Full text of "NASA Technical Reports Server (NTRS) 19720024852: Nondestructive test of regenerative chambers"

See other formats

NASA CR- 1 20980

BAC Report No. 8654-953003

(NASA-CR-120980)

kegei ibbmxvb « b » ce

Malone, et ai l^e-L
-1972 132 p

nobdesteociive test of
Final Report G.ft.

Co.) Jul.

CSC L 14D

G3/15

N72-32502

Unclas

43502

FINAL REPORT

NONDESTRUCTIVE TESTS OF

NATIONAL AERONAUTICS AND SPACE ADMINISTRATION

JULY 1972

CONTRACT NAS 3-14376

Bell Aerospace Company. .... textroni

POST OFFICE BOX ONE

BUFFALO, NEW YORK 14240

NOTICE

This report was prepared as an account of Government-sponsored work. Neither the Unites States,
nor the National Aeronautics and Space Administration (NASA), nor any person acting on behalf
of NASA:

A. ) Makes any warranty or representation, expressed or implied, with respect to the accuracy,

completeness, or usefulness of the information contained in this report, or that the use of
any information, apparatus, method, or process disclosed in this report may not infringe
privately-owned rights; or

B. ) Assumes any liabilities with respect to the use of, or for damages resulting from the use

of, any information, apparatus, method or process disclosed in this report.

As used above, “person acting on behalf of NASA” includes any employee or contractor of NASA,
or employee of such contractor, to the extent that such employee or contractor of NASA or em-
ployee of such contractor prepares, disseminates, or provides access to any information pursuant to
his employment or contract with NASA, or his employment with such contractor.

2. Government Accession No.

1. Report No.

NASA CR-1 20980

4. Title and Subtitle

NONDESTRUCTIVE TESTS OF REGENERATIVE CHAMBERS

3. Recipient's Catalog No.

5. Report Date

July 1972

6. Performing Organization Code

7. Author(s)

G. A. Malone, R. Stauffis and R. Wood

9. Performing Organization Name and Address

Bell Aerospace Company
Division of Textron
P.O. Box 1

Buffalo, New York 14240

12. Sponsoring Agency Name and Address

National Aeronautics and Space Administration
Washington, D. C. 20546

8. Performing Organization Report No.
BAC 8654-953003

10. Work Unit No.

11. Contract or Grant No.
NAS 3-14376

13. Type of Report and Period Covered
Contractor Report

14. Sponsoring Agency Code

15. Supplementary Notes

Project Manager, Rudolph A. Duscha, Chemicaland Nuclear Rocket Procurement Section,
NASA Lewis Research Center, Cleveland Ohio

16. Abstract

Flat test panels simulating internally cooled regenerative thrust chamber walls were fabricated by electroforming, brazing and diffusion
bonding to evaluate the feasibility of nondestructive evaluation techniques to detect bonds of various strength integrities. Ultrasonics,
holography, and acoustic emission were investigated and found to yield useful and informative data regarding the presence of bond
defects in these structures.

<-o, OR ILLUSTRATIONS reproduced
CO in BLACK. AND WHITE

* If S

** l

17. Key Words (Suggested by Author(s))

Nondestructive tests
Regenerative chambers
Metal bond integrity

18. Distribution Statement

Unclassified - Unlimited

19. Security Classif. (of this report)

20. Security Classif. (of this page)

21. No. of Pages

Unclassified

202

For sale by the National Technical Information Service, Springfield, Virginia 22151

NASA-C-1 68 (Rev. 6-71)

NASA CR- 120980

BAC Report No. 8654-953003

FINAL REPORT

NONDESTRUCTIVE TESTS OF REGENERATIVE CHAMBERS

G. A. Malone
R. Stauffis
R. Wood

Prepared for

National Aeronautics and Space Administration

July 1972

Technical Management
NASA, Lewis Research Center
Cleveland, Ohio

Chemical Propulsion Division

Rudolph A Duscha

Bell Aerospace Company
P. O. Box 1

Buffalo, New York 14240

FOREWORD

The work reported herein was performed for NASA/LeRC under Contract NAS 3-14376 titled
“Nondestructive Tests of Regenerative Chambers.” Mr. R. A. Duscha was the NASA/LeRC project manager.

The authors wish to acknowledge important contributions to this report by several individuals. These
include the treatment of brazing and diffusion bonding by F. Pruett, electroforming assistance by J. Nowak,
nondestructive evaluation assistance by L. Vecchies, and metallographic analysis by C. Wright.

Mr. G. A. Malone, co-author of this report, was Program Manager. Messers, R. Stauffis and R. Woods
were responsible for the nondestructive evaluation work conducted.

ABSTRACT

Flat test panels simulating internally cooled regenerative thrust chamber walls were fabricated by
electroforming, brazing, and diffusion bonding to evaluate the feasibility of nondestructive evaluation
techniques to detect bonds of various strength integrities. Ultrasonics, holography, and acoustic emission
were investigated and found to yield useful and informative data regarding the presence of bond defects
in these structures.

iii

CONTENTS

Section Page

I SUMMARY 1

II INTRODUCTION 2

III TEST PANEL FABRICATION 3

A. Test Panel Design and Bond Patterns 3

B. Materials 5

C. Preliminary Bond Strength Evaluation 5

1. Electroformed Test Specimens 5

2. Brazed Test Specimens 20

3. Diffusion Bonded Test Specimens 21

D. Electric Discharge Machining Baseplates 23

E. Electroforming Test Panels 23

F. Fabrication of Braze Bonded Test Specimens 33

G. Fabrication of Diffusion Bonded Test Specimens 45

H. Post-bond Processing and Discussion 45

IV NONDESTRUCTIVE EVALUATION 57

A. Test Selection 57

B. Test Techniques 57

1 . Ultrasonics 60

2. Acoustic Emission 62

3. Elolography 64

4. Spectrum Analysis 66

5. Test Sequence 72

C. Test Results and Discussion 73

1. General 73

2. Ultrasonic 73

3. Acoustic Emission 75

4. Holography 79

5. Spectrum Analysis 87

D. Infrared Feasibility Study 99

1. Tests Methods 99

2. Testing Techniques and Results 99

V DESTRUCTIVE EVALUATION 104

A. Failure Testing 104

B. Metallographic Analysis 104

C. Bond Strength Determination 106

D. Correlation of Results 106

VI CONCLUSIONS AND RECOMMENDATIONS 108

VII REFERENCES 110

ILLUSTRATIONS

Figure Page

1 Panel Assembly Designs 4

2 Electroformed Full Bond Patterns 6

3 Electroformed Weak Bond Patterns ... 7

4 Electroformed Non-Bond Patterns 8

5 Brazed Full Bond Patterns 9

6 Brazed Weak Bond Patterns .. . . 10

7 Brazed Non-Bond Patterns 11

8 Reference Bond Defect Applied to Selected Brazed and Diffusion

Bonded Test Panels 12

9 Diffusion Full Bond Patterns 13

10 Diffusion Weak Bond Patterns . 14

11 Diffusion Non-Bond Patterns 15

12 Typical Configuration of Lap-Shear Specimens 17

13 Typical Electroform Bonded Lap-Shear Specimens 19

14 Braze Pack and Thermocouple Positioning in Vacuum Furnace . 22

15 Vacuum Hot Press - Door Open 25

1 6 Electrode for Electric Discharge Machining Channel Pattern in Nickel Baseplates 26

1 7 Typical Electric Discharged Machined Nickel 200 Baseplate 27

18 Electroforming Fixture for Deposition on Nickel Coverplates 28

1 9 Typical Electroformed Plate After Initial Surface Grinding to Remove

Edge Buildup 30

20 Masking Patterns for Electroforming Weak and Disbond Area on Channel Lands 31

21 Appearance of Bond Patterns After 0.0025 Inch Deposit 32

22 Coating Braze Foil with Photoresist for Photofabrication of Bond Patterns 34

23 Application of Bond Pattern Photomaster for Exposure of

Photoresist on Braze Foil 35

24 Development of Photoresist Bond Patterns for Brazing 36

25 Chemical Etching of Braze Foil Bond Flaw Patterns . . 37

26 Full, Weak, and Non-bond Braze Foil Patterns Before Chemical Etching 38

27 Photofabricated Copper Braze Pattern Located on Nickel

Baseplate Using Reference Holes 39

28 Assembly of a Full Bond Braze Test Panel 40

29 Etched Full Bond Braze Foil Pattern on Nickel 200 Baseplate 41

30 Etched Weak Bond Braze Foil Pattern No. 1 on Nickel 200 Baseplate 42

31 Etched Weak Bond Braze Foil Pattern No. 2 on Nickel 200 Baseplate 43

32 Etched Non-Bond Braze Foil Pattern on Nickel 200 Baseplate 44

33 Vacuum Brazing Furnace 46

34 Chemical Etching to Produce Diffusion Bond Pattern in Nickel 200

Baseplates 47

35 Weak Diffusion Bond Pattern No. 1 on Baseplate Before Etching 48

36 Weak Diffusion Bond Pattern No. 2 on Baseplate Before Etching . . 49

37 Weak Diffusion Bond Pattern No. 3 on Baseplate Before Etching 50

38 Diffusion Nonbond Pattern on Baseplate Before Etching 51

39 Ultrasonic Technique 61

40 Ultrasonic Reflector Method 63

41 Acoustic Emission Technique 65

ILLUSTRATIONS (CONTD)

Figure Pa ^ e

42 Illustration of Holographic Methods 67

43 Effect of Pressure on Hologram 68

44 Holographic Interferometry Technique 69

45 Spectrum Analysis Test Method Approaches 70

46 Typical LUtrasonic Results on Electroformed Panels 74

47 Overlay of Ultrasonic Result on the Hologram for Panel 4E 80

48 Overlay of Ultrasonic Result on the Hologram for Panel 7E 80

49 Overlay of Ultrasonic Result on Hologram for Panel 13E 81

50 Overlay of Ultrasonic Result on Hologram for Panel 15E 81

5 1 Overlay of Ultrasonic Result on Hologram for Panel 2 1 B 82

52 Overlay of Ultrasonic Result on Hologram for Panel 22B 83

53 Overlay of Ultrasonic Result on Hologram for Panel 34D 83

54 Overlay of Ultrasonic Result on Hologram for Panel 40D 84

55 Overlay of Ultrasonic Result on Hologram for Panel 49D 84

56 Effect of Thickness on Hologram 85

57 Effect of Flatness on Hologram 86

58 Set-Up Pulse-Echo/Top Surface Interface, Panel No. 3 89

59 Set-Up Pulse-Echo/Top Surface Interface, Panel No. 8 89

60 Set-Up Pulse-Echo/Top Surface Interface, Panel No. 3 89

61 Set-Up Pulse-Echo/Top Surface Interface, Panel No. 3 89

62 Pulse-Echo/Channel Interface, Panel No. 3 90

63 Pulse-Echo/Channel Interface, Panel No. 8 90

64 Pulse-Echo/ Back Surface Interface Reflection, Panel No. 3 91

65 Pulse-Echo/Back Surface Interface Reflection, Panel No. 8 91

66 Pulse-Echo/Back Surface Interface Reflection, Panel No. 37 91

67 Pulse-Echo/Back Surface Interface Reflection, Panel No. 49 91

68 Pulse-Echo/Back Surface Interface Reflection, Panel No. 20 91

69 Pulse-Echo/Back Surface Interface, Panel No. 3 92

70 Pulse-Echo/ Back Surface Interface, Panel No. 8 92

71 Pulse-Echo/Back Surface Interface, Panel No. 37 92

72 Pulse-Echo/Back Surface Interface, Panel No. 49 92

73 Pulse-Echo/Back Surface Interface, Panel No. 3 93

74 Pulse-Echo/Back Surface Interface, Panel No. 8 93

75 Pulse-Echo/Back Surface Interface, Panel No. 37 93

76 Pulse-Echo/Bond Interface, Panel No. 20 94

77 Pulse-Echo/Bond Interface, Panel No. 20 • 94

78 Pulse-Echo, Panel No. 28 95

79 Pulse-Echo, Panel No. 28 95

80 Pulse-Echo, Panel No. 28 95

81 Pulse-Echo, Panel No. 28 95

82 Through - Transmission/Land/Bond Composite, Setup 96

83 Through - Transmission/Land/Bond Composite, Panel No. 3 96

84 Through - Transmission/Land/Bond Composite, Panel No. 8 96

85 Through - Transmission/Land/Bond Composite, Panel No. 20 96

86 Through - Transmission/Land/Bond Composite, Panel No. 28 96

87 Through - Transmission/Land/Bond Composite, Panel No. 37 96

88 Through - Transmission/Land/Bond Composite, Panel No. 49 96

ILLUSTRATIONS (CONTD)

Figure

Page

. . 97

Q 1

y i

y z

Q75

Pif /-Vi otiH C'oir'h /T atirl Tntprfjirp. Panp.1 Mo 3

y^t

rilCIl dIlU V^dieii' .LdliU IXILClla^C/, r aiit/i i>u. o

Pif r*V» ntiH /T otiH Tntprfuop Panpl Mfl 9D .

rilCIl dllU ^dLCJl/ .LdllU UllCl idee, JTaiic;i

Pif r*Vi an H r^Qtpli /T utirl TntorfsiPP Ponpl ^do 9 8

.men dnu edieu/Ldnu imciidee, jr diici inu. z.o

Pi + r*Vi nnH r~ , a'fr*Vi /T anrl Trif'PrfciPP Panpl Wp 9 7

y /

riicn ana v^dien'.Lanu imeridee, rdiici i>u. j i

Pif r»V» q nH /T onH TtitprfdPP Ponpl Kfn 4Q

men dim ^dieii/mnu inieiidec, idiiei i>e».

Pliotnaranh of T inp Rran TrKnprtion Hn ihument

100

101

102

IR Line Scan Recording and Photograph of Nickel Bond Sample

A t*<=*a Qcon P r*ot*H inty of Mir*trpl RonH RptntTlp. No 7 . . .

101

103

r\lCd OL'dll IavLUI UlUg Ul i^iL^xvL/1 uunu uainpiv •

Photomicrographs of Electroformed-Thermally Diffused Test Panels

105

vii

TABLES

Number Page

I Test Panels to be Fabricated 3

II Material Mechanical Properties and Chemical Analyses 16

III Chemical and Mechanical Properties of 0.250 Inch Thick Nickel 200 Alloy 16

IV Evaluation of Processes to Produce Electro formed Bonds 18

V Braze Bond Lap-Shear Strengths and Brazing Parameters 20

VI Diffusion Bond Lap-Shear Strengths and Bonding 21

VII Electrodeposit Mechanical Properties 24

VIII Final Thickness Profile of Electroformed Test Panels 53

IX Final Thickness Profile of Brazed Test Panels 54

X Final Thickness Profile of Diffusion Bonded Test Panels 56

XI Nondestructive Test Rating Chart 58

XII Selected Methods Chart 59

XIII Spectrum Analysis Instrument Settings 71

XIV Nondestructive Evaluatuion - Electroform 76

XV Nondestructive Evaluation - Braze 77

XVI Nondestructive Evaluation - Diffusuion Bond 78

XVII Bond Strengths and Acoustic Emission Data 107

I. SUMMARY

Flat test panels were fabricated to contain nickel 200 baseplates with integral manifolding and
coolant passages enclosed by means of bonded coverplates, The bonding methods used were electro-
forming, brazing, and diffusion bonding. Inconel 600 composed the coverplates in the latter two methods.
Of the fifteen panels bonded by each method, three groups of panels were produced to contain planned
bond integrities of full, weak, and nonbond. In addition, the effect of thermal diffusion after electro-
forming was investigated on separate specimens.

Bond strengths were verified by testing lap-shear specimens prior to fabrication of nondestructive
evaluation test panels.

Potential nondestructive evaluation techniques were screened to determine those methods of
greater promise in detecting bond defects in chamber wall bondlines. Ultrasonics, holography, acoustic
emission, and spectrum analysis were ultimately selected. The bonded panels were subjected to these
evaluation techniques and rated on ability to evaluate bond integrity. The identity of the test panels
as to bond integrity was concealed by code numbering so that personnel performing the nondestructive
evaluation would have no prior knowledge of planned bond patterns or types.

The test panels were later tested to verify achieved bond strength by destructive testing and
metallographic analysis. Ultrasonics, holography, and acoustic emission were each useful in differentiating
the various bond types. It was found that these evaluation techniques could be employed most ad-
vantageously to ascertain bond types when applied sequentially, rather than as completely independent
methods.

II. INTRODUCTION

A conventional regeneratively cooled rocket thrust chamber is usually constructed by bonding a
liner (inner wall) to a shell (outer wall) by means of electroforming the shell to the liner, brazing, or dif-
fusion bonding. Integral coolant passages are introduced by milling or electric discharge machining the
required configuration into the liner or shell material before bonding. In the case of electroforming,
these passages may be introduced by a, variation in which the channel lands are electroformed on the
liner.

Fabrication of such devices is critical from a standpoint that detection of inferior bonds or leakage
paths for the coolant must be detected as early as possible in the fabrication process. Otherwise, ex-
pensive and time consuming manifold joining, flow, proof, and hot fire testing will be uneconomically
expended on a possibly defective piece of hardware.

Several nondestructive evaluation techniques appeared feasible for detecting these flaws or potential
failure areas before completion of fabrication.

Since the hardware is capable of being pressurized, introduction of controlled hydrostatic stresses
can be utilized to produce strains or finite bond movements possibly detectable by other less common
nondestructive means such as holography or acoustic emission. Using this philosophy, the subject pro-
gram was conducted to determine and demonstrate the usefulness of these nondestructive methods in
evaluating the integrity of a metal to metal bond.

III. TEST PANEL FABRICATION

Flat test panels were fabricated containing Nickel 200 baseplates with integral cooling passages. Cover-
plates of electroformed nickel or Inconel 600 were bonded over these baseplates to enclose the cooling pas-
sages. Bond defect patterns were introduced on selected panels and full bonds applied to the remainder for
subsequent evaluation nondestructively and destructively as discussed in this and later sections of the report.

