ARMY RESEARCH LABORATORY - DTIC

1. IntroductionThree final candidate nonchromate conversion coatings were selected from agroup of six original candidates, based upon feedback from the Bradley FightingVehicle Systems (BFVS) Environmental Management Team (EMT) and fromprevious EMT-sponsored studies performed by the U.S. Army ResearchLaboratory (ARL) [1-3]. Criteria for final consideration included corrosion andadhesion performance, as well as economic feasibility and scaleabilty toextremely large baths capable of treating an entire Bradley vehicle hull.Aluminum alloys 5083, 7039, and 6061 were selected by the committee forinvestigation with the chromate conversion coating alternatives. As in the priorstudy, all candidate vendors agreed to have their products evaluated andsupplied the pretreatment materials. In an effort to maintain consistency andequal application conditions, all vendors traveled to Concurrent TechnologiesCorporation in Johnstown, PA,* and supervised the pretreatment as well as thecoating stages of specimen preparation.2. Experimental ProcedureAluminum panels (39 each nominally 10 cm x 15 cm x 0.6 cm) of alloys 5083H131 and 7039-T64 were machined from rolled armor plate stock. A similarquantity of aluminum 6061-T6 industry standard test coupons were alsoobtained. All coupons were clearly labeled using a mechanical die topermanently indent the experimental designation. Thirteen panels with eachconversion coating combination were prepared. From each set of 13 panels, 10were sent to ARL—three for American Society for Testing and Materials (ASTM)standard B117 [4] salt fog, three for GM 9540P [5] cyclic corrosion, one forlaboratory adhesion (ASTM D3359 A and B [6] and ASTM D4541 [7]), and threefor exposure at the Aberdeen Test Center (ATC) automotive test track [8,9]. Thethree remaining panels were sent for mounting and fielded durability exposureon actual Bradley vehicles stationed at Camp Roberts, CA. In addition to panels,actual components used on Bradleys were treated with the candidates anddistributed for laboratory, test track, and field exposure. The componentsconsisted of headlight guards, taillight guards, antenna brackets, and accesscovers. All of the Bradley components were grit blasted prior to thepretreatment stage to clean and remove prior coatings. A detailed breakdown ofthe panel and component distribution as well as numbering schemes are listed inConcurrent Technologies Corporation, 100 CTC Drive, Johnstown, PA 15904.

Tables 1-3. Personnel from each of the pretreatment vendors were presentduring the surface prep, pretreatment, and coating application stages(Tables 4-6) to ensure equal and optimal conditions for the pretreatmentcandidates. Following the cleaning and pretreatment stages, all of the panels andcomponents were coated using standard Chemical Agent Resistant Coating(CARC) consisting of MIL-P-53022 [10] epoxy primer and MIL-C-53039 [11]topcoat. One component, an access cover for the Bradley bilge pump, wastopcoated with seafoam green MIL-C-22750 [12] which is used for interiorBradley surfaces.Salt fog testing in accordance with ASTM B117 [4] was used to screen theCARC-coated panels. The solution used was the standard 5% NaCl. All panelswere photographed prior to testing, upon significant changes, and at failure.The panels, (three each) for each conversion coating, were exposed for 3000 hr ofsalt fog. These panels were "X" scribed using a standard carbide-tipped,hardened steel scribe. Figure 1 shows a representative photo of initial specimenappearance after scribing (all painted panels appeared visually identical beforetesting). Final detailed ratings for the 3000-hr duration were assessed usingASTM D1654A [13] which quantitatively indicates the damage caused by pittingor delamination outwards from the scribe (Table 7).A cyclic corrosion test chamber (CCTC) was used to evaluate the CARC-coatedtest panels and Bradley headlight guards. For each conversion coating tested,three primed and topcoated CARC panels were subjected to CCTC testing. As insalt fog, the panels were X scribed. The scribed panels were placed into thechamber (Figure 2) and tested using GM Standard Test 9540P [5], method B,which provides a more realistic accelerated environmental test than conventionalsalt fog [14]. The standard 0.9% NaCl, 0.1% CaCl2, 0.25% NaHCOa test solutionwas used. The 9540P test consists of 18 separate stages that include thefollowing: saltwater spray, humidity, drying, ambient, and heated drying. Theenvironmental conditions and duration of each stage for one complete 9540P [5]cycle are provided in Table 8. In addition, standard plain carbon steel calibrationcoupons described in 9540P [5] and supplied by GM were initially weighed andsubsequently monitored for mass loss at intervals set by the specification. Masslosses measured for steel coupons used for this test were within acceptableparameters stated in the GM specification (Table 9). The panels werephotographed or digitally scanned prior to testing, upon significantobservations, and at the suspension of the testing (120 cycles). As with B117 [4]salt fog, the extent of damage was assessed using ASTM D1654 [13].Three scribed test panels of each alloy/pretreatment and actual Bradleycomponents were attached to selected locations on a Family of Medium TacticalVehicles (FMTV) test vehicle at ATC for exposure to the Munson automotivetest track facility [8]. Figures 3 and 4 illustrate the mounting schemes used fortest panels and Bradley components. The Munson facility, based upon General

