ML20027A543
| ML20027A543 | |
| Person / Time | |
|---|---|
| Site: | Wolf Creek  |
| Issue date: | 11/24/1978 |
| From: | Koester G KANSAS GAS & ELECTRIC CO. |
| To: | |
| Shared Package | |
| ML20027A542 | List: |
| References | |
| NUDOCS 7812200349 | |
| Download: ML20027A543 (11) | |
Text
.
FII;AL REPORT ON
_ SPOT UELDED STRUTS FOR S!;TTPPS PROJECTS TABLE OF CO';TE!;TS SECTION TITLE PACE 1.0 Introduction 1
2.0 Material Description
--~~~~~~~~~~'1' 3.0 Material Deficiency and its Safety Implication L
4.0 Technical Evaluation and Criteria 2
5.0 Test Program at Jobsites 3
6.0 Additional Plan of Action 3
7.0 Test Program at Unistrut Corporation 4
8.0 Corrective Measures 5
9.0 Conclusions 5
TABLES:
1.
Test Results of Callaway and Wolf Creek Jobsites 6
2.
Test Specimen Details 7
3.
Shear Load Test Sus: y 8
4.
Retest of Specicen 9
SKETCH:
1.
Test Set-Up 10 7812200 397
1.0
_INIF0 DUCTION In accordance with the require ents of 10CFR 50.55e, this report is prepared to provide a final summary of the deficiency in spot resistance welding of electrical raceway support material, pur-chased under Specification No. 10466-E-037.
In July 1978, SNUPPS Project was advised of a problem related to inadequate spot resistance welding on the strut supplied by Uni-strut to several other projects, where the adjoining channels were separating during fabrication of supports. No such failure was reported on either of SNUPPS jobsites.
Immediately, tests were initiated at both SNUPPS sites. A total of 76 randocly selected samples from different strut types were subjected to a shear test to verify the veld strength. No failure was reported.
The specification was revised to include shear test for any new material purchase as a precaution. Additionally, 380 randomly selected samples from both sites were tested at Unistrut test laboratory to obtain more data. Only one Unistrut Part No. P-5501, Coil No. 7J-4661 from Wolf Creek site, was found unsatisfactory.
Immediately, actions were taken to re=ove this strut from the site.
2.0 MATERIAL DESCRIPTION Basic strut channels'are roll forced from hot dipped galvanized #12 gauge, 0.105 inch thick strip steel. These channels are connected to each other in various configurations, such as back-to-back, back-to-side, or side-to-side, by spot resistance welding. This caterial is purchased from Unistrut Corporation, Wayne, Michigan, as standard catalog items under the Specification 10466-E-037, used in field fab-ricated supports for electrical raceways.
3.0 MATERIAL DEFICIENCY AND SAFETY IMPLICATIONS The strut is used in the electrical raceway support system design.
The connection provided by spot welds between the strut sections is relied upon to caintain the integrity of the built-up section to carry the postulated loads, within the design =argins. Thus, inadequate Page 1
fusion at the spot welding may result in inadequate strength and may adversely affect the safe operation of the plant, under severe loading conditions.
After revie ing the nature of the deficiency experienced on other projects, the following was concluded:
The deficiency was related to spot welding technique and/or a.
procedure, b.
The quality of the spot welding con not be evaluated without actual test to verify the strength of the veld, Destructive shear test will provide the failure locd data c.
for the veld, and will positively verify the soundness of the spot weld.
4.0 TECHNICAL EVALUATION
AND CRITERIA Based on the "F.ecoz=anded Practices for Resistance Welding" AWS C1.1, by American Welding Society, AIST Specification for the Design of Cold-Forced Steel Structural.Membees, Part I, Section 4, specifies allowable shear strength per spot veld to be equal to 1.65 kips, with a factor of safety of 2.5 for 0.109 inch thickness. High factor of safety is specified to account for the dependence of the quality of spot weld on many variables such as tips of electrodes, pressure, voltage, current, presence cf impurities, finish of the caterial, caintenance of the equip =ent.
