After much work this weekend I was able to get my first splint off to the 3d printer. I’m very happy with how it turned out but it seems that whomever was scanned in for the model I downloaded either has pretty small hands, or there was an issue with scaling when importing the file, but because it doesn’t seem to be off by a large factor I would assume that it is due to the hand size of the model.
But to back track, at the end of last class I was having issues with combining bodies, and 3-d printing. In my old definition I had created solids by converting surfaces to meshes and using weaverbird’s thicken mesh feature, but when I saved these meshes to .stl, I had errors that I talked about in my last post, due to bodies being inside one another. While I did not talk about it in my last post I had also experimented with trying to create solid closed breps by using offset surface and lofting between the edges of the surface. But I got errors due to list issues where points were referencing the incorrect point on the offset surface. I thought that this an essentially unsolvable feature, but I found that I the error had arisen from a join function that connected the curves together incorrectly so after removing this and instead joining the many surfaces of the loft together afterwards I was able to get clean lofted surfaces.
With this change I was able to get clean .stl files that passed when I uploaded them to shapeways website, even if the wall thicknesses were very slightly too thin. While shapeways prices for the part were very cheap, lead times were too long to produce a part in time for Thursday, so instead I used 3duniprint.com, a local 3d printer I had used previously. He was able to print the part in high detail resin in 36 hours, even if the price was slightly higher than shapeways. He has helped me an many of my classmates out of jams in the past even going as far as hand delivering a model at midnight the night before it was due, I would recommend him to anyone especially his resin prints.
In addition to the work I did to the grasshopper definition, I also did some research on splint design, the thumb splint you see above was based on arthritis splint that I had found on an occupational therapist’s site that specializes in hand splints for arthritis users.
Because these users need custom splints, but don’t have the immediate need or have the swelling problems associated with those who use splints to heal from traumatic injuries, they are perfect candidates for these 3d printed custom splints.
As I continue to refine my grasshopper definition I am going to continue to make it more and more automated so that it requires less tweaking when running it with new geometry. Currently there are a few issues that make you adjust the sizing of the voronoi each time and sometimes causes the exterior support to come out crooked.
EDIT: So while I am trapped in my apartment waiting for the people from conedison to come and turn my gas on I was able to fix the “crooked error.” As you can see below with the finger splint, the previous definition sometimes had errors if the point at center of the surface had a “weird” normal in comparison to the geometry of the splint
I was able to create a fix by populating the surface with 10000 points, evaluating the surface at each of these points and then finding the average normal vector to find the average normal surface to project onto.