RE: Mark Hales on Technique: Shifting the weight about
Discussion
NJH said:
How on earth does a tyre know the difference between download caused by aerodynamic force and download caused by accelerating mass aka weight transfer?
Note this is not a philosophical question.
A singular tyre 'knows' the difference as the download associated with weight transfer is at the track / tyre interface and is caused by acceleration or braking. This is not the case for aerodynamic download, which is through the hub (or wheel / tyre interface to be more accurate)Note this is not a philosophical question.
Gemaeden said:
Okay. I've made mistakes before and no doubt I'll make more in the future, but hopefully I'll gain a better understanding from the responses I get.
If someone can show me the error of my assumption through data rather than theory, or anecdotal descriptions of feel, then I'll be happy to change my belief and stop upsetting people.
Indeed a good post - and you're not upsetting anyone (at least not me..)If someone can show me the error of my assumption through data rather than theory, or anecdotal descriptions of feel, then I'll be happy to change my belief and stop upsetting people.
I'll try / a few thoughts, and I hope they're coherent:
First - the maximum lateral acceleration of the car will not be improved by weight transfer. It will be at it's max when the tyres are all at their limits which implies a 50/50 distribution of load for a 50/50 mass distribution in a steady state/neutral throttle (being careful with terms). Intuitively that fits reasonably with the friction circle and the idea that you have maximum cornering ability at a reasonably neutral throttle.
Second, the size of the friction circle is not static. You can prove that with the miliken numbers - download x friction coef gives you your grip; while it is not linear, adding download increases grip (witness the popularity and prevalance of wings - they simply provide mass free download) So, to some extent this answers your question regarding friction circle plots. There is a trade off - up to some point the weight transfer increasing the size of the circle is more advantageous than what it takes away in lateral ability - a smaller slice of a bigger pie. At some point that trend will reverse. Obviously at the point the tyre locks there is zero lateral capability. Off the brakes it's all lateral, but a smaller circle
Note also that this happens in reverse at the rear - braking does not increase the overall lateral capablility of the car - in fact it reduces it - due to the non-linearity you've picked up on, the combination of a highly loaded tyre and a lightly loaded tyre is less than 2 evenly loaded tyres. So we just change the balance. All the yaw in the world won't help you if you're beyond the lateral limits, so you don't see a 'spike' in lateral capability, it's more about being able to use what you have; Equally if you had a car that oversteers badly, trail braking would be no help at all.
Next, take into account the specifics of the formula ford car. Particularly it has an open diff and traction issues. You could set it up to be neutral, but it would then spin its wheels on corner exit. To improve its ability to put the power to the road, you need to increase mechanical grip at the rear - soften the rear suspension and particularly remove roll stiffness. Problem with that is you try to control all the roll at the front; that transfers a lot of load at the front compared to the rear, reduces the ability of the fronts to generate grip and you have a car that in a steady state does not want to turn - it has less grip at the front. aka understeer. When turning the fronts will saturate before the rears, so you quickly reach a point where turning the steering more won't help - but the rears aren't doing their share, so you aren't turning optimally. In order to do that you have to (temporarily) put some more grip back on the front to enable the car to turn, ie trail braking/shifting some weight onto the fronts. You're using it not to get more lateral capability out of the car, but to overcome a setup that is bad for turning - you live with / drive around that setup in order to optimise another aspect of the car (traction)
Gemaeden said:
I mentioned weight transfer being a red herring, as I didn't see it as increasing grip per se, merely distorting the sidewalls.
I have to see that as a second order - I can't really see how download would reduce the stiffness of the sidewall - more logically to me, download gives you more grip which enables more sideload on the tyre (download x friction coef again).. which leads to more sidewall deflection and a greater slip angle for peak lateral. At least, that makes sense to me. I may be wrong.Gemaeden said:
My reasoning was framed by looking at Formula Ford and such like non-aero friction circle graphs.
If putting weight on the nose improved turn in, then why is there not a scattering of plots showing good lateral acceleration at near maximum longitudinal braking g on such graphs?
Surely the increase in yaw rate would show up as lateral acceleration in this area.
What I see in these graphs is like a scribbled 'T' shape, whereas what I would expect to see is some fanning out at the bottom of the 'T' if grip were increased by weight on the nose.
