G-Sum - questions for the proponents
#1
Lifetime Rennlist Member
Thread Starter
G-Sum - questions for the proponents
I have been playing with data and decided to do G-Sum analysis on my data from last year. Before I jump in, I want to make sure I understand exactly what this is. My unerstanding is that you take the combined g (which is the square root of the sum of squares of lat g and long g) and integrate that over a section of track. Then the theory is that the max G-Sum shows the fastest lap or section and then you can see where you should improve?
Is this correct? Are there any G-Sum advocates still around here?
Thanks
Is this correct? Are there any G-Sum advocates still around here?
Thanks
#3
Well, it has to be more complicated and analysed than that... You could pump up the
G-sum numbers by simply doing a couple of straight-line max braking exercises
down the main straight, and then accelerating back up to speed... The trick is
that not all G-load is appropriate or beneficial G-Load.
G-sum numbers by simply doing a couple of straight-line max braking exercises
down the main straight, and then accelerating back up to speed... The trick is
that not all G-load is appropriate or beneficial G-Load.
#4
Lifetime Rennlist Member
Thread Starter
The car is a Stohr DSR. 1000 cc making about 185 HP. Little bit of downforce and gets around the track OK.
Joe,
I agree with the defect in the concept of G-Sum. I am going to run it on my '04 Run Offs data. Exclude lap #1 as it is a partial race lap, and the final lap as I had an off. Otherwise there was certainly no extra braking and such. Seems that should be a good data set to see what G-Sum really discloses. My feeling has always been that it will not show anything that other data does not demonstrate better, but that was always challenged by some here. CC was going to do the G-Sum calc on my data but backed off. So here we are in the dead of winter and I am now willing to set up the calculations in the Motec software to do the calcs. Just want to make sure I am doing the right math according to the proponents of G-Sum then try to get some agreement as to what it should tell us.
Joe,
I agree with the defect in the concept of G-Sum. I am going to run it on my '04 Run Offs data. Exclude lap #1 as it is a partial race lap, and the final lap as I had an off. Otherwise there was certainly no extra braking and such. Seems that should be a good data set to see what G-Sum really discloses. My feeling has always been that it will not show anything that other data does not demonstrate better, but that was always challenged by some here. CC was going to do the G-Sum calc on my data but backed off. So here we are in the dead of winter and I am now willing to set up the calculations in the Motec software to do the calcs. Just want to make sure I am doing the right math according to the proponents of G-Sum then try to get some agreement as to what it should tell us.
#7
Mark, you might want to go back to CC's Famous...
"Slam on your Brakes when entering the Corner".... post.
About mid way thru this thread a computer wizzard who use to work for one of the F-1 teams was posting. He seemed to understand the DAS technical formulas, quite well. Seemed to be a nice guy, too !
"Slam on your Brakes when entering the Corner".... post.
About mid way thru this thread a computer wizzard who use to work for one of the F-1 teams was posting. He seemed to understand the DAS technical formulas, quite well. Seemed to be a nice guy, too !
Trending Topics
#8
Lifetime Rennlist Member
Thread Starter
Originally Posted by trumperZ06
Mark, you might want to go back to CC's Famous...
"Slam on your Brakes when entering the Corner".... post.
About mid way thru this thread a computer wizzard who use to work for one of the F-1 teams was posting. He seemed to understand the DAS technical formulas, quite well. Seemed to be a nice guy, too !
"Slam on your Brakes when entering the Corner".... post.
About mid way thru this thread a computer wizzard who use to work for one of the F-1 teams was posting. He seemed to understand the DAS technical formulas, quite well. Seemed to be a nice guy, too !
The problem with G-Sum (well there are many problems with it) is that it appears to be made up on this site. I have spoken to a couple of pro data engineers and they have never heard the term. CC had previously described it as the integral of combined G, so that is what I went with.
Now, if that has any value what-so-ever, it should be a good predictor of lap time, don't you think? I mean, if my highest G-Sum lap is not my fastest, and the second highest is not the second fastest, etc, then what is it that G-Sum is indicating?
#9
Rennlist
Basic Site Sponsor
Basic Site Sponsor
I could never figure out the graphs. They always gave me a headache after pouring over them for a while. I guess that my *** isn't cut out for DAS. I'll have to stay with the SOP method or hire a race engineer. Actually, I really like video & what Tony's software does for it. That seems to make the most sense to me.
__________________
Larry Herman
2016 Ford Transit Connect Titanium LWB
2018 Tesla Model 3 - Electricity can be fun!
Retired Club Racer & National PCA Instructor
Past Flames:
1994 RS America Club Racer
2004 GT3 Track Car
1984 911 Carrera Club Racer
1974 914/4 2.0 Track Car
CLICK HERE to see some of my ancient racing videos.
Larry Herman
2016 Ford Transit Connect Titanium LWB
2018 Tesla Model 3 - Electricity can be fun!
Retired Club Racer & National PCA Instructor
Past Flames:
1994 RS America Club Racer
2004 GT3 Track Car
1984 911 Carrera Club Racer
1974 914/4 2.0 Track Car
CLICK HERE to see some of my ancient racing videos.
