M N E
n E C
- M -
e E R o S m L
P e4
- ' 5 A C
S L O
n-E 1
I P e I J
- S I
S c E n- D a N e Y
i I 3 -
o- E l
L N p - - I L s D R
' 1 L
A i O L e .,_ ~- D R
.D 2
~
O L
N d M
A a o et i
/
p,
' 5 L '.,
N O
I a - 7_ g 0 e
a
, s 3 T 5 2 1 1 A .R3 S2g S M
I m I T
S E
O.
,Ill , -
O 9 e
H
)
R
(
a 2
A J-e n
- 8 8
/
a 0 1
B L t
C R
W O
tt L
9 t{ t y
- t a
e n
l a
P n L )
(
c i
o t
a :
i v
S r
e _- R
)
D
(
I a S d Y o 2
/
OLAe f
h ,
e n
t M a e lp
'c
_ N M )_ t r
u e
A n o _ {_
_ _ g N
k c
a r
K i
_ n c t id a _
n e
e c
a C m _
_ b e
f r
u s
A i r t _ u l s _- p r
a n
R _ r
_ e E _
d n
u 7 i t
e n
C L e
p
% 'S i g
n n
Vt ip '
, ia E
_ t d
-_ e n o
_ k C c c
_ o b t r n
_ c o A -
it
_ e c c e
_ a s F _
f S
r u
s s
o R
r
_ ) C
_ (
o )
U -
(
b S Q {p
( ; g 2
O
/
(
M
1 1\l1 iI1 0 0 l
l e 3 A
J-f l
e 8 8
/
0 a 1 B 0 R
W 4
C 9o 2
' 8 1
~
' 6 1
n 9o i
o ' 4 1
t c
S i d ' 2 I
e 1 S P r yo t Y d o o ' 0 1
/,
L a R A L o o o 8 0
N n
A i o o ou '
6 K i t
a pm C t i
n oo o 4
A I R
C 2 E - - - - - - - - 0 C 2 8 6 4 2 1 8 6 4 2 0 A 1 1 1 1 F 6* al5 _% G* S iE" n R
U S
0.
I l ll lll!llll ;l i i -l
o 1 l
l e 3 A
J-t t
e 8 8
/
a 0 1
B 0 R
W O g
" 2 8
1 6
1 n [
i o 4 1
t i
c S
I d
e i 2 1
r S P $ t Y d i 10 / ,
L a R OA L o o o '8 N m A u o b,
'6 K i m
x $<
C a '4 A M R '2 C
E - - - - - - - -
0 C 2 8 6 4 2 1
6 4 2 0 A 1 1 1 1 5" g F ^-o=3!r %G3 g "o_
R U
S O
, ,l
I 2 l
l e r 3
A J-o n
e i v
a
)
t s
n 8
8
/
a u Be r
h " e m
0 1
B t c
s s
R n W C a e F
r o " s i
t s a A m
r B o B f L e (
D 3
M A
R G E P
O I P
C R R P
N G d e
3 "
N o I
P C I r e
P t u t j
a p a D m D y E o t y
r t
r e
D1 1
C e p
w m a
AE E o r
l F o e
RS P I
P l
l i
a G e
GA P C
i a
r i p
EH R t e
a P
DP N M i
1 { ! :
3 l
l e 3 A
J-l l
e 8 8
/
a e 0
1 B J R n W aw s o e n 3 )
s y i o 8 3 5
t 8 l
a c 4- 8 4-n e r P A r N R e o C p c o /
M i e t G
A P n n o oi o n E R
d R e zit U cac ecr
- t k )
G c a
i t
2 N O r sfir adioc 7 (9 5
k C ) l 1 po
) 4l-4 6R C P M 3 l
l 3 a 9 u 1 o3 8 t
5 4C 3 /B B.
G W -
N h Pt -
h 884RnRoCG ELB
- B) g I
PE u Niw4 Ci /R d G
( t o l aaN l
O/aGU GcENaC nE P D I
r h nn i EfiR( (
O iiI 1
T ggR RdUCL 3 DC I
a R r r PU o N R C v.
ii l
E i t POOENM( s e DP E n /E iBB N. B. R e
r E - - - -
r AI e G aBB6 PLLR RP f m
GC u e e e e e ER c r
DN i
C _
O- 2 l
e 1 W
l G-l l
e 8 8
a /
0 B 1 R
p,w W i
a r
t e
i r .
s c s n n t
n i t
oo i t
e a m i c a u l
p l pak o pv a s s Ae e l
e r e n v ywb l o
O e r a e - b i D ol f r a
r t
s ,
t S t r ae o e v
e u
T l
A- ui cf e Q
i FL l
e gvb e r
e ek S OU h t
- R s a D
R E ES f o t n n Le m
a C A
CR e I
- r e r g t N t a r u - - o s r a AM t
S C P P CA . . . .
FR I
NG I
GO I R SP O
s aM
_ 3 l
l e 1 W
G-H e r e t y 8 8
/
a d n
i l
i 0
1 B u b
a R e t s W aw i p
C p
d W e T d k B e c B d a r L n c r a y o p f x l l
s e i a e y t s
. l nd y t
a ee r
l a
e e n d r f n g
e ma u p a
w s c x e a r e n T b i
c g R .
a r i n d e
AS t a f od a
p o
E T d eo l e
H N y sl t
r a c v T E e bi n e p d M o t a o d FP n s t
r a o p d n a n OO EL l
i a d o e m d t i
eau o
r l TE e i aw i
f i l AV t a t eo r e v a
TE l
M D s Ve SD . . .
1 ,
iII !l l i a
o 4
C 1 l
l e d W
G-8 H e e p 8 p
a o t
c i p
/
0 B l e
i d c 1
R v e r
i W
aw d e p m a
s t o n i y s
y d d d e r os l
a e i f
i n s r f d a s e d a g e s
v el o n d pl i
s n oa r a a l i e s et r e e d t ve i
n s y
es pe di en g
n l orr t h o M e a n
a l
e a vk pi t ew c
F t
n ec d a ny T P e r oi tl d c ci R E m o t b AS e e h dn n ta p l t e s E T ea e
i p my m-HN t e t e s C e es T E l
i n el i S i ci b ey M f n h at a pl FP f
t g o l a s i
pn a
OO EL l n
e y
c a
bi n yf d
eme a
e r go k TE t u c
r t er ct AV i n c na asy r
TE i F A E s t
Cs SD . . . .
o
,l l l
O 5 _
e 1
l d W .
l e G- _
l l
e r e
t a d 9
8 a t s e
/
3-t e n t B E b h p e i c S R
g P Wt d C gw I P
n o Agm e r A u i
r S no p L e ec .
d tr E E t e E i r nsr .
M T c al u s s S -
S ad -
S
)
s Wte ec e n
A S
E s
e n
i aAmo r
eM r h
g d
h e
L h t S ep s u c o
i g r -
N u r as c ob i t a
e I
A od d f s r T t n l snl y W -
A S wu e weh saan A S
d e s R o o x nt r a t
s s I
R O I
( b r uh g eR I f
o t e e n O E
l l F a e f ut Ph f l
a s h g
e w t
T l
S o E oaE h e u I I S ml Mt r /
s e o WR Y d' o t
f S oa t G l
E s
t s t AC L l en A h e e s n ngmv e n i L A s r FE N wtu i i i
h g
e e A sh d t a l E C db u n w 0 n , eol ev o i
a o N
i 4 ae g co wre t t s
l nt s CA I
6 3-l aa r edd s e n t d VT s ee n s t t l B ee sso i l
a e RP W mva i sUUC n c W EE I e n n o i o d s ed SC s aa c - - - s u ga -
NC B h t i n F Alo I A e e . .
O .
rl l' #
O 6 l
e .-
1 W
l s G-t t
e k c t o 8 8
a a r y /
0 B a c a . 1 i
r w . R E
l r
e a W aw P I
t i
r i
t n d e
P c e n L f r u E o f e s E t s t T n s m n S e e u e m n c r m N p h i i r
g c e O o u n p B
l e o a x R v t h e A e n t d i r e C .
w e p A
l R a o w i
p OI R O i
i t
n l
y l o d F r s e E
i S n e d k d T I i v e c l
e I S d s s a w WR Y e e u r c
L s r s AC A N
u d u a t
a l
a s e
L a e i
x n FC E A 0
t a c o
d c
a r
h g a.
d E N 5 e
r p J l
o ui 6 r f
o re CA I
3- u W d n s s
t t wr i VT B t c A a e o c RP W a r S y p h n l e s s
EE e I
f r g SC e p m o
a u mes h o os NC i
P S C T S a I A e . . . .
O
7 l
e 1 W
l G-e 8
- e G d e
p 8 a N k i
p
/
0 1
B I S
r o d R e W aw S w k t E s c n C i a e s r t O y c s R l a n i s
P n u n D a f o o 0
c N 4 s d A2 6 n e 3- o w N V i t
W m r s C 3
, I o t s
A I
E 1 L 3
)
R f
e t e
R , E0 / O d e
S G W e p E ( g i T L E s r a
p I A R r l h
WR I i RU a p w c n
AC EN o T e h i
L A r s 6 FE M y
s 1
E C a f G
l r
t s o N N e
v e n .
CA I
I P o t
e b s i
i p d
- i t
c EE .
R w R ni ud SC W l
r O o e l
B o e r NC W R F w p I A e e .
l ,
l
O 8 l
l e 1 W
G-H e t h
t i
8 8
/
a r o )
w s
0 1
B p 2 e R e 7 O r y
g 5
4-h m
e W
o R c l
o C
/
s d G n o E i o
t h
t R a e U m
- m N
(
i t
3 s
. i s e l
o s y e r
V l a t u
n c O 1 a a r
6 e f 0 r 1 u C S G t
c W N E a r
T O R f l
a I U C n T N R o A
i t
o N.
t i
y d
C I
s n
B B dara c
L o L d u P i t
d e c u e p c P b v o a A i r
t o
r l
ee t r
n p vt B o m e e B C i Db L e e e O
1
9 l
e s 1
W G-l n
t e d s
i o 8 8
t n t /
a s u i c 0 1
B n o b o d e R r W O
i t r p a e u w a l
a o e r
v l a
e i c e g r g n e a i
s n k n e a g e e
c s l s
i l s s e e B s B s s L a s a
r o o f
o t t
d d e d S e s e p u s .
N o a u s O
l e t a m a a ra t
I v d T e d g d y o A e t
r e
t n
ep r
C d o p mC I
L C o r iRr P p eN E p P P l I a x re A P i
r e h C e et B R t
a po Pi n i
B. N M L . e e
0 3 l
e W G-l e n 9 t
t i o 8
/
3-a t a
z S B i r
R C
pw t e A c )
a r s .
a eC h ) sA y
cC iR aF a R l
nI iTr aP t
e D , BC aI R BR mP Lr / N dE oA s n 1 f
( E t a1 r
o ts3P CM O p n Wi n e
r e (D T a r
y it ms r
l aa t
a s e netid ny mtr p ed mo x mo etr r
S up eic r ee f p E ser ep pP eCmo x u r L mlapeE i c p B a r c n7 Pi r a l MA i gpo6 P oo a I
Ci r AR r p t t a - R - a C t e
RE 1 8 1 D N M GV I OL . . . . .
RE PD .
O
l l
4 W
G-9 8
e H n
/
3-S e i o r ag l .
g R H t a l oi t n n C a t a i ni i A l
a n c v ed l l n B i c i e
r i
l e nh ee a r n o r
i d _
O h oef b l
e o h n e
h f n ib t
a r mt un t
i o re n o i t witcid i
n i sn eiaci r o l
e e s o i p pr w nimc r - n o
r i
t c ooisd u r i niio pd c
g n .t i
c s ao d nl s S s N k f uCe r se de hf r
eias d n ntis k c
ai xinwtsee ne l
r wtu ot enr O c a a s d d h oed i I
T r. nggh a en u sl t
n r h c -
S c a umsefcegl t E
h y spi ost spr k t t elyae erd utmua t i ox i U a Q w.eAow g n nmlaske i clar c n r S t e ig7n h 4 wo mexlaai l ioo c .
esr a sair txc n
, i t f R apu5 nc0 C uio4 A a l p p t R pi h i
i sp eb xgr di enf 6 o wal a r eie5 f vcd eCsT eega ami i
xf oe3 T et /s e l
k h t yB ce pW i
S r rr tnGe .
pl d ceAuc G A hf e eeEl nio appr i r ir r o P
mR mop CcPMicd I N t r
ui n uU i
at r ece t
oE I
T f n cN s eh sM -
oo o nni i
r t nS E t iDiidO AA E d c t -
M ee er i _
Nd . . . . -
S -
R C )
A 1
(
e -
l ,
5 W
G-O 9 8
/
3-m s g S s - . g n R eR sn i l
C nC t i l e A ud E sgG uuE so h/ t i
d c
e s n e e r b ds m.
.o d
AW r t e
R U
r p
r u nar oc e vW r o sooin cav er l
)
eA N e d d d t e ru S n i n eni th e u
n t
c l
ee u mesef r mlabi f
i t
u r t eso e p
i f r i
pe d n
o t
s sh st i
p n ophe eivs t h i
C o st c xt ut cl e as C
(
r s edi g tt S inp miae W l t
d es n uet c nis N T f t sk i o u O
I os no f o ecd r ae c a T t c nd A hs rcpt r ec O. S E e o s E nr al s ue
.e u e i
f f nt F d d a e U r eo o y l
Q h sw n a n t
ue vs i
eift f S h se wd S f s R
t spe o r sola oa oys C sl s a e G A smh e el nslaml n i
N T s or t staos s nil i
yraGa i
I S
s c u f 8
. t a ed abR t
s T A al ao8 7 I E r c tomdiAAoI nf oPf P eiN4 r E h m t depe r e M uh F c .
ex N e . . e S
R C ) )
A 2
(
3
(
O
, l l lll t
O 9
8 9
1 5
1 h
c T t a
r a CY g M ET JI i n
OIV t
e e
RT M s PC ee -
P A t ei r t ot UE i
h8 O9 z GC deus bL s
,1 ,
t r
u RD t n Su eS b u 5 1
yK b
O ON sRm b
mh J TA .
e rCu ur c R AS a l
pao l RT sC C oM EL t
nl e NU el nt e ES ot p a GER m B
o M E C A D l a
E t e
TN M P S ,
G G
S O
4V b
g l l
O 9
8 9
1 5
1 h
cr a l M S m a
- E o1 nS e U I t i
nr s
- = S ro f ef o
=
=
S I
vore I S
=
gu I Y n de r o
T ico f e
E aro t
d
= up l
F o A gd en C
- O -
S
/ Ct r
a E M
Y Rn Nt e Sg R x Aibn
= f oe f
ou O yy yr t
=
T cc co A an ue at a u r
_ - L qu qe
- =
U eq en de r de G
E Af Ag
=
R - -
P
_ G G
S O
O 9
8 9
1 5
1 3
h n 8 cr o 1 a fCi o Ea t e
M s d d sea ecr i
u
= nt g i
G
=
e vad ce g2
.1 i r e1 c peb t
R
= -
l fl f
edu et I
Sed 1
l- S eie f
z I
i o u f
= E dmi n s a r n G.
V I ytn ed i o g.
