RE: Roll Centers and CG

["Bergman, Micke" <M.Bergman@domain.elided>]

OK. Lot's of good (and some not so good) stuff about roll centers and CG.
Most important is of course the discussion, which hopefully helps people
understand what it's about and others might get new ideas to try out.

My input....

1) In an Alfa (or any production or race car) you simply can NOT get the CG
to low. Lowering it always improves cornering force (keeping everything else
the same).

This is because 
a) load transfer will decrease resulting increased tire grip (tire physics)
b) vehicle will roll less improving wheel angles

Now roll centers have been discussed in general but it is very important to
separate front and rear end as they have completely different problems. In
general lowering RC without other changes increases roll as the distance
between RC and CG grows. Total weight transfer remains (almost) constant and
weight transfer at the axle with lowered RC will decrease.

There are no 100% rules but as a general direction the roll centers have to
be above the ground and preferably not much. Why car makers decide to do
differently is too complicated to explain here.

As many people already stated having roll centers above CG doesn't work even
if it sounds interesting. Try it if you like - might be a nice show car!

Alfa's (all RWD models) problem is that the height difference of front and
rear roll centers is very big. As you lower them it gets even worse with a
front roll center of -10..0 cm and a rear of 220-260 mm (116 cars, the 105
is even worse)

This means that in cornering about 20% of the rear weight transfer goes
through the RC and not the springs. This calls for soft rear springs (and no
sway bar - which Alfa put on for marketing reasons). It also means that 20%
of lateral force is "unsprung" reducing grip on a rough surface.

The low front RC has the effect that extremely stiff springs (with sway
bars) must be used to reduce body roll. This mainly creates additional
stress on the chassis (and extreme stress for the people trying to find
stiff enough torsion bars).

Theoretically the roll centers don't have a big input on the cornering
forces. On rough or bumpy surfaces yes, but on a smooth race track not. They
affect on load transfer at the axles and the body roll, which together are
most important. This means in practice that roll centers and springs must
work together. Modifying the roll centers without changing suspension is
like pissing against the wind.

What I have done is raised the front RC by 60 mm and lowered the rear one by
20 mm. Then matched springs to this setup. Camber gain is increased and rear
camber and toe changed too. Works fine and on a 100 sec lap my GTV 2.0 is
about 10 sec faster than any std Alfa on street tires.

About the "dive limiter". 
This is a device needed because of the high rear RC. The RC requires springs
so soft they can't control dive enough to prevent a decent air dam from
touching the ground. 
To correct this the dive limiter can be used with great success. 
If it reduces roll? 
In corner braking YES - otherwise NO. 
This because in braking the rear rises increasing the distance between CG
and RC increasing roll (and weight transfer) If the rear doesn't rise this
distance remains constant. I haven't tried but I guess the impact isn't very
big. Now as a mind workout you can figure out what it does with progressive
springs!

About lowering the 116 rear RC in praxis. 
Difficult - hell no. 
Just move the center pivot lower (on your own risk). Moving the chassis
supports - not needed. Two reasons 
a) the Watt linkage tolerates a slight angle without problems 
b) as the car (hopefully) is lowered the chassis supports are already lower
than stock.


Cheers,
Micke

--
to be removed from alfa, see /bin/digest-subs.cgi
or email "unsubscribe alfa" to majordomo@domain.elided