The online racing simulator
iRacing
(13603 posts, closed, started )
Thanks for the reply MadCat360. Do we know what kind of Brake bias is normal in the RL Skips? This isn't so much of an issue in the iRacing version if the BB is around baseline setting FWIW. Although I'm only talking about using a couple of notches back.

You are coming right off the brakes there when it happens though, whereas my issue is more mid way -where I need to pause on the brakes if RFB.

Yeah, I've not had pure power oversteer there.

As always with RL footage, I'm struck by how small corrections for the things you can obviously feel in the car, that we can't really see, keep it all under control.
Quote from jtw62074 :If it was engine braking, the behavior would manifest itself right away at small slip angles rather than just large ones. In the first part of the video where I'm blipping and releasing the throttle, the car straightens up when the throttle is released (understeer yaw moment). The trouble is that this fundamental behavior reversed course once the slip angle grew enough. The understeer moment decreased, passed through 0, and then changed sign.

There's one key bit about the final blip that is different... your steering input. You straighten the wheel whereas in the others you keep it pointed into the turn.

Not exactly sure how that might skew things, but induced understeer was something that was thrown around a while back. Could the constant steering input be inducing understeer?
Quote from Postman Pat :Thanks for the reply MadCat360. Do we know what kind of Brake bias is normal in the RL Skips? This isn't so much of an issue in the iRacing version if the BB is around baseline setting FWIW. Although I'm only talking about using a couple of notches back.

You are coming right off the brakes there when it happens though, whereas my issue is more mid way -where I need to pause on the brakes if RFB.

I don't know for sure. I'm sure each car is a little different, kinda like the ARB settings are usually different from car to car.

Based on iRacing, it feels like the brake bias is around 60-62% front. But who the heck knows. The MX-5 runs a completely ludicrous 85% front in iRacing's baseline and it still feels like it's using the handbrake right as you turn in. As a result the thing stops like crap now. But at least it's controllable.

Anyway, looking at what Todd wrote (thanks for taking the time dude), it seems like it would be pretty far forward (for the car). I can't ever recall locking a rear tire in the Skippy. I've locked the right rear tire a few time going over the crest of turn 8's braking zone at Laguna in the Miata I drove this year, and also locking the rears going over a similar crest in the braking zone of turn 10 at Thunderhill.

The Skippy will oversteer any time you release the brakes too quickly. Doesn't really matter if you going from 40% to 20% brake pressure or from 10% to 0% - any fast release will cause the car to rotate quicker if you're cornering hard. My problems usually lie in the last 10% of the brake release, so that's why I usually get it in turn 2 at Laguna. The brake pressure you need to carry to the middle of the corner is just barely more than resting your foot on the pedal. At least, that's what it feels like after locking it down so hard during threshold braking. The brake pedal is incredibly stiff. Even so, using maybe 5% brake pressure, if you come off that pedal too fast, it will bite and try to spin you.

On the other hand, the Miata you can release as fast as you want and it won't have a problem. As can be seen in this video, I can snap off the brake as fast as I like. http://www.youtube.com/watch?v=28ijoZwpmbU In reality it's a vacuum assisted pedal, so it's probably not as aggressive of a release as it looks, but still. I've never felt TBO from the Miata unless I just plain used too much brake, like you'd expect traditionally. In my first race this year I came upon a spin in turn 14 of Thunderhill, which is a heavy trail braking corner that needs a lot of rotation. I saw the spin and I increased brake pressure just a bit as a reaction, and I got way loose. If that was the Skippy I'd have developed a tremendous push.
Just after writing that last post I think the "release brake and spin" stuff finally crystalized in my head. It may be that this is much simpler than what I suggested in the last post. I'll try to explain. What we're going to see here is that this can happen with no load sensitivity at all. Load sensitivity would alter it, but in the complete absence of load sensitivity this "release brake and spin" stuff can still happen.

What we'll do is look at a car that is cornering at two different states seperately. First, a car that's cornering without any braking applied at all. Let's call this "pure cornering" from here on out. The second case would be a car that's cornering and braking at the same time, which we'll call "combined cornering" since it's doing a combination of two things at once (braking and cornering). Cool? Okeedokee then.

In the combined cornering case the car has forward weight transfer. In the pure cornering case it does not. So let's take a look at a car in both situations. I'm using a little program I wrote to spit out tire loads given lateral (cornering) and longitudinal (braking) accelerations. The weight of the car is the same as the real skip barber car, but I'm using 50/50 weight distribution because this translates directly into yaw moments so we can see push/spin at the limit without doing more calculations. Also I'm using 50% front total lateral load transfer which is probably too low for a Skip Barber car. The fundamental behavior is what we're trying to look at here, not necessarily the specifics of the Skip Barber F2000. Here are the specs I'm using:

Weight 1500 lbs
Track width 61 inches
center of gravity height 12 inches
front/rear weight distribution 50%/50%

Ok, first the pure cornering case. I'll keep the accelerations somewhat lower than the real car can probably do so we don't fall into the trap of thinking this is something that only happens when we're at the limit of traction. We'll start by cornering in a left turn at 1g which lets us skip a calculation or two and keep things simple. In the data below, the wheel loads are laid out like this:

