rec.autos.simulators

I know this question has been talked about a lot, but I haven't yet
seen an answer that allows me to understand what those beast are and
how they work. And adjusting them without knowing how they work leads
to nothing and is driving me nuts. So: what the heck are those
clutches, where are they installed, and what do they do? Likewise for
the ramp angles.

Let me tell you how a standard differential gear looks like, and you
tell me where the clutches and ramp angles come into play. Deal? [I
don't have a proper dictionary for automotive terminology, and I know
the names of most parts only in German, so I have to paraphrase a bit
at times. Bear with me. Info about the proper English terminology is
appreciated.]

In the diff box, there is a huge gear that is driven by a smaller gear
on the end of the driveshaft. In the center of the huge gear, there is
a casing that contains four small conical gears. Two of those gears
are connected to the two axleshafts. The other two gears sit on a
separate pinion shaft.

The whole thing looks (roughly :) like this:
[Switch to Courier of another fixed pitch font, or it will look even
less comprehensible...]

                 |\
                 ||XXXXX|XXXXXX
                 ||     |     X
                 || \\zz|zz// X
                 ||  0  |  0  X
=================||==0  |  0==X===============
                 ||  0  |  0  X
                 || //zz|zz\\ X
                 ||     |     X
                 ||XXXXX|XXXXXX
                 |/

The bigger part on the left side is the big gear I wrote about.
The things made of "=" signs are the axleshafts.
The part made of "X"s is the casing.
"\zz|zz/" is one of the conical gears with a part of its pinion shaft.
"\
 0
 0
 0
 /" is one of the gears at the end of the axleshafts.

Now, where do the clutches come in? And to what do the ramp angles
refer?

Thanks for any insights!

--
Wolfgang Preiss   \ E-mail copies of replies to this posting are welcome.

<snip the cool ascii art n stuff>

I'm not sure of the physical layout of a 60's era differential, but I
understand how it works.
Ramp Angles...
the two numbers for ramp angles, 85/30, correspond to power/coast values for
when the diff locks up.  Higher values on the power side means quicker
lockup when you apply throttle, lower values on the power side means lazier
lockup when you apply throttle.
On the coast side, lower values means that when you lift off the throttle,
the diff is slow to lockup, higher values means the diff locks at the
slightest hint of throttle lift. (I find this undesireable)
example...15/85 values would mean you pretty much gotta floor the gas to get
the car to move, yet as soon as you begin to lift off the gas the diff would
begin to lockup and try to slow the car.
Clutches...
this sets how tightly the diff locks up between the rear tires.  More
clutches means that the rear tires are allowed less ability to spin at
different rates (like when turning a corner).  Less clutches allows for a
little more play between the tires.  If you find that it's too hard to
control a powerslide or trailbraking, try less clutches.

Hope that helps

Andre

On Sat, 31 Jul 1999 07:03:12 -0700, "5th Doctor"

http://simracing.com/alison/gpl/ferrari.htm

Where Alison Hine and Dave Kaemmer are discussing GPL diff. settings.  Give the
article a peek if you haven't; it's food for thought, even if your car's not
red.

Cheers,

Steve B.

"Note that in a Salisbury-type differential such as those used by GPL's cars, a
higher ramp angle (e.g. 85 degrees) gives less locking, while a lower ramp
angle (e.g. 30 degrees) gives more locking. On the other hand, higher preload -
more clutches - gives more locking."

remove "edy" from address for email

--

Header address intentionally scrambled to ward off the spamming hordes.

cisko [AT] ix [DOT] netcom [DOT] com

        See Carroll Smith's book, "Drive to Win" for a complete
explanation with pictures.  Much other useful stuff.

                bob

1- Almost every setup I've ever seen uses a ramp angle of 85/30. From your
explanation,I take this to mean that the rear locks up slower when applying
gas as well as locking up slowly when getting off the gas and on the brakes.
Or would one be better off using a lower power number?

2- The more clutches the tighter the rear. Yet most of the latest setups
have gone to 4 clutches rather than 2 for the Lotus and Ferrari. From your
explanation,I would think that 2 would be a better choice.

     At the start of the race when I apply the gas (no matter how easy),the
car pulls ***ly to the left. I always get left behind. Once I'm
going,the steering is as it should be. This problem was never this bad until
I installed the patch and started using a MSFF wheel rather than my T2. Slow
corners that require 1st or 2nd gear usually wind up with *** oversteer
when I apply the power.

Joel Willstein

A usual "open" diff, which we have in our normal road cars,
distributes the power to the wheels in such a way that - when going
through a curve - the inner wheel can rotate less frequently than the
outer wheel, which has to cover a greater distance.

