tyre wear index to lateral g force - a magic formula!

Looking for new road/trackday tyres for my car I came across lots of references to numbers like "120" , "180" basically in describing the  compound softness.

Apparently   its an index developed in the USA. to measure grip in standardised way.. Relying on Wiki as ever its here   https://en.wikipedia...Quality_Grading

In the treadwear section the methodology of using a special control tyre and the candidate tyre  is explained.

The interesting bit is the formula to convert the index , 100, 180 etc, to actual grip. It is mu=2.25/TW^0.15. where TW is the treadwear index

So going from a TW of 180 to 100 in the formula suggests a 9% increase in grip , or at least  friction co - efficient..

That is huge grip increase just from compound change. Some tyres in the  MSAUK list 1b, i.e "road " tyres for racing, sold in  UK come in compounds of 100 to 180 so , if it is true I gain 9% extra grip by buying a variant of the same tyre?

This seems a bit too good to be true?? a clue might be here in the Wiki article "The assigning of UTQG grades is done solely by the tire manufacturer. In many cases, this has resulted in the UTQG grading system to be more of a marketing tool than was originally intended".

There is some evidence of treadwear ratings being gamed by tyre makers. Some classes of motorsport specify a minimum treadwear number eg autocross in the US have a category where tyres used must be at least 200 TW. Guess what - there are some pretty grippy tyres out there with TW200 on the sidewall.

Interesting, the formula converges to a mu of 2.25. I wonder if that is close to the mu for a top fuel drag slick.

Edited by gruntguru, 07 June 2021 - 20:57.

Have a read of SAE 942484. Back then (1994) the actual mu of a dragster tire was only 1.5, but the vehicle could accelerate at 3-4 g because of some weird stuff, which I am less than convinced by. I think the paper is on the right track, but i don't see how you can exert a vertical force into the ground greater than the axle weight for very long.

I've measured instantaneous longitudinal mu of 1.2 on very boring tires in an ABS stop. Perhaps some of that unexpected grip was the mechanism in that paper. For boring tires on the Flattrac machine long mu and lat mu are often the same or close enough, crude explanation being that when the patch of rubber is sliding it doesn't really care whether it is moving forward or sliding sideways.

I will have a read of that paper thanks.

TF dragsters have ~1000 lb of downforce from exhaust jet effect alone. Some of the jet effect is used for forward thrust by angling the nozzles (pipes). When you add in the inertial spike as the CG is lifted by tyre growth there is temporarily a lot of "DF" at launch. At higher speeds the wing takes over with reputedly 12,000 lb of DF at 325 mph. (car weighs 2,330 lb)

Edited by gruntguru, 08 June 2021 - 04:24.

Have a read of SAE 942484. Back then (1994) the actual mu of a dragster tire was only 1.5, but the vehicle could accelerate at 3-4 g because of some weird stuff, which I am less than convinced by. I think the paper is on the right track, but i don't see how you can exert a vertical force into the ground greater than the axle weight for very long.

 

I've measured instantaneous longitudinal mu of 1.2 on very boring tires in an ABS stop. Perhaps some of that unexpected grip was the mechanism in that paper. For boring tires on the Flattrac machine long mu and lat mu are often the same or close enough, crude explanation being that when the patch of rubber is sliding it doesn't really care whether it is moving forward or sliding sideways.

It was a good day when I realized that camber was not just something used to optimize tire temps. It was a way for me to compromise lateral and longitudinal grip to provide the biggest benefit for a given situation.

I am not an SAE member so I can only read the synopsis of 942484, but it seems to indicate that a drag car can pull 4 g while the " natural" tire grip is only 2g.

As I cant read it I have question - does the way a top fuel car leaps up and sort of arches its back on launch act a spring which as it unwinds can push the rear ned down beyond pure  weight.?

A bit like peak loads form vibrations in an assembly far exceeding the static load ?

4G is average acceleration over the course. Launch acceleration is 8G (due to the effect you describe plus tyre growth at launch). Acceleration would then reduce (to perhaps 2.5 - 3.0 G?) then climb again as the wing builds DF.

Accelerometer graph here https://dragnews.com...te-the-hardest/ doesn't confirm the high launch acceleration.

Edited by gruntguru, 09 June 2021 - 21:50.

I thought the exhaust thrust issue was taken apart here some time ago and found to be wanting.

