square "pipe" on numerous Brep Edges

Replies to This Discussion

Permalink Reply by peter fotiadis on April 27, 2015 at 10:06am

Peace of cake ... actually not quite (see the setting for the offset corner type > sharp doesn't work).

PS: somehow we reinvent the wheel with that one (how the brep is created?)

I'll be back, best, Arnie S

Attachments:

• squarepipe_V1.gh, 22 KB

Permalink Reply by AdnFrs on April 27, 2015 at 12:06pm

Thank you Peter!

This looks like it works very well. Do you mind walking me through your thought process on this so I can learn a bit? 

Would you be able to explain the SecBrepPlane command and how you made it? Would it be possible to accomplish this without creating a custom component?

Thank you once again!

Permalink Reply by peter fotiadis on April 27, 2015 at 12:19pm

Er... I did nothing actually since the only way to work that thing is with the round corner style (which I suspect IS NOT desired).

See PlanB as well (via Sweep1) that also doesn't work > but I took that @#$#@ sweep1 personally > more soon (It works in Rhino, mind):

As regards the SecBrepPlane "command": it's C# code (I had plans  to include 2 ways to do it ... AND some other things as well: that's the reason that I've used the C# [see for instance the very poor sweep1 options available in the GH component] but let's wait for the trad V3 update).

It's possible to do it with plain components (I'll include a Plan C option to the V3).

In the mean time get the V2.

more soon

Attachments:

• squarepipe_V2.gh, 41 KB

Permalink Reply by peter fotiadis on April 27, 2015 at 12:29pm

THIS is unacceptable: now it works with the sharp corner style (I swear that it didn't).

Moral: Karma

Attachments:

• squarepipe_V2A.gh, 45 KB

Permalink Reply by AdnFrs on April 27, 2015 at 12:41pm

Strange. It worked for me too. Some sort of Glitch I guess.
I have one question. When I increase the number of segments in PFrames to 200 (desired amount) the extrusions become distorted. Is this because of the direction of the extrusion? I can't seem to think why this would happen? Thoughts?

Thank you so much for all of your help. This is great!

Attachments:

• PFrames 87.JPG, 78 KB
• PFrames 200.JPG, 100 KB

Permalink Reply by peter fotiadis on April 27, 2015 at 1:11pm

Strange * Strange I must say (This IS NOT great):

round corners (in offset):

sharp corners (bool op fails):

It's about time for the Plan Z: to handle this entirely with C# - and use a 2nd Method for the sections (sorry about that ... but all what matters is the proper result, he he).

What do you mean "distorted" ? In what sense? (the direction of the extrusion [PS: I'll add a flip option] is irrelevant).

1M Q: why all the cases that appear easy-busy are the most challenging? (Murphy's law N666).

Permalink Reply by AdnFrs on April 27, 2015 at 1:25pm

By distorted I mean that when I increase the number of segments in PFrames to 200 the extrusion of the sections on the far right and far left of the object do not have any thickness. The below jpegs illustrate this, on the bottom image where the segments is 200 the many of the sections do not have a thickness. The extrusion thickness seems to decrease as the sections turn more parallel to the 'ground'.

Does that make sense?

Permalink Reply by peter fotiadis on April 28, 2015 at 1:31am

Distorted == you have some perspective in your plan view (planes are perpendicular to Plane.WorldXY since the circle is on that plane).

Dealing with this entirely with code (Plan Z): It has to do with what actually happens when offsetting the sections inwards (works OK) or outwards (it doesn't [occasionally] works ... except if the corner style is round). I have C# code that deals with that situation (it's a bit complex to explain what happens in outwards offsets, mind - but don't bother with freaky C# things > just > stick to the end result, he he).

Until V3 is 100% working in all situations > study these  - blue the section, yellow the offset :

Inwards (corner style sharp):

outwards (corner style sharp) :

Permalink Reply by peter fotiadis on April 28, 2015 at 9:07am

Find a couple of minutes to finish this (was some fit curve thingy when offsetting outwards, forget about it):

You have 4 options (Note: as delivered > all "links" to GH components are deactivated: meaning that the C# does all the work as a standalone "app"):

Plan B: Sweep1 > it doesn't work > "Karma" without doubt, he he.

Plan Z: use C# as a stand alone ring maker (Lot's of information is provided in case of failure(s) [for instance: outwards offset + sharp corners + fitOffsetCurve false]). If you opt for that, delete all the rest. To allow the C# to do the rings you just toggle createBreps true:

Plan A: toggle createBreps false and use gates to redirect flow to Plan A components.

Plan C: by pass C#, make your sections via components and connect the resulting List to the (top) Gate that controls the sections List. Not recommended for a variety of reasons ... but if you insist ...

best, Peter

Attachments:

• squarepipe_V3.gh, 349 KB

Permalink Reply by AdnFrs on April 28, 2015 at 11:24am

Wow! Thank you so much for all of your help/time. I am trying to patch plan C into A as per your instructions. I think this will help me understand the processes better. However the Data receiver (in C) that you direct me to connect to A is not collecting any data. What should I attach it too? 

thanks!

Permalink Reply by peter fotiadis on April 28, 2015 at 12:13pm

Elementary (my dear Watson): "some" things were deactivated in order to pass control to that freaky C# thingy (and a flatten tree was ... er ... "forgotten", he he )

Get Plan C as standalone (but I don't recommend that: use it only for getting the gist of the whole approach).

Now ... are you ready for some really WOW things with these rings?

BTW: Plan B is actually the right/smart thing to do (with C#) but that @#$@ Sweep1 doesn't work ... I really can't understand the reason.

best

Attachments:

• squarepipe_V3A.gh, 349 KB

Permalink Reply by Pieter Segeren on April 29, 2015 at 8:17am

Hi AdnFrs - Peter,
I think there are allready several discussions that are about the "square pipe" (try the search box on the right).
For this case I assumed that it would be okay if I made the planar intersections with your Brep to be just a four-segment-closed-polyline. If that was indeed an allowed assumption, you can try this: 

Attachments:

• SomeMoreSquarePipes_PSG.gh, 37 KB

• ‹ Previous
• 1
• 2
• Next ›
• Page