Coupled GA-SA Optimization Algorithm and Application for Compressive Curved Damper Springs of Automotive Lockup Clutches

Abstract

An efficient global optimization algorithm has been developed by combining GA (genetic algorithm) and SA (simulated annealing). It is observed that the coupled optimization algorithm can converge faster than each single algorithm. It is applied for the design optimization of damper coil springs that absorb the torsional impact and vibration imposed on the clutch of an automatic transmission. Additionally, an user friendly computer program is developed in order to design various types of torque converter damper springs automatically. Dynamic characteristics and torque capacity of damper springs varies according to a spring constant that depends on a wire diameter, a coil diameter, the number of turns, a coil length, etc. The torque variation according to the spring constant characteristics of a curved compressive coil spring is calculated and compared with that of a straight one using the coupled simulated annealing and genetic algorithm.