OK, In the interest of eliciting more positive feedback, I thought I would seek assistance from this board on what I am currently working on. I know many would say just drive etc., but I live in Chicago and it's winter, and my track time is severely limited so I am doing the best I can in other areas (like DAS). Here is my though process.

I am trying to analyze specific turns in my most frequented tracks to see what turns I need to work on. I think most of us have an idea what turns we handle pretty well and which need more work. I want to quantify (to be sure my perceptions are right) this and put a little more "meat on the bones". I am thinking of calculating max g-sum as you go through the turn and then looking at actual data to see what percent of g-sum was I able to achieve for each portion of the turn. In theory, this should indicate that "I really need to work on my entry to turn 3" or "trail brake deeper at turn 6" or maybe "achieve maximum lat g's in the sweeper" etc.

Yes, there will be line limitations and other constraints, but I think this is good at identifying two things, 1. how close am I maintaining the limit throughout the complete turn, and 2. how much variation am I having lap to lap.

In order to calculate the max g-sum through the turn, it's fairly easy for the entry and pause phases, because most of my tracks are relatively flat (so please skip elevation and camber changes for right now, also lets ignore down force effects) and I was running street tires with a 1.1 g-sum max. The difficulty comes in track out and this is where I am looking for the most help.

At track out, you are acceleration limited so you cannot achieve 1.1 g-sum after the lat g's drop. Also, the +long g's you can generate is a function of speed (gearing, torque, etc.) so I am thinking you need to create a formula or look up table for +long g as a function of speed so the g-sum max drops as you are going faster.

I hope I have explained this reasonably well. Any better thoughts or suggestions?

 

You know what might be a help in using your data to improve your driving? Put your equipment into a really good driver's car, like Greg Fishman, and ride with him while you are collecting data. Then you could correlate the data with your observations of where and how he is going fast, and where he can/is maximizing his G forces. I think that once you have learned what it takes on the track vs what the data says, you will take a quantum leap in your understanding of what you need to do. I also feel that this may help you become a better judge of what the data really says.

 

Well, I would think that looking at your actual torque curve output (at the wheels) per gear should give a pretty good idea of what accel you could generate with a given vehicle mass. Maybe I'm oversimplifying it?

But I prefer to look at the root question you're trying to answer, rather than surface questions. You've expressed concern about your ability to optimize the acceleration of the car during the track out phase of a corner. This is a very relevant question, not surprising at all given your history with high-hp cars. Not to knock you for having those cars - like I don't wish I had them? But this is perhaps one area you lose the experience gained from driving underpowered cars, optimizing what little you have.

So lets take advantage of being snowed in; now IMO is the best of all times to go back to the kart track, with the hot shoes, and work the kart hard - in the sense of having them beat you out of corners, head-to-head, and learn how to develop the feel of the car tracking out properly. I've always felt that I got a better feel for that from the karts than I ever got in my IT car. Too many variables and still too much torque in the IT car; seems like karts are so crappy off the pipe that you can gain the appreciation for the correct or optimal track out even more quickly, listening to that motor bog. Then you develop the right feel for it, and that can translate back into whatever you drive.

Just a thought.

 

CC said:
At track out, you are acceleration limited so you cannot achieve 1.1 g-sum after the lat g's drop. Also, the +long g's you can generate is a function of speed (gearing, torque, etc.) so I am thinking you need to create a formula or look up table for +long g as a function of speed so the g-sum max drops as you are going faster.


I have been working on this a great deal since the last time we spoke about it. I think what you say above is not really true. I think if you do it right you CAN have max Gsum thru almost all of the track out portion of the corner.

Let's look at the portion of the turn from the apex (really the WOT point) all the way to 100% straight and headed down to the next straight (call this point the end of the turn). At the WOT point you should be at max Gsum and at the end of the turn the Gsum will equal the Long G since the Lat G is zero. Let's use .4G for our example.

What you want is to carry the max Gsum toward the end of the turn for as long as you can.

In the entry of the corner, you have available to you decel capability of 1G (easily). This means that getting the car to max Gsum is 'easy' (not from a control standpoint, but from the standpoint of simply being able to do it).

In the exit part of the corner, you have only .4G of Long G available to you. This means that in order to be at max Gsum, you have to be cornering more strongly, say at .75 G.

