17 replies to this topic

[shaun979]

I just started looking at some data after recently putting on a decent system on some karts. The G-G plot is unlike any car I've driven of examined in that their longitudinal G under braking is so low relative to the peak laterals in a turn. With cars ranging from production road car to touring to formula, it is usually within a 5-25 % of each other (lat vs long), whereas with these karts with some fairly hard compound tires, we're barely crossing 1G in braking, but pulling 1.8 - 2.4 G laterally in turns. These are karts with good front and rear brakes, and the drivers are good too (not speaking for myself). The brakes have enough power to lock wheels if the driver so chooses, and there is enough modulation. Drivers speak of lock at either end being the braking limit, and when I drive the karts, I feel the exact same thing running similar laptimes.

The ellipticity of the G-G is even more surprising when considering that in most turns, these things have the inside rear tire either totally unloaded or very lightly loaded, leaving only 3 tires to take the load, whilst under straight braking all 4 are used. Do kart tires have negative load sensitivity in the load range (~100lbs at rest) they run? Or is it just the tire construction? I'm still waiting for my kart scales to arrive to get static loads and then measure CG with driver in kart, measure tire patch, calculate load transfers, to get some answers maybe.

I know there are some experienced karters here from Australia and Europe, and I hope they can add if they've seen the same in their data and possible reasons for it?

Hope this is not too far off topic.


Best regards

[shaun979]

Both X (lat acc) and Y axis (long acc) go from -3 to 3 G

[mariner]

As I know nothing about karts this may be a very dumb question but do the karts have a lot of front negative camber built in ( or does the caster angle vary this)?

If so it could be partially that the front tires work better in lateral G than longtitudinal G due to the negative camber benefits to tire grip. Presumably there is enough weight on the front to make the front brakes do much of the retardation work. This could acentuate the difference between the two results.

Just a thought.

[shaun979]

As I know nothing about karts this may be a very dumb question but do the karts have a lot of front negative camber built in ( or does the caster angle vary this)?

If so it could be partially that the front tires work better in lateral G than longtitudinal G due to the negative camber benefits to tire grip. Presumably there is enough weight on the front to make the front brakes do much of the retardation work. This could acentuate the difference between the two results.

Just a thought.

That's not a dumb question. Camber and caster are adjustable independently, and camber does vary with steer angle and caster level, but with the setups we run, the steering angle is small enough that there is no significant change to wheel camber in a turn vs straight... especially to make that much of a difference!

There is 40-45% front weight on these things at rest, more will transfer under braking, so I think the front should be able to do much of the braking.

[Greg Locock]

Is it that the 'suspension' rates are so hard that the weight transfer unloads one end of the kart too quickly for the driver to adjust to it? It seems to be a bit of a clue that the g-g plot is more or less symmetrical about the x axis, perhaps indicating that the limit is one axle's worth of grip.

[gruntguru]

I just started looking at some data after recently putting on a decent system on some karts. The G-G plot is unlike any car I've driven of examined in that their longitudinal G under braking is so low relative to the peak laterals in a turn. With cars ranging from production road car to touring to formula, it is usually within a 5-25 % of each other (lat vs long), whereas with these karts with some fairly hard compound tires, we're barely crossing 1G in braking, but pulling 1.8 - 2.4 G laterally in turns. These are karts with good front and rear brakes, and the drivers are good too (not speaking for myself). The brakes have enough power to lock wheels if the driver so chooses, and there is enough modulation. Drivers speak of lock at either end being the braking limit, and when I drive the karts, I feel the exact same thing running similar laptimes.

The ellipticity of the G-G is even more surprising when considering that in most turns, these things have the inside rear tire either totally unloaded or very lightly loaded, leaving only 3 tires to take the load, whilst under straight braking all 4 are used. Do kart tires have negative load sensitivity in the load range (~100lbs at rest) they run? Or is it just the tire construction? I'm still waiting for my kart scales to arrive to get static loads and then measure CG with driver in kart, measure tire patch, calculate load transfers, to get some answers maybe.

