I am trying to simulate a conjugate heat flow through various materials. The simulation i am looking at is the 30mm VIP -83.3C inside 23.3C outside.
I want to replicate a freezer, so i have set one initial boundary subdomain at -83.3C to replicate the inside of freezer.
I have created a wall around each material and the subdomain that is adiabatic.
I then created a boundary condition with a Natural convection of 23.3C on the ‘outside’ to replicate the room temperature.
Is this the correct way to set up the simulation to see/check the heat transfer through each material and also to see temperature on the inside (-83.3C material) material.
Hello @rd2007 , and thanks for posting your question here!
I do not quite understand the reason of the large box outside to mimic atmospheric temperature. Do you think it may be sufficient to define a room temperature at the end of the material line, and remove the large box to improve computational demand?
So we have the large box as the ambient room temperature as we do not want to set that face as a ‘defined’ temperature.
The reason for this is we are trying to simulate the freezer running. The freezer detailed the inside temperature and has a probe measuring ambient. The ambient being 23.3C, inside being -83.3C, when the face temperature was measured it was slightly less that the 23.3C so we don’t have a ‘defined’ temperature for that face.
In this case, initial temperatures will only act as initial values as you’re using a transient/time-dependent simulation. As the time goes, the temperature will rise and rise according to the ambient temperature defined outside. If this is intended, then yes the simulation setup seems alright from my side.
In case you want to fix the cold temperature, then you should be using a wall boundary condition with a fixed temperature option.
Also, you may want to replicate thin parts by using Thin Layer Resistance - which should help you improve the computational demand.
I ran a few simulations using the ‘Steady State’ rather than transient.
I got some results from these but they are not as i’d have thought.
I refined the mesh adding mesh refinements and also inflate boundary layer.
When doing this on 3/4 simulations they fail, either due to velocity field diverging or the gauge pressure diverging.
The one sim that does work with these refinements is the 30mm VIP 50mm PU the other 3 seem to fail.
