Hi,
I am trying to analyze welds in ansys by modeling the weld as couplings (in a shell model) and postprocessing the data in an excel file to calculate stresses in the weld
In the post processing, I select each node (from a set of coupling having 2 nodes, I select one node) and write out the Forces (Fx, Fy and Fz) on the node.
So I have a text file which has node numbers with location and forces Fx, Fy and Fz at the node.
Now using these forces at each node, I analyze what forces would cause shear in the weld and what force will cause tension/compression. The mesh is uniform at the couplings meaning that the element length is same throughout which means when I calculate the stress I use the force and divide it by length of element and throat of weld.
I find the shear stress (tau1 and tau2) - two directions causing shear of weld, and a normal stress (sigma) in the third direction.
Now to get the Von-mises stress, I use the formula sqrt [sigma^2+3*(tau1^2+tau2^2)] and compare with yield strength whereas I compare shear stress with the sy/sqrt(3) (0.57Sy).
Is the approach reasonable??
Shiraz
Sr. Engineer
I am trying to analyze welds in ansys by modeling the weld as couplings (in a shell model) and postprocessing the data in an excel file to calculate stresses in the weld
In the post processing, I select each node (from a set of coupling having 2 nodes, I select one node) and write out the Forces (Fx, Fy and Fz) on the node.
So I have a text file which has node numbers with location and forces Fx, Fy and Fz at the node.
Now using these forces at each node, I analyze what forces would cause shear in the weld and what force will cause tension/compression. The mesh is uniform at the couplings meaning that the element length is same throughout which means when I calculate the stress I use the force and divide it by length of element and throat of weld.
I find the shear stress (tau1 and tau2) - two directions causing shear of weld, and a normal stress (sigma) in the third direction.
Now to get the Von-mises stress, I use the formula sqrt [sigma^2+3*(tau1^2+tau2^2)] and compare with yield strength whereas I compare shear stress with the sy/sqrt(3) (0.57Sy).
Is the approach reasonable??
Shiraz
Sr. Engineer
