Learning Forum

[m.caragiuli]

 Hello,

my question is about the contacts setting between bodies. In details, if I have a brush which should rotate about a pin is it correct to use just a revolute joint (in order to constrain the lateral movements and all the displacements and rotation other than one single rotation) ? Or I need to create the contact between the brush and the pin as no separation and other contacts to prevent penetration between the brush and the "wall" at the edges of the pin? I was thinking about applying a remote displacement to the brush and set to 0 the displacements and the rotations except that one occuring along the axis of the pin. So how many contacts whould I create?

I would appreciate any help I can get.

Thanks!

[peteroznewman]

Hello,

There are often three or more ways to build a model.  In the case of a pin and a hole, the simplest model is to add a revolute joint between the pin and the hole. If the pin is on a structure that moves, then the hole can follow the motion of the pin as well as be able to rotate about the pin.

If you use a remote displacement on the hole face, that connects it to ground, so the motion of the pin cannot enter into the solution.

Using No Separation between the two cylindrical surfaces leaves two DOF free, rotation and axial displacement, so it is not equivalent to a revolute joint.

[m.caragiuli]

Hi Peter,

thank you for the tips. I've already tried to use the joints but my simulation diverges. Maybe there is something wrong with the contacts. Let me explain the model:

there are two flexible bodies  (upper green and lower brown) connected by a rod with the brushes at the ends. Actually the rod has two components since the lower parte can slide within the upper part. However each rod is a rigid solid body with the brush. 

A force applied at the lower body should move backwards ad upwards the lower body to get close to the upper one. Thus, the lower rod should translate along and within the upper one. 

I thought about the following connections:

-a translational joint between the faces of the two rods (the lower is the mobile body)

- one revolute joint at the interface between the pin and the lower brush (the hole) 

-one revolute joint between the upper brush and the pin

I have constrained the movement of the superior face of the upper body.

Do I wrong anything?

Should I create a contact between the two bodies to make them slide one over the other (but they don't touch)?

Should I make a frictionless contact between the two rods?

Is there some missing connection?

[peteroznewman]

A translation joint between the rigid cylinder and shaft is an overconstraint because the rod is free to rotate on its axis.  Change the translation joint to a cylindrical joint.

An alternative is to leave the translation joint, but change the revolute on the end of the rod to the brown part to a spherical joint. That allows the brown part one more rotational degree of freedom than having a revolute and a cylindrical joint on the rod.

[m.caragiuli]

I don't understand why do you want to add the rotation degree of freedom to the cylinder. I don't think the rod should rotate about its axis, only translate to enter within the hollow cylinder. Isn't it?

 

I have jut tried to leave the translation joint and change the revolute on the brown part to a spherical joint. The simulation diverges with this error "An internal solution magnitude limit was exceeded. Please check your Environment for inappropriate load values or insufficient supports.  Please see the Troubleshooting section of the Help System for more information."

 

[peteroznewman]

The cylinder you show is just one side of the full model, there is another on the other side, right?

You must have other connections besides a single cylinder. What are they?

[m.caragiuli]

this time I've considered just half of the model and suppressed all the teeth except  the rear upper and lower two teeth and periodontal ligaments. In this way once everything works I can add other bodies. I've created fixed supports on the outer surface of the periodontal ligaments. Bonded contact between them and teeth, and teeth and the two splints. The lateral pins are part of the splints, one single body. Then I used all joints to connect the pins at the end of the cylender and the translation joint to connect the two cylinders. The joints are body body, rigid behaviour.

I though about creating a frictionless contact between the two cylenders to allow the translation, but if I use a joint I don't think I should use also a contact. Moreover, I thought about constraining the contact between the brush and the "walls" of the pin, but I think that a revolute joint does the same thing.

Large deflection is on. To create the simmetry region I've split the model in two simmetrical halves and applied the region to the lateral faces of the upper and lower splints.

Am I neglecting some connection?

Thank you for your kind support!

[peteroznewman]

When you manually add a Symmetry Region, you can pick the faces that are on the cut boundary, but the Symmetry Region does not automatically select the normal direction to those faces. It defaults to the X-axis, which might be totally wrong.  Please check that.

No, you don't need contact if you used a translation joint between two bodies.

[m.caragiuli]

Hi Peter,

sorry for the delay, but I have tried to perform the simulations following all your tips, however I still have some doubts about the the nature of the induced movement. I had applied a force on the surface of the lower pin. This force should make the lower splint and all the related bodies move backward and upward so that the two surfaces of the splints( the bottom surface of the upper splint and the top surface of the lower splint become in contact to slide on each other). So I don't know if the problem occur on the coordinate system of each joint or on the type of joint. Do you have any suggestions? Just to be clear I set the coordinate system of each joint so that the z axis of the revolute joint points along the longitudinal axis of the pin and outwards, while the cylindrical joint has the z axis along its longitudinal axis.