Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.

Trouble understanding Heat Flux integrals

Please login with a confirmed email address before reporting spam

Hello!

I am working on a model that involves heat flux, with conduction, convection and radiation. In the model I have to compare the losses over the outher boundaries. After running the model, which is 2D axisymmetric, I use a line integral over the boundaries, to obtain a surface integral in 3D. However, the results do not make sense to me, which makes me think that I do not understand the specific integrals correctly. I included a picture of the boundaries and the corresponding values found.

What I do not understand:

  1. Why does it say that there is no convective heat flux when I included convective heat flux in the heat transfer physics interface?

  2. Why is the value for conductive heat transfer so large, when the temperature difference would mean a much larger convective and radiative heat flux?

  3. Why does the total energy flux (which gives the same values as the total heat flux) only include the conductive heat flux?

Can someone tell me how to interpret the given results? It would be greatly appreciated.

Kind regards, Steggink



4 Replies Last Post 19 nov 2019, 03:50 GMT-5

Please login with a confirmed email address before reporting spam

Posted: 5 years ago 1 nov 2019, 11:26 GMT-4

I think I found where the problem lies. Namely, I am using a single layer material on the upper most boundary (again, see picture attached to last post), which has its results in a different result node. However this gives rise to two new questions:

  1. How are boundary conditions, such as Heat Flux, imposed? On the material in the domain, or on the single layer material on the same boundary? Or in other words: does a boundary condition such as Heat Flux (or Surface-to-Ambient Radiation) 'see' that there is a thin layer in between?
  2. When using the line integral in the results, on a boundary that has a thin film material, how do I know which side of the thin layer material is accounted for? Since I can only select the boundary, not a side of the thin layer material.

Any help would be greatly appreciated.

I think I found where the problem lies. Namely, I am using a single layer material on the upper most boundary (again, see picture attached to last post), which has its results in a different result node. However this gives rise to two new questions: 1. How are boundary conditions, such as Heat Flux, imposed? On the material in the domain, or on the single layer material on the same boundary? Or in other words: does a boundary condition such as Heat Flux (or Surface-to-Ambient Radiation) 'see' that there is a thin layer in between? 2. When using the line integral in the results, on a boundary that has a thin film material, how do I know which side of the thin layer material is accounted for? Since I can only select the boundary, not a side of the thin layer material. Any help would be greatly appreciated.

Nicolas Huc COMSOL Employee

Please login with a confirmed email address before reporting spam

Posted: 5 years ago 5 nov 2019, 10:47 GMT-5
Updated: 5 years ago 5 nov 2019, 10:47 GMT-5

Dear Steggink,

I believe that you observations are fine. Here are some details about the postprocessing variables: 1. The convective heat flux boundary condition set a heat flux at the boundary that is related to the convection in the fluid domain next to the boundary. But since the velocity is 0 at the boundary so there is no heat transfer by convection at the boundary. 2. & 3. The conductive heat flux corresponds to the total heat flux that enters the solid since there is only conduction in the solid. This corresponds to all the energy received by conduction, convection or radiation from the exterior at the boundary.

Regarding your second message, the heat flux and surface-to-ambient are aware of the layer and are applied on the exterior side of the layers. It is possible to distinguish the upside and downside of a boundary by using the correponding variable. For example the temperature on the upside is ht.Tu, on the downside it is ht.Td. The flux variables also have a u and d version to distinguish the 2 sides of a layer.

I hope it helps, don't hesitate to contact the support and share a model if we can help further.

Best regards,

Nicolas, COMSOL

Dear Steggink, I believe that you observations are fine. Here are some details about the postprocessing variables: 1. The convective heat flux boundary condition set a heat flux at the boundary that is related to the convection in the fluid domain next to the boundary. But since the velocity is 0 at the boundary so there is no heat transfer by convection *at* the boundary. 2. & 3. The conductive heat flux corresponds to the total heat flux that enters the solid since there is only conduction in the solid. This corresponds to all the energy received by conduction, convection or radiation from the exterior at the boundary. Regarding your second message, the heat flux and surface-to-ambient are aware of the layer and are applied on the exterior side of the layers. It is possible to distinguish the upside and downside of a boundary by using the correponding variable. For example the temperature on the upside is ht.Tu, on the downside it is ht.Td. The flux variables also have a u and d version to distinguish the 2 sides of a layer. I hope it helps, don't hesitate to contact the support and share a model if we can help further. Best regards, Nicolas, COMSOL

Please login with a confirmed email address before reporting spam

Posted: 5 years ago 14 nov 2019, 03:26 GMT-5
Updated: 5 years ago 14 nov 2019, 04:14 GMT-5

Dear Nicolas,

Thank you very much for your reply. It is only now that I see it, so sorry for replying a little late.

