I am trying to bend a 100x20x1.35mm coupon, with up to 9mm displacement in a 3 point bend set up in ABAQUS. I use kinematic hardening with plastic properties. It seems when I use a lower displacement (e.g. 0.1mm) the model is able to be solved , but increasing the displacement or adding additional steps with increasing displacement seems to cause the model to abort.
I have tried mesh refinement, increased number of increments, and using multiple steps increasing the displacement, but to no avail. (Using 9mm initially gives the error messages "Time increment required is less than the minimum specified"/ "Too many steps made for this increment"/ "Unavailable foundation functionality". )
Would be grateful if anyone could provide any insight into this problem. Thanks
Increase the time increments to 1000000 and make intial 0.001 and maximum also 0.001 it can be solved.
How about decrease the initial step size? check your material parameters.
It has been four days, so perhaps you have solved this, Katherine?
I use Solidworks nonlinear and have had similar problems with the same three-point bend. Check that your supports are "fixed" and the sample bar is constrained to bend only in the plane you expect. If it has other freedoms, you may have the "free-body problem"...basically the specimen is slipping in an unexpected direction. Also check your contact definitions are appropriate - there has to be some slide of sample over supports.
The specific error messages you are getting don't match those in SW so I'm speculating. The error messages in SW are not very helpful either. I was eventually able to take the deformation to the start of push-through of specimen between the supports (start of declining force) but could not find a way to go further.
If you get this problem resolved, let us know what worked.
I have had a similar problem in SW. On my case the reason was the non linear property table. The strain was too big. I extrapoled some values to have more date and I got the problem solved. May be it works for you too.
To Alberto Naranjo - yes, that is a bug in Solidworks Nonlinear. The algorithm stops when it runs past the end of the available stress vs strain data, in the table you supply to define the material properties. The SW manual I have, claims the program extrapolates stiffness from the final two points, and I was told this in the classroom course too, but that's not what I found.
I had to manually extrapolate the stress-strain look-up table, same as you. But I still can't get my test bar to "pop" through the supports as it would in real world if pushed far enough. The available SW solvers won't handle the declining force together with contacts (this from memory, a year ago now).
None of this may be relevant to ABAQUS, but Katherine Wang might find it useful to check whether her input stress-strain curve is extensive enough.
I would very much like to hear whether ABAQUS can solve this kind of large-displacement problem with contacts. The reason is, my three point bend was just a simplification of a T-shaped intrauterine device (IUD) being pushed through a funnel-like loading device to fold the arms against the legs of the T. The loading device aids folding of the T into the insertion tube used to place the IUD through the cervix into the uterus. When I couldn't get that simulation to run far enough, I simplified to the three point bend case. My conclusion (so far) is SW Nonlinear can't do this problem.
Steve Nazar, thanks for your input and advice; I have not yet managed to solved this with the top pin, but removing it and applying velocity has managed to displace it. However any higher than 1 in the direction of loading results in abnormal load distribution. The boundary conditions and step conditions have been made according to suggestions however there are still some issues arising. I will update once I have solved the model.
Thanks for the update, Katherine.
Yes, the problem gets easier without the defined contacts - just the bridge beam and the two supports, with load applied at the center but no contact with another body there.
Perhaps those of us who have worked on this problem (in somewhat different software packages) could share some of our graphics that show the state when the algorithm crashes? I have some rather odd graphics I'm puzzling over, too.
I had put my own difficulties down to my beginner ignorance, so it is interesting others are having similar problems even using better analysis packages than mine.
I wonder too if this problem has been discussed in more depth elsewhere. It is a very obvious analysis problem.
Hi all, thanks for the suggestions and help. I have managed to solve the problem and get it to successfully displace 9mm. The trick was to apply individual boundary conditions for each defined DOF. E.g. if I am constraining the two nodes for two holding pins in all DOF (3 DOF, X,Y, Z), and constraining a node of the loading pin in two DOF (X, Y, - and not Z being the direction of load); I would create a total of 8 BCs, each with an individual BC in one direction. As opposed to 3 BC's. Hope this helps others.
Thanks, Katharine. I'm not sure what 9 mm displacement means in the context of the other dimensions. What was the stationary "pins'" span and the diameters?
If you attempt to deform further, how far will the simulation run?
My problem with IUD / loading device simulation amounts to the three-point bend case where the moving "pin" is pushed all the way through the space between the supports. In Solidworks Nonlinear I have been able to get to the stage where force is decreasing, but no further. I'm interested in whether ABAQUS would be a better tool for my problem.
- Badges
- Science topic