A. TEST PANEL DESIGN AND BOND PATTERNS

The test panel design shown in Figure 1 was selected on the basis that the structure closely simulated
production type regeneratively cooled chamber walls and provided sufficient bonding surfaces for variation
of bond defect sizes and patterns. In this design, space was provided for location of pressure fittings and
transducers, necessary in the performance of some tests, without interference with data acquisition.

Pressurization manifolds were included at each end of the panel to allow passage of fluids through the
panel to assure passages were not completely blocked by inadvertent braze runout, to remove channel filler
materials necessary in the electroform method of bonding, and to purge air from the passages during hydro-
static pressurization work.

Material thickness selection was based on mechanical properties of the materials of construction, the
channel land width, and the ratio of channel width to land width. The coverplate and baseplate were both
designed to a thickness of one-eighth inch, which corresponds to the land width. The channel width was pur-
posely made twice that of the land in order to introduce peel, in addition to normal shear, as a mode of
failure in final destructive testing. With this design, it was anticipated that thermally bonded panels with full
bonds would fail in destructive test at approximately 6,000 psi pressure (41 .37 x 10 6 N/m. 2 )

Table I lists the test panels to be fabricated by method of bonding and type of bond required (i.e. -
full, weak, and nonbond). Each method of bonding, except the electroform-thermal diffusion combina-
tion, requires fifteen panels. Of these, five panels are of each bond type.

TABLE I

TEST PANELS TO BE FABRICATED

No. of

Type of

Materials of

Panels

Fabrication

Bond

Construction

Diffusion Bonding

Full

Inconel 600/Nickel 200

Diffusion Bonding

Controlled Nonbond

Inconel 600/Nickel 200

Diffusion Bonding

Controlled Weak Bond

Inconel 600/Nickel 200

Braze Bonding

Full

Inconel 600/Nickel 200

Braze Bonding

Controlled Nonbond

Inconel 600/Nickel 200

Braze Bonding

Controlled Weak Bond

Inconel 600/Nickel 200

Electroforming

Full

EF Nickel/Nickel 200

Electroforming

Controlled Nonbond

EF Nickel/Nickel 200

Electroforming

Controlled Weak Bond

EF Nickel/Nickel 200

Electroforming and
Thermal Diffusion

Full

EF Nickel/Nickel 200

Electroforming and
Thermal Diffusion

Controlled Nonbond

EF Nickel/Nickel 200

•2

Electroforming and
Thermal Diffusion

Controlled Weak Bond

EF Nickel/Nickel 200

Rather than utilize one bond defect pattern on each group of five panels for a particular bond type,
it was determined that more useful information would be possible if two or more patterns were used. By
this procedure, bond defect size and ratio of defect to full bond were varied within a bond type panel group.

The patterns approved for electroformed bonds appear in Figures 2, 3, and 4. Two full bond panels
were purposely planned to contain small reference defects to determine nondestructive evaluation sensitivity
in locating small isolated flaws and confirming the bond line position in such techniques as ultrasonics. Three
weak bond patterns were used to provide different total areas of weak bond and various sizes of individual
defects. These same patterns were applied to the nonbond panel group in order to enable planned direct
correlation of defect size, shape, and area between weak and nonbonds. This was possible in the electro-
forming method of bonding due to the fact that electrodeposit bond strength can be varied while maintain-
ing complete bond coverage over a given surface.

Braze bond patterns are shown in Figures 5, 6, and 7. With the exception of the nonbond group, a
reference defect (Figure 8) was applied to selected test panels to provide a means of accurately detecting the
bond line and determining the nondestructive method sensitivity in defining small defect patterns. Patterns
were varied on panels in the weak bond group to provide differing defect sizes, shapes, and total defect area.
The weak bond patterns and non-bond patterns are not similar because the bond type is achieved by planned
control of full bond area. The use of brazing stop-off compounds was not employed because of possible
interference with the nondestructive evaluation methods. Planned weak and non-bond regions were pur-
posely designed large to accommodate some expected braze flow.

The diffusion bond patterns, Figures 9, 10, and 1 1, are similar to the braze patterns in that reference
defects are provided on selected panels and bond strength is controlled by the area of planned full bonds.

B. MATERIALS

The materials of construction required for fabricating the test panels consisted of Nickel 200 for
baseplates, Inconel 600 for thermally bonded coverplates, and nickel anodes for electroformed coverplates.
The nickel anodes obtained were rolled, depolarized and of a sufficient purity for the intended application.
Properties obtained from the electrodeposits are discussed later in this section of the report.

The Nickel 200 and Inconel 600 sheets were purchased to appropriate commercial or military speci-
fications and certified analyses obtained to assure conformance. Metallographic sections indicated no defects
which would influence the test data or performance of contractual requirements. The certifications supplied
the information shown in Table II.

C. PRELIMINARY BOND STRENGTH EVALUATION

To establish and verify the electroforming, brazing, and diffusion bonding process capabilities for
fabricating full size test panels, preliminary test pieces were fabricated for bond strength evaluation. Test
results demonstrated that the desired bond integrities could be obtained.

1 . Electroformed Test Specimens

Lap-shear test specimens as shown in Figure 12 were produced to evaluate the effect of nickel
surface activation treatments on the bond strengths achieved after subsequent electroforming. Since the
mechanical strength of electroformed nickel as used in the project was much greater than that of the Nickel
200 alloy making up the baseplates, all electroforming to determine bond strength was conducted on 0.250-
inch (6.35 x 10~ 3 m.) thick Nickel 200 stock. This was done to assure that failure in the lap-shear test
would occur in the vicinity of the bond line. Certified analyses on the 0.250-inch (6.35 x 10* 3 m.) thick
Nickel 200 stock confirmed that this material was of acceptable quality and similar to the Nickel 200
to be used in the full-size test panels. Table III provides the certified test data.

ELECTROFORMED WEAK BOND-PATTERN I ELECTROFORMED WEAK BONDS-PATTERN 2 ELECTROFORMED WEAK BONDS-PATTERN 3

Figure 3. Electroformed Weak Bond Patterns

SAMPLE CODE NO. 20 (TEST)
21 (TEST)

BRAZED NON- BOND PATTERN

SAMPLE CODE NO

LAND NO.

. 26 (REFERENCE)
IT (TEST)

30 (TEST)

33 (TEST)

38 (TEST)

I 2 3 4 5 6 7 8

BLACK AREAS ARE PLANNED
NON-BOND DEFECTS

Figure 7. Brazed Non-Bond Patterns
11

Figure 8. Reference Bond Defect Applied to Selected Brazed and Diffusion Bonded

Test Panels (Scale in Inches)

DIFFUSION NON-BOND PATTERN

Figure 11. Diffusion Non-Bond Patterns

TABLE II

MATERIAL MECHANICAL PROPERTIES AND CHEMICAL ANALYSES

Material:

Condition:

Specification:

Size:

Heat No.:

Nickel 200 Sheet
Cold Rolled Annealed
ASTM B-162
1 /8 in. x 48 in. x 48 in.
533A

Inconel 600 Sheet

Cold Rolled Annealed and Pickled

MIL-N-6840A

1 /8 in. x 36 in. x 48 in.

NX 2907

Chemistry %

Actual

Required

Actual

Required

0.05

0.15 Max.

0.03

0.15 Max.

0.28

0.35 Max.

0.23

1 .00 Max.

0.05

0.40 Max.

9.58

6.0-10.0

0.005

0.01 Max.

0.007

0.015 Max.

0.05

0.35 Max.

0.25

0.50 Max.

0.02

0.25 Max.

0.22

0.50 Max.

15.81

14.0-17.0

99.52

99.0 Min.

73.85

72.0 Min.

Mechanical

Properties

Tensile Str. (psi)

63,000

92,000

Yield Str. (psi)

25,000

45,500

Elongation,

% in 2 in.

TABLE III

CHEMICAL AND MECHANICAL PROPERTIES OF 0.250 INCH
THICK NICKEL 200 ALLOY

Chemical Analysis (%)

C Mn Fe

0.03 0.26 0.03

0.005

0.03

0.01

99.61

Mechanical Properties

Yield Strength (psi)

31,000

Ultimate Strength (psi)

61,000

Elongation, %

Hardness, Rg

Condition: Cold Rolled, Annealed Sheet

Shear

Nickel 200 baseplates were cut to a size of 1 .0 by 8.0 inches (0.0254 x 0.203 m.) and the edges surface
ground for uniformity. All surfaces were chemically etched to remove oxides and other imperfections
which might affect bonding. A minimum of three specimens for each bond type (i.e., -full, weak, and
non-bond) were prepared for electroforming. The actual procedures used to produce each bond type
are discussed in Subsection E, Electroforming.

Electrodeposition was performed in a 200 gallon (0.757 m. 3 ) nickel sulfamate electroforming
bath at a current density of 60 amperes per square foot (5.574 amperes/meter 2 ) and an electrolyte
temperature of 1 10° to 120°F (317.2° to 322. 1°K). This current density was higher than that used to
produce the full size test panels. The current density only affects the rate of electroforming and mechan-
ical properties of the deposit. Bond strength is related to proper activation of the nickel substrate and
cathodic protection of this surface until the instant electrodeposition begins. It is essential that all traces
of oxides and organic contaminants be removed from the surface before bonding.

After electroforming, the specimens were surface ground to remove external roughness normal for
thick electrodeposits. Overgrowth (nodules) on the edges were removed by mechanical milling. Notches
shown in Figure 12 were milled into the coverplate and baseplate on selected specimens. The remaining test
bars were retained for use as test samples in practical application studies on some of the nondestructive evalu-
ation methods under consideration early in the project. Notches for lap-shear testing were then electric dis-
charge machined in these specimens and the shear strength determined. Bond strength results for those speci-
mens produced by the processes evaluated for test panel fabrication are shown in Table IV. It was noted that
producing notches for lap shear testing by electric discharge machining resulted in higher bond strength results
than were encountered on specimens with mechanically milled notches. Figure 1 3 illustrates typical lap-shear
specimens after test.

TABLE IV

EVALUATION OF PROCESSES TO PRODUCE ELECTROFORMED BONDS

Sample

No.

Bond

Type

Width

Length

Area

Breaking Load

Shear Strength

Inches

Meters

Inches

Meters

In . 2

Meters 2

Lb.

Newtons

psi

N/m 2

Full

0.996

0.0253

0.110

0.00279

0.1,096

7.071 x 10' s

4,850

21,573

44,250

305.1 x 10 6

Full

0.974

0.0247

0.125

0.00317

0.1218

7.858 x 10“ s

6,250

27,800

52,100

359.2 x 10 6

Full

0.997

0.0253

0.050

0.00127

0.0499

3.219 x 10' 5

2,265

10,075

45,400

313.0 x 10 6

Full

Retained for Sonic Frequency Analysis Evaluation

Full

0.993

0.0252

0.080

0.00203

0.0794

5.122 x 10~ s

3,450

15,346

43,350

298.9 x 10 6

Full

1.003

0.0255

0.117

0.00297

0.1178

7.600 x 10~ s

5,550

24,686

47,100

324.7 x 10 6

Weak

0.975

0.0248

0.095

0.00241

0.0926

5.974 x 10' 5

4,000

17,792

43,200

297.9 x 10 6

Weak

0.993

0.0252

0.095

0.00241

0.0943

6.084 x 10' s

4,325

19,238

45,850

316.1 x 10 6

Weak

Retained for Sonic Frequency Analysis Evaluation

1 5*

Weak

0.947

0.0247

0.100

0.00254

0.0974

6.284 x 10' s

2,700

12,010

27,800

191.7 x 10 6

Weak

Broke in Milling

Weak

Broke in Milling

Weak

0.981

0.0249

0.097

0.00246

0.0951

6.135 x 10~ s

200

470

3.24 x 10 6

22*

Weak

0.991

0.0252

0.120

0.00305

0.1189

7.671 x 10' s

2,750

12,232

23,120

159.4 x 10 6

Weak

Retained for Sonic Frequency Analysis Evaluation

Nonbond

Retained for Sonic Frequency Analysis Evaluation

Nonbond

0.995

0.0253

0.086

0.00218

0.0856

5.522 x 10~ 5

625

2,780

7,300**

50.3 x 10 6

Nonbond

0.997

0.0253

0.098

0.00249

0.0977

6.303 x 10~ 5

1,050

4,670

10,750**

74.1 x 10 6

* Notched by Electric Discharge Machining

Values Due to Full Bond on Specimen Edges (1/8 inch or 0.003175 meters, each edge).

2. Brazed Test Specimens

Lap-shear specimens were fabricated and tested to establish the optimum brazing parameters for
producing the various bond strength levels required on the full size test panels for nondestructive evaluation.
The samples consisted of Inconel 600 coverplates and Nickel 200 baseplates with overall dimensions of 5
inches by 3.1 inches (0.127 x 0.0787 m.). Each component material was 0.125 inch (3.17 x 10" 3 m.) thick.
The final braze metal selected was OFHC copper, 5 inches by 3.1 inches by 0.003 inch thick (0.1270 x
0.0787 m.; 7.62 x 10~ 5 m. thick). The photofabrication process (photo-resist masking and chemical etching)
was used to produce the defect patterns in the weak and non-bond samples. Table V lists the bond strength
results and brazing parameters for each bond type.

TABLE V

BRAZE BOND LAP-SHEAR STRENGTHS AND BRAZING PARAMETERS*

Bond

Type

Braze Temp.

Time At
Temp.
(Sec.)

Lap-Shear Strength

Strip 1

Strip 2

Strip 3

°F

°K

psi

IM/m 2

psi

N/m 2

psi

N/m 2

Full

2000

1366

43,650

301.0 x 10 6

47,200

325.4 x 10 6

44,450

306.5 x 10 6

Weak

2000

1366

25,200

173.7 x 10 6

47,950

330.6 x 10 6

36,650

252.7 x 10 6

Nonbond

2000

1366

36,500

251.7 x 10 6

15,750

108.6 x 10 6

21,000

144.8 x 10 6

* Vacuum was maintained at 5 x 1CT 5 Torr (6.665 x 10 3 N/m 2 )

The nickel and Inconel sheet stocks were sheared to size and belt sanded (240 grit) to improve the
surface condition. It was noted that the plates contained variations in surface flatness, as well as normal thick-
ness variations to 0.010 inch (2.54 x 10' 4 m.). Since plates were, in extreme cases, as much as 0.020 inch
(5.08 x 10" 4 m.) out-of-flat, brazing would be difficult to control.

To alleviate this condition, the Inconel 600 plates were stress relief annealed at 1600°F (1 144°K)
for three hours in vacuum using a hot press with an applied pressure of 750 psig (5.171 x 10 6 N/m 2 ). As
many as twelve plates were cycled at one time. Plates were electrophoretically coated with 0.03 micron
aluminum oxide particles to prevent bonding during the anneal. The Nickel 200 baseplates were similarly
stress relieved, but the annealing temperature was reduced to 1300°F (978°K) for the same period of time.

Stress relieving did not fully flatten the plates. Surface grinding was evaluated as a possible means of
obtaining flatness within 0.001 inch (2.54 x 10’ 5 m.). This was not satisfactory; bowing was still pronounced.

The process by which best flatness was obtained was as follows:

(a) Stress relief anneal plates with hot press flattening load, and

(b) Surface lap plates on both sides.

The resulting flatness variations in Nickel 200 and Inconel 600 ranged from 0.0005 inch to 0.0060 inch
(1.27 x 1CT 5 to 1.52 x 1 Or 4 m.) in the free-standing condition.

Initially it was planned to electrodeposit the copper braze metal on the Nickel 200 plates. An acid
copper electrolyte was used to deposit 0.002 inch to 0.003 inch (5.08"x 10~ s to 7.62 x 10~ 5 m.) of copper
for brazing. Because of edge buildup in the deposition process, it was necessary to polish the excess copper

away to obtain a flat matching surface for the Inconel 600 coverplate. A bright acid leveling bath was eval-
uated and found to solve the edge buildup problem. However, upon brazing a test coupon with this copper,
porous area developed. This may have been due to trace quantities of leveling agent incorporated in the
copper deposit.

Copper foil of 0.003 inch (7.62 x 10~ 5 m.) thickness was cut to the Nickel 200 plate size and
brazed successfully. To produce the defect patterns, the copper foil was coated with a commercial photo-
resist compound and cured. Photomasters of the defect patterns for weak bond and non-bond were applied
over the photoresist and and exposed. Chemical etching was used to remove the unprotected copper and
the pattern was produced. The remaining photoresist compound was removed with solvent.

The photofabricated copper foil patterns were applied over the Nickel 200 plates. Inconel cover-
plates were placed over the braze foils and brazing was conducted in a vacuum furnace. A baseplate and dead
weight of alloy D6AC were used to sandwich the plates being brazed. Aluminum oxide coated separator
sheets were inserted between braze specimens to prevent undesired bonding. The alloy D6AC was drilled
through the sides to accommodate thermocouples, Figure 14. This assured that the entire braze pack could
be brought to a temperature of 1900°F (131 1°K) and quickly elevated to 2000°F (1365°K) to accomplish
brazing (copper melting point is 1980°F or 1355°K).

After brazing, the composite panels were cut into three 1 .0 inch (0.0254 m.) wide strips, each 5
inches (0.127 m.) long and 0.250 inch (6.35 x 10" 3 m.) thick. Lap-shear notches were electric discharge
machined on each side to provide a bond with dimensions of one inch by 0.060 inch (0.0254 x 0.001 52 m.)
to be tested.

3. Diffusion Bonded Test Specimens

Specimens for evaluating bond strengths from the diffusion bonding process were identical
in size to those used in the brazing work. The photofabrication process was used to produce accurate defect
patterns in the Nickel 200 baseplates of the weak bond and non-bond specimens. Table VI lists diffusion
bonding parameters and the corresponding bond strengths obtained.

