Predictive model of the solder paste stencil printing process by response surface methodology
Chun-Sheng Chen,Hai Wang,Yung-Chin Kao,Po-Jen Lu,Wei-Ren Chen
DOI: https://doi.org/10.1108/ssmt-08-2021-0056
2022-02-25
Abstract:Purpose This paper aims to establish the predictive equations of height, area and volume of printed solder paste during solder paste stencil printing (SPSP) process in surface mount technology (SMT) to better understand the effect of process parameters on the printing quality. Design/methodology/approach An experiment plan is proposed based on the response surface method (RSM). Experiments with 30 different combinations of process parameters are performed using a solder paste printer. After printing, the volume, area and height of the printed SAC105 solder paste are measured by a solder paste inspection machine. Using RSM, the predictive equations associated with the printing parameters and the printing quality of the solder paste are formed. Findings The optimal printing parameters are 175.08 N printing pressure, 250 mm/s printing speed, 0.1 mm snap-off height and 15.7 mm/s stencil snap-off speed if the target height of solder paste is 100 μm. As the target printing area of solder paste is 1.1 mm × 1.3 mm, the optimized values of the printing parameters are 140.29 N, 100.52 mm/s, 0.63 mm and 20.25 mm/s. When both the target printing height and area are optimized together, the optimal values for the four parameters are 86.67 N, 225.76 mm/s, 0.15 mm and 1.82 mm/s. Originality/value A simple RSM-based experimental method is proposed to formulate the predictive polynomial equations for height, area and volume of printed solder paste in terms of important SPSP parameters. The predictive equation model can be applied to the actual SPSP process, allowing engineers to quickly predict the best printing parameters during parameter setting to improve production efficiency and quality.
materials science, multidisciplinary,engineering, electrical & electronic,metallurgy & metallurgical engineering, manufacturing