Front left tire-----Front right tire
Rear left tire-----Rear right tire

Cool? Ok then. Here's the pure cornering case:

Pure cornering:
Lateral acceleration : 1g
Longitudinal acceleration: 0g
wheel loads:
227-----523
227-----523

Ok, we can treat this for simplicity as though there are two torques acting here. One from the front axle and the other from the rear. The front axle is trying to twist the car to the left which we'll call positive "yaw torque," and the rear is trying to twist it back the other way which we'll call negative "yaw torque." If we add the wheel loads of the front two tires together, then the rear two tires together, in our simplified case we can just treat these for the sake of this example as the yaw torques themselves. So what are the yaw torques?

yaw torque front axle
+750
yaw torque rear axle
-750

The total yaw torque on the car is just 0. (+750 - 750 = 0.) This means the car is rotating at a constant velocity in our corner. So it's probably just zooming around the turn normally and might be pushing or neutral, but at least it's not accelerating into a spin. It's just your standard normal turn.

Ok, now for case two. Here we'll add 0.5g braking:

Combined cornering:
Lateral acceleration: 1g
Longitudinal acceleration: -0.5g
276-----571
179-----474

This gives us yaw torques:

yaw torque front axle
+847
yaw torque rear axle
-650

Total yaw torque (847-650) = +197. This is an oversteer moment now so the slip angle is going to increase from here, perhaps even spin. This is just your standard forward weight transfer under braking trying to spin the car around that you might normally expect. More weight on the front tires = higher yaw moment in the oversteer direction.

So where are we now? The pure cornering car is neutral while the combined braking/cornering car is oversteer. As I went over in the previous post, in reality with our braking we are reducing the lateral forces. In the case of the Skippy this changes the combined cornering scenario by reducing the front lateral forces more than the rears by enough to end up with understeer instead of oversteer (negative yaw moment instead of positive).

In this case we might have something like the following. So far, for simplicity we assumed the wheel loads were equal to the tire forces. Under actual braking we'll end up changing the lateral forces, so instead of 203/645/106/547 tire lateral forces we might wind up reducing the front lateral forces to around 60% of what they would have been and the rears to 90%, winding up with something like this:

Combined cornering:
Lateral acceleration: 0.7g
Longitudinal acceleration: -0.5g

tire forces
166-----342
161-----426
yaw torque front axle
+508
yaw torque rear axle
-587

Total yaw torque (508-587) = -79

This isn't really right because these numbers would change the lateral acceleration, which would change the wheel loads, but the point is that there is enough braking here to reduce the front lateral forces so far that we end up with understeer under braking and cornering because of the effect of the brake bias on the lateral forces of the tires. Here the understeer or push is shown by the -79 total yaw torque. So even though the numbers here are not really right, the principle is still the same: Enough forward brake bias will give you understeer. The front tires do not actually have to lock to create rather dramatic push, as a couple of you found out on the real Skip Barber car.

We now have three states instead of two:

1) Pure cornering (no brakes)
2) Combined cornering and braking without the brakes reducing the lateral forces at the tires, resulting in an oversteer moment
3) Combined cornering and braking with the brakes reducing the lateral forces at the tires, resulting in an understeer moment

Here's the kicker and where the "lift off the brakes and spin" moment happens: When you are braking and cornering you are really in state #3 where the lateral forces are reduced enough at the front in relation to the rear to make the car push instead of spin. However, at the instant you release the brakes, the tire loads are still in state #2. The tire forces then jump to that as well.

You jump from state 3 with the brakes applied to (almost) state 2 when we instantly release the brakes. I.e., the weight is still transfered forward even though the brakes have been released and the longitudinal acceleration has returned to 0 because the springs are still compressed due to the pitch of the car. The orientation of the body has not yet changed to pure cornering. The nose is still down.

So we get a quick burst of oversteer when we release the brakes. As the nose starts rising back up we are transitioning from state 2 back to state 1 where there is understeer again.

The shock absorber settings will influence how quickly these state transitions happen, but right now I'm thinking they really aren't the cause. The real trouble to me at this point seems to be lying in the brake bias itself. If state 2 was neutral and the brake bias of state 3 was adjusted in a way to also be neutral, you would transition from braking/cornering to pure cornering when releasing the brakes without changing the yaw moment or at least not letting it ever go positive. In other words, the tail wouldn't kick out and it wouldn't matter much that you released the brakes.

What's interesting about this to me is, if my thinking is right here, that this basic thing should happen independently of whether or not there is load sensitivity in the tires, and that this shows up even if we completely neglect the shocks. It's possible that the shocks might make this situation worse or better than this, however, but the underlying thing here seems to me to be more about brake bias than anything else.

What about releasing the brakes quickly versus slowly? As we saw before, releasing the brakes instantly takes us from state 3 (push) to state 2 (spin). However, if we do it a little bit more slowly the oversteer moment will not jump immediately to +197. It'll be lower than this because while we're transitioning from state 3 toward state 2 we are keeping the front tire lateral forces lower than the rears through the brakes, like in state 3 but not quite as much.

