Menu

Discussion Forum

New Discussion
Filtra

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 The information provided may be out of date.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.

displacement at a node as a variable?

Posted 9 apr 2012, 03:07 GMT-4 Structural Mechanics Version 4.2 4 Replies

Lakshmi Selvakumaran
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Hi,

I am trying to model a composite and I have problems when i try to use Periodic boundary conditions and your help would be much appreciated. I would like to enforce periodic boundary conditions between two boundaries excluding the first node and would like to add the value of the displacement of the first node to it.
The formulation can be found in a poster available in this link. I am trying to reproduce the same and am unsure how to go about it.

www2.ul.ie/pdf/741534352.pdf

So my question here is:

is it possible to 1) specify these conditions only for specific nodes rather than the whole boundary and 2) use the displacement at a point as a variable within the model?

Thanks.

Best
Lakshmi.
4 Replies Last Post 9 apr 2012, 09:18 GMT-4
Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 9 apr 2012, 04:03 GMT-4
Hi

if you define a boundary, i.e. edge i 2D as you domain to apply a condition, then you can modulate the BC value along the edge with the variable "s" (going from 0 to 1 along the arrow direction, turn on 2D details graphics label view), or you might se the x,y,z coordinates to modulate your BC value.

Be aware that point wise loads (apart from 1D nodels) are mostly singularities (you cannot precisely define the derivatives, hence the solver has issues with the Jacobian derivation) and should b eavoided, excet for specia cases and when you really understand the effects involved

COMSOL assumes loads on domain and boundaries,and only exceptionally on lower dimension entities, but its fully allowed to "load" lower dimansional entity

--
Good luck
Ivar
Hi if you define a boundary, i.e. edge i 2D as you domain to apply a condition, then you can modulate the BC value along the edge with the variable "s" (going from 0 to 1 along the arrow direction, turn on 2D details graphics label view), or you might se the x,y,z coordinates to modulate your BC value. Be aware that point wise loads (apart from 1D nodels) are mostly singularities (you cannot precisely define the derivatives, hence the solver has issues with the Jacobian derivation) and should b eavoided, excet for specia cases and when you really understand the effects involved COMSOL assumes loads on domain and boundaries,and only exceptionally on lower dimension entities, but its fully allowed to "load" lower dimansional entity -- Good luck Ivar
Lakshmi Selvakumaran
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 9 apr 2012, 05:50 GMT-4
Thanks a lot Ivar. Just to further clarify, could I use periodic boundary conditions on the x, y and z boundaries and limiting the application to s nodes along the boundary? and could I specify the displacement at a node by specifying it as u(x,y,z) where (x,y,z) represents the co-ordinates?

Thanks again.
Thanks a lot Ivar. Just to further clarify, could I use periodic boundary conditions on the x, y and z boundaries and limiting the application to s nodes along the boundary? and could I specify the displacement at a node by specifying it as u(x,y,z) where (x,y,z) represents the co-ordinates? Thanks again.
Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 9 apr 2012, 07:54 GMT-4
Hi

normally you can also do this on pairs an pariodic conditions, at least I do not see why not, provided you know what you are doing ;) (I do not have a clear example of exactly what you are trying).

on the other side, take care to understand corectly the dependnet "variables" u,v,w in fact these are fields and do implicitely vary as u(x,y,z,t) or r whatever is applicable w.r.t. your model settings. But the (x,y,z) is implicit and arrive via the entity selections.
1
Therefore if you select an edge (boundary in 2D) and want the i.e. intop1(u) along this edge, but not for lts say s<0.1 then you can write intop1(u*(s>0.1). take care with the aveop() as this one normalises over the full edge length and would give you

AVERAGE = aveop1(u*(s>0.1)) = intop1(u*(s>0.1)) / intop1(1)

which IS NOT EXACTLY the same as fully written out

AVE = intop1(u*(s>0.1)) / intop1(1*(s>0.1))

Hope you catch the subtility ;) in anycase just try it out on a simple model i.e. HT or solid

--
Good luck
Ivar
Hi normally you can also do this on pairs an pariodic conditions, at least I do not see why not, provided you know what you are doing ;) (I do not have a clear example of exactly what you are trying). on the other side, take care to understand corectly the dependnet "variables" u,v,w in fact these are fields and do implicitely vary as u(x,y,z,t) or r whatever is applicable w.r.t. your model settings. But the (x,y,z) is implicit and arrive via the entity selections. 1 Therefore if you select an edge (boundary in 2D) and want the i.e. intop1(u) along this edge, but not for lts say s0.1). take care with the aveop() as this one normalises over the full edge length and would give you AVERAGE = aveop1(u*(s>0.1)) = intop1(u*(s>0.1)) / intop1(1) which IS NOT EXACTLY the same as fully written out AVE = intop1(u*(s>0.1)) / intop1(1*(s>0.1)) Hope you catch the subtility ;) in anycase just try it out on a simple model i.e. HT or solid -- Good luck Ivar
Lakshmi Selvakumaran
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 9 apr 2012, 09:18 GMT-4
Thanks a lot Ivar. I will try it out.


Thanks a lot Ivar. I will try it out.

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.

Suggested Content