Six Sigma Robust Design of Composite Hand for LCD Glass Transfer Robot

Abstract

This research studied robust design of composite hand for LTR (LCD glass Transfer Robot). 1st DOE (Design of Experiment) was conducted to find out vital few Xs. 108 experiments were performed and their results were statistically analyzed. Pareto chart analysis shows that the geometric parameters (height and width of composite beam) are more important than material parameters (E1, E2) or stacking sequence angle. Also, the stacking sequence of mid-layer is more important than that of outer-layer. The main effect plots show that the maximum deflection of LTR hand is minimized with increasing height, width of beam and layer thickness. 2nd DOE was conducted to obtain RSM (Response Surface Method) equation. 25 experiments were conducted. The CCD (Central Composite Design) technique with four factors was used. The coefficient of determination (R2) for the calculated RSM equation was 0.989. Optimum design was conducted using the RSM equation. Multi-island genetic algorithm was used to optimize. Optimum values for beam height, beam width, layer thickness and beam length were 24.9mm, 186.6mm, 0.15mm and 2402.4mm, respectively. An approximate value of 0.77mm in deflection was expected to be maximized under the optimum conditions. Six sigma robust design was conducted to find out guideline for control range of design parameter. To acquire six sigma level reliability, the standard deviation of design parameter should be controlled within 2% of average design value