Assessing tolerable viscosity differences in polymer-to-polymer welding

The welding of dissimilar polymers is becoming increasingly common. There are a couple key drivers to this trend. First, many new plastics applications must be used in environments where adhesives cannot be easily applied, such as for medical products where finding adhesives that pass FDA regulations is nearly impossible. Another key area where dissimilar polymers are joined is for highly engineered applications in which a material is needed for its very specific — and expensive — engineered properties. Often, the expensive material will need to be joined to a lower-cost material to provide some nearby structure that does not require the same specialized properties. A third driver is the need to save space for small or very compact assemblies. Even a thin layer of adhesive can be a significant barrier to performance for some products, such as small electronics. Lastly, an assembly may require the joining of components that must be manufactured via different processes, such as joining an injection-molded frame to an extruded bag. Due to the different plastics production processes used, the two parts can have very different thermal properties even if they are the same type of polymer. One of the most critical properties affecting dissimilar polymer joining is viscosity. If the polymers’ viscosity is not similar enough, intermolecular diffusion cannot occur. The goal of this work is to set clear boundaries on the tolerable limits of the viscosity mismatch that can be accommodated and still result in intermolecular diffusion. Polymers of the same grade of material were heated conductively using a hot plate to different temperatures, then the melt layers were pressed together in the standard hot plate welding approach. The parts were then tensile tested and cross-sectioned to determine relative strength and whether intermolecular diffusion occurred. By using the same grade of material heated to different temperatures, the independent variables are minimized, and the effect of viscosity can be clearly identified. The test data was then analyzed for statistically significant shifts to identify the maximum allowable difference in viscosity.