Note: This discussion is about an older version of the COMSOL Multiphysics® software. The information provided may be out of date.

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.

Analytic Pipe flow

Phillip COMSOL Employee

Please login with a confirmed email address before reporting spam

Hi everyone!

My problem:

I built a model of a geothermal pipe. Basically it is a U-formed pipe wich is surrounded by (warm) earth and with water flowing through it. Within the pipe the water becomes warmer.
I let Comsol calculate the laminar flow in the pipe, the heattransport in the fluid and in the solid. This works good as long as the pipe is not too long (max. depth of the U: 10m). But i want to extrude the geometry up to a depth of 100m and this ends up in to much mesh-points and degrees of freedom to calculate it.

My idea to solve the problem:

I thought about leaving away the laminar flow part by just giving Comsol the analytic (parabolic) pipe flow solution for the heat transport in fluids. The advantage is that Comsol would not have to calculate it.
Does anyone know how to get Comsol to use the analytical solution of a laminar pipe flow? Especially the turning part of the U-pipe makes me wondering how that could be done.

5 Replies Last Post May 15, 2012, 4:13 a.m. EDT
Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Mar 30, 2011, 3:24 p.m. EDT
Hi

I would try to section the pipe into three areas, the two straight ones, and the half circle (u), and then use a cylindrical coordinate frame. But OK this works easily for a 2D, more complex in 3D. And a pipe in 2D is a "long plate" and not a circular section, so its not giving a correct model, or one with fringe effects.

Have you tried to calculate the fluid flow separately (manual segregated solver approach) then store it, and use it as initial values for the HT part ? assuming the velocity is not changing with the temperature. At least valid as first approximation, in my view

--
Good luck
Ivar
Hi I would try to section the pipe into three areas, the two straight ones, and the half circle (u), and then use a cylindrical coordinate frame. But OK this works easily for a 2D, more complex in 3D. And a pipe in 2D is a "long plate" and not a circular section, so its not giving a correct model, or one with fringe effects. Have you tried to calculate the fluid flow separately (manual segregated solver approach) then store it, and use it as initial values for the HT part ? assuming the velocity is not changing with the temperature. At least valid as first approximation, in my view -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Mar 31, 2011, 3:52 a.m. EDT
Hi Phillip

Long pipes are really difficult to handle with CFD and you idea good one but hard to define.

There is also another type of solution. In this flow is defined by "thin conductive shell" and not calculated at all. I cannot explain idea because I haven't made this... check if you get can idea form this.

Model can be found from:

cid-17c9234737d68a3c.skydrive.live.com/self.aspx/.Public/lattia%5E_lammitys%5E_3D%5E_ver2.mph

and explanation and pictures from:

comsolfi.blogspot.com/2010/04/lattialammityksen-simuloinneista.html (in finnish, sorry)

Note that this is made with Comsol 3.5 so I don't know how it works with 4.1.

Best regard

Tero Hietanen

Hi Phillip Long pipes are really difficult to handle with CFD and you idea good one but hard to define. There is also another type of solution. In this flow is defined by "thin conductive shell" and not calculated at all. I cannot explain idea because I haven't made this... check if you get can idea form this. Model can be found from: http://cid-17c9234737d68a3c.skydrive.live.com/self.aspx/.Public/lattia%5E_lammitys%5E_3D%5E_ver2.mph and explanation and pictures from: http://comsolfi.blogspot.com/2010/04/lattialammityksen-simuloinneista.html (in finnish, sorry) Note that this is made with Comsol 3.5 so I don't know how it works with 4.1. Best regard Tero Hietanen

Phillip COMSOL Employee

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Apr 6, 2011, 5:11 a.m. EDT
Thank you all very much for your really fast replies and your helpful suggestions.
I forgot to say that my problem is 3D and there is no chance to reduce it to 2D.
(It is a geothermal problem and the underground will be heterogenous.)

The example of Niklas Rom could help because this would solve
the flow problem and reduce the problem effectively.

Thanks again and best regards,
Phillip Oberdorfer
Thank you all very much for your really fast replies and your helpful suggestions. I forgot to say that my problem is 3D and there is no chance to reduce it to 2D. (It is a geothermal problem and the underground will be heterogenous.) The example of Niklas Rom could help because this would solve the flow problem and reduce the problem effectively. Thanks again and best regards, Phillip Oberdorfer

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago May 23, 2011, 8:20 a.m. EDT
Hii Phillip Oberdorfer....I am working on same problem as yours and I am new to using Comsol and this problem........I have done modeling in rhinoceros and tried to import it to Comsol.....but it does´nt predict the geometry in right manner.......please help me in modeling the system
I have to simulate the system for 35 m depth of pipes and want to find outlet temperature for fixed inlet temperature, homogeneous or heterogeneous ground with time varying temperature and rectangular body made of cement surrounding the pipes....with time varying properties.......As you have already done for 10m depth pipe ...Please help me with the idea and steps I should follow for solving this problem .......if possible give me some of your work or tutorials......any contribution will be a great help to me....


Thank you for your time.....

Akshay
Hii Phillip Oberdorfer....I am working on same problem as yours and I am new to using Comsol and this problem........I have done modeling in rhinoceros and tried to import it to Comsol.....but it does´nt predict the geometry in right manner.......please help me in modeling the system I have to simulate the system for 35 m depth of pipes and want to find outlet temperature for fixed inlet temperature, homogeneous or heterogeneous ground with time varying temperature and rectangular body made of cement surrounding the pipes....with time varying properties.......As you have already done for 10m depth pipe ...Please help me with the idea and steps I should follow for solving this problem .......if possible give me some of your work or tutorials......any contribution will be a great help to me.... Thank you for your time..... Akshay

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago May 15, 2012, 4:13 a.m. EDT
Hi all pipe flow modelers. We have now released a product that should be of great interest to you, the Pipe Flow Module. The lead image on this release page is an example from the model library.
www.comsol.com/products/4.3/

Niklas Rom, COMSOL
Hi all pipe flow modelers. We have now released a product that should be of great interest to you, the Pipe Flow Module. The lead image on this release page is an example from the model library. http://www.comsol.com/products/4.3/ Niklas Rom, COMSOL

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.