Magnetohydrodynamic Propulsion Drive

Cool name, right? Magnetohydrodynamic drives use a crossed electric and magnetic field to propel ions, such as salt water ions. That simple. Think the caterpillar drive from the Hunt for the Red October, or the Japanese boat, "Yamoto". I first came accross this idea at "Evil Mad Scientist Laboratories" 3 years ago. I found this article shortly afterwards and decided to build one! I apologize for the cruddy pictures-it's a cell phone camera.

The gray strips on the inside are aluminum electrodes. The big bricks on the left and right side are 1" x 1/2" x 4" NIB magnets. The magnet field on the inside maxed out my sensor, which goes up to 1000 Gauss. I'm guessing the inner gap is around 1T, considering the size of these things. The middle box is made of fiberglass; it needs to be to hold the magnets apart.

The black coating is "Plasti-Dip", possibly one of my favorite things. It's a liquid rubber that drys to form a rubberized coating on pretty much any surface. It's used by R/C hobbyists to waterproof electronics, such as servos and receivers. I used it to coat the MHD drive.

It comes in both a can and a spray. See the big spill on the can? Funny story...so I bought this thinking it would be easy to just dip the magnets in one at a time and get a nice coat on them. Turns out the can is steel. The magnet just jumped into the can and made a huge mess- my hands were black for about a week. (That's when I bought the spray.)

My setup was a tub of salt water and a 13.8V 25A DC power supply. Here are some videos of it in action:



The effects of electrolysis are pretty obvious; I can't think of any way to counter this. The aluminum electrodes wear out pretty quickly. Reversing the electric field (and thus the flow) every now and then seems to help with electrode wear. Titanium or platinum electrodes would probably help a lot, but are also expensive. It was a cool project. Future work will include:
1. Incorporating a similar (probably bigger/longer magnets) into a custom R/C boat or submarine.
2. Using onboard power. I'm thinking LiFePO4 power cells- the high discharge rate would be very beneficial. I think I'll stay away from LiPo's this time. Considering all the H2 and O2 produced, I think there will be enough flammable stuff around.
3. Experiment with ways to reduce the electrolysis and increase electrode life.

Learning Lessons

Well, after over 4 months since ordering, the remaining 3 stators are here. And what did I get for my trouble? Express shipping to the wrong address and a 15% coupon. Wow, gee, thanks...

Yeah, I'm pissed off. The express shipping meant that they couldn't change the address before it was delivered. Now I might not be able to get them for another 3 months. Anyone know what MIT does with mail sent to closed dorm addresses over the summer?

Needless to say I am most dissatisfied with GoBrushless. One of the most incompetent merchants I've ever dealt with. However, it wasn't completely their fault because their supplier in China is mostly to blame. Though the shipping error was completely their fault.

Anyways, I'll get them eventually. I'm glad I didn't have to do a total redesign, or cut up a motor ((like one of these) for a stator, the latter of which would have resulted in at least a slight redesign.This experience taught me the value of flexibility; unfortunately, brushless motors aren't very flexible. I still should have designed for more available parts (stators); GoBrushless is currently the only source for 65mm stators. Believe me, I know. I've spent hours and hours scouring Google for another source- nothing.

In summary, here are the lessons I've learned:
1. Always ask about stock status of a part before ordering.
2. Stators (or anything) from China could take many months.
3. Design for available parts/design for flexibility.
4. For what you pay for custom USA parts, you get back in time. For example, if I had gone with ProtoLam, I could have designed exactly the stator I wanted and I probably would have had all four of them in less than a month, but it probably would have cost four times as much. (There's also a quality advantage to USA parts, but that's a different story.) However, in the real engineering world, the net cost of waiting four times as long would have outweighed the higher production costs.
5. Repeatedly to the point of exhaustion remind the merchant of the correct shipping address.

I found out something else, too. These stators, and some other ones GoBrushless sells, are the same (as in from same supplier) as the ones in Scorpion motors. However, Scorpion doesn't officially have a 65mm motor out yet, though apparently that's coming soon along with 85mm ones. Scorpion is also developing wind generator motors, so they could potentially be a source for large stators. I bet they'd sell you stators directly if you contacted them.

Finally

I HAVE RECEIVED WORD THAT THE STATORS ARE IN THE USA!!!!

That is all.

ELB mid/late May

The first thing you should notice (well, at least if you go to MIT) is that mid/late May is when I should be studying for finals. Hmmm, evil finals or fun machine work... Finals won out eventually, but until then, I put off studying as long as possible to finish these hubs.

Now on to the mill work. I needed to put a round groove into the press-fit surface that would accept a 3mm rollpin. So I used a ball end mill. I had no idea people made these things; they're pretty cool.

I don't have pictures, but during this process I did a tool change on every part in order to drill and power-tap 10-32 grubs screw holes for axle locking. The next step was to flip the part upright and bore the four wire holes. This required an extra long drill because the regular length drill couldn't get the last 3/4" or so.

 Test gauge used to zero the X and Y axis. Mark has a bunch of these.

Really long drill bit. I'm not sure why it looks like I'm drilling into the vice here...weird camera angle I guess.

 I got two finished this day before the shop closed.

The 4 completed hubs. Note the two shorter ones for the rear.