TABLE VI

DIFFUSION BOND LAP-SHEAR STRENGTHS AND BONDING PARAMETERS*

Bond

Type

Bonding

Time

(Minutes)

Temperature

Bond Surface
Load

Shear Strength

Strip 1

Strip 2

Strip 3

°K

psi

N/m 2

psi

N/m 2

psi

N/m 2

psi

N/m 2

Full

180

2100

1422

750

5.171 x I0 6

37,000

255.1 x 10 6

61,200

422.0 x 10 6

28,750

198.2 x 10 6

Weak

180

2100

1422

750

5.171 x 10 6

8,250

56.9 x 10 6

16,850

116.2 x 10 6

24,850

171.3 x 10 6

Nonbond

180

2100

1422

750

5:171 x 10 6

1,550

10.7 x 10 6

3,250

22.4 x 10 6

15,850

109.3 x 10 6

*Vacuum ranged from 4 x 10 -s to 5 x 10 4 Torr (5.332 x 10 3 to 6.665 x 1CT 2 N/m 2 )

Surface flatness was more critical on plates used for diffusion bonding than for the other fab-
ricating techniques. For this reason, all plates subjected to surface lapping were selected for flatness suitable
to diffusion bond. Surface preparation of the nickel was the same as used for brazing. However, to produce
the weak and non-bond patters, etching was used to relieve the Nickel 200 surface profile at predetermined
areas.

For chemical etching, the photoresist method used to fabricate braze foil patterns was applied.
The same defect photomaster patterns were used. To bond the photoresist compound (due to application
of a more severe and longer etch), nickel surfaces were phosphatized by immersion in a phosphoric acid

solution at 180°F (355°K). Photoresist was applied and exposed to the bond pattern photomaster (trans-
parency). After developing the photosensitive coating, chemical etching removed 0.003 to 0.004 inch
(7.62 x 1CT 5 to 10.16 x 1(T 5 m.) of nickel to create the planned flaw pattern. The etchant was nitric
acid containing ferric chloride. The photoresist was removed by solvent and the phosphatized layer was
stripped by a brief immersion in nitric acid.

The Inconel 600 plates were deoxidized and surface cleaned by dipping in nitric acid-ferric
chloride etch at room temperature. The etch rate was sufficiently slow that no thickness changes were ob-
served. After rinsing, a nickel plating strike was applied to the plates from a Wood’s nickel solution. This
was primarily used to afford a bonded layer of “pure” nickel as a diffusion aid on the Inconel.

Diffusion bonding was accomplished in a vacuum hot press. The furnace, Figure 15, has
resistance heated tungsten elements and molybdenum platens. The bottom platen was covered with an alumi-
nium oxide coated steel plate as a diffusion barrier. Over this was placed a Nickel 200 plate, etched side
exposed. The Inconel 600 coverplate was placed on top of the nickel plate. An aluminum oxide coated
separator was positioned on the Inconel, and the entire assembly was centered on the bottom platen. The
top platen was lowered to make contact with the assembly, but no pressure was applied to the ram.

The furnace was closed and evacuated to a vacuum of 5 x 1(T 4 torr (6.665 x 10" 3 N/m 2 .). At a
temperature of 1000° F (81 1°K), a load of 750 psig (5.171 x 10 6 N/m 2 .) was applied on the parts being
bonded. The diffusion bonding was performed at 2100°F (1422°K) for a three hour holding period at
temperature. After bonding, each plate was cut into three strips and notched for lap-shear testing in the
same manner as the brazed specimens.

D. ELECTRIC DISCHARGE MACHINING BASEPLATES

Fifty-one Nickel 200 base plates were sheared from 0.125 inch (3.175 x lCT 3 m.) thickness stock to
dimensions of 7.75 inches (0.1968 m.) long by 4 inches (0.1016 m.) wide. These plates were assigned code
numbers by stamping in the upper left corner on the side opposite that to be machined for channels. All
plates were electric discharge machined using a graphite composite electrode as shown in Figure 16. Dimen-
sional checks showed that all critical channel and manifold tolerances were met. Mounting holes for fixtur-
ing of the plates for electroforming were drilled into each end of every plate. These holes also served as
indexing positions for a template to be used later to machine the bolt hole pattern and pressurization ports
necessary in nondestructive evaluation with holography and acoustic emission. A typical base plate with
the channel pattern is shown in Figure 17.

The individual base plates were cleaned by vapor degreasing and alkaline cleaning to remove dielectric
fluid from the electric discharge machining operation.

E. ELECTROFORMING TEST PANELS

Mechanical properties of nickel deposits produced by the electrolyte operating parameters planned
for panel production were first obtained, Table VII

Twenty-one electric discharge machined baseplates were randomly selected for bonding by electro-
forming. This included the six plates for panels to be later subjected to thermal diffusion treatment to
determine if bond flaws could be repaired by this technique and verified by nondestructive evaluation.

Five base plates were randomly selected from this group to receive a full bond. A pentagonal shaped
fixture was fabricated from stainless steel sheet to support the plates during cleaning, masking, and electro-

TABLE VII

ELECTRODEPOSIT MECHANICAL PROPERTIES* (NICKEL SULFAMATE ELECTROLYTE)

Electrolyte Composition

Temp.

Current

Density

Mechanical Properties

Units

Metal

Ni Cl 2

Units

Ultimate

Strength

Yield

Strength

% Elong.
In . 2

°F

°K

Amp/Ft 2

Amp/M 2

oz/gal

9:4

1.2

4.0

120

psi

97,100

65,100

kg/m 3

70.4

9.0

322

2.79

N/m 2

669 x 10 6

449 x 10 6

oz/gal

9.0

1.4

4.1

110

psi

100,000

67,200

kg/m 3

67.4

10.5

316

2.79

N/m 2

689 x 10 6

463 x 10 6

oz/gal

9.2

1.1

3.9

105

psi

90,900

60,700

kg/m 3

68.9

8.2

314

2.79

N/m 2

627 x 10 6

418 x 10 6

*Data represents an average for six tests under each electroforming condition.

forming operations. Figure 18 illustrates this fixture with plates attached. The fixture was designed to pro-
vide close proximity between panels so as to minimize edge built-up along the sides. However, build-up on
the plate ends was not restricted due to the subsequent necessity to surface grind for proper sealing surface
for the pressure fittings used in non-destructive tests.

The plates were individually immersed in a nitric acid-ferric chloride solution to remove recast metal
from the electric discharge machining. Each baseplate was dipped in molten plater’s wax (melting range of
180° to 185°F or 355° to 358°K) to fill the channels. By repeated dipping, wax layers were built up to
completely fill the channels and manifolds. The excess wax was then trimmed flat to the panel surface using
a plexiglass scraping tool. Any residual wax films on the channel lands were removed by scrubbing the plates
with a bristle brush and a cleaning compound composed of Alconox (a commercial detergent cleaner) and
Shipley’s No. 1 1 Scrub Cleaner (a fine pumice-type compound). The panels were then rinsed and examined
for a water break free surface to confirm complete wax removal from surfaces to receive a full electroform
bond.

The panels were mounted on the electroforming fixture and the Nickel 200 activated for full bonding
by anodic-cathodic treatment in a 30% by weight solution of sulfuric acid. The panels were made anodes at
a current density of 50 amperes/ft 2 (4.64 amperes/m 2 ) for two minutes and cathodes at a current density of
100 amperes/ft 2 (9.29 amperes/m 2 .) for two minutes. While still wet with sulfuric acid, the fixture and panels
were transferred into a 200 gallon (0.757 m 3 .) nickel sulfamate electrolyte with current applied. The fixture
was mechanically rotated to assure each plate receiving equal exposure to the nickel anodes and to agitate the
cathode film of electrolyte sufficiently to dislodge hydrogen gas which tends to codeposit with nickel during
electroforming.

Agitation of the electrolyte was aided by compressed air (oil-less pump) bubbled through the bath
and by a high volume pump with an outlet directed at the parts being electroformed. The bath temperature
was maintained between 90° and 100°F (305° and 31 1°K) and the current density held at 30 amperes per
square foot (2.79 amperes/m 2 .) of cathode surface. A one mil (0.001 inch or 2.54 x 10" 5 m.) build-up of
fully bonded nickel was made. The fixture and panels were withdrawn from the electrolyte in order to
make the wax conductive so bridging of nickel over the channels could occur.

The wax was made conductive by the silver spray reduction method. A two nozzle spray gun was
used to apply simultaneously a stream of silver nitrate solution and an organic reducing agent such as formal-
dehyde. On impingement of these two chemicals the silver nitrate is reduced to metallic silver. Since metallic
silver on the Nickel 200 surfaces would be a contaminant leading to poor bonds, a way of removing this
material without disturbing the conductive layer on the wax was needed. The silver was efficiently removed

±1

1 {

# m

* I

J ! ^ J

i 1

Figure 15. Vacuum Hot Press Door Open

Figure 16. Electrode for Electric Discharge Machining Channel Pattern in Nickel Baseplates

(Scale in Inches)

Figure 17. Typical Electric Discharge Machined Nickel 200 Baseplate (Scale in Inches)

■ '

_jf r — ^

from the nickel by reverse plating in a separate tank of nickel sulfamate solution. After rinsing, the parts
could then be activated in sulfuric acid and electroformed as previously described.

After electroforming a build-up of 0.025 inch (6.35 x lCT 4 m.) and grinding, the panels had the
general appearance shown in Figure 1 9. Thickness of deposit at the bottom was slightly thicker due to
length of the anodes employed. The channel pattern is faithfully reproduced throughout the electroforming
due to initial conductivity differences in the nickel lands and the conductivized wax. Surface grinding and
polishing removed all visible pattern.

The expected condition of porosity in thin deposits over the silver conductivized layer on wax in the
channels was found. This was corrected by surface grinding the porous regions, filling the pores with a silver-
butyl acetate preparation, dressing excess preparation from the surface, and resuming the electroform opera-
tion. After wax removal, the residual preparation was flushed out with solvent. The effectiveness of this
procedure was evident when the panels were later exposed to pressures exceeding 1500 psig (10.34 x 10 6
N/m. 2 ) with no leakage.

The nonbond pattern panels were produced in the same manner as the full bond with the exception
of the initial 0.001 inch (2.54 x l(T 5 m.) deposit. The Nickel 200 surface was activated and a deposit of
0.001 inch nickel applied. Plater’s tape was applied over the lands and wax filled channels. Using a sharp
blade and metal straight edge, the pre-determined pattern was scribed into the tape.

After removal of the tape from the non-bond areas, the panels appeared as shown in Figure 20.

These plates were etched in nitric acid-ferric chloride solution until 0.0015 inch (3.81 x 1(T 5 m.) of metal
was removed. This exposed the original Nickel 200 surface in a pattern exactly as desired for disbonds.

For the ensuing electroform operation, the exposed areas were passivated to prevent bonding by dipping in
a sodium dichromate solution. The edges of the passivated areas were scratched with a sharp blade to pro-
vide places for “tack” bonding so the disbonds would not inadvertently separate. The etched disbond areas
were rebuilt by depositing 0.001 5 inch (3.81 x 10 rS m.) of nickel. Masking was removed and the entire
panel surface built-up by another 0.001 inch (2.54 x 1CT 5 m.) to lock the disbonds in place.

After conductivizing (described earlier), the entire nickel surface was electroformed to a thickness
of 0.025 inch. (6.35 x 1(T 4 m.) Appearance of the panel at this point is shown in Figure 21. The bond
pattern is readily visible. Roughness exists over the channels due to the silver Film and plating conditions
maintained in early deposition — low agitation and low current density were employed so as not to disturb
the silver bridge over the wax. The pattern and roughness were removed by surface grinding; copious
amounts of cooling fluid were applied so as not to disturb the wax. It was desired to leave the wax in the
channels until the pattern was removed by grinding, and the surface examined for defects or porosity. This
made subsequent repairs or build-up easier.

The weak bond panels were made exactly as the non-bond panels with the exception that the weak
bond regions were not passivated. Instead, the etched weak bond pattern in the 0.001 inch (2.54 x 1(T 5 m.)
build-up was subjected only to a dip in 30% by weight sulphuric acid (no current application), and the
plates were transferred to the electroforming bath with no current applied. After a delay of two minutes,
direct current was applied to start electroforming. The length of delay in starting the plating is proportional
to the strength of the bond obtained.

The electroformed-thermally diffused panels were produced by the same processes as the electro-
formed test panels. After electroforming, the panels were examined by ultrasonics for reference data be-
fore thermal treatment. The thermal cycle consisted of heating in vacuum to 1800°F (1255°K) for one-
half hour.

Figure 19. Typical Electroformed Plate After Initial Surface Grinding to Remove Edge Build-up

FABRICATION OF BRAZE BONDED TEST SPECIMENS

Fifteen Nickel 200 baseplates and a similar number of Inconel 600 coverplates were sheared from
0.125 inch (3.17 x 1CT 3 m.) thick sheet stock. All material was in the annealed and pickled condition.

0.125 inch (3.17 x 1(T 3 m.) diameter holes were drilled in the top right corners of each Inconel cover plate
to provide a means of suspending plates in chemical cleaning and plating solutions later in the specimen
fabrication.

It was observed that all baseplates and coverplates contained variations in surface flatness after
shearing. This same condition was noted in the preliminary work to determine strength values for various
bond integrities. The additional operation of electric discharge machining the channels and manifolds intro-
duced increased flatness deviation. This required modification of the previously developed procedure for
obtaining satisfactory flatness as follows:

( 1 ) Stress relief anneal sheared plates.

(2) Electric discharge machine (Nickel 200 plates only).

(3) Chemically etch plates in nitric acid - ferric chloride solution to remove all recast metal and
surface contaminants.

(4) Repeat the stress relief anneal.

(5) Surface lap the plates on both sides.

(6) Re-etch chemically to remove any surface contamination.

Resulting flatness variation in Nickel 200 and Inconel 600 ranged from about 0.001 inch to 0.006 inch
(2.54 x 1 CT S m. to 1.52 x 10 -4 m.).

Copper foil of 0.003 inch (7.62 x lOr 5 m.) thickness was cut to the Nickel 200 plate size and brazed
successfully. To produce the channel patterns, the copper foil was coated with a photoresist, Figure 22.

After curing, a photomaster of the channel and manifold pattern was applied over the photoresist and ex-
posed, Figure 23. Figure 24 illustrates several foil sections of various bond patterns after developing of the
exposed photoresist. The etching operation is illustrated in Figure 25.

Tabs were incorporated in the pattern design to maintain position of individual copper sections of the
pattern after etching of the foil. Figure 26 shows patterns for the full bond (being held), weak bond, and
non-bond (adjacent to the measuring scale). It will be noted that an inch long detail pattern of bond dis-
continuities exists on the weak and non-bond patterns. This is for the purpose of evaluating NDT method
sensitivity to flaw size.

The photofabricated copper foil pattern was applied over the Nickel 200 baseplate so as to line up
copper braze with the channel lands. For this purpose, reference holes were drilled in the nickel and etched
in the copper, Figure 27. An Inconel cover plate was placed over the braze foil, Figure 28. Typical braze
patterns for various bond integrities are shown in Figures 29 through 32.

The brazing was conducted in a vacuum furnace. A baseplate and dead weight of Alloy D6AC were
used to sandwich the plates to be brazed. Aluminum oxide coated stainless separator sheets were used be-
tween braze specimens to prevent undesired bonding. The Alloy D6AC was drilled through the sides to ac-
commodate thermocouples. This assured that the entire braze pack could be brought to a temperature of

d to resist

Figure 24. Development of Photoresist Bond Patterns for Brazing

Figure 27. Photofabricated Copper Braze Pattern Located on Nickel Baseplate

Using Reference Holes

igure 29, Etched Full Bond Braze Foil Pattern on Nickel 200 Baseplate

(Scale in Inches)

Figure 30. Etched Weak Bond Braze Foil Pattern Wo. 1 on Nickel 200 Baseplate

(Scale in Inches)

urc 31. Etched Weak Bond Braze Foil Pattern Mo. 2 on Nickel 200 Bascp

(Ccale in inches)

Figure 32. Etched Non- Bond Braze Foil pattern or Nickel 200 Baseplate

(Scale in Inches)

1900°F (131 1°K) and quickly elevated to 2000°F (1366°K) to convert the copper to an alloy with the nic-
kel and Inconel. The braze furnace is shown in Figure 33.

Blockage of channel passages by braze run-out was checked by pressurizing one of the two fitting
ports after filling the other with water. No blockage was expected due to the resistance of the copper nickel
alloy to flow. However, some run-out in the vicinity of the pressurization port (at one end usually) was en-
countered. This proved no serious problem because the exact extent of blockage could be determined by
ultrasonic “C” scan. Tne plugged end was placed down in a tank of hot water and the air in the channels
was displaced with liquid for subsequent hydrostatic tests during non-destructive evaluation.

G. FABRICATION OF DIFFUSION BONDED TEST SPECIMENS

Surface preparation of the Nickel 200 and Inconel 600 plates was the same as used for brazing. How-
ever, the weak and nonbond patterns were produced on the Nickel 200 baseplates by chemical etching to
relieve the channel lands at predetermined regions. The photofabrication process was again employed to
produce the flaw patterns which would permit selective acid attack on the surface to be removed to a speci-
fic depth of 0.003 inch to 0.004 inch. (7.62 x 1 CT S m. to 10.16 x 1 CT 5 m.) The etching operation is shown
in Figure 34.

The weak bond patterns are illustrated in Figures 35, 36, and 37. The baseplates in these pictures
have the photo resist applied, exposed, and developed to allow etching to occur in selected areas. Figure
38 shows the nonbond pattern.