See my wordy pontificating above - we're just using weight transfer to enable us to get the car to optimum yaw. If we weren't setup to understeer, we could do that with the steering wheel. As we are understeering, we need another trick to get the rears to optimum yaw. That does infer a bit of deceleration, but it's not dramatic as someone else noted, so it won't be rammed forward on the circle.If putting weight on the nose improved turn in, then why is there not a scattering of plots showing good lateral acceleration at near maximum longitudinal braking g on such graphs?
Surely the increase in yaw rate would show up as lateral acceleration in this area.
What I see in these graphs is like a scribbled 'T' shape, whereas what I would expect to see is some fanning out at the bottom of the 'T' if grip were increased by weight on the nose.
Gemaeden said:
NJH said:
How on earth does a tyre know the difference between download caused by aerodynamic force and download caused by accelerating mass aka weight transfer?
Note this is not a philosophical question.
A singular tyre 'knows' the difference as the download associated with weight transfer is at the track / tyre interface and is caused by acceleration or braking. This is not the case for aerodynamic download, which is through the hub (or wheel / tyre interface to be more accurate)Note this is not a philosophical question.
Aero download is acted on the car, through the springs/dampers into the hubs and so to the contact patch - as you say.
'Weight transfer' is caused by good old newtonian physics - the car wants to keep moving, but the external force - braking at the tyre contact patch is resisting that. It is indeed acted through the contact patch - if the centre of mass was inline with the contact patch, there would be no weight transfer.
It isn't however, the COM/COG is above, so you have a forward push high up, and a backward push low down. The entire system wants to rotate nose down - watch the nose of the car dive when you jump on the brakes.. That rotation is opposed by the wheels acting on the body via the suspension which puts download on the tyres - just as in the aero situation, albeit with a bit of retardation too..
Edited by upsidedownmark on Friday 21st July 09:35
Thanks for taking the time to reply. Since posting I've done further research which has improved my understanding.
Where I might be going out on a limb, as far as most people's understanding goes, is to say that maximum (non-aero) grip is actually generated when the car is stationary.
My reasoning for this is that when the car is stationary, the tyres have more time to deform and fill the gaps at the tyre road interface. The better filling of the gap, the more grip. This is why cars with soft tyres or those that have been standing for a long time are stickier.
This could be tested by applying a lateral or longitudinal force through the c of g of a stationary car and comparing the measurement with measurements in motion. It's something I don't have the facilities to do.
The fact that the c of g is kept as low as possible on most performance cars to aid cornering points to this being correct.
If weight transfer were at all advantageous then there would be no reason to keep the c of g low. While it may add an advantage in certain circumstances (which I doubt) it is not of an advantage during a complete lap.
My point is that a driver senses the weight transfer and the improved turn in, but in pretty well every description of trail braking on the web the effects of braking and slowing are not given as the rightful cause of improved turn in. It is attributed to weight transfer, which is an effect rather than a cause.
When I look at things from the point of view of basic physics and the first law of thermodynamics, that of energy conservation, what happens is that velocity component of the kinetic energy (ke=1/2mv^2) that is trying to make the car go straight on (velocity is a vector) is converted into heat energy in the brakes.
Newton's second law of motion F=ma is thus applied. As the velocity reduces the vector force (steering) required to accelerate the car laterally reduces. That is what allows the car to turn more sharply. If the original steering is not reduced, the lateral acceleration will increase and it is sensed as oversteer.
It is changing the velocity that has a relatively huge effect as it is squared, which is why I pointed this out in my first post.
When I look at Newton's third law of motion. I see that any increased braking effected by the front tyres (due to weight transfer) is opposed by an equal and opposite force. i.e. The greater the load on the front tyres, the greater the force they will have to resist, which means that there is no net benefit. In fact....
...thanks to the first law of thermodynamics again, some of the equal and opposite force results in frictional forces within the tyre and suspension. So the braking force that is absorbed by them is turned to heat. Which is presumably why Milliken saw a reduction in the coefficient of friction with increasing load. (See my first post).
It also explains why vehicles with low centres of gravity brake in a shorter distance. As loads are distributed more equally between the tyres and therefore there are fewer losses in efficiency due to internal tyre (and suspension) friction.
I would also like to say that I like to apply Occam's Razor to physics questions. To my mind this is the simplest explanation that fits, and so is the one I am tempted to believe is right. Until someone can show me where I've made a mistake, or provide an even more simple explanation of course.
upsidedownmark said:
First - the maximum lateral acceleration of the car will not be improved by weight transfer. It will be at it's max when the tyres are all at their limits which implies a 50/50 distribution of load for a 50/50 mass distribution in a steady state/neutral throttle (being careful with terms). Intuitively that fits reasonably with the friction circle and the idea that you have maximum cornering ability at a reasonably neutral throttle.
