Complex Geometry in Dynamo


#1

Revit is often chosen as the primary documenting tool in many architectural offices and not a design tool as it’s widely acknowledged that Revit is extremely limited in dealing with complex conceptual modelling geometry. Autodesk’s attempts at introducing a dedicated conceptual modelling environment in the form of FormIt has also failed drastically as there hasn’t been a single new modelling feature or enhancements over the past two years. Apart from the splitting surfaces, Formit has even less modelling capabilities as the already crippled Revit and doesn’t have any freeform organic capabilities. Several free freeform apps are available in the Apple app store that can do more and handle very complex geometry than a dedicated conceptual tool like FormIt. Dynamo which was supposed to be the big game changer for architects using Revit, seems to head into an unexpected direction as well. While it’s evident that in almost 20 years of the existence of Revit, Autodesk hasn’t catered not once to contemporary design architects and this apparent dislike towards architects seems to impair Dynamo’s modelling development efforts as well in favor of the usual nitty gritty day to day practical enhancements, which resembles Revit’s development for the past 20 years.

One can only speculate that Autodesk hates architects or innovation in architecture in general. This was also evident at Autodesk University where not a single class was dedicated towards conceptual design in Dynamo let alone Revit. Apart of the countless classed on documentation, construction and practical uses with Dynamo, the few classes that did talk about geometry, where all focused-on Rhino/Grasshopper and getting Grasshopper geometry into Revit. As a matter of fact, there were classes where the speakers explicitly discouraged the use of Dynamo for geometry creation. How weird would it if you would go into McDonalds and they encourage you to buy a Whopper instead of a Big Mac.

There is nothing wrong for using Rhino or any other modelling tools for conceptual design that suits your needs. However, it’s almost inconceivable that Autodesk which does make excellent modeling tools for the industrial, civil, automotive, gaming, mechanical industry in packages like Alias, Fusion, Inventor, Maya, doesn’t have anything for the AEC industry. Unlike Archicad that has a partnership with Mcneal, Autodesk doesn’t have relations with Mcneel yet it relies entirely on Rhino/Grasshopper for geometry creation.


#2

For what it’s worth and perhaps I’m bias here, Formit, Dynamo and Revit are all better for conceptual design than you’re giving them credit for. McNeal’s software does have some advantages, and none of the Autodesk products perfectly replicate them all. However there are plenty of advantages to the Autodesk eco system as well, arguably more valuable ones as Autodesk offerings are light years ahead in design development, construction documentation and construction administration aspects where the vast majority of time is spent. Add in the additional faster revisions, better interoperability, and collaboration functionality (all of which are significantly lacking on the Mcneal’s software), and the advantages get even greater.

I do hear the frustration though. The great thing about the content in class I think you were referring to (though I’m admittedly not 100% certain which one you are referencing) is that it taught how to efficiently integrate Rhino content into the process, allowing the Rhino guys to go nuts without setting the bulk of the rest of the team back three months with a massing change that can’t be readily transferred.

Out of curiosity, did you happen to also attend AS128363: From First Line to Final Design: VR as the Ultimate Creative Medium, or BLD124675: Project Fractal, FormIt, and Dynamo Studio: The Perfect Trio During Early Design Phase? While I wasn’t at AU I have watched both and they presented some pretty good concepts and some cool developments on the conceptual end which I flagged in my notebook for further study (assuming I ever get the time).


#3

@robertobatista Would it be possible to share examples of complex geometry that Grasshopper can quickly generate/solve with a simple workflow (Not complex forms that might have taken days, weeks or months to create).
Trying to create the same could be an interesting exercise in demonstrating Dynamo’s abilities and/or limitations.


#4

While I think its unexplainable why all other Autodesk 3d software and other non Autodesk BIM platforms like Archicad, Microstation, Vectorworks, Digital Project, Allplan etc. all have the standard tools needed for complex geometry creation and Revit as the only software package doesn’t, I can somehow understand that it might be too difficult to make Revit comply with today’s standards and therefore Dynamo and Formit could have been Autodesk’s tools to bridge the gap. Most people can live with that, if these packages indeed provided those missing tools. However it’s strange to see why Formit after so many years hasn’t got any expansion of their modelling toolset beyond the basic extrude and revolve which is already present in Revit. Formit doesn’t have to be in sync with all of the layering under the hood which Revit does and Dynamo to a certain degree, yet it a double curved loft is impossible in Formit. Dynamo on the other hand made great strides around release 0.7, but seems to be hindered by the push for more practical day by day feature and since then we haven’t seen any expansion of geometry creation tools as well.

As for Autodesk University, I absolutely don’t mind a few classes on collaboration between Rhino and Revit. The reality is that most if not all firms do their conceptual designs in Rhino/Grasshopper and classes on interoperability could be essential. However, it would have been nice to also see best practices or classes on ways to design with Dynamo exclusively. At the same token, if the constant message is to only use Dynamo in DD or CD, then the classes are in line with the apparent philosophy behind the tool. Even their own Autodesk Pavilion was done in Grasshopper which sends a clear message https://www.youtube.com/watch?v=rfpj6yCX7VI. The same firm that was so advert against the use using Dynamo, also questioned the development of Dynashape as Grasshopper already has Kangaroo for centuries and Dynamo should stick to automating boring stuff. “leave the fun stuff for Grasshopper”. According to them (Woods Bagot) you should not put your eggs into one basket, yet ironically, they put all their design creation into one basket as well which is Rhino.

