[CFD] Convection (Heat Transfer Coefficient) Boundary Conditions

[@jimmyalmacddissi9426]

I’m an active engineer doing CFD and heat transfer for thermal systems design. This is an excellent video and I really liked the format. I’ve been looking for a good description of this type of problem for a while. Well done!

[@DrFatinMunirah]

Happy to hear regarding Finite Element Method

[@fluidmechanics101]

Yes, I would like to get around to this at some point. Finite Volume is far more popular for CFD, so that has been my focus for now

[@arashzandian3540]

Hi Aidan; thanks for your awesome CFD videos. I watched many of your videos - especially the ones about different turbulence and heat transfer models - to refresh my memory for many fluid mechanics and CFD topics and to get ready for a job interview for a CFD Engineer position at Tesla. I'm fortunate to say that I got an offer from them and am soon to start my career at this position. As a token of appreciation, I would like to support your channel. Please let me know how a viewer like me can help out the channel. Cheers

[@fluidmechanics101]

That's fantastic! Congratulations Arash 😊 if you want to support the channel, you can support in whatever way you feel comfortable with. Liking / subscribing / sharing the videos with friends and colleagues is great and always appreciated. If you want to contribute financially (to help me buy more research papers, so I can make more videos) then you can always donate / buy me a coffee / follow on Patreon. All the links are in the video description and any support is much appreciated. Best of luck in your career at Tesla!

[@Dz8598]

This really confirms my hypothesis /suspicion regarding how this boundary condition works. You are a savior, thank you

[@Silavite]

Splendid explanation, thank you for the video! (One side note: At 13:27 I couldn't help but smile a bit when I heard 25 - 30 °C described as a, "hot day.")

[@tejesdas3896]

Fantastic presentation, so lucid. An example is always better. Thank you.

[@SHUBHAMMISHRA]

Yes. It is very useful. Specially the way you had shown the fluent and OpenFoam settings is awesome.

[@kailashrangaraj]

This is one of the best videos, that I can ever watch

[@ramkumarvenkatachalam4993]

Hi Aidan, Really appreciate it. Got the clarity. So useful as I was trying to simulate a Conjugate heat transfer problem with convection boundary condition. Cheers!

[@МаусЦзэдун]

The disadvantage of this boundary condition is that the average coefficient on the surface is assigned. If it is important to know the local temperature distribution, it is better to use a direct calculation of the heat transfer coefficient.

[@rafaortwein6646]

Nice lecture. It would be nice to see an example with meshed air and the convection plume resolved.

[@amirulislam4462]

This video is very helpful for my project.....thank you so much.

[@sanjaykumaryadav3186]

Yes sir it was really helpful. Thank you sir.

[@Unknown-kp3gh]

Thank you for every video. Please can you make video about evaporation-condensation phenomena inside a heat pipe (closed pipe).

[@arjunprasanthU]

Excellent video sir. I have a battery that is discharging. As it discharges, it generates heat which is dissipated into the cooling medium. While the heat generated in the battery is captured by meshing it and using an MSMD plug in that introduces a heat source term in the battery component related cell zones, can I still use the "External heat transfer coefficient to capture heat dissipation ?"

[@hardiksharda9673]

Amazing talk!😀 Thank you!

[@soumiabaalicherif6475]

hello, thank you for your videos really good work, i have quastion about heat transfer coefficient inner suface how can'i calculate Tw and Tref please, i need your help

[@fluidmechanics101]

TW is the local wall temperature. Tref will be the mass flow average temperature across the pipe. (This is the standard definition for internal flows). My video on heat transfer coefficient might help?

[@178airil]

Thanks for the explanation. Just out of curiosity, if the internal heat transfer coefficient inside CFD is post calculated from the solution of temperature, how CFD calculate the heat transfer between the cells. Does it compute based on the temperature wall function and thermal diffussity alpha = k/rho/cp. Is this alpha user input? Does CFD treat these fluid properties as constant value or temperature dependent? In 1D from what I know to calculate inside temperature of flowing fluid user has to to calculate/ input the internal heat transfer as part of thermal resistance based on nusselt correlation

[@fluidmechanics101]

Yes! It is based on the wall function and the thermal diffusivity ☺️ if your mesh is resolved through to the thermal sub-layer then it will be based on the molecular conductivity of the fluid

[@veeraraghavanbalaji7029]

Very useful information!!

[@alesh576]

Thanks man , you are a legend. so nice explained.

[@mohammedziou]

Yes it is very useful, thank you

[@kimia_arooz]

Very useful, thank you

[@EngLhag]

Nice recap. It's sad that my former heat transfer simulations didn't have a constant heat flux coefficient for me to use because I had fluctuations in fluid velocity XD A complete mess.