I think we determined via simple Newtonian calculation that the massflows through the exhausts were never going to merit the grandiose DF claims being made.

I recall doing a calculation and 1,000 lb was the right order of magnitude. Can't recall the thread.

I think mass flow is about 10 kg/s so exhaust velocity would need to be 500 m/s to produce 5,000 N  - about 1,000 lb.

That 500 m/s velocity is of course well above speed of sound. DOHC did some quick maths and came up with a much lower number here. He uses 300 m/s.

That link tries to take us to the moderator panel and includes some IP info

Thank you Gruntguru for the G force track.

On the exhaust downforce thing one point I would ad which may be relevant is that the shattering noise of Top Fuel  car is partially  due to the excess nitro burning off in the exhausts. Hence the huge sheets of white flame seen in a night run.

That means a simple cubic capacity*revs* VE calculation may not be correct.

I know sometime ago the relative low power of extra F1 noise was discussed but Top Fuel noise is very different, it literally shakes your whole body if you are within 100 ft. 

The point about the speed of sound is valid. You'd also get mach choking at the tailpipe, I doubt they'd risk that just to get downforce.

 

I know sometime ago the relative low power of extra F1 noise was discussed but Top Fuel noise is very different, it literally shakes your whole body if you are within 100 ft. 

I went to an NHRA event when I was about 5 years old and hated it, purely because of the sensory experience. It was unpleasant for my tiny body.

The point about the speed of sound is valid. You'd also get mach choking at the tailpipe, I doubt they'd risk that just to get downforce.

You don't know many Top Fuel guys, do you?

You'd be right there - it is one the motorsports I've never seen in real life. So now you've got me thinking, do they dyno their engines? Or is the cost of engine per run so high that they do all their development on the track?

Nobody has ever sucessfuly dyno'ed a Top Fuel engine. Long ago somebody tried to hook on up to a aero engine dyno that could take up to 4,000 bhp but it didn't really work and Top Fuel power soon soared up to 8,00o bhp.

So most power was estimated using acceleration vs weight with a guess at drag - as getting coast down data is hard with a big chute on the back!

Then modern strain gauge technology was applied to teh drive shaft and real data could be captured. This led to finding out they really can deliver 11,000 bhp. 

Here   

I am an addict of Top Fuel and Funy Cars , the rest of drag racing doesn't grab me that much but the sheer mechical violence of a Top Fuel run is awesome. TV or the  internet cannot describe actually being there and feeling the noise. Not wise to watch without ear plugs but just once it is worth it.

.

Edited by mariner, 12 June 2021 - 09:19.

You don't know many Top Fuel guys, do you?

So is the exhaust velocity in the pipes supersonic or not?

A top fuel dragster runs a 2" fuel line. If that helps narrowing down the flow calculations (mostly what it can not be above due to Reynolds number) The compressor uses 600hp at full chat if that helps narrow down the amount of air getting stuffed in there.

The Air/fuel ratio can be as low as 1.7/1 so if they do not get ignition they usually go hydraulic and lift the head of.

Top Fuel is a must see once in a life time for anyone interested in racing/engines/wild sports/sound

Edited by MatsNorway, 12 June 2021 - 16:16.

Do you see mach diamonds in the exhaust? If not then it isn't supersonic. http://www.aerospace...ion/q0224.shtml

I have some old karting data that shows lateral G forces of 1.4G. This has to be about the max limit for pure force through tyres.  Though my data is old I seriously doubt whether there has been any further tyre development between then and now.

This is for full slicks mounted on wheels that optimise the rubber usage including wheel width that about limits tread walking under load so true mu values result.  Needless to say there are no aero tricks or exhaust loadings that can affect G loads.  Karts frame/seat locations are also optimised so as to get the best tyre temperatures across treads and right to left.

I used to have some good F Ford on track data but that is too long ago for me to go find but I recall that data pretty much validating, but not equaling the kart data.

There was a recent tyre comparison published in Oz that gave volumes of good data for DOT approved road car tyres. The best dry braking tyres were Maxxis Premtra 5 that showed 1.04 G forces from 100 kph to full stop.  There were six tyre makes that all were above .96G.  The test Vehicle was a Hyundai Tucson so you can expect better test numbers from tyres mounted on higher performance vehicles with less compliant suspensions. At least these are real life numbers.

Regards

Some dynomometer testing has been done using one cylinder of a top fuel engine.

