I need to summarize what we have all learned here about proper driving, because I am clearly not intelligent enough to master this material. So here goes:

1) Threshold brake to the apex on almost every corner.
2) Early apex in order to apply earlier throttle.
3) Apply heavy filtering to your g data and plot the derivative, called jerk. Only be relentlessly plotting filtered data, can anyone ever come close to ColorChange's jerk rating.

Did I miss anything?

 

Here is my g and jerk. I did lateral instead of combined, but it is the same idea as what Color posted. You can see how smooth I am and the jerk plot shows how well I manage the car. I am pretty well at the limit and am pretty sure no one - probably not even Shumi - could do it better. Gee (sum), I hope no one asks what the raw data looks like.

 

I'm not sure that's the same thing that I was saying.

First, a spread between .8 and 1.0 is fairly huge. In a 75 MPH turn (400 foot radius), that's a speed differential of 8 MPH. Racers would kill for such a huge speed advantage in coming out of a corner.

Second, without knowing the course, I have no idea of knowing whether or not you should be getting similar G-sum peaks. For example, maybe you have a bit of camber in your favor for some turns, and you are getting G-sum peaks of 1.0 on those turns. A better driver/line might be able to get G-sum peaks of 1.2.

I'm just saying that conceptually, it seems difficult to believe that my car is going to corner the same irrespective of where my right foot is.

 

Brian: Yes, .8-1.0 is pretty large as I am in a new car, having trouble with brakes, and not a great driver.

Look at my video to see the course. It is pretty flat and only minor camber on a few turns. The driver could only pull 1.2 g-sum if there was banking/camber I wasn't using or taking a widely different line that happened to have much higher friction coefficient (or switched to R compound tires). The track has pretty consistent surface so this is highly unlikely at the track.

I will post g-sum data from other tracks when I drive them and unless something is happening at the turns (elevation, camber, ...) the max g-sum number is very similar, track to track, just like my graph shows.

Conceptually, the car has a fixed amount of grip (ignoring aero). You can use it for accel/decel and lat, in any combination. You are forced to trade lat g's for long g's if you are on the fc (g-sum max). That's it.

 

Sunday - Gee, that's schmoother than a baby's butt. Is that due to all the snow on Mid-O ??? (display says Jan 6, 2004...unless you're one of those metric date weenies)

You need to supply us with the fully ripe, unfiltered Jerk output.

 

CC - It looks like we can agree on something. If the DAS cannot show you the proper line, why bother. Many have suggested a competent instructor preferably one with experience in your model of car. If there are different lines for the corner, an instructor will show you all of them and let you experiment (at least that's what I do). You can then use DAS to optimize. Since the G-Sum permits multiple solutions which produce equal results on the graph, it cannot be useful in determining the fastest line. By the way, the fastest line in any corner is not cast in concrete. Varying track conditions will affect the optimum line. Racers experiment in practice to see which of several lines is the "good" one for that particular day.

As to your graph I can't derive smoothness information from it.

Please, tell me that we are not (as Mark implies) back to the trail brake to the apex thing again....

On ABS - In JUne we ran a test. Was set up a course on a very large skid pad. The idea was to accelerate to 75 MPH and at a set of cones bring the car to a stop in as short a distance as possible. We left plenty of room to stabilize the entry speed.

We observed markedly shorter stopping distances when drivers used threshold braking as compared to ABS. I am not talking about a few feet which might be within the margin for error, I am talking about 15-20 percent shorter. Cars used for the test were a GT3, a GT2, a 996TT, and several non Porsches. Of interest was the fact that Hondas seem to have an overly sensitive ABS system, they showed the largest differential in stopping distance between ABS and threshold braking.

Regards,

 

Could it be possible that similar peaks in your graphs are merely showing your driving limits rather than the cars limits? I personally find it hard to believe that anyone on this board is 'at the limit', since if we could do that with consistency, we'd all be pro drivers.

 

Why thank you Professor. No snow - just those silly Aussies and the way they show dates. No snow needed, I am just sooooo smooth becasue the data proves it. Raw data? Well, if you insist - I will show mine even though Color won't. Just look at the jerk plot. Looks just like the filtered one, doesn't it?

 

Mark - the lower plot looks like the audio from a rolling stones concert. Is that smooth?

 

Bob: Not sure what we agree upon.

Since the G-Sum permits multiple solutions which produce equal results on the graph, it cannot be useful in determining the fastest line.

No, there is only 1 way to drive any 1 line at the limit! And, the DAS is great at analyzing which line is faster. It can’t suggest the fastest line on it’s own, it’s up to the driver to select the line, drive it near the limit, and the DAS will show wheather it was faster or not.

As to your graph I can't derive smoothness information from it.

If you watch my video, you can see that I don’t shift like an idiot. Therefore, if shifting causes a 0.4 smoothness peak, taking a turn at below 0.1 is pretty good. Does that help?

The DAS can give an indication on lines chosen by looking at the radius. When they add 3 dimensional WAS GPS (should be ready soon), it will be way easy to analyze lines. I can’t wait for that.

I will share my braking data when I get it.

Pesky
When the car shows the same max g-sum numbers on the skid pad at the limit as on the track at the limit, the number is known. Quite simply, I often skid and if the g-sum never gets past 1.0 during a skid (or high slip angle), guess what? That’s the max!

 

Color - Lateral G figures can be misleading. I borrowed a DL1 (very neat machine using inertial sensors and GPS) and saw readings over 1.2 G. I doubt my car is capable of that on a level surface. Throw in a slight banking for drainage and bingo - more G's. That may make lat G's misleading if some turns are banked while others are off camber. Bumps also mess things up it's hard to get useful readings on bump sections.

Rgds,

 

OK guys (you know who you are) pay up.
I told you that I could get Color to waste his time on that stupid jerk/snap idea. He even renamed it smoothness. I'm pretty sure that if it is useless data, then we can get Color to believe it means something. I won that bet.

 

Mark, what's next, the snap crackle and pop theory? I think it only applies to Japanese cars though ...

Is there a relationship between "jerk" and Professor Tester's "jack"?

 

So...uh...Mr. Driver - Looking at the scales on on the vertical axis of the left-hand side, it looks like there is a lot of extraneous, noisy spooey on the 'jerk' data that is more than an order of magnitude higher.

Were there 'tire worms' on your tires, or did you drive thru the marbles on that lap ?

 

Professor! Brilliant analysis. I would never have considered tire worms - it's too cold for them up here. We don't have termites either BTW. I am more accustomed to the larger version of the tire worm called a "snow snake". They are white of course and cannot be seen. They do snap at your skis and knock you off balance though resulting in a lack of smoothness. I can see how a tire worm could do something similar. Next trrack event, I plan to make a close examination of my tires since I suspect the tire worm, like the killer bee, has moved north. I recall skidding my car in spite of my perfect driving. You may have resolved a mystery for me. Note, that I am not confusing tire worms with the customary "brickyard barnacles" which cling to slow moving cars owned by HWFMR members.

Signed,

Relieved (take that any way you want)