Now I just need to press fit them. Unfortunately, I didn't have all four stators in (grrr...I still don't) yet, so I only did one. First, I had to machine a little tube that would provide a good surface for pressing. You can see it underneath the stator in this picture:

Ta-Da! One finished. 

You can see the rollpin in place in the above picture. That little bugger was hard to get in...it kept bending. Anyways, that's it for now. I'm taking June off, but I'll be back to work on this in July as a side project to teaching high school kids about engineering through an Edgerton Center class. I'm hoping to get the whole board finished by early August! 

Stuff left to do:

1. YELL AT GoBrushless UNTIL THEY GIVE ME MY OTHER STATORS

2. Wind

3. Solder the motor controller boards

4. Machine the rotors

5. Plastic mill/laser work (hubcaps/spacers/hall effect sensor board)

6. Bore out the tires

7. Assemble a wheel and test it 

8. Mill/waterjet the battery box components, assemble, and test waterproof-ness

9. Find out what Franco did with the radio control

10. Wire everything up

11. Don't kill myself the first time I ride it.

Yikes, that's a lot of work left.

ELB early May

The hubs are the parts that adapt from the axle to the stator. Some people combine the axle and the hub into the same piece (think BWD), but I couldn't do that because longboards have trucks. Trucks are somewhat complicated hunks of metal that allow skate/long/mountain-boards to turn. They also have axles embedded in them meant for non-hub motor wheels, which means I have to have an adapter if I want to use stock trucks. I probably could have made my own, but that would have required a lot of design time and money (large chunks of aluminum are expensive), and I'm a firm believer in "simpler is better" when it comes to engineering. Read the design review for more details on the trucks I used, but for now, here's a picture:

Anyways, back to hubs:

 Front hub design. The rear differs only in length thanks to some stupid brackets welded on the axles I'm using- see the pic of the trucks above.

They are made out of T3-2024 aluminum, which has high shear strength. The hub slips over most of the axle and then threads onto the last half inch. Two grub screws bite into two flat spots ground onto the axle, thus keeping the hub from unscrewing and falling off. The largest diameter surface is a press-fit for the stator, and the groove in that surface is for a rollpin (Ohhh! That's what the round groove in the inner surface of the stators is for...). The thin grooves on the bearing surfaces are for snap rings, i.e. bearing retainers. The four holes that go through the thickest part of the hub are my way of avoiding having to grind a flat spot in the hub under a bearing for wires to route through. The drawback of this plan is that I have to have a massive bearing on the deck-facing side of the motor. Oh well, at least the larger bearing will handle shocks from the horrible Cambridge streets/sidewalks better.

I began by cutting down the stock into pieces slightly longer than the total length of the front hubs. The plan was to make four identical hubs and then cut off some of the end on two of them to make the rear hubs.

 My mess. The second step was to do the boring and threading. Two are finished with that step in this picture.

Mark (the Edgerton Student Shop guy) wanted to use the CNC lathe for the outer surfaces, so I of course agreed (soo much less work and more precise). After about three minutes of CAM, the lathe was ready to go. Check it out:

One of the half finished hubs. S'up Mark.

Four half done! This took maybe 30 minutes. It would have been much quicker if I didn't have to be so careful about not taking too much off of the bearing surface.

Next step was to flip them around and do the other half.

Left: one of the massive 1" x 2" bearings. Top: One hub with the surfaces done.

4 with finished surfaces. You can see the inner threads in this pic.

The next step was cutting snap-ring grooves, one 5/8" and one 1" in each. Mark has an awesome bookcase full of engineering reference guides, and one of them happened to have everything you can imagine about snap-rings in it. He also has some cutting tools specifically designed to cut very very thin grooves.

 So shiny... Note the snap-ring groves.

The next steps were all mill work, and were done later in May.

ELB April

Due to other school projects, I didn't really start building until April. I began by practicing winding the stators.

My winding station. Giant vice with paper for protecting the epoxy. The piece of PVC is how I get leverage.

The stators are GoBrushless 18T, 65mm x 34mm. They have a green epoxy coating for insulation and wire-enamel protection. Unfortunately, that epoxy likes to chip off the tips of the teeth, which means 5-minute to the rescue. So strike two against these stators. What was strike one you ask? The fact that I ordered four of these back in January and STILL haven't gotten the other three in. GRRRRR

It was at this point that I realized that the giant spool of 16AWG wire I bought wasn't going to work because I couldn't fit 15 turns of it on a tooth (I got 14! so close...). So I went to 17AWG:

I tried a bunch of different schemes to get the best possible packing, including that ugly loop sticking out the side. Needless to say I abandoned that idea.

Success! That's 1/2 of 1 phase wound with 15t of 17AWG wire per tooth. These pics were the product of a couple hours of trial and error (including messing around with 16AWG), but now I know how to wind these motors. 

The next step was working on the hubs, probably the most complicated part of this whole project.

Electric Longboard - Pre-blog work

About the project: While the name of this blog is Rocket Science, it was originally created as a build-log for the Electric Longboard (ELB) project. Inspired by the hub motor stylings of the BWD Scooter, one of Shane's builds, and Charles' many builds, the ELB is a 4WD longboard with custom in-wheel hub motors. I came up with the general idea in November and completed the bulk of the design work over IAP, MIT's awesome-we-don't-have-real-classes-in-January period. I turned the project into a personal (and later, a duo with Franco M.) research project through the Edgerton Center.

Check out the updated design review!