@import url(http://kodlab.seas.upenn.edu/pub/skins/sinorca/basic.css); @import url(http://kodlab.seas.upenn.edu/pub/skins/sinorca/layout.css); @import url(http://kodlab.seas.upenn.edu/pub/skins/sinorca/sinorca.css);
My task this summer was almost exclusively the mechanical design of the new X-RHex robot. There were of course many unexpected difficulties, many of which arose from two major realities: A lack of experience and an extremely difficult size constraint.
The first logical step in any design process is to fully define the constraints and functional requirements such a design must meet. In this case the goal was to construct a new RHex the same weight or lighter, exactly half of the height, the same leg profile, much stronger mechanically, and with increased modularity and accessibility for maintenance. Stress testing with the first revisions of the CAD design showed our frame capable of withstanding 3x the worst case dynamic load it would ever experience on a regular basis with minimal fatigue.
There were other major differences in the design approach I took for X-RHex versus that of RHex. Probably the most significant (for which Kevin Galloway receives all credit) is the concept of having the motor mounting be a structural member, attached rigidly to the frame at top and bottom. In conjunction with all who worked on X-RHex, we also decided to make the electronics (motor controllers and power) consolidated into a single modular stack, one of which placed on either end of the robot adjacent to the battery. These stacks could be easily removed from the bottom of the robot. This all was in support of the requirement that every electrical component and all motors could be removed without disassembling any other part of the frame. Finally, I also made an effort to make the shell and shielding for the robot seamlessly integrated into the design of the frame, easily removed where needed and far more structural in purpose than the previous RHex style plating. Carbon fiber channels down the length of either side of the robot lend a great deal of strength to the aluminum frame itself. Side loading batteries make loading and unloading much easier as well.
Now that the construction itself is complete, we are currently stuck in a seemingly endless cycle of solving small unexpected problems; things that need some persuasion to fit and loose ends that were forgotten in the process. Probably the greatest success thus far, in my mind, is the one to one gear solution for absolute magnetic encoding of the motor output shafts, which after testing proved to be fully functional. As with any construction process I expect that there will soon be a revision of X-RHex in preparation for building another in the future with improved ease of access and assembly in mind.
Copyright Kodlab, 2017