Donut/Torus Redesign and Upgrade

Back in September, I saw a design for a torus made of paper on the internet, and decided to take on the project. After four hours of meticulous cutting and fitting the pieces together, I came out with this. A neat feature that I discovered accidentally is that it folds flat!

Last week I was reminded of this project at Peter Mui's Fixit Clinic that was hosted by The Saturday Thing and decided to CAD the design up. Unfortunately my image-to-.dxf software didn't like the file I had for the paper torus, so I ended up making the shapes freehand in AutoCAD. This took about an hour, but time lost in design was made up in cutting time. I scaled the drawing up so the outside diameter of the torus became about 9 inches. The slots are designed for .093" acrylic, and luckily we had a big sheet on hand. To my pleasant surprise, I hadn't screwed anything up in the design and everything fit perfectly! After taking a few pictures of the clear creation, we shut off the lights in the lab, took some of the Aurora Bear bases, and lit it from below.

Stator Wrapping Complete!

It's a milestone in an engineer's life when he completes the wrapping of his first stator for a custom brushless D motor. I'm proud to announce that I have hit that milestone, and my thumbs are killing me. First step was to make a diagram for wrapping, which can be found in this post. Over the course of the next 10 hours (not consecutive), thumb and finger health decreased steadily and a few pictures were taken. To help you understand the numbers in the captions, one "Phase" consists of two pairs of two strands (four strands total) each wrapped around two teeth (four teeth total). The A phase consists of two teeth wound clockwise and two teeth wound counterclockwise. Put three of these phases together in the right order with the right orientation, and you've got yourself a motor.

1/4 of phase A wrapped

1/2 of phase A and 1/2 of B wrapped

Phases A and B complete, 1/2 of Phase C

One more tooth to go!

DONE. GORGEOUS.

Finishing this felt amazing (as I loudly proclaimed in N52 when I completed the last turn) despite my throbbing fingers. I've technically been working on this project since late August, and I'm finally getting somewhere. Shortly after I posted this photo to facebook, Shane and Charles displayed their amazement with the quality of the wraps, which is apparently pretty good. Which was what I was aiming for. The next step will be machining the endcaps and spacers to house the stator and rotor. Too bad I'm going back to school in a week and I'm helping the MIT Marine Robotics Team with some stuff until then.

Looks like it's time to get reacquainted with the machine shop. Let the hub motor season begin.

Stator Wrapping and Xbox Case Mod Preview

I finally got off the ground with (name TBA) scooter's drive system. After many months of planning and procrastination, I began wrapping my stator... and boy do my thumbs hurt. Shane told me before I began that there is a spectrum of quality of stator wrapping from neat and meticulous to quick and dirty. I chose to pursue the neat and meticulous path because a more tightly wound stator would be a more powerful one. A side effect of this path is that my right thumb and forefinger are now throbbing from pulling as hard as I could on wire for an hour straight. I am not looking forward to wrapping another 11 teeth, but it must be done!

This just popped into my head: I might call this project Time Scooter. Because I've been working on it since August. We'll see if the name sticks.

Here are some pictures from the day:

Wrapping configuration diagram

Wrapping two strands at the same time is tough.

Another project I've been doing on the side this past week is creating a clear case for my xbox made from acrylic. Unfortunately 3-402 has been (hopefully temporarily) closed before MIT's IAP starts, at which point it will be occupied full time by a class. I haven't been able to make any new parts, but I have started to model the case in SolidWorks. So far I have taken my Xbox apart four times, designed and made a prototype of the motherboard mount... and that's about it. I was aiming to complete this project before school starts again, but that might not happen because of the fabrication delay.

Let's see what this thing looks like on the inside.

What an Xbox looks like on the inside.

Might be tough to see (it's clear!) but here we have the first version of the motherboard mounting piece.

Hard drive module, disassembled

Ideally I will buy a ~120 Gb 2.5" hard drive and format it to xbox's native system so I don't have to spend exorbitant amounts of money for a tiny amount of space. The current drive was $80 for a 20 Gb USED, which I find to be absolutely ridiculous.

ALASKA UPDATE!
I just got off the phone with Ed Moriarty, who is currently in McGrath, Alaska, and he says the Aurora Bears are a huge success! The kids up there are having a great time making the bear assemblies, and even the superintendent of schools showed up to make one!

End Year Post: Aurora Bearealis

Happy New Year, everyone! I'll spare the little details, but boy, 2011 was quite an adventure. A year ago I would have never thought I'd be where I am now, and it's still amazing to me how lucky I've been.

