Rf2 aero.

[Simulation_Player]

Hello fellow sim enthusiast,
I would like to discuss on how rf2 does things for its aero simulation, share my thoughts and hear your thoughts as well.
So the problem starts with RH being taken as average of left and right side, on straight line its no issues but in corners it is quite problematic and makes the setup bit counterintuitive.
for example
with stiffer ARBs the car RH at front and rear goes down and RH goes up with lower ARBs
on F1 cars RH going down at front is very good BUT if rear goes low as well this reduces rake and these cars performs best with proper rake (in my finding 30-50mm rake is good for FPRO for example).
So if i remove only the rear arb this way rear stays high (front also goes high but not as much as rear so net increase in DF), this way i don't have to compromise for low speed rear grip and high speed rake.
If you look at individual corners RH this looks very ridiculous, the RH on corner with stiffer V.S softer arbs are nowhere near same BUT as long their avg is same u get same DF.

So my thoughts, would it better if we take RH at 3 points (left , centre and right) and these points have individual effect on aero ?

 

[Lazza]

Hello fellow sim enthusiast,
I would like to discuss on how rf2 does things for its aero simulation, share my thoughts and hear your thoughts as well.
So the problem starts with RH being taken as average of left and right side, on straight line its no issues but in corners it is quite problematic and makes the setup bit counterintuitive.
for example
with stiffer ARBs the car RH at front and rear goes down and RH goes up with lower ARBs
on F1 cars RH going down at front is very good BUT if rear goes low as well this reduces rake and these cars performs best with proper rake (in my finding 30-50mm rake is good for FPRO for example).
So if i remove only the rear arb this way rear stays high (front also goes high but not as much as rear so net increase in DF), this way i don't have to compromise for low speed rear grip and high speed rake.
If you look at individual corners RH this looks very ridiculous, the RH on corner with stiffer V.S softer arbs are nowhere near same BUT as long their avg is same u get same DF.

So my thoughts, would it better if we take RH at 3 points (left , centre and right) and these points have individual effect on aero ?

It could be improved, yes. But is it really that bad? No. Would it be easy to make things worse by attempting a "simple" fix that isn't properly thought out? Yes.

 

[Simulation_Player]

It could be improved, yes. But is it really that bad? No. Would it be easy to make things worse by attempting a "simple" fix that isn't properly thought out? Yes.

I'm not saying my fix is the way to be clear, also this isn't exactly regarding roll sensitivity
i was wondering by making rh separate would make setup that are more realistic, because i honestly don't think avg rh is ok considering rf2 is heavily focused on modern aero race cars, setup takes quite unrealistic turn imho.

 

[Lazza]

I'm not saying my fix is the way to be clear, also this isn't exactly regarding roll sensitivity
i was wondering by making rh separate would make setup that are more realistic, because i honestly don't think avg rh is ok considering rf2 is heavily focused on modern aero race cars, setup takes quite unrealistic turn imho.

Do you have logs from your scenario above? Showing the effect of lower rear ARB? (ie. bigger rear ride height difference, same avg ride height; same downforce?)

Also, how much should a lateral ride height variation (or degrees roll, same thing really) affect downforce? Flat vs your high ARB roll vs your low ARB roll?

 

[Simulation_Player]

Do you have logs from your scenario above? Showing the effect of lower rear ARB? (ie. bigger rear ride height difference, same avg ride height; same downforce?)

Also, how much should a lateral ride height variation (or degrees roll, same thing really) affect downforce? Flat vs your high ARB roll vs your low ARB roll?

I have logs yes, i will post them later. ofc car is stiffly sprung even in softest setup.....so the RH gain from detached rear arb only is small 1-2 mm but this is in corners where the rake matters the most...plus u gained tons of traction in slow speeds, win win situation. eventual low speed understeer can be cured by stiffer corner springs which has added benefits of additing more rake in corners without as significant hit to mechanical grip as stiffer rear ARB.
Also how much roll affect df , no one has numbers for that....but effect is real i have confirmed via multiple sources (physics devs, real engineers ), df shouldn't be the same. it affects rear wing, front wing/splitter even more...everything.
Video done by F1 aerodynamicist :

 

[Robin Pansar]

Also how much roll affect df , no one has numbers for that.

Downforce data of no roll angle vs roll angle on a Dallara GP2 car:
0 deg roll angle:

1 deg roll angle:

In general I don't think there's much aero roll sensitivity in the game. No parameter to set it as far as I know either.

 

[Lazza]

I have logs yes, i will post them later. ofc car is stiffly sprung even in softest setup.....so the RH gain from detached rear arb only is small 1-2 mm but this is in corners where the rake matters the most...plus u gained tons of traction in slow speeds, win win situation. eventual low speed understeer can be cured by stiffer corner springs which has added benefits of additing more rake in corners without as significant hit to mechanical grip as stiffer rear ARB.
Also how much roll affect df , no one has numbers for that....but effect is real i have confirmed via multiple sources (physics devs, real engineers ), df shouldn't be the same. it affects rear wing, front wing/splitter even more...everything.
Video done by F1 aerodynamicist :

I didn't ask if the effect is real, nor suggest it wasn't.

The magnitude of the effect is important, and in a relevant context.

In general I don't think there's much aero roll sensitivity in the game. No parameter to set it as far as I know either.

There is no parameter to directly influence roll aero sensitivity.

