Multi-Response Evolutionary Operations With Taguchi Parameter Designs For Automatic Electrostatic Painting System

Evolutionary Operations (EVOP) or Response Surface Methodology (RSM) is a method for finding the optimal conditions of industrial processes. In general, this approach will search for the proper conditions under a consideration of a uni-response system. This work determines the efficiency of a proposed sequential algorithm in the context of RSM for automatic optimisation of metallic painting parameters for aluminium alloy wheels via Taguchi parameter design. The study categories the area on wheel into four zones and there are then four responses. The metallic painting system in this work uses an automatic electrostatic painting (AEP) spray gun and there are seven parameters (controllable predictor variables) that effect the metallic paint thickness including colour shade. To improve metallic painting parameters via Taguchi design we do study the interaction of each parameter to create linear graph of interaction for all parameters. This brings an orthogonal array of the specific experiments. By referred past records we can design lower and upper levels of each parameter. The purpose of experimental designs is to find the relationship among parameters to paint thickness and colour shade on the wheel. All experimental data will be used to fit a multiple regression model for paint thickness in each zone. A sequence of runs is then carried out by moving in the direction of steepest ascent to approach the better responses. Optimisation strategy in this work will be applied on each significant path in each zone. The results suggest that the proposed levels pf predictor variables from the method of steepest ascent seems to be more efficient on the AEP surface when compared with the same preset levels. The EVOP based on Taguchi design works well on both the average and the standard deviation of the thickness when the responses are determined in areas zoned Z1, Z2 and Z4. Although the average and the standard deviation of the responses in zone Z3 after the EVOP are not satisfied, it is at the same level, on average, when compared. Under a consideration of customer’s requirement on paint thickness and colour shade of the wheel, experimental analyses on metallic painting parameters also bring the reduction of metallic paint consumption in each wheel during the manufacturing process.

Keywords: evolutionary operations, response surface, Taguchi parameter design, multiple regression, steepest ascent

Corresponding author: E-mail: lpongch@engr.tu.ac.th