generative modeling for Rhino
I am trying to achieve a sort of surface minimisation to use within form finding experiments, similar to what is seen in this Kangaroo video:
After downloading kangaroo and having a look at the example files, and also looking on some of the other discussions on here I can't seem to find anything that will help me. I was just wondering if it is possible to achieve this effect without having the WeaverBird Components added? - As I am using a University computer that doesnt have this installed.
Yes, it is possible to do surface minimisation without Weaverbird - just make a starting mesh with plenty of faces in Rhino. There is some discussion of this in this thread:
Thanks for the discussion link. I have had a look through and although people seem to be using weaverbird on the definitions I have tried to adapt it so it will work without using this. I lofted some circles together to create tubes and then joined them all together and converted it into a mesh. I then inputted this into the definition you posted to the other thread - and tried to adjust it so it would work. However when I try to run it, the final mesh button fails to collect anything. I think it is a problem to do with what is plugged into the connections part of the springs button.
Would you mind having a look for me?
Many thanks in advance!
Thanks, I tried what you suggested to extract the curves from the mesh. This works when plugging it into the connections part, however after trying to run the simulation I get a runtime message from the Kangaroo engine:
'Solution exception: Index was out of range. Must be non-negative and less than the size of the collection. Parameter name: Index.'
And again the end mesh button fails to collect data. I cant seem to see what is going wrong - one odd thing I noticed though is when I preview the Anchor Points, they seem to be weirdly in the middle of the tube - this cant be right?
I will keep going with it but any suggestons are very much appreciated!
I suspect this is due to how you are building your mesh from separate parts, and the vertices not matching up or being duplicated.
remember that there is no need for your starting mesh to be curved - it is generally easiest to start with the simplest possible mesh with the right topology (number of holes and handles). So usually just simple boxes - try making a mesh box with 10 faces in each direction and copy it several times and boolean them together.
The main part of the mesh will round out when you start relaxing it. The rounded ends will take a little more care, but would be possible.
Thanks so much for all your help.
Wiktor - Thanks for the example file, yes that is the effect I am trying to achieve. I donwloaded your example and set the mesh in my definition as the first box mesh - but I am still getting the same runtime error - would you mind having a look? Just to clarify, the boolean toggle should be set to false when running the simulation yes?
Thanks! I have just managed to get weaverbird on the computer I am using at Uni. I have run the definition you posted and also the one Daniel posted here: http://www.grasshopper3d.com/forum/topics/how-to-create-nodesbone.
However for both I get the same message before the definition opens (3.jpg) - something about a chunk being missing?
Then when I try to to run the definition some odd shapes are coming out...not what i expected. I have included screenshots of both.
Thank you so much for your help and sorry for being a pain with this!