Motor's (GM's) road test, combines vehicle durability tests with acceleratedcorrosion events to simulate conditions that a fielded Army vehicle would likelyencounter (Tables 10-13). The test panels and components were exposed to testtrack conditions for seven phases that represented seven equivalent years ofexposure in a fielded Army environment. Interestingly, seven years correspondsto the overhaul schedule for Bradleys in service but in this case, the seven phasesof exposure happened to be the actual exposure time remaining on the FMTVtest vehicle. Each phase consisted of 15 driving and corrosion cycles described inTable 14. At several points within each phase, a high-temperature, highhumidity exposure chamber was used to accelerate corrosion conditionsfollowing the road tests. As in GM 9540P [5], the corrosiveness of theenvironment was monitored using standard plain carbon steel mass losscoupons. These coupons were placed at several locations on the vehicle andwere used to monitor the corrosion rates from each phase of testing. Thecoupons allowed testers to adjust the number of humidity chamber exposuresduring each phase to increase or decrease the severity of the test [9]. Panels andcomponents were assessed during the exposure duration for degradation due tocorrosion and loss of adhesion. At the conclusion of the test track exposure,additional laboratory adhesion measurements were performed.Three unscribed test panels of each alloy/pretreatment and actual Bradleycomponents were attached to selected locations on an actual Bradley test vehicleat Camp Roberts for 94 days. During the 94-day exposure, the vehicles traveled93 test track and 4000 durability miles. Figure 5 illustrates the mounting locationused for test panels on the Bradley test vehicle.Paint adhesion for both primed and topcoated panels was determined using awet adhesion test (Method 6301.2 of Federal Test Method Standard No. 141C [15]as specified in MIL-C-81706 [16]). In this test, a standard adhesive tape is used tocheck adhesion on painted specimens after soaking for 24 hr in deionized water.After soaking, each panel is removed and quickly dried. Two parallel scribes,1 in apart, are made within the first minute after removal. Tape is uniformlyapplied across the scribes and then immediately removed. Upon removal, anyevidence of paint separation is noted by visual observation of both the panel andthe tape. MIL-C-81706 [16] describes adhesion based on a pass or fail system. Toreceive a "pass" rating, there must be no separation of the paint from thesubstrate or between layers of the paint. Additionally, a more detailed rating inaccordance with ASTM D3359A [6] was used (Table 15).Dry adhesion measurements were obtained in accordance with ASTM D3359B[6]. This method employs a 6 x 6 grid of perpendicular scribes spaced at 2-mmintervals. Standard tape is uniformly applied over the cross-hatched area andthen immediately removed. Once again, upon removal, any evidence of paintseparation is noted by visual observation of both the panel and the tape. Therating method for ASTM D3359B [6] is described in detail (Table 16). Dry

