ML20113D319

Final Rept on Evaluation of Spot-Welded Matls Used in Support Sys for Electrical Conduit & Cable Trays at Diablo Canyon Power Plant

ML20113D319
Person / Time
Site:	Diablo Canyon
Issue date:	07/01/1983
From:	BECHTEL GROUP, INC.
To:
Shared Package
ML20105B784	List: ML20113D238 ML20113D319 ML20113D327
References
FOIA-84-21 NUDOCS 8501230059
Download: ML20113D319 (19)

Text

. ._ __

.g_

'g . .

ENCLOSURE 1 4

4 FINAL REPORT 4

ON THE EVALUATION OF SPOT-WELDED MATERIALS USED IN 4

SUPPORT SYSTEMS i

FOR ELECTRICAL CONDUIT AND CABLE IRAYS AT DIABLO CANYON POWER PLANT 4

July I,1983 l

\

r f

e i

i j

) )

i e

85012 0059 840627 in - _

PDR FOIA PDR N~i

~

3 GARDE 04-21 t i

I l DIABLO. CANTON PROJECT  !

PACIPIC GAS AND ELECTRIC COMPANY RECNTEL POWER CORPORATION 16 !

7-TABLE OF CONTENTS Section Title 1.0 Introduction and Summary 2.0 Test Program 3.0 Evaluation of Test Results M c*1 cniAL*G r 7(54 gg gg y g Table Title A L'## - pg,j .

1 Test Results - A1202 Sections SD#[0/'d'#

2 Test Results - R1202 Sections 17 6 3 Test Results - H1202 Sections NO Figure Title 1 Back-to-Back Superstrut Members i

2 Test Specimen Configuration 3 Typical Load vs. Deflection Curve .

~

I~ /b . '

wr/m-u to M. HA u

l:

. .-* c e

r. '

1.0 INTRODUCTION

AND SUMMIJtY 1.1 Purpose This report describes the testing and subsequent analyses performed to assess the adequacy of spot-welded Superstrut channels installed in the Diablo Canyon Power Plant, Units 1 and 2 (Diablo Canyon). This testing was undertaken to supplement the information submitted in PGendE's February 25, 1983 response to the NRC regarding the adequacy of spot-welded materials in use at Diablo Canyon.

1.2 Backaround Electrical conduit and cable tray (raceway) supports at Diablo Canyon use composite, spot-welded channels manufactured by the Superstrut Division of Midland-Ross Corporation in Oakland, California. The

, flexural capacities of the composite channels are specified in a catalog provided by the Superstrut Division. The flexural capacities given in the catalog were used in designing the Diablo Canyon raceway supports. In order to develop the flerural capacity of the composite sections, shear between the individual strut members must be resisted j l

by the spot welds.

f In December 1982, NRC personnel conducted a facility inspection at the Superstrut8 Division plant in Oakland, California. This inspection was t made in response to an allegation regarding the Superstrut Division's control of welding processes used in the manufacture of composite 3417a/0142A/es t

O

. channel struts. The NRC reported in January 1983 that no evidence could be found of a program to verify the published capabilities of the composite channels.

The Superstrut channels used at the Diablo Canyon were procurred from  !

the Superstrut Division's Oakland plant as commercial grade material and were manufactured in accordance with industry practice. As a result of the 1GLC's findings, the Diablo Canyon Project (Project) undertook efforts to demonstrate the adequacy of these commercial grade materials to meet the Diablo Canyon licensing commitments. An investigation into the quality of the spot-welded Superstrut channels was made by the Project. The results of this investigation were summarized in a submittal to the )GLC dated February 25, 1983.

Upon completion of this investigation, the Project began a testing program to provide further confidence in the adequacy of the strut material as used at Diablo Canyon.