=
T :
i l
t r a a i se vR C
i o bet c e O -
E t
r o
i l
an t e rg rn J t ede og i
= B a r
rnd ea f
t g
= e hl o uu
= O n e
t e oe at ns pP n
l
= T G itniu i
aa e
=
C y r
mnq rei ou E r n leT
= J S u ev t n hc vd en eo O d Dct e Da
= R e
=
s P U - -
= -
P G
G S
O
l OVERVIEW OF W MODEL51 STEAM GENERATOR #
O ,
PWR STEAM C4NERATOR -
- 1V%O". ..,o.
w4 sere u.. ,o.
Ma%AAv WER $.88LL C :,'l1n ae,,o. .
l L!
P EEDW A'8 A ssutST asifivtpaaf@n 6And O
LDREA 8*sEL4 a $U fh r;,n s,,.wg h; l
) ...e.ss 4 vb I
O 9
8 9
1 5 l 1
h S cr N a
M O
I y n
o T lt n
i t
c
=
C % a c l e
= E 0 i f
e s
= P 7 i n
g e S
t
= u i b
= N o s t u
b n I
a o i n
E i
=
t t b n
e a o
= -
N I
m id c R
=
L e n d e i n
= E r e u O S d
e g a m R o
= A t
a n a r p s o B i o f
= c f i
i t
t c o s
- = F r a e s e
i s
= O p t
e z a s d i s b
_ = Y e s m s d a
= R a a et e t n d
= c re e
=
A t o t s
or pe v r
M w e eff i
e T B Rd S
=
=
M U
- P S - - - -
_ G G
S O
t
O 9
8 9
/
1 5
1 f
h o n c y i
.r l it t s
M i
b tt n
t sn e
{a 61 s y ye m C e
l e1 am E.g a
l
= Rio l np v
- n aiu o cn c at n f r
p ni is Rn f
oe oq t s a
i e m te
= 9a tm cp t
cele It n oc l iosb ei fo e ror
= fu fr dp S sb v
f Eq Etn Rm Rp n
= iaC' e o ala N nE 1 c 1 c
ii Eu en nte 1
r Dq kg I
Ay ne B d c iea it bn Ne asi c
O -
=
O nn ae oZ Rhit Rht oi tned nd ae
= R nq u aa n cic r s i
r D oe dwn dw n ee s,a
= N ir tf i.i st un oo un oo ltu aq ev bda
- i uu l
i i
/i o
=
=
U aAr Rta st Rta st dn eh r
ps
= O c ie ie cc e Atn sr sr ne Tb
= R aerr l yp r l yp r at Uo t
au a e ae vw r p
= E Dc Ai nt n Ain nt de An
-- D N
= - - - -
P G
G S
O
\
1
O ~
9 8
9 1
5 1
h c
r a
M
=
=
=
DIN d e
= NB s d
e s
u
=
AO m y o
d e e r
NR a
t c u l
=
e p e d OD p t
= m e d e s c
= I N e n o T U c t i r
=
n n s p O -
I SO i e
r e
m b e n o
I e p u UR i
_ = p i t t c
x u QS f
= e q o e p
e t s
= CIS d l
l a e s n e
AY i
= i f i c e e t
= e
= AL v n
e m a
m a
- TA i F d S S AN I
=
= DA - - - -
=
=
=
P
_ G G
S O
l l l
O 9
8 9
1 5
1 h
c r
a l M e S b
=
E t u
V e
= I T :
o t
a d
C
- t i d la E e
=- J v o
s n
v o
B t
m i o s
=
y O
t e
r a t s t
n c i r e
=
t s i e c N n d v e m O =- - O e m
i d
n e s
f d
e gs en s
g
=
I e so n T
i t
c c d eit i
= e t e e d
=
A p s
r o
p f
e s s
ba uu i n
=
D 0 e r
d i tq te f
t n
I e m s e
=
L 5 5 m r y ey u
A n i f
r s
a f
i t
t,t t
sg n m u
V n a n e e re c
= h o ut o
= E t e C M d I Sin D
= D r o
=
=
N M - - - - -
P G
G S
O
O 9
8 9
. 1 5 l 1
h D c
r N a U M
O
=
F S A E
G P
I e
d Y g n
a a T
t s
e a T 2 g w
C a
=
rd an n o )C st d E ea s i Cas nn F gS w is oa O
t t
= w1 s s i e i gg
= E ntcwrap k o Rht g ,op en D i o ao e i
ewie n Ri t t
=
F gtnRmd e
t a( t dt t e p i
on i
= r l P g ph Rcieh n ee
= O di c t i r t hi t c h re Y
d BCn nVCi u r
- o ng r
- SvC eeC e r peliu
= R bASGC pDSC u
bSS u
= A U S
T M
= M - - -
- U S
P G
G S
O
9 y 8 r 9 t o
1 a
r 5 o 1 b h a c L r t a s M F e w
- O h
=
E '_,
t r
=
k o
=
= G N c
=
= A T N i
f i
S E
= c
- a AIM P
=
=
=
WC
/E l
l e.
e
= GP t t
= NS I B a
TS
==
=
TI I
Pi
= LG AE CL-
= I PT
= YO
=
=
TH
=
P G
G S
L
O 9
8 9
1 5
1 h 0S cr 9
/M M
a 0A 9 E DT 0
0 NR x 1
=
NA wx
- . e x y m0 i 8 %
OA . : . . o a s
T D I s
- " o L
C D '*
i0 6 l l
a E N W
T A y O
E E i0 h
p D N g . i .
4 a r
= F LI g
o
, . l
= " l O E i0 2 e t
a
> . .
- o ,
M Y S
= T A I
= - L B I
0 8
2 0
0 BR 0 = u 0 0
=
1 AO fg!:gE$E BF
=
G PL G
S O
O ~
9 8
9 1
5 1
h cr a
M N
0 O
I 0
1
= T CV
=
EM 8 0
TA s
s EE o
= L DT 0 6
l l
a
= FR Wy O
OR h p
=-
= YT .0 4 ar
- = TA g o
= LA I
I 0 t l
l a
= BR 2 e M
= -
= B F - - - - 0
= O 0 8 6 4 2 0 R 1 0 0 0 0 0
=
= -
P 8Ef3yEf5
=
=
P G
G S
O
O 9
8 9
1 5
1 h
c r
a L M L A 0 0
1 WA
= -
LM 0
8%
AA *x s
=
UE , x *, " s o
_ IC T
- 5x x#
O S
L l
l a
=
ARR x_
x x
x w
,x x
W y
DA x x %xx 0 p h
O = NA , u
- ( a r
g AD *
- d -
l o
= D - x x -
0 t l
a e
= 2
- ES x M
= TS -
- u :
0 0
0 0 0 0
= I T
0 1
- k0=
6 4 E *E.j ld 2
S .
= -
E P
G
_ G S
O
o 9 y 8 r 9 t o
1 a
r -
5 o 1 b h a c L r L t a s M L e A 0 0 h w
WV 1 t r
o
= LM
- N _
AA x* x
, 0 8% i c
UE x n
- s s
, f i
c
=
T T C
- wxx
=
x , 0 L
o P
a _
x A.R R
l 6l -
e.
x x a
xx =. W DT l
x p* y l e
o x x h NA ,0 p t x* x
- t 4 a a
- AA x x
r g
o B
_ D - x x l
l a .
= ES xx "
,0t
= 2 e -
"x =
TS x
M
AO
= ML I
0 0
i g
i e
0 o 0
=
T 0 8 M 2 1
= S # {0 53 y Eh 8
=
E
=
P G
G S
o I!!l llI l
O 9
8 9
1 5
1 h S c S S a
M r
OM LA
~
=
L E 0 0
=
L T 1 AR . .E '
- WR 0 L A y4 . , 8%
=
A A s
D
^, t- *
. . _. so
= U ' 5 p4 ,6l0 L
= T D -
- a l
a C N ,
s t a-
. _ W y
A A oO O = .i
,0 h p
a
=
D E ,
Po$
., _4 r g
NN
= l o
A L l
I
= .0 a
' s ,. ) . 2t e
= D E .
- 8
- s. s M
=
E S _
= T A - -
," i 1 l l
0
= AB 0 0
0 8
o s e 0 2
o
= MR I
1
- .!di v$ - ,Ju
=
= TO
- =
SF E
P
- G G
S O
O -
9 8
9 1
5 1
h c
r l
a L M L A S 0 0
WM 1
=
LA
= - AE g,, 0 8%
= UT = , .o s
s T R aa ,
, J,**..
g,* + ^
o
= CR ..
^
o*,7 . 0 6l L
A T l
= ' a
=
- a
, , %[ .
t ,o n W
= DA ,
e,a. oo a y O =
=
NA ,
g% .o
*
0 4 p h
a
_ A R
- a- r g
= o
=- DO * = +.
=,e*= 0 a l
l
E F , *
2 t M
e
T S
o A S n -
MO n u ,
" 0
=
I 0 0 0 c 0
= T L 0 1
M 6 4 2
= S :
- j g3 ]iE%ml om i
=
E
=
P G
G S
l i O.
, l
O 9
8 9
1 5
1 h
c r e e a b v u o M l yt r p
b g S nn i r
i m
N ok i
t s o o
= f t O eca )l dt s
=- I pm Dl a ec S sy nl Ow mfe U ib P-(h ee ed Cs i
L Es o n gu sgge C
- o o dp i
t r na
= N tde ch et it ns O cn t ea ea e% e
- O =-
=
C f f
als d 0 f4 C r
c /wg sn
= N sa l
o>
y C ait
_ ic t
e S t t
= O s oe t
ig l
i r dr api
_ I alsa ba o
= T eac l
ats f do
_ A ba D ef fi d ga t
o tD
= O a
_ D I
hnd cia r r
p/w g P mO oP it L ra g hn gt i
w t ue uere
=
A Ccd it Hp i
L o
Ath
=
=
V
= - - - -
=
P G
G S
O
O 9
8 .
9 1
5 1 l h
c r
D
' t i
s a T e s
o M N g e p a r s e
= O t sn e w
h d C ao wta i
s t
p e
d n
== S N
/
gd na t
e a
nd e r a
sls mla n a t
= i r t m O t g e n i i i
= I i
ode t s
e cei m dig
,s
= S d ee h p s sh y gC U t aev r t p ei t oE
L m e e mw l ox C i t s d a s he O
=
N s
eo r
r C E
s n r e pl rp om O ef n fom mo
_ =
C dly nl a i
mc i
c e
t s a p cr
= ui yec n eo
=
=
=
N O
I T
l lap rs ee i
t i
o a
r t
t e s-o ne e
ff dlee xi b
e sn
= n- ad ve m atw A
l e a t po
=
gt er e e o mp D
Ce p ipTB d
o se
=
I L Ed Wre - -
Cr
- = A
_ =
V - - -
P G
G S
O
O 9
8 9
1 5
1 l h
c D r
t n
o e r
a T f m
M N s hs m go l
a
=
O at np ge t e e ut ad C i sd e nro v
t oe t
= u S Ce d op sb
=
N Et r rm oi ei bt r O x ep o f o on ro dt pc pr e es I
=
l S mn t
oe ga ni x nt
= U o i ym ia t
L c no c ad y
= C f ot i
ay t
oic roa l
a rc O = N ni r un d cru
= O i oa t v ce cui o nnc aicb aqa t af cba
=
= C t e one r
grer i n bnog
_ pe nf e
= N rcm i z
iai l
s o sl i
aiz
=
= O t e ri uc io scn a rns
=
I noe de- t o
= T i sp epo l
uidt e oenvrov t
A st s vst - f
= y s e o rga l
= D l agm e pnn ig e
s np
=
I L nr a m s,i om Ala s I us Uci
= A
=
V - - -
=
P G
G S
O
O 9
8 9
1 5 l 1
h S
c N r
a A n o
M L i t
c
- P n u
onve
- a t oit G f oa f
s Dic t ce
= N ys a Oef P pde I
t a )
= L D d s i
l i) ny
= P bL O ni l d
= aE P a/egv e
=
M b o
r (P
(
n L ni g Eli cg t A :
o pt o peu
= S t a dm i
ih t t cp Pmf f
o ae f
l p
= nl i
ee d s s wn I t S a ag t O d rn ed eeoa
==
=
I F t i
l y o' r
r e
dw f
ol a f
r gah d npne a
rmot c O
i
= b y a g pt t ie dode
= N i
l e
n iCc z e l
iu br t e cee t O r iEf s b" at adsd n a
= I C g e oo f mbe a r T E se se i
r t
= C Ein "d pn
= A y r
r l
de e eu l eit o etamew s
= U u det r dc du e e
=
L S o xf Meo c" on Mfu selah Uest vc u b
= A d e
= V U s
= E - - -
P G
G S
O
O g\
9 8
9 1
5 1
)
h c FL g r
a OE i n
gt M Y (P gi lum T T i
PL I ,
== LI BIM I 3
= 7 0
=
AL 0
- = L
= ' 0 1
= B G E P
= ON 0 R GI '
8
= '
= P
= F U G 0 p h
t e
O
= 6 D OL e
== NP 0 T
u r
= OC 4
I
=
= TE
= AG '
0 2
NN
= I I
= MD - - 0
= RE 0 0 0 0 0 EE 0 8 6 4
== TC 1
- O EX
=
P DE G
G S
g/
O l
O 9
8 9
1 5
1 h
c r
n a l i o
t M c N e p
O I
s n
T s i r
C b e
f o
E t u d P e t e
S v c i
t e N
I c
e l
e s
f R e s d i s
O O g u n e F l p i o b u
t S d n
c e
t f
o E
I a p s e y n G
l f
i i
p o
E t n
e % m a
T d 0
0 s
A I A
- R l a
o 1
_ T G - -
S P
G G
_ S r,
O l
O 9
8 9
1 5
1 h
cr E g a u M
V I lp T d
=- C n a
Cta na Eht t
= E g gs c F n n e
= E i g
i s mte D g u ead s
= l u ntde
= L p isub otl
= L d nis t
c ootu A a e emwrve p
O
=- NE
- D i t
gb nu t i
s s oi n eta c t
t ar fe
=
=
OT ce ev )
%ind 0d ce e L
TC I ti
= et c E 0n ege i
CE de ff P
(
1 f a eth E P oe ) otahf y d t o
- D nd
_ P S itla O o %
SN l
- i i
tt
= I bud P i ai ct5 l
= N S I
bv ai (
=
l ub e 6 aiapy oi rd p vlen s l E n
=
=
0 B Pn i Erio
= 0U - -
= 1 T
=
P G
G S
O
lII {
e O
9 y 8 r 9 t o
1 a
r 5 n o o 1
D i t
b a
h c
r L
L T E c e L p t a s M AC g n
s n e TE i t
i w
= AP c e %
0 ht r
= HS n t
e h 0 t o TN i
a d w1o N
== d c I
E E g n et i n a ml e f E isg eb i
c
= TB g um e g n hi s c P a -
=
NL s s i
poc e
t o
l
= c -
AL de e
p np o e
==
I l nb s is l O RW A
ae gb i
n t c a ee t t
e
= a US ps f
it nu o s o B y
GB E t
= ce ) t icnl
= ti ev L i
l i
/
gs
= OU de etc E P
b a i na pec NT l
= ff ( b oed
) o r mn
- E D aa
=
=
G NIT V tl i
l i
bu y
a &
l P e P
= sm aro I aid ( r e D df
=
LC bv oi
= h &
l ien r
= PE rd Pi n p w P F p MF
= :
a l
=
= AE G o
P SD -
G G
S 1
O
ll ,
yI lll o .
