Adaptive Components and Parametric Ceiling

[Anonymous]

I'm trying to use adaptive Components on rods that I will orient vertically onto a parametric conceptual mass so that it looks like a parametric ceiling made up rods extruding downwards. The problem I'm having is I'm not sure how to constrain the adaptive component to the vertical plane so that it only moves in the y direction. Have a look at this link to see what I'm trying to model (http://www.ondiseno.com/noticia_en.php?id=5797). I would greatly appreciate some help!

[barthbradley]

An approach that might work, would be to put the rod into a pattern-based family and then use that family in a divided surface. 

Actually, that won't work because the rods are different lengths (e.g. not a repetitive pattern). Not sure I understand the problem though. You want to post what you've got so far? 

[ToanDN]

@Anonymous
Tick Always vertical in the Adaptive family.

[barthbradley · Accepted solution]

Actually, if the component is hosted to an adaptive point workplane, the adaptive point itself has "Orient To" settings. 

This will explain: https://knowledge.autodesk.com/support/revit-products/learn-explore/caas/CloudHelp/cloudhelp/2015/EN...

[FAIR59]

You can adapt this tip on  https://forums.autodesk.com/t5/revit-architecture-forum/share-your-favorite-tips-tricks-and-workflow... ( post 64 )  to your needs

[Sahay_R]

@FAIR59 are you talking about the Revit Attractor Points post?

(PS - are the posts numbered? I have never been able to figure out which one is Post # XYZ.....)

[FAIR59]

yes, that is the post I'm referring to.

In my browser i can the post numbering under the user-profile of the sender

[Alfredo_Medina]

@Anonymous

I wonder if you figured out how to make that ceiling. That is an interesting challenge. Did you get it done?

[Viveka_CD]

Yes, @FAIR59 is talking about post 64 by @Anonymous

[Viveka_CD]

Hi @Anonymous

Can you post a screenshot of your progress? I see the experts on this thread looking forward to hearing from you.

[barthbradley]

@Viveka_CD: I think the OP gave up and went with a "popcorn" ceiling instead. Ha! 

You know, I'm thinking if some of the concepts in this video could be applied toward creating the OP's ceiling.  What do you think @Alfredo_Medina? 

https://youtu.be/4EcRbVIU-pc

[Alfredo_Medina]

Yes, it requires some splines, surfaces, and a repeater. And, I think it has some randomness.... Sounds like a good recipe for making my next Youtube video about this. 

[ToanDN · Accepted solution]

Getting the bottom of the light tubes to follow the the "topo" ceiling is easy.  Getting the top of the them to align with a flat plane is difficult and is not covered here.  I just used a section box to crop the 3D view for now.  Without manually aligning the top to a plane, I can think of export the model to SAT, re-import it in as an in place generic model, and use a void to chop the top flat.  Perhaps some folks can come up with a more elegant solution (@Alfredo_Medina specifically). 

Anyways, below are the steps to get the ceiling modeled.  They are fairly straight forward.  All the reference points and the grid sizes in the family can be parametric so the form of the topo can be adjusted by parameters ( I wasn't doing it here though).  I chose to use an intersecting grids instead of an UV grids for an uniform spacing between tubes on a plan.

  

[Anonymous · Accepted solution]

I used a method similar to this I think. I first created a curved mesh surface with conceptual massing then converted the mesh to nodes. I then went to components and went generic adaptive component and created a family. This adaptive component is constrained to a point that can be attached to the individual nodes of the mesh created in conceptual mass (make sure that you constrain the dimensions appropriately so the adaptive component doesn't warp out of shape when dimensions are adjusted). So then all you do is click on the little sphere I. Adaptive Components that has the xyz planes attached to it and orient to orthogonal in the properties menu. Then when you import the adaptive component into your conceptual mass, you can place the adaptive component by connecting it to the node points. The orient to orthogonal is important because without this option, the component would just attach randomly to the mesh. Where can I upload pictures what I've done?

[Anonymous]

Sorry one thing I forgot to mention was to convert the point (the little sphere) to adaptive Components so it has the little number above it.

[Viveka_CD]

Hi @Anonymous

You can upload your progress pictures right here. (scroll down when you reply to this post and see attachments - choose file to attach the images)

Looking forward.

Regards,

[Alfredo_Medina]

This is great, @ToanDN  , I don't think it is necessary to try to align the top of the cylinders to the surface. In that link in post #1 in this thread, they say that the cylinders are made of cardboard. Most likely they are just regular cylinders, probably recycled from paper rolls, just hanging there, jsut eas you made them. Excellent.

The only thing I would add (if I get to do my video) is some randomness, because it seems to me that the images in that link show that the length of the cylinders is not constant. And probably the diameter varies, too, I'm not sure.

[barthbradley]

---

@Anonymous wrote:
So then all you do is click on the little sphere I. Adaptive Components that has the xyz planes attached to it and orient to orthogonal in the properties menu. Then when you import the adaptive component into your conceptual mass, you can place the adaptive component by connecting it to the node points. The orient to orthogonal is important because without this option, the component would just attach randomly to the mesh. 

---

"... orient to orthogonal"?  You did that by setting the Adaptive Point's "Orient to" parameter to "Global xyz"; right? Or, did you find another way to keep the AP's tubes always upright in the host? 

[barthbradley]

@ToanDN: I think I understand what you did, differently from @Alfredo_Medina. It looks to me like you repeated a fixed height column over a divided "curvy" surface, and then cut away everything above the "ceiling" plane.  Right? 

@Alfredo_Medina: I would love to see how you are able to make these repeater columns random diameters as well. That would be wicked cool! 

Thanks for the Fun Thread, @Anonymous!

Cheers. 

[ToanDN]

The randomness can be done via a Dynamo script.  There is a node generating random number that can assigned to a parameter.  You can amplify the randomness by ^2 or ^n the value if need be.  Couple it with the diameter with a formula and the result are total random sized tubes.

I have done it for fun to generate the one below.  !st one is the original.  2nd one is with random lifting of the panels.  3rd one is with random thicknesses.  I didn't amplify them so the results are subtle but you get the point.