Algorithm for Estimating Bonding Process Variation by Applying Statistical Analysis

Support System for Prediction of Solid State Diffusion Bonding (Part 3)

Abstract

An algorithm for estimating the bonding process variation is developed to establish the selection system for the optimal solid state bonding conditions. The bonding process variation (dispersion) due to surface roughness is discussed by applying the statistical analysis. The distributions of void spacing and void volume are estimated by the overlap method which was proposed in our previous paper (part 1). The standard deviations are calculated from the distributions. The numerical calculations are performed by taking into account the standard deviations. Also, the hazard rate is newly introduced to comprehend the bonding dispersion. Experimental bonding tests suggest that the algorithm developed in the present study can predict the bonding process variation.