With combustion occuring in the exhaust pipes, you would expect the exhaust temperature to be at least 1000*C and the speed of sound more than 700 m/s.

 

I think mass flow is about 10 kg/s so exhaust velocity would need to be 500 m/s to produce 5,000 N  - about 1,000 lb.

So is the exhaust velocity in the pipes supersonic or not?

That question is well out of my wheelhouse. I have no idea, although I believe I have seen pics of Fuel cars with compression diamonds visible in the exhaust.

My comment was more as to whether or not they would risk the test. If those guys have an inkling it will pick up 0.001 sec they'll try it. If you're not blowing things up on a fairly regular basis, you're probably not trying hard enough.

As far as I know, all of their full power running is on track. I don't believe there is a dyno built that could actually measure the power of a TF engine. The flow bench work that they do on *everything* is damned impressive.

As a long time racer the lap times set on full race crossply slicks on 10" wide rims in 1990 are far slower than 'semi slick' 'DOT" tyres on a 7" rim in 2000.

And these days they are a deal faster again.

I sell Nankang semi slicks. And get so many enquiries from people wanting to drive on the street with them. Or want drag race sizes that are not generally available. I still hear the crap about 10"  tyres or 12" tyres as they think wider will be faster. Which they are not. Case construction, bead construction and rubber compounds is the go fast items. Though all decent tyres these days are turning on the rims. I mark them with paint pen and have seen the outside has turned more than the inside. Back in the 70s generic aftermarket wheels often had knurling on the rim edge to stop it.

Hoosier do make some tyres that crossover from Drags to road race. I tried some and could not get any sense from them but many others have gone very quickly on them. Both on road courses and Drags. On the same tyre!

I have sold a few Nankang for street drags and had no complaint, though for slower styles of cars. And these days 12s are slow.

Tread wear numbers are an indication of speed from the tyre, but far from the be all end all.

With all of these tyres inc the drag race styles of DOT they make speed differently as well as how many heat cycles you will get from them.

Most of the road race styles are a very stiff case and are also quite heavy. Of those some take longer to 'come on'   or 'go off'.

The tread wear may give an indication of how long they last. Many want harder ones so they last longer. Though that is generally not a good idea as they simply go hard and will not work anymore.

Then we can start on Speedway tyres!!

That question is well out of my wheelhouse. I have no idea, although I believe I have seen pics of Fuel cars with compression diamonds visible in the exhaust.

My comment was more as to whether or not they would risk the test. If those guys have an inkling it will pick up 0.001 sec they'll try it. If you're not blowing things up on a fairly regular basis, you're probably not trying hard enough.

As far as I know, all of their full power running is on track. I don't believe there is a dyno built that could actually measure the power of a TF engine. The flow bench work that they do on *everything* is damned impressive.

Agreed with the work they do for power. And to run it on the dyno for 1 or two pulls will wear it out!

While the speed is impressive the engine rebuild between runs is not.

On You Tube their is a long series called 'The Surfers' about the exploits of a couple of slightly 'alternative' Californians who built a top fuel car in the 60s with Mike Sirokin driving that won some big races. Worth watching even if you have no real interest in drag racing.

I have seen a couple of articles with Don Garlits who was a leader in what he did but he simply cannot live with the waste in modern racing

I followed up on some of the SAE tire articles  and found one from Gave Hullam " understanding race tires" 9039268 in 1998.

Two things in it were new to me. 

Firstly  he seems to argue that a race tire will develop less vertical load with rotratioanl speed because the radius is reduced by the flat contact patch. Therefore the sum of centrifugal forces in the top half of the tire , with no flattening, will exceed those in the lower half which has a slightly less average radius. 

Now that sounds plausible but I have never heard that point made before in my , albeit limited, tire readings.

Secondly he argues that negative camber results in side thrust with tire tread heating because there is no sideways distortion of the tire only an unequal force caused by one side of the tread having a smaller radius than the other. So as the car rolls onto the outside tire you get cornering force without  tire heating.

Again I can see the logic of no, or at least less , tread distortion but I thought the limit on negative camber was excessive inner edge temperature  as per the F1 fuss a few years ago which resulted in Pirelii limiting negative camber allowed.

For both of these am I a) missing something and b) have I misunderstood what he was saying ?

I followed up on some of the SAE tire articles  and found one from Gave Hullam " understanding race tires" 9039268 in 1998.

