Poor-sha

I talked about this idea I had in a couple of threads on the forum. This is the first part in a series where I use PDR video and data to analyze track incidents. This first episode is an oversteer situation exiting turn 7 at Road Atlanta that results in contact with the tire wall. Big thanks to the driver for sharing the data and being willing to let me make this video public.

fleming23

It is so hard to not lift when the car does something unnatural. I've done it countless times, and probably gotten lucky more than a handful of times. Thanks for putting this together! It is often tough to make those split-second decisions to save, or prevent, a crash, but it helps to see the what and why to try and make better conscious choices lap after lap.

I just watched the short clip of the car spinning maybe 10 times, and compared it (visually) to a video of my laps and the only visual cue I see is that he appears to be cutting the track shorter than necessary and could have tracked out to the rumble strip. I don't think he would have touched the outside gator had the car not spun. This maybe corresponds with your analysis in that the driver had more steering angle in than maybe was needed, coupled with additional speed and G-load.

Whis9

RocketRod

Thanks for sharing your analysis!

Racingswh

I have destroyed a car just like that as well. Backed it in though since it went all the way around. Nice to see it broken down like this with data. Thanks for doing this!!

Poor-sha

Yeah, not lifting is very counter intuitive and difficult to learn. In this case it was ~0.2s from the time he started opening to the wheel to the time he started lifting. That's all reflex and not cognitive processing. People tend to scoff at me for saying this but it's one area where even the most basic driving game or sim really helps. As long as it has the physics model to unweight the backend when you pop off the gas and you're using a wheel with pedals you can help train yourself to to not just lift in real life.

Bad_AX

This is definitely worthwhile. Watching this video several times causes me to relive the first major spin I had on track at VIR turn 12. The data analysis really breaks down the contributing factors. The driver is applying too much throttle for the steering angle with highly predictable results. The driver is overly focused on setting up for the pass and not paying enough attention to the immediate vehicle dynamics. Ask me how I know ;-) Unfortunately, he did not have much runoff area. On balance, I think it is hard(er) to learn without exceeding the limits and experiencing loss of control. I was fortunate on my first "big off" to have a driving coach that insisted on breaking down the mechanics of the spin and then getting right back out on the track and getting back up to speed to reinforce the lesson.

Thank you for posting this and thanks to the driver for sharing the video and data.

Happy New Year!

fatbillybob

I think you got a great idea for a learning series. But is the pdr reliable and accurate to do that? I have a hard time trusting these kinds of data things. Your data based conclusion is trailing throttle oversteer...basically guy lets off gas unsettles car as car yaws countersteer without effect hits wall. That all sounds great and your data analysis sounds convincing. I suspect the driver you told that to is still puzzled. How can he go against data? Is the interpretation right? Is the timing of all the data parts right? I got a gps based video camera that seems at times to not be able to synch the video with the audio. What software glitch is causing that? Does the PDR have those kinds of issues which could cause you to draw the wrong conclusion?

I don't do squiggly lines but I want to learn. Looking at the video It really did not look like he did anything wrong. His statement of doing similar to the lap before seems consistent to me. Except he was going faster (based on your data stream) and probably wanted to chase the gtr down. Redmist we all get that. Video data seems to bare out what he says...he is driving that turn like he always does. Video data shows steering angle at one point I can freeze the video at 3 degrees and he is going straight down track nicely. All that time prior and after this 3 degree point video shows rpms close to 3000rpm. Video does not show rpm drop until well after driver knows he is in trouble and goes well into yaw. Then steering angle shows big left correction. He did everything right as any driving instructor would say back blows out counter steer. So why didn't he catch it? I say from the comfort of my keyboard armchair quarterbacking that he was faster than before but driving the same way as before that was the error. The TTO was just an aftermath symptom. The closer you go to the edge the more you have to be ready to react to the edge or anticipate your response to going over the edge. That is why most can't do qualifying laps for an entire race. The driver therefore ran out of talent going faster not feeling the early signs of breakaway of the car or failing to realize the limit of the car, and thus counter steers too slowly and too late. He was reacting to the car instead of getting ahead of the car. He did not realize his extra speed would require more track out. He kept his steering input to point the car down the straight to use less road and keep pace with the GTR. He could have driven the same at the higher speed if he used more road which he had plenty of on the left track out. Then he could have driven the same way. So there are 3 things that could have been done besides drive slower, counter faster, counter with anticipation, use more road.

Finally, what influence is the track? I have never been to RoadAtlanta but will race there in a couple of months. I'll have a 20 min practice session then qualify and race it. So I have to learn quick. There are lots of black marks on the tarmac where the driver hits the wall. It seems like this occurs often. Why? Video never does a turn justice. What is that turn really like? Some turns are deceptive and look like you can drive anywhere but have a set line for speed. It looks like the apex is sort of uphill and over then the exit is off camber and down encouraging a spin? It looks like both the driver and the GTR need to be more left by the curbing to minimize the off camber, use all the road and launch out the turn. People who use the gas pedal as a light switch or big horsepower could have more problems here. I would guess that this turn 7 is a trap. If racing to defend the classic inside line your opponent can beat you on the outside because the defensive line is slower than you think.