The bottom platen of the diffusion bonding hot press was covered with an aluminum oxide coated
steel plate as a diffusion barrier. Over this was placed a Nickel 200 channeled baseplate with exposed channels
facing up. The Inconel coverplate was placed on the Nickel 200 and aligned by means of referencing holes
drilled at end locations where subsequent pressure fixture bolts would be inserted.

Another aluminum oxide coated separator was placed over the Inconel and the entire assembly was
centered on the bottom furnace platen. The top platen was lowered to make contact with the assembly, but
no pressure was applied to the ram.

The furnace was closed and evacuated to a vacuum of 5 x 10" 4 Torr (6.665 x 1C T 2 N/m. 2 ). At a
temperature of 1000°F, a load of 750 psig (5.171 x 10 6 N/m. 2 ) was applied on the parts being bonded.

The bonding was performed at 2100°F (1422°K) for a three hour holding period at this temperature

H. POST BOND PROCESSING AND DISCUSSION

During fabrication of the electroformed panels, it was necessary to stop the deposition process at
intervals to remove edge build-up. This was performed by grinding to obtain a flat surface considered
desirable in ultrasonic and holographic inspection. The bondline was expected to be out-of-flat due the
the variation in surface flatness of the nickel 200 baseplates. Flatness variation diagonally across the electro-
formed panels was as high as 0.020 inch (5.08 x l(T 4 m.) in some cases. This contributed to difficulty in
obtaining a desirable gate setting to determine flaws ultrasonically on the electroformed panels with only
0.020 inch (5.08 x l(T 4 m.) of deposit build-up. As the deposit thickness was increased, this effect was of
less significance.

All electroformed panels were ultrasonic examined after a build-up of 0.020 inch to 0.030 inch
(5.08 x l(T 4 m. to 7.62 x l(T 4 m.) of nickel as a coverplate. Reference panel No. 3 (full bond) and 8 (weak
bond) were subjected to a pressure of approximately 1270 psig (8.76 x 10 6 N/m. 2 ) for determination of

l&J

yii

f I %•

¥ I

it '

■i .

igure 33. Vacuum Brazing Furnace

Figure 35. Weak Diffusion Bond Pattern No. 1 on Baseplate Before Etching

(Scale in Inches)

Figure 36. Weak Diffusion Bond Pattern No. 2 on Baseplate Before Etching

(Scale in Inches)

VilVJf

Figure 37. Weak Diffusion Bond Pattern No. 3 on Baseplate Before Etching

{Scale in Inches)

Figure 38. Diffusion N inbond Pattern on Baseplate Before Etching

(Beale in Inches)

acoustic emission response. Panel No. 3 developed a bulge next to Land No. 8. Panel No. 8 emitted noise
and developed pin-hole leakage in the channel next to Land No. 8. The bulge was pressed flat on Panel
No. 3 and an electroform repair corrected the leakage on Panel No. 8. Re-examination of both panels by
ultrasonics gave no indication of detectable defect propagation. Since Panel No. 8 was a weak bond re-
ference panel which may have been influenced by the pressure testing, it was decided to change the refer-
ence designation to a companion specimen, Panel No. 2.

During the surface grinding operations, Panel No, 9 was inadvertently ground out-of-parallel due
to a fixturing problem. This panel was successfully repaired by electroforming and tested after full build-
up to about 0.060 inch (1.52 x lCT 3 m.) of nickel as coverplate.

Residual stress from the electrodeposited nickel also contributed to out-of-flatness and subsequent
thickness variations after flat surface grinding. This stress is tensile in nature and normal for nickel electro-
deposits from electrolytes not containing stress reducers. Thickness and flatness variability could have
been corrected by stress relieving, but this was undesireable due to possible thermal changes to planned
bond integrities. The resulting thickness profile (diagonally across the panel face) for each electroformed
panel is furnished in Table VIII.

After testing at a build-up of 0.060 inch (1.52 x l(T 3 m.) of nickel, Panel No. 1 1 was electroformed
to a final thickness of 0.1 10 inch (2.79 x 10" 3 m.) for comparison with brazed and diffusion bonded panels
having coverplates of similar thickness.

The electroformed-thermally diffused panels underwent blistering during the thermal treatment at
1800°F (1255°K). This effect was similar to laminations occurring at plating restarts and possible reasons
for this condition are discussed in Section V.

Braze bond patterns were successfully achieved as planned. Several panels exhibited minor varia-
tions or defects not planned. However, these panels provided useful information in the investigation. Panel
Nos. 23 and 33 leaked in the pressure fitting regions due to inadequate braze flow. Silver soldering, at a
temperature below that at which brazing occurred, was successful in correcting this condition.

The temperature at which brazing was performed resulted in some grain growth in the Nickel 200.
This would account for premature yielding of the channel webs during pressurization for nondestructive
evaluation. In order to minimize the effect of this condition on performance of holographic evaluation
and to enable destructive evaluation to be conducted, the Nickel 200 baseplates were strengthened by
electroforming a full bond layer of higher strength nickel on the backside of each brazed panel. There
were no indications that this operation adversely affected ultrasonics, holography, or acoustic emission
results.

In some instances, braze flow was greater than anticipated and planned non-bond or weak bond
regions had greater bond strength than desired. This did not appear to be a serious problem because the
metallographic analysis and nondestructive evaluation results generally confirmed the true bond condition
in existence.

The coverplate and baseplates for brazing contained previously discussed flatness variations corrected
to a serviceable condition by grinding and lapping. The resulting plate parallelism was varied and total panel
thicknesses after brazing and electroforming structural backing are shown in Table IX.

Planned bond patterns were achieved on the diffusion bonded panels, but to a lesser degree than
on the other methods. This is due to the difficulty of maintaining nonbond without the use of diffusion
stop-off compounds. Such compounds were purposefully avoided to prevent undesireable detection by the
nondestructive methods under evaluation.

2 r^ocooLnor^coLDCNomoooir)
a, incDtDCOLncouocDto^-r^^rcDCD^'

_£ ooooooooooooooo
2 qpppppqpqqqoooo
ooooooooooooooo

o CMCO^root— ncMrf^-cDr^r^coror^
C COCOCOCOCOCDCDCOCOLOCDLOCOCDLf)
— ppppppoopoppppp
ooooooooooooooo

jn ocor^ioinoLfir^iocMLOocomr^
J; cDr^^rLncDr^r^mr^^j-cDLOCDLDLO

« ooooooooooooooo
5 pppppppppppppoo
0 0 0 0*00000000000

o cococor-iflhO)(\jc)(Oinc 3 )^»-fNj
c OCDLOCOCOCDCOCOCDLOCOLOCDCOCO
— ppppppppppppppp
ooooooooooooooo

S 2 ooLorvCNOr^oor^ocNr^omr^
a> iDcoo^-LrjcDtncDin^riD^cDLDLO

« 2 ? ooooooooooooooo
^ pppppppopppoooo
ooooooooooooooo*

■g 05 (D in oo o n m 0 cm in o ra ro m

£ LOCDCDLOCDCDCDCDCDLOCOmCDCDCD

— pppppppppppppoo
do’dddddddddd o o* o

w» OOCOOOLnOOOLDOOCOLDmOCO
£ CNCNCOCOCNCOCNCOCNCNCNCNCMCVICO

+* oococooooocooocooocooocooocooo

Q> OOOOOOOOOOOOOOO
2 pppppppppppppoo

o’dddddddddddddd

<D ■.

■C Or^OOCOOCDOCOOr^OOOOCD^—
C ^^^^CN^^^CMCNCNJCMCNCNCO

d o d o* o’ o o o d o o o d o o

£ OOLOOOOOOOONOlflCOCOlflnOlfl
m CMCNCMCOCMCMi-COCNCMCMCMCNCMCM

cofocococQcooocorococoroococo

2 pppppppppppppoo
dodo o' dod o' dodo do

o 0 ) 0000050 ) 1000001 ^* 001^000
C CNCMCNCOCMCNCNCOCNCNCMCNCNCNCN

« ooiflifloooconointninioooco

m CMCMCMOQCNOOCNCOCMCMCNCSICNCMCM

'M cococococorooocooooooooooococo
a) ooooooooooooooo
5 poop o poop opoopp
oodddddddddddo’d

*£ ooocooo«“r^ocooooooocooo

2 CNCNCMCOCMCOCMCOCNCMCMCMC^CMCN

oodddddodddddd o’

CNICO'^LDCOr^OOOO’— CMCO^fLO

TABLE IX

FINAL THICKNESS PROFILE OF BRAZED TEST PANELS*

Meters

0.00602

0.00620

0.00650

0.00582

0.00612

0.00622

0.00607

0.00612

0.00577

0.00584

0.00569

0.00610

0.00630

0.00544

Inches

r^^fcoo)r-mo)^-r^.o^toco^r

C0^-liTCN^f^-00^CNir0CN^r}-'r-

CN <N CN CN CN CN CN CN CN C\ | CN 1 CN 1 CNJ 1 CN

ddddddddddddod

Meters

0.00589

0.00622

0.00643

0.00569

0.00630

0.00607

0.00599

0.00523

0.00587

0.00582

0.00607

0.00617

0.00549

* CO

Inches

CNLnoo^oooa>coco*-a}a)cocQ

oo^locn^cococoococnoo^i-*-

CNCNCNCNCNCNCNCNCNCNCNCNCNCN

ddddddddddddod

Meters

0.00592

0.00635

0.00653

0.00559

0.00635

0.00630

0.00612

0.00605

0.00584

0.00577

0.00612

0.00630

0.00533

Inches

(nONOOCO«-COOOM-(DO

coLOLncNLO'^'^-cocorocN^r^i--

CNCNCNCNCNCNCNCNCNCNCNCNCNCN

ddddddddddddod

Meters

0.00300

0.00262

0.00272

0.00287

0.00241

0.00224

0.00239

0.00234

0.00221

0.00239

0.00254

0.00221

(/I

Inches

cocor^comco^j-cNr^^r^roor^
»**00«-0)00c350)000)0)0000
-'-.--qqqqqqqr-^o
d o o o o d d d o o o o o o

Meters

0.00302

0.00249

0.00269

0.00277

0.00241

0.00218

0.00249

0.00221

0.00241

0.00236

0.00269

0.00262

0.00211

Inches

CT>cotoa>incooor^r^.ir)oocoooco
<-oooo>coCT>coooa>ooooo
o ^ r- o q o o o o o <— r- o

ddddddddddddod

Meters

0.00302

0.00244

0.00274

0.00267

0.00239

0.00226

0.00257

0.00218

0.00221

0.00251

0.00244

0.00264

0.00208

Inches

CDCDOOLO^fCDt— cOr^-CDCOrf^-CN
^-O)OOO)COOOOCOO)0OOCO
^qr-t-qor-qqqq^r-q

o d d d d d d d d d d d d d

Test

No.

r^co^-CNCO^rcoooO'— comoom

t-t-CNCNCMCNCMCNCOCOCOCOCV)^

</>

-Q

~Q>

— to

> O
a) V

_Q i-

cd a)

■a •-

0) Q.

-Q a

‘C CO
O —

00 -T-J

<u c
■o ra

8 g

o .E

-C -o

^ CD
"O to
C 1/5

o g
a S
S3 «

i- M-

O ^

n ®
.2 o

4 -* CD
to C
O * 4 —

CL O
*

Table X presents measured thickness variations resulting from flatness grinding operations required
to achieve planned bond patterns. Electroformed full bond backing was applied to the backside of the panels
to minimize channel web yielding during tests.

It was originally planned that a section from an experimental chamber be subjected to nondestructive
evaluation. This segment was sectioned for evaluation but was not tested. The section contained channels
of various depths and a porous region existed in the liner which had a thickness varying from 0.030 inch
to 0.019 inch (7.62 x 10" 4 m. to 4.83 x l(T 4 m.). Repair of this section to provide useful test data did not
appear feasible at this time. Test results on flat specimens indicated liner deformation was likely to occur
at pressures anticipated for acoustic emission evaluation of the liner-to-shell bond.

TABLE X

on necessary flatness grinding and lapping operations.)

IV. NONDESTRUCTIVE EVALUATION

A. TEST SELECTION

The initial phase of the contract required selecting a maximum of four nondestructive techniques
with which to conduct a feasibility test program on fabricated panels. In order to achieve a completely
objective program, a comprehensive literature search was performed to evaluate all of the candidate non-
destructive test techniques for their application to the testing of regenerative chambers. The primary ob-
jective was to find a test, or test combination, with the ability to detect good, bad, and weak bond con-
ditions in the narrow joint lines produced by electroforming, diffusion bonding, and brazing. Many tech-
nical articles were reviewed; however, it was generally concluded that very little information was available
which was applicable to the component geometry and fabrication methods under consideration. Refer-
ences 1-27 in Section VII list the relevant literature reviewed. It was thus considered pertinent to conduct
practical feasibility evaluations using those techniques which appeared to have the most probability of
success based on literature supplemented by fundamental test theory.

For schedule reasons, it was decided to utilize test coupons which had been prepared to demon-
strate the ability to produce good, non, and intermediate bond integrity. The coupons used were in every
case fabricated by the electroforming technique. These were chosen since the weak bond condition could
be created easily, with bond occurring over the entire surface rather than by isolated areas of total non-
bond. Totally bonded weak bond is the most critical condition to determine by nondestructive testing.
Representative specimens were evaluated by the techniques and organizations shown below:

Test

Ultrasonic

Holographic

Sonic Spectrum Analysis
Acoustic Emission
Infrared

Test Conducted By

Bell Aerospace Company
Bell Aerospace Company
Aerotech Laboratories
Dunegan Research Corp.
Automation Industries

Based on this work, Table XI was compiled to provide a rating of all tests considered for this
program. It is important to note that the comments, limitations and recommendations made in this and
subsequent tables of this report are related only to the materials and bond characteristics investigated
under this contract and should not be considered in any other context.

The first four test methods in Table XI were subsequently proposed to NASA and after discussion
these were approved. A provisional test sequence was also established. The selection, reasons and antici-
pated capabilities are summarized in Table XII.

It was considered that part of the limitations associated with the infrared evaluation might be
attributed to the thickness of the test specimen. In order to assure that no method was unjustly elimi-
nated it was agreed to subject three of the electroformed test panels, when only 0.020 in. (5.08 x 10" 4 m.)
thickness of electroform was present, to infrared testing in addition to the other tests.

B. TEST TECHNIQUES

Following the selection of the four test methods, preliminary development was performed to op-
timize and standardize the actual test techniques.

TABLE XI

NONDESTRUCTIVE TEST RATING CHART*

Test Method

1.1 Ultrasonics

1.2 Spectrum Analysis

1.3 Acoustic Emission

1.4 Holography

1.5 Infrared

1.6 Radiography

1.7 Eddy-Sonic

1.8 Acoustic Imaging

1.9 Radiation Gaging

1.10 Liquid Crystals

1.11 Neutron Radiography

1.12 Eddy Current

1.13 Microwave

1.14 Fluoroscopy

1.15 Magnetic Particle

Rating Subject

Availability

A off shelf

B requires modification

C not readily available

Cost

A below $15,000

B $15,000 -$25,000

C above $25,000

c 11

A 12

Ease of Use

A good

B fair

C difficult

B 5

Reliability

A good

B fair

C poor

B 2

c 6

Ease of Results

Interpretation A easy

B requires expertise

C difficult

A 14

B 4

c :

Applicability to Bonding
Techniques on Nickel A all three

B two

C one or zero

c 8

c 9

c 10

c 13

Application on Thrust

Chamber A fully

B partially

C not at all

B 1

B 3

B 7

Potential to Detect

Incipient Flaws A good

After Firing B fair

C poor

Points Rating A 2 points

B 1 point
C 0 points

Recommended for Further Evaluation

Yes

Index of Notes

*NOTE: The grading system used here is intended to judge the test methods applicability to this program only and should not

be considered to present the general merits or limitations of the test discussed.

1 .

Requires expensive tooling and recording devices.

Applicable to thin nickel 0.020 in. approximately.

To be determined.

Applicable to brazes only.

Spot check only.

10.

Insufficient sensitivity.

Requirements not established.

11.

Cost high size is currently too small for this application.

Laboratory use only.

12.

Investigation costs high compared to chances of success.

To be determined.

13.

Poor sensitivity and repeatability.

To be proven. Tooling requirements unknown.

14.

Fully bonded weak bond strength not determinable.

TABLE XII

SELECTED METHODS CHART

fctf

5 CO

£3 ft

>* w

sa h

73 «*
cd On

73
p

O cd
X> ^
c (U jH
3 <j <D

^ 3 'g

(S S w

O g O

. 2 •§ 3

cd co '55

T| ‘2 M

P o
O s >

73' >
2 8
« o

■g 5
g C
J 5 o

P o

o ^

C <D
CO X)

S o

CL _

® § g

I § S3 *

£ U W o

S C/3

cd <d

« a

73 <L>

P £

O C/3

■§ 2? g

3 u >

9 § |

IfS
2 «« >■
<U 2 3

•J: ^ x

1 1 1

2 «2 CL

£ Q &

to «

0) <D
M

*2 ° 3
a 73 .»
too 0> M
c .3 73
■■2 3 «>

i ?*

cd X

8 ta
2- 3

* CL X «■

W cs 2 H

rr m *— < r^i

£ »
s ^

.2* M

u o

2 ‘a 1
5 a

t3 X

| §■
m &

CL o
X o
«« S 3

X CD
CL X

0 a*

C *M

1 1

6 ^

p 2^

73 .5

j=S

•3 2T

C cd

O **3

•g

« c
8 O
a g
» 6
H _

2 3

• co

G T3

£ 04
I *-* CO

>• a

X> H

3 £ 73

s» « s

c 4 J M

2 £ 2

,g « g

C CO CD

3 2

CL a)

£ a

o .
H h

p g>

3 X
O «
o P

xj o 4>

P O

S Sjs

2 f 3

rC -f ^

,ts 4 ) (D
£ £ £

<D CD

p *2

> £ 2

1 'S

^ cr

£ o

cd t-i
<n .52

2 p

S -2

g g
5 £
x o

> 2
X 2
81

p is

.2 *S

5 g

8 g

73 .2

co cd

o >

-m <D
Qi 1>

3 B

CL cd
O £

o *3

CL ^3
W cd

£ c

P o

W) P
.3 ed
3 -4->

^ O

Q c

£ -g

1 §•

co &0

too •£

.£ B
t3 §
.2 3

«w co
cd

J3 73
CL C

«M 2

o >9

CL O
cd P
P to
C P

g g

cd ^

CL 73
P

8 2

73 .1

2 «

X» p

-g 2

O CD

O >

too .

