User login

Navigation

You are here

Geometric nonlinearity

I am trying to write my own code for geometrically nonlinear analysis of mixed structures (cables, trusses, beams...). I would like to have it as general as possible ie. applicable also to 2D and 3D solids. I am trying to follow nonincremental Newton Raphson procedure from Ansys theory manual, but my computation are not converging.. Can I kindly ask you for advice ?

Here is my flow:

=== Linear analysis ===

(1) inv Kt *F = du

resluts are correct comparing to Ansys

=== Nonlinear analysis ===

while (F_ext - F_int) < 1e-3

(2) u = u +du

(3) computing new transformation matrix T and new tangent stiffness matrix Kt ( = Ke + Kg) in global coordinate system based on coordinates updated by du

(4) computing nodal vector of internal forces using only elastic stiffness matrix from (3): Ke * u = F_int

(5) computing residulas: F = F_ext - F_int

(6) computing increment of deformations using tangent stiffness from (3) and residual forces from (5): inv Kt * F = du

(7) go to (2)

end

 

Is my flow correct ? Can you plese advise me how to improve it, or make it correct and general ?

Comments

hello   peto24ap

Why do you use line 4 method for computing internal force? I think this is the problem. Use ordinary method for internal force, Just like the    method introduced in  " introduction to finite elements in engineering" ,by chandrupatla or every other fnite element book, to make your internal force vector.

look for this article "COMPOSITE MATERIALS IN  COMPRESSION " by  Jens Lycke Wind, In this thechnical report you can find an algorithm for Newton raphson.

I used it before for load base problem but i could not write a code for material nonlinearity which is under displacement BC, and displacement control.

Ask me if there is some thing to ask, our ways are similar. maybe we can help each other in the modeling

M.J

Subscribe to Comments for "Geometric nonlinearity"

Recent comments

More comments

Syndicate

Subscribe to Syndicate