ML20086M961

Evaluation of Containment Coating Acceptance Tests

ML20086M961
Person / Time
Site:	Wolf Creek
Issue date:	12/31/1983
From:	Daru V BECHTEL GROUP, INC.
To:
Shared Package
ML20086M950	List: ML20086M947 ML20086M961
References
3871, NUDOCS 8402170162
Download: ML20086M961 (18)

Text

,'.,

Attachment to KMLNRC 84-013 EVALUATION 2/10/84 0F C0flTAlfiMENT COATING ACCEPTANCE

, TESTS

~

g.

w s

4 Report for llolf Creek Project 4

./

Prepared for J. H. Smith By __ -

V. C. Daru

9- T Approved T F, \,M.fp.M M. F. Stucnfielav Area Manager Materials and Quality Services Department Research and' Engineering BECHTEL GROUP, I;iC GAITHERSBURG Job No. 10881-021 Log No. 3871 <

December 1983 m

w .

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

ABSTRACT The Containment Coating Acceptance Test Program was designed to provide assurance that the coatings applied prior to July, 1980, to the surface categories outlined below,.will perform satisfactorily in service, and meet e

1 the, requirements of Specification No. 10466~A-125(Q), Technical Specification for Surface Preparation, Furnishing and Application of i

Special Field Coatings and Surfacers for the Standardized Nuclear Unit ,

Power Plant System (SNUPPS). The categories of affected areas are:

i

(1) Containment Wall Liner ' Plate (2) Reactor Cavity Pit Liner Plate 4

l (3) Containment Dome Liner Plate

~

, (4) Structural Steel Members -

(5) Maintenance Truss I

J n

i 1

t L

o e

c 1

~

4

- ;1 l- .

l

EVALUATION OF C0tiTAltiMENT C0ATING ACCEPTANCE TESTS

,, CONTENTS Page ABSTRACT i

C0!1TEliTS ii LIST OF TABLES iii 1.0 ItiTRODUCTION I

2.0 CONCLUSION

S 2 3.0 ANALYSIS OF SAFETY IMPLICATI0t!S 6 4.0

SUMMARY

OF It1VESTIGATI0tl 8 APPEtiDIX 'A' 14 I

-ii-

NJ,

% J LIST OF TABLES -

TABLES Page TABLE I PHASE II SUICIARY OF FIELD TEST RESULTS 13

-iii-k

- l

.1. 0 INTRODUCTION 1.1 Prior to July 1980, the Constructor at the Wolf Creek site performed safety related coating work with an unapproved procedure. Submission of this procedure t, and approval by the Architect / Engineer (A/E) is a requirement of Specification No. 10466-A-125(Q), to which the Constructor was working. The Constructor subsequently submitted the proceuure for review, it was brought to an acceptable level, and approved by the A/E.

1.2 There are five categories of material on which coatings work to the un-approved proci. dure was performed by the Constructor:

1. Containment wall liner plate

2. Reactor cavity pit liner plate

3. Containment dome liner plate

4. Structural steel members

5. Maintenance truss 1.3- All five categories of material received an initial application of inorganic zine primer by suppliers to the requirements of Specification 10466-A-122(Q), Shop Priming of Structural Steel with Inorganic Zinc Ethyl Silicate Coating. Dimetcoat 6 (D-6), an inorganic zinc primer, made by Ameron Protective Coatings, was applied to the containment wall liner plate, dome and reactor cavity pit, while Carbo Zinc 11 (CZ-ll) made by Carboline Co., was applied to the structural steel and Maintenance Truss.

1.4 During review of the Constructor's unapproved procedure and verification documents, it was noted that zinc salts, which occurred on the weathered inorganic zinc primer, were removcd by abrasive Brush-off Blasting to SSPC-SP7, and reprimed with the same inorganic zine primer, prior to finish coating ir, accordance with Coating System 101 of Specification No. 10466-A-125(Q).

1.5 Further analysis revealed that there was no supporting evidence- that Carboline or Ameron had recommended, tested, or approved brush blasting as a method of cleaning for nuclear safety related coatings.

1.6 In addition, coating inspection reports were found to be in non-compliance with the requirements of ANSI N101.4. In order to verify that these deficiencies would have no impact on the safety related functions of the plant, Specification No. Technical 10881-A-021(Q)

Specification for Containment Coating A: ceptance for -Kansas Gas and Electric Co., was prepared by the A/E to provide criteria for testing and

. evaluating the acceptability of the coatings. applied prior to July, 1980.

