Study On The Surface Integrity Of Titanium Alloy TC4 In Ultra High Speed Grinding Process

Titanium alloy is one of typical difficult-to-machine materials, and is widely used in industries due to its excellent properties. But there are many disadvantages such as high grinding force, high grinding temperature in the grinding process of titanium alloy, result in surface burn, residual stress and worse quality. Ultra-high speed grinding can improve the surface quality and productivities by high grinding wheel speed, which has low grinding temperature, low grinding force, low deformation. By comparison of advantages of conventional grinding and ultra-high speed grinding, ultra-high speed grinding is regarded to an efficiency method to machine titanium alloys.From literatures review, it is known that there are a few studies on grinding of titanium alloy. The purpose of this study aim at ultra-high speed grinding technology on titanium alloy grinding on 314m/s grinder designed by Hunan University. By analysis of surface integrity of titanium alloy, some conclusions are drawn as following:(1) It reduced the grinding force and cling to grinding wheel while titanium alloy is machined by ultra-high speed grinding method, therefore, better surface roughness was obtained.(2) The process parameters affect grinding force per area greatly in the process of ultra-high speed grinding of titanium alloy. Normal and tangible grinding force per area becomes bigger with increasing of grinding depth and workpiece speed. But grinding force per area decrease at high grinding wheel speed.(3) The white layer and heat-affected layer exist on the surface/subsurface of titanium alloy, and the thickness of white layer is thin. The bigger grinding depth, the thicker white layer is. The change if micro hardness is consistent with thickness of white layer and heat-affected layer.(4) It predicted the surface roughness of titanium alloy by method of BP neural network which grinding wheel speed, grinding depth and workpiece speed are used as input parameter. The simulation results shows BP neural network method can be used to predict surface roughness, and it is basically consistent with experimental results.