[@prakashthirunavukkarasu29]

Thank you very much for this excellent talk. I would like to know what is Tref here. I understood that its a reference temperature but what exactly it is that we obtain from the CFD code?

[@fluidmechanics101]

Tref is a reference temperature that you have to apply as an input to the CFD calculations, not an output. If you watch my video 'heat transfer coefficient' you should find the information you are looking for

[@prakashthirunavukkarasu29]

@@fluidmechanics101 Thank you Dr. Aidan

[@foambuilder]

Thank you for the video. I have a question: if I would like to run a laplacianFOAM simulation, how could I impose a free convection BC?

[@fluidmechanics101]

Good question, I'm not up-to-date with OpenFOAM, so I'm not sure. They may already have the boundary condition implemented. I would have a look through the tutorials

[@foambuilder]

@@fluidmechanics101 till now I tried different methods, but in laplacianFOAM I don't know how to solve this issiues!!!

[@xuxu119]

Thank you.

[@98pedroneto]

Another great video, thanks! Also, I was curious to ask, is there is a similar boundary condition for mass transport, using a mass transfer coefficient?

[@fluidmechanics101]

I don't think so because the wall physically blocks the transport of mass from the inside of the pipe to the outside. With energy (temperature) the pipe wall conducts the energy from the inside to the outside of the pipe, so an external heat transfer coefficient makes sense

[@98pedroneto]

@@fluidmechanics101 Thanks for the reply! But what about in the case of a gas-liquid interface, do you think that in that case the boundary condition would make more sense?

[@fluidmechanics101]

I think you would need a careful mass transfer condition in your multiphase model (not a boundary condition)

[@ravirajuvysyaraju8211]

@@fluidmechanics101 Hi , I am following your videos from quite a long time! They are really good. I am wondering can you make a video on mass transport using Ansys fluent. It will be really helpful for me and many people out there. There is very less info on mass transport in open source.

[@pritamyawale989]

More boundary conditions please

[@gustavohenriquecosta2111]

Hi Aidan, thanks for your videos, you've been of great help for me as I'm starting to learn CFD. As of today, I'm trying to simulate a case where i can obtain the external HTC for a hot pipe surrounded by ambient air and then i want to compare the HTC values calculated with OpenFOAM with the values calculated using empirical correlations, so i can validate the source code. (assuming fixed temperatures for both the pipe and air) Is it possible to do this? I'm not sure if i understood your video right, but what i got is that the CFD code can't calculate external HTC?

[@fluidmechanics101]

You definitely can. You just need to mesh the air around the pipe. The reason for applying a HTC is that (for large problems) the mesh can get quite large and you want to save cells, to have a faster simulation (so apply a HTC rather than meshing)

[@gustavohenriquecosta2111]

@@fluidmechanics101 First of all, thank you so much for taking your time to answer me! Oh, okay, i think i understand it now. But, watching your video about HTC i came to the question, how to extract the HTC from my simulation? Shall i look for the heat flux and then calculate HTC or is the solver able to provide me a Nusselt number? I'll be using buyoantFoam in OpenFOAM v10.

[@fluidmechanics101]

Pull out the heat flux, and then calculate the HTC yourself in paraview or MATLAB, so that you can be sure you know what you are doing (dividing by the correct reference temperature)

[@gustavohenriquecosta2111]

@@fluidmechanics101 wow, thank you sooo much for this! you helped me a lot.

[@ramkumars2329]

excellent video.. thank you!

[@achyuth_rj]

Hi Aidan, great content again! I am literally binge watching your videos preparing for a CFD interview :D A query: after computing the heat flux, how can we use that to understand the structure of the plume outside the pipe?

[@achyuth_rj]

do we directly extrapolate structure of thermal BL from cooling law?

[@fluidmechanics101]

You could set up a separate CFD model for the outside of the pipe, with the boundary condition on the pipe surface set using the results of your CFD model for the inside of the pipe. This is useful in some circumstances where you want to look at parts of a system in detail and are happy to have different 'sub models' of different parts of the system you are looking at

[@achyuth_rj]

@@fluidmechanics101 Got it! Thank you for taking the time to reply.

[@kanchaitheinglim8165]

Hi, I create the air enclosure over my system. Should I set the boundary condition at 6 faces of air enclosure?

[@fluidmechanics101]

I think I understand what you mean. You could always run a quick simulation with a very coarse mesh just to check your boundary conditions?

[@kanchaitheinglim8165]

@@fluidmechanics101 yes, I think 6 faces of air should be open, but I am not sure they should be set by pressure outlet, constant temperature.