An independent testing agency, KTA-Tator, Inc. of Pittsburgh, PA, was em-ployed by the Client, and assigned responsibility for performing the tests to the requirements of Specification 10881-A-021(Q).

. 1_

L _

I

1

. l e

1.7 The Containment Coating Acceptance Test program was divided into two phases, o Phase I consisted of testing coated panels in the laboratory and at the site. These panels were prepared to establisn acceptance criteria for the containment coatings applied prior to. July, 1980, by compari-son of adhesion test values obtained in the laboratory and at the

, site, followed by testing of duplicate laboratory test panels to the project Desion Basis Accident (DBA) criteria at Oak Ridge National Laboratory (ORNL).

The three coating systems used for preparing the laboratory test panels were as follows:

a. Ameron D-6/Ameron Epoxy Finish 90 for Containment Liner Plate,

b. Carboline CZ-ll/ Carboline Epoxy Finish 191HB for Structural Steel members.

c. Carboline CZ-ll/ Carboline Epoxy Primer 193LF/ Carboline Epoxy Finish 191HB for the Maintenance Truss.

Test panels were prepared under a controlled environment using four known variables for the coating systems a and b above. These variables were brush blasting, relative humidity, thinning and curetime.

o Phase II consisted of the following:

a. A statistical sampling plan, with a reliability factor of 95 percent, was designed to test and evaluate the quality of applied Ameron coatings for the Containment Vertical Liner Plate and the Containment Dome Liner Plate.

b. A non-statistical verification plan was designed to test and evaluate the quality of applied Ameron coatings for the Reactor Cavity Pit and applied Carboline coatings for the Maintenance Truss and Structural Steel members.

1.8 The results from both phases of the test program were published in a report

.. issued by KTA-Tator, Inc. ' Coatings Evaluation Report, Number 21777-23, Revision 2.

2.0 CONCLUSION

S 2.1 The conclusions contained herein are the A/E's interpretation of data and results obtained from the Phase I and Phase II test programE. reported in the KTA Coatings Evaluation Report Number 21777-23, Revision 2~

b i

2.2 PHASE I TEST PROGRAM

a. Acceptance Criteria Adhesion test values obtained from the Phase I test panels, prepared at 88-100 percent relative humidity, and 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> DBA test results of the duplicated laboratory panels were used to establish the acceptance criteria for the Phase II tests. Acceptance of the coating systems performance under DBA conditions was based on the 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> test period which is consistent with the criteria established by AriSI fl101.2-(1972).

o Elcometer Adhesion Test The 300 psi Elcometer adhesion test value selected for the Phase II test program is 100 psi higher than the minimum value recommended by AfiSI fl101.2 (1972), which is endorsed by Regulatory Guide 1.54. The lowest adhesion value that qualified, under DBA conditions on the duplicate panel was 216 psi. Even though the average adhesion values obtained from the Phase I test panels, prepared at less than 100 percent relative humidity, were acceptable, the duplicate laboratory panels failed to meet DBA acceptance criteria. Test data indicated that the failure was a result of insufficient curing time for the inorganic zinc _ primer prior to application of the topcoat.

o Cross-Hatch and Knife Tests The lowest value for the cross-hatch test method obtained under the Phase I, which passed DBA on the duplicate panel, was 4. The lowest value for the knife test method obtained under the- Phase I, which passed DBA on the duplicate panel was 5. The minimum criteria for both the cross-hatch and knife tests methods was established as 5.

It was felt that use of all three adhesion test method would Provide a highter probability of finding the failed coating.

b. 29 Day Immersion Tests The extended 29 day immersion tests decreased the performance of both the Carboline and Ameron coating systems that passed the 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> DBA conditions. The decrease in performance was noted by the formation of intact blisters. -There were no detachments reported under the 29 day extended immersion test for the coating systems whose adhesion and DBA test results were used to establish the acceptance criteria for the Phase II program. Detachment occurred for coating systems where-i adequate curing was not obtained at lower humidities. The 29 day extended immersion test provides valuable information regarding the serviceability and performance of the coatings that may be encountered after DBA conditions.

