Ripstik Assembly

The Ripstik was completed on the morning of February 2nd, just in time for XFair. A month later (3/2/15), I'm finally documenting the experience and realizing my delayed motors still haven't arrived!

Anyway, XFair was pretty fun. It took place in the midst of the Boston snowstorm, so I didn't even attempt to ride my deathtrap on the slick floorspace. Instead I took off the coverplates to show fair-goers how the internal mechanism worked.

photo cred: John Chow

There's still a lot of work left to do before the snows melt and before this will be rideable. My first (and only) trial run had me richocheting off walls in the MITERS hallway, for two main reasons. One, my torsion rod was too squishy to supply any meaningful force. If I placed a foot on one end of the board, I could step on and twist the other half with barely any resistance. That had to change.

Way back in freshman year, I found a slim bar extrusion of spring steel that could possibly substitute. It was too narrow to usefully put screws through, so I milled two clamps to house the bar.

original torsion bar

ripstik with new torsion bar installed

I also discovered during the trial ride that my ripstik was uncomfortably tall. Turns out when you learn how to ride on something an inch plus half shorter, anything else becomes weird. I removed the upper half of my caster assembly and merged its functions with the angle block, which took off 1/2". Maybe that will be good enough. 

Top Plate, Nutstrips, and Caster Parts

Ripstik: slowly and steadily coming together! 

The shiny blue shim steel I'm using for the top plates (aesthetic purposes) is really thin (0.03"), so it was sandwiched between sacrificial plywood and 1/8" 6061 aluminum to make sure it didn't warp while waterjetting, 

In addition to that, I'm milling angles on my tube cutouts so that they properly attach to the bent chassis.

The top plates will attach to the chassis with nutstrips, pieces of 0.5" square extrusion with 1/4-20 tapped holes every inch. Marking and powertapping all these holes took a while.

 Here I'm testing a quick caster-lock idea. A caster consists of an upper half and a lower half, with a common pivot shaft (1). If I have a slab that can slide through slots in both halves, I can lock and unlock the caster.

Once convinced that my lock idea would pan out, I proceeded to make all the parts.

Milling the angle blocks to 21.5°. These blocks connect the caster assemblies to the chassis.

kingpin/bearing assembly

freshly-waterjetted caster parts

    

TIG Welding Aluminum (poop welds)

I needed to fix cracks in my ripstik chassis caused by the bending brake. The front portion, in particular, really needed fixing. A well-intentioned friend found it cooling down in the oven, decided I probably kept it in the oven for a reason, and left it in while baking cookies at 350°F. (This is close to the tempering temperature for 2024) My metal ended up heat-hardening and even after another stint in the oven still ended up with massive cracks.

Some panels even had full blown tears on the edges. I definitely didn't want to rest any weight on cracked panels, so I asked the Media Lab shop guys for advice. They pointed me to the TIG welder, taught me how it works, and told me to practice on scrap.

It took me three days of practicing and a bunch of scrap metal before I felt ready to try the real thing. I've never welded before, so starting out on TIG and aluminum was especially difficult. Luckily for me I had plenty of practice material.

   

Judging the distance I needed to keep between the torch and my workpiece was pretty hard. I crashed a few times, which meant I had to stop and regrind the tungsten shaft.

Oops.

Once I was confident with my abilities on the scrapped interior shells, I started fixing cracks on the real thing. The result was not pretty, but it was effective.

photo cred: Melody Liu

+ me at the welding table!

Ripstik Chassis and Baking Aluminum

I'm rushing to get a deliverable out by Feb 2nd, so I've been lagging on documentation. But here's some of what I've been doing with the ripstik chassis:

I'm using eighth-inch 2024 aluminum plate for the body, originally going with a 2 layer shell design because I wasn't convinced that one layer would hold a person. I modeled sheet metal folding in Solidworks and waterjetted the flat pattern.

I chose 2024 instead of the more common 6061 aluminum alloy to facilitate bending. 6061 (especially the T-6 variety available to me) is extremely susceptible to cracking when bending. 2024 T-3 is doable, and is still readily machinable. T-3 and T-6 refers to the level of temper in the standard sheet. For me, lower temper is better for bending. 

Then, I put my cutouts on the bending brake and put in angles. I ended up needing a larger bending radius than I had planned for; turns out that the specific sheet I was using needs a bit more than the "thickness of part" rule of thumb.

In addition, turns out my 0.125" plate was just barely within the maximum thickness allowed in this brake. I had to throw around my body weight on the lever to get even a slight bend, and I wasn't looking forward to spending hours and hours to get the angles I wanted.

The internet tells me 2024 aluminum has an annealing temp (temperature at which it softens) of around 750°F, and a hardening temperature of around 300°F. My oven gets to 575°F+ on broil, so I stuck my plates in there for 2 hours, let cool to room temp (1 hour), and tried again.

This strategy worked reasonably well (beware of well-intentioned friends leaving your aluminum in the oven while baking cookies!) Only the narrow tube-holding plate cracked severely after the pseudo-annealing, and I was okay with cutting those off and waterjetting replacements.

Aluminum 2024 Recipe

        1. While oven is at room temperature, place aluminum on top shelves

        2. Set oven to broil

        3. Open the windows if necessary. (broiling aluminum smells much like hot glue)

        4. Wait 2 hours

        5. Turn oven off. Without removing aluminum, let rest for 1 hour

        6. Remove aluminum. May be warm to the touch.

LEGO Minifig

Photo post-processing done with Paint.NET's posterize feature 

they have hoodies!

To celebrate LEGO's collaboration with MIT's Media Lab, we were all invited to come eat dinner and play with blocks. One of the stations they had set up was "Make Yourself in Minifig Form," with suitcases full of pants, tops, heads, and accessories.

When I was a kid, LEGO girls only had one hairstyle! (Though this particular hair was probably Harry Potter's) And I didn't realize LEGO now made parts that fitted on the shoulder. Or that neck-parts existed. I remember making tiny Kleenex capes and scarves, but these are real plastic!

So I huddled at the "Kids Corner" with kids ages 3-43, only half listening to the Ninjago presentation on-stage (also really interesting,) searching for myself. A yellow zip-up jacket. Slate-colored jeans. Blue hoodie. Glasses face. Black messy hair. Green cross-body bag. Yup, that's pretty much me. Just add coffee.

LEGOs are great.

Me + coffee cup

good-cop/bad-cop head makes for interesting changes of expression

My photoshoot setup.
Everything's possible if you just believe in the macro setting!