If you released the brakes slowly enough you would never hit state 2 where you have full forward weight transfer and no brakes. You'd slowly transition from state 3 (mild understeer) to state 1 (neutral) and all the inbetween states would be somewhere between mild understeer and neutral. The car wouldn't spin. It looks to me like this might be where the fast or slow release of the brakes might be coming into play. You have to avoid jumping too close to state 2 and try to go from state 3 to state 1 instead. Either that or fiddle with brake bias or use gas and brake at the same time to dynamically alter the brake bias throughout all this.

I'm really glad you asked about this. I never gave it this much thought before, but this seems to be the situation as far as I can tell. It also seems that if this happens in a sim, it isn't necessarily a matter of "bad physics" or a bad tire model, but rather a case of "this car and its tires need a bit more attention in this area and could be improved"... It seems to me that the iRacing car goes a bit overboard with this, but the fundamental behavior isn't necessarily wrong. Some small adjustments to the tires on the car could probably get it to be closer to the real car in this area.
Quote from Mattesa :There's one key bit about the final blip that is different... your steering input. You straighten the wheel whereas in the others you keep it pointed into the turn.

Not exactly sure how that might skew things, but induced understeer was something that was thrown around a while back. Could the constant steering input be inducing understeer?

To some extent yes, induced understeer magnifies the effect. However, this is really further evidence that the force curves drop off too much after the peak rather than the opposite. Induced understeer is a geometric effect (tire lateral force vector pointing more rearwards and less sideways), but is significantly magnified by any drop off in force occurring in this region of the tire force curve at slip angles beyond the force peak. No force drop off means very little induced understeer. In that case countersteering to correct a slide will do what it's supposed to: Slow down the yaw rotation rather than speed it up. This goes hand in hand with steering into a spin in order to save it.

However, I'd still argue against induced understeer being the cause, rather than something that magnifies it and thus might overshadow it, for another reason: As far as I was able to tell, this was happening with all four tires pointed straight ahead at any slip angle somewhere between the peak and around double the number of degrees passed it. I.e., with all four wheels pointing forward (0 steering), and peak lateral force coming in at (just making this number up) 7 degrees, the car was understeer at 7-12 degrees and oversteer at anything over that. The trouble was, the further the car rotated the faster it accelerated into a spin. This makes catching a slide ridiculously more difficult than it is in reality.

(By the way, I'm just making up those slip angles to illustrate. I don't know what the real numbers were.)

This stuff is much easier to see when you make a sim, or have one handy like rFactor or Racer or something where you can play with the tires yourself. I'd encourage anyone interested in this to try making tires that drop off very quickly after the peaks and see what this does to the car. As you steer into a spin the car straightens up. As you countersteer you accelerate the spin. It's the opposite of what happens in reality and is the main reason why some sims are so hard to catch slides in. People can argue about lack of g-forces or bad force feedback all day long, but they're missing this key point: If the tires do this the cars will be just that much harder to catch on top of the missing g-forces and bad FFB. In many cases, significantly so.

Live For Speed doesn't have this problem.
Quote from Mattesa :I'm not quite understanding the issue here, so I'm going to prod at your post a bit of you don't mind:

steady state cornering was pretty strong understeer.

- Okay...

As the rear slip angles increased, the understeer moment decreased slightly, then decreased more rapidly at higher slip angles

- Isn't this expected? As the rear slips more than the front you get less understeer?

and eventually reversed directions. I.e, understeer moment turned into oversteer moment which should not generally happen.

- It seems to make sense though? From steady state you increase speed leading to higher slip in the rears (definition of oversteer?) then eventually gets to a point where the car is yawing into the direction of the turn.

Can you explain what situations and how to replicate what you're talking about in game?

Mattesa, check out the attachment. It's something I made for one of the iRacing forum discussions on this.

This is showing what I believed to be happening with the FGT and HPD cars. Here we see lateral force versus slip angle for the front and rear tires separately. More specifically, these would be yaw moments at the front and rear axles in pure slip which are shaped the same way and have the same features shown here.

In my video I started out blipping the throttle while keeping the steering constant. The throttle blip increased the slip angle (traction circle effect at rear tires) momentarily. With the throttle then released the car straightens back up pretty quickly. This means the peak force at the rear tires is higher than the peak force at the front tires. We could say the rear tires are big fat suckers that stick better than the fronts do.

The trouble was that this reversed at some larger slip angle. If the car was extremely close to neutral I wouldn't have had a big problem with it, but this thing pushed like a tank as can be seen in the video. To reverse course and swap the understeer moment to oversteer moment like that, the fat rear tires would have to have less grip than the fronts do at those larger slip angles. If both the front and rear tires dropped off after the slip angle peaks at about the same rate you wouldn't expect anything but understeer. The car should still straighten up when you release the throttle, or at least slow down in rotation, but it didn't. It was as though the rear and front tire force curves (more specifically the yaw moment curves) crossed over each other and the rears became like little tiny tires that had even less grip than the front tires did, and around you went...

