Why are torsion bars inherently inferior?

[ianc]

Hi folks,

I've read over the past few years here a lot of comments about how older 911's are great, but ultimately can't compete with newer cars in the handling dept. This seems reasonable on the face of it, given the leaps technology has made in 20 years, but lately after driving my GF's 2005 330i rather extensively (nowhere near a track of course), it seems like my 87's suspension is better in ALMOST every respect. The BMW kind of 'hops' over bumps or holes in the pavement, and the car will lose its set and line when cornering moderately hard. My 22/29 Carrera remains calm and poised in these situations however. The 911 is more harsh\noisy of course, but that can be a plus or minus at different times depending on use. Mine is a DD, but I'm alone in it 70-80% of the time.

I don't know what kind of suspension the 330i has, but I would suspect it uses struts.

Just what is it about newer suspensions (964's and 993's in particular, but all cars in general) that makes them inherently better (or more efficient\capable) than the TB's Dr. Porsche first gave us?

I would be most interested in thoughts on other suspension geometries and their strengths\weaknesses compared to TB's. IIRC, Tyson has offered quite a bit on suspensions in the past, but never specifically on this question.

Open for comment, so please do....

ianc

[RWebb]

The 'best' suspension is usually one that uses unequal length A-arms ('wishbones') in terms of performance. But there are packaging considerations also - with 2 big A-arms - where do you put the motor? the luggage? The McPherson and Chapman stru designs try to preserve the nice things about the suspension while still affording some room for the things the suspension is designed to carry.

The Nice Things:
One thing you want the suspension to do is to maintain the tire's contact patch and attitude as the car rolls, and as the wheel travels up and down i response to bumps in the road. The unequal length A-arm design is very good at this. Other designs less so. The early 911 has a particular problem with toe and camber changes as the the wheel travels up and down.

Now, none of this really relates to torsion bars per se. It just so happens that torsion bars give excellent packaging.

The only real issue with torsion bars per se, is that you cannot change the spring rates at the limits like you can by making coil spring elements taper at the ends.

It isn't so much the coils in the 964/993/996/997... that make them superior it is the suspension geometry.

That's the main outline of the problem. As usual, there is a lot more - some of it related to yaw, ... but yawn...

[Porschekid962]

Firstly you must take into account "who" is telling you that torsion bars are inferior. Secondly look at what top flight racecars of our time are running, you will find quite a mix. Certainly equal length and unequal length a arms are the stuff of choice but why? Because its simple, plotting setups is easier and racecars these days are built for it. That doesnt mean the Tbar is out or obsolete. I believe it was either Renault or Ferrari that were using torsion bars on the front ends of their formula one cars for the past few years. Coilovers are and seem to be much easier to "tune", change of ride height, change of springs, revalving (in the case of a push or pullrod activated damper) whereas a torsion bar becomes a bit more buried into the setup. If you have a suspension setup that has been tweaked and tuned for ages much like the Porsche front a arm with torsion bars is, not only do plenty of options arise but different setups emerge for different kinds of racing

Most people toss t bars because someone told them too or there are more expensive options that are seeminly better for coilovers or who knows what. There is plenty of bench racing going on and a lot of real world knowledge. The nice thing about t bars in a porsche is that it is putting weight at the lowest point in the chassis, they are relatively inexpensive and there are lots of options for them. Going to coilovers usually requires modifications of one sort or another. Lesser tire clearance, structural enhancement on the rear end and cost.

You can rant and rave all day about different suspension setups, if you want to inform yourself more on the topic pickup Alan Staniforths "Suspension" book. Covers all types of setups and go's through the pro's and con's of all plus has input from some heavy hitting car designers of the somewhat recent era, Tony Southgate, Gordon Murray and some other chap.

[ZOA NOM]

Formula 1 uses torsion bars. I'd say that's a pretty good endorsement.

[Won]

Originally Posted by ZOANAS

Formula 1 uses torsion bars. I'd say that's a pretty good endorsement.