adhesion measurements using this method were performed on all of the exposedpanels and components in addition to the initial set reserved solely for adhesionpurposes.Pull-off adhesion measurements in accordance with ASTM D4541 [7] wereperformed on selected laboratory adhesion and exposed panels and components.For the pull-off adhesion test, a loading fixture commonly referred to as a "dolly"is secured normal to the coating surface using an adhesive. The adhesive usedfor the treated and untreated panels was cyanoacrylate. After allowing theadhesive to cure for 24 hr in laboratory conditions (Table 17), the attached dollywas inserted into the test apparatus. The load applied by the apparatus wasgradually increased and monitored on the gauge until a plug of coating wasdetached. The failure value (in pounds per square inch [psi]) and the failuremode were characterized and recorded. The pull-off test apparatus and dollyconfiguration are illustrated in Figure 6. Examples of different failure modesobtained from actual panels are provided in Figures 7-9. For the pull-off test, thespecimen must be of sufficient minimum thickness to ensure that the coaxial loadapplied during the removal stage does not distort the substrate material andcause a bulging or "trampolining effect." On thin specimens such as the Al 6061test panels, the resultant bulge causes the coating to radially peel away outwardsfrom the center instead of uniformly pulling away in pure tension and thusresults in significantly lower readings than for identically prepared thickspecimens. Of the panels evaluated in the test matrix, only the 5083 aluminumarmor test panels at 0.375 in and the 7039 panels at 0.250 in had adequatethickness for valid pull-off test procedures. All of the Bradley components withthe exception of the bilge pump cover had sufficient thickness to accommodatethe pull-off procedure.An additional method used to assess post exposure adhesion on the nonscribedBradley-mounted Al 6061 test panels was the mandrel bend test. The apparatusused was the "conical" type in accordance with ASTM D522 [17] method A. Thisprocedure examined cured coatings of uniform thickness on sheet metal bybending them over a conical mandrel. As in the pull-off test, thickness issuesnecessitated the exclusion of some specimen substrates. In this test, only the Al6061 panels (1/16 in) were thin enough to be used for this procedure. Theapparatus used for the bending procedure is shown in Figure 10. Immediatelyafter bending, the specimen coatings were visually examined for cracking. Themandrel diameter at which cracking ceased as determined in the plot in Figure11(a) is taken as the resistance to cracking value. Figure 11(b) illustrates thecorrection factor for coating thickness. The total elongation of the coating can becalculated using the measurements and the sum of the elongation from thefigures as follows:E ei tci,

where:E total elongation, %;d elongation from Figure 11(a), %;t thickness, mils, andci correction factor from Figure 11(b).Coating elongation calculations for panels cracked down the entire length werenot possible. However, coating/substrate delamination measurements weretaken and compared among the specimens.3. Results3.1Salt FogBased upon salt fog performance from a previous study [1], the laboratory saltfog panels were exposed for 3000 hr and assigned one of the ASTM D1654 [13]rating codes in Table 7. In addition, the panels were scraped along the fulllengths of the scribes to confirm the extent of the creepback damage uponconclusion of the exposure. The creepback ratings at 3000 hr are plotted inFigures 12-14. For Al 5083, corrosion performance was excellent for BrentOxsilan AL-0500 and Alodine 5200 pretreatments at or near a perfect 10. Incontrast, blistering under the paint and loss of adhesion was severe for theOrganosilane-pretreated panels (Figure 15). For Al 7039 panels, Organosilanehad the best overall performance followed by Alodine 5200. Unfortunately, therewas wide scatter among the data for these panels with ratings ranging from 1 to10. The Brent treatment was consistently low in performance with ratings of2 across all three panels. For Al 6061 panels, there was a wide disparity betweenthe pretreatment candidates. The Brent process was clearly superior with aperfect 10 or no damage vs. extensive blistering and creepback at 16 mm orfurther on the Alodine 5200 and Organosilane panels rendering assessments of 0and 1. Post-exposure dry- and pull-off-adhesion measurements were alsoperformed on the panels, and the results are summarized in sections 3.6 and 3.7.3.2Cyclic Corrosion Test Chamber (CCTC) (GM 9540P [5])The painted panels and components were all subjected to 120 cycles of GM 9540P[5]. The assessment used for 9540P is identical to the assessment for ASTM B117[4] salt fog for painted specimens (Table 7). The creepback ratings at 120 cyclesare plotted in Figures 16-19. As in salt fog, the failure mode for the painted