I wonder if this has anything to do with sidewall compliance - substantial in the lateral direction but minimal in the longitudinal, coupled with a chassis/suspension that is quite rigid in all modes except longitudinal torsion (which again offers mainly cornering/roll compliance).

Having said that, you would still expect to see spikes in the longitudinal G. Lack of compliance should only affect the ability to maintain max braking grip.

No chance there's a problem with your accelerometer I suppose? Are you sure the data includes some straight-line braking?

[Tony Matthews]

No chance there's a problem with your accelerometer I suppose?

Try turning it through 90 degrees!

[mariner]

One other thought

- I beleive tires usually generate more longtitudial grip than lateral grip because the nominal contact patch has less "breakaway" area fore and aft than sideways. The extreme version being a low pressure dragster tire.

- Kart tires are of very small diameter so the fore and aft patch area may be much less relative to the effective sideways grip area than on a larger rolling radius tire.

I am not sure I expressed that very well but if you extrapolated the reducing tire rolling radius to near zero I think you would see the effect in exaggerated from.

[dosco]

What are the factors that limit the braking performance?

Or asked differently, what sort of braking system would you need in order to achieve the braking accelerations that match the lateral (cornering) accelerations?

In my limited karting experience, I would say that the brake design (disc diameter, pad size, and pad material) would be a major contributor. Maybe also the hydraulic pressure.

[mep]

Are you sure your measured datas are correct?
Thats the first thing you should check.
I think 2.4g is a bit much for a cart.
Maybe the tilting of the cart has some influence.

[gruntguru]

Acceleration G's are similar to braking G's. Is this a traction-limited 4WD kart? Or is the accelerometer tilted and favouring acceleration readings at the expense of braking?

[shaun979]

Hi guys,

Installation is flat and certainly not favoring acceleration. It is a RWD kart and as lapped is just slightly traction limited because we're still trying to accelerate while turning to exit quick. The gearing is short (topping out 60+ mph, ~36 hp ) so the karts accelerate quite hard.

There are 3 big braking zones per lap where we brake as hard as possible because there is a good amount of speed to shed. I forgot to mention that the front brakes are controlled by the right hand, and the rear brakes by the left foot. I am going to try and get some higher level drivers to drive this kart and see what the best they can generate in braking is. Others and myself have already tried many times to get decel better, even ignoring the sector or overall lap, just exploring the brakes, but have been unable to. In the sessions where we drive together, there isn't anyone braking considerably harder than the rest, and there are some very good national and international race karters in the group. The brake point as we observe onboard and trackside, are for the most part identical. I don't think it is likely at all, that the 4 or so of us who have tried to brake better and generate better data are sucking so bad at it, but hey you just never know.. I'll keep looking for that super-braker.

I think the points about tire sidewall and patch are possible causes too. Braking loads affecting control are also possible although most of it is taken up by the lower body bracing against a very solid heelstop, leaving really only upperbody mass that must be supported against the steering wheel. The upperbody is fairly reclined too, so that helps the driver deal with it somewhat, but yes it is taxing on the core. Dosco, if the brakes are able to lock either end at will currently, with decent control, what reason is there to believe that the brakes are a limit to getting much closer to braking at 2G?

==

I don't think there is anything wrong with the accelerometer. Better tires on other courses have generated much higher lateral sustained accelerations. I haven't even collected data on the days where the track is heavily rubbered in and the entire field's laptimes improve by 3% as a whole. I'm not really interested in the exact peak accelerations, spikes and what not, but more the strong ellipticity of the G-G.

For those who question the accelerometer and its installation, below is some very heavily filtered GPS-derived acceleration data of the exact same laps that the accelerometer data was pulled from. The GPS filter is built into the software and I cannot unfilter it. The ellipticity remains, and while the lat and long acc are both diminished, lat acc still consistently and aggressively crosses 2.0.

Same -3 to 3 G on both axes...