(

c. Effect of Brush-off-Blasting The use of Bruch-Off Blasting, cleaned to SSPC SP7 between the two coats of inorganic zinc,. prior to application of the finish coat, neither affects nor changes the adhesive qualities and the performance of either the Carbolin'e or Ameron coating systems under DBA conditions. The overall performance was found to be comparable to a single coat of inorganic zinc. The residual dry film thickness left after the Brush-off-61ast did not affect the adhesion values.

d. Effect of Thinning The addition of either 12.5 or 30 percent thinner during the reapplication of inorganic zinc did not affect the adhesion properties or the DBA; '3st results.

e. Effect of Relative Humidity' ~'

Although the relative humidity (RH) during the laboratory panel preparation was not controlled to the precise ranges specified in Specification 10881-A-021(Q), due to unventilated RH cabinets, the data obtained was adequate to bracket the documented field conditions for the cabinets, the data obtained was adequate to bracket the documented field conditions for the Phase II program.

Inorganic zine curec at icw RH prior to the application of the finish coat resulted in inconsistent data with both passing and failing results under DBA conditions, even though most of the adhesion test values on the duplicate panels were found to be within acceptable limits.

The Ameron coating system began to achieve consistently good results at 88 percent RH and above, wnile the Carboline coating system became consistent at 100 percent RH. This indicates that Ameron D-6, with amine stabilizer, cured faster than acid stabilized Carboline CZ-11.

In both cases, the test results indicated that in order to perform satisfactorily under DBA conditions, high humidity and longer curing time for inorganic zine is necessary prior to application of the finish coat.

f. Effect of Dry Film Thickness,es (DFT)

Although the DFT's for the _ inorganic zinc primer and the finish coats on the test panels were not maintained 'within the limits of Specification No. 10881-A-021(Q), the DFT's were found to be closer to the actual CFT's measured during the Phase II Program. No_ definitive trend could be established to indicate'a correlation between the passing and failing DFT's tested undar DBA conditions. The overriding.

factors appear to be cure time and relative humidity. DFT's tested which successfully passed the DBA conditions were as follows:

A h--

Carboline Coating System:

IOZ Primer (CZ-ll_) Epoxy Finish (191HB)

DFT range 2.0 to 6.4 mils Maximum 9.0 mils DFT range 2.5 to 3.'4 mils Maximum 10.7 mils Ameron Coating System:

10Z Primer (D-6) Epoxy Finish (90)

DFT range 3.3 to 5.1 mils Maximum 11.0 mils

g. Adhesion Test Results Adhesion test results of the test panels made at the jobsita were found to be higher than the laboratory test panels. The Ameron coating system, in both cases,showed more consistency in adhesion test results than the Carboline coating system. The high values are attributed to tha lower DFT of the inorganic zinc applied during the reapplication and/or open air and steam curing, versus the confined laboratory cabinet curing. Both laboratory and site adnesion test results represented good consistency in disbonding. The nature of disconding in the majority of cases, was cohesive. Both the Carboline zinc primer and Ameron epoxy finish were 1000 cohesively di:; bonded.

h. Brush-off-Blasting Versus near White Metal Blasting Comparison of adhesion test results and DBA performance of t)7 coats of inorganic zinc with Brush-off Blast between coats, cured at 100 percent RH, versus a single coat application of inorganic zinc after near Unite Metal Blasting to SP10, beneath tre application of a finish coat, revealed that they are comparable.

Comaarison of these results against the results obtained from the penels prepared at less than 100 percent RH indicated that adecuate curing of inorganic zinc is essential for the performance of the coating system.

2.3 PHASE II TEST PROGRAM

a. Containment Liner (Verticle).and Dome Liner Plate (Categories 1 and 3.

Ameron coating system D6/SP7/06/90 that was applied prior to July 1980 was found to be acceptable based on the following:

(1) . Talysis of adhesion test results reported in KTA Coating Evaluation Report Number 21777-23, Revision 2.

-5 .

\

l i l

., i

, I i

(2) Statistical sampling plan where no .: failures were discovered at 247 designated. test grids. See Appendix 'A'.

(3) Established acceptance criteria for Phase II. See Ta'le b I. All adhesion test results of 247 test grids as reported in the KTA report were fou'nd to be exceed the established acceptance >

criteria for Phase II. 'This indicates that a complete cure of i

inorganic zine was achieved prior to application of finish . coat.

