Feb 18, 2017

2016 EC Clubhouse Construction

A short post about the wooden clubhouse my buddy Elena (left) and I lead for this year's EC REX! (Elena's in charge; I was first mate :P )

Our goal was to make a small (smol!) hangout space alongside the larger East Campus Fort, to which we connected our structure with a rope bridge!

hangout space getting love during the East Side Party kickoff event
Our structure (left), attached to the larger fort project (right)
We (along with fort team!) submitted design drawings and structural calculations to a PE and architecture firm to get their stamps for the temporary structure, and also got stamps from Cambridge Fire Dept.

Construction started with digging holes and adding gravel for foundation cinderblocks. We spent a long time setting up a level surface that compensated for the uneven dirt ground. Even more time was spent moving the foundation locations around to avoid all the tree roots!

This took a night, then half another day to fix my mistakes
Our two person team had a more leisurely pace than the 8 person team leading the Fort build. Both teams ended up finishing within 10 days, perfectly on time for the party!
Put up the central tower by Day 2

Three towers up!

By the end of Day 3 we had floors on the entire second storey.

Starting on the tower again

Adding joists to the third floor.
Once the overall structure was completed, Elena and I split up the rest of the tasks. Elena added in the railings and I started working on the rope ladder.

The rope ladder was our access point for the second floor, and was attached to the spandrel via eyebolts and to the ground via anchors. The anchors get driven 2 feet into the ground, then a load pulling on the steel cable wedges the anchors at an angle.

Had to soak the ground to drive in ground anchors. These will be impossible to move.
Most of my time was spent working on rope ladders and bridges, and by the time I finished Elena had completed the rest of the clubhouse around me! The REX Chairs ordered a bunch of beanbags to make the hangout space cozier.
Note how we don't actually touch the tree. Trees in the courtyard are historic and we aren't allowed to hurt them!

Maya was super excellent and knotted the rope ladder together, then tensioned it up. In planning phase I miscalculated how angled I wanted the rope ladder to be; I set the horizontal distance too far and the structure ended up being more saggy ideal. We helped make climbing easier by adding a wooden handhold/step, as well as reinforcing the OSB shear wall to accommodate people stepping on it. Up above on the second floor, we screwed in 2x4 hand-holds onto the floor to help people exit/enter the ladder.

Art thanking our sponsors! East Campus and Clubhouse team are really happy we had such talented artists joining in
Elena and I saw the sponsor art, got super excited, and immediately asked the REX chairs if they would ask people to make the rest of our shear walls pretty. The chairs sent out a call for artists to the 100+ mailing list of people helping with construction and REX events, and we got the most beautiful art decorating the OSB shear walls.

Feb 16, 2017

[2.70] Seek and Geek #2: Pipewrench

If you haven't noticed, I really like pipewrenches.

Me on the "About Me" tab
Time to geek out about pipewrenches!

Better than roses for Valentine's Day (speaking from experience)
 A pipe wrench is an adjustable wrenches designed to be used on soft round pipes and fittings. One jaw is fixed and the other is effectively pinned with plenty of slop, such that pushing forwards on the handle pulls the jaws together.

Sliding jaw (left) and tooth block screwed to stationary jaw (right) are both replaceable

The teeth on the two jaws point in opposing directions; they dig into the soft pipe when rotated. The steel teeth blocks are also replaceable (so are the jaws), so a cast-steel or more-modern-aluminum will last you a long time.

Fun fact: the pipewrench was invented in Cambridge, MA by Daniel Stillson. Pipe wrenches are also known as Stillson wrenches.
Adjustment comes from turning a thumb-nut against a rack. This rack acts as a leadscrew but is also flat (ease of manufacturing since you can start with flat stock)

artofmanliness's beautiful cartoon

 Another cool fact about pipe wrenches is that the sliding jaw not only moves axially relative to the stationary jaw, it also is designed to have limited angular movement. The slop in the rack allows the sliding jaw to angle forward when gripping a pipe, providing some flexibility in positioning and added compliance for grip.

Feb 15, 2017

[2.70] Planar Exact-Constraint Exploration (plus bonus photoelasticity!)

I was watching this youtube video about visualizing stress concentrations in acrylic using polarized light, and I thought it would be really neat if I could see how each of the constraints hold load in my exact constraint (EC) system.

What's going on in the video is a technique called photoelasticity, which is helpful for visualizing stress strain analysis for complicated geometry or loading conditions (or both!). Acrylic (and many other transparent materials) exhibit birefringence under stress, where the magnitude of the refractive indices throughout the material correspond to the magnitude of stresses at each point.

I hope to play with two items of MechE science with this EC - I can experiment with different constraint configurations by moving the dowels around, and I can observe the relative stresses on the dowels imposed by the load.

Exact constraint is the principle of only using as many constraints as there are available degrees of freedom. An object on a plane has three - two translations and one rotation - so the goal here is to achieve only three points of contact. Using exact constraint lets you avoid having your parts bind or behave in unpredictable ways.

