Influence of test methodology and probe geometry on nanoscale fatigue failure of diamond-like carbon film

Abstract

The aim of this paper is to investigate the mechanism of nanoscale fatigue using nano-impact and multiple-loading cycle nanoindentation tests, and compare it to previously reported findings of nanoscale fatigue using integrated stiffness and depth sensing approach. Two different film loading mechanisms, loading history and indenter shapes are compared to comprehend the influence of test methodology on the nanoscale fatigue failure mechanisms of a DLC film. An amorphous 100nm thick DLC film was deposited on a 500μm silicon substrate using sputtering of graphite target in pure argon atmosphere. Nano-impact and multiple-load cycle indentations were performed in the load range of 100μN to 1000μN and 0.1mN to 100mN, respectively. Both test types were conducted using conical and Berkovich indenters. Results indicate that for the case of a conical indenter, the combination of nano-impact and multiple-loading cycle nanoindentation tests provides information on the life and failure mechanism of the DLC film, which is comparable to the previously reported findings using the integrated stiffness and depth sensing approach. However, the comparison of results is sensitive to the applied load, loading mechanism, test-type and probe geometry. The loading mechanism and load history are therefore critical which also lead to two different definitions of film failure. The choice of exact test methodology, load and probe geometry should therefore be dictated by the in-service tribological conditions, and where necessary both test methodologies can be used to provide better insights of failure mechanism. Molecular dynamics (MD) simulations of the elastic response of nanoindentation are reported, which indicate that the elastic modulus of the film measured using MD simulation was higher than that experimentally measured. This difference is attributed to the factors related to the presence of material defects, crystal structure, residual stress, indenter geometry and loading/unloading rate differences between the MD and experimental results.