2015 Formula Student Germany

[Z]

Thijs,

I repost my Launch-V-T graph below to better explain this.

My main point is that Zurich's Acceleration runs must have been much closer to my curve-B (= Team Toothless ) than to curves C, or C' (= Team Testosterone).

So all the excuses that several generations of students have been making, namely that low 3 second times are impossible without special dragster tyres, specially stickyed-up dragstrips, and super-mega-powerful turbo-engines, are nonsense. (I remember many such "It's impossible!!!" excuses when I insisted that low-3's were feasible back in 2005.)



Based on the 0.3 metre jump start you mention (= "infinite" acceleration at T=0), I would say that Zurich's V-T curve starts off above the left-side of my curve-B, and probably stays above it for the first ~half-second (ie. the first 5 to 10 metres). Zurich's curve will then track reasonably close to curve-B up to about T = 1.5 seconds, then dip below it to T = 2.5 seconds, then above it again for its slightly faster top speed.

Put simply, Zurich's curve will be a rounded-off version (plus DAQ noise) of my curve, which was drawn as a number of straightline segments to simplify things. Importantly, any "area" that is under Zurich's curve but is "outside" my curve-B MUST be matched by an equal area being lost inside my curve-B. This is because the total area under each curve must equal the total run length of 75 metres. (This assumes FSG's Acceleration track was truly 75 metres long. There have been athletic events chasing World Records that have, ahem..., favourably adjusted their track lengths... But I doubt FSG would do that? ).

Anyway, most important message is that any good RWD-car, E or C, should be able to do similar low 3 second times. Or better. No magical super-sticky dragster tyres are needed.

So from now on, any Acceleration time of 4 seconds or more = LOSER!

NO MORE EXCUSES!!!

Z

[Thijs]

Z,

I still feel you're brushing over the differences in the graph I posted below a little easily.
Of course this is an estimate, I'm not part of the Zürich team, but I've been with the Delft team long enough to know how electric 4WD 80kW cars move about.

We are of course in agreement that low 3's are possible, so yay I'm still not sure how you would do that with a combustion car though. Zürich showed what you can do on this surface with almost perfect controls (just the right tiny amount of wheelspin).
Delft showed that even with a car with slightly better specs (bit less weight, bit higher top speed), you can lose 3 tenths if your controllers don't work well (their wheels spun quite a bit more).
How would a C-car ever match the smooth torque delivery of individually controlled electric motors?
With higher R%, I'm sure they can get a bit closer to 1.5g at the start, rather than 1g where most of them are now, but try to accelerate at 2.5g, and they'll still end up like every car in this pretty awesome video (that unfortunately doesn't show the Zurich getaway...):
https://www.youtube.com/watch?v=fVrC565ynzQ

Anyway, let's see if anyone at FSA can go below 3.3!

Thijs

[JulianH]

I asked this year's team about the data. Apparently it is on a USB-Drive somewhere
As soon as I get it, I will post it here to hopefully end the discussion about curves

Cheers

[Thijs]

hmmm curves

would be nice Julian

[BeunMan]

Question from someone not knowing a lot about 'mechanics': What are the power losses in friction, heat buildup and wheel slip and what is the actual transfer (%?) to the ground (resulting in forward momentum)?

Also, air resistance and tire friction are significant at higher speeds, do they matter in any way (even without wings) to the end result?

[tromoly]

Thinking out loud, others elsewhere in the forum have said that an acceptable Mu for simple tire calcs is about 1.5, if that's the case then by using F = Mu * W and F = m * a (with a couple g_c thrown in for unit conversion) would mean that even with 100% of vehicle weight on the driving tires the maximum acceleration would be 1.5G, neglecting any additional aero download onto the tires of course.

Side note, about this time last year an acceleration discussion started in an event thread and there was talk of spinning it off to its own thread but never happened, anyone else think a dedicated acceleration thread should be created?

