Physics The Physics discussion thread

This statement alone proves the issue with your car....you. If you don't know the function of DCAMBER you can't claim Kunos is using fake values to provide real behavior, because you don't know what those values represent.
Instead of being so smug why don't you just tell me since you feel I am ignorant.

I am telling you that even the camber values Kunos use fo the C7 are outside the envelope or real world values used to achieve the behavior they have achieved.

Stereo was saying that instead of changing the car's actual camber measurement (i.e. leave the suspension completely unchanged from the measurements you took), that you can change the tires' optimal camber to provide the "correct" feel even with unchanged suspension. Since every tire is different and has different characteristics, this isn't an unrealistic change to make. To adjust the Camber/Grip relationship, you adjust DCAMBER 0 and 1. This formula can allow you to visualize changes (it will give you a graph of the curve...IIRC y axis is a grip multiplier, x axis is camber in radians): 1+(x*DCAMBER_0+(x^2)*DCAMBER_1)
This is totally outside of the mechanism used to adjust camber on a real chassis - in fact in real life the tire DOES stay the same - you are in fact adjusting camber of the CHASSSIS to work with the tire not the other way round.

So evidently it is... (because if the size is different that means the construction is different and therefore every characteristic of the tire, regardless of compound, will be different in some way)
Not when you install actual C7 wheels and tires on the car - so again, having to change each tire to simulate chassis changes proves my point actually.

And yes suspension points are based on the wheel center.
In AC how are the a-arm pickup points located on the chassis? Are they relative to the axle center which is relative to the CofG?
 
I checked the C7 Stingray tires and their peak grip is at -2.4 degrees so I'm back to wondering what "the feel response as expected" actually means.

Suspension is placed:
1) WHEELBASE sets distance from front to rear wheels
2) TRACK sets distance from left to right wheels
3) BASEY sets height relative to CG at 'as input' suspension height (eg. -0.1 means the wheel center is 10cm below the CG)
4) ROD_LENGTH sets actual ride height
Then every suspension pickup is relative to the center of the wheel.
 
Instead of being so smug why don't you just tell me since you feel I am ignorant.

I am telling you that even the camber values Kunos use fo the C7 are outside the envelope or real world values used to achieve the behavior they have achieved.


This is totally outside of the mechanism used to adjust camber on a real chassis - in fact in real life the tire DOES stay the same - you are in fact adjusting camber of the CHASSSIS to work with the tire not the other way round.


Not when you install actual C7 wheels and tires on the car - so again, having to change each tire to simulate chassis changes proves my point actually.


In AC how are the a-arm pickup points located on the chassis? Are they relative to the axle center which is relative to the CofG?
You just said that you had to change the tire sizes going from C7 to C5...that means the tires are different, no?

Even if they are the exact same tires, the grip/camber relationship has absolutely nothing to do with the car, it's a function of tire construction. Kunos' tires may just not be perfectly correct (which might make sense as you said the C7 also feels different than what you expect)... -though a -2.4 optimum lines up with pretty much everything I've seen, so maybe play with relaxation lengths (which seem too short on Kunos cars) or something.

And no, all pickup points are relative to the wheel center.
 
I checked the C7 Stingray tires and their peak grip is at -2.4 degrees so I'm back to wondering what "the feel response as expected" actually means.

Suspension is placed:
1) WHEELBASE sets distance from front to rear wheels
2) TRACK sets distance from left to right wheels
3) BASEY sets height relative to CG at 'as input' suspension height (eg. -0.1 means the wheel center is 10cm below the CG)
4) ROD_LENGTH sets actual ride height
Then every suspension pickup is relative to the center of the wheel.

Wheelbase is a static relation front to back with the wheels straight - however, track is quite variable.

In the case of the Corvette, the front (and rear track for that matter) track can vary by as much as 20mm - there is about 15mm of adjustment on each side via lower control arm pickups. Camber is adjusted by kicking out the lower a-arms.

So my question is at what adjustment are Kunos basing their chassis basis on? A setup with zero front and rear camber and zero toe?

And what height? A stock factory height?

To build and entire chassis using the center of the wheel as the base is really obtuse - especially when the fixed component - the chassis is using a a formula to arrive at the pickup points.

