trumperZ06

Well now.... maybe... Chris has cleared up the majic of " G-SUM ".

It seems that we are looking for deviations... rather than trying to MAX a graph. So do we need an experienced pilot to " set the bar " and then compare our graph to their's, or do we break down various laps ( specificly corners), to try and determine where we are SLOW ?

IMO... it looks like DAS may be more useful for the experienced driver... looking to find those last 10ths of a second... while the newbe's should concentrate on the basics. Until you are able to lap consistently, it's going to be hard to determine what is causing the deviations. It could be a combination of driver imput, late braking driver error, wrong line, early apex... etc.

RJay

So I've stay out of this thread as I was observing a suitable period of morning for the loss of out dear departed tiny Tim. ( Sorry we're coming up to the holidays and A Christmas Carol is my favorite story of all time)

I was just wondering. Do any loggers actually display G-Sum or lat/long G on the in cockpit display? I' don't think my AIM can do this sadly. One thing I think would be extremely helpful is if I could take advantage of my shift lights to indicate building lat g instead of revs. I've got this problem with being a bit of a chicken$hit in a few particular high speed corners. When I look at the logs, I see I'm pulling a couple of tenths less than I probably could. Nevertheless, I have a hard time keeping my right foot down or carrying in more speed. I think the lights might help in these few situations where I just can't seem to make myself do it. Just an actual indicator (as opposed to the butt-o-meter) that theres more on the table, its okay, you really can be a little quicker here, just a corner of the eye indicator. Might be helpful for braking as well.

Another question for the technically armed. Is the capability of a tire essentially the same regardless of direction? I.e if I can get 1.3 G laterally out of a tire, should I be able to 1.3 G under braking as well? Or is this relationship assymetrical?

RedlineMan

So, here's what I'm taking from this;

- Tim was not correct. As much as I am able to understand exactly what he was trying to prove, he was wrong in his insistence that MAX everything was important.

- MAX needs to be in quotation marks. It is a highly theoretical value that changes with conditions. Seems to me you could have a smooth graph, and be slow, or a choppy plot and be fast? How do you decide what "max" even is? Max would be the best you were able to muster, or your best average?

- Further, seeing a deviation from "max" is only actionable if you can discern that there were no external variables influencing the numbers.

- This sort of analysis would not be very helpful in choppy, "highly active" sections of track like Chris highlighted earlier. Can you imagine the plots for Madness? The changes are so drastic as to make the Delta too fragmented. It seems to me that this data would be far more usefull for portions of track that were reasonably "steady state" or geometrically regular. Coincidentally, it is not the slow portions where you generally can "make" time, but only lose it. It is the faster segments where you need to ratchet up your courage and fine tune your methods.

- It seems to me that getting the clear track to lay down a bunch of laps to create averages from which to highlight anomalies would be VERY difficult for most of us. Having the scratch to rent a track to get clear laps is beyond me, for sure. What's a testing budget?

Forgive me if any of this is dumb, but I've kind of maxed my mental plots here.

SundayDriver

Overlays are great but why would we care about g data in this case. Overlaying speed tells you so much more. You can see if someone in the same car is carrying 2mph more in a corner. You can see the effects of different braking and what that means to speed. This is one of the most powerful tools out there.

kurt M

RJ, get yourself a small lead weight and paint it red. Hang it from the rearview mirror on foot long string. When you see it the pointing to 9:00 O clock you are doing a right-hander well. Add another weight and string and hang this one from the A pillar. When this one points to 3:00 O clock you are hitting a left-hander well. If ether one hits you in the helmet while coming out of a corner you have enough tire and motor.

Just an opinion.
Max while driving a textbook line for the car being driven, Tim is correct. The problem is two fold... What is max and what is the correct line? Find out the answer to both then use historical and future data to see if you and you car are able to achieve the golden point or singularity of using all of the track and car. Make a change to anything, look and see if the change improved or reduced your or the car abilities and times. The key is to look and understand all the variables and not to focus on one aspect only.

If you are cutting with an ax you don't need to measure with a micrometer as chalk works fine and is better suited to the task at hand. Data analyses = micrometers. Use it when you can drive that accurately and can act on the information. It has a place in race and performance driving but a very specialized and rarified one. It can be a good part of a training system but it is not as "someone" might think a stand alone training tool. Just as ABS and PSM will not get you through a corner by themselves.

SundayDriver

Yes, many systems can display g's - at least lat and long. Stack and Pi both do this. Additionally, my MoTeC can display virtually any math channel so it could do combined g's as well.

But there is a huge problem with g data. It is a very noisy signal as you are not only measuring the g data you expect but also every little flex and twist of the chassis. The g signals need to be rather heavily filtered. Even with that, you may show 1.2 g's one moment and 1.4 the next all while pulling the same speed on the same corner. So which is right. You really need to look at a chart to visually smooth the data.

On the other hand, speed data has almost no noise and every data system I have used can be set to display minimum corner speed as you exit the corner. So do you want to try to read g data in real time as you navigate a corner, or see minimum speed after you have exited the turn?

There are great tools to be found in data acquisition systems but, IMO (and that of the pros I have worked with) you are not going to find it in g data.

