rec.autos.simulators

   I actually love GPL in a "this is the closest to a true physics
model out there and thank you sooo much Papy" way.Hear me when I say
this is in no way putting down this tremendous effort towards realism
in racing sims.I find that the "feel" of GPL is very good and that the
driving model is deadspot on.What bothers me though is the setup
variables and how they dont seem to do what their supposd to.

  I have read a couple of books on the subject of car setup and feel I
know how to set the basics for a good neuteral handling car.What I
mean is Springs,Rollbars,ride height.But the things I try just dont
give me the results I expect.

  For instance Camber and rollbar settings.Someone other than Papy has
told us GPL doesnt model dynamic camber change with roll(Doug
Arnao??).What this means is no matter what you do to prevent roll of
the chassis the camber will stay the same.I can live with that since
the person states its too cpu intensive at the moment.What I cant live
with is that it didnt say this in the manual that came with the
game.Many GPL sites say to set rollbars you have to tighten until the
readings across the tires are even.GPL does not model dynamic camber
change with roll so how can changing the roll of the chassis affect
camber??I have gone from both extremes in the rollbar settings and the
temps across the tires stays the same.Put zero camber on both rear
wheels and burnout for awhile and youll notive that the temps across
the rears are even indicating that the squat of the chassis and
resulting camber change is not modeled.Again I can live without the
dynamic camber change for now:)Papy please tell us this in the
manual.This way we can set the camber and forget it.

  Another thing is weight transfer and ride height.Ride height is the
only variable we can change that significantly alters weight
transfer.Lower the car and it grips much better(if it doesnt hit the
bumpstops)but the temps dont show this happening.I went from 4in ride
height to 2.75 and the temps from left to right remained the same.This
has nothing to do with dynamic camber so I question the setup menu
alot in GPL

   More Examples:  
  Toe-in has no effect on straightline speed but manual says it does

  Ride height doesnt alter the tire temps but instead just adds grip
  as you lower the chassis.

   No dynamic camber change with roll so no changing the amount of

   roll the chassis does will have effect.

  I only write this because I am hoping someone who is compleytly
knowlwdgeable on these facts can write a Real setup guide for these
simulated GPL cars.

  For example camber again:In GPL,Camber is not changed by any factor
other then how many G's a driver pulls.Just make sure that the tires
are even across.Try to go with symetric camber and live with the loss
of grip since asymetrical camber causes unstable braking.

  Toe-in doesnt have any effect on overall speed so set for handling
   only.

  It would be great if everyone adds to this list so we can all start
to take advantage of the setup menus as the fast guys do:)

Hmm. They may not have simulated the shape of the contact patches, but I
am sure they have simulated the size of it (simulating the grip related to
camber)... The contact patch size in relation to roll is very fundamental,
and I can't imagine this excellent sim not to have this in there...

If someone said that GPL doesn't simulate dynamic camber changes, I would
guess that this is related to something else, like, for instance the
camber change you would see over the suspension travel...

If camber changes with roll aren't simulated you would have just as much
grip on the outside tires if you rode a high curb on the inside... Try it.
I think you lose a lot of grip...

---Asgeir---

On Mon, 20 Mar 2000 10:07:35 +0100, Asgeir Nesoen

  See this is exactly what I mean.Extreme accuracy racing sim
companies should tell the buyers how well everything is modeled so we
wouldnt have these questions.GPL-2 (hopeing)could have a physics
read-me and say stuff like --Toe-in is now modeled completly accurate
so any variance from zero will slow the chassis down.Dynamic camber
change and roll center change with roll is now modeled.High frequency
bumps are now modeled.Camber thrust has been modeled in.Under carriage
downforce has been modeled so Rake will take its effect.Etc...

   What I am guessing they mean by "GPL doesnt model dynamic camber
change with roll" is this.In real life,as a car goes around a turn the
chassis will roll and depending on how much roll it does,camber will
change accordingly.In GPL the  cars roll  but this doesnt cause a
camber change.So a car with lots of roll and a car with very little
roll will have the same camber readings.

   I think I read somenwhere that the camber actually changes due to
how much cornering force is being applied.Pull less G's(dont think GPL
cars even pulled one G)and youll have less camber change.I wish I knew
if this was right.

Just as a matter of interest for you, Skeeter, the ride height for the Lotus
49 (as debuted at Zandvoort) without driver, was approximately 3.75 inches
at the centreline of the front wheels and 4.5 inches at the centreline of
the rear wheels. (wheels being the original 15 inch diameter).

--
Regards,
Bruce Kennewell,
Canberra, Australia.
---------------------------

Increased lateral G's requires increasing weight transfer to the outside
wheels. This generally causes the suspension to compress more on the
outside, and the compression often causes camber changes.
(That is, if the camber varies with suspension motion then it will vary
with lateral G because of the consequent suspension motion.)
--
Matthew V. Jessick         Motorsims

Vehicle Dynamics Engineer  (972)910-8866 Ext.125, Fax: (972)910-8216

I see your point now, Skeeter.

But If camber change is deducted from the G forces in stead of the actual roll,
the the camber change is calculated, but from different input. Simulating the
physics involved in a racing car on todays computers must necessarily tak a LOT
of shortcuts. My guess is that calculating camber on Gs pulled instead of
calculating roll, suspension geometry changes and finally camber changes at the
contactpatch is way, no, WAY to intensive for a meager PIII 550...

And letting people know exactly where every shortcut is in the physics model
would mean giving a lot of information away for free, wouldn't it? I mean,
these guys are protecting their physics engines like their kids...

