Request for Lagrangian particle trajectory analysis in PeleC

[ShuyingJiang-bachelor]

Hello everyone, I want to add massless tracer particles to my PeleC simulation for particle trajectory analysis. I would like to know if PeleC can achieve such functionality. Below are the detailed specifications of this analysis:

1. The particles should be Lagrangian point-mass particles with zero mass, meaning their velocity always equals the local fluid velocity at their position;
2. In my understanding, the initial positions of particles should be specified. During the simulation, for each timestep t, Newton’s second law should be used to calculate the particle’s next position at t+dt based on its velocity, followed by iterative updates. Yes, this should not be limited to post-processing, as my flow field is unsteady;
3. Under this treatment, the exported particle positions represent their trajectories, and the physical quantities (e.g., temperature, pressure, heat release rate, species mass fractions) at each position correspond to the trajectory’s properties;
4. If possible, I hope to track multiple particles simultaneously for characterization.

Can PeleC implement such functionality? (I chose PeleC because my case involves supersonic flow.)
Thank you very much for your help!

[baperry2]

PeleC does not have built in support massless tracer particles, although it does support particle evolution for liquid spray droplets through AMReX's particle capability. One could add tracer particles based on that capability with a bit of work/coding. In the past, we've examined tracer particles for unsteady flows through post-processing by dumping lots of plt files with short time intervals, but I agree that is not an ideal solution.

[ShuyingJiang-bachelor]

Thank you very much. As you mentioned, the solution approach should start from the existing mature liquid particle module. My thoughts are:

1. Introduce liquid particles into my simulation case but disable their associated processes (e.g., evaporation, reactions);
2. Some additional work may still be needed to adjust the mass properties;
3. Subsequently, other analyses can leverage the liquid particle framework;
4. The method you proposed—"examined tracer particles for unsteady flows through post-processing by dumping lots of plt files with short time intervals"—is what I plan to implement, though it is not a perfect solution.

Once again, I truly appreciate your comprehensive response. I still have one question:
a. Does PeleC have a solid Lagrangian particle module? In my understanding, solid particles might be more suitable as tracers compared to liquid particles, and from the perspective of modification effort, solid particles seem more appropriate.

[baperry2]

I'm not sure whether it would be easier to hack into the liquid spray capability or create a separate tracer capability using AMReX particles with that as an example.