I 1.3 Summary This report presents the results of testa made to determine the shear ,

capacity of spot welds in Superstrut channels. These tests were performed to dpfine the capabilities of the composite channels in use at Diablo Canyon. This report provides a description of the test scope, procedure, and data evaluation.

o

• v After completion of a statistical evaluation of the test results, acceptance criteria were established. Upon reviewing the raceway support designs against acceptance criteria, it has been verified that the applications of composite, spot-welded Superstrut channels at Diablo Canyon are acceptable as installed.

2.0 TEST PROGRAM The test program was comprised of two main activities: a) sample selection and b) specimen preparation and testing. All aspects of the I

program were performed in accordance with the " Outline for Sampling, Preparation, and Testing of Spot Welded Superstrut Members for Diablo l Canyon Power Plant", (Attachment 1). Salient aspects of this procedure  :

are described below.

v 2.1 Sample Selection Development of the sampling program and subsequent data evaluation were performed in conjunction with Jack 1. Benjamin and Associates, Inc.

The following considerations provide the basis for the sample size:

l 1

i a) The number of samples of Superstrut channel included in the test

! j' l

program

. is equivalent to approximately I weld per 500 f t of

}l l channel. Ibis test frequency is consistent with that used during i

manufacture. Diablo Canyon's raceway supports contain i approximately 100,000 ft of strut material. Therefore, at I weld 3417a/0141A/es "

. .g ', ,. j i

l

. l test per 500 f t, a comparable number of tests is approximately '

200. ~This results in testing samples from approzinately 1.3% of all supports containing composite struts.

b) Three types of composite Supet strut are installed in the plant.

They are all back-to-back and are identical in section width and l

naterial thickness, varying only in member depth: A-type (3-1/2 i,

t j

i

,, in. deep); E-type (4-7/8 in. deep); and B-type (6-1/2 in. deep),

Figure 1. From a review of the support details for the plant, it was estimated that 60% of all supports use A-type sections, 30% use E-type, and 10% use H-type. For large populations of materials, industry suggests 30 tests as a sufficient sample size for i statistical evaluation. Therefore, 30 H-type strut members (the 1

amallest percentage population) were selected for testing and s proportionally larger sets were selected of the other two sizest 90 E-type and 150 A-type for a total of 270 tests. This sample

size exceeds the sampling target of 200 tests and 1.3% of all supports containing composite struts.

t i

I Specific samples to be tested were selected from all raceway supports at Diablo Canyon by random number generation. This technique provides l a statistically unbiased set of samples representing the total population of , composite channels in the plant. Some supports selected i by the random number generation were omitted from the sample set.

{ These could not be used as samples because they did not contain a spot-welded composite channel, were inaccessable, or the channel was too short for testing. .

l 3417a/0142A/es e

. l When the selected samples were removed in the field, it was found that only about 15% of the raceway supports containing E- and B-type members had spot-welded composite sections. Based on the adjusted populations, new sample sizes were determined. Since the population of B-type members was much smaller than originally estimated,10 samples were selected for this statistically small sample. For the larger population of E-type members, 30 samples were selected. These sets represent 6.5% and 4.1% respectively of the supports containing spot welded E- and H-type sections. These final sample sets provide a higher percentage of tested samples than the percentages originally targeted, and are, therefore, statistically acceptable for the purpose of developing spot-weld strength criteria.

2.2 Specimen Preparation and Testina 1

Specimen preparation and testing was conducted and documented in Bechtel's Material and Quality Services Testing Laboratory in I

i accordance with approved QA Program Procedures. 162 A-type members, 34 E-type members, and 9 H-type members were tested to failure. For each i

specimen, shear load was applied to a single spot weld by applying a tensile load to the specimen at a slow uniform rate until failure of the spot weld occurred, Figure 2. Tables 1, 2, and 3 contain the test results. All welds tested had strengths greater than 1600 lbs. When i

performing the thear tests, substantial elongation of the spot welds

was observed, indicating the failure of the specimen to be ductile I l

l 4

3417a/0162A/es

j.

• rather than brittle. Ten additional specimens (7 A-type, 2 E-type and 1 H-type) were prepared and tested from which load-deflection data were recorded. A typical load-deflection curve is presented in Figure 3.