Here we largely agreed, particularly about maximum grip (both laterally and longitudinally I understand) being available when the weight is evenly distributed. Where I might be going out on a limb, as far as most people's understanding goes, is to say that maximum (non-aero) grip is actually generated when the car is stationary.
My reasoning for this is that when the car is stationary, the tyres have more time to deform and fill the gaps at the tyre road interface. The better filling of the gap, the more grip. This is why cars with soft tyres or those that have been standing for a long time are stickier.
This could be tested by applying a lateral or longitudinal force through the c of g of a stationary car and comparing the measurement with measurements in motion. It's something I don't have the facilities to do.
The fact that the c of g is kept as low as possible on most performance cars to aid cornering points to this being correct.
upsidedownmark said:
Second, the size of the friction circle is not static. You can prove that with the miliken numbers - download x friction coef gives you your grip; while it is not linear, adding download increases grip (witness the popularity and prevalance of wings - they simply provide mass free download) So, to some extent this answers your question regarding friction circle plots.
It may be semantics, or a complete misunderstanding somewhere, but I would argue that the friction circle is static, ie. It has ultimate limits. But I would agree, and this may well be the point that you are trying to make that a vehicle's point on it varies with download (as well as steering and and speed)upsidedownmark said:
There is a trade off - up to some point the weight transfer increasing the size of the circle is more advantageous than what it takes away in lateral ability - a smaller slice of a bigger pie. At some point that trend will reverse. Obviously at the point the tyre locks there is zero lateral capability. Off the brakes it's all lateral, but a smaller circle.
Here I disagree, particularly as it seems to contradict your earlier point that " the maximum lateral acceleration of the car will not be improved by weight transfer".If weight transfer were at all advantageous then there would be no reason to keep the c of g low. While it may add an advantage in certain circumstances (which I doubt) it is not of an advantage during a complete lap.
upsidedownmark said:
Note also that this happens in reverse at the rear - braking does not increase the overall lateral capablility of the car - in fact it reduces it - due to the non-linearity you've picked up on, the combination of a highly loaded tyre and a lightly loaded tyre is less than 2 evenly loaded tyres. So we just change the balance. All the yaw in the world won't help you if you're beyond the lateral limits, so you don't see a 'spike' in lateral capability, it's more about being able to use what you have; Equally if you had a car that oversteers badly, trail braking would be no help at all.
Here we agree, as this is essentially what I was saying in my first post. Braking reduces the lateral capability of the car. It causes understeer by loading the front tyres. However the effect of the reduction of speed means that even though the car has reduced 'capability' it no longer needs so much capability as the braking has slowed it down. My point is that a driver senses the weight transfer and the improved turn in, but in pretty well every description of trail braking on the web the effects of braking and slowing are not given as the rightful cause of improved turn in. It is attributed to weight transfer, which is an effect rather than a cause.
When I look at things from the point of view of basic physics and the first law of thermodynamics, that of energy conservation, what happens is that velocity component of the kinetic energy (ke=1/2mv^2) that is trying to make the car go straight on (velocity is a vector) is converted into heat energy in the brakes.
Newton's second law of motion F=ma is thus applied. As the velocity reduces the vector force (steering) required to accelerate the car laterally reduces. That is what allows the car to turn more sharply. If the original steering is not reduced, the lateral acceleration will increase and it is sensed as oversteer.
It is changing the velocity that has a relatively huge effect as it is squared, which is why I pointed this out in my first post.
When I look at Newton's third law of motion. I see that any increased braking effected by the front tyres (due to weight transfer) is opposed by an equal and opposite force. i.e. The greater the load on the front tyres, the greater the force they will have to resist, which means that there is no net benefit. In fact....
...thanks to the first law of thermodynamics again, some of the equal and opposite force results in frictional forces within the tyre and suspension. So the braking force that is absorbed by them is turned to heat. Which is presumably why Milliken saw a reduction in the coefficient of friction with increasing load. (See my first post).
It also explains why vehicles with low centres of gravity brake in a shorter distance. As loads are distributed more equally between the tyres and therefore there are fewer losses in efficiency due to internal tyre (and suspension) friction.
