Performance Electronics Traction Control and Launch Control Help

[NHPhil]

Hey fellow engineers,

Our team switched over to a PE3 ECU this year (which has been a fantastic decision at every step). However, we are pretty confused where to start with setting up traction and launch control. A couple specific questions and some open-ended ones below:

1. What types of switches do your teams use for the TC Arm and LC Arm/Start buttons? We aren't sure if toggle or momentary are appropriate.
2. Do you use TC and or LC in any of the FSAE events? If you're willing to share, which ones?
3. Tips/suggestions/good resources (not necessarily PE specific) for tuning traction control tables?
4. How much do you use these tools during test driving?

Would love any help you can provide.
Phil

[Z]

Hi Phil,

Welcome to the Forum.

I have negligible knowledge of electronic T/L-control, but since your post above was looking a little lonely I thought I might get the ball rolling.
~~~o0o~~~

TRACTION/LAUNCH CONTROL DONE CHEAP (x3).
================================
1. Select a higher gear.

Since your car probably has a 5 or 6 speed gearbox, you can simply select the next gear up (ie. with lower numerical ratio). This sends less torque to the rear-axle, which means the driver can stomp on the skinny-pedal whenever they want and ... NO WHEELSPIN!
~o0o~

2. Fit smaller, cheaper engine.

Similar results to above, but cheaper and lighter, so better!
~o0o~

3. MORE REAR%!!!

It has always baffled me why so many FSAEers fit a high-powered engine, typically a 600-four, then turbo it for even more power, and then add a great big rat's-nest of electronics to TURN OFF THE POWER ... just when they most need it.

So, build your car with 60 - 65R%. Then, whenever you stomp on the skinny-pedal, it just goes forward, FAST. No blue smoke from rear-tyres. No wasted fuel. Just high forward acceleration.

So easy!

Z

[Charles Kaneb]

1) "Select a higher gear" works only by reducing the force available at the rear tires. You will only be faster that way if there is no effective throttle modulation, and you have the choice between (say) 350# forward force and 450# forward force at tires that would take a 380# forward force at the peak of the force vs slip angle at your load.
It's much better to use the throttle or clutch to carefully manage the amount of force applied at the tires by changing the speed; a slip-percentage-sensitive tire may lose 3% of forward force for 10% of extra slip beyond the peak.
There is another reason the drag-racers use what appear to be extremely short gears - 4.11 or 4.56 even when there's enough power (torque*alpha) to get to the force vs slip percentage peak off the line with 2.73s. A 4.56:1 rear end and a 2.2:1 1st gear will have the crankshaft rotating ten times for one turn of the tires. If the tires start to "go up in smoke" past the slip percentage giving peak force, the engine speed will need to increase. With the same effective flywheel-and-crankshaft inertia in both cases, the shorter-geared car will go past the peak by less per tenth of a second during the spin-up, keeping you closer to the peak force while you back off.

2) A short first gear and heavy flywheel may give you enough stored rotational kinetic energy to get a very strong launch with an underpowered engine. The calculations for the ideal flywheel mass for the acceleration event and the ideal flywheel mass for the straightaways of the endurance event should give different results, though...

3) The best F-R static weight distributions for the acceleration event and the skidpad event are different. If you've got a good tire model, and "adequate" power, you can figure out whether it's worth mounting ballast to change your center of mass location, and where.

I would say "RTFM" to find the documentation for the PE3 traction control, but I just went on the Performance Electronics website, downloaded the PE3 manual, flipped to the index, went down to the line for "Traction Control", and found a blank space where the page number was supposed to be. I'd suggest calling them tomorrow during business hours in Cincinnati, Ohio, USA.

[Z]

Charles,

My above suggestions for cheap ways of doing T/L-Control are all intended for its use, primarily, in the AutoX and Enduro events. These are the events with highest scores, and these scores are most adversely affected by uncontrolled wheelspin. Namely, wheelspin coming out of a low speed corner => means whole car spins-out and collects N x cones => means many lost points.
~o0o~

"1) "Select a higher gear" ...
It's much better to use the throttle or clutch to carefully manage the amount of force applied at the tires by changing the speed..."

