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 The information provided may be out of date.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.

Dear Mr.Ivar or others, would you give me some more help about my 2-D collapsible channel?

Please login with a confirmed email address before reporting spam

Hi, everyone!
I am driving crazy by Comsol! I am lost in some many errors and other mistakes!

I attached my mph file, can you look at it , and give me some suggestions!

It keeping showing the error: failed to find consistent values'

I want to solve the time dependent problem.

Also I am confused why I have three studies, study1, study2, study3, which one I should use.

Thank you so much!



6 Replies Last Post Mar 16, 2011, 9:56 p.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 15, 2011, 7:29 p.m. EDT
Hi

There are many things to say here, first you are jumping into a 3 physics model without obviously really understanding each one, that is a really tough start, you should start with an easier example in each of the physics and then start to couple them. It takes a few years to learn structural, fluids and computational physics at the university, so do not expect to master COMSOL in a few days, but if you start simple you will quickly catch up and become an expert, I'm sure. COMSOL has also some excellent courses absolutely worth to follow, and you have several videos on the web site

A few comments:
1) your v4.1 is "old" you should see with your local COMSOL rep how to upgrade to a newer patch, it would help for the solving process too
2) your spf is more or less OK, select in the study 1 only the spf "Use in this study", deselect the two others an solve the single physics. I had to reset the solver by Right clicking on Study 1- Solver Configuration - Delete Solvers and set the time stepping to (0,0.01,1) (1 sec is enough). You can also see what is happening by turning on the plots during solving: further down in the solver nodes select "Fully Coupled" and open Results while solving and select Plot
3) the Solid physics will not work like that, as it is set up for a Domain i.e. the same geometry as your fluid, but I believe in your mind its only the "boundary" the collapsing part that is a "thin shell plate". Two ways around: either you use a shell physics, or you give the collapsing region a "thickness i.e. a "surface" and you select only that one
You do (perhaps) not need to simulate the full tube boundary just the collapsing region (leave the rest for later
4) To test it out, set a known pressure (with a numerical value) and check that the tube wall is changing shape by the load in i.e. study 2 by selecting ONLY the structural physics you are using
5) then only replace the numerical value of the pressure by the pressure of the fluid, set up study 3 to solve for BOTH spf and the structural case, this should give a reaction of the fluid on the wall
6) to now you have only the fluid reacting on the wall, but as the wall moves it should change the shape of the tube for the fluid flow, this requires the ALE physics, to make it driven by the structural and to have the SPF driven by the ALE mesh. But that becomes longish to expalin here, worth to take a few exercises and run a few of the model Exchange examples

Have fun COMSOLING

--
Good luck
Ivar
Hi There are many things to say here, first you are jumping into a 3 physics model without obviously really understanding each one, that is a really tough start, you should start with an easier example in each of the physics and then start to couple them. It takes a few years to learn structural, fluids and computational physics at the university, so do not expect to master COMSOL in a few days, but if you start simple you will quickly catch up and become an expert, I'm sure. COMSOL has also some excellent courses absolutely worth to follow, and you have several videos on the web site A few comments: 1) your v4.1 is "old" you should see with your local COMSOL rep how to upgrade to a newer patch, it would help for the solving process too 2) your spf is more or less OK, select in the study 1 only the spf "Use in this study", deselect the two others an solve the single physics. I had to reset the solver by Right clicking on Study 1- Solver Configuration - Delete Solvers and set the time stepping to (0,0.01,1) (1 sec is enough). You can also see what is happening by turning on the plots during solving: further down in the solver nodes select "Fully Coupled" and open Results while solving and select Plot 3) the Solid physics will not work like that, as it is set up for a Domain i.e. the same geometry as your fluid, but I believe in your mind its only the "boundary" the collapsing part that is a "thin shell plate". Two ways around: either you use a shell physics, or you give the collapsing region a "thickness i.e. a "surface" and you select only that one You do (perhaps) not need to simulate the full tube boundary just the collapsing region (leave the rest for later 4) To test it out, set a known pressure (with a numerical value) and check that the tube wall is changing shape by the load in i.e. study 2 by selecting ONLY the structural physics you are using 5) then only replace the numerical value of the pressure by the pressure of the fluid, set up study 3 to solve for BOTH spf and the structural case, this should give a reaction of the fluid on the wall 6) to now you have only the fluid reacting on the wall, but as the wall moves it should change the shape of the tube for the fluid flow, this requires the ALE physics, to make it driven by the structural and to have the SPF driven by the ALE mesh. But that becomes longish to expalin here, worth to take a few exercises and run a few of the model Exchange examples Have fun COMSOLING -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Mar 15, 2011, 11:24 p.m. EDT
Dear Mr. Ivar,

