Advise : run a steady state study to verify the model before running the longer calculation transient study.
Check the intial temperature definition.
It might be that in the boundary conditions definitions, only the external faces are defined at 22°C, but nothing is specified for the internal temperature of the part(s). Then the model could use zero Kelvin (-273.15°C) as intial temperature for the internal body, and give a resulting temperature lower than the (thought as) initial temperature.
Hope this helps,
Thanks for the reply Regis.
When I choose the "transient" option in properties, I selected "Thermal 1", where every single face was at 22°C. Then in the "Thermal 2" (Transient one), I set only a couple faces with different values (such as 100°C and 850°C) while leaving the rest of the faces empty, but that still resulted in negative values. I thought that calling the initial temperature of 22°C from Thermal 1 would make everything initially at 22°C and I wouldn't need to apply it on the faces again?
Also, what's the use of having the choice between "initial temperature" and "temperature" after I choose the "transient" option in "Thermal 2"? Isn't my initial temperature coming from "Thermal 1" with 22°C?
Also, when I apply 22°C (or any value) on all the faces on my "Thermal 2" (Transient part), it seems to override the 850°C and 100°C parts) even if I didn't apply the 22°C on those objects...
Hm, as I use the probe technique, I find that the structure is indeed at 22°C initially (so no need to reapply that again in the transient). Everything seems reasonable except for one part: There is a small area around the 850°C part that goes to -171°C, which doesn't make sense at all. Everything else is resonable except where it goes to -171°C (which also brings areas around that to lower temperatures like 1°C) and I've checked that everything is defined (either initially or in transient properties) so I don't know why that's happening
In Thermal2, for the t = 0 s, there would still be some internal nodes with no temperature definition.
It could be that the nodes in question are not taken in the mesh/result in Thermal1.
- Copy-paste the mesh of Thermal 1 to Thermal 2. The nodes of the meshes of Thermal1 and Thermal2 should match in number/positions.
- Check that steady state Thermal 1 has all (internal) points/nodes to 22°C (Probe->On selected entities-> select the body(ies)->Update). Not only faces.
- If there are several bodies, check that the "850°C" body is included also in the Thermal 1 Parts.
- Is that body included and meshed in Thermal 1 (Mesh->Create Mesh Plot, and compare the meshes Thermal1 / Thermal 2)?
- If a face is used in two boundary conditions definitions, usually only one of them is used.
- External faces should have a boundary condition. Either insulated (no heat flux), or a convection, or a radiation, or a temperature applied, or a symmetrical condition, ...
- Is the "850" part a separate body in the geometry ? Has the other body in contact with it, a splitted face at the contact area ?
I duplicated the "Thermal 1" to "Thermal 2" so the mesh and all the parts are exactly the same (including the 850°C one). I checked that all internal nodes are also 22°C. So literally everything in "Thermal 1" is set to 22°C. The 850°C body is a separate body that is bonded to the rest of the structure.
In "Thermal 2", I changed it to "Transient" with initial conditions from "Thermal 1". Then I selected the body that had the 850°C and set it to 850°C. The rest of the structure I didn't apply anything because I assumed that it would be set at 22°C since the initial condition from "Thermal 1" initially sets everything at 22°C.
Nvm I figured it out, thanks for all your help!