The connection provided by spot welds between the strut sections is relied upon to caintain the integrity of the built up sections. The calculated shear in a spot weld in a member depends upon variables such as loading, sectional properties, end conditions, use of strut as beam, brace, colu n or tie, etc.
Factor of safety used for the bending stresses in the struts is 1.65 in the design of the supports.
Therefore, it is reasonable to consider a shear value different than that specified oy AWS.
Ccnservative weld strength require =ent for SNUPPS Project is calculated to be a caxicu: of 1,200 lbs only. There-fore, for the analysis of the test data, failure load of 1,200 lbs or below is considered inadequate.
Page 2
4 It was also decided that when failure load is below 3,200 lbs.
for a test specimen, two adjacent velds shall be tested to provide additional data. A value of 3,200 lbs. failure load is choser. as a conservative design base which includes safety factor of 2.5+.
?
5.0 TEST PROGRAM AT JOBSITES - (Callaway and L'olf Cresc) 5.1 Testine Forney Compressive Testing Machine (which is used for concrete a.
testing) with specially fabricated fixtures was selected to a
perform this test.
b.
Sa ples were selected at random from every strut loE received
~
for different purchase orders.
Destructive shear test was performed and dailure loads were c.
recorded.
d.
Suc=ary of the test data is tabulated in Table No. 1.
/.
5.2 Review of Test Data Out of 76 samples tested, only one weld was below 3,200 lbs., failure load being 1,600 lbs., with adjacent weld at 3,400 1.bs.
Including the above weld, there were 3 welds between 3,400-4,000 lbs.,' tt a at 4,000 lbs., and 71 were above 4,000 lbs. Highest value has 9,0'00 lbs.
5.3 Conclusion of Initial Test This data did not indicate any problems with struts at,S;UPPS sites, and provided reasonable assurance abo 6: the adequate quality of spot welds.
In addition, the specificatiba was revised to' require additional testing for future shipments.
6.0 ADDITIONAL PLAN OF ACTION The following plan of action was decided in the coeting between Uni-strut, SNUPPS and Bechtel on August 17, 1978:
a.
Unistrut indicated that ' starting in 1973, all eatbrials fabricated and delivered to the SNUPPS jobsites had been subjected to pull tests, tha results of which were traceabl so individual coils,of taterials delivered. This data was reviewed and cc.tsidered accep-table, subject to final Bechtel audit.
(This data was reviewed at Unistrut by a Bechtel representative on October 17, 1978, and l
vas considered acceptable.)
Page 3
- ? ;-
=..
)-
~
' 'i n.
. - -J.,
N f'
?.
~
b.
Representative ' samples from 1976-1977 strut material shall be-c subjected to shear test, as the pull test d-ta was not availabic or'Erac4able.
- c. ' Materials to be shipped in the future shall be required to have shear test, per revised specification.
7.0 TF.ST PROCPJb! AT UNISTRUT CORP.. k'AYNE. MICHICAN 1
7.1 Criteria Basic criteria vas. set per Paragraph. 4.
7._2 Samples Quantity of samples was based on one weld per 100 feet of strut. An effort was made to select the samples from all different coil nu=bers i
e
~
previously delivered to Callaway and Wolf Creek jobsites. All samples were taken from stock; fabrication shop or scrap pile. Table No. 2 lists numbers of samples from each site, by strut type and number of cpils.
3" long test specimens (380 total) were p~repared and identi-r
(
fied by Unistrut Q.A. department.
7.3 Testing Shear tests were performed using compressive shear test set up per j_
d Sketch No. 1.
The tests were witnessed by a Bechtel representative.
9 7.4 Review of Test Data Sum =ary of the tert data is tabulated in Table No. 3.