Using Kunos method there are 1000's of ways to create a baseline chassis of the very same car - the Basey number is a fixed number in the suspension.ini file - but the height of the wheel center relative to the CofG is going to depend on the current height setting of the car. This assumes a car does note have height adjustment built into the chassis.

So the next question is how are Kunos managing camber adjustments for the configured car? If the wheel camber is altered, are they unassuming the pickup points remain the same and the wheel angle just miraculously changes, as if by a magic ball joint on the end of the axle?
 
You just said that you had to change the tire sizes going from C7 to C5...that means the tires are different, no?
Yes and no. Yes for where the size is changed and no for where the tires are used as is.

Even if they are the exact same tires, the grip/camber relationship has absolutely nothing to do with the car, it's a function of tire construction. Kunos' tires may just not be perfectly correct (which might make sense as you said the C7 also feels different than what you expect)... -though a -2.4 optimum lines up with pretty much everything I've seen, so maybe play with relaxation lengths (which seem too short on Kunos cars) or something.
And this is the same as real life - when I buy Hoosier tires the tire guy tells me the tire needs -2.5* of camber to work well, he doesn't worry about the car, just the ideal camber he wants.

It would seem that if the physics engine was well behaved it should not matter what car a given tire file is applied to, no matter what size the tire - and I believe I have read that Kunos have stated tire files should be able to be applied to various any car that needs tires that size/compound.

And no, all pickup points are relative to the wheel center.
Which again seems odd and not very accurate way of doing it as the suspension file uses these fixed numbers that would be different for the very same car depending on the current attitude of the suspension.

All you have to do is jack a car up and then lower it and the wheel center changes relative to the suspension pickup points as the car settles.

It just seems this method is prone to errors in measurement.

Either way, this is the method and these are the numbers Kunos has decided to use
 
So the next question is how are Kunos managing camber adjustments for the configured car? If the wheel camber is altered, are they unassuming the pickup points remain the same and the wheel angle just miraculously changes, as if by a magic ball joint on the end of the axle?
Yes, it's effectively changing the angle the axle is mounted inside the spindle without changing the position or shape of any other components.


It doesn't matter what height you use as long as you consistently take all measurements at the same one. Stock ride height is probably the easiest for a car you don't own; since you own it, full droop might be the easiest to do consistently.
 
Yes and no. Yes for where the size is changed and no for where the tires are used as is.
I mean on the real car. Do the C5s and C7s use different tire sizes?

And this is the same as real life - when I buy Hoosier tires the tire guy tells me the tire needs -2.5* of camber to work well, he doesn't worry about the car, just the ideal camber he wants.
Well that's just a blanket statement by the salesman...a -2.5˚ static setting on one car will result in sometimes vastly different mid-corner angles in another car. Regardless, this is irrelevant to the point at hand.

It would seem that if the physics engine was well behaved it should not matter what car a given tire file is applied to, no matter what size the tire - and I believe I have read that Kunos have stated tire files should be able to be applied to various any car that needs tires that size/compound.
Kunos has not stated that. They've stated that you can use their tires as a base for yours. It doesn't mean that you will be able to directly carry over their tires. As I've said, if changing the static camber on your car in game to -3.5 vs. -2 or whatever the regular number is fixes the handling for you, you want to keep that -2 value and adjust the DCAMBER numbers to amend the handling. Otherwise you can even just adjust the tire grip coefficient or its relationship with load to affect behavior.

Which again seems odd and not very accurate way of doing it as the suspension file uses these fixed numbers that would be different for the very same car depending on the current attitude of the suspension.
All you have to do is jack a car up and then lower it and the wheel center changes relative to the suspension pickup points as the car settles.
Which is why you have the rod length values to dictate the attitude of the suspension as-measured...how are you currently calculating them?

It just seems this method is prone to errors in measurement.
It is not provided it is used correctly.

Either way, this is the method and these are the numbers Kunos has decided to use
I've commented within the quote...it's easier.
 
Which again seems odd and not very accurate way of doing it as the suspension file uses these fixed numbers that would be different for the very same car depending on the current attitude of the suspension.

All you have to do is jack a car up and then lower it and the wheel center changes relative to the suspension pickup points as the car settles.

