Is 911 Engineering Actually Backward?

The dynamics are not that complicated. You have 4 tires which generate grip based upon the pressure holding them to the pavement. To generate the most grip you would want proportional pressure to all 4. This will rarely occur dynamically for any car. For steady state cornering the outside pair of tires function nearly alone, but for acceleration or braking all 4 tires can be used. The front tires experience additional pressure when braking and the rear tires experience additional pressure when accelerating.

So for braking some amount of rear-bias is best. Conversely for acceleration with AWD some amount of front-bias is best. For maximum acceleration with 2WD you want 100% of the pressure on the drive wheels. RWD: lifting front wheels is best. FWD: 100% front bias would still suck. Of course nearly all engines are hopelessly weak compared to what tires can handle, so optimizing for acceleration is generally pointless.

Sure, I'll put my two cents in... Wasn't the 928 designed as a replacement for the 911? Haven't they acknowledged that there are more ideal layouts? The thing is purists HATE the idea of going away from Dr Ferdinand's layout. I applaud Porsche for staying "retro" until retro became cool again-all the while making it work. Its a lot like how America threw a fit when mazda was going to make the next FWD!!! Mustang. The public wouldn't have it. Besides, the world would be boring if all cars were alike.

HStockSolo wrote: The dynamics are not that complicated. You have 4 tires which generate grip based upon the pressure holding them to the pavement. To generate the most grip you would want proportional pressure to all 4. This will rarely occur dynamically for any car.

Right. The solution is obvious; we need a car with an engine that is free to move about the chassis dynamically as we require to keep the forces balanced. Kinda like motorcycle sidecar racing, when the guy leans around to get the most out of the rig.

That may not be practical, but if you used a turbojet engine you could use some form of thrust vectoring to achieve the same ends - some outlets to the sides, and of course, some thrust reversers to enhance braking... excuse me while I head out into the garage!

rwdsport wrote: Also, we (engineers at least) know exactly how helicopters fly, thats just a silly comment.

Yup, I know they beat the air into submission.

HStockSolo wrote: Of course nearly all engines are hopelessly weak compared to what tires can handle, so optimizing for acceleration is generally pointless.

This is what I was thinking. Except that NOW all engines are weak, when the 911 design first came out and they were rolling around on that technology, was this the case?

We're also ignoring the polar moment of inertia- it's not just static weight bias, but how that weight is distributed. I understand that cobras were 45/55, most modern mid-engined supercars are 40/60 (within 5% of that), miatas and BMW's are roughly 50/50, I think corvette's are roughly 50/50.

Thing is, (one of) the reason(s) that 911's turn so sweetly, Elise's, Cayman's, Boxsters, (M/R, R/R) is that there isn't a giant mass over the front wheels, making initial direction change very quickly (low directional inertia over the front wheels). Add to that a rearward weight bias which biases rearward more than a 50/50 car for braking and acceleration, you've got a good recipe for all parts of the track. Except steady state cornering, and you overcome that with wider rear tires. Where I think the 911 bites you (over the mid engined cars of the group) is that they have a pretty long lever arm from the steering wheels to the center of mass- the pendulum thingy. The F/R 50/50 cars are probably more forgiving because although they don't turn in with quite the ferocity of the mid and rear engined cars, they have a pretty high polar moment because their weight is more evenly spread front to rear, causing a situation where spinning isn't as likely, but they may not have quite the turn-in, the braking, or acceleration traction.

Again- on the 911 front, I'm going with development as the key- they have attributes that are great for certain parts of the track: braking, acceleration, and corner exit acceleration. At the expense of corner entry stability (I'd imagine trail braking to be a bit dangerous), and steady state cornering speed. These are obviously compared to an equal car with different layout.

On corner exit a rear engined car can take advantage of almost as much anti-squat as a live axle car and still retain the unsprung weight advantage and dynamic handling offered by IRS during entry and mid-corner. Having that extra anti-squat mechanical force applied to the drive wheels allows the driver to get back on the throttle very early in slow corners.

So..
- Accelerates out of a corner like a live axle (100%+ anti-squat) so more power to the ground earlier.
- Retains IRS which allows dynamic traction adjustment and unsprung weight savings (stability and acceleration)
- Exceptional braking performance due to rear bias
- Pendulum at rear allows for easy stable direction change with small throttle adjustments... great in slow corners or fast sweepers.

The only real downside is that it requires some extra precision/timing and smooth transitions on corner entry. That can be a little daunting for people slow to realize a bit of momentum building up in a non-ideal direction somewhere out near the rear bumper

The 911 was a semi-unique approach to a set of problems that ultimately has a better answer but it's success isn't all that surprising.

