Seite - 46 - in 3D Printing of Metals

Metals 2016,6, 284 Figure9. (a)COFcurvesofTi64samplesasa functionofnumberof sliding laps. (b)MeanCOFsof EBM-builtandas-castTi64samples. Figure9bshowsthemeanCOFvaluescalculatedfromthreegroupsofexperimentaldata. It is revealedthat the thin1mmsamplehasahigherCOFthanthe10mmsample,andthe lowestbeing the as-cast sample. The larger deviation ofCOFvalues in 0.5mmsample ismainly attributed to theobviousmartensiteprotrusionswithin itsmicrostructure.Moreover,weobservedthat themean COFof the 20mmsamplewas nearly the same as the as-cast sample, while COFs of the rest of theEBM-built samples increasedwith thedecrease in sample thickness. Interestingly, the thinnest EBM-built samplewith thehighestmeanCOFhas the lowest specificwearrate (asseen inFigure10a). Moreover, thespecificwearrate increasedwith increase insample thickness.Figure10breveals the 2-Ddepthprofilesof the samples’wear tracks correspondingly. Thedecrease in specificwear rate withdecrease in sample thickness ismainlydue to the increasedhardness as a result of thefiner microstructureand thepresenceofα′ in the thinnerEBM-built samples. Ingeneral, an increase in hardnesswill correspondingly lead toan increase in shear strengthof thematerial, and itpossibly results inahigher frictioncoefficient [22]. Ahigher shear strengthdefinesahigherability to resist plastic shearingduringslidingandassuchahigherwearresistance [22]. Therefore, itmaysuggest that the thinnerEBM-built sample,whichhasahigherhardnessbecauseof the fast cooling rateas comparedtotheas-cast sample,exhibitshighwearresistanceand,assuch, reduces theroughening of thesurfaceduringrubbingand, thus,produces lesserweardebris. This, consequently, results ina lowerspecificwearratebutahighmeanCOF. Figure10. (a) Specificwear ratesandmicrohardnessvaluesofEBM-built andas-castTi64 samples; (b)Surfaceprofilesmeasuredacross thewear tracksof thewornTi64samples. 46