Simple Frame analysis not working! I know this is simple!

[kc97]

Hello...
I am new to ABAQUS and have a fairly straight-forward probelm (I hope!). i'm trying to model a simple 2D frame made of steel I-sections. I initially modelled it with rectangular sections without difficulty. However, when I changed the sections to I-Sections, it wont run! I have tried changing iteration parameters, to no avail. Section orierntation is (0,0,-1). Any other tips? It is pinned at the bases and has a signle point load applied.

I'd greatly appreciate any insigiht!

Thanks,
K

 

[OptiEng]

hi Kc97, When you say, it won't run, what do you mean? What are the error messages that come up in your monitor window?

 

[kc97]

Hello Opti...thanks for replying.

The error message reads:

"Too many attempts made for this increment

Abaqus/Standard Analysis exited with an error - Please see the message file for possible error messages if the file exists."

And the .msg file says:
" ***WARNING: SOLVER PROBLEM. NUMERICAL SINGULARITY WHEN PROCESSING NODE
FRAME-1.1378 D.O.F. 6 RATIO = 9.63575E+009."

Thanks

 

[OptiEng]

Is this the same node you have applied your concerntrated load on? This might be causing a problem reaching a solution. Are you considering geometric non-linearity (i.e. selecting the 'nlgeom' in your step module)? How fine is your mesh, in the areas of importance? Hope this helps.

 

[kc97]

Every node displays this error (i.e. about numerical singularity). Not just the one that load is applied to. In any case, there is no issue when I use a rectangular section.

I am using what I think is a very fine mesh (elements are 50mm long throughout the frame, whcih is 7450mm high and 30,000mm wide).

Thanks for any advice you can give!

I was wodnering if it might be something to do with the orientation?

K

 

[OptiEng]

I am not an expert on numerical singularities, but I know they are common when modelling crack problems and can lead to convergence issues. If you can post a cae file or picture, maybe we can offer some more advice?

 

[tgrimley]

I've also had it caused by rigid body motion. Does it also warn about unconnected regions?

 

[corus]

Depends on the element type, you might have pinned trusses instead of rigid beams and so the thing just collapses.

Tata

 

[kc97]

Hello,

No I have beam elements (B22 or B21 depending on the analysis) throughout the frame.

Thanks,
K

 

[corus]

Section properties missing so it has no stiffness?

Tata

 

[kc97]

No I have included section properties for an I-section (h, b1, b2, I, etc....)

This is the problem...it works when the section properties are for a rectangle!

THanks for time! I really appreciate it...