In analysis of machining processes, finite element analysis is widely used to predict forces, stress distributions, temperatures and chip formation. However, constitutive models are not always available and simulation of cutting processes with complex tool geometries can lead to extensive computation time. This article presents an approach to determine constitutive parameters of the Johnson-Cook's flow stress model by inverse modelling as well as a methodology to predict process forces and temperatures for complex three-dimensional tools using Oxley's machining theory. In the first part of this study, an analytically based computer code combined with a particle swarm optimization (PSO) algorithm is used to identify constitutive models for 70MnVS4 and an aluminium-alloyed ultra-high-carbon steel (UHC-steel) from orthogonal milling experiments. In the second part, Oxley's predictive machining theory is coupled with a multi-dexel based material removal model. Contact zone information (width of cut, undeformed chip thickness, rake angle and cutting speed) are calculated for incremental segments on the cutting edge and used as input parameters for force and temperature calculations. Subsequently, process forces are predicted for machining using the inverse determined constitutive models and compared to actual force measurements. The suggested methodology has advantages regarding the computation time compared to finite element analyses.