The forums were in an uproar over this and debates on it sprung up everywhere. Half the people said "adjust your FFB" or "you can't feel the g-forces" or "big deal, the car spins out because it's powerful and you're hitting the gas and at low speed you're in first gear and have no traction" or "you just need to adapt to the NTM" and on and on, totally unable to see what we were talking about.

There's something in vehicle dynamics called the Mass Moment Method (MMM for short) where you can plot out the yaw moments on a 2D graph as a function of all combinations of front and rear slip angles. There is a definite characteristic to a car that pushes: As the front and rear slip angles increase to and beyond the peak force values, the yaw moments all converge on a single point or narrow band on the graph somewhere in the understeer moment area. In other words, if the tires peak at 5 degrees and the car is pushing, it will also be pushing at 10 degrees, 15, 45, etc., all the way out to 90 degrees.

More precisely, the yaw acceleration should remain in the push direction, meaning the yaw velocity of the spin should reduce towards zero all by itself regardless of what you're doing with the steering, the way it does in the first part of my video at slip angles just a little bit over the force peaks. I don't have a problem with the yaw acceleration trailing off a bit and changing, but in the video the yaw acceleration not only changes, it completely reverses direction at some larger slip angle. I've never seen an MMM diagram that looked anything like that, regardless of the slip angle.

The only thing I could think of that would cause it was to make the rear force curves drop off more rapidly than the fronts did. At some slip angle they will cross each other and the yaw moment would reverse just like in the video. I tried this in my hobby sim and it worked.

But then think for a minute what this would mean: At some large slip angle, those giant rear tires that created so much more force than the front tires did, creating push at the limit, would lose all that and produce less force than the front tires did, even if the front and rear tires were at exactly the same slip angle. Why would a 335mm tire at 30 degrees slip angle produce less force than a much smaller tire would at that same slip angle? It always seemed rather silly to me and I pointed it out several times. Thankfully they ended up fixing it or at least improving it significantly on both the FGT and HPD in the next patch. Hooray.

Meanwhile many of the people would argue that it's an illusion created by bad force feedback or a lack of g-forces or something along those lines. Doesn't matter. Give them a real car with tire force curves that do that and they'll be complaining about the real car just as much as I complained about the simulated ones.
Attached images
TireForceCross1.jpg
Why would it need to have different slip angles characteristics on front and rear to do that? Wouldn't simply just having too much drop off do exactly the same? I mean, once you get the rear of the car into big enough slip angle the car will spin because the rear will have less grip than the front. No matter if in four wheel slide or just oversteer.(unless you steer into the turn) Even in four wheel slide the car will slide in four wheel slide just fine as long as the rears produce enough lateral Gs to keep the rotation from accelerating.

In a steady state four wheel slide the front tires rotate slower than the rears because of the rear of the car needs to travel longer distance. So in other words when the radius increases the lateral acceleration decreases (a=v^2 / r) and thus the (the need for) lateral force also decreases (F=ma). The outside rear tire therefore has the largest slip angle and biggest load in four wheel slide situation. So technically in four wheel slide even when the rear has less drop off (as long as there is drop off) than the front the car will still spin at some point when the front lateral Gs exceed the rear lateral Gs because in four wheel slide especially when not steering into the corner the rears will always have more slip angle than the rears?

So in the beginning when the slip angle or the provocation of the car into the spin is too small the car will just understeer because the rears reach the peak first (and don't go over it) and therefore the fronts always produce equal or less Gs in that situation. But when the rear goes too much over the peak and when the rear produces less lateral Gs the car spins because the rotation of the car starts to accelerate. Just like in that ford gt video you posted. When you get the rear to step out enough for the rear to go over the lateral peak and some the car spins because the drop off is causing too much difference in lateral G ability between front and rear. I always try to avoid using technical terms.

Why would it need more drop off in the rear for this to happen? Wouldn't just having too much drop off alone cause this issue of first there being understeer, then extremely small sweet spot of neutral four wheel slide and then just icy spins? I mean the more drop off the smaller the window of four wheel slide and the more extreme the understeer and oversteer situations? I can understand that obviously having more drop off on rear will make the situation worse and the car easier to spin but isn't the real issue the too much drop off?

Also without knowing anything about tires (other than that they taste odd) shouldn't a wider tire have little more drop off in lateral grip at the same slip angle than a narrower tire simply because the tire carcass being wider and thus being more rigid (because its rigidity and second moment of area being bigger). At least when looking at the tire twisting around its contact patch? Like when being stationary and turning the wheel in a car the contact patch of the tire (and the tire carcass itself) "twists" and with wider tires the twisting causing less deformation in the tire. So isn't the principle of wider rears having more drop off technically correct even if it may be overdone in iracing?

I really don't really see why it would need different slip curves for front and rear to get that result though. Doesn't just having too much drop off take care of that?
Sweet post Todd. That's really insightful. Dropping the longitudinal force to zero before the nose has a chance to come up really makes sense that it produces such a dramatic effect.