9 y 8 r 9 t o
1 a 5
G r o
1 N I
b a
h cr TE 2 L -
a CB 0 2 3
t s
e M EU 0 0 w TT = 3 h E E =
) 6 t 3 ) r D 0 5 0 o V
- ,I 0
(
0
(
= N c
GT )
) )
7 7 7 i _
NCE f
i 0 0 (
0 c
( (
I
) ) ) a T F 9 9 9 P CE e.
s 0 0 (
0 -
( (
n 7
= ED P o 0
p
= l l
e o p i
t p SA
- p =
t t
m e a .
L e l p B
- = NG I
u s
E P
l p
m n FN s a a m o
= I ,
a s i
=
OG l a
n 9
0 s
m o
t c
e YG i o =
m o d p
s
= TU I
t i
d D d n
n a i n
L PL d O a r I a P r
% 0 B D t u %
3 0
5 0
=
A N h o 1 _
=
B t O A i
W - - -
P R G P G
S o
l
e
. O -
9 y 8 r 9 t o
1 a
r 5 o 1 b h a c L r t a s M e w
e h
= l b t r
N t u
N o -
=-
O I
e v c T i t
c i
f i
c P f e a M e P
e
= d U t o e S O e n l l
e S
t O -
A OXO b u b u t t
a t
B d
O R e e v d E i t
c a T e r g
- S f d
e d
e
U =
- L C X O
=
=
=
=
P G
G S
e O
e 0
9 y 8 r 9 t o
1 a
r 5 o 1 b h a a c l e L r g i t t a n s n ht s M
i l
p ino u g e m nie o gb n w L ai s ie i
d h
= t ar u e t r
= A sd ce det ahc t
e o
=
I cd N TE i
t r t e x t
a ce g n pd ere c
=-
NM mle e s ei s e d su v na op i i f
i
- = E E t s
y e oes e r
ie c
a
=
UH sbu t
enh ues b t
ar cr P -
= QC yt uit t o dngs io d ee
= E S
S b h dc n o ni i nd e
l l
e ea o c di "e Cg G
O / t
- t e in t
ao n n Eg t a
=
CN I
c edn ciu o icr oz t nl p as u
B TI L l ea s derr in pu e AP hi -
=
)
e s sf f t it
=- p% Cnyu wim MM l
0 Ei ,e b l
E T A m4d i
= el are nd t" ae bg
= aeib un S S
- sot c vd t _
nr etu i
a% ep ee hg l
= Y i hsmo cg S t0s i
au
-- n2n I (i Wbmwoit Ep l
= - - -
=
P G
G S
e O
l !l
O 9 y 8 r 9 t o
1 a 5
S r o
1 E b h
c
%B L a
r a
0 0 U t s
M 1 T e w
= DE V ,
0 h
t NI ,
1 r o
AT N C
c
- %E 0F 6
o m*1 o
- a 0
0
. 9 0
g i
f i
c
=- n) a
=
4 E a .
- 8. iL 0eE d P
,D eP
=
0O c(
xd
. 7.E ol l
l e.
8 a 0ofhs e O = 2W FT 7
0
= ye t r t
t a
=
)
ih B O R l
=- W * " . 6.b i T
- 0aC bo E
= NO r
=-
OF 5. P 0 _
= SN I
=
=
RO - - - -
4 AIT 0 8 6 4 2 0 0
= PC 1 0 0 0 0 0
= ME Eh$o s E. $8
$ 1
= OP
=
CS P
N I
G G
S O
O 9
8 9
1 5
1 l
h c
r K g s a R u ae M O tl ap t sm oo h d
=
- W t sa n l msg mt a n L ac gi
=
A t eet i nm u C e l
= t ps
- sd s ms I
T ode a a Y mlubu sns oe L not l ow aib t
e A icu i t t rv net c oi N t e pd et par c e us e O
A s qn nee f
eida S
- ind
%ege s%rg i c0e
= -
N 0 0et h t0 a1 fd
= O 1 hf t o mso ea I
hd g
= S ie%
t t
sen yvi r U wt5 c6 sie L ne t ct epy %es C vsln Enio 0fl 4ec f u
=
N
=
O
= -
C - -
P G
G S
O l
9 8
9 1
5 1
h cr t a n et M
l e nn co S ma s
=
N l eme pro
= pod O ur seg f
= I T o pnn
=
A t dht ee L et su
=
U mtel ra c o un
- M I
f rl e f
o e ah
- S pvew t
c a
- D = -
O s nds p
i t
y o am ne L m l i
- i i
b R t ase e a s
_ = A l
ulth mss uc h
t pt a
c m i
=
=
C i seg r d
y u g i
l n g a l
l s
==
E r eai l n t s i zi n c T g e l
t cp o i D s g s N uim pt t
OClul a mlya d
= O asld o nf o e
s P EP F M
Caoh U
=
=
=
P G
- G S
O
4 9
8 9
1 5
1 l
h c
r K n e
s e n e
)
r a R h ly b
u o yev M O wid t i
t ka l
l a d cb e o w W rv u q
e pd
= o y d seh N wle ee a r r nd%n e
i
- l aa t rag u0
= O I i c m a sd e %o2 0r
= T t yi x sef eb o 0rn ua A 1 l
aor mtu s sh L n p e ee r ot t a t or
= U tr a os p
hvb ci s cm t
ene s
ore e t
t
=
M I
pi p n neun f
l c ab O S uo oe s s
== ae%
i si tde c
sp t
eerg t b 0 o u (4
_ = S t 3ia u m t
=
l stl n e s
=
N s s u nny gv e O e r s ieb
/
gh nib t
c tu
= a d i
=
I S n og i l
nwe ped lpe mfd
= U in t
vn add ee
= L l are ut i
miu atco s
na r
=
C ms s err %eg
= iu f l f u 0 he N
l
= Sc l
Aes 4wd
= O
=
C - - -
P G
G S
O
O 9
8 9
1 5
1 h l n c d od r
a S e r pan M E a p
u n I
e oVo rI i T
I r
p s g Gxar i t
V 1 2
e r
gd t n ns I
ie krpon T 1 d ew opm C
1 se A d n
i i vv ee WAe fd ao e l
=
E af rr 3f a f
oC inc c on
= -
D 8. t yR eia ps s
O 1 dN Sim v r
=
sC u r e O -
C eR t
se Ero f d Mrr d
iN
= -
D u y t
cn au Sf oe p N Gb Anf
- pd
=
A gd ee i
mn a nioo i tt E dan Rwe ed ne
= -
D I
s di v ee l ue ate vl d e
t ei m am pap a U e sr t p o ixlo,n c
d i f
iee
=
G i u vd en am ro C t r
vte
= . G Ra Dc E aCen PEde G g e
M S
= E R - -
A -
= R P
G
_ G S
O l
O
~
ys t r i
l i
t o
bc e -
aa t e
i l
e t eR i
R m s r
m e Eet i r
D a m'f o o y2 r5 c t o3 W NWa r b u
a r
t a9 r9 E ft g o 9 o I
V t oho gr8 9 oS t a b 8 yr bn bolt ao E nip 2 tSaL 9 dctt g
OR eL y 2-dR e gs 1
eos n tDeh i
mfoi tB tCnu 5 n ws M e e l i
nAib t 1 sRha e tW A v nbo e r st s em ooaN h PrSoNd R
G rtli i en e u c O ip pceN meRIF PeMl r
n C o M
r a ia f
i c n ch l
R sE o p PR ac i P dnD l e.
nI ml a e (N o tte nc ovr i C a l
B i
a a ue t
t s
n e l
aI .
M vr Eof M
0 i l
s s n
b 9 2 ai o 8 8 t t 2 0 s c 2 1 e pe BD R o T Wt Ldi s
n r n e f
oac i s h rv e cr e u as qen n s i i h e r
f o
ct y t
e nt i ei l ndb i
t onel a i c pr e e e a pd e s
ni tnh r
a i O c e hg n
. r s c ut s i
ou i v
n t
cM r on e o o e e hr i
sh h D o E st n m i
s c c n R r
.J V i t
s e v
ma r I o n f
a T e o me g B r
S.D C y n r
p oes CRi g n r E i
l t
om e
g e r pi i
b yy ei r n g J amg r r ee r
a a B i l
on t
oo t n e n an a O e r c n i m
a aaigi n emta i
MM l l r
g u u n g M h ei o ggE n t t t ysyro t
t P
r RR e ef E o s c c ej e A f it y
r r jo o r R i t
n s fo nl l
ib aa e ei oi a
r r
P P G ar y s u an ia l
e l l c cit v s e LC O qma R u ui a NR R Toie r
E NNDM PN P pm ODN l
l
)
O
_ t s
c n i
n e o m
_ s e a r i
r u
_ ~
t l
u q e
D e r
_ c n E
i v
r o S
i e t c
s S n e p :
r E i r s o R f o i n f t
D s t n t n e D n e
e r m A m r
u e v
e c o O S r f o
r p
~
E i
u U q y m s e c i S r a u
d n
d l
S d n q a l
e en I
a e n s teWio Y a d o l
e Sl a tc a s e R st i i r t s ys ep e d a O i t
r n
a l u
a V e l a ems r l n r n T c s v e eia v al i
e A n n o o s e r s u eC OtS id m
a L i e
n e sl mi iS s zCd i
U t t u t r
g o
a a c n e v q e zSl e s o aias s r -
P r G i f i i
i t
c i
n r
PNIGWCDF y E l a a m e
h c
i l
t R u x f f e Qe i
b a E T Y
i l
e R E E. K . . .
ODN
~
7 ,I ,
O d o
C l
a s n n
? o o n i t
a i
t o s n i
s i
t t s
o o a e t n
p l
u u i g i
q s n c e r
s t i
t e e l u r o
y c s R r o
i t
c e r
p p
r t a u a a r h s e
l u p c r r l g y a o c e r t r e s
f u o s
u e i s
s N t e d r a
p s s n C b R
t i n n f a l
e a o o N O
t t
p n g a
H l
u d s e s o n l n r i
t a
i v a m o o c t
s u i a
r t c
f a
l a m h n
g r e r v n c e e o o i p f y e t r x t c d P e e r n
D l c r
a p a e
l a d s u -
t s e i c o dr i s e h n s i s n m T o e r ht a s s a e r
g s s s s sd p o e e e e e r d o id a r d d d P
y o i v
i v
i v v n l
e v
i l
t ib D
P o
r P
o r o r o r a t e
ia P P s D l
e w R o E
D H . . . . .
O N
'W' '
y
- t. ,
e 4
4 0 C o
I c l O j i
$N C i o l z c -
4 o ~
E-M co y
O C q C o 3
'm m
?
E c) 8 a i
.g <
a e 8 t u o *' '
l--- C k f
$ 5 $
o U-E i
D C o u
- ~
D o e
eg a L G f e m-
.O $ a O M 2 /
a - 1
. @D W 8 $ C W c O <C o 4 m
? 8 l
z D 3'o o 0 o
I j
.$ C C
.o s e g.
1
.O o -
03 ..
o C
~
e > sx $
C
$c
'.25 @ !
o V E C a M 4 3 8 -D o a & O ~$ m O .
a a ~ o m < < g 3 .
E 1 R E @ .5
- C b b a 5 d' .9 S ~
o - E ~
'C
@ m @
- a. O ~
8 p $C W Sr .
c O
=
$ to
- k O
e 5
e h
3 W s a
- 2 /
m a 3 .
C D w x <
O&z . - . , ,
l l
- 1
- o. '
e o f f
t a
S s
l l
i n n 8 i o o 8 -
t a s i g
t Y t n
e r
o R e
F s p 3
r e e s t
o o r R t r s f P o d
l a p n m s n u e i o
a r n a n R t
a g
o n t o
r i s
r a l a n P t e i u e y a p m n e s
it l
i c P a e n r b
ia i
l S A P le b R u .
E P . . . .
- o. D N
e
O L E LA C
- L CN AI A CNT I
HS N CI S OE S T A e N E
r HL O u
I UE T SSC t SSE EE P c RVS u P N N
. M I O
r A I R T A t
G CI ERK S O R PES P STA I t Y NRT n
I T C I
W e L E
_O I
B Y N ET K mL e
A I
E R
DVS I
OITA CCT g E D
A N
a N GITK O
n N CS a
S M T N
I N
E M K E S GA AT N
A M
O -
{< .
0 s
n s o e i
c t
a n O .
c a r
i i
f e o
l l
a T u
Q g
- i n .
e -
s e
t a
n i e i t
r e
i l
t t e i v p O
i u S .
t c
O s s A t n s e
e e e i a
m n l n d m i t
v i
r a
r g i o i a C o i p i t
c t i
e o t t u r P
r a S E q A C y
t n
t s M E C w it a S T S e l
i b e C A U I
P N i
a s l
R e e r
E P . . . . .
ODN i l
O t xi e e n a i
s L
h t s
mra l e
c 6 f
t d e a wanu ud i I l l e c t s S s e o i
f u b I
uh cN o i c
ot s i n e r n e n n oh o o nii st j s e) n an a a r
gn o r
t i
m l t
oi uED i
U l
t a e m4 c g d h f o
o0 n N t
on h r i s
an l
g n of oo n s w n o e o c u e n mni i i i
t r u e s v p or t c e i
P d
n mia t t
c n i
d R a 10 cn p u e r r o s s
sf e t t l
c s ful d e e i pa e on p o r e o egsdaf i
c o r
h g
o t r i C s r S t or a an a mla P o S s
i s
n s n l s ,
l an o
k r 7 o8 f o w91 e a ainog i
r -
e ei t I a a t t l
n i sd aw i t n6 P R rya t ei r mr 8 et r a ke aS l
i t
a t e o v s h t e9 n1 En u n "t S s pe d
t oi r a ar o7 s e gl e bh e n J p8 t s r. r d 8t bimd n r e9 S e elai f n
o a l n mn oin t
e -
ee n avt actt te h or 1
C i l a s o i i t
c si e
p t
s( t t eior pe e ms y e l
u a mteNi c pl sd M s z s u t
a n l l u a - i r
c my r n e a
i aies nr e t
- g l o
r e
e g C n si t o t r ait f "
P S t
u i i s s dS s pS aS r y
s f ah t LS uL S deCS i
l i
t le s i r mr e g T y ,
h g NCd NC PCInPCRC b t t .i i ia n l
l e a i
n b el c ogeo l
h -
R S t
e r el g Plat s e r i
_ E t s C v H . . .
oN D a C
~
xi m e e e n t t d n h s l i
b t e l wan ud o a i i f
r n s u d .
r e o n d e uh u n t o t. s o u n e n n nii s o r
i r
e n d o gn ). o h f n g y oi i t a u d m4 ac g s o a e r oC on r h e r
h r l i u t g nof noo t S
n c d i
p no'mni or i
t t
u r
S C t e
d mia t f s C a o
i P n 0 tc p -
r a 1 n c r n d l
c s ful d i o g e i pa e m s n t
e o e r
ai f l a sd i e i e
S t gsd i e u b or a an a r el s s s s t pts i a
mev i l s n s n , t r
e a ainog masit o p
t l
sd a w i h s e i
n i l ei r e
t oS tnF eAl r
et ra b S xS a a te o v ks t
s pe d C e si c S s er glr de eC d ap to9 t f t s n oeht 8 s eel ai f e ct r9 i
n o
C a l i
n mn a oin mmc u rud u u s a 1
s si o t ,
c -
s e1 e
oce noeh t .
ct ms p y s( t t u a l l i
a n l l
t s z s u e d p c gf i t r
p r
g o
I l
e a aies nr e s a s al ooA s -
r e c C n st t :
h ah oe w P t a s nt a e y
S s f u :
i ah i
t L mL hc s v t it l s i r mr e g .i T y ,
h g N oN i
ib a
l e
n t n
l b el t
c i
l h
P fr P t ear e oge i
o i
le a R S t
e r el g Plat s e r i E
D t
s a
C v H . . .
