Inquiry Regarding Differences Between DSMC and background_mcc in Elastic Collision Handling and Particle Merging Algorithm in WarpX

[zhongerting]

Dear WarpX Team,

Hello! I am currently working on PIC-DSMC coupled simulations in WarpX. I have added elastic collisions between electrons and ground-state atoms to my simulation case, implementing them using both the DSMC and MCC methods, respectively. During the simulations, I observed the following:

In simulations using background_mcc, the number of electron macro-particles and the total number of microscopic particles first increase and then decrease. However, the maximum number of macro-particles reaches about 10^6 (over a 300 ns simulation).

In simulations using DSMC, the number of microscopic electron particles consistently decreases. Within a very short time (3 ns), the number of electron macro-particles increases to 10^7, and the number of atomic macro-particles also rises to 10^7. This ultimately leads to excessively long computation times per step.

In both cases, I considered only elastic collisions, 9 excitation processes, and ionization of ground-state atoms, totaling 11 types of interactions. The cross-sections used in both simulations are identical. For excitation in DSMC, the two_product_reactionprocess was applied.

I would like to ask:

What are the differences between DSMC and background_mccin handling elastic collisions, which could lead to such disparities in electron density and macro-particle numbers?

Additionally, how does the particle merging algorithm work in WarpX? Is it possible to configure it to merge particles with similar velocity directions and magnitudes?

Thank you for your time and support!

[zhongerting]

input_excerpt.txt