' The measurement of DFT's at these test sites revealed that Ameron zine primer and epoxy finish coat thicknesses did exceed' the ,

required tolerances of Specification No. 10466-A-125(Q).

, b. Reactor Cavity Pit Liner (Category 2)

Ameron coating system 06/90 applied prior to July, 1980 was 4

found to be acceptable based on results reported in the KTA report for 24 verification test sites. See Table I. DFT readings did not conform to the requirement of. Specification No. 10466-A-125(Q).

These readings, -however, are well within the'DFT ranges that have j been tested under DBA conditions and documented by Ameron.

i .

c. Structural Steel (Category 4) 1 ' Carboline coating' system CZ-11/SP7/CZ-ll/191HB was found _to be t

acceptable on_ 69 out of_72 steel members tested. Three members

tested .for adhesion did riot meet.the established acceptance criteria i

for Phase II coating work. The test data indicates that proper curing of inorganic zine was achieved on 69 steel members while inorganic zine on 3 members was not cured prior to application of the finish . coat. 97 out of 141 DFT readings did not meet

, Specification No. 10466-A-125(Q) requirements. 'These- DFT deviations were found in the Carboline inorganic zinc primer and epoxy finish.

d. Maintenance Truss (Category 5) l Carbcline coating system CZ-11/193LF/191HB was found -to be i

unacceptable on the basis of DBA and adhesion test results. This coating system did not . meet ANSIN101.2(1972) and Specification i 10881-A-021(Q) criteria of a maximum blister size of number 4 few, 3

nor did it meet the established acceptance criteria for the Phase II.

3.0 ANALYSIS OF SAFETY. IMPLICATIONS I 3 '.1 Based on the evaluation- of the Phase I and Phase II test results- i

• contained 'in the. KTA -_ report, it is concluded that the Ameron and Carboline coating systems applied tc materials and items in Categories l' thru 5 are not' reportable under the criteria of 10CFR50 Appendix B.

+

" 6-i ,

_ , , - . , . . .#.. - , - , , ,, , . . , .,m m, . -, ,

3 3.2 With the exception of coatings applied to Category 5 and three structural steel members in Category 4, all coatings will perform satisfactorily in service in accordance with the commitment of Section 6.1.2 of FSAR for the following reasons:

a. In all cases ~ adhes' ion ~ test results exceeded the established acceptance criteria.as' required by Paragraph 7.21 of Specifi:stion No. 10881-A-021(Q). '

b. In all cases, adhesion test results exceed the required 200 psi adhesion value of' ANSI N101.2(1972).

c. In all cases, adhesion test results indicated that Brush-off-Blasting does not have a detrimental affect on the performance of coating.

d. In all cases, . adhesion test results indicated that the inorganic zine applied at the jobsite half been cured prior to the application of the finish coat. The statistical sampling process further supports this conclusion for Categories 1 and 3. In Category 1, a total of 123 randomly selected sites were tested and no nonconformances were found. Therefore, there is 95 percent confidence that the actual reliabi.lity level of Category 1 coating statistically falls between 97.59 'and 100 percent. In Category 3, a total of 124 test sites were tested and no nonconformances were

, found.- Therefore, there is 95 percent confidence that the actual reliability level of Category 3 coating statistically falls between 97.61 and 100 percent.

Furthermore as a unit, when Category 1 and Category 3 are combined, there is 95 percent confidence that the W. tical and Dome Liner Plate coating has at least 98.79 to 100 percent of the required quality.

3.3 With regard to the coatir.g on 3 Structural Steel members in Cagetory 4, the defective coatings will be removed and replaced in accordance_ with

, Specification No. 10466-A-125(Q). The unqualified coatings on Category 5 will be enclosed by a protective screen to confine debris, in the event there is a coating failure, and to minimize potential, future maintenance problems during-plant cperation.

3.4 In order to satisfy the documentation requirements additional testing will be_ scheduled to record test results of-higher DFT's noted during the Phase II test program for Categories 1, 3 and 4. These tests will include adhesion and DBA in accordance with Section 4.0 of ANSIN101.2 (1972). These test results, will be included in a supplemental KTA report.

