Evaluation of Bonds in Armor Plate and Other MaterialsUsing Infrared Nondestructive Testing Techniques

Paul E. J. Vogel

Reported here are the results of using an ir technique in evaluating the bond in hot-roll bonded armor

plate, a coating bond, and the bonds in the three-layer construction of a small missile motor. When a

specimen is exposed to heat, the heat flows into it at a rate that can be determined. If the flow is inter-

rupted by a void, the inclusion of a material of different conductivity, or any similar thermal barrier,

a difference in surface temperature will result that can be defined with an ir radiometer or thermograph.Lack of bond, while not presenting a gross void, is shown here to be a sufficient impediment to be dis-

tinguishable.

IntroductionThe evaluation of many types of bonds has presented

a real problem when checked by the earlier nondestruc-tive testing methods. The problems were presentedby acoustically mismatched materials and couplingdifficulties in ultrasonic testing, poor conductors andbond depth in eddy current work, bond orientationwhen approached by x ray, and similar obstacles toother techniques. In 1963, the Army Materials Re-search Agency, now the Army Materials and MechanicsResearch Center, explored the feasibility of the applica-tion of ir testing methods to bonded materials. As aresult of that study, ir techniques are now routinelyconsidered as an approach to each new materials test-ing problem.

Armor PlateThe bond integrity in dual hardness, hot-rolled

bonded, composite armor is a major factor in the ballis-tic limit of the armor. Variations in the bond can oc-cur during manufacture or as a result of being impactedby a projectile. In either case, the unbond is usuallyconcealed, and although impact may sometimes causesurface cracks to appear, the cracks do not necessarilyreveal a bond separation. The ir approach was in-dicated here because of its speed and resolution, andbecause its noncontact nature permits use during hot-rolling of the plates in production.

Thermal conductivity through a bond is a functionof bond integrity. If thermal energy is introducedfrom one side of a composite plate, delamination orvarying degrees of bonding influences the heat migration

The author is with the Army Materials and Mechanics Re-search Center, Watertown, Massachusetts.

Received 4 April 1968.

and results in a nonuniform temperature pattern ap-pearing on the opposite side. This two-sided tech-nique was used in evaluating the armor plate specimensbecause both sides were readily accessible and thebonded plate was of uniform thickness. Other heat-ing techniques might often be more appropriate, as isdescribed later, depending upon the material, its thick-ness, configuration, and accessibility.

The test setup for the armor specimen is shown in Fig.1. Uniform heat flow and constant temperature gra-dients were maintained in the specimens by balancingthe heat being injected at the back by a controlledintensity heat lamp, with front surface cooling accom-plished by blowers.

If the armor had been mounted on its vehicle so thatonly one surface could be used for testing, one of manyalternative heating systems would be to scan the platewith a hot spot, such as produced by focused, high in-tensity projector lamp. The radiometer is scanned tofollow that spot by a distance corresponding to the timedelay determined by the first layer thickness and ther-mal conductivity. Instrument output then displaysany temperature differences as a function of thermaldiffusion through the bond interface. In this case, ahigher surface temperature would indicate an unbondbecause the heat would be trapped on the surface.Conversely, with rear heating and front inspection, alower temperature would show where a flaw preventedthe thermal flow to the front.

The ir thermograph' is an 20-cm ac radiometricdevice equipped with an opto-mechanical scanningsystem. The sensing element is an immersed thermis-tor bolometer detector for noncontact temperaturemeasurements in a target temperature range from-170'C to +250'C. In its usual operation, the in-strument produces on Polaroid film a visible display ofthe ir emitted by a 10° X 20° target area by employing

September 1968 / Vol. 7, No. 9 / APPLIED OPTICS 1739

Fig. 1. Test setup for armor inspection. Shown are, (1) T4-Ccamera, (2) front surface cooling blower, (3) specimen, (4) rear

surface heater.

Fig. 2. Photograph of armor plate specimen showing cracksradiating from points of impact of bullets. The area shown is

10 cm wide.

Fig. 3. Thermogram of armor plate specimen. Darker areasindicate an unbonded condition.

high resolution raster scanning in a manner similar tothe construction of a television image. The detectorviews a 1 X 1 mrad field and can detect a temperaturedifference as small as 0.10C at 250C. The instrumentcan also be used as a single line scanner or to constantlymonitor one spot, and it can display on an oscilloscopeor recorder.