352/380 speci-mens exceeded 3,200 lbs. shear load. An additional 15 specimens had if shear loads between 2,400-3,200 lbs. An additional 10 specimens had
[
shear loads between 1,200-2,400 lbs. Only 3 speci= ens were below I
1,200 lbs., which were considered inadequate. Two of these were froa j
the same Coil No. 7J-4561, Unistrut P-5501. These welds failed at saucut, but three other samples fro = the same coil were above 4,000 4 -
lbs. One speciten beles 1,200 lbs., was from Coil No. 7X-4444, Uni-
[-
' strut P-1001/ ~There was a total of 15 samples from this coil, from r-which only one was b*elow 1,200 lbs., two were above 2,400 lbs., and twelve were above 3,200 lbs. There were 15 sc=ples from Coil No.
7K-4444 (same coil as 7X-4444, but different operator),14 were above 3,200 lbs., and only one was at 2,350 lbs.
Page 4 lL 5,
p
K.
\\
x ya.
-1 Additional data for the welds adiacent to the scecimen with a failure load of below 3,200 lbs., are tabulated in Table No. 4.
Average shear load (average of the original specimen and two adjacent welds), was
~ Q ~Eigher than that of the original specimen. Two exceeded 3,200 lbs.,
.g s
\\
None of which was from Coil No. 7K-4444.
N s
7.5 Statistical Analysis The test data was grouped according to the welding machine which pro-duces a given shape. Statistical analyses were performed on each sampig'to include computation of scmple statistics (e.g. mean and
+ s.T standard deviation), plotting histograms, and fitting probability density functions to the data.
The factor of safety against failure, computed as, Minimum specified weld strength Factor of safety =
,aximum design load A
was greater than 2 for all samples; which is greater than the factor of safety for the strut material of 1.65.
The minimum specified wcld c r.
strength was determined for a 90t confidence level; i.e. lot failures (minimum specified weld strength) in the lower tail of the probability m
4 density function.
7.6 Conclusion of Shear Tests at Unistrut Based on the above, one Coil No. 7J-4661, Unistrut p-5501 at Wolf Creek
+
site, was considered inadequate. All other pro-1978 strut material ex-coeds the design requirement; failure in Coil No. 7X-4444 is considered to be an isolated case.
8.0 CORECTIVZ MZASURES a.
Specification Mo. 10466-E-037 has been revised to include additional testing of the spot weld at the manufacturing plant. Documen tation is required to be shipped uith the strut.
s b.
Unistrut p-5501 Coil No. 7J-4661 is being removed from the Wolf Creek job site and 5.111 be replaced with acceptable material.
9.0 COMCLUSIONS
The problem described in this report was the result of:
+
Inadequate test program prior to 1978 by Unistrut to verify the a.
strength of the spot weld.
b.
Inadequate maintenance of uelding equipment by Unistrut.
Supplier's actions including the added Quality Assurar.ce and test programs per revised specifications, will provide adequate assurance of effectively precluding reoccurrence of this problem.in the future.
i g
--m
, ~ ~,. - -,,
Table a. 1
- Shear Load Test Su==ary for Callaway and Wolf Creek Jobsites
. n.stru:
..,o.
O:-
SF-P L3/.D3 f1bs.)
Part ':3.
Snecim a
{ 1.2C0
!<,2.400 l<3.200 l>3.200 P-1001 Callaway 60 60 Wolf Creek 8
1 7
(n. )
A0ino 68 3
1 67 P-3501 Callaway 0
0 Wolf Creek 4
4 (3) TOT /J.,
4 4
P-5001 Callaway 0
0 L'olf Crcek 4
4 (C) TOTA 1.
4 4
TOTAL (A+E*C) 76 1*
75 Note:
- l.
Wh c ' t adjacent ucid was subjected to shear test, it exceeded 3,2 00 lbs.
2.
At Callaway, cinieu: and ta>:i=u= shear loads were 4,000 lbs.
and 9,000 lbs., respectively.
3.
At Wolf Creek, cinitus and ca::icus shaar loads were 3,000 lbs. and 7,200 lbs., respectively.