 

Two things in it were new to me. 

 

Firstly  he seems to argue that a race tire will develop less vertical load with rotratioanl speed because the radius is reduced by the flat contact patch. Therefore the sum of centrifugal forces in the top half of the tire , with no flattening, will exceed those in the lower half which has a slightly less average radius. 

If the sum of the vertical loads on the four contact patches is less than the weight of the car it will fall through the road under gravity.

Edited by gruntguru, 30 June 2021 - 02:46.

I agree with Gruntguru but to help hre is teh relevant data 

 TIRE SPRING RATE CHANGE WITH SPEED - At speed the vertical momentum loss also affects the tire deflection and ride height (axle height). In the example above the load on the road is approximately 2380 Ibf (the • P*A term, 2000, plus the vertical momentum force, 380). The static tire load at the same deflection is 2000 Ibf. Tire spring rate is higher at speed. The tire deflection for the 2380 Ibf ground load tire at 220 ft/sec is, dR = 1 3.5*(1 -Cos(9.7)) = 0.1 93 in (Eq. 9) At static conditions the contact patch length of a 2380 Ibf load at 40 psi is, L = 2380/(40*1 1) =5.41 in (Eq. 10) the contact angle is, ß = Arcsin(5.41/27) = 1 1 .56 degrees (Eq. 1 1 ) and the static tire deflection is, dR = 1 3.5*(1 -Cos(1 1 .56)) = 0.274 in (Eq. 1 2) At a speed of 150 MPH the tire deflection reduces 0.081 inches from the static conditions to 0.193 inches. "Average" static tire spring rate with 2380 Ibf load is 8696 lbf/in (2380/.274), which increases to 12,332 lbf/in (2380/.193) at 150 MPH. Not all of the vertical impact force calculated is lost to the ground. Some energy goes into side wall and tread bending and is partially recov- ered. The actual tire spring rate may not change the theoretical amount. Theoretical tire deflection is below actual, so the theoretical tread impact angles are low affording some unknown compensation. 

I am sorry I cant link to the articel  but its under copyright.

My own simple explanation of why there is not a vertical load due to the contact patch is that the contact patch may be closer the the wheel centre at the mid point but since the tread circumference is fixed but flexible the ENDS of the contact patch are further from the wheel centre thus cancelling oute h shorter mid point. Clearly there is still a change in ride height as sped i.e centrifugal force grows but its not an actual imbalance causing the 380lb vertical load he calculates 

But I am not an engineer and this is an official SAE paper so presumably peer group reviewed?

Judging by the many SAE papers I've read, I don't think peer review is even welcomed much less required, most read more like PR fluff than hard science.

SAE papers are not reviewed. In the 80s and before they were fantastic, if dense, sources of useful information. These days half of them are just adverts and you'll notice a conspicuous lack of useful data and code.

I'm scratching my head about the suck down of the tire giving extra force, as grunt says, the FBD of the whole car seems to be awry.

I'm guessing that the extra 380 lb is a virtual force that accounts for the reduced tread deflection.

Do you see mach diamonds in the exhaust? If not then it isn't supersonic. http://www.aerospace...ion/q0224.shtml

Would there be mach diamonds for pulsating exhausts too? Sounds reasonable to get them easier with stable flow.

btw. Top fuel does not only shake your body, it shakes your eyeballs. It hits the body to the core, the instinct to run for cover kicks in.

Edited by MatsNorway, 30 June 2021 - 10:02.

From my experience having to use road tires in racing, tread wear numbers are not comparable brand-to-brand.  They are model-to-model within a brand though.

From my experience having to use road tires in racing, tread wear numbers are not comparable brand-to-brand.  They are model-to-model within a brand though.

Not many classes on a road tyre. And if they are usually a control tyre. 

With HQ Holden racing initially a full tread tyre [Goodyear Ducaro] would chunk big pieces off them. They allowed [what was happening anyway] to buff the tyre down to just above mininum tread and then the tyres would live. BUT being a Badday had very few useable heat cycles. They went off and were useless. Many ended up on car trailers.

Subsequently they went to Bridgestone RE92 which was far better with heat cycles. Then to Kuhmos which were so hard were not great. But wore ok.

Now they are on Kenda which appears to work ok and last ok.

This over a near 30 year period. HQs are getting rare now so more expensive for parts and panels. I do notice they play nicer now!