In this particular application, it was used to produce athermogram, on Polaroid film, and it was set in its 3-min scan option that skips alternate lines of raster.The 6-min scan option would have given 50% moreinformation by filling the alternate lines, but thisminute detail was not required to define the thermal gra-dations.

Because the specimens had the typical mottled ap-pearance of unfinished steel, a glare-free black paint

was sprayed on both surfaces to level the emissivityand absorption. Other coatings that are thermoset-ting,2 water soluble, or stripable are available.

One typical specimen is shown in the conventionalphotograph in Fig. 2. This has been impacted at B,C, and D, and penetrated at A with a fan-shaped portionof the front surface detached to the left of A. Cracksradiated as outlined in white, and except for the piecementioned above, no delaminations were visible orevidenced by any bulging. The thermogram of thisspecimen, shown in Fig. 3, was made with a rear surfacetemperature of 450C, and a grey scale setting represent-ing, black to white, a temperature differential of 6C.

A comparison of Fig. 2 and 3 show definite thermalpatterns conforming to the surface cracks. The blackarea is the cold extreme and is somewhat blacker thanthe cold end of the grey scale setting. This specimenwas sliced along the line drawn across Fig. 3 in order todetermine microscopically the significance of the ther-mogram gradations.

The exposed edge of the lower section is related tothe thermogram in Fig. 4. The black one-fourth ofthe edge corresponds to the upward running crack fromthe bullet hole, because the material beyond that crackseparated into two plates after cutting. The rate ofheat flow is displayed as being inversely proportional tothe crack width, and the temperature range from the

-d Fig. 4. Crack and thermogram related. The heavy line on thethermogram is where the specimen was cutt to reveal the crack.

Fig. 5. 500X photomicrograph of the closed crack at left ofunbonded area.

1740 APPLIED OPTICS / Vol. 7, No. 9 / September 1968

Fig. 6. Rubber coating on steel. Dark areas (arrows) show lack

of bond.

Fig. 7. Rubber coating after cold-patchingArrows indicate a line of lack of adhesion of

original coating.

unbonded areas.the patch to the

widest crack into the good bond area is about 30 C.The area over the last 1.25 cm at the left of the crack isdirectly above impact point B.

A 50OX photomicrograph of this area, Fig. 5, showsthe interlocking irregularities at the interface that are,for all practical purposes, a close physical contact.Assuming no crack enlargement during cutting andetching, the thermogram is displaying a contactingunbond ranging from 0.125 mm at the dogleg to 0.025mm midway to the end.

The possibility that the display resulted from the per-turbance of the nearby void was considered, but as thegrey persisted through a number of heating and coolingmodes, no doubt remained that the ir technique wasseeing a contacting unbond. Differences in thicknessof the diffusion layer that ideally is on the order of0.125 mm to 0.200 mm at the bond of the two plates ofalmost identical composition, have not yet been suc-cessfully identified by ir or other NDT techniques thatare fast enough for continuous production line testing orfield inspection.

Coating BondFor an example of the bonding of a coating, the

specimen was an irregularly shaped steel plate about1 cm thick, backed by 4 cm of rubber and having therubber wrap around to the front surface decreasing to athickness of about 0.625 mm. The thickness of the

rubber on the back precluded an effective two-sidedtechnique, so the specimen was heated and scawtedfrom the front. Otherwise, the test setup was the sameas on the armor plate.

To demonstrate a different method of creating athermogram, the specimen was heat-soaked at 700C,then cooled on the surface of interest by blowing ambi-ent air. In this approach, any unbonded coating wouldcool more rapidly and would appear darker on thethermogram. Two such areas appear as indicated bythe arrow in Fig. 6. The rubber was stripped fromthese areas and replaced by cold-patching which did notadhere to one edge of the original coating along the lineof a sharp bend in the steel as shown in Fig. 7.

In Fig. 6, there are some abrupt changes in the tone ofthe raster that do not appear in Fig. 7. In Fig. 6, thespecimen was cooled much more rapidly than in Fig. 7so that the target temperature range during the scan-ning period exceeded the picture temperature range thathad been set in the electronics unit. This was donedeliberately to illustrate how the thermogram may beadjusted to span extreme changes in target energylevels by gradually advancing or retarding the elec-tronic temperature offset (ETO) potentiometer withoutchanging the attenuation control which is stepped.This technique of chasing-with-ETO allows smallradiation variations to be closely examined in bandsalong a specimen having a large total variation duringthe period of thermographing.

