Slow simulation speed under sunlight (photo-emission flag = 1)

Hello all,

I have read topics on the forum previously opened about the issue, however, couldn’t get proper solution except reduction of capacitance (Csat). My current simulation is at around ~200s (i5 computer) although 3 days have passed. My circuit is not too complex, several nodes are connected to each other with resistors.

I did created a mesh with average quality of 0.74, did quite optimization with the GMSH tools and algorithms and it is the maximum value I could get for a telecommunication spacecraft model.

I created bi-maxwellian electron populations and a back-tracking vol distribution for protons, is there any recommendations for the problem to increase speed of the simulation under sunlight and photo-emission case? Nearly the simulation cases I made have absolute potentials with positive values under the sunlight, did not have any issue with eclipse case.

Best Regards
Arif Armagan Gozutok

I ended and re-started the simulation several times. I guess I miss something, is there any recommendation to increase the simulation speed under photo-emission case? :smile:

Hello Arif

A late reply might be better than no reply. There are many possible reasons, but what comes to mind is to decrease the photoElectronDensification variable a few order of magnitudes. To know if this would help, you can look at the simulation output (number of particles per population) and see if photoelectrons dominate. Spis is a bit too aggressive with wanting to simulate photoelectrons and secondary electrons.

Another thing that comes to mind is the quality of the mesh. An average quality is not always useful, it’s reality the worst-quality tetrahedrons that gives rise to issues. Maybe some optimization of the mesh can be done (outside of spis, and then imported as a .msh)

Best regards,

Dear Fredrik,

Hi, thank you for reply! I’m happy to hear it and I will investigate the parameter “photoElectronDensification” as well on further projects, maybe a fine tuning will do the trick. However, the overall area of spacecraft was quite big (around 260 square meters, the illuminated area was around 105 square meters, including +70 square meters of solar array).

Actually, the model I generate was too basic 3D satellite, consisting of boxes and plates, I use the GMSH as an optimizer and haven’t used any mesh refinement on any parts on surface elements during modelling. I will consider the model’s worst mesh/tetrahedron quality and will improve apart from the SPIS.

Best regards