Measuring Grip

[aviatorflightsims]

Hi,

Just stumbled on this site and it looks great! I have a question I have been trying to get an answer for for ages and hope you guys might be able to help. How does a team determine the level of grip of a tyre -- for example on the limits through a fast turn? Is it determined by the lateral force component measured in something like ATLAS or MoTeC (miniature spikes in the telemetry) or is there another way? Thanks for your insight.

[A13EX_f]

a combination of accelerometers and data from the steering input requested by the driver from this they would be able to tell if the car is under-steering / over-steering etc. more often the driver will also know the limit of the tyre grip when the rear end of the car steps out and he has to put in opposite lock, this will also show up on the sensors measuring the steering input.

[Jersey Tom]

First step to measuring something is defining what exactly it is. My own feelings on this topic can be found in my sig.

[McMrocks]

do you mean the friction coefficient? it is f and it tells about the relationship between the gravitation force and the friction force.

with this number you can calculate the corner speed if you have the corner radius

[aviatorflightsims]

Sorry -- yes when trying to determine how a car which experiences grip/no grip/grip etc -- on the limit through a turn -- how that limit can be expressed in telemetry data. So in effect taking out the driver from the feedback experience, and descerning it from analysing the telemetry.

[Jersey Tom]

Sorry -- yes when trying to determine how a car which experiences grip/no grip/grip etc -- on the limit through a turn

Well, what do you mean by "grip" and what do you mean by "no grip."

[DaveW]

Sorry -- yes when trying to determine how a car which experiences grip/no grip/grip etc -- on the limit through a turn -- how that limit can be expressed in telemetry data. So in effect taking out the driver from the feedback experience, and descerning it from analysing the telemetry.

May I ask the type of vehicle? Tin top, perhaps?

[aviatorflightsims]

Yes, exactly that! Basically a Corvette in iRacing.

[Jersey Tom]

I think you have to be very specific in what exactly it is you want to measure or evaluate. "Grip" is a word that I try to avoid at all costs in real world engineering.

[aviatorflightsims]

It's really the point at which the tyres slide over the surface, losing traction, then regains traction. I'm basically trying to evaluate the point at which the tyre model is on its limit, and this is the key point at which a tyre model in a sim is usefully judged.

[DaveW]

Yes, exactly that! Basically a Corvette in iRacing.

In iRacing... That is impressive. I think I may have an explanation.

If a real vehicle is excited on a rig in roll, I always see two "roll" modes. The lower of these involves sprung mass roll (of course), but coupled with a lateral deflection phased so that the roll centre is below the ground. The motion requires the tyres to deflect laterally. The tyres are not rotating on a rig, so cannot "relax" the load away. For a long time I thought that the "mode" was not important for a car with rotating wheels, & I have other evidence to support that view.

Now consider a real vehicle loaded up in a turn, with the tyres deflected both vertically and, crucially, laterally. The dynamic state of the vehicle is rather similar to the state the vehicle would be in on a rig, preloaded to the start of the mode I described above. When one or more tyres let go (start sliding), the "mode" is released, so a dynamic response will be added to the subsequent vehicle motion. The forces involved can be quite large if the roll inertia is large (hence my question).

Here is the tricky bit. The motion will be sinusoidal and a cycle should end at the starting condition with the tyres loaded up again. If the motion is large enough and is insufficiently damped, then it can repeat indefinitely (be unstable), until the turn is released. That unstable motion can certainly happen in Touring Cars.

OK. That is my rather clumsy description of what might be happening to your vehicle. The reality is actually rather more complex (it involves both roll & pitch for example, & I suspect the tyres also impose a destabilizing vertical input).

It is up to you to devise a way of "measuring" it.

[Jersey Tom]

Biting my tongue here...

Edit - now see Dave's post. I agree, up to you to figure out how you think you can measure it. Think about what the car and driver are doing specifically at that area you're interested in... think about the data channels you have available and what they mean... and see what you come up with.

Wish I could talk about it more.

[bill shoe]

Hi, How does a team determine the level of grip of a tyre -- for example on the limits through a fast turn?.

They hire a vehicle dynamics evaluator, such as a tire evauator, to drive the car. The evaluator feels subtle aspects of car performance that can't be discerned with lots of DAQ and PhD's. The evaluator writes a short report and gives it to the team. This is the actual information in question. Sims are just a way of estimating or approximating this information. All else being equal you want a better evaluation driver and a better sim person.

If you want the information itself ("level of grip of a tyre") then the eval driver is far more accurate and efficient. If you are trying to develop a better sim model then the evaluator is useful but the sim person has to figure out the correlations. If DaveW is only at the hypothesis level with this then you know it's difficult.

[Greg Locock]

Well I suppose that is one way of doing it.

Alternatively if you measure the ground velocity in x and y at each end of the car, and the steering wheel angle, and the latacc and longacc at the cg, you can certainly see a lot about what each axle is up to, and hence what the average tire on each axle is doing. That probably won't help much with the sort of limit behaviour DaveW is writing about, but if that was happening in a race I'd be inclined to sort the problem out rather than worrying about the exact instantaneous mu (or grip) of the tire while it is happening, although that number may be an important part of the system model.

[Max Trenkle]

How to measure 'grip' of a tire:

Well tire 'grip' is useful in two directions: longitudinally & laterally. The first being forward/backward, and the second being left/right. The combination of these two directions is what allows the car to travel in curves.

Tires have a coefficient of friction... that changes depending on the forces in the Z direction (up/down).
Forces in the Z direction can be related to weight of the car, aerodynamics forces, unsprung mass of the car, etc. All of the different forces affect the tires differently when coming from different sources. This you can learn more about in further study of vehicle dynamics.

In a corner, your limiting factor is lateral acceleration. Let the 'grip' forces be define in Y direction (left/right).

ForceY = Coefficient of Friction * ForceZ

Note that more ForceZ will increase ForceY with a constant Coef of Friction. However, take note that earlier I said that the Coef can change depending on the properties of the tire and how they respond to changes in load.

This is why Aerodynamics play such an important role in going fast. Maximum Mechanical 'grip' in lateral acceleration is around 1.7 g's. Maximum Aerodynamics 'grip' can exceed 4-5 g's. Tires are a big deal.

So, in conclusion:
Tires are complicated, and dynamic. They don't have a 'max grip'. Well they do... but it's only limited by factors that are wayyyyyyyy out of this scope.
Good luck, you have a lot of reading to do.