Most FSAE drivers have very little training, and such throttle/clutch "modulation" is a step too far for them. Even harder towards the end of an Enduro, after spending a long time wrestling with the badly designed "Armstrong" steering.

It is much easier to drive through the low-speed corners in a slightly "highish" gear, and not worry about the rear tyres breaking free. Also less risk of cock-ups due to a missed, or wrong-way, gear-shift, which can result in DNF.
~o0o~

"2) A short first gear and heavy flywheel may give you enough stored rotational kinetic energy to get a very strong launch [in Accel-Event] with an underpowered engine."

Yep!

"The calculations for the ideal flywheel mass for the acceleration ... and endurance event should give different results..."

Not really. Not if you aim to build a lightweight, low-power, high-aero-DF car. Such a car would be close to the ideal "constant speed car" that runs at a fairly steady 55+ kph on all parts of the track. Sure, it would go a bit faster on the longer straights, and a bit slower through the hairpins. But overall it could have the fastest laptime, while also having the slowest top-speed. So awesome mid-straight acceleration in Enduro is not really needed.

And the above concept is best suited to win Fuel. And Cost, because of its small, cheap engine.

But ... roll-the-dice in Design!
~o0o~

"3) The best F-R static weight distributions for the acceleration event and the skidpad event are different."

Says who? Not according to my reckoning. Not according to many of Bill's posts from about a year ago (IIRC, on the Static and Dynamic Sections of Forum).

This FSAE-only ideological mantra that says "...it is impossible for a fat-kid and a skinny-kid to balance on a see-saw" is bizarre. Where are the theoretical numbers? Where is the real-world evidence?

Anyway, regardless of whether your car has the lightweight, low-powered, engine above, or an awesome, turboed-tyre-shredding, 600-four, if you build the car with more R%, then it will ACCELERATE FASTER. And WITH LESS WHEELSPIN, even with no T/L-control!

And, if built right, it can also win SkidPad.

Z

[FrederikWe]

In regard to 3)

What are your assumptions when you say if it is "build right" it can also win Skid Pad?

In my understanding for rear wheel driven cars the best layout for aceration would be a ~65% R. Just at the point where the load on the front wheels at max. acceleration would be close to 0, leaving just enough to make minor steering corections.

Going to Skid Pad with such a car the lateral performance potential of the rear axle would be lower because of the declining friction coefficent of the tyres resulting in oversteer.
Even with a adapted ARB setup you could maybe get a almost neutral behaviour, but the total lateral performance must be lower than in a comparable car with a 50:50 layout.
Am I missing an important point in this thought?

Freddy

[Flight909]

In regard to 3)

In my understanding for rear wheel driven cars the best layout for aceration would be a ~65% R. Just at the point where the load on the front wheels at max. acceleration would be close to 0, leaving just enough to make minor steering corections.

Look at very old pictures from FSUK - I remember seeing pictures of teams adjusting CGH to make better acceleration. Make high ride height will help to put more distribution on rear in longitudinal acceleration. I don't know if anybody did this since?

[MCoach]

Looks like everyone came in here to pay the troll toll...
More rear % is really just missing the point, don't let him distract you.
There are better ways to go about accelerating a car.


Definitely RTFM. Emailing the company could help too.

I think the switches would come down to a personal choice and to just pick one and go and if it doesn't work out, switch it out for the other one.

In terms of tuning, test test test test test. Test on different surfaces too. Test on the closest surface and conditions you can get to the track you will perform on first.
Have multiple options. Have traction control maps for different surfaces. Low mu / wet, asphalt, concrete...if you have enough time and tires to do so.

[Z]

What are your assumptions when you say if it is "build right" it can also win Skid Pad?
... declining friction coefficent ... resulting in oversteer.
... adapted ARB setup ... neutral behaviour, but ... performance must be lower than ... comparable car with a 50:50 layout.
Am I missing an important point in this thought?

Frederik,

Yes, "important point" is highlighted above, and repeated at bottom of this post.

Two ways to look at this:
~~~o0o~~~

