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 The information provided may be out of date.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.

Normal and tangential directions in comsol 4.0a

Andrew Prudil Nuclear Materials

Please login with a confirmed email address before reporting spam

Can some explain to me how the new system for formal and tangential directions works?

I'm trying to add a load to a boundary in a solid mechanics module. This was super easy to do in 3.5a, but it isn't clear to me in COMSOL 4.0a what is really happening.

In COMSOL 4.0a, when I add a boundary load, I am able to select "Boundary System 1 (sys1)" for the coordinate system. This boundary system is an item under definitions however I don't know what any of it means. I am working in a 2D axi-symmetric geometry and this window produces (t1,n,to).

When I try to apply a forces using "Boundary Load", my only options are in the 't1' and 'to' directions but no normal direction. I have been producing boundary arrow plots trying to figure out what those mean but it still in not clear.

I have been able to plot the normal vector using (nR, nZ) and (nr,nz) but there is no (n,t) option under the boundary load so I don't know how I am supposed to do this.

I have tried reading the help and documentation but it still isn't clear to me. Can anyone explain this? Does anyone have a model from 4.0a where this is clear?

12 Replies Last Post 13 giu 2012, 17:14 GMT-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: 1 decade ago 2 ott 2010, 08:15 GMT-4
Hi

I do not have the feeling it has changed from 3.5a, apart from the presentation:

if I take the 4.0a in 2D axi I see "n" for normal to the selected boundary, "t1" for the first tangeant, defined along axis "2" by default and "to" the out-of-plane tangent (perpendicular to the screen, or along "phy" the rotation angle around the 2D axi axis).

the thing is that COMSOL has axis 1, 2, 3 defined by default as x, y, z (for 3D)
but in 2D-axi it's r, z (implicit phy for the third),
while in standard 2D it's x, y and z out of the plane

--
Good luck
Ivar
Hi I do not have the feeling it has changed from 3.5a, apart from the presentation: if I take the 4.0a in 2D axi I see "n" for normal to the selected boundary, "t1" for the first tangeant, defined along axis "2" by default and "to" the out-of-plane tangent (perpendicular to the screen, or along "phy" the rotation angle around the 2D axi axis). the thing is that COMSOL has axis 1, 2, 3 defined by default as x, y, z (for 3D) but in 2D-axi it's r, z (implicit phy for the third), while in standard 2D it's x, y and z out of the plane -- Good luck Ivar

Andrew Prudil Nuclear Materials

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2 ott 2010, 10:51 GMT-4
That is helpful. But I still don't see a box to USE the normal direction for a boundary load. I've attached a picture to show what I'm talking about. The only options are to apply forces in the t1 and to directions, which as you have described are actually tangential and out of the plane directions.
That is helpful. But I still don't see a box to USE the normal direction for a boundary load. I've attached a picture to show what I'm talking about. The only options are to apply forces in the t1 and to directions, which as you have described are actually tangential and out of the plane directions.


Andrew Prudil Nuclear Materials

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2 ott 2010, 10:58 GMT-4
I looked a bit more, and it only looks like this for the 2D axially symmetric case. For both the "solid mechanics" and "thermal stress" physics modes in 2D axially symmetric the normal direction is not available.

Thoughts? I'm assuming this is a bug in the physics mode GUI? I would imagine it is supposed to look like it does for the 2D Cartesian coordinate case. Is there a way of fixing this or do I have to wait for a patch or version 4.1?
I looked a bit more, and it only looks like this for the 2D axially symmetric case. For both the "solid mechanics" and "thermal stress" physics modes in 2D axially symmetric the normal direction is not available. Thoughts? I'm assuming this is a bug in the physics mode GUI? I would imagine it is supposed to look like it does for the 2D Cartesian coordinate case. Is there a way of fixing this or do I have to wait for a patch or version 4.1?

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 3 ott 2010, 03:43 GMT-4
Hi

I do not understand why you define a pressure (by default a boundary NORMAL item on the tangeantial vector direction !, it shoudl be always on "n" for me.

