Rogue Strandbeest Mk2 Build [COMPLETE]

[Brandon_Green]

Started assembling frame pieces, I initially cut the pieces the exact size and they wouldn’t fit into the jig, so I had to trim off 5-7 thousandths

I realized quickly that leaving the bearing in while fully welding was a bad idea, so I switched to using the top bearing only and removing it after tacking

I think the jig worked out perfectly, at least so far all the welded pieces look the same. Here are those pieces layed out on the next jig, they will be welded to some angle iron which will be the main backbone of the frame. Still have lots of crankshaft pieces to machine

[Brandon_Green]

At a current working rate of about 6 hours per month and 12 to 18 hours of work left to do, it should take somewhere between 1 day to 3 months to finish

Got the crankshaft plates drilled, these will be much smaller, 1" x 1/8" flat bar, vs the mk1 1.5" x 3/16" flat bar, but since the length of each piece is also smaller I’m hoping it will be strong enough. Drilling and then reaming all the pieces at once worked great, although some kant twist clamps would have made it easier to keep the pieces aligned between ops.

[Brandon_Green]

Got the main frame sections welded together, there was some slight warpage in the angle iron after welding, but I think it will be ok

Had to make jig holes 0.025 wider since the bearing holders ended up being slightly wider, eventually I will learn to always cut things with some clearance instead of cutting jig holes exactly 1.5" and attempting to make the bearing holders to exactly 1.5". Made sure to only take notes on my own material

Here are the crankshaft pieces, next step will be welding these in place, will use a 3d printed jig to hold them 120 degrees offset from each other

I thought I was going to have to cut apart the temporary alignment rod to get it out, but I was able to abuse some long extensions and hammer it out

Looks close enough to aligned I think, the concern is that its aligned well enough that once the individual crankshaft pieces are welded together that they are in line enough that it can spin freely without binding

[John_Marlow]

Brandon, this is an amazing build. The breadth of your various skills is also impressive. I can’t wait to see the finished item.

[Brandon_Green]

Welded the crankshaft into the frame, the part the struts will connect to and pivot on are bolted in with a 10-32 bolt, this is the only way I could figure out to assemble it without having split bearings. A simple 3d printed jig made it easy to align each crankshaft 120 degrees offset from the next

It’s hard to tell in the following video, but the bearing blocks are not perfect aligned, so some of them deflect ~10 thousandths when spinning the crankshaft. That’s why it has a lot of drag and stops spinning so quickly when it has considerable mass. But it should be good enough, the ball bearings do a good enough job so you can’t feel the drag.

If I was to try again I would try one of the following to get the bearing blocks perfectly aligned

• Annealing, heat backbone of frame to red hot after welding while its still in jig to relieve stresses and see if it straightens itself
• Bolt together bearing block supports to frame, instead of welding bearing blocks directly to frame (which introduces warping of frame), bolt them together so that there is some adjustability to align them while bolting. The clamping force of a 3/8 bolt should be in the 3 ton range which I think would be plenty

[Brandon_Green]

Since the entire mechanism only works on flat ground, the feet can be any size circle centered on the correct point, and it will simply raise the entire mechanism up by the radius. This was an error I made during the Mk1 build, where I thought the Jansen linkage dimensions meant that the point of ground contact had to be exactly what the linkage said.

I found some smooth rubber sheets (SBR) to use instead of the diamond pattern tread I used on Mk1. The feet will also be narrower at only 1.25" wide

These 3d printed rubber holders just bolt onto the bottom of the round plywood feet. Went with PLA since thats what I have an excess supply of, even thought its the least durable, I dont expect to leave this out in the sun or rain.

I’m attempting to just glue the rubber straight to the PLA with some Gorilla brand Clear Grip Contact Adhesive, the small scale test I did seemed like it may work

[Brandon_Green]

Good news, assembled a leg on the frame for the first time and it looks like I didn’t make any critical errors with strut lengths, mechanism moves smoothly through a full rotation

Here it is held in rough position how it will mount on the bike, now that I know the final height (21 5/8") of the center of the crankshaft from the ground I can cut the struts to bolt it to the main axle (which is 12 3/4" from the ground)

[Brandon_Green]

Realized after welding what I thought would be final weld that I made a rather large error when choosing a piece of angle iron to be the main part of the frame. My thinking was it would be plenty for the vertical and forward loads, but I failed to consider that the load is being transferred to it through a ~14" lever arm. By just grabbing the mounts I can twist the frame almost an inch, angle has almost no torsional rigidity. Luckily adding metal and making it heavy is easy, so I will just weld a 1.5" diameter round tube onto the angle to give it strength, it will just not look as nice as if I had properly engineered it in the beginning. Will delay completion by a few days at least.