N C
O .
- 30 I
+ I le k r b bai s - te o.
MvD o r
c p d
0mr 2 e a %oCm R H s r d
l te e Dr
' e i
v F An o
e C v
a
+
~
W -
t s ro r r L ut B le e
, B a r d bf i e e i
' n 0al ip 5 n e 6 r o
T G
-am 4 0V A g
n r >
e ip c u
p d ,.
a q n s
.O M
r T a
r r
o r g .I a ie f e n ec t e s f
i n n i po*
r r c a e lp i
t o
a c
S n'<
a SF f~ o F
a 5'
W b
- r o )e L p v d or ie r
a e
r m
-ir g
Y>
X d
F _
4 ec ixi e FMR (
c P B
d i e .
-_ 0 f le t e N 2 n b ng o im oe ud '
C GMW S
mS g ne c teC ipo r r ysC -
I SF Sn i r
te a
W 5o -
O
h
-'v 7 - - - - - ,
,r
- .N * #[ h , f- ,t . , 1 e wy h ,, .
-f sr
- - p.
g ;l / '. .< ; 19 p, -, ,a g,- ,~ , . ,a,,o. . , ,
4, s ,
> > . . . . t a r;.- =
m g:;. w , , ,,~
w
' - ~ ' NM '
y ;. .
r . ,- n. - e., . #.*,e _
, .. g I, ~ k
)[ . e .
\ / pl n 1 / N ,t .i:
Q '.yp,,,,,,,.. e - _p, ,
.'( ,
,. l' b
~k ; .,
I w h- b k.. . , . . ,'8 %7
- f t ,, ,
,[}(
~ ..
[ '
o_a .'.[m l*. . ._';, - ,' '
i h', .
M.. ~M,._.C3lP8ha. '.'.' ;ggg,7, j ,s . l'
, - . , _ . . - .* * * . %s-a m .1 i
%Y~* p e f & s. . , , "' j'9 .}! .G[lo.
(( L ,v gem s or* * - o #
' ^ i c:) 4
.. ,l y ' 'l ~
- J. Q &]. -
? l D; .Q, $?N , .%L-2,. .w. -.- & G%22[L '"&*Lw. '<yj*l[ OMQ#,e,yb.$ 1 a w .v.
s.g m n, *pQ. ..;>! fp Qq,
~. i. m m~ ,,,~_-w_.'"
-n_
t
- w-__
- W y.;...,,
g 4. m.43
, p a ,%y y z.
Ltp f,;
- f.%m .py. u.4
- 4. J ' =d q n:{
l
'G7;khhkkhh ( M g. E
^
.1
> kNhke$- m md,n p :, k p %.~, h ia l y n m,.,. em Wy hd@w.ewn (A.w ,,,g' :.'.<.;,, <vss.,<s@s.;, m-s#fr d*fW..M:3S AK an sw M; nclgt@$p:l' Me:f %m d h yyn J;pR.Q7,y g u pg L. , s
.ny y ~
C ,5 e, n W' ;m.-Mg,%u:*L'""" s ,va f 92'Crheswi*p'il G9dk;hq]vida.h - -
w ON4MN w%mq:p9 Q4 @,71 M %. w
.m:
-- .f.# et @m md!re g w '.3p
~, ,1;f W L./ M4W% .J W C M
ww w.~ ": i @,m .,r:-
- . >c.u ;,g g s rCgM
.= a .,
cf',f,C
...i .;
ti- w y:?.
{
-d Q:w M$iMQ 4k WRl7j Wp% y I bQ '*,,Z^ %, q'G [g,,,g, s . e y J M M.. @.. sy: hM
%a % m ; f -
,a g
.gn h 7 %.'Erl>i *===
= w.. . , *
?
-x y W ,c . ,..
a,__
, ,,A. 4 , @a g
d, 6,.. *,.]
8m lz2h h kh[;pl,N, ww aa I
' * ==.a h*-M g.Sai e ~
m
- s sg ,~
EM___ *
~
[d/p 5 ?
a:, ,
f@k~.7 i9t,,b, $jj , *
't
@We@ =wM.Yi.H eg " a A.f
~ . . -
-- n>~ -
f m.4;G-- -M ' - - .m-Wiw C) u w.w . " --
,,.n
":i M['s N b g.mw2M= g "MM @%
k@p@hs( C~4NM".$hs ..@%v# 2 M Q
w6 N
Gi J-m{ 9 .MM, yyL@
w 3m g1 n
-w m -w--,..
- ~ ~ + ~
% w, g$
n"JW~* @
i m -.
m a-i
- .a%
p' %m l
\,j m V s .~?+. m=.,,_
4: r&a.. w"M-e3 m .
TOWTQ .%w?9h,hf( 0ZWu@[ %is 'P:'** @ -rL' - ,R ,. ,
-~
F Y ( w[{'fhl2 --! , . ,, Y,q l
m- A 17 n+Y .-
w m .,,,=,p s..e. , . _ --m - n..n.
9 g- A -- _
+., .-
hr 'rph N Y '* " *
- d ,%
>dt % --~I.8a=.'-- k . 7 PT WM.u &.-- . _M T%
_q N,
W w %.mmy qvtg6W .= ;.7.J w ;pws";rmn, . . ..,,c.6, .
h kYN h g (g Mgy m$h -- Rh og y&
~
. L b y i -n Q
w .~.-r.m 4?
%y we&,;W,,,,;a.n?) e .s'5 # ,,aeG.,.me _ :. g E+ S. 9 - u . L p - - -
( .
. - ~ _ .
hh k W i -
lL
+ . ,; f m. i ',' x . z., : =. m,lC e m,Aw-a .%; w. ,1.,,
w p' ,x%
.,c m._ mw.:A -
~ :. .===e -
}- ,y v
- i l
Q w
Q<
ru -. h. -4.a k . .- r. 5 w%w -
s e
k % #4MIa'*@, D -- - #~'*- - , =ier .,,,- --s v.yG3pgi~ e q ~'q.gd" p,
y 4 .
p . . - :~ z . . 9 ,
,. y' ** N.hik ,. h.d ~ '
.EN..nk. .
~ - w.. A w
'b a (h w i h' f 7
Ja. r
.W e@J.p'[g.w.<.[., _"" DMDg"2gp
. *:~.p.r,eq
~
h@4.yrg+ M1 em tN
$sw$ pdcF -; r.,qs= - , , ~' %"n ,' ~
_'g-QOM mn , qw .,~
k&** p. A w .
w em m-w o v - +
o ,
m~
e c enw.sr,M , + w a.w n e- .,-.- y.w .- .
> mm ,4.m r., .
wis,rt Nw.h.,,c%., . ?! y f og f
ks y usykSD[Qh $pyK yst,&lY.k.p%f PMC.2MdL6 Nh~ f?h ikid ~ 'h'4%
,.o ,: % A.g"Qyf.x f.
YN n.WtEa.&@k.Whh.
'f.hnkm+L"EQMA%2t2~6tb&%r-aMk a h .a )q iempy.M p$y weQ_.),
faM,bM, d.2 a m M"i . N @ ~ h %a 4
_ . - _ - _ _ _ - _ - _ _ _ _ _ _ _ - _ - _ _ _ -__ __ ___ A
.! f 4 '- [ } .
i
-...e,. V .J. '
.,.},, ,T s<;, - . " , - ,j F 7 3 "W N ' "T ~ " ~" ~ " " ' [Y ' ' AN . ,T
' , " " R T W C,P " F "" N
$ ^-
V f j', k,, ;
~
- 4 $ .
I i <
?
~
. J,,e&$'g j $
a d a c e n e s> t % # esid @ k' h.$ '
o i ALg& N WW
- h. W. .w%if: M=. -
M..., m~ :
ff.;l%Qs .
5: :'?...W.+ .W. 2,. T,.W, .4. ye-
'%V M a. n_ ... s .~ u A 'N :r ;-,~
w w-@g,+ r [~.
igZd k:
- : : -. p
, x W . .,.., ...
- e gy,J.
,.e
- %- % u '
y w .
w= -.. 3~ .Qin- ; .. - , k
>!MEFAw1 yW,.,AT'_
p._
_ _ ,3 - - _ ___ __,--T
,1,.e- -
ww = mw _ _ m
'l , - tw:
, . Y$$
% v w
I1 .;.y.h_2' L ?' - e I s- a j _ . * ,- -
w ~
i-A --- --- ~
,, .r' , ' , ,md
.y' nw w'r-. u p.x ~,- .
) *h g As .- a c i
- .. .._ _ _- g%.- - -- . ..
O T""T~m E
.=
^,..., *
- . 76,)$ .
- p. n' s, _
a#,
h,, -
- . . - .. t W 1. m . .n
, ft .
4
. r. - .
~,:-%
. .y m- - -.
- c1 __
. 'bor:#.
{%^
MS
- O . ,
[
l 1
k *'
,M ,hg g g ' ,,.,
<
- y'g[, W.
,% J. geg - J p a g
- u;, b.f.j p i s
( 9';
- R).k,t? @';,, ',
j ', ~,l;: i sf. 5 y & ; -n.
, w ,
- ?. ", 'r , l
? @bM((' D ,[ j;IUS d'Nd[;' - h N T[ N P,' i, "
- Q['%".' Y[f 5 ,
w
%.)W.$s.
.MM 3ki
$R %. .NW #
$ M.SGO.,N. M S. .'$ ~i -
m' G 1,,, s ! 'm. 9, n . .
y.K',;,b..wl s n, 7, 'a
. m m .
'y.*
,rW Y . ; , p ' %l[ ;.- . -
s .o .%,,,.. ,
V,. l:.L <'*-
^~ 4 1 ' :'o, ;; C.
c.-t* * , f f 9 c ' q.
'& . ', E -
Y.Idj_* N *j h'.y.)
7 &;gi ' f,*df'y. ' ' . :.- g .3.R.N * !,.
. M h,,'Y 4
,y p tf;/s a f ' .f * , 5
.s x :.;a. M g' -
n*
x : L w l'h,1.. h ! l_.y;;$."
.+ e
. hN[p .'d '*M d f. 4 ,h M b$ M .y y .5;N, M
[p j
w'S n
,r x .-
.k.
+ +'M* ., ~$v .n, a p.
.ak. : .u,!,'I
. . .+a
. . .np : n:8hd+i
>h7 q .u,.;.%m% .w .wh~s
- 3.r
- n. ., L ,
wJNy)g
, ~n .
k.,, l& A .m$G]&y$f,. : ?.:- w f %m, Q &? f Q
=J V -
lu%.&, +s.yjgQy&p.V p e,). %, ,' , ,
hT.?.:.
a utgA A $ m w r g .,g w,s c 3 f,., c m- &$gl y, gwg(m$r?m-s ~.d kun,h%%ng qpkfWWyy.n pa@rt-q#pw.;MW W r%m m
@ m5 ? 9.n W
& yW $,
,Mn4- Ue % ?.q w ait 29 ,n m$ l AN fff 'bSP'.T;w'l W V Gl a. 4 g y.DN d)'#M$7,N@ h t
.c.
%$k ibph {. M Jgg Nh- @@ 1 w..ws
'yp[hae 2 %g[y,M t ngn%, *T w wA-: k.g e nu wp w wm.ndw4 uwm . .
-. er .wmm QQMa@ .+ ,. -
a( Yg g sQ $, a %w, m' jig R . y. w_;Q\Q.y.n
- q. .
g1V e f.Gz _; -fQ,@ y k4.nh ~n' AYDEMh.$,f h_,dTih-h
- a. -. A.a~. ~
k 1
$^fM ,,,,,,,.g ,%
J.8 k>) \ -
- [N.
%P v; .
5
- w.<,p Wih'Q
- '
E 14 w W n) -N
'? A n e
? W '* O $
[h,.
gry;y ye G w _ . " _ " . _ , , ,; 6 '**
r*V.Rb t W -
.c *~~-~~ v i n - , - / q.gp&
w ,a my p by. o >u . ,
4
' '. A yy q.
.o n e M.
Ep,.j, y ,,
J y @Mw.v
. .W
-wp*. h.. mkW - w. g. gw Ms.3 . /8@- - , ,. .
.e x@@%.. ,,..o. f%u#. -iffn..w w k .94# . . -n-c ,9%$xp@n%
% L. "
yLv4 a CMM w 5
.w~,. -
p wg --' o__ - WF mbe'Q w
.'b.w!!Q F - *p,N/ %
ht%
f.hmQt[;
li .'.'."*%.
.O
. . n, ; v. ..w*
' * * . d,',..
's. :a .
- ' , s* *ms' .,.,.m%'&e
- .~. .. 1 M~
y
., -a M - e.
e
~
%.~
>3" N m, #. .,,[
%NT , ,a 'yO f W % .' <
- J*6
% .. %,b 4 * /ps.**., d#
p' = b y
'd
9.d' e.
- .b "f
+4* %: '.0* .k N.hf h [" [ k' W.e 1 <
4,c ym. & e ,
y v -e. s '.a%v
- ~ ' o, v f 1y 4
g p -* . e h
43m hh fJMp '
s- *].
,m **WN k b .g 'h ?r k, w
. . .t; -
'M '
. ' y, . ,,'P / - ( , , . " . .;7
.. .L .
9 - . =
LO' Y
, @WL e 1; ?p'Yh j s Y: ,. ..
m p. a 3
- . g,' - .
., .<.i..m
- 53i g M gg. .1
.I Wg J
i*%@.d7
~*
%+f
", 3 dg '
.4/
,4 = . p ';g
.,, .'. g. '
'.4.m k
%, js h f.. , v5. ,s.Y[" .'s ge
- O
- h
\
q .u g?
M ~w ,Y.. g[
. ~
_ . e ,
k', '
- n. . .
. w. .
w"f, m '/
w[4 v3 , r h,
,.9 f h . "* .
a 4 '9g d
4 W
.. .a ..-
- * * g-g ,,,4;
. r ~"
r #y p&q@w%gK g m ou ws] .
-J- y 'f s cw.
g, h b'.
pyr P m 'e c,a, r e .
.. w .t g@.WO' ,<& .
- f. "a ,, - '.
($w r
d6,f g g E/A K,b ; .&$4., l w
]Wg gs l
q sw
~ ?- 41} ' , ,.
.; g...-- .,m.-
., m --c., .y. --
,o , p . . _
- . rs , . , , - - - _ . . , , ,g ' .
ii if gm i di I I
stM %ih wMtw w&im#h @g,,:. ,, .,iNR4VVkl Y
- i$W. , kiE@r?#Ph@, .., ,p.4'.s&mmy gj "~ ff-
}
m: ww :9r p or D'y,i t) . .o y
'~
.Q
.y y' #yo } W; es J
- ~ y*' w
y . , .,
. .k;g
~
g' yp
.c w1 g 4, ..
3
. 11
. i u -
w.
g.>.-
.. y^
y . . .
.J ~ < .t* ... g' l e -*+.- n i
, .. w i
- I" O
.. ip .
,.g%x-
- q b P I .e g, g, nmmw~ . ,. ,,s<.,. ,.n ~=- .
s-O u.~,
a ,. .,
387.88,,M.,.2, et . . :~ .
y m. #..o.r, i;M w ..
g..