In reply to Giant Purple Snorklewacker:

Never driven one in anger, but I assume you want to make sure you're pointed straight on hard braking?

DWNSHFT wrote: I could have made it more clear that I'm not doggin' on the 911. I love 'em. I'm just trying to wrap my mind around two seemingly contradictory "facts:" 1. 911s are "bad" because there is "too much" weight in the back, and 2. Ferrari put more weight in the back of their brand new supercar than in the front. Clearly there is more at work here than meets the eye. David

Think of a car as a see-saw with 2 fulcrums (front axle, rear axle). Although both the Ferrari and the 911 have a rearward bias (like 40/60), it's not the same. The Ferrari weight mass is still concentrated between the two fulcrums. The 911's weight mass is concentrated behind the rear-most fulcrum. That's what makes the handling so different.

Javelin wrote:

DWNSHFT wrote: I could have made it more clear that I'm not doggin' on the 911. I love 'em. I'm just trying to wrap my mind around two seemingly contradictory "facts:" 1. 911s are "bad" because there is "too much" weight in the back, and 2. Ferrari put more weight in the back of their brand new supercar than in the front. Clearly there is more at work here than meets the eye. David

Think of a car as a see-saw with 2 fulcrums (front axle, rear axle). Although both the Ferrari and the 911 have a rearward bias (like 40/60), it's not the same. The Ferrari weight mass is still concentrated *between* the two fulcrums. The 911's weight mass is concentrated *behind* the rear-most fulcrum. That's what makes the handling so different.

Not sure that is true. we are talking about a front engined ferrari.. so I would think that the engine, which would be the heaviest part, would be towards the front of the car.

Teh E36 M3 wrote: In reply to Giant Purple Snorklewacker: Never driven one in anger, but I assume you want to make sure you're pointed straight on hard braking?

Not necessarily - you can brake all the way to the apex - you just have to be very smooth and fast coming off them and accurately aimed to exit because you need to go very quickly back to the throttle to transfer weight - but (think steep sloped curves, not spikes) smoothly. People get in to trouble when they pause too long or cannot commit to the throttle because they are not correctly positioned. I know a lot of guys who left foot brake to make the whole thing smooth but I don't because I don't have ABS and end up flat spotting inside front tires when they get light or hit a mid corner bump. (or maybe I just suck at LFB).

Generally, new students are taught to always get the braking done before turning and they learn to blend the inputs with seat time.

In reply to mad_machine:

The Ferrari is like the SA RX-7, "front mid-engine". The entire engine sits aft of the front wheels.

ok.. I was not sure how much was behind the wheels.

Knurled wrote:

mad_machine wrote: I had a 72 911. An E. Except for doing something stupid like lifting or braking while cornering.. I never had an issue with it's handling.

Note that those are bad ideas to do in ANY car, really. Try those tricks in an early RX-7 with the good spring rates and the 18mm rear sway bar, and you'll also be looking at where you came from before you're done saying "Oh SH..."

At least it was not a widow maker like the pre 69 cars with their shorter wheelbases

And I *want* one of those, just because of their cussedness

My son found this out last year during an autocross held on a very slock (but dry) black top surface at the Luzerne County, PA safety training course. He tried to adjust speed (slowing down) in a turn when his 85 FB RX7 swapped ends. Surprise!!!

Better he learn that on course than on the street. He has never made that mistake again.

Javelin wrote: Think of a car as a see-saw with 2 fulcrums (front axle, rear axle). Although both the Ferrari and the 911 have a rearward bias (like 40/60), it's not the same. The Ferrari weight mass is still concentrated *between* the two fulcrums. The 911's weight mass is concentrated *behind* the rear-most fulcrum. That's what makes the handling so different.

I think Javelin has maybe hit it. Let's use trail-braking as an extreme example. Car is all loaded up on the outside, with the front wheels extra weighted from the trail-braking, and the rear wheels extra light. The back end of the car is close to breaking traction. With a mid-engine car (whether front-mid or rear-mid) the mass of the engine is pushing against both the front and the rear tires. But with a 911 rear-engine, the engine (and transmission) mass isn't pushing much against the front tires; the rear tires are having to do most of the work of resisting the momentum of the mass trying to spin the car. The distance the entire engine is >behind< the rear wheels is like a lever arm, multiplying the force on the rear wheels.

