Tyre friction/interaction with road surface

green serpent

green serpent

Registered
Hi guys,
I know this has been discussed ad nauseum, but just wanted to put my thoughts out there. I've put around 600 hours into rf2 now, and to me at least it is the best sim - if I didn't think that I wouldn't play it so much.

However, occasionally I jump into a car and the sensation of realism isn't there. For example, the Cobra. I love driving this car, the way that it behaves is great, and in fact, it's one of my favorite cars in the sim. But to me, the ffb doesn't feel REAL.

I fired up AC yesterday after a long, long time of not playing it. Even without getting used to it at all, as soon as I drove out of the pits the connection with the car was immediate. Physics wise, I felt there was far too much grip (even with the least grippy settings), and certain other physics issues really annoy me (hence why I don't play the game very often), but jeez was the ffb great. It practically felt like I was driving a real car.

I fired rf2 back up, and as always the physics were a revelation, and the ffb was very good and communicative (I can actually drive on the limit better in rf2), but it lacked that real sensation when compared with AC. For example in AC, the ffb was very light in easy corners, and extremely strong (relatively speaking) with hard cornering. I don't seem to get as much dynamic range in rf2. It's more than that though, and I can't really put my finger on it, it felt damped and solid, and I just knew exactly where the front tires had to be pointed at all times (the path of least resistance), due to the feedback I was receiving.

I feel that this must be due to rf2 lacking ffb from the tyres. As an example, many of you will know that if you lower the caster below a certain point, you'll feel the force feedback drop off (and in some cases pull the other way, which is due to the aligning torque). But with low caster settings the ffb doesn't just drop away and go light, it's literally non-existent. Once you turn the front tyres past a certain point, there is zero ffb. This indicates to me that there is no friction between the tyres and the road surface being sent through the steering rack. Am I wrong in thinking this?

This is getting a bit long, so I guess my thoughts are, would rf2 feel more like real life if it generated it's ffb from the steering rack and the tyres combined? Would it provide a greater connection to the car and road and be more immersive? Would it make it a more realistic simulator, and is it ever likely to happen?
 
As the tires are what connect the ground to the suspension, how could the FFB even be imagined if the tire interaction is not taken into account? Maybe the solution it's much more simple than the issue : green track, hard tires, cold tires, green track, FFB settings.
 
green serpent said:
I feel that this must be due to rf2 lacking ffb from the tyres. As an example, many of you will know that if you lower the caster below a certain point, you'll feel the force feedback drop off (and in some cases pull the other way, which is due to the aligning torque). But with low caster settings the ffb doesn't just drop away and go light, it's literally non-existent. Once you turn the front tyres past a certain point, there is zero ffb. This indicates to me that there is no friction between the tyres and the road surface being sent through the steering rack. Am I wrong in thinking this?

This is getting a bit long, so I guess my thoughts are, would rf2 feel more like real life if it generated it's ffb from the steering rack and the tyres combined? Would it provide a greater connection to the car and road and be more immersive? Would it make it a more realistic simulator, and is it ever likely to happen?
Click to expand...

That bold text doesn't make much sense after what you said ... the fact that the ffb goes "zero" after certain point (slip angle) means that the ffb is generated from both things... if you remove the mechanical trail effect reducing caster, what you get is the pure self aligning torque from the tyres... and that torque can go negative after certain slip angle.
 
"But with low caster settings the ffb doesn't just drop away and go light, it's literally non-existent. Once you turn the front tyres past a certain point, there is zero ffb. This indicates to me that there is no friction between the tyres and the road surface being sent through the steering rack. Am I wrong in thinking this?"

Ok admittedly I was wrong here. With low caster settings the ffb goes light, but if you boost up the ffb multi, you can defiantly feel friction between the tyres and the road surface. If you sit in one spot or drive slowly and turn left/right, you can feel the friction. My bad.
 
green serpent said:
Once you turn the front tyres past a certain point, there is zero ffb. This indicates to me that there is no friction between the tyres and the road surface being sent through the steering rack. Am I wrong in thinking this?
Click to expand...

Do you know the correlation of Mz / Alpha (ie slip angle) vs Fy / Alpha? Peak Mz does not correspond to Peak Fy. In most cases, Peak Fy happens at a larger Alpha than peak Mz, meaning the point at which the tires are at a slip angle that provides the most grip, the self aligning torque is on a downward slope (ie past its peak).

Basically if you only drive to the "peak" ffb, assuming the tires are modelled correctly, you're not even remotely close to using the maximum grip afforded by the tires at that particular moment.
 
borbor said:
meaning the point at which the tires are at a slip angle that provides the most grip, the self aligning torque is on a downward slope (ie past its peak).
Click to expand...
Yes, I did actually know that.
I can feel the ffb drop away once I steer at a certain angle (provided the caster is low enough).

