cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
Showing results for 
Search instead for 
Did you mean: 

Trouble with sketch constraints

Amethyst

Trouble with sketch constraints

I've been struggling for hours on keeping a fairly simple sketch together.

What isn't simple is the fact that I have a family table and one of the instances has it's CSYS rotated 180 degrees.

I probably could break out of the family table and get the sketch to work in it's model that would have the rotation completed prior to the sketch. It's the rotation that unglues the sketch.

The strange thing is that all the other sketches don't seem to have an issue with the Csys rotation. I have a twin to this sketch that has no issues. I created it from an earlier model and I've been trying to see why it succeeds and the other fails.

My sketch has 2 lines on the ends which reference edges. In between this is an angled line which joins the 2 end lines with a radii on each side. The 2 radii are constrained to end at the edge of each of the edge referenced lines.

The one other complication is that the reference edges are offset with this other family tree instance.

After the family tree verifies as a failure I go into this instance. As I mentioned I see the twin sketch succeeds, and the immediately following sketch shows as a failure. I go into the sketch and when I look at it's references it tells me they are not updated. I update the unresolved references and solve the sketch. When I look at the results I see that the reference edges are in their true position but the tilted line in the center along with the 2 connecting radii are not rotated into position. The radii are skewed and don't connect anything. My guess is that the biggest issue is that the tilted line will not rotate into position.

I've tried everything to make this sketch stick together. I've made the radii as a circle as well as a round. I've dimensioned the angled line to a horizontal reference curve, the reference edge line as well as a construction line that I placed. I've tried locking the entities. I've tried using splines instead of radii to connect the tilted line to the end reference edge lines. Nothing seems to stick.

Is there a trick to keeping a sketch together after a csys gets rotated? How can you be sure to get all sketch entities to rotate after the CSYS rotation?

I presently don't have access to my CAD PC, but I should be able to upload the model should anyone have questions on it.

4 REPLIES 4

Re: Trouble with sketch constraints

You are trying to do the one thing Creo, Wildfire, and Pro|E are simply not equipped to handle. Better to simple pattern the sketch about an axis than to associate it to a different CSYS.

I didn't read all of the text above, but the bottom line is that things flip way to easily in Creo. It is a problem in every respect with no simple solutions. This happens in sections; patterns; and assembly constraints. Something PTC has never sought to enhance.

Re: Trouble with sketch constraints

Creo certainly is a strange paradox. It does the difficult things easy and makes the easy things difficult.

I don't mean this disrespectfully but I have been stumbling into a number of 1 things that Creo can't handle. It seems like anything associated with manipulation of existing data.

I appreciate that you've pointed this out to me.

It's just strange that only 1 sketch isn't surviving the CSys rotation.

Re: Trouble with sketch constraints

It is a pretty accurate observation, Paul. Another that is highly sensitive is projected curves in sketches such as text. If you upsize the text in the master, projection looses its references easily.

Re: Trouble with sketch constraints

Antonius,

Thank you for the additional warning. This too might be something that could bite us down the road as we perform the projection prior to scaling up.

Last week we discovered that our essential part profile sketch somehow lost it's associativity after we modified a sketch offset from the original.

A couple weeks ago we discovered that Creo only does a true mirror in one direction.

We are discovering awful software quirks in the midst of a rush of a series of jobs. We are losing days because of the quirks.

We could abandon a practice that we see might not work right away but how do we even know it is a practice and not a settings issue.

The limitations we find with moving solid parts is what brought us to use an additional CSYS to get the desired rotations.

I will submit a ticket to PTC and at least let them know that this is a very limiting issue.