Exactly. Torsion bars offer very tight packaging and reduced unsprung weight. They're easy to swap out too (think front suspension of a 911).

I believe the 330i has MacPherson front and multi-link rear suspension.

What makes the newer suspension superior is better overall design like the use of "multi" link system and revised pick-up points so the geometry is optimized through wider range of travel. As is the case with older 911s, the suspension attachment points using torsion bars are very limited by design due to the length and shape of the bars, unless you go with push-rod/rocker arrangement.

[Ole Petter]

A lot of front wheel drive cars use torsion bars in the back. That obviously is for packaging...

As for the 911, the RSR/935's of old used coil overs. So with upped horsepower, bigger wings and more tyre patch the coil overs gives greater tuneability for different tracks and different aerodynamic demands. Added downforce requires stiffer suspension.

Torsion bars should be just great for a road-biased 911 with a wide variety of use. For a dedicated racer I would think the added adjustability is a bonus. But the coil-over RSR/935's have much more solidly mounted suspension, with roll-cages mounted at the suspension pic-up points and solid bushings etc...

So it is not just a question of which is best... The frist RS models had torsion bars. A well built 911 racer with torsion bars can be just as "good" as a 911 racer with coil overs. But for a RSR-class racer they may be a neccessity. For a street driven car, I don't see the point. And the torsion bar has an advantage in that upping the thickness doesn't "destroy" the suspension.

As for here in Norway, i don't know of any "old" cars that has been equipped with coil-overs...

Ole

[911nut]

Suspension design is really about achieving the geometry the engineer wants within the constraints of the package. F1 cars have used coilovers with their pushrod suspensions but it's easier to package torsion bars and lighter too; the torsion bars they use are tiny.

The 935 used coilovers for weight reduction. Since the a-arms weren't being subjected to the bending forces of the torsion bars they could be made lighter. Titanium springs were also used for the weight reduction. This should lead us to conclude that, from a kinematics standpoint, that coilovers are superior to a torsion bar suspension due to quicker response as it's components are lighter (however I don't think that it's all that important on a street car).

In "Porsche 911 Story" Frere wrote that the factory switched to coilovers to improve ride quality (road harshness) and in an McPherson arrangement for packaging (the driveshafts for the Carrera 4 had to be accommodated).

From an ownership standpoint I prefer torsion bars. Ride height is adjustable with hand tools and time vs. a $500 sport spring package on my 965.

[304065]

The answer is very simple.

You can't get enough spring rate in the tail end of a 911 with torsion bars. Even if you could make one that would do what you want, it would be very heavy and might fill the available space in the torsion bar tube.

Look at what the factory did in the early '70s with their own race effort, they ditched torsion bars very early in the development cycle.

[Steve@Rennsport]

Originally Posted by john_cramer

The answer is very simple.

You can't get enough spring rate in the tail end of a 911 with torsion bars. Even if you could make one that would do what you want, it would be very heavy and might fill the available space in the torsion bar tube.

Look at what the factory did in the early '70s with their own race effort, they ditched torsion bars very early in the development cycle.

Absolutely right on,...

Its all about getting sufficient spring rates for the application. For example, while 24mm & 33mm bars appear to be very large, they are quite modest spring rates for racing purposes and coil springs are required for such usage.

[Bill Verburg]

Why are torsion bars inherently inferior?

They're not. Each type of spring has advantages and disadvantages, strengths and weaknesses. Specifically wrt to 911s as stated above on 935s the switch to coil assist and straight coilover was to get adequate spring rates. But the advantages of coil over to adjust height, corner balance etc was also much appreciated.

Don't confuse the change in suspension design with spring type. The later suspensions, particularly in back, were developed to open room through the middle of the car for AWD, and to provide more favorable toe and camber patterns, that part has nothing to do w/ spring type.

