Hi all,
Back to some physics programming, I'm trying to get Richard Chaney's
idea (see http://www.chaney.karoo.net/) of using a quaternion for
rotation, and angular velocities (adjusted using precession) and a
location/velocity for the rigid bodies to work.
It contains some easy & fast to use formulas for forces placed on a
point on the body (for example, aerodynamic drag). This leads to a
force on the center of mass, and a torque.
Now I wonder, before I haul over my source; can I just add resulting
forces and torques to get 1 end-result, and apply that to the body?
I mean like:
init:
totalForce.SetToZero()
totalTorque.SetToZero() (both are 3D vectors)
calc_forces:
wing_front_force=(x,y,z)
totalForce+=wing_front_force (in body coords)
// Calc torque because of wing_front_force using T=F*r
// where 'r' is a vector from the CM to the point
// of force application
totalTorque+=wing_torque
wing_back_force=...
totalForce+=wing_back_force
totalTorque+=wing_back_torque
integrate:
// Apply force and torque to body
cm_acceleration=totalForce/mass
cm_angular_acceleration=... (not really totalTorque/inertia, but
using some more careful formulas because of precession)
I read at one point, IIRC, that you can see a total force system of a
rigid body as having a resultant force and torque, so you can combine
all forces and torques to come up with just 1 force and 1 torque.
Would this be the way?
BTW The Giancarlo Genta book (Motor Vehicle Dynamics), is it any good
on explaining the Pacejka Magic Formula? The formula seems great, but
I can't find too much raw info on it (to use in a simulation), and the
book, priced at US$98, is quite pricy if all I can find on the formula
is in an appendix, perhaps (seeing I also have RCVD and Gillespie's
Fundamentals of Vehicle Dynamics).
Thanks for any ideas,
Ruud van Gaal, GPL Rank +53.25
Pencil art : http://www.marketgraph.nl/gallery/
Car simulation: http://www.marketgraph.nl/gallery/racer/
What you say is true. The acceleration of the center of gravity is given
by the sum of all forces acting on the rigid body, while the change of
its angular momentum (the product of the tensor of inertia and the
vector of angular velocity) is given by the sum of all the torques
acting on the body. These torques include the ones produced by the
forces acting on the body (rxF), plus some additional torques (e.g. the
torque of the engine as produced on the wheel), which cannot be
represented by a force and a point of its application. Be sure to
calculate everything with respect to the center of gravity, though.
-Gregor
Ruud van Gaal wrote:
>
> I read at one point, IIRC, that you can see a total force system of a
> rigid body as having a resultant force and torque, so you can combine
> all forces and torques to come up with just 1 force and 1 torque.
> Would this be the way?
>
On Wed, 31 Jan 2001 17:03:04 +0100, Gregor Veble
<[email hidden]> wrote:
>What you say is true. The acceleration of the center of gravity is given
>by the sum of all forces acting on the rigid body, while the change of
>its angular momentum (the product of the tensor of inertia and the
>vector of angular velocity) is given by the sum of all the torques
>acting on the body.
Thanks, I can go ahead and try to convert the lot then, until it runs
nearly like it's running now.
> These torques include the ones produced by the
>forces acting on the body (rxF), plus some additional torques (e.g. the
>torque of the engine as produced on the wheel), which cannot be
>represented by a force and a point of its application.
How would this be represented then? (I currently just apply the engine
torque to the wheel, not applying that torque to the sprung mass
(body) itself in any way)
Thanks again,
Ruud van Gaal, GPL Rank +53.25
Pencil art : http://www.marketgraph.nl/gallery/
Car simulation: http://www.marketgraph.nl/gallery/racer/
Quoted text (8 lines)
>> These torques include the ones produced by the >>forces acting on the body (rxF), plus some additional torques (e.g. the >>torque of the engine as produced on the wheel), which cannot be >>represented by a force and a point of its application. > >How would this be represented then? (I currently just apply the engine >torque to the wheel, not applying that torque to the sprung mass >(body) itself in any way)"J. Todd Wasson" wrote:
Just like the Newton's law says that if one body exerts a force, the
other exertst the exactly opposite force back on the first one, this can
be extended to torques as well. Therefore, when you apply the torque on
the wheel, the wheel exerts the exact opposite torque back on the car
body. It's important to notice, that, unlike forces, the pure torques do
not have a specific origin associated with them, only their orientation
is important; this means that you can directly apply the torque as
produced back on the body by the wheels as if acting at the body's
center of gravity; no transformaion (apart from the rotation into the
body system, if that's how your calculations are done) is needed.
Todd extended that idea also to the engine torque. You can consider the
engine as another rotating mass that exerts torque on the body along the
appropriate axis, and that's why the wheelie tendency is very much
dependent on the engine orientation.
Also, without going into too much detail, I can tell you that you can
consider the body as a sort of 'sink & source' for torques; that is to
say, you don't need to worry too much about how the torque actually gets
to the wheels, if you input it along the appropriate axis into the body,
and then 'take it out' on the wheels (as per the above Newton's law
example), it's the exact way of doing it as far as the effects on the
body, wheels and engine go, even when you take the losses into account.
You will, however, still need to somehow be able to couple the engine
and wheels rotational velocities and inertias, as they are rigidly
coupled (with an additional degree of freedom in the differential). But
as far as torques go, that's the simplest and proper way to do it. I
would go into doing it all only after what you currently set out to do
works, as debugging will be much more messy if everything is taken into
account from the start.
