The goal is to replace the rear wheel of a bicycle with a strandbeest walking mechanism before November 10th and see what happens. Similar to this but hopefully something much faster / smoother / maneuverable (Going to try 3 legs instead of 4). In the future we may be able to race multiple walking propelled machines against each other.
Anyone who is interested in helping with the build is welcome to come lend a hand, build days will be on the calendar, there will be a lot of lathe and machining work if you are looking for an excuse to practice on the lathe.
The first build day is this Saturday October 13th 2pm, where at a minimum I will be CNC cutting all the pieces for the leg linkages
Previously I built a proof of concept to test out the delrin + smooth rod joints (there are a 6 joints per leg where each joint has 3 or more bushings), a press fit of ~15 thousandths makes the delrin bushing very rigidly mounted to the plywood with no adhesives needed
After seeing how over built the prototype was I changed the design and made the plywood significantly thinner, just 0.75" min plywood width for linkages vs 1.25" seen in prototype, which also reduced overall mass by over 25%. Instead of doing any destructive tests of the revised thinner design I’m just going forward with the full scale build guessing that it will be strong enough, the legs may or may not snap when put under load, (it only costs $45 for another sheet of plywood to remake the legs, so not a huge risk)
I have a set of over/under reamers that you can use for those 0.500" holes in that delrin. The sizes are 0.499" and 0.501". Let me know if you want to use them.
I assume the largest risk comes in any defects or voids in the home depot quality plywood causing unexpected weakness, the foot and upper triangle are probably under the most bending stress (since I made them not triangular on a whim to reduce mass), the other linkages should only face compression / tension
Couldn’t find many plywood beam strength test videos on youtube, but here is a shear test where you can see it fail at the point where there were defects in the plywood as you would expect
I wonder how much stronger you could make the ply with a layer of fiberglass/epoxy on each side… I’m thinking a lot. Could possibly apply it before cutting with the CNC, but expect to dull the blade and create some interesting dust.
The compressive strength parallel to face of 1/2" birch plywood is ~4,000 psi from what I could find, the cross section of a linkage in the design pictured above is 1.11 in^2, so it should be able to support over 2 tons through each linkage in pure compression, my guess is the dynamic load should not exceed 300-400 pounds of force and where the 1/2" axle is press fit into the plywood with only 5/8" of plywood around it would be the weak point.
My concern is buckling. I’ve spent the past hour both mildly inebriated and also pulling out books on buckling strength. It’s a bigger deal for me because of the thinner plywood. Also plywood is weird; orthotropic and stuff. Anyway, something slightly comprehensive is in the works. I need it for my own work anyway.
Calculating buckling starts to get into estimating bending moments and moments of inertia, which is beyond me. I’m just hoping that side loading (turning or leaning bike so that force is applied outside plane of the joints) will be minimal enough to not have to consider
In other interesting stats, the currently planned cut will use 7.6 sq ft out of the 32 sq ft of plywood (23% of the plywood going into finished product), all the plywood pieces should weigh 10-12 pounds. With all the other metal axles and frame hopefully can stay below 25 pounds for the entire mechanism. The frame (connecting mechanism to the bike) may end up being simply 1.25" or 1.5" steel angle instead of CNC cutting more plywood, I’m thinking the steel angle will be simpler to attach to the bicycle.
Suddenly the legs dont seem as big when I overlay them on a 26" bike, the legs are going to be moving very quickly if it can achieve anything more than walking speed
We got all the plywood cut out and the bushing and crank bearing holders machined on the lathe. I should be able to glue and screw all the linkages together over the next few days so that we can start on the crankshaft Wednesday night
I was surprised how off the estimate in vcarve was, actual cut time was 1/3 what it predicted (only 25 minutes of cutting time vs 1.5 hours predicted)
I dont think the CNC router table spoil board is very flat currently (on the edges can see where it has pulled up from where it was glued previously and there is some gap between it and router table), on some areas it cut through fine but on most of the board it left a significant amount, which just took longer to go over everything with a jigsaw and clean up edges before sanding
Making bushings out 1.5 inch delrin, luckily the through bore in the lathe fits a 1.5" shaft, so I didn’t have to cut the rod down nearly as much as I did the first time.
Should have realized before cutting everything out, but I didn’t notice the plywood from home depot, while being 1/2 Birch was not the same as the prototype plywood I used(far fewer plies, probably not as strong), will carry on and see how this turns out.
I also didn’t double check the thickness before cutting, and this was 10 thou thicker than I expected, meaning I will need to sand down each face that meets to add enough clearance
Leg assembly is going to be rate limited by the number of clamps I have, my initial plan was to use #8 wood screws to clamp pieces together during glue up, but after trying that on one piece I think the holes would weaken the already sketchily thin plywood too much.
There is some noticeable ‘flex’ in the leg if I put the foot on the ground and hold the furthest joint and press in the middle with all my weight, I’m just going to call that extra ‘suspension’ a feature for now, can always bolt or glue on some extra material to stiffen the long section of the leg if needed later.
Will be a work day tomorrow Wednesday 10/17 around ~8:30 to 11:30, any member interested in helping with the build is welcome, will be lots of bridgeport and lathe tasks including:
Create jig to align strut holes for glue up (strut ‘K’), jigsaw out plywood and use bridgeport to drill two 1/2" holes 17.168" apart
Rough cut 3/16" x 1.25" steel flat bar for crankshaft to 5.75", use bridgeport to drill holes 4.160" apart, one reamed 1/2 hole, one for #10 bolt, hand grind corners
Cut steel rod for center of crankshaft, face ends to final dimension
Cut smooth rod for ends of crankshaft, drill and tap 1" into each end for #10-32 bolt
Use bridgeport and boring head to embiggen 11mm sprocket ID to 1/2 inch
On at least one side, weld rod to the flat bar, being sure rod is perfectly perpendicular to flat bar
Assemble jig to hold two sides of crankshaft perfectly parallel, weld exactly 120 degrees apart
My progress has been limited to having part of a Perl 6 units of measure library working. Way overkill for the analysis, but I don’t want stupid lroblems to come up when analyzing 200 links.