T(h)or's GT86 hammer (Europe) - essential track setup

[Futaba]

Did you end up needing the reflash adapter for the Tactrix OBD2 to flash your ECU on the 86?

[DarkSunrise]

Originally Posted by Tor

I didn't upload any video from last time on Nürburgring GP because it's boring to watch as I was fighting with massive understeer.

Agree from the videos your setup understeers (probably the softer rear rate), but your driving style might exacerbate it as well.

At the 35 second mark, it looks like you overbrake for the corner, then are full throttle before the apex. All the weight transfers off the front and the front pushes wide on exit. You might find it understeers less if you carry more speed into the corner, lift a bit (or trail-brake) to tuck the nose into the apex, then get on the power from apex to exit. I think that will allow you to carry more speed out of the corner.

BTW nice car control skills in the wet driving video! Couple of good saves in there.

[nikitopo]

Originally Posted by Tor

I agree to a certain extent. Though I think it's a matter of personal preference and goals.

The fun factor may be just as high with a stock car, or at least without worrying about setup and just drive with whatever you get. In this case you are right, that one can focus on driving technique and learn to work around behavioural shortcomings of the car.

But for me part of the fun is optimising the car so that it behaves the way I want it and to make it as fast as possible with what is possible with the setup. In the meantime I work on my technique as well to work with the setup the way it is at the moment. The video I posted are all clips of trying to get the car to stop understeering, which succeed to some extend. In the Dunlop curve (the cambered 180), if I just put my foot down the car would plow. By modulating the throttle I got it to rotate. A mid corner left foot dap on the brakes might have done the same but I was too chicken to try.

The combination of using different techniques and adjusting the car is what I find fascinating. Lap times in itself is not so interesting, but a way to measure if what is done different works or not. This is in part why I have little interest in the Nordschleife, because it's so long and, I think, has so many variable that it's difficult to make back-to-back lap time comparisons.

Although we won't set records with the car, there is a substantial 86/BRZ community here, and there is somewhat respect credits to be gained by setting a good lap time.

Yeah but with tire pressures you will get lost. At least this is what happened to me. Too many parameters for a human being. Professional drivers have a whole team of mechanics to think all these details.

Maybe you can use the method of a needle type thermometer. Instead of doing an indirect measurement (tire pressure), you make a direct measurement (tire temperature). You measure after the race the middle and the sides of the tire. If the temperature difference is not very large, then you were OK with the pressures. If you see big difference, then it means that the tire was over- or under- inflated and adjust the next time.



You said that you have the AD08R tires (I have the same). They are good tires and can withstand heat and pressure. I don't think you'll need to adjust pressures a lot and you'll focus afterwards more on driving

[Tor]

Originally Posted by DarkSunrise

Agree from the videos your setup understeers (probably the softer rear rate), but your driving style might exacerbate it as well.

Thank you, that is very useful. Nürburgring GP is pretty difficult (for me) because it has a lot of camber and elevation changes and the track is so wide, that it's difficult to find the right line. Zolder in comparison is narrow and flat and, dare I say, easy in comparison.

In many turns, I turn in too early too so there is definitely lot to improve.

Originally Posted by nikitopo

Yeah but with tire pressures you will get lost. At least this is what happened to me. Too many parameters for a human being. Professional drivers have a whole team of mechanics to think all these details.
...
I don't think you'll need to adjust pressures a lot and you'll focus afterwards more on driving

@nikitopo , I think it's as complicated as one want's to make it. At Zolder I just set a desired pressure, like 2.1 bar. Then drove 3 rounds and went into the pit. Now one tire may have 2.3 another 2.2 etc. Then I would lower them back to 2.1 so they were all the same again and go back out.

On Nürburgring GP, I guess I'm heating up the left front significantly more than the other tires, so it would need a lower cold pressure.

I also tried to take temperatures, but with a laser thermometer it was too inaccurate to be useful. A tyre pyrometer is on my wish list though.