It's not quite as dramatic as iRacing seems to think it is though. You might want to outline that stuff on the Skippy iR forum. I think they would like to hear it, although, based on how some of your other threads turned out, certain people in the community might not agree... there really are a ton of fanboys on there that won't even consider the fact that iRacing might have it wrong even slightly.
@Hyperactive:

Just a quick one this time. Been here way too much today already

My thinking is that the rate of drop off would have to be different in order to get this happening where when the steering angle is 0 (almost identical slip angles at all tires then) the balance changes from push to spin. The trouble here with "isn't drop off enough by itself" is that the front tire forces would also be dropping too. Not just the rears. So to use "g" instead of lateral force with a simple tire that rises linearly, peaks at 2 degree points, then drops at constant slope to a ridiculously low value unbelievably quickly:

Tire A:
slip angle
1------------0.2g
2------------0.4g
3------------0.6g
4------------0.8g
5------------1.0g <---peak
6------------1.0g <---peak
7------------0.8g
8------------0.6g
9------------0.4g
10------------0.2g

Obviously as you've said, if the rear tires peak and go still higher in slip angle to say 8 degrees, the lateral g for the rear drops to 0.6g. But so does the front if the steering is 0 and the slip angle is the same. That's why I centered the steering. The balance hasn't changed here. Only the overall lateral acceleration of the whole car has dropped.

The argument can be made (and was in the comments section on my video), that since I straightened the steering the rear slip angles are not really the same. They're a little bit higher. That's fine. If the tires are all the same and the tire force curves were really like this the car could accelerate into the spin. That's why I said if the car was almost a neutral steer car I wouldn't be bothered so much by this. But really, this car is a strong pusher. So the tire above might be our front tire, and the rear tire force limit is higher and might look like this instead:

slip angle
1-------------0.3g
2-------------0.6g
3-------------0.9g
4-------------1.2g
5-------------1.4g <---peak
6-------------1.4g <---peak
7-------------1.2g
8-------------1.0g
9-------------0.8g
10------------0.6g
11------------0.4g

Here the slope of the drop off is the same as the front tire (0.2g per degree) while the initial part rises more quickly and peaks higher as I'm sure you'd agree is obvious from the video.

If we start out a normal turn with the fronts peaked at 5 deg and 1.0g, the rear slip angles are a little over 3 degrees to balance it at 1.0g. We can now kick the back end out with a throttle blip up to 8 degrees and it the car should still return to the original 5 and 3 degrees. This is what happens in the first part of the video.

Ok, now let's kick the car passed the peak and straighten the steering. How much slip angle difference is there between the front and rear tires? 1 degree or so? Not sure off the top of my head, but let's just call it 1 degree. If we slide the rear out to 8 degrees, the fronts are at 7 like this:

1.0g rear (8 degrees) and 0.8g front (7 degrees). Still understeer.

And if we slide the car 1 more degree:

0.8g rear (9 degrees) and 0.6g front (8 degrees). Still understeer.

And another:

0.6g rear (10 degrees) and 0.4g front (9 degrees). Still understeer.

There is in every case here the exact same balance: 0.2g per degree understeer moment, so long as we keep the steering straight and our rear slip angle is 1 degree greater than the front. So I don't think the argument really holds true. Even with a ridiculous drop off on both tires you aren't doing much to the yaw moment passed the peak.

Of course if you countersteered in a slide with these tires and kept the fronts at 5 or 6 degrees, the car would just spin faster, which is one of my other points that I think you agree with already. And of course then as the slope of the drop off increases, steering into the spin becomes more and more likely to be a better way to catch a slide. That does illustrate excessive drop off.

In my video if I had kept the front slip angles at their peaks by countersteering and the car accelerated the spin, it would have shown tire force drop off just like you said. This is why I centered the steering right away in order to show what appeared to be greatly different slopes.

I tried to do this in my sim and it was pretty tough. The front/rear tires had to be quite dramatically different after the peaks. Much more than I was expecting really. It was kind of hard to make it do what the GT did...
I want to add something to that last post. I assumed the rear slip angle was 1 degree more than the front. However, your point starts to become more apparent if we go farther than this and say it's 2 degrees difference instead:

1.0g rear (8 degrees) and 1.0g front (6 degrees). Drift/neutral
0.8g rear (9 degrees) and 0.8g front (7 degrees). Drift/neutral
0.6g rear (10 degrees) and 0.6g front (8 degrees). Drift/neutral

Now the balance has changed from push to drift/neutral and if we go any further (slip angle difference > 2 degrees) the balance would have changed and the yaw moment would have reversed just as you suggested. However, I've got a really tough time believing this is going to happen with slopes that are identical or just a little bit different in practice. The rear tire in this example goes from 0 to 100% in 5 degrees and back down to 0 grip in the next 10 degrees. You'd have to go so far out in the slip angle direction along the rear tire force curve that you find a point that's lower than a point on the front. The slip angle difference to do this on a car as pushy as the GT might be enormous.

Maybe we should try this with some more typical tire data. Unfortunately most real tire data shows very little drop after the peak, if any at all, which kind of illustrates several of my issues including this one all at once...
Quote from jtw62074 :@Hyperactive:

Just a quick one this time. Been here way too much today already

My thinking is that the rate of drop off would have to be different in order to get this happening where when the steering angle is 0 (almost identical slip angles at all tires then) ...