Anyway that's what I'm thinking. I'm no instructor and I don't live at the pointy end of the grid either. At best I'm a Runoffs mid-packer, so if others see it differently I'm willing to learn.

dclafleur

I'm not able to watch the video at the moment but if my memory is correct Turn 7 is somewhat straightforward type 1 turn, brake hard, apex late, on the throttle early to try and gain MPH on the back straight. Someone feel free to correct me if I'm wrong on this its been a few years since I drove road Atlanta.

truth.b

sub'd to view at lunch

moose.b3

Thank you for posting this. I appreciate the lesson.

GoatHead

Thanks, and keep them coming.

village idiot

Can you post a summary of what happened for those of us that can't watch the whole video in the office?

Part of the big problem is that he kept adding left steering angle. When the front has no grip, you have oversteer, especially on throttle. I can't watch the whole data analysis at work. I'll check it out later but I'm guessing he got on the gas earlier and throttle steered a bit, which is why he pulled left. Then it started coming around, he got understandably scared and our natural reaction is to turn away from the wall coming at us not realizing that more steering angle = less car turning. Either that or he had the car rotating on entry. That's generally how you crash to the inside.

Small steering inputs are key. It takes a long time to change our reactions from OMG! TURN! to small steering inputs, but it definitely helps control the car. I made good progress with that last year but have a ways to go.




The real problem with comparing data from a previous lap is that there is no guarantee that the previous lap is a good comparison, especially with ABS/TC (he's in race mode, which makes it clearer). The C7 chassis, ABS and incredible amount of mechanical grip does a lot to mask mistakes, generally.

shizon'00

Awesome stuff and this is going to be a good series. I had a moment similar to this from a recovery standpoint, but for different reasons. I was at Indy last summer instructing with MORPCA at IMS. My wife and I were both driving the C7Z and had brought a few sets of used tires. Long story short, delaminated cup 2 tires means no rear grip and it happened somewhere during this session. This is complicated by a transition in T13, right hander, from the in-field back on the oval where the outside rear tire gets light as the outside front hits the transition at a diagonal (think of balancing the car on the outside front and inside rear). It's a very unique dynamic I hadn't felt before, but it unsettled the car enough to start some power-on oversteer towards the inside wall. A little steering correction and a small lift of the throttle to stop the rear wheels from accelerating but keep weight in the rear saved me from a very expensive mistake as I think safer barrier might cost more than the car! It's all very clear in the data.

Another good example was that C7Z that hit the inside wall at the ring. No data, but you can hear the same basic concept and see slow hands in the the steered wheel angle.

A pinch + unloading of the rear in the recovery stage. I think everyone has done this at some point in their track career. I often tell students that "opening the wheel" could just be a couple degrees and it's all the difference in the world.

Sean, I'd be curious if there's anything in the shock pot data to look during recovery?

Great stuff! Also, big thanks to the provider of the video/data so we can all learn!

YYZ06

Good stuff! Thanks for posting. I'll be looking forward to the next installment.

TCorzett

Thank you to those who are sharing their video/data so the rest of us can learn... and thank you for putting together the analysis!

NSFW

Thanks to whoever shared the video and data, and thanks for a great analysis.

My crashes to date have all been in lower-risk stuff (snowboarding, mountain biking) but I try to look back in terms of "what could I have done to save that?" and "how did I get to the point that I needed a save?"

This makes me want to invest in data sooner rather than later, because without the data traces it would be really hard to recognize (and learn from) what went wrong on the way in to that corner.

And also a racing sim, because I don't know how else to develop the right reflexes to pull of a save in a situation like that. At least not without sacrificing a lot of cars.

C6WNC

Poor-sha.....Really good analysis. This type of mistake is more common at T7 than most realize. Higher corner entry speed, not unwinding the wheel on track out (pinched it down) and a little too much throttle then off throttle. Just look at the video and note the skid marks on the asphalt, they all go to the right. Most would think an off would occur tracking out to the left with too early a corner entry plus too much speed. The other corner that will bite you at RA is T5. On corner exit and under acceleration up the hill the weight transfer is all in the rear. Drivers are in too big a hurry to move over to the left to setup for T6. The front gets light and the driver puts too much steering input. When the weight starts to transfer back the front grabs and the result in snap oversteer and you meet the wall on the left. ..Look at the skid marks on the asphalt as they all go left. A few years ago, a good friend was going to RA for the first time and I told him about T5. It happened and he backed his GT3 into the left wall.

Fatbillybob...If you can get to RA early enough, a track walk would be a big help to understanding some of the corners. If that is not possible you can drive through the infield area and get a closer look at the track and the different corners. When I did a track walk I was amazed at how much banking was in T6. I alway thought it was fairly flat.

Learning from videos and analysis like this is very valuable. We learn expensive lessons when it goes wrong, which I experienced a year ago at VIR. I had a GoPro video and Aim Solo, and it was valuable watching the whole thing unfold. I had a hot lap going with a presumptive 2:01:00 going into T11 but entered too hot for my heavily heat cycled hoosiers to hold enough grip. I drifted on to the FIA gators approaching Oak Tree and lost all my grip...two wheels off and car would not rotate, I had a meeting with the tire wall. "Ambition exceeded adhesion" in this case. I learned from my error and I am a better driver for it.

Poor-sha thanks for the contributions and insight in this forum. I don't post much, but alway enjoy reading your comments or threads, along with some others on CF.

Bob

CanadianVetster

Poor-sha

Episode 2 is up on the channel