T

.-74

. D 4.0 StJ!EARY OF IfiVESTIGATION 4.1 General

!! hen two coats of inorg.tnic zine primer are applied with Brush-off Blasting in between, thet second coat of primer must adhere to the previous coat. The epoxy finish coat must also adhere to the two coats of inorganic zinc. Phase I of-the test program examined whether or not there is adequate bond strength between each coat of inorganic zinc, and between the inorganic zinc and the epoxy finish coat. These adhesive prcpcrties between coats, and between the coating system and the substrate are measurable by test. Adhesive strength is a result of the following factors:

a. Partial solubility and/or wetting of the previous coat.

b. tiechanical interlocking to a rough surface on the previous coat.

c. Bonding due to chemical or polar attractions of each generic type of resin used in the coating vehicle.

It is also recognized that self-curing, ethyl silicate-based, inorganic zinc is primarily dependent on moisture in the atacsphere to complete its cure by a hydrolytic reaction.

4.2 Description of Deficiencies The following deficiencies and non-conforming items were found during the initial review of the Constructor's coating procedure and his supportive coating work records.

a. The coating procedtre was not submitted for the A/E review prior to its use, as required by the Specification fio. 10466-A-125(Q).

b. No evidence or supportive documents were provided to verify implementation of the coating suppliers' published procedures and recommendations.

c. No DBA test data for supportive documents were provided for Brush-off-Blasting to indicate that the applied coating system would perform satisfactorily under the project DBA.

d. No records were provided'on the amount of thinning used during the reapplication of'the inorganic zinc primer after brush-off-blasting,

e. ho records were provided to verify that the second coat _of inorganic zinc coat was cured prior to the apolication of finish coat.

f. Environmental condition 1 records during the application were incomplete.

(

1 i

I

g. No records were available which described the qualification of personnel a, plying the coatir.3, as required by ANSI N101.4 (1972).

4.3 Phase I Test Program

a. Investigation Phase I of the test-program was developed to examine whether or not the anomolies identifed below would change bond strength characteristics and. affect the performance of the coating systems under DBA environments. The anamolies were:

(1) Brush-off-blasting to SSSP SP7, between the two coats of inorganic z'inc.

(2) Variations in relative humidities (35 to 100 percent, as documented) and time interval prior to application of finish coat.

(3) Variations in thinning (J2~.5 and 30 percent)

b. Laboratory Tests The laboratory test panels were prepared with different variables in an attempt to bracket as closely as possible, the known conditions at the site at the time coatings were applied. These variables were required so that appropriate disposition of the deficiencies noted in Section 4.2 could be made, and acceptance criteria for Phase II coating work could be defined. The variables used were as follows:

(1) A single coat application of inorganic zine and epoxy finish was applied, as required by Specification No. 10466-A-125(Q),

with Brush-off-Blasting to SP-7 of laboratory weathered inorganic zinc followed by a reapplication of inorganic zinc primer, and an epoxy finish coat.

(2) The reapplied inorganic zinc was cured at different relative humidities, ranging from 35 to 100 percent, prior to the application of the epoxy finish coat. The cure time at relative humidities less than 100 percent ranged from 16 to 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> and 48 to 72- hours. The cure time for 100 percent relative humidity was approximately 2 weeks, to simulate the complete cure conditions documented _at the site.

(3) The reapplication of inorganic. zinc primer was performed at 2.0 to 4.0 and 3.0 to 6.0 mils dry film thickness. The epoxy finish coat was held to 4.0 to 6.0. mils, as required by Specification No. 10466-A-125(Q).

(4) The reapplication of. inorganic zinc primer was performed at 12.5'and 30 percent thinning.

e g.

k.__.______._

P 3 Two sets of laboratory test panels were prepared. One set of panels was used for adhesion tests, while the other set of panels was used at Oak Ridge Nuclear Laboratory (ORNL) for DBA testing. Adhesion test methods, .used-to obtain the bond strengtn of the coating were in accordance with ASTN standards. E1cometer adhesion test Cross-Hatch adhesion test, and~ Knife adhesion test methods were used.

Acceptance criteria were established by using the adhesion test results and the successful performance of the aformentioned coating application variables under SNUPPS project DBA conditions.

Acceptance criteria for Elcometer adhesion tests and DBA ccnditions were established in a::ccrdance with ANSI N101.2 (1972). .

