Normal boundary intersection method based on principal components and Taguchi’s signal-to-noise ratio applied to the multiobjective optimization of 12L14 free machining steel turning process

Mathematical modelling and optimization of surface quality and productivity in turning process of AISI 12L14 free-cutting Steel

“…As a result, there are many interacting parameters responsible for improving different machinability aspects including tool life, cutting force, surface finish and formation of small and stable built-up edges (Bouchelaghem et al, 2010;Kishawy et al, 2012;Peruchi et al, 2014;Song & Zuo, 2014;Xu et al, 2012). In all machining processes, including turning, controlling the workpiece-tool interaction associated with the workpiece material is extremely difficult, especially when combining these parameters to study their effects on machinability and/or surface characteristics and their impact on service life (Costa et al, 2016;Fnides et al, 2011). Due to this complexity of combinations and their impact, we need to find a model for optimizing these parameters to expect a better quality and productivity.…”

“…Generally, the quest is to optimize the cutting parameters, i.e. : feed rate, depth of cut and cutting speed, etc., in order to attain the minimum of surface roughness as an index of surface quality (Costa et al, 2016;Peruchi et al, 2014).The maximum of material removal rate and the minimum of cutting power as productivity and cost efficiency criterion (FNIDES et al, 2011;Kishawy et al, 2012;Kulekci et al, 2016). It is about performing a multi-objective optimization where the responses are correlated and may present conflicting results if each objective function is proceeded individually.…”

“…Despite its successful use for optimizing many nonlinear multi-objective problems, this method may lead to inappropriate results when the objectives are conflicting (Xu et al, 2012). To with draw this shortcoming, Costa et al (2016) applied successfully a hybrid method combining Taguchi's signal-to-noise ratio (SNR) and principal component analysis (PCA) methods with the normal boundary intersection (NBI), to optimize the mean roughness, the total roughness and the material removal rate of AISI 12L14 free machining steel in turning process (Das & Dennis, 1998;Dureja et al, 2010;Tebassi et al, 2016). They used response surface methodology RSM to generate the objective functions.…”

“…Amongst the few scientific works on AISI 12L14, dealing with the main subject of the present paper, we can cite the followings. Costa et al (2016) by the use of a novel RSM and SNR-PCA-NBI Taguchi based approach, was able to find a single solution giving the optimum values of Vc, f and ap in order to minimize roughness (Ra, Rt) and maximize material removal rate. Gomes et al (2012) proposed a new PCA based GCM approach, in the same context (Salgado Jr et al, 2012).…”

Multivariate mean square error for the multiobjective optimization of AISI 52100 hardened steel turning with wiper ceramic inserts tool: a comparative study

“…However, when there are correlated responses being analyzed, those methods can be inadequate. If the variancecovariance structure among the responses is not considered, the optimization may lead to unsatisfactory results [4,5]. The presence of such correlation, according to [6], can influence the optimization results, since it can create errors in the regression coefficients and unbalance the mathematical models.…”