Quoted text (17 lines)
> >How would this be represented then? (I currently just apply the engine > >torque to the wheel, not applying that torque to the sprung mass > >(body) itself in any way) > > I think you just add the engine torque along the appropriate axis. I saw a > motorcycle drag race simulator similar to the program at my site, that wanted > the engine's location and direction of rotation. Funny how the engine torque > could either increase or decrease the tendency of a bike to wheelie! > > I'll be back to hear Gregor, in case I'm missing (another) something :0) > > Todd Wasson > --- > Performance Simulations > Drag Racing and Top Speed Prediction > Software > http://PerformanceSimulations.ComI saw a nifty show about Robbie Knievel's
motorcycle jump over a fountain in Las Vegas.
He explained how he could use the front or
rear brakes o slow either wheel and thereby
pitch the vehicle slightly while flying.
(Slowing the rotational inertia of the wheel
increases the rotation of the overall vehicle
slightly in reaction. The overall angular momentum
is conserved.)
Which is pretty cool but still pretty crazy ;)
- Matt
On Sat, 03 Feb 2001 04:30:34 GMT, "Matthew V. Jessick"
<[email hidden]> wrote:
Hm, cool, I must try this this afternoon. ;-)
Ruud van Gaal, GPL Rank +53.25
Pencil art : http://www.marketgraph.nl/gallery/
Car simulation: http://www.marketgraph.nl/gallery/racer/
Quoted text (11 lines)
>I saw a nifty show about Robbie Knievel's >motorcycle jump over a fountain in Las Vegas. >He explained how he could use the front or >rear brakes o slow either wheel and thereby >pitch the vehicle slightly while flying. >(Slowing the rotational inertia of the wheel >increases the rotation of the overall vehicle >slightly in reaction. The overall angular momentum >is conserved.) > >Which is pretty cool but still pretty crazy ;)
"Matthew V. Jessick" wrote:
It works in GPL too - when I go over Fugplatz I try
to rev up the engine in the air because with the increase
in the rear wheel's rotational momentum the body of the
car pitches up slightly, allowing me to land the car
on the rear wheels just before the front wheels, which
is makes for a much more stable landing than lifting off
throttle which pitches the nose straight down into the
ground.
Quoted text (13 lines)
> > I saw a nifty show about Robbie Knievel's > motorcycle jump over a fountain in Las Vegas. > He explained how he could use the front or > rear brakes o slow either wheel and thereby > pitch the vehicle slightly while flying. > (Slowing the rotational inertia of the wheel > increases the rotation of the overall vehicle > slightly in reaction. The overall angular momentum > is conserved.) > > Which is pretty cool but still pretty crazy ;) >"Grant Reeve" <[email hidden]> wrote:
John Surtees talks about the importance of landing on
the rear wheels in the book 'Six Days In August' although
he didn't accomplish that by revving up the motor in the
air (he preferred rather less air time than we seem to,
thinking that the power isn't being delivered to the
ground with the rear wheels in the air). He described
his technique as compressing the suspension and then
reaccelerating as you reach the rise, kind of like getting
an XR100 to wheelie.
As one would expect there appears to have been quite
a difference in thought about the which wheels to
land on as Denny Hulme gave the exact opposite
advice (the interested party was a young David
Hobbs) reasoning that regaining steering input
was more important than worrying about the rear
coming around.
The book itself is quite fascinating as the author
spent an entire month with 'Big John' around the
time of the 1967 German GP. Surtees mentions
that the lighter more powerful Honda 'Electron'
motor gave 394 HP at best and that all versions
in '67 only revved to 10,200 RPM, making maximum
power at 9,500 RPM and simply not pulling at all
below 8,750 RPM. Too bad the GPL engine
model seems to use a second order curve
to model the power band.
Also of interest was that he geared for 190 MPH
at Spa and 172 MPH at the 'Ring. They also give
the fuel loads for the German GP as:
BRM: 40 Gallons
Brabham: 32 Gallons
Cooper: 37 Gallons
Lotus: 35 Gallons
Honda: 48 Gallons
The Japanese it appears were quite the opposite
of Colin Chapman when it came to estimating fuel
consumption. ;-)
- George
Quoted text (8 lines)
> It works in GPL too - when I go over Fugplatz I try > to rev up the engine in the air because with the increase > in the rear wheel's rotational momentum the body of the > car pitches up slightly, allowing me to land the car > on the rear wheels just before the front wheels, which > is makes for a much more stable landing than lifting off > throttle which pitches the nose straight down into the > ground.On Sat, 03 Feb 2001 21:51:35 GMT, Grant Reeve
<[email hidden]> wrote:
>
>
>"Matthew V. Jessick" wrote:
>> He explained how he could use the front or
>> rear brakes o slow either wheel and thereby
>> pitch the vehicle slightly while flying....
Some people use a quick brake before going off into the air. That way
you compress the front suspension, then get off the brakes, and you
get a pitching torque which gets you landing on the rear wheels (if
you time it right! It's so easy to screw up yet-another-lap on the
'Ring :) ).
Ofcourse, using the brakes seems a waste of precious GPLRank time. :)
Ruud van Gaal, GPL Rank +53.25
Pencil art : http://www.marketgraph.nl/gallery/
Car simulation: http://www.marketgraph.nl/gallery/racer/