It just seems this method is prone to errors in measurement.p

Either way, this is the method and these are the numbers Kunos has decided to use

It's great that the more you post the more you keep raising some blog's already sky high credibility. Keep it up.
Edit: sorry for OT leaving now but you are in good hands...listen to these guys and you'll probably end up with something that works
 
Yes, it's effectively changing the angle the axle is mounted inside the spindle without changing the position or shape of any other components.
I guess this is the compromise they settled for.

It doesn't matter what height you use as long as you consistently take all measurements at the same one. Stock ride height is probably the easiest for a car you don't own; since you own it, full droop might be the easiest to do consistently.

Well the C7 suspension looks to be fully lowered in the suspension app. Not so sure about their steering rack/tie rods front and rear though - i will measure a C7 but on the C5 they are closer to the axle center line than the C7 suspension layout shows in their app.
 
It's great that the more you post the more you keep raising some blog's already sky high credibility. Keep it up.
Edit: sorry for OT leaving now but you are in good hands...listen to these guys and you'll probably end up with something that works

Must be an inside joke as the significance of this statement is lost on me :) hehe
 
I guess this is the compromise they settled for.



Well the C7 suspension looks to be fully lowered in the suspension app. Not so sure about their steering rack/tie rods front and rear though - i will measure a C7 but on the C5 they are closer to the axle center line than the C7 suspension layout shows in their app.
Just checked again and wheel end of the steering rack tie rods of the Kunos C7 are off by a factor of about about 120mm and the upper control arm wheel points front and rear are off about 35mm.

Well at least the game can feel somewhat like a real car driving, even if the physics engine is using input data dissimilar to the chassis being modeled.
 
What matters for you is your car, not what the C7 has. If the C7 has wrong measurements, go and open a physics bug report in their forum, but here we are talking about your car.

If you don't have real tyre data to match, then best thing you can do is like mclarenf1papa said, adjust your tyres according to the handling you expect. If you get your desired handling at 3.5deg of camber, then make the appropiate changes so you match that handling at your target cambers.
 
What matters for you is your car, not what the C7 has. If the C7 has wrong measurements, go and open a physics bug report in their forum, but here we are talking about your car.

If you don't have real tyre data to match, then best thing you can do is like mclarenf1papa said, adjust your tyres according to the handling you expect. If you get your desired handling at 3.5deg of camber, then make the appropiate changes so you match that handling at your target cambers.

Well, the C7 does matter - as the C7 and C6 are just iterative changes chassis wise from the C5. So it would stand to reason that if the C7 was accurately modeled in AC, and the simulation engine was truly modular, then using it as a starting point to help keep my C5 heading in the right direction is appropriate.

In fact I should be able to just pluck sections of the code for the C7 and plug it right into the C5.

The problem I see (and again I may be wrong, but the more I dig the less I believe I am wrong) is that only SOME of the chassis data matters, the final "feel" comes down to tweaks - which is not really the simulators should work.

See, many C5/C6/C7 parts are interchangeable, and people do augment their C5 and C6's with upgrades and components derived from the latest model. Wheels, tires brakes and shocks come to mind.

The part that honestly bothers me is that the very chassis dynamic should rely on the a-arm movement in the physics engine, which would mean the wheel relative to the outer a-arm attachment points on the wheel spindle cannot change - the idea of a pivoting ball joint at the end of the axle to allow for dynamic camber is a crock - essentially the tire model is taking over for the wheel and suspension - the tire is the only thing that can move relative to the spindle.

The idea for using AC as the sim for the mod, is that as in the real world, the C5 (and C6) could be updated by players by adding components - known components in a modular fashion, but if each and every iteration of each and every car needs "tweaking" then it is not a modular physics engine.

And as I show, the C5/C6/C7 chassis is tunes in real life by moving the a-arm pickups - something that AC has no way of catering for.

Maybe AC was designed for BMW's and other euro "sport" cars that don't have any facility for camber adjustment in stock form.

But when it comes to a chassis that is adjustable, the engine seems to not cater for it - which is a pity :(

Of course this could all come down to the fact that Kunos have not shared enough critical documentation with regard to the engine - maybe if they did share all the critical documentation then the stupid ignorant sim community (those of us that spent money on the game and DLC) wouldn't keep getting it so horribly wrong :)
 
Last edited:
The problem I see (and again I may be wrong, but the more I dig the less I believe I am wrong) is that only SOME of the chassis data matters, the final "feel" comes down to tweaks - which is not really the simulators should work.

