Playing with my widget i can view front and rear wheel speed and the difference whilst riding.
With the default 125 i can observe that front braking with any useful but controlled degree of force (not locking the wheel) and the bike upright, straight and stable causes a difference of up to 3m/s difference in wheel speed with the front rotating slower than the rear. Not tried it with any of the more powerful bikes yet.
This seems to suggest either the front is always sliding to some degree or the rear is spinning up. Pretty sure i am not lifting the rear tyre and the throttle is shut, all braking assists are off.
Not sure it means anything, i am just wondering what causes this to happen, any ideas?
From my knowelage rear wheel just slides. I read about it in some motorcycle book. At high speeds for example rear wheel spins faster to the degree that speedometer from front wheel can show up to 20 Km/h lower speed than from the rear wheel.
Thanks, i was wondering how it could occur without the bike tying itself in a knot. Strangely the default 990 does not seem to exhibit it to the same degree and seems to be about 1m/s difference with similar braking force.
Just a thought, but are the wheels the same size? 17inch rear to 16 inch front maybe? or even rim width would have an effect I'd have thought? In that scenario the front wheel would surely spin faster? :-\
Just a thought. ;)
Hawk.
Normal riding the wheel speeds are reported as identical - i guess they must be 'normalised' somehow to allow for the variations in contact diameter. Difference only shows when braking.
Essentially, any tyre must slide to generate a force. As G.Clooney used to say: no sliding, no party (not 100% sure of the quote).
So when you open the gas, the rear will slide. Even when you don't feel it, even on a Fiat Panda.
Same when you brake: the tyre will slide, in order to generate the necessary force (to slow down the entire bike).
If a tyre is not sliding, then it's neither accelerating (i.e. generating a force that pushes the bike forward) nor braking.
Same goes when cornering: the tyre slides (laterally) to generate a cornering force.
Side note: one does not need to lock a tyre to be sliding, but most people do not realise that.
MaX.
P.S.
For the extremely picky ones around here (just me ?): the above is not absolutely true, tyres can generate little amount of forces even without sliding (e.g. rolling resistance, camber thrust, ...).
But for significant amounts of force (be it braking or accelerating), the tyre will slide.
Thanks, makes sense now you have explained it. :)
Do you have a source that explains it? Don't think I understood.
Quote from: RIDER on September 01, 2015, 04:57:37 PM
Do you have a source that explains it? Don't think I understood.
Just to grasp the idea, look at the black line on that graph:
(http://www.racer.nl/images/pacejka.jpg)
It gives tha ratio Fx/Fz as function of the slip ratio.
Fx is the longitudinal force the tyre can generate (to accelerate or brake). Fz is the normal load (how much "weight" is on the tyre).
The slip ratio is equal to (wheel rotation speed times the wheel radius - V) / V, where V is the speed at which the bike (actually, the contact patch) is moving with respect to the world.
- Slip ratio < 0 ==> wheel is rotating slower than the bike velocity, usually when breaking (-1 = wheel is locked)
- Slip ratio = 0 ==> no slipping (i.e. pure rolling motion)
- Slip ratio > 0 ==> some slipping, usually when accelerating
As you can see in the graph above, when the slip ratio is zero, the black line shows zero for Fx/Fz: in order to have the tyre generating a force, the slip ratio must be bigger (or smaller) than zero.
MaX.
The idea seems simple but impossible, that's why I feel like im missing something. Would the same idea affect tracked vehicles?
Quote from: RIDER on September 01, 2015, 10:49:11 PM
The idea seems simple but impossible, that's why I feel like im missing something.
Impossible ?! What do you mean ?
It's not an idea, it's how a tyre behaves. I mean: you put a tyre on a complex machine capable of measuring these properties and the graph above is what comes out.
Quote from: RIDER on September 01, 2015, 10:49:11 PM
Would the same idea affect tracked vehicles?
Yes, it's a property of the tyres in general. It applies to cars, bicycles and all.
I think it even applies to train wheels (no rubber, metal-metal contact).
MaX.
Maybe a simpler concept is that you must have friction to give the driving (or retarding) mechanism between the tyre and surface, to get friction you need two sliding surfaces in contact, therefore the tyre must be sliding to some degree.
Maybe too simple?
A bit too simple (as there's also static friction for example), but yes that's more or less it.
MaX.
You even have a little bit of slip when a tire is just rolling. I won't explain this but when you think of the tire (maybe in 2D) on the ground it's a circle with a flat spot. Since the hub of the wheel has a constant RPM and the velocity at the tire is proportional to the radius and this rpm the velocity at the surface of the tire must change at the flat spot. So the tire speed in the middle of this spot is less then the surface speed of the tire not on the ground. So the speed of the bike will be somewhere between this speed of the free spinning wheel and the speed the tire has at the point of coontact to the street.
So one parameter for this effect is how much tire pressure you're using. An ideal stiff tire wouldn't show this effect at all.
Quote from: SKD on September 02, 2015, 11:42:53 AM
You even have a little bit of slip when a tire is just rolling.
Right, but that depends on the exact definition of "just rolling". As soon as you take into account flexible tyres, all becomes a bit messier and one has to be very precise.
In general: pure rolling = no slipping (i.e. contact patch is not slipping with respect to the ground/track). But if there's tyre deformation, this does not mean that "tyre radius * omega = V" (as you explained, due to the deformation).
MaX.
That's the most interesting thing ive learned in months. I wonder about how much were talking. Enough that it adds to the difficulty of anticipating/feeling a tire breaking traction in low grip situations? Or can it really never be felt?
