Error while importing temperature profile in charge solver ("Poisson solver encountered an unexpected numeric error")

Hello ,
thermal_tuning_waveguide.ldev (8.3 MB)T3.mat (22.6 KB) :kindly help me resolve the issue.

Dear @s.de,

As the error message is indicating, the issue here is due to there being conductor materials in your simulation that are not assigned as contacts. CHARGE simulates the transport in semiconductor materials, and conductors can only be included if used as contacts. Here, the geometries “d1” and “wg2” (which don’t have electrical boundary condition assigned to them) use the material named “Doped Si” which in your file is defined as a conductor.

Kind regards

Dear @kghaffari,

Thank you so much for your reply. I no longer have the above problem but I have enountered a new one.

PN_phase_shifter.ldev (4.8 MB) T7.mat (29.1 KB)

The charge and Temperature.mat files are given above.When I run the charge solver without importing the temperature it works well.But when I import the temperature it shows the above error.
Can you kindly help me resolve the issue.

Kind regards,

Dear @s.de,

Please note that when using the heat source object, the imported data needs to be in W/m3 (here it seems you are attempting to input temperature). Also, if the CHARGE solver is used when using this object (which is the case in your file) you would need to set its temperature dependence option to coupled (Please refer to this article for more information about this object):

In case the issue persists, it would be helpful if you could also share the file you use to generate the imported heat source so we can further investigate this.

Kind regards

dear @kghaffari,

Thank you for your reply.I have imported the temerature profile to charge by using coupled temperature dependence as you have showed.

But it is still taking the temperature as 300K in the charge solver. How to import the temperature profile to charge successfully?

Kind regards,

Dear @s.de,

I now realize that you have been using the import temperature object (and not the import heat object). Therefore, your initial settings of using a non-isothermal mode and importing temperature data into the object is correct. Sorry for the mistake.

I suspect that the issue is related to the generated temperature distribution (T7.mat). If you could please share the .ldev simulation you use to generate the data, I would be glad to look into it.

Kind regards

T7.mat (28.4 KB) thermal wg.ldev (9.4 MB) charge wg.ldev (3.3 MB)

Dear @kghaffari,

Thank you so much for your response. Here I have attached the thermal and charge simulations.I took the settings in the non-isothermal mode . Now, I have a different error that has been attached here. Kindly take a look into it.

Regards,

Dear @s.de,

I was able to resolve the issue by expanding the monitor that is used to export the temperature (“wg2_1”). This seems to better encompass the structure in CHARGE, making obtaining convergence easier.

Please find the necessary files attached (simulations are saved in layout mode). Looking into the HEAT simulation you would notice that I have also increased the mesh refinement to improve accuracy (the lock mesh option needs to be unchecked for the change to take effect).

thermal wg_modified.ldev (8.5 MB)
T7_modified.mat (102.6 KB)
PN_phase_shifter_modified.ldev (5.0 MB)

Please note that in your updated CHARGE simulation (“charge wg .ldev”), the mesh order set for the objects (specifically “anode”, “cathode”, and “rib”), leads to the silicon region being overridden by the contacts. Please refer to this article for more information about mesh order. Also in the HEAT simulation, the second material needed for your convection boundary condition seems to be missing.

Finally, I would recommend looking into our CHARGE and HEAT courses which can be useful in getting you started with these products.

Kind regards

Dear @kghaffari,

Thank you so much for correcting and modifying the files.The simulations are working now :).I will get back to you if I have more queries.

Kind regards,