Ansys Fluent: Introduction to Natural Convection | Tutorial

[@abhinandanbhardwaj1369]

Please make 3D video as well. And more videos like this.

[@cfdkareem]

Working on a 3D example now! Should be up in the next week. Thanks for watching.

[@vijethanchalkar5340]

Thanks!

[@christianmuller5143]

yes a 3D Model would be great. I addition a cooler on the enclosure with forced convection would be a great add on. Is it possible to take the surface temperture of the enclosure as an heat surce of a new thermal calculation? How would I do so? I general it would be great, if you show how to connect different thermal analysis (or a thermal analysis with an fluent analysis)!

[@randomstuff8510]

Sir please make a video for the heat sink analysis (3D analysis) with natural convection having a TDP of 5w. Sir please please make this video as soon as possible Or tell me the inlet and outlet conditions which I need to put for natural convection Thanks in advance

[@cfdkareem]

Hello, the inlet and outlet conditions for natural convention will be the same for 3D as the 2D, shown in the video. Each boundary will be a pressure inlet/outlet with pressure=0.

[@leolkliu]

Hi sir, why didn't you consider to enable "Radiation" for this natural convection case? As I know, radiation would contribute significant power dissipation as well in natural convection cooling.

[@cfdkareem]

Hello, great question! Radiation usually only becomes a significant source of heat flux if there is a significant temperature delta between two bodies. For most applications of air cooling radiation will be ignored. You can add radiation using one of the models in fluent if you believe it will be a significant source of cooling for your problem. Thanks!

[@leolkliu]

@cfdkareem Thank you for your quick response. I am new to Fluent. I would do more cases for Thermal analysis. 😀

[@cfdkareem]

@leolkliu thanks for the suggestion! I was planning on some more thermal analysis videos in the next few weeks, including radiation. Do you have a specific problem you are interested in?

[@leolkliu]

@@cfdkareem Do you have an email box? Perhaps we can have some discussions or share more info. Thank you.

[@franklineze8636]

You mentioned the inflation layers are important for capturing the heat flow, but you didn't use it for the earlier 3D forced convection video. Could there be a reason for that?

[@cfdkareem]

Yeah with the student/educator license you are limited by mesh size. Inflation layers would add significant size to the mesh, but should definitely be included for a real study.

[@franklineze8636]

@@cfdkareem oh, alright. Thank you!

[@abdullaal-tarmoom3437]

hello karee, can you validate the results for this case?

[@cfdkareem]

Hello, this case was created solely for tutorial purposes and does not have any corresponding experimental data. Some estimations about the validity of the results can be gathered by solving analytical equations for the 1D heat transfer through the heated body.

[@Lucifer0396]

Hello your videos are very helpful. I just had one doubt with respect your video with heat sink under forced convection, if we want natural convection for this case what changes we need to make?

[@cfdkareem]

Hello, to get natural convection you will have to change the density of air to "incompressible ideal gas" and make sure then gravity is set to "-9.81". If you want to do natural convection with a heat sink you can copy the set-up from this video for a 3D domain.

[@randomstuff8510]

Then what should we take at the inlet and outlet should we go with velocity or pressure and what values should we consider Please reply

[@АлександраБогданова-м9ц]

Hello! Can you help me please? An error occurs Inhomogeneous process distribution on multiple machines. Or processes per machine not an exact multiple of GPGPU's per machine. Or not enough GPGPU's per machine. GPGPU computing disabled. I went through all the possible combinations of Solver Processes and Solver GPGPUs per Machine 1- 1, ..., 1- 4, 4-1))) (When you press Setup). Clear generated data also doesn't help. Thank you for video! It's very interesting

[@cfdkareem]

Hello, I would try to remove gpgpu entirely. GPU computation in fluent doesn't give a significant speed up in simulation for most cases. It is usually recommended to keep it off unless there is a specific physics model that is known to provide a speed up in computation time.

[@АлександраБогданова-м9ц]

Thank you very much for the hint. When I set it to zero, the errors disappeared, but an empty screen remained. How I solved it: I cleared the generated data and updating the project. After that, I had it displayed.@@cfdkareem

[@lucasasth7458]

Hello! Great videos!! Do you have any experience with PyFluent? If yes, can you make a video abou it? Thank you!

[@cfdkareem]

Hi Lucas, I have experience coupling Fluent and Python prior to the introduction of PyFluent. I'll try it out and see how it works. Is there a specific model/problem you are trying to use PyFluent for?

[@lucasasth7458]

​@@cfdkareem Yes! It could be any conduction or convection problem. It could be in a simple 2D geometry, such as a plate, for illustrative purposes. In my specific case, I would like to automate the solution of the model in Python so that I can run it in a loop, changing only some parameters of the model, and rerun the simulation. My application involves an inverse problem, which requires running the 'direct problem' numerous times. Thank you very much for the response !!

[@ammarlaichi8474]

Hello Mr, I have question about mixed convection of water or nanofluids in circular duct, the emperature raises in each cross section great than 10 degree, but boussinesq hypothesis not valid in this case, What do we do now?

[@cfdkareem]

If you have the data for density vs. temperature you can set the material property to match instead of using an approximate like ideal gas or boussinesq

[@ammarlaichi8474]

but in this case we use water as working flluid@@cfdkareem

[@ammarlaichi8474]

in operating condition we activate gravity, density of water and inlet temperature in operating temperature, is it correct?@@cfdkareem

[@cfdkareem]

Yes that is correct. If you have a temperature difference and a variable density then you will have natural convection.@@ammarlaichi8474

[@omarabdelhamid3726]

Do you see Icepak better for this simulation?

[@cfdkareem]

Icepak runs Fluent solver in the background. So from an accuracy standpoint it wouldn't change much, but could save some time on the setup. I prefer to work in Fluent because I find it more flexible