Numerical generation of grinding wheel surfaces based on time series method

Abstract

Surface integrity has a significant influence on the functional behavior of engineering components. An effective grinding wheel topography model is the foundation to realize the controllable design and manufacturing of grinding workpiece surface integrity. In this study, a moving average model (MA) based on time series method was proposed to generate the grinding wheel surface topography. 3D grinding wheel topographies were measured by an LSM 700 laser scanning confocal microscope (CLSM). The Johnson transformation system was applied to generate the non-Gaussian sequence with predetermined statistical parameters. To address the convergence and memory requirement problems, the non-linear conjugate gradient method (NCGM) with an exact line search by the secant method was employed to solve the system of nonlinear equations. The results showed a good agreement between the measured and the generated surfaces in terms of the autocorrelation function (ACF) and statistical parameters was achieved. It is worth noting that the computing time of the model developed in this study is only determined by the autocorrelation lengths which are almost independent of the size of the grinding wheel. This technique could be more efficient when applied to grinding wheels with large size. Therefore, this method may serve as an effective way to solve the grinding wheel topography reconstruction problem.