ArmyrHex

So I thought I had this milling thing under control…. Not so much. Today everything went wrong.

First I forgot that my stock aluminium piece I was using today was 10mm wider than I have modeled it in Fusion. Que my 4mm endmill digging in with full engagement into the piece instead of shaving off about 1mm at the time from the edges and promptly breaking off.

Went to the shelf of endmills to purchase and found that there is one last 4mm carbide endmill. Mounting in in the machine I slip with the wrenches. That was apparently enough to break the brittle carbide mill clean off.

Went back another time and found another end-mill of a different kind. Mounted it and started. Apparently it is true that end-mills for steel don’t work for aluminium. After a promising start I suddenly hear a snap. And that, my friends, was mill number 3 for the night.

So having wasted most of an evening and some 300sek in broken end-mills I finally went back and milled the piece out with my trusty 8mm endmill that worked so well the week before. The fine detail was lost of course, but at least I got something that looked sort of like my CAD model.

The end-result, two pieces that at least fit together. Could have been worse.

So the Delrin experiment went well! My parts were so thin and Delrin so soft that the pieces got deformed by the vice, but over-all it worked as expected. Time to graduate to Aluminium!

What do you know? A beauty, isn’t it?

Unfortunately, the tolerances were off and the bearing didn’t fit the first time, but that was an easy fix.

Also, I milled two pieces, but learned about peck drilling the hard way. If you try to drill a 2.5mm hole 25mm deep in one go you break your drill…. So now I know. For those who do not know, peck drilling means that you drill a few mm at the time and then pull the drill out to clear out all of the removed material and reapply alchohol mist for lubrication before going down again.

So to build a robot, you need to design a robot. Since RHex was designed by universities (Using military money, of course), you can actually find documentation.. Not great documentation, but still: https://kodlab.seas.upenn.edu/uploads/Main/xrhextechreport.pdf

So the original weighs in at something like 10kg and cost like a small car. I decided to aim for a 2.5kg machine. So what sort of motor would I need? Top-of-the line brushless DC motors are out of my budget range, but six of these solid little ones would give me a pretty good kick. This would give me about 10kg (98N) of max upward force. For a 2.5kg robot that leads to a vertical acceleration of about 3g when all legs are pushing at the same time.

These motors come with an encoder built-in which is convenient: https://www.pololu.com/product/4846

To drive them, this little driver with 5A peak should be able to keep up, I hope: https://www.pololu.com/product/1212

Below is the design for the metal frame and the motors in place. The red little pieces fit a ball bearing for stability and an optobreaker to be able to calibrate absolute position of the legs. The space between the motors should fit the DC motor drivers, an Arduino Mega and enough batteries to power the entire thing.

Use this link to download the model for Fusion 360: https://a360.co/2NWdxsP