PDA

View Full Version : 3D Printed Endoskeleton



bradr
05-21-2015, 06:52 PM
My objective is to completely 3D print the endoskeleton pieces and compare against aluminum frames. I downloaded the 3d model of HR-OS1 from https://sketchfab.com/models/28d2a6f8729d4d508820bcfd8f943bab. (Awesome and thorough wiki by the way). Using Rhinoceros on mac I was able to separate the pieces and get them ready to print. I started with the F1 Bracket (http://www.trossenrobotics.com/bioloid-metal-f1-bracket) and below is the result.


http://bradrisse.com/f1-bracket/20150521_162002.jpg

http://bradrisse.com/f1-bracket/20150521_162057.jpg

I don't have the aluminum bracket to compare, but the strength of the ABS plastic is surprisingly strong. I tried to break it by pushing the ends together.

http://bradrisse.com/f1-bracket/20150521_162229.jpg

My printer settings were Infill: 40%, Shells: 2, Layer Height: .2mm for the one pictured. Im currently attempting Infill: 100%, Shells: 1, Layer Height: .1mm. I has to use a drill bit to make the holes wide enough to fit M3 screws, but wasn't difficult.

I believe this will not only reduce the cost if you have access to a 3-d printer, but will also have some other benefits like weight, color, customization, spare parts, and cost.

Aluminum is already pretty light, but this bracket feels like a feather. Right now I'm printing with white filament, but I plan on printing a variety of color options. I would also like to try some prototype legs by adding a few more servos and lengthening them a bit. Printing extras will be a snap and overall the cost should be lower.

Some potential cons would be the plastic breaks under tension, degrades over time, access to a 3D printer and takes time to print the parts.

Will update with remaining parts and how they come out and how long it takes to print all the pieces. One thing I noticed about the 3d model is it seems to have the old body plates.

KevinO
05-22-2015, 12:57 AM
Respectively speaking, the fact you can actually bend and more than likely twist the bracket shows that this will fail. It won't be able to stand if there is any flex, and your last picture shows a ton of flex.

bradr
05-22-2015, 01:59 AM
I agree. This Darwin clone had the same issue http://www.instructables.com/id/Robot-Cloning-by-DIY-3d-printers/?ALLSTEPS. He ended up using aluminum for top leg as it kept breaking under stress. Ill try to beef up the sides a couple mm and inset the screws. The servo inside the bracket might reduce the flex.

tician
05-22-2015, 11:34 AM
...it's been a while since bioloid-mike has been around these parts...

Anyway, it is probably better that you make the endoskeleton/frame and exoskeleton/covers a single, larger, integrated structural unit like the 3D printed 'igus' version of the nimbro-op. Plastic generally is a lot less durable than aluminum, and printed plastic is even less durable than injection molded plastic even if the exact same material. You cannot simply transfer a metal design to plastic and expect it to be anything near functionally equivalent.