Inconsistent potential values with batch simulations

SPIS
1

I am trying to run a batch of simulations for a SC with varying sun directions on SPIS 6.2.6. On its surface, the SC has a particle detector with a large face biased at -8V. The main global parameters changes I made are as follows:

-1 electron and ion population, both with densities and temps at 10 cm-3 and 10eV

-Photoemission turned on (3)

-Varying Sun direction (sunX, sunY)

I ran one simulation through the SPIS GUI to test and make a control. Then I used this simulation to take the circuit, mesh, and groups files to create a folder for my batch runs to take from, as per the documentation and from premade scripts. This script was launched via terminal.

The batch simulations ran well, but when I checked the final SC potential, it is way off from the potential I had from the control run (comparing to the batch run with the same sun direction). The batch run gave me a potential of ~6V, and the control run had ~10V.

The batch runs are using the same exact files from the control run, so I am not sure why they are giving me different results. Could someone help explain why launching a simulation from the GUI gives different results from launching from the terminal? I can provide more details if necessary.

2

Hi Joseph, not sure what’s up. Identical simulations giving inconsistent result are either unstable or not actually identical, so perhaps you came across a bug with running spis -b and feel free to report that.

In the meantime, I just want to say that I usually set up a batch of simulations by using spis -p. This requires a full ready simulation project, and doesn’t allow for changes in e.g. the mesh. But if you clone that project and make batch changes in circuit.txt or global parameters, there are no issues.