An expert system for setting parameters in machining processes
Introduction
Setting machining parameters is a complex task because a validated model of the process is required, then a kind of a search procedure is used on the model to establish a proper set of input parameters. There are several forms to generate models; however, mathematical models are not black boxes. Neural networks could be a very precise model; however, they are black boxes where a explicit formulation about the correlation between variables and the effects on output response is not evident.
Linear regression could provide a good decision criteria about the impact of the input machining parameters on output response; however, in some cases, could generate poor models for predictions. Setting parameters of a process is equivalent to optimize a function, we establish a criteria about the function to maximize its response or minimize a cost derived from the function, per example; then a optimization tools can be applied.
Symbolic regression with genetic programming (GP) are commonly used for this purpose; however, there are several drawbacks like complexity to manipulate tree data structures, tendency to generate very large tree structures and consequently high CPU time consuming; otherwise, GP could generate compact formulas eliminating some variables that could be considered as useless.
In this paper the expert system proposed works with turning and milling processes. With little adaptations can be used in another processes inclusive not machining ones. Some machining systems are a challenge because there are a variety of parameters and materials that define the performance of the process. Turning is one of the most used process in machining and setting parameters in this process is made by experience, following tables from builders, or by trial and error. More advanced approaches have been used like experimental design and robust experiments (Motorcu, 2010), response surface methodology (Bhushan, 2013, Chauhan and Dass, 2012, Villeta et al., 2012), analysis of variance or ANOVA (Aouici, Yallese, & Fnides, 2011), grey relational theory or grey analysis (Ranganathan & Senthilvelan, 2011) and Taguchi methods (Gaitonde et al., 2009, Hanafi et al., 2012, Maniraj et al., 2012). Other approaches requires a model to be used with optimization tools; regression models (Liang, Ye, & Zhang, 2012), neural networks (Senthilkumaar, Selvarani, & Arunachalam, 2012) are commonly used.
Using these models and by genetic algorithms (Jangra, Jain, & Grover, 2010) and particle swarm optimization (Raja & Baskar, 2011) the machining parameters of turning processes can be set. Milling is another machining process more used too like turning. Design of experiments, Taguchi methods (Ji et al., 2012, Kadirgama et al., 2012), analysis of variance (Gopalsamy et al., 2009, Mustafa, 2011, Yang et al., 2009) and response surface methodology (Mangaraj & Singh, 2011) are the most used approaches for setting machining parameters. Optimization approaches like particle swarm optimization (Raja & Baskar, 2012), Kriging interpolation search techniques (Lebaal, Schlegel, & Folea, 2012) have been used for setting the parameters of milling processes too. Our proposal is based on symbolic regression, However, this approach using genetic programming has been few used (Raja & Baskar, 2010).
Considering setting of parameters as an optimization problem, different criteria has been considered. Usually surface roughness is the principal response to determine the efficiency of the machining process turning or milling, however consumption time, cutting temperature, tool wear are example of responses that could be considered to improve the machining process. In this paper, improve the surface roughness will be the criteria for setting the parameters.
This paper is divided in five sections; this section is the introduction, Section 2 is a description of an hybrid system used for setting parameters. Section 3 is a description of the application where a comparison with linear regression and genetic programming is made. Section 4 is the application of the proceed hybrid system for setting parameters on the machining process given in the last section. Finally, Section 5 includes conclusions and future work.
Section snippets
An hybrid system for modelling and optimization of machining processes
Setting of parameters is important in machining process because could generate an improvement on quality on machined pieces (this reflected mainly by surface roughness), material removal rate (MRR) reduction of time processing, tool wear ratio (TWR) and cutting temperature, and others which importance depends of the process (Fig. 1). Tables of machine fabricants or information given by the provider of the material to be machined gives some clues about the most suitable set of parameter that
A case of application for setting parameters
As an illustration of the proposal, a setting of parameters of two processes will be made considering two different materials, aluminium and steel. An experimental design is made considering one type of aluminium and two types of steel. First, the machining of the pieces was made using a turning machine model Okuma LB15 (Fig. 3). A tool type DNMG 432 PG of grade RC8025 was used for steel and a tool type DNMG 432 GP of grade CQ23 was used for aluminium. Three kinds of materials were machined
Expert system procedure and results
A mathematical model using symbolic regression are generated considering the same conditions mentioned above, the same size of the population, 50 individuals are selected during 200 generations; ten runs are made per run. Considering the solutions shown in Table 10, Table 11, Table 12 for milling in every run, a criteria of low complexity, low error, high and low PRESS can be taken here. The same criteria is applied for Table 13, Table 14, Table 15 for turning.
For milling, three models
Conclusion
An expert system is proposed to assist machining processes users in order to generate a set of machining parameters that improves the process considering the minimization of surface roughness; however, other criteria could be considered and that incorporation is easy. The expert systems requires of experimental data, however, historical records could be considered too constraining the space of search but under working conditions. A set of models are generated from experimental data. A model is
References (25)
Optimization of cutting parameters for minimizing power consumption and maximizing tool life during machining of Al alloy SiC particle composites
Journal of Cleaner Production
(2013)- et al.
Optimization of cutting conditions for sustainable machining of PEEK-CF30 using TiN tools
Journal of Cleaner Production
(2012) - et al.
Optimization of dry machining parameters for high-purity graphite in end milling process via design of experiments methods
Journal of Materials Processing Technology
(2009) - et al.
Modeling and optimization of hard turning of X38CrMoV5-1 steel with CBN tool: machining parameters effects on flank wear and surface roughness
Journal of Mechanical Science and Technology
(2011) - et al.
Optimization of machining parameters in turning of titanium (grade-5) alloy using response surface methodology
Materials and Manufacturing Processes
(2012) Montgomery introduction to linear regression analysis
(2007)- et al.
Multiperformance optimization in turning of free-machining steel using Taguchi method and utility concept
Journal of Materials Engineering and Performance
(2009) - et al.
Taguchi method and ANOVA: an approach for process parameters optimization of hard machining while machining hardened steel
Journal of Scientific & Industrial Research
(2009) - et al.
Optimization of multiple-machining characteristics in wire electrical discharge machining of punching die using Grey relational analysis
Journal of Scientific & Industrial Research
(2010) - et al.
Machining performance optimization in end ED milling and mechanical grinding compound process
Materials and Manufacturing Processes
(2012)