My EC is the simplest version you could make - pins in holes. I wanted something configurable, so I set up a acrylic-lasercut grid of 5/16" holes to accommodate wooden dowels. The smaller holes on each side are for mounting the EC to a standard hacksaw frame.

Pins in holes
The standard hacksaw makes a convenient experimental fixture for applying tension; I just have to attach my EC to the blade. The dullest hacksaw blade on the workbench got shortened (cannibalized with a hacksaw!) and two additional holes.

Saw blade taped onto table with VHB

The hacksaw then got shoved into a plastic u-channel thing I had in my room and the entire assembly was clamped to my desk. I elevated the assembly on a pair of V-blocks to get a better angle from my laptop screen.

I borrowed a camera polarizing filter from a friend, and now we're ready for some science!

Spinning the polarizing filters makes patterns visible!
I tensioned up the hacksaw and observed the birefringence patterns of just the grid plate. There are a few cool things here:
  • You can immediately tell which screw got clamped down with a lock nut vs. the regular hex nut
  • There are some nicks at the bottom of the acrylic plate that I didn't notice before
  • You can see all the defects in the lasercut holes. Maybe there was some uneven melting?
  • Since the regularly-spaced holes break up the neutral axis, you see some stress fields deflecting diagonally. At the same time, you get a cross-shaped pattern of dark spots between each hole which was not what I expected (I thought you'd get something more like magnetic-field-line-shapes)
How I thought stress patterns would go

How stress patterns appear to go

I'll probably revisit this idea after I read up on math :P

Back to the actual assignment - Exact constraint! Here it holds a pair of pliers with one pin on the bottom and two above, where it takes advantage of moment to hold things in place.

It supports this random item from my desk using two pins on the bottom and one constraining the side.

Cool birefringence patterns when I put force on the bottom two pins holding random-item-from-desk.

Also cool patterns happen when I exert a moment on the entire plate!

Feb 13, 2017

Prosthesis-Climbing-Shoe Prototypes


My January-independent-term project was making proof-of-concept climbing and hiking shoe attachments for prostheses! Specifically, one of those below-knee carbon fiber prostheses for athletes.

This kind of foot
They're typically used for running, generally on a track. Most attempts to make these suitable for other athletic activities (hiking, climbing, etc) require permanent modifications to the leg or try to fit the foot to standard shoes. Ossur + Nike have a removable-sole one that works as a high-performance trail runner, but theirs requires a proprietary fastening system that wouldn't work on generic blade-feet.

Ossur's Flex-Run w/ Nike Sole
Plenty of climbing-specific prostheses exist on the market. These range from ones mimicking biological footprints to highly-specialized points. There are also a bunch of proof-of-concept ones: climbers' own custom creations and student research projects (mine's going to be another silly student research project.)

This project started from a conversation with Hugh Herr that went something along these lines: Say I wanted to go climbing. Right now I have to carry a set of normal walking legs, then I need a set for the approach - that's a good hike, and then I have climbing legs. I'd also have to carry all the wrenches and stuff to remove them. That's a lot of weight and used-up space.

A set of attachments that could all fit on the same prosthesis - like shoes - would be more compact (and also probably cheaper!) Even more ideal would be a set of attachments that required minimal tools to put on and remove.

Like the ripstik project, this one also got funding by ProjX (The branch of TechX student group that funds projects) to get started over January, demoed in early February, and will probably become another ongoing, intermittently-worked-on item in my life.


I wanted to see how well a sock-like climbing attachment would perform using high-friction sticky rubber. Ideally, this attachment would be easy to put on and take off when in an unloaded state but stay put when loaded and in use. The attachment should also have a mechanism preventing it from flying out when the leg is swinging freely.

The project originally considered three concepts: a hiking "approach" shoe, a specialized climbing shoe, and ice crampons. In the interest of time, I decided to focus on the climbing/hiking shoes and cut out the crampon-shoe for this January project. A working friction-mechanism on a climbing shoe is highly likely to work well on the other attachments.

First set of drawings pitching the project to funding
I briefly entertained the idea of using cams clamps or other locking mechanisms as my fallback for keeping the shoes attached to a swinging foot, but I wanted to see how the attachments performed without these additional features first.

Instead, I decided to try including an elastic rubber strap that hooks onto the back of the prosthesis and applies elastic pressure (while hopefully maintaining enough static friction to not just roll down the slope.) This elastic strap ended up being entirely useless, but now I know.

Considering how retaining strap would work
Deciding what types of rubber to use where
I bought a durometer sample rubber pack to figure out what stiffnesses of polyurethane rubbers corresponded to the properties I wanted. My super-sticky, highly compliant rubber "sock" was made from 70Oo polyurethane, reinforced by strips of 40A rubber. I also bought 60A rubber hoping to use it as an abrasion-resistant toecap, but that rubber ended up not working well at all with my adhesives.
Concept design on a climbing foot with rubber types chosen
Climbing rubber came from Five-Ten's stealth-C4 resole kits. I had originally contacted resellers of all sorts of climbing shoe brands asking for rubber sheet samples - turns out my concept of a "small sheet" was 1 sq. ft. and theirs was 40 sq. ft.
This is plenty.

