ML19326C471

Mill Test Rept for Tendon 32H14, Prepared by Suzuki Metal Industry Co

ML19326C471
Person / Time
Site:	Arkansas Nuclear
Issue date:	04/06/1972
From:	AFFILIATION NOT ASSIGNED
To:
Shared Package
ML19326C456	List: ML19326C455 ML19326C463 ML19326C468 ML19326C471
References
NUDOCS 8004230595
Download: ML19326C471 (30)

Text

e ,3

-

9 4 /j Xj ( (

-

.

4 ATTACHMENT NO. 3 MILL TEST REPORT FOR TENDON 32H14 .

.;.

N e

8004230 7 7 5

- ,

'

f t. . ... , -

G . 6

.

• .- -

. .

( ( f 0. 39.(- (

. , /y- t my

.

. nS pOCi.lon Cerh.f.MCdie

/ '

.

Uncoated Stress Relieved Wire For Prestressed Cone:ete hiessrs Mitcuhjshi Corporat.ictu ,

Order No. *!37-19713

.

dis SUZUKl METAL INDUSTRY CO., LTD.

Quantity 313 coiln '/99,693 1.ba.4# '

,, lil-:An 01TICE & No.1-in Al: AllANT-KITA Dest.lnat. ion w h'I rl O S t en AKAnANE wouKs KITA.utt. TOKYO. JAl'AN

, luspection Date i.pril G, ]972 ,4AltASillNO WORKS NO. 7-7:4) Il!G ASillN AR AS!IINO Shipping h1 ark NAH ASnlNO. CinnA. JAl'AN M'

'

gy, l'.'s- 37 67- B 0g.h - A d

d I'. P. hlanager, Inspection Dept.

E.2 iE

.

s. .'.I!I' f,. /. 1 a. ..,...

a ,

. J. 2 i ll J A P.d; 68E .0021-04

,

ii../1 t i;0. :4 19?) G ')2072 iA). 696-100) 1002-1007 1 Specification : A.S.T.h! A421 -(,5 hh and Prescon's .;pec. iArhannan Job) to Chemical C<nnposition Carbon  : 0.72~0. 03 f,s hlangane:<c  : 0.40 ~1.10fd Phosphorus max.  : 0.01094 Sulfur max.  : 0.05094 Silic m  : 0.10~0.35f4

.

(2) Wire Nominal Diameter  :

yg inch Tolerance of Diameter inch

0.002, -0.001 Application . Type g Tensile Strength  : min. psi 2e.0,000 Yield Strength  : min. Usi 210.Ou0 lilongation in 10"  : min. . f4 Cast I. aid free on'h".9abstantially flat surface, shall form an arc of a circle not less than 12 feet in diameter.

,

2 Test Results .

(O Chemical Composition . f4 i:ha..rge' No - C. Si. I hin. P. S. Cn. InspCclion No.

_ .

."._"3. _." "'"l_ . _ . .

.[ . _ . . _ . . . -.. . _ . . . . -

N 11-2192] - 0.75 0.2S 0.G2 0.019 0.014- O , 02 -- - 695-1001 - --

?

0. 7 ft ._:. _O . '.yL._ O . 7 !L_ O. 019-. 'O . G07_.I 0, 0;'.

- - -

-- G-22072

'

. ~ - -._ ._ . .. .. . . . .- . . . . _ . . . .. .- -- **i

. s -

, .- (

.

RESPONSES TO NRC QUESTIONS ON ARKANSAS NUCLEAR ONE - UNIT 1 CONTAINMENT BUILDING THIRD YEAR TENDON SURVEILLANCE QUESTION I:

The information provided to-date indicates the following varia-tions in atrength may cause concern (see pages 2-1, 2-3, and 4-1 of November 4, 1977 report):

tendon 31H40) tendon 31H39) liftoff forces lower than required tendon 31H50)

, tendon 32H14 - Some wire samples failed at less than guarantee minimum ultimate strength.

This information involves only hoop tendons.

The information given on page 4-2, however, is different. It indicates that the average lift-off force (per wire) for all vertical tendons is approximately +0.41% higher than expected, but that the average value of lift-off forces for both hoop and dome tendons is lower than expected for dome tendons -0.68%

(approximately);' for hoop tendons - 1.1% (approximately)1.

Another contradiction can be found in information provided in Figures 4.1, 4.2, and 4.3. It can be seen that, per wire, tie lift-off force was lower than expected for:

V95 (Vertical) 1D20; 3D21; 2D08 (Dome) 32H24; 32H40; 32H32; 31H39; 31H41; 31H50; 31H40 (Hoops)

I-1. Explain these contradictions and correct them.

