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.

Beam Cross Section (bcs) interface returns odd results

Please login with a confirmed email address before reporting spam

I was trying out the Beam Cross Section (bcs) interface in 2D to obtain some data I need for my graduation project.
I need to calculate Shear Correction Factor (ky), Max Shear Stress Factor (muy) Torsional constant (J) and Torsional Section Modulus(Wt) for a non-standard cross section.
To try out the interface I imported the geometry of a pipe with outer diameter 25.4mm and 2mm wall thickness, which is easy to calculate from formulas and so I could compare the results and check if I was doing it right. The formulas are found in COMSOLs user guide for the structural mechanics module.
However, the returned data for Torsional Section Modulus and Shear Correction Factor ar far from what is expected for that cross section: Ky should be around 1.9 but returns 0.5 and Wt should be in the 1.5e-6 order but returns 9.5e-8. Other data like area, moment of inertia, section height return correct numbers.
I attach the file I'm using if you can check if there´s any problems with the geometry, since using solid cross sections returns expected values (ex: a solid circular bar or square bar), and hollow cross sections return weird values (ex: hollow rectangular cross section or hollow round section).
Hope all this makes sense. I also attach a file with the beam interface were you can see the values COMSOL itself calculated on another study in table 3. Just in case I'm doing something worng ;) . Hope you can help me out.


4 Replies Last Post Nov 21, 2016, 10:26 a.m. EST
Henrik Sönnerlind COMSOL Employee

Please login with a confirmed email address before reporting spam

Posted: 8 years ago Nov 20, 2016, 10:56 a.m. EST
Hi,

A quote from the doc which I hope will solve your problem:

"If the section is not simply connected, add one Hole node for each internal hole. In that node, select all boundaries around the hole."

Regards,
Henrik
Hi, A quote from the doc which I hope will solve your problem: "If the section is not simply connected, add one Hole node for each internal hole. In that node, select all boundaries around the hole." Regards, Henrik

Please login with a confirmed email address before reporting spam

Posted: 8 years ago Nov 20, 2016, 4:09 p.m. EST
Hi, thanks for your help.

With the suggested change, the Wt value is now correct (1596.3mm^3) consistent with User Guide formulas, but still, Ky remained the same 0.5 and should be:

Ky = 1 + di/do = 1 + 21.4mm/25.4mm = 1.84 very far away from formula value

and Shear Stress Factor uy:

uy = A*(do^3-di^3)/[12*(do-di)*Izz] = 147.03mm^2*(25.4^3mm^3-21.4^3mm^3)/[12*(4mm)*10137mm^4] = 1.9903 and the BCS returns 2.1306 which is pretty close but still makes me wonder if it's trust worthy.

From what I see in the formula, Ky could never be less than 1, am I correct?

Thanks again for your help, have a great day,

Alfredo

**NOTE: I tried attaching the file in my home computer but it gives me a "file extension error", so if you need it I could try again tomorrow in a computer with COMSOL installed in my university.
Hi, thanks for your help. With the suggested change, the Wt value is now correct (1596.3mm^3) consistent with User Guide formulas, but still, Ky remained the same 0.5 and should be: Ky = 1 + di/do = 1 + 21.4mm/25.4mm = 1.84 very far away from formula value and Shear Stress Factor uy: uy = A*(do^3-di^3)/[12*(do-di)*Izz] = 147.03mm^2*(25.4^3mm^3-21.4^3mm^3)/[12*(4mm)*10137mm^4] = 1.9903 and the BCS returns 2.1306 which is pretty close but still makes me wonder if it's trust worthy. From what I see in the formula, Ky could never be less than 1, am I correct? Thanks again for your help, have a great day, Alfredo **NOTE: I tried attaching the file in my home computer but it gives me a "file extension error", so if you need it I could try again tomorrow in a computer with COMSOL installed in my university.

Henrik Sönnerlind COMSOL Employee

Please login with a confirmed email address before reporting spam

Posted: 8 years ago Nov 21, 2016, 7:31 a.m. EST
Hi,

The correct (approximate) shear correction factor is (1 + di/do)^(-1), so the current expression in the theory section of the Beam interface is erroneous. Thanks for pointing that out. There are actually even better approximations, but they depend not only on the geometry, but also on the material data (Poisson's ratio).

If you look at all other cross sections in the documentation, you will see that the values are <1 as expected.

The values that you get from the Beam Cross Section interface are 'exact' according to the theory, down to discretization errors, whereas the values that you see for the Common Sections are based on approximate formulae. The difference between 1.9903 and 2.1306 shows this effect.

Regards,
Henrik

Hi, The correct (approximate) shear correction factor is (1 + di/do)^(-1), so the current expression in the theory section of the Beam interface is erroneous. Thanks for pointing that out. There are actually even better approximations, but they depend not only on the geometry, but also on the material data (Poisson's ratio). If you look at all other cross sections in the documentation, you will see that the values are

Please login with a confirmed email address before reporting spam

Posted: 8 years ago Nov 21, 2016, 10:26 a.m. EST
Hi,

I had initially thought the difference in uy was due to the difference in solution methods, but the huge error in ky made me wonder....

So, 1.84^-1 = 0.54 ~ 0.5 of the beam interface.

Thank you so, so, very much for your help, have a great day,

Alfredo
Hi, I had initially thought the difference in uy was due to the difference in solution methods, but the huge error in ky made me wonder.... So, 1.84^-1 = 0.54 ~ 0.5 of the beam interface. Thank you so, so, very much for your help, have a great day, Alfredo

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.