Cars Honda NSX 3.6.5

[Trava]

NT01 tires in the S2000 video have the same temperature rating (A) as the RE010, whereas traction rating is greater for the NT01 which should come as no surprise.

[6:31 AM]
Arch:

Source used for temperature vs mu curve is primarily "Influence of Tire Core and Surface Temperature on Lateral Tire Characteristics" by Christian Angrick, Sebastiaan van Putten, and Günther Prokop. You are free to provide better data for similar tires.

[6:36 AM]
Arch:
Add to the post
"RE71R also has a "temperature rating" of A and peaks at around 53c, which is lower than the current RE010 data in the mod. I would not take it very seriously. Most tires like this peak around 60c or so."

 

[Some1]

I would argue that even Senna himself couldn't drive the NSX in Assetto Corsa on the limit like in real life, not because the NSX is inaccurate, but because he's missing important feedback from real life g-forces, for example (butt-feeling?). Unless he has a kickass simulator setup?
That, and the fact that he is, sadly, dead.

 

[gobes]

Referenced study shows peak friction occurs at 65C to 75C. However the coefficient of friction occurring at 60C is equal to the maximum temperature measured in the study, ~115C. Therefore the tires should be able to produce the same lateral force at 115C as they do at 60C, and lose grip gradually after their 'peak'. Regardless of the data gathered, the car still does not perform equally in-game as it does in real life timed scenarios such as BM Tsukuba laps.

 

[Chris Kennedy]

hadn't tried type s until today (lots of km in 92r, 94r on nordschleife and normal nsx on LAC, high force etc). it does seem a handful to the point of not being enjoyable on nordschleife, but still managed to get to galgenkopf with a few errors and then tried pushing on the exit and lost it, but eh, the other variants drive amazing for me, most engaging and enjoyable cars in AC. Going straight in with the type s on one of the most challenging circuits without having spent a lot of time with the other variants seems like its not going to end well if the type s has a reputation irl anyway. To say that all variants of this mod lose grip suddenly and cant be caught is ridiculous, the 92r can be driven just like the irl video on nords and you can inutitively countersteer with your foot still on the gas which wasn't true of the previous version of the mod.

was definitely plenty of feedback through the wheel (g27), was making small corrections at flugplatz, schwedenkreuz at the top of 4th, low end of 5th no problem. i am using increased tyre volume on these cars which i am sure is helpful, i made a post with how to do it without messing up other cars here: https://www.racedepartment.com/threads/honda-nsx.98746/page-37#post-3354241

 

[Trava]

Referenced study shows peak friction occurs at 65C to 75C. However the coefficient of friction occurring at 60C is equal to the maximum temperature measured in the study, ~115C. Therefore the tires should be able to produce the same lateral force at 115C as they do at 60C, and lose grip gradually after their 'peak'.

[9:54 AM]
Arch:
"Here is the webplotdigitizer output for that data, the dropoff above peak is higher than you are claiming"
20,062739331556532; 0,6456575682382132
29,32291762399289; 0,7419354838709675
39,96556689029808; 0,8392059553349873
49,628404252060164; 0,904714640198511
59,97500229758295; 1,0009925558312653
70,52991452991459; 1,0535980148883373
79,91508133443617; 1,0436724565756823
88,9066568636461; 1,027791563275434
99,77673620684374; 1,0039702233250618
108,77149771773426; 0,975186104218362
119,84229390681; 0,9384615384615382

"The current data, which has been adjusted to take the difference in surface and carcass temperature into account, still has more grip at above-peak temperatures than the test is indicating. Generally the high temperature values will be relatively accurate as the carcass and core temperatures equalize, so if anything, the current model has too much grip above peak. FWIW the data lines up nicely to some graphs from Toyota as well."

 

[Trezoitao38]

Referenced study shows peak friction occurs at 65C to 75C. However the coefficient of friction occurring at 60C is equal to the maximum temperature measured in the study, ~115C. Therefore the tires should be able to produce the same lateral force at 115C as they do at 60C, and lose grip gradually after their 'peak'. Regardless of the data gathered, the car still does not perform equally in-game as it does in real life timed scenarios such as BM Tsukuba laps.

The problem seems to me more related with that. First time I pick any of the NSX the first kms I think the mod is ok. But, when the tyre gains heat just a small fraction higher the car become instable.