I just drew a picture about slip angles... In four wheel slide the car rotates around its center of mass right? Doesn't that mean that the rear slip angle is with (0 steering) always more than at the front? The car isn't going in straight line sideways but it is going sideways around a circle (with little power applied to keep the speed and radius constant)

In the pictures is a porsche 911 GT2. Rearwards center of mass. flat 6 engine... The thicker green dotted line represents the center of mass' path around the center of the circle. The other two green lines are the front left and right rear tires' paths. The straight dotted lines are the direction of the slide of the tires at that moment.

First a tighter turn and then a bigger radius turn. Car slides in four wheel slide. The tires are angled so that the longer side of the rectangle is in line with the tires (tires pointing straight ahead, 0 steering, 0 toe). The rear of the car goes around bigger radius than the front. Car travels anti-clockwise towards left in the pic. The slip angle on the right rear is bigger than on the front no matter how I draw the pic .
Attached images
c1.jpg
c2.jpg
Quote from Hyperactive :I just drew a picture about slip angles... In four wheel slide the car rotates around its center of mass right? Doesn't that mean that the rear slip angle is with (0 steering) always more than at the front? The car isn't going in straight line sideways but it is going sideways around a circle (with little power applied to keep the speed and radius constant)

umm, yes. That's discussed in the rest of my post

As an aside: The drawings look very nice (how did you make those anyway?). The rear slip angles are going in the wrong direction though. The box won't be aligned that way. Instead, the center of the box would be tangent to the path at 0 slip angle rather than the inside rear tire, then rotate further clockwise as slip angle increases.

EDIT: <Slaps self in forehead> Is your car moving counterclockwise by chance?

Why of course it is! You said so right here:
Quote :Car travels anti-clockwise towards left in the pic.

You have the drawings right, nevermind
Quote from jtw62074 :If it was engine braking, the behavior would manifest itself right away at small slip angles rather than just large ones. In the first part of the video where I'm blipping and releasing the throttle, the car straightens up when the throttle is released (understeer yaw moment). The trouble is that this fundamental behavior reversed course once the slip angle grew enough.

hmmm good point
Nice to see a constructive discussion of vehicle dynamics for a change!

I'm really not sure what to make of the latest changes to iRacing, at first I really struggled to get to grips and subsequently managed to dip my SR below 1.0 and have now been demoted to a C license. After finally deciding to treat the cars as being completely new and different (and stopping entering races on untested combos and finding it surprising that I'm spinning off all over the place) I've found the new cars are actually quite driveable so long as you keep corrections to a minimum. I find you've got to be fairly aggressive with the cars on the NTM to actually make them understeer.

I really can't get on with the MX5 Roadster or the SRF on the new tyre physics. I never realised the MX5 Cup was actually different to the Roadster. The Cup now feels much better than the Roadster, which I think is at least in part down to the different fixed sets.
Quote from ajp71 :Nice to see a constructive discussion of vehicle dynamics for a change!


I really can't get on with the MX5 Roadster or the SRF on the new tyre physics. I never realised the MX5 Cup was actually different to the Roadster. The Cup now feels much better than the Roadster, which I think is at least in part down to the different fixed sets.

Same with me. Before the new physics I drove the Roadster alone, but now its not really drive-able for me. Now I can only drive the cup.

as to the rest of the current discussion here. Kudos to you guys who understand stuff like that.
Quote from jtw62074 :umm, yes. That's discussed in the rest of my post

The point I'm trying to make is that a four wheel slide situation does not exist where the front and rear tires have similar slip angles because the rear slip angles should always be higher. I may be missing some important bit here though.You say the rear slip angles are "They're a little bit higher". But aren't the rear slip angles drastically higher in such situation?

According to my pic even with really small slide or really big slide the slip angles on front and rear are very different which means the front and rear tires are functioning at completely different parts of the slip angle and thus the drop off is causing that huge difference in lateral grip. Even if in such situation the need for lateral grip to keep such system stable would be less than that because even the rear has much higher slip angle the radius of its path is also bigger which means the lateral g required by it is smaller.

First this the tire:
slip angle
1------------0.2g
2------------0.4g
3------------0.6g
4------------0.8g
5------------1.0g <---peak
6------------1.0g <---peak
7------------0.8g
8------------0.6g
9------------0.4g
10------------0.2g

Let's put the car (rwd car with 50% weight f/r weight dist.that weighs 1300kg) into 5 degree four wheel slide (=yaw). Attachment 1. Car's track is 2m wide and wheelbase is 4 metres longs. The car drives around in constant circles and the circle radius is 100 metres measured from the center of mass of the car.

With 0 steering we get slip angles of:
front left: 3,8 degrees
rf: 3,9 degrees
rear left: 6,1 degrees
rr: 6,2 degrees

So in such situation the lateral grip provided by each tire are (without load sensitivety):
front left: ~0.75g
fr: ~0.75g
rear left: ~0.95g
rr: ~0.95g

Looking at the car's tires' slip angles and what Gs they produce it looks like the car is understeering (50% weight dist.).