[JulianH]

Hey guys,

I just got the data from Zurich. It's their fourth Acceleration Run at FSG.

They crossed the starting line roughly 0.22s after the real start (with a speed of ~11kph) and crossed the finishline after 3.52s with a topspeed of 111kph (the FSG speed trap seems to be quite accurate!).

Have fun discussing it

Cheers
Julian

[Thijs]

Thanks Julian,

I had expected the car to pick up the acceleration rate a bit, making use of wings, as it approached its power limit. The fact that the car still got to 3.3s without doing that probably means I overestimated drivetrain losses.
It still looks very close to my estimate though. I predicted crossing the starting line at 'almost 3m/s', or 10.8km/h (it's apparently 10.66km/h in real life) and crossing the finish line at 3.49s.
The acceleration rate looks almost constant up to about 66km/h and averages 1.42g, which is what my graph started out at, and which supports my main point that 2.5g off the line is not how Zürich got to 3.3s.

Either way: impressive stuff from the guys from Switzerland!

Thijs

[Z]

Julian,

Thanks for your above post with data. Haven't had time to look at it closely now, but I guess it confirms what I have written below. Will comment again later.

The following is a rant I wrote earlier today, just before my "credit expired" so I couldn't post it, and I had to go and do other stuff...
~o0o~

Thijs,

Firstly, you have many, many years of hard grumpiness training ahead of you before you can rightly call yourself a "Grumpy Old Fart". I suggest running naked through the blackberry bushes every morning for the next, ooh..., several decades, just to properly develop your grumpiness skills...

Back to Acceleration Event.

... By saying that your particular curve must be pretty close [to Zurich's], you're implying that your assumption of 2.5g off the line is also right (a key assumption in your thesis of how to get to sub 3.0s acceleration runs for RWD cars). And it's not. I'm not saying it's not possible in principle (although I also doubt that), ...

Actually, my key assumptions are given in the boxes at the right of my "Launch V-T" sketch (above), together with the assumption that the distance covered is 75 metres, indicated by the equal 75 m area under all curves in the sketch. The initial 2.5 G acceleration is somewhat incidental to the story (more below).

More importantly, my key AIM of that sketch was to show the usefulness of Velocity-Time curves for understanding these things.

The most important point here is that the AREA under the curve MUST ALWAYS be the constant 75 metres. On the other hand, the shape of the curve, and hence the car's Velocity at any given Time, can vary quite a bit, but ONLY as long as the "area" constraint is met. And momentary Accelerations (= slope of curve) are the most fickle, and can vary hugely without having a great impact on the final result. To see this, imagine a lot of "noise" on the curve, which can give huge variations in Acceleration, but only smaller variations in Velocity, and perhaps NO variation in Area under the curve (= Distance).

This usefulness of V-T curves, and the perhaps misleading nature of Acceleration-Time curves, can be seen in the A-T curve you posted (and thanks for that). To get useful information out of the A-T curve you must integrate it once to get Velocity, then integrate again for the Distance covered. I can see that your two A-T curves, red and blue, have the same area under them, so they suggest (misleadingly *) the same final Velocity (= 30 m/s). But trying to "see" the Distance covered at any time with an unaided eye is much harder, perhaps impossible.

(*) Worst problem is that a "flying start" appears in a A-T curve as an infinitely high Acceleration at the left of the curve, that lasts for dT = zero seconds. But that "zero-width spike" has FINITE AREA under it! See below for an example of how the integral of Infinite-Acceleration x Zero-Time = Finite-Velocity. Or look-up "Dirac Delta function". There are many examples like this in Mechanics, so beware.

Bottom line here, A-T curves can bounce around all over the place, but these large variations can make very little actual difference to the car's Velocity or Distance on track.

As for the "2.5 G off the line" (or more!), well, that is not my opinion, it is a much repeated historical fact. Please re-read what I wrote about launch techniques (specifically about pro-lift, and unsteady aero-downforce...), which was the most important point of my previous posts... (more on this below).