So, what causes most the problems in Nordschleife is lenght of the track. Final sections become a nightmare to drive the car. Specially long corners.

It is more obvious in Type S-Zero, but occurs the same in the others cars in a smaller proporction.

I've tried yesterday other road cars of Assetto Corsa, F40, BMW M3 E92 and the mod of Porsche 993 Carrera in Suzuka, all of them with street tyres, the less gripiest one. The F40 looses control when the turbo kicks in, but I can feel the exact moment when happens and react to counter steering properly. I still can do the same with all NSX, include S-Type, but when the tires warm up just a little bit more, the car suddenly looses control.

This mod reminds me of the discussions about GTR 1. When, as it is a simulator, many people with no experience with real cars expected it to be difficult, as in general many simulators of the time were. GTR 1 had a physics problem when the car lost traction on the rear wheels. Some real GT drivers at the time who used the simulator to training activated traction control to make the simulator more realistic. In the sequel, GTR 2 corrected precisely this error and changed the perception of many people about what a racing simulator is. Discussions similar to this have also occurred in earlier versions of Iracing and other simulators.

Maybe for NSX modders it's better to leave this car aside for now. And test these physics features on other cars and see how they react. How would they react to Nascar cars making long oval turns? This would give an idea of how the parameters of this physics work.

 

[D-Fazer]

The problem seems to me more related with that. First time I pick any of the NSX the first kms I think the mod is ok. But, when the tyre gains heat just a small fraction higher the car become instable.

So, what causes most the problems in Nordschleife is lenght of the track. Final sections become a nightmare to drive the car. Specially long corners.

Also with this track in real life, you can lost your brakes before the end of the first lap, maybe 2 laps if you're good enough with brake management

Just because these old sports cars have steel caliper that make brake oil boil

I've experienced it by myself

Be happy to don't have that feature also because you'll be even more disapointed with a road car with OEM tyres on race track..

 

[mclarenf1papa]

The problem seems to me more related with that. First time I pick any of the NSX the first kms I think the mod is ok. But, when the tyre gains heat just a small fraction higher the car become instable.

So, what causes most the problems in Nordschleife is lenght of the track. Final sections become a nightmare to drive the car. Specially long corners.

It is more obvious in Type S-Zero, but occurs the same in the others cars in a smaller proporction.

I've tried yesterday other road cars of Assetto Corsa, F40, BMW M3 E92 and the mod of Porsche 993 Carrera in Suzuka, all of them with street tyres, the less gripiest one. The F40 looses control when the turbo kicks in, but I can feel the exact moment when happens and react to counter steering properly. I still can do the same with all NSX, include S-Type, but when the tires warm up just a little bit more, the car suddenly looses control.

This mod reminds me of the discussions about GTR 1. When, as it is a simulator, many people with no experience with real cars expected it to be difficult, as in general many simulators of the time were. GTR 1 had a physics problem when the car lost traction on the rear wheels. Some real GT drivers at the time who used the simulator to training activated traction control to make the simulator more realistic. In the sequel, GTR 2 corrected precisely this error and changed the perception of many people about what a racing simulator is. Discussions similar to this have also occurred in earlier versions of Iracing and other simulators.

Maybe for NSX modders it's better to leave this car aside for now. And test these physics features on other cars and see how they react. How would they react to Nascar cars making long oval turns? This would give an idea of how the parameters of this physics work.

Nascar tires operate at much higher temperatures, so it’d be fine.

 

[mclarenf1papa]

Referenced study shows peak friction occurs at 65C to 75C. However the coefficient of friction occurring at 60C is equal to the maximum temperature measured in the study, ~115C. Therefore the tires should be able to produce the same lateral force at 115C as they do at 60C, and lose grip gradually after their 'peak'. Regardless of the data gathered, the car still does not perform equally in-game as it does in real life timed scenarios such as BM Tsukuba laps.

I have no idea where you got that interpretation from, but it is wrong.

The tires have 9% grip loss at 115C, that's a very, very noticeable amount.