1 Data was forwarded to the consulting firm of Jack 1. Benjamin &

-Associates for a complete statistical evaluation (Attachment 2).

3.0 EVALUATION OF TEST RESULTS J

Statistical analysis of the data consisted of the following: (s) y i

comparison of data from the three strut types to determine whether the  :

j spot-weld strength could be assumed to be from the same population of  !

spot welds and thus combined; (b) determination of the validity of the  !

{ normal distribution to reasonably represent the spot- weld strength j

populations in the plant. These tests were made at the 0.05 i

4 significance level which determines that there is a 95% chance of not 1, i.

j rejecting a hypothesis when it is true.

l f Based en the results, it is concluded that spot-weld populations of the l

i A- and H-types any be combined. The E-type population, however, has a

different mean and standard deviation, and therefore, should not be i

combined with the other populations. Results also show that the A- and '

H-thedatafitthenormaldistribution.

s The E-type does not fit the normal distribution over its entire range, but can be conservatively l l

considered a normal distribution at the lower end.

I i

1 l

3417a/0162A/eg l _ - - - - - - - - - - ~

It was observed during the testing that the individual spot welds elongated considerably, indicating significant ductility in their Ioad-deformation characteristics. Subsequent tests performed on ten I additional samples provided load-deformation curves as a measure of the inherent ductility of the system. This important property allows '

redistribution of loading to take place among the participating spot welds for the entire system to achieve its maximum resisting capability. The reevaluation of the as-built raceway system has taken into consideration the ductile behavior of the welds.

Statistical results of shear capacity tests are as follows:

Types A & H Type E Mean 5,383 (1bs) 3,156 (Ibs)

Standard Deviation 1,458 (Ibs) 1,069 (1bs) 10th Percentile 3.514 (1bs) 1,786 (1bs)

The 10th percentile determines the strength which is exceeded by 90% of all welds, when the strengths are normally distributed. For this application, considering the demonstrated ductility, the shear capacity j corresponding to the 10th percentile is very conservative and adequate 1

for this evaluation.

I The shear allowables 4 for the two sets are:

't l

Type Allowable Shear Per Spot Weld A and E 3500 (Ibs)

L E 1775 (1bs) '

I l I 3417a/0142A/eg

Baving established an acceptance criteria for the shear loads in Superstrut channel spot welds, the raceway suppc te designs were reviewed. The shear flow at the junction of the individual strut members was calculated, distributed between the spot welds, and compared to the allowable shear loads cited above. The calculated shear loads were within the allowable shear loads.

This testing program, in conjunction with a review of the calculations, confirmed that the spot-welded Superstrut channels used in the Diablo Canyon raceway supports are acceptable.

l l

l  ;

b l

i

/ l t

t l

l f

I i

3417a/0142A/eg

r.  ;

LIST OF ATTACHMENTS

1. " Outline for Sampling, Preparation, and Testing of Spot-Welded Superstrut Members for Diablo Canyon Power Plant," May,1983.

2.

" Statistical Analysis of Shear Strength Data for Support Strut Spot Welds Diablo Canyon Power Plant," Report by J. Benjamin & Associates, June, 1983.

e e

7--

~

TABLE 1 TEST RESULTS - A1202 SECTIONS Shear Load Shear Load Specimen Number G Failure (Ibs) Specimen Number G Failure (1bs)

A1202 - S60 5820 A1202 - S36A 5120 A1202 - S78 4860 A1202 - 593 5060 A1202 - 570 6080 A1202 - S45 6020 A1202 - $42A 7560 A1202 - S337 5500 A1202 - S431 6760 A1202 - S391 5420

