Developing a Rework Process for Underfilled Electronics Components via Integration of TRIZ and Cluster Analysis

This paper integrates the theory of inventive problem solving and cluster analysis to develop effective rework processes for underfilled electronic components. First, the function analysis is used to clarify the functional disadvantages such as underfill material attaching to the component, the printed circuit board (PCB), and the solder joints during the rework process. The cause-effect contradiction chain analysis then helps to identify engineering contradictions. Second, this research proposes clustered contradiction matrix. The optimal number of clusters for the inventive principles is determined by Ward's method of hierarchical approach. The K-means method is then used to allocate the engineering parameters to the appropriate cluster to achieve superior clustering performance. The priority of inventive principles considered as solution triggers is based on their corresponding frequency of appearances. Three solutions focused laser beam to remove electronics component, dispensable component bracket, and interchangeable PCB are proposed to rework the underfilled electronics components. Finally, the idealities of solutions are assessed and the most ideal solution interchangeable PCB is nominated for the U.S. patent.