Prediction of thermal damage in superfinish hard machined surfaces

A new concept of using single-step dry machining of hardened steels has been proposed at Purdue as a finish process to replace grinding and other finish processes. This concept has been proven to be feasible and disclosed in US Patent 5878496 [Liu and Mittal, 1999). The objective of this research is to develop the capability to predict the thermal damage beneath hard machined surfaces so that optimal machining conditions can be obtained. This research is divided into two major parts. In the first part, the thermal damage models of the workpiece material are constructed, through heat treatment experiment, for the two damage types, namely, re-tempering and re-quenching, in hard machined surfaces. The re-tempering model describes the material hardness change as a function of material thermal history, while the re-quenching model expresses the hardness based on material peak temperature. Since both thermal damage and work hardening effects exist in hard machined surfaces, and influence hardness distribution, a decoupling method is developed to separate and study the two effects individually. It is found that work hardening effect is negligible when a relatively sharp tool is used in hard machining. In the second part, a finite element (FE) model is constructed to simulate hard machining process to obtain cutting temperature. The modeling effort focuses on obtaining the real flow stress properties at elevated temperatures, through applying the developed tempering model to eliminate tempering effect in experimentally obtained strain-stress data. Meanwhile, the effort also involves the development of sub-models on the friction, heat generation, heat conduction, and annealing effect. Orthogonal machining of hardened AISI 52100 steel is simulated by the proposed FE model, and verified by experiment on cutting forces, chip morphologies, and the formation of white layer in chips. Next, the FE model is applied to simulate realistic 3D hard machining, and material thermal histories are obtained and fed into the thermal damage models. Subsequently, the distributions of thermal damage beneath machined surfaces are predicted. Excellent agreement is found between the prediction and the experiment