The effect of coating to substrate hardness ratio on the behaviour of coating failure caused by single pass scratch test


Student thesis: Master's Thesis

View graph of relations


  • Kin Ho LAU


Awarding Institution
Award date3 Oct 2006


The demand of hard coatings in industries is increasing because of its advantages such as high wear resistance. It is important to pick the right coatings for the components to achieve the best adhesion. One key parameter affecting the adhesion is the hardness ratio between coating and substrate and there is currently no research output focused on this relationship in scratch tests. Therefore, the present work aims to investigate the influence of the coating to substrate hardness ratio on the resulting scratch failure modes by using single pass scratch tester. Four different kinds of coatings, Al, W, TiN and Ti1-xAlxN coatings were deposited on three types of M42 high speed steel (HSS) with different hardness, aluminium and Tungsten carbide substrates by using closed field unbalanced magnetron sputtering (CFUBMS). As a result, a range of samples with different coating to substrate hardness ratio were obtained. Hysitron Ubi 1 nano-hardness tester was used to find the nano-hardness value of the coatings. Teer scratch tester ST-3001 was used to perform scratch tests on all the samples and the failure modes on the scratch tracks were observed under scanning electron microscopy (SEM). After data collection, scratch failure maps based on the correlation between their hardness ratio and the failure modes were built. Totally 8 different failure mechanisms were found in the coatings. They included discontinuous chip removal, continuous chip removal, tensile cracks, through thickness tensile cracks, external transverse crack, edge spallation, continuous edge spallation and breakthrough of coating. Results showed that soft Al coatings did not demonstrate meaningful correlation because of its poor adhesion. It was observed that W, TiN and Ti1-xAlxN coatings behaved similarly, their hardness ratios have similar effect on the failure modes. The relationship of them can be divided into regions of no failure, failure (crack/spallation) and coating breakthrough. The critical loading at a particular failure mode occurred decreased when the hardness ratio increased. It appeared that the relationship between hardness ratio and failure modes is independent on the kind of hard coatings used. Therefore, the map may be generalized for different kinds of hard coatings design for further work. When considering the Hertzian contact stress added on the substrates instead of the normal load, it clearly showed three different regions in the graph which each region represented different substrates. Only coatings on steel substrates provided good correlation between hardness ratio and failure modes. It demonstrated that the adhesive strength of the coating decreases with the Young’s modulus of the substrate. Therefore, we focused on studying the coatings with steel substrates which have same Young’s modulus. Their graph of normal load against hardness ratio can be used to find a suitable coating on the substrate and to determine the adhesion of the coatings. In conclusion, the effect of hardness ratio on scratch failure modes can apply on steel substrates which have same Young’s modulus. However, it cannot infer that the same relationship can be applied on other substrates due to the difference of the material properties.

    Research areas

  • Hardness, Testing, Coatings