Particulate Surface Adhesion Sandbox
One of the projects I am currently working on involves modeling the gas-surface or plasma-surface interface. This has dual applications. First, in the world of contamination control, it is not completely clear how to model dust particulate attachment to surfaces. Most codes, including my CTSP utilize a fit to experimental measurements of Klavins and Lee (1987) to calculate what fraction of particulates of a certain size detach given certain acceleration. However, it is not really clear how to integrate this model with non-static environments, such as the vibrations or aerodynamic flow encountered in spacecraft fairings. It is also not clear how to model particles bouncing off surfaces. Many codes, again including CTSP, utilize a coefficient of restitution to assign how much of the incident velocity is retained on rebound. There is no good experimental data for this coefficient, and hence my analyses are typically for a range of values to bracket the predictions. Not very first-principle based!
Secondly, in the world of plasma plume interactions (or plasma processing), researchers are often interested in predicting surface erosion due to sputtering. Here we can use models such as that of Yamamura, however, this model does not allow us to model the dynamic evolution of surface morphology. Essentially, we assume that the surface is flat, and using the model, compute the total mass loss. But sputtering yield is a function of the incident angle, and hence it is crucial to resolve the “cratering” that may occur. This effect can be modeled using Molecular Dynamics (MD) but only at microscopic scales not practical to engineering simulations. My idea here is thus to combine concepts from methods such as Discrete Element Method (DEM) with PIC to model the surface layer dynamically.
The video below shows what happens when the surface adhesion and distance is increased and horizontal flow is applied.
To Do / Bugs
- Check for back-side impact – all particles now bounce in the positive normal direction
- Add interparticle collisions
- Attach particles at the touch position, not circle centroid
Let me know what you think and if there are any particular effects you would like to see included.
Free molecular flow in a cylindrical pipe – with multithreading
HTML5 for Scientific Computing