I haven’t seen the classes you’ve mentioned above. Though second one was given by engineers if i’m not mistaken. Nothing wrong with that btw, but in general engineers have different needs compared to Architects which is why I can assume that they are capable to stay within one ecosystem


#5

That would be indeed interesting. If I would take the samples shown at AU then I personally don’t think Dynamo would have any problems recreating them. But then again, the difficulty lies in the process of going through all the iteration quickly before getting to the final massing. Apparently then Dynamo falls short as it’s to slow and less agile than Grasshopper

Class: The Long Game
Maintaining Complex Interoperable Workflows Through Multiple Phases and Teams


image

Class: Panelling with the intent to fabricate

Class:
AS126493-L Creating Transparency in France with Dynamo

The geometry that I think can be challenging in Dynamo are the following
Blending two surface that are tangently connected. image
Examples where these “node” would he extremely useful.image


Patch Boundary would be challenging as well
image
image

Multiple various radius fillets. This would be huge if Dynamo could do this.

Rebuild surface or adding segments or points to surface.
Soft selection
image


#6

@robertobatista Have you explored T-Splines?
Here is a quick example, in case you haven’t.


#7

Sketchup is vastly more prolific than Rhino. By a factor of 30+.

Rhino has no stats on number of users I have seen, but has over a million subscribers to their newsletter. They estimate that the bulk of the subscribers there don’t own the software, and many use the software rarely (less than once a year). I know roughly 50 people or so in my last office subscribed, but only a dozen or so used the software after the initial trial.

The ‘best guess’ at the number of architects in the world is 1/2000 people. Current world population is 7,600,000,000, so 3,800,000 architects. Even if we cut that number in a third it is still more than the Rhino newsletter.

Meanwhile sketchup has 35 million users a year acroding to info presented at their basecamp summit. Even if you say that’s 3 computers per user (home, work, laptop) you still have 12 million, or 12 times the rough Rhino newsletter distribution. Note that the format tool set are captures most of the sketchup tools.

Side note: there are currently less architects than needed as only 20% of buildings are designed by an architect - meaning everyone needs to work 5 times as much (200 hour work weeks for all!) or five times faster (Dynamo). Do more, with less. Or as my boss used to say: Draft faster.

There is a type of design firm that Rhino is tailored for. Their buildings are uniquely that style and it is easily modeled with Rhino. Good that there is a solution and it can be tied to the post conceptual design tools. If you’re designing entirely with grasshopper I’d say Dynamo is a suitable alternative though.


#8

That is true. The ease of use and pricing is why it’s so big.

Rhino is mostly used by firms that do contemporary architecture. Not necessarily the Zaha Hadid type of work but work that needs some sort of research and form finding in the conceptual design phase. Rhino is also more used in European and Asian countries where contemporary architecture is more common that the USA. So in Autodesk’s defense, I do understand why their priorities are not into enhancing and improving Revit’s modeling tools since the average US Revit users, produces off the shelf strip malls and brick colonials. At the same token, Nike produces high end tennis shoes for a handful of pro’s that can be bought off the rack in any sport store for the average Sunday recreational athlete. Autodesk has this approach as well with all their other 3d modelling software. Movies like Avatar or Shrek can be made in 3ds max, but the vast majority of 3ds max users use it for simple renderings. So in that sense there is a strange division in philosophy within Autodesk.

I hope there will be more case studies on this instead of the push to not use dynamo for any conceptual work due to the limitations. .


#9

I believe statistically most of those are still done via AutoCAD, if they aren’t being built off the same plans from the one they did in 1996. I’d laugh if it weren’t true.


#10

A few more quick examples…
T-Spline Build Pipes …

T-Spline Surface…

Soft Selection…


Dynamo generated Voronoi 3D Printed Arm Cast
#11

I don’t think Autodesk hates architects.
Why would one even consider this, it’s an awfull thought, stop hurting yourself, it’s stupid. No one ever got himself clean from rolling in the mud.
Rhino and Grasshopper have some more years of development under the belt.


#12

Great samples. Wish there were tutorials on T-splines.

Is it possible to use a different profile like a rectangular one instead of a round one with T-splines?


#13


#14

@robertobatista Changed the title of this post, hope you don’t mind :slight_smile:


#15

I don’t think the issue is only with the complexity. I can have a long list of what makes Grasshopper a better tool than DynamoBIM for iterative design and a lot of it has to do with the GUI design and the ease of customization and development but the one which is related to this discussion is the difference between the geometry libraries. Rhino’s geometry library is light-weight and fast. Dynamo uses ASM geometry library which is heavy and slow and uses much more resources than Grasshopper. That has it’s own reasons but at the end of the day from a user/developer point of view you need a geometry library that gets the job done.

Generate 10000s points and run a couple of 1000 intersections in Dynamo. It will take for a couple of minutes and you will run out of RAM! Do the same operation in Grasshopper and you won’t even know that you’re dealing with 10000 points. I had to update my laptop to be able to develop for Dynamo after developing around 4 years for Grasshopper with that laptop!

T-Splines is great but won’t solve the source of the performance issues with ASM and Dynamo itself.

Related:


#16

While it’s no secret the fact that Revit’s modelling engine is amongst the worst in the industry, it is understandable that Autodesk who bought Revit years ago and inherent all the goods and bad, just isn’t capable to fix the problem and to make the product better.

Fortunately, with Dynamo, Autodesk doesn’t have to invent the wheel and can start from scratch without all the trails and errors. After all, you have Grasshopper, Generative Design , Marionette. that set the bar in terms of GUI and geometry library long time ago. How is it still possible with all of this knowledge and the huge amount of resources Autodesk has compared TOthe other companies, that they managed to create a tool that is the worst of the bunch?