Learning Forum

[a.p.c.sedilot]

Hello,

 

I am trying to simulate a photodiode with InGaAsP quantum well , I setted up the mqw solver directly in the charge solver. When trying to run the simulation, the first part (mqw solver) goes well but when it comes to charge itself, it start to run and then after 5min I have got this error : "

job aborted:
[ranks] message
[0] process exited without calling finalize

device-engine-msmpi.exe ended prematurely and may have crashed. exit code 0xc0000005

"

I have tryed to run the simulation without the mqw solver and then this message doesn't show up anymore.

 

Could you help me understand what it means and how to fix it please ?

[kghaffari]

Hi,

Thank you for posting your question. The message unfortunately doesn't point to where the problem can be coming from. However, since the issue occurs in CHARGE/MQW coupled mode only, I have the following recommendations:

• Please go to CHARGE>Results tab and make sure that "electrostatics" is enabled. This option is required for correct coupling between CHARGE and MQW.
• Please make sure that no materials are defined as "alloy". Even alloys should be defined as "semiconductor" (when being created from the material database).
• If the quantum well structure is defined through the GUI, it is possible that switching to script could benefit the convergence behavior. An example of constructing the geometry through script is available in this article (specifically in QCSE_EAM_charge_mqw_coupled.ldev).
• In case the issue persists, please provide screenshots of overal geometry, material list, CHARGE Mesh tab, Results tab, Advanced tab, and MQW tab, in case I can provide further recommendations based on those.

Best regards, Khash