Help with running a CHT on a big Radiator model

I am trying to achieve a CHT on a radiator built and operated in the 60s of a nuclear reactor program in the US. Molten salt through the tubes and air across the tubes, you can find the model here:

And you can find the Simscale project here:

I set everything up in Simscale, but I can only run the simulation for maximum 3e4 seconds, and I need way more to achieve convergence, probably around 2e5 (around 72 hours real time). The mesh is quite large, currently it sits at 26 million cells.

Can anyone help me with running this model? There is also another big model I need to run, but one thing at a time :slight_smile:

/Malcolm Akner

Hi Malcolm, this is an interesting project!

To begin with, I want to clarify that the ‘Maximum runtime’ limitation is real time seconds , and it is independent from the simulation’s timesteps. It represents the amount of seconds you are allowing the case to run for, and after these pass, it automatically stops. The limitation is indeed approximately 8 hours for community plan users.

My initial question is, why do you expect that it will take 72 real hours for the simulation to reach convergence? Do you have a specific goal?

Best regards,
Fillia

Hi! :slight_smile:
Thank you, I have worked hard on the model and to understand the theory behind the turbulence models used!
Previous simulations done on a heat exchanger (CHT) of similar mesh-size took well over 24h to simulate. This radiator includes air which has a density variation and I want to make sure that my residuals are low. Based on the completed 8 hour simulation (timestep = 1, endtime = 400) the flow barely makes it a 1/3 of the pipe-lengths, and I want to reach a steady state - this is why I want a much longer sim-time, but 72h might be pushing it. Still, I need more than I currently have access to.

Who could I contact to run this case for me? I’ve been told that this is something that happens routinely on Simscale but so far I’ve had no luck.

/Malcolm Akner