.~4 ;- _ _
' */,,. ~;b. ? ,,
$ My,
-~
2f 2, ,.
N . ,,,
I'$9,' , !?%? ,.'&,f N,- }h ,y -
. 3, J. .
- i.,~'.1. 4 n,cC.s% n%.,lt,}
.e . =
~. 4.L e ; .o . , .
"g,",
~'
h
..%+ . .k. .- f-.'- $a*. p. "h.*
y.
~ r. . 9. .
, p h%- % [ 1. , ';;j,% Q M ' !
'Sb . > :;*q y d i k ':
'. N L )(.$.% K S z.k s W ~ %
v l ?.
gf?G,-?,'y%@
},; . .. . .a a'W'-i < ' *,
' '$ ,* n., '
- f. ,
s ,=.
[
e h 4, ~
.,. M
. i' l D ,..
,.- ...,h;,p[ ~w' .i ~
A. 'lw%fe if WV4SI y.r. , m , O
J 7 ,
, ( '
- i J
f 4hhE = (, .' ,
^
g h.'V j% fgj, hk ? ' u.o2__N ',,
#a,
. g -
"r ww 1 Dl_ g,- e f.' gg l' .
y ., p+3
+
ma. ,. W4? <
-d V
e # 4 .I f
,my g.%
s (' g.
,4 5
%.gwg
. D'la.A}p%yes%
+ $ ._
;s . a k s* yff.a. g. ,
,!n ofg>
a s.
&wny: ,
Xhin9fJeun, >iG
-umut '^lNf :@
>m Q. l
*W. w-96
.ngry.
% p .', s >
w
?.
/ ~..
durg%w.iryY g;f;& y
"_ 4, : sa w Aw, _., 6 n; p ;; * -
u h*/ I 1 ,,m 1 u - r 4 s., : a +
' A% ,e
' ' p/ ** . ..
;g
(_;
- h N 'U{
a
; 4-:c:w i y.d w,3
%' cM,r^%** .. N.b.,
s .II,D a i WN Ql , ' * ? ' '$' ' %j. ' v D$ m k%$f w Ti@?,: D % ,. 4'l
...,.,K
- l. * %'
4 h' ?$e $.$, yylp'.,)
.KQ
" ' 3 L%e,.y e
- a. c. ww
. ~ .
. kNgk";;;' -4 t.
-we.ww .- . !. h.
s rg Y.hY. m ty%p
- S !
QMka's. gi ( p $$ 9 h ,. x ( %.Y.
-. !!}
= . .- E5'C3h
~- I
' hM% I
!f i.!
EY M'
#rE2MMb.c pp@li!E., %,
w . e ,.y -- e% u,n,, o -
- w m e , g k. ,:m, i
* ~. M , , .
WYSI !!j t ~_% - , l' -
- ;j F+ m:n f:5-%>
;;i r"- 1 n
O 3 .. gg ,, - . 3' R I i 7
, e "W%(t,.
af,r 3 , .. e e
%p?A g b+;
?a :~
a l d Y * .. $,
. Qyf #'s#g g y,, p -
. +s V
3 b p,e.
~ ~r ;
C , i e e ' ? t !. . /C a
.h'
- p. .
a
> , .I 1 . .
'V
$ 5 - , "
_ , vg V
.'7 4 ') d',p, F' .- ' -
T1T ,g~-'g,g -
.k *$ * %.'5 '
f 4, *b h ,'g4P'Vg h. w w k, m:1m'mpw &. T . b; s N w w' Nwm m gh:e*m n. p,y ekm m m:c N W m . wpa , w- am - .aw;m.Mw.p.n@n,g.pyw=m',n.i
,3m ges4 h g e,g,(u.,4 O 'kh3 h h,kf $tg h hh hj w f9 L
, h 's. ; 4g g (y q y$yy(.(
ia;q,4 w w
@g,.
M,. ,.,.. ,
- t16,.n@Q((<g$3 w$ e n~~. A L. % D ,. i
~
v .. :
'ef9 0
~.
g Q*. * *
, n~ .
- c' *C m[#'
I't", 7 t{.k1 W,tJ 3.M ' ~ s
~ ~ c
- n. ' FT,V
%.;' , J ,!E,d , *L' w?.g. <
v sq QCQ[ d a n. m .
- 4. . 4~ , ., 2, %.... E
.e
%D.
' n.
r:.A5 (' ~l.n,,?g . DCWE - rW W@, ,t ,%
$ap
&1 W) g,fel.WMlyk % p a~y.s # 4R y m@oys E fth aj m.
j
- am g
- [w*.r.,,m- 7.iCT nn c.
, g'*"w n)~*fs
m s u&
,bVam W3 +w n, p.) iM JK,; L.% ** % .,:,s. y'
'** + ~
y
'^
eQ&S 94.m wasp
- Didish.!
SIM$bEfId w Nfc [>;g$- P ':5'Ryg,
. wuw.Mu, P. g(
Myd@ $ d.n.Mifg0.76 ; :=: m...' .
,. y-
- m. n ;e p W:.1, I 5, % . 'j- i'ie-;r.%..2,,,t Y fc.n
>,,s a .
.. e
.,g. ;; s.aia.;jis ei fWA x - m_ 5.,[$7?
h,M, r:- n: + W$M $Wsf W'MsL*Q , ~ ~
' I:
e M'"l f?C ~ * !=: mC r% - O 1i
?
.M 'M E ODA ~P . h' 'M~f
$li b . ~"
~<, 'J ??
W vv h~ nw ..
'#n ,. .
$36~ ,UM,i". .--
"<*'hp_ h "
3D,* ,, M
. ~
rs. WyJ 4 *- 4 4 sis t muluSJ;)
'; pg;'tEfg.b. ' "'* y' by[7 m ::
QfA *:e'.', ;;~2 .; y; D e ?J S :: - < um , y" vgj F,,,,y.s~,cYM't ny y
/q- .e, :.: a e !:
m.e p .. . -
-. . , c ..
W ki , terh
" [h i ***' *! #
' " ' 3 t e f*
, \ -W *:, k *> ., .
v
"'i , %sg:
- l:
4 f, ? - w.wd.jh. Th*r ., r-i*i My ;.; Mk y r( ' tqp m -4. ,. 31 1 te--wdc.g :: K ,,,,'
" i e- -y .
9 ... :: h$5 N. ~.
- a h. h.
W';; D.4W. *j h _
- ~.
km4
; *' l yt r gd- _1 p :.:
~
[Tl ,
.- 3.
jy g t' ws w.i;grf,y _ w 3E y [ f n.n{ k r 8 ..< , S
- &h*'
ces*Q..,s. 4
.e ,::..
- - w. n m. m..< .
x,- -
,j $8 O '
,,A . $*I..=*.
13 t i;, .1 cM
#L ,1
,pl?5i f I y __ - u..
4-o,l** s w% = usa. a.
.et. k
! * k
&. ._ e, Pf$(ej a ;-
F{.b:.* ' ay?.. ' 9 4 g & 4G~
++. l C _
l t
(. i$fk'TkhkhYbkh Nf3w'6'? y ? b h ON .U.);i,$'6' l$ wahrn x4 k u! :;I h k'. M p ngg mpes % w.
- 33. .. h hk? n +N$ h, Y Y hlhlhifhh?hbhhk b5N $.~ if$ #$bh{ yn n m y#'
Esh w w -- m
~ rD , am":4" :: om
- ~
Qy3 "l'- } , - [: a C-;g2,, .
- w.
y Q :i;.;,_' ~ . m%.cwe-$Ihi f_ fF , p' J4'e nt tf,;f s A- r ce l cy..,. =:.g .-..vkgf,gw.f - li! il! 3.f/fd, cp:'
- w. 2m e, n y .q = ~.
v$ , {' f,q5 f%@, .,jg-el'(. .x:.:Q:.. j* 9;.
's *s ;{ h f <@.
%'Od 17.J ~ t j'$ ij k 6g$dy,r,ll
' E $%",. ' %
fr.m : . ys-n ; eV h
. ..g '; ;
M ~. Lg,n%. .
- . ji * $y.@g$j j 9*
k:.
&s bwin ed up w * - %
n 6l&g e.WP $g" !! E
~
e d r N 29 37 m,. sm e M a [. 1-4 IhdIj
.m.,
.e .
,, -y ..
t .., x g s.. h, '%
.y a m x (e9w e
w :, r- x . p sag lW - J ll
.* * ,, . CLY **4;
.EM c- . ii
- 2. ,n,g
, w - . A w
n -
~:
N
'.f -
5 :: { e-L p- 7. . 71
-e.:Pi
#1'
*f '
., g, g.
7 gana a:. -. y.
- v rn m
.ng n; @.,,r.4 4m I
.w.g. 3, ,e bi _.
a N
, u
~ '
. in. .
W;.c W ' em , l
= .
. .w
**C
., es 9 k!b M' 'M
;";,_ > -- - .r s l s.
V I l t l
l. i
% y& .w .......:,.7..;
. :. x..: :..:.,,,... 1
- =; .= = m ',-- .
'b ,hhf$5 hhh
. ~ . 51
@f & pnpr" ?Wm'N,h.hf pMgMpv kkUY
$m fh hkh 'Y A $'s" p,hhfl h?sygpFNEEEfMyM;k$l, n @gMV q$ M NfM w uMiDw@g $!N d,f 9' W %; % %w$$
a ?V ' W .7a w^M .
- ,v m- M D ii d waE M %
yn W r WR !! W.TNPf 'M,\
- ,Of : ' .
r... l G u. - - w sw
.; ~~
5 .g%;.g bA g%h;,, s . . .
? b.a.I A'
- g. (r',5 Uhr h I!l N 'N ~I a v n .e e.e. -
#. ..n . to iii W w w.m l e,m_eny + s.
- y. ..n n= e:w nn ;
w L ' +x m.e s
. .- m.a. w , .
esn -m .m. :: i .4 em. 3 < Yl .., 5h hYb
~
.'- h ? h u# +,a 2 na gm . ' r,. . :,c u
kggb n
. x.,
wii! mbi{hf$ w.;7egy aw Eu
%.p b vbbf 7Y , r h
@ y' w ~.
.w e kg z, m9,.9, , ,w . ,.o .., =.. g-@,mde.~
. w . .. 7
~y
$ d .
l l ws a=:ie . 4v s4rg pr.mu mgg< 2 ,u r ea paa a
~
~
... - qT s y#p g E::
i ei - g L 0 _ esti$ 9 ['}
*l~ -
F if I.'. ( p .- y. .. u,
. E L
._ . 4m
.. .,e,},x . . . . .
. A r.
' $ ?;!..
c f Q.h- > t:A -
.. g 1. 3..
$ * #'; Yf 2 - t f
'n ,, ,. .d gummmenss
. " fM.-.(
f'S?/ @ p..{ J g3 .- O 4.3 ' , ,..- v Q k@%y?i;' %Qp'
.; 91;,.- W . M '"f
- =*-="*E' #
M y
' 3 .. * %f ,
- g. w ,.~ .
. +;
n eg. y a g eM" pm6l m;,i. }tj
,e [ g gi j [- 4C1E3.i ?"Mf.
m
. . ~ . ,.
4 /,, , o .
O t u c
)
d s t u r s g W~ d( ed uy B 0 4 0 1 2 -3 4 6 1 1 2 g ;n_[}{f l o e it e - ~ - - r p K l
- . ~ _ -
c - i F i -
,.p mo_ r
" 7 y- 4* 6 w M .
> 4Q Al o e r a :4 C v n m
g - a l u , o W - C ~ . L 0 5 0 5 0
, 6 4 3 0 4 3 5 - F e >
4 r u e r
, t u H
z t c u r s , o ft t c u r s
%i 1
M r c d . i M o r c m, f e x d e o , a r . i , u e y s E q L a p ~ . a 0 5 0 0 5 0 e M 6 4 3 0 t r u 6 4 3 c , d l g, u r i e j ; t s o r F . y_ , i c M c yh ,j ~ h n
, gf i d ,._
n - e _ l x o e
-g; %
- i e n M 'e , s t S L. F y Ebh1 T a n g} : l e j a -
+
r o c. ' \ . s t b , u S r .f f a i C D
O - z H M ' 1 S a S ,
- I f C o C s
l e n i
. * , i g e n v A a :. I $e W l
t c a n a r i e . e I f d e tu Ri g e n e I i v o tL c , f e 8 7 6 5 4 3 2 1 0 9 8
- /
f - f5 4 4 4 4 4 4 4 4 4 3 3 E4 gEg9 *<y- [
0 o . l
.a
.i n G. i: -
?
f
.: ~
d u - t
.i
'p g d i
n n a ' o h t s L d i n o n o W o i f t a oS d l i F e i t is o _ ff i f i c P s nS B d d s a oC l l 3 e C i F n
.i d o t
- l a
e n i n o otC .a r u s a i i n o i V t a r a l a it t e m e i t r e a rin M e i t a n n n e e r u s M ai e r e f o V a r i o r ie e m a e .. _:
^
- ~
Vsd t h m f i s nr r u M w u o o us l a l
, e P ca r e i x
l m g ae ni P F V W F i i M n R n a d l i F e i CM A j i i l o F R M h e h t e d e i i i i
=:.
