GR 86 torque curve for Assetto Corsa
 

I wanted to virtually compare the on-track performance of 2022 against 2017-2020 and 2013-2016.
To achieve that, I've decided to create a virtual '17 and a virtual '22 based on the '13 that comes in the game.
I'm assuming the main differences are just the torque curve and the final drive, and that the final drive of 2022 is the same as 2013-2016.

Here's a screenshot from https://acstuff.ru/u/torque-helper/ vs the torque curve in the in-car display:

https://i.imgur.com/tp2WTFl.png

Here are the numbers I entered into that tool:

Here are the same numbers, subtracting 13% for drivetrain loses (I don't know why the tool doesn't allow that),
which matches what AC did to the 2013 Toyota 86 compared to the spec sheet:
These are the values I'd enter into power.lut for this car.

Is that torque curve official? There’s a strange drop after 3k rpm

This is the same torque curve that the real car shows in the instrument cluster.
It's been discussed on multiple other threads before.
Good news is that the new torque dip is much smaller than the old one.

Looks really close, wouldn't be more than 1 ot 2 nm off I would guess!

I think potentially there is a slightly sharper curve down after 7k than is what is represented on the image, but I might be imagining it because your running it through to 0 at 8.

Is that forced in AC, ie do you have to run it our to 0 at the top gearing?

Thanks!
IMO these numbers might be interesting even for folks who only have general interest in the car, but no interest in AC.

Don't know if it's strictly required, but this is how all other configs I looked at have.

I found out that RaceChrono has a mode that allows you to log data from Assetto Corsa
(RaceChrono can't stop amazing me), so I couldn't help but log some data.

I did a few laps at Laguna Seca in the '13-'16 GT86 that comes with in the game.
According to my calculations, the built-in torque curve produced ~167 wheel hp,
which is typical for these cars irl.
Car setup:

• -1.7º camber all around (tweaked the car config to allow such camber adjustments)
• 0 toe front, 0.1 toe rear per side
• Street tires
• TC off
• Otherwise default
• Track conditions: "Ideal"

My opt lap in the GT86 was 1:48.63

Then I also did a few laps in the same car with the same setup, but with
the torque curve from the earlier post (~197 wheel hp, matches what Toyota folks say).
My opt lap in the "GR 86" was 1:46.05

I can certainly polish my laps slightly, but I don't think I can
easily shave off more than a few tenths from these opt laps.

Here's the comparison between speed traces of the best laps in each car:
https://i.imgur.com/h9GLUnS.png

If you trust such a testing approach, looks like we'll be
seeing 4–5 mph of difference at the end of longer straights,
and ~2.5s of difference per lap at Laguna Seca.

It's important to note that my 1:48.63 lap in the GT86 is over
2 seconds faster than the 86DC Stock class record (1:50.78).
This can be explained by many factors, such as:

• The virtual model of the track is different (fewer bumps?)
• The virtual track had "ideal conditions"
• The virtual Street tires generate more grip than GT Radial SX2

As another data point, I'm seeing corner speeds in the sim 1–2 mph
higher than my car does irl in the 86DC Street class.
(wider tires than 86DC Stock, coilovers)

Having said that, I think the "4–5 mph speed, 2.5s per lap" difference
is the right ballpark.

Using your power.lut values, I got a similar differential in times, although my speeds were slower (I used an otherwise-unmodified car, street tires, TC on, `fast` track conditions).

I thought assetto corsa used some of the truest representation of many tracks since they used laser scanning, albeit not sure how much for laguna. Maybe better times could also be due to having less fear with higher cornering speeds. I know I push my speeds higher in assetto than I would in rl...of course many runs have resulted in wrecks lol

Very cool. Thanks for the info!

There's only so much you can do with laser scanning.
You can't laser scan how grippy the surface is :)

Also tracks change. Notably, the inside of the first apex, and the right runoff area in the
downhill part of The Corkscrew are now more drivable irl than what it was a few years ago.
The track in AC doesn't have those, so I presume it was laser scanned before those changes.

I've heard CotA is getting bumpier each year, that's another example of the limitation of laser scanning.
Don't know to what degree it applies to Laguna.

I'm trying my best to drive the same way in AC as I would in real life, leaving roughly as much "safety runoff".
Definitely not enough difference to explain multiple seconds of difference.