HCb Posted March 17, 2015 Share Posted March 17, 2015 I'm trying to re-design my homemade CNC plasma table. I'm still trying to learn Inventor and this seemed like a good project to try to improve my Inventor knowledge and skills. I've got a problem I cannot resolve. I've re-designed (re-drawn) parts and started the whole assembly over from scratch a couple of times. I cannot get a constraint to work out for me. I want to see if someone can tell me what I'm doing wrong, please. I'm posting the files. I hope I got them all. I've started putting smaller parts together in what I believe are coherent units as assemblies. Since before I made the subassemblies all the way through several complete re-starts, I cannot get one part to constrain to another. Specifically: CRP Long Bearing HCB Brng:2 (as a part of CRP Long Bearing HCB Assby:4) can not be constrained with a tangent, outside constraint to X Axis CR Rail:1 (as a part of X Axis Rail:2) top side. I get a "assembly cannot be solved" error. There are four "bearings" modeled/drawn like this and which "ride" on top of the X Axis CR Rail:2. I cannot get any of them to sustain/support/tolerate this tangential constraint. The "bearings" on the other X axis (X Axis Rail:1) work fine. The X Axis Rails are constrained to be flush at the top, the parts all around should be identical. But, somehow, the "bearings" on the X Axis Rail:2 will not tolerate a tangential constraint to the "bearings" which ride on them. Like I said, I've re-drawn components (even completely re-drawing from scratch the CRP Long Bearings (which had come from manufacturer drawings originally)), and re-started the whole assembly several times. I don't understand why the constraints won't work. I'm still learning the software. I'm still trying to learn the paradigm of assembly constraints. Maybe I'm doing something stupid. I've looked at the problem so long, I've lost objectivity. I'd appreciate some help, please. Again, I hope I've attached all the relevant files. I do apologize for the size. Thank you. --HC _CNC Plasma 5.zip _CNC Plasma 7.zip _CNC Plasma 6.zip _CNC Plasma 4.zip _CNC Plasma 3.zip _CNC Plasma 2.zip _CNC Plasma 1.zip Quote Link to comment Share on other sites More sharing options...
JD Mather Posted March 17, 2015 Share Posted March 17, 2015 Every assembly should have one grounded or fully constrained component - some of yours do not. Use symmetry about the Origin workplanes. I very seldom use Tangent constraints. (even if that is what the parts "do") Mated workplanes are more robust. Think about the math the computer has to solve - you want to keep it a simple and bullet-proof as possible. The larger the assembly (more components) the more important this becomes. I didn't bother to try to find the true source of the error (it would be interesting to know as it probably indicates something is quite right). Quote Link to comment Share on other sites More sharing options...
JD Mather Posted March 17, 2015 Share Posted March 17, 2015 Given that I do not see any Horizontal or Vertical (or Tangent) constraints in this sketch - I would expect to find mis-aligned faces somewhere in the assembly. If you have done everything right - the missing constraints would not mean anything, but based on my experience when I see this, I find..... Quote Link to comment Share on other sites More sharing options...
HCb Posted March 18, 2015 Author Share Posted March 18, 2015 Every assembly should have one grounded or fully constrained component - some of yours do not. Use symmetry about the Origin workplanes. I very seldom use Tangent constraints. (even if that is what the parts "do") Mated workplanes are more robust. Think about the math the computer has to solve - you want to keep it a simple and bullet-proof as possible. The larger the assembly (more components) the more important this becomes. I didn't bother to try to find the true source of the error (it would be interesting to know as it probably indicates something is quite right). Every assembly should have one grounded or fully constrained component - some of yours do not. Use symmetry about the Origin workplanes. I very seldom use Tangent constraints. (even if that is what the parts "do") Mated workplanes are more robust. Think about the math the computer has to solve - you want to keep it a simple and bullet-proof as possible. The larger the assembly (more components) the more important this becomes. I didn't bother to try to find the true source of the error (it would be interesting to know as it probably indicates something is quite right). JD, thank you for taking the time to look at my files and give your input. It is appreciated. I was not sure about the grounded component in each assembly because I feared that if I grounded one part in an assembly it would prevent the whole assembly from "moving" (either in placement constraints or in motion simulation) in parent assemblies. I haven't played with it both ways to see. Obviously you wouldn't mention it if doing so was wrong. I'll go back and re-work the assemblies and add a grounded component to each. I'm not sure what "symmetry about the origin work planes" means. Do you mean in regards to sketches or to assemblies? I'll look through my book and do some online searches. I had not thought about the math involved in the constraints because, honestly, what this program does is way beyond my abilities in math and geometry (and I think I'm pretty good at that stuff...but this is Nth level magic). I'm not sure I'm understanding "mated workplanes". Would that be to create a workplane through the "bearings" and then mate that to the surface along which they are to "run", with an offset? In your second post, you have a PNG of a screenshot showing the profile is not fully constrained. Jeez. It's awful. I opened the files back up and found that the profile for each of the three parts based on it need 1,019 or so constraints. I originally got a file from 3dContentCenter (I think it was) but instead of a sketch and extrusion, it just had a solid (as I recall). I projected the loop to get a sketch which showed to be fully constrained. Somewhere in the mix, I lost the basis for the projected loop (I guess) and, yes, it's completely unconstrained. I began trying to constrain the shape but after I got to about "only" 500 missing constraints I started over with the original file which showed the shape to be completely constrained. I re-created all three pieces based on this fully constrained sketch and re-assembled the assemblies I posted here. I could still not get the other tangent constraints to work. So, I drew a really simple square tube to simulate the aluminum extrusion and re-created the three part files based on that sketch. I then re-created the assemblies using those basic shapes and was able to make the tangent constraints I had been unable to make before. There is something wrong with my part files (the underlying solids) which I got from the content center. I'm not sure what mechanism is used to get the part files from the manufacturer to that site (I believe these files were provided by the manufacturer in this case, but I could be wrong), but maybe there's something not right in there. Maybe I screwed something up. Regardless, when I used original sketches to make the extrusions (very simplistic shapes, very easy to constrain), the whole assembly worked. I'll revisit the artwork/part file problem at some point because I'd like to know where things went wrong Anyway, thanks again and I'll go "do my homework" on the symmetry and mated workplanes. --HC Quote Link to comment Share on other sites More sharing options...
HCb Posted March 18, 2015 Author Share Posted March 18, 2015 JD, I played with some work planes. If I understood you correctly then the attached files will show the use of work planes to allow the simulation of tangent constraints. I did it two ways; one with work planes in the sub-assembly, one with the work planes in the parent assembly. The WP in the sub-assembly worked easier for me as the part (sub-assembly) rotates to meet the needs of the constraints whereas the WP in the parent assembly moved around on the sub-assembly. This means that when the SA needs to be rotated to come together properly, it won't do it per the constraints I place on it when the WP are added in the parent assembly. Works great with the WP in the SA. The only downside to having the WP in the SA (that I found so far) is that the visibility of the WP can only be turned off at the SA level, affecting all SA instances in the parent assembly (which is behavior I would expect, of course). But it means that, in order to break constraints, tinker on something, then re-establish/re-create the constraints with the WP, I have to turn visibility on or off at the SA level. Turning it off in the parent assembly breaks connectivity (I forget the precise term) to the SA which, I think, defeats the purpose of having an SA. WP visibility isn't a huge big deal and the SA works better/assembles easier (with correct orientation of the SA) with the WP in the SA. Test Tang Assby1.zip --HC Test Test Assby2.zip Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.