This is a Good Thing, but it has a downside to it: as soon as one of
the wheels loses grip (f.i. the inner wheel in a curve), all the power
goes to this wheel. In an extreme case, we have one wheel spinning
madly, while the car stands still.

Using NO diff is not an option on road courses, since the car wouldn't
want to turn at all with such a setup. This leads to a lot of tire
wear and a very ugly behavior in tight turns. Yet, the car would
accelerate very nicely in a straight line. Champ Cars on large ovals
do without a diff and use "stagger" (rear wheels of different sizes, a
larger one on the outside) to facilitate turning.

What we want is a diff that allows us to accelerate out of turns
without having the inner wheel spinning, yet does not make the car
understeer madly in turns. This is where a "limited slip" diff comes
into play.

In an open diff, the axle shafts can turn (more or less) independently
from the diff housing. This goes so far that, when the car is jacked
up, you can turn one wheel clockwise, and the other wheel will turn
counter-clockwise. The diff housing will not move at all during this
operation.

The infamous clutches will prevent this from happening. They connect
the axle shafts to the diff housing, but allow some slip. The more
clutches you pack in there, the less slip they will allow. The static
"clutch preload" does not depend on the torque of accelerating /
braking. Therefore, if you put too little preload in there, you still
will have wheel spin. If you dial in too much preload, OTOH, you get
basically a completely locked diff, which is even less desirable.

Fortunately, some folks (Smith mentions Mike Endean and Hewland)
developed the concept of "ramps". It is difficult to describe how
those ramps actually look, but they work like this: imagine a wedge
that is located between the axle gears and the diff housing. When you
apply power, the wedge will be pressed into the gap and make the two
parts turn simultaneously - with a certain amount of slip.

The force with which the wedge, or ramp, links the two parts depends
on the torque applied (the more torque, the greater the force, the
less slip allowed) and on the angle of the ramp - the ramp angle. A
30 angle will transmit the greatest force and give the least slip.
(even sharper wegdes would supposedly give even higher pressure and
locking power, but are not used, apparently.) A 90 angle is not
really a wedge or ramp anymore, but a stop, and will not exert any
pressure.

The real nice aspect of this design is that you can use two different
wedges, one for accelerating, and one for engine braking. When
accelerating, the engine gives torque to the axles and the wheels, and
the power, or drive side ramp comes into effect. When engine braking,
the wheels are giving torque to the engine, and the coast side ramp
starts working.

Therefore, you could use a setup like 30/90 (30 power side ramp angle
/ 90 coast side ramp angle) to get very little diff slip while
accelerating, and a completely open diff while coasting. I assume that
the transition between "almost fully locked" and "open" (and vice
versa) would be rather harsh, though, and might unsettle the car in
mid-corner when you step on the throttle again. The clutches are used
to give a certain amount of lock all the time and smoothen the
transition between coast and power phase.

Excuse me now, I have some testing to do at Monza. :)

--
Wolfgang Preiss   \ E-mail copies of replies to this posting are welcome.

Maybe your setup??  I dunno.  I don't use a FF wheel.  My suspension is
symetrical, I don't use the clutch for starts, and I can usually launch the
car straight.

Here's my take on this (I hope I'm close to being on the right track, so to speak -
comments please!):

1. More/faster locking of the differential results in more/faster oversteer
response.

2. Decrease the power side ramp angles to quicken the locking and thus increase turn
exit oversteer.
    Increase the power side ramp angle to slow the locking and thus decrease turn
exit oversteer.

3. Decrease the coast side ramp angles to quicken the locking and thus increase turn
entry oversteer.
    Increase the coast side ramp angle to slow the locking and thus decrease turn
entry oversteer.
    (I change an 85/30 setting to an 85/45 setting when I'm into strong oversteer on
turn entry).

4. Add clutches to increase mid corner oversteer.
    Remove clutches to decrease mid corner oversteer.

5. Do all the above before adjusting anything else (??)

I'm not quite sure of the mid-corner contribution of the ramps as your note
indicates that they
change both the force (locking?) and the slip (speed of locking?).

So I went to Monza and took a close look at my current ride, a Ferrari
with a modified Alison Hine setup (modified means I changed the
gearing and adjusted the roll bars and braking bias according to my
personal needs, but left the rest unchanged. This setup used to be
"Ali Silverstone neutral", not the Monza setup which has still too
much understeer for my liking.)