RESPONSE

The expected average prestress force per wire is a prediction of the force per wire averaged over all tendons. The average minimum design prestress force per wire is the minimum force per wire, averaged over all tendons, at which the containment is expected to have adequate prestress to withstand the design loadings. The expected average prestress force oer wire must be equal to or

. greater than the average uninim tm design prestress force per wire during the design life for the plant. For purposes of the tendon

~

-

,

-

(

.. .

surveillance test, the average minimum design prestress was taken as the minimum acceptance criterion for tendon lift-off force.

The concern expressed about the lift-off forces for tendons 31D40, ,

31H39, and 31HSO was brought about because the lift-off forces were too low to meet the acceptance criterion used during the first year tendon surveillance in 1975. During investigation of that potentially reportable condition, the lift-off force accep-tance criterion was found to be more conservative than necessary.

A more up-to-date method for calculating normalizing factors also came to light during this investigation. Once the acceptance

-

criterion and normalizing factor calculations were revised (see response to Question III- 6), . the criterion of the operating procedure was met. Thus, all tendon lift-off forces were found to be acceptable.

The average value of lift-off force for each tendon group was found to agree with the expected force per wire to within il.1%

That is excellent agreement of measured forces with predictions.

Since the expected average prestress force per wire is based on an average over all tendons, havint a number of tendons with some-what lower force than the expected average prestress force is antici-pated and is not a contradiction. (See Attachment No. 1 for discussion of expected average prestress force. )

Explain the significance of lower lift-off forces for

~

I-2.

the stability of the structure.

_.

. . -

9

• ( (

.

RESPONSE: ,

The lift-off forces measured du:ing the surveillance test hiid a satisfactory average value for each tendon group, and all indivi-dual tendon lift-off forces were greater than the recalculated minimum acceptance value (average minimum design prestress).

See response to Question III-6. Therefore, the tendon lift-off forces are satisfactory, and no adverse structural effect is indicated.

+

I-3. Explain the significance of ultimate strength of wire samples lower than the guaranteed minimura strength for the stability of the structure.

RESPONSE

At the end of 40 years all anticipated losses will have occured and the average stress in the tendon will be 55% of its ultimate strength. The containment is designed for a pressure of 1.5 P.

Therefore the containment has an ultimate capacity to resist the accident condition, excluding the effect of the liner plate and reinforcing steel, of 1.5 x 1l00=2.73 times the design pressure given in the FSAR. If the cal 'ty of the liner pl. ate and the reinforcing steel be included the factor would be in excess of 3.0.

Twenty wires have been tested during the first two surveillances for tensile strength. Only one wire broke below the Ainimum guaranteed strength. Ar discussed in the response to question II this is attributed to the method of wire removal.

. . ( I

.

For the purpose of demonstrating the effect of a reduction in wire ultimate strength on the safety of the containment the above sample will be used. ,

The reduction of ultimate wire strength for the single wire is 14%. When added to the total wires tested the average ultimate strength, using the guaranteed minimum strengtl. for the other 19 wires, the reduction becomes less than 1%. As can be seen this has virtually no effect on the ultimate capacity of the containment.

QUESTION II:

In your letter dated August 24, 1977, you stated that 46 tendons have the same heat of material as tendon 32H14. In the November 4, 1977 report, the problem of the heat is not mentioned. Explain this contradiction and possible omission of the heat problem.

RESPONSE

Based on additional tests, both physical and chemical, and the results from physical tests done on the same heat of material from another tendon during the first year surveillance, it was determined that the low breaks were a result of deforming the wire during wire removal. The Bauschinger effect discusses the problem with erratic results where the wire has been stressed beyond yield in one direction and then tested to failure in the opposite direction. (See Appendix I of the Report.) ASTM 421 states that samples used in tensile test having a permanent radius of curvature less than 12 feet shall not be used for

I tensile tests.

l l

. . ( I

.

Since it was not established that this heat of material is below the minimum guaranteed ultimate strength and previous tests

'

conducted on another tendon of the same heat during the first year surveillance were acceptable, it was not considered a reportable item and therefore not included in the third year tendon surveillance report.

For the purpose of verification a wire from tendon 32H14 will be removed during the fifth year surveillance for test purposes.

QUESTION III:

The documents furnished suggest that the accep ance values and the normalizing factors used in the one-year surveillance in 1975 be modified and that after such a revision the normalized tendon lift-off forces will exceed the required minimum values.

