Swoopy headache
Swoopy headache
(OP)
I am working on a swoopy part (sorry no pictures available) that is constructed as follows:
1. Create some 3D curves of the edges.
2. Use the fill feature to connect these edges.
3. On one side use the freeform feature to tweak that side.
4. Use variable fillets to put nice rounds on the whole thing.
Here is the problem.
When creating a sketch by passing a plane through the resulting body it is extremely difficult to get the ends of intersection curves to match up and stay matched up where the variable radius fillets join the fill feature derived surfaces.
Anybody have thoughts on how to tame this beast?
1. Create some 3D curves of the edges.
2. Use the fill feature to connect these edges.
3. On one side use the freeform feature to tweak that side.
4. Use variable fillets to put nice rounds on the whole thing.
Here is the problem.
When creating a sketch by passing a plane through the resulting body it is extremely difficult to get the ends of intersection curves to match up and stay matched up where the variable radius fillets join the fill feature derived surfaces.
Anybody have thoughts on how to tame this beast?
TOP
CSWP, BSSE
www.engtran.com www.niswug.org
"Node news is good news."






RE: Swoopy headache
Scott Baugh, CSWP![[pc2] pc2](https://www.tipmaster.com/images/pc2.gif)
FAQ731-376: Eng-Tips.com Forum Policieswww.scottjbaugh.com
RE: Swoopy headache
Rob Stupplebeen
RE: Swoopy headache
Scott Baugh, CSWP![[pc2] pc2](https://www.tipmaster.com/images/pc2.gif)
FAQ731-376: Eng-Tips.com Forum Policieswww.scottjbaugh.com
RE: Swoopy headache
Rob Stupplebeen
RE: Swoopy headache
There are a couple of issues.
TOP
CSWP, BSSE
www.engtran.com www.niswug.org
"Node news is good news."
RE: Swoopy headache
Instead of selecting individual faces to create the intersection curve, select the surface/solid body (in the feature tree). This should point to the surface/solid body as it is at the time of the sketch, as opposed to specific faces that happen to intersect your sketch plane. As a result, this ought to allow substantial editing upstream without failed sketch elements. Ought.
Jeff Mowry
www.industrialdesignhaus.com
A people governed by fear cannot value freedom.
RE: Swoopy headache
Rob Stupplebeen
RE: Swoopy headache
I do that when I can. Sometimes the curve on the face won't allow it's endpoint to do a pierce with the curves edge. I think this is because of the mixture of fill surface and variable radius fillet. One is algorithmic and the other is analytic (rolling ball) so the intersection is calculated and may not be on either surface.
Jeff,
I never thought to go back to the solid bodies. I'll have to do a lot of cleanup and delete the curves that I don't need, but I'll give it a shot.
TOP
CSWP, BSSE
www.engtran.com www.niswug.org
"Node news is good news."
RE: Swoopy headache
TOP
CSWP, BSSE
www.engtran.com www.niswug.org
"Node news is good news."
RE: Swoopy headache
TOP
CSWP, BSSE
www.engtran.com www.niswug.org
"Node news is good news."
RE: Swoopy headache
Imagine you want to split Face A with Sketch1 and Sketch2. First split the face with Sketch2 and then with Sketch1 but make sure that Sketch2 references Sketch1.
Seems like if the Sketches and dependencies their are not in the same order as the Split features that use them there is a little confusion in the ordering and naming of vertices and edges. If you are creating a model this in not very likely, but when working on an old model where the Sketch position in the tree is no so apparent this can happen. That's my story and I'm sticking to it.
Thanks for the help. I don't think there is a solution for the edges not being able to take a pierce relation.
TOP
CSWP, BSSE
www.engtran.com www.niswug.org
"Node news is good news."
RE: Swoopy headache
I noticed face/surface confusion amp up (considerably) in v2008. By comparison, v2007 was much better at surviving the CTRL-Q rebuilds of a surfacing-heavy model. Nearly every surface Trim, by contrast, in v2008 would fail to remember which side of the surfaces to keep and which to discard with a hard rebuild. I never found a rational explanation or solution to this problem, and it's caused me many, many lost hours of work.
Good luck with this current issue. By the way, I think you can also select a feature (such as Variable Fillet) for your intersection curve generation as well (not sure). Might help if the solid bodies are still a mess.
By the way, where are your Variable Fillets being used? If you're filling between surfaces, you know you can tell the Fill to be tangent to those reference surfaces from which you're deriving your edges, right? And since this is true, you wouldn't need a Variable Fillet on those edges, right? Just curious. Variable Fillets are quite twitchy.
Jeff Mowry
www.industrialdesignhaus.com
A people governed by fear cannot value freedom.
RE: Swoopy headache
But this ordering issue that I had is one of the weaknesses of the feature tree. Several 3rd party vendors, have, over the years, come up with software that will map the true dependencies between features. I think I use the Parent/Child dialog box more than many. You can actually do some operations there via a RMB. But the P/C dialog is confusing and the feature tree really doesn't give the true picture as far as creation order. In it's day it was miles ahead of Pro/E, but now it is showing it's age.
One trick to find out what is really important in a feature is to bring up the delete dialog, but just don't delete. Even there you don't know which children will actually be deleted and which will survive delete with injuries.
I enjoy troubleshooting as much as the next guy, maybe more. But I was having serious brain cramp and you guys were a big help.
TOP
CSWP, BSSE
www.engtran.com www.niswug.org
"Node news is good news."
RE: Swoopy headache
Earlier this year, I completed a very surface-heavy design for a hand tool (using client-specified v2008). I modeled all the main bodies in a single part, and then used the Save Bodies feature to save each body into its own part file for later detailing. This is a high-risk way to do things, since if a feature fails high in the parent part's tree, you've got perhaps thousands of down-stream features at risk of being trashed.
This did, finally, happen. A surface Trim decided to fail in such a way as to not let me reselect which side of the surface to keep and which to discard. I simply rolled above this feature, completed the Trim (a newly-created feature worked fine, but the existing feature--inexplicably--could not be repaired), and set to repairing all the down-stream damage. Most of this involved re-associating the downstream features with the newly-trimmed surface body. But there were certain features which were trashed and irretrievable. But neither could I delete them, though the downstream features really didn't depend on these. I finally had to shove these in a folder called "hosed" and leave all the red failures as part of my part file. It would seem SolidWorks itself doesn't understand all the parent/child relationships, since I'd completely made all failed upstream features irrelevant to everything downstream--and yet couldn't delete those features without losing hundreds of downstream features. Frustrating.
Jeff Mowry
www.industrialdesignhaus.com
A people governed by fear cannot value freedom.
RE: Swoopy headache
TOP
CSWP, BSSE
www.engtran.com www.niswug.org
"Node news is good news."