UNCLASSIFIED AD NUMBER - DTICSAMSO USAF ltr, 28 Feb 1972 THIS PAGE IS UNCLASSIFIED Ig p~. -) I -0. SUSLA. Ca. 90045 oG=NERAL DYNAMICS ASTRONAUTICS A2136. (Pri 9-61 GENERAL DYNAMICS]

UNCLASSIFIED

AD NUMBER

AD831537

NEW LIMITATION CHANGE

TOApproved for public release, distributionunlimited

FROMDistribution authorized to U.S. Gov't.agencies and their contractors; CriticalTechnology; SEP 1962. Other requests shallbe referred to Space and Missile SystemsOrganization, Los Angeles, CA.

AUTHORITY

SAMSO USAF ltr, 28 Feb 1972

THIS PAGE IS UNCLASSIFIED

Ig

p~. -) I�-0.

SUSLA. Ca. 90045

oG=NERAL DYNAMICSASTRONAUTICS

A2136. (Pri 9-61

GENERAL DYNAMICS] ASTRONAUTICS

) Report No. ERR-AN-196 vr

AN EVALUATION OF THE

MECHANICAL PROPERTIES OF- ADHESIVES

AT CRYOGENIC TEMPERATURES AND THEIR

CORRELATION WITH MOLECULAR STRUCTURE

J. Hertz

Materials Research

This document is subjectto special e:,port controls endeah it• r Ial to foreign

La wade onlywith pi',.or approval of;SHq.SA.?SO, LA., Ca. 90045Attn: SP(SD

Prepared byGENERAL DYNAMICS IASTRONAUTICS

A Division of General Dynamics CorporationSan Diego, California

September 1962

This research was supported under GeneralDynamics IAstronautics

Research Program No. 111-9206

G••EA[DYNAMICSS-1 k A "8TICS

C i T 11962

L4

GENERAL DYNAMICS ASTRONAUTICSI

CONTENTS

List of, Illustrations v

List of Tables vii

Acknowledgment ix

Introduction xi

Abstract xiii

1. EXPERIMENTAL PROCEDURE 12. DISCUSSION OF RESULTS 53. ConCLUSIONS

37References

39Bibliography

41

I

I r ~ ~ ~ '--tt-

GENERAL DYNAMICS [ASTRONAUTICS IILLUSTRATIONS

1. Effect of surface preparation on the tensile-shear strength of Metlbond 406. 6

2. Tensile-shear strength of Metlbond 406 vs. loading-rate. 7

3. Effect of thermal cycling on the room-temperature tensile-shear strengthof Metlbond 406. 8

4. Effect of specimen width on the tensile-shear strength of Metlbond 406. 9

5. Tensile-shear strength of Metlbond 406 vs. temperature with variousplastic adherends. 10

J. Effect of various surface preparations on the-tensile-shear strength ofFM-1000. 11

7. Tensile-shear strength of AF-41 and Tacky AF-41 vs. temperaturo. 12

8. Tensile-shear strength of Metlbond 409 (Tacky) vs. temperature. 13

9. Tensile-shear strength of Narmco 3170-7133 (X-284) vs. temperature. 14

10. Tensile-shear strength of various epoxy-polyamide adhesives-vs.temperature. i5

11. Effect of resin's molecular weight on tensile-shear strength. 17

12. Total linear thermal expansions reported by NBS Cryogenics Engineering

Laboratory. 18

13. Ultrasonic test equipment for determination of adhesive modulus properties. '19

14. Total linear thermal expansion of some adherends. 20

15. Total linear thermal expansion of some adhesive systems. 21

16. Total linear thermal expansion of adhesives and adherends tested atAstronautics. 22

17. Tensile-shear strength of Coast Pro-Seal 777 vs. temperature 23

Iv

-i 1

rGENERAL DYNAMICS ASTRONAUTICS

18. Static and dynamic tensile-shear strengths of-adhesives tested-atHughes Aircraft Co.. 24

19. Butt-tensile strength of various adhesives vs. temperature. 25

20. Impact strength of variousadhesives vs. temperature. 26'

21. Pi-tension tests of sandwich pane'lshavlng titanium faces and~flberglasscore. 27

22. Pi-tension tests of sandwich-panels having stainless steel faces and-fiberglass core. 28

23. Shear properties of honeycomb panels bonded with Metlbond 406. 29

24C Spectrum of Metlbond 406. / 33

25. Spectrum of -AF-41. 33-

26. Spectrum of FM-1000. 33

27. Spectrum of AF-40. 33

28. Spectrum of Resiweld No. 4, Component A. 34

29. Spectrum- of Resiwela N1. 4, Component B. 34

30. Spectrum of Narinco 3135, Component A. 34

31. Spectrumof Narmco 3135, Component B. 34

32. Spectium of Metlbond'4041. 35

33. Splctrum of AF.32. 35

34. 'Spectrum of APCO 1219. 35

viK

GENERAL DYNAMICS ASTRONAUTICS

TABLES

I. Surface preparations. 2

H. Young's modulus of Metlbond 406 obtained by tensile-testibg of molded 16--specimens.

vii

{ p

2.3

GENERAL`DYNAMICS ASTRONAUTICS

ACKNOWLEDGMENT

The author wishes _ _express his appreciation to the many individuals who were in-volved in the-program to evaluate adhesives at cryogenic temperatures. The effortsof Mr. J. L. Percy in constructing the ultrasonic test equipment-and the efforts ofMr. M. D. Campbell in determining the thermal expansionproperties of a series ofadhesives and adhfrend§ is gratefully acknowledged. In addition, tfie author thanksthe many technicians who were instrumental in'bonding and testing the. test specimens.

ix

€I

GENERAL DYNAMICS j ASTRONAUTICS

:INTRODUCTION

The Materials Research Group dtVGdneal P3 micsJAstronatitics, A ,Di••v n .o ei, li•

eral Dynarnics Corporation, condu6ted a,:Resear,&hV 3ngineering Authorizatloi, (REA)

program in 1961 entitled, "An Evaluatio, n :of'rte"Mechanical Propertie, of Adhesi esat

Cryogenic Tenperdatures aad'Their Correlation With Molecular ýStructure." This was

a continuationooftEA Program No. 1-1'1-910'6'which was accomplishedAnn 1960. Theobjectives. S•,thi is.ogi'am were;to oýtain niechanical and physicalproperty;.datha onvarious bPtAc clasý:es of structural, adhesiVest over the 'temperature range of -4230 to

78 0F, andoK.tp~orrelate thisdatadithin the mblecular structure of-'the adhesivesystems.,

Sincd A.;stronautics began this work in 1960, many otber organizations have entered

into testingof adhesives at cryogeni temperatures. This increase, in. testing has re-sultedprimarily from the awarding of major spade contracts lsaturn S-IV, Saturn S-Il,

and Saturn S-I-B). Each of these vehicles requires the use of.cryogenid propellants.In addition, small research contracts have been awarded by the National Aeronautics

and Space Administration in this area. Data which has been generated.by other organ-

izdtions, and which is availablh, has been included inthis report. WNherethe data gen-eratedoby other sources is too extensive-to be included, ithas-been referenced .wherever

applicable. An up-to-date survey of all literature available on the properties of adhe-

sives at cryogenic temperatures Is listed in the Bibliography. Ii causemany of the

designs generated for various stages of Saturn required the us oof honeycomb structure,

a large effort was placed on evaluating bonded honeycomb structures. Although it wasnot'the initial intention of this REA1 to evaluate composite structures, it was found that

-this type of data was more important for present design considerations than the normal

mechanical property testson metal-to-metal joints.

xi

X ýENERh DYNAMICS I-ASTRONAUTICS

/

ABSTRACT /

/

Stoe -eflLe-Lhe4r,,it-ýensfle,, and impact tag,4were run dn a series of Adhesives from-

,423? to 780F. Pi-tensioni plate-shearifnd finigue tests were run on a seriesofsland•€lchoneycomb,panels •havin~g stp_,?ess,.stel1 and titanium faces, fiberglass-ýhoney-Ob~mb;66re,,, bonded withepo~xy-nylon and fllied epoxy..phenolic adhesyes. The effect§ýrt#merfs, ekln~ing cycles i thermal cycling, a~nd lbiaitng'rates o4, ihe tensile!-shear

strength of adhesives was determined at, cryogenic temperatures4 jThermal expansion and tensile testing was conducted on nrdded and cast adhesives

'from -423° to 78,0F. Infrafed'curves were run on the adhosives evaluated. in-this-Research Engineering Authdrization (REA) report, as wal as those- evaluated in REA1i1-9106.

0/ I/ I

//

// /"I

/

/

/

j/ xrl/

9 GENERAL-DYNAMICS IASTRONAU TICS I

EXPERIMENTAL -PROCEDUREf

-Tensile-shear~ (1/2-in, overlap): cou~pons were m~achinedzand fabricated from the follow-lng niaterials: 0.020-in. EFH 301 stainless steel, 0.020-in. 5Al-2.5Sn titanium alloy,and 0. 125 -in.- Conolon 522 (ep,4-y.ý-ibe~rgl ass lami nate). The, lap-shear coupons wereprepared..by b~ndirzg 4 in. x 1 in. strips to obtain an over-all. test specimen of 7-1/2 in'.x -11n. Theý ends -of the spec m'ens contained either spot-welded or bonded doublers;for reinforcement in'che grip areas and 1/12-in, holes for accommodating Oin;ype grips.

