How to Avoid Ghost Particle Error When Using Hertz Model in MFIX-DEM

orange · January 11, 2025, 11:18am

Hello everyone,

I’m simulating a 50μm silica particle impacting a layer of 0.5μm polystyrene particles at 20 m/s in MFIX using the Hertz model. However, I keep encountering this error message:

Error: There exists a NORMAL particle in a GHOST cell. Particle may have gone OUT-OF-BOUNDS.

I’ve tried reducing the NEIGHBOR_SEARCH_N to 1 as recommended, but the issue remains. I want to keep the original physical properties of the particles, because I have do the same experiment. Any advice on how to fix this without altering the properties? The case file has been attached. It will report an error within 1 minute when restarted.

bug.zip (34.5 MB)

Thank you for any insights or suggestions!

ebreard6 · January 11, 2025, 6:51pm

this is an odd one. I tried with mfix-24.4. Particles disappear (keeps running though). I don’t have the same issue with the Spring Dashpot…

orange · January 16, 2025, 3:05am

I apologize for omitting earlier that I’m using MFiX-24.3.1 and 23.4.1. Both versions will report errors, as illustrated in the attached figure. I opted for the Hertz model over the spring damping model in order to utilize authentic physical properties.