... I think you're also not taking into account that cars are allowed to start 0.3m before the start line. This makes a massive difference (the car crosses the start line already going almost 3m/s)...

Yes, it does make a big difference. It makes it much easier to get a low time!

If my sketch's curve-B had that 0.3 metre "flying start", then it would need ONLY ~1.5 G initial acceleration up until it was 60 kW power limited (ie. at T = ~1+ sec), and yet its final ET would be LESS than 3.3 seconds. This because its initial "acceleration" at the start of the timed section would be "infinite", but only for zero dT (ie. at T = 0 the curve rises vertically to an initial velocity of V = 3 m/s, after which it slopes up-to-right at a much more gentle 15 m/s per second).
~o0o~

... I'm still not sure how you would do [low-3s] with a combustion car though...
Zürich showed ... perfect controls (just the right tiny amount of wheelspin).
Delft showed ... you can lose 3 tenths if your controllers don't work well (their wheels spun quite a bit more).
How would a C-car ever match the smooth torque delivery of individually controlled electric motors?
... they'll still end up like every car in this pretty awesome video ...
https://www.youtube.com/watch?v=fVrC565ynzQ

Tristan also suggests that "You need a really good launch control to get sub 3.5...

And Tromoly also suggests that "... an acceptable Mu for simple tire calcs is about 1.5, if that's the case then ... the maximum [startline] acceleration would be 1.5G... "

Now, Thijs, thank you for that YouTube link, ... because it shows how PATHETICALLY PISS-POOR ALL FS-Teams' cars are! (See, that's how you do Grumpy! )

All your cars look like the V8-utes (ie. with ~30%R) that young Australian country-boys have fun with on a muddy paddock after the local B&S Ball. (Cultural note: B&S Balls are a rite-of-passage for rural Australian youth, with the B&S standing for Bachelor & Spinster, ... or maybe Beer & Sex, followed by circle-work with the utes...)

All your FS-cars most obviously have ~NO GRIP, because they show ~NO FORWARD PROGRESS!!! Geez, one car even shows large amounts of rear-wheel floppy toe-wobble as its feeble Fx forces kick in (see at 1:08)! Mutter, grumble...

So, ... get yourselves down to the local dragstrip and learn how to do this properly. The Toothless Hillbillies there are really very friendly. and they will help you. They will tell you to STOP WASTING YOUR TIME talking about "controlling" what you currently DO NOT HAVE.

FIRST OF ALL YOU NEED THE "MECHANICAL GRUNT" to get your cars moving quickly, and only much later will you have to worry about "controlling" it.

So, first make sure that ALL OF THE CAR'S Fz IS ON THE DRIVING WHEELS (ie. NOT your current "dynamic-R%" = ~60%!).

It is easy to see when you have this, because the front of your car MUST JUMP UP OFF THE GROUND when "coming out of the hole".

Your rear-tyres' sidewalls MUST wrinkle as the car's rearward-weight-transfer, plus the PRO-LIFT, plus the unsteady-aero, pushes the driven rear-wheels DOWN INTO THE GROUND.

In short, your cars should LEAP across the start-line, like a startled antelope...

And during the extensive Launch-testing you should be doing, you should have numerous drivetrain, suspension, etc., FAILURES due to the awesome forces involved.

And then, after you have strengthened all those weaknesses, maybe more than ~35% of you can also finish Enduro?

Anyway, let's see if anyone at FSA can go below 3.3!

Yes!!! Oh-so easy, and FUN!

Z

[JulianH]

Thijs, since we got a 4WD, the wings are always in minimum drag position and do not produce a massive amount of downforce.
I saw that Delft started the run with a closed DRS and opened it up after roughly one second. We tried that last year and the effect was not really there. (But you risk that your DRS is not working correctly ).

Back in 2012 with the 2WD car, we drove with a closed rear wing which gave us the mentioned effect. Since you now need the grip on the front axle, a closed rear wing hurts more than it helps.