SimScale CAE Forum

View deformed geometry in a simulation


#1

Hi all,

I have a really naive question - when I view my simulation results, I can see the displacement in a colored heatmap but I cannot see the actual deformed shape of the geometry. How can I see that?

FYI - the project is https://www.simscale.com/projects/yw5aj/test_astm/ and it is the “Simulation 2” that I was looking at.

Thank you!

Shawn


#2

Hi @yw5aj,

the “Warp by Vector” filter in the post-processing environment will do the job. Simply apply it on your result field and you can then visualize the deformation field via an actual deformed geometry.

Best,

David


#3

Hi @dheiny ,

Thank you! That’s exactly what I was looking for.

It maybe me running the simulation wrong, but when I run my uniaxial tension of a elastic-plastic material, the deformed body looks very problematic. I went into my https://www.simscale.com/projects/yw5aj/test_astm/ , went to simulation 2, solution fields, added WarpByVector to displacement fields.

I have a few years experience w/ ABAQUS and ANSYS but have never used Code_Aster or Salome-Meca before, so it could be just me not being familiar with SimScale’s solvers or postprocessers. I’ll try to figure things out, and any feedback would be really appreciated :slight_smile:

Thank you!

Shawn


#4

Hi @yw5aj,

I’ve had a look and a few things to be addressed:

What is the material that you are assigning? You have a very low Young’s Modulus assigned (E = 724 N/m2) - for Example steel would be E = 200e9 N/m2

Also for the displacement you are applying, I would rather recommend a “fixed value” BC. Consider the value of displacement that should be applied - by setting dx to 50, you are applying a displacement of 50 meters which is why the deformed shape looks odd.

@saienrugdeo was doing a similar project a while back and his project might help you with the setup of your own analysis

You’ll find it here:
https://www.simscale.com/projects/saienrugdeo/tensile_test__comparison/

Best,
Anna


#5

Hi @yw5aj,

Attached is your displacement compared to actual geometry

It is pretty simple, just go to "warp by vector"and then click on desired “Field (von-misses stresses/deformation)” and also add “opacity” as seen in the picture.

As anna pointed out, an Youngs modulus of 754 N/m^2 is very low compared to other materials like Al, Fe etc.

Hope my suggestion on warp by vector was helpful.
Cheers !


#6

Ah - thank you! I was having the old habit from ABAQUS and using the unit-free numbers. Unfortunately, I guess the OnShape geometry import has automatically converted mm to meters, so I should also convert my 724 MPa to 724e6 Pa (it’s a polymer) and the distance should be 50e-3 = 0.05 m, instead of 50.

Thank you again for your time - I really appreciate it. And also I appreciate your sharing of the similar project. I’ll fix it and give it another shot! :slight_smile:

Best,
Shawn


#7

Fixed the unit issue and it’s working now! Thank you so much for helping. The results look really good! The necking behavior is actually reflected.

Thanks again :slight_smile:

Shawn


#8

Hi @yw5aj,

indeed, you currently don’t have explicit control about how units found in the CAD model should be treated. Providing more control over units when importing CAD models is on the roadmap. If you’re envisioning a specific way of doing this, feel free to share it in the Vote for features section!

Best,

David


#9

Thanks again for your response @dheiny!

I have a new question wrt. this project (and please let me know if I should open a new thread): I ran a simulation with reaction force in the output request (added it to the “Solution fields” in the “Solution control”). However, when I came to the post-processor, I do not know how to plot the reaction force over time. I can color code it as if it were stress fields, but obviously this is only suitable for observing field distribution like stress, instead of force over time. I know there must be a way to do it - could you kindly share a pointer to related documentation? Optimally, I’d love to learn it in both the online post-processor and the offline paraview.

Thanks!

Shawn


#10

Hi @yw5aj,
there are several ways how you can do that.

  • Online: In the result control in addition to the solution fields where you added the reaction force, there are items called edge-, area- and volume calculation as well as point data. All of those create plots from some processed data (sum, min/max, average) on the specified entities. There you could add an area calculation on the fixation faces and compute the sum of the reaction forces, the total reaction force. This will create a plot with the reaction forces after the calculation is finished. Unfortunately at the moment you have to re-compute the full calculation to add such a plot after the initial run was finished.
    Here is the corresponding documentation link: https://www.simscale.com/docs/content/simulation/resultControl/fieldCalculations.html
    And a related forum post by @ahmedhussain18 where those items are also explained: [Result Control] Calculation methods on specific entities and their differences
    And a public project by @BenLewis, where several result control items have been used to compute, amongst others, the reaction forces: https://www.simscale.com/projects/BenLewis/spring_-_custom_macahines/

  • Offline: In local ParaView there are several ways how you can do that. For example you can use the extract blocks filter to reduce the computation to only a part of the model (fixation faces for example) and then use a calculator filter where you can simply use the SUM operator to compute the total force. This forum conversation might also be helpful: How to get nodal solution results?

Hope this helps!

Best,
Richard


'Dynamic Nonlinear Impact Analysis of Skull with and without Helmet' simulation project by ahmedhussain18
#11

@rszoeke Thank you so much for your detailed response! This works perfect.

My favorite part of SimScale is really the support from community whenever I have a question. :slight_smile:


#12

You’re welcome!
That’s the whole point of the community, learning from each other!
Best,
Richard