Study of break down of thin hard surface coatings with emphasis on finite element simulations

The first part of the Ph.D. work presents the identification of a multi-linear stress strain law for AISI M2 tool steel (High Speed Steel - HSS) which enables the determination of the critical stress and strain components which lead to the failure of Titanium Nitride (TiN) coating, coated on to it. The procedure of identification relies on an inverse method where Rockwell indentation experimental characteristics (profiles obtained under successive load indentations by means of a profilometer) are matched with the profiles obtained by Finite Element (FE)-simulation of the indentation process using LS-Dyna non-linear implicit FE-code and the Ximcor inverse modelling/optimization program. Radial cracks, observed in the coating due to indentation, tends to arrest at characteristic length (Lc). The stress/strain condition in the coating at these lengths are the critical stress/strain components "just" required to break the coating and it is calculated explicitly. Lc for a 1500N load indentation is measured by the use of Scanning Electron Microscopy (SEM). By using the identified stress-strain law in a 1500N load indentation FE-simulation, the substrate strains (small strains) are computed at Lc. They are then used in calculating the stresses (nett maximum tensile stress) in the coating assuming it to be elastic and under plane stress condition. The obtained results are compared to earlier works and they serve as limiting values/failure criterion. The result (nett maximum tensile stress) is used in finding the load bearing capacity of the TiN/HSS coating-substrate system under normal loading and sliding of abrasive particles. The second part of the Ph.D. work presents the investigation of elastic stresses developed in the work piece material due to impact and sliding of abrasive particles in tribological contact situations. LS-Dyna implicit finite element analysis is used to investigate these contact stresses in the work piece material under imposed Hertzian pressure loading. At first the predicted stress field is compared with analytical solutions under both normal and frictional tractions. The model is parametric and built in LS-Ingrid in terms of contact width ?a?. Application of Hertz pressure with frictional tractions is one of the issues of this study. The study is a preliminary step in the extension of the model where the work piece will be modelled with a thin hard layer/coating (TiN) under the same contact situations. A study of the load bearing capacity of TiN/HSS system compared to the uncoated one is done. The nett maximum tensile stress calculated from the first part of the Ph.D. work is used as a check in the study. An investigation of the optimal thickness of TiN coating on the HSS steel is discussed. The third Part of the Ph.D. utilizes the know-how of inverse modelling in Part I and the modelling method as in Part II to present the idea of a new way of identifying the residual stress and the Young?s modulus of a hard thin coating like TiN on elasto-plastic substrates like HSS. The method assumes a) ultra micro/nano indentation experiments (range of 10 - 200 mN) using Ultra Micro Indentation Systems (UMIS) a trademark product of Commonwealth Scientific and Industrial Research Organization (CSIRO), Australia, b) axi-symmetric FE-simulation of the same and c) inverse modeling technique. The basic idea of the inverse modeling is to combine the experiments and the FE-simulations through an optimization scheme, so as to identify the optimum design variables/parameters (the residual stress and the Young?s modulus of the coating) which minimize the error between the experimental data and the FE-data. The experimental data and the FE-data are the profiles obtained under a selected critical load which can be obtained by Atomic Force Microscope (AFM) scanning. The FE-simulation involves a coupled thermal and structural analysis. The thermal part creates a equi-biaxial residual stress in the coating and the structural part only involves the mechanical loading. The chapter only presents the idea, i.e., the numerical part and the experimental data is presently assumed as synthetic data. The project was finalised in 2004. (Anatha Ram, Joachim Danckert, Torben Gade Faurholdt, Department of Production, AAU)