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.

What is the difference between "Rigid Connector" and "Prescribed Displacement" boundary condition in Solid Mechanics Comsol?

Please login with a confirmed email address before reporting spam

To me, "Rigid Connector" and "Prescribed Displacement" look the same at first but when I looked at the Rigid connector equation I see the formula for moment of inertia or center of mass and I see no formula for the prescribed displacement. Could anybody explain in what situations we should use which one?

4 Replies Last Post 31 mai 2016, 02:33 UTC−4
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: 9 years ago 30 mai 2016, 17:25 UTC−4
Hi again ;)

These are very different. The Prescribed Displacement is a general boundary Condition where you impose a given displacement to some or all DoFs i.e. a "Fixed BC" is the same as a Prescribed Displacement with all three displacements active and set to "0".

A Rigid Connector makes the selected surfaces FULLY RIGID (with no deformation) and you define a "Center of Rotation" (by default in the middle of the selected boundary, but you can place it manually anywhere) and you apply displacements and forces and moments TO THE CENTRE OF ROTATION, then these are distributed onto the Rigid Connector Surfaces. A RC BC is often used as an interface (Attachment) to a Rigid Body model, or to a Shell, or as a simple point load mass/force/Moment load.

A few limitations:
the RC mass is not considered for eigenfrequency analysis (was not previously might come once in newer versions I still hope ;)
The RC mass is not considered when applying a Body load such as gravity load, you must manually add a RC Force with the corresponding mass. A gravity Body load does take into account an "added mass BC" though another entry you will find in the BC list.

To test out such features, use toy-models.

For me, in my time zone, its getting rather late, so I'll stop here :)

By the way COMSOL proposes excellent courses the 2 day intro, and then the 1 day for the different modules, absolutely worth to follow to get a cick-start !

And by the way since your are new here, I understand, for your info, I'm not by COMSOL just a convinced user :)

--
Have fun COMSOLing
Ivar
Hi again ;) These are very different. The Prescribed Displacement is a general boundary Condition where you impose a given displacement to some or all DoFs i.e. a "Fixed BC" is the same as a Prescribed Displacement with all three displacements active and set to "0". A Rigid Connector makes the selected surfaces FULLY RIGID (with no deformation) and you define a "Center of Rotation" (by default in the middle of the selected boundary, but you can place it manually anywhere) and you apply displacements and forces and moments TO THE CENTRE OF ROTATION, then these are distributed onto the Rigid Connector Surfaces. A RC BC is often used as an interface (Attachment) to a Rigid Body model, or to a Shell, or as a simple point load mass/force/Moment load. A few limitations: the RC mass is not considered for eigenfrequency analysis (was not previously might come once in newer versions I still hope ;) The RC mass is not considered when applying a Body load such as gravity load, you must manually add a RC Force with the corresponding mass. A gravity Body load does take into account an "added mass BC" though another entry you will find in the BC list. To test out such features, use toy-models. For me, in my time zone, its getting rather late, so I'll stop here :) By the way COMSOL proposes excellent courses the 2 day intro, and then the 1 day for the different modules, absolutely worth to follow to get a cick-start ! And by the way since your are new here, I understand, for your info, I'm not by COMSOL just a convinced user :) -- Have fun COMSOLing Ivar

Please login with a confirmed email address before reporting spam

Posted: 9 years ago 30 mai 2016, 18:23 UTC−4
Thank you, Ivar. I know you are not from Comsol, but I consider you Comsol HERO. Almost all the posts in Comsol that at least I visit, are answered by you. I wish you were the Author of the Comsol manual. Comsol manual looks pleasing but not very functional in my opinion. Too many inconsistencies and somehow vague in some parts.
In my opinion, they should publish the manual online so that people can search online rather than too many pdf files that should be searched separately.
Those pdf files are good for those who work every day with Comsol, not me that just use it few times a year.
I have toy models but they are not multiphysics just one solid mechanic physics!! Do you have a toy model that have many physics inside it and also willing to share? :)
Thank you, Ivar. I know you are not from Comsol, but I consider you Comsol HERO. Almost all the posts in Comsol that at least I visit, are answered by you. I wish you were the Author of the Comsol manual. Comsol manual looks pleasing but not very functional in my opinion. Too many inconsistencies and somehow vague in some parts. In my opinion, they should publish the manual online so that people can search online rather than too many pdf files that should be searched separately. Those pdf files are good for those who work every day with Comsol, not me that just use it few times a year. I have toy models but they are not multiphysics just one solid mechanic physics!! Do you have a toy model that have many physics inside it and also willing to share? :)

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: 9 years ago 31 mai 2016, 02:04 UTC−4
Hi

Get a good indexer (I'm, using the crappy MS one but it works, had much more success with the Google one a decade ago, but it has gone too expensive for me, I invest in COMSOL modules ;)

I would rather say that their documentation is complete but there are so many things to say and so many entries, depending on what one is doing that I find it difficult to suggest another way to write their PDF's.
One thing to say with the COMSOL developers, that I have not found by others, is that if you argument clearly, with examples illustrating your approach and your suggestions to simplify some actions, they do listen and you find many things in the next release.
But COMSOL's developers must also take the large image, other users needs different approaches, or even use the software in a different way and COMSOL must remain compatible for any type of physics interactions, so we users must also adapt to that.
COMSOL documentation is sometimes oriented to show the commonalities, history over the last century has made scientists to transform the same PDE slightly so that each physicist could put his name on the Formula, even if it was a simple transform of an existing one, I have discovered many of these, to human tricks to get famous, since I started to use COMSOL for Multiphysics, therefore the theory is some times presented slightly differently from classical text books, but its to highlight the commonalities (simple example: Chemical Diffusion theory and Heat transfer, same PDE but different names and formulations for the same difficulties to set up a model that converges correctly, if you know one and see the commonalities you know the other one, same methodology same approach!
Anyhow taking a step back may give us other ways to approach a model and new insights. That is why I like so much the COMSOL conference, I would only appreciate that the presenters give more example how they set up the program to get their results, rather than to concentrate on the originality of their final results, and including distributing "toy models" illustrating their COMSOL approach.