But if we put the car into 7,5 degree four wheel slide the situation changes. Just changing the yaw angle 2.5 degrees (attachment 2) :

With 0 steering we get slip angles of:
front left: 6.3 degrees
fr: 6.4 degrees
rear left: 8.6 degrees
rr: 8.7 degrees

So in such situation the lateral grip provided by each tire are (without load sensitivety):
front left: ~1g
fr: ~1g
rear left: ~0.5g (nice to have real data for these "extreme" slip angles)
rr: ~0.5g

That's one oversteery biatch right there.


The rear slip angles are much higher...?

I think I'll go and make an excel out of this thing to find out where that critical yaw angle is. Need to find some tire data too, preferrably fake one .


Quote from jtw62074 :As an aside: The drawings look very nice (how did you make those anyway?).

I used catia. It is very easy to illustrate 2d mechanics problems with it just by making a drawing with it and then dragging the bits around to see how the thing moves. Usually it takes longer to start catia than it takes to draw such pictures.

E: fixed it all and added two more illustrations of slip angles at 11 and 25 degrees.
Attached images
sl5.jpg
sl75.jpg
sl11.jpg
sl25.jpg
Quote from jtw62074 :I've often wondered what the real Skippy does, so I'm really happy that you got into this discussion and explained what it feels like in the real car, Madcat. Thank you.

one thing about the real skippy though is a trainer car that they probably dont allow you to do a lot of setup work on right?
in theory it might be just the same but they only ever allow you out with a brake bias far forward
also it might come down to something as simple as youre steering too much when you release the throttle and the front tyres jump from a state of being laterally traction limited by braking to peak lateral force in an instant
@hyperactive

My quick estimation of a 1 degree larger rear than front slip angle was probably way too low for the case in my video. I agree with that of course. What I was trying to point out was that the front tires would also be dropping off, not just the rears, which takes away a great deal of the effect I thought you were referring to. Usually when I run across this thinking, people seem to be only considering the rear drop off and forget about the front entirely. My mistake if you weren't.

On your point about not needing different slopes: Strictly speaking you're right about that of course. The slopes don't need to be different as can be shown with a ridiculous example tire like I came up with where all the force disappears in 10 degrees slip angle. I meant it more as a matter of practicality with real tires. With the exception of what appeared to be some kind of Nascar tire at -8 degrees camber, I haven't seen a tire force curve that drops off more than probably 5-8% at quite high slip angles (in the dry, the wet is another matter), and even that is somewhat questionable because of the noise involved in the raw sampling and how much the forces can change due to temperature during the sweep test. I.e., the curve can appear to drop off due to temperature effects as the tire is being steered up to the high slip angle. Usually on the return trip to the lower slip angles you end up with a much flatter force curve, which is more like what a real car is probably running on most of the time. When this is taken into consideration, most real tires are dead flat after the force peaks right out to 90 degrees, believe it or not, and some continue a slow rise after the nominal peak. We need to keep in mind that these tire plots we see are not always processed the same way, and drum tests usually show more drop off than flat belt tests do. So when somebody pulls out a graph that shows a lot of drop off, I'm immediately skeptical. My thinking and view on this was largely shaped by conversations I've had with Doug Milliken on this while we worked together on tire testing for VRC Pro.

If we take a rear tire that peaks at a 10% higher lateral force than the front does and let the slopes after the peaks be identical, at least in scale, and something along the lines of a slope you might find with a real tire like the FGT's, my gut says the slip angle difference needed would be beyond what you would get at cornering radii on the scale of what was shown in my video. So if both tires dropped as much as 9% after the peaks, even 90 degrees slip angle difference front to rear wouldn't be enough to reverse the yaw moment. So if I picked 1 degree or 10 or 20 slip angle difference it might not have mattered much. This was coming partly from practical experience with my own sim. This then led me to the conclusion that in practice the slopes would really need to be quite different. In my sim it took quite a bit more of a difference in slopes than I originally thought it would in order to get behavior roughly similar to the FGT and HPD.

The way to proceed from here, as you seem to already be going toward, is to grab some real tire data and see just how large a slip angle difference you'd need in order to get the yaw moment reversal happening. We'd need a rear tire that produces a great deal more force than the front so we have something like the FGT: A stronger pusher at the limit.
Big smelly crap that walks! I just noticed I had the set the diameter to 100 instead of the radius being 100... need to fix those numbers slip angle numbers as they are slightly off
Quote from Shotglass :one thing about the real skippy though is a trainer car that they probably dont allow you to do a lot of setup work on right?

I imagine that's right, but don't know for sure.

Quote :
in theory it might be just the same but they only ever allow you out with a brake bias far forward
also it might come down to something as simple as youre steering too much when you release the throttle and the front tyres jump from a state of being laterally traction limited by braking to peak lateral force in an instant

Bingo. Plus the weight transfer that's already there. This is going from state 3 to state 2 (or whatever it was).
Quote from Hyperactive :Big smelly crap that walks! I just noticed I had the set the diameter to 100 instead of the radius being 100... need to fix those numbers slip angle numbers as they are slightly off

Good to see, but it must be said that this doesn't invalidate your point in the slightest

The low speed spin problem was precisely that. It seemed to get worse the slower you were going, at very small turn radii. Even smaller than 50 or 100 radius.