Couldn’t find the sine bar in machine shop so I had to eyeball making a 26.2 degree cut, these are the vertical supports that bolt to the bike axle.

Will wait to drill holes for axle until I get all the legs on and get an exact height, would like to adjust it up and down slightly so I can avoid needing a chain tensioner

[MathewBusby]

Have you considered doubling the angles to make a box instead of adding tube? I imagine the acute gap between tube and angle is hard to weld.

Double the angle meaning use the same angle flipped over and welding at opposite corners to make a square tube.

[Brandon_Green]

Got a fairly large diameter but thin (1/16") wall tube to use, I should probably look up the actual formulas for torsional rigidty, but I’m not an engineer, this looks like it should improve things

[Brandon_Green]

Seems like I could have made some kind of game of thrones meme here, getting close to first test ride

[Brandon_Green]

Welding on a tube to reinforce torsional rigidity worked amazingly well, went from being able to grab the mounting plates and move them back and forth easily (twisting angle iron) to feeling very solid

Decided to go ahead and paint it at this stage and just hope won’t have to do anymore welding on it

I’m just going to ignore where I forgot to come back and complete the weld after retacking it on this side, the other 3 sides are fully welded, so hopefully that is strong enough

[Brandon_Green]

After 3d printing lots of spacers, went ahead and put the legs onto the frame. This will allow determining the exact height that it will bolt to the bicycle at.

Just have to slide the rod in and start putting each leg on, lots of silicone grease was used in all the joints

I printed some different height spacers because the left and right sides had slightly different distances between them

Turns out all the frame work turned out close enough to the initial 3d model that it all came together as expected and everything moves freely

Next will be holding it up to bike to mark and drill the holes where it bolts to the bike axle, I am hoping that I will get lucky and the chain will not need a tensioner, but I have a couple tensioners on standby incase one is needed.

Will need some large warning stickers or something to not touch while in motion, there are a ton of pinch points

Close up of how the diff attaches, there was plenty of room to just slide the coupling on, then put the key in, and set the diff there and slide the couplings back and tighten the set screws.

[steve_a]

You sir, are a steely-eyed mechanical builder! This is beautiful!

[Lampy]

cannot wait to see you ride that beast!

[ESmith]

Will need some large warning stickers or something to not touch while in motion, there are a ton of pinch points

Something like this might send the message:

Of course that’s not as entertaining as the semi-standard ROTATING SHAFT HAZARD graphic:

[Brandon_Green]

The great thing about 3d printing a lot of parts, can simply open up the cad file, tweak some sizes, and reprint, here the top sprocket was 0.250" inches out of line with the bottom (amazing close for previously just eyeballing this alignment)

So I just had to add some distance and am reprinting that part. I’m assuming that the extra torque (bending housing left to right) on the diff housing generating by moving the sprocket even further from the center won’t be significant

[Brandon_Green]

Had some expert help today and was able to finish Mk2

I randomly missed getting the perfect spacing by about 1/8", so I had to add a tensioner

I got the tensioner part from amazon (just lucked into matching the red/black theme) and mounted it to the frame by threading a 5/16-24 hold and tightening a bolt to the proper amount to let tensioner move freely but not wobble, then a jam nut on the outside keeps it from loosening. So far it looks like the 1/8" thick plate that I tapped is just thick enough for this work. Avoids having to weld a nut on after I already painted everything.

For the attachment of the frame to the bike I ended up just jamming a spacer in and zip tying it, much simpler than the adjustable bolt based attachment on Mk1, I can simply print multiple heights to adjust the angle of the frame to the bike.

[Brandon_Green]

First short test ride:

[mblatz]

Tour de France!! Tour de France!!