Design sketches after playing with rubber samples

I had the pleasure of talking to Laura Shumaker about her experiences at SFT Climbing (she maintains the most amazing design blog for SFT, which also has some manufacturing and testing insights.) She helped confirm my assumptions about the material properties of the special-climbing rubber, and suggested some adhesive brands that worked nicely with both fabrics and rubbers.


(In which Ava takes her non-existent knowledge of clothing design or footwear fabrication and tries to make shoes)

For this iteration, everything was just hand cut. Figuring out lasercutter parameters for the various rubber types is a job for next time.

I thought that if I punched the corners first, the rubber would be less inclined to tear. This probably would have worked better had I used a larger-diameter punch, or at least a hollow punch meant for soft things and not metals. The rubber tore at the corners anyway, but I fixed that by sanding larger-diameter inside corners with a dremel.

I went ahead and made two shoe soles using the same sock pattern. For the hiking shoe, I did a bit of reading about how real people design tread patterns to improve multi-directional grip and let water/mud optimally escape to aid traction, then promptly ignored all of that and just cut out rectangles. This rubber was a stiff but still high-traction mystery sheet. 

Mystery-rubber sole + treads got glued to the flat sock-rubber pattern. I got a frying pan and random billets to use as a clamp.

Grinding, folding, and gluing the rubber sock pattern was a long exercise in trial and error. I quickly realized I should've built up the sock using several layers of thinner rubber instead of attempting to form it from a single piece. Trying to curl the rubber into a toebox and not have the joins peel apart the adhesive did not work super well, even after reinforcement with fabric (random pants fabric I had) and grinding channels along the bend edges. 

Marking out where I wanted the shoe upper to end and the approximate location of retaining strap
Rubber dust got everywhere. Everywhere.

Bending and gluing toebox

toooo many clamps, and yet often not enough clamps

Started using pcord to compress everything while the glue dried

Some creative uses of string-clamps and real-clamps


Did I start a project just to ensure I'd go to the gym more? Whooooo knows.
I took the climbing foot to BKB-Somerville and played around on the bouldering wall.

First experiment was exploring toebox performance with the prosthesis on the wall. I picked some holds near the ground so I could apply weight by hand (I have two biological legs, so figuring out how to rigidly attach a prosthesis to myself was a no go)

The toe seems to grip well, despite having the side-rubber-wings peel up a bit. I didn't have a problem finding traction using the flat front or the corners of the shoe, but I expect this foot would have problems with cracks and narrow&vertical pockets. But overall this shoe reasonably performed all the footwork positions regular shoes do.

Common footwork technique examples

Second experiment was to try climbing with the shoe. For this bit I made a small HDPE form and lashed it to a flip flop.

A really silly test setup

This setup was far from ideal - neither the compliant flip flop nor the shape of the too-short plastic insert helped me put bodyweight at the toebox. In addition, my foot is wider than the prosthesis, so I couldn't properly utilize some of its nicer edging qualities.

Climbing with floppy-foot - notice how far the toebox is able to bend with this setup

In the test climb videos, you can see how I avoid using the toe on some smaller holds. I also pretend that my foot only consists of the surface covered by the shoe and deliberately test different foot positions, which results in some wonky moves. But the shoe didn't slip off the plastic, and it felt pretty good!

I took the hiking shoe out on a sunny day, got some glamour shots of it with the prosthesis on some rocks (below), and then ran around on it with the flipflop. This field test highlighted how fragile this concept shoe was - I managed to destroy it after only a few minutes on relatively flat ground. Next time, I shouldn't rely on the sock inner rubber to make any structural bonds and should make a sewn-fabric skeleton instead.

The hiking sole worked as well as your average sneaker on dry ground, but the shoe became unusable before I could test it on anything muddy or wet.

See the adhesive failing? Rubber is a tricky material

"Conclusions and Future Stuff"

Techfair came and went - I shared tables with friends' startup company and showed demo videos of the feet. I'm excited to also show off this work to Hugh soon and see if he would test a more robust version of these shoes.

Conclusions - 
  • While the soft rubber "sock" nicely achieved the objective of holding the shoe in place when pressure was applied, it was thicker than necessary and would have been better as several layers of thinner rubber bonded together.
  • The Renia Colle de Cologne adhesive worked well on both fabric and rubber, but was not magic. It only prevented peeling when the materials had large contact areas (not good for joining corners)
  • The retaining band did essentially nothing. Next iteration should use plastic clips or similar light-load solutions for countering shoe inertia when the foot swings in the air. Luckily, the sticky-rubber idea worked well for holding things in place under load, so I now know that the new retaining mechanism only needs to deal with the shoe's own weight.
  • I should form a pointed, rigid toecap for the next climbing shoe to make it more talon-like.
  • The next versions of these shoes should incorporate more rigid materials and less rubbers - nearly everything could have been improved with greater stiffness.

  • I can climb 5.7 to 5.9 and V0 gym grades in flipflops, but it's terrible.