Your exlplanations are very clear. I have just one thing that still troubles me.

  1. You say that the conductive heat flux should contain the convective, conductive and radiative losses (so that I should only take the total heat flux for total losses? Up untill now, I was adding the conductive heat loss and the absolute value of the radiative loss). Why then, does it seem like the radiative flux goes in an opposite direction (aka it has a minus sign)? Since in my model, the inside always has a higher temperature than the ambient. Also, for slightly different parameters, I obtain results with a higher value for the radiative heat flux integral than for the conductive one. When in reality the radiative heat loss would be contained in the total heat flux, this would seem odd to me. What is the flaw in my interpretation here?

Since I make use of a CKL, I unfortunately cannot make use of the support, so your answers are really helpfull!

Thanks again.

Dear Nicolas, Thank you very much for your reply. It is only now that I see it, so sorry for replying a little late. Your exlplanations are very clear. I have just one thing that still troubles me. 1. You say that the conductive heat flux should contain the convective, conductive and radiative losses (so that I should only take the total heat flux for total losses? Up untill now, I was adding the conductive heat loss and the absolute value of the radiative loss). Why then, does it seem like the radiative flux goes in an opposite direction (aka it has a minus sign)? Since in my model, the inside always has a higher temperature than the ambient. Also, for slightly different parameters, I obtain results with a higher value for the radiative heat flux integral than for the conductive one. When in reality the radiative heat loss would be contained in the total heat flux, this would seem odd to me. What is the flaw in my interpretation here? Since I make use of a CKL, I unfortunately cannot make use of the support, so your answers are really helpfull! Thanks again.

Please login with a confirmed email address before reporting spam

Posted: 5 years ago 19 nov 2019, 03:50 GMT-5

Dear mister Huc, or anyone else,

I am still having trouble understanding my results. In a follow up of the first model, I obtain the results that are shown in the attached picture. I have labeled the normal total heat flux that Comsol gives me when integrating over the outer boundaries (the roof, air outflow, wall sections, air inflow and bottom, from left to right) with colors. In this model, a 20 kW heat source is applied at the bottom surface (boundary heat source). These do not make sense to me for the following reasons:

  1. The total of all losses is much larger than the 20 kW supplied, which is the only heat source.
  2. The result for the bottom isn't 0, which it should be since it is thermally insulated, but also not the 20 kW which is prescribed for this boundary (although it is supposed to be directed upward).
  3. The sign of the loss at the air outlet is opposite to the loss through the roofs, even though there is an outflow of heated air (which should thus be a loss in heat) so that the sign should be equal to the loss through the roof. Equally, the result at the cold air inflow is not as I had expected.
  4. The loss through the walls is diproportionally large.

I included the MPH file of the model below.

Can anyone point me to where the problem(s) lie(s)? Thanks a lot in advance.

Kind regards, Steggink

Dear mister Huc, or anyone else, I am still having trouble understanding my results. In a follow up of the first model, I obtain the results that are shown in the attached picture. I have labeled the normal total heat flux that Comsol gives me when integrating over the outer boundaries (the roof, air outflow, wall sections, air inflow and bottom, from left to right) with colors. In this model, a 20 kW heat source is applied at the bottom surface (boundary heat source). These do not make sense to me for the following reasons: 1. The total of all losses is much larger than the 20 kW supplied, which is the only heat source. 2. The result for the bottom isn't 0, which it should be since it is thermally insulated, but also not the 20 kW which is prescribed for this boundary (although it is supposed to be directed upward). 3. The sign of the loss at the air outlet is opposite to the loss through the roofs, even though there is an outflow of heated air (which should thus be a loss in heat) so that the sign should be equal to the loss through the roof. Equally, the result at the cold air inflow is not as I had expected. 5. The loss through the walls is diproportionally large. I included the MPH file of the model below. Can anyone point me to where the problem(s) lie(s)? Thanks a lot in advance. Kind regards, Steggink

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.