We also have a Hyundia Excel class. Near standard cars on a control tyre.

A top fuel dragster runs a 2" fuel line. If that helps narrowing down the flow calculations (mostly what it can not be above due to Reynolds number) The compressor uses 600hp at full chat if that helps narrow down the amount of air getting stuffed in there.

 

The Air/fuel ratio can be as low as 1.7/1 so if they do not get ignition they usually go hydraulic and lift the head of.

 

Top Fuel is a must see once in a life time for anyone interested in racing/engines/wild sports/sound

Dont forget Funny cars, I believe they use bigger engines so possibly more power. In a far shorter wheelbase car with a 'body' over it. And very similar times. With the driver behind the engine!

Well sometimes the bodies are over the cars, they often resemble convertibles or less when the engine explodes and throws the body away. Bodies that are tens of thousands in carbon fibre

Thanks to my son ploughing through JSTOR and even darker parts of academic libraries I have now read the 1994 "drag tyre magic" paper.- he calls it tire momentum theory.

If I understand it correctly(?) the "trick " is simple. The effective contact point of the tyre is not at the bottom of the tread but up on the front side. The huge sidewall ripple drops the rim down and so the tread spreading forward results in large vertical thrust vector from the rim torque. If the front , heavily compressed i.e. loaded contact point, was at 45 degrees from the horizontal the vertical load from the engine torque would be 50% and the horizontal component also 50%. 

Given the tyre has a 1.5 G capability it can give more thrust than load so the huge 5g horizontal drive force at the track comes more from the rim torque than the vehicle weight.

It’s something like downforce as its " free" but it requires huge torque to work. As TF dragster has 4,000 ft lb of engine torque and 3.2:1 axle ratio it all works. 

There is one extra tyre ripple effect which is that eh effective wheel radius is very small. Maybe only 5 inches since the rim pulls at the sidewall and reduces the effective diameter.

To use some crude numbers the tyre can give 1.5G and you need 5G so a ratio of 3.33. The car weight is 1 ratio so you need 2.33 extra. With a 2,500lb car 2.33*2,500lb is 5,880lb The axle torque is 4,000 ft lb* 3.2 for 13,200. So 5880/13200 is 45%. Now that isn’t enough because the static tread radius is 18 inches so 13,200/(18/12) is 8,800 – not nearly enough for 5g.

The other factor is the effective radius of the wrinkle wall is estimated at 5 inches so its really 13,200/(5/12) = 31680 ft lbs. So 5880/31680 = 0.186 vertical share of tyre tread torque.

 I hope I have done justice to the author with my very simplified example

Edited by mariner, 02 July 2021 - 13:37.

How to put that into a FBD? I get close if I imagine an opposing force vector pulling down on the rim near its back edge. This opposing vector is the result of the stationary contct patch being accelerated up off the road - that change of momentum requires a force.

Would people accept that on average the net downward force exerted at the tires is exactly equal to the car's weight, if there is no aero?

Yes a FBD of the entire vehicle confirms that MUST be the case.

My head is thinking there MIGHT be a scenario where the peak contact pressure near the front of the contact patch (generated by the tread striking the ground with a significant vertical velocity component) might also have a larger horizontal slip ratio than the average for the contact patch and therefore be generating most of the thrust. Could this generate a higher total thrust than obtained by multiplying mu x (average contact pressure) x (total contact area)?

I don’t pretend to understand all of this but I think you have to start with the observed data. A TF car clearly reaches 100 mph in under one second from both in-car and trackside data logging. That is 4.6G. Nobody seems to believe that tyres can get beyond 1.4G so over 3G needs explaining. 

At an average of 50 mph wing force is not large and nobody seems to assign more than a few hundred lbs to exhaust downforce. So they might get 1.4G near to 2G for a 2,500 lb car but nearly 3g of measured acceleration is missing and has, logically, to come from INSIDE the car.

The tyre force bit is easy -ish . The axle torque at tread face purely horizontal in an ordinary tyre but if you sliced that tyre horizontally across the axle centreline the tread force would be 100% vertical. So at 45 degree thrust angle it would split 50/50 and so on.

As a TF engine  produces 11,000 bhp at 9,000 rpm  it must have about 6,000 ft /lb of torque. The gear ratio is 3.2:1 so the axle torque is 20,000 ft lb or eight times the weight of the car. Even at a 20 degree tread face thrust angle that is a huge downward component.

