Question about Clothworks

I was trying to drape a fabric (e.g. plastic tarps) over some arches. But I can’t figure out how to get the edges to “grab” the arch. If I make the fabric big enough, the ends hand over and it doesn’t all fall down. As you can see from this one, the hang over wasn’t enough (and hanging over isn’t what I want). I don’t see how pins would work in this case.

Maybe you can create a ‘bended plane’ that aligns with arches and try pinning lots of points at the borders of fabric?
Author of the plugin @Anton_S will probably have a better idea/comment on this.

Edit: Made a test, not bad i think. If you create more fabric (wavy plane maybe?) between arches, it’ll be better i believe.
Here is the .skp file: drape_arch.skp (990.5 KB)


You can also assign a pin type to the bordering arcs. They will lock down vertices overlapping their bounding box.


I think what Anton is trying to say is that you can actually let an object be a ‘pin’. You don’t have to assign hundreds of pins separately. The ‘pinned’ object can then be moved just like a regular pin, which I find very helpful! See the video below.


Ahh can’t wait to finally try this at the weekend good tip.

1 Like

Thanks to each of you for your comments. I haven’t been able to make an object pin work yet. I have’t figured out how to connect the object to the cloth. Anton indicates the pin object and cloth must have overlapping bounding boxes. I haven’t understood exactly what “over lapping” means yet for Clothworks. Do the points of contact between the pin object and the cloth have to have identical [x,y,z] vertex locations? Have the same gridding, etc? I couldn’t duplicate n•1’s results when I tried to reverse engineer his model from the video. The cloth just falls without being influenced by the arcs that are designated as pins.

I manually built the cloth from the vertexes of the arcs, so I know the vertexes match up. But when I gridded the cloth, there were a bunch more vertex points added to the cloth that have no corresponding arc vertexes.

Any suggestions? Thank you.

The bounding box is the blue box that surrounds the Pin component. The larger the box is the more vertices it grabs.
As you see here it only grabs one on the corner, then I make it larger and it pins more and so on. The actual geometry of the pin is really only so you can see the pins.


That’s a good demo of how pins work. Interestingly, since the pin is set up to always face the camera, you can get different results if you scale the pin by a side handle and orbit the camera.


Thanks Box and DaveR. I see what you are saying. I didn’t realize resizing the pin connected it to more vertices. I’m not sure that the Pin orientation being connected to the camera is easy to work with.

This doesn’t exactly answer my question, however. I was trying to use the “Make Pin” designation as n•1’s example seems to show. The pin icon doesn’t seem to appear in that usage. The motion constraints provided by the pin icon on the tool bar apparently aren’t as general as an object’s constraints accessible by designating an object as a constraint with “Make Pin” in the contextual drop-down menu.

Or maybe I’m just not interpreting n•1’s example correctly.

Thank you.

With some more investigation, I think I’ve answered my own question. I appears that when “Make Pin” is used, any of the “Cloth” object’s vertices that are within the bounding box of the “Make Pin” object will be frozen/constrained. It appears that the interpretation of “within” is strict and doesn’t include vertices that are on the edges or faces of the bounding box.

Also, it appears that when using “Make Pin”, the bounding box orientation is not connected to the camera location, but to the object that had been designated a “Pin.” This might be useful in some cases. A capability to select arbitrary vertices & designate them as Pin’s might be nice.

1 Like

This topic was automatically closed 91 days after the last reply. New replies are no longer allowed.