Modding tools and documentation

[neoraptor]

Check the model I used (file is attached to this message) with this applet : http://www.geogebra.org/webstart/geogebra.html (simply open the file there).

My problem is to find how it is referenced as there is 3 different reference system for the fork : the chassis, upper fork and lower fork.

I don't manage to put everything in the right order for now :S

It also looks like the fork is mounted reverse on the scooter according to the screenshot from PiBoSo.


I am happy for any help as I already look into it for a few hours without finding the correct reference for the front. 

[C21]

It also looks like the fork is mounted reverse on the scooter according to the screenshot from PiBoSo.

But the fork is mounted reverse on the scooter isn´t it?

I am happy for any help as I already look into it for a few hours without finding the correct reference for the front.

Right now i don´t have a clue either.
maybe Max has...

[HornetMaX]

It also looks like the fork is mounted reverse on the scooter according to the screenshot from PiBoSo.

But the fork is mounted reverse on the scooter isn´t it?

Upside-down or normal fork does not make a difference for GPB: the fork will still have an unsprung mass and a sprung one (integrated in the steering head mass).

I am happy for any help as I already look into it for a few hours without finding the correct reference for the front.

Right now i don´t have a clue either.
maybe Max has...

No time yet to have a look at what neoraptor has done (looks interesting).
However I never played seriously with 3d modeling and, to be honest, I don't plan to do it neither: my life is already not long enough without tinkering with 3ds

Once I start the physics modding doc, I may want to understand/crossreference the stuff that "links" physics and 3d model.

MaX.

[neoraptor]

I was not speaking about upside-down or normal fork.
If we look at the screenshot from Piboso, the fork offset seems negative (go back toward the rear) compared to the chassisSteer point (which is where the contact between the chassis and the fork is done).

It is like the triple clamp is not mounted toward the front, but toward the rear wheel). 

[HornetMaX]

I was not speaking about upside-down or normal fork.
If we look at the screenshot from Piboso, the fork offset seems negative (go back toward the rear) compared to the chassisSteer point (which is where the contact between the chassis and the fork is done).

It is like the triple clamp is not mounted toward the front, but toward the rear wheel). 

I haven't verified, but if it is true, it is indeed an unusual setup, at least for a bike (maybe ok for a scooter, don't know).

RC bikes for example are typically like this: it gives a sort of cart-wheel self-stabilizing effect.

MaX.

[HornetMaX]

Some other random questions for Piboso:

• In the main bike .cfg file, wheel0/1 section, what is the parameter "Friction" (1.5 for the 990) ?
• In the .tyre file, what are R0 and Fz0 ?
• In the .tyre file, do you have a reference article describing the Pacejka parameters that you use ? (guess they are the ones before section compound0)
• Within compoud0 section: what is affected by Below/AboveTempFactor ?
• Within compoud0 section: what are NumKeys, Key0 and Key1 ?
• Within compoud0 section: what is curve_type ? Releate to the Magic Formula I guess ...
• Within compoud0 section: what are resistance_dry and resistance_wet ?

MaX.

[HornetMaX]

neoraptor, I think there's one bug in your drawing above: you have a line between fwheel' and chassisSteer and it is likely in your mind this is the steering axis. I think that the steering axis should be between fwheel' and steerJoint'.

My interpretation of the points is the following:

• chassis_steer: origin point for the reference system of the steering head. This reference system (let's call it Rh, h for head) has its origin in steer_chassis and it is rotated by an angle equal to the rake angle (so that the X axis will be parallel to the steering axis). The coords of the other steering points (steerJoint, frontUpper, frontLower) are relative to Rh.
• steer_joint: point that defines the steering axis. As it is in Rh, the equation of the steering line in Rh will be y = const, where const is the Y coord of steerJoint in Rh (i.e. it's Y in the geom file).
• front_upper, front_lower: these describe the front fork movement (fully compressed, fully extended). They lay on the steering axis (now ).

MaX.

[PiBoSo]

"chassis_steer" is the position of the steer link in chassis coordinates. It is rotated by the rake angle.
"steer_joint" is the position of the same link relative to steer coordinates.
Same for "rsusp_joint", that is the position of the pivot in swingarm coordinates.

[HornetMaX]

"chassis_steer" is the position of the steer link in chassis coordinates. It is rotated by the rake angle.
"steer_joint" is the position of the same link relative to steer coordinates.
Same for "rsusp_joint", that is the position of the pivot in swingarm coordinates.

