Modeling minimal surfaces architecture project

Does anybody know the script of this prototypes? I just want the gh files.
Thanks for you help


Are you going to give it a go yourself and ask for assistance as you go, or are you expecting for someone on here to model them for you for free and send you a gh file? Someone can correct me if I’m wrong, but the latter is not what these forums are for.

(Edit: I’m not trying to be funny here, we are all here to help you, but just asking for gh files from a picture for free seems a bit off to me).


Those aren’t minimal surfaces. You’re probably better off learning some SubD modelling.


FWIW from the architectural press: “The proposal for the RTV Headquarters in Zurich by Oliver Dibrova (Asymptote) is an exercise on minimal surfaces and their porous qualities. Inspired by liquid crystals, these surfaces behave as spatial elements for programmatic interweaving, creating highly differentiated spaces. Various conditions of “soap” films are created as the surface changes from states of solid to liquid at different temperatures, dividing the program into administrative areas, TV studios, radio studios, an open public zone, and internal working spaces.”

Can’t quickly track the architect or the firm he was working for at the time (Studio Mamem Domingo ETH-Zurich), but they (or their client) would be the best people to ask for a free handout of their intellectual property.


The original designer might say that, but still, the shapes shown in those images are absolutely, definitely not minimal surfaces.
That term has a specific and clearly defined mathematical meaning - zero mean curvature, and locally area minimizing.

I’m guessing they meant they were loosely inspired by some images they saw of topologically complex triply periodic minimal surfaces, but the part of that they’re copying in those forms is the topology, not the minimality. Hence why I suggested looking at SubD.


Sorry Daniel, I wasn’t intending to impugn the veracity of your statement, merely to show how the OP may have associated the illustration with minimal surfaces.


1 Like

Daniel is correct, I would say learning some SubD is the way to go. You can easily create surfaces like in your images by drawing individual mesh faces, joining, convert to SubD, and offsetting them as a starting point.


Oh man,unlike your statements,your quotation shows you’re so funny…
My main purpose is not copy of these images by gh.file, furthermore I’m not an expert of grasshopper too;Hence I just want to analysis the design process for my master thesis.
And If I could get the file as you said in terms of modeling,undoubtedly I’d consider his/her copy right as well.
So don’t judge too soon before knowing my reasons.
Good luck sir…

Thanks for your advice,I will have seen those tutorials soon.

Thank you sir.I would appreciate your time devoted in terms of mentioning the designer and explain the details.

“I just want the gh files”.

“My main purpose is not copy of these images by gh.file.”

What is the purpose again?

1 Like

Anyway, here is the process in grasshopper for something similar that uses simple meshes as inputs, I guess this could be adapted to make something related to the image although you would have to work on variable thickness and aspects such as contours if desired. The subdivide and offset operations have to be scripted one liners:

I hope this puts you on the right path at least to learning working with SubD as Daniel suggests. I don’t judge you, but your reasons are still not clear. As this is literally only 5 components I’m fine sharing it though if it helps! (6.1 KB)


I assume there is no specific reason here to explain my purposes for you or prove my honesty, whereas I just was looking for some help.
Anyway thank you for your help and sharing your data,I’m going to look at SubD tutorials as you all said in order to get to my desire level of modeling.

No worries, sorry if I came across rude earlier. John.

Hi Jeremy,
I’m the architect who designed the proposal for Zürich West around 2006.
To answer your questions. The initial minimal surfaces (batwing surface) were generated with surface evolver. Surface Evolver

Grasshopper wasn’t evolved enough back then, not sure if it already existed actually.

Someone on this thread mentioned subdies which is absolutely a way to model it if you don’t need the surfaces to be “perfect minimal surfaces”.

If the design evolves beyond the concept phase I’d advise to use proper parametric modeling.


1 Like