Railway disc brake is vulnerable to surface damages including fade, wear, squeal, thermal cracks and fatigue being just few of them. To counteract these negative consequences, reliable thermal model that can accommodate space and time variables is essential. The aim of this study is to develop new non-axisymmetric moving heat source and compare its efficiency with pre-existing traditional models. Factors responsible for temperature spatial and temporal variation are identified first and then programmed in ANSYS APDL similar capability to a FORTRAN. Heat flux and convection coefficients are calculated by empirical equations and stored in parameters and arrays for later use, based on small time and pad angular increment. The modelling is to successfully solve the problems in traditional models by estimating surface temperature difference as high as 49 °C, within acceptable computation time. Besides, its consideration of radial distance reported variations from traditional models as high as 10% and 60% in moving heat source and axisymmetric, respectively. And, it is also verified with the literature within acceptable variation. Finally, it is suggested that the model can be applied in conducting pad geometry optimization, thermal stress and fatigue life of disc brake.