And once I am in the car, I guarantee I don't think about the temperature or pressure.

Originally Posted by Futaba

Did you end up needing the reflash adapter for the Tactrix OBD2 to flash your ECU on the 86?

@Futaba , no adapters are necessary for the 86/BRZ.

[Tor]

Originally Posted by nikitopo

Professional drivers have a whole team of mechanics to think all these details.

Or you get one of these to reduce the number of hands required:



It's used for 1/2 the price new. I get it sent for approval. Pay afterwards or send it back if I'm not happy. Seems like a good deal.

[Tor]

Got it today! I am so excited I have to share a picture.

Works perfectly and the condition is good. Looks like it was used a lot, which doesn't bother me - it's an instrument, not a showpiece.

[Twinguin]

This is seriously cool! Let us know, how it works in daily (or rather trackday) usage!

[Tor]

Well, this car is turning into a science experiment.

I just came home from Nürburgring GP circuit. And I have so much data to look at that I might have to call in sick tomorrow. Work is really an annoying thing that gets in the way of all fun.

One stint on the GP circuit (about 15 mins):

Tires:
@Twinguin this thing is the sh!t.

Tire pressure before and after (actual cold pressure was 2.06 Bar, as it the tires cooled to ambient 26 deg C before going on the track) - snowflake is cold, sun is hot:



Not surprisingly the left-hand side tire pressures went up more.

Inside temperatures:



Middle temperatures:



Outside temperatures:



To be honest I didn't even look at it in details yet. 2.45 Bar (35 psi) hot is probably a bit too high. I didn't even look at the temperatures yet. Any opinion?

Edit (looked):
2.45 (35 psi) doesn't look too bad, I think. And is not completely out of line with what Yokohama recommends (32-34 psi).

Hot, taken less than a minute after coming off track:

Left front     Right front
2.45 Bar       2.40 Bar
O  M  I        I   M  O
67 63 64       63 60 60

Left rear      Right rear
2.45 Bar       2.40 Bar
O  M  I        I   M  O
60 58 62       60 58 57

I, inside. M middle. O, outside

Right-side looks like near perfect. Left-rear needs a bit more pressure to get rid of the middle 58 deg. Left front overheats on the outside - now what the hell do I do about that? @MaximeT , help please!

Maybe I need more camber up front (have -2.5 deg)? But what will happen to the right-front then? I could try to dial it up to -2.8 or -3 deg (max avaiable -3.5 deg). Probably it will make it less understeery too (more about that below).

Engine:

http://datazap.me/u/tor/track-0?log=...lo=16-27-34-37

My own interpretation, with explanations:

IAM (Long-term timing being pulled) = 1 or 100%, meaning no timing is being pulled
FBKC (instant timing being pulled due to knock) = Negligible. 10 events total with the highest being -1.4 deg as I get back on the throttle (tip-in, normal).
FLKC (what causes IAM to drop) = Generally less than -1 (which is good), in fact, mostly less than -0.5. Two or 3 times it goes below -1. I think this is normal when tracking too, and it doesn't cause IAM to drop (good).

I think the timing looks pretty robust for track usage, with the fuel I am using (Shell V power Euro 100 RON octane - 93 US I think?).

LTFT ("Long term fuel trims") are fuel the ECU is adding based on learning over time. At low throttle angles, the engine uses the Lambda probe to compare the fuel/air mixture to what is desired. It then stores the average fuel trims and applies them when aggressive input throttle is being used.

My long term fuel trims are a bit high above 5000 rpm causing the engine to run a bit richer than desired. And it means my "Mass Air Flow scale" (the calibration of the sensor that measures how much air is ingested) is slightly off (Still, it's a lot better than it was stock!!).

I calibrated it already. But I think 2 things were not optimum.
1) The logs used for the calibration was done at a low 13-15 deg C (German summer lol).
2) I think I smoothed them too much on the lean side. Which is now causing the ECU to add extra fuel.