Quote from: RIDER on September 02, 2015, 05:49:21 PM
That's the most interesting thing ive learned in months. I wonder about how much were talking. Enough that it adds to the difficulty of anticipating/feeling a tire breaking traction in low grip situations? Or can it really never be felt?
I've never been on a track so I can hardly tell for sure. I'm fairly confident it's much easier to feel a small lateral slip (side slip) than a small longitudinal one.
I'm under the impression that at the front you barely feel the longitudinal slip under braking, and when you feel the lateral one is probably already too late :)
At the rear it's easy to feel large longitudinal slips (both braking and accelerating) and even small lateral slip (rear going side-way).
To give an idea, the murasama rear quali tyre has its longitudianl force (no side slip, no camber, nominal vertical load) that reaches his max at a slip of 0.15 (i.e. 15% difference between "bike" speed and wheel speed). As soon as you have even a modest slip, the force can be fairly big. I guess that's why I think we do't feel the slip (except in extreme cases).
MaX.
Sometimes you can hear the rear tyre squeak, accelerating hard off the grid ;) You can always feel the rear sliding and squirming around on the power, its really not a big deal when you are on the bike.. big controlled slides are a different story though lol :o
on the brakes you can guarantee the rear will be moving around too! ;D or chattering because i downshifted like a tw@t ::)
Bad question, I was asking if you can feel slip levels that are way too low to feel (obviously not). And sense the slip increases gradually up to a sudden break in traction, it's easier to anticipate rather than more difficult.
You do feel the front sliding laterally easy but usually when trail braking so you aren't on the edge of the tire.
So whenever a bike is leaned over and moving forward, the tires are slipping laterally due to the small cornering load?
Quote from: RIDER on September 02, 2015, 09:27:47 PM
So whenever a bike is leaned over and moving forward, the tires are slipping laterally due to the small cornering load?
Yes, even if in principle you could turn without slipping (laterally) thanks to the camber thrust (a lateral force generated by a tyre with a camber angle, Camber thrust (https://en.wikipedia.org/wiki/Camber_thrust)).
Camber thrust is nice because it happens immediately as you camber, while lateral forces generated by slipping have some sort of delay (relaxation length, GPB takes it into account).
But when you take a corner hard enough, camber thrust is not enough so you'll have to slip.
MaX.
Wow, its amazing the level of detail GPB has within the physics and everything... Really is something special 8)
Sort of related..
Had the default 125 on Victoria and noted that at turn 2 at a constant lean, turn rate and speed i lose the front when the bike hits some small bumps toward the exit of the corner even though it was fine earlier in the turn. Obviously taking the corner slower works OK but it does seem that you should be able to go a lot faster. :)
The bike appears to have very little compliance for bumps when fully leaned over. In this situation is the tyre lateral stiffness the only form of 'suspension' (is it modelled) or are the frame/swingarm/forks stiffness modeled to allow the frame system to flex a bit and soak up the smaller bumps.
The magnitude of the bumps do not appear as is they should be sufficient to cause the wheel to skip (and cause a fall) but they do with the 125.
Quote from: h106frp on September 03, 2015, 08:47:09 AM
The bike appears to have very little compliance for bumps when fully leaned over. In this situation is the tyre lateral stiffness the only form of 'suspension' (is it modelled) or are the frame/swingarm/forks stiffness modeled to allow the frame system to flex a bit and soak up the smaller bumps.
I think tyre stiffness is modeled in GPB. Are the values correct ? Hard to tell but I think they have been changed in beta4 or 5 to be closer to reality.
And yes, chassis/fork/swing-arm flexibility should help a bit too, in principle.
MaX.
Not exactly what you are talking about, but changing some mass points can help a little in the 125cc: http://forum.piboso.com/index.php?topic=2263.msg32177#msg32177 (try the attached geom file)
I fully agree about the chassis/fork/swing arm change the things. Also, from what I try, the rear suspension mass is a point that helps with the prevention of loosing the front in some bumps when leaning constant.
Steer data seems, in principle, an important point on that too, but we doesn´t know what is kdamping and kYaw (unless I´m missing something).
In a bike testing, founded these values to help on bump/leaning issues.
Lock = 30
Damper = 9
DamperPower = 0.9
spg0 = -280
spg1 = -3
sdg0 = -70
sdg1 = 0.3
sig0 = 0
sig1 = 0
KYaw = 20
KDamping0 = 6
KDamping1 = 0.15
spg, sdg and sig are related to speed and lean angle? Do you have some info about how to calculate them?
Quote from: Alone on September 03, 2015, 11:43:44 AM
spg, sdg and sig are related to speed and lean angle? Do you have some info about how to calculate them?
They are virtual rider parameters: they are the gains of the PID (Proportional, Integral, Derivative) controller that stabilizes the bike dynamics in terms of lean angle.
spg0, spg1 are for the proportional action
sig0, sig1 are for the integral action
sdg0, sdg1 are for the derivative action
The overall proportional gain is something like "spg0 + spg1*V", where V is the bike speed: that makes the gains vary depending on the bike speed.
There's no easy way to compute them: one would need a linearized model of the bike at different speeds, some knowledge about feedback loops, PID controllers etc.
Side note: when I see how fast the steering moves in GPB without any input change (i.e. in the telemtry, how "noisy" the steering signal is), I always think that there may be something wrong in the virtual rider ... bad gains or bad PID implementation ... I don't think on a real bike the steering moves like that.
MaX.
Thanks Max!
Thanks, I will give the geometry file a try first.