Time Varying Source

[skilduff]

Hi all,

I've been trying to implement a time and spatially varying heat source to the code, but it is not behaving as expected. I've made similar changes to my code as this post: #1904

The main difference in my code is that my power varies with time. I obtain the time from these lines of code in flow_sources.cpp:
const auto dt = config->GetDelta_UnstTimeND();
const su2double t = config->GetTimeIter() * dt

I then calculate the source's volumetric heat flux for the current time and add it to the energy residual:
residual[nDim+1] = -Volume * Source

I have compared the values calculated in my code's source term (in W/m^3) and they match the analytical model I am comparing it against. However, when I run the simulation, the temperature increase does not match the temperature increase from the analytical model (it is off by several orders of magnitude). I think the issue may be arising from calculating and adding to the residual, but I am not so sure.

Any help would be appreciated.

Thanks

[NAnand-TUD]

Hello, If you can point me to the branch where you have implemented this, it would be easier to give you some feedback.

[pcarruscag]

That's correct, have you printed some values from the code to see if they match expectations?
Is your simulation non-dimensional? Perhaps you are missing some dimensional factors.

[NAnand-TUD]

Right, got it. Disregard my comment @skilduff .

[skilduff]

My simulation is dimensional. I have printed Spe (the source term in W/m^3) and compared it to my analytical model; the results are the same. However, my analytical model also predicts an increase in temperature of ~250 K, whereas my simulation shows the temperature does not increase even 1 K.

[pcarruscag]

Have you done some simple verification? For example an adiabatic channel flow with your source term and verify that enthalpy out = enthalpy in + integrated source

[skilduff]

Yes, I am trying to validate the model by simulating an essentially 0 velocity flow in a 1m x 1m square.

I am using this equation:

$$\frac{d}{dt}\rho e = S$$

I am also assuming constant density and a calorically perfect gas: $$e = C_vT$$