The case of the missing table legs

Replies to This Discussion

Permalink Reply by Ethan on October 21, 2015 at 2:08pm

when you fillet the curve you lose the vertices that are driving your legs. see the example :)

Attachments:

• TubeTable2.gh, 14 KB

Permalink Reply by Kim hauer on October 21, 2015 at 3:00pm

I don't want mitered table rail corners. I want all intersections filleted just like the single leg which attaches to the table rails. :)

Permalink Reply by martyn hogg on October 21, 2015 at 3:25pm

When you fille the polygon curve you lose the vertices apart from one which is probably the end of the curve. That is why you also lose the legs because the vertices drive the legs.

Just double up the polygon component and fillet one. Use the one without the fillet for the vertices.

Attachments:

• TubeTable.gh, 18 KB

Permalink Reply by Kim hauer on October 21, 2015 at 4:18pm

Thanks martyn:

Yes that works when using 4 Table Sides (legs) but for some reason increasing or decreasing Table Sides (legs), makes the Table Rails now disappear...yikes!

Permalink Reply by Joseph Oster on October 21, 2015 at 4:51pm

I'm not sure what I did differently but it seems to work for me (attached).

P.S.  Parts get lost in the 'Mesh Union (MUnion)' - use 'Solid Union (SUnion)' instead!

Attachments:

• TubeTable_2015Oct21a.gh, 21 KB

Permalink Reply by martyn hogg on October 21, 2015 at 5:23pm

I just tried it and at first, with 5 legs, the rails disappeared and then after changing the slider, they reappeared!

My SUnion didn't work though!

Perhaps this would be better done with just lines and the excellent Exoskeleton component?

Attachments:

• TubeTable.gh, 23 KB
• TubeTableExo.gh, 16 KB

Permalink Reply by Joseph Oster on October 21, 2015 at 5:27pm

Did you use my code?  Looks rock solid to me, from 3 to 10 legs.

P.S.  I just tried yours and it breaks - your code is different; you added a 'Crv CP' instead of using the 'DeBrep' vertices ('V') for the 'Cyl B' input.

Permalink Reply by martyn hogg on October 22, 2015 at 2:47am

I just used the CrvCP to get the table leg points back onto the filleted polygon.

Maybe the SUnion is a bit flaky because joining cylinders with identical rads creates those infinite slithers inthe corners. Your fix below with the varying rads solves this!

Permalink Reply by Joseph Oster on October 22, 2015 at 8:01am

OK, I understand now that you used 'Crv CP' for the same purpose that I used 'Pull' point.  So besides the 'x-0.001' expression I used, the big deal was using the same radius for the fillet as the radius of the table legs and 'Pipe'.

Perseverance furthers.

-- I Ching

Permalink Reply by Kim hauer on October 21, 2015 at 6:43pm

If I'm not mistaken both martyn and joseph use Mesh union in each of their revised scripts,
but for some reason only joseph's works on my computer..... I'm visually not able to see any other difference.
(it would be great to have a compare feature, would came in handy after spending to much time the darn computer:)  )

martyn I also like your Exo alternative .....cool!

Permalink Reply by Joseph Oster on October 21, 2015 at 7:10pm

Yes!  A compare feature for GH code would be AWSOME!!

I didn't notice the need to replace 'MUnion' with 'SUnion' until after I posted my code because I hadn't disabled the preview for the components that feed the union.  I changed the screen shot but can't change an uploaded file without deleting the post and re-posting.

I was just about to upload it but now it breaks on polygons of 3, 6, 8, 9 or 10 sides!!???  How bizarre and frustrating...  Maybe I wasn't looking closely enough before.  Probably the 'Pipe' around the table top, I've had 'SUnion' trouble with pipes before.

I just noticed the negative slider value being used for 'Length' of the table legs - that's odd.

Attachments:

• TubeTable_2015Oct21b.gh, 22 KB

Permalink Reply by Joseph Oster on October 21, 2015 at 7:44pm

OK, here's something a little different that is working for all polygons...

1. polygon 'Fillet Radius (Rf)', 'Pipe' radius and 'Cylinder' radius are all set by the same slider.
2. an expression is used on 'Cylinder' radius (the table legs) to subtract 0.01 from that value. (it fails otherwise)  (P.S. 0.001 also works)
3. the vertices of the polygon without the radius, determined by 'Discontinuity (Disc)' points, are 'Pull'ed to the radiused polygon and used as the centers for the cylinders (table legs).

These steps seem to line things up such that 'SUnion' always works!?

Attachments:

• TubeTable_2015Oct21c.gh, 19 KB

• ‹ Previous
• 1
• 2
• Next ›
• Page