Goals for 2012:

1) Create something new that has the potential to improve someone's lifestyle
2) Do something that is universally regarded as "totally awesome, dude"

(Goals may be changed, removed, or added over the course of the year)

On to the engineering part of the post:

The past week was incredibly hectic. Ed Moriarty is headed to Alaska on Tuesday to promote interest in science and engineering in young kids. To help him achieve this goal, I designed and mass-produced what he calls "Aurora Bearealis" which combines the Aurora Borealis (northern lights) with UAF's Aurora Bear logo.

Over the past 96 hours, I've spent about 50 of them at the Edgerton Center making this project come to life. I've also consumed about a pound of boneless chicken wings and a whole pizza. That's not a lot for four days of food, but then again I really didn't find time to eat. Big thanks to Ed, Tommy, Bruno, Kenny, and the several drop-ins who took an interest in the project and helped in any way they could. Here are a few shots that were taken when I wasn't CADing or eating.

A tricky part to cut efficiently.

Approximately 150 full assemblies. In a box.

A video of the final product. Turn down your volume; the video was taken in the laser room with ventilation going at full blast.

After several days of dreaming in vector format spending more time at MIT than at home, we cranked out over 190 full assemblies, which, as I said earlier, took about two full days' time. This includes transport, setting up parts to cut in AutoCAD, cleaning machines, preparing materials for cutting, establishing cut settings, actual cutting, taking materials out of machines, cleaning machines again, testing prototypes, sorting parts, and packing them into sets of 10 or 20. It was exhausting but extremely rewarding. As put by Ed, these bears are going to make a lot of kids in Alaska really happy because they now have a really cool night light during the darkest time of the year, and the kids can say "I made that!"

TL;DR I am now one with the laser cutters in 3-402.

SolidWorks and Engineer Magazine

This Saturday, I didn't build anything. Anything physical, that is. Instead I spent several hours learning how to use SolidWorks, which MIT kindly provided for me for free. After several grueling hours of learning the ins and outs of the software and making some lovely extruded polygons, I set out to model my scooter.

Mate. Mate mate. Mate mate mate.

Exploded

This modeling will probably eat up most of my time until I get lazy and just think about the design in my head. That actually isn't such a bad thing, since head-thinking has worked out pretty well so far.

Last but not least, I was interviewed by Mark Dwortzan for Engineer, a semiannual publication about the College of Engineering. The story will be in the spring issue, which will be on the shelves around March (?) of 2012.

New Scooter

Here's to actually doing projects and having fun doing them. The long anticipated construction of New Scooter (official name TBA, not to be confused with Pneu Scooter) finally got its butt off the couch and went to the whiteboard for some calculations.

So lets say that I want to be able to climb a 10% grade on my scooter.

F = mg sin(theta)
F = 75 kg * 9.8 m/s^2 * .1
F = 73 N

This means that the force of gravity pulling me back down the hill is 73 N, and that I want 73 N pushing me up the hill. Lets say I want to climb the hill at 5 m/s.

P = linear force * linear velocity
P = 73 N * 5 m/s
P = 365 W

Torque = Force * radius, and I have a 6" diameter wheel.


T = F * r
T = 73 N * .076 m
T = 5.55 Nm

I'm planning on building a 33 volt LiFePO4 battery.


365 W = 33 v * 11.1 A

Now for my new favorite equation: Tea is for nibbler. I mean T = 4NIBLR

That is, Torque = 4 * # of wraps * Current * Strength of magnetic field * Length of stator * Radius of stator

For this calculation, we will estimate that the B field is about 1 T. This is an estimation, but it's really not far off of the actual value.

T = 4NIBLR
N = T/4IBLR
N = 5.55 Nm / 4 * 11.1 A * 1 T * .0254 m * .034 m
N = 144 wraps per phase

Now lets say the maximum current through quadruple 22 guage (four strands in parallel) wire is 40 amps peak. Replacing the 11.1 A in the equation with 40 A, we get N = 40 turns per phase.

If A = 11.1, N = 144 to get me up the hill.
If A = 40, N = 40 to get me up the hill.

Now for the torque constant, in Nm/A.