What this means to me:

We have gone over the entry phase to death already, but here's one more recap: The turn in is done near max decel capability and then cornering force is rapidly added as decel drops away. At some point, the minimum necessary speed is reached. I THINK (but am not sure) that this SHOULD be at max Lat and 0 Long G (which is not possible in real life, I am pretty sure). Immediately throttle is applied, and Long G builds quickly to the max of .4. During this short phase the cornering force must drop.

Now we get to the sticking point. I am saying that it is wrong to begin unwinding the wheel when the throttle goes to WOT. I say you don't unwind the wheel until you have run out of reason to turn anymore. If you were able to unwind the wheel prior to the end of the turn, you did not maximize speed thru the exit phase, and probably not thru the entry phase.

Also note that the transition from decel to accel will decrease the load on the front tires, which means they will require more slip angle to generate peak cornering force, which supports my theory.

I am not advocating using any more steering angle than necessary at any time. Not unwinding the wheel will become necessary if the corner is taken on the limit. All of this points out that the slow-in, fast-out method of making the exit of the corner straighter than the entry just does not make sense.

At some point in the corner, the Gsum IS going to have to drop from max down to max Long G. I guess the question is how long should this take.

All of this neglects the effect of oversteer. If the car has what I think is the correct degree of oversteer, the driver may have to unwind the wheel in order to continue on the desired line. In this case, though, the Gsum plot is going to look the same.

Larry suggested using the faster driver as a rabbit. Without a doubt, this is the easiest way to see exactly what you are not doing as well as the other driver. The faster driver however may still be (and probably is) sub-optimal. This approach will only work in the long run if you have a string of ever faster drivers to emulate. Once you become the fastest driver in the world you no longer have data to look at.

Because of this and other more practical realities I suggest to CC that he DOES learn to use his own data to go faster. I also suggest that he never miss an opportunity to use another driver's data.

924Racer's idea of figuring out the accel potential is good. It would be even easier to do a series of accel runs from different speed while taking data. Then you could easily put together a table of expected accel based on RPM and gear.

I hope this is useful.

Chris Cervelli
Premier Motorsports

 

Chris, I have never had access to data, but my butt-o-meter tells me a lot about maximizing my g forces. Maybe when I get back into Club racing I'll have the opportunity to do some monitored testing.

Anyway, I feel that there is a small amount of pure maximum lateral acceleration in every turn, the moment you lift off the brakes and transition to the gas. And the total g force achieveable is determined by the limiting grip at either the front or rear of the car. If you are turning in on the brakes, ever lightening them as the cornering force builds, and at the moment that the rear starts to slide, transitioning quickly to the gas, I feel that you are achieving maximum grip at both ends of the car at that moment.

Also, in the part of the turn leading up to this point, your max g sum is limited by the grip of the front tires. After that point, your max g sum is limited by the back of the car (unless you have terminal understeer). The back of the car should be trying to "come around" from the moment you go to WOT, until the cornering forces end at trackout. You have to unwind the wheel in order to keep the front of the car ahead of the back. You are maximizing the grip at the back of the car, even though you are unwinding the wheel all the way.

At trackout, you fairly abruptly straighten the wheel to end the cornering forces because the corner is over. In my example, you are drifting the back of the car all the way from transition to trackout, maximizing the g forces that the rear tires can generate. The fronts would be somewhat under that limit. I think that this is somewhat in line with what you are saying.

 

Doesn't your cornering force depend on the layout of the track? There are many variables... such as track surface, tire condition, temp. & moisture that can change.... even from lap to lap.

Possible examples are: an off camber corner being tricky ( maybe even downhill too), and you are trying to balance your car by using both the throttle and steering imput.

A positive camber corner generally allows greater corner entry speed... and if you set corner entry up right, you can put the throttle down very early. This is where your car takes a set, and you hold the steering wheel constant until track out point is reached. This is why I think that relying on one method.... can be confusing.

I like both Larry's & Chris's explainations... although I don't think they are saying exactly the same thing. Where is John Hajny's (RedLine Man) imput on this ???

 

Thanks guys. Chris … it looks like you’re right. Duh! I should have checked the data. On most turns, the g-sum can be held right to track out and then drops dramatically. Here is a chart that shows it.

