pls help rotational constraints

[kathy]

Dear Gurus:


I am stuck in simulation constraints. I have a simple cylinder. I fix all 6 dof on top surface, then I apply a load around cylinder, then I put constraint 2 on the bottom surface. I want the bottom surface rotate around X axies, here is how I set the constraint 2: Translation: X=Y=Z=0; rotational: X=1.5 radian, Y=Z=0.


Then my simulation resultis wrong! The bottom surface won't rotate around X axis to 1.5 radians. Can anybody tell me what is wrong with the constraint?


Thanks a lot!


Kathy

 

[rmckinley]

When your second surface rotates, components of it must also translate. It sounds like you may want to constrain translation ofthe surface in X and allow freedom in all other degrees. Also, when you input a value for prescribed translation, you are enforcing a specific amount of displacement in whichcasea force load is not needed. What I ususally do to affect a rotation of a surface is to create a momentload using "total load at point." To do this, I create a point on the axis of rotation and select it.Then I select the surface that is to move.


Hope this is helpful.

 

[kathy]

HI rmckinley
Thank you very much for the tip. I still don't understand it. Does it means that i have to prescribed a non zero displacement constraint in Y and Z and set displacement in other direction free? Then it becomes Y= a number, Z=a number, X is free or zero? rotational: X=1.5 radian, Y=Z=0. remain the same? Will it be hard to calculate all of the Y and Z translational displacement for all points in the bottom surface. The second thing you talked about is that we have to balance the translation of Y and Z by a point moment? Do I understand this part right?


I would really appreciate your comment.


Thanks a lot!

Kathy

 

[giannibruno]

Hi, sorry for my english


If you want the top to have a know displacement (that means the cilinder is already under a not know force that causes the displacement itself) use a cylindrical coordinate system that has to be placed in the top surface with the z axis on the cylinder axis. Then define the constraint using the cylindrical coo. sys. You can define DOF as angle and displacement along z. Free all rotation and insert angle displacement. Mind what kind of unit are you using (degree or radiant?)!


Remenber that placing a displacement means putting a load that you do not know as value but as effect!


Feel free to reply


Ciao

 

[rmckinley]

kathy, I don't fully understand your application but if you are trying to cause a specific amount of rotation on the surface to find out what the force is, you must use a cylindrical coordinate system (or else there will always be some translation). In this case, you will not have to apply a load but rather you will find out what the resultant load must be by measuring the reaction force developed in your constraint. Otherwise, if your load in known, you might try this: Constrain your first surface in all degrees. Then apply the load on the top surface of the cylinder -- without constraining the other surface. I might be of more help if I could see your model. Is the load around the cylinder a moment load? If so, you might want to use a "total load at point." With this technique, a cylindrical coordinate system may not be necessary but can be helpful. Do you really want to apply a load around the cylinder? If so, you may get erroneous results if the outside surface to which you apply the load borders directly on the bottom surface that is constrained.


Hope this helps

 

[kathy]

giannibruno and rmckinley
Thank you so much for the great help! I really appreciate it! It works in cylinderal sys now.

cheers,


Kathy

 

[james95777]

Kathy,


Are you using solid elements or shell elements?


Nodes on solid elements haveonly translational degrees of freedom, and no rotational degrees of freedom. So, rotational constraints on solid elements aren't a valid boundary condition, and Mechanica disregards them. It only gives you that option because at that point to allow for the possibility of that edge being a shell element.


Nodes on shell elements, however, can have rotational degrees of freedom. Even if you use shell elements, you would actually need to specify translational constraints on the base of the cylinder. While the base of the cylinder is rotating, the nodes are actually translating. The only node that is rotating would be the one at the axis of the cylinder. Having said that, I think it would be hard to create a solid cylinder with shell elements.

 

[Dorasatan]

[No message]