non-linear geometry 2

[chairats]

Do anyone know the meaning of non-linear geometry (Nlgeom option in the *step command). I'm trying to do finite deformation of structure on the soil.

I was succeeded in doing the normal analysis but when I put Nlgeom option into the *step command. I got the following errors in the msg file:

***ERROR: TIME INCREMENT NEEDED IS LESS THAN THE MINIMUM SPECIFIED. THE
ANALYSIS TERMINATES.

However, I use the minimum time increment equal to 0. in the static command:

*Static
10., 500., 0.,50.
**

Any suggestions will be very appreciated

Top

 

[orthofem]

Th error message you cite is a rather typical one from ABAQUS, and to diagnose the root cause often requires the specific knowledge of your area of expertise (i.e. soil mechanics).

In any nonlinear analysis, whether it involves geometric nonlinearity (NLGEOM), contact, material nonlinearity, etc., the solution is, as you know, a sequence of substeps, equilibrium iterations and, in some cases severe discontinuity iterations. When, ABAQUS detects, in any given iteration, that it is not converging to a solution, the solving algorithm will utilize a method known as bisection. The step size is cut in half, and the sequence of iterations begins again. ABAQUS has a default minimum step size (note that a zero step size has no physical meaning) and if it fails to converge at that minimum, your job will quit as you observed. You can change the default minimum step size, but it's generally better to try and understand (and fix) the physical cause of the non-convergence.

Take a good look at your status file (*.sta) to see how far along your job got before it stopped. Also look at your message file (*.msg) and search through it for ERROR messages that might help you identify the underlying cause.

Orthofem

 

[Drej]

My experience of ABAQUS tells me that this error message is usually down to prescribed loads being too high - causing the material to be stretched beyond the specified mechanical properties. I'm not 100% sure what the NLGEOM command does, other than lets ABAQUS know that your geometry is non-linear; ABAQUS then "prepares" for this by looking for non-linearities in the matrices formed (I think!).

Hope this helps,
-- D R E J --

 

[eigenvector]

Hi Chairats,
It has been a while since you asked this Q. but I am new to this forum and I will answer anyway.
As DREJ pointed out NLGEOM makes sure the analysis is second order (Geometric nonlinearity is considered
) Even I have had the problem you mention before and it was mainly due to a large time step increment. I think that your analysis did not even complete one full time step and you can check this through the .msg and .sta files. I suggest that you reduce the initial time step increment (much less than 10 you specified) and let the rest of the time steps be computed automatically. Hope this helps. Good luck,
Eigen