Grasshopper

algorithmic modeling for Rhino

Hi everyone,

I'm having trouble with the network surface component. Ideally I'm hoping to use this as a final step for some more complex definitions, but for right now, I can't even get it to work with two sets of curves, linked directly in from rhino (and flattened, so no data structure issues).

I'm attaching a screenshot of what I'm working with with. I keep getting a "null" error when I plug in a panel, but I have no idea what's causing it. Sometimes I can get some of the surface to form (as shown in the image below), but I don't understand what's making it fail when it does, and have not yet been able to get it to work for the entire curve network.

Any suggestions?

Thanks,

Jane

Views: 4908

Attachments:

Replies to This Discussion

No code, no help.  Can't see your data in a .pdf image.  Works for me.  Like many here, I normally don't reply with code if none is supplied with the question - but I needed the exercise.  Result in this case is the same as connecting the curves to 'Loft'...

Attachments:

Well, let's see. First, if you took a look at either the images, or the content in my post, you would have seen that at this point in the troubleshooting process it's just a network surface component with two different curve components plugged in. So code is not the issue. Second, your "exercise" doesn't actually help, or address the question. Thanks so much for all of your help though, and do take care. 

Hello
Joseph is right, he means code and data. Data is surely part of the problem. I imagine that the component use rhinocommon networks surface which has more options. So using a c# component with networksuface with autosort or no autosort could help. If not you could use patch, surface deform ... What are your constraints for this surface? Continuity on the border, precision ....?

Well, let's see...  I'm guessing that "Jane Smith" is not your real name and that you're male, not female.  Poor attitude for someone seeking help.  Of course I looked at your useless images.  As Laurent said, they tell us nothing about the curves, their direction, sequence and other important details.  So sorry I bothered to reply at all.

Read this first! How to get help on this forum.

Joseph... you are the one with the poor attitude here. Your help was useless and your tone was rude from the beginning. Please don't pollute this high quality online experience with your sub-par commentary. 

Well Jane

You don't need a gazillion profiles in order to do "unusual" random nurbs (if random is your goal) - unless these profiles mean something, that is.

For instance this creates a few profiles (random z on points per profile etc etc) and then "distorts" the curves by dividing them again and applying an added random z "noise".

BTW:  you can use Surf from points as well for such things. 

Notify if you need a demo (BUT ... using solely code) 

Ok after having a little play i think you problem maybe the order in which the input curves you are using are selected, as i too can break the Network Srf component if i use window selection some times or if i pick them in a random order. 

If you have lots of crvs to select and you don't want to manually select them in order, just draw line down one edge of the selected curves then get the end points of the selected curves and use Sort Along Crv Component

and a item component to re order them.

Thanks for all of the suggestions - and point taken about including files with posts. To that end, I'm uploading both .3dm and gh files.

Peter - the geometry is actually specific, although I know it looks very chaotic! I'm looking to be able to control a few different aspects - degree of "randomness" in different zones, a few different levels of base plane curvature, overall size, spacing, degree of potential randomness and domain constraints in x vs. y vs. z axis, etc. I'll have a look at what you've posted.

Tom - The end goal is fabrication of molds for cast panels. They need to perform in a specific way, which is what is driving the geometry - right now it's at the prototyping / testing stage, which is why I'm looking to control as many parameters as possible. I know I'm asking it to deal with a lot of geometry, but I'd like to see how far I can push it - so if you have a suggestion as to a better method, please let me know. 

Laurent - I apologize, but I'm not sure what you mean by constraints? I did try all of the possible inputs for "C", but nothing seemed to make a difference. You mentioned a C# component - I haven't worked with C# before, although I would be willing if it would be helpful in this scenario. Is the component that you mention something that already exists, or something I would need to write myself?

Matt- Thanks for your suggestion, I did not think to reorder the curves - but that makes sense. I'll give it a shot.

And Joseph... why in the world does my gender have anything to do with anything? Unless you assume that because I pushed back a tiny little bit against the aggressive and snarky tone of your initial comment, I couldn't possibly be female? Doesn't it then follow that you only took that tone in the first place thinking that, given my gender, I would be a passive target? And you wonder why I don't use my real name on the discussion board. Good grief...

Attachments:

You are lucky I bothered to reply at all in the first place.  The customary practice here is to just ignore threads where people haven't bothered to post code or read the Read this first! guidelines.  It becomes tiresome trying to be polite about that.  I considered your reply to be snarky so responded in kind.

I'm looking to be able to control a few different aspects - degree of "randomness" in different zones, a few different levels of base plane curvature, overall size, spacing, degree of potential randomness and domain constraints in x vs. y vs. z axis, etc. I'll have a look at what you've posted

Well ... you'll need some def for that, he he. That said I did an other thingy that is a "super set" of the C# captured above  (where you can control  or distort the profiles [Loft] or better the Points [Surf from Points] with a vast variety of ways (including chaotic ones) some time ago but as usual is done only via code mostly because the things that does are only doable with code. Anyway I'll post it here just for fun ... if I can find it (forgot the def name, the directory, the workstation and the reason for doing it, he he).

Here is a solution with mesh, I manage to make patch work but without enough precision. More work needed. But as you begin with a mesh I propose a solution with quad mesh and smooth with weavebird Catmull&Clark. 

You have to specify

the minimum height of a plane (-1 here)

The number of subdivision in X and Y (100 x 100)

the subdivision for CatmulClark (smooth the surface if 1, 2 or 3 an also augment the number of facets)

Antother with 5 760 000 facets, but too much discredization so it didn't smooth the Delaunay mesh. 

Attachments:

RSS

About

Translate

Search

Photos

  • Add Photos
  • View All

Videos

  • Add Videos
  • View All

© 2024   Created by Scott Davidson.   Powered by

Badges  |  Report an Issue  |  Terms of Service