Predicting the effect of machining and hardness of work piece parameters in roughness of final surface of medium Carbon Steel (1060) by the neural network

Surface roughness in the machined pieces is as one of the main and effective parameters on the quality of the final product; therefore selecting the parameters which influence roughness of final surface is so important. In this research the effect of main parameters of machining on the surface roughness of extended Carbon Steel (1060) in the machining process is studied using the artificial neural network that is one of the most applicable tools in the case of artificial intelligence. Cutting condition (the speed of cutting, depth of cut, feed rate and cutting tool position) is considered as the network inlets and roughness of final surface is considered as the network outlet. Also the under study piece has been investigated at (Brinell hardness 165, 218, 247).The created model for Brinell hardness of 165, 218, and 247 is instructed and the instruction accuracy of the created network is about 99.665%. Also the efficancy of created models is studied that prediction of the surface roughness for hardness amount of 165, 218 and 247 is shown in one piece in the following table. Then using simulate annealing algorithm it is made an effort to optimize the created model and at last the best state and form for the medium Carbon Steel (1060) from 10 performance stages.

Keywords:

extended carbon steel (1060) , surface roughness , artificial neural networks , machining , simulate annealing(SA)