Optimisation of the milling process is often based on simplified models of the cutting forces generated during the machining operations. Most of the simulation algorithms use simple linear relationship between cutting forces and the undeformed section of the chip. The coefficients of those models (often called specific pressure) are considered to be constant with respect to the other parameters. The variation of those coefficients as a function of the cutting speed or the chip thickness has been experimentally observed but most of the authors neglect this effect for simplicity sake. This paper proposes an experimental method to take the variation of these cutting coefficients with respect to cutting parameters into account. Analysis of variance is performed to identify the significant parameters. A design of experiment approach is followed to select the adequate experimental points to ensure maximal precision with the minimal number of machining tests.