Also, something that would actually increase the slip angle difference further is the spinning motion of the car itself. Your plots are steady state which is perfect for finding out at least what the minimum slip angle difference would be. As we add additional yaw velocity to spin the car (so we have some rate of change to the slip angles), the difference grows beyond that. So your mistake reduces the slip angle differences a bit, but the spinning motion that's been left out increases them again somewhat, so it may not be as far off as you might be thinking.
Quote from ajp71 :Nice to see a constructive discussion of vehicle dynamics for a change!

I'm really not sure what to make of the latest changes to iRacing, at first I really struggled to get to grips and subsequently managed to dip my SR below 1.0 and have now been demoted to a C license. After finally deciding to treat the cars as being completely new and different (and stopping entering races on untested combos and finding it surprising that I'm spinning off all over the place) I've found the new cars are actually quite driveable so long as you keep corrections to a minimum. I find you've got to be fairly aggressive with the cars on the NTM to actually make them understeer.

I really can't get on with the MX5 Roadster or the SRF on the new tyre physics. I never realised the MX5 Cup was actually different to the Roadster. The Cup now feels much better than the Roadster, which I think is at least in part down to the different fixed sets.

I haven't driven iRacing since shortly after the latest patch. I lose interest and take a break for a couple months occasionally.

I enjoy some of the cars and not so much others. My favorite as of the last patch was the HPD. That felt soooo much better than before. I like a car you can be aggressive with and don't have to constantly battle just to keep pointing somewhat forward. The FGT was a lot more enjoyable too. Enough to make me annotate my video with 'this has been fixed' anyway.
@hyperactive

I see your changes and the new plots. That looks more like it. Your point about not needing constant slopes is still correct though. The important part you showed was that the rear slip angles are bigger than the fronts. How much so then isn't too terribly important.

My feel for how much it probably would be comes from doing similar maneuvers in my hobby sim while displaying the slip angles. I want to say that in "slow" (whatever that means) slides that 5 degrees difference is a lot. That may be a memory failure though. Your data would be more reliable than that
Quote from jtw62074 :@hyperactive

I see your changes and the new plots. That looks more like it. Your point about not needing constant slopes is still correct though. The important part you showed was that the rear slip angles are bigger than the fronts. How much so then isn't too terribly important.

It looks like the yaw angle of the car doesn't change the slip angle difference between front and rear. The cornering radius does though. The tighter the corner the bigger the slip angle difference between front and rear. Which explains why people call the issue of there being too much drop off as "a slow speed problem" or "lack of grip at slow speeds". The radiuses of corners are of course smaller and the corners are tighter at slower speeds and that's where the slip angle difference really makes bad tires look bad. Of course the effect and help of downforce is minimized and the power of the engine is maximised at slow speeds as well which makes the issue even more visible

Here are few excel graphs about the corner radius effect on slip angle difference between right side front and rears. The tighter the turn the bigger the difference in slip angles between front and rear. The effect of yaw is pretty minimal. X is yaw angle of the car and Y is the slip angles.

I'm still saying that the bad tires alone are able to make it an issue but I need to check it with some better tire data. Your imaginery tire data was pretty horrible. I'm piecing together an example using the rfactor corvette c6r tire files (which felt pretty similar to the iracing one when I tried it) to see what happens (I think that does not have very good slip curves so it should be good candidate for this experiment). I'm even adding "yaw aero losses" to the mix. The rf car does not lose any aero downforce when the car yaws so that was really easy to add .

Anyways when you tried to slide the ford gt the cornering radius of the car decreased because of engine braking and tire slip... which increased the slip angle difference between front and rear. So unlike I said earlier it is not the increase of slip angle itself that adds to the difference of slip angles between front and rear but the car slowing down and the turning radius getting smaller that causes the slip angle difference between front and rear.

The big problem with my excel sheet is of course the limitation that it is just a stationary situation and the movement of the car and the inertias are totally missing.
Attached images
30.jpg
45.jpg
100.jpg
Quote from Hyperactive :It looks like the yaw angle of the car doesn't change the slip angle difference between front and rear. The cornering radius does though.

actually not as such and thats imho a pretty terrible and confusing way to think about it
the correct way that makes sense both in terms of how you think about slip and mathematically is to consider how the velocity vector of a tyre changes during cornering
namely additionally to the forwards vector (essentially just the speed of the car) you also get a vector perpendicular to the car that represents the speed at which the car rotates around its cog
assuming normal cornering for the front wheel that vector is pointing inwars (thus reducing the slip angle) and for the rear wheels its pointing outwards (increasing the slip angle)

naturally those velocities grow as the car turns torugh a tigher radius and roates round its cog in a shorter time
This thread is closed

iRacing
(13603 posts, closed, started )
FGED GREDG RDFGDR GSFDG