The depth profiles of the thermophysical properties of alloy systems, for example, shape memory alloys (NiTi), steel, and tool steel, can vary considerably due to rolling, surface machining, heat treatment, mechanical wear, and erosion. The same is true for coated tool steel samples, which show variations of the effective thermal depth profiles due to the effects of substrate preparation and deposition of the coatings, for example, plasma-etching, arc erosion, nitriding, chemical vapor deposition (CVD), physical vapor deposition (PVD), sputter deposition, and plasma spraying. In this work we present a large variety of measured effective thermal depth profiles. In a first step, we identify the effects of coating deposition and substrate preparation on the measured depth profiles. In a second step, we identify and try to quantify the effects of mechanical wear and erosion of both coated and uncoated surface. To this finality, the signals, which have been measured with the help of IR radiometry as a function of the modulation frequency, have been calibrated with reference signals measured for homogeneous samples of glassy carbon. The normalized amplitudes and phases have been approximated using layer models, mainly the two- and three-layer model with an opaque first layer, with respect to both the visible and the IR spectrum. Additionally, the signals measured for different coatings have been normalized against each other. By this latter calibration procedure, even smaller details and differences of coating deposition and substrate preparation can be identified, as well as the effects of wear and surface erosion. The virgin coated samples normally can well be described by the two-layer model, and the thermal transport parameters of the coatings as a whole can be determined quantitatively with rather good reliability (Ref. 1). The deviations from the two-layer model, which can be related to details of the deposition process, for example, to gradient layers or bond layers, are described by thermal diffusion times, which are orders of magnitude below the thermal diffusion time of the coating as a whole, so that they are negligible in the numerical simulation of heating or cooling processes. In many cases, however, the quantitative description based on layer models with only a small number of adjustable parameters is not possible, since the discrepancies between the reality of measurement and such a model are too large. This is the case for: (i) combined arc erosion and plasma nitriding of surfaces, which shows roughness effects with two different characteristic lengths (Ref. 2); (ii) CVD coatings, which show effects of preferred growth of some grains; (iii) strongly eroded coatings of cutting tools; and (iv) NiTi surfaces after rolling, surface machining, surface polishing, etc., where mechanical treatment can produce layer systems of the effective thermal properties with both reduced and increased effusivity values (Ref. 3).
Read full abstract