Missile MotorsThe construction of missile and rocket motors using a

solid propellant presents another variation in bondinginspection. Bond defects may occur at the first inter-face, case-to-liner, or the second, liner-to-propellant(Fig. 8). While ultrasonic testing could show a firstinterface defect, it would have little success at the secondinterface because of the mismatch of materials. Theflaw orientation virtually rules out x-ray, and othermethods are similarly eliminated by reason of flaw depthor the electrical and nonmagnetic properties of the

FIRST INTERFACEBOND DEFECT

LINER

SECOND INTERFACEBOND DEFECT

METAL CASE

PROPELLANT

Fig. 8. Cross section of a cylindrical missile motor case showingschematically the two types of bond defects that can occur.

September 1968 / Vol. 7, No. 9 / APPLIED OPTICS 1741

r EMPLOY

..1I

'TURE

:EE T

.~~~~~

20 SEC.

>TIME

Fig. 9. Relative temperatures over first interface (left) and sec-ond interface defects in a missile motor case. The 20 sec. be-tween markers represents one complete circumferential scan.

materials. Infrared techniques have therefore beenwidely adopted for inspection of bonding of dissimilarmaterials such as this.

In this instance, a specimen with known bond defectswas prepared by the U. S. Army Missile Command.When the motor is assembled to its warhead, fusing,flight stabilization apparatus, etc., access to the insideis difficult. The inert propellant furnished had ap-proximately the same physical characteristics as the livepropellant, including the characteristic of many plasticmaterials to flow at moderately elevated temperatures.Because such plastic flow could distort the core con-figuration and result in erratic flight, lowlevel heatingnecessitates use of a radiometer that is sensitive at nearambient temperatures.

The specimen was mounted on a revolving spit androtated at 3-rpm under a 250-W heat lamp. The sameradiometer was focused on a single point 900 after theheat source so that the surface temperature was beingread 5 see after heat injection. Only a single line scanaround the motor was wanted in this test, so the in-strument was set in its spot radiometer option with anX-Y recorder display.

Figure 9 shows a recording of one scan. The ordinaterepresents the thermal emission of the surface. A

highly reflective strip longitudinally along the cylindersurface served as a reference marker because of itslower emissivity and subsurface temperature. Thetime between the markers on the X axis represents 20sec, or one revolution of the specimen, or one completescan of its circumference.

The recording begins with the drop in radiant energyat the marker, followed by the high amplitude of thefirst interface defect, a lower peak at the second inter-face defect, and a return to the marker. These peakscan be explained as the retention of more heat in thecase over unbonded areas resulting from the interruptedheat flow into the liner, in the first instance, and intothe propellant in the second. The defect indicationsare broader than the 2.5-cm bond defects because of thelateral heat conduction in the steel.

ConclusionThe three cases cited here are considered to be repre-

sentative of a wide range of bond inspection problemsand were selected to show basic differences in heating,scanning, and display techniques. As with any otherNDT method, a familiarity with the material(s), theirtype flaws, and the basic laws relating to the NDTapproach being used, is required.

Unlike other NDT methods, the basics of thermo-dynamics are almost instinctive, as distinguished fromlearned, so that the using inspector or technician doesnot require an extensive theoretical training in order topractice the method. Additionally, he can readilyinterpret the thermographic display by visual com-parison of the tested area in terms of the possible flawsas related to the specimen construction or configuration.

The noncontacting feature makes possible the examin-ing of specimens that are hot, cold, remote, electricallyenergized, etc., frequently without the need to introduceany thermal energy as when the heat or cold of pro-cessing or operation is present.

Infrared not only fills important gaps in the art ofnondestructive testing, but it can also augment, verify,or replace other techniques.

References1. R. W. Astheimer and E. M. Wormser, J. Opt. Soc. Amer.

49, 2 (1954).2. N. R. Burrowes, "Emissivity Equalization by Thermo-

setting Coatings." Transactions of the Infrared Sessions,Feb. 65 Convention of the American Society for Non-destructive Testing.

October 1968 will featureLight Scattering Spores

1742 APPLIED OPTICS / Vol. 7, No. 9 / September 1968