Best regards

[mariner]

Not sure if this will help/work but

Here is an aerial view of the Rye House Kart track in the UK, which also has a helmet video run which allows a fairly precise lap position versus elapsed time count.There is one long straight ( at the back) where I would think full braking is used.

http://www.timpenton.../rye_house.aspx

The track is listd as "about 840 metres long and 40 seconds per lap. So it should be possible from the aerial view to get a rough idea of the corner radii and the length of each straight. I guess if you know enough about karts to guess the max speed then you could do so approx calcualtions on deacceleration and corner speed which would allow an independent sesne check on your puzzle. I doubt it is very precise vs data on board but if a rough check suggests the same relationship then your data is liley correct.

[Greg Locock]

You can back calculate track profiles from accelerometer data, or if you use Google earth you can map the circuit, apply your speed trace, and that gives you lateral accelerations. It is quite time consuming but possible, I have done it for Beckley Park GoKart Circuit in Geelong.

[dosco]

Installation is flat and certainly not favoring acceleration. It is a RWD kart and as lapped is just slightly traction limited because we're still trying to accelerate while turning to exit quick. The gearing is short (topping out 60+ mph, ~36 hp ) so the karts accelerate quite hard.

Dosco, if the brakes are able to lock either end at will currently, with decent control, what reason is there to believe that the brakes are a limit to getting much closer to braking at 2G?

Well, as I said before I am no expert.

Looking at it from first principles, at the tire the braking is a function of the braking force applied (at the disc), the coefficient of friction of the tire, and the vertical force (e.g. weight) applied at the tire.

If you can lock the brakes, then it's down to the tire's coefficient of friction and weight on the axle. You could add weight ... is there a minimum or maximum as defined by the rules? This might not be an option as addition of weight may negatively impact other performance areas. Last option is the coefficient of friction of the tire ... not sure you can do a lot there either. I have heard of soaking the rubber in "Simple Green" and other chemicals ....

[munks]

Is it that the 'suspension' rates are so hard that the weight transfer unloads one end of the kart too quickly for the driver to adjust to it? It seems to be a bit of a clue that the g-g plot is more or less symmetrical about the x axis, perhaps indicating that the limit is one axle's worth of grip.

That's a good observation, but it doesn't quite explain why the G-G plot is elliptical. After all, with the lateral G's so high, surely the cornering forces would cause even more severe weight transfer than in the longitudinal case.

Personally I tend to agree with the idea that it has to do with the tire size, shape, and orientation: basically, some combination of the front camber and small diameter tires. I also suspect it is nearly impossible to threshold brake, as the tires' longitudinal frequency response is probably quite high.

[Tenmantaylor]

I also think it's a result of kart tyres having a tread width more or less equal to their height giving much more grip laterally than longitudinally. The best example of the opposite I can think of is a racing bike tyre. They have a contact patch which is almost the exact 90 degree of a kart being a very tall, narrow tyre (and only have two of them). Racing bikes generate many more Gs in accel and braking than cornering, a vertically elliptical G readout over a lap I would think. Obviously bikes have a contact patch this way to allow cornering on two wheels but still it proves bigger tyre diameter allows greater longitudinal forces to be generated.

[munks]

I also think it's a result of kart tyres having a tread width more or less equal to their height giving much more grip laterally than longitudinally. The best example of the opposite I can think of is a racing bike tyre. They have a contact patch which is almost the exact 90 degree of a kart being a very tall, narrow tyre (and only have two of them). Racing bikes generate many more Gs in accel and braking than cornering, a vertically elliptical G readout over a lap I would think. Obviously bikes have a contact patch this way to allow cornering on two wheels but still it proves bigger tyre diameter allows greater longitudinal forces to be generated.

Could be. It has occurred to me that other wide tyres used in full-size car racing series probably exhibit similar lateral/longitudinal differences, but we don't notice them in the G-G diagram due to the longer wheelbase compared to the wheel track. On karts, I mean the tyres are almost at the corners of a square. In F1, there is a long wheelbase with the CG biased to the rear which makes braking somewhat optimally distributed.