73 73

P &

(D O

^ 3

o «

2 £

r. i-i

% «

(D O

— < M

X) CL

1 -3

2 CL

CD .C

*> 73 'O
OOP

x o o

too O

g £

o >,

on x>

c S
«Sf

P 73 o

.5 g> o

o P tJ
ut « ed

V "Z O

p £ P

^ cd

too

,o P

o c

CL/ #-

M H
CL <5

J3 di

' JN CO

73 CL 73

>» 2
X) i 2

. p
o
x>

52 £

cd 5/5 3

CO ^ p

CD P o
M — 1 .2
7 <D t/i

— .cS -S3

■ TO

: O

73 £ .*3
‘> O 73

v-i cS O
CL cd X

> O
2 a
CL xi

H %

p <u
P o
a> -r*
73 co
CD g
73 2

■8

4 >

„ c
o 2

CL P
O o -

£ CD

c»

73 ^
P ^
0 ) O

Oh — *

4 ) too
73 P
P

’O P

<L» M

to *2 «?
H o £)
■a m m
tL CL CL

o 2

O CD
M •«

CL >*
co g

p 2

■M O

8 e

73 2
p p

O lyj

CD 73

x 8 S

P o
*w g

<D to
+-* CD
73 **

X 'P

H 'I

>»

ffi.

*2 o w
2 § •“
a o a

s;^

77 co P

1 g> a
£ «<

r-> +-*

* co

.tS 0>

^ <D

§ i

CO ^

p 2

P Oh

P H-.

77 , O

5 a

PL 73

1. Ultrasonics

Ultrasonic inspection is based on emitted and received short wave length and relatively low
energy vibrations which are transmitted through the material under evaluation. In many ways the behavior
of the sound within the material and at joint lines or flaw locations can be considered analogous to that of
light passing thru different media interfaces. As such this sound may be closely monitored by choice of
the angle of sound incidence, focusing, and the selection of frequency dependent on sensitivity/resolution
and penetration desired. In general higher frequencies are used for increased sensitivity while the lower
frequencies yield greater penetration.

There are several means of employing the sound which are dependent on joint configuration
and the inherent nature of the flaw associated with the fabrication process.

The more commonly used of these are:

a. Pulse-echo longitudinal wave which introduces the sound normal to the entry surface
and detects delaminations and bond defects parallel to the entry surface.

b. Pulse-echo shear wave which introduces the sound at any specified angle and detects
defects oriented off-normal to the surface such as lack of fusion in weldments and longi-
tudinal seams in raw materials.

c. Thru-transmission which uses two transducers (one transmitter and one receiver) located
on opposite sides of the part, and also detects delaminations and bond defects especially
where several interlaminar bonded layers are involved.

d. Reflector method which is similar to the thru-transmission method with the exception
that a smooth surface replaces the receiving transducer and acts as a reflector, returning
the sound back to the transmitting transducer. This method is useful on thin materials
in which the front and back surface signals cannot be electronically separated on the
scope (resolved). It detects delaminations and bond defects.

All of the aforementioned techniques can employ focused transducers which concentrate the
sound similar to a light lens and thus increase sensitivity. A “gate”, which permits monitoring and record-
ing of the results at any selected interface, can also be used in conjunction with these techniques. A “C”
scan recorder provides a planar X-Y axis facsimile of the gated interface.

The perferred technique for inspecting bonds of this nature is the pulse-echo longitudinal
wave method, using a high frequency (15 - 25 MHz) focused transducer. Using this approach, it was possible
to gate the bond interface and achieve a high sensitivity level. The technique shown in Figure 39 was es-
tablished for evaluation of all full thickness electroform (0.060 in. or 1 .52 x 10” 3 m. thickness) diffusion
(0.120 in. or 3.04 x 10T 3 m. thickness) and braze (0.120 in. or 3.04 x l(T 3 m. thickness) bonds.

With the aim of determining bond quality at the earliest possible fabrication stage, it was de-
cided to attempt ultrasonic inspection of the electroform panels at the 0.025 in. (6.35 x 10“ 4 m.) buildup
stage. However, it was determined that the pulse-echo method previously established could not be used due to
resolution limitations. These were principally associated with gating and differentiating entry and bond inter-
face signals. Instead, a technique based on the reflector method previously described was employed. This

<x>

Figure 39

technique consists of introducing 10 MHz focused sound (longitudinal wave) which is transmitted
through the panel to a smooth glass reflector where the sound is reflected back through the panel to the
transducer. Any changes affecting attenuation in the sound due to thickness, surface irregularies, or bond
variance can be seen at the reflector as negative amplitude changes. Gating the reflector for these negative
changes provides a “C” scan similar to the pulse-echo method with the exception that factors other than
bond variance are also recorded. Figure 40 graphically displays the technique used for this study. The 25
to 45 percent trigger setting (recording threshold level) is indicated, since this was varied between panels
and greatly influenced the resulting sensitivity. The need to change this recorder threshold setting occurred
due to variations in surface finish, thickness, and parallelism. The test was only sensitive at the 45 percent
setting where more ideal panel conditions existed. A further technique innovation used in testing the 0.025
in. (6.35 x 10" 4 m.) thick panels by the reflector method was the introduction of 500 psi (3.45 x 10 6 N/m. 2 )
internal pressure. The theory behind this approach was that the capability of detecting intimate contact
nonbonds open to pressurization would be enhanced due to the increase in the nonbond separation width.
The choice of 500 psi (3.45 x 10 6 N/m. 2 ) maximum was made for two reasons:

a. To minimize the risk of panel distortion, and

b. To avoid destroying low level acoustic emission data.

The study was conducted using the reflector method by first obtaining a “C” scan, pressuriz-
ing the panel, repositioning it exactly as before, and obtaining a second dynamic scan without changing any
instrument settings. The scans were than visually compared for any nonbond growth resulting from the
pressure.

2. Acoustic Emission

Unlike ultrasonics, acoustic emission does not introduce sound in any manner, but instead
only listens. This listening may be performed within selected frequency ranges and be permanently re-
corded. Sound emissions from material deformation travel to a broadband transducer via surface, longi-
tudinal and shear waves, where they are converted to an amplifiable electrical signal. Since the test is de-
pendent on noise emitted during material movement, the application of acoustic emission requires some
form of stressing to create these inaudible emissions of sound energy. These emissions may be monitored
and recorded as:

a. Summation - graphically

b. Rate - graphically

c. Real-time-oscilliscope

d. Direct - tape recorder

Since all acoustic emission feasibility work was performed on lapshear specimens, several un-
knowns existed with respect to instrumentation settings and test pressurization methods. These consisted
of:

a. Gain settings (decibel selection)

b. Filter requirements (frequency range)

c. Pressure (stress)

d. Recording method

Figure 40. Ultrasonic Reflector Method

Based on limited information available from a test run by Dunegan Research Inc., and a trial
panel tested at BAC, it was decided to attempt testing on panel 3 using an 80 db gain setting, 1-3 KHz
filter setting and bottled nitrogen gas as a means of pressurization to 1 ,500 psi (10.34 x 10 6 N/m. 2 ) maxi-
mum. Panel 3 was chosen for safety reasons as it represented an almost full strength bond. The digital
counter was set in an intermediate range (x 100) for a summation count in the memory mode. The
memory /summation recording method was chosen since manual pressurization could not be applied con-
sistently enough for acceptable rate recordings. Furthermore, the count obtained periodically (from the
memory recorder) would enable simple rate graphs to be constructed. The result of this test on Panel 3,
at 0.025 in.(6.35 x 10' 4 m.) buildup, was that no noise emission was detected either on the X-Y recorder
or the audio monitor. Some evidence of leakage was noted. As a result, it was decided to conduct the
next test on Panel 8, also at 0.025 in.(6.35 x 10' 4 m.) buildup and containing weak bonds (i.e., one which
was expected to emit sound), using hydrostatic pressure so that pressures up to 3,500 psi (24.13 x 10 6
N/m. 2 ) could be achieved. In addition, it was decided to remove all filters. During the test some noise
was emitted in reaching 2,500 psi (1 7.24 x 10 6 N/m. 2 ). This was so low that it was decided to decrease
the Multiplier level from x 100 to x 10 in order to obtain full scale Y axis deflection. After reaching 3,500
psi (24.13 x 10 6 N/m. 2 ), a short evaluation of gain was made by increasing the setting-first to 90DB and
then to 95DB. At these settings, pump and background interference was experienced. The panel was found
to show some evidence of distortion over the channel region, and thus it was concluded that all further
acoustic emission tests would be conducted at full build up.

As a result of this work the test technique shown in Figure 41 was established for testing the
remaining panels.

3. Holography

There are three basic techniques for producing holograms. These are:

a. Real-time (dynamic flaw detection)

b. Time-lapse (static, captured image, flaw detection)

c. Time-average (inherent and load induced stress patterns)

The principles of approach with these techniques are briefly outlined below:

a. To produce a real-time image, a hologram. is taken of the test object and developed be-
fore the object is deformed by stress. If the hologram is then viewed so that the recon-
structed virtual image is superimposed on the object, any stress-induced deformation
will appear as a dynamic display of fringe patterns. Data concerning the object is not
restricted to static object displacements, and fringes can be formed or modified by any
localized motion or deformation of the surface. Changes in the topology or shape of
the surface cause instantaneous variations in the fringe geometry.

b. To produce a time-lapse hologram an initial exposure is made but not developed. The
object is then slightly deformed statically, and a second hologram is recorded on the
same film. When developed and reconstructed, this double exposure displays fringes
superimposed over the object image. These fringes are the result of interference between
the first and second object beams which represent the two positions of the object. These
fringes present a highly accurate recording of the deformation undergone by the test
specimen. Surface regions above sub-surface discontinuities are deformed differently

ACOUSTIC EMISSION TECHNIQUE

Transducer Location

from the rest of the surface as a result of the applied stress, and hence, produce
anomalies in the fringe patterns which are readily discernible.

c. Similarly, a time-average hologram may be produced by exposing the object while it is
undergoing dynamic vibrational movements. The reconstructed image displays fringes
that accurately reveal the dynamic deformation of the test specimen.

During the test method selection it was postulated that real-time holography would be con-
ducted. The optimum test response pressure was to be established on each panel by viewing the fringe re-
sponse as the pressure was increased. It was also intended to record the real-time fringe patterns on a
polariod photograph. Although this approach was made on the initial 0.060 in.(l .524 x 10' 3 m.) thick
electroformed panels after acoustic emission testing, the results were disappointing in that image quality/
repeatability was poor and the fringe patterns were almost impossible to consistently capture on a photo-
graph.

In view of these difficulties, the other holographic techniques of time-lapse and time-average
were than evaluated practically. Time-lapse was found to produce excellent quality holograms which were
both repeatable and reproducible. Time-average was found unsuccessful, and it was therefore decided to
standardize on the time-lapse technique. Figure 42 shows typical examples of the results obtained. The
choice of time-lapse, however, meant that a standard test pressure must also be established. Electroform
Panel 3 was tested at several pressures, Figure 43, and review of these led to the selection of 500 psi (3.45
x 10 6 N/m. 2 ) gas pressure for the test, since it was the lowest pressure showing good definition.

Subsequent testing revealed that although this pressure worked well on all electroformed
panels, the thicker (0.120 in. or 3.048 x 10' 3 m.) diffusion and braze bonded panels required higher pres-
sured (up to 1 ,450 psi or 1 0.00 x 1 0 6 N/m. 2 ) to attain similar results. Figure 44 details the finalized holo-
graphic technique.

4. Spectrum Analysis

This test differs from the previous three methods in that it is not currently used for nonde-
structive evaluation. The equipment is designed to analyze transducers, not materials. However, it was
considered that by using the equipment in conjunction with a transducer (having a known spectrum) and
varying the material (i.e., bond), the resulting spectrum “shifts” might characterize the bond and provide
a feasible test application. Since the method is basically that of a specialized ultrasonic technique, many
approaches are feasible. Three such approaches were selected for study under this contract. These were:

a. Pulse-echo longitudinal wave using focused immersion transducers at 10, 15 and 25 MHz.

b. Thru-transmission longitudinal wave using a focused immersion 10 MHz transducer.

c. Contact pitch and catch using a fixed angle dual 5 MHz thickness measuring type
transducer.

Figure 45 portrays the three approaches and illustrates the typical reflectoscope cathode ray
tube (CRT) response from each method.

The instrumentation used to conduct the study consisted of a Sperry 72 1 Reflectoscope using
the HRL or 10N pulser-receiver to set up the test, a Hewlett Packard oscilloscope to display the RF sound
trace (to select which interface to gate), a Branson/Aerotech UTA (for gain control and gate selection), and
a Hewlett Packard analyzer (for frequency control and display).

[ffl

ILLUSTRATION OF HOLOGRAPHIC METHODS

250 psi 3E
(1.72 x 10 fi N/m. 2 )

750 psi
(5.17 x 10 6 N/m. 2 )

500 psi 3E
(3.45 x 10 6 N/m. 2 )

1000 psi
(6.89 x 10 6 N/m. 2 )

EFFECT OF PRESSURE ON HOLOGRAM

Figure 43

HOLOGRAPHIC INTERFERROMETRY TECHNIQUE
JODON HS 1C SYSTEM

Regulated to 500 psi
For Electroformed Process;

13R0 ' 1440 for Braze and Diffusion
Bond Process.

Method I

Pulse-Echo Longitudinal

CRT Response

Transducer

Top

Surface

Interface

Sound Direction

Back

Surface

Interface

Channel Interface
Land-Bond Interface

T-Transmitter R -Receiver
Method 2 Through Transmission

Method 3 Pitch and Catch

T R

Figure 45. Spectrum Analysis Test Method Approaches

All immersion transducers were 3/4 in. (1.905 x 10‘ 2 m.) dia., lithium sulfate 1.5 in. (3.810
x 10' 2 m.) focal point, with the following serial numbers:

Freq.

Supplier

A/N

Sperry

13799

Branson

CF 27940

Branson

KF 8916

The pitch and catch transducer was a Sperry type SRL-Z model 50B1265/ECN 5 MHz.
The settings used for these instruments are given in Table XIII.

TABLE XIII

SPECTRUM ANALYSIS INSTRUMENT SETTINGS

Oscilloscope Unit

180A/1801A/1821A

Magnifier

10X

Display

Int.

Ext Input

Position A

0.1 V/Div

Display

Alt B

Position B

0.5 V/Div

Polarity

- Negative

Input

Delay

Delayed 1 0 ju sec

Sweep

Auto Ext

Scope

+ Positive AC

UTA Unit

Coupler

Attenuation

0-20 DB

Damping

Gain

50 Max

Gate Delay

Gate Width

Rep Rate

Max

Pulser-Trigger

Int.

Spectrum Analyzer

141T/8553B/5552A

Center Freq

Bandwidth

300 KHz

Scan Width per Div

Input Attenuation

Clipper

Approx 10%

Scan Time

2 MS/per Div.

Linear

1 MV/per Div.

Linear Sensitivity

Video Filter

Off

Scan Mode

Int.

Scan Trigger

Line

Note: The instrument settings with a “V” indicate those which were variable.

Applying spectrum analysis to selected panels required some standardization, since many
factors in theory could be monitored from the spectrum envelope. All of these factors were not necess-
arily influenced by the bond interface and therefore could lead to misintrepretation of the information.
The surface condition, thickness, attentuation factors, grain size, and velocity differentials (impedance
mismatch) all influence the resulting spectrum trace. Practical experimentation indicated that the major
change occurred to the center frequency as a result of impedance mismatch. For this reason, it was de-
cided to restrict the evaluation to comparison of center frequency shift only. In order to achieve this,
it was necessary to vary certain instrument settings (marked V) to provide a constant amplitude envelope
for correlation.

5. Test Sequence

Following the establishment of Test Techniques for all methods, it was considered essential
to determine test sequence. Based on theoretical data, the following approach was anticipated.

(a) Ultrasonics - To establish basic integrity and provide a 1 :1 ratio map of the bondline.

(b) Holography - To view real-time displacement and stress patterns during the acoustic
emission test and thus locate emission sources.

(d) Spectrum Analysis - To determine bond characterization potential on selected areas of
panels based on the other test results.

Items b and c were initially planned to be conducted simultaneously using the real-time holographic
method to 1 500 psi (10.34 x 10 6 N/m. 2 ) maximum gas pressure and acoustic emission to the same
pressure.

However, problems were encountered which made this approach impractical. It was dis-
covered that the real-time holographic method was not sufficiently reproducible to permit taking high
quality consistent polaroid photographs. This was solved by using the time-lapse technique. It was then
found that 1500 psi (10.34 x 10 6 N/m. 2 ) maximum air pressure was not sufficient to adequately stress
the panels for the acoustic emission test. This was resolved by using hydrostatic pressure applied by a
hand pump (in stepped pressure increments of 500 psi or 3.45 x 10 6 N/m. 2 ) to 3,500 psi (24.13 x 10 6
N/m. 2 ) maximum. 3,500 psi (24.1 3 x 10 6 N/m. 2 ) yielded too much surface movement during the time-
lapse technique and lost the hologram entirely. Coupling these changes with the irreversibility phenomenon
in acoustic emission (Kaiser effect), it was decided to separate the tests in an attempt to minimize data loss
and maintain comparability.

