I've struck a problem withg using keyboard control for steering in my
sim.
I need to use keyboard steering and I've basically got the left &
right keys increasing the steering angle to about 30 degrees. If the
user has no keys pressed then the steering travels back to 0 degrees.
If I use Beckmans magic formula (Phors Part 29...),
Flateral = weight * 31 * SlipAngle * (1 + POW(ABS(9.625 *
SlipAngle),2.375))
then the front-wheels slip angle is only momentarily at its peak, it
then passes through the max-force slip angle & tapers off, leaving the
car with extreme understeer.
If I however limit the max steering angle to +/- 8 degrees it steers
ok, but once the car starts to slide there isn't enough steering angle
to control the oversteer & it's useless.
It appears to me that one can't use keyboard control with slip
formulae that taper off after the peak, if I use a simple flat
limiting it works ok.
(Well sort of works, it still won't slide out with full steering lock
while accelerating from 0.)
This is my point of view:
- If you HAVE to implement keyboard control (i.e. if u are working for
a commercial product)... try to implement some sort of "magic steer"
that will turn the wheel just enough to be on the slip angle peak
value.
- If you are working to a "sim for fun" (like I'm doing), just forget
about the keyboard! There's no point in implementing real physic and
then "destroy it" with key control?.. would it be possible to drive a
real car with the keys???
Stefano Casillo
[email hidden] (Steve Hovelroud) wrote in message news:<[email hidden]>...
Quoted text (25 lines)
> I've struck a problem withg using keyboard control for steering in my > sim. > > I need to use keyboard steering and I've basically got the left & > right keys increasing the steering angle to about 30 degrees. If the > user has no keys pressed then the steering travels back to 0 degrees. > > If I use Beckmans magic formula (Phors Part 29...), > > Flateral = weight * 31 * SlipAngle * (1 + POW(ABS(9.625 * > SlipAngle),2.375)) > > then the front-wheels slip angle is only momentarily at its peak, it > then passes through the max-force slip angle & tapers off, leaving the > car with extreme understeer. > > If I however limit the max steering angle to +/- 8 degrees it steers > ok, but once the car starts to slide there isn't enough steering angle > to control the oversteer & it's useless. > > It appears to me that one can't use keyboard control with slip > formulae that taper off after the peak, if I use a simple flat > limiting it works ok. > (Well sort of works, it still won't slide out with full steering lock > while accelerating from 0.)[email hidden] (Stefano Casillo) wrote in message news:<[email hidden]>...
Yeah, it is mainly for fun (some way to have fun eh...) and I was
thinking along the same lines as you. There's no way to make a serious
sim with keyboard control.. I might just finish off my fun driving
game then move to a more serious "SIM" later on.
Another question, re Slip, I got to thinking last night, assuming no
weight transfer in lat or long direction & no long slip, if the car
was going fast, & one slowly started to turn the steering wheel in one
direction, if you turned it very slowly (and didn't hit anything),
wouldn't the rear/front slip angles be nearly the same, & if this was
the case, due to the front wheels steering, & therefor have a
fractionally higher slip angle (just a bit), wouldn't the front end
always lose it's grip before the rear end & therefor always
understeer??
(Sentence too long.. breath ;)
I thought I understood this "SLIP" thing but other people in RAS are
saying the car should eventually spin out???
Quoted text (15 lines)
> This is my point of view: > > - If you HAVE to implement keyboard control (i.e. if u are working for > a commercial product)... try to implement some sort of "magic steer" > that will turn the wheel just enough to be on the slip angle peak > value. > > - If you are working to a "sim for fun" (like I'm doing), just forget > about the keyboard! There's no point in implementing real physic and > then "destroy it" with key control?.. would it be possible to drive a > real car with the keys??? > > > Stefano Casillo >
The grip also depends on the weight on the contact patch. If you're
cornering as described below, but you come off the throttle, then weight
goes forward, rears lose grip, and the back end steps out.