Look...you're wrong. To date, IER has done 6 cars with even more extensive data than what you have access to (regardless of the fact that you physically have the car). One of these cars was tested by a driver that professionally races it in real life (which in and of itself implies a much more in-depth understanding of "feel" than someone that doesn't race) and effectively said that the only difference to real life is that he wasn't sitting in the actual car driving around the actual track (and the brake pedal he was using allowed for more brake pressure than the real thing - so easier to lock the brakes). We're not questioning the accuracy of Kunos' cars...they may not have had enough data and improvised to make the car accurate based on their knowledge. We're arguing for the accuracy of the underlying physics engine, which is undeniably sound.
 
I've commented within the quote...it's easier.

Well its not easier to reply to but here goes, my reply is underlined below yours.

mclarenf1papa
RC45 said:
Yes and no. Yes for where the size is changed and no for where the tires are used as is.
I mean on the real car. Do the C5s and C7s use different tire sizes?
It does not matter - a stock C5 uses 17" and 18" wheels while the C7 uses either 18"/19" or 19"20" depending on the package - but to be clear in real life people can and do run C6 and C7 wheels as direct swaps onto C5's.
So if AC was as modular and accurate as claimed I should be able to take the entire C7 tire data set and apply it to the C5 to simulate swapping out wheels. So again, the tire size is not a factor unless the C7 tire data is fudged to make the C7 feel like a C7.



And this is the same as real life - when I buy Hoosier tires the tire guy tells me the tire needs -2.5* of camber to work well, he doesn't worry about the car, just the ideal camber he wants.
Well that's just a blanket statement by the salesman...a -2.5˚ static setting on one car will result in sometimes vastly different mid-corner angles in another car. Regardless, this is irrelevant to the point at hand.
Salesman? Who said anything about salesman? I have spoken directly to Hoosier and they have said they want -2.5 or -3.0 depending on tire - they may then offer advice that "if you are driving X-car, then a static camber of Y* will give us dynamic camber of 3.0* to work with.
Please do not assume that you are the only person that understands going fast in real life.
So I guess we need to know what data Hankook shared with Kunos for the C7 versus the DCAMBER) data that is in the tire file - I assume they used Hankook tire data since that's the texture they used.


It would seem that if the physics engine was well behaved it should not matter what car a given tire file is applied to, no matter what size the tire - and I believe I have read that Kunos have stated tire files should be able to be applied to various any car that needs tires that size/compound.
Kunos has not stated that. They've stated that you can use their tires as a base for yours. It doesn't mean that you will be able to directly carry over their tires. As I've said, if changing the static camber on your car in game to -3.5 vs. -2 or whatever the regular number is fixes the handling for you, you want to keep that -2 value and adjust the DCAMBER numbers to amend the handling. Otherwise you can even just adjust the tire grip coefficient or its relationship with load to affect behavior.
So you are agreeing then that the AC sim engine needs to each tire tweaked per car even if the tire is the SAME TIRE because the sim engine is not very good?
That's fine - just say that AC needs fudged numbers to work - no shame in that, but don't pretend it all accurate.


Which again seems odd and not very accurate way of doing it as the suspension file uses these fixed numbers that would be different for the very same car depending on the current attitude of the suspension.
All you have to do is jack a car up and then lower it and the wheel center changes relative to the suspension pickup points as the car settles.
Which is why you have the rod length values to dictate the attitude of the suspension as-measured...how are you currently calculating them?
What is the relationship between the Pickup_x points in the car.ini and the BaseY value in the suspension.ini?

If the rod length is calculated and the BaseY is calculated where is the Pickup point plane derived from?


It just seems this method is prone to errors in measurement.
It is not provided it is used correctly.

Either way, this is the method and these are the numbers Kunos has decided to use
 
Last edited:

Latest News

Online or Offline racing?

  • 100% online racing

    Votes: 78 7.0%
  • 75% online 25% offline

    Votes: 122 10.9%
  • 50% online 50% offline

    Votes: 164 14.6%
  • 25% online 75% offline

    Votes: 308 27.5%
  • 100% offline racing

    Votes: 444 39.6%
  • Something else, explain in comment

    Votes: 4 0.4%
Back
Top