For me lecturer or the COMSOL courses are where you learn and interact the most, the lecturer are all advanced COMSOL users if not developers, (including their workshops, webinars and COMSOL Days ...) and by interacting at the COMSOL conferences, these are coming up soon (I'll do everything I can to be in Munich this autumn ;)

Using COMSOL or any advanced FEM and modeling program only a few times per year is terribly difficult, there are too man things/settings implications ... to remember, but the general multi-physics is something one can learn und use everyday, even without COMSOL.

I make several toy models per hour, but these are simple tests, with one, two or more physics, before I attack a major model, but these are very specific and not documented so not really useful outside my short term need. These toy models are also their to validate my approaches, are global energy preserved, have I combined correctly the BC, and not forgotten one ...
Another issue I have, as you I read, is that I do mostly confidential applied research for external clients, and my "toy-models" are too clearly indicating what I'm doing, these are difficult to give away like that, I would need to leave them alone a couple of years and come back to them later, but I have thousands of models and no time to document them in detail or sort them really... perhaps something for my activity when the company tells me it's my time off, formal retirement is approaching, in a few year ;)

--
Good luck
Ivar
Hi Get a good indexer (I'm, using the crappy MS one but it works, had much more success with the Google one a decade ago, but it has gone too expensive for me, I invest in COMSOL modules ;) I would rather say that their documentation is complete but there are so many things to say and so many entries, depending on what one is doing that I find it difficult to suggest another way to write their PDF's. One thing to say with the COMSOL developers, that I have not found by others, is that if you argument clearly, with examples illustrating your approach and your suggestions to simplify some actions, they do listen and you find many things in the next release. But COMSOL's developers must also take the large image, other users needs different approaches, or even use the software in a different way and COMSOL must remain compatible for any type of physics interactions, so we users must also adapt to that. COMSOL documentation is sometimes oriented to show the commonalities, history over the last century has made scientists to transform the same PDE slightly so that each physicist could put his name on the Formula, even if it was a simple transform of an existing one, I have discovered many of these, to human tricks to get famous, since I started to use COMSOL for Multiphysics, therefore the theory is some times presented slightly differently from classical text books, but its to highlight the commonalities (simple example: Chemical Diffusion theory and Heat transfer, same PDE but different names and formulations for the same difficulties to set up a model that converges correctly, if you know one and see the commonalities you know the other one, same methodology same approach! Anyhow taking a step back may give us other ways to approach a model and new insights. That is why I like so much the COMSOL conference, I would only appreciate that the presenters give more example how they set up the program to get their results, rather than to concentrate on the originality of their final results, and including distributing "toy models" illustrating their COMSOL approach. For me lecturer or the COMSOL courses are where you learn and interact the most, the lecturer are all advanced COMSOL users if not developers, (including their workshops, webinars and COMSOL Days ...) and by interacting at the COMSOL conferences, these are coming up soon (I'll do everything I can to be in Munich this autumn ;) Using COMSOL or any advanced FEM and modeling program only a few times per year is terribly difficult, there are too man things/settings implications ... to remember, but the general multi-physics is something one can learn und use everyday, even without COMSOL. I make several toy models per hour, but these are simple tests, with one, two or more physics, before I attack a major model, but these are very specific and not documented so not really useful outside my short term need. These toy models are also their to validate my approaches, are global energy preserved, have I combined correctly the BC, and not forgotten one ... Another issue I have, as you I read, is that I do mostly confidential applied research for external clients, and my "toy-models" are too clearly indicating what I'm doing, these are difficult to give away like that, I would need to leave them alone a couple of years and come back to them later, but I have thousands of models and no time to document them in detail or sort them really... perhaps something for my activity when the company tells me it's my time off, formal retirement is approaching, in a few year ;) -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 9 years ago 31 mai 2016, 02:33 UTC−4
Thank you Ivar for clarification, In comsol conferences do they share the files for their models? I am asking because I could only find the pdf of presentations of previous conferences in the comsol website but not the files.
Regarding your example of "...Chemical Diffusion theory and Heat transfer, same PDE..." you are absolutely right, and I noticed that too. For so many years we should learn the Euler name for any type of physics and equation and math ......Even the comsol charge us for different packages that have the same PDE ;)
I personally have played with Abaqus and Ansys too. I like Comsol because I see equations and the interface are so much easier to learn and more attractive. In industry they only like Ansys for multiphysics and talking about comsol is an insult!
About the toy model, I totally understand :)
Thanks again
Thank you Ivar for clarification, In comsol conferences do they share the files for their models? I am asking because I could only find the pdf of presentations of previous conferences in the comsol website but not the files. Regarding your example of "...Chemical Diffusion theory and Heat transfer, same PDE..." you are absolutely right, and I noticed that too. For so many years we should learn the Euler name for any type of physics and equation and math ......Even the comsol charge us for different packages that have the same PDE ;) I personally have played with Abaqus and Ansys too. I like Comsol because I see equations and the interface are so much easier to learn and more attractive. In industry they only like Ansys for multiphysics and talking about comsol is an insult! About the toy model, I totally understand :) Thanks again

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.