Boundary Conditions and Springs

[principiante]

Hi guys,

I'm trying to get the correct bearing reaction forces from the simulation of a torque rod. There are two bearings. Bearing A fixes the part on the rotation axis but allows for rotation (roller bearing). Bearing B inhibits the roation (basically a connection to the ground). The torque is applied on the other end of the rod and occurs within a short time period. The setup is displayed in the attacched image.

I use Abaqus Explicit for the simulation. The problem is that the reaction force at bearing B (F_By) is much higher than it should be judging from an experiment. When I add a spring (Abaqus: *SPRINGA) between bearing B and the part, the reaction force there (F_By) completely changes and is closer to the experiment. If I only use bearing A with a complete fixation (encastre) the reaction moment is correct. Why is the reaction force at bearing B overestimated and how can this problem be solved?

Thank you in advance for your suggestions!

P.S.: For the boundary conditions I use for each bearing a reference point that is coupled to the surface where the bearing and the part touch each other. The geometries of the bearings themselves are not modeled. I use C3D8R elements.

 

[FEA way]

First of all I think that you don't need Explicit for that, I suggest switching to Standard (implicit solver). Can you show a picture of your model with boundary conditions ? The way these two bearing are connected (interaction) may be crucial here. Do you use contact (if yes then which type) or tie constraint between these bearings ?

 

[principiante]

Hi FEA way,

There are no contact constraints in this model. The hollow rod is connected to the plate in the back of the picture with a tie constraint on the touching surfaces to enable the transfer of the torque. Bearing A is realized with a reference point and kinematic coupling to the surface where the bearing and the part touch each other. Bearing B is realized with a reference point and coupling with continuum distributing to the surface. I didn't use kinematic distributing for the coupling at bearing B because Abaqus put out a warning when I added a spring between the bearing and the part that a rigid connection is ill advised with a spring and this message disappeared with continuum distributing.

Even though I am doing the simulation as a student I can only show this simplified version due to general confidentiality reasons. Sorry for that.

 

[FEA way]

Try using distributing coupling for bearing A too. These coupling constraints should provide more realistic results since it allows for some small relative motion,

 

[rb1957]

if you have hard constraints (like at A and at B) the model is over-constrained. You're saying that these two surfaces don't deform relative to each other.

having a hard constraint at A and a finite stiffness constraint at B (your spring) is better. Now the two surfaces can have (tiny) elastic deformation.

You should be able to see this in the output. With two hard constraints there should be no relative displacement between the two surfaces (sets of nodes), with the spring you should see some displacements.

another day in paradise, or is paradise one day closer ?

 

[Tmoose]

Hi principiante,

I see you have not been back since June 18, the day of your original post.
Perhaps you received answers sufficient to solve your problem.

As is so often the case, this old man needs pictures and drawings of your arrangement with arrows and explanations to help understand what it is you are REALLY asking.

I often take a few screenshots of the model, etc from various orientations, save them as png files, and cut, past and annotate them with MSPaint etc. For a start, Boundary conditions, which component is the torque rod, and Bearing A and bearing B.
Extra time, maybe 5 minutes, often more.
Very often the creation of such image documents simplifies and clarifies my understanding of the topic as well, with the result communication in the future is much improved.

As a respondent I often annotate posted images in an effort to help clarify my response.