My real life experience indicates to me that while driving within the limits of the tires and steering the tyres along the path of least resistance, the feedback through the wheel is rather light. Then, during hard cornering, the feedback through the wheel becomes extremely heavy, relative to moderate cornering forces. I'm sure a better wheel would give me much better dynamic range, however even with a g27, AC does a pretty good job of replicating real life ffb.

My point is not to knock rf2 ffb, I'm simply asking the question wether rf2 would be even better if it also got ffb directly from the tyres (ie like iracing/AC).
 
green serpent said:
Yes, I did actually know that.
I can feel the ffb drop away once I steer at a certain angle (provided the caster is low enough).

My real life experience indicates to me that while driving within the limits of the tires and steering the tyres along the path of least resistance, the feedback through the wheel is rather light. Then, during hard cornering, the feedback through the wheel becomes extremely heavy, relative to moderate cornering forces. I'm sure a better wheel would give me much better dynamic range, however even with a g27, AC does a pretty good job of replicating real life ffb.

My point is not to knock rf2 ffb, I'm simply asking the question wether rf2 would be even better if it also got ffb directly from the tyres (ie like iracing/AC).
Click to expand...

Do you try to play with this parameter in your UserData/player/Controller.json (put 1 to test)?
"Steering resistance type":0, (0=use damping, 1=use friction)
 
MaD_King said:
Do you try to play with this parameter in your UserData/player/Controller.json (put 1 to test)?
"Steering resistance type":0, (0=use damping, 1=use friction)
Click to expand...
Parameters like this make me doubt about the whole FFB accuracy.

It is first stated that the FFB is calculated at the steering shaft based on the different forces acting in the wheels through the suspension and steering linkage.

Later we see this apparently distorsioning parameter where you can decide resistance type...

It would seem that the raw FFB calculation is afterwards postprocessed somehow with several controller.json modifiers in order to get a better result, which is bad decision IMO. Only several parameters to compensate for hardware limitations should be considered. Changing resistance type from friction (independent from speed) to damping (speed dependant) is unjustifiable from a simulation POV.

Sorry to repeat the same sentence in different threads but when you sit on a real car you get the FFB that you get. You cannot do anything to change it. So it should be in rF2. As I said, some multipliers can be added to suit a child's arm strength or to avoid clipping. It is difficult, however, to justify so many fiddling to FFB not based on anything except for the perception of how should it feel. FFB should be treated as another physical parameter to be measured and validated in its accuracy.
 
Last edited:
I guess I am just frustrated. I am fed up with the way the g27 feels, really would like to upgrade but just can't afford it right now, however I fire up AC and honestly it doesn't feel too bad. Perhaps low to mid range wheels struggle with the complexities and subtleties of the rf2 physically deforming model, and fair better with simpler models, giving a better feel. Would love to do a back to back comparison with a high end wheel.
 
Hi all ^_^

Just to let you know Green seprent, I had a G27, now i'm on a T500RS ... It's a new world !!!
FFB is 30% stronger, then you'll feel all the "little things" you miss with your G27. (Friction limits, tyre problems, wheel weight, the way the car weight moves from left to right in chicanes...)
I don't know if you'll be faster (i've been a bit faster with the T500), but the feeling with high end wheels is soooooo good :p
I don't even want to try a fanatec, I'm affraid that I'll be disapointed by my T500 T_T

Cheers ^_^
 
i do get that light feeling on many of the ISI cars.. especially the f2 thats just not right at all. Normally if available adding a load of caster can help. Funny enough most mods do not get this light feeling at slip angle. T300 here. For me this lies rf2 ffb issue. Otherwise its quite detailed.

p.s AC does have good ffb you can feel a lot with what the cars doing and gives good sense of weight distribution and traction loss.
 
lokidan said:
i do get that light feeling on many of the ISI cars.. especially the f2 thats just not right at all. Normally if available adding a load of caster can help. Funny enough most mods do not get this light feeling at slip angle. T300 here. For me this lies rf2 ffb issue. Otherwise its quite detailed.

p.s AC does have good ffb you can feel a lot with what the cars doing and gives good sense of weight distribution and traction loss.
Click to expand...

Feeling light doesn't mean that it's implemented wrong. Remember that a T300 still only produces less than 5 Nm peak torque whereas racing cars produce over 15 Nm. So for example the F2 car with its low caster will not get close to its peak forces in slow corners. Divide that force by three and it will feel somewhat weak even on a T300 unless you drive it on an oval track with constant high-speed corners.
 
Is RF2 meant just for high end wheels only?

No......

Why does rf2 try to simulate real world forces on a wheel when a majority of users dont have the wheels to kick out that torque?!

It just does not feel natural at all. Could compare to other sims that work amazingly with the t300. but i won't...

