Hmm... The paper didn't actually use the phrase "molecular adhesion," but
rather "adhesion," so perhaps Gillespie is using a different definition than
I'm thinking of.
Here's an exerpt from the paper:
(This paper was concerning development of a numerical tire model specifically
dealing with the relationships between tire heating, sliding velocity, and
grip. The key here is that the focus of this is in the *sliding* portion of
the print.)
"The development of the contact model is based on studies of mechanical and
thermal effects in the contact interface. This investigation is conducted on
the microscale, on which two different friction mechanisms can be
distinguished. Hysteretic friction results from the internal damping of ***
in case of periodic excitation. Therefore, the loss modulus is the decisive
characteristic of *** for this friction component. As the energy
dissipation takes place in layers of the *** block adjacent to the contact
surface, hysteretic friction is a bulk effect."
"The second part of *** friction is adhesive friction. It is a consequence
of inter*** bonds that are made up between the *** and the friction
surface on a molecular scale. In case of a relative motion of the two friction
partners, these bonds are stretched and finally ruptured. The stored bond
enery is dissipated. Adhesive friction is the minor part of *** friction.
Hence, the characteristics of hysteretic friction are the basis for the
macroscopic contact model that will be briefly described in the following."
Later on the paper says it leaves out the adhesive friction analysis entirely
due to its relative insignificance.
It should be noted though that this is referring primarily to what is happening
in the portion of the print that is sliding, which in the case of a tire
operating at the limit, might very well be the majority of the contact patch or
a significant portion of it at least.
I haven't read Gillespie's books, so can't comment on the relationship between
this and what he described. Perhaps he was referring to the non-sliding region
of the contact patch, where perhaps the situation is reversed. I haven't
thought or read about it enough yet to comment.
As I mentioned before, I work a bit with an engineer and a couple champion
1/8th scale gas RC drivers that have discussed the same effect with me. These
cars behave opposite to what you just described. They have a solid rear axle
(and front one-way bearings for selective 4WD) and also benefit from lifting
the rear wheel when increased oversteer is desired. Note though that "lifting
the inside rear wheel" probably means "lowering the load on the inside rear
wheel."
They typically adjust this on the RC cars by altering caster and are knowingly
altering the inside rear load, or "lifting" it in varying amounts. Increasing
caster causes more inside rear lift, and the car responds with increased
oversteer on throttle (and subsequent understeer off throttle due to higher
aligning torque fighting the servos and possibly bending the suspension a bit.)
This sounds to be the exact opposite of what you've described with go-karts,
which itself is the opposite of what I've heard other people say. I'm a bit
confused :-) From other's descriptions, it seemed that leaning outside the
corner or altering the setup in a way that increased inside rear tire lift
caused an increase in oversteer on throttle, the same effect that we see in RC
cars.
I'll take your word for it though :-)
Locked axles are a little tricky to make generalizations about because there a
couple of things happening like you described. First of all, the solid axle is
trying to straighten the car, causing understeer of course. The car, generally
speaking, should probably have increased understeer with less load transfer
from this effect as you described.
On the other hand, this forward/rearward force that's being developed (causing
understeer) uses up some of the friction circle, leaving less side force. So
whether reducing lateral load transfer (less wheel lift) at the rear causes
understeer or oversteer depends on which one of these two effects dominates.
I'd think that this would depend to a large degree on the rear track width /
center of gravity height ratio. It could theoretically go either way. With
the extremely low CG RC cars, inside rear lift is generally (always?) an
oversteer effect on-throttle. I'd imagine the CG is in proportion much higher
on a go-kart, so perhaps that would explain why the opposite might be
happening.
I'm no expert, mind you, but that's my take on it right now :-P
Todd Wasson
Racing Software
http://www.racesimcentral.net/
My car sim
http://www.racesimcentral.net/
