Tyre friction/interaction with road surface

Yes. That is what I am saying.
In the geometric construction I have applied the slip angles coming from telemetry to those parallel rotated tires and to rear tires. The 4 perpendicular lines to those travelling directions at each tire intersect exactly in the centre of rotation of the vehicle. The obtained radius also matches the value shown in the telemetry. This has been tested for two different speeds with different slio angles. Too much coincidence.

If front tires wouldn't have been parallel they wouldn't intersect as it happens in the case of reference.

 

I don't understand why you're drawing conclusions on the direction of the wheels from calculated slip angles when you can directly log toe. Have you looked at the toe telemetry?

 

You are right. I looked up for tire rotation or something like that and since I couldn't find it I checked if parallel wheel situation would correctly match slip angle results, which it did. Hence I reached the conclusion that for Formula ISI wheels rotated in parallel.
I just made the graphs using toe parameters form telemetry. Much easier than drawing with Photoshop!!!! Thanks

Here the results for Formula ISI, Flat 6 and Karts. In the end I was more or less right with Formula ISI: Ackermann effect is very small as you can see in the graph. For Flat6 which has a higher steering lock, the difference grows bigger due to it, but it is more interesting to see a negative Ackermann applied. Karts do show a clear non linear curve throughout the whole steering range. I was wrong in my simplified model asssumption.

It should be noted that Formula ISI at 210 degrees (8.2 deg at tires) would require an Ackermann geometry causing a 0.49 degrees difference between both tires so that 4 axes intersect altogether. In the graph, the difference between both tire toe parameters is 0.126 degrees for that steering position I tested. This is the reason why parallel tires would seem to properly match real slip angle values. In reality the difference between them is 0,1 degrees which is really difficult to appreciate. I cannot say, BTW if those Ackerman configurations are the typical one for those cars.

My conclusions: Tire behaviour and vehicle dynamics simulation seems to be pretty accurate.
FFB will depend a lot on this tire configuration. If the same high caster settings that were analysed were applied in a real car, a huge difference in force feedback should also be perceived. I think people's major problem resides in ISI's decision to scale FFB output for each car based on a criteria which could be very arguable. trying to define a nominal FFB level for a car regardless its setup is pretty much impossible. In those downleveled cases if you happen to use a low FFB setup you will need to increase steering torque sebsitvity or FFB multiplier tobe able to pereive lower end forces.

 

So there seems to be proper steering ackerman simulation, but Formula ISI has lots of anti ackerman?

 

Sorry. I forgot to mention which color is each.
The black ones showing Antiackerman are Flat6 FL and FR toes: FL tire toes more than FR tire in a right turn and opposite.
The red/blue ones showing a very light Ackermann are Formula ISI's FL and FR toes.
The green one's are of course ISI's Kart F1. Here the non linarity is huge. There are other Ackermann settings available to choose.

 

Different cars will of course have different amounts of ackerman. Different cars also have different amount of toe-in too which you want to zero for every test. Otherwise your ackeman values have some toe-in there making them unsuitable for any comparison.

It gets more complicated if the car has bumpsteer.

 

I only half agree with this statement Paul. Certainly lifting the throttle is a great way to get the car rotated, and I think rf2 does the best job of all the sims at changing the line of the car based on front/rear weight distribution. But lately, it's been bugging me to no end having the front of the car rotate inward under very light power. I am not here to question the realism of this or question the physics, I am simply saying that it's hard to get a feel for and somewhat unpredictable.

I will use the caterham 7 as an example, only because I have been driving it heaps lately (and other isi cars tend to do the same behavior).

Imagine you are turning into a corner, the front tires are well within the limit, you apply a bit too much throttle, and at the exact same time you feed in counter-steer. You'll get into a fairly stable, easy to control slide. No worries.

Now imagine the same corner, but now the front tires are at or slightly past optimum slip (at this point with very low caster settings, the ffb would have dropped away). You apply only very light throttle, and the front and rear tires seem to break away simultaneously, the rear steps out, and the front of the car rotates toward the inside of the track. To make matters worse, the counter steer which worked so well in the first example is now making things worse, the more you counter steer, the more the front of the car seems to 'tuck in' toward the inside of the track.

I've watched many 'racing fail' you tube videos, and I'm not questioning the realism here. What I'm saying is, if you push the front tires too hard in rf2, it punishes you severely. For example, all other things equal, if you have turned the front tires too much, the simple act of dialing out steering input can cause a gain of grip at the front and cause over steer, which is basically impossible to correct.

In a car like the caterham or skip barber, I can drive them like a mad man, sliding through entire corners etc etc and still getting half decent lap times. What I struggle to do is drive them quickly, while keeping the front tires within the limits. I find it harder to drive these cars smoothly and within the limits at reasonable speed, than I do sliding them all over the place and setting PB lap times.