Page 6
Tnble No. 2 Test Specimen Details:
UhiON - (CAIJ.AWAY )
KANSAS - (WOI.F CREEK)
Unistrut Total Total lio.
Total No. Of 7 No. O f Total No. O f No. Of Part No.
Quantity Of Srecimen Quantity Specimen Coils Quantity Specimen Coils P-1001 20,000 216 10,000 91 7
10,000 125 8
"-5501 8,000 68 2,000 12 5
6,000 56 6
P-5001 9,500 96 3,500 23 2
6,000 73 29 TOTAI.
37,500 380 15,500 126 14 22,000 254 43 s
E Note:
(1) Coil numbers selected for the sampic at Callaway were different from those at Wolf Creek.
(2) All the sampics were randomly taken from the respective colt number and Unistrut part number.
(3) p-1001, 20 foot long sections were made with five head welding machine.
Sampics representing each wcld head were taken.
e
+
t
.. _ _ _... _, _ _ _, a-Table No. 3
- Shear Load Test Su==ary I
Unistrut No. Of SHEAR LOADS (1bs.)'
Part No.
Specimen 4 1,200 l42,400 1 <3,200'
>3,200
?-1001 Callaway 91 0
5 5
81 Wolf Creek 125 1
1 2
121 (A) TOTAL 216 1
6 7
202 i
P-5501 Callaway 12 0
- 1 5
6
. Wolf Creek 56 2
1 2
51 (B) TOTAL 68 2
2 7'
57 P-5001 Callaway 23 0
.0 1
22 Wolf Creek 73 0
2 0-71 (C) TOTAL 96 0
2 1
93 3
380 3*
10 15 352 TOTAL (A+B+C)
(1) 2 were from Coil No. 7J-4661, Part No. P-5501. Total of 6 sa=ples fro = this coil were tested. 3 were above 4,000 lbs., I was 1,400 lbs., and 2 failed at sawcut.
- (2) I was from Ceil No. 7X-4444 (?-1001). A-total n' 15 samples were tested, twelve were above 3,200 were above 2,400 lbs, and only one was below
_,_su There were 15 sa=ples from coil No. 7K-4444 (same coil as 7X-4444, but different operator),14 were above 3,200 lbs., and one was at 2,350 lbs. When adjacent welds were tested, average load exceeded 3,200 lbs.
4 5
Page 8 i
i
~ _, _ -.
..._,,_,_____,_.m.
._. - _...,. _.. _ _. ~,,... _,
. - ~ -- ---.
Table No. 4 - Retest of Specimen, Below 3,200 lbs,' Shear Load.
Unistrut No. Of AVERACE SHEAR LOAD (lbs. )
Part No.
Specimen 4 1,200 42,400
<3,200
>3,200 P-1001 Callaway 10 0
5 4
1 Wolf Creek 4
1
'l 1
1 (A) TOTAL 14 1
6 5
2 P-5501 Callaway 6
0 0
6 0
-Wolf Creek 5,
2 1
2 0
(B) TOTAL 11 2
1 8
0 P-5001 Callaway 1
0 0
1
'O Wolf Crack 2
0 2
0 0
(C) TOTAL 3
0 2
1 0
IOTAL (A+B+C) 28 3
9 14 2
Note:
1.
Average shear load is the average of two adjacent welds and the original specimen.
2.
Three specimens with a shear load below 1,200 lbs., are the sa=e three specimens described in Table No. 3.
Page 9
9 e
4 e
Lc A D v
sis s ".c.c.sv, Tac, 0
}
um
- c. A p sc t tJ T0 6Atf fLATC to omc uomu.
h i
Tc,e wa s.a u N L D.L!
3l l1.
p I
_ Te sT secur w i
t l
ScTToru tVASHOL L
!l i'
6 pot W E.L D l
u i ;-l__ 4 lll I
d t
SAse (i.ATc \\
,1 j
j L___I i
I
- SKElCH.
1 L'
Page 10