In wire bonding technology the use of varying substrates and surface types always requires the determination of optimized bond parameters to achieve high bond quality. Underproduction conditions bondability checks have to be performed before the new material is given to the line. Each bondability check consists of many trials with e. g. increasing US-energy or bond force. Then pull tests have to be carried out and the best result has to be found. Usually these efforts take a long time and give no guarantee to find the optimized parameters for any condition. A new approach is possible using the Enhanced Evolutionary Strategy. Individuals are represented by the different randomly generated combinations of machine parameters (mutation). The best individual survives (selection) and the following generation of individuals is derived from these values. The need of different generations to determine the best is time-consuming. The evolution can be enhanced using a regression analysis. The analysis computes the maximum (or minimum) of the so called process response surface.
展开▼