So, I took the drive home as an opportunity to make 100 km of logging at a suitable 20-25 deg C ambient temperature, so I can try to get the calibration right. That's another set of data to look at and work on.

Driving!:
The important part. Of course, I took video too. I had a great free run all laps, which I did consistently around 2:40 (2:39,5 best, 2:41,77 worst). The times are not really great, but I am very satisfied anyway because I changed the way I attacked the corners a lot. I braked a lot deeper into the corners and went on low maintenance throttle and waited until after the apex before putting the foot down. It seemed to work wonders for my previously too throttle-aggressive self-induced oversteer. So I guess hitting the same lap times as before is a good beginning. When I have time, I will edit and upload the video.

Edit: Brakes:
The XP10 pads worked wonderfully. I did the pedal dance this time again and it really feels more predictable. It's a pain in the ass to do it. I have to start the car several times to get it to work after the engine has cooled down slightly when waiting to go out. I might have to get a PDNanny.
Of course, I took temperatures too with a laser thermometer (after the tires). Both front and rear disc were around 170 deg C. The last 1 1/2 lap I didn't go full speed on the straights and didn't brake that hard. I kept the speed up in the corners though to keep the tire temperatures.

[Twinguin]

If you call that a science experiment, then it's a damn cool one! Like!

Also, on the GP track I never like that they tell you right at the last corner when time's up. No real time to have a cool down lap for the brakes then. ANy problems with that when you stapped to take the tyre temperatures?

[Tor]

Yes, that is pretty annoying. Some sort of "last lap flag" would be nice.

As I was bedding the rear pads on the same occasion, I started cool down on what I thought would be the last lap. I must have gotten the time wrong or maybe they were a bit generous because I ended up doing two cool down laps. I still took the corners as fast as I could but accelerated and braked less. Hence, I came in with hot tires but cool brakes.

[Wayno]

The reason LTFT changes is because the DI:PI ratio changes. LTFT is doing exactly what it needs to do to compensate for the DI:PI split. It's so high because you messed with the maf scale in areas that CL hits. You would have been better off keeping the Stg1 maf scale and only touching OL above 3.3V, you could even let the tool spit out a maf scale for you,which you can't do for that scale you have now.

http://datazap.me/u/tor/tor-102-2?lo...zoom=2335-2459

You can revert to this one, and use the exact values the mafscaling tool provides above 3.6V, which would produce something like this.

64.7
83.0
107.3
127.0
144.6
175.8
205.7
247.5
305.0

Another delusion people will have you believe is that AFR has to follow commanded AFR, so you need a MAF scale tuned to an inch of its life and lumpy as fuck. You don't.

Also, LTFT at idle is 100% absolutely irrelevant. If you give people logs when the car is idle and LTFT is X, people are going to look and go "LTFT is high", and make you focus on irrelevant stuff that wastes your time.

If you want to see perfect LTFT in OL (where you spend 100% of the time on track), go back to your very first logs which the vanilla Stg1 MAF scale.

http://datazap.me/u/tor/ww1082-stg-1?log=0&data=19-34

Take the tune attached, drive 100km, then log on the street and pull timing out where needed, keeping it absolutely smooth in the process. Ignore knock on the track, there's no point taking timing out based on track data, you'll just de-tune your car that way. Don't just blanket remove timing on the whole map when there's no knock or knock only in one area.

Here's a log on of Stg1 on MT with regular DI/PI ratios

http://datazap.me/u/ashenfall/stg1v1...&zoom=862-1101

Here's a log of Stg1 on AT where I must have been running with 100% DI, but same MAF scale, same AFR as 20% PI, but different LTFT which makes it happen.

http://datazap.me/u/wayne/204-stg1-9...zoom=1845-2048

Here's another one, it apparently adjusted and leaned out after this.

http://datazap.me/u/primetime/log-14...9&zoom=513-672

Another MT

http://datazap.me/u/lipton54321/stg1...zoom=4592-4786

Another AT

http://datazap.me/u/ebonyfox/98-stg1...zoom=9158-9512
http://datazap.me/u/ebonyfox/stg-1-9...12&zoom=59-237


You want to be hitting around 11.7 on OFT or 11.6 on tactrix at 7000 rpm, +- 0.1.

