Tie constraints (inherited by CalculiX from Abaqus) work like welded/bonded contact in other software (prevent any relative motion - sliding and separation) but are implemented differently. They don’t use the contact algorithm. Unlike the tied contact type (also available in Abaqus) that has the same effect but uses contact elements and thus increase the size of the model.
Regular (e.g. hard or linear) contact in CalculiX allows for both separation and sliding. You can control the behavior in the normal and tangential directions separately (instead of predefined contact types available in some other FEA software). The latter allows you to specify a friction coefficient (otherwise, it’s frictionless). However, CalculiX doesn’t have rough contact (no sliding) and no separation modes available in Abaqus. Those require more manual workarounds.
It was actually created in a newer version. It will open in 2.1.11 dev.
this ran for me based on your original model without the pretension added, but with contact tied (needs to be changed to an overclosure method) , Gunnar’s additions show the direction to completion and it will run quick if you reduce the element order.
No - have never used Creo. My FEA experience to date has been
Brief experience of the old Algor DesignCheck
a little know system called 3g.author which has some very nice ease of use features, but the company was bought by Autodesk and the product killed some years ago. This did use P & H refinement (automatic) if I recall correctly.
have tried FEMDesigner, but have always struggled to get it to solve anything except the examples provided with it.
have tried to use CAELinux , even bought a book which claimed to be a training manual for Salome - but it was not helpful.
is there any example test case from external software references? it seems tied contact implementation in CalculiX have advantages over another. Stiffness in normal and transversal direction can be set, thus make it possible to model fully bonded, separation (normal), sliding (shear) and partial composite actions in flat and curved surfaces.
The idea of the compression-only constraint is that you don’t have to model the rigid surface. Internally, it uses GAP elements (simple contact elements - spring-like) whose opposite nodes are grounded. They can be even visualized. You just set their (spring) stiffness.
Surface interaction settings give you this flexibility. Surface behavior is for the normal direction, Friction is for the tangential direction. GUI options are described in the PrePoMax documentation but to learn how all those settings work, you can refer to the CalculiX User’s Manual: https://www.dhondt.de/ccx_2.22.pdf
Even Abaqus documentation and tutorials can be helpful due to similar syntax.
i has been discussed in CalculiX forum, it may have similar to cohesive zone model except in limit or damage.
maybe a new thread is needed to discuss separately, simple example to do comparison taken from CalculiX example as officially recommended. All results should be consistent with constraint generated by ‘areampc 123’ and ‘areampc slide’ command in CGX. Additional of external software e.g Abaqus or Ansys as reference is highly appreciating to clarify about the implementation.
Finally spotted something - Step numbers from set up do not match step numbers in results!!
There was a STEP 0 in results with only 1 increment in the results (but only showed briefly and now I can’t access it again). Now back to only Step #1 showing with 8 increments. I can’t believe this can be made so difficult.
I don’t know where to check for convergence files - but the monitor window when running did show step 1 then step 2. When I revert to it now, it only shows step 1 (with 10 increments).