Thank you so much! I know I am a little rush for comsol. In fact, I already try to learn comsol for almost two months, and my professor really wants me to get something. I will do as what you said, and see how far I can go.

Thank you!
Dear Mr. Ivar, Thank you so much! I know I am a little rush for comsol. In fact, I already try to learn comsol for almost two months, and my professor really wants me to get something. I will do as what you said, and see how far I can go. Thank you!

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 16, 2011, 2:16 a.m. EDT
Hi
the best, to avoid getting frustrated, is to advance in steps, take one physics at the time, check the videos, AND after a few weeks take the first introduction COMSOL course, then continue by yourself a few week and take the more advanced courses, one by one (Ideal case, I would love to do exactly that, one can always dream ;)

--
Good luck
Ivar
Hi the best, to avoid getting frustrated, is to advance in steps, take one physics at the time, check the videos, AND after a few weeks take the first introduction COMSOL course, then continue by yourself a few week and take the more advanced courses, one by one (Ideal case, I would love to do exactly that, one can always dream ;) -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Mar 16, 2011, 5:23 p.m. EDT

Hi
the best, to avoid getting frustrated, is to advance in steps, take one physics at the time, check the videos, AND after a few weeks take the first introduction COMSOL course, then continue by yourself a few week and take the more advanced courses, one by one (Ideal case, I would love to do exactly that, one can always dream ;)

--
Good luck
Ivar


Re: Dear Mr.Ivar or others, would you give me some more help about my 2-D collapsible channel?

Dear Mr.Ivar?

Really thank you so much! Before I started my project, I already watched videos and examples. My professor really wants me to start something, I did it the whole day and last night until 3am. I still get error. As you said, I first forget about moving mesh, and just do the Laminar flow and solid mechanics, because I think they really related with each other through the pressure in the membrane. Can you check it for me again, I am so sorry that I keep bothering you, but I really feel so sad, facing to my comsol!

Thank you so so much! Thank you !
[QUOTE] Hi the best, to avoid getting frustrated, is to advance in steps, take one physics at the time, check the videos, AND after a few weeks take the first introduction COMSOL course, then continue by yourself a few week and take the more advanced courses, one by one (Ideal case, I would love to do exactly that, one can always dream ;) -- Good luck Ivar [/QUOTE] Re: Dear Mr.Ivar or others, would you give me some more help about my 2-D collapsible channel? Dear Mr.Ivar? Really thank you so much! Before I started my project, I already watched videos and examples. My professor really wants me to start something, I did it the whole day and last night until 3am. I still get error. As you said, I first forget about moving mesh, and just do the Laminar flow and solid mechanics, because I think they really related with each other through the pressure in the membrane. Can you check it for me again, I am so sorry that I keep bothering you, but I really feel so sad, facing to my comsol! Thank you so so much! Thank you !