,. A1202 - S124 5000 A1202 - S344 4520 A1202 - S340 4020 A1202 - $395 6100 A1202 - S131 5660 A1202 - S208 5840 A1202 - S156 5780 A1202 - S192 6860 A1202 - S381 5480 A1202 - S543 6040 A1202 - S256 3620 A1202 - S331 4960 A1202 - $369 4440 A1202 - 5339 6220 A1202 - S216 7480 A1202 - S399 5580 A1202 - S65 5060 A1202 - S67 2900 A1202 - S443 6920 A1202 - S444 4520 A1202 - S304 2820 A1202 - S171 6840 A1202 - S639 5060 A1202 - S683 4880 A1202 - S329 6360 A1202 - 5403 5000 A1202 - S273 4440 A1202 - 5396 3860 A1202 - S$97 5740 A1202 - S394 4980 A1202 - S236 4920 A1202 - S584 4740

'A1202 - 5241 5280 A1202 - S654 4440 A1202 - S393 4560 A1202 - S335 4480 A1202 - $165 6880 A1202 - 8378 5020 A1202 ,S472 6100 A1202 - 5214 6900 A1202 - S352 ,

5040 A1202 - S118 5020 A1202 - $6 5580 A1202 - 5298 5700 A1207 - $170 5160 A1202 - S451 5060 3429a/0105A/eg

l TABLE 1 (CONT'D )

TEST EISULTS - A1202 SECTIONS Shear Load Shear Load Specimen Number 8 Failure (Ibs) Specimen Number G Failure (Ibs)

A1202 - S221 8100 A1202 - S154 4880 A1202 - $21 5880 A1202 - S292 3520 A1202 - S104 6900 A1202 - S361 2360 A1202 - S146 5800 A1202 - S418 3940 A1202 - Sill 4180 A1202 - $426 6060  :

A1202 - S148 4820 A1202 - S441 3280 A1202 - S205 3820 A1202 - S655 4020 A1202 - S254 6480 A1202 - S272 3140 A1202 - S495 2760 A1202 - S248 5680 A1202 - S479 5040 A1202 - S406A 5280 A1202 - S312 4420 A1202 - S191 5940 A1202 - S373 5620 A1202 - S416 4680 A1202 - $182 3040 A1202 - S193 5440 A1202 - S547 4140 A1202 - S782 6360 A1202 - S468 5320 A1202 - $386 4460 A1202 - S164 4700 A1202 - S346 7780 A1202 - 5115 5060 A1202 - $218 5380 A1202 - 5694 6840 A1202 - 5250 4380 A1202 - S279 4740 A1202 - S242 5720 A1202 - S372 4920 A1202 - S180 5300  !

A1202 - S612 5040 A1202 - S289 3640 A1202 - S219 [

5860 A1202 -5415 3360 l A1202 - S538 4400 A1202 - 582 3480 A1202 - 8407 4440 A1202 - S630 4900 A1202 - S144 l 6400 A1202 - 8637 5140 A1202 - S632 4860 A1202 - $513 7540 {

3429a/0105A/eg )

W k' ,

TABLE 1 (CONT'D)

TEST RESULTS - A1202 SECTIONS Shear Imad Shear Load Specimen Number G Failure (1bs) Specimen Number G Failure (Ibs)

A1202 - $430 6060 A1202 - S38 4720 A1202 - 5812 3760 A1202 - $19 6700 A1202 - S25 4020 A1202 - S13 7600 A1202 - S23 4100 A1202 - S290 6120 A1202 - S516 5580 A1202 - S743 6060 A1202 - S708 5320 A1202 - S617 6440 A1202 - S498 4920 A1202 - S134 3680 A1202 - S502 3840 A1202 - S716 4040 A1202 - S685 6200 A1202 - S609 6560 A1202 - S590 5120 A1202 - S742 3220 A1202 - S233 4460 A1202 - S781 2940 A1202 - S92 6600 A1202 - S334 6000 A1202 - S440 8620 A1202 - 5114 4920 A1202 - S583 8040 A1202 - S187 5280 A1202 - S641 4420 A1202 - 5520 5240 A1202 - S769 5240 A1202 - S580 9200 i

A1202 - S644A 5420 A1202 - S796 2280 A1202 - S532 4080 A1202 - 5706 6160 A1202 - S351 5880 A1202 - S455 6800 l A1202 - S494 3920 A1202 - S725 3060 A1202 - S562 5180 A1202 - 5696 7060 A1202 - S515 11400 A1202 - S76 7820 A1202 - S582 / 9620 A1202 - S534 6520 A1202 - S35 5300 A1202 - 866 2920 A1202 - 8528 5640 . A1202 - S524 5340-A1202 - 8731 8720 A1202 - 8718 6080 A1202 - 5923 9000 A1202 - 5777 5600 3429a/0105A/eg I

s .