Running leaner than say 11.8 is great for the street, for one pull. Shit on track.
Your fuel is only E5 so not really enough E% for any great cooling benefit.

[Twinguin]

Originally Posted by Tor

Yes, that is pretty annoying. Some sort of "last lap flag" would be nice.

[ ... ] I must have gotten the time wrong or maybe they were a bit generous because I ended up doing two cool down laps. [ ... ]

That's what I noticed as well: Sometimes they let you drive a few minutes more which results in 1 or even 2 more laps than normally. If they would at least tell you on the straight, that would be nearly a full lap to cool down. I mean, the guy at the pit lane exit could just wave a chequered flag! Just to quote Mr J. Clarkson: How hard could it be?

[Tor]

Thanks a lot for the explanation Wayne and for the custom ROM.

Messing with the MAF was quite a good learning experience. I might give it another try just to see what happens. If it all goes haywire, I'll revert to the MAF you provided.

Regarding AFR and timing these are my concerns:

Originally Posted by Wayno

If you want to see perfect LTFT in OL (where you spend 100% of the time on track), go back to your very first logs which the vanilla Stg1 MAF scale.

http://datazap.me/u/tor/ww1082-stg-1?log=0&data=19-34

The 108.2 98 octane tune ran great! Perhaps the best of the tunes so far. Even though the LTFT in log above may not be quite representative of where it would have ended up though. As due to my noobness, I logged with only 50 km after flashing.

Still, I think running lean maybe produced the best performance of the tunes I flashed so far and the engine itself seemed happy. The "rich.au" MAF scale felt like it had the least power, even with more timing, for comparison.

What concerns me about running very lean is the already very hot engine bay and the heat of the header and especially the cat potentially burning ignition coils while tracking. I assume the EGT will rise when going leaner but don't know if it's a substantial difference?

Your fuel is only E5 so not really enough E% for any great cooling benefit.

Actually, I sent an email to Shell and asked. V-power is E0. Apparently, I got it wrong with the 5%. It's a quota where they have to sell 5% E on average. So all the 95E10 sales makes up for the the 100E0 V-power.

Take the tune attached, drive 100km, then log on the street and pull timing out where needed, keeping it absolutely smooth in the process. Ignore knock on the track, there's no point taking timing out based on track data, you'll just de-tune your car that way. Don't just blanket remove timing on the whole map when there's no knock or knock only in one area.

Thanks a lot for putting this ROM together for me!

Regarding timing, I also tried flashing the 102 table already, and it had FLKC above 6000 rpm.
http://datazap.me/u/tor/custom-maf-s...&data=19-26-27

I also need some compromise as there are tracks where I can only get 98 octane. So I’d rather have the timing slightly on the conservative side. Also, I don’t mind losing a bit street performance for keeping IAM=1 on track.

I think I mentioned in another post that I already have more timing than your 100 octane tables. I actually made a mix of the 98, 100, 102 table.
Below 3600 rpm, the 98 octane table
From 4000 to 5600, 100 octane table
Above 5600 it's the 102 table, but with 0.35 less timing.

This is how it looks, compared to the 102 table:

[Tor]

Originally Posted by Twinguin

That's what I noticed as well: Sometimes they let you drive a few minutes more which results in 1 or even 2 more laps than normally. If they would at least tell you on the straight, that would be nearly a full lap to cool down. I mean, the guy at the pit lane exit could just wave a chequered flag! Just to quote Mr J. Clarkson: How hard could it be?

That's actually a very simple solution to the problem. I wonder if it would be possible to put up as a suggestion somewhere.