Comments on: When in doubt… Check the mates in your subassemblies! http://gabijack.com/2008/09/when-in-doubt-check-the-mates-in-your-subassemblies/ Gabi's Blog on solid modeling, design and engineering Tue, 07 Sep 2010 19:25:29 +0000 http://wordpress.org/?v=5211 hourly 1 By: Markko http://gabijack.com/2008/09/when-in-doubt-check-the-mates-in-your-subassemblies/comment-page-1/#comment-274 Markko Mon, 29 Mar 2010 10:01:37 +0000 http://gabijack.com/?p=123#comment-274 I just had the same kind of thing with an adjustable drop tube on a medical device. The 2 limitdistance mates in 2 nearly identical subassemblies has had me running around in circles until I found your site. The other 70-ish mates kept failing left, right and centre, and suppressing/deleting them just pushed the errors somewhere else. Maybe SW could get their act together on what is effectively an essential feature in a subassembly to make things work like 'real world' components. The only way the limitdistance works is to put the part in the top level assembly, which effectively means that the whole thing has to be put together before the restrictions apply - not very 'real world' On thing I noticed though - it might depend on the 'level' of the subassembly too. I had a limitangle in a small subassembly, which worked fine in the next up (small) assembly, but failed miserably when this was placed in the top level assembly. I found that it would not drag, or when it did, the other components would lose their (perfectly acceptable) mates and fail. Just to bleat a bit more, I'd be happy if SW stopped putting new features in their software for a bit, and started making the (very capable) stuff they already have work properly/stably. Wouldn't go down well with the marketers, but then marketers aren't engineers, are they? Rant done. Thanks for the tips I just had the same kind of thing with an adjustable drop tube on a medical device. The 2 limitdistance mates in 2 nearly identical subassemblies has had me running around in circles until I found your site. The other 70-ish mates kept failing left, right and centre, and suppressing/deleting them just pushed the errors somewhere else. Maybe SW could get their act together on what is effectively an essential feature in a subassembly to make things work like ‘real world’ components. The only way the limitdistance works is to put the part in the top level assembly, which effectively means that the whole thing has to be put together before the restrictions apply – not very ‘real world’

On thing I noticed though – it might depend on the ‘level’ of the subassembly too. I had a limitangle in a small subassembly, which worked fine in the next up (small) assembly, but failed miserably when this was placed in the top level assembly. I found that it would not drag, or when it did, the other components would lose their (perfectly acceptable) mates and fail.

Just to bleat a bit more, I’d be happy if SW stopped putting new features in their software for a bit, and started making the (very capable) stuff they already have work properly/stably. Wouldn’t go down well with the marketers, but then marketers aren’t engineers, are they? Rant done.

Thanks for the tips

]]> By: Andrew Rodgers http://gabijack.com/2008/09/when-in-doubt-check-the-mates-in-your-subassemblies/comment-page-1/#comment-273 Andrew Rodgers Mon, 09 Mar 2009 20:18:19 +0000 http://gabijack.com/?p=123#comment-273 I've had exactly the same problem for months with an assembly that includes two pairs of cylinders. Everything is fine as long as I suppress one of each pair. I've become convinced that the way in which SolidWorks solves the limit distance mates is the problem. I’ve had exactly the same problem for months with an assembly that includes two pairs of cylinders. Everything is fine as long as I suppress one of each pair. I’ve become convinced that the way in which SolidWorks solves the limit distance mates is the problem.

]]> By: Michael Lord http://gabijack.com/2008/09/when-in-doubt-check-the-mates-in-your-subassemblies/comment-page-1/#comment-272 Michael Lord Mon, 29 Sep 2008 07:15:57 +0000 http://gabijack.com/?p=123#comment-272 I get this same problem quite offer when using Limit Mates. I use quite a few gas struts, accuators and hinges. Everthing will moves quite smoothly then out of no where a warning and error mates appear. This is common if I use a folding hinge (similar as you have) and a gas strut pushing. I find (most times) Ctrl Q rebuilds and errors are corrected. I get this same problem quite offer when using Limit Mates. I use quite a few gas struts, accuators and hinges. Everthing will moves quite smoothly then out of no where a warning and error mates appear. This is common if I use a folding hinge (similar as you have) and a gas strut pushing. I find (most times) Ctrl Q rebuilds and errors are corrected.