panels was blistering along the scribe. For Al 5083, most panels sustained thefull 120-cycle duration without any damage.An exception was oneOrganosilane panel rated a 6 which had some blistering along the scribe. As insalt fog, the best performing pretreatments for Al 7039 were Organosilanefollowed closely by Alodine 5200 with much less scatter among the data and,similarly, the Brent pretreated panels had significantly greater damage. For Al6061, overall damage was less severe than for salt fog and once again the Brentpretreatment excelled for this alloy. Performance for Brent panels ranged from 8to 10, Organosilane was next best with scattered ratings ranging from 5 to 10,and Alodine 5200 rated poorest with data ranging from 4 to 5. Headlight guardsfrom actual Bradleys, composed of Al 5083, scribed on three flat surfaces each,were also exposed. For these components, all three pretreatments endured thefull 120-cycle duration with no corrosion damage. As in salt fog, post-exposuredry- and pull-off-adhesion measurements were performed on the panels andcomponents, and the results are summarized in sections 3.6 and 3.7.3.3ATC Automotive Test TrackPanels and components were all scribed and subjected to seven phases or"equivalent years" of exposure at the ATC test track. As in salt fog andGM 9540P [5], the assessment used for characterization of corrosion damage wasASTM D1654 [13] (Table 7). It became immediately apparent that the test trackenvironment was significantly more severe than the chamber-based methods.Damage measured at just four phases was already severe for several of the testpanels (Figure 20). The extent of the damage at four-phases exposure is plottedin Figures 21-23. For the 5083 panels the Brent Oxsilan AL-0500 and Alodine5200 were undamaged; however, the corrosion damage on the Organosilanetreated panels was severe and consistent across all three panels—one eachmounted on the sides and the tailgate—all rating a 4. The 7039 panels alsoshowed significant damage at four-phases but much of it came fromdelamination problems as was seen in Figure 20. Much of the adhesion problemsappeared in the Brent and in the Alodine 5200 panels with two of the Alodinesdegraded to 5 ratings and one of the Brent panels degraded to a 0. TheOrganosilane-treated 7039 panels fared well and rated from 8 to 9. The 6061panels all showed significant blistering from their scribe with Organosilanesrating 3 to 5, Brent Oxsilan AL-0500 rating 3 to 4, and Alodine 5200 rating 1 to 3.Encouragingly, the Bradley taillight guards and bilge pump covers fared betterthan the panels at four phases with little or no damage. At this point in the testtrack exposure, selected panels were removed from the test vehicle for review bythe Bradley Environmental Management Team.The remaining panels and components continued exposure to seven test trackphases. At the conclusion of seven phases, the remaining panels andcomponents were evaluated and are plotted in Figures 24-27. For the remaining

panels, the relative trends noted among the pretreatments for each alloyremained consistent with the four-phase evaluations. The only noted differencefor the panels was the advancement of the blistering and delamination from thescribes. In contrast with the four-phase observations, the Bradley componentsboth exhibited corrosion damage. For the Organosilane-treated Al 5083 taillightguard, the blistering creep was severe and rated a 0. In addition, there wassevere chipping and paint loss near the mounting point which was not visible foreither of the other two treatments, both of which scored perfect 10 ratings(Figures 28-29). The results for the Al 5083 bilge cover differed in that theOrganosilane cover rated best with a 10 followed by Alodine 5200 and Brentwhich rated 9 and 5, respectively (Figure 30). As in salt fog and GM 9540P [5],post-exposure dry- and pull-off-adhesion measurements were performed on thepanels and components, and the results are summarized in sections 3.6 and 3.7.3.4Fielded Bradley Vehicle Exposure—Camp Roberts, CAThe panels and components exposed at Camp Roberts upon completion of thedurability miles were sent to ARL for examination. None of the panels orcomponents were scribed before exposure; however, panels and componentswere examined closely for blistering or delamination. While there was somechalking of the 383 green CARC MIL-C-53039 [11] topcoat as a result ofultraviolet exposure from sunlight, none of the panels or components returnedexhibited blistering or delamination problems.

Measuring Adhesion by Tape Test." ASTM D3359, West Conshohocken, PA, 1987) wet adhesion, ASTM D3359B dry adhesion, ASTM D4541 (ASTM. "Standard Test Method for Pull-Off Strength of Coated Specimens Subjected to Corrosive Environments." ASTM D4541, West Conshohocken PA, 1989) pull-off adhesion,