The use of the spectrum analyzer was never intended as an overall scanning test, but purely
as a spot check for bond character (relative strength) in areas highlighted by the other test techniques.

In view of the many variables and approaches discussed under the test technique section, and the need for
an extensive development program found during the initial tests conducted, it was agreed with the NASA
Program Manager to restrict testing to a limited number of panels and evaluate more than one approach.

As a result the original planned test sequence was not followed. The actual sequence used

was:

(a) Ultrasonics - Initial map of bond line.

(b) Acoustic Emission - 0 to 3500 psi or 0 to 24.13 x 10 6 N/m. 2 (in 500 psi or 3.45 x 10 6
N/m. 2 increments) for bond integrity.

(c) Holography - 500 psi (3.45 x 10 6 N/m. 2 ) on electroformed panels, and 1 500 psi (10.34
x 10 6 N/m. 2 ) maximum on brazed and diffusion bonded panels to establish relationship
with the preceding tests.

(d) Spectrum Analysis - Separate study on selected panels only.

Note: This finalized sequence should not be misconstrued as the ideal order. Instead, it
reflects the practical approach taken to prove test feasibility. Further consideration is given
to test sequences under the discussion section.

C. RESULTS AND DISCUSSION

The results of three tests, ultrasonic, acoustic emission and holography on all panels are presented
grouped in accordance with the fabrication process in Appendix A. Where a result is indicated by Panel
number alone in the following text, the illustration is contained under that “A” designation in the appendix.
The fourth method, spectrum analysis, is presented as a separate entity, as is the limited infrared evaluation.

1 . General

In considering the merits and correlation of the selected test methods it is essential to take
into account both the panel design limitations and fabrication anomalies encountered during the program.
These have in all cases been resolved or explained with respect to their source. Certain individual test
panel variation, such as flatness and thickness parallelism, have made complete comparison of fabrication
processes by test method difficult, since a measure of “nonstandardization” was often a resultant feature.

In discussing the merits of each test method these factors will be dealt with in more detail. An opinion
of their effect on test results is offered.

The overall objective of the contract, which was to establish feasibility of new test methods,
is considered to have been achieved. Three of the four selected methods showed high merit. These were
ultrasonic, holography and acoustic emission. A more detailed discussion by test method follows. In
order to facilitate the evaluation, Tables XIV, XV and XVI have been prepared. These tables give a synop-
sis of the results illustrated in Appendix A.

2. Ultrasonic

The pulse-echo technique used on all bonds proved to be an excellent screening method for
nonbond condition. As was anticipated, it did prove unsuccessful in determining the true weak bond con-
dition in the electroformed panels. The best correlation with planned defects was evident in braze bonded
samples. The diffusion bond process presented more difficulty in achieving planned defects as discussed
in Section III. Ultrasonic evaluation proved a valuable tool in demonstrating the true nature of the actual
bond patterns achieved. This was later verified by metallographic analysis. In general, the reflector tech-
nique used on the thin (0.025 in. or 6.35 x KT 4 m. thick) electroformed panels produced a less sensitive
test for detecting nonbonds, Figures 46(i) and 46(ii). However, in the case of Panel 12, Figures 46 (iii) and
46 (iv), it did show nonbond which was not found at the full buildup thickness by pulse-echo. The pre-
sence of this planned defect (at the base of Land8) was clearly confirmed by hologram Panel 1 2. The

reason for the pulse-echo not seeing this defect is due to the surface finish.

Surface finish was also a major factor restricting the clarity of the reflector technique. It proved
to be a deterrent to ultrasonic evaluation of the electroformed panels at the thicknesses studied. At least

M W

-It-

»v

part of the apparent lack of response obtained from the 500 psi (3.45 x 10 6 N/m. 2 ) pressure tests, e.g.,
Figures 46 (i) and 46 (v), was due to surface condition coupled with the fact that only thin electroformed
samples were evaluated using this less sensitive reflector method. Another factor which may have influ-
enced the pressure study was noted during acoustic emission and destructive testing. The 2: 1 ratio of
channel to land favored a baseplate and/or coverplate yielding prior to bond line disturbance. It is there-
fore considered, that the use of pressure to increase sensitivity warrants a more critical evaluation.

3. Acoustic Emission

An intercomparison of the results shown in Tables XIV, XV and XVI does not indicate the
high rating and applicability which is predicted for this test. In order to assess the true promise of this
test, it is necessary to consider the electroform results separately, because it is the only fabrication tech-
nique to give three strength conditions — full, weak and nonbond. The other fabrication methods pro-
duced varying degrees of bond/nonbond which resulted in the acoustic emission study being more related
to critical defect size. Another experience encountered on the brazed and some diffusion bonded sam-
ples, but not the electrofonned, was the yielding of the base plate. This probably affected stressing of
the bond interfaces, and hence affected the emission data. Consideration of the electroform data shows
a distinct trend as related to true weak bonds. A weak bond, regardless of defect pattern, consistently
created high emission counts. This is thought to be attributable to the steady yielding of the entire weak
bond areas. The full strength bonds, as expected, emitted the least noise, with one exception Panel 1,
which appeared to represent a weak bond condition. Examination of the metallurgical results showed that
some edge rupturing of the initial electrodeposit layer occurred. This did not propagate or cause disrup-
tion of the actual bond line, and probably accounted for the increased emissions. The nonbond panels
generally gave intermediate emission counts. The explanation for this is that all nonbonds were terminated
purposely by full strength bonds, and thus had a smaller area of strong material being strained — thereby
emitting less noise — than the entire weak bond area. The emission count only came from the increased
yield local to defect edges and occasionally where higher counts existed from the propagation at these
points.

Although not directly comparable to these results, the tests on brazed and diffusion bonded
panels support these theories. For example, where braze samples failed at the joint line and crack propa-
gation occurred high emission counts were recorded. When gross plate distortion occurred without signifi-
cant bond line failure only low counts were recorded. This indicates the test was discriminant to the de-
sired bond line information. In the case of diffusion bonded Panel 40, which was a full bond containing
a “reference defect”, the emission count was approximately double that of the full bonds without such a
defect. This was due to the small areas of unbond within the defect pattern yielding or propagating. The
same trend was also shown in Panel 31 which was the brazed full bond with “reference defect”.

The results on the weak electroform bonds indicate that, with further development, this test
could be capable of classifying bond integrity. Careful study of all results shows several areas which re-
quire further evaluation. These include:

(a) Determination of maximum pressure that will still maintain the test truly nondestructive.

(b) Development of capability to differentiate between weak bonds and propagating non-
bonds.

Another feature of the acoustic emission results which is worthy of continued study and
evaluation is the characteristic shape of curves obtained by graph of total count against pressure. Of the
thirteen tests run on electroformed panels, only one (Panel 15) does not categorize. The rate of emitted
noise shown in these graphs varies distinctly between full strength bonds and weak or nonbond samples.
The weak bonds themselves show an increased rate compared to nonbond.

Electroform Nondestructive Evaluation (Correlation with Attained Pattern)

1 —

' —

▼—

’ —

*—

* —

' —

« —

T—

' —

* —

T—

' —

* —

r—

* —

5 f

oo"

E E

LU 3
O O
X O

1 00

r—

r->

cm"

X "O
O O
O O

Q il O O O x u_

’■M

4->

*4_I

_CD

V- *

k—

’*5

V-

h 2
n .2

JZ •-
7 D

> x

> 03

X <£

CD §

to o

a> "5

O ’

CO '

r _ a>

« t! 4-

O C Q>
U 10 T 3

t: ~o

5 TS
S g .
JS o g

03 to q

t to 03
O 3 E
O XX

—

_cd

E t
o o
CO u

-0*0-0
O O O
o o o

a a o

-Q

5 ®

_Q £

-a

c a

ill

2 S

E 03 ^
CO $ c

-a

-Q

*o *o
o o
o o
(3 (3

’co

x
c .

■BE

00 C

g O
2 “o

-o

to c

O i- iu IU co

-a

-o

+-<

« o
>n
= c

0 )

■M

CD 3

+~>

"cD

*CD

_CD

03 03

E E

+->

■f-*

"03

C0 00 co CL CD 2 c. oo O m

< < < < oSS <

C 3

H-

-a

. r-»
X —
03 03
C C
c 2
jo Q-

x c
c/> <S
< *£

^ —

0 )

_ ru

? a
c c

C CD

8 oi

D C

CO —

*J 3

_C 0

t/i y:
CD C 3
03
> M-
— CD

CD TJ

| ai
QJ 5

>1
0 ) O

X "co .
F 03 3

o 5 £

ill

^ 3 .

• 4 — >

»—

' —

0 M

'a>

-a

H-

J—

4 ->

03
■ 4— 1

r—

“O

4- 1

4 -»

+-»

■M

4 -*

4 -»

+-»

■M

+->

■ 4 -»

•

0 .

-X

~~T

“

0 )

CT5

'Ct

’£1

T—

t—

r—

■' —

' —

v—

' —

* —

r—

*—

T-

T—

r—

r^’

t—

»—

co'

</)

-X

—

I 4-»

4—

m §

O Q

CT)

4-*

r—

*3 O

(A W

T—

t —

» —

r—

x —

r—

L < '

4— 1

4-4

1—

■4-4

4-4

4-»

I-

_©

4->

° cT

^ o

4-<

4—

X
©
4— 1

4-*

v£

4-4

o .2

CT)

4—*

4~»

CT)

4-4

o *=

O >

Z -c

4-4

4-<

4—

4-'

<4—

fsl

M—

_©

more bond

4-4

■a

X q3

4— 1
CO

-7-5

—

— «

O cn

4-»

o £

C D

■C

4-4

+-»

4—

■o

4-*

c o

4—4

CT)

_©

■a

—

4—

4-*

"©

_©

< •

■*— *
O
©

•

4—

-a

4*4

4-»

4—*

4-4

4->

4—

4-*

4-4

4— 1

4—4

■

•

-3

-X

—

T7T

c •

anel

Mo.

1^*

<3-

Diffusion Bond Nondestructive Evaluation (Correlation with Attained Pattern)

4. Holography

Comparison of holograms with the matching ultrasonic “C” scans shows that excellent corre-
lation was obtained between the two techniques, Figures 47 thru 55. This indicates the ability of the tech-
nique to distinguish between bond and nonbond conditions. More important, is that the holograms on
weak electroformed bonds also show positive results (e.g., Panels 5 and 9). This establishes two of the
four test methods as demonstrating feasibility for determining the weak bond condition. The results of
destructive testing showed that holography was also capable of determining the high stress concentration
areas which were ultimately the failure regions. This type of information used during initial design research
programs for space hardware could provide valuable information to the design engineers.

Study of the results also shows that more understanding of variables affecting holographic
image quality and interpretability is required. Some limited work was conducted in the areas of surface
flatness requirements, material thickness and pressure variations. Figure 56 is a hologram of the 0.060 in.
(1.524 x 10' 3 m.) thick electroformed Panel 11 at 500 psi (3.45 x 10 6 N/m. 2 ), showing good definition.

The same panel after additional buildup to 0.120 in. (3.048 x 10’ 3 m.) is shown along side. 500 psi (3.45
x 10 6 N/m. 2 ) was again applied and the definition is considerably reduced. The fact that pressure can effect
and in fact compensate for thickness is indicated by comparing the holograms obtained on diffusion bonded
Panel 36 (taken at 500 psi (3.45 x 10 6 /m. 2 )) with those on Panels 34 and 43 (taken at 1440 psi or 9.93
x 10 6 N/m. 2 ). Changes in thickness uniformity within the same panel also produced differences, with the
thicker sections being less interpretable than the thinner ones due to less surface movement for a given pres-
sure. Figure 57 is a hologram indicating the zig-zag pattern attributable to the greater thickness on the right
hand side. The adjacent hologram shows the complete absence of this pattern on the same panel after sur-
face grinding flat. The results on test panels also are highly indicative that thickness and pressure variances
influence definition, (compare holograms on Panels 1 1 , 22, 3 1 , and 44). A more detailed study is required
to fully explain all these variations. The fringe patterns also vary to some extent. The hologram on Panel
22 shows the bands going transverse to the panel width when in all other cases the bands tended to follow
the major axis of the lands.

An important factor which requires further investigation is the effect of test sequence on the
results. As a consequence of the established sequence, holography was always performed subsequent to
acoustic emission. This means that some prestressing to considerably higher pressure (at least 2x) than
that used for the holograms had always occurred prior to holography. It is not known whether stress
patterns and detection sensitivity were enhanced by this factor. The results definitely indicate that in some
cases permanent deformation had already occurred, i.e., Panels 17 and 33 on which fringe patterns could
not be obtained, due to gross movement. This means that in at least these cases the test had been destruc-
tive. It is essential to ensure that effective results are obtainable without any suspicion of permanent de-
formation. In view of this, a closer evaluation of the acoustic emission data showed that little or no signi-
ficant data would be lost by performing holography, to a maximum of approximately 1000 psi (6.89 x 10 6
N/m. 2 ), prior to acoustic emission testing. Any future work would be carefully scrutinized in this area
with two major objectives of test succession in mind. These are:

(a) Review test approach with the ideal of conducting holography and acoustic emission
concurrently.

(b) If this is not possible carefully evaluate configuration design and conduct holography
at as low a pressure as possible prior to acoustic emission at proof pressure level.

OVERLAY OF ULTRASONIC RESULT ON OVERLAY OF ULTRASONIC RESULT ON

THE HOLOGRAM FOR PANEL 4E THE HOLOGRAM FOR PANEL 7E

HOLOGRAM FOR PANEL 13E HOLOGRAM FOR PANEL 15E

(C<C

OVERLAY OF ULTRASONIC RESULT ON HOLOGRAM
FOR PANEL 21B

if}

-J

HOLOGRAM FOR PANEL 40D HOLOGRAM FOR PANEL 49D

PANEL 11. EFFECT OF THICKNESS ON HOLOGRAM

5. Spectrum Analysis

The fourth method was considerably more experimental in nature and did not show the ob-
vious application that the other methods did. It did nevertheless show sufficient merit to have justified
its inclusion in the selected methods.

The study was conducted on only two panels selected from each of the three fabrication
methods. These panels were:

No.

Fabrication Method

Type

Electroformed

Full Bond

Electroformed

Weak Bond

Brazed

Full Bond
Nonbond

Diffusion Bonded

Full Bond

Diffusion Bonded

Nonbond

These panels had electroformed backing to prevent bulging during Acoustic
Emission pressurization and thus limited the ultrasonic approaches.

Nonbond is listed here instead of weak since the brazing and diffusion bonding
process did not provide bond line variance other than full and nonbond areas.

The spectrum envelope photographs of each test are contained in the figures listed below.

(a) Pulse-Echo Method Figures 58 thru 81

(b) Thru-Transmission Method Figures 82 thru 88

The majority of the work was conducted using the pulse-echo longitudinal wave method with 10, 15 and
25 MHz frequency transducers. In each case an envelope of the front surface reflection was obtained ini-
tially to show the transducers' characteristics and to set center frequency (Figures 58 thru 61). Subse-
quently, an envelope of each interface (i.e., channel, bond and back surface) was obtained where possible
for comparison to the set center frequency (Figures 62 thru 81). Deviations from center were measured
and designated as “actual” peak frequency. Reviewing the amount and direction of center frequency
shifts reveals the following:

a. Figures 62 and 63 display an increase in frequency when gated to the channel interface
where a good strong reflection is possible. Complete explanation is not evident for this
other than possibly focusing was more ideal at that interface.

b. Most of the photographs depict a decrease in frequency. This was expected due to
attentuation factors at each interface.

c. Figures 76 and 79 display the greatest frequency shifts on brazed Panels 20 and 28,
respectively. This is understandable, since the brazing fabrication method introduced

a thicker dissimilar bond interface material by using a copper sheet which creates higher
impedance mismatch and attenuation.

d. Figures 76 and 79 along with Figures .69 thru 72 also indicate which transducer approached
the ideal frequency for the material and thickness involved. In this case the nominal 15
MHz transducer (actually operating at 17 MHz) offered the greatest frequency shifts.

The thru-transmission method could only be gated to a composite signal representing sound transmission
through all interfaces and therefore did not permit individual interface selection and hence evaluation. The
test was also limited to 10 MHz which was the highest frequency available capable of sufficient sound trans-
mittal for monitoring. In this case the setup frequency, Figure 82, was obtained with water acting as the
only attenuative media. Very little frequency shift was noted, Figures 83 through 88, and no significant
conclusions could be drawn.

The pitch and catch method used the back reflection from a 0.1 195 in. (3.035 x 10" 3 m.) thick
nickel plate as its reference, Figure 89, since it is a contact test method and no sound reflection from the top
surface was recordable. Review of Figures 90 thru 98 provide no conclusive trends. The only trend obvious
is that where bonds exist, there is an increase in center frequency from the reference Figure 89, and where
nonbond is present, there is a decrease. It is also evident that no difference is detectable between full and
weak bonds. The three methods employed represent a diversified approach but should not be considered
as the only possible ultrasonic approaches. Many other ultrasonic applications are possible including the
“delta concept” which may provide the technology break-through in establishing spectrum analysis as a
NDE tool worthy of development within its own right.

The results obtained to date offer no conclusive proof that spectrum analysis is capable of
determining bond integrity.

Set-Up Pulse-Echo/Top Surface lnterface/2 MHz per Div. Typ. F . 16 , Sec 3000 Speed Film (Typical)

*r- r "'v

|||g§§fl

Panel No. 8 (Weak) 25 MHz Transducer Panel No. 3 (Full) 10 MHz Transducer

14 MHz Center Freq. 14 Actual 10 MHz Center Freq. 10 Actual

Pulse-Echo/Channel interface

Panel No. 3 (Full)

1 5 MHz Center Freq Figure 62

16 Actual

Note: Both Increase
in Center Freq.

Panel No. 8 (Weak)

14 MHz Center Freq. Figure 63

15 Actual

y y-.