This approach is questionable, si nce one of the goals of succes-sive surveillance operations is to establish historic continuity in the evaluation of the safety of the structure. A modification of the basic criteria will destroy this continuity. It is imper-ative first to establish the significance of not satisfying the original criteria and normalizing procedures. To clarify this problem, answer the following questions:

III-1. It is indicated in your letter of August 24, 1977 that the lowest value recorded and accepted as valid for the ultimate strength of the wire samples was 229 ksi; the highest was 246 ksi. The required ultimate strength of the wires required in ASTM A-421-74 is 240 ksi. The maximum deviation was, therefore, approximately -4.6%.

It is also indicated that the stress at elongation of 1% was less than the required 192 ksi. In addition, some samples with ultimate strength lower than 240 ksi had an elongation smaller than that required by the specification of 4.0%. The wire manufacturers have a great amount of statistical information available.

Therefore, present a study of the significance of these deviations. (See pages 5-2 and 5-3, Tables 5.1, 5.2, 5.3, and Appendix G.)

.

. . ( (

, RESPONSE:

The following firms were contacted for information on wire.

1. Prescon - San Antonio, Texas

2. INRYCO - Melrose Park, Illinois Prescon, in addition to their Japanese source, uses Armco wire on occasion. INRYCO uses both Armco and Florida Wire and Cable

-

as their domestic wire sources, in addition to occasionally

"

using a foreign source. INRYCO indicated that they prefer to use domestic material suppliers.

.

This was discussed with Armco and Florida Wire anc Cable. Both companies indicated that they use their own company specifica-tions to buy rod stock from other producers. When they commence wire drawing operations, tensile specimens are taken from the beginning and end of each wire spool.

Both companies claim to have very low rejection rates. As a consequence, they only keep raw test results on file, and have not performed any statistical analyses. Furthermore, both companies have other product lines so that when any tendon wire is rejected, it is downgraded and diverted to another product because it is too expensive to attempt any reprocessing of the tendon wire back through their system.

.

-

.

.

- - ( (

In view of the lack of statistical data from these sources we cannot do a study on the significance of the deviations from ASTM A421 of the tendon wire samples tested during the last surveillance inspection.

However some statistical data furnished by the vendor when he submitted his original bids is available. In their proposal under a section giving engineering data they discussed a series

,

of tests covering 330 to 340 coils of wire representing 25 heats of material. They found from the results of tensile test that

.

the standard deviation was approximately 1% to 2%.

.

III-2. There is a confusion between " sample" length and " gage" length. For example in the letter dated August 24, 1977, the expression "100 inch sample" is used; also on page 2-3 of the report the expression "10 inch samples" and "100 inch samples" are used. In Tables 5.1, 5.2, and 5.3 in Appendix H the expression " measure gage length" is used throughout. Clarify these contra-dictions.

RESPONSE

The terms " san.gle length" and " gage length" are used inter-changeably. In either case, the distance between the gripping mechanisms on each end of a tensile specimen is designated.

III-3. Explain the meaning of the expression " guaranteed minimum ultimate strength" (pages 5-3 and 5-4). What are the tolerances and permitted deviati>ns on the

" guaranteed" values?

.-_ ___--

__

.

'

( I

RESPONSE

.

" Guaranteed minimum ultimatet s'rength" means that a coil of wire is not acceptable for tendon fabrication if a tensile test indicates ultimate strength of less than 240 ksi for any test sample from the coil.

As discussed in the response to Question III-1, statistical data on rejection rates has not been maintained by the manufac-turers contacted. However, data are available for the heat of material used in fabricating tendon 32H14.

Tensile tests were made on 31 samples with a low break of 245.0 ksi and a high break of 258.0 ksi.

III-4. Clarif dures.y the meaning of " improper" wire removal proce-If the reason for the difficulties is in the explain why this procedure did not damage the wire in vertical and dome tendons. See page 3 in Appendix I.

The mention of "Bauschinger Effect" may be inappropriate since this effect has been mostly investigated for uni-axial tensile tests or torsion and the problems presented in this documentation are, much more complex (see page 5-4 of the report and page 3 of Appendix I .)

Describe the future wire removal procedure as planned at this time.

RESPONSE

In this case the " improper" wire removal procedure caused per -

l manent bending of the tendon wire from the hoop and dome tendons .

The wire from the vertical tendon was removed without permanent deformation.

The tensile tests for samples from the dome tendon indicate ultim~ite strength of at least 240 ksi. That proves nothing one say or the other concerning the Bauschinger effect

_

,

.

. , ( (

.

explanation for low tensile test results. Qualitatively, a reduction of ultimate tensile strength would be expected as a result of permanent bending. Quantitatively, the reduction may or mey not be great enough to cause low ultimate strength test results.

