Learning Forum

[DG89]

Hi everybody,

[Karthik Remella]

Hello What boundary conditions are you using at the other boundaries? There is a hot spot in your domain towards the bottom. Any reasons for that? Can you plot the corresponding temperature plot here?
Karthik

[DG89]

Hi Karthik thank you for answer. At the other boundaries I did not use any boundary conditions (are they considered adiabatic by Ansys?). There is no reason for the hotspot. I share the temperature plot, as required. I do not see anything strange...

[DG89]

Any idea?

[Karthik Remella]

Hello Apologies for my delayed response!
Please specify boundary conditions on all boundaries. It is a good practice and Ansys will respect all the exact conditions you wish to impose. Please impose all boundary conditions and see if the results make sense.
Karthik

[DG89]

Hello Apologies for my delayed response, in turn!
I tried to impose all boundary conditions, but the problem still survive. I also tried to change model (I choosed a very simple one, that is a prism), the materials (default material), the display options (elemental mean), but I did'nt obtain benefits.

[Karthik Remella]

Hello What is your entire geometry? Is this a 2D or 3D run? Can you please help me understand the problem you are attempting to solve?
Karthik

[DG89]

Ok, let's start with a more simple object:
A prism, 0.1m hight, 1m long, 1m large
standard steel material
3d solid body
Steady standard thermal analysis
Convection in two opposit faces, while the others four are perfeclty insulated.

The PDF shows some windows of setting parameters and some results. As you can see, even if the object is very simple the directional heat flux along the z axix (please, refer to the images in the file) is not uniform in the y direction, and the same occurs for the temperature. Moreover, the direction flux along the x is very "strange", while I'm expecting it whould be zero, according to the inputs. Curiously the higthest and the lowest values are the same, in absolute value.

[peteroznewman]

Try to use more than 2 elements through the thickness. I built your model with 9 elements through the thickness and get a perfectly uniform heat flux in the Z direction.
The Heat Flux in the X direction is numerically equal to zero.
The temperature plot looks normal.

[DG89]

I'm tried to follow your suggestion, but my result is still different from your.
Firstly I changed the mesh:
Here below I post the screenshot of the settings:
And finally the results:

[peteroznewman]

Archive your latest model and attach it to your reply. I will rerun that in ANSYS 2021 R2. Maybe the 7 year old solver doesn't give the same result as the latest solver.

[DG89]

Ok, I solved the problem:
the geometric model had not only solid bodies but also two surfaces which, peraps, influenced the final result. I suppressed them and now the solution is similar to your.

[DG89]

However, if I pass to a more complex geometry the problem persists.
As regard the heat flux in the z direction, it is interesting to note that the absolute value of the maximum and minimum are the same; it seems to be a simmetry in results.

[DG89]

Good afternoon,
May anyeone try to fix my second problem? Thankyou.

[DG89]

I remember that the problem is the unsymmetry in the directional heat flux along the z axis, despite the boudary conditions and the geometry.