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 16, 2011, 5:46 p.m. EDT
Hi

that looks already much better ;) a few things though:

you have deleted the material data, you should add 2 materials, one for the fluid, one for the solid/rubber/plastic ... without it will not solve (OK you have added user material data for the solid, but nothing for the fluid), always check the main steps:

0) decide on dimensions, and main physics + solver. Then define: 1) geometry, 2) material, 3) physics with its properties and BCs, 3) mesh, 4) return to recheck materials before solving, 5) set up the solvers and probes if required, 6) solving, 7)post-process, and in between some parameters, definitions, variables etc

In the SPF I suppose that boundary 6 is a wall and not aprt of the outlet (it was late yesterday you said ;)

Then your "solid" material has "0" thickness at the tube border, I would suggest that you give it also some larger thickness at these attachments, and you "fix" these small diameter thickness. i.e.e for ellipse 2, subtract "-th" on "a" and on "b". Then in the geometry, after ellipse 2, try a: union + a split and + a delete entity: domains. Then you can delete the rest of the objects not needed

Another point: you are in 2D, so the physics is normally for a "depth or in paper thickness of by default 1[m] (this applies anyhow to the fluid. Now you have define a thickness of 0.5[mm] for the structure, if that will give numerical results OK, the physics will not be fully correct, because a thin channel of 0.5[mm] "thick" and 10[mm] high will see significant wall effects (in reality) but these are not modeled in 2D as this is in the depth (Z) direction.

I would use 10[mm] depth, then one could say that the vertical and the depth values are the same (it will only have effect on total force calculations, no effect so long you couple densities

Boundary loads on the solid section: you apply an external pressure load on the top membrane side, and a difference on the bottom. This will depend on the "gauge pressure" you use for the SPF. I would say it s better to apply only fluid pressure on the bottom and use a gauge pressure that is as wanted (i.e 1[atm] for ground level atmospheric pressure), then apply a full pressure on the outside of the structural part. But this is "habits", youre way, if sorted out correctly, would probably work jsut as OK too.

--
Good luck
Ivar
Hi that looks already much better ;) a few things though: you have deleted the material data, you should add 2 materials, one for the fluid, one for the solid/rubber/plastic ... without it will not solve (OK you have added user material data for the solid, but nothing for the fluid), always check the main steps: 0) decide on dimensions, and main physics + solver. Then define: 1) geometry, 2) material, 3) physics with its properties and BCs, 3) mesh, 4) return to recheck materials before solving, 5) set up the solvers and probes if required, 6) solving, 7)post-process, and in between some parameters, definitions, variables etc In the SPF I suppose that boundary 6 is a wall and not aprt of the outlet (it was late yesterday you said ;) Then your "solid" material has "0" thickness at the tube border, I would suggest that you give it also some larger thickness at these attachments, and you "fix" these small diameter thickness. i.e.e for ellipse 2, subtract "-th" on "a" and on "b". Then in the geometry, after ellipse 2, try a: union + a split and + a delete entity: domains. Then you can delete the rest of the objects not needed Another point: you are in 2D, so the physics is normally for a "depth or in paper thickness of by default 1[m] (this applies anyhow to the fluid. Now you have define a thickness of 0.5[mm] for the structure, if that will give numerical results OK, the physics will not be fully correct, because a thin channel of 0.5[mm] "thick" and 10[mm] high will see significant wall effects (in reality) but these are not modeled in 2D as this is in the depth (Z) direction. I would use 10[mm] depth, then one could say that the vertical and the depth values are the same (it will only have effect on total force calculations, no effect so long you couple densities Boundary loads on the solid section: you apply an external pressure load on the top membrane side, and a difference on the bottom. This will depend on the "gauge pressure" you use for the SPF. I would say it s better to apply only fluid pressure on the bottom and use a gauge pressure that is as wanted (i.e 1[atm] for ground level atmospheric pressure), then apply a full pressure on the outside of the structural part. But this is "habits", youre way, if sorted out correctly, would probably work jsut as OK too. -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Mar 16, 2011, 9:56 p.m. EDT
Dear Mr. Ivar,

I don't know what word should I use to tell you how nice and helpful you are. Thank you so much! I will try my best to fix it. I feel hope and encouraged!

thank you!
Dear Mr. Ivar, I don't know what word should I use to tell you how nice and helpful you are. Thank you so much! I will try my best to fix it. I feel hope and encouraged! thank you!

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.