Where is the Center of Gravity

[randywebb]

Nobody seems to know for sure; hence this thread.

Walt Fricke once posted that his notes indicated it was 20.7 inches above the ground; no statement as to where it is on the longitudinal axis.

[randywebb]

In response to a question on a 1973 car, Ed Mayo (PCA Tech. person) gave this answer:

It's been a number of years since I went through that exercise since it involved some time consuming gyrations on a set of scales to accomplish all th is. I believe the newer electronic scales actually have a way of computing this much easier now, but since I still have my old technology electronic scales I can't confirm this. Trusting to my memory here I believe the fore/aft point was the jack point, which makes sense doesn't it, and that vertically it was about 18 " high. Sorry I can't be more precise here.

Ed Mayo - PCA WebSite - 11/19/2006

[javadog]

Randy,

The height is a little tough, as you have to weigh the car while tilted on an incline. The fore/aft number is easy. Use the front/rear weight distribution and the wheelbase.

By the way, what 911 are you asking about?

JR

[SandyI]

For a 911, I thought the center of gravity is centered between the two front wheels. That's why it has such a distinct pendulum feel when you throttle steer. And the reason the rear end comes around with such force if you lose it.

Or not.

[javadog]

Nope.

JR

[SandyI]

But if it isn't between the two front wheels -- where is it? It can't be in the middle, under the drivers and passenger seats. Are you saying it's between the front wheels and the seats, like under the dashboard?

[joetiii]

According to Vic Elford, the pivot point on the 911 is right over the front wheels which is why the car is a phenomenal car to drive in rally conditions.

The center of gravity is something completely different and that which I know very little about except for the fact that lower is better.

[SandyI]

My mistake, I was thinking about the pivot point.

[afterburn 549]

Maybe everyone is mixed up between balance pt and the exact cg . for a certain is up a bit from the balance pt.......

[randywebb]

There is also roll center, and the Cp, Center of Pressure...

I'm talking generally about the early cars, tho the others will be similar (except for the '72, where the factory says the oil tank relocation moved the CG forward by 6").

[jmz]

the earths core?

[SandyI]

Originally posted by jmz
the earths core?

[YTNUKLR]

CG does not need to be in a "structure," like under the seats. Center of gravity can exist in space, I would imagine it be approximately where Ed Mayo states, probably just in between the seats and ~20 inches off the ground.

The universe is really the center of gravity, the earth is just a piece of the universe that has gravity

[javadog]

Okay,

If I remember correctly, Randy has a '73 911. If you do the math for a car like this (42%front/58%rear weight distribution and an 89.5 inch wheelebase) the fore/aft center of gravity works out to around 3 to 4 inches in front of the jacking point. Obviously, this will vary with the fuel load. Interestingly enough, if you look at the occupant placement in the front seats, they sit just aft of the center of gravity. I'd wager that with a full tank and a driver on board, the C of G is close to that of the unladen car. Anyway, what this means is that half of the mass of the car is ahead of this point and half is behind this point.

In an ideal world, the left/right location of the center of gravity would be located on the centerline of the car but this doesn't happen with an average 911, at least with a driver on board. The C of G will be located a little left of center. On one of my 930s, with my butt in the seat, I end up with about 120 pounds more on the left side, which moves the C of G about an inch or so to the left of the centerline of the car. Yes, the right front brake likes to lock up a little early.

I have no idea about the height of the C of G, as I've never measured it. I have heard estimates of anywhere from 16 to 20 inches, which feels about right.

The center of pressure is ahead of the center of gravity, although I don't recall ever seeing a reference to it's exact location. This is what makes a 911 such a pain to drive in a gusty cross wind. It's also why I have front and rear spoilers on all my 911s, as they move the center of pressure to the rear somewhat and minimize the cross wind sensitivity.

Roll centers are a completely different topic and are related to the geometry of the suspension on both ends. The roll center at the front and rear of most cars is at a different height.

What Vic Elford was talking about wasn't related to the center of gravity of the car. He just like the ability of the early 911 rally cars he drove to be steered with the throttle, in many cases way beyond what you would want in a track car. He was talking about swinging the rear completely around in tight turns. This was something the 911 did well, what with so much of it's weight in the rear and his genius level of car control.

JR

[randywebb]

I think Bill V. posted this and suggested that the CG of a 911 should be about where that of a 996 is.