]]> By: Gabi http://gabijack.com/2008/09/when-in-doubt-check-the-mates-in-your-subassemblies/comment-page-1/#comment-271 Gabi Fri, 26 Sep 2008 12:37:24 +0000 http://gabijack.com/?p=123#comment-271 Thanks for the info! BTW, I was in a hurry yesterday when I answered your first comment and I swear I saw you had two of the bars constrained. Perhaps it's because in my mechanism your horizontal bar is vertical for me, so it seemed to me that you had a vertical and a horizontal bar constrained. Anyway, I took a better look at your mechanism and you are right, it is the same. You are also right about the angle and the whole sketch getting overdefined if you try to make that angle a value of 10, for instance. However, I think this is not the reason why my assembly would become overdefined. See, try to model this using blocks or, better yet, model some bars of the same lenght of those you drew there and assemble them together. Then move the driver (L1) clockwise and observe how the mechanism behaves. When the driver and the horizontal bar are at an angle of around 30, the mechanism locks. A similar thing happens if you move the driver counterclockwise. This is the way the mechanism behaves, but unless you try to apply an ANGLE mate of less than 30 between the driver and the horizontal bar, the assembly shouldn't become overdefined. The mate that was applied in my assembly and in the one with the cylinders was a LIMIT mate. It limits the range of motion of two components in the assembly within two values, but it doesn't constrain them to be any of those two values. Try applying a LimitAngle mate between the driver and the horizontal bar and you'll see that this doesn't overdefine the assembly either, even if you specify the limits as 0 to 50, for instance. All that you'll see is that the driver can oscillate between 50 and 30 and that's it. According to the limitangle mate, it could go all the way to 0, but because of the way the mechanism is designed, it will continue to lock at 30. Well, this is what I think, based on what I see, but I could be wrong too, of course. If this is right, you could have one of these mechanisms pushing the flat component without causing the assembly to become overdefined, but when you add a second one (both of them with limit mates) that's when it becomes overdefined and I think it is because SW interprets it as if you were trying to apply the same mate twice. Like I said, I could be wrong. Thanks for the info! BTW, I was in a hurry yesterday when I answered your first comment and I swear I saw you had two of the bars constrained. Perhaps it’s because in my mechanism your horizontal bar is vertical for me, so it seemed to me that you had a vertical and a horizontal bar constrained. Anyway, I took a better look at your mechanism and you are right, it is the same. You are also right about the angle and the whole sketch getting overdefined if you try to make that angle a value of 10, for instance. However, I think this is not the reason why my assembly would become overdefined. See, try to model this using blocks or, better yet, model some bars of the same lenght of those you drew there and assemble them together. Then move the driver (L1) clockwise and observe how the mechanism behaves. When the driver and the horizontal bar are at an angle of around 30, the mechanism locks. A similar thing happens if you move the driver counterclockwise. This is the way the mechanism behaves, but unless you try to apply an ANGLE mate of less than 30 between the driver and the horizontal bar, the assembly shouldn’t become overdefined.

The mate that was applied in my assembly and in the one with the cylinders was a LIMIT mate. It limits the range of motion of two components in the assembly within two values, but it doesn’t constrain them to be any of those two values. Try applying a LimitAngle mate between the driver and the horizontal bar and you’ll see that this doesn’t overdefine the assembly either, even if you specify the limits as 0 to 50, for instance. All that you’ll see is that the driver can oscillate between 50 and 30 and that’s it. According to the limitangle mate, it could go all the way to 0, but because of the way the mechanism is designed, it will continue to lock at 30. Well, this is what I think, based on what I see, but I could be wrong too, of course.

If this is right, you could have one of these mechanisms pushing the flat component without causing the assembly to become overdefined, but when you add a second one (both of them with limit mates) that’s when it becomes overdefined and I think it is because SW interprets it as if you were trying to apply the same mate twice.

Like I said, I could be wrong.

]]> By: CrazyCAD http://gabijack.com/2008/09/when-in-doubt-check-the-mates-in-your-subassemblies/comment-page-1/#comment-270 CrazyCAD Fri, 26 Sep 2008 04:55:25 +0000 http://gabijack.com/?p=123#comment-270 this is what i found as the linkage for the wing flap mechanism. looks like two four-bar mechanisms coupled to each other. <a href="http://i36.tinypic.com/23gzpfm.jpg" rel="nofollow">http://i36.tinypic.com/23gzpfm.jpg</a> this is what i found as the linkage for the wing flap mechanism. looks like two four-bar mechanisms coupled to each other.

http://i36.tinypic.com/23gzpfm.jpg

]]>