Tutorials that start and end with existing objects?

Those are all good suggestions.
From the looks of most all the learning material out there, grasshopper is Heavily focused on architecture. And that’s really cool, but I wanted to explore it a bit in terms of what it could do for product design, but in the Middle of the design workflow.

You can apply anything you learn to anything else. That is the beauty of this software and things like it. I’ve applied facade definitions to jewelry, or generic geometric explorations to products. Once you understand the software well then you will see the only difference really is the scale of the object. A surface is a surface and a mesh is a mesh, doesn’t matter the actual application. In terms of not seeing product specific definitions as much as arch, it is because product design fields are much more restrictive in what they can share. Workflows, manufacturing methods, material applications are usually patented and protected internal to companies. Architecture rarely has these restrictions. Furthermore, usually at the point where design moves into manufacturing it also moves to other (usually much more expensive) softwares specific to the manufactures make methods.



The problem with Grasshopper is the fact that it misses basic surface tools. Its extremly difficult to create a pattern in surface representation, when a shape is curved and constraint to certain rules, such as being demoldable, continous and smooth and if everything must be curved/crowned. (Which are quite common constraints in product design) .

Now, usually in architecture you don’t have these limitations, which makes almost anything coming from that direction useless. Because they simply ignore that.

Furthermore, a lot of tutorials and example resources are made by people still being students themselves. The problem with that is obvious.

Furthermore, like @Michael_Pryor said, another problem is that industrial design is usually restricted and confidential. I simply have no chance in showing or sharing any of my professional work here, since it would violate against my previous and current employment contract. Partially I don’t even own the data or tools used anymore.

But you can use Grasshopper very well in Product Design and there are exclusive use cases.

However it takes a while until you can use it in a really useful way. I believe the following list shows what you should definitly learn

At some point you’ll definitly need to…

  • know when and when not to use Grasshopper, the reality is always semi-automated
  • know different ways of mapping surfaces on a freefrom surface and as a bonus you’ll need to this efficently. Learn mapping on Trimmed/Faced surfaces and Breps/Multiple surfaces.
  • understand surface modelling in first place, this is your workflow.
  • understand anything about data management
  • be able to write code and you’ll need to know Rhinocommon (It its required!)
  • understand any of the math involved. As a plus knowing Bezier, Nurbs-Math can really help to create custom surface tools on demand

So the learning curve will definitly be steep! Essentially you can use any tutorials if you watch them differently. What type of problem is being solved, and how can I apply this for my projects?

Since the Grasshopper universe if filled with all these simplistic tutorials, it is very hard to filter out what is useful and what is not.
I think you can skip any tutorials using Meshes, being done by very young people and you should definitly not look for a specific pattern, but rather follow tutorials about how components work.
Get the terminology and search this forum for solutions. And last but not least, just play around as much as possible. Knowing and understanding the tool always helps

1 Like

I’ve been struggling with GH also. Most of the tutorials are more of an exercise in following instructions than learning GH. Programming doesn’t come naturally to me. The best source I have found that actually explains the fundamentals and how to approach a design probem is Gediminas Kirdeikis on YouTube. What I ultimately figured out was that I needed to try and create a really simple project on my own, do some research, and figure out what components I needed to use - sometimes through trial and error. Eventually I realized that the key to being successful in GH is understanding data management (list, trees, indexes, etc) and I’m still trying to get through that. I haven’t found any tutorials that teach the design process. Basically they are just instructional manuals.

I agree partly. What I don’t agree with is the losing flexibility part. In some not obvious ways you can still get the flexibility by working with creative workflows.

Workflows such as using stream gates with logic can still open up possibilites. And other ways, such as, I felt limited by the graph mappers a few weeks back, but I stepped back and said hey I can make my own substitute to create more complex mapping curves for scaling loft sections individually. I put a series of points (say 10 or however many) in Rhino within the coords of 0,0 and 1,1, put them in a points container in GH. So then I was able to drag those points manually instead of a graph mapper, and those moves would adjust the scaling of each section.

Don’t be fooled into thinking of limits of the software, or at least if you do think as such, be sure to second guess your conclusion and think of potential workflows that could break down the “limits” that you may be falsely assuming are there.