[Tim K]

Most scales do have the ability to calculate the height of the C of G (i.e. Intercomp). However, what's kind of a pain is that the suspension is supposed to be restrained as it will move when one end is inclined and weight is removed from that end. The rear dampers can be replaced with solid links. But the front struts? I dunno. Maybe a quick and dirty solution would be to use tie-down straps to keep that end from unloading.

If you do not have the fancy scales, this formula can be used:

h= [ ((L*(L^2-H^2)^(1/2)) / WH ( Wr - Wri) ] + r

h= height of the center of gravity off the ground
L= Wheelbase
W= Total vehicle weight
Wr= Rear (or front) axle weight at level
Wri=Rear (or front) axle weight at incline
H= height of the inclined axle off of the ground
r= Rolling radius of the wheels

The problem with some other formulae that are around (Fred Puhn - How to make your car handle) is that they require that the inclination angle be measured. Its a real pain doing it accurately. Then the trig functions require that you carry so many digits that you loose a lot through rounding errors.

If you have even one accurate scale, you can take the necessary measurements one wheel at a time while using spacers at the other wheels.

As it was mentioned, the driver accounts for quite a bit. Also, some even take these measurements with no fluids in the car and their effects are added after the initial calculation.

Hope this helps.

Tim K

Edited for error

[Walt Fricke]

I'd really like to see someone's results from their efforts at measuring the vertical CG, preferably also referencing it to the floor pan or some such, and the height of the reference plane above the ground (since that way perhaps a guy could adjust for the different tire heights and ride height due to suspension manipulation for his own steed). What Randy is remembering has to be something I mined from Ludvigsen, and it most likely was for a 993 or later car.

Of course, someone's stripped down race car is not going to be quite the same as someone's full interior street car, sunroof versus non, full roll cage, etc. But it would be a lot better than just guessing. I'd like to figure out how to equalize the roll moments front and rear, and I think fairly accurate CG location including height is one essential ingredient. It certainly is in figuring roll rates and the like.

Figuring where horizontally the CG is is trivial if you have your corner weights.

Locking the front suspension whilst jacking up the rear is tricky. I am thinking a guy could remove the front strut dust covers. Then, with a driver in the car (though in theory once you know the CG location without driver, if you can figure the driver's cg and weight you ought to be able to just add that in), measure how far it is from the strut top to the strut mount on each side. Then make two U shaped open sided pipes of exactly those heights and just wide enough to slip over the silver rod. Then do so and start jacking.

And I think it is generally easier (certainly is for me) to jack the rear up rather than the front. I like Tim's formula, though with a digital level (which you can zero on the car to measure angle change easily) maybe angle measurement is less of a problem. Those read to a tenth of a degree, versus the old standbys that just read to a degree. In any method of measuring I think you get more accuracy the higher you can jack/lift one end of the car, and that may be tough to deal with. Because as you jack the rear up the car wants to roll forward on the scales up front. And scales aren't made to resist much loading in that plane.

Walt Fricke

[914efi]

Is it really true that moving the oil tank forward by ~24" at a weight of ~20 lbs moved the cg forward by 6"?

[javadog]

Originally posted by 914efi
Is it really true that moving the oil tank forward by ~24" at a weight of ~20 lbs moved the cg forward by 6"?

Nope. It'll move, but not nearly that much, even though the weight is probably higher than 20 lbs and the distance moved is probably greater than 24 inches. I haven't measured either one.

Let's use our earlier example. If we start with a car that has a 42%front/58%rear weight bias, if we moved the center of gravity 6 inches forward, we'd end up with a weight bias of around 48.7f/51.3r. I've never seen anything close to that on a 911.

Looking at a few samples from Road & Track tests of the period (as tested weight distribution, with fuel and driver):

'68 911 Coupe, sportomatic 41/59
'69 911E Coupe, roof rack 44/56
'69 911S Coupe 43/57
'71 911T Coupe, sportomatic 43/57
'72 911E Targa 42/58
'73 911S Coupe, sunroof 41/59
'74 911 Carrera 43/57
'75 911 Carerra, sunroof, reactors 43/57

So, there's probably more variation in CG location from model to model, and as a result of different options, than there would be from moving the oil tank.

JR

[cgarr]

I hate math so I did it the easy way:

http://www.thedirtforum.com/centergravity.htm