z e t l l i s i
= s l i
i a
.tiv V W P m - - - - - - - - -
- s a P F F F V r o 0 9 8 7 6 5 4 3 2 1 0 o N M R N 1 0 O O 0 O O 0 O O 0 ic PW co 3. $ mC; ; 3 c-i l O 1l l i( ll
O i f e a c . t t d d s e n r r u o e y o w a s e e e l d d r o m u n o s a C m s d E y a e l t t x M n s i S a u m A c i i f m r o t o n r f g o d n i s y e i o l f s t a e e a t s h e n u t r c e a n n s c s e s e o i e r p er r u e r b r O t ut c t u t s f o c u c u n u r u o t r t s t r m i s o s t t a or r c o r n e ci c m r
. i m i m o s m e m e f e* r n n d ru e n i
u i g io n i o r e p i - q e s n t s y n m e r r e i r c a e p u aah t e ru pr u p msn l l et at a c d n n t d c u l r e ru c e r r g i n ao r v p o o l e ot t r e i t t s os e d r d o r P C n o t y S e o r r a t s xt s e c ac e s S ei i _ l i ci Md _ i i b le N m S m W S ia _ i n _ l a _ R e S t C C E t s . . . . D a O N C
I !!! jl ! i l i , i c - t a n O m d i a -. m d t e M - a n n h s y r a e i b g m l o dt , b d r s en u u e pn n e c p v gi o gm o n o - i d o r nt c s p p i s e ei u o i t p up d t a a q ) c s s ye t n n i f n o l i t ni l al e l a i eX I i cen m u t a m n t s n u Q r e c t s r o a ro o i t d n s )8 D n ii ut c t s a o qe e8 n mS eS s p/ 7 o g eC r e y( n m DN i f t Co df s a i n eB i R y ul w
)
l ad c f i i a c o n o Ni t c nqwa ulel i T r nt a d l i i I t db Oi~. i f a c C n al i a e dt esi n Q( o g u l e me n )9 r t t -- n8foi l a er - Q u t e d e h n p2 s oip8 o/r m s (1P ebu m , o h : i i l t 8 r s Ct sb af Pe dS aI s ut R o4 e r r - t s e e E e t n psi - d n v S y i n Mor e nS Ce sE n/n I t m 2 o//iD I eT a TGiN v ddU )a i l s v S p - t I
/
E D i t C S p A m ei h i l s 0 9 p 4-mUEi e d r R c t nan 9 8 N ( A C dat o r m NefoU a o8 o/ ni8l l l i( 2 sN oVr Ve - m e J gt s e c o s n e e xe a pl c aar gi txt r g d : u c un init o o r e U s e A Cmdi r c adidi n t P r y C d a
- r e h v c
/
h t s dfe o r cpl eainWepmm r oe aph f ie it il E e i g o er r upi po t L t CppP b i l a M k c a e e r i l h qA wAC e s j g S b s R a . O e i E A T r H e . . . ON D ll , -
- i , l i! I l O l i
V s n e o i c i t d n an c o e i i f t p i l a p ac uf i A Ql a i . t e eu a do )9 uQ r dl e dr m/ 8 e c n a m (1 o a n or no d n
) cd oP s r t a
c sd d a n e i p nPe : s S t od hs ad ' T n T N t d r nn N eS
)
n o o 0A 0 eoa S eA i t ( c 3 8 bi nQ ma oP t At t Oi a c 2 8r
- eit ni nmeiw f
c Ae mmad o mN aian n i l a Q u t e Wtn I i gx m, i hCo por m e s f oWh i t nEcxE aT s r ee f t s e ) gi t oiE vr , b r eiet tt S y i t nw d ee t E mem wi - S I
/
I i i t v i t r ud e l h omcA uf aTr eR m m Ao E ch D c us l e ouU l N A t r i l nVS n ds1 - C ( sb nnno l Cs m a e r eu p o s o o s. ho T- a v r g d ( r r T n o o el l e G G G. pWWT eN a nSC i r t P y C iV r wd L n a it il b E en R a N P p a S i l e M - - - - - R S E A . . . ON D
O s d u s e t nt n d e t o o n a e c r p y p e m o n o n t c o r o n n de. d p i i t c t a e o e n cl a t e d n t r s i m c c n nP e a n e o a n d , o p i t t
~
m o w n m ( s a r r ot i l a e e s t s f r a f m d n n ei n h p e e o l l t i pd o i u c m m V go r b q n a i c e x n o r e,t s e d o m p t r c C e n oe i i t f l r d u e a n mfr e o s pc n dh t i r n ue t s pn a e p n t e o c p e u s mr od n i r s og t e p c c r y o r a n ed p rn h r f
)
n A f 3 o , o r i eaig a t ) t t s u p r e io i oh a d e1 t a u ei r n d Ot ic n l a0 n 3 c f s t l oat d r v i r i nd n i l f a a oG i f i l t n e rec agc).et , e u l i E a eh l t c x) Q t u mr t oa e i i aR en e . m V r q e r o faf /s a s,he ot e c e x dnNU e r f t u y r e si r w;t i i t s l s y i d a , b a qnt l aer f anne pi t l S I n e c 0-2 e e r or i cd cf i t o em, A. e c l oa pet n3 n I
/
n cet nt E p nc i ac i sE a8 o i D i e f i si e c r N p d iB t ae r c nl i i glpzi og , ( e op n t A uE r d i m a gI mfer r gs ps r xi x se ed d p a n e (e r - s sf ,o s n s e pse gi as g e r n n ar yEI S o r s an d eidisd e i z e sd P h ir i uihvt a s sDI / t y t v e h v u s t n e uNE i l t g oh r c u r a cowl r e oor opdGD i b n P wsPl f I n vP( Cty nSN b ia I le er R t E S . . . . . . . ON D
. .l
l1I! l ! , ] 1!!l
~
,d e n
) c a c
e a
- O t e-r o i z
s f r s ed g
,s f a t _
E r g h i nles n i Ato,c n p l a wd e l
) T u i
n m t l eb d E rt s i a r e s c mh eewe r l t
,s n t t s
i p s ga n gin o m u o s toisn e - o t pl a o i (eE t a m l a - i ( c sD r o r ) l I b eprl i e dN t s s l a L- / d z ct zi l or n s c r a t a l i V ht f o o l a c i t c el cnr v oim r e a m a or osf ,et s x mi r d e s u r p c f dd a c i d Eter e l p d os e n l n Di Nc c n )r ec n a t d e t el nw ad c a n e p r n aa n n d mlei i t eo me yi e e neh ns p hi t a r t alp s ed A t o c o/r s e i t p t s e l x pe l
)
n o t of i i f e r poh ex r up r w e us uw al i t d eau l h d S ser a O a _ i c n l b t o nicvot e a s de 0 "n , f i qt f l a a b aie f c r sm Q u l ddi xn g(1 o i t ( s d eu a ra l nneiga epsl mV aa ys nc n o s r d s ei l l t e s x psl ei x end n p t i l i t angepss aS rewms i eyrap i i y ei S d yl t i r l a nopd ,as I S n l ib u a i r l t c r i i mta u u su s,d r i I
/ e di as cid ci a e acs ci l
D E p en es ol t c xl undd epbi i auf o N p . r oni t i r u pal l r t s s ( l poa e A n o al yV ed r l p "n t l i m Vs n s t n a r - x e r s,s s s i t aX I s x nned e cngdinamp i . g i or S dnes a eiii reeer r i oinnege o P h enr uu u f i in el n pl l it y t g pi eqqql f slat ci pzl zi t pv e . l peh eeel u erapoou e i i b a n AD w R R R L- QSTA nbSd i le e r - R t . S E . . . . . . . . . ON D _ Ill !
O s e g d e gt n s se i n' n x r edi s i s e u
- t n r i i
t m au d d e r t al r i a s s mgnd d mn r t e oie a - r ep cint a y ot ot ea f n l r n e p a l r e pam i d (t in c l l o i f n s e C "s e i , i l i V c a n g yie n pe c o an t s i x b c n a e s r i ot a a de e w i d ct o t ct r m n s g )n i f f u n as e sA r n o p e r o r i s n o n t n ot e p df ot a r i i om c u dais at y at ed A l b c S d I y e n r i s d t o bl aif I r ee c c o a l
)
n n ei r m o ni t io a t aC u" e r p q u yac O ic t a l l qoe f d gr e i l i t mr r a p f V e t c n r ) u fo t nf on ila d oyeo r n pn i Q u x a ec a i n aain pia s a
~
t o n m i t t m d n ooictec r o t r i r a r t n - t e s n ecf r ad t eono i c pn/ei l ii y ~e r t u pe c d S I p s u peidiSmin aee r o mra S p s
./
I i odi ed r n y: n pid l E D A n oGa r n s n pab n oims o n N ,.c o rat s yi t mco i
- i
( t as n k e nn l t a s ac aoe/md y eot t m a r e n eTi t e s ni i f r oi c s at el g r e e muaim f o s mc ci t u r r u e e qu on n P r e l e Hi oimof l i y n p a mh qvlpd no u o0 R e(c o t md qc1 i l i b k iA mnanoce i a a I le R e . ODN E W e e . . .
; i
O ot t c d f e n e ad s n n o o i t s ct c s a r et ef f n n een t it n n ep o t e mpo mm pi i u c u qr y qeo t ef c d u n to a ac r e i t S mar o pe x d mta O i t r y e w i a c h h t nt g M et ui i n qml ni el i o n f r og i t e f t n c e osi z s ei a r G r t s s c at t e R f f eb c e n . eae l f m a I E eicd t ht hnd r n _ g r o ec e n _ P r e e d : t a y m a e at u po n v i l t p e i l _ ib i L t c aoi t ia u vl c _ l e k s j e e e vt e _ R q a b oe e E E T O Tdd OD
.N O
e h t e e r t nh a m o et l i p m g mle. tos t a u o n ed nd qr o go e e ef T e s m a d
'u j
mms p u U n s gt oie n o e n s qub op v i i r ey el l
's re s i g
i ed w w t n d . i n u ost ed n gS hi t mn e a a e n xly eitr u s e i r r o u s d taa c r ut e x y qc e e i s nMed s e n yt u n i r t s f e i h diod s o od n t y ct t u et a i w s e rash Td n l d a e al s s
%%% s e o .e h f 0 0 0 ot t r s e e: 1 1 1 on e gm Q '
et h d r e+++ s l I n n r zm t di n df d oo a na c o e oio s c Cy t 1
- r d
epi mndn e r s Ei r 9 e si auqe mae Mae 0 o l t n S n 4 t e e b n d o y e Al i sN nd mn r t r
,Et s yi oC d
u m y yo e e tu to i e e s c r vh t at c s mliTi nt fe o t l S x e r t nlpe ap pe a r et wf oe s e sf Ci miu s o oidi l i r i t a u ci e eqARD u t ct s q ed r h es e c a mM r n ndo e e ee , opeb an r o e y ao s s s e psf C gl r l aruidn e ac l . gi R n u u u tc i l l otc nh t r P ar ot r a i r E wPPP e epets i yte g l i ebMo e r c hn i l t i i n n t nh t aSl i l l h h ei t nff e c U ocAfo e at e i bt t TT iane le i s R m p x Ei u E . . . . gNE D q
O , i e ' Y f w OL EP A J i T m M 1 DS E I I VD T S, r Y S ^ N O I R T E V N C I E 9
%S C E
R N I C e I im N T O
~
S A ^ R ] i m T - T L - [\ U R E S V f L U P
,._O
O f e i n u f E m a n e . i o b e t c l g a r n a t e r a n e I h t s n d 5 0 e . n4 m a , b z p dH y c
~
i u e aM po n q yb r 52 e t u E kc dn 2 q e g e z o a r, r i s O n h ,t e f f s r s i w d s me uma f t o t ou o o ni adiedh b u ci e ra h e l f e v t h h S a m "2 i d d v d t t s n o r f y ww/wii e os d ww y 1 l d d e ol ndd sh oie s s S t u t u r aa r r l a an n r aht r c c k k c u n b aaodb t aa i x r S q s r w bb t clel i an o r r c c s s e o dd t - a l ei r c l mM a - r c r aaf ed h de eC u r l an r gi o t n s r e vd ar oo r R e cl giSuC t g nbb ot r c e x ei e s s c n2 22 tsi d ont aG e s n af P ar m u e ra i l I r yte t T it n l iI i r a PRT 0 b iane l t e p M t 2 3 6 - - - 1 - - - - - e m s R p x e Ei u E T . = . . ONE D q .
O i oa t el r a 1 g n
~
i a vF vAe
%%% 9%% %%% %%%
5 i i z s
- 5. i69 7 9 34 2. 51 175 75 24.8 840 e h 4 Dmt or 1 1 1
2 - 3 11 5 22 d n
%ffo r a
n i o t c t e e f d o t c ew pa CCC CCC CCC CCC f e yF l SSS SSS SSS SSS e de T s e n c e l u f i n O w a 888 999 222 333 111 111 222 222 s ri l F e c u dl r s e n a C r r y u e t d c g u d u n S t S d 146 246 146 245 ip g n s n ix n t r a a r a r hh a mM an T T C r o f gi otc r o : n P ar e o ye r it l t n g r e i s i bt I n ev i l u iane a se 222 222 222 222 c lemp h l uc n o R PeR Ei u C C ODNE q l
Frequency Spectrums of the.. Ultrasonic Inspection Systems Similarly Calibrated ..
, u _:
^ *
- 0.8 - Adjusted
's -
Gain { 06 - [\ . e -
\
E 04 - [ \ E ,1 \ . . e 0.2 - f \ 0.0 i i h I
. 0 1 2 3 4- ,
- u. Frequency, MHz Transfer Function of a Defect
'1.4 -
1.2 -- {1.0 3 g . 0.8 - c -
= , 0. s -
2 0.4 -
' ~
0.2 - I I I ' 0.0 - O 1 2 3 4 - Frequency, MHz Frequency Spectrums of the Return Pulse from the Two Systems . 1.0 - e 0 .8 - Discrepancy due to transfer i
$ function of defects
@ 06 c -
, ,\
j O.4 -
\
/
T
= 0.2
/ \
/ N 0.0 O 1 2 3 4 Frequency, MHz
Conclusions:
This illustrates how the defect transfer function interacts with'similarly calibrated equipment frequency spectrums to produce the
' discrepancies shown earlier in experimental results. ,
b
\l i l y n o
a i t r R o O n p t o -
~
o aym l cc d i t n e _ s et s e u u yg oqs n _ s a c ar nz e i B e f h i t oi s l e l h t cd d e t t i wen o d u pa m o od s b eni n n o s-c s n M aiet i c i t p l e t n b h e o l a c omt t c aco on e i t f e l a i f d c t p e me r ow i s y o D o bi ra t n nt o o f l f h g p
- i pe v c i
s ni s e l l l o i a i an ar miroicp m a d o d O W otaci t m n y d t d e r o e p d mle o d e e n yr et o s r o g i co s t s u n i P nf e e einh c a e g e r l e
- u st t n
l e u i a d q e o y e qi eh u a r g i n r t s d u g fr T o e n h a c m bl s e at o c t i s d e . c e t f f l a x ey d mnt a t c e n i r r i veo ne r o e o mM t a Wo re ot p yam r st i t y a t s f f e e v a i s n an r o eh s mp R u o n w e gi ot oI r i u o .i e m r c cT TS o f q n g t c c i P ar e u I
/ne o d a a d-yt dg :
e U Ti n p i t l a r f r o si n t f i n s p n f u l i I o o i w bt A D i i iane nc p eal r c S T l Rem p aa r r r u h t f mev oe s e Ei u TT P odd D . . . . D q O NE
'll ' _
, .--~__v, .c n w _,
1.2 ~ Q M dd 1.0 - l ..
-- Experiment LL i
\
i- '8 -
'\
4 3 0.6 - s.
= \
0.a -
\
t
. \
0.2 -
\
\
0 O e - 10 i Y 20 30 1 Distance from Corner, mm ' i 900 corner reflection measurement for shear-shear-shear wave i data through 20-mm-thick steel at 5 MHz ! m 1.0
- h. '
\
Model 8
) -- - xper ment 0.8 -
i iE N A g 0.6 -
/
R / e . 2
$ 0.4 -
c- f i g , 0.2 - I
\ i
/ l '
\/
O I ! l 0 50 100 150 200 Distance Along Block, mm Centerline beam pattern for 450 longitudinal wave transmission through a 133-mm-thick steel at 1 MHz O 1
-_-__.___----_-__-_-.-__a..__ - _ _ _ - _ _ _ _ _ _ . _ _ . - - . _ . - - . _ - _ _ _ _ . _ - . . . -_____.--x.-
N
, o ,a
# 0
_E. 2 uO N
. E E. N -
i m_co b N
- g cv C xW- _
E
-m Eom .ox
- o, e -b ~~DHC
* 'A M -
L-O o t
- C ;
- o l .8 e o toC YO "i OU O #
e vhe .o c)).E
- o _e
* - o A I 1 8
Et c) o op - ? e - C =: o - 1_
- e. o =
~
- O N O _ R e I E
=s C
e m
, , , , , , ,l .. i , o :- W N -
2 2 @ $ 8 @ k=
~ 8 e W N 2 m-(ap) soua;a;ag o; eartelag apnn!!dwy
~
c}Io2 o I e,, I o c) _0., -
*1$ jW
) c m x-9
'w- _ -
8 E e $_ $ co o a g
' O.
r - g -= - o, O b. u) o n _N_ _o c) '
- A o-e 52m 2
5 .E c S-m oN -
$. "to -
{u G N
$h e
t i i i i i I 5
.9 oo e e i
- 8 y.
