Motorcycle/Rider simulation. Problems controlling oscillations.

[lagoausente]

Usually speed changes don't affect lean angle, even if a wheelie is done while leaned over, the bike retains it's lean angle (if it doesn't slide, and rider doesn't compensate prior to the wheelie). Since braking puts a force on the side of the front tire, some motorcycles get an inwards steering torque during braking, and this can reduce lean angle if the rider doesn't compensate with opposing torque inputs, while other bikes (or perhaps the tire profile) seem to not experience this inwards torque effect due to braking.

I think you are wrong;
Lean depends mainly on centrifugal force. Gyroscopic efect from wheels helps the bike to maintain stability but the lean changes come from centrifugal force.
The tyres are the supporting point, the drawn they describe are determined by the handlebar. The upper zone of the bike and the rider suffer the centrifugal force, and mainly this one determines the leaning.
At the same time centrigual force is depends on one parameter, the angular speed (or radian frequency as you like).
Note that at a constant radian frequency, the more radius that a circle has, more linear speed is needed, and the less radius less linear speed.
Example: Sweden and Spain have the same angular speed, they complete 360º on 24 hours, but Spain has a higher linear speed then Sweden. As an extreme, a point on the Earth at 1km from the north pole, also turns 360º in 24 hours but near zero linear speed.

That explains that on a easy curve a rider can get the max lean at a much higher speed that on sharp bends, as more sharp, less speed en vice versa.
what I want to say with all of this?
We have two parameters that determines centrifugal force and so the lean:
1) linear speed
2)Radius
At a constant radius if you increase linear speed will increase centrifugal force and so the bike will tend to rise and vice versa.
At constant linear speed, decreasing the radius will increase centrifugal force, increasing the radius will decrease centrifugal force and so the lean. This last one is the "countersteering" effect, turning to left lean to right.
There are a lot other parameters, like the brakes, the countersteering effect of the own front wheel (only 10% of total) , the rider weight changes etc, but that should must be considered as extras, not the mainly behaiviour of the bike.

 

[rcgldr]

Usually speed changes don't affect lean angle ... (other than braking on some motorcycles which causes the front tire to turn inwards).

I think you are wrong; Lean depends mainly on centrifugal force. ... At a constant radius.

Lean angle changes require a torque along the roll axis, which is due to a combination of gravitational and reaction force to centripetal acceleration at the center of mass and the forces (vertical and horizontal (centripetal)) at the contact patch of the tires.

On a motorcycle that tends to hold a lean angle when the rider relaxes on the handlebars, speed changes result in radius changes but the lean angle is essentially held since that is the tendency of such motorcycles.

 

[lagoausente]

Lean angle changes require a torque along the roll axis, which is due to a combination of gravitational and reaction force to centripetal acceleration at the center of mass and the forces (vertical and horizontal (centripetal)) at the contact patch of the tires.

Agree

On a motorcycle that tends to hold a lean angle when the rider relaxes on the handlebars, speed changes result in radius changes but the lean angle is essentially held since that is the tendency of such motorcycles.

Not agree, linear speed changes affect the angular speed but not the circle that the bike is drwaing. (not before lean change)
On bikes you describe,when they are keeping the lean, the handlebar and front wheel gets an angle in respect to the rear by itself. So even the rider relax the torque the bike has there own angle between both wheels. On a curve to right, the handlebar will be a little to the right. The smaller circle (sharper curves) more to right will be the handlebar and vice versa.
On a bigger circle to have the same lean the bike must have more linear speed, on that bigger circle the handlebar will be a little more to center (by itself bike behaviour).
Now suppose we have the bike running at 100kmh at a circle, if you increase the linear speed, centripetal force will increase. If your theory is ok, for the bike keeping the lean must increase the circle radius, and so the handlebar (on a curve to right) will have to go a little to center. But the front wheel gyroscopic effect will avoid it.
The own gyroscopic wheel effect (10% of total of the byke coutersteering effect) makes that if you torque the handlebar to left, then the wheel itself will tend to lean to right (see the pdf linked). So on the situation described, is supposed that on the situation I described before, when linear speed increase, to keep the lean the front wheel must go a little to center (left) so gyroscopic wheel effect will make the wheel lean to right. Since it´s only 10% of total lean force of the bike, the wheel will find some intertia to change it´s actual position. Result must be that the increasing the speed the bike tends to rise up (if no torque is applied by the rider because of the own wheels intertia that offers stability).
Racing experience tells that, look at the pdf:
http://perso.wanadoo.es/jcgtortosa/pilotaje/manual%20de%20pilotaje.pdf
On page 42:

En la FASE IV del giro, momento de aceleración progresiva, con la distancia entre ejes totalmente recuperada y la rueda trasera empujando desde atrás, la tendencia de la moto es a levantarse de la tumbada de manera progresiva. Aún así, si el piloto necesita poner la moto vertical con más rapidez, deberá aplicar contramanillar en el sentido inverso al que lo hizo para entrar a la curva y con la progresividad que requiera la situación.

google translate:

In the Phase IV of the turn, when progressive acceleration, with the wheelbase now fully recovered and the rear wheel by pushing from behind, the tendency of the bike is to get out of lying in a progressive manner. Still, if the pilot needs to get the bike vertical faster, contramanillar be applied in reverse than it did to enter the curve and the escalation that the situation requires.

The pdf is based on years of experience on Racing competition and from page 40 it describes each detail of the bike behaviour regaring countersteering and all the phases before and after the lean. Google translator que give a help.

PD: google translator is not very good: contramanillar means countersteering.

 

[rcgldr]

Linear speed changes affect the angular speed but not the circle that the bike is drawing.

Depends on the motorcycle, some tend to hold lean angle even under acceleration, even if the bike does a wheelie, depending on rider position.

Now suppose we have the bike running at 100kmh at a circle, if you increase the linear speed, centripetal force will increase.

Only if there isn't some inherent tendency for the steering to straighten out the small amount required to hold a lean angle, in response to acceleration, which is a potential response depending on the motorcycle. When braking the force on the side of the contact patch can generate an inwards torque along the steering axis, causing the bike to straighten up; this effect varies between motorcycles and tire profiles.

For a racing motorcycle, the rider is hanging off, and the offset center of mass results in a tendency to lean further inwards, so the rider is never truly relaxing on the handlebars, just reducing the amount of steering input to a very small amount of torque on the handlebars when holding a lean angle.

The own gyroscopic wheel effect (10% of total of the byke coutersteering effect) makes that if you torque the handlebar to left, then the wheel itself will tend to lean to right.

... and the rear tire, will generate a yaw response to the right in response to a lean to the right, creating a small torque to the right at the contact patch of the front tire, somewhat opposing the left torque applied by the rider, but this is a small effect.

That 10% effect depends on the speed. As mentioned, once above some relatively high speed, the gyroscopic effect would tend to cause under-correction, called capsize speed in mathematical models. However, tires have a round profile, and the contact patch being on the side of the tires when leaning counters the capsize factor, and the end result for most bicycles and motorcyles at high speeds is a tendency to remain stable or at least change lean angle at a very slow rate. Camber thrust and torque forces at the contact patches further complicate matters.

Link to wiki article that includes info about capsize speed, but the not the opposing factor from contact patch with a round tire profile:

http://en.wikipedia.org/wiki/Talk:Bicycle_and_motorcycle_dynamics#Eigenvalues