TABLE 2 TEST RESULTS - E1202 SECTIONS Shear Load Shear Load Specimen Number 4 Failure (1bs) Specimen Number 4 Failure (Ibs)

E1202 - 5350 4

2980 E1202 - S686 2740 3 E1202 - 8420 3380 11202 - 5660 3120 j E1202 - S397 2380 E1202 - 8610 2700 11202 - S414 2760 11202 - S673 3420 11202 - S793 2120 11202 - S326 4300 11202 - 5778 1620 ~ E1202 - 5540 4020 11202 - S349 3420 11202 - S1037 2700 E1202'- 5829 2560 11202 - 51013 5760

.11202 -'5970 2200 11202 - S959 3000 E1202 - $1005 3120 E1202 - 5961 2620

11202 - S971 1940 -

11202 - S924 2520 E1202 - 51072 1980 - E1202 - S931 3060 11202 - S991 3440 11202 - S981 6060 l E1202 - 5936 2800 E1202 - S825 5120 E1202 - S818 2660 E1202 - S948 2600 E1202 - S1021 2700 E1202 - S1003 5700 E1202 - S977 3280 E1202 - S1031 2540 i

t 4

a I

/ r i '!

t F

f i

3429a/0105A/eS

. . . , _. . - . _ ~ _~ . . . . - - . _ _ . , _ _ - , _ _ _ _ , . _ , . . . - . .

p TABLE 3 -

TEST RESULTS - R1202 SECTIONS Shear Load Shear Load Specimen Number G Failure (Ibs) Speelsen Number G Failure (Ibs)

H1202 - 51027 4780 E1202 - S353 4860 B1202 - S984 5320 H1202 - S844 4920 ,

E1202 - 5843 9700 R1202 - S827 5600 H1202 - S983 7680 H1202 - S994 4280 E1202 - S1091 3920 h

l' 3429a/0105A/eg l

F E

,9 .-I

• e e

4 e

i, 7F"

"(TYP.)"

l  ; 6-1/2"

, 4e7/8"

4 31/4" A1202 E1202 H1202 FIGURE 1 l

DIABLO CANYON NUCLEAR POWER PLANT l

t I

BACK TO BACK SUPERSTRUT MEMBERS r

l .

HOLES TO BE DRILLED -

PRIOR TO SAW CUT SAW CUT THRU OPPOSING MEMBER ON EITHER SIDE OF CENTRALLY LOCATED SPOT WELD, AS SHOWN.

TEST 1 SPOT WELD AT APPROX.

CENTER OF SAMPLE

\

i l *

, 's I s  %

. , .. arP.,

, ..... ~

b:

t 6

• \

3 APP ROX. 7"

,E.,Y ... P.,

,.7 -

FIGURE 2 DI ABLO CANYON NUCLEAR POWER PLANT TEST SPECIMEN CONFIGURATION

.# D'

• l l

l

, I

puLTIMATE -%PACITY (5300 LBS

5-l 1

x 8

23 c BREAK POINT d (2520 LBS) ,

2- r i

1- f I

t 0 '

0 0.b ' O.04 0.N 0.'08 0'10 0.'12 0.'14 0.'16 0.18 DEFLECTION (INCHES) ,

9 f

e I

l' 9

FIGURE 3 l

DIABLO CANYON NUCLEAR POWER PLANT I

'i TYPICAL LOAD VS.

DEFLECTION CURVE y - - - , , w --h -