If you're on the throttle (in a RWD car) then the rears "use up" some of
their grip to provide longitudinal force (think of the traction circle). The
rear tyres therefore have less "room" to grip on the lateral axis and so
start to slide before the fronts.
Quoted text (13 lines)
> Another question, re Slip, I got to thinking last night, assuming no > weight transfer in lat or long direction & no long slip, if the car > was going fast, & one slowly started to turn the steering wheel in one > direction, if you turned it very slowly (and didn't hit anything), > wouldn't the rear/front slip angles be nearly the same, & if this was > the case, due to the front wheels steering, & therefor have a > fractionally higher slip angle (just a bit), wouldn't the front end > always lose it's grip before the rear end & therefor always > understeer?? > (Sentence too long.. breath ;) > > I thought I understood this "SLIP" thing but other people in RAS are > saying the car should eventually spin out???[email hidden] (Stefano Casillo) wrote in message news:<[email hidden]>...
It seems to me that if the racing line is known then you can quickly
figure out if the user is steering too much or too little, and adjust
the steering automatically.
I am assuming that this is what N2002 does when using the keyboard,
works quite nicely ....
Colin
Quoted text (11 lines)
> This is my point of view: > > - If you HAVE to implement keyboard control (i.e. if u are working for > a commercial product)... try to implement some sort of "magic steer" > that will turn the wheel just enough to be on the slip angle peak > value. > > - If you are working to a "sim for fun" (like I'm doing), just forget > about the keyboard! There's no point in implementing real physic and > then "destroy it" with key control?.. would it be possible to drive a > real car with the keys???[email hidden] (Colin Reed) wrote in message news:<[email hidden]>...
So, by this do you mean that the program should vary the steering
angle to get the car CLOSER to the optimal racing line, thereby
over-riding the user?Wouldn't this then defeat the whole purpose of
having him chose what racing line he wanted?, or am I missing the
point?
I was thinking of varying the max steering angle depending on slip
angle so with large slip angles I could have large max steering
angles, but when the slip angle was within say 10 degrees then limit
the steering to perhaps 10 degrees. This way if the car got completely
out of shape, the driver would have the maximum steering angle with
which to compensate.... just a thought.
Quoted text (21 lines)
> [email hidden] (Stefano Casillo) wrote in message news:<[email hidden]>... > > This is my point of view: > > > > - If you HAVE to implement keyboard control (i.e. if u are working for > > a commercial product)... try to implement some sort of "magic steer" > > that will turn the wheel just enough to be on the slip angle peak > > value. > > > > - If you are working to a "sim for fun" (like I'm doing), just forget > > about the keyboard! There's no point in implementing real physic and > > then "destroy it" with key control?.. would it be possible to drive a > > real car with the keys??? > > It seems to me that if the racing line is known then you can quickly > figure out if the user is steering too much or too little, and adjust > the steering automatically. > > I am assuming that this is what N2002 does when using the keyboard, > works quite nicely .... > > ColinSure,
I think this is the way all the Crammond's Gp series implemented
keyboard control and "low sensitivity" analogic controls...
It can be done... but it takes work, time, data structures (driving
line.. u'll need that anyway for the AI... but when u start a
simulator u dont want to think about that immediatly..do u?)...
My line is this: I just don't care about people willing to play with
keyboard.. Im not loosing money on them like a commercial product
would... and I don't think key control is meaningful for a racing
simulation... simple...
P.S. It would be interesting to know how GT keycontrol is done.. it
feels pretty good to me...
P.P.S. Colin.. I cant remember it now... did raceDemo implement a key
control?
[email hidden] (Colin Reed) wrote in message news:<[email hidden]>...
