[CFDlearnerY]

Hello,

I am actually evaluating if OpenFoam is the right choice for my case.

What I want to simulate is as follows: I have a pipe that has an insulation layer around it. Furthermore, there are 2 small pipes in direct contact with the main pipe through which cold water flows (for cooling). These 2 pipes are inside the insulation layer and are parallel to the main pipe. A fluid (liquid) flows inside the main pipe and the whole system is in an environment with for 30 °C. I also need to take radiation from the sun into account.

So I have 3 fluids (fluid inside the main pipe, water and air outside the insulation layer) and 3 solids (main pipe, insulation and the small cooling pipes), but non of these phases have mass transfer to each other, only heat transfer (Is it a multiphase case?). I want to calculate heat transfer and temperatures in transient condition.

My questions are: Can I use OpenFoam for this case? if yes with which solver? How should I handle it when I have 2 incompressible fluids and one compressible (Air)?

Thank you very much for your help!

[jmt]

You can use chtMultiRegionFoam. This solver allows an arbitrary number of fluid and solid regions which are coupled via a temperature and flux matching condition.

I have spent time validating this solver and find it useful. Out of the box, the fluid and solid regions are tightly-coupled, so you may have very slow solid convergence unless you implement a decoupling scheme (which I and others have done).

[CFDlearnerY]

Thank you very much for your reply. I know that this solver is designed for compressible fluids. How can I define incompressible fluid then? Another question is, should I consider the air around the pipe as compressible fluid or incompressible? Since it is natural convection

Originally Posted by jmt

You can use chtMultiRegionFoam. This solver allows an arbitrary number of fluid and solid regions which are coupled via a temperature and flux matching condition.

I have spent time validating this solver and find it useful. Out of the box, the fluid and solid regions are tightly-coupled, so you may have very slow solid convergence unless you implement a decoupling scheme (which I and others have done).

[dlahaye]

Could you please elaborate on the decoupling procedure? Thx!

[jmt]

Originally Posted by CFDlearnerY

Thank you very much for your reply. I know that this solver is designed for compressible fluids. How can I define incompressible fluid then? Another question is, should I consider the air around the pipe as compressible fluid or incompressible? Since it is natural convection

For the compressible fluid, you can set density to constant in the thermophysical model. In chtMultiRegionFoam, each region has a separate folder containing the relevant physical models and solver options. This can be tedious to configure but is extremely flexible in defining the problem.

How large are your temperature and density gradients? This may help you decide whether to treat the air around the pipe as compressible or not. For natural convection, I suppose you will need to consider buoyancy, so perhaps compressible is the best approach.

[jmt]

Originally Posted by dlahaye

Could you please elaborate on the decoupling procedure? Thx!

Regarding decoupling, there are any number of ways you could implement this. Personally, I used the approach of Konle et al. in which the fluid region is periodically frozen and the solid regions are solved with a larger time-step (governed by the solid time-scale). This has been effective.

You could also consider the approaches described on the Chalmers site in which information is exchanged between the fluid and solid regions at some larger interval as opposed to every time-step.

[dlahaye]

Thx! Much appreciated.

[CFDlearnerY]

The environments temperature is around 30 °C. The fluid flowing inside the main pipe is 6 °C and water flowing through cooling pipes is 4°C. The pipe has a pretty thick insulation layer and I expect that out surface of the insulation will become very warm due to radiation from the sun.

Thank you very much for your help

Originally Posted by jmt

For the compressible fluid, you can set density to constant in the thermophysical model. In chtMultiRegionFoam, each region has a separate folder containing the relevant physical models and solver options. This can be tedious to configure but is extremely flexible in defining the problem.

How large are your temperature and density gradients? This may help you decide whether to treat the air around the pipe as compressible or not. For natural convection, I suppose you will need to consider buoyancy, so perhaps compressible is the best approach.

[jmt]

Great. I think compressible is appropriate.

Please let me know if you have further questions. I might be able to help.