[ljowdy]

Using a Formula 1 car as a standard to state that "if it's good enough for them then it's good enough for me" doesn't really mean anything. Look at a Formual 1 car as it goes through turns, the suspension hardly moves.
Formula 1 suspension engineers have said that there is more suspension in the tires than in the car. In other words, the tires "are" the suspension.

I have to agree that the coil overs are the way to go because they can be tuned easier than torsion bars. (spring rates and shock adjustments or replacement)

[304065]

I agree, F1 and karts rely on the sidewall as a spring.

I got the Intercomp catalog the other day, I see they sell a strain-gauge type device for measuring sidewall rate. Clever.

[Grady Clay]

Great thread.

I don’t have much to add.

An important issue not regularly discussed if the ‘spring rate’ of the tire.
That changes dramatically with large diameter wheels, short profile tires and rigid sidewalls.

The suspension design, spring rate, camber change and toe change have to be designed for this situation.

Best,
Grady

[dd74]

Originally Posted by Steve @ Rennsport Systems

For example, while 24mm & 33mm bars appear to be very large, they are quite modest spring rates for racing purposes and coil springs are required for such usage.

OK, so if one wants higher spring rates, but wants to/has to stay with torsion bars, what are their options for higher spring rates? Shocks possibly?

Or as Grady suggests, the tires?

[ZOA NOM]

Originally Posted by dd74

OK, so if one wants higher spring rates, but wants to/has to stay with torsion bars, what are their options for higher spring rates? Shocks possibly?

Or as Grady suggests, the tires?

Lose weight. The lighter the car, the higher the effective spring rate will be for the same torsion bar.

[Tim K]

Originally Posted by ZOANAS

Lose weight. The lighter the car, the higher the effective spring rate will be for the same torsion bar.

Basically. The natural frequency of the sprung mass (body) increase as the mass itself decreases. We all know that there are other advantages as well

Originally Posted by dd74

OK, so if one wants higher spring rates, but wants to/has to stay with torsion bars, what are their options for higher spring rates? Shocks possibly?

Or as Grady suggests, the tires?

A tire with a lower "spring rate" will lower the total rate of the system. I think this could be helpful depending on the surface that the vehicle is riding on. The tire is a spring that is in series with the vehicles suspension. It does not affect the movement of the sprung mass (the body) relative to the unsprung mass (the suspension).

Tim K

[gtc]

Remember that a coil spring is just a long, wound up torsion bar...

I'm curious now as to how much unsprung weight was lost when the switch to coil springs was made in the road cars.

[Chuck Moreland]

Originally Posted by gtc

Remember that a coil spring is just a long, wound up torsion bar...

I'm curious now as to how much unsprung weight was lost when the switch to coil springs was made in the road cars.

None, all else being equal coil springs add unsprung weight.

A tbar is 100% sprung weight. A coil is roughly 50% unsprung, depending on the motion ratio.

[304065]

Originally Posted by dd74

OK, so if one wants higher spring rates, but wants to/has to stay with torsion bars, what are their options for higher spring rates? Shocks possibly?

Or as Grady suggests, the tires?

If you want a higher spring rate out of the tire, +1 or +2 on the wheel and use an ultra-low profile tire to maintain the same rolling radius. This is why you see 911SCs in G class (or whatever the new designation is now that PCA has inverted the class designations) rolling on 275-35-17 tires on Fikse FM10s. The availability of performance tires in the larger diameters helps.

But watch the inertia of the larger wheel, it might not spin up and down as well, and also watch total unsprung weight, these matter a lot.

Or you can use helper springs on the shocks. . . check your rule book.

[RWebb]

re JC's & SW's posts ("You can't get enough spring rate in the tail end of a 911 with torsion bars.
* * *
right on... can't get enough..."
_________________________________

That is very true and Porsche's "Cheatin' Loophole Genius Race Engineer" found the solution -- it was added coil-overs. after which they went to pure coil-overs w/no torsion bars.

But that didn't really seem to be the original question. He asked about superiority and made the comparison to street cars, where one wants a relatively supple suspension, and not ultimate spring rates.