I see your point, but rf2 has failed already at grabbing the market it needs to be more accessible hopefully s397 has learnt that lesson.

Obviously while still keeping the great physics.
 
Last edited:
lokidan said:
Is RF2 meant just for high end wheels only?

No......

Why does rf2 try to simulate real world forces on a wheel when a majority of users dont have the wheels to kick out that torque?!

It just does not feel natural at all. Could compare to other sims that work amazingly with the t300. but i won't...

I see your point, but rf2 has failed already at grabbing the market it needs to be more accessible hopefully s397 has learnt that lesson.

Obviously while still keeping the great physics.
Click to expand...

The other sims (AMS, iRacing, AC) have the same FFB philosophy, if there are differences is likely due to the way pneumatic trail is implemented, and due to possibility to tweak the output with damping, smoothing etc.

I have a T300 too and personally I much prefer rF2's FFB over AC's, for example.
 
i agree rf2 ffb mainly is far superior.. Its just that heavy feel you would expect going into a fast or hard turn when the tyre walls are being properly tested, is when i expect some forces from the ffb - not go light?!. I think the newer cars are better like the dallara. But they all are much better with the magic caster fix.
 
lokidan said:
i agree rf2 ffb mainly is far superior.. Its just that heavy feel you would expect going into a fast or hard turn when the tyre walls are being properly tested, is when i expect some forces from the ffb - not go light?!. I think the newer cars are better like the dallara. But they all are much better with the magic caster fix.
Click to expand...

It depends on the suspension geometry. If it has little mechanical trail (usually means low caster) you get more "feeling" from the self aligning torque of the tyre. And like @borbor said, when the tyre is at peak lateral forces, the self aligning torque is on a downward slope (sometimes it can be even negative).
 
Euskotracks said:
Parameters like this make me doubt about the whole FFB accuracy.

It is first stated that the FFB is calculated at the steering shaft based on the different forces acting in the wheels through the suspension and steering linkage.

Later we see this apparently distorsioning parameter where you can decide resistance type...

It would seem that the raw FFB calculation is afterwards postprocessed somehow with several controller.json modifiers in order to get a better result, which is bad decision IMO. Only several parameters to compensate for hardware limitations should be considered. Changing resistance type from friction (independent from speed) to damping (speed dependant) is unjustifiable from a simulation POV.

Sorry to repeat the same sentence in different threads but when you sit on a real car you get the FFB that you get. You cannot do anything to change it. So it should be in rF2. As I said, some multipliers can be added to suit a child's arm strength or to avoid clipping. It is difficult, however, to justify so many fiddling to FFB not based on anything except for the perception of how should it feel. FFB should be treated as another physical parameter to be measured and validated in its accuracy.
Click to expand...

AMS gives me this experience. Every car feels natural and proportional to each other right out of the box. Just set one master "strength" setting and you're done. Oh, and for those mod makers who cannot figure-out how to make their car proportional to the stock content, there is a realfeel.ini strength variable that you can play with. I have not had to adjust it for any Reiza vehicles, but do for some mods. Problems solved.

rF2 FFB is more sophisticated than AMS, but I think the problem is that even the ISI content is not all programmed consistently. They were experimenting with different ideas and approaches over the years, and the latest cars are really nice, but they never went back to redress all the science experiments of the earlier vehicles.
 
Had an F1 2016 session with my mates last night, and after they had left I fired up rf2 for a comparison. Initially the ffb in F1 2016 felt basically like AC, (ie easy to control the car and 'natural'), rf2 by comparison felt unstable and 'loose'. But after a while the level of detail in rf2 saw me pushing the car in ways that I simply can't do in F1 2016 or AC for that matter.

One place were rf2 really shines is the loading up of the ffb under brakes. I was much faster in rf2 than F1 2016 (I know it's not the same car etc etc) but I think I gained a lot of time in the braking zones by bleeding the brake off gradually into the apex.

The other thing I noticed, is that even when I set smoothing to it's max setting (32), F1 2016 felt far more smoothed. Rf2 sends so much more info through the wheel, and it can be overwhelming and harder to interpret, the first few laps are always a write off, but then something clicks and I'm comfortably driving at a pace not possible for me in other sims. All the right signals are being sent to the wheel, but there is so much going on in rf2 (tyres skipping over the track surface, tyres deforming, car is moving so much more etc etc)... there's so much information that you need to process, which can be good and bad.

I know comparing F1 2016 to rf2 is apples and oranges, but my point is, everything in other games (F1, AC) feels very damped and smooth, rf2 feels very raw and complicated in comparison, which can be difficult, but ultimately is better.

A question about AMS: I have always found that the ffb doesn't load up and get heavier under braking at all? If it does, it's not to the degree of rf2.
 
You must log in or register to reply here.
Back
Top