 

[Mascot]

Not sure if it's been mentioned before but the speedo needle soon maxes out its travel before top speed is reached. A quick search suggests that the real dial scales to 280kph but on this car it only goes up to 180kph.
(Hondas - haven't checked the Acuras)

 

[D-Fazer]

Not sure if it's been mentioned before but the speedo needle soon maxes out its travel before top speed is reached. A quick search suggests that the real dial scales to 280kph but on this car it only goes up to 180kph.
(Hondas - haven't checked the Acuras)

View attachment 615441 View attachment 615442

Japanes car speedo is like that...

You can see the same on the R34 GTR and tons of jdm cars

 

[Mascot]

Japanes car speedo is like that...

You can see the same on the R34 GTR and tons of jdm cars

Ah, OK - thanks. I thought there might be a simple explanation. Not a big JDM follower.

 

[gobes]

I have no idea where you got that interpretation from, but it is wrong.

View attachment 615430
The tires have 9% grip loss at 115C, that's a very, very noticeable amount.

Is it a very, very noticeable amount? From 40C to 60C the difference in friction occurring is greater than that between 60C and 115C, yet from driving the car in-game you wouldn’t realize this. The sudden drop-off in grip under higher temperatures (90C) and load doesn’t seem to line up with my interpretation of the data in the study.

Perhaps I am wrong in my conclusion due to lack of skill or knowledge, but the Arch tire physics are very different from all the others out there both in a good (data / research) and bad (actual performance) way. Maybe in this case worse performance is better?

 

[Trava]

Is it a very, very noticeable amount? From 40C to 60C the difference in friction occurring is greater than that between 60C and 115C, yet from driving the car in-game you wouldn’t realize this. The sudden drop-off in grip under higher temperatures (90C) and load doesn’t seem to line up with my interpretation of the data in the study.

[3:49 PM]
Arch:

Actual data
70,52991452991459; 1
79,91508133443617; 0.990
88,9066568636461; 0.975
99,77673620684374; 0.952
108,77149771773426; 0.925
119,84229390681; 0.890

In game
70 | 0.998
75 | 0.995
80 | 0.99
90 | 0.9725
100 | 0.95
110 | 0.9225
120 | 0.8925

The supposedly more correct carcass temp adjusted result came out with lower grip above peak, but I have tweaked the curve to be closer to the experimental data based on the assumption that carcass temperature will have stabilized at higher temperatures. The low temperature region has increased mu due to the more data-accurate curve being very divergent in practical testing. I may still revise this as the model has changed.

Actual data, under peak
20,062739331556532; 0.612
29,32291762399289; 0.704
39,96556689029808; 0.796
49,628404252060164; 0.858
59,97500229758295; 0.950

In game, under peak
0 | 0.7
10 | 0.77
20 | 0.84
30 | 0.9
40 | 0.95
45 | 0.97
50 | 0.9825
55 | 0.9925
60 | 0.997

This can also be a factor causing some momentary handling issues if the front is under peak but rear is over it, although in most cases problems arise when both axles are above peak already. I also had some issues with the experimental performance of so low grip under peak. I'll revisit it.

 

[SpaceCadet]

wow u can drive a car thx for the info

I like how you are a fan of arch physics, then mclaren f1 papa says that arch tires are complete unacceptable **** and you like well now i agree with this message after a bunch of people said the same thing and you told them how bad drivers they are and so on

it's just a matter of time before you die in the s zero
after a few laps on the tsukuba, the car literally starts to just turn around at low speed for absolutely no reason just from turning the steering wheel 40 degrees in first corner and you all go on talking nonsense knowing that it drives like crap

it's also funny to see that apparently someone thinks that to drive this car you need to be a professional racer
as someone with irl time attack experience I can tell you for sure it's not some god's work to control oversteer even in a mid-engined car and in general there is no problem with this, the problem is the behavior of the tires under lateral load but these are just my thoughts

Despite everything, I'm glad that this mod exists and I get some driving pleasure from it.

 

[gobes]

[3:49 PM]
Arch:

Actual data
70,52991452991459; 1
79,91508133443617; 0.990
88,9066568636461; 0.975
99,77673620684374; 0.952
108,77149771773426; 0.925
119,84229390681; 0.890

In game
70 | 0.998
75 | 0.995
80 | 0.99
90 | 0.9725
100 | 0.95
110 | 0.9225
120 | 0.8925

The supposedly more correct carcass temp adjusted result came out with lower grip above peak, but I have tweaked the curve to be closer to the experimental data based on the assumption that carcass temperature will have stabilized at higher temperatures. The low temperature region has increased mu due to the more data-accurate curve being very divergent in practical testing. I may still revise this as the model has changed.