There are some interesting things going on here, turn 8 at Putnam, street tires, dry track, 600hp.




A. If you look at g-sum (blue line), I enter the turn pretty well, nice high smooth transition from braking to turning. A better driver would hold the g-sum up higher than I do shortly after A where I let it drop to a little over 1.0g. (my bad)
B. It looks to my like I had some throttle induced oversteer in the shaded zone. If you look, the accel g’s (red line) drop very quickly while the rpm’s rise at the same time speed drops. If you look at a normal decal slope (point D) and compare that to the slope of point E, it looks like I lost the rear end, had to lighten up and then get back on the throttle.
C. It looks like I took a little too much curb on track out. Again, the steep slope drop of accel g’s looks like I spun the tires hitting the curbs.

Overall, it does show that max g-sum can be held until track out for this turn and looking at the rest of my data, it looks like it holds for most turns.

 

You must still be waiting for the CAN interface. Don't hold your breath.

What is the math for corner radius?

Chris

 

Let's assume you are on the right line, since nothing else will work unless that is the case.

So, what is that data telling you? What part of the corner should you work on, and what should you do differently?

 

Is there any DAS system that's advanced enough to tell you this? Let's consider for the second that if you drove a constant radius turn through a corner, you would achieve max G-sum for the entire corner. So, achieving max G-sum throughout the corner isn't necessarily proof that you are driving the right line.

 

No, a DAS system can not tell you this. And that is why I am assuming he is on the right line. If he is not, then none of the data nor advice has any real meaning.

 

Chris: Here is the formula for corner radius. Yes, I am really waiting for the can interface ... and yes I know better than to hold my breath although they told me it would be soon

ABS((Speed_1^2)/(15*Acc_1))

Mark:
Yes, the current DAS does not indicate an accurate line. When the have the GPS it will do a pretty good job.

Mark and Brian:
My goal right now is to drive the car consistently near the limits. Once ... if ever ... I can do that, then I will be able to focus heavily on line selection. Right now I can't do the first part.

 

Mark, that's why I suggested that CC get data from a really good driver so that it will be relevant. And he should be in the car so that he can see the line and the driving techniques that produces that data.

 

The thing about data analysis, and why I have a hard time seeing the utility at my lowly level, is that it seems to be a circular type of problem. I mean the data might tell you that the car can go faster, but unless you have the actual driving skill to balance the car properly and make it go faster, it doesn't really matter much. We all know that putting the car on the edge all the way through the corner and maximizing exit speed is the goal. Doing that takes experience and practice so you can feel the grip that is there and use it. Looking at the graph, it seems to me that you did not carry enough speed into the corner and the g-sum fell off, but as the radius tightened it came back up. Once back on the gas, all that hp is tough to modulate and the rear got loose dropping the g-sum and costing exit speed. I think it takes a lot of skill to go barreling into a corner so fast that the car will barely turn in, but that's what you have to do to keep everything maxed out. Then you have to carefully manage that insane entry speed with balanced inputs as the car wants to swap ends to mid corner and beyond. Like I said, looking at a mathematical representation of what's happening after the fact doesn't show you how to do it, just that it can be done better and we already know that so what's the point. I do think it is fun to think about driving though and if it helps you remember what was happening so you can relive it .. then that's probably not a bad thing.

 

Tim,

What are YOU using corner radius for? I have used it at times. While it does jump around because of the noise in the g data, it is a pretty good indicator of steering smoothness if you filter the data.

When I was at the same stage of learning to drive the car at the limits, I found that data was the most valuable if I could overlay it with a faster driver's data. That really is the ultimate use of the tool at that stage. Many times, however, I didn't have that data so I would try to see what the raw data was telling me, as you are doing now. The point I was driving at is this - your interpretation of the data indicates that you are not maximizing g-sum at exit. (While I personally don't think that is the best way to get around a track quickly, I will respect that max g-sum is your goal). So how do you increase g-sum in that phase? Either take a tighter line (which serves no purpose to going faster) or accelerate earlier.

The data tells you that you should be thinking "Throttle, throttle, throttle" as you get to the apex of the corner. That is the improvement the data is telling you to make. AND, since you have posted videos in the past, that is one of the comments you got from many people who tried to help.