#11
Rennlist Member
Not a proponent, just another guy dumb enough to have read all of those threads. I don't remember everything, but I do know that a major assumption for the analysis was that you would drive the "right" line. So - analyzing G-Sum alone would not be conclusive unless you knew that you were on the right line.
There are probably other caveats, etc, but that's just the one that I remember
There are probably other caveats, etc, but that's just the one that I remember
#12
It is very simple.
Forget about the lap integration. Plot the Gsum line for the corner in question. It should quickly rise to the peak decel value (provided there is significant braking for this corner) and stay at or near this value until almost the track out. If it does not, then you are losing time in the corner by not using all the grip the car has to offer.
This most common problem is having a big dip during the transition from braking to cornering. Hence the 'brake hard to the apex' thing. It is not that simple, but it does express the goal in obvious terms.
It only tells you how effectively you are driving the line you have chosen. By doing lap to lap comparisons you can find the best line. If a lower Gsum yields a better split time, then that line is better.
It is a very quick and easy way to show the driver's strengths and weaknesses. Without it, you do not have a way (other than by looking at inconsistency in split times, which takes some time and can be misleading) to evaluate performance unless you have a faster driver to compare against.
I don't understand the bias against DAS. Everyone agrees that you can't be a top notch engine builder unless you have a dyno, and you can't be a top notch chassis guy if you don't have scales. How can you be a top notch driver without DAS?
Chris Cervelli
Premier Motorsports
Forget about the lap integration. Plot the Gsum line for the corner in question. It should quickly rise to the peak decel value (provided there is significant braking for this corner) and stay at or near this value until almost the track out. If it does not, then you are losing time in the corner by not using all the grip the car has to offer.
This most common problem is having a big dip during the transition from braking to cornering. Hence the 'brake hard to the apex' thing. It is not that simple, but it does express the goal in obvious terms.
It only tells you how effectively you are driving the line you have chosen. By doing lap to lap comparisons you can find the best line. If a lower Gsum yields a better split time, then that line is better.
It is a very quick and easy way to show the driver's strengths and weaknesses. Without it, you do not have a way (other than by looking at inconsistency in split times, which takes some time and can be misleading) to evaluate performance unless you have a faster driver to compare against.
I don't understand the bias against DAS. Everyone agrees that you can't be a top notch engine builder unless you have a dyno, and you can't be a top notch chassis guy if you don't have scales. How can you be a top notch driver without DAS?
Chris Cervelli
Premier Motorsports
#13
Nordschleife Master
Originally Posted by Premier Motorsp
I don't understand the bias against DAS. Everyone agrees that you can't be a top notch engine builder unless you have a dyno, and you can't be a top notch chassis guy if you don't have scales. How can you be a top notch driver without DAS?
Personally, I think DAS vs. non-DAS relates more to how people learn. Some drivers are more feel based, and some more technique based (just like in golf or horse riding). When I had my DL90, I did like it, though I think the data would have been more useful when used with video. I plan on upgrading to a DL1 over the winter, as I do think data could help me improve.
#14
Rennlist
Basic Site Sponsor
Basic Site Sponsor
Yeah, but I now understand more about what you want from Gsum from what you just wrote than I did after reading dozens of threads and 40 or 50 pages of diatribe, though I still need those graphs dissected and explained in detail. Chris, it sounds like you're the man for the job.
#15
Lifetime Rennlist Member
Thread Starter
Originally Posted by Premier Motorsp
It is very simple.
Forget about the lap integration. Plot the Gsum line for the corner in question. It should quickly rise to the peak decel value (provided there is significant braking for this corner) and stay at or near this value until almost the track out. If it does not, then you are losing time in the corner by not using all the grip the car has to offer.
Forget about the lap integration. Plot the Gsum line for the corner in question. It should quickly rise to the peak decel value (provided there is significant braking for this corner) and stay at or near this value until almost the track out. If it does not, then you are losing time in the corner by not using all the grip the car has to offer.
Originally Posted by Premier Motorsp
It only tells you how effectively you are driving the line you have chosen. By doing lap to lap comparisons you can find the best line. If a lower Gsum yields a better split time, then that line is better.
It is a very quick and easy way to show the driver's strengths and weaknesses. Without it, you do not have a way (other than by looking at inconsistency in split times, which takes some time and can be misleading) to evaluate performance unless you have a faster driver to compare against.
It is a very quick and easy way to show the driver's strengths and weaknesses. Without it, you do not have a way (other than by looking at inconsistency in split times, which takes some time and can be misleading) to evaluate performance unless you have a faster driver to compare against.
1. Look at the Combined G chart to see if you are using all the grip.
2. Then go for lower lap times.
I appreciate the opportunity to try to understand this and question it without being called Low Character.
BTW - I am clearly with you on the value of data aq. It can show all sorts of things that the driver can not feel or only suspects. But, IMO, it is not as good as video for newer drivers.