For future tests, wire' removal procedures will be specified on a per,casq hasis and any wire removal procedure causing permanent wire deformation will be considered as unacceptable. s III-5. In several locations in the report, it is indic'ated that some buttonheads are "offsize." The original design of buttonheads by the manufacturers is very sophisticated. Therefore, explain what is meant by "offsize" and justify the statement. that this did not influence the strength of the tendone. (See page 2-1 and Appendix D.) Discuss other imperfections of the buttonheads, such as cracking, offset, etc.

RESPONSE

"offsize" buttonhead refers to a buttonhead which is either larger or smaller than that shown for the GO-NO GO GAGE. This inspection is not intended to be a basis for acceptance or rejection of the tendon but rather a recording of the as-built condition.

After a tendon is detensioned and inspected it is retensioned to 80% of the minimum guaranteed ultimate strength based on the  ;

total number of effective wires in the tendon. This is in accor- i l

dance with Section 8.1.5 of the Operating Procedure 1304.91, l Appendix A of the report. The tendon is then shimmed to a force

. - -

. ,

.

, ( l

.

equal to/or larger than the lift-off force determined at the time of detensioning the tendon in accordance with Section 8.'. 6 of the above procedure.

By subjecting the tendon to a stress in excess of the required lift-off force without a buttonhead failure, the adequacy of the wire anchorage has been established.

In Section 3.2 anchorage components of the Three Year Surveil-lance Report, it is noted (paga 3-2) that no split buttonheads were found.

" Offset" buttonheads refers to buttonheads which are not concen-tric with the wire.

III-6. On page 2-2, it is stated that with two revisions of the basic criteria, all the tendons are considered acceptable. This statement seems to be inappropriate since, as indicated above, the use of new criteria '

may be objectionable in itself. Therefore, qualify this statement.

RESPONSE

Containments are designed based on the average tendon which is subjected to an average elastic loss. Losses due to friction, elastic concrete creep and shrinkage and steel relaxation are l

considered when originally determining the number of tendons required. l The amount of existing prestress is determined by obtaining lif t-off measure-

!

ments at the tendon end anchor. The end anchor force is subjected to all losses except friction. Based on past field experience and te'sts performed on concrete for treep and shrinkage and on steel wire for relaxation, it has been '

l

- - . , .

.. . . __ - -

.

. ( (

.

determined that approximately 70% of the losses occur during the first year and the following expressions can be used to determine the average wire force which should exist after one year and at the end of life. If exactly the required amount of

~

prestress has been supplied, then the end of life value and the minimum required value will be the same if the surveillance acceptance criteria is the same as the original design criteria.

Wire force 1 year after post-tensioning:

[ Creep, Wire _ Wire _,.fInitial_[Averagei h[

_

l force area stress elastic / _(.70)L shrinkage & (1)

_ _ (loss ( relaxation loss)l _

.

,

Wire force end of life:

e

[ Creep, Wire,~/ Initial) [Avdrage)

._

Wire h~

_ elastic l 1 shrinkage & (2) force - area

_

stress j loss

) (relaxation loss Since all tendons cannot be anchored at exactly the same value

and at the same time, then a correction for the initial anchorage value and the elastic loss must be made so that the value will be typical of the entire average tendon population. This is

!

done by. determining a normalizing factor by which the lift-off j value is multiplied when making corrections:

i

,

. - .

! Average initial Average _

anchor force -

elastic Normalizing = .all tendons . lord factor Actual -

. -

(3)

' Actual-l tendon -

elastic

<-

lift-off 1 loss -

--

_. . - _ , _ _ _ _, ,- -. __ . . . _ _ _ _ - . _ ._ _-.

-

.

.

-

. ( (

.

When reducing the data from the first surveillance for presenta-tion, a technique similar to Eqs. (1) through (3) was used. How-ever, when applying Eqs. (1) and (2), the average elastic loss was not subtracted and the required values were too high. When using Eq. (3), again, the' average elastic loss was not subtracted 3 in the numerator and the normalizing factors were too high .

When reducing the data from the s w ond surveillance, the elastic loss was considered.

In addition, the method of determining friction which used the tendon end values was compared with presently used methods which use a weighted-average value and it was determined that the friction forces were slightly overestimated, and therefore .ess tendons than were put in, would actually satisfy the design criteria.

-

Since these tendons were put in, then it is acceptabic to have a lower minimum required wire force level . All these changes were incorporated and included in the second surveillance report.

For additional information, see Attachment 1.

Since the 1.ower limit is an average value for all tendons sur-veyed it follows that some tendons may fall below the average minimum design prestress at some point in time. Therefore it is important that this value be set to allow for individual tendons to fall.below the average minimium design prestress by an acceptable margin.