TD in DC

John,

Your summary is what I personally have taken from this thread, except perhaps the thought that the G-Sum analysis might not be helpful in "choppy, 'highly active' sections of track." For the experienced driver, I think that it would be just as useful, if not more useful, in those parts of the track than in more "traditional" parts of the track. Provided that you both (1) drive well enough to know what to expect from yourself and the car and (2) you have the skills and experience to implement changes, you could use the analysis in a weird part of the track to analyze different lines by experimenting and attempting to minimize G-Sum (or at least minimizing G-Sum "peaks"). Indeed, without the analysis, you might be fooling yourself about what is the fastest way through the weird section. By using the analysis, you might learn that the "way everyone does it" is a good way, but not the best way. Once you have determined the best line, you could still use the analysis to see mistakes or timing adjustments you can make in order to maximize speed on that line. You will be looking for deviations from the values you know you should have based on what you are doing at that time rather than for deviations from "constant" values.

Just my two cents, but I am struggling through with everyone else.

TD

Premier Motorsp

Sunday is right. If you have speed data from a faster driver to compare with, then looking at Gsum is secondary to that. If the car was really well instrumented, looking at Gsum would become less necessary yet, as the faster driver's technique to gain speed would be obvious from the other graphs.

Getting data from a faster driver is not so easy once you are racing.

John's points:

Tim was correct. You ARE trying to reach the max and keep it there, but determining what the max is on a given section of track is next to impossible. So a slavish devotion to a specific number does not make sense. Generally a driver who is trying is going to reach the limit somewhere in the corner. At that point Gsum will be highest. It is reasonable to assume that this Gsum could be maintained thru the braking and somewhat past the apex. I understand that this is not always true. Data is useless without interpretation!

Yes, the deviation from max is actionable only if the track is not causing the problem. Set up changes possibly can fix car problems and driving changes are obviously possible as well. You can always accept the fact that you cannot reach what you think the max should be. This shouldn't prevent you from trying to do better.

Madness and my Sears Point example make the analysis much more difficult. You can break up the section into smaller chunks that don't have huge geography changes. Maybe on sections like these a speed trace comparison would be easier.

The beauty of this is that you DON'T have to run a bunch of laps to figure out the max. If you hit 1.7 somewhere in the corner, you probably have a chance of doing all of most of the corner at 1.7. You can find this by doing just one quick lap. Common sense is necessary. If you pulled 2.0 G for a second because you ran into somebody's door, then maybe that lap shouldn't be used for comparison.

Sunday, what would you do if someone is kicking your *** in an identical car and they will not share data with you? How are you going to go faster?

Technical nitpicks:

The pendulum won't hit 3:00 or 9:00 no matter how hard you corner. Even if you are at 100G lateral, you still have 1G pulling down on the pendulum.

Tires generally will pull pretty much the same G in any direction. Cars are not always capable of this. Usually you do better in Lon G because the wheelbase is longer than the track.

G data does need to be heavily filtered, especially if you are using to dictate driving style. You can't make control inputs at 100Hz, but you can easily log G at 100 Hz.

Chris Cervelli
Premier Motorsports

RJay

Thanks gentlemen! Wasn't thinking of actually looking directly at the numbers as much as being able to set trigger levels on lights simply to give me a rough idea when I'm NOT pulling the lats that I think I am. No lights == try a little harder, you've got some in the bank!

BTW, not to go OT, but you been on TV yet Sunday? Watched the DSRs this weekend and thought of you.

kurt M

Pull one lat G in a corner will produce 45 deg angles on pendulum as you said there is one G pulling sideways and one pulling down. The 3 O clock and 9 O clock references were absurdum ad infinitum and a poor attempt at humor

Exit speed, segment and lap times rule.

SundayDriver

Chris,

Here is what I have done and would do:
1) Follow that car in practice/qualifying to see where and why they are faster.
2) Find someone faster to drive your car to get their data. There are always some very talented pros around who can do this becuase huge driving talent does not equal financial success in racing. For example, I know of a few Mid Ohio instructors who do this at MO.
3) Get coaching from a pro.
4) Look inside yourself and ask, for each part of every corner - am I terrifying myself? If not, then there is probably more speed there.
5) Look to the data. If you look at you data after every day, then you will have a feel for what it should look like. I think the ability to use data to find more speed is one of those 'art' things. It is not always scientific. But tools like eclectic lap (where the system builds a theoretical lap based on best segment times) can give you a pretty good target for improvemet.

Of all people, I certainly do not discount the value of Data Aq but I value the other items even more. Is this the right way to do it? I don't know, but it is the way I have done it, and it works for me and I have seen it work for many others as well.

TD in DC

hmmm. So a plot of driving experience versus terror looks like an inverted bell curve.

You start out terrifying yourself because you have no idea what you are doing. Then you learn how to drive and eventually get to the point where you are subconsciously competent and not scared anywhere because you are in complete control. Then, you start wringing out every bit of speed possible until you are back to where you started because you are terrified through the corners and farting couches of the same material as your seats everytime you return to the paddock.

Very zen like.

No wonder this sport is so addictive.

kurt M

Ignorance is bliss. Terror comes not from inexperience but from experience.

I like "Am I terrorising myself in every corner" I am not there yet but there is a set of corners at VIR that I have yet to take a breath while driving them.

M758

Chris those words the key aspect of this entire g-sum good vs bad.


One can't just look at g-sum to be fast. Split times and overall lap times play key role as well.

Geo

This is exactly what I was wondering. G plots still seem like trivia to me when you can plot speeds and look at sector times. Chris even said you'd have to look at sector times for some things, so why not just start there?