Actual data, under peak
20,062739331556532; 0.612
29,32291762399289; 0.704
39,96556689029808; 0.796
49,628404252060164; 0.858
59,97500229758295; 0.950

In game, under peak
0 | 0.7
10 | 0.77
20 | 0.84
30 | 0.9
40 | 0.95
45 | 0.97
50 | 0.9825
55 | 0.9925
60 | 0.997

This can also be a factor causing some momentary handling issues if the front is under peak but rear is over it, although in most cases problems arise when both axles are above peak already. I also had some issues with the experimental performance of so low grip under peak. I'll revisit it.

Data does line up with how it handles in game with greater grip drop-off from 80-90C. Just seems too steep of a change and I am curious how this conclusion was reached specifically for the RE010.

 

[Trava]

Data does line up with how it handles in game with greater grip drop-off from 80-90C. Just seems too steep of a change and I am curious how this conclusion was reached specifically for the RE010.

[5:15 PM]
Arch:
The tires are from the same manufacturer, with a similar treadwear rating, tread pattern and use-case. FWIW I also have some data from Toyota that shows much the same for a similar tire, although a lower peak.

 

[gobes]

Ok, I’m satisfied with your answers. Obviously a lot of research and time and testing went into this so one can only presume it’s level of accuracy is very high. I was wrong in thinking the car should more grip and easier handling, though I wish it did for my own enjoyment. I will continue to adopt a new driving style for this car and follow for future updates. Thanks Trawa for forwarding the messages from Arch.

 

[Trava]

and you told them how bad drivers they are and so on

it's just a matter of time before you die in the s zero
after a few laps on the tsukuba, the car literally starts to just turn around at low speed for absolutely no reason just from turning the steering wheel 40 degrees in first corner and you all go on talking nonsense knowing that it drives like crap

Watch local idiot skill issues racedepartment | Streamable

Watch "local idiot skill issues racedepartment" on Streamable.

streamable.com

S Zero, default setup except -5psi rear tires. I'm not sure where all this "car is impossible to catch and undrivable" complaining comes from.

Ok, I’m satisfied with your answers. Obviously a lot of research and time and testing went into this so one can only presume it’s level of accuracy is very high. I was wrong in thinking the car should more grip and easier handling, though I wish it did for my own enjoyment. I will continue to adopt a new driving style for this car and follow for future updates. Thanks Trawa for forwarding the messages from Arch.

[6:26 PM]
Arch:
I'm always trying to improve the model and will continue revising it based on available data, but I think a paradigm shift is needed for something like this. The heating model is much more accurate and different from a typical consumer sim one. The Tyres app does not tell the big picture; grip is split between carcass (not shown) and surface, while core is a heatsink and controls pressure. tyres_app_ext can display carcass temperature.

 

[mclarenf1papa]

Is it a very, very noticeable amount? From 40C to 60C the difference in friction occurring is greater than that between 60C and 115C, yet from driving the car in-game you wouldn’t realize this. The sudden drop-off in grip under higher temperatures (90C) and load doesn’t seem to line up with my interpretation of the data in the study.

View attachment 615460

Perhaps I am wrong in my conclusion due to lack of skill or knowledge, but the Arch tire physics are very different from all the others out there both in a good (data / research) and bad (actual performance) way. Maybe in this case worse performance is better?

This is mainly just a balance thing. NSX is pretty heavy at the rear and is RWD, so rears will come in before the fronts (inducing strong understeer as the slope of the curve on the cold side is high). But they'll also overheat before the fronts, probably just as the fronts are coming up to optimal, which will induce a oversteer (but not as much as the understeer induced at the start). Might have to dynamically adjust your expectations of the car as you do more laps. You're probably getting used to the understeer caused by the cold tires, but once they heat up, you can't drive it the same way.

2% of tire grip is more than enough to swing balance one way or the other, it doesn't take much.

That said, the paper does say that due to the excitation frequency of sandpaper vs asphalt, the temperature sensitivity shown will be slightly higher than on asphalt. Toyota data was also belt-tested so same would be true.