I'm trying to solve a beam bending problem with large deformations and preloaded state.

Model problem:

A beam, 1 meter long, of circular cross section is terminated at both ends, the load is applied at the middle of the beam.

The first step is to apply the load and unload the deformed state (inistate)

The second step is to change the geometry (upgeom) and load the deformed state.

The solutions in the first and second step should be identical.

Result:

The solutions in the first and second steps diverge significantly for small loads.

As the loads increase, the solutions begin to converge.

What's scary is the error (last picture) in the results on the second step - as if the file with the results of step 1 is being read incorrectly.

However, when nlgeom and upgeom are disabled, the solutions converge.

At the same time, when zero displacements along the X-axis (in the load direction) are disabled, the solutions converge in any case.

Can you tell me what the problem is?

And how can this problem be solved?

Macro downloaded in the attachment.

]]>I've just downloaded the Polyumod to try and model an inflatable PA66 shell.

After defining the material from the material database (PA66 2012) and submitting the simulation I get an error:

"39888 elements have zero transverse shear stiffness. Please check the input data. You may use the *TRANSVERSE SHEAR STIFFNESS option to directly specify values of transverse shear stiffness, or use the *TRANSVERSE SHEAR option to specify material transverse shear moduli for Abaqus to compute the transverse shear stiffness. The elements have been identified in element set ErrElemZeroTransShearStiff."

when modeling the same shell but as a solid extruded part the error is not occurring.

Is there a solution to this error?

Thank you.

]]>I'm trying to numerically simulate the compression of a **polyurethane foam** shock absorber for car.

I use ABACUS/Explicit

The constants for the equation of state were obtained from experimental data for compression, at a "characteristic" (small) compression rate using the MCalibration program.

While resolving, an error message was received:

The ratio of deformation speed to wave speed exceeds 1.0000 in at least one element.

One of the possible causes of the error is the incorrect calculation of the constants for the polyurethane foam equation of state.

*** using the "standard" equation, which is in ABAQUS

I am ready to send the model and experimental data to anyone who has the desire and opportunity to devote time to this task.

307]]>I am trying to verify my implementation of a viscoelastic Materialmodel in ABAQUS via modeling an uniaxial loading/unloading test.

The specimen is a 100x10x1 mm thick solid material (3D Model).

The Top surface is fixed and the load is applied as a displacement boundary condition on the bottom surface.

When I do the caluclations with a purely hyperelastic material, the strain is homogenous in the middle of the specimen, as expected. However, after I add the PRF Model with two nonlinear viscoelastic Elements using the BB-Model, the strains become inhomogenous. The model becomes kinda spotty, meaning there are periodic regions with higher strains (as can be seen in the Figure). This happens after an elongation of about 50%.

277I tried playing around with stabilization in the step-controlls, but unfortunatly that did not really help.

I am greatful for any advice.

Thanks in advance

Martin

]]>

I would like to model a closed cavity with incompressible hydrostatic fluid inside.

The cavity is deformed in the interaction between the internal fluid being which is increased in volume by adding new water (at a known rate) and the external elastic interactions. The pressure is being measured. I know how to model the problem using fluid elements, using *static, and cavity interaction, defining pressure as an input. However, I would like to have volume as a control parameter.

I looked around in the manuals but I havent found any straightforward so far. Any ideas guiding me to the right direction?

Thanks a lot in advance...

A

]]>I'm modelling in Abaqus/CAE 2020 a cylindrical structure that has to be supported by multiple nonlinear springs at its shell surface. The nonlinear springs need to to represent the characteristics of soil layers, so that at each soil layer a new spring is needed.

My problem is that, as soon as a define more than one nonlinear spring, my simulation is aborted.

So far I have managed to run my simulation with one nonlinear spring defined as:

*Spring, elset=s1_py-spring, orientation=zyl, nonlinear

1

0, 0

535.1478059, 0.001

1042.39801, 0.002

1499.382057, 0.003

1892.613097, 0.004

2217.841801, 0.005

2478.12813, 0.006

2681.012773, 0.007

2835.924671, 0.008

2952.353536, 0.009

3038.824607, 0.01

The problem arises as soon as a define a second nonlinear spring like this:

*Spring, elset=s2_py-spring, orientation=zyl, nonlinear

1

0, 0

133.6711142, 0.001

259.7325047, 0.002

372.2139596, 0.003

467.7267022, 0.004

545.4769863, 0.005

606.6212133, 0.006

653.4140725, 0.007

688.4829661, 0.008

714.355594, 0.009

733.2230855, 0.01

I get the following error message:

Error in job LBA: in keyword *CONFLICTS, file "LBA.inp", line 5795: Keyword *Conflicts is generated by Abaqus/CAE to highlight changes made via the Keywords Editor that conflict with changes made subsequently via the native GUI

Line 5795 is the line where I defined the second spring as nonlinear. This has to be done via keyword editing since nonlinear springs cannot be defined in the GUI. This problem does not come up when I do just one spring.

I can't use one spring since the springs model the soil at each layer.

Does anyone have experience modelling multiple nonlinear springs?

Any suggestions would be greatly appreciated.

]]>I have an inverse calibration loadcase wich runs without problems. It runs with a dynamic explicit solver (Abaqus). The problem is the missing Fitness value of the Loadcase. I would say, without a fitness value Mcalibration cannot consider this load case in the optimizations. Am I wrong?

My inverse model is a sphere indenting a plate. It consists of a solid indenter controlled via a reference point with a force amplitude as preset. Only CF2 and U2 are used as history outputs. RF2 is switched off, because it is always 0 (force preset). The Abaqus Model is 2D and uses CAX3 elements.

The image shows a screenshot of the loadcase in Mcalibration. The upper right chart shows the mentioned load case.

178Why is the fitness value missing? What could be the problem? Is this fitness value necessary to properly consider the load case during optimization?

Thank you!

Regards

Niels

]]>Any advice on how to include damage to achieve element deletion in a hyperelastic material? I am implementing a VUMAT of a modified Yeoh model and I want it to account for damage.

Thank you.

]]>