.

, ( (

.

-

Operating Procedure 1304.91 does not specify a margin therefore when the low lift-off values for the tendons discussed in Section 2.1 of the report were discovered, an engineering evaluation was made. The evaluation showed that when using the current method for calculating friction, as mentioned above, the average minimum prestress per wire could be lowered and still satisfy the den s..

criteria. The method for calculating friction is similar to that given in BC-TOP 5. (Example for Hoop Tendons.)

With the two changes discussed above the acceptance criteria given in the Operating Procedure was met,.

III-7. On page 2-2 of the Report, it is indicated that the "available" equipment could not perform adequately the wire continuity test. Explain why the needed equipment was not available for this important test.

RESPONSE

The wire continuity test could not be performed adequately because the stressing washer could not be pushed along the tendon the 2 to 4 inches necessary to allow grasping each wire with a special tool for pulling on each wire. A number of different ways were tried to push on the stressing washer; none of them worked. All the equipment corisidered standard in the tendon surveillance business was on hand, but it simply wasn't adequate. In addition, a hydraulic vibrator was used, but it was unsuccessful also.

In all cases where the~ stressing washer could not be pushed far enough to allow the test to be run, the wires were checked by one of the two methods listed on page 2-2 so that the continuity of all wires was verified.

.

-

( (

III-8. "No corrosion" is indicated on page 2-2. Explain the procedure establishing this negative condition, espe-cially whether the procedure used permitted the dis-

RESPONSE

covery of any small pitting. See also Appendix G.

The anchorage hardware was cleaned with solvent, brushes, and rags, then it was visually inspected for corrosion. Corroded areas with diameter or width greater than about 1/50 inches would have been detected.

,

III-9. Discuss in detail the testing of wires. You should cover all important facets of the testing procedure including, but not limited to:

a. failing inside of gripping jaws. Influence of the shape of the jaws.

b.

the influence of curvature and twisting of tendons as installed on the resistance of wire- to removing, which may have predeformed the wires in a complex way.

c. rate of speed of the testing machines.

d. ambient temperature during testing.

e. possible eccentric loading by the machines, bend-ing or twisting of the wire samples,

f. influence of the type of machine.

g.

comparison of testing methods used by the manu-facturers, by Pittsburgh Testing Laboratory, by Hales Testing Laboratory and by any other organ-ization participating in testing of wire samples.

Conformance with ASTM specifications.

h. influence of the temperature of filler and its pressure.

i. influence of the average temperature of tendons

.

in place.

j.

machine calibration.

.

i

(

1

RESPONSE

.

III-9a The procedure used for field testing of wires is given in the Attachment No. 2, " Procedure for Testing of 100-Inch Gage Length Wire specimens."

The wires tested by Pittsburgh Testing Laboratory were gripped in serrated jaws. Wire failure inside the gripping jaws indicates that the jaws caused stress concentration and the shape of the jaws affects any stress concentration. Therefore, the break strength in these tests is, possibly lower than the actual ultimate strength of the wire material, b

The tests by Hales Testing Laboratory were conducted

. .

on buttonheaded wire samples '.ollowing a procedure in accordance with ASTM A-421.

III-9b. The force needed to remove the wires is less than 3000 lbs. (25% of ultimate) . Therefore, deformation because of curvature and twisting of the tendon is not likely.

III-9c. In all cases, the testing machine speed was in the range which is normal for steel tensile tests as specified in ASTM E8. Therefore, rate of loading had no influence on test results.

_

,

. - _ _ - _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . - _ _ - _ _ .

_

'

. < (

III-9d. Although ambient temperature was not recorded, all wire tests were done with ambient temperature between 50*F and 85*F. Therefore, ambient temperature had no influence on test results.

III-9e. Since a buttonheaded wire is self-centering loading from bending and twisting did not bring about premature wire failure. Bending or twisting for the tests by Pittsburgh Testing Laboratory would not be applicable for the reasons discussed in the response to III-9a.

III-9f. The only influence which the type of testing machine could have is in the calibration. The hydraulic cylinder used for the field tests was calibrated

.

before and after the wire tests, and the testing

, machines used by Hales and by Pittsburgh were certi-fied as having been calibrated according to good materials testing laboratory practice as specified in ASTM E4.

III-9g. All tensile tests were done in accordance with ASTM A-421 with two exceptions for the field tests:

(1) Sample length of 100 inches was used instead of i

10 inch length. The longer ~ sample length was used  ;

i to comply with the requirement for the longest prac- l tical sample as stated in Regulatory Guide 1.35. I

)

1

.

._

.

