PDA

View Full Version : Fastest Racing Line



gsbaker
07-13-2004, 11:21 AM
It's a bit geeky, but the article "Trail Blazing" in the current (August 2004) issue of Racecar Engineering is an interesting look at different racing lines, and which yields the best sector time.

It examines the classic, parabolic, elliptic and hyperbolic lines.

Conclusion? The classic racing line, while offering the highest corner speed, does not yield the lowest time. That honor goes to the parabolic line, which resembles a trail braking entry into a pinched corner with early throttle out. There is no late apexing--the entry and exit lines are mirror images about the apex.

Interesting stuff. Anyone tried this?

------------------
Gregg Baker, P.E.
Isaac, LLC
http://www.isaacdirect.com

Dyno
07-13-2004, 03:07 PM
I'm assuming the "classic" line is a couple of straight lines (the entry and exit) connected tangentially to a constant radius.

If you think about it, we don't (or can't) really drive that way. It would require one to instantaneously move the steeering wheel from the straight-ahead position to the position that defines the arc.

Given that, it's all about how quickly you move the steering wheel away from straight ahead. Slow rotation of the steering wheel while braking is a good description of trail braking.

So, I think we all drive more-or-less "parabolic-like".

gsbaker
07-13-2004, 03:30 PM
Originally posted by Dyno:
I'm assuming the "classic" line is a couple of straight lines (the entry and exit) connected tangentially to a constant radius...

Exactly.


<font face=\"Verdana, Arial\" size=\"2\">So, I think we all drive more-or-less \"parabolic-like\".</font>

I agree, it's only natural. What struck me, however, was just how far this could be taken, almost going to a rain line.

It was suggested that some F1 drivers are doing this as a matter of routine, so I'm going to have to pay closer attention on SpeedTV.

------------------
Gregg Baker, P.E.
Isaac, LLC
http://www.isaacdirect.com

RSTPerformance
07-13-2004, 08:20 PM
Depends possibly on the car??? Maybe this scenario I am about to post would be better as another post but I figure here is as good as any...

At Lime Rock in the last race I ran with a LSD for the first time (at Lime Rock)... I found that "trail braking" was very different and/or difficult especially in "Big Bend." In general I think the fastest way around a turn seemed to be that at turn in you are at full throttle again "pulling" the car through the turn. In big bend this isn't possible as the car pushes way to much at the second apex. Braking so much as that you are full throttle for the entire turn kills your entry speed and will certainly allow any car to pass you in the braking zone of big bend. Big bend is a turn where trail braking seems to be key. I had a very hard time with this as the LSD also caused the car to push drastically in the first half of the turn, especially if you tried trail braking. Many scarry moments, only a couple wheels off though.

How do other front wheel drivers with a Welded diff or LSD handle 2 apex turns where trail braking is key?

Raymond
RST Performance Racing

------------------
http://rstperformance.bizland.com/rstsignature.jpg
RST Performance Racing
www.rstperformance.com (http://www.rstperformance.com)
1st and 2nd 2003 ITB NARRC Championship
1st and 6th 2003 ITB NERRC Championship
3rd 2003 ITB ARRC Sprint Race
4th 2003 ITB ARRC Endoro
1st 2003 AS NERRC and NARRC Championships

Dyno
07-14-2004, 08:28 AM
Originally posted by gsbaker:
It was suggested that some F1 drivers are doing this as a matter of routine, so I'm going to have to pay closer attention on SpeedTV.



I think it would look like this...

Since a parabola has an ever-decreasing radius down to the apex, at which point the radius immediately starts to increase, I think you would never see the steering wheel take a "set". The driver would slowly rotate the steering wheel while driving all the way to the apex and then immediately start to slowly unwind.

I can see how a computer would call this the optimum solution. It would look great plotted on the friction circle, as opposed to the classic line, which never explores the braking+turning or turning+accelerating quadrants.

Easy enough for a computer, or maybe a ShueyComputer, but I think it is difficult for a mere mortal because the car is NEVER in steady-state cornering in this scenario. It never "takes a set". The force vector is always moving around.

