Artificial neural network based approach for modeling surface roughness in turning hybrid composite under compressed air cooling condition

Abstract

Modern manufacturing industries are continuously seeking for products which will be light weight, robust, less costly and possess good quality in terms of surface finish and dimensional accuracy. To attain the needs, material engineers are constantly striving to develop new metal alloys as well as composite materials. Composite materials‘ light weight, high specific strength and high specific modulus are being regarded as some gifted properties which are largely facilitating their applications in different engineering sectors. Kevlar is an organic fiber which possesses excellent specific strength and specific modulus but very poor machinability whereas glass is an inorganic fiber that possesses good machinability along with good blending capability with Kevlar while reinforced in the same matrix. So, to take the advantages of both the contrasting fibers, a hybrid composite can be formed which is going to increase the machinability of Kevlar and retain reasonably good mechanical properties as well. High speed machining is often considered as an accurate manufacturing process for making fiber reinforced plastic (FRP) products and it is established that with high cutting speed and feed productivity increases but high tool wear also takes place hence cost increases. To optimize the situation, different machining environments have been evolved to cut FRPs. Application of conventional cutting fluid, high pressure oil or minimum quantity lubricant is strictly prohibited for machining FRP materials since these machining environments change the FRP composites‘ properties. In this context, only cryogenic cooling during machining and machining under compressed cooling condition is allowed for FRP materials. It is found compressed air cooling environment has been very effective in machining FRPs when surface roughness and cutting force are taken into consideration. In this research work, turning operation of hybrid composite i.e. Kevlar and glass reinforced polyester was performed under both dry and compressed air cooling condition. Investigation was carried upon due to compare the performances of two machining environments. Cutting parameters in the machining process were cutting speed, feed and depth of cut and measured responses were surface roughness and cutting force. Finally, a predictive model of surface roughness was developed using artificial neural network (ANN) which has been validated against the experimentally found results.