Panel No. 3 (Full)
9.75 Actual

No. 37 (Full)
8 Actual

Pulse-Echo/Back Surface Interface/ 17 MHz Center Freq

v .4

No. 8 (Weak) No. 49 (Non-Bond)

16 Actual 15 Actual Note: Decrease in Center Freq

Pulse-Echo/Back Surface Interface (Center Frequency Marked)

•*->

(TJ LT3
a. t-

No. 8 (Weak) 15 Actual

14 MHz Center Fre Figure 74

Pulse-Echo/Bond Interface/Panel No. 20 (Full)

17 MHz Center Freq.

14 Actual Figure 76

15 MHz Center Freq.
15 Actual

Figure 77

18 ajn6i-i

lenpvsi

Z HIA) 9 1 B8jvpa|ie-i

08 ajnBi-j |anp V 9l

* -- ZHIAI91 Puog

QL ajnBjj

lenpy Zl

z HI/\ia puog-uo N

8 L a^6i j

Z HIA) 0L

I enpv 6

puoa-uoN.

(puog-uoN) 2Z ‘ON lauBj/oipg-asind

FOLDOUT FRAME

Pitch and Catch/5 MHz Center Freq./0.5 MHz per Div. Typ.

Pitch and Catch/Bond/Land lnterface/5 MHz Center Freq

No. 3 (Full) 5.25 Actual

No. 28 (Non-Bond) 4.5 Actual

■ • j

INFRARED FEASIBILITY STUDY

1. Test Methods

As stated earlier under method selection, three thin electroformed panels (0.020 in. or 5.08
x 10" 4 m. electroform buildup) were subjected to Infrared nondestructive testing to establish if the con-
clusions of the literature search were correct. These tests were performed using the following equipment
and techniques:

Line Scan Test

a. Liquid Nitrogen Cooled Pointing Radiometer, Model PPL 87 A30 1 .

b. Two 1 000 W Bulbs with Focusing Mirror, Model HFW 87A2000.

c. Scan Table (horizontal scan - vertical index), Model TM 87 A 1 8 1 8.

d. Recorder (X-Y).

Area Scan Test

a. Liquid Nitrogen Cooled Area Scanning Radiometer, SPM5 Camera, CR1 Recorder

Control Console.

b. Heat Source

Temp Test Area Scan System,

c. Scanning Table Model TA87A280

d. Recorder.

2. Testing Techniques and Results

a. Line Scan Testing

The inspection side of the sample was painted black prior to testing. This is done to
eliminate light reflection from the heat source.

The sample was then placed on the scanning stand and positioned in front of the infrared
camera and the heat reflecting mirrors. The test setup is shown in Figure 99.

Several tests were performed using different scanning speeds and time delays between
heating the sample and reading the surface temperature with the pointing radiometer. Slight surface
temperature changes were noted only when scanning Panel 7 at 1/4 inch per second, with a monitoring
delay time of 3 seconds. No temperature differences could be distinguished on the other panels submitted.
Unfortunately, the temperature changes noted on Panel 7 were not of sufficient magnitude to provide any
indication of the bond condition in this panel. Line scans of Panel 7 at the maximum obtainable sensitivity
are shown in Figure 100. No indication of bond condition is shown on these scans.

b. Area Scan Testing

Having established what appeared to be the most sensitive test using the pointing radio-
meter, Panel 7 was then tested on the area scanning inspection system using these parameters. Even with
changes m the scanning speed and the delay time, it was not possible to produce any recordings indicating

defects. A typical example of the results obtained using the Infrared area scanning system may be seen
in Figure 101 .

Based on the results obtained, it must be concluded that Infrared inspection does not
appear to be a feasible method of inspecting nickel bond samples of this specific process and material
type for disbonds. Apparently, the heat transfer characteristics and material thickness prevent a sufficient
heat buildup over an unbond to be detected with these techniques.

102

AREA SCAM RECORDING OF NICKEL BOND SAMPLE NO. 7
Figure 101

103

V. DESTRUCTIVE EVALUATION

A. FAILURE TESTING

After nondestructive evaluation, all test panels were subjected to destructive testing. Failure was con-
sidered to occur if : 1 ) the panel bulged to an extent where pressure could not be hydrostatically increased
because of volume displacement due to base metal yielding, 2) visible bulging indicated bonds had fractured
or were peeling, and 3) pressurization had reached the limit of the equipment (or failure of pressure seals
had occurred).

Where possible the portable hydraulic handpump (Enerpak, 10,000 psig or 68.95 x 10 6 N/m 2
rated) was used to pressurize to destruction. Acoustic emission data was obtained on selected panels
during destructive testing, but only to the maximum pressure of 10,000 psig (68.95 x 10 6 N/m 2 ). Some
panels did not fail at pressures to 10,000 psig (68.95 x 10 6 N/m 2 ). These were repressurized at higher
pressures in separate test facility. Pressures were recorded at which visible failure began to occur and are
shown for each panel in the Appendix. The area of bulge in failure was traced for later calculation of
bond strengths. This area is outlined on panel diagrams in the Appendix.

B. METALLOGRAPHIC ANALYSIS

Sections were made from each failed panel and reference panels (not subject to intentional destruc-
tive pressurization) for verification of bond integrity and comparison with results derived from nondestruc-
tive evaluation. Where possible, sections were made which illustrated the failure region (bond disturbance)
and the unfailed region (each bond type which did not fail). In the Appendix, the photomicrographs are
shown for each test panel, as well as the location by land number for each section. The direction of viewing
is shown on the panel drawings to orient the bondline micrograph to the planned bond pattern. All photo-
micrographs were made at 100X magnification.

Judgment of bond type as weak, nonbond, or full bond was based on the disturbance, or lack of
disturbance, of the Nickel 200 channel lands since this is the weaker of the fabrication materials. In the
case of braze bond failures, disturbance of the braze alloy or Nickel 200 materials were considered in deter-
mining bond integrity. It was noted that weak and full bonding occurred in planned nonbond regions on
several of the brazed and diffusion bonded panels.

Two of the electroformed panels exposed to thermal treatment for joint diffusion evaluation were
sectioned and metallographically examined, Figure 102. The photomicrographs of the full bond joint of
Panel No. 47 indicate segregation occurred at the bond line in the layer of electroform which bridged the
channel. The bottom picture shows failure in an adjacent land bond which indicates this segregation from
high temperature exposure is detrimental to bond strength.

Such segregation was not evident at the joint line on Weak Bond Panel No. 16. This was a planned
weak bond which experienced diffusion after a treatment at 1800°F (1255°K). However, the bottom
picture shows separation at a plating restart surface on the same panel after exposure to 1800°F (1255°K).

Some difficulty was experienced with the silver conductivizing process at the time this panel was
produced. Re-conductivizing was required, and it is possible that surface contaminants were not com-
pletely removed by the cleaning process between silvering operations.

104

Full Bond Panel No. 47

Weak Bond Panel No. 1 6

Diffusion at bond lines appear to be impeded
by segregation following the thermal exposure
at 1 800°F ( 1 255°K). (Magnification 100X,
Reduced.)

Failure of bond line in “full bond” panel after
exposure to 180Q°F (1255°K) and subsequent
pressurization - below 1000 psig (6.89 x
10 6 N/m. 2 ). (Magnification 100X, Reduced).

Some diffusion occurred in this planned weak
bond area. No segregation was evident. (Mag-
nification 100X, Reduced).

Lamination occurring at electroform restart
after exposure to 1800°F (1255°K). This
condition existed below each blister ob-
served after thermal treatment. (Magni-
fication 10QX, Reduced.)

Figure 102. Photogmicrographs of Electro formed-Thermally Diffused Test Panels

105

In the processing of both panels in Figure 102, it was necessary to surface grind for flatness during
the electroform build-up. After grinding, the plates were acid etched to remove surface contaminants from
the grinding process. The laminations were associated with blisters which indicates regional residues or con-
taminants, such as imbedded grinding material, may not have been completely removed in the acid treatment.
This condition was not apparent on the other panels in the program, and bond strengths were obtained as
planned.

It is evident that surface grinding may influence bond integrity of plating restarts after exposure to
elevated temperature. Investigation of procedures to assure complete removal of these residues was beyond
the scope of this program, but should be considered in future work where electroform diffusion is involved.

C. BOND STRENGTH CALCULATIONS

Bond strengths were determined by mapping the failure bulge of each panel at full scale and graphic-
ally integrating the effective area of pressure reaction on the channels contained in the bulge. The area of
the lands within the bulge was calculated and applied to the formula:

Bond Strength =

Area of Channels x Failure Pressure
Area of Lands

Results of these calculations are shown in Table XVII. Acoustic emission counts to failure (or to a
maximum of 10,000 psig or 68.95 x 10 6 N/m. 2 pressurization) are also shown. An examination of this
data indicates the degree of success achieved in obtaining the planned bond strengths.

The electroformed weak bonds were generally stronger than planned. The high strength of the bond
on Panel No. 14 was probably due to the small percent of nonbond land area and the relatively small size of
flaws planned. Weak bonds produced the higher emission count during failure testing and full bonds the least
count.

Brazed bond strengths were as expected. The values obtained in general reflect the percent of full
bond area planned to obtain the three bond integrities. No correlation of acoustic emission data was noted.
Diffusion bond strengths were as expected, except for Panel Nos. 41 and 42 which contained bond or lack of
bond not planned. Acoustic emission data is not conclusive regarding bond strength during failure pressuri-
zation.

D. CORRELATION OF RESULTS

Results are correlated in the Appendix. The planned flaw patterns and fabricated patterns achieved
were satisfactory with but few exceptions. These exceptions were primarily on weak and nonbond diffusion
bond panels where more bonding was obtained than planned. Calculated bond strengths were relatively con-
sistent with planned bond type with the general exception of diffusion weak and nonbonds which differed
by non-bond area only. The correlation of nondestructive evaluation results and metallographic sections
to verify actual bonds achieved was good.

For further information on the correlation of results, reference to Section IV, Subsection C, is
suggested.

106

TABLE XVII

BOND STRENGTHS AND ACOUSTIC EMISSION DATA

Failure Pressure

Calculated Bond Strength

Acoustic Emission

Bonding

Process

Panel

No.

Bond

Type

Pressure Range

psi

N/m. 2 (x 10 6 )

psi

N/m. 2 (x 10 6 )

Counts

psi

N/m. 2 (x 10 6 )

Electroform

■

14,000 +

96.53+

32,800 Nom.

226.16 Nom.

6,500

3500 to 9500

24.13 to 65.50

IMI

14,900

102.74

35,910

247.60

Full

14,000+

96.53+

32,800 Nom.

226.16 Nom.

900

3500 to '9500

24.13 to 65.50

Full

14,000+

96.53+

32,800 Nom.

226.16 Nom.

- .

Weak

15,000

103.42

29,520

203.54

• -

Weak

12,000

82.74

29,800

205.47

32,100

3500 to 9500

24.13 to 65.50

Weak

8,600

59.30

19,140

131.97

Weak

11,000

75.84

25,660

176.93

30,900

3500 to 9500

24.13 to 65.50

Nonbond

8,500

58.61

15,150

104.46

10,600

3500 to 8500

24.13 to 58.61

Nonbond

4,000

27.58

7,890

54.40

3500 to 4000

24.1 3 to 27.58

Nonbond

10,200

70.33

22,950

158.24

3500 to 9500

24.13 to 65.50

Nonbond

5,500

37.92

11,575

79.81

11,700

3500 to 5500

24.13 to 37.92

Braze

Full

9,200

63.43

23,270

160.45

219,100*

3500 to 9200

24.13 to 63.43

Full

6,000

41.37

16,910

116.59

1,200

3500 to 6000

24.13 to 41.37

Full

7,500

51.71

15,765

108.70

—

Full

48.26

1 5,000

103.42

3,400

3500 to 7000

24.13 to 48.26

Full

8,600

59.30

21,750

149.97

18,600

3500 to 8600

24.13 to 59.30

Weak

15.86

4,820

33.23

1,800

0 to 2300

0 to 15.85

Weak

14.48

5,465

37.68

1 ,850

0 to 2100

0 to 14.48

Weak

24.13

8,310

57.30

250

0 to 3500

0 to 24.13

Weak

17.24

6,030

41.58

1,800

0 to 2500

0 to 17.24

Nonbond

1,900

13.10

4,155

28.65

3,600

0 to 1900

0 to 13.10

Nonbond

(Ref)

1,000

6.89

2,275

15.69

0 to 1000

0 to 6.89

Nonbond

1,200

8.27

2,690

18.55

99,700

0 to 1200

0 to 8.27

Nonbond

500

3.45

1,090

7.52

4,390

0 to 500

0 to 3.45

Nonbond

1,000

6.89

2,296

15.83

100

0 to 1000

0 to 6.89

Diffusion

Full

1 3,200

91.01

27,825

191.85

100

3500 to 9500

24.13 to 65.50

Full

12,300

84.81

29,350

202.37

500

3500 to 9500

24.1 3 to 65.50

Full

15,400

106.18

37,100

255.80

800

3500 to 9500

24.13 to 65.50

Full

14,000

96.53

32,660

225.19

400

3500 to 9500

24.13 to 65.50

Weak

8,800

60.68

20,770

143.21

—

Weak

7,800

53.78

22,090

152.31

Weak

7,000

48.26

16,575

114.28

Weak

8,000

55.16

16,000

110.32

—

Nonbond

12,000

82.74

26,300

181.34

Nonbond

400

2.76

875

6.03

4,700

0 to 400

0 to 2.76

Nonbond

8,750

60.33

18,290

126.11

—

Nonbond

5,500

37.92

1 2,830

88.46

8,000

3500 to 5000

24.13 to 37.92

*PaneJ 20 leaked at the pressure fitting under high pressure; this possibly contributed to the high counts recorded.

107

VI. CONCLUSIONS AND RECOMMENDATIONS

The overall objective of the program was to establish the feasibility of nondestructive evaluation
methods to detect bond integrity in regeneratively cooled chamber walls. This was to be achieved using
simulated wall sections containing bonds produced by brazing, diffusion bonding, and electroforming.

Three of the four selected nondestructive evaluation methods demonstrated feasibility. These were
holography, acoustic emission, and ultrasonics. Generally, each nondestructive method was useful in provid-
ing specific information essential to complete evaluation of a bonded metallic structure.

Ultrasonics was beneficial as an initial screening approach to detect nonbonds which might prohibit
use of nondestructive methods requiring pressurization. It is recommended that ultrasonics be utilized as
an initial screening technique for the detection of bond from nonbond in all fabrication processes where
practical. Ultrasonics was unable to detect strength differences in electroformed bonds.

Acoustic emission and holography require stressing of the test piece to produce a detectable change
in the bondline. It is essential that this stress (pressure) be maintained below proof pressure for the hardware
design to assure that the test remains nondestructive.

Based on results from the flat test panels, both holography and acoustic emission demonstrated a
potential for detecting the weak bond condition as distinct from full or nonbond. As anticipated, holography
was capable of locating the defective regions, whereas acoustic emission indicated overall condition only.
Acoustic emission was unable to differentiate between emission counts from weak bonds and propagating
nonbonds as explained below:

The planned weak bond and nonbond panels contained multiple defect areas of
various sizes. This complexity prevented the establishment of a fundamental
data base for interpreting the full significance of emission counts for individual
panel conditions.

It is suggested that an approach using a simplified panel design and defect pattern should provide the data
necessary to better define the relative capabilities of holography and acoustic emission.

It is also recognized that emission source locators can be added to further define test panel flaw
conditions. Displaying the test data in this manner will result in obtaining a more definitive inter-relationship
of acoustic emission with the holographic and ultrasonic data.

A factor which may have influenced the holographic results was test sequence. Acoustic emission
was always conducted prior to holography to minimize anticipated data loss (Kaiser-effect). The higher
pressures (3500 psi) used in the acoustic emission test may have been sufficient to cause plastic deformation
in some of the panels containing large non-bond areas. This permanent deformation could have enhanced
the sensitivity of the holographic non-destructive test. This pre-stress may have established a new base-
line condition for the holographic test. This should be considered, along with the other recommendations,
in establishing the optimum test sequence for continued study.

The fourth test method investigated, spectrum analysis, did not provide the anticipated bondline
characterization. Of the three bonding processes evaluated, only brazing showed any positive application.

108

It is significant that this bonding method was the only one to provide a dissimilar bondline material of finite
thickness (relative to inspection techniques). Three ultrasonic approaches were evaluated. Other techniques
are available which may provide positive results. Any future work would require an extensive development
program on equipment and technique prior to application to thrust chambers.

Infrared, a method initially considered as a primary candidate, was briefly evaluated. Materials in
the configuration under evaluation did not respond to this test.

Of the three bonding techniques, electroforming provided the more useful bonds for evaluation in
this program. This was due to the fact that weak bonds of varying strength and size could be more easily
achieved.

VII. REFERENCES

1 . Guidance to Nondestructive Testing Techniques, U.S. Army Material Command, April 1970.

2. Martin, George, “Exploratory Development of Nondestructive Testing Techniques,” North American
Rockwell, September 1968.

3. Martin, George, Moore, John F., and Coate, F.M., “Exploratory Development of Nondestructive Testing
Techniques for Diffusion Bonded Surfaces,” North American Rockwell, AFML-TR-70-188, September
1970.

4. Oaks, A.E., “New and Refined Nondestructive Techniques for Graphite Billets and Shapes,” General
Electric Company, AFML-TR-70-212, February 1970.

5. Kraska, I.R. and Kamm, H.W., “Evaluation of Sonic Methods for Inspecting Adhesive Bonded Honey-
comb Structures,” General American Transportation Corporation, AFML-TR-69-283, August 1970.

6. Sessler, T.G., “The Effect of Stress Fields on Ultrasonics Energy Reflected from Discontinuities in
Solids,” Syracuse University Research Corporation, Syracuse, New York, June 1971.

