Silicon based heater in DEVICE Heat Solver


I am trying to run the heat solver on my heater that is structured based on silicon, oxide, and metal.

Could you please look at the errors and let me know what the issue is?

Thanks, HaniHeater-version2.ldev (5.8 MB)

Hi Hani,

My apologies for the tardy reply. I have taken a look at your file and the problem with the current setup is that there are no thermal boundary conditions. The simulation needs at least one reference (temperature) point in order to find a unique solution. This reference can come from a fixed temperature boundary condition or from a convective or radiative boundary condition.

Another point that I wanted to mention is that unlike the CHARGE solver where usually the simulation volume is quite small since we need to only include the active region, in HEAT simulations you usually need to include a large simulation volume to properly account for the entire volume over which the heat will spread. In your setup the simulation volume is very small, however the heat from from the heater will most likely spread beyond this volume and you will need to include that in your simulation.

For example, in this KB page ( you can see that the simulation does not include the heater and the waveguide only rather it includes a large portion of the substrate and the cladding so that the entire volume over which the heat spreads is simulated.

Hi Aalam,

Thanks for your response. I have kept choosing a boundary condition that would fit with my heater but it is not working. I am trying to heat the silver on the top and see the change of phase or rather the refractive index change as a function of temperature.
Could you please help me achieve that?Heater-version2.ldev (5.8 MB)

I believe the problem in your setup is coming from the fact that you are only simulation the heater itself. When you are performing a thermal simulation, in most cases you have to consider a large volume surrounding the heating element to account for the spreading of the heat in all direction. For example, in the simple thermally tuned waveguide example in KB ( you will notice that the simulation volume not only contains the heater but also the oxide cladding, silicon substrate, and the air on top. In your file also you have to look at the bigger picture and include the regions surrounding the heater. This will enable you to apply appropriate thermal boundary conditions. For example, if you make the simulation volume large enough to include the air surrounding your device then that will allow you to use a convective boundary condition. Also, if you can include the substrate if your simulation then you can (in most cases) assume that the bottom of the substrate is at room temperature. This will allow you to apply a fixed temperature boundary condition.