The Carboline coating system CZ-11/SP7/CZ-ll/191HB successfully passed DBA test at 100 percent relative humidity, while Ameron coating system D6/SP7/D-6/90 successfully passed DBA test at 88". RH and above. The adhesion test results of these coating systems were used to establish the acceptance criterfa for the Phase II coating work. Acceptance criteria for the Phase II coating work are shown in the Section 2.2a of this report.

c. Tests at the Jobsite The jobsite test panels were prepared so that a comparison of adhesion characteristics of the Carboline'and Ameron coating systems with the controlled laboratory test panels could be made. These panels were prepared using the same coating systems and variables as the laboratory panels, with the exception of relative humidity, which was not controlled.

The Elcometer adhesion test values for the field panels were significantly higher than those obtained from the laboratory test panels. However, both sets of panels were found to have corresponding values in nature. The report also indicated that the jobsite panels had generally thicker residual inorganic zine than the laboratory panels, and the dry film thickness of the reapplied zine was lower than the laboratory panels with no readings found to be outside the specified limits.

4.4 Phase II Test Program

a. This phase of the program was established to provide DFT- measure-ments and adhesion tests at the designated test grids of the coating applied by the Constructor prior to July 1980. The DFT was measured with both Magnetic-Pull-Off (ron-destructive) and Tooke (destructive) gauges. The measurements were required to establish the ' level cf compliance with the thickness ~ range'specified for coat-ing system 101 of the Specification No. 10466-A-125(Q). The adhesion tests were performed using the same method utilized for the Phase I work. The Phase II test program was divided into five different categories as follows:

L

-e h

6 - -

Category 1 - Containment Liner (Vertical) Plate Category 2 - Reactor Cavity Pit Liner Category 3 - Containment Dome Liner Plate Category 4 - Structural Steel Category 5 - Maintenance Truss

. b. Containment Liner (Vertical) Plate and Containment Dome Liner Plate (Categories 1 and 3)

Ameron Coating System D-6/SP7/D-6/90 was applied to Categories 1 and

3. The surface area of Categories 1 and 3 was divided into 24-inch square grids. The grids which were accessible for testing, were included in a random sampling process for selection of sample test sites. Appendix A describes the random sampling process. One hundred twenty-four (124) sample test grids were statistically selected for each category. Each selected test site was tested, at or near the center of the grid for adhesion and dry film thicknesses. If the test was indeterminate, another test was performed in the same grid. Pass or failure of coating was based on the acceptance criteria established by Phase I of the program. If no more than two test grids failed for each category, the coating system was considered qualified. This statistical method provides a 95 percent confidence level that, at the most, 5 percent of the coating is defective.

123 grids out of 124 tested for Cagegory 1, while 124 were tested for Category 3. One test grid in Category 1 was impossible to test, and therefore, it was deleted. The average adhesion test results and DFT readings are outlined in Table I.

c. Reactor Cavity Pit Liner Plate (Category 2)

The reactor cavity pit liner was enated with Ameron coating system D-6/90 directly over steel prepared by Near White Metal Blasting to SSPC-SP10. This coating system was tested to confirm that the applied coating met the requirements of the Specification No. 10466-A-125(Q). A total of 24 test sites were selected on a non-statisti-cal basis. The test results obtained were also found to be very similar to the results obtained from laboratory applied Ameron coating systems which were cured at 100 percent relative humidity and DBA tested. The average adhesion test results and DFT readings are shown in Table I for comparison with the established acceptance criteria.

d. Structural Steel (Category 4)

. . Structural steel inside the containment was coated witn the Carboline coating system CZ-ll/SP7/CZ-ll/191HB. A total of 72 structural steci members were tested on a non-statistical basis.

The test method was designed to provide .nearly 100 percent confidence level for all structural steel. Specification No.10881-A-021(Q) permitted the use of one of the three adhesion test methods

[

-ll-L ,

.s.

]

for each test location. A correlation was developed by KTA to con-firm that the single adhesion test would establish satisfactorily the required adhesion test results. The cross-hatch adhesion test-method was selected. The average adhesiom test results and DFT readings are shown in' Table I.-

e. 'Mair.tenance Truss (Category 5)

The maintenance truss was coated with Carboline coating system CZ-11/191LF/191HB. Testing of this category was designed to endorse qualification of the applied Carboline coating system to the project

.DBA conditions. The tests were performed on a non-statistical basis. KTA selected a total of eight adhesion test sites for each of the three truss sections. The members tested were:

(1) Any face of the two top chords (2) Any face of the two bottom chords (3) Any-face of one diagonal member from each of.the four sides of the truss.