Butt-tensile coupons were prepared from 3/4-in, round stock 321 stail1ess steel.After bonding, the coupon had-an over-all len~gth of 2 in. and a bond area of -0.442sq.ln. The coupons~were threaded on each end with a 3/4-10 thread.

Impact specimens were prepared from 321 stainless steel flat stock, per Federal'Test Method ,Standard'No. It7 (Tentative'Standard Method 1051. 1-T). They consistedof -two stainless steel" blocks (1 in. x 1 in. x 3/8 In. anid 1-3/4 in.,x 1 in. X 3/4 in.),bonded together aincfresulting in 1 sq,~ In. of bonded area.

P1-tension specimens were made per Fort Woi'th 8ppecifcation (Refezbfnce 1)*.Faces were--eIther 0.032 in. EFH 301 staftA02es stelo~r 000i.5125nttnualloy and.were, bonded'to 1/2-In, thick Hexce1 T'1R?.3/16 In. core o .-. ?l.c~t

Lx ~jxef) density. Specimens are 2-In. dia. discs -n tehuve a bond area- of -p1(3. 14),sq. in.

Plate-shear specimens were made Per MIL-STD-40btA. The spec~iv'ens- were 5..670,in. x 2 In. and had faces of 0. 032-in.. E FH 301 stalnleFA s&tl which.werf, *b6nde.,'3 to1/2-in, thick Hexcel HRP-3/16 in. core of 5.7 pcf density.

*References are listed on Page 39. ;


Table 1. Surface preparations.STAINLESS STEEL & T!ANIUM ADHERENDS

1. (ASTRONAUTICS ETCH) WIPE SPECIMENS WITH TRICHLOROETHYLENE. IMMERSE SPECIMEN FOR 10 MIN. AT 1507F

IN THE FOLLOWING SOLUTION:

100 GM HYDROCHLORIC ACID (CONC.)

,4 GM HYDROGEN PEROXIDE (30%)

20 GM FORMALIN (40%)

90 GM WATER

RINSE SPECIMENS IN COLD TAP WATER. FOLLOW THIS WITH A DISTILLED WATER RINSE & AIR DRY. THIS IS

FOLLOWED BY 5 TO 10 MIN. OF ETCH AT 140-160'F IN THE FOLLOWING SOLUTION:

100 GM SULFURIC ACID (CONC.)

10 GM SODIUM DICHROMATE

'30 GM WATER

REPEAT RINSE AND DRYING PROCEDURES AS ABOVE.

2. WIPE SPECIMENS WITH TRICHLOIROETHYLENE. IMMERSE SPECIMENS FOR 5 TO 10 MIN. AT 200F IN THE FOLLOW-

ING SOLUTION:

263 GM PREBOND 700

1 GAL. WATER'

RINSE SPECIMENS IN COLD TAP WATER. FOLLOW THIS WITH A DISTILLED WATER RINSE & AIR DRY.

3. SAME AS FIRST PROCEDURE EXCEPT USE 300 GM WATER IN SECOND ETCH SOLUTION.

ALUMINUM ADHERENDS

WIPE SPECIMENS WITH TPICHLOROETHYLENE. IMMERSE SPECIMENS FOR 10 MIN. AT 150"F IN THE FOLLOWING

SOLUTION:

30 GM WATER

10 GM SULFURIC ACID (CONC.)

4 GM SODIUM DICHROMATE

RINSE SPECIMENS IN COLD TAP WATER. FOLLOW THIS WITH A DISTILLED WATER RINSE & AIR DRY.

,PLASTIC LAMINATE ADHERENDS

SAND LIGHTLY & WIPE CLEAN WITH TRICHLOROETHYLENE.

accomplished according to manufacturer's recommendations. P1-tension and plate-shear coupons were not etched prior to bonding. Tensile-shear specimens were loadedat a rate of 600 to 700 psi of the shear area per minute (ASTM D 1002-52T) exceptwhen evaluating the effect of loading rate on lap-shear strength. The butt-tensile cou-pons were loaded at 350 lb./min. (ppm); the pi-tensionspecimens were loaded at both

2

li . . .. . - - . . .

) GENERAL DYNAMICS j ASTRONAUTICS

500 and 4, 000 ppm: and the plate-shear specim6 is were pu'.ied at 0.0285 in./min.head travel. Impact specimens were -broken using a hammer velocity of 17 fps.

The tensile-shear, butt-tensile, p1-tension, and plate-shear specimens were testedusing a 50i 000-lb. Baldwin-Emery SR-4 Testing Machine, Model FGTj and a 30, 000-lb. Tinius Olsen electromatic testing machine. Impact speciimens were tested using aSonntag Universal impact machine, Model SI-1P. The proper temperature environ-ment was provided by immersingthe specimen in alcohol and dry. ice (-100F), liquidnitrogen (-3200F), or liquid hydrogen (-4230F). Details of the experimental procedureincluding temperature measurement, liquid level -indication, and safety precautionsaregiven in a paper by Christian and Watson. (Ref. 2).

Alignment fixtures (Ref. 3) were utilized for bonding of-the tensile-shear, butt-tensile and-impact specimens. Spherical joints were utilized in the tensile-testingmachines to ensure alignment during testingi Prior to test the tensile-shear couponswere measured optically to determine the length_ and width of the overlap area. Thick-ness measurements of the adherends and the bonded ardei for each specimen were madewith a micrometer to determine the adhesive thickness of each test coupon. After test-ing, all specimens that failed in the bonded area rather than in theadhorends were7) visually examined. .The type and extent of each type of failure was noted. For thisreport a cohesion failure is defined as one that occurs within the adhesive, while an

adhesion failure is defined as one that occurs in a weak -boundary layer. In the caseswhere a primer was utilized and failure occurredcbetween adhesive and primer, thefailure is considered to be one of adhesion. In the cases where the adhesive is a sup-ported system and failure occurs between- the resin and the reinforcement, the failureis considered to be one of cohesion.

Thermal expansion specimens were machined out of molded or cast adhesives.These specimens were tested utilizing a modified Leitz dilatometer and a thlin-sheet,quartz tube dilatometer. The modified Leitz dilatometer results in a continuous tem-perature vs. expansion curve, whereas the thin-sheetdilatometer was used only forobtaining the total linear thermal expansion of the specimens over a selected temper-ature range.

Infrared curves on the adhesives were run on a Perkin-Elmer Modelv21 double-beaminfrared spectrophotometer. Specimens of the unsupported film systems were pre-

pared by stretching or heat-pressing of the tapes between sheets of Teflon in order toobtain specimens thin enough for infrared study.

3


DISCUSSION OF RESULTS

Tensile-shear testing at cryogenic temperatures was initiated-in REA 111-9106ý'andcontinued in 1961. The tensile-shear testing was primarily concerned with determin-ing the effects of various parameters on tensile-shear strength. The parametersevaluated included surface preparation, loading rate, thermal cycling, and specimenwidth.

Figure 1 indicatesthe effect of surface preparation of the adherends on the tensile-shear strength of Metlbond 406. The adherend utilized for this program was 0.020-in.EFH 301 stainless steel. Metlbond 406,was selected for additional evaluation since ithad resulted in-the highest strengths at bryogenicdtemperatureswhen tested in 1960.Values obtained with the Astronautics concentrated chromate etch were reported in

ERR-AN-032. Two other etching cycles were examined this year. The; Forest Prod-Sucts Laboratory's stainless, ste~i etch (see Table I) is identical to the first part of the

Astronautics etch butis followed by a dilute chromate rather than a concentrated chro-mate etch. The FPL etched specimens when tested at -423°F failed at approximately70% of the tensile-shear strength obtained with the Astronautics etch. The -use of anepoxy-nylon primer (Bloomingdale Rubber Co. FM-1009-8), in conjunction with theAstronautics etch, resulted in strengths approximately 50% lower than those obtainedwhen no primer was utilized. This confirms the testing of primers accomplished inthe initial REA. The use of an alkaline surface preparation (Bloomingdale Rubber Co.Pre-Bond 700) resulted in an average tensile-shear strength at -423 0F, approximately7% higher than those obtained with the Astronautics e6th. It should be pointed out, how-

ever, that the batch of Metlbond 406 utilized in the 1960 program was-different from theone utilized in evaluating the FPL stainless steel etch, Pre-Bond 700, and FM-1009-8.There may be a drastic difference in tensile-shear strength obtained with one batch ofadhesive from that obtained with another batch. This phenomenon has been evident interts conducted at higher temperatures by Narmco Research and Development, and isquite a common occurrence in the testing of adhesives.

