This content has been marked as final. Show 7 replies
That is a tour de force in surface modeling! You're very creative using SW.
I got it to knit into a single solid body by twiddling with the edges & curves used to define the surface lofts for the taper feature. I suppressed some of your features that I didn't think were essential to the model.
I used Insert>delete>face to get rid of the bottom side of the Extrude-Lands because they didn't appear in your flat version. Was that a good guess?
1. I owe you a beer or several! Maybe I should try to make it to a Cosug meeting one of these days.
2. Perhaps I did a bad thing. Prior to posting the question, I deleted all other features to the part not related to this taper feature before uploading it. So I set to repeat what you did into my original part. Here's what the part should look like with the extra features: http://www.ascent-design.com/data/Beam_OML.jpg
3. I've never used or even heard of DeleteFace before, and so looked it up. To clean up the bottom edge of the protruding Extrude-Lands and Box features, I made a surface offset=0.000 early in the model and later used it for SurfaceCut to clean it up; a straightforward move.
4. I re-pointed all 37 projected curves onto the new DeleteFace surface to rebuild all those Lofts - I figure you did the same.
5. The part would finally build, although at this time the screw patterns have failed.
I consider this a big leap forward. The new part is at:
Here is my take on your modeling challenge.
Don't be so quick to create the solid. The surfaces give you a lot of options. I waited until after creating the solid to add the inside fillets.
Add all your holes at the end after creating the solid.
You can take the two helper surfaces that are hidden and do a RMB and Delete them so they do not show up in your model. The Delete is a feature that can be suppressed if you need the surfaces back in the future.
I suspect some of the surfacing wizards here can show a more efficient modeling technique then what I used. Surfaces are your friend and can solve a lot of modeling challenges that are difficult with solids alone...... :-)
Beam_OML_AMW.zip 3.9 MB
An alternate way of handling your fillets.
Another way for you to get rid of the extra bosses on the bottom of your part in your original model is to use a Surface Cut. Insert > Cut > With Surface. On your model it fails, which is a sympton of your current feature order and modeling technique.
Beam_OML_AMW_2.zip 5.1 MB
Perhaps you downloaded after I uploaded a revised part in response to Gerald.
- Of course the hole patterns are moved last when using a surface approach.
- The Cut with Surface operation has been completely robust. It helped clean up the tilted boxes that you guy's were not originally privy to.
- Because Gerald's delete face changed the part back to surface, I too wondered why not stay surface. However it worked just fine.
- Adding fillets last as you suggest is certainly more robust. I should know that by now....
- What does RMB mean? Why would I delete the helper surfaces versus just hiding them?
My last fix is to get the Hole Wizard hole pattern to build, and await some inevitable change from my client. It's my first time trying a whole wizard on a 3D sketch, I go thru contraining points to the outer surface, then putting them concentric to a external reference sketch circle. There must be a more robust way. It might be obvious that this thing is the chassis in a large assembly already build around the Flat proof model.
Thanks again, Ralph
I used your original model... :-)
RMB stands for Right Mouse Button.
Deleting the surfaces helps clean up the model, which can be important when exporting models as IGES or STEP for milling. Your extra helper surfaces will not end up in the translation. The extra surfaces can cause confusion to downstream users be they SolidWorks users or users of translated data.
Lots of ways to go about modeling your part I am sure. It is interesting to see the solutions different people come up with.
...I should try to make it to a Cosug meeting one of these days. You're always welcome! Next meeting is 06/07/07 - Vic Leventhal is our featured guest. If you can't make it to a user group meeting, the SWUGN Tech Events are terrific. I picked up a lot of surfacing tips at the one in Kansas City. See http://www.swugn.org/pages/sum..._Technical_Summit.html
...I've never used or even heard of DeleteFace before... Your solution using a zero offset face and a Surface Cut is fine. Delete Face is a cute trick that sometimes is a real time saver in modeling. It might be interesting to do a time study comparison to see which rebuilds faster. I agree with Anna's sentiment that extra surfaces are not always a blessing.
...I re-pointed all 37 projected curves onto the new DeleteFace surface to rebuild all those Lofts - I figure you did the same. Yes, that was a time burner. As I proceeded thru the tedium, I was thinking about different ways to produce the taper. Anna's idea of chamfering or filleting a solid has definite apeal.
It sounds like you're on track to deliver a "final" model. Congratulations! 90% of the stuff I do is prismatic and I'm happy about that (I consider myself to be a rookie when it comes to surfacing). Matt Lombard is a master of this stuff. His book http://www.amazon.ca/SolidWork...aul-Tran/dp/0470080132 is worth a read. Your approach in this model follows his notion of "hybid" modeling - a combination surface and solid modeling. The point is to use the technique that solves the problem; the trick is to understand what the kernel is doing so that you get the desired result with minimum mouse hammering.