Flexible Assembly problem

[ksauter]

I have a drawer slide which has 3 sliding sections with limit mates. It works beautifully in its own window. I assemble 2 drawer slides to one drawer as a subassembly. The slides each work beautifully in this assembly as well. I assemble the drawer (with the slides attached) into a cabinet assembly. I want to attach both slides to the cabinet in the next assembly, just like in real life, and be able to pull the drawer out, with both slides staying attached to both the drawer and the cabinet, again just like in real life, so the drawer slides allow the drawer to move and also limit its movement.

I can attach one slide to the cabinet and pull the drawer out with that one slide working beautifully, just as intended. As soon as I attach the other slide to the cabinet, SW says the assembly is overconstrained, which is hogwash. This happens in real life all the time. I just want BOTH slides to work properly in the cabinet assembly. How do I convince SW this is a legitmate mate?

 

[ShaggyPE]

This is because solidworks is treating the two slides as the same... including the extended position. Even though the position is the same, which one is the driver and which is the passenger? You will need to create two configurations of your slide assembly. Put one on one side and the other on the other.

Also, I thought SW fixed this issue on newer versions. Which version are you using?

-Dustin
Professional Engineer
Certified SolidWorks Professional
Certified COSMOSWorks Designer Specialist
Certified SolidWorks Advanced Sheet Metal Specialist

 

[ksauter]

I'm using SW2008 SP0. Making the parts different configurations still gets me the same results. SW says "the two parts are already fully defined" after I apply the mate. That's a bit of a lie, because I was able to pull the undefined drawer slide out of position, apply the mate, and it slid into place properly. Then, when the assembly rebuilt, it gave me a problem again. So this isn't a solution. There has to be a way to do this.

 

[CorBlimeyLimey]

Instead of mating to both sides of the drawer to both sides of the cabinet, mate a central plane of the drawer sub-assy to a central plane of the cabinet sub-assy. The slightest discrepancy in the widths being mated could cause the over-constrained condition.

 

[SincoTC]

Just don't get me started about limit and tangent mates!!

I went through agonies in SW05/06 trying to get these things to work correctly, then it got subsatantailly better on SW07 and seeminly did away with the necessity of separate configurations for multiple instances of flexible sub-assemblies, however, they were always rather fragile and would occasionally flip states and break every mate, you just hoped that you could close without saving and they would be OK on re-opening!

I too made a double slide-drawer way back, the different configurations didn't work until I suppressed the limit mates on the second slides config and then in the assembly, I slaved the slide parts to the first limited one by coincident mates on relevant planes/added axes or concentric mates on suitable holes (but not same elements used by the limit mates). Then in SW07 it all seemed to work OK with just two slides with the default config. Now I've just opened it on 09 to refresh my memory and the mates are all broken again, arghh!!




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

SW2009x64 SP2.1 Pentium P4 3.6Ghz, 4Gb Ram ATI FireGL V7100 Driver: 8.323.0.0
SW2009x32 SP1.1 Pentium P4 3.6Ghz, 2Gb Ram NVIDIA Quadro FX 500 Driver: 6.14.11.7751

 

[ksauter]

This seems to be an issue with SW. I'll have to play with the mates to get what I want, and maybe I will, maybe I won't. I'll just figure out how to deal with it. I'd like to slap the crap out of the mate-flipping gremlin in SW, though, because that annoys the hell out of me. If I want a mate flipped, I'll do it myself, thank you. At least ask me first and let me shut it off if I want to.

I'm making a list of enhancement requests. At least the SW guys listen to requests, unlike PTC.

 

[CorBlimeyLimey]

If I want a mate flipped, I'll do it myself, thank you. At least ask me first and let me shut it off if I want to.

I'm not sure if it was in SW08-SP0, but in SW09 there is an option for just that. Actually I think it was incorporated in a later SP.

 

[ksauter]

That's exactly what I mean. I mention it and the next thing I know, it's in the program. SW listens. If it were PTC, it would be 15 years and it still wouldn't be there.

 

[ksauter]

I feel bad about bashing PTC. Proe is a good program and I'll use it if a customer wants it, but it does take a long, long, looooooooooong time before they fix the simplest, stupidest little problems, if they ever get around to it.

 

[mjcole]

They are still trying as hard as hell to keep the Menu Manager active until WF20 if they make it that far. I used to work there but no job offered after graduation cause of economy. I now work for a SW VAR and SW 3D sketch gives way more control and ease of use than ISDX.