I suspect that you have not catched that the boundary coordinates are not fixed in "global" space, but defined "per boundary" you select, or that COMSOL consideres.

In Comsol there are many "implicit" assumptions, i.e. all physics apply to a "finite element" hence a "dV" an elementary "small" volume. This explains why a body load or gravity is by default defined as g_const*solid.rho (where g_const is the gravity acceleration 9.81[m/s^2]). This because implicitely we define expression to be integrated over the TOTAL volume summing up all small FEM items of volume dV to get to aour final CAD geometry V.
The same for boundary loads, expressed in pressure [Pa] units as these apply to the boundary (dA) of the small volume elements (dV), hence implicitely to be integrated over the boundary surface ...

--
Good luck
Ivar
Hi I do not understand why you define a pressure (by default a boundary NORMAL item on the tangeantial vector direction !, it shoudl be always on "n" for me. I suspect that you have not catched that the boundary coordinates are not fixed in "global" space, but defined "per boundary" you select, or that COMSOL consideres. In Comsol there are many "implicit" assumptions, i.e. all physics apply to a "finite element" hence a "dV" an elementary "small" volume. This explains why a body load or gravity is by default defined as g_const*solid.rho (where g_const is the gravity acceleration 9.81[m/s^2]). This because implicitely we define expression to be integrated over the TOTAL volume summing up all small FEM items of volume dV to get to aour final CAD geometry V. The same for boundary loads, expressed in pressure [Pa] units as these apply to the boundary (dA) of the small volume elements (dV), hence implicitely to be integrated over the boundary surface ... -- Good luck Ivar

Andrew Prudil Nuclear Materials

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 4 ott 2010, 10:56 GMT-4

I do not understand why you define a pressure (by default a boundary NORMAL item on the tangeantial vector direction !, it shoudl be always on "n" for me.


Ivar,
I agree 100% that a gas pressure force generally acts normal to the boundary. The problem is that when you select the 2D axially symmetric option, there is no normal vector for me to apply it on! If you look at the picture closely, you'll see that I have selected "Boundary Frame" which should allow me the option of applying forces in the normal and tangential directions (as it does if I select 2D Cartesian) but there is only the into tangential and out of the page directions. Furthermore, if you look at the equation mode, it doesn't look complete.

I think I'm either miss understanding how something or this is a bug in the COMSOL GUI.
[QUOTE] I do not understand why you define a pressure (by default a boundary NORMAL item on the tangeantial vector direction !, it shoudl be always on "n" for me. [/QUOTE] Ivar, I agree 100% that a gas pressure force generally acts normal to the boundary. The problem is that when you select the 2D axially symmetric option, there is no normal vector for me to apply it on! If you look at the picture closely, you'll see that I have selected "Boundary Frame" which should allow me the option of applying forces in the normal and tangential directions (as it does if I select 2D Cartesian) but there is only the into tangential and out of the page directions. Furthermore, if you look at the equation mode, it doesn't look complete. I think I'm either miss understanding how something or this is a bug in the COMSOL GUI.

Andrew Prudil Nuclear Materials

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 4 ott 2010, 12:52 GMT-4
I got a very helpful private message from one of the COMSOL staff with a workaround. I figured I'd post it in case anyone finds this thread with similar problems.

I suspect this is a bug in the 2D axisymmetric case. We are investigating it.

In the meantime, there is a workaround. In the 2D-axisymmetric case the outward-pointing normal vector has components nr and nz in the r and z directions, so you can take advantage of that to specify your load, e.g. provide components -po*nr and -po*nz for your distributed pressure of magnitude po.

