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.

input interior boundary force in structural mechanics

Hi:
I want to extracting reaction force on interior boundary and input it in the same boundary as an external force to let comsol to imitate bounded contact type in ansys.
My model is simple, just a 2m*1m* 1m concrete lying on the middle of a 3m*1m*0.2m steel with stress on all external boundaries of the concrete , I can sliceplot z level 0.01 to get the reaction force on the flat interior boundary, and export it as a txt file with coordinate form.Then I import the txt file into function, and start to update the model. However, the model failed in mesh initialization. Does anybody know how to successfully import a function?

6 Replies Last Post May 15, 2011, 3:03 AM EDT
Posted: 6 years ago May 10, 2011, 2:07 PM EDT
Hi

what is the difference with your case and a standard COMSOL continuous boundary condition then ?

--
Good luck
Ivar
Hi what is the difference with your case and a standard COMSOL continuous boundary condition then ? -- Good luck Ivar

Posted: 6 years ago May 11, 2011, 6:19 PM EDT
Now I have figured out they are exactly the same! I am doing a project that use same model to compare the result of comsol and ansys in some specific contact condition, such as bonded, frictionless and frcitional. During the process I have found that with same force, same structured mesh, The maxium Von mises stress I have got from comsol is usually larger than ansys while the displacement in 2 softwares is exactly the same! So I wonder if comsol and ansys may have some difference in processing or selection maxium von mises stress?

Second question, Now I am using comsol to perform frictional contact in a 2D model just like the 3D model I mentioned before. I define the static coefficient as 0.2, However the result is the same as no friction . I have viewed the sliding_wedge model in the help document and tried different friction coefficient, it seemed to work well with different displacement. However, when I changed the spring force, a displacement related force, to constant force, the displacement on the square is the same as no friction, which means no matter how large the static coefficient is, the displacement won't change. Now I am totally confused, could you please give me some advice?
Now I have figured out they are exactly the same! I am doing a project that use same model to compare the result of comsol and ansys in some specific contact condition, such as bonded, frictionless and frcitional. During the process I have found that with same force, same structured mesh, The maxium Von mises stress I have got from comsol is usually larger than ansys while the displacement in 2 softwares is exactly the same! So I wonder if comsol and ansys may have some difference in processing or selection maxium von mises stress? Second question, Now I am using comsol to perform frictional contact in a 2D model just like the 3D model I mentioned before. I define the static coefficient as 0.2, However the result is the same as no friction . I have viewed the sliding_wedge model in the help document and tried different friction coefficient, it seemed to work well with different displacement. However, when I changed the spring force, a displacement related force, to constant force, the displacement on the square is the same as no friction, which means no matter how large the static coefficient is, the displacement won't change. Now I am totally confused, could you please give me some advice?

Nagi Elabbasi Certified Consultant
Posted: 6 years ago May 12, 2011, 12:25 AM EDT
The results with and without friction can be the same if there is no tendency for the contacting bodies to slip relative to each other. You should check the tangential displacements at the two surfaces in the frictionless case. More likely however there is an input error, since it is hard to create a realistic frictionless contact problem where there is no relative tangential slip.

Nagi Elabbasi
Veryst Engineering
The results with and without friction can be the same if there is no tendency for the contacting bodies to slip relative to each other. You should check the tangential displacements at the two surfaces in the frictionless case. More likely however there is an input error, since it is hard to create a realistic frictionless contact problem where there is no relative tangential slip. Nagi Elabbasi Veryst Engineering

Posted: 6 years ago May 12, 2011, 5:23 AM EDT
I have set the frictionless contact for the model, and the result is the same as that in ansys. I have a normal pressure on the vertical surface of the contact region. The contact surface of square would deform under the pressure, which leads to a slide of partial surface against the rigid surface of steel. The friction coefficient is against this slip trend. Thus the contact surface would deform more slightly than other position without friction force. And the square deforms into a trapezium. However, in comsol no matter how I adjust the static coefficient of friction, the square deforms into a smaller square, just like a frictionless contact. I am wondering why comsol would neglect the friction coefficient.
I have set the frictionless contact for the model, and the result is the same as that in ansys. I have a normal pressure on the vertical surface of the contact region. The contact surface of square would deform under the pressure, which leads to a slide of partial surface against the rigid surface of steel. The friction coefficient is against this slip trend. Thus the contact surface would deform more slightly than other position without friction force. And the square deforms into a trapezium. However, in comsol no matter how I adjust the static coefficient of friction, the square deforms into a smaller square, just like a frictionless contact. I am wondering why comsol would neglect the friction coefficient.

Posted: 6 years ago May 14, 2011, 6:25 PM EDT
Now I have figure out why I could not solve the frictional contact problem . And I wish to share my solution. I have not set the parametric solver in the solver setting! To solve a frictional contact problem, it is essential to use parametric solver to increase the force incrementally. For example,the boundary load is set as F*para and para is defined as 0.1 0.3 0.6 0.8 1 in the parametric solver to increase the load from0.1F to F. Moreover, manual scaling and relative tolerance should be set carefully to solve the nonlinear problem successfully.(sometimes the solution time is challenging your level of patience. To get the result in a simple 3D model, the solver runs for as long as 5 hours! while ansys only needs several minutes.I wish further improvement in the frictional model of comsol.)The sliding wedge model in the structural mechanics really helps me a lot.
Now I have figure out why I could not solve the frictional contact problem . And I wish to share my solution. I have not set the parametric solver in the solver setting! To solve a frictional contact problem, it is essential to use parametric solver to increase the force incrementally. For example,the boundary load is set as F*para and para is defined as 0.1 0.3 0.6 0.8 1 in the parametric solver to increase the load from0.1F to F. Moreover, manual scaling and relative tolerance should be set carefully to solve the nonlinear problem successfully.(sometimes the solution time is challenging your level of patience. To get the result in a simple 3D model, the solver runs for as long as 5 hours! while ansys only needs several minutes.I wish further improvement in the frictional model of comsol.)The sliding wedge model in the structural mechanics really helps me a lot.

Posted: 6 years ago May 15, 2011, 3:03 AM EDT
Hi

Interesting, what you are reporting here
But I see no reason why COMSOL should not solve the model in a few minutes also.
I agree it's worth investigating. For you it could also be worth to ask "support" then you will get the clear response how to get the solution also quickly with COMSOL.
Pls report it back, as it is of general interest for us out here, I believe

--
Good luck
Ivar
Hi Interesting, what you are reporting here But I see no reason why COMSOL should not solve the model in a few minutes also. I agree it's worth investigating. For you it could also be worth to ask "support" then you will get the clear response how to get the solution also quickly with COMSOL. Pls report it back, as it is of general interest for us out here, I believe -- Good luck 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.