.o
, -e e e , a o a 3 5 g . - -
o _ c. p
=g (SP) soua;ajag ol arrielag apnt!;dwy o{
*e .Do n3 o O 9 C. '5 J' 8 e z w0 =.
>= O N. _ e
~
c .=_ ES o a
~
? 'O g~
l i 1 I i I O N O e e e N D (SP) soua>a. jag cl ar::elay apnpidwy k (I i l
o
~
g l l l
~
/t -e j
E : I.I , ggE g
/
- e. /f . : : : E
^
;!! ,;; sai= ;g / I 9RS I -
.) /f I , \ .y -
oe-a = .j [- _ . g l-lll l \ _ , ; 9 ll.i l i
//
/
j
/ :
,e- f i
> 2
) .
\ li;t -
.i ff.f " l '
2 i
/ ~> -
~
2
/ *! :
/l .l i l 1= 2 =
ff I j I [ 5 E i 2
^
E I\ 'N('[(
\
s
. t I
L 2 E
~
'I $ '
.f
~
\ \I f -
'I l l 2 . i
)L.. % 2
,lI l r - 3 8
/
/ /.. C. . d.j 8
l llI l' g / g - ~ g I - E - E Ii f x
//
\
N s
*l E I y li '
l '
- g /y/
o ! . I o g
'S 9 y e o o o y e
@ 8 @ 8 S
.. ... ,,,e v . ., i . e ,,..,,,,,.....n.,,
3 - 5 -
. E o oy 1 I i E3 \ \3 E l: I
~gE \
K .E j ,- , f l 5
, 7
/ !E E j}
- I e
omEe- Ece . j _ m ; o iI l -
- I I i i- : : :
- I l
I
- ; 2 II i ! l [
. . p z , . . . . -
>e a,E
- C
\
I e i l
- 2 ej 2
- l l I O
* ,e g
3 . ; _ l*" j o l l - ~ g : t :
. / u
\,
. v -
/ o
/
a 7 f 4\ u i -
/ /
- u j .
/
~
s s
/ / $
\ 1. -
e
.=
l l 1 I a
' f O $ I lE
\
\
\
s 3
'\
3 I 5 S S l I 2 R e " 8-e e v eo ronmemy .an.i.e O 4
_-_ ; . _ - - _ - - . . - - - - - _ .-_ _ = _ _ _ __ 0 hm;
- o a 7,WL: l .J j,- - o-LI.
- E.E E E' E E -l l .- .S EEE E E I .# 1 -
o, , .v.m ~ ee ,E e Il
- I =
o. f
- L.
l-l l l 1 l a. . g .: I l-
!f!I I o I l=
l r f.J f /:. i
/ j- -
g l
, ll .
l t ./ - . w u
' .- /. .Ij- '. . s :
I - ( ,5
.lI f 1.. .. f. . ' ' , . E I. ~
*2- .
I ;; - I' I : l
/ */..) . -
o
).
\ . :
I :l \ I '.j' . .. . *
- ~ g
/ &
./ \ c; w
.7 /'
) l Uc i
//. go.
, , t =-
o o- o o o o cr -
- - ~ a o =-
i af eo .pniiicwv .an.i.,
. i O '
o l . i - o
. . 7 .
\ :
U. ( Ill - l .
. l =: l
- s
.- l , t. --
./ . I 1 - . .,
o P l I .:,-
. +
' a -
1 n l . o
/.-
8-t I g w 1:i)
\ sls.
a I. i l . . l 4 l - w 5 l . i - e2* sg.E. O l
,. I s -
I ;4 l
'. /
y~ .)
- - - I f .
1
- i ' *
**i l
5* w
.( / .. .I E
I
,c 0' h)
,8
- i I
,i 5 k
r
- \
.Q ,>
M
- g 4
I I i I "l. . a 3 ./ :
- ./ .... .
... . s. : ,
)
i
; . .- l o- . . :;
- i e l' l - ,: - ,
m I I c E eu c
- I '. ~ &
1 i le: < ' . . .,, ,, <> =. ;=~.. .~~. e
. .l .
; . g . . . .,..o~.. = 1 i
gl
.c : s l C .i
./
c1 .i\ l\n < y j 2 .
, .. . .i = 2 .i ,
3 $! E * $g E *l
' h /k[ .
lt lll 5
.) o i I g t o
o O C O C C I ~o o o o O
.C ~ n' o '
~ , 8 . '
I g .
,o .. ni,,o., ..n .,. . ,, ..oni ,o my ..,l.,.,
l
O' g n i t m t n s s e i v b e r g u m o i g s p r edn s d i u p n q n o a a e i h bs r r n t s e o o a p f i t s p a d a p c wt e s l ane t e t e i f i r l a f e m o l p e i r s p r v e ai w m l a t c qu o a - h c t n m t s e c g e r a n n o oe o r i i m i t wh t p pw t s e r e p c ef o e t u e ae i l t x s a e oerv f t l t n i t s d d n d i s r e n O h ni i s e a t r o om c ae b p i m r h d f t e et pi d t e d ah et e e h t r x e ag e dw l t t n - p a si s o d n m d n og e wt n a o y i l a an ab c n o i d t l f u v i d a i sd s s t n et a i s n i s y nh s y y c x e o c alu n i t r a e cB c a l a 'an l a e b - v ni n u dd a a t mM s e sE r g aq an r gi o h a i n m u os g n g ge r otc wl a i n i nf r P ar l e m m r 'e e i r l et i l e e r ye d ei x ct d c d t i l t t n s o t e a n a of e o ne u I i SM i bt M D m Mf a Mc ia e n t le m R p t a Ei u S . . . . . ODNE q ~ ' l
) ! !jl!il i g
~
l e O b - t a ' t p r - n e e e c t n m c e p a c. i u e r l l q a a e s s m l l t s . l aia n r r r e o ot e f m s f a e e l b em i r u g n ai n q a t ph t e r h e r y c c o c e . cf n g ay e r a e c u l aenu r q e e r v s t s q f a O l t r e ne er t r e h n i f mf r n a s ee e c s r t c s u n i a s m p qc e m e w e r r e t s o ef o t s ys y l y sf t d s d t s u a op dh i S t Ce c d ns x t c n ay ._ i r a E x a b c . mM t a s n Me sr s b d w n e a o u an r o n c Ae t a o r r q gi ot c o s t n e r a e r P ar i s t c e em r ar B N fr u yte f r t i n il bt i l c D e Cp ua an - - - i l e n _ Rem o e p - Ei u C . . O D q NE 9
, ll
O ~ d n d e o u C o p E m M s o S t c c A f e l e o t f e b u s c y o l t u d s n r e e o r u f t q n n e r e e f k h s r r o t e r t e d p n w n e n a d c o kr n f i a o t c o t n e w r o s I X n o O - i t a i n n o p e d h o i t c r i t i t c e e l a n s t i v o e s r t S k w l a r c e t o E i h l z s n p M t z i s d y e e r S r. o y u m A on l a S t i r R f r n o e fo l i x e p C
/ t a r d s t
a x G kr o e w e E e mM an k r e R o w mr u n r gi o o t e U et a t otc t n d v c l N u r P ar yte W p m t f a e s ar r gc u d n i l i t I n e r o r e r p f o bt ia e n u C D P U o C t lem R p u Ei u F . . . . . O D q NE i .
g o 3 S E I T I i g V I l l
~
l l T C CC S S A I P P I C S I P m ar . g o r P y t i l i b ia l e R _ E gN D e l l
O e s a t t a um S e c yo r A l . m l l f n f e i t ar e i r s e mp r d e s d e oe p o i v nt a o h t t l l t . a oe e e s t n r v m e . i t n o e r u a st d n i g o d u sd r e n a i q n - t c c a ah e i e e o r pi s b st p p e . s s dl s ot s s n f o nt e t v e r i i i i s y o s s u E ec e n e s r pe v mo h e l a yn ei t i t i n o e b e i r t a d e n c l i b t r d h ut r a A i e u e - qs at c p
- s s s -
en s,ar e a s u1 r o a R b re - eme ep c O ap l y9 0 ht h s n d a l t ar o a4 d i n e a n f u t a d t c mN r o e i f p o D i l a oe ns t hgan c ce ro b r Oe r ya au e n S f gC S. i l nr m f r E p P o I i uf p b oe i o s o u o h eS I l od U d oe l t sf r e e s d R I t r oCh u p h n e e o t s t o r u h nn eV
,i d zf y y t r P
l si t s m n l al t ei msa ec i CR i ad n e a a u i i S r e o Anb a r r g o o i s s S. a
.i t
c I P of t a c m P r R : p yU me mac l 1 1 y e a l a r p ad C t il l v c ne ei S l i i x o h py cT At h ot t I b i a C eU EC Tt P le R S I j b e E P Oh t . . . . OND
MD )S
%t e C A
m m 0a l D 5P % t _ / - aC 0 5 )
/s +
- )
SR E 4_ m - m eV A A f M I_ 0 (m e, r 0 (m uP S _ A ' _5e z j - 5 e z d e
/, 4 i
i
' S S ee ' i_
. t _
t C r / /_ t c s4 c .. oh f e B.
'I - e rT .
e ) f e
'.0D P( C I
1 A t 0 D E 1 i D C J L
%A k .C. '
MS 1 5 0 1 5 SI P c.5D. ' 5 e o 5 *.,eo c y
- oo A n t- oboO E hi M t S d A Cna C m '
=
e m e A r u s C)3 D d ', t c n ) 3 % e c ' e a o . 0 t o I f e o - ) r s ) mN r N 5 e E
- m P m d
r a oe a M t 0 (m l C 405 (m n a f t S 5 e S
/ e r a l P w z /
t A z l p el ( A A _ i S I P /j I i S p PP ( s '
/
#[
t c ( I t c r a nl l t 4_ e s B // e h u B r,r#t f t f s ol s e e l u / e h iC t R igI _10D s e 0D 1 t o
,, ; l e '
CS eI 1 L I C I' R E C o m 1 I A C 1 5 1 5 te P C S S S Dn i I P cO;:- o I P cO.~ E y O A O
O y C
~ A 2 3 l D
g - i n 0 1 t r o p e R e h l t e C v f e o A L D f 2 3 n I f
% O i
o 0 t t 2 u c s C n O ut e F a l aP C se a er - - A D 29 e I 73 th % 5 aT Rl l na C i oro A t f 2 3 I D. c % e ev l 0 t A 5 e e 1 5 O DL .
. uOO=
O .
]
l l O Amplitude of Response of Sharp and Smooth l Planar Defects' (Lower Crack Tip Response) for 10 to 25 mm in Height 0 100
^
1 5 I
- 5 !
j 5 E E 50 6 5 E E { E 35
,y eE c to 10
%,N . -
7go_ aai
. .i E.
0 (! .,,
-. ' ~N . . _ . _ .
e0 - 20 (xg 43 44 34 35 5S 5e i 20 - 45 _ 10 kb ' g ,- -
. g g
o /. ~
.E co
-25 -
5 o
-30 1 0 50 100 150 200 Position in Depth of the Defect, mm ,
i .O
O 3 o N - l e t a
/
l P ^ - r ' f o g - y~~ K X s ' e r u d e C _ l c A C o e D) l A
%sa
)
r - e Dl a P t t 0 o %i r y i 1 p t d m r o 0 s 5 u e ' m o P r E dn l t o 0 c - C 2we Mi O l e l l a EN S A ( e h M( S , s r S r a A l o / M f - l- - s / l t / u /
/
s - s) / l e l en r a
/
/
R a iue cd p
/
/
/
T eeo pcr .
/
i I R Sou r E f I R P( ' 1 1 1
- - - - - _ _ - - _ _ - _ - _ ~ - _ -
0 5 0 C 1 0 0 S I P g gCS ce $ E t 8 o E g g e eo % E~5.E_e .n O
-l u O d l t s u s e r-f d t c e e e f i i g n e e f r o e l g s d r n s e o i s i t c c L e s a N l w e h P e e h i t s i v t w
- n e r
t a e l d o l c la a i a n e t aty
- d i c t a cl i n
r n s e u i i e i s l g l db t b i s n a n a n o y v ib i r l i s e er o g d a a t h p n n n d o t i o u a e n n n g i s o r 1 1 t a e r oo r e s y 1 C i c e t n c i t d 1 S o w e e n l a C I s d t u S P s t cl s n e O n I P e a el a ed i s A l t a t y f e c pe eh d n
- l a c i v d e e y i
l p i a s d t n a R p t c e e a d u e h ht r oh e a o d j t we t i t c oe p b mtw f a a r e t u s d t a e a d r D t u ea e e u bl f p l b t r e t a a l l a n l ae t i v l i e I m n u t i t c e o o wh e 0 o e p m s si t b c e m s t t w s m 3 a b m o ae l u g a a s n s o i c l h r P s toib ei a c g y t o a s r t r r f h e b n s gl a o P y o h L L h t C l s o n e r p e S a p i i l t s N N n I l z ro r u o i i b P P P Cis S c A ia t le Rl l t a EC S . . . . .
- ONP Ds i l h
O n e s a u g e n ro s i yt b i e f c l e n - d i t n e a s e a o ar e t s c b u r u c m y r qe q t i i t r a o n an i t n g a n twhc i t e o l u s s rp i t oh g e t u el g e aiE aii R p n D n dl l
/
i ot uC e i t c oN 0 7 s sS I d o r sd t yP a pocne 9 4 el st i e C n - t i bh o R l i t i atc a Ci n a n t sl l n e v /bl i t s opd u C e l ei i t s a Gorr d sS l z a n E n e eP I z nid Rd an r o O r e c a e n U nu mna Ci n n t nilv a N oul p S n 2 - i e e n r t r r h e r s .P I o ht s v s o o i t k wepr ca _ i i nt i ph t f o i S. f t o pa t ai n ef r t aU k i ni e ouh opc pt ecic r f e ov T s h at si e ps n Ri si nk r t b alr et c e l u oio t R yl omc cih Sl a awI S nil c ond ppe e i e .ml st st i g a C a hnlaSfe yUS. s hei tcf l e t al d eo I l i h b C ol r t s n i a m vt ogo f dnat ectat at es e rm a r r d h n eateRt e g o r Pi n ts d hs n d reit ei n e t o eWsuh e P y i e a n n e i l i gi v ul adddBAT e i v o r i v l vt n b s o s r o o o r t i l l i s o u e r n r PDPEiPCP n i t b ia C c e re p - - - le S jd m R I b n o E P Ou c . . . . . . . ODN
O CEC OECD/NEA CSNI* JRC
'SC MANAGING BOARD
' CHAIRMAN: R. NICHOLS (DK)
- VICE-CHAIRMEN: S. BUSH (PNL USA),,
LONG TERM PROGRAMME' ADVISORY GROUP . G. DEUSTER (lZFP, FRG) B. WATK!NS (UKAEA. UK) MANAGER: S. CRUTZEN (JRC ISPRA, NDE LASS) J. MUSCARA (NRC, USA) SECRETARY: N. MCDONALD (DECD/NEA) RG ETF RL EVALUATION TASK DA. REFEREE GROUP. REFERENCE USORATORY FORCE OPERATING AGENT CEC, NDE LABS, J6C-ISPRA FROM JRC ISPRA R. NICHOLS 4 ~ CEC' JRC ISPRA P.JEHENSON S. CRUTZEN 1)'n c. (. < DESTRUCTNE EXAMINAT10N C00RDINAT10N REAL CONTAMINATED STRUCTURES TESTS DEC S. CRuTzEN (JRC-iSFRA) RCS S. CRunEN pRC-iSPRA) DATA ANALYSIS COORDINATION
' Fuu.-SCALE v5SSa TESTS DAC S.CRuizEN gRC iSFRA)
FSV K. KUSSMAUL (MPA FRG) N0ZZLES AND DISStMILAR METAL WELDS NDW' . E. BORLOO pRC-ISFRA) PARAMETRIC STUDIES ROUND ROBIN TESTS ON AUSTENmC STERS ASI- S. DOCTOR (PNL USA), H. HERKENRATH (JRC) . EouiPMENT FauRES STEAM GENERATOR INTEGRITY TESTS EFM R. DENIS (JRC IS?RA)
""Iq ,q-
. C. SIRAC (CEA-F) amSONic TESw M00aunG .