[@fluidmechanics101]

Ah yes, you will need outflow on one of the faces so that you can conserve mass. Many CFD codes are clever enough to switch the face around to allow outflow on a pressure outlet. Sometimes this is called an opening boundary condition? If not, you might have to specify pressure inlets and outlets separately

[@erhanarslan1473]

Thank you so much.

[@AyaFahham-ol2qx]

Love you🎉❤

[@EclecticVibe]

Hello Mr.Aiden, this question is not regarding the convective coefficient. I am wondering if you have a video on compressible flows. Thank you.

[@fluidmechanics101]

Hi Mandeep, I don't have a video on compressible flow yet. Hopefully I will have one soon 🙂 it is definitely on my list of videos to make

[@EclecticVibe]

@@fluidmechanics101 Thank you Aiden... definitely looking forward to it. Really appreciate the amazing work you are putting into the video.

[@muhammadsaad5055]

Hello! I am modeling natural convection in internal flow. There is glass and ground, with radiations on the glass. I took glass and ground both as convection boundary conditions with operating and freestream temperature as same. So, that means the flux is zero. I took the ground as convection because I wanted to have the heat absorption effect on the ground, Is it the right approach to use ground with certain thickness as a convection boundary condition?

[@tingting7555]

Thanks mate!

[@joaocoutinho1725]

Hey Aidan! Great content as usual. I am new to CFD, but could we determine the h coefficient by doing only the fluid-air mesh (with an allow with the pipe format), and give a constant Temperature condition on the walls = Tsurface? And the output would be the h coefficient determined with Tref?

[@fluidmechanics101]

I don't quite understand your question but it sounds right to me ☺️

[@joaocoutinho1725]

@@fluidmechanics101 ahaha sorry maybe my English or CFD knowledge is not the best. Imagine I have a cylindric box at 20C and a external flow (forced convection) with a velocity of 1m/s and Tfluid=-18. I want to know the convection coefficient so I can use that value on the simulation with only the cylindric body. Can I use other simulation, only with the fluid domain (without the solid domain) to study the flow and therefore the convection coefficient?

[@fluidmechanics101]

Yes absolutely! That is a great way to do it. If you simulate the outside and use this to calculate a heat transfer coefficient, you can then apply this heat transfer coefficient onto the external walls of your cylindrical domain. Chances are the heat transfer coefficient will probably be between 5 and 10 W /m2 K, it usually is ...

[@ursamajorable]

Excellent. Can you make a lecture on solving Natural Convection Heat Transfer in OpenFOAM?

[@shashidharreddy2959]

What if the pipe is at the focus of a parabolic trough? The pipe receives non uniform heat flux in the form of concentrated sun light. And there there are convective losses. Ansys Flent doesn't allow to specify both heat flux and external convective heat transfer coefficient at the same time. What to do in this case?

[@sanjaykumaryadav3186]

how do i calculate if it is complex geometrical shape instead of circular cylinder. thank you

[@fluidmechanics101]

You may not have an empirical correlation for complex geometries. You can either model it as a simple shape or run your own CFD simulation of a sub-model and derive your own heat transfer coefficient 😄 quite tricky but it can be done!

[@SHUBHAMMISHRA]

I am looking for simulation details of phase change material (PCM)

[@hamnaabdulrab7717]

Can you please tell mw how can I drive the convection for biot number

[@NAYAN-t3e]

But what to do if we have to capture the effects from point to point on exposed surface boundary ? I.e. To use local variation of heat transfer on the exposed surface with outside ?? As directly mentioning the theoretically calculated value will only give average value which supposed to be constant over the complete interfaced boundary surface exposed. As this is simple for regular geometries like pipe, plate, etc. But how to tell to CFD code that my 'h' coefficient is varying with position like x, D or theta??Is including code for that necessary ??

[@fluidmechanics101]

Yep, you will need to include some code for that. Exactly how you do the code will depend on the CFD code you are using. In CFX you can use CCL, in fluent you need a UDF etc.

[@NAYAN-t3e]

@@fluidmechanics101 How can we do it in comsol multiphysics or Matlab ??

[@fluidmechanics101]

Good question! I haven't used Comsol before. Could you email their customer support / help desk?

[@trooper2708]

Sir can you help me decide correct boundary condition for a solar collector which is of glass and acts as a solar wall beneath which air is heated due to green house effect.

[@fluidmechanics101]

I think there is some guidance in the Fluent manual for solar loading. Can't remember off the top of my head though!

[@roopakbaliyan5935]

Can anyone told me how to calculate local heat transfer coefficient on inner side of pipe.

[@martinneumuller9561]

Thank you! In mathematics this type of boundary condition is also known as "Robin boundary condition", see en.wikipedia.org/wiki/Robin_boundary_condition