Notes
1/ you can plot the (nr,nz) vector using a line arrow plot to visually check the above.
2/ if you need access to the tangential vector you can build it easily: it has components -nz and nr or nz and -nr (depending which direction you want it to point).
I got a very helpful private message from one of the COMSOL staff with a workaround. I figured I'd post it in case anyone finds this thread with similar problems. [quote]I suspect this is a bug in the 2D axisymmetric case. We are investigating it. In the meantime, there is a workaround. In the 2D-axisymmetric case the outward-pointing normal vector has components nr and nz in the r and z directions, so you can take advantage of that to specify your load, e.g. provide components -po*nr and -po*nz for your distributed pressure of magnitude po. Notes 1/ you can plot the (nr,nz) vector using a line arrow plot to visually check the above. 2/ if you need access to the tangential vector you can build it easily: it has components -nz and nr or nz and -nr (depending which direction you want it to point). [/quote]

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 5 ott 2010, 02:35 GMT-4
Hi

go your point (finally), indeed loks like a bug, V4.1 is on its way for the COMSOL conf in Boston I have heard, so then it's most probably fixed.

But I have noticed now that it applies to ALL boundary coordinate systems, systematically !!
Really worth to get it fixed ;)

--
Good luck
Ivar
Hi go your point (finally), indeed loks like a bug, V4.1 is on its way for the COMSOL conf in Boston I have heard, so then it's most probably fixed. But I have noticed now that it applies to ALL boundary coordinate systems, systematically !! Really worth to get it fixed ;) -- Good luck Ivar

Andrew Prudil Nuclear Materials

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 5 ott 2010, 09:27 GMT-4
Cool. What is the word about when the rest of us will get 4.1? Is there something more definitive then "this fall"?
Cool. What is the word about when the rest of us will get 4.1? Is there something more definitive then "this fall"?

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 5 ott 2010, 10:16 GMT-4
Hi

Well traditionally we get them at the conferences, Boston is soon, Paris shortly thereafter, so probably they are burning the CD already. We will see :)

--
Good luck
Ivar
Hi Well traditionally we get them at the conferences, Boston is soon, Paris shortly thereafter, so probably they are burning the CD already. We will see :) -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 20 apr 2011, 11:10 GMT-4
How are the tangential and normal directions defined in general. I guess the normal is the outward normal. How are the two tangential directions defined?
How are the tangential and normal directions defined in general. I guess the normal is the outward normal. How are the two tangential directions defined?

Andrew Prudil Nuclear Materials

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 20 apr 2011, 12:17 GMT-4
The definitions are discussed in the documentation. Your best bet it likely to produce a boundary arrow plot using components of the vector in question. That way you can be sure that you have the appropriate normal/tangent on your geometry
The definitions are discussed in the documentation. Your best bet it likely to produce a boundary arrow plot using components of the vector in question. That way you can be sure that you have the appropriate normal/tangent on your geometry

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 13 giu 2012, 17:14 GMT-4

Hi

I do not have the feeling it has changed from 3.5a, apart from the presentation:

if I take the 4.0a in 2D axi I see "n" for normal to the selected boundary, "t1" for the first tangeant, defined along axis "2" by default and "to" the out-of-plane tangent (perpendicular to the screen, or along "phy" the rotation angle around the 2D axi axis).

the thing is that COMSOL has axis 1, 2, 3 defined by default as x, y, z (for 3D)
but in 2D-axi it's r, z (implicit phy for the third),
while in standard 2D it's x, y and z out of the plane

--
Good luck
Ivar


Really helpful. Thanks Ivar.
[QUOTE] Hi I do not have the feeling it has changed from 3.5a, apart from the presentation: if I take the 4.0a in 2D axi I see "n" for normal to the selected boundary, "t1" for the first tangeant, defined along axis "2" by default and "to" the out-of-plane tangent (perpendicular to the screen, or along "phy" the rotation angle around the 2D axi axis). the thing is that COMSOL has axis 1, 2, 3 defined by default as x, y, z (for 3D) but in 2D-axi it's r, z (implicit phy for the third), while in standard 2D it's x, y and z out of the plane -- Good luck Ivar [/QUOTE] Really helpful. Thanks Ivar.

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.