What constraint do I use for a component only held by friction?

[cadkins]

I have created a simple rotating disc and a small mass (a block) and I want to sit the small mass on top of the rotating disc and apply friction, so that when the disc reaches a ceratin angular velocity the mass will slide off .

How do I constrain the two together so that the block is almost free and that only the friction between the two components holds them together?

Thanks.

[sdensberger]

Huh, I'm surprised your model ran as well...can you post your model?

[sdensberger]

I'll have to look at your model, but I believe the bonded nature is only for shear laods; the block should still be allowed to separate from the disk.

[sdensberger]

I suggest looking into Mechanism (not Simulate) and do this from a rigid body dynamics point of view. I don't have any experience with Mechanism, so hopefully someone else will chim in.

[JonathanHodgson]

Cameron,

What question are you trying to answer?

[mlindqvist · Accepted solution]

If it's a mechanism dynamics analysis I would use a planar constraint. Friction between disc and block. The disc accelerates, at a certain speed the block slides off the disc.

/Mats L/

Video Link : 5470

[cadkins]

Apologies for the confusion; my fault.

I was asking a question mechanism dynaimcs analysis, but thank you everyone for your help.

I didn't see a category for Mechanisms so I posted in Simulate with a tag of Mechanisms.

For future reference should I post Mechanisms quaetions in Parametric?

Question for Mats Lindqvist,

Do I need to constrain the block in either of the translation axes or just face to face? and to clarify "3-D contact" does not need to be used?

Cameron

[mlindqvist]

I used a "planar constraint" with friction applied to the rotational axis (this is done in assembly mode, edit the assembly constraint definition and then you can enter/edit dynamic properties for each joint axis.

For the translational axes I have entered/enabled regeneration values, so I can click "regenerate" to get the assembly back to the original positions. But in mechanism, I have not entered any drivers/forces on the translational joint axes, they are free to move.

I did not use 3D contact. It is probably possible, and more realistic to use 3D contact, but this is a lot more computationally expensive. You might also need to add rounds to the edges of the block, since 3D contacts don't like sharp corners. A workaround in this case could be to add spheres at each corner of the block, sphere against planar surface is (I think, can anyone confirm this?) less computationally expensive than an arbitrary shape surface against a planar surface. If you run a 3D contact analysis, you might have issues when a corner of the block slides past the edge of the disc. As I mentioned, 3D contact's don't like discontinuities.

Hope this helps...

/Mats L/

[cadkins]

Hi MatsLindqvist‌,

Sorry to dig up old bones here but I've come back to this. I originally bypassed my whole problem by using a slider joint and applying friction along the rotational axis, in the direction of the centripetal force and it solved the problem fine. However my dissertation adviser didn't really like this as it uses predetermined knowledge of which direction it is going to slide off because the slider joint only allows one degree of freedom in the translation axis.

What I need to do is be able to use the planar joint with friction, however it still eludes me. The planar connection would be with the turntable face and a bottom face of the mass. You can't apply friction on the translation axes but only on the rotational axis. The problem is the rotational axis is normal to the planar axis, whereas with the slider joint when I got it working, the rotational axis with the friction is parallel to the planes.

How did you manage to apply friction in the right direction with a planar joint? Help would be much appreciated as it's part of my dissertation.

Also what effect would the contact radius option have on this?

Many Thanks,

Cameron

[cadkins]

Thanks for the help Mats, its cleared up some issues for me.

I'm still having problems with it, the mass isn't sliding and instead its rotating with the block as if it is fixed to it. I have set one translation starting plane to 0mm and the other translational stating point to 100mm (with regeneration on). I have then turned on the roatioanl cnostraint and enabled friction giving static friction of 0.74 and kinetic friction of 0.6.

It could be the way I'm running my analysis? When I open my mechanism analysis I leave the type on "position", however this greys out "Ext. Loads" which I notice if you set type to "Dynamic" you can enable friction and gravity?

Thanks again

Cameron

P.s. I have attached my assembly to this post. I use Creo 2.0.