THE OPTIMAL DESIGN OF ELECTRIC CABLE CONDUCTOR AND INSULATOR RESISTANCE USING TAGUCHI MULTIRESPON METHOD
A Parkhan(1), H A Kwintanada(2), I D Widodo(3)
1 Industrial Engineering Department, Universitas Islam Indonesia, Yogyakarta, indonesia
2 Industrial Engineering Department, Universitas Islam Indonesia, Yogyakarta, Indonesia,
3 Industrial Engineering Department, Universitas Islam Indonesia, Yogyakarta, Indonesia
Abstract
The quality of the electrical cable can be determined by resistance of insulator and conductor. The higher the resistance of the insulator and the lower the resistance of the conductor, the higher the quality of the cable. NYM 2x1.5 mm2 cable produced by Corporate X has an isolator resistance of 84.25 Ω and conductor resistance of 11.976 Ω. Even though the products meets Indonesia National Standard (SNI), but to win the competition, the company needs to improve the quality of their products with competitive prices. There are some factors influencing the resistance of insulators and conductor. They are core extruder speed (A), non-core extruder speed (B), cooling water temperature (C), PVC melting temperature (C), cabling speed (D), copper elongation (E), PVC density (F) and human error (F). In this research, design of experiment, based on Taguchi method, is used to improve cable resistance. Because of having two responses, the optimization process uses the TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) method. The result showed that the combination of levels of these factors could produce average insulator resistance of 95.70 Ω (increasing by 8.90 Ω), and the average resistance of the conductor 11.908 Ω (decreases down by 0.057 Ω). This optimal condition was obtained by combining the levels of factor A1 B1 C1 D2 E2 F1 G1, i.e. the system works at 140 rpm cabling speed, non core extruder speed 2.9 m/s, core extruder speed 2.9 m/s, cooling water temperature 20 0C, copper elongation 150 - 250, temperature for melting PVC 160 0C-165 0C and PVC density 154,95-155 kg/m3. Sensitivity analysis to accommodate the dynamic weighting of product quality attributes indicated that the changes in the order of priority attribute weights on a formal scale system used to convert linguistic forms into fuzzy numbers were not sensitive to optimal factor level combinations
Keywords: Factor Level Combination, Insulator and Conductor Resistance, Sensitivity Analysis, Taguchi, TOPSIS
Topic: Industrial Engineering