My point in this long winded post is simply that I wish rf2 had a no nonsense way of letting the driver know when the front tires are at/over the limit, considering how detrimental it can be. Whether it be by improvements to the ffb or by using some kind of canned effect as an assist. Most people out there are running a g27 and imo opinion it's very hard to know when you've exceeded front grip/scrubbing, wheres in other sims its seems like it's a non issue.

 

I should reinforce the fact that I think in general, rf2 ffb is probably the best of all the sims. I am more thinking about new users and people with low end wheels here.

 

Have you tried if different steering torque sensitivity settings help?
What about damping settings? damping should probably be at 0 for g27.

 

I think higher steering torque sensitivity might help to that oversteer/ lack of wheel speed problem.

 

This happens when you are having lot of understeer and lots of steering lock, then if rear steps out slip angle of front tyres increases even more. Now slip angle of the front tyres are far from where the max steering force would be. As you start to counter steer you will reach the point (amount of slipangle) where the max steering force comes and it will make the car rotate even faster.
But this is rarely problem with high end wheels that can turn through this quickly to proper amount of counter steer.

 

Exactly. When you've turned the front tires past a certain point (ie front tires are scrubbing), and then the rear steps out, the car seems virtually impossible to correct. Basically you are along for the ride, and counter steering at this point is essentially counter-productive, because it should have occurred a second ago.

The only way to really know whether the front tires are pointed within acceptable limits for a given corner is to drop down the caster, and when the ffb drops away, you know that you've pushed a little too far. Any over steer at that stage would be quite difficult to properly control. The problem with that is you can't set the caster low enough on many cars to achieve this feeling. If I could feel the grip drop away with higher caster settings, obviously that would be ideal.

Yes, I think a little bit helps with oversteer. Though I had an idea which I'm not sure is valid but could help with the understeer thing, essentially it's the opposite of boosting low end forces. If high end forces were boosted relative to mid range forces, maybe it would be possible to feel the difference between when the self aligning torque is at it's maximum, compared to when it slightly drops away after a certain front slip angle.

 

That's really interesting. If the Accuforce struggles, what hope does the g27 have!

I don't think you can judge the physics or ffb of rf2 properly with a low end wheel. For example the physics might be telling the front tires and thus steering shaft to rotate at x speed, but if your wheel can't keep up with that speed, then it's negatively effecting the physics of the game.

 

Yes, you can do that by lowering the steering torque sensitivity.

 

I honestly can't perceive much if any difference by altering the STS. I think the g27 just lacks the dynamic range needed to produce the forces I'm looking for (ie light in easy corners, quite heavy in hard cornering). I have a mate who might be getting a Fanatec CSL soon, can't wait to give it a go.

 

He is probably wanting to feel the drop off of forces when going close to the limit of front tyres.
Have to say that i don't remember experiencing much torque drop off with my osw either when driving that caterham, but for some reason don't have those problems that green serpent is having.
green serpent, if you are having too much understeering and you use LSD set coast to minimum.
Or drive without lsd.

 

Yes I would love to feel that, as it is what I think RL would be like. Like mentioned, currently the only way to get that sensation is to lower caster. Perhaps IRL self aligning torque wouldn't drop away too much with a higher caster setting, and rf2 reflects that.

Yes, that was explained to me. I personally don't think my expectation is unrealistic. I know from real life experience that going around a tight corner at a descent speed takes both hands and a decent amount of strength, whereas flowing through some easy bends can be done with one finger on the wheel. I don't expect a cheap wheel to be able to match real life forces, however at the moment I am unable to perceive much of a difference in cornering forces between 'easy' corners, and 'hard' corners, or put differently, throughout the various stages of lateral tire load.

Vertical loads are a different story (ie weight transferring from front to rear.) rf2 is easily the best in that respect.

 

Yeah, definitely using LSD before driving is bad habit.

 

Is this reflected in telemetry from contrasting corners, or does it just not come through in the FFB for you? If telemetry shows a difference, then lowering STS should improve how much you can feel that difference. If there is no (or very little) difference in telemetry, you'll always struggle to feel the difference without a good wheel.

 

What you are suggesting is the RIGHT way to know if there is a problem with the FFB device or if the FFB itself isn't showing expected values.

From what I have checked in my tests, steering torque calculation seems to be OK. However, some people trend to think that. FFB should feel as they expect it regardless the setup being used, which has been proved to affect significantly.

The lack of systematic procedure in order to properly calibrate the FFB is a constant in this forums. Everything is done empirically based on what they expect to feel (in most cases without having ever driven the car being used) instead of comparing it to what the car physics dictate.