Then again, do you know where this is a very common occurrence? Autocrossing! Imagine doing a tight 180 around a single cone. The turn is so short distance-wise, by the time you've cranked in enough steering lock to get around the stupid thing, you're at the apex and already need to start unwinding. Very parabola-like.

Geeky indeed.



[This message has been edited by Dyno (edited July 14, 2004).]

gsbaker
07-14-2004, 10:47 AM
Originally posted by Dyno:
I can see how a computer would call this the optimum solution. It would look great plotted on the friction circle....

Yup. That's exactly what the article was getting at, keeping the car on the edge of the circle.


<font face=\"Verdana, Arial\" size=\"2\">Easy enough for a computer, or maybe a ShueyComputer, but I think it is difficult for a mere mortal because the car is NEVER in steady-state cornering in this scenario. It never \"takes a set\". The force vector is always moving around.</font>

And the force vectors (the set of vectors at any/every location on the curve) always pass through a single focus point, which moves depending on the angle of the turn.

Fun stuff to tinker with, but I'm happy to just keep the car out of the weeds.

------------------
Gregg Baker, P.E.
Isaac, LLC
http://www.isaacdirect.com

Mike Spencer
07-23-2004, 09:52 AM
I *JUST* picked up the issue and haven't really read the article yet. However, it seems to me that they are probably computer modeling instead of doing real-world testing. I believe the results would be different.

I'm not saying what they call the "fast" line wouldn't be faster, but the computer model assumes (wrongly) that traction is equal on all parts of the track. The "racing line" has more rubber down on the track and has more traction than other areas.

It *WOULD* be interesting to see what would actually happen out on the track....


------------------
Mike Spencer
NC Region
ITA/7 RX-7 (finally)
1990 RX-7 Convertible (street car)

grjones1
07-23-2004, 11:26 PM
[How do other front wheel drivers with a Welded diff or LSD handle 2 apex turns where trail braking is key?

Raymond,
I don't know the "geeky" description, but sometimes I find cracking the rear loose to line up the second apex is the only way to go. An inside pass at Oak Tree sometimes calls for the same thing. Sound viable?
G. Jones

emwavey
07-26-2004, 11:53 AM
Finally someone who is factoring time it takes to turn the wheel. I posed this question a while ago in H-T, and no one understood what I was talking about... they kept misunderstanding and thinking I was talking about "lag-time" more in a stock car, with how long it takes for the suspension and tires to fully understand the inputs.

So how does actual steering input factor into your driving. More quickly, more likely to brake loose, more slowly and well you're just too slow?...

-dave
8)

Mike Spencer
07-28-2004, 03:27 PM
Originally posted by gsbaker:
Interesting stuff. Anyone tried this?



Gregg -

I finally got around to reading the article this morning. The misunderstanding is probably ALL my fault, but I thought they were breaking new ground. When I was done reading my initial reaction was, "So?"

The bottom line (no pun intended) seems to be that trail-braking is faster than what they call the "classic" line. Personally, I don't believe I have EVER seen anyone brake in a perfectly straight line, drive at a constant velocity through the turn at the "optimum cornering speed" until they are pointed down the next straightaway and then (and ONLY then) get on the gas.

My take on it is that everyone trail-brakes TO SOME EXTENT. (Probably less so on corner entry and moreso and corner exit. (Although, on corner EXIT wouldn't that be trail-accellerating? But I digress...) In that respect I thought the article was somewhat simplistic (or perhaps less valuable than I first thought).

In another respect, I thought they did a wonderful job of taking the "everyman" explanation of trail-braking that Carroll Smith wrote (I forget which book) and complicated the bujeezus out of it!

That said, I may try to modify my corner entry a little in the future just to see what happens.

I don't mean to be too critical. I thing "Racecar" is a great publication. I wouldn't pay the exorbitant price otherwise. http://Forum.ImprovedTouring.com/it/biggrin.gif

Just my $0.02

------------------
Mike Spencer
NC Region
ITA/7 RX-7 (finally)
1990 RX-7 Convertible (street car)