7. Whaley, H.L. and Adler, Laszlo, “Flaw Characterization by Ultrasonic Frequency Analysis,” Oakridge
National Lab Metals and Ceramics Division, Materials Evaluation, August 1971 .

8. Frederick, J.R., “Dislocation Mechanisms as Sources of Acoustic Emission,” University of Michigan,
Ann Arbor, Michigan, Army Materials and Mechanics Research Center, Watertown, Mass., June 1971.

9. Hagemaier, D.J., McFaul, H.J., and Moon, D., “Nondestructive Testing of Graphite Fiber Composite
Structures,” Douglas Aircraft Co., McDonnel Douglas Corporation Materials Evaluation, June 1971.

10. Dunegan, H.L., Harris, D.O., and Tetleman, A.S., “Detection of Fatigue Crack Growth by Acoustic
Emission Technique,” Lawrence Radiation Laboratory University of California, Livermore, California,
Material Evaluation, October 1970.

1 1. Chambers, R.H., “Time and Frequency Domain Analysis of Acoustic Emission During Fatigue Failure,”
University of Arizona, Tucson, Arizona, Army Materials and Mechanics Research Center, Watertown,
Massachusetts, June 1971.

12. Hartbower, C.E., “Development of a Nondestructive Testing Technique to Determine Flaw Criticality
Based on Stress Wave Emission,” Aerojet, Sacramento, California, Army Materials and Mechanics
Research Center, Watertown, Massachusetts, June 1971.

13. Romrell, D.M. and Bunnell, L.R., “Monitoring of Crack Growth in Ceramic by Acoustic Emission,”
Materials Evaluation, December 1970.

14. Acoustic Emission Research/Dev., Volume 22, Number 5, Pages 20 - 24, May 1971 .

110

15. Vest, C.M. and Sweeney, D.W., “Applications of Holographic Interferometry to Nondestructive Testing,”
Radar and Optics Laboratory and Department of Mechanical Engineering; The University of Michigan,
Army Materials and Mechanics Research Center, June 1971.

16. Kersch, L.A., “Advanced Concepts of Holographic Nondestructive Testing,” GCO Incorporated, Ann
Arbor, Michigan Materials Evaluation, June 1971.

17. Harris, W.J. and Clauss, Francis J., “Inspecting Bonded Structures by Laser Holography,” Metal Progress,
August 1971.

18. Waters, J.P., “The Uses of Holography for NDT,” United Aircraft Research Lab, East Hartford,
Connecticut, June 1971.

19. Sneeringer, J.W., Hacke, K.P., and Roehrs, R.J., “Practical Problems Related to the Thermal Infrared
Nondestructive Testing of a Bonded Structure,” McDonnell Aircraft Company, St. Louis, Missouri,
Materials Evaluation, April 1971 /Volume XXIV, No. 4.

20. Inteieri, A.J., “An IR NDT Bond Inspection Systeqi for Rotor Blade Honeycomb Box Assemblies,”

The Boeing Company, Vertol Div., Philadelphia, Penn., Materials Evaluation, June 1970/Vol. XXVIII,

No. -7.

21. Martin, George and Moore, J.F., “Research and Development of Nondestructive Testing Techniques
for Composites,” North American Aviation Inc., AFML-TR-68-202, June 1968.

22. Lavoie, F.J., “Acoustic Holography, A New Dimension in Seeing with Sound,” Machine Design,
September 1971.

23. Schneiderman, R., Nuclear Research Corporation, Brochure; Future and Termed Exciting for Nuclear
NDT and Measuring Control Device, Metal Working News, October 1967.

24. Special Report/A Guide to Nondestructive Testing, Materials Engineering, June 1969.

25. Materials Evaluation April

1970

- Vol. XXVIII,

No. 4

May

1970

- Vol. XXVIII,

No. 5

June

1970

- Vol. XXVIII,

No. 6

July

1971

- Vol. XXIX^ '

No. 7

26. Karplus, H.B., Semmler, R.A., and Arneson, B.E., “Evaluation of Nondestructive Testing Techniques of
Diffusion Coatings,” AFML-TR-67-358, May 1968.

27. Way, F.C., “Determination of the Performance of Holographic Nondestructive Testing Systems,”

Pratt & Whitney Aircraft, Florida Research and Development Center, West Palm Beach, Florida
Materials Evaluation, July 1971.

Ill

DISTRIBUTION LIST FOR FINAL REPORT

NAS3-14376

Bell CR 120980

REPORT

COPIES

R D RECIPIENT

National Aeronautics & Space Administration
Lewis Research Center
2 1 000 Brookpark Road
Cleveland, Ohio 44135

Attn: Contracting Officer, MS 500-313 /'

E.A. Bourke, MS 500-203 /

Technical Report Control Office, MS 5 - 5 /
Technology Utilization Office, MS 3-16'’

AFSC Liaison Office, 501-3 ;

Library /

Office of Reliability & Quality Assurance,

MS 500-1 11

J.W. Gregory Chief, MS 500-203

R.A. Duscha Project Manager, MS 500-209

Alex Vary, MS 106-1 r

R.F. Lark, MS 49-1

W.E. Russell, MS 14-1

Director, Shuttle Technology Office, RS
Office of Aeronautics & Space Technology
NASA Headquarters \

Washington, D...C. 20546 \

* \

Director Space Propulsion and Power, RP

Office of Aeronautics & Space Technology
NASA Headquarters
Washington, D. C. 20546

Director, Launch Vehicles & Propulsion, SV\
Office of Space Science &

NASA Headquarters
Washington, D. C. 20546

Director, Materials & Structures Division, RW
Office of Aeronautics & Space Technology
NASA Headquarters
Washington, D. C. 20546

Director, Advanced Manned Missions, MT
Office of Manned Space Flight
NASA Headquarters
Washington, D. C. 20546

DESIGNEE

REPORT

COPIES

p fT RECEIPIENT

20 National Technical Information Service

Springfield, Virginia 22151

1 National Aeronautics & Space Administration

Ames Research Center
Nloffett Field, California 94035
Attn. Library

1 National Aeronautics & Space Administration

Flight Research Center
P.O. Box 273

Edwards, California 93523
Attn : Library

1 Director, Technology Utilization Division

Office of Technology Utilization
NASA Headquarters /

Washington, D. C. 20546

1 Office of the Director of Defense
Research & Engineering
Washington, D. C. 20301

Attn: Office of Asst. Dir. (Chem Technology)

2 NASA Scientific and Technical Information Facility
P.O. Box 33

College Park, Maryland 20740
Attn: NASA Representative

1 National Aeronautics & Space Administration

Goddard Space Flight Center
Greenbelt, Maryland 20771
Attn : Library

1 National Aeronautics & Space Administration

John F. Kennedy Space Center
Cocoa Beach, Florida 32931
Attn : Library

National Aeronautics & Space Administration
Langley Research Center
Langley Station
Hampton, Virginia 23365
Attn : Library

DESIGNEE

Hans M. Mark
Mission Analysis
Division

Merland L
Moseson,
Code 620

Dr. Kurt H.
Debus

E. Cortwright
Director

Lewis Thurston, Jr.

REPORT

COPIES

R D RECIPIENT

National Aeronautics & Space Administration
Manned Spacecraft Center
Houston, Texas 77001
Attn: Library

National Aeronautics & Space Administration
George C. Marshall Space Flight Center
Huntsville, Alabama 35912
Attn: Library

Jet Propulsion Laboratory
4800 Oak Grove Drive
Pasadena, California 91103
Attn: Library

Defense Documentation Center
Cameron Station
Building 5 /

5010 Duke Street J

Alexandria, Virginia 22314
Attn: TI SI A /

,•/

RTD (RTNP)

Bolling Air Force Base
Washington, D. G. 20332,

Arnold Engineering Development Center
Air Force Systems Command
Tullahoma, Tennessee 37389
Attn: Library \

Advanced Research Projects Agency
Washington, D.C. 20525
Attn: Library

' Aeronautical Systems Division
Air Force Systems Command
Wright-Patterson Air Force Base,
Dayton, Ohio
Attn: Library

Air Force Missile Test Center
Patrick Air Force Base, Florida
Attn: Library

DESIGNEE

J.G. Thiobodaux, Jr.
/ Chief, Propulsion &

' Power Division

Hans G. Paul
James Thomas

Henry Burlage, Jr.
Duane Dipprey

Dr. H.K. Doetsch

D.L. Schmidt
Code ARSCNC-2

L.J. Ullian

REPORT

COPIES

RECIPIENT

DESIGNEE

R D

Capt. S.W. Bowen
SCLT

Donald Penn
Robert Wiswell

1 Air Force Rocket Propulsion Laboratory (RPM)

Edwards, California 93523
Attn : Library

1 Air Force FTC (FTAT-2) Donald Ross

Edwards Air Force Base, California 93523
Attn: Library

1 Air Force Office of Scientific Research SREP, Dr. J.F. Masi

Washington, D. C. 20333
Attn : Library

1 Space & Missile Systems Organization

Air Force Unit Post Office
Los Angeles, California 90045
Attn: Technical Data Center

Air Force Systems Command
Andrews Air Force Base
Washington, D. C. 20332
Attn: Library

Air Force Rocket Propulsion Laboratory (RPR)
Edwards, California 93523
Attn : Library

Office of Research Analyses (QAR)

Holloman Air Force Base, New Mexico 88330
Attn: Library
RRRD

U. S. Air Force
Washington, D. C.

Attn : Library

Commanding Officer

U. S. Army Research Office (Durham)

Box CM, Duke Station
Durham, North Carolina 27706
Attn : Library

U. S. Army Missile Command
Redstone Scientific Information Center
Redstone Arsenal, Alabama 35808
Attn: Document Section

Col. C.K.
Stambaugh,
Code AFRST

Dr. W. Wharton

REPORT

COPIES

DESIGNEE

R D RECIPIENT

1 Bureau of Naval Weapons

Department of the Navy
Washington, D. C.

Attn: Library

J. Kay,

Code RTMS-41

Commander

U. S. Naval Missile Center
Point Mugu, California 93041
Attn: Technical Library

Commander

U V S. Naval Weapons Center
China Lake, California 93557
Attn: Library

Commanding Officer
Naval Research Branch Office
1030 E. Green, Street
Pasadena, California 91101
Attn: Library

Director (Code 61 80)

U. S. Naval Research Laboratory
Washington, D. C. 2039X)

Attn: Library

Picatinny Arsenal - \

Dover, New Jersey 07801
Attn: Library \

Air Force Aero Propulsion Laboratory
Research & Technology Division \

Air Force Systems Command \

United States Air Force \

Wright-Patterson AFB, Ohio 45433
Attn: APRP (Library)

i? Electronics Division
Aerojet-General Corporation
P.O. Box 296
Azusa, California 91703
Attn: Library

H.W. Carhart
J.M. Krafft

I. Forsten

R. Quigley

REPORT

COPIES

R D RECIPIENT DESIGNEE

1 Space Division

Aerojet-General Corporation
^200 East Flair Drive
EhMonte, California 91734
Attn:. Library

1 Aerojet Ordance and Manufacturing

Aerojet-General Corporation
11711 South Woodruff Avenue
Fullerton, California 90241
Attn: Library v

1 1 Aerojet Liquid Rocket Company

P.O.Box 15847
Sacramento, California 95813
Attn: Technical Library 2482-201 5 A

1 Aeronutronic Division of Philco Ford Corp.

Ford Road

Newport Beach, California 92663
Attn: Technical Information Department

1 Aerospace Corporation

2400 E. El Segundo Blvd.

Los Angeles, California 90045
Attn: Library-Documents

1 Arthur D. Little; Inc.

20 Acorn Park'

Cambridge, Massachusetts 02140
Attn : Library

1 Astropower Laboratory

McDonnell-Douglas Aircraft Company
2121 Paularino

Newport Beach, California 92163
Attn : Library

1 DCIC, Battelle Memorial Institute J.F. Lynch

Columbus Lab. Room 11-9012
505 King Avenue
Columbus, Ohio 43201
Attn: Library

R.J. LaBotz

Dr. L.H. Linder

J.G. Wilder

A.C. Tobey

REPORT

COPIES

RECIPIENT

DESIGNEE

ARO, Incorporated

Arnold Engineering Development Center
Arnold AF Station, Tennessee 37389
\ Attn: Library

Susquehanna Corporation
Atlantic Research Division
\Shirley Highway & Edsall Road
Alexandria, Virginia 22314
Attn: Library

Beech Aircraft Corporation
Boulder Facility
Box 631

Boulder, Colorado
Attn: Library \

Bell Aerospace Co.\ /

Box 1 \.

Buffalo, New York 1 4240*

Attn: Library

Instruments & Life Support Division
Bendix Corporation-

P.O. Box 4508 /

Davenport, Iowa ,;r 5280S
Attn: Library,/

Bellcomm /

955 L’Enfant Plaza, S. W.
Washington, D. C.

Attn: ^ library

Boeing Company
Space Division
P/O. Box 868

/Seattle, Washington 98124
/ Attn: Library

Boeing Company
1625 K Street. N. W.
Washington, D. C. 20006

Douglas Pope

John M. Brueger

H.S. London

J.D. Alexander
C.F. Tiffany

Boeing Company
P.O. Box 1680
Huntsville, Alabama 35801

Ted Snow

REPORT

COPIES

R jD RECIPIENT

1 Chemical Propulsion Information Agency

Applied Physics Laboratory
8621 Georgia Avenue
Silver Spring, Maryland 209 1 0

1 Chrysler Corporation

Missile Division
P.O. Box 2628
Detroit, Michigan
Attn: Library

1 Chrysler Corporation

Space Division
P.O. Box 29200
New Orleans, Louisiana 70129
Attn: Librarian

1 Grumman Aircraft Engineering Corporation

Bethpage, Long Island, New York
Attn : Library

1 Hercules Powder Company .

Allegheny Ballistics Laboratory
P.O. Box 210

Cumberland , Maryland 21501
Attn : Library

1 Honeywell Inc.

Aerospace Division
2600 Ridgeway Road
Minneapolis, Minnesota
Attn: Library

1 IIT Research Institute

Technology Center
Chicago, Illinois 60616
Attn : Library

1 Kidde Aer-Space Division

Walter Kidde & Company, Inc.

567 Main Street
Bellville, New Jersey

1 Ling-Temco-Vought Corporation

P.O. Box 5907
Dallas, Texas 75222
Attn : Library

DESIGNEE
Tom Reedy

John Gates

Joseph Gavin

C.K. Hersh
R.J. Hanville

REPORT
COPIES
R_ D

RECIPIENT

Lockheed Missiles and Space Company
P.O. Box 504

Sunnyvale, California 94087
Attn: Library

Lockheed Propulsion Company
P.O. Box 111

Redlands, California 92374 .

Attn: Library, Thackwell

Marquardt Corporation
16555 Saticoy Street
' Box 201 3 - South Annex
Van Nuys, California 91409

Martin-Marietta Corporation (Baltimore Division)
Baltimore, Maryland 21203
Attn: Library

Denver Division /

Martin-Marietta Corporation
P.O. Box 179 \/

Denver, Colorado 80201
Attn: Library y-

Orlando Division

Martin-M arietta Corporation \

Box 5827 \

Orlando, jFlorida \

Attn: Library \

Western Division \

McDonnell Douglas Astronautics
5301 Bolsa Avenue \

/Huntington Beach, California 92647 \

y Attn: Library

McDonnell Douglas Aircraft Corporation
P.O. Box 516

Lambert Field, Missouri 63166
Attn: Library

DESIGNEE

H.L. Thackwell

L.R. Bell, Jr.

Dr. Morganthaler
F.R. Schwartzberg

J. Fern

R.W. Hallet
G.W. Burge
P. Klevatt

R.A. Herzmark

REPORT

COPIES

RECIPIENT

DESIGNEE

R D

1 1

Rocketdyne Division
North American Rockwell Inc.

6633 Canoga Avenue

Canoga Park, California 91304

Attn: Library, Department 596-306

Space & Information Systems Division
North American Rockwell
12214 Lakewood Blvd.

Downey, California
Attn : Library

Northrop Space Laboratories
3401 West Broadway
Hawthorne, California
Attn : Library

Purdue University
Lafayette, Indiana 47907
Attn: Library (Technical)

Radio Corporation of America
Astro-Electronics Products
Princeton, New Jersey
Attn : Library

Rocket Research Corporation
Willow Road At 1 1 6th Street
Redmond, Washington 98052
Attn : Library

Stanford Research Institute
333 Ravenswood Avenue
Menlo Park, California 94025
Attn : Library

Thiokol Chemical Corporation
Redstone Division
Huntsville, Alabama
Attn : Library

TRW Systems Inc.

1 Space Park

Redondo Beach, California 90278
Attn: Tech. Lib. Doc. Acquisitions

Donald Fulton

Dr. William Howard

Dr. Bruce Reese

F. McCullough, Jr.

Dr. Gerald Marksman

John Goodloe

Curtis Watts

REPORT

COPIES

RECIPIENT

TRW

TAPCO Division
23555 Euclid Avenue
Cleveland, Ohio 441117

\ United Aircraft Corporation
\ Corporation Library
\400 Main Street

East Hartford, Connecticut 06108
Attn: Library

United Aircraft Corporation
Pratt & Whitney Division /

Florida Research & Development Center
P.O.Box 269 1\

West Palm BeaclvEdorida 33402

Attn: Library

United Aircraft Corporation
United Technology Centers.
P.O. Box 358 '\ t

Sunnyvale California 94038
Attn: Library/

Vickers Incorporated
Box 302./

Troy, Michigan

Vought Astronautics
Box 5907
(Dallas, Texas
Attn: Library

DESIGNEE

Frank Stattler

Dr. David Rix
Erie Martin
Frank Owen
Wm. E. Taylor

Dr. Schmitke

Dr. David Altman

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Books

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "NASA Technical Reports Server (NTRS) 19720024852: Nondestructive test of regenerative chambers"

See other formats