building a traction control system

[mopoet178]

I am a senior Mechanical Engineering student at CU Boulder, currently enrolled in System Dynamics. I was thinking for the project this semester my group could design and build a rudimentary traction control system for my FZR.

I am not looking to do anything fancy with gyros and ABS, just wheel speed sensors and an ignition retard. I hope this project will continue to inspire me as a design engineer and teach more about motorcycles and different solutions to the problem of going fast.

Does anyone know any good literature regarding motorcycle traction control theory? There's a lot I don't know about this stuff: what percentage wheel slip relative to the front wheel speed to allow, how smooth/abrupt to program the ignition response in order not to upset the bike chassis mid turn, do I need a steering head angle sensor, what kind of wheel sensors do I need, how many times per second I need to collect wheel speed data, stuff like that...

I have lots of spare parts and electronics and parts to make a good test module, all kinds of good data acquisition equipment at my disposal, and two other good engineering minds working with me. We can build some pretty kick-ass electronics, and machine anything we need practically. I think we can pull this off.

Ultimately I would like to mount all this on my 1989 FZR600 and actively test and use it at the track. This will, however, depend on my personal confidence in our work since it's my butt on the line.

Any guidance or constructive criticism is welcome. Like I said, the biggest obstacle right now for me is information. I don't fully understand the problem yet, so I need some good books to fill me in on what really causes wheel spin and what steps to take to control it. Now is the time for research.

Any help appreciated. If you really know what you are talking about I'd probably like to pick your brain over a brew.

-Marsh

[rybo]

Marsh,

For rudimentary traction control maybe a better approach would be a crank speed sensor.

The front and rear wheel speed algorithm gets really complicated in a hurry since the expected difference could vary greatly depending on lean angle, if the bike is in a position with the front or rear wheel is off the ground etc. This is the reason the advanced systems use gyros.

Some of the basic TC systems that I'm aware of being commercially available use the crank position sensor to count revolutions (or fractions thereof) and then have an algorithm of expected rate of crank acceleration vs. actual rate of crank acceleration to determine when to either reduce throttle (fly by wire bikes) or retard ignition to get the rear tire back to a grip state.

This is the extent to which I can speak knowledgably on this subject....but should provide some food for thought.

[Matty]

Marsh,

For rudimentary traction control maybe a better approach would be a crank speed sensor.

The front and rear wheel speed algorithm gets really complicated in a hurry since the expected difference could vary greatly depending on lean angle, if the bike is in a position with the front or rear wheel is off the ground etc. This is the reason the advanced systems use gyros.

Some of the basic TC systems that I'm aware of being commercially available use the crank position sensor to count revolutions (or fractions thereof) and then have an algorithm of expected rate of crank acceleration vs. actual rate of crank acceleration to determine when to either reduce throttle (fly by wire bikes) or retard ignition to get the rear tire back to a grip state.

This is the extent to which I can speak knowledgably on this subject....but should provide some food for thought.

that just hurt my brain....

[CYCLE_MONKEY]

To go one further, and probably more accurate, how about strain gauges on the wheel spokes? If you have strain, you're accelerating or decellerating. If you're in either of those modes, and suddenly the strain lessens or goes neutral, the wheel is slipping, and you can program the TC to intervene. This would eliminate any error from the tire changing effective circumference as the bike leans.

Or instead of strain gauges, how about a very precise and finely calibrated speed sensor on the wheel hub, and then another on the rim, and synch them. That way, if you start seeing speed errors between them, you've got "strain", and can use it like the above example. It would prevent the need for slip rings to get the strain gauge sensor data from the wheel to the ECU while it's rolling.

The only form of "TC" I'd like to see is the throttle controlling the actual engine RPM like a rheostat with the ECU controlling the throttle position, timing, etc. to achieve that. NOT the throttle plate controling speed balancing it against a load as has always been the case. Why? Because, if the load on the engine lessens with the same throttle opening, the engine speeds up. With the throttle controlling actual engine RPM, that wouldn't happen. If you were a ham-fisted rider and whacked the throttle open, it could still spit you off, unlike a true TC system, yet it would comensate for a sudden decrease in traction mid-corner.

[mopoet178]

Good info, Rybo. This is the kind of stuff I need. I know it's gonna get really math intensive, and I kind of look forward to it. This kind of stuff is why i study engineering.

I know I cant get all the info I need on a public online forum, so what I am really trying to get out of this post is a starter list of good sources. Who knows some good books, good websites, good knowledgeable people to talk to that will aid me in designing and implementing my own TC?

Keep it coming, guys!

-Marsh

[OUTLAWD]

it would be pretty slick to make the fizzer throttle by wire while you are at it

then you could have kind of a stand alone TC setup

[irdave]

Yeah, I think all the actual "real" stuff is rate of change of the crank. Wheel speed sensors are seen a little on street stuff, but not actual race stuff.

Brewer has one of my books on single track vehicle dynamics right now- but it has a rather large section on data (Tony Foale's book- but it's currently out of print- I think you can get it in .pdf format.) You're in Boulder- get in touch with Claude Rouelle at OptimumG. He should (used to) do a bunch of work with Formula SAE and GO TO HIS SEMINAR ON VEHICLE DYNAMICS!!!! As a student it's crazy cheap and sooo worth it. But all they do is data and simulation.

On another note, I looked into designing and building a motorcycle frame as my senior design project (ME) and had 10 people lined up to work on it- and funding. School said no way... after they said,"Is there any engineering in that?" Really. Asshats. Hope your project goes better.