Numerical Modeling of Wave Soldering in PCB

Abstract

Manufacturing of electronic boards (commonly referred as PCB) is a highly automated process that requires an accurate control of the various processing variables. Amongst the soldering processes, wave soldering is one of the most often used. In this, the various electronic components are provisionally inserted onto the PCB, and a low velocity jet of melted solder is directed to the moving board. Due to capillarity effects the solder adheres to the component/board interface and the process is completed. This methodology is most often used for small components. The adjusting of the operating parameters (solder nozzle orientation and velocity) is often carried out on a trial and error basis resulting in a time consuming process that is at odds with the increasing demand for smaller production series that the electronics industry is faced with. In addition the number of defects (mostly from missing components that are washed away by the impacting jet) is more likely to occur when thinner substrates are used in the PCB manufacturing. The present paper describes the application of a Computational Fluid Dynamics model to describe the interaction of the solder jet with the PCB and the integrated circuits. The model includes the conservation equations for mass, momentum and energy in a transient time frame. The jet and surrounding ambient atmosphere are modeled as two separate fluids and the interface is tracked by a VOF model. By adjusting the computational mesh refinement the interface is captured with accuracy. The drag forces occurring in the various components are computed from the pressure data field. The model allows the optimization of the wave operating parameters as a function of the component type of and its layout in the PCB.

Experimental measurements of the shear force on surface mount components simulating the wave soldering process

“…The 0603 component was selected as, because of its small size, it statistically represented one of the most prone to losses in the manufacturing line. For this purpose, a computational simulation was carried out, based on a numerical model developed previously by the authors (Arcipreste et al, 2014).…”