Quoted text (21 lines)
> [email hidden] (Stefano Casillo) wrote in message news:<[email hidden]>... > > This is my point of view: > > > > - If you HAVE to implement keyboard control (i.e. if u are working for > > a commercial product)... try to implement some sort of "magic steer" > > that will turn the wheel just enough to be on the slip angle peak > > value. > > > > - If you are working to a "sim for fun" (like I'm doing), just forget > > about the keyboard! There's no point in implementing real physic and > > then "destroy it" with key control?.. would it be possible to drive a > > real car with the keys??? > > It seems to me that if the racing line is known then you can quickly > figure out if the user is steering too much or too little, and adjust > the steering automatically. > > I am assuming that this is what N2002 does when using the keyboard, > works quite nicely .... > > ColinNo,
Because the rear slip angle will increase together with the veichle
slip angle...
So, assuming no weight transfer, no load sensitivity...and so on..
You steer the front wheels..create slip angle on the front...unbalance
the car and create a yaw... the rear wheel create a slip angle in
response to the yaw...
now.. do the rear wheels have enough weight to balance the yaw? If
they have...u'll get understeer.. if they dont.. u'll get oversteer...
[email hidden] (Steve Hovelroud) wrote in message news:<[email hidden]>...
Quoted text (36 lines)
> [email hidden] (Stefano Casillo) wrote in message news:<[email hidden]>... > > This is my point of view: > > > > - If you HAVE to implement keyboard control (i.e. if u are working for > > a commercial product)... try to implement some sort of "magic steer" > > that will turn the wheel just enough to be on the slip angle peak > > value. > > > > - If you are working to a "sim for fun" (like I'm doing), just forget > > about the keyboard! There's no point in implementing real physic and > > then "destroy it" with key control?.. would it be possible to drive a > > real car with the keys??? > > > > > > Stefano Casillo > > > > > Yeah, it is mainly for fun (some way to have fun eh...) and I was > thinking along the same lines as you. There's no way to make a serious > sim with keyboard control.. I might just finish off my fun driving > game then move to a more serious "SIM" later on. > > Another question, re Slip, I got to thinking last night, assuming no > weight transfer in lat or long direction & no long slip, if the car > was going fast, & one slowly started to turn the steering wheel in one > direction, if you turned it very slowly (and didn't hit anything), > wouldn't the rear/front slip angles be nearly the same, & if this was > the case, due to the front wheels steering, & therefor have a > fractionally higher slip angle (just a bit), wouldn't the front end > always lose it's grip before the rear end & therefor always > understeer?? > (Sentence too long.. breath ;) > > I thought I understood this "SLIP" thing but other people in RAS are > saying the car should eventually spin out???Indeed, Crammond is the king of making games playable with a keyboard. I've
got a friend who doesn't own a wheel who was able to keep me on my toes
using a keyboard in GP3. ISI still don't have a clue how to implement it.
They promised they'd get it right for F12002 but they haven't (yet).
On Sat, 11 May 2002 16:28:50 +1000, "Damien Smith"
<[email hidden]> wrote:
>Indeed, Crammond is the king of making games playable with a keyboard. I've
>got a friend who doesn't own a wheel who was able to keep me on my toes
>using a keyboard in GP3. ISI still don't have a clue how to implement it.
>They promised they'd get it right for F12002 but they haven't (yet).
Typical of the era both 'companies' are living in.
IF playing with keyboard is OK
THEN controls are totally fudged
In English: it's nice that Crammond has gotten it playable with a
keyboard, but that means the player is helped tremendously; imagine
driving your real car with a keyboard.
But I guess, driving with a keyboard is useful to some people, just
not when you're really interested in driving simulations.
Ruud van Gaal
Free car sim: http://www.racer.nl/
Pencil art : http://www.marketgraph.nl/gallery/
Thanks to everyone who responded, my 1st sim (this one) is mainly for
fun & realism isn't the "TOP" priority. I'm yet to do any AI but I
have thought about it and my next subjet to attack is the 3rd (Z)
dimension. I'm going for an on/off road, multi surface type
hills/jump/road sim game & am using this as a learning experience.
It's already been ongoing for over a year, but I'm slowly getting
there... (not much free time with which to prog in..)
