Non-contact dispensing in the electronics packaging industry is becoming the most widespread dispensing method. The crucial research direction is to achieve accurate flow control for non-contact dispensing technology. However, the whole dispensing process contains fluid-solid coupling and its nonlinear and unstable flow, which makes it difficult to obtain an accurate prediction model for the quality of adhesive points. In this paper, based on the existing theory and the idea of minimum error backpropagation, we compare the experimental data with the theoretical value of adhesive point by subsequent experiments. Finally, a simple and effective adhesive point quality prediction model for non-Newtonian fluids in the dispensing process is obtained. The Taguchi method was used to prove that the maximum error between the predicted adhesive point quality and the experimental value is <10%, which is highly accurate. The Minitab software was used to analyze Taguchi's experimental data, and the sensitivity of dispensing quality to process parameters was ranked as follows: driving voltage > adhesive supply pressure > nozzle heating temperature; this study can be of guidance to the design of dispensing valve structure and dispensing process parameters, and to achieve high speed and high precision flow control.
Read full abstract