People used to assume that AutoCAD wasn’t good at complex 3D, but they just didn’t imagine workflows to give far more dynamic results.

And BTW I’m coming from 10+ years of Solidworks more recently, and so I do see that I can’t just point and click to make sketches then extrude them, but I can sketch a curve in Rhino, put it in a curve container in GH, then extrude/sweep/loft that, then go back and adjust the curve all I want, as if it were in a solidworks sketch. Just about the only thing you can’t do is use parametric dimensions, but you can achieve the same results, just the interface isn’t as automatic, so you have to be even more creative than you already have to be with Solidworks, but the end results are far more wide open and flexible.

Becaue you can’t. The design process is something you have to understand yourself, which you only do by putting the actual work into it like you did. Get a real-life example, try to solve it using grasshopper. For that you need an understanding of the mechanics of grasshopper, and for that, “instruction manuals” on how to do specific things are helpful. They let you piece together your own solution from different sources. But the annoying work of trying it out, failing, and trying again is on you to do.

1 Like

Yes, or direct programming. Often real programming is seen as Rocket Science. In reality its much easier to directly code pieces by text, than dealing with the data-management-hell of Grasshopper.

I mean just think of anything in life which is hard to learn. A Youtube tutorial or book always just gives an entry point. Learning how to hold the pen, doesn’t allow to write something good.

Grasshopper made programming accessable, but at the same time it is a tradeoff.
Learning how to morph a pattern on a shape is another thing. I have read that some people claiming they have problems with applying a pattern on difficult or poly-surface shapes. But there are hundrets of solutions to these problem. It also highly depends on the pattern or the shape. It is impossible to list all of them, since there is no general approach.

Furthermore people always expect one-click solution made in 1 hour. When I was doing professional work in Grasshopper, it took me weeks to find perfect solutions to some problems. And it took years to become better. That is just how it is.


Don’t be fooled into thinking of limits of the software, or at least if you do think as such, be sure to second guess your conclusion and think of potential workflows that could break down the “limits” that you may be falsely assuming are there.

Thanks for your response and describing your experience.
Let me answer to your statment above.
I dont think grasshopper is limited in any way and thats in my view the tricky thing because YOU GET ENDLESS OPTION = FLEXIBILITY on the other hand you have to decide WHAT CONNECTION YOU NEED TO MAKE = AUTOMATION.
You establish a system and the more nodes you connect the more automation and especially dependecies you get.
I start coding because its more flexible in establishing a system instead of node based scripting.
Node based scripting is not only for geometry its also for animation rigging shading compositing etc. its in my view only a logic evolution to overcome the limits of a drop down menue because options and tools pop out like muchrooms and they dont fit into a menue bar or drop down.

Anyway i love grasshopper and its the reason why i start coding not only its very visual and not to abstract but also of the community and company which is similar to blender very open highly skilled in knowledge (im not too smart so its only my point of viw and the more i know the more i realise how stupid i am) and they share there knowlwedge.

Well I definitely disagree with saying you can’t teach the design process. Good tutorials don’t edit out mistakes, but include how you fix them. That’s the design process. How many times have you tried to fix someone else’s geometry in Rhino and it’s just missing a simple boundary. And shoot…networkSrf is failing. Zoom in. Hmm, looks like it’s joined. So in a tutorial, they’ll just say to fill it with a network surface, and Maybe try a 1 rail sweep. But when neither of those work, you do things like explode the polysurface, join it again. Try again. No? Double check tolerances. Select bad? Hmm, recreate the neighboring fillets. And bam, you’ve fixed it.
If no one shows things like that, then you never learn how to fix them.

That’s a big problem I see with many tutorials. They often show you the end of what you are making, but don’t tell you Why they are doing something, what it’s for, or generally the direction they are headed. It’s almost like you’re just playing follow the leader until you get to the end and you’ve made the same thing they did.

Anyway, this discussion is great, and clearly I’m not the only person who is/was having troubles finding a place for grasshopper in their workflow. Although, based on some of what I’m hearing, I’m not sure it will ever make it into my workflow. When someone says “Hey, can you make this pattern on this surface” they expect it to not take more than a day or two.
But if you actually organize your nodes and files well, you can likely reuse previously built scripts to accomplish something like that in a day. Often times it’s looking at creating 6 different patterns in 2 different overall shapes.
I have ways of doing that, but in polys. Other processes are getting more parametric also, but won’t be production ready for a few months.