I was surprised to find the setting for the ramp angles set to 85/30.
This should mean "little lock under power / lots of lock while
coasting" according to what I've learned from C. Smith's book. More or
less the opposite of what I expected to be a good setting...

Since many people claim that the numbers in GPL actually mean the
opposite to what Smith says in his book, I did a little skidpad
testing. I reduced the number of clutches to the minimum (in order not
to dilute the results regarding the ramp angles) and tested four
different extreme configurations.

[As an aside: each ramp can have 30, 45, 60, or 85 of angle. It's 85
instead of 90 since the 90 angle parts tended to chip, according to
Smith. I find the pulldown menu in GPL somewhat confusing for a
beginner: since all combinations are allowed, *two* pulldown lists of
four items each would have been handier than one with 16 options.]

I started doing donuts on the front stretch at Monza and found the
following:

85/30 - only inside wheel leaves skidmarks
85/85 - again, skid marks only by the inside wheel
30/85 - both wheels leave marks
30/30 - both wheels leave skid marks

This conclusively proves two things:

1. The first number refers to the power side, the second to the
coasting side. Just as one would expect in reality - but this is
software, so I thought it deserved a special mention. :)

2. A higher number means less lock-up of the diff, a lower number more
lock-up. Sorry, 5thDoc, but you have to re-evaluate your info. Smith
is right.

Well, now I took the car with a 30/85 diff for a spin - literally.
While braking for Parabolica on the outlap, the rear of the car
overtook me. And no, this usually doesn't happen to me ( not anymore,
not all the time, at least.) At first, I was - unpleasantly -
surprised. But it really isn't that surprising when you think about
it: a locked diff means the car is going in a straight line - which is
what you want when braking for Parabolica! An open diff, OTOH, makes
turning easier. But I don't want to turn when braking at full speed.
(If you've seen today's German GP, you know what good a spinning car
at 300 km/h is... ;)

But the change made it much easier to induce throttle oversteer in
mid-corner. A bit *too* easy, to be frank, but a step in the right
direction. I drove a bit with a 45/60 ratio and will explore some
other ratios as well. Suffice it to say that this setting has quite
some effect on the car's handling. Try playing with it!

--
Wolfgang Preiss   \ E-mail copies of replies to this posting are welcome.

Under power, yes.

This depends entirely on how you approach the corner. The lower the
coast side ramp angle is, the more stable the car will be while
(engine) braking in a straight line. And the less willing will it be
to turn. If you are trail braking into the turn, a lower angle will
mean more understeer. I agree with the first part of your statements -
it's the part after the "thus" that I'm not sure about.

This is under dispute among the authorities. Dave Kaemmer agrees with
you, Alison Hine disagrees. I have no opinion yet (and am no authority
either.) :)

compare http://simracing.com/alison/gpl/ferrari.htm

After fiddling with the ramp angles this afternoon, I tend to agree.
They are pretty influential.

I will try to make my terminology industry-standard compliant in the
future. :) It seems I have to resort to a crude ASCII rendering
again... This is my impression of the diagram on page 7-28 in Smith's
book:

                      _______
                     |       |d
                     |       |||
                     \       |||
                     _\ b    |||
                    /a\\     |||_______e______
                    \_//     |||
                      / c    |||
                     /       |||
                     |       |||
                     |_______|

The huge lump in the middle is the side gear ring, aka ramp body.
(a) - the planet bevel gear, which is supposed to be round. :)
(b) - the coast side ramp.
(c) - the power side ramp. (Both ramps have the same angle in this
      example. Too bad ther aren't "/" signs with different angles. :)
(d) - the clutch packs.
(e) - the axle.

Now, when the car accelerates, the bevel gear is pressed against the
power side ramp, which in turn is pressed to the right against the
clutches. This maximizes the lock of the diff:

                      _______
                     |       |d
                     | ----> |||
                     \       |||
                      \ b    |||
                     _ \     |||_______e______
                    /a\/     |||
                    \_/ c    |||
                     /       |||
                     | ----> |||
                     |_______|

When the car is coasting, the same happens on the other side:
                      _______
                     |       |d
                     | ----> |||
                     \       |||
                    /a\ b    |||
                    \_/\     |||_______e______
                       /     |||
                      / c    |||
                     /       |||
                     | ----> |||
                     |_______|

Now, what I wanted to say is simply that, when there are little or no
clutches, the transition between coasting and accelerating in mid
corner might be a bit harsh. In the worst case, it could go from "full
lock" to "no lock" (when the bevel gear is in the middle) to "full
lock" again within fractions of a second. I just feel this could be
problematical.

Cheers,

Steve B.

remove "edy" from address for email