Sliding joint along curved path

[gkillaspy29]

Hi, someone please, please help me with this. It's driving me nuts, and has been for days. Even if the first reply is "It's not possible, stop wasting your time" that will give me some semblance of sanity back.

I'm trying to get the van door to slide along it's guide rails. Like a real van door. I've tried all kinds of joints, contact sets, motion links..

It's as close as I've got it so far, using contact sets. But as soon as I grab the door and move it...fast, slow, careful, erratic - whatever, it jumps straight out of the guide rails that the contact sets should be keeping it in! 

Any and all suggestions welcome, I'll try to respond and eliminate any suggestions I've tried already.

Thanks.

Public link:

https://a360.co/2JMsSLj

[dieselguy65 · Accepted solution]

i have been down this road plenty. it doesnt look possible other than using contact sets.

its been requested for along time now to make it possible.

as you have found, contact sets pretty much suck.

[gkillaspy29]

@dieselguy65 wrote:

i have been down this road plenty. it doesnt look possible other than using contact sets.

its been requested for along time now to make it possible.

as you have found, contact sets pretty much suck.

---> I have found that. Thanks for letting me know I'm not alone! 

---

[danialk]

I realize this has been dormant for some time, but am working with this type of scenario now. Has there been a way created to do this?

[TrippyLighting]

Nope!

[trevor.jack]

My requirement is slightly different:  I need a curved component to slide through two slots/mounts which rotate to allow the movement.  There is a T foil (ie a vertical foil with a horizontal foil on the bottom) which needs to slide up and down through a hull.  But the vertical strut, which would normally be a straight section, is curved for some of the length, at the top, so that sliding through the mounts induces the bottom to "cant" outboard.  Perhaps there is a simple solution using contact sets.  But I haven't found it.  If the "vertical" strut has constant curvature, a revolute joint works fine.  This can be seen in this screencast:  https://autode.sk/2uoepvs.  But that's no good for the case where the strut is not of constant curvature, as shown in this screencast: https://autode.sk/2Yg3mCy .

[etfrench]

You may be able to simulate it by using two joints, one for the straight section and one for the curved.  The origins of each would need a rigid joint between them.  You may need to create a hidden copy of the rotating mount in order to create the revolute joint for the curved section.

[trevor.jack]

Thanks for your suggestion, etfrench.  But I don't think that's going to work.  The curved section does not move as a revolute joint.  The centre of the radius starts at the height of the deck of the hull (ie same height as the bottom of the curved section).  This is clear as the tangent there is coincident with the straight part of the foil.  But when the curved section is entirely within the hull, the centre of curvature must be at half the depth of the hull as it has tangents at both the top and bottom mounts.

[vmu620]

the solution is:

if you want to make the animation and you have no other option Load your step files into blender and animate the curved rails manually.  blender is a free software that lets you easily animate along curved rail motion paths, but it may not be backwards compatible if you wanted to go back and forth with your geometry. 

good luck

[johnson2423]

Ouch!  Most current option / suggestion is..... use another software. (blender).  😛

[johnson2423]

OK, I solved my own problem.  "Sliding along a curve".  My steps were:

1. Make component one, make component two.

2. Make a copy of components one and two. 

3. Join copy of cmt1 to cmt1, but use a revolving joint.  Make limits to the joint so that is can turn 1/2 or turn once.

4. Make copy of cmt2  to cmt2, but use a sliding joint.  Make limits to the joint so that it can only translate so far.

5. Use a motion link between turn and translate so that one component moves translates while the other rotates.  

Thus, one body turns along a curve or a helix.  Sounds blah.  It wasn't so bad and it is showing me what I needed to see.

[sccFW676]

This is ridiculous. A 6 year old request (that's not unreasonable), and still no solution. Yet the cost of F360 keeps going up.

I'm out. I'd rather pay more for software.

@sccFW676 - this post has been edited due to Community Rules & Etiquette violation.

[TrippyLighting]

I have not gone through the example attached to the original post, but the tangency relationship that has been released maybe 2 years ago might help solve the problem:

@sccFW676 instead of blowing a gasket, it might be a better option to start a new thread, post a concrete problem and model and see if there is a solution. For obvious reasons developers don't have the time to  search ten thousands of forum posts to see if newly developed functions might solve a particular, 6-year old user problem.   

[g-andresen]

Hi,

please share „your“ file fore reply

günther