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.

Torsion problem simulation

Please login with a confirmed email address before reporting spam

Hello guys, I am new to COMSOL and I want to simulate torsion of circular shaft in comsol, As there is not any option available for direct torque application, I applied opposite (point) forces in opposite points on the end face equivalent to the torque. But there was high stress concentration around area near points. So can anyone tell me how to simulate this problem properly so that stress concentration will not be significant as in the case for point force? Thank you in advance


6 Replies Last Post Nov 19, 2020, 6:10 a.m. EST
Edgar J. Kaiser Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 4 years ago Sep 11, 2020, 10:00 a.m. EDT
Updated: 4 years ago Sep 11, 2020, 11:14 a.m. EDT

Hi,

one option is to apply a tangential distributed boundary load L to the cylinder face. If the cylinder axis is in x=y=0 and along the z-axis it can look like Lx = const * y, Ly = -const * x, Lz = 0. Adapt the const to achieve the desired torque. There may be other options too.

Cheers Edgar

-------------------
Edgar J. Kaiser
emPhys Physical Technology
www.emphys.com
Hi, one option is to apply a tangential distributed boundary load L to the cylinder face. If the cylinder axis is in x=y=0 and along the z-axis it can look like Lx = const * y, Ly = -const * x, Lz = 0. Adapt the const to achieve the desired torque. There may be other options too. Cheers Edgar

Please login with a confirmed email address before reporting spam

Posted: 4 years ago Sep 12, 2020, 1:32 a.m. EDT

Hi Edgar, I have cylinder of radius = 25mm, as you said earlier about constant, I have calculated its value and defined as parameter. While applying load, I have specified Fx = constrad and Fy = -constrad. But I am getting error in solution saying operator (range) is not valid in analysis. I am attaching screenshots of it. PLease have a look and help me.

Hi Edgar, I have cylinder of radius = 25mm, as you said earlier about constant, I have calculated its value and defined as parameter. While applying load, I have specified Fx = const*rad and Fy = -const*rad. But I am getting error in solution saying operator (range) is not valid in analysis. I am attaching screenshots of it. PLease have a look and help me.


Edgar J. Kaiser Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 4 years ago Sep 12, 2020, 2:57 a.m. EDT

You can try the example I gave you, it works. You can't use range() to define variables. The boundary load must be defined as a 'Force per unit area' to make it a distributed load.

-------------------
Edgar J. Kaiser
emPhys Physical Technology
www.emphys.com
You can try the example I gave you, it works. You can't use range() to define variables. The boundary load must be defined as a 'Force per unit area' to make it a distributed load.

Please login with a confirmed email address before reporting spam

Posted: 4 years ago Sep 12, 2020, 5:14 a.m. EDT

Hi Edgar, I think you forgot to upload the example file which you were to send me. Can you please send it ?

Hi Edgar, I think you forgot to upload the example file which you were to send me. Can you please send it ?

Edgar J. Kaiser Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 4 years ago Sep 12, 2020, 6:37 a.m. EDT

There isn't an example file. My posts tell what to do.

-------------------
Edgar J. Kaiser
emPhys Physical Technology
www.emphys.com
There isn't an example file. My posts tell what to do.

Please login with a confirmed email address before reporting spam

Posted: 4 years ago Nov 19, 2020, 6:10 a.m. EST

Hi, i recently work on this problem also.

If your shaft is in elastic material, i think i will be work. for boudary condition: - fixed constraint at one side and rigid connector at another side - in rigid connector, active the applied moment around the rotation of your geometry. The total torque will be distributed uniformely on your boundary surface/ edge or points. - If you want to apply a constraint rotation motion, you can define it in section of rigid connector. But since the resolution is more easier that base on applied load than applied displacement, you can profit the option of 'global equation' to determine iteratively the good load according to every steps of displacement. - if you need to achieve a large rotation angle, which means a big shear strain, the non linearity geometry will be recommanded to be activated. Then you may need to use a auxiliary study with some configuration of your solver (based on blog of comsol, the constant Newton raphson method is recommanded with a dumping coef of 0.7-0.9) . - I did keep the model for the moment, but a I can show you a gif generated by comsol 5.5. the Bar is partially shown with a 90° sector in order to affiche the rotation. the angle large more 360° is achieved with elastic mat and NL geo activated.

Hi, i recently work on this problem also. If your shaft is in elastic material, i think i will be work. for boudary condition: - fixed constraint at one side and rigid connector at another side - in rigid connector, active the applied moment around the rotation of your geometry. The total torque will be distributed uniformely on your boundary surface/ edge or points. - If you want to apply a constraint rotation motion, you can define it in section of rigid connector. But since the resolution is more easier that base on applied load than applied displacement, you can profit the option of 'global equation' to determine iteratively the good load according to every steps of displacement. - if you need to achieve a large rotation angle, which means a big shear strain, the non linearity geometry will be recommanded to be activated. Then you may need to use a auxiliary study with some configuration of your solver (based on blog of comsol, the constant Newton raphson method is recommanded with a dumping coef of 0.7-0.9) . - I did keep the model for the moment, but a I can show you a gif generated by comsol 5.5. the Bar is partially shown with a 90° sector in order to affiche the rotation. the angle large more 360° is achieved with elastic mat and NL geo activated.

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.