Figure 2 indicates the effect of loading rate on the tensile-shear strength of Metl-bond 406 joints using 0.020-in. EFH 301 stainless steel adherends. The effect ofloading rate was conducted at -320° and -423°F, and covered the range of loading rate

)5

00 00LA*1

GENERAL DYNAMICS [ASTRONAUTICS

10,000

8,000

0.00,

~6,000-~-

'~4,000 _ _ _ _

ADHERENDS: 0.020-IN. 301 (EFH) CRES

SO ASTRONAUTICS CONCENTRATED CHROMATE ETCHt• • FPL STAINLESS STEEL ETCH

0 PRE-BOND 700-A ASTRONAUTICS ETCH PLUS FM-1009-8 PRIMER

0

-450 -375 -300 -225 -150 -75 0 75 150

TEMPERATURE (0 F)

Figure 1. Effect of surface preparation on the tensile-sheai strength ofMetlbond 406.

from 1,000 to 10,000 ppm. At both temperatures the general trend was a-gradual in-creaseintenslle-shear strength with increase iniloading rate. There was approxi-mnately a 40% increase in tensile-shear strength when specimens were loaded at 10, QOOppm over the strengths obtained when the specimens were lbaded at 1,000 ppm.

McClintock and Hiza (Ref. 4) have shown that thermal cycling has a deleterious ef-fect on the tensile-shear strength at room tempbrature of a filled epoxy adhesive. Todetermine if this might also be the case with the epoxy-nylon systems, some stainlesssteel specimens were bonded with Metlbond 406. Five specimens were cycled 10 timesbetween -423* and 787F, five were cycled 10 times between -320°F and 787F, and fivewere cycled 10 times between -1000 and 78F. They were then all pulled at 787F andcompared to specimens which had not been subjected to thermal cycling. Figure 3 isa graphical presentation of the results obtained. It can be seen that there was no ap-preciable change in tensile-shear strength as a result of thermal cycling.

0


io,000

86,000-

4,00

I I

zult"ADHERENDS: 0.020"IN, EFH 301 CRES

2,000 0 -"320°0 FI -423°F

NOTE: SPECtMENS ETCHED WITH ASTRONAUTICSCONO-ENTRATED CHROMATE ETCH

02,500 5,000 -7,500 /0,000-,LOADINGDRATE (PPM)

Figure 2. Tensile-shear strength of Metlbond 406'vs. loading rate.

It has generally been accepted that the width of the overlap join't~does not affect the,strength of the joint. However, testing, of specimens 38 in. x 4 in. wide, having 1/2-in. and 3/4-in. overlap joints bonded with Metlbond 406, resulted in strength valuesconsiderably lower than those obtained with the standard -specimen 1 in. wide. It,-wasfelt that" the extreme length of the specimen.and the equipment used for holding aridloading the specimens might result in some eccentricity in the joints,ý Tests •were-re-peated at room temperature by Narrnco Materials Division, And at room temperatureand at -423°F at Astronautics -with..smalle r specimens, (8 'to 12,in. long). Results areplotted inFigure 4 and seem to verify that joints bonded with Metlbond 406 result inlower strength per unit, area with increase in the joint width. It is felt that- additionaleffort is required in t~his field, both experimentally and analytically, with Met'lbond406 and other adhesive systems.

In REA 111-9106 (Ref. 3), it was determined that when Metlbond 406 was used forbonding tensile-shear specimens having polyester-fiberglass or phenolic -fiberglass

LODIGRAE(PM

Fiur 2 enil-ser trnghofMelbn 46 s.ladngrae


10,000 ADHERENDS: 0.020-IN. EFH 301 CRES

NOTES: (I) ASTRONAUTICS CONCENTRATED CHROMATEETCýH

(2) CYCLED 10 TINES-FROM 78TO RESPECTIVE_______0_TEMPERATURES PRIOR TO TESTING AT 780

044 000 ..

2,000 - -

-6O 450 -375 -o0 -22 1 5q -• 75 150

-TEMPERATURE (F)

•,: • , /.Figure 3, Effect of thei-nal cycling on the room-temperature tensile-shear :

• •strength of•Metlbond 406.

adherends, failure always ýoccurred -within the adherends rather than in the adhesive.•,,t It Was felt that a truer value Of the joint strength of Metlbond 406 might be obtained by

testing laminates which had higher interlam'inar shear strength. Figure 5 shows the

tensile-shear values obtained-when using Conolon 522 (epoxy-fiberglass laminate) ad-herends; Values,' obtaindd previously have also been,,included forcomparison purposes.Once again all fa -ilures occurred in the adhierends rather than in the adhesive, even•Cthough the+ sress levels at which failure -occurred were much higher than those-ob-

S~tained previously.

S~Figure,6 shows the effect of various surface preparations on the tensile-shear• ~strength of FM-1000 epoxy-nylon adliesive. Since original specimens tested In 1960-} had been supplied by Bloomingdale Rubber Co., and because these specimens had

uneven glue lines, it was decided to test some specimens bonded at Astronautics. Us-ing, the same etch procedure as had been used initially (Astronautics etch), specimenstested at -3206F resulted'in an average tensile-shear strength approximately 30%8

w 4-


10,000 __ __

8,000

ADHERENDS: 0.020-IN. EFH 301 CRFS

0 I-IN. WIDTH0 1.5-IN. WIDTH

< ,•d& 2-IN. WIDTHS3/4-IN. OVERLAPS

|-IN. WIDTHS/ -"0 4-1N. WIDTH•, 4,000 u ","7'

2,000o•

S NOTE: DARKENED SYMBOLS DENOTE SPECIMENSMANUFACTURED AND TESTED AT NARMCO.ALL SPECIMENS ARE I/2"IN. OVERLAPS

,UNLESS OTHERWISE NOTED.

I -450 -375 -300 -225 -150 -75 0 75 150

TEMPERATUREN (N)

Figure 4. Effect of specimen width on the -tensile-shear strength of Metlbond 406.

higher than that obtained in 1960. Specimens prepared with the Astronautics etch andusing the FM-;009-8 primer were tested at -423°F and: resulted in values similar to,thoseobtained with the Metlbond 406 joints which had been similarly prepared (seeFigure 1). Specimens etched with Pre-Bond 700"once again proved to give the higheststrengths with stainless steel adherends at -4237F.

Figures 7 and 8 show the tensile-shear strengths obtained with two new epoxy-nylonsystems: Tacky AF-41 and Tacky Metlbond 409. The advantage of utilizing tackyunsupported tapes is most apparent when bonding large vertical surfaces. With theother unsupported tapes it Is necessary to utilize tacking agents, heat tacking, ormechanical means for holdingthe tape in place during bonding. Bonding Is more diffi-cult and there Is a possibility of obtaining poor bonds In local areas. Figure 7 indi-cates that curing temperature is critical when utilizing AF-41. By curing at 390&F in-stead of 350°F there was approximately a 35% decrease in tensile-shear strength at3200 and -423°F. The values obtained with Tacky Metlbond 409 are extremely promis-ing, and further testing with this adhesive system is indicated.

A IN 1 ,-9I

GENERAL DYNAMICS;IASTRONAUTICS

ADHERENDS:

@CONOLON 522 (EPOXY-FIBERGLASS LAMINATE)-E CONOLON 527 (PHEOLIEC-RFIBERGLASS, AIAE

ACONOLON 506 (PHEOLYESERFIBERGLASS, LAMINATE)

Z 6,000

4,000

2,000 ---

01

-450 -375 -300 -225 -150 -75 0 75 150

TEMPERATURE (0 F)

Figure 5. Tensile-shear strength of Metlbond 406 vs. temperature with various

plastic adherends.

-Because it is advantageous to utilize room-temperature curing adhesives, someeffort was- placed on evaluating this type of system e'•ren though they-were definitelyInferior toWthe heat-cured systems. Figure 9 shows the tensile-shear strengths ob-tainedwith Narmco 3170-7133 (X-284) adhesive with both 0.020-in. EFH 301 stainlesssteel and 0.020-in. 5A1-2.5Sn titanium alloy adherends. This system was developedby Narmco Research and Development under NASA contract No. NAS 8-1565, Develop-ment of Adhesives for Very Low Temperatures. It, consists of Narmco 3135 (epoxy-polyamide system) filled 33% by weight With finely ground nylon. This system, whenevaluated by Narmco, resulted in excellent strengths with 7075-T6 aluminum adher-ends; however, the values obtained at Astr6nautics with stainless steel and titaniumadherends were poor. Tests were run only at -3200 and -423 0F to conserve funds.

Figure 10 shows the tensile-shear results obtained with various ratios of Epon 828(epoxy)/Versamid 125 (polyamide). The 50-50 ratiD by weight resuilted in the highest

100


-~~~~~ ~10,000 _____ _____ _____ _ ____

j ~~8,000 T-6 ,0 00 - _ _ _

.00, k.

S~ ADHERENDS: 0.020-IN. EFH 301 CRES

0 ASTRONAUTICS ETCH (SPECIMENS PREPARE[)2,000 A PRE-BOND 700

* ASTRONAUTICS ETCH (SPECIMENS PREPARED

AT ASTRONAUTICS)* ASTRONAUTICS ETCH PLUS'FM- 1009-8 PRIMER

-, -450 -375 -300 -225 -150 -75 0 75 150

)] TEMPERATURE (0 F)

Figure,6. Effect of various surface preparations, on the tenslle-sheai-strength of FM-1000.

values. The results previously obtained on N4r, mco 3135 have also been included onthis graph. In most instances, there was a iarge.spread in tensile-shear data. Thisis more common in room-temperature curing liquid adhesives because it is-more diffi-

cult to control glue line thickness, mixing of components, gas bubbles in the glu line,etc. Only specimens bonded with Narmco 3135 were tested at -100° and -423@F. It isthought that had the other systems been evaluated at these temperatures they wouldresult in curves similar In shape to that obtained'with the Narmco 3135.