Kt = T/I = 4NBLR


For 11.1 A we get .5 Nm/A
For 40 A we get .14 Nm/A

1 Nm/A = 1 V/(radian/s)


33 v/(.5 v/rad/s) = 66 rad/s
and
33 v/ (.14 v/rad/s) = 236 rad/s

To calculate speed

66 rad/s * 1 rev/2 pi radians * 60 s/1 min * 60 min/1 hr * 2 pi (3 in)/1 rev * 1 ft/12 in * 1 mile/5280 ft = 11.2 mph

This is too slow. Well actually, it's a pretty safe speed, and at about twice the speed of me comfortably jogging somewhere, I'd be pretty happy going that speed. However, I'd like to go a little faster, so let's calculate my speed if I used the 40 A + N = 40 configuration. By multiplying by 236/66 we get 40 mph. As crazy as I am, I'd like to be able to build more vehicles after this project, so let's see if we can get something a bit slower.

At the advice of Shane, I wrapped the stator to see how many wraps I could reasonably fit onto one tooth.

AutoCAD predicts 95 wraps will fit if I wrap really tightly and ensure that each wire fits neatly in the valley of the two under it.

Realistically, I could fit 72 wraps of single strand 22 guage wire around one tooth, leaving room for some airflow and allowing for error in other teeth's wraps.

If we plug in 72 as N in our favorite equation, we get Kt = .248 Nm/A, which is right where it should be. According to Shane, most electric kick scooters have a Kt of between .2 and .3 Nm/A. The higher side of this range will give a scooter more torque, but go slower, and a Kt on the lower side will go faster, but accelerate slower.

Kt = .248 Nm/A


5.55 Nm / .248 Nm/A = 22.4 A


33 v/ .248 v/rad/s) = 133 rad/s

Going the long way again:

130 rad/s * 1 rev/2 pi radians * 60 s/1 min * 60 min/1 hr * 2 pi (3 in)/1 rev * 1 ft/12 in * 1 mile/5280 ft = about 22 mph

This is about right. Fast enough to actually be used as transportation somewhere, and torquey enough to get me up to speed quickly and scare other people who want to try my little toy.

Hopefully I'll wrap the stator sometime in the next two weeks, and then get to the construction of the body, battery, and turning of the rotor can and spacers.

Let the scooter building season begin...

Scooter Battery Replacement

The original Deathscooter battery was kindly made for the Engineering Design Workshop by Shane Colton. Unfortunately, Shane was pressed for time, so he accidentally wired the battery backwards. We discovered this after we connected the motor controller to the battery and the connection exploded. The wires' colors were 'reversed' via heatshrink tubing, and two new deans connectors were soldered to where the old ones used to be. Oh, and the battery didn't have any balance wires either. This battery should not exist. And so in August I built a new battery pack (same specs: A123 8S2P 4400 mAh 26.4V) to replace the old one. On that fine summer day, I did everything except attach the connectors, which is, you know, kind of important.

2.5 months later, I finally got around to finishing the pack. The reason I hadn't done this sooner is that the battery pack that Shane built hadn't exploded yet, and was still performing well despite my extensive usage sometimes requiring multiple ~70% capacity charges per day. The reason I got around to doing it is the darn thing was sitting in N52 almost fully assembled, and I should really be using the pack with balance wires.

Alright here we go. The first step is to put these little tiny connectors on each of the balance leads and put them in a plastic holder so it can be plugged into the charger. A note for anyone trying to build their own pack: do NOT try to strip, solder, or do anything else to two balance leads at the same time. If you try to cut or strip them both at the same time, they will short together on whichever metal cutting tool you have chosen for the task. You and your pack will not be happy. Strip one wire at a time, solder the connecter, crimp, place in the holder. Repeat as necessary, in that order.

Soldered the deans connector onto the main wires. Pack is operational.

- black white brown backwards rainbow +

And we have our fully operational balanced pack!

Looks like we're done for the day. NOPE.

The pack doesn't fit inside the battery bay on the scooter because the wire sticks out the top instead of the side on part of the pack. The segment of wires in front of the conveniently (and accidentally) placed cinnamon toast crunch box needs to be moved from the top to the side.

Carefully slicing open the pack, I am terrified of cutting a wire or worse, shorting two wires with my knife blade. Luckily neither of those things happened.

Black wire was pushed over to the side, and the pack is ready to be sealed up. Wait...where's my duct tape? Crap. Well, floor mate Julia to the rescue... 

Probably the most colorful pack I will ever make.

It fits!

And so Deathscooter was made a bit more colorful. 

Note: I will still not ride this machine until I get a new front wheel, which is in the mail as of tomorrow (hopefully).