Grasshopper Parametric Curved Surfaces

[Lara4747 (Lara)]

Hi! I’m working on a first year uni architecture project, and I want to create these types of curved surfaces in grasshopper. I know it’s possible in rhino, but it needs to be made in Grasshopper for my assignment. It would also be great to have some kind of parametric aspect where the no. vertices can be adjusted, etc. Any help would be greatly appreciated!

[jaime.miranda]

Hi Lara!,

I recommend you take a look on this video series (Kangaroo - Mesh Relaxation - Tutorial 3 on Vimeo) by Daniel Piker on mesh relaxation using Kangaroo & Grasshopper.

Be warned, some gh components might have changed in name or merged with another (This was released more than 10 years ago!) but the concepts and mostly workflow should give you direction to arrive where you want to go.

Remember to attach relevant files as well as a start in what you want to achieve, i.e. your best attempt at tackling the problem on your own otherwise, not many may be willing to start it for you

Give it a go!, come back to the forum anytime for help-

Best!

Jaime

[PeterFotiadis (Pfotiad0)]

A few tips (K2 for Kangaroo):

1. The objects in your pics are called tensile membranes (so for examples [via K2] in this or the old - dead- Forum search threads using these keys).
2. K2 works best if the start Mesh (or Meshes) is/are as “regular/rational” as possible.
3. A Mesh has naked edges (adjacent to one Face) and clothed ones (adjacent to two Faces). Use different spring forces for these 2 “groups” - as we do in real life membrane design. See last attached pic (the weird hinged plate and the perim cables attached to that).
4. Exploit various options related with Anchors (i.e. vertices “locked” in place [or almost locked: case plastic Anchors]). Go to some membrane makers sites (like BirdAir) and study concepts and examples.
5. Avoid using cables (i.e. Lines) from “Anchors” (in Mesh) to real Anchors (in some tube top pt or other). This obviously approximates reality BUT requires code - so forget that part for the moment.
6. If you want results related with real-life (well … in some abstract way) study Mesh Connectivity. This is used in order to work (post relaxation) with Anchor plates and the likes (depending on the plate angle these MAY be hinged combos).
7. In quite a few real-life cases we define the start mesh as a flat one (then K2 does the rest). For instance Membranes of that type started from a flat Mesh:

8. You need to understand the “basic components” of similar solutions in order to prepare data for relaxation.
   
   
   
   
   
   
   
   

[Birk_Binnard]

Here’s a simple approach using circular arcs:

ArcSrf.gh (8.6 KB)

[Lara4747 (Lara)]

Thank you so much Jaime! I’ll be sure to check that out

[Lara4747 (Lara)]

Hi Birk, that’s awesome thank you so much - I’ll give that a go

[Lara4747 (Lara)]

Hi Peter, that’s amazing thank you so much - I’ll definitely start experimenting with those and this is super helpful

[PeterFotiadis (Pfotiad0)]

Always remember: with regard the relaxation part 99.99% of the whole tensile membrane challenge is the definition of the source Mesh/Meshes. For instance IF doing it flat (i.e. say on ground) is not suitable … try to solve it with the simplest possible way. Even topologies that appear complex are in fact very simple. For example the source Meshes shown below are just subdivided Meshes from a (almost flat) Polyline: using variable cables from mesh vertices to real Anchors (this means that the Anchors ARE NOT the Mesh vertices) allows a great simplification of the whole source definition puzzle - BUT that requires code as anything in AEC real-life designs.

BTW: you can attempt to do other type of things as well (i.e. other than classic “rational” tensile AEC stuff). See 2 freaky ones: for some future EXPO using a WIP elastic BirdAir fabric that may or may not enter production. Reds are cables in their start pos while greens are the same fixed to their final position.

[jaime.miranda]

That’s wild! , I’ve always wanted to see what you and Don Julio are up to on your professional day to day.

Best regards Peter, awesome stuff as always!,
Jaime

[PeterFotiadis (Pfotiad0)]

@jaime.miranda

professional day

For these - ugly - days no R/GH I’'m afraid (AECOSim [BIM], Catia, NX on duty):

BTW: membranes are one thing but the components are the real McCoy. For these you need feauture driven modeling and solid MCAD app(s).

[jaime.miranda]

Man, I don’t want to hijack this thread, but I need to ask so here I go…
(warning, bit of a long read!, do we have rules against long posting? )

I’ve seen many attempts at companies in my line of work to reinvent the wheel around NOT buying either Catia or Siemens NX or any other robust (and expensive) feature driven software, with most of them either completely missing the point or achieving varying degrees of success , with the most consistent compromise arriving in rhino’s ballpark due to the flexibility to build what you are missing around it, but that’s another topic…

Maybe it’s been a matter of lack of experience or a software limitation on our side, but I’ve seen in most of the models you’ve shared, the best of both worlds (feature driven solid modelling and what seems to be NURBS(?) is that NX’s proprietary kind of B-rep stuff?), so I guess my question is regarding your workflow… so I take it that siemens NX has a similar thing to Catia’s power copies(?) and a way to instance them all at the same time, is this C# based?

I don’t want to stray from Rhino, but it would be sweet to populate that clean, rationalized, single patched curved assembly based surface, in something that’s able to keep the relationships of the parts without having to build them myself a tiny bit different every time…

Anyways, back to the asylum now.
.
.
.
also, this little number from that awesome rhino meme chat gallery:

[PeterFotiadis (Pfotiad0)]

@jaime.miranda

I don’t want to stray from Rhino

Well … the whole AEC/BIM/PLM/CAD/MCAD/Whatever thing has as follows:

1. Assuming that you know what you are doing (and where/when).
2. Assuming that your Karma is OK.
3. Assuming that you have some sort of fatalistic aspect on things.

Then your bike is a proper Panigale (V2 or V4). Meaning lot’s of dollars, unreliable to the max, frustrating at times, tricky to service … and … lethal.

Exactly like learning/using/paying Catia/NX (not to mention AECOSim).

[jaime.miranda]

Man, big red pill fan. I guess I’ll trouble you more often then!