Work accomplished by Hiza and Barrick at the Cryogenic Engineering Laboratory ofthe National Bureau of Standards (Ref. 5) has shown the effect of the molecular weightof epoxy resin on the tensile-shear strength obtait'.ed with, the cured epoxy system.The epoxy resins checked were' derived from the pcilycondensation of epichlorohydrlnand bisphenol-A. Figure 11 shows the tensile-e.hear values obtained with 0.125-in.

2024 aluminum adherends when the epoxy resins wer'e cured with stoichiomedtric [

11 11

V. t : = • • = • .

SGENEFAL -DYNAMICS-I ASTRONAUTICS

10.,000 I

8'000 A*1ATOATETHNOMLCR

ADRERENDS: 0.020-IN. EFII 301 CRES

A AF-41,-ASTRONAUTICS ETCH, 390' CUREN AF-41. ASTRONAUTIC'S ET.CH,'NORMIAL CURE

- TACKY AF-4.1,PRE-BOND-700 ETCH, NORMALCURE0,6,000 " . __CUR

z

2,000 -

0)

I -

-450 -375 -300 -225 .4075 0 75 150

TEMPERATUIk.E(°F)

Figure 7' Tensile-shear strength of AF-41-and Tacky AF-41 vs.. temperature.

-amounts-of-meta-phenylenediamine. The values obtained at 3000K (81 0F)-indicate agradual increase in tensile-shear strength with increase in molecular weight. Thissame dependence had previously been shown by N. A. de Bruyne (Ref. 6) with aluminumlap joints using phthalic anhydride as a curing agent. At 760K (-3230 F) Hiza andBarrick found that tensile-shear strength increased with molecular weight up to approx-imately a molecular weight of 700 and then gradually decreased with an additional in-crease in molecular weigI They postulated that, at 3000K, residual stress in the lapjoint was primarily dependent on chemical shrinkage, while at W76K the residual stresswas due primarily to thermal contraction. Figure 12 shows the graphically integratedthermal expansivity of the materials evaluated at NBS. Hiza and Barrick postulatedthat the uncured resin is absorbed by'the adherend and that the adhesive strength isrelative to the magnitude of the adsorptive forces; the strength of adsorption of epoxyresins increasing with molecular weight. As the chemical curing reaction progresses,the mobility of the resin molecules is reduced, thus limiting additional adsorption.The cohesive -strength is developed as a result of the curing process, forming a tough,

12

- -- -

GENERAL DYNAMICS I-ASTRONAUTICS

10.000

8.000- 1- .. L1 .... ______

8.00

z6',000 •

2,000

0 o I:I

-450 -375 -300o -225 -450 75 0 75 150

TEMPERATURE (°F)

Figure 8. Tensile-shear strength of Metlbond 409 (Tacky) vs. temperature.

crosslinked structure.. Hiza-and Barrick assumed that-forces of adsorption. increase.as the temperature is decreased. Most polymeric materials also tend to increase incohesive strength at low temperatures, therefore, it would be reasonable to expecttensile-shear bonds at 760K to be stronger than those at 300°K, provided that residualthermal stresses are not limiting. Becuses Hiza and Barrick consider the-residualthermal stresses to consist of stresses resulting from chemical shrinkage and-thermalcontraction, the net effect of residual stresses is dependent on the service temperatureand the magnitude of thermal contraction of the adherenr~material. At 300°K the ther-mal effect is small since the resins were cured at a comparatively low temperature(1 hour Et 1100C). At 76 0K the differential thermal contraction between adhesive andadherend, adds stresses to those already present at 300°K. Looking at Figure 12, one

can reae.4y see why the high molecular weight resins suffer the greatest loss in strength.

It should be noted that Hiza and Barrick only considered thermal stresses. Thismay be misleading because there are a number of other stresses which may be giving

13

GENERAL DYNAMICS ASTRONAUTICS C2,500

ADHERENDS:* 0.020-IN. EFH 301-CRES8 0.020-IN. 5 AI-2.5 SnITITANIUM ALLOY

NOTE: SPECIMENS ETCHED WITH PRE-BOND 700

2,000"

II41

W 1,000

500

TEMPERATURE (°F) °

00

Figure 9. Tensile-shear strength of Narmco 3-170-7133 (X-284) vs. temperature. ,

interactions, and what appears to be a correlation may in fact be far from a correla-tion. There was n o attempt made to determine what is the change in modulus of the :adhesives at the different test temperatures. Since each resin system checked wouldhave a different ultimate cohesive strength and a different shear modulus at each ofth(3 test temperatures, neglecting these properties and trying to correlate molecular•weight to tensile-shear strength by use of thermal stresses alone appears to be over--simplifying the problem.

To obtain a truer picture of the stress concentration resulting in tensile-shearjoints, an attempt was made to obtain the mechanical properties as well as the thermalexpansion of molded adhesives. The first attempt was to obtain streas-strain curvesat the various test temperatures on molded tensile coupons. In addition to being verybrittle when tested at -320°F, the values obtained on the molded adhesives were veryinconsistent. To obtain more consiste-nt and useful data an attempt was made to ob-tain modulus data ultrasonically by modifying a setup which had been developed',-,Deitz and coworkers at Massachusetts Institute of Technology (Rlef. 7). The equlpmeut

14

1 a.

_ _ _

GENERAL DYNAMICS jASTRONAUTICS2,500 f

2,000 _ _ _ _ _ _'I

Z 1150

000

I Vol

-1•0 0 - 22N -180 -7• 0 7M12adeivsv. temeraure

-45 -35.30 -22- -15 - 7 0 -5 150_

TMEAUREPON 2/ESMD 2)6:5

(see Figure 13) and specimens to be evaluated have been built, but this REA expiredbefore the work could be completed. The equipment for obtaining the necessary res-onance properties, the resonant frequency, and half-power bandwidth, consists ofthree major parts: a driving system to set up longitudinal ultrasonic vibrations in thespecimens, a detecting system to indicate the amplitude of vibrations, and a frequency-measuring system. The specimen consists of a thin adhesive layer Joining two cyiin-drical bars end to end. Relations have been derived by Nolle and Westervelt (Ref. 8)to calculate Young's modulus and viscous modulus knowing the resonance propertiesof the cemented bar and a continuous bar of the same length and material. Table 11contains the modulus data obtained at 78°, -i00°, and -320°F when using Molded ten-sile specimens of Metlbond 406.

Total linear thermal expansion vs. temperature curves were run on some typicaladherends and adhesives over the temperature range of -200R to +25°C. These curves

15

GENERAL DYNAMIC$] ASTRONAUTICS

Table Ii. Young's modulus of Metbond 406 obtained~by tensile-testingof molded speciztiens.

SPECIMEN NO. TEMPERATURE (IF) YOUNG'S MODULUS-(PSI)'

1 78 43,400

2 78 44;300

3 78 49,000

AVERAGE 45,600

4 -100 291;000

5 -100 354,000

•6 -100 296,090

7 -100 * 235,000

AVERAGE 294,000

8 -320 525,000

9 -320 1,177,000

10 -320 810,000

11 -320 866,000

AVERAGE 770,000

are shown •In Figures 14 and, 15, Figure 16 shows the total linear thermal expansionof these adhesives and adherends between 76* and 300°K. Examination of -this graphseems to indicate that for each individual adhesive system, the highest tensfle-shear-strengths should~be obtained with the 2024-T3 aaluminum alloy adherends. The jointsutilizing 3,01'full-hard stainless steel adherends should result in higher strengths thanthose using 5A1-2.5Sn titanium alloy adherends; In most cases, this order'has been

,followed; however, it should be pointed out that the strength will be dependent on theability to propare each of the adherends adequately so that in no case Is failure adhe-sive in nature.

Many of the common sealants and putties act as structural adhesives at cryogenictemperatures. These would be used more frequently for cryogenic adhesive applica-tions if these materials had higher room-temperature adhesive properties. Figure 17shows some tensile-shear strengths obtained with Coast Pro Seal 777 polyurethanesealant using 0.020-in.. EFH 301 stainless steel adherends. The values obtained-wereconsiderably lower than those obtained previously by the Applied Manufacturing Re-search Group (Ref. 9). The discrepancy in resuits may be a result of two etch proce- iidures. Coast Pro Seal 777 also has been known to be affected by storage conditions"and this mayaccount for the large difference in tensile-shear results.

16 01 __ _________

'----__________ ' " - -,-' - - - -.- -.-

GENERAL DYNAMICS IASTRONAUTICS

5,000

4,000- -

z3,000 ____

W

T 2,0O00

z ADHERENDS: 0.125-IN. 2024 ALUMINUM

oI-.l 76°K1 300 0 K

1,000 NOTES: THESE SPECIMENS;WERE BONDEDWITH EPOXY RESINS

CURED WITH META-PHENYLENEDIAMINE.

