Internal General beam Element

disladisla
Hello,

I’m comparing the input values required in the Internal General beam Element and the ccx 3D Timoshenko beam element U1.

Ccx is using an advanced coded version of Timoshenko Beam Element (U1/B3D2) [109]
That implementation has implicit that the beam cross-section is symmetrical, or the element local coordinate axes are selected to pass through the cross-section shear center.
Timoshenko shear coefficient accounts for the different shapes of beam cross-sections. (I assume I11 and I22 should then be given with respect to the shear center).

Ccx general beam is defined by I11, I12 ,I22, A, k (Timoshenko shear coefficient)

¿What about the internal general beam element?

Internal general beam is defined by I11, I22, A, J (Torsion constant) and Perimeter
¿Is it the same formulation but just written in terms of J ?. ¿Is the general section limited to symmetrical cross sections?
¿Are you assuming rectangular cross section as shown in the GUI picture?


Thanks in advance.

[109] Yunhua, Luo, An Efficient 3D Timoshenko Beam Element with Consistent Shape Functions. Adv. Theor. Appl. Mech. 1 , no. 3, 95-106 (2008).



Comments

  • VictorVictor
    The internal solver's beams assume shear center = centroid, so they're not really correct for asymmetric sections. The general section doesn't necessarily correspond to any rectangle though because you can define I and A in a way that isn't a rectangle.

    Perimeter is only used for some kinds of load and isn't part of the stiffness.
  • disladisla
    Thank you Victor.
  • disladisla
    Hi,

    Following the above discussion,

    If Perimeter value is not involved in the stiffness matrix, would I be wrong if I assume that value as 1 [m2/m] by default so I can indistinctly convert in-between linear load (KN/m) and Traction load (KN/m2)?.

    I see in the manual that heat fluxes, convection, radiation are Perimeter dependent, but I’m not interested in thermal analysis right now.

    If assuming P=1 [m2/m] by default could affect other results, I would kindly ask to implement linear loads KN/m as a function of position in the internal solver.

    Thanks in advance and regards
  • VictorVictor
    I think that should be OK but be careful of any other loads or constraints applied to those beam 1D faces since many of them use surface area which is obtained from perimeter. For example, elastic support uses it to weight the springs, RBE3 uses it to weight the MPCs.

    The display also uses perimeter to set the clipping planes so if it's wildly different from the scale of the model, it might look a bit glitchy.
  • disladisla
    Thank you Victor.
Sign In or Register to comment.

Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

Sign In Register