Average adhesion test results and DFT readings are shown in Table I.

e 6

e e

o

E  :

~ 2 TABLE I .

. a --

PHASE II -Suanary of Field Test Results Category 1 Category 2 Category 3 Category 4 Category 5 Total Locations 123 24 124 149 '37 Thickness Measurements 123 24 124 141 24 Pr. Ave. 3.5 mils 3.3 mils 4.1 mils 3.8 mils 3.4 mils y~ <1.6 mils 0 0 1 (Ave. 1.5) 1 (Ave._l.5) 1 (Ave. 1.0)

>4 2 mils 25 (Ave. 5.5) 3 (Ave. 5.7) 50 (Ave. 5.8) 32 (Ave. 5.3) 0 Int. Ave. flA flA flA

<l.6 mils flA flA IIA 2.5 mils flA flA

. > 4.2 mils NA i NA flA 3 (Ave. 1.2) flA 0 Fin., Ave. -7.3 mils ,6.4 mils 6.4 mils 9.6 mils

-<l.6 mils, _ 4 (Ave. 3.0)' 3.3. mils - -

~ .1 (Ave. 3.0) 11 (Ave. 2.8) 0 13 (Ave.2.5)

>7.2 mils 7*

59 (Ave. 10.1) 6 (Ave. 8.6) 36 (Ave. 8.7) 97 (Ave. 11.4) 0 l

Elcometer -

No. of Tests 251 48 248 48 Average psi 48 l782.(306-1000) 743 (538-1000) 828-(538-1000) 878 (564-1000)

Failures _0 0 0

'381 (106-890) 0 16 (Ave. 200)

(<300 psi)

Cross-Hatch

.No. of Tests 123 24 124 140-Average' 37' 9.2 (6.8-10) 8.9 (6.5-10) 8.9 (6.3-10) 9.1 (1.4-10)

Failures (<,5) 0 0 0 3.6(0-9.5) 2 (1.4,3.1) 23 (Ave. 1.1)

Knife-No. of Tests 123 24 12<

Average 49 28

.7.5 (5-10) 7.1 (6-9) 7.9 (6_-10) 7.2 (4-9)

Failures (<5) 0- 0 0 3.3 (0-7) 1 (Ave. 4)

. 6 18 (Ave. 2) t -

, o APPENDIX 'A' SAMPLING PLAN

-o A'samp'ing plan based on the result 5~ of inspection of a single sample drawn from the lot is called a single sampling' plan. .It is characterized by two_

numbers, n and c, where n is the sample size and c is the acceptance number.

Let X be the number of.defectives in the sample. The lot is accepted if X I. c, and the lot is rejected if X > C. Let E be the consumer's. risk, i.e., he is willing to take in having a poor-quality lot accepted by sampling plan. In other words,# is the probcbility of the sampling-plan accepting a lot that has -

a tolerance-fraction defective p. Let us row derive the probability of accepting a lot submitted with fraction defective p. 'If the lot size is N and we are sampling without replacement,- then the probability distribution of

.the number of defectives k in a sampleLof,siz6 n is hypergeometric. In symbols, N-NP N

0) P -[X= k) =(NP k n-k n In practice, the lost size N runs into hundreds, thousands, or even larger. It can be proven that the hypergeometric distributio'n with parameters N, n, and p approaches the binomial distribution with parameters n and p. That is, n) k n-k (2) P [ X= k) =(k/ p (1 p) for k = o, 1, . . . . . ., n.

Hence, k n-k (3) =P [X I c ] =

[

k=o (n) k/ p (1-p)

The testing of containment coating was set up as a single sampling plan program-that yields results of 95-percent confidence that, at most, 5 percentaof the coatings.are defective. In-this' case, the risk is 5 percent because of 95-percent confidence and the tolerance fraction defective is~p = 5 percent. By.

using the equation (3), the.valueslof n and c can be determined. For example, n = 59 and c = o, n = 93 and c = 1, and n = 124 and'c = 2.

s

'1, 1

1

,-