SPECIMENS WERE TESTED AT THE CRYOGENIC ENGINEERINGLABORATORY, NATIONAL BUREAU OF STANDARDS

0 250 500 750 1,000 1,250 1,500 1,750 2,000

MOLECULAR WEIGHTOF RESINS

Figure 11. Effect of resin's molecular weight ontensile-shear strength.

Hughes Aircraft Co. has done considerable work in the area of sealants and adhe-sives-at cryogenic temperatures (Ref. 10). Figure 18Mshows-the static and-dynamicshear strengths obtained with DC-731 (silicone sealant),and Metlbond 406. HughesAircraft Co. has recently obtained a small research contract from NASA for the de-velopment of sealants for low-temperature applications. Progress on this contract(NAS 8-2428) is available, but data is too comprehensive to be included in this report.

In addition to the testing of tensile-shear specimens, a series of butt-tensile testswas also conducted. It can be seen from Figure 19 that thebutt-tensAle strength ofspecimens bonded with epoxy-nylon adhesives is considerably higher than 'the strengthsobtained with other classes of adhesives. The values obtained with the butt-tensilecoupons are considerably higher than the values obtained with tensile-shear coupons.This would indicate that there is less stress concentration in a butt-tensile specimenthan in a tensile-shear specimen. Results obtained with Metlbond 406 at -423°F weremore consistent than the values obtained with FM-1000. Examination of the FM-1000bonded specimens after failure revealed two distinct phases of the adhesive. One phase

J 17

'GENERAL DYNAMICS j ASTRONAUTICS

I DEAPA: DIET"YLAMINOPROPYLAMINE- ". MPD: META-PHENYLENEDIAMINE

"RFSINAIPPROX. WM 400,DEAPA6

" . R IL.1�N MVMAN 808, MPD , ,, , ,,

RESIN MWý 1038, MPD

R•ESIN MW -2638, MPD

RESIN M 384, MPD .

RESIN APPROJ,4MW V -

400, M0%AI 203, DEAPA

ALUMINUM

I; I -. I I I2 4 6 LU 12

76AL/L] ,X1o

3

300

Figure 12. Total linear thermal expansions reported by NBSýCryogenics O'Engineering Laboratory.

consisted of small glossy-pieces of resin; The larger the particle size, the lower thebutt-tensile-strength. Similar findings-were observed whenwfailed tensile-shear cou-pons bondedwith %FM-1000 were examined. The order of decreasing strength observedin thebutt-tensile tests when testing different classes of adhesives was almost identi-cal to that observed with thetensfle-shear coupons (see Ref. 11).

Figure 20 compares the impact strength of a, series of adhesives at -320° and 78°F.Once again, the epoxy-nylon adhesives resulted in higher strengths than all the otheradhesive systems evaluated. This would indicate that not only are they stronger atcryogenic temperatures, but that they are also considerably tougher. The Metibond406 resulted in much higher strengths than the FM-1000, indicating that although theyare, both epoxy-nylons there is adifference in properties resulting from differences inmethod of preparation, catalyst, and ratio of components. All impact specimens weretested with a Sonntag tester, using the normal hammer at a head velocity of 17 fps.

I Initial testing of the Metlbond 406 bonded specimens With a head velocity of 11 fps failedto break the specimens.

180

K -7

SGENERAL DYNAMICS j ASTRONAUTICS

In,

Figure 13. Ultrasonic test equipment for determination of adhesive modulus

properties.

,19


-0

!5.

00

S-20

A5.0 Al -2.5 Sn TITANIUM ALLOY0301 FH STAINLESS STEEL82024 T:- 3 BARE ALUMINUM

.30 t

-40 1_•200 -150 -100 -50 0 50 100

TEMPERATURE (°C)

Figure 14. Total linear thermal expansion of some adherends.

20

'Aw GENERAL DYNAMICS ASTRONAUTICS

100

0,65:35 EPON 828/VERSAMID 125ý-A- 50:50 EPON 828/VERSAMID 125U ARMSTRONG A-4" METLBOND 406,

-100 1

- fi

.1-501

0-100 00 0 5

TEMPERATURE (0 C)

Figure 15. Total linear thermal expansion of some adhesive systems.

21

• 4


METLBOND 406

ARMSTRONG A-4

50:60 EPON I ____828VERSAMID 125

c65:35 EPON __-__S8IViRSMID 125

5.0 Al -2.• soTITANIUM ALLOY

301- FH STAINLESS STEEL

2024-T3 BARE ALUMINUM,

02 4 6 0' 12 14

AL/L]7 X 103

3O00

Figire 16. Total linear thermal expansion of adhesives and adherends O }tested at Astronautics.

'Pi-46nsion testing of sandwich structures having both-stainless steel and-titaniumfaces bonded to 5.7-pcf fiberglass honeycomb core has been conducted over thetem-'perAture range. of ý4230-to '780F(see Figures 21 and 22). In~most cases, failure oc-curred primarily.in the core. The best adhesives for bonding sandwich- structures foruse at cryogenic temperatures, have proved to be the epoxy-nylon unsupported tapes and.the filled epoxy-phenolic supported tapes. The pi-tension tests have shown that the-3207F strengths are much higher than those obtained at 787F. The -423°F pi-tensionresults are somewhat lower than those obtained at -320!F, but are still considerablyhigher than the room-temperature strengths; therefore, room-temperature tensionallowables on the honeycomb core cahj be safely used when subjecting the core to cryo-genic environments.

Plateshear testing'of sandwich panels have also been tested from -4230 to 78 0F(see Figure 23). Again, an increase in strength was noted when going from 78° to-3200 F; however, average strength at -423°F was lower than the room-temperaturestrength. This may be a result of the small number of specimens tested. It appearsthat there Is a need for additional evaluation in this area.

22

_ _ __ _ _ --. __.

0 GENERALDYNAMtCS ASTRONAUTICS

10,000 -..

*' .STRONAU-r1C.. Co•NCENTRATED CHROMATE ETCHA APPLIED MANUFACTURING RESEARCH ETCH

ADHERENDS: 0.020-IN. EFH 301 CRES

8,000

6,000

SIN

S4,000•

WN

z

I-.

2,000 N

0

-450. -375 -360 -225 -1 50 -75 0 75 150S~TEMPERATURE (OF)

Figure 17. Tensile-shear strength of Coast Pro-Seal 777 vs. temperature.

Fatigue tests were run on~pit-tensien specimens having 0.032-in.,EFH-301 stainlesssteel faces, Hexcel, HRP-3/,16 core (5.5 pot) and epoxy-nylon adhesives. Specimenswere fatigued at -423°F, at a rate of 5 to 10 cps from 0 to 600 psi. As many as 23,j000cycles were run on one specimen without failure. Th, s specimen was then slowlybrought to a stress level of 900 psi before it failed.

One purpose of this REA was to attempt to correlate the adhesive strengths at cryo-genic temperatures with some basic physical properties of the resin system or thebasic molecular structure of the resin. As stated previously, thermal expansion datafailed to reveal any correlation between bond strength and resin system. Wor'k accom-plished at Narmco (Ref. 12) has also led to this conclusion. In fact, when investigatingthe thermal expansion data of two epoxy-nylon systems, it was found that the systemcontaining the most nylon showed less contraction than the system containing an ap-preciably lower amount of nylon. This-occurred even though nylon has a much highercoefficient of expansion than the particular epoxy resin used in the adhesives. However,

4N

w2

~' 4,OOC ____ ____ _____ _A__ _ _ _ _

GENERAL DYNAMICjS ASTRONAUTICS

1O,000.. ADHERENDS: ALUMINUM

0 DC*731 STATIC SHEAR (INSTRON UNIVERSAL TESTER)E DC-731. DYNAMIC SPEAR (BARRY SHOCK MACHINE)A METLBOND 406, F:ATIC SHEAR

8,000___ . METLBOND 406. DYNAMIC SHEAR

zID-6,oo 0 -

I-.

26,000

0I,,.4 -375 .300o -225 -150 -75 0 75 150

TEMIPERATURE (OF)

Figure 18. Static and dynamic tensle-shear strengths of adhesives teted

at'Hughes Atrcrao Co.

other factors ,such as degree of crosslinking, degree of cure, method of adhesive proc-essing, crystallinity, etc., can influence the thermal expansion of adhesive systems.

One of the problems with attempting-to correlate adhesive , strength with molecularstructure is that adhesive systems in most cases are proprietary,. Another difficultyis that the raw materials used in the adhesives&, as well as the compounded adhesivesthemselves, vary from batch to batch. Degree, of cure, length of'storage prior tobonding, method of manufacture, and many other factors affect the strength of thebond. Therefore, it is possible-to start with the same basic components and find thatthe final properties of the adhesives vary considerably, particularly at temperatureextremes. It Is not unusual to find a, large variance in bond strength of an adhesivefrom batch to batch. •

24I

,C GENERAL DYNAMICS ASTRONAUTICS

36,000

ADHERENbS: 321 STAINLESS STEELII* METLBOND 406* METLBOND 4041 AFTER PRMING WITH

30,000' -- METLBOND 4041,TYPE IIA FM-100O* HT-424

,- NOTE: SPECIMENS ETCHED WITH FPL STAINLESSSTEEL ETCH

•24,000 j ,

18,000 -------

12,000 - " - . .S-- .+- ... . - 4. - ---- .

