Modelling Challenge

Recommended Posts

I had to create this shape today: A circle which morphs into a rectangle. The shape had to be absolutely symmetrical. Sounds easy but is it? Who can do it?

I tried multiple extrude and it didn´t look smoth

Can you show us?

I have TWO possible theories, but I'll let others give it a shot first!

Loft. two rails. Two nurbs surfaces. Make sure direction is same for both shapes. The green = nurbs in same direction, the pink: nurbs in differing directions…

Edited by mjm
better screen shot of Image of result added

Here's my attempt. There's some variability without the knowing the exact proportions but I'm pretty sure I know how its done. It was the curves on the surface and the highlighted isocurves that gave it away. I don't think its done by lofting.

I'll hold off posting the method for now.

Kevin

1 hour ago, Kevin McAllister said:

Here's my attempt. There's some variability without the knowing the exact proportions but I'm pretty sure I know how its done. It was the curves on the surface and the highlighted isocurves that gave it away. I don't think its done by lofting.

I'll hold off posting the method for now.

You are very close, in fact it looks like you have cracked it. And yes, you are right, it's not done via a loft. The loft will not give you any precise control over the alignment of the curves. Is that a true circle at the bottom though?

Edited by Stephan Moenninghoff

Here's my attempt!

Edited by markdd

8 hours ago, markdd said:

Here's my attempt!

OK, not bad. The top must be a rectangle though, not a square. If you can do that I would consider it solved.

Edited by Stephan Moenninghoff

2 hours ago, Stephan Moenninghoff said:

The top must be a rectangle though, not a square.

Ok. Try this!

What I can't do is make the corners straight. There always seems to be an outward facing curve to them. Yours and Kevin's seem to be slightly curved in.....

That's the best i can do for now!

I think my solution might not be very straight forward. Can't get the inward facing curves.

Curious for the solution.

Here is my version, and as I expected mine is not very straight forward.

Thanks for showing your solution. Maybe we should this more often here. I like those quick modeling challenges. Always learn a lot.

I think I also discovered a bug in the process.

Btw. I used the german version but I hope it's half way clear how I did it.

Here is the bug I encountered. Flipping normals causes the surface to disappear.

Edited by elepp

17 hours ago, Stephan Moenninghoff said:

You are very close, in fact it looks like you have cracked it. And yes, you are right, it's not done via a loft. The loft will not give you any precise control over the alignment of the curves. Is that a true circle at the bottom though?

The bottom is not a true circle in mine.

59 minutes ago, elepp said:

Here is my version, and as I expected mine is not very straight forward.

I had a version done this way too. It seemed like the first thing to try.

4 hours ago, Stephan Moenninghoff said:

@elepp and @markdd I think you both have solved it. In Mark's example I am still not sure the bottom is a true circle. If it isn't, it can't be considered as solved!

The sides don't curve inwards BTW. It may look like they do depending on the view but the sides are straight. Here is how I did it. Curious to learn your methods.

Ah, the Deform tool.... I considered that. Mine started as a loft from a square to a rectangle. I also used the Fillet Edge tool with a variable radius edge. The one anomaly I found is that it only worked for the first corner and then I had to reverse it and change the order to do the other three corners. (This might be a normals bug too @Jim Wilson ? )

This was a fun challenge.

Kevin

1 minute ago, Kevin McAllister said:

The one anomaly I found is that it only worked for the first corner and then I had to reverse it and change the order to do the other three corners. (This might be a normals bug too @Jim Wilson ? )

Not sure about normals being at fault (I didn't normally pay them much mind but theyre coming up more often in troubleshooting it seems), but having to do certain corners or certain edges first always got to me. I'll submit this one as I was able to replicate that fairly quickly.

I did it Kevin’s way too! I also saw the same normals “bug”.

I need to learn more about Normals and their use within Vectorworks. Could anybody point me to a resource that could help explain their concept and usefulness in CAD?

I like @elepp's surface method. It may be more work but it creates an even cleaner geometry than mine. I try to solve everything using solids because the surface modeller always creates hollow objects which may create problems down the road (Push/Pull does not play nice with most surface objects). I have a few more challenges up my sleeve. I'll dig them up when I have a few minutes. Thanks for contributing, guys. This was fun!

Edited by Stephan Moenninghoff

What a great tutorial. Thanks folks.

1 hour ago, Stephan Moenninghoff said:

I try to solve everything using solids because the surface modeller always creates hollow objects which may create problems down the road (Push/Pull does not play nice with most surface objects).

I normally use Stitch and Trim or the Create Solid checkbox in the Loft Dialog to turn my surfaces into solids. In newer versions the solids geometry is pretty clean. If you find anything that misbehaves I would consider bug submitting it.

Kevin

Thanks @Stephan Moenninghoff for the challenge and all of you for discussion and solutions!

Cool, great tutorials and a lot of fun to learn something about modeling beside the normal hacker work 😉

18 hours ago, Stephan Moenninghoff said:

I like @elepp's surface method. It may be more work but it creates an even cleaner geometry than mine. I try to solve everything using solids because the surface modeller always creates hollow objects which may create problems down the road (Push/Pull does not play nice with most surface objects). I have a few more challenges up my sleeve. I'll dig them up when I have a few minutes. Thanks for contributing, guys. This was fun!

That's true and annoying at the moment. If I use push/pull on the top surface it disappears. Best way is to add a solid on top and combine the to volumes. It's also not very intuitiv that you can apply the boolean function of adding volumes together on nurbs surfaces.

27 minutes ago, elepp said:

That's true and annoying at the moment. If I use push/pull on the top surface it disappears. Best way is to add a solid on top and combine the to volumes. It's also not very intuitiv that you can apply the boolean function of adding volumes together on nurbs surfaces.

Yes, I don't think that's actually a bug that would merit logging. More like a limitation in what surface modelling can and cannot achieve.

Esp @Stephan Moenninghoff @elepp @Kevin McAllister, and also @all others - Many thanks! Challenge is a great idea.

I have had so many problems using the variable fillet in past, that I just ignore it. UNTIL NOW that is.

Maybe a tutorial on this needs added to the video list?

Wish - radius waypoints on edge by length or stations (document units) in addition to percentage.

Great to be prodded into accepting, trying, retrying tools and processes.  Many thanks!

-B

47 minutes ago, Benson Shaw said:

I have had so many problems using the variable fillet in past, that I just ignore it. UNTIL NOW that is.

^ Same for me too...

47 minutes ago, Benson Shaw said:

Maybe a tutorial on this needs added to the video list?

Wish - radius waypoints on edge by length or stations (document units) in addition to percentage.

^ I would love a live preview like in Rhino, where you can drag/add/subtract radii points in relation to the edge you wish to fillet.

Kevin

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.

Reply to this topic...

×   Pasted as rich text.   Restore formatting

Only 75 emoji are allowed.

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×

7150 Riverwood Drive, Columbia, Maryland 21046, USA   |   Contact Us:   410-290-5114

© 2018 Vectorworks, Inc. All Rights Reserved. Vectorworks, Inc. is part of the Nemetschek Group.

×

• KBASE
• MARIONETTE

×
• Create New...