You could say that GPL is marvellous because it has very accurately modelled
physics, but I would state the opposite, GPL is marvellous because Papyrus has
been so good at disguising the shortcuts they HAD to do in the physics model.

It is absolutely possible to calculate every single aspect of a racing car down
to the tiniest detail, but not at 36ths of a second... The shortcuts Papy made
are almost impossible to spot... Incredible...

---Asgeir---

Sadly... They should licence it like QuakeIII and Unreal

<Snip>

Make that 288 times per second. 8 times the optimal framerate.

--
Olav K. Malmin
remove spam when replying

At least one realtime vehicle model (ours - shameless plug!) is available
for license, and has been licensed twice so far, once for a game machine
(never made it to production[sigh]), other time for <www.smsonline.com>.

When we do analysis (not real time) modeling, for car design, we often use
_much_ smaller time steps than this -- for example, to properly model wheel
spin-up and spin-down.

-- Doug

                Milliken Research Associates Inc.

<snipped>

I think you are confused over the definition of camber change. Camber
refers to the relationship between the wheel and the chassis, not the
wheel and the road surface. So camber change refers to the change in wheel
to chassis angle with suspension travel.

So when GPL doesn't model camber change, that means when the chassis rolls
and the suspension moves up or down, the camber angle of the wheel in
relation to the chassis does not change as is would in a real car with an
unequal length A-arm type suspension. This does not mean the angle between
the wheel and road doesn't change...it does. When the chassis rolls the
wheel leans by the same amount it GPL instead of by a slightly different
amount like it would in a real '67 GP car, that difference being the
amount of camber change.

What makes camber change desirable in the real world is you can run less
static camper for better straight line grip and still get the desired
wheel to road angle in the corner. Kind of makes straight axle suspensions
look good but that's another story.

So, the lack of camber change in GPL still means you need to set wheel
cambers for the best compromise between cornering grip when the chassis
has rolled, and straight line grip for acceleration and braking. This is
why GPL still models car physics so well even with out camber change.

Good job Papy. Now where are the new sims based on this physics engine?

Marty

This statement is incorrect.  GPL does, indeed, model camber change with
bump.  What it does not model is the change in the *shape* of the
camber-vs.-suspension-deflection curve at different ride heights.

In GPL, if you set your ride height to 5", and set your static camber
to -2.0 degrees, then push the chassis down 1", your camber will change by
some amount, X.  If you then set your ride height to 4" and your static
camber to -2.0 degrees and push the chassis down 1", the camber will still
change by X.  In real life it would change by an amount probably close to X,
but not exactly X (how close would depend on the suspension design).

Randy

    Now there's something to find out..   What is the travel length of the
suspension.   If I set the SRH to 2.5 inches, I know the chassis can only
travel at max 2.5 inches down before it hits the pavement.  Maybe I should
ask, "what is the geometry of each cars suspension in regards to suspension
travel?

On Wed, 22 Mar 2000 16:41:11 -0500, "Randy Cassidy"

  Hello Randy,I assume your Papyrus's Randy Cassidy.I know that
Papyrus is working on different projects but was wondering if you
could do one more tiny patch for GPL.What I think would greatly
benefit setting up the cars in GPL is an indicater of when the chassis
hits the bumpstops.In a real car you can feel the chassis when it hit
the bumpstops and notice the corresponding loss of grip at that end of
the car.In GPL we get no indication of this but instead just lose the
grip at the axle it happens at.Its hard to tell if the sudden loss of
grip at the fronts is due to the weight shifting off of them on
accelleration or the fact that they just hit their bumpstops.

You could easily use the unused portions of the screen for
this.Divided into 4 they could light up when the wheel at that corner
of the car hits its bumpstop.It would allow ease of trying different
wheel rate/ride height since it could show when the suspension
bottoms.

   Example:_Wheel rate at 50F/75R and a 2.5 rideheight and you see
that you hit the bumpstops.Try 55F/80R and 2.5in RH or 50F/75R and 3in
RH...you get the point:)Just wondering if alot of people would ask for
this,would it be made?

    isn't that like downgrading GPL in to a Game?

  Why would this make GPL into a game?First off it is a game but we
can call it a sim due to its extreme realism in simulating racing:)

  Like I said a real car gives this very important feedback so you can
setup the chassis not to do this.If I take curva grande and my
suspension bottoms,I wanna know about it.I dont want to be thinking
that I came into fast and overcooked the turn.Once I know I WILL fix
it and then there will be no more signs showing me this.Now how is
that making GPL into a game?

On Thu, 23 Mar 2000 04:13:47 GMT, "Mike Barlow"

%   Why would this make GPL into a game?First off it is a game but we
% can call it a sim due to its extreme realism in simulating racing:)

%   Like I said a real car gives this very important feedback so you can
% setup the chassis not to do this.If I take curva grande and my
% suspension bottoms,I wanna know about it.I dont want to be thinking
% that I came into fast and overcooked the turn.Once I know I WILL fix
% it and then there will be no more signs showing me this.Now how is
% that making GPL into a game?

All we need is a "thump" sound or something to key us into the fact that
we have hit the bumpstop.

However, with FF there are some very subtle cues that one has hit the
bump stops.
--
**************************** Michael E. Carver *************************
     Upside out, or inside down...False alarm the only game in town.

=-=-=-=-=-=-=-=-=-=-=-=-=-=-=<[ /./.  [-  < ]>=-=-=-=-=-=-=-=-=-=-=-=-=-=