Hey everyone!
I recently started using CalculiX and am trying to model a shaft with two bearings. To do this, I want to define a bearing stiffness matrix as supports.
I am using PrePoMax as a pre-processor and have tried to include rotational stiffnesses in the .inp file, as PrePoMax does not support this. However, this isn’t working very well.
Are there any solutions for this problem?
Thanks in advance.
You could use discrete spring elements. Translational ones will be easy, but for the rotational stiffnesses you will need some workarounds with rigid body constraints (indeed not directly supported in PrePoMax and require some keyword edits, so you are in the right place to ask): Torsional Spring - SPRING2?
Springs can be nonlinear too, if needed.
Thanks for your reply,
I tried connecting the bearing surfaces using *Rigid Body with reference points and attached spring elements to them. I made a modal analysis with a frequency step to investigate the Eigenfrequencies at 10000 rpm, but I’m getting frequencies that differ significantly from results I have from another simulation software (Bearinx from Schaeffler), what made me rethink my model.
Here a short example of how I implementet it in my .inp file:
*Rigid body, Nset=Internal-1_support1, Ref node=10733, Rot node=10734 (nodes of Reference Points)
*Rigid body, Nset=Internal-1_support2, Ref node=10735, Rot node=10736
*ELEMENT, TYPE=SPRING1, ELSET=E_SUPPORT2_X
1000001, 10735
*SPRING, ELSET=E_SUPPORT2_X
1
2.13485E+08
*ELEMENT, TYPE=SPRING1, ELSET=L1_RX
100007, 10734
*SPRING, ELSET=L1_RX
4
9854.776
Any ideas why the results do not seem accurate? Is this approach not suitable to model a bearing or making a frequency analysis?
The idea is to connect two ROT NODEs with a spring so you should try with SPRING2 or SPRINGA elements (two-node springs). They can’t be defined in PrePoMax, so you will need further keyword edits for that.
I will try that. Thank you. I will let you know when I am able to create a useable model.
Quick update: I have managed to implement a bearing stiffness matrix for angular contact ball bearings. I used point springs and connected the reference points with rigid body constraints. The center offset and spring stiffnesses can be calculated in such a way that the stiffness matrix is implemented. The central reference point also has an offset to represent the off-diagonal entries. I got really good results comparing to Bearinx.
I did not have to edit the input file manually!
Great, can you share maybe the two models/results?
Sorry I can’t share any models due to confidentiality reasons. But I got about 0,6 % deviation between my models regarding the first bending eigenfrequencies. But I can provide the calculation for the position and spring force if needed.
Well, I’m actually interested in the formulas. Unfortunately, I can’t make much out from the picture—probably because of the perspective. I’d appreciate it if you could share them. They don’t have to be particularly polished.
