Why is rack & pinion system more desirable than steering box system?

[sugarwood]

What are the pros and cons of either system?

Why do sports cars tend to have the former?
If R&P is more precise, why do they make the other kind?


How the steering system works | How a Car Works

[T77911S]

a steering box may be cheaper, could be more durable.

just my speculation on why RnP is better because I have never thought about til now:
with a box you have an arm that connects to it that then connects to a rod to control the front wheels.
as that arm moves from center it would move the front wheels less per turn input from the steering wheel UNLESS it could be compensated for this in the steering box.
with a RNP the rack moves the same amount left or right based on the same steering input no matter how close the wheels are to their maximum turning point.
this gives a more precise steering input no matter how far the wheels are turned.
also Porsche used RNP so it has to be better..

draw a circle. draw a line from the top down to the center. draw another line at 1 oclock, 2 oclock and 3 oclock. all to the center. then draw a horizontal 2 inch line to the right. the difference between the 12 oclock line and the 1 oclock line is greater than the distance between the 2 oclock line and the 3 oclock line.

all this also applies to your suspension and the lower control arm.
a lower control arm at the 4 oclock position will add negative camber when the wheel moves up because it pushes the wheel out from the car,
a lower control arm at the 3 oclock position will add POSITIVE camber when the wheel moves up, which is bad when it comes to cornering. (for the outside wheel).

[manbridge 74]

Slop.

[Bleedsblue]

I f'ing hate steering boxes. My experience with them is solely related to BMWs, but in short even the tightest, fresh box still loses some of the "feels" and directness of a good R&P. For cars like the E34/E39 (V8 only) BMW 5 series, the choice of a box was mainly a packaging one, and inventive DIYers have spent a lot of blood and sweat on trying to convert to R&P.

The box actually remains my single quibble/dislike on the E34 chassis BMW.

[sugarwood]

Ok, so more moving parts with the steering box.? One extra ball joint on each side?
Is that also called the "floating ball" system?

[1QuickS]

"Floating ball" is actually "recirculating ball". The picture shows a serpentine coil of ball bearings that wrap around the shaft driven by the steering column. As the wrap of balls moves along the length of the steering shaft (worm shaft) they act as a linear gear that meshes with the gear attached to the shaft that the linkage to the steering road are attached (sector shaft).

All the balls in their cage, the fit of the balls in the rolled groove in the steering shaft, the sliding clearances of the cage that holds the balls with the main housing and the backlash in the mesh between the ball cage (ball nut) and the sector shaft are only controlled by tolerances in manufacturing. These naturally relax with time without any method of adjusting tolerances out. All these are contributors to lack of feel in steering input to steering response.



The rack & pinion has a much simpler design with fewer parts resulting in less "lost motion" from steering input to steering of the wheels. Also, as the picture shows, the Porsche rack & pinion (at least the 911s that I am familiar with) have the added feature of a rack that has a spring loaded follower which effectively deflects the rack into mesh with the pinion which minimizes backlash or clearance in the gear mesh; this is observed in the sectioned image at the left.

[Tidybuoy]

My guess, and I'm not an engineer so this is an uneducated guess.....Space. I think less space is required for the rack & pinion. Similar to the torsion bar suspension taking up less space than springs & shocks.

[Bleedsblue]

This is getting interesting! So many technical minds on this board, I love it.

On BMW steering boxes (recirculating ball, I think all or at least the majority are made by ZF), there is actually a nut to "adjust" the box. Though I admit I've no idea what is being tightened, it actually can make a significant difference in the "sloppiness"/response.

I actually thought the recirculating ball box was a smaller overall system than R&P, Vern, but I may be wrong. On the BMW 5 series I was mentioning, there was basically no room for a latitudinal R&P because of the oil pans of both the "big" straight sixes and the V8s. Then on the E39 5 series, BMW worked out a way to fit the R&P in the "small" six cylinder cars, but had to keep the box for the V8 cars. Packaging complexities aren't all about the actual size of the steering system though, of course.

[1QuickS]

Yeah, my 2002s had a screw for adjusting the steering box but that only removes endplay from the sector shaft. There are many more wearing bits & interfaces that wear without adjustment features to compensate.

Steering boxes mount differently than racks so convenience of location may be BMW's choice but many sedans do use R&P; certainly BMW could figure it out if they wanted to. Another aspect of their compact design is the attachment points of a steering box are closer together than those used to mount a rack. This means (in generalities) that the structure supporting the steering box must be of heavier design than for a rack. A heavy structure will deflect less than a light structure and this deflection is another contributor to loss of steering feel. However, a steering box is usually mounted on the side of the engine bay which will typically have a stronger structure to support suspension loads.

As a point of reference: Look at all the F1 cars, Sports Prototypes, Ferraris, Lambos, Loti, etc. using steering boxes to understand which is preferable for performance applications.

[cmonref]

The Porsche steering gearbox is also adjustable thru the shims at the base of the block that holds the rack against the pinion. Haven't tried to access it with the gearbox in the car, however.

[Flieger]

Worm and sector boxes have lots more friction inherent in them and so have much lower "reverse efficiency" meaning less road feel comes back to the steering wheel.

Steering boxes were a legacy from the days of solid front axles that just had one tie rod tying the left and right steering arms together. The pitman arm would attach somewhere on that link.

Racks make much more sense for independent suspension, partly because the tie rods can follow a similar arc to the control arms, reducing bump steer.

[James Brown]

weight also, pick up a old GM box, ugg. P&P can be made very light

[evan9eleven]

Originally Posted by cmonref

The Porsche steering gearbox is also adjustable thru the shims at the base of the block that holds the rack against the pinion. Haven't tried to access it with the gearbox in the car, however.

can you elaborate, maybe with pics? tia.

[1QuickS]

See post #6 for picture of follower resting on back side of rack which deflects rack to minimize gear backlash with pinion.

[chris_seven]

Steering Components (Automobile)

[sc_rufctr]

IMO There's no clear or real advantage to either system except that rack & pinion takes up less space in the engine bay.

And being more modern is probably cheaper to produce.

[T77911S]

if you notice all (I cant think of any that dont) big trucks use a steering box.

your sports cars use the RnP. goes back to what I said about the arm coming off the box, #5 in the pic above.
my jeep has a box and seems to steer ok, my mazda truck has box and I think its about 10 turns lock to lock.
an RnP has the same ratio at full lock as it does in the center as where I don't think the box does.