IMOD J. COFFEY (CEGB UK) E#8" E*i" 333 RA - R. MURGATROYD.(UKAEA) O' . t. 3
d
)
T S A ( l s e t e O S i c a r t i p n L s t e Ni s P , u ,ht r A or a n t c e n o o k 4 s D e r e . RH i n d u m i o S t SH a t r g c n o A i : r b s r P y i o e i t l l R d l i d a b ia C n u e L le R S I o o E P R C ON D
0 t c e b y e s s dl a e d or e i i d n t t t u l aa s e i y c n m t t n ed i i l s a t o e b e m pwa i l ab l a f r o ef l ei r h n r r a e p t u d v e n t n vd a c o
. an ye i t
h a t f i e c ta d l id b e p h e ad t w pn s n s a l aa i ef cl e d e r un h ei a r i c v e e t t e r e O c w ot r e t n a e m i s n pb mtn e
)
d f ye t l e a r u
't i
t a pt r s n n n ei mo op a e o d mr cm m ( c i ocs i of i o t i t o t 4 dd e o d t d s e c e n nt e n e n o y g i gf f g id y d s e i m i _ a r g i t c d u s n ea t u e e d u s e t D S Dh t o r A S s S t D P t c g y ec e t y l l i t vf i t r e n vi t y e v i l l l i i e e i t i l i b c a i i t t t b ia C a p cd e m a elu b a c e le S j e r b a j i j i _ ON l R I a b z a v e b E P C Oi s P Oe R O D
O e it ms k c ce6859 2 102 473 1 4 0 a e S 111 4 p 1 31 34 3 r c 1 S n
#_ oc ie s
s of re e 1,1 , 1, 1 rr or p y M M,M .M F F ca s T T, T, T, M t cFFF M AAI M,T,AA, T T,TM , sa ep f e t r sp e D ra a L l t dt A us h nh T nl t g u t ag u Oaa T rf r gi i sol ara holr a l a D tet i t cwto t t o t t gW u to t N n a l s//tb t t t tbso/tb Ai a ahhu hhhu arhu R = O s i C g gS r e/uu t t g uuu g gS C//u W gS GI A e t a s.o o sooo t t sso
) i arr arrr aar du dt M CWW CWWW CCW k
't S n ,
oin kc cs ( n s ca e e cy r r cg 4m i t t se et i ni c imr e uf m ipoe l t ce o i gl
.r a t a a t
g cSf i r e rG f ai P o r Ao y y y cat hn at l a y i r d t i tMr Ce y mh km l ui t r it l a l i e i e e l t l Cm cd hm i b ia S ba ms aie i b . t aa a le R Smu I p a rB a CW r i l e == E PS C P R TM ON D
0 ~n o i s s d l e w r l i a i m s i s d r d e n m a m s u l e s z s 9 z i 8 o ht 9 9 8 n . 1 9 9 n r 8 o S. 9 e9 1 U g O s n 8 9, m1r n e i 1 mei r i i t v e r e u mps 9
)
d i t c i v r r mts
- s
- m-u y 8
9 t a a ma 9 s es r 1 n u c 8 e r u t e l o msi s l
- l l
9i a c omwt s s o1 d u u t c em pyot s f as st
- t r -
( f e st t l l t r e st eteeT eh h ast t 4 t l zl e yu gg t f c st oc gt eh g a . z u - i r y u n l p S. oteit t r ed e el r os oost cf l t o l i r r a ei a eU nSb wwc mmd i i m i l r a wb wc a r t c z2 z ei tc p a r i a g - o .h n a l o A t i a e P r i n S. o eC* *
- P* *
- R*
y eU ot s wu l i l t l n sOto i b ia C u TA - - - le R S I t t a E P S . . . ON D
O . . n r r e v p o oe r o f s u c m s o l ei s fa n c a t l a ro b n pt c e o ear l r a n i c n p o f t e r hc m n e t e oc e a t e w m w t c. et e d o i r h i vl u g s, oy ql i e r l el l a z l pb r a z l l t e r o wus i i d on of , h f Cs g c o f n n oipoi i h s k wl a t n c i t O s s e e p a eg o p s, T i gy t msl e n e a o e vi s s i l o a f of oe r n d s r py v t e e n ot mcn er i o h e C r m t e e i d e u s u i S m m n aqe s n pt o e rp n r i o A l o a ido f r ei r o t t m a c F v e ei tr ant u ac a r g o e p r D c l at r e y v n e r P s d a e n t E te g it l y I n e i viitol i b : x n L c c i b t a e ei ia e p ,d le w k s e pb o e u e j l R a b s or c pc E N T Oipn St y n i OND
O I r f o t r s i s o p e n s a- e r i l e b t d i c f a i u t r i s d . G l i - n b a e ) o b c n i t c o n g e r i e i s p p r e s e p d n I
)
p x g a e n d g i r e M e s n e a B i d t n a a r i - o e g k s R p ) n n i ( t owe s g )n E R n n e s go n i t O e si a v i t ei o pr e i i r m d st e u ng i o r
)n geds u b e s c r
c on pi o D o t ao agr i I S e I s uWm dh o& o F k h gBC s c a d e gWn( ( a o l ed n i t n&i3 t s r 1 T E p r b a u t eB s( e ef fs M vi k p l i an d r 3Wi l S T s a A s l t v a da f e RC o.e1- ( l t A ma a n yn ar e g ai u e f oe yt r s n r gt oi r e t pme t du or c u ryl a v i t r r C r s e r wle p t SOSCC o P ci e p it y io n i e n oq u i v m A p l i b c t i t C e eo R c R u ia ep i v r P - - - - S lesn t R I c E we A . . . . ODNN
O i l b 7 0 4 i s r s g a 4- h n t s s i e R i l k C b n o / t a a r t a G s d t E e e a R o l d U t i a - y N d t i t ( e e l i d i b e l p r a c p o b n a f o e d r i r s s e p e at n wen t p c e r ) x r e u 0 e e op l e l 6 g t s i a C n e n f A i t mm o e d S a a c e n r e r b a N p a/ s p s ( s y o m a O l t u d m "e te h s u c s e t s o r n y m r I f t a s t em A R n r n s e y R P P k e oe pt yt s s y E d f o m t a i mro r e s n u o s e n s_ f t a e S t e s u f r i i t i t oe i d wvg i k t n l i b ci t r e s s e ~e o e a a o e k a pi r f s a yi n T l t b s p ma i P a r e t o n ch ai r r t i t a r I I t e p ni si r dI S gt t oi r 3 n o a e l e w g PrC oi d ed e r r r s y n mp A u Ws e ed i io t e eI R l i a " a mle n l t DIS i b c P F Ab E ia ep w les o R n c O i E we . * * .
- ODNN
O ) ) ) ) ) ~) ) ) ) ) 2 3 2 3 3 4 4 4 5 t y e0 u - 0 0 0 0 0
- - - - - 0 0- 0 0 - -
aEEEEEE EEE l i r e 1 o mcV u n 5 4 5 7 6 5 1 9 5 1 (5 (1 (7 (3 (1 ( ( ( ( 2 6 1 a a i r t ( P b r n o n i r p . o B mkn i i t a c R 2 5 3 4 6 9 1 8 7 0 1 e p
~
s n ) ) ) ) ) ) ) ) ) ) I n 4 4 5 5 6 6 6 6 6 8 r o e0 u 0 0 0 0 0 0 0 0 0 f o i t e aEEEEEEEEEE l s c e n cV1 7 8 9 0 5 0 3 7 0 pa (3 (2 ( 3( 1( (9 (6( 5( 41 ( 5 m st r n o O t e I p s d y l ei mkn S W a R 1 2 3 4 5 6 7 8 9 0 _ 1 3-: e y e l e i l t n n n r s i c o oitoe s e F a k s c i t c c t a V n n a O ej e w s mar T f mjnd nI e t r or eo ow i ei nt u rsd ai r gt e o t s e e r F e r aar s el s et i r oi r s y r S s u s u y aonoe t e v huA P C r g s c r n t _ y i o n n s e nWC eP oS 'y n e l t it i e r r e e r Gr C i b c ae k P P gcto mt or b m n e rv a a a wn ru r c c ed wh i i p - les g R In a oime et e eot at s n E w R L HESRSRPSI ONN D e l
O R k a ye tt r - l iB 3 4 4 2 5 4 3 el 0 0 0 0 0 0 0 Mie - - - - - - - bh E E E E E E E eat 2 1 5 5 1 0 4 rb . . oon 3 7 3 1 3 8 7 Cre Pv i G S E C N A T R k - O n 1 2 3 4 5 6 7 P a M R I N O I T C E S y G tt . N l i 7 8 8 9 0 0 1 I l el 0 0 0 0 1 1 1 P aMi - - - - - - - I u b E E E E E E E P nea 1 4 3 8 9 O nrb . 2 9 M' Aoo 1 2 1 7 7 1 3 E C r T P S Y S W F E 3 s
- r s E o e E t n ) s s N a i e e e O r l d n n s C e oi k i i e O n e t s n l l n e g t a i g r su t e n l a eo g o n m h n( e r i n o a c i g a t o i e s l e r h c i t t i v u c u t c s d el s s s c e ga i u S o p rv r d p s g
t m a e m u h n p p u p n ) p co p m p , m i p e si u u u d n it p n p i i e da m e P s v l o n v n n i po r e i e i ( d r ms ui f v r v p i d i
- s r - r t e t e d e d n
e n Wn n r
- n -
m i F e E m i i r n b r n l o b o t r t i u ni y o u o a t oa l m t m t mt p n W mn p W W O - C o F E oo C c S u F E F E W F E < l l
O ei t m _s a dm l u el
)
l a
)( c l
aI t i
)
l a
)
l a
)
l a yx SE wov u uI a u u u s s t s s 's t i ids i i i l d an td v vno v v v un ( ( ar ( ( ( ca b a d - i I II y I I I py I e I I h s I I I yr I c I I s s I I I T o g a eg a - sf s ssn l s s s e sr s ssi c s s s t au a aar a a a a l s l l l u o l l l C C( C CCd N C C 'C S T ' N E E C M N E A R T I R U k O Q n 1 2 3 4 5 6 7 P E .a M R R I I N S O I I - T D C N E A S N y G O tt N I 7 I T l i 8 8 9 0 0 1 l el 0 0 0 0 1 1 1 w P A aMi - - - - - - - I C u b E E E E E E E P I nea 1 4 3 8 9 2 9 O M E T F I S S nrb Aoo C r P 1 2 i 7 7 1 3 S A Y L S C W I F X E ' N 3 O s r E
- I s E
T C o e N E t n ) s s O a i e e e S r l d n n s C e oi k i i e 0 E n e n M e g t s l l n S g r t a i A su t e n l a eo g o n m h n( e r i n o a c i g a t o i e s l e r h c i t c t s i v u c u t d el s s s c e ga i u S o p rv r d p s g t m a e m u h n p p u p n ) p co p m p m i p e si u u u d n it p n p i i e da m e P s v l o n v n n i po r e i e i ( r ms f v r v d ui i d i
- p s r - r t e t e d e d n
e n Wn n r
- n -
i F e i i r _ m b E m l o b o . n r n t r t u y i a ni o u o _ t oa l m t m t mt p n W mn p W W W O _ o F oo u F F F C E Cc S E E E
, ! i
O 4
~
r e i s r u s y t c l e ) f a b c l a u l a i t r n n i a c a v e) ) A m a l E es c O i r A h e r l a u t c c a r o e R a eh ) l P e t a n gee s r o h e n sl a r o r e r f t t Giu ( t s o e r f u s s eh ge n l i x 2 s t c e t a d o e m(WnG i t s( S n f u t s t o1 f a n m u t el k s a l P a r o m B ohaWu y( G T m j s ma f l a A h c o1 u d n ai r gt r e o la m R agn a oi r n s s P e eior PSZG P C r r r y n i o e v e v d o i io l i t t t c o o o _ b c a ep e C C G - - - - _ i l es l _ R n i e E w S . . . o _ D e _ NN
O n i o t s a c m i f . i s e s t s l a y c . S t n e k
- a. d n f l o a o -
p o
- C t e
g t ht g O i n t n a . i w r u s k l p t n r n e e a t n m p p R t e e e u s A i s r g m R n a o P o n f r c i s s n g g n e k n n o n T s a o i k i k i t p i l y ma s a n n a e l p ai r r e n r r c p gt oi r C r i o m m x e u P s : e e s a w l y n i t io l u r e t s y t sy f e l i b c t ia ep n c n e S S A - e g les R n o n E we C I ODNN i
O n h d o i l t a d n t e e c a a s d d e i i v d d g u o n e n G r p a w a e
~
t h n d f a i v c n e r o r i a c r s e r f o
's t i n s c n s a r b n
e p i t od r e
, el ci l
i o c a c n w t a S ed n a o s d n pn r c e n I s a u r i r e nt S s at d i 8 n e s u m e l ed v n . 8 i c d .m s 9 l a u o ni n i c a t a e s 1 i t r
/r c e
B I f e 0 s X e r f 3 s u h ed t n d nt h g k - o r n oo c n s t sC e e i ai o oi n i O bmgm t i c m R cE l i v e ro 9 r pt a r M e S 8 v s - i pn n fo S i c f 9 o oa s , E s 1 a o c sA Nl a r a Mn t
, l p e e 2 latn n de - i S e 1 r e
u i l e o v gi n t r l c A m - e v o r d r i ni a u n n p ol o 1 1 o u m u p e dn ey i t l t c G s e i gm n k g p i e m s a/m e o s r c u e v A m r pe do T s a n l owe gampnfo/ l f s d s gpb n o e ma i e n n ot i i ol l n k v r i k g4 h i i t l l l e ei i ai r r e r e r st a wio v w gt oi r o Dfo oe ge t n r zI t ad e e e mD pe t P C r W : n wSibi ec n t i u y i o n eo vi t s e mag sl appx i l t t d l t i t E i t a r e e r h ppe u MMoPA i i t b c F aN OC ia ep M cd e n les j S b u R n i E we A Ofo . . . . . . . ODNN l
O
~
s e i n l e t d i n u e. G m n u t s o c n i t o a c D l p . e h p 8 s c n a o I o r e s t s d p a i e p s s B y a A a l a B l a t a n k u D a O i s t p S A R R e D
- c P n R e o P f
o r c C N o n F n n o k o o i s s R R n a a T C
/
C p
/ x k E G G e s E E e T
a M R R h ma kr S U U t air A N N e r gt oi e r o t r o h h t e W l P C r p i s i s p y n l l o p b b m e o _ t ii l i t r u u u _ b c S P P C ia ep tu les R n I u E w F . . . . ODNN e O
.}}