I ()

(2) The extension under 16ad was measured with a device having

'

the smallest division equal to 0.0003 in/in of gage length instead of 0.0001 in/in specified in ASTM A-421 for reasons of practical field application.

Tests performed by Pittsburgh Testing Laboratory used jaws to grasp the wires while Hales Testing Laboratory installed button-heads on the wires and the tests were performed with much better results.

• III-9h. The filler temperature during installation is high enough to make the filler quite fluid (115 F to 2000F) but not hot enough to affect the steel heat treatment. After installation, the filler temperature is in the range 300F to 100 F; that would have no affect on the steel heat treatment either. The filler pressure has no effect on the mechanical properties of the tendon wire, 111-91. The average temperature of tendons in place was within the range 300F to 1000F at all times. Therefore, the in-place temperature

-had no influence on wire strength.

III-9j. Wire test machine calibration was addressed in response to Question'III-9f.

III-10. Add to information provided regarding filler (page 3-1 and Appendix C) a discussion clarifying the problem of shrinkage of the filler when cooling (outside of trumpets). What procedures have been used to check on possible voids at intermediate points in tendons due to the cooling shrinkage?

-_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

.

- ( l

<- 1

RESPONSE

The filler material is installed under a pressure of about 100 psi and at a temperature of around 115'F. The sheathing is vented at the high points and drained at the low points thus ensuring that the sheathing is filled and the wires fully coated. As the

How-filler cools it will shrink about--5% by volume and voids may occur.

ever the wires will remain coated. There are no provisions for physically determining the presence of a void within the sheath-ing. However, during the surveillance, wires are removed from each family of tendons and inspected for corrosion as well as the tensile and elongation tests. See Section 7.2 of the Operat-ing Procedure 1304.91, Rev. 2. All of the inspection wires in the third year surveillance were classified as #1 corrosion level From bright metal, see Appendix G, Wire Inspection Data Sheets.

this it is concluded that the filler material is performing its intended purpose, that of inhibiting corrosion of the tendon wires.

III-ll. Indicate whether local bending of bearing plates has been checked.

RESPONSE

The bearing plates have been visually inspected, and no deforma-tion has been observed.

III-12. Indicate how accurately the absence of pitting in anchor hardware has been observed, and the minimum size observable.

-. -- _

--. ,

.

.

4 (

"

RESPONSE: '

See response to Question III-8.

III-13. In Appendix B the normalizing formulae are presented.

Discuss the possibility that factors neglected in these formulae may be more important than factors which have been included. (See Appendix B and page 4-1.) Some of the neglected factors are:

-

First surveillance formula does not include con-crete creep and concrete shrinkage or thermal effects and concrete placing variation.

-

Integration along the tendon may introduce ques-tionable properties of materials.

-

Subsequent surveillance formula do not appear to include the following: thermal effects, shrinkage, detailed effects on creep, bearing plate displace-ments, changes in concrete, Young's elasticity modulus and Poisson's ratio, jack orientation, cracking of concrete.

-

Indicate the tolerances in normalizing factors.

RESPONSE

.

Attachment 1 gives extensive information on how the containment posttensioning requirements are originally determined. Also discussed is the method of determining predicted values, normalizing factors and a rational acceptance criteria.

III-14. Tables 4-1 and 4-2, pages 6-1 and 6-2 and Appendix B indicate that the elongations have been measured between the outside face of the bearing plate and the inside face of the anchor head. Discuss whether the precision of these measurements is sufficient to establish the adequacy and the correctness of lift-off forces.

l l

l 1

!

. . - -

.-

.

-

I (

_ RESPONSE: -

Since the lift-off force is measured independently using a cali-brated hydraulic ram, the precision of elongation measurement has no effect on " adequacy and correctness" of lift-off forces and used as an order of magnitude verification.

III-15. Discuss ram calibration. In page 4-3 of the report it is indicated that load cell calibration is " traceable" to the National Bureau of Standards. This expression is too vague. Indicate whether the NBS did in effect calibrate the load cells and if not, what is the signif-icance of this procedure for the reliability of the calibration. It is also indicated in the same page that the uncertainty of the measurements is O.K. (two times standard deviation). Define the standard devia-tion for such a small number of measurements and dis-cuss the basis for accepting an uncertainty of twice the standard deviation in this case. Discuss also whether the participation of NBS, University of Cali-fornia, University of Arkansas, Wiss, Janney, Elstner

'

and Assoc., Inc., and Zabel Calibration Service in different phases of the calibration operations leads to compatible results. (See Appendix F.)

