Prediction of the tribological parameters controlling the tool wear is one of the most complex research axes in the metal cutting literature. The nature of the friction on the tool-chip interface is the main process which influences the distribution of stresses and temperatures which in turn activates the thermomechanical process governing tool wear. Under extreme conditions of temperature, strain rates and pressure, occurring especially in high speed machining, the adhered contact phenomenon is highly localized especially near to the tool tip and extremely non-linear due to strong influence of the secondary shear zone (SSZ) and the nature of bonds between asperities of tool and chip. In addition, the analysis based on post-mortem examinations of chips and worn tools, especially in hard metal cutting alloy and high speed machining, show that the adhered friction with intimate contact and no relative motion of the material chip on the interface, are the principal cause of appearance of the plastic deformation layers (SSZ). This localized shear zone plays a role of intensive heat sources interacting with the tool side and, in turn, activates diffusive and abrasive wear. In this work, a hybrid model combining analytical and numerical approaches is performed in order to solve the non-linear thermomechanical problem on the chip and predicts the nature of friction contact, i.e. fully sliding, sticking/sliding or fully sticking contact, and these for a given distribution of asperities, characterizing the ratio between the real and apparent contact areas A(r)/A(n) on the interface, and a given global friction coefficient (mu) over bar, characterizing the ratio of the experimental cutting forces. The shear stress generated in the primary shear band, the energy produced in the secondary shear zone, the local friction coefficient and the friction energy produced on the sliding part of contact, the proportion of the sticking/sliding interfaces, and geometrical parameters of the chip, are obtained by analytical means. The analysis of this model is based on experimental data and applied for a large cutting speed (1 ms(-1) <= V <= 60 ms(-1)) and able to give some correlation about the distribution of adhesions marks on the contact with respect to the distribution of the tribological parameters at the interface. This model can be used in improving tool wear prediction and the estimation of tool-life