6,ooo. . . . : ='

6,000 - - -

0

0450 -375 -300 -225 -150 -75 0 75 150'

TEMPE,. -JRE (OF)

Figure 19. Butt-tensile strength of various adhesives vs. temperature.

25

-.... -- u+. i

v.

GENERALDYNAMICS[ASTRONAUTICS

25

ADHERENDS: 321 STAINLESS STEELS50:;50 EPON 828/VERSAMID 125

* METLBOND 4069 HT-424

NOE FM- 1000 -20 NOTE: SPECIMENS ETCHED WITH FPL STAINLESS

STEEL ETCH

10'-

01 1-.. . . . F- . . . .. . . . .. . . . I %-

-450 -375 -300 -225 -T50 R 75 0 75 150

TMPERATURE (OF)

Figure 20. Impact strength of various adhesives vs. temperature.

26


1,500 . .FACES: 0.05 -IN. 5.5 Al - 2.5 LC TITANIUM ALLOYCORE: HEXCEL 3/16-IN. HRP, GFB WEB, 5.5 PCF

0 METLBOND 406a HT-424

1,200

900

600

1-4

IiJ300_ _

-450 -375 -300 -225 -150 -75 0 75 150

TEMPERATURE (F)

Figure 21. P1-tension tests of sandwich panels having titanium faces andfiberglass core.

27

S('1

0 i•


1,500FACES: 0.032-IN. EFH 301 CRESCORE. HEXCEL 3/16-IN. HRP, GF-13, 5.7 PCF WITH

METLBOND 406 ADHESIVE & HEXCEL 3/16-IN. HRPGFB WEB, 5.5 PCF WITH OTHER ADHESIVES.

1,200

I-.b%%NT

600 __

30-

7 1-450 -375 -396 -225 -1.50 -75 0 75 150

TEMPERATURE (°F)

Figure 22. P1-tension tests of sandwich panels having stainless steel facesandfiberglass core.

280, 28

GENERAL DYNAMICS. ASTRONAUTICS

20,000

2 1 8 .0 0 0 " ,, ,

0

"• 16,000

14,000

. 600

400

FACES: 0.032-IN. EFH 301 CRES"CORE: HEXCEL 3/16-IN. HRP, GF-13• 5.7 PCF

""6 ULTIMATE SHEAR STRENGTHI SHEAR MODULUS (CORE)

200 1 1-450 -375 -300 -225 -150 .75 0 75 150

TEMPERATURE (0 F)

Figure 23. Shear properties of honeycomb panels bonded with Metibond 406.

29 'V

[J

'GENERAL DYNAMICS] ASTRONAUTICS

ij • Infrared curves were run on the adhesives evaluated in this prqgmiam (see Figures

24 through, 34). Of prime interest were the four epoxy-nylon unsupported tapes (Metl-

.boj~dnd 406, AF-40, AF-41, and FM-1000), and the two epoxy-polyadmide systems"(Narmco 3135 and Resiweld No. 4). The generic name nylon refers to a member ofth'poliyamide family and is defined by the DuPont Co. as follows: "A generic termfor anyilong-chain synthetic plymeric amide which has recurring amide groups as an.intgrai par of the main polymeric, chain and which is, capable of being formed into ahiial entjin which -thestructural elements are oriented in the direction of the axis."

jh-eý n:yloin-materials commercially available are- made by self-condensation of4- " ,.Tiinojds or the .reaction &f dibasic acids with diamines. Those most commonly

-use ar[ pýolyhexamethylene-adtpamide (nylon-6/6), polyhexamethylene-sebacimidej(4y.ny1 /ii-.0)-i, nga1IkOXy methyil substituted polyhexamethylene-adipamide (nylon-B),

poyaproot., " py (.nyln-d6om , and polyamide from il-aminodecanoic acid(nylon-ll)•:Another plyamide commonly used is the "Versamid" type which is made by the con-densatioiof difinerized-vegetable oil, acids and suitable polyamino compounds. The-Verid type is, theone commonly -used-inf the room-temperature curing epoxy-polya••ies-(Narnco d _3135, Resiweld i No. 4, etc.). The condensation reactions for the'form, ki ofio fth•e• ,nyqon-6/6,and the Versamid type polyarnides are as follows (whereR R;a.•ibe hydr•.t or n-other lifioleic aciddimer group,, Ref. 13):

*• +,:.•l :i, ; ! i '"" '!"1'"4occccJ 4N(H,)~~H~. H 0 0 HO O0H HO 0S.. . .CH C . . ...

h

'ADIPIC'ACID- HEXAMETHYLENE- POLYHEXAMETHYLENE-ADIPAMIDE (NYLON-6/6)

0 I

C-OOH' C-OOH Cooir C-N-CH2-CHI-N-R(Ci(2 )7 (C, 2), ý(CHl)7 (CH2 )7

CH CH Cl, CH O H H

CII CH HC HC.I-(CH2 )7 -COOH HC CH-(CH2 )7 -C-N-CH 2 -CH 2 -N-R

CI12 CHI HC CII-CH2 -CH WCH-(CH 2 )4-CH3 HC CH-CH2 -CH CH-(CH2 )4 -CH 3I i' \ / 4 \ /

CHl CH CH CHi I ICH2 NH

CH + (CH 2)5 .. (CH 2)5 H2N-CH 2 C 2 (CHO)I I 'ETHYLENE DIAMINE I

(CH2)4 Ci 3 Cl13 CH3

CH3

-9, 12 9, 11 LINOLEIC ACID DIMER VERSAMWD-TYPE POLYAMIDELINOLEIC LINOLEIC ACID

ACID (ISOMERIZED)

30

-I

4 -- --- ---.. . . . . . .- t • • -- fi•• , . . ... . . . -~T • • • ... ,• • .•

GENERAL DYNAMICSi ASTRONAUTICS

When--.'n.nes having higher functionality than ethylene diamine are used, the resin-ous product will-have-active amine hydrogens (rather than-arnide hydrogens) capable ofreacting with epoxy-resins just as ordinary polyamines, Because of the bulk of thispolymer, the Versamid-type polyamide would not be-expected to-result in hydrogenbonding (steric hindrance effect), and in most.cases would result in an amine-type re-action with an epoxy.

CH -CHR 1-CH-CH\o/ \/2

CH -CH-RI-cH-CH + 2R 2-NH CH -CH-I 1 -CH-CH 0 CHi,-CH-R -CH-CH-2 22 1 2 1 11 2 1 1 1 -12

o 601 NH OH OH NH NH OH OH N-R2

I 12 InIR2 R4 CH2-CH-R 1-CH-CH 2

OH 0/

DIEPOXIDE OR PRIMARY PRIMARY CONEN- SECOND STAGE IN CROS"LININGEPOXY RESIN AMINE SATION PRODUCT

CONTAINING SEC-ONDARY AMINOGROUPS

In contrast to the preceding type of reaction, the nylon-6/6, nylon-6, etc., arevery "•near and are capable of aligning themselves so that hydrogen bonding is

#7) prevaent:

NH CO-IIN CO .HN CO-tIN Co..

CO.II N C NH-CO NH..O..IIN

,. UN ?CO-HN CO -, CO-HN CO..

..OCI Nit-Co N .I iN

NYLON-6 NYLON-6/6

The hydrogen bond is a weaker crosslinking bond; it has a dissociation energy of

about 2 to 5 Kcals/mole, compared to about 50 Kcals/mole for a C-C bond. The in-corporation of a sufficient number of hydrogen bonds as crosslinkages results in de-sirable properties at room temperature and imparts to the adhesive-the necessarystrength at very low temperatures without the penalty of brittleness produced by co-valent linkages. In the case of the unsupported epoxy-nylon tapes, enough catalyst isgenerally added to fully polymerize the epoxy. The epoxy can actually be considered

31

--------------------

Ig - -

GENERAL DYNAMICS ;FASTRONAUTICS

to act asa reactive diluent in contrast tothe reaction with the Versamid-type polya-rnide. This results in the epoxy-polyamide(Versamid type),being more brittle than theepoxy-nylon tnsupported tapes. ItXalso accounts for the toughness of the epoxy-nylonunsupported tapes at crydgenic tempieratures.

Examination- of -the infrared curves of the eOoxy-nylon unsupported tapes revealedthat they hadalmost identical spectra. There was no -attempt-to obtain relative per-centages of individual groupings be causeat tý,e tirme there was no means available ofaccurately controlling the specimen thickness. The.Metlbond 406 and AF-41 appear touse~dicyandiamide as a,2catalyst sinde only in these two systems did the cyano groupingat 4.5'to 4.6 microns appear. ItXis believed that FM-1000 uses diamino diphenyl sul-fone as a catalyst since the Sulfone grouping, at 8.7 microns is apparent-in its spectrum.The catalyst used in the AF-40 was not readily apparent. Because of the limitedimountt of information on'the exact composition of the -epoxy-nylon unsupported tapes

there can be no definite conclusions, drawn as to the effect of.catalyst type -on~the.cry!-genic properties.

