Surrogate Model-Based Structure Optimization of Jetting System

Abstract

Jetting technology has become popular in microelectronic packaging due to its high efficiency and noncontact property. Its performance is significantly affected by the structure design. However, it is very difficult and time-consuming to conduct the optimal design under the traditional approaches, especially when more than one performance criteria is required. A surrogate model-based design method is proposed to optimize structure parameters under two different performance criteria. The surrogate model can serve for the quick evaluation of the process performance, upon which the process can be optimally designed via iterative search. As two performance indexes are required in the jetting process, a biobjective optimization method is used to design multiple parameters concurrently. Under this new design approach, the design can be easier to complete, and the jetting performance is significantly improved in both simulation study and experiment testing.