Would it be possible to post your files even defeatured a bit using Engineering.com?

You can try creating an identical configuration to avoid the over constrain. I think havine the sub assembly twice with same config may be causinng the overdefined error.

Michael

 

[handleman]

SolidWorks (at least last time I tried it) counts a limit mate as though it's a "fully defining" mate, even though it allows motion. If you add a limit mate between any two components and then try to add a distance mate that falls within the limits, SW will say it's overdefined. That's essentially what you're doing when you add your second slide.

-handleman, CSWP (The new, easy test)

 

[EEnd]

One solution might be to create a configuration with the limit mates suppressed and use that for the second slide. Then you can create mates in the assembly where they are both used which keep the slide without limit mates aligned to the slide with.

Eric

 

[ksauter]

I did exactly that and SW still tells me the assembly is overconstrained as soon as I attach the second configuration (with the suppressed limit mates) to the cabinet. I'll keep trying and report back if I figure out something that works. Right now I have more important things to deal with.

 

[handleman]

Can you zip and upload the assembly in its "overconstrained" state?

-handleman, CSWP (The new, easy test)

 

[ksauter]

This is for a customer, so I'll have to create a simplified assembly. I'll do that in the next few days and try to upload it.

 

[SincoTC]

I've just rebuilt my test drawer assembly, with the method I originally used and described above (the same as EEend's solution).

I have attached the assembly, but just realised that you probably can't open it as it's in SW2009 so I'll try and describe it!

One slide has Limit Mates and a Symmetric mate to control the extension and four coincident mates on planes to position the slides in the other axes. The other slide has an added configuration I appropriately called "No Limits" and has the Limit and Symmetric mates suppressed.

I inserted the two slides into the bench assembly, initially both with the default limit mate configurations. After setting them both to flexible, I mated them into the bench's draw recess, taking care to ensure that this mating is only to each slide assembly's Fixed (f) element and that each is correctly orientated so that they will extend in the correct direction and then check they work. Now set one slide's configuration to "No Limits" and then slave this slide's ball carrier and moving member to their equivalents on the limited slide using coincident mates on their respective end faces, I then rebuilt and checked that they now moved in unison and obeyed the limits. Finally, I added the drawer, mating it only to master slide's moving element (inner face and outer end) plus planes for the vertical position and was relieved to find it all still worked!

I hope this helps.



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

SW2009x64 SP2.1 Pentium P4 3.6Ghz, 4Gb Ram ATI FireGL V7100 Driver: 8.323.0.0
SW2009x32 SP1.1 Pentium P4 3.6Ghz, 2Gb Ram NVIDIA Quadro FX 500 Driver: 6.14.11.7751

 

[ksauter]

So it sounds like you're mating the slides to the cabinet first, changing one slide to the 'no limits' version, tying them to each other so they move together, making sure that works, then mating the drawer to only ONE of the slides, and everything works. I'll try that. It shouldn't be this hard to make this kind of common operation work.

 

[SincoTC]

That's correct, it's the method I had to resort to in SW2005 to get my assembly model to work, in those days, two flexible sub assemblies HAD to have different configs to co-exist in an assembly even if they weren't connected in any way!

I believe this limitation was lifted in SW2006, when I modified my model so that both the slides had the same config and were connected by the drawer (like the real word) and they worked OK. We didn't pursue this development, so I had no reason to open it again until I checked it out the other day and found all the mates showed errors, I'm fairly certain that it was OK in SW2007, so this behaviour looks like a regression to the earlier limitations, possibly the relaxation caused bigger problems in other areas.

I think that SW should allow some kind of Master/Slave checkbox to enable flexible assemblies to be used in combinations like these without infringing their precious overcontraint criteria! I had no end of trouble with three hinges on a door in 07 and as it's now in production, I haven't the need (or the heart) to open again in 09!

I'm trying to work out suitable enhancements to these limit and tangent mates to put into an ER. I can see the logic behind the "overconstraint messages", however, you're quite right, it shouldn't be this hard!!

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

SW2009x64 SP2.1 Pentium P4 3.6Ghz, 4Gb Ram ATI FireGL V7100 Driver: 8.323.0.0
SW2009x32 SP1.1 Pentium P4 3.6Ghz, 2Gb Ram NVIDIA Quadro FX 500 Driver: 6.14.11.7751