Remember the golden rule of 911s: never lift off the gas in a corner. Why? If you do you transfer weight forward, off the rear wheels, reducing their traction. But there is still the same >mass< back there, trying to swing the back end around. So there becomes a huge disparity between rear mass and the rear traction to hold it in place. Which is sortofa technical description of a major spin. ;-)

A good driver can definitely trail-brake a 911. The first car I drove on track was a stock 911SC and because it under-steered in slow turns the best way to get it to turn was to trail-brake some to get the back end to rotate. Then, when the back end has pointed the front end toward the apex you can go for gas and the back end squats down, hooks up, and you come screamin' out of the turn. Of course, get it wrong and you will spin horribly. You don't trail-brake as much in a 911, for sure. I used to eat up 911s in my 914 in decreasing-radius turns because I could trail-brake so much harder.

Good discussion. Keep going!

David

Javelin wrote: In reply to mad_machine: The Ferrari is like the SA RX-7, "front mid-engine". The entire engine sits aft of the front wheels.

I really, really hate this marketing-speak.

Front-engine, mid-engine, and rear-engine are SAE definitions.

"Front-mid engine" is a 3-martini lunch definition.

I mean, okay, congratulations, the engine's in the same spot as in an Econoline.

(And, interestingly enough, EVERY RX-7 I have ever seen did not have the engine behind the front axle centerline. Hell, I'm looking at one right now (and nursing a busted tooth, and cursing Racing Beat for something that is probably not their fault, but I WON THE BATTLE) and the axle centerline is about in line with the front stationary gear. And this is after I set the engine back an inch on the subframe, via shifting the subframe forwards an inch for more caster. (My motor mounts are un-pretty and a constant source of problems)

I've been toying the idea of a one-off, mid engined car to run in B-Mod (similar dimensions to the LeGrands and stuff that dominate now, but built today rather than up to 30 years ago, with tech and design theories in use today, rather than up to 30 years ago), and this got me to thinking, a car with a considerable amount of rearward bias would work great in autocross, since the speeds don't get that high (compared to roadracing, anyway), the extra rear bias would help on the start, and everything that makes a 911 such a treat to drive at slower speeds should make a similarly-biased car rock in autocross...

one thing I'm not sure I quite understand fully, is it the static weight distribution that makes 911s nasty towards the limit in steady state corners (for the inexperienced, anyway), or is it the length of the lever arm? like, if one were designing an autocross car and wanted that super nimble feel, would one go for more static weight on the rear wheels, or would one simply shift the engine back a few inches until the desired feel was achieved?

Knurled wrote: Let's see if you can sense what is about to happen before it happens.

Uh, he braked too late? What does that have to do with it being a Porsche?

HiTempguy wrote:

Knurled wrote: Let's see if you can sense what is about to happen before it happens.

Uh, he braked too late? What does that have to do with it being a Porsche?

it bit?

I thought it was "Concentrating mass at either end of the car is bad". I've read authors who claim that the Cayman/Boxster handles better than a 911 because mass is concentrated in the center of the car.

I am going to add... from what I have been led to believe, a car with mass at one end is easier to recover from a spin than one with it's weight in the middle.

All that weight tends to lead once the spin starts

HiTempguy wrote:

Knurled wrote: Let's see if you can sense what is about to happen before it happens.

Uh, he braked too late? What does that have to do with it being a Porsche?

It's not really about "braking too late"... he was cresting a rise while in a corner and the back end swung around like a hammer. No time to correct and then whoa down for the righthander.

Poor notes? Maybe.

Well, how "bad" is a blonde vs a brunette? I mean, different people are comfortable with different setups, so person A may go faster with a front engine car and person B goes faster with a rear engine setup. Mid-engine? Somewhere between the two. It's going to be very very hard to quantify what's "better" since it's both subjective and subject to every driver's different driving techniques...

Teh E36 M3 wrote: ...(one of) the reason(s) that 911's turn so sweetly, Elise's, Cayman's, Boxsters, (M/R, R/R) is that there isn't a giant mass over the front wheels, making initial direction change very quickly (low directional inertia over the front wheels). Add to that a rearward weight bias which biases rearward more than a 50/50 car for braking and acceleration, you've got a good recipe for all parts of the track. Except steady state cornering, and you overcome that with wider rear tires. ...

Well put, though as I noted, driver preference can either enhance or hinder the above, turning it into an advantage, or scaring the driver.

Javelin wrote: ...The 944 had a "perfect" 50/50 distribution, making it neutral in cornering and transitions. This, however, means the 944 doesn't excel in any one area....

And what's important to note is, just how often on a track is a car truely at "neutral cornering." The answer is never. If the driver is very slightly on the gas, weight's biased rearward, and if he's even slightly off it, it's biased toward the front. So the whole 50/50 myth of perfection is just that, because ONLY when cornering at a constant speed (and not even then if you split hairs) does the car do well, and how often is a track car at a constant speed? Never.