I would also like to say that I like to apply Occam's Razor to physics questions. To my mind this is the simplest explanation that fits, and so is the one I am tempted to believe is right. Until someone can show me where I've made a mistake, or provide an even more simple explanation of course.
RobM77 said:
CABC said:
NDNDNDND said:
Mike showed me how you can 'thump the throttle' as Mark describes before the apex, loading up the rear tyres enough that you could slingshot through the corner at full throttle. Felt amazing and was much faster!
i've been playing around with this myself. I guess you can 'thump' the throttle in a 5 as it's quite low powered, whereas something with 400hp would need a 'firm press', over a fraction of a second. Maybe Rob can help here.I've been fortunate enough to be messing around with a Morgan too, and i have to say it's been great fun. You can feel rotation and the back bedding down far more clearly than in an Elise for example. Not saying that the Morgan is in anyway advanced, but on a smooth road it really is informative and a good training tool.
Often these sorts of problems are dealt with by setup, but quite often they can lend themselves to driving techniques as well, on top of those setup changes. For example, the trick with setting up front drive cars is sometimes to reduce rear grip below what is optimal, because this allows the car to be driven in a way which results in it being faster overall. This also manifests in a slightly different driving technique for FWD - I used to aim to get a very slight amount of oversteer before the apex, because even though this was slightly slower through that phase of the bend, I could get on the power harder and sooner, which overall in terms of lap time was a benefit.
Driving techniques that produce high yaw rates from deliberately increased tyre slip do not increase the max speed in the particular corner, but can do on the exit of that corner, and hence, if followed by a straight, can result in a high overall average speed.
Case in point here are the 4wd inlet restricted wrc cars on tight turns ie:
2006_Monte_hairpin
On that turn, those drivers are using the handbrake to literally turn the car 180 deg, but in all cases, the car pretty much comes to a complete halt. ie, the apex speed of the actual corner is very low, far lower than if they had taken the pure racing line around it.
But, a wrc car makes peak power at very low engine speed, and with 4wd, accelerates very hard, so because the car has been turned so early, they make the next straight relatively longer, and hence the average speed down the straight is higher. In effect they compromise the turn to the benefit of the straight!
Of course, WRC is a sport against the clock, rather than other cars. Do that in a race with other cars and they'd just overtake you in the turn. This is one of the reasons it's unusual to see cars meet their ultimate qualifying times in races (among other reasons of course)
A rather timely article as I recently finished reading Damon Hill's Watching The Wheels and looked up Damon's exploits on bikes and happen to come across a video of Damon's foray into FF https://www.youtube.com/watch?v=lL5OpfndQms
What a fantastic sight of watching a train of FF's drifting around Paddock Hill Bend.
Thank's Mark. You're one of my favourite writers in the car/motorsport publications and always a pleasure whenever I've come across your freelanced articles. Nice to have you onboard at PH.
What a fantastic sight of watching a train of FF's drifting around Paddock Hill Bend.
Thank's Mark. You're one of my favourite writers in the car/motorsport publications and always a pleasure whenever I've come across your freelanced articles. Nice to have you onboard at PH.
Gemaeden said:
NJH said:
How on earth does a tyre know the difference between download caused by aerodynamic force and download caused by accelerating mass aka weight transfer?
Note this is not a philosophical question.
A singular tyre 'knows' the difference as the download associated with weight transfer is at the track / tyre interface and is caused by acceleration or braking. This is not the case for aerodynamic download, which is through the hub (or wheel / tyre interface to be more accurate)Note this is not a philosophical question.
How do you think the weight is transmitted to the tyre under braking/acceleration? Magic braking only force pixies?
The loads are transferred by the same mechanical systems that connect the tyres to the car.
Well one of our contributors to this thread quotes Occams Razor but seems to revel in doing a great job of applying the exact opposite.
The clue is in the original article, and Mark's follow up. Its inertia and Newtons second law, everything we have been talking about comes from that, utilising weight transfer is nothing more than increasing the F by utilising some more A in the F=MA.
Of course if one doesn't understand what mass actually is then of course the simplicity of the basic physics can seem very elusive (another clue, I have said what it is above).
The clue is in the original article, and Mark's follow up. Its inertia and Newtons second law, everything we have been talking about comes from that, utilising weight transfer is nothing more than increasing the F by utilising some more A in the F=MA.
Of course if one doesn't understand what mass actually is then of course the simplicity of the basic physics can seem very elusive (another clue, I have said what it is above).
Gassing Station | General Gassing | Top of Page | What's New | My Stuff