Aligning sketch items on a construction line
@Anonymous - Colinear should work, as long as all the items are in the same sketch, and both items are lines. If you share your design we can take a look. However, a couple of comments:
- colinear (or any sketch constraint) will only affect the two lines being constrained. Unless you have your "shapes" well-constrained/dimensioned within themselves, then a constraint or dimension may distort your shape
- you could, as suggested here, define each shape as a sketch inside of its own component, and use Joints or Align to align them - this is one way around the problem above, but takes more work
Jeff Strater
Engineering Director
in the screencast below, I show how to go about using the first method. The basic idea is to dimension/constrain one of the shapes so that it is fully self-constrained. To test that, I usually just pick up and drag one of the lines. That shows you whether it is constrained, and also gives clues about what is unconstrained. A trick is that you can hit ESC to abort the drag. So, drag, ESC, add some dimensions, drag again, etc. Once it moves nicely, then colinear should work OK.
Jeff Strater
Engineering Director
Accepted solution
this is weird. I thought I posted this before. Maybe the spam filter grabbed it. See the screencast below for some tips on how to constrain each sketch shape. The idea is you try to drag it, see what happens, use ESC to cancel the drag, add more dimensions, etc. Once it is self-constrained, you can use colinear to constrain it to something else
Jeff Strater
Engineering Director
sorry for the delay, @Anonymous . It looks like your screencast didn't make it into the post. Sometimes posting screencasts is finicky. If nothing else, just paste a link to it here. It's probably something simple, it would be good to figure out what is different with your design.
Jeff Strater
Engineering Director
Accepted solution
Thanks, @Anonymous - that looks like a solver bug to me. As you showed, that upper section is fully self-constrained, so the colinear constraint should work to position that section with the line. If you have that version of the design available, I'd like to see it. I can try reproducing it from the original design as well.
The only thing I can suggest until I can see the design is to maybe Fix the construction line, to kind of kick the solver into one solution over another. Also wonder whether the first colinear constraint might be affecting the second, but that seems unlikely.
Jeff Strater
Engineering Director
I think I am starting to understand the problem somewhat. I tried to reproduce your dimensioning scheme today, from a still shot of the video:
sorry, hit "Post" too early. Anyway, I was trying to reproduce this set of dimensions in the earlier version of the model. And, I was unable to do so. I got it to this point:
and was unable to add the circled dimensions, without over-constraining the sketch, so they get added as driven dimensions.
I noticed a few strange things in your design:
There are a couple of dimensions that exist on the same geometry, one driven, one driving. I don't think that is a problem, but I also don't think that I can say for certain that it is NOT a problem, either. I also noticed that the angle dimensions indicated were different from each other. Not sure if that was intentional, but given how symmetric this section is, I was surprised that those angles are different.
So, what I think is happening is not that the colinear constraint is failing, but that the dimension system in this block of stuff is already in not good shape, and the colinear constraint just reveals it. My recommendation would be to delete a bunch of these dimensions, and try to re-dimension the geometry again, and see if you can get a minimal set of dimensions. I'll try to do that now, and see what I can come up with.
Jeff Strater
Engineering Director
this is the minimal set of dimensions I could come up with that is in line with your scheme, but works with the colinear constraint to a fixed vertical line:
There are plenty of dimensioning approaches that would also work. The next step I did was to relate the various duplicate values to each other using an equation:
This should make it a bit more stable - make sure that these values don't diverge from each other.
Jeff Strater
Engineering Director