Metal Removal Process Optimisation using Taguchi Method - Simplex Algorithm (TM-SA) with Case Study Applications

Çankaya University Journal of Science and Engineering Volume 12, No. 2 (2015) 033-058 Metal Removal Process Optimisation using Taguchi Method - Simplex Algorithm (TM SA) with Case Study Applications Oluwaseyi A. Ajibade1, Johnson O. Agunsoye1, Sunday A. Oke2* 1 Department of Metallurgical and Materials Engineering, University of Lagos, Nigeria, 2 Department of Mechanical Engineering, University of Lagos, Nigeria, e-mail: , , Abstract: In the metal removal process industry, the current practice to optimise cutting parameters adopts a conventional method. It is based on trial and error, in which the machine operator uses experience, coupled with handbook guidelines to determine optimal parametric values of choice. This method is not accurate, is time-consuming and costly. Therefore, there is a need for a method that is scientific, costeffective and precise. Keeping this in mind, a different direction for process optimisation is taken by employing the combined Taguchi method-simplex algorithm (TM-SA) for optimal parametric setting of manufacturing processes. The process parameters were optimised and the efficiency and robustness of the method described in four literature cases. These cases involve high-speed flat-end milling, forming in hydrodynamic deep drawing, cup deep drawing and abrasive assisted drilling. The computations showed that the TM-SA exhibited superior results in one of the cases and equivalent results in others. This implies that the proposed approach could comparably serve as an optimisation framework with significant advantages of reducing experimental costs and allowing variable usages with the requirement of functional derivation. It is also easy to use. The novelty of this article is the application of a distinctly new method in optimisation for cost reduction and variable usages for the metal removal process. Potential applications of the proposed approach by material type is its usage in machining mild steel, grey cast iron, brass and aluminium with HSS and carbon steel, respectively, used as tools. Keywords: Optimisation, parameters, Taguchi method, simplex algorithm, metal removal process. 1. Introduction Nowadays, it is common to observe that many manufacturing processes such as metal removal [1,2] face huge problems including over-production of defective products, unnecessary transportation during parts processing, waiting to receive instructions from superiors on ISSN 1309 – 6788 © 2015 Çankaya University 34 O. A. Ajibade et al. actions to take. Others may be inventory build-up, over-processing of parts and underutilisation of labour. The metal removal process is a popular one in manufacturing industries with many applications in mechanical and chemical systems. Turning, milling, drilling, broaching, hoving and sawing are the major examples of mechanical metal removal processes. In addition, chemical machining, thermal, touch-cutting and electric discharge machining are significant processes in chemical metal removal. The major aim of many metal removal processes is to remove metals as quickly as possible with bias for low production time and production cost [3]. Generally, in the metal removal process industry, a number of indices are used to evaluate the performance of the process and hence many processes are referred to as multi-performance based. For example, [3] identified two indices for a high-speed end milling process as tool life and metal removal rate. These two indices were correlated with cutting parameters, including milling type, spindle speed fed per tooth, radial depth-of-cut and axial depth-of-cut. Thus, these types of performance problems exist in mechanical and chemical metal removal processes. At present, in industries, operator’s experience is used in determining the optimal values of the parameters involved in the metal removal process. This experience is often coupled with handbook values provided by machine manufacturers. Unfortunately, such practices are subjected to errors and there is no guarantee that the values obtained from such practices are near optimal results. As a consequence, it is important to find out simple and effective optimisation methods that could serve the purpose of a fast evaluation of metal removal process parameters. A metal removal optimisation problem refers to one which requires the objective function to either be minimised or maximised in conjunction with a set of constraints. Thus, in this paper, the authors take a different direction for process optimisation by applying a combined Taguchi method-simplex algorithm (TM-SA) for the optimal parameter setting of a manufacturing process. The details of the method used with four case examples are provided. In the domain of metal removal process parametric optimisation, the central theme has been the use of the Taguchi method (TM). However, with the growing interest of optimisation as a competitive tool towards sustained manufacturing [4,5,6], it is evident that the problem of obtaining improved optimisation results must be addressed urgently in view of the increasingly harsh business environment. Partnering TM and simplex algorithm promises improved optimisation results and must be pursued to the advantage of metal removal processes worldwide. Generally, in optimisation studies of metal removal processes, the overall objective is to develop and implement a methodology to predict the optimal values of CUJSE 12, No. 2 (2015) Metal Removal Process Optimisation using TM-SA 35 parameters bearing in mind the factors like production rates, lead time and cost-related objectives of the metal removal operations [7] and the energy efficiency problem [8]. Consequently, improved methodologies for the prediction of such parameters are important and necessary pursuits in the area of metal removal process. However, the most pronounced optimisation approach which may appeal to both machinists and researchers is the TM in that it reduces the cost of production. Four case studies were conducted, involving a drilling case, a case study on milling and two cases on deep drawing. A new approach, TM-SA, is then proposed and tested using the same four case studies, as carried out for the simplex algorithm (SA). Here, it is shown that for the four case studies, the three methods, TM, SA and TM-SA yield comparable results. The major reason for combining TM and SA is to take advantage of experimentation cost, while allowing a large number of variables, which practically exist in industries. To achieve the goal of improved optimisation, a new methodology, TM-SA is proposed, and the testing and comparison are in three stages. For the first stage, Taguchi method only is used and applied to a set of case studies involving experimental data from the literature. In the second instance, the SA alone is applied in the testing of the four case studies. The third stage involves the integration of TM and SA, as TM-SA in the prediction of the optimal values in all the fo (...truncated)