Profile characterisation and response surface modelling of laser surface hardened Cr12 mould steel

Abstract

Abstract For laser surface hardening (LSH) process, it is of great significance to obtain bead information from hardened layer as it is closely related to mechanical properties of hardened area. Correlating process parameters to fewer geometry features would greatly reduce the modeling complexity. In this paper, a three-factor-five-level central composite design (CCD) is proposed, based on response surface methodology (RSM), to model, analyse and optimize the influences of three process parameters, namely: laser power (LP), scanning speed (SS), and defocusing distance (DD) on the bead profile of laser surface hardened of Cr12 mould steel. Transverse cross-sectional profiles were firstly modelled by parabola function, in contrast to sine function, ellipse function and Gaussian function. The profile focal length (FL) as well as the hardened depth (HD) were selected as characteristic lengths for the bead profile. The results show that the response models interpret the hardened depth and focal length successfully and satisfactorily. It is observed that both responses are determined by three input variables with quadratic correlations. The predicted results are in good agreement with experimental counterparts. Validation experiment was finally conducted for testing the generalisation of fitted regression models. As the maximum relative prediction errors for both responses are respectively 3.91% and -7.35%, good generalisation of developed models has been accomplished.