Hello,
I would like to compute de the stress tensor of a Timoshenko beam at its Gauss points, to be able to implement an elastoplastic law in my finite element calculations.
Firstly,I know the displacement field at any point of my beam thanks to the relation u(x) = N(x) U, where U is the matrix of degrees of freedom at the nodes of my beam tU = (ux1, uy1 , uz1, θx1, θy1, θz1, ux2, uy2, uz2, θx2, θy2, θz2)
Then, I took as an expression of N the form given in this article https://www.researchgate.net/publication/236659875_Shape_functions_of_three-dimensional_Timoshenko_beam_element#fullTextFileContent , which corresponds to a Timoshenko model.
I deduce the deformations for small strains with ε = 1/2 (grad(u) +tgrad(u)), I obtained the equation shown in the picture.
I then apply Hooke's law to find the stress.
I then obtain that for a traction test (ux2 = constant, the other components of U are zero), the displacement field and the strain tensor are constant on my beam in particular along a cross-section, with only εxx non-zero, on the other hand the stress tensor has non-zero components other than σxx.
I conclude that my model shows that the cross sections are non-deformable, with therefore additional "virtual" forces, which prevent the beam subjected to traction along x, from being refined along y and z in accordance with the Poisson effect . On the other hand, I would like to have a "natural" behavior where the beam is refined according to y and z.
Do you have any articles for this?
Thanks a lot