RESPONSE

Tendon Stressing Rams The rams used for measuring lift-off force and restressing the tendons were calibrated using a load cell belonging to AP&L. j l

In January 1977, the load cell was calibrated by University of California using a testing machine which was certified as having calibration last done in November 1976 using load cells and a  ;

i dynamometer ring calibrated by NBS in November 1969, April 1971, and April 1975. Calibration of the rams allows determination of the multiplying factor for converting hydraulic pressure measurement to ram force. To accomplish a ram calibration, nine

.

.

,

.

( (

sets of pressure vs. force readings were taken for each of nine pressures. The scatter of those data is the measurement uncer-tainty. The standard deviation is one possible measure of that uncertainty. To calculate standard deviation, the usual algorithm was used:

2 S= IX -

(IX)2 h n-1 where S is the standard, deviation, X is a data point value, and

-

n is the number of samples.

i By stating the uncertainty as two standard deviations, the impli-cation is made that any single measurement made with that ram has a 95% chance of being within a band of 12S about the indicated force.

Wire Test Ram The wire test ram used in the field was calibrated before the test by University of Arkansas and after the test by Wiss, Janey, Elstner and Assoc., Inc. The two calibrations agreed within .. 3%.

That agreement by two independent calibrations should be taken as verification of the accuracy of ram calibration. Zabel Cali-bration Service calibrated the load machine used by Wiss, Janey, Elstner and Assoc., Inc. to calibrate the wire test ram.

l l

III-16. On page 5-1 of the report it is stated that 13 wires from 23 surveillance tendons showed anomalies which were judged to be inconsequential. This is presented

.

.

( (

.

.

in some detail in 8 paragraphs on the same page. How-ever, this information appears to indicate lax quality control during erection and not sufficient attention provided during previous surveillance operations.

Discuss this problem.

RESPONSE

The two wires identified as missing from tendon 31H40 were reported discontinuous and removed during the one year surveillance. The construction records of tendons 21H45, 32H40 and 32H48 showed that a wire was removed for inspection during the installation process. During initial installation the stressing operation was suspended for an extended period due to a failure of the stressing ram. Upon resumption of the stressing operation a 1

wire was removed from some tendons for corrosion inspection. Of the remaining seven discontinuous wires, three failures were directly related to construction or initial posttensioning.

The remaining four wires failed either during the three year time interval'following initial posttensioning or during surveil-lance retensioning. This represents a failure rate of 4/4255 or 0.09% of the population inspected. The observed failures are therefore judged inconsequential to the overall tendon system.

(See response tc Question I-3)

III-17. On page 5-4 of the report a brief discussion is given of the metallurgical investigation, presented in a more detailed way in Appendix I. The appropriateness of including the use of 12 inch sheave and the "Bauschinger" effect as explanations of the weakness of some wire samples has been quectioned above.

a. Discuss the manufacturing, testing, and quali+.y controls of Suzuki, Limited, also the possible effect of transporting the wires through long distance shipping.

-_- _ - _ - _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _

(  !

b.

Explain whether the fact presented in page 2 (Appendix I) that one wire showed evidence of rust on wire didthe failed not, end, whereas the other failed to be expected,corresponds to a normal condition, or may indicate some special circumstance causing different behavior of the two wires.

RESPONSE

III-17a Suzuki has been asked to furnish information on their manufacturing processes, testing procedures, standards of acceptability and quality control practices. This information has not been received as of this submittal.

While the method of transporting the wire may have some effect on the wire, quality control procedures upon receipt and prior to fabrication would have disclosed coils of wire which did not meet the specification requirements.

A sample was taken from each end of each coil of wire and tested prior to releasing the coil for fabrication. Therefore the effects of improper shipping methods would have been revealed an,d the wire rejected prior to fabrication.

III-17b The evidence of rust on the one wire fragment was a result of handling and storage after the wire test.

a It had no significance with respect to the wire test results.

c III-18.

Appendix A describes the Surveillance Procedure. State whether this procedure is in accordance 4_ith Regulatory Guide 1.35, with the ASME/ACI Pressure Vessel Code,

_

Section III, Division 2, with the Technical Specifica- i tions and with Bechtel's Topical Report BC-TOP 5 '

h

i (

l

,

RESPONSE

The Surveillance Procedure described in Appendix A was checked for compliance with the Technical Specifications and submitted i

to the Plant Safety Committee during the preparations for the l

!

surveillance. The procedures are in accordance with Regulatory Guide 1.35 and Bechtel's Topical Report BC-TOP 5.

III-19. Appendix E presents data sheets on retensioning, re-tensioning force and elongations. Discuss the possible errors in this operation and the tolerances which should be used to evaluate the results. On page 4-3, it is stated that Ram #1 failed, was repaired and recalibrated. Discuss the consistency of results obtained by using a recalibrated Ram. Evaluate the possible errors. See Question .15 above on calibration.