The infrared curves 6n Narmco 3135A and-Resiweld No. 4-(Component A), are al-most identical to the published curve on Epon-828, -while the- spectra on Narmco-3135Band Resiweld No. 4- (Component B) are almost identical to the published spectra on' -2Versamid 125.

32

- 32

. ._______________ ____ ___________________

20,000 10,000 5,000 4,000 3.000 2,500 2,000 1,500 1,400 1,300 1.200 i,1(oo 1,000 950 9600 850 600 750 700

4 4 4

Hz~f~f ; 11 .IljZ I I I -I - -H 1I ~ ~ ,. ti.~~t~1f4

t4 L±L I,-

A, Hi fl q

*r 44-1 4-1 i ll

-4-444 0

biH itf Ili.c-T.I

2345 6 7 8 9 10 12ý 1 3 ý14

WAVELENGTH (MICRONS)

Figure 25. Spectrum of AF-bo 406.

A.7.1~4_m -T 1 11______

900 850- 800 750 700 650 20.000 10,.00. 5.000. 4,000 3,000 2,500 2100 1,500 1,400 1,300 1,200

Tri -t4 11111 11 1 fI I

Pt +It

j~ ~ Ii I H' 11AT 4

11- 12 13 14 15 ;zIf 1 1 i ,'

if- I I

+++ t-I I-t- R fIIl IIllMI 1111 41 IA I

ti 4 12 13 14M 15P 1t 2 3T 411 5 1 7 8

Fir 27 1pect

c0 -- -,,I t W 1 W T M ! 1 , 1ý ! . - ýj .. .

GENERAL DYNAMICS] ASTRONAUTICS

2.500 2,000 1,500 1,400 1,300 1,200 Y1,100 1,000 950 900 850 800 750 700 650at I 1ý I I' I I1 1l ! 1 11 1

'llt II I IT rN T T IIII

+T 1- 1 H I. I IV ,&

ii I fiI i I T t lil ITT I T t iT I

J I I I Ii f t Illt li t 1 1 1 11 il I I IIif I l I fi l I I I 1 11 1 I

ToI I1,'1I o l M T I I I i l III I lt I I I I I I I II P I I

11 P !t I... o 1 .1 11 l t - I I i . I I 1 1 o l 1 1 11 T

11 '11 ! i I IT T 411 1 11 I II oii I Is I TIi1 f l

A I I 11 11+ 11I ,:MT IfillI il11trT - 1QT1-7T

45 6 7 8 9 10 11 12 13 14 15


Tigure 26. Spectrum of FM-10OO.

2,50 2,00 1.500.1.400 1.300 1,200 1.100 1,000 950 900 850 800 750 700 650

IIT TT

IT 1 ITII 1.I If

ii; 14 i- i-- il-- i lt-- I I

4- -1- ýl T . 4 1 t I, f t M l t j [ ll I* .. . ..

4 5 6 7 8 9 10 11 12 13 14 15


Figure 27. Spectrum of AF-40.

33

_ _ coo

GENERAL DYNAMICS jASTRONAUTICS

20.000 10.000 5.000 4,000 3.Cm 2,500 .2.000 1,500.1.400 1,300 1,200 1,100 1.000 950 900 w5 80

I I, t l

'111414 1-11 1.411jH1.41.141.. 1 1 1'

~iTTI-411tfl ir~........u 41 . 1w ..... i T~h~~8~

41 1111~ 411 l~4TIP Jr1 Ii I~ Ii ftlm tt I t~~r---l t~l t "'- ill, t I., rt-

H -H .149 1.1 4+11'1 I~ 11- 1 1ii t i , I-tv irr 1 iti. llI

ý4*.w.1 .Htrtl-rrr -'t I I ýR7r I4 tf; W K . 1tI. I 1 -, 11 2I 34 A4 41 6. 7 814. 9:~4 10 17 12


Figure 28. Spectrum of Resiwveld No. 4, Component B.

N.' __T 4

K`--r-f-m i 7tm 1" '

850 800 750 700 650 20.000 10,000 5.000.4,0.00 3.000. 2.5-00. .2000. 1,5 00 1,400 1,300 1,200

:Jl iii ''p IT I - V , I i

IW I-:: Is

~~;:: UI., .. I..

14L 15

I~~ lilt---

-t1t 41I* -la U!

12 13 14 15 1 2 3 4 5 6 7 8


Figure 30. Spectrum of Narmco 31351,

2,500 2,000 1.500 1,400 1,300 1,200 1,100 1,000 950 900 850 Bo0 750 700 650

sit I, 1, 91: 9 1 91 '

t 11111 1

NOW 111'111 il I

it 1- 11 1l T I 1 111!ilIm

t, if t' JAM9 ~

If

TI 944 T, I.ý 1- .I1'i

I It it U9 W

*9 II ' I ,

4 56 7 8 9 10 11 12 13 14 15


Figure 30. Spectrum of Narmco 3135, Component A.

000 2.500 2,000 1,500 1.400 1,300 1.200 1,100 1,000 950 900 850 800 750 700 650

tt*tli 1 t It IT

1T9, I T

1- tH 11 111- . fil 1

I A t I I fl llMir ti Iititii till til1T ,+ 11

til till fil

f' Tr *Mti- liT_1TffJ- - it

4 TI 11 91 11 121 fil 14 1

WAVEENGT (MICRONS)1 f

Figur 31.Spectum o Narmo 315,...mponnt.B

"IT t" -I -- c0

20,000 10,000 5,000 4,000 3,000 2,500 2.000 1,800 1,600 1,400 , 1,200 1,100 1,000 950 900 850 B00 750 700

0 80 L*--: . -'J~ 'L

I A

z40 -I A~~

L1I U U. aHU.HjI

Z 20--1 k 4-

1 2 3 4 5 6 7 8 9 10 11l 12 13 M14WAVELENGTH (MICRONS)

Figure 32. Spectrum of Metibond 4041.

20,000 10,000 5,000 4,000 3,000 2,500 2,000 1,500 1.400 1,300 1,200 1,100 1,000 950 900 850 800 750 700

.....~.. . . ... IF l ilt III if~*~

tn i T t'ldV71 7j1 li ti ll ;;T, '1' I it ; I il I

-fljI. , F -. ý ... It I- IIH lilHilltdi'11; tt1Hl~ '1l l'! 'p IKA -H-III H -1 . I

900 850 B00 750 700 650 20.000 10,000 5,000 4,000 3,000 2,500 2,000 1,500 1,400 1,300 1,20(

;:q7 .1. .

S44 I 11 Ull P. Il'!1,

i I,' lyI

11 1 3 1 512 3 4 5 6 7 8

VAVELENGTH (MICRONS

Figure 34, Spectrumc

950 900 850 800 750 700 650

Jil 1 1111 ii f~ IiI Il 111 M A i

11 12 13 14 15

I~k

GENERAL DYNAMICS IASTRONAUTICS

2,500 2,000 1,500 1,400 1,300 1,200 1.100 1.000 950 900 850 800 750 700 650

11li ft I -.ill ,.II--t t

[fil .1-,1i-

i,1 IIH-V-IiW1I-VI liltI'l 1 , 11 l I l t N t 11 .

it I l i

WITH T ''I .1. 11 ill I

111 11 6-k l fil if 9 10 1 12 13 14!1

Figur 34. Spctu of11 APCO1 1219.tT

ITI

351


:3CONCLUSIONS

1. The Pre-B-Dnd 700 cieaning cycle resulted in the highest strengths at cryogenic

temperatures with stainless stbel joints bonded with FM-10(Vand Metlbond 406.2. An increase in the loading rate from 1,000 to 10,000 ppm at -320* and -423 0 F i

will result inapproximately a 40% increase in tensile-shear results. 7

3. Thermal cycling from 780 F to cryogenic temperatures shows no appreciablechange in the room-temperature tensile-shear strength of Metlbond 406.

4. Contrary to popular belief, test results indicate that the width of overlap will ap-

preciably affect tensile-shear strength.

5. The use of primers with epoxy-nylon adhesives will lower the tensile-shear

strength of the adhesive joints at cryogenic temperatures.

6. The bvaluation of Narmco 3170-7133 with stainless steel and titanium adherends

at -3200 and -4230 F resulted in very low strengths when compared to the values

obtained~at Narmco with 7075-T6 aluminum adherends.

7. No correlation of expansion data and tensile-shear strength at cryogenic tempera-tures was possible with the adhesive systems-evaluated.

8. Butt-tensile and impact tests indicated that the epoxy-nylon unsupported tape

adhesives-Were the best systems for metal-to-metal joints at cryogenictemperature.

9. Pi-tension and plate shear tests at cryogenic temperatures indicated that fiberglasshoneycomb core (5.5-5.7 pcf) will normally fail before the epoxy-nylon or filled

epoxy-phenolic adhesives.