That’s the easy bit the hard bit is how the force is all inside the car without any external reaction like downforce?

Mr Newton taught us that each action must have an opposite, and equal, reaction so where is the other end of the 20,000ft lb of tread torque.?

A simple suggestion its inside the cylinder heads which is why a TF engine has many huge cylinder head bolts. If they need to be so strong then applying Newton there must be an opposite reaction to the head bolt load and it’s at the tyre and partially downwards so increasing downforce?

Edited by mariner, 04 July 2021 - 10:32.

Getting from 1.4G to 3G is not that difficult.

1. I doubt the 1.4 mu number for TF slicks. 1.4 mu is seen in many classess of racing and is even approached by some street legal semi-slicks. So - the number is higher for drag slicks. Then you have "track prep" - significantly increasing mu. Try walking on a prepped track - shoes stick to it. So lets say mu = 1.8

2. Exhaust DF reputedly 1000 lb on a 2000 lb car +50% so mu = 2.7

3. The cg of the car climbs significantly at launch - lets say 100mm in 0.1 sec. That requires a vertical acceleration of 2G - effectively tripling the vertical force on the contact patch. 2.7 x 3 = 8.1 G (for 0.1 seconds and 2.7 thereafter)

External forces applied to the vehicle as a whole must be reacted by other external forces. Head bolt tension is an internal force.

Gruntguru , I entirely see your logic but  - 100 mph in under one second is, 88ft/sec at 60 mph *100/60 = 147 ft/sec/32.2 = 4.55 G for the  first second not 3 G and I struggle to get all 4.6G via exhaust downforce etc. 

To be honest I am not sure that I fully believe my explanations above but the alternative does seem to be that , as you say, 1.4G is plain wrong for a drag tyre. That implies the vertical force vs lateral force curves beloved of set up models don't apply.

So see if the following is an explanation of 4.6G.

Take the "sticky tape test" used to show why a tyre can generate more lateral grip than the vertical load. You put 5 cm on a flat surface and try and pull it off horizontally - very hard. Now pull vertically and it peels off easily. So its about the inter molecular attraction of millions of molecules sideways but few vertically etc.

Now try adding three extra layers on top of the one in contact with the surface. If tyre grip is mainly a function of vertical force it should be three times harder to pull it of sideways but it isn't because the molecules in contact between the surfaces didn't increase. Now make the contact strip 20 cm not 5 cm  - if it is harder to pull off then the conclusion of these tests is that tyre grip has a surface area component independent of vertical load and that's why the TF car gets 4.6G - the 1.4G limit isn't a limit at all??

BTW to avoid  relying on my amateur explanations here is a preview copy of the article .

You can't see the tread impact angle he describes as key in the preview but it is calculated as 35.4  degrees so the maths people may be able to reproduce his logic.

https://www.sae.org/...942484/preview/

Edited by mariner, 05 July 2021 - 09:41.

Gruntguru , I entirely see your logic but  - 100 mph in under one second is, 88ft/sec at 60 mph *100/60 = 147 ft/sec/32.2 = 4.55 G for the  first second not 3 G and I struggle to get all 4.6G via exhaust downforce etc. 

 

To be honest I am not sure that I fully believe my explanations above but the alternative does seem to be that , as you say, 1.4G is plain wrong for a drag tyre. That implies the vertical force vs lateral force curves beloved of set up models don't apply.

 

So see if the following is an explanation of 4.6G.

 

Take the "sticky tape test" used to show why a tyre can generate more lateral grip than the vertical load. You put 5 cm on a flat surface and try and pull it off horizontally - very hard. Now pull vertically and it peels off easily. So its about the inter molecular attraction of millions of molecules sideways but few vertically etc.

 

Now try adding three extra layers on top of the one in contact with the surface. If tyre grip is mainly a function of vertical force it should be three times harder to pull it of sideways but it isn't because the molecules in contact between the surfaces didn't increase. Now make the contact strip 20 cm not 5 cm  - if it is harder to pull off then the conclusion of these tests is that tyre grip has a surface area component independent of vertical load and that's why the TF car gets 4.6G - the 1.4G limit isn't a limit at all??

   I think this is right - the tyres are very sticky and the drag strip surface is also covered in sticky  "traction compound".  .  I think the basic laws of physics would say that a tyre alone  cannot develop downforce.