THEORETICAL.
============
This approach might start with consideration of "Tyre-Load-Sensitivity". Broadly speaking, this says that the lower the contact pressure between road and tyre-rubber, then the higher the "Coefficient of Friction", or "Mu", between road and rubber.

Now, some (many? most?) FS/FSAE-teams might try to "optimise" this TLS behaviour in order to build the "optimal" car. Unfortunately, such teams NEVER finish the car! This simply because to achieve the "optimum" TLS they must have either infinitely large tyres, or else a zero-mass car. Or both. Clearly, a never ending design cycle follows...

But, more importantly, winning FS is NOT JUST about TLS.

So, let's forget about "optimalisation", and just aim to build a car that is significantly faster than all other current FS cars. Let's start by aiming for a car with significantly lower mass than the competition. So NOT a "comparable" car. Say we aim for, ooohh, 120 kg, or less? So even with a heavy-ish driver the total mass is under 200 kg.

(FWIW, I would do such a car as "all-steel" (err, + some plywood), with NO CF/titanium, etc. But mainly NO UNNECESSARY JUNK. So small air-cooled single, two-speed-drivetrain at most, NO starter-motor + hi-amp-battery, NO clutch, etc...)

Next, let's put some big, sticky tyres on the car. Nowadays a suitable choice might be the 8" wide, 10" rim-diameter, Hoosiers. Let's assume, in round numbers, a R% = 60%. So we have less than 60 kg load on each rear-tyre, and less than 40 kg on each front-tyre.

Importantly, we now do compare the above loads to the opposition (just to see how UN-comparable our car is). Most other top cars are at least ~40 kg heavier. So even if these other cars, at ~240 kg total with same mass driver, have 50:50 mass-distribution, they will have the same 60 kg load on all their tyres, as we have on our rear-tyres. So any TLS effect is the SAME on all their tyres, as it is on our rear-tyres. So no gain or loss from TLS in terms of the ultimate grip level.

But, as you noted, the TLS effect means our more lightly loaded front-tyres have even more grip. So...
~o0o~

BUT-1 (!) What about "handling balance"?

You have already given the answer to this, namely, just lift the inside-front-wheel during hard cornering. Note that with 66.6...%R (= 2/3), and inside-front lifted OFF the ground, all three wheels still on the ground are equally loaded (at 1/3 total-Fz each)..., so handling is "balanced".

An even better solution is to fit SMALLER front-tyres, and keep both of them ON the ground. Say, the 6" wide Hoosiers.

Ahh, the car becomes even lighter!
~o0o~

BUT-2 (!) There is still AERO to consider!!!

I leave the detailed calcs to you, but I am sure you will agree that with the same aero-downforce, a lighter car is always faster around corners than a heavier car. So better in SP, AutoX, Enduro.
~~~o0o~~~

PRACTICAL.
==========
Or you can study the prior-art.

The designers of the cars below chose very high R% because of their very powerful engines. Hence all these cars have neck-snapping forwards acceleration. They also go very fast around corners. With balanced handling!

(I cut & pasted these pics from another of my posts from a few years ago. Edit: Some pics now gone.)

Typical F1 Turbo era car. Renault started it, but this first 1977(?) car wasn't too successful, and had nowhere near the 1,000++ hp of the later turbo cars.


F1 Tyrrell P-34 "six-wheeler" (1976/7) Moderately successful (1 win, several other podiums...). Design was mainly to reduce aero-drag by tucking front wheels into bodywork. Williams and March did 4R:2F-tyred cars, and Ferrari did 2 x tyres on each rear corner, similar to the rear-engined (~600 hp), Auto Unions of the skinny-tyred 1930s era.


Yep, fat-kids and skinny-kids CAN BALANCE on the see-saw!
~o0o~

BOTTOM-LINE(s).
=============
The current FS/FSAE practice of "optimalisation" via "sensitivity-analysis" does NOT get you off the top of a small hill. Or, "You cannot cross a chasm by taking small steps."

If you start with a 50:50 car, and then consider the TLS effects on the handling balance of comparable mass cars with comparable tyres, then you are pretty much stuck forever on top of the same small hill you started at. And anything more than ~40 hp just produces much blue smoke from your rear-tyres, with little forward progress.

However, if you really want to blow the opposition away, really push the envelope, then you have to "search the entire solution-space", even the places on the other side of that chasm. This way you can find the car that scores highest points OVERALL. And this may be a car that occasionally lifts its inside-front-wheel off the ground. Or not, depending on tyre choice.

Z

[BillCobb]

In regard to 3)

What are your assumptions when you say if it is "build right" it can also win Skid Pad?

In my understanding for rear wheel driven cars the best layout for acceration would be a ~65% R. Just at the point where the load on the front wheels at max. acceleration would be close to 0, leaving just enough to make minor steering corections.

Going to Skid Pad with such a car the lateral performance potential of the rear axle would be lower because of the declining friction coefficent of the tyres resulting in oversteer.
Even with a adapted ARB setup you could maybe get a almost neutral behaviour, but the total lateral performance must be lower than in a comparable car with a 50:50 layout.
Am I missing an important point in this thought?

Freddy

Sorry to jump in late in the show, but the problem(s) discussed here are the direct result of building a car from a lame recipe and then asking it to perform tasks that appear to be in direct conflict with each other. This is not a Catch-22 situation, though.

Instead, a car meeting both of these X and Y max accelerations is easily synthesized from requirements which are flowed down to subsystem and then to hardware requirements. If you would care to look, almost all very high performance production cars are designed by this process and the hardware set that achieves their goals does not include soft compliant front steering systems, bizarre TLLTDs, absurd roll steer and camber kinematics or 3 wheeled tire contact.

Try designing a car instead of opening your Erector Set (Mechano to some of you) box and grabbing up all the nuts and bolts and wheels and pulleys. If you still can't figure it out, find a different book to read.

[MCoach]

These discussions really belong in a different thread.
They miss the point the author was trying to reach.