10. No correlation between molecular structure and adhesivo strength at cryogenic

temperatures could be obtained because of the variation of adhesives from batchto batch, the proprietary nature of the adhesives, and the variations of degree ofcure, length of storage prior to bonding, method of manufacture, etc.

0 3

SRECEDING PAGE BLANKPAGE P' "k"'

• .

SGENERAL DYNAMICS ASTRONAUTICS L

REFERENCES

1. Standard Test Procedures for Reinforced Plastic and Sandwich Materiais. Ao P. fPenton, General DynamicsI Fort Worth Report No. FZM-2352, August-1961.

2. Improved Cryostat and Accessories for Tensile Testing at -423°-F. J. Fo Wats~n.and J. L. Christian, ASTM Bulletin.

3. Cryogenic Adhesive Evaluation Study. J. Hertz, Astronautics Report EAR- ¢-AN-032, January 1961.

4. Strength of Structural Adhesives at Temperatures Down to Mindis 4240°F. B.j Mo IMcClintock and M. J. Hiza, CEL, NBS, WADC Technical Rep0rt 59•260i,-November 1959.

5. Cryogenic Adhesive Properties of Bisphenol-A Epoxy Resins. M- J. HimzaandP. L Barrick, CEL, NBS, SPE Transactions, April 1961.

6. Journal of Applied Chemistry. N. A. de Bruyne, January 1956.

7. The Measurement of Dynamic Modulus in Adhesive Joints at VUltrasoricFreVuenqle" sA.G.H. Dietz, P. J. Closmann, G.M. Kavanagh, and J.N. RossenMT, presetat the fifty-third annual meeting of the ASTM, 26-30 June, 1950.

8. A Resonant-Bar Method for Determining the Elastic Properties of Thin nLalndia

W. A. Nolle and P.J. Westervelt, Journal of Applied Physics, Vol. 21,,, Ndj,4,,April 1950.

9. Adhesives for Low Temperature. N. N. Glemser, Astronautics memo toG.Lemke, February 1961.

10. Adhesives for Lunar Probe. H. M. Gonzalez, Hughes Aircraft Co., January1961.

11. Epoxy-Nylon Adhesives for Low-Temperature Applications. J. Hertz, Advancesin Cryogenic Engineering, Vol. 7, p. 336, 1961.

12. Development of Adhesives for Very Low Temperature Application. M. B. Smithand S. E. Susman, Narmco Research and Development, Telecomputing Corp.,Annual Summary Report, NASA Contract NAS 8-1565, May 1962.

O~? 339

R EC DING,•~ IAGE" BLANK -

GENERAL DYNAMICS ASTRONAUTICS 4

13. Epoxy Resins. H. Lee and K. Neville, McGraw-Hill Book Co, Inc., 1957.

14. Polyamide Resins. D. E. Floyd, Reinhold Publishing Corp., 1958.

15. Personal communications with M. B. Smith and S. E. Susman of Narmco Researchj - and Development, Telecomputing Corp.

^Il

40:

GENERAL DYNAMICSI ASTRONAUICS

BIBLIOGRAPHY

Adhesive Properties Minus 3007F to Plus 700°F. North American Aviation, AF33(600)-28469, ASTIA 151529, September 1957.

Epoxy Resins as Cryogenic Structural Adhesives. M. J. Hiza and R. M. McClintock,

NBS, Cryogenics Engineering Laboratory; also Modern Plastics, 35, 172, 1958; also

NBS Report 5093, 1957; also Low Temperature Strength of Epoxy Resin Adhesives,

NBS Technical News Bulletin, Vol. 24, No. 5, p. 84, 1958.

Strength of Structural Adhesives at Temperatures Down to Minus 4240F. W. M. Frost,

WADC Technical Report 59-260, November 1959. The Strength of Ten Structural Ad-hesives at Temp,.ratures Down to -424°F. Advances in Cryogenic Engineering, Vol. 5,p. 375, 1959.

Bonding Plastic to Metal for High Strength at Low Temperature. C. E. Eppinger andW. J. Love, Advances in Cryogenic'Engineering, Vol. 4, p. 123, 1958.

Externally Bonded and Sealed Insulation for Liquid-Hydrogen-Fueled Rocket Tanks.T. F. Gelder and V. H. Gray, NASA, Lewis Research Center, Advances in Cryo-genic Engineering, Vol. 5, p. 131, 1959.

Cryogenic Adhesive Evaluation Study. J. Hertz, Astronautics Report ERR-AN-032,January 1961.

Structural Adhesives for Cryogenic Applications. J. Hertz, Adhesives Age, August &1961.

Manufacturer's Literature, Narmco Research and Development, Telecomputing Corp.,J. Lincoln and B. Smith, 1961.

Epoxy-Nylon Adhesives for Low-Temperature Applications. J. Hertz, Advances in t_.

Cryogenic Engineering, Vol. 7, p. 336, 1961.

41

GENERAL DYNAMICS- ASTRONAUTICS

Properties of Foams, Adhesives and Plastic Films at Cryogenic Temperatures.R. N. Miller, et aL, Lockheed-Georgia Co., paper presented at the WashingtonMeeting of American Chemical Society, Div. of Organic Coatings and Plastics Chem-istryp March 1962, Vol. 22, No. 1.

Development-of Adhesives'for Very Low Temperature Application. M. B. Smith andS. E. Susman, Narmco Research and Development, Telecomputing Corp., AnnualSummary Report, NASA Contract No. NAS 8-1565, May 1962.

Development of Adhesives for Very Low Temperature- Application. M. B. Smith andS. E. Susman, Narmco Research and Development, Telecomputing Corp., QuarterlySummary Report, NASA Contract No. NAS 8-1565, July 1962.

Adhesive Bondi'ig of Metals. G. Epstein, Reinhold Publishing Corp., New York,

N.Y., 1954.

Interim Status Report on the Development of Wet-Wall Insulation for Saturn S-IV.Douglas Aircrbft Co., Report 38797, August 1961.

Adhesive for Lunar Probe. H. M. Gonzalez, Hughes Aircraft Co., January 1961. QCryogenicAdhesive Properties of Btsphend-A Epoxy Resins. P. L. Barrick andM. J. Hiza, NBS, Cryogenics Engineering Laboratory, SPE Transactions, April 1961.

Adhesives for Low Temperature. N. N. Glemser, AstronaUtics memo to G. Lemke,February 1961.

Development of Organic Sealants for Applications at Very Low Temperatures. R. I.Akawie, C. J. Bahnn, C. L. Segal, and I. M. Zelman, Hughes Aircraft Co., First,Second, and Third-Quarterly Progress Reports, NASA Contract NAS8-2428, Novem-ber 1961-April 1962.

Selection of Adhesives for Cryogenic Insulation. N. W. Steele, Boeing Airplane Co.,December 1961.

Effect of Temperatures From -70° to 600°F on Strength of Adhesive-Bonded LapShear Specimens of Clad 24S-T3 Aluminum Alloy and~of Cotton- and Glass-FabricPlastic Laminates. R. F. Blomquist, H. W. Eickner, and W. Z. Olson, ForestProducts Laboratory, NACA TN2717, June 1952.

42

M7


Research on Structural Adhesive Properties Over a Wide Temperature Range. H. R.

Merriman, The Glenn L. Martin Co., WADC Tech. Report 56-320, April 1957.

Strength Properties of Metal-Bonding Adhesives at Temperatures from -100 0 F to+ 800°F. H. W. Eickner, Forest Products: Laboratory, WADC Tech. Report 59-152,

November 1958.

Shear Strength of Adhesives in Stainless Steel and Aluminum Lap Joints at Temper-atures From -1000 to + 1, 000°F. R. M. Lulling and W. Z. Olsen, Forest ProductsLaboratory, ASD-TR61-497, April 1962.

Study of Adhesives for Insulation Panels. J. Hertz, Astronautics Report MRG-276,

November 1961.

Polyurethane Adhesive at -423 0F. J. Hertz, Astronautics Report MRG-95, August1959.

Organic Films for Repair of Leaks Through Cracked Spot Welds. J. Hertz, Astro-nautics Report MRG-281, January 1962.

Testing of Fiberglass Honeycomb Cores in Support of a Centaur Non-Vacuum TypeBulkhead Design. J. Hertz, Astro sautics Report MRG-326, July 1962.

Adhesive Testing. G. D. Peddle, Astronautics memo to J. Hertz, May 1962.

Thirty Day Evaluation of Foams and Honeycomb for Centaur Intermediate Bulkhead..M. D. Campbellh J. F. Haskins, J. Hertz, H. Jones, and J. L. Percy, AstronauticsReport MRG-312, April 1962.

Evaluation of Repairs for Cracked Spotwelds Applicable to Centaur Intermediate '-

Bulkheads. A. Bleich, L. Girton, and J. Hertz, Astronautics Report MRG-317,

May 1962.

43

Documents

UNCLASSIFIED AD NUMBER - DTICSAMSO USAF ltr, 28 Feb 1972 THIS PAGE IS UNCLASSIFIED Ig p~. -) I -0. SUSLA. Ca. 90045 oG=NERAL DYNAMICS ASTRONAUTICS A2136. (Pri 9-61 GENERAL DYNAMICS]