Assembly Path/ Clearance?

[MikeHalloran]

Imagine two flywheels, with very short shafts extending toward each other. Say the shafts are joined by a two-piece coupling that splits on a plane that includes a diameter and the shaft axis, like a Victaulic pipe coupling. There is no axial clearance to speak of around the coupling; it's sort of in a big groove in the machine.

Easy to disassemble, right? Just withdraw the coupling halves radially.

Then the boss says he wants to use an existing part to link the structures surrounding the flywheels. Said existing part has a rigid hoop pretty much in the plane of the coupling, so the coupling halves can't be withdrawn radially more than a few inches. There is a gap between the hoop and the flywheels, so maybe if the coupling halves could be withdrawn a bit, tilted diagonally, and withdrawn some more, they could come out... if there's clearance between the flywheel and the coupling halves (which look sort of like engine main bearing caps).

Is there a way that SW can be helpful in figuring out if the coupling halves can squeeze between the solid objects in the way, without knowing in advance what 3D path is necessary to get them through the gap?



Mike Halloran
Pembroke Pines, FL, USA

 

[Eltron]

You can try to use physical dynamics to figure it out. Check the Help for Physical Dynamics.

Dan

http://www.eltronresearch.com

 

[MikeHalloran]

Thanks, that looks promising.



Mike Halloran
Pembroke Pines, FL, USA

 

[MikeHalloran]

"Physical Dynamics is unable to find solutions for this model. Physical Dynamics has been disabled."

sigh.



Mike Halloran
Pembroke Pines, FL, USA

 

[dgowans]

Mike,

You may have to simplify your model some, such that SWX doesn't have to think so hard to do the collision detection. By simplify I don't mean modify the geometry at all, because that would be pretty much meaningless, but to try to combine any components that you want to avoid hitting with the coupling halves into a single part. Then you'll only be doing a collision detection between 2 parts, rather than on a whole assembly.

When I've done this type of thing (not easy, BTW) I've found that I needed to have the moving part semi-constrained so that your motion isn't completely random when dragging the part to be inserted/removed. Even if you do prove that the whole thing can be (dis)assembled, this is no guarantee that it's going to be easy to do in the physical world. Any opportunity to just prototype the whole mess and figure it out that way?

 

[MikeHalloran]

Re prototyping, it may be possible. One of the pieces is part of an engine we sell, one is a complicated bracket that bolts up to it, and the coupling half is not too big to fake out of wax or wood. Of course, the tools in our "Engineering Lab" comprise whatever I carried in, in my pocket.

Re simplifying the model, is it possible to copy three objects from an assembly (one of which is itself an assembly), and paste them into a blank assembly, while retaining their spatial relation to each other? I can copy them individually, but then they come in aligned per their own local coordinate system, and not tied to anything else, ready to be constrained with mates, but that's not what I want to do.




Mike Halloran
Pembroke Pines, FL, USA

 

[dgowans]

Mike,

You could always do a "Save As" (check the save as copy toggle) on the assembly, open your copied assembly and delete what you don't want. You may have a few mates crap out, but you can fix the parts of interest in position.

 

[CorBlimeyLimey]

Or use Pack-and-Go to create a copy of the assy replete with copies of all relevant parts. Then you can play around without fear of messing up the original files.

 

[rockguy]

Or select the objects you want in the tree, right click and select "form new sub assembly". I think that will work.

Rob Rodriguez CSWP

http://www.axiscadsolutions.com

http://www.robrodriguez.com

http://designsmarter.typepad.com/rob_rodriguez/

Eastern Region SWUGN Representative

http://www.swugn.org

http://www.nvtswug.com

SW 2007 SP 2.0

 

[MikeHalloran]

Many thanks to all.



Mike Halloran
Pembroke Pines, FL, USA

 

[TVP45]

Mike,
I had a similar kind of problem of interlocking 6 spiral coils by twisting, tilting, and sliding. The complexity simply overwhelmed the machine/software capability and each step in physical dynamics would take maybe 2 minutes or so. I ended up handling it by doing one coordinate at a time together with judicious cuts and suppressions to eliminate 90% of the assembly. That is, I would slide a portion of two coils along the axis till they touched. Then I would judge where the tilt interference was tightest and do a tilt there with perhaps a new cut. etc. etc. My best guess was that this was conservative by maybe 15-20%. But, if I had it to do over again, and management actually believed in 3D, I would have used prototypes.

 

[MikeHalloran]

I couldn't get the "form new subassembly" to work, because the objects are themselves parts of different subassemblies, and are not close to each other in the tree.

I was able to make a scratch copy of the assembly, remove about a third of it, and suppress the remainder of the irrelevant stuff, and manipulate the view to get a better picture. It don't fit.

I was hoping SW would make this sort of stuff easier.



Mike Halloran
Pembroke Pines, FL, USA

 

[SincoTC]

If you can justify the cost, or the price of getting it wrong is too high, you could get a bureau to Rapid Prototype the parts for you direct from your SW models (or exported STL files). Then you can juggle them around by hand to see what's possible or where a small chamfer or modification may make it all happen!

Not seeing the parts I don't know if they would be strong enough, but some of the newer RP 3D printers can produce models as strong as ABS and maybe they could be scaled down to reduce build time and material costs.

Trevor Clarke. (R & D) Scientific Instruments.Somerset. UK

SW2007x64 SP3.0 Pentium P4 3.6Ghz, 4Gb Ram ATI FireGL V7100 Driver: 8.323.0.0
SW2007x32 SP4.0 Pentium P4 3.6Ghz, 2Gb Ram NVIDIA Quadro FX 500 Driver: 6.14.10.7756

 

[MikeHalloran]

I sliced up the part that has to come out, turning two pieces into six. It's an ugly solution... but not as ugly as what it replaces.



Mike Halloran
Pembroke Pines, FL, USA