(See also Appendix F.)

RESPONSE

Retensioning forces are subject to the following errors:

Nominal force Systematic Uncertainty Error (%) (%)

1000 lb/ wire -6.6 or less 0.5 80% of Ultimate 0.6 or less 0.5 Lift-off force -0.4 or less 0.5 The systematic errors are taken from the calibration curves

.

shown in Figure F.1 of the Surveillance Report. The uncertainty is two times the standard deviation of the calibration data as discussed in the response to Question III-15.

.

.- t i Elongations were resolved to .020 inch least count and are con- l 1

sidered accurate to within i the least count (i.e., 1.020 inch).

Ram #1 was. repaired and recalibrated, but was not used for tendon work after that. As a point of information however, it could have been used if needed with accuracy equivalent to its per-formance before the failure and repair.

III-20. Appendix H presents data sheets on tendon wire tests.

Discuss the possible errors and tolerances to be applied to results. See also Question 9 above on wire testing.

RESPONSE: . _ , , . .

_

The forces for the wire tests are subject to random errors in using the hydraulic cylinder and to any errors in reading the pressure gage.

Random errors (uncertainty) are of no consequence since the calibration data as given by the University of Arkansas contain no deviation and no scatter was encountered. Pressure gage reading errors are judged to be within 120 psi which is equivalent to 36 lb. force. The error from that source is then equal to or less than 136 lb. or +0.3% of the guaranteed ultimate strength.

The displacement measurement for determining the elastic limit is accurate within 1.001 inch. Since the maximum length measured with this device was 1 inch, the accuracy for that measurement was 10.1%. The displacement measurement for ultimate elongation was accurate within 11/32 inch which is 10.03% of the gage length.

,

_ _ __

.

'

(

III-21. The list of references presented in the report (Appendix I, page 4a) includes only references which date from 1958, 1963, and 1971. Present some references which are more up to date, especially on "Bauschinger Ef fect . "

RESPONSE

A computer search was conducted using the following files:

a) NTIS (National Technical Information Service) b) Metadex (ASM Metal Abstracts) c) Engineering Index No additional published data were found regarding the Bauschinger Effect and its influence on the mechanical properties of wire.

Several references were found where the Bauschinger Effect had been investigated with regard to its influence on the mechanical propertiec of various steel alloys and non-ferrous alloys.

-

,

Abstracts, when available, were reviewed and papers selected which contain useful information.

References on Bauschinger Effect

1. Nakajima, K., et al, Bauschinger Effect in Pipe Forming, Transactions of the Japanese Iron and Steel Institute, Vol. 15, 1975, pp 1-10.

__ - _________ - _ - _ _ _ _ _ _ _ _ _ _ - - - _

i

.

2. ,

Wilson, E. G., Stress -Varied Creep of 20% Cr-P5% Ni-Nb Stabilized Austenitic Stainless Steel, Proceedings of Meeting at the University of Sheffield, England, Septem-ber 20-22, 1972, pp 111-121.

3. Dietrich, Hermann and Schmidt, Werner, Effect of Minor Permanent Deformations on the Mechanical Properties of Steels, TEW Technical Bulletin, Vol. 1, June 1975, Krefeld, German, pp 85-93.

4. Gupta, S. P., Kodali, Surya, P., ~

Effect of Pre-Strain on the Bauschinger Effect, Scr. Metall., Vol. 10, February 1976,

.

pp 111-114.

5. Pepe, J. J.,

The Cocuression and Tensile Behavior of a Cold-Worked ~High Strength Steel, Report No. WVT-TR-74004, February 1975, Watervliet Arsenal, New York.

III-22.

Indicate (Appendix I) whether the chemical analysis agrees with Suzuki results.

RESPONSE

The chemical analysis done by Anamet Laboratories agrees very closely with tF chemical analysis contained in the Inspection Certificate furnished by Suzuki Metal Industry Co. , Ltd. as a part of the documentation package. A copy of this documentation is enclosed as Attachment 3.

I l

l l

- .

_ _ _ _ - --__ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ - - _ _ _ _ _

. -

( (.

III-23. Conclusions are-presented on page 2-4 of the November 4, 1977-report. Modify these conclusions in accordance with the answers to these questions.

RESPONSE

During the process of at.3wering these questions, no information came to light to cause a change of the fundamental test conclu-sion. No abnormt.1 degradation of the containment structure is indicated.for Arkansas Nuclear O.ne - Unit No. 1.

i

,

b

,

. )

'

.

,

4

'

. _ _ . _

- , - _ _ -. _ -