Troubleshooting Solidworks Assembly Stuck: Tips for Smooth Motion

[Curl]

Solidworks Assembly is "Stuck"

I made an assembly of a few parts, and I mated them how they should. When I move one part, I need the whole device to move around how it would in real life, but this requires a few parts to revolve/spin about different axes, which seems to confuse Solidworks. Basically the assembly seems "stuck" although the motion I want is totally possible (I can manually set it to this position without problem).

I tried to nudge it and re-mate, and I get this error message. Apparently Solidworks can't "move" my part to position, however it is possible, you just have to twist it around another axis (as defined by a different mate), then it will fit.

Is Solidworks just dumb or is there some ways of getting around this?

 

[xxChrisxx]

It depends on how you've mated everything up. Solidworks 07 got terribly confused when I did an engine, I was messy with my mates and it jammed up every time I tried to rotate it more than 360 degrees. It's difficult to give a diagnosis as to why it happens because I never found out.

I just rebuilt the assembly with more methodical mates, that seemed to solve it. Have you had the symptom where one of the compoents to flips 180 out of position if you try to force the motion when it jams?

 

[Mech_Engineer]

I've had good luck with being able to make assemblies and sub-assemblies move in SW 2009, 2010, and 2011. You should take a look at your mating conditions, it looks like one of them (a parallel condition) breaks when you try to move the component.

 

[Topher925]

It sounds like your assembly is over defined. Double check all of your mates to make sure they are the right type and mate the correct surfaces. Somewhere, there's a contradiction.

 

[Curl]

Okay I changed something to the width mate and its moving now, however I get this error which is pis5ing me off exceedingly. I have 2 parts which have circular openings, and I want to add the concentric mate, and this is what its saying.

If I rotate the thing a bit, it will give a new value for the "distance apart". There's no way I have an ellipse, I'm sure they're circles so how could this happen?

EDIT: ROFL apparently one of them is not circular. How can a solid of revolution have non-circular profile? This is hilarious.

 

[Topher925]

Ar you sure the part is not just a different radius rather than elliptical? The perspective view that CAD programs give can be deceiving.

 

[Curl]

Ar you sure the part is not just a different radius rather than elliptical? The perspective view that CAD programs give can be deceiving.

Doesn't matter if the radius is different, I just want them concentric.

I guess I'll start over from scratch since some sort of whitchcraft is afoot in these models.

 

[Pooty]

Yeah, same as Topher I've had problems when my mates are overdefined. Sometimes, rebuilding your assembly a few times to find the most efficient mating process really helps. I know this doesn't help very much but... You should show a picture of your assembly. I love SolidWorks and I'm always interested with what someone is doing in it.

 

[xxChrisxx]

How can a solid of revolution have non-circular profile? This is hilarious.

Regarding this, you know that computers use discreet calculations. As such a solid revolution isn't acutally circular. The easiest way to describe it is that when AutoCAD plots a circle, it's acutally plotting a polygon with a very high number of sides. Solidworks does that same thing for revolved objects.

Doesn't matter if the radius is different, I just want them concentric.
I guess I'll start over from scratch since some sort of whitchcraft is afoot in these models.

It may be a modelling tolerance issue. I can't rememeber if solidworks has the option (it should have though) but you can set it to have a varying level of fidelity to the 'requested' part. Higher fidelity = slower performance.

This could be your issue.


Also that centre part looks like a sphere, I've not seen a 3d modelling package (except maybe Catia, but I've never used it only seen a demonstration by a pro) that doesn't have some headaches with spheres when you try to mate them.

 

[dawin]

You need parts to revolve/spin when something else moves? Have you tried gear mates?

Rather than make the circles concentric, why not try mating the temporary axes to be collinear/coincident @ an angle/insert what you need here. May be easier.