I am of the opinion this has much less influence than it's given credit for. I've had discussions with both Sim_Player and Nieubermesch (sp?) on this in the past, but will happily keep rolling the questions out in the hope of some sort of conclusion based on something tangible...

 

[Simulation_Player]

Downforce data of no roll angle vs roll angle on a Dallara GP2 car:
0 deg roll angle:

1 deg roll angle:

In general I don't think there's much aero roll sensitivity in the game. No parameter to set it as far as I know either.

Fantastic info, thank you for posting.

 

[Simulation_Player]

I didn't ask if the effect is real, nor suggest it wasn't.

The magnitude of the effect is important, and in a relevant context.



There is no parameter to directly influence roll aero sensitivity.

I am of the opinion this has much less influence than it's given credit for. I've had discussions with both Sim_Player and Nieubermesch (sp?) on this in the past, but will happily keep rolling the questions out in the hope of some sort of conclusion based on something tangible...

magnitude of that effect will vary from car to car, again you will have to be extremely lucky to find data about it.
Robin posted a very good picture showing the magnitude, truly you don't loose 500 kgs per degrees of roll but if it can be implemented in rf2 it will surely add more realistic route in setup philosophy.
Along with mechanical reason why roll control in rf2 is not as important as IRL...would love to get that sorted as well. From what i've heard even on new tyres camber isn't as important, which ARB has effect on....so all these "small" effects add up.

 

[Lazza]

magnitude of that effect will vary from car to car, again you will have to be extremely lucky to find data about it.
Robin posted a very good picture showing the magnitude, truly you don't loose 500 kgs per degrees of roll but if it can be implemented in rf2 it will surely add more realistic route in setup philosophy.
Along with mechanical reason why roll control in rf2 is not as important as IRL...would love to get that sorted as well. From what i've heard even on new tyres camber isn't as important, which ARB has effect on....so all these "small" effects add up.

Well half this post I agree with, that's why I said it earlier (aero could be improved). I imagine this would be a new approach basically, because the component separation starts to not make sense - especially if future damage model improvements will mean different interactions front to rear of the car (not just, for example, entire front wing removal - which already can't be configured correctly in terms of downstream aero effects). Easiest to think of solution is an aero map with combinations of parts wholly/partly missing/damaged, across different scenarios. I would imagine some sort of table generation a bit like the tyres (though much simplified), based on some parameters.

I would also expect work of this sort to coincide with a licensed vehicle showing behaviour that deviates from the information provided by the manufacturer, and there's the rub - maybe the difference hasn't been enough to worry about so far?

The aero sits on top of the mechanical simulation, and I doubt (without knowing for sure) the rF2 cars capture all the trickery that happens in modern formula cars. So there's a danger of putting an uber-accurate aero model on top of a not-so-uber-accurate mechanical model - if you end up with overall performance being further from reality, is it really beneficial?

Plus, matching a virtual car's cornering performance to the real one includes any effect of aero roll. And the real car is designed around a certain roll rate, so the aerodynamicists don't maximise downforce in a straight line and roll the dice in cornering. Robin's tables above I find interesting with the sheer difference in those scenarios - but I don't know the usual working roll in those cars either.

As far as FPro, assuming something like 1700mm track you'd need something like 30mm RH delta side to side for 1° roll. So it's pretty substantial. I suspect aero performance variation with changing roll is within margin of error for the simulated car overall.

 

[Simulation_Player]

Well half this post I agree with, that's why I said it earlier (aero could be improved). I imagine this would be a new approach basically, because the component separation starts to not make sense - especially if future damage model improvements will mean different interactions front to rear of the car (not just, for example, entire front wing removal - which already can't be configured correctly in terms of downstream aero effects). Easiest to think of solution is an aero map with combinations of parts wholly/partly missing/damaged, across different scenarios. I would imagine some sort of table generation a bit like the tyres (though much simplified), based on some parameters.

I would also expect work of this sort to coincide with a licensed vehicle showing behaviour that deviates from the information provided by the manufacturer, and there's the rub - maybe the difference hasn't been enough to worry about so far?

The aero sits on top of the mechanical simulation, and I doubt (without knowing for sure) the rF2 cars capture all the trickery that happens in modern formula cars. So there's a danger of putting an uber-accurate aero model on top of a not-so-uber-accurate mechanical model - if you end up with overall performance being further from reality, is it really beneficial?

Plus, matching a virtual car's cornering performance to the real one includes any effect of aero roll. And the real car is designed around a certain roll rate, so the aerodynamicists don't maximise downforce in a straight line and roll the dice in cornering. Robin's tables above I find interesting with the sheer difference in those scenarios - but I don't know the usual working roll in those cars either.

As far as FPro, assuming something like 1700mm track you'd need something like 30mm RH delta side to side for 1° roll. So it's pretty substantial. I suspect aero performance variation with changing roll is within margin of error for the simulated car overall.

would love to see new approach done with aero on RF2 , If im not wrong rf2 various effects aero is set by 3 parameters which generates a curve, i always found that quite restricting.
In iirc AC u can set a table for every mm of RH....this to me seems better solution IMO.
I agree that we wouldn't want a superior aero solution on top of "simplistic" mechanical physics.
Roll sensitivity is good but if devs think the work is too much for the realism it brings then i'm fine with, they know way more than me how to improve every aspect of their simulation.