If you really wanted to, you could make something fully parametric. Someone here mentioned you couldn’t, and although that may be true in Practice, it’s not actually true. You could do it, but It’s just not really Worth it. There are more appropriate software for doing things like that.

Hopefully I’ll see some people showing examples of how they used grasshopper in actually released product design. Maybe that will expand my view on how it is used. Otherwise, it looks like I’ll just practice with a few patterns here and there when I have time.

Out of curiosity, what is behind this? Besides creating patterns on polysurfaces, what else should a product design plugin have? or in other words, what kind of problems/solutions would you expect? Independently (or not) of the parametric or procedural modeling, I mean the things you master and want a software to help you.

1 Like

Yes I sadly know this, but this is because they don’t understand the complexity of it. Sorry, whats the alternative to it? Manuell modelling? Using the crappy array tools like Alias, Catia or Icem offers? Thats the problem with that. Or outsourcing it to a cheaper country where 20 people draw on it? I have seen all of it, and it all was insane and poorly made.

I mean you may solve these things in one day or in two weeks. You never know. But then you also really need to be expert. Almost any professionals using Grasshopper I worked together, have learned that over years. How many product designer I have meet telling me they want to know how it works, then they never really tried, stopped it and claimed lack of resources or tools to be the problem. The skills involved are completely different to modelling, but it demands a special mindset. The job has more in common with being a software developer or engineer, than being artist.

I read the ModeLAB primer but I didn’t get a practical grasp of how to work with data trees until I had to use grasshopper to build real world objects.

I haven’t seen a lot of tutorials that use buildable objects to illustrate data trees.

1 Like

I would propose that you model a pattern completly within a manual process. Then you copy the exact steps in Grasshopper. Because thats what it in the end is. Its a visual scripting engine to automate CAD work.

1 Like

Good points on both posts on how to get better with GH, thanks!. What I was trying to communicate (and obviously did a poor job of) was trying show that the upfront planning is where much of the learning occurs. I doubt that anyone doing a YouTube tutorial in GH just finished their morning cup of coffee, sat down in front the PC, fired up GH, and began pulling in components, tying them together and had a finished product in about 40 minutes. There is always a plan beforehand on how to proceed from a strategic standpoint before doing any actual work. As an example, when I did my first vase in GH, my first thought was “How did I do it in Rhino?”. Well, I had done a revolve around a curve which was doable in Rhino, but lacked much of the parametric ability I wanted. Same goes for a polar array with a series of curves, but again a lack of parameticability (is that a word?) or maybe extruding a closed curve then baking it into a SubD and adding edges that can be manipulated. Finally, how about taking a closed curve (ellipse, polygon, gear, etc.) making several copies in the z plane and trying to manipulate them? This is where ended up and it eventually worked well. Once I started trying to do the manipulation, I realized I had to look at each step as a separate sub-operation and see what individual components were available and what was needed. Finding out why a component didn’t work was more instructional than finding one that did work. What little programming I have done was always through this approach of breaking down the design into small steps, solving them, then moving forward. All I am asking for is for the tutorial authors to consider discussing some design narrative in the beginning of the tutorial or where a complex operation takes place so the conceptual part of the design is discussed. Sorry for the rambling, but to me this important for new users to grasp GH design.

1 Like

You are just repeating the same thing, but on a different level. What if neither tolerance nor exploding a polysurface is the right solution to find the issue? Without an inherit understanding, you will not be able to apply this process to another situation. And you never get a deeper understand just by watching someone else do it.
Its like watching all competitions on a balance beam ever organized - without putting the hours on the actual apparatus, you will fail on your first trys.

People want to know, they dont want to learn.
In all the workshops/seminars/tutorials I ever held, it was only those who put in the extra hours afterwards who really gained something.


That sounds like a really niche complaint about a very broad document…

I’m not saying the ModeLAB primer is deficient or that it needs to be all things to all readers.

I’m saying that I think tutorials that go through several different steps of tree management on a path to modeling parts that have to work in the real world would be a great way to teach trees, and that most tree tutorials seem to be a little too short.