Thanks !

But how are steer coordinates (and swingarm coordinates) defined then ?

I know I can backward deduce them (knowing that steer_joint and chassis_steer are the same point in the two diff coord systems), but ...

MaX.

[PiBoSo]

"chassis_steer" is the position of the steer link in chassis coordinates. It is rotated by the rake angle.
"steer_joint" is the position of the same link relative to steer coordinates.
Same for "rsusp_joint", that is the position of the pivot in swingarm coordinates.

Thanks !

But how are steer coordinates (and swingarm coordinates) defined then ?

I know I can backward deduce them (knowing that steer_joint and chassis_steer are the same point in the two diff coord systems), but ...

MaX.

steer_world_matrix = steer_local_matrix * chassis_joint_world_matrix
steer_local_matrix = no rotation - steer_joint
chassis_joint_world_matrix = chassis_joint_local_matrix * chassis_world_matrix
chassis_joint_local_matrix = rake angle + chassis_steer

[HornetMaX]

"chassis_steer" is the position of the steer link in chassis coordinates. It is rotated by the rake angle.
"steer_joint" is the position of the same link relative to steer coordinates.
Same for "rsusp_joint", that is the position of the pivot in swingarm coordinates.

Thanks !

But how are steer coordinates (and swingarm coordinates) defined then ?

I know I can backward deduce them (knowing that steer_joint and chassis_steer are the same point in the two diff coord systems), but ...

MaX.

steer_world_matrix = steer_local_matrix * chassis_joint_world_matrix
steer_local_matrix = no rotation - steer_joint
chassis_joint_world_matrix = chassis_joint_local_matrix * chassis_world_matrix
chassis_joint_local_matrix = rake angle + chassis_steer

Now I'm even more confused, doh !

chassis_joint_local_matrix = rake angle + chassis_steer

So this means the X axis goes through the chassis_steer point and has an angle of let's say 24deg (Rake) wrt to vertical. This X axis is the steering axis.
Y axis (the other axis in our 2d view) is in the bike's longitudinal plane and is orthogonal to X and goes through chassis_steer point too too.
Correct ?

chassis_joint_world_matrix = chassis_joint_local_matrix * chassis_world_matrix

OK, this accounts for chassis orientation in space (roll, pitch, yaw). For our purposes, we can ignore it I guess.

steer_local_matrix = no rotation - steer_joint

Not sure I get it. Does this mean that the origin of this reference system is such that steer_joint actually corresponds to chassis_steer ?
And are X,Y,Z axis of this reference are oriented as:

• X orthogonal to the two fork tubes (rider's left when on the bike)
• Y parallel to fork tubes, pointing "up"
• Z orthogonal to X and Y (hence orthogonal to the plane of the two fork tubes, pointing "forward")

Is this what you mean ?

I'm no longer sure what front_upper, front_lower are, as they don't seem to be parallel to the steering axis ...

MaX.

[HornetMaX]

@neoraptor:

plenty of thing I don't understand yet (front_lower/upper, front_length, rear_upper/lower, rear_length).

In the .ggb file, the sliders control the rake angle, the swingarm pivot and the swingarm length correctly (pivot is useless tough, as min/max points are identical).



MaX.

EDIT [MaX]: Ignore, that model is very wrong !

[HornetMaX]

I think I've got it almost right now, I only miss this: are front_upper and front_lower the same point when front_length = 0 ? So front_lower (and fwheel) travel along the direction defined by the segment front_uppper .. front_lower ?

MaX.

[PiBoSo]

I think I've got it almost right now, I only miss this: are front_upper and front_lower the same point when front_length = 0 ? So front_lower (and fwheel) travel along the direction defined by the segment front_uppper .. front_lower ?

MaX.

"front_lower" is the position of the bottom of the front suspensions in unsprung forks coordinates.

[HornetMaX]

I think I've got it almost right now, I only miss this: are front_upper and front_lower the same point when front_length = 0 ? So front_lower (and fwheel) travel along the direction defined by the segment front_uppper .. front_lower ?

MaX.

"front_lower" is the position of the bottom of the front suspensions in unsprung forks coordinates.

Arghh ... guess a proper drawing would solve that for good. Anyway ...

So steer_joint and front_upper are in steer coords while front_lower and fWheel are in unsprung fork coords ?

And how do you position the unsprung part ? front_bottom (in chassis coords) must be laying on a line passing by front_upper (in chassis coords) and parallel to the steering axis ?

MaX