F1 Steering Rack, a closer look

All that has to do with the power train, gearbox, clutch, fuels and lubricants, etc. Generally the mechanical side of Formula One.
Brian.G
Brian.G
311
Joined: Fri Dec 10, 2010 10:52 pm
Location: Ireland

F1 Steering Rack, a closer look

Post

Another part under the spot light, this time the F1 Steering Rack.

In this thread I will be unlocking a few secrets hidden within this part - The F1 Steering Rack and Control Valve, mainly its construction, operation, and some other features.

Before we begin I must give a quick shout out to Craig Scarborough @ScarbsTech. As you can imagine coming up with these posts takes a lot of looking online for interesting F1 parts. Spotted this part on Race to the Finish F1 parts website but being an Engine person couldn't identify what part of the car it was from. A message to Craig set things straight that it was steering related.

First off, a quick rundown on power steering and how it works. While this is basic to start off, it may be no harm read it all the same since sometimes understandings and assumptions of how something works can be incorrect.

As we all know, most steering systems use a rack and pinion system - that is, the steering wheel is connected to a shaft which carries a pinion gear at the end. The pinion engages with a rack - on the ends of this rack are tie rods which connect to the hub carriers. When the pinion is rotated the rack moves in a linear fashion within the steering rack housing thus steering the wheels. The image below shows a basic manual steering rack layout,

Image

Moving on from that, nearly all cars now have whats called power assisted steering. These systems can be hydraulic assist or electric assist. For the purposes of this thread, and taking into account the Redbull part in question is hydraulic assist, that is the system we shall be concentrating on.

Hydraulic assisted power steering uses a few components.

1.Hydraulic Pump to provide system flow/pressure.
2.Spool Control Valve.
3.Rack capable of turning the oil pressure in the system into linear motion to steer the wheels.

The image below shows a basic hydraulic assisted power steering rack layout,

Image

As we can see the spool valve directs the fluid to chambers either side of a piston connected to the rack. This fluid when pumped into either the right or left chamber pushes on the piston and causes the rack to move in that direction. For example, if it is pumped into the left chamber above, it will cause the rack to move in a linear fashion to the right.

Digging a bit deeper into the spool valve and pinion operation it is probably apparent that the above diagram is just a representation of the layout. The diagram above shows the pinion shaft in the steering rack as one piece surrounded by the spool valve.
If the spool valve was directly mounted onto the pinion shaft and the pinion shaft one piece, then the rotational movement from steering wheel to pinion would mirror the workings of the manual steering rack described earlier and any hydraulic assist would be pointless since all the work is still being done anyway with the rack and pinion.

So to combat this, a rotary motion delay needs to be built into the pinion shaft between the spool and the pinion.

This delay then means that the spool has time to ''see'' what rotational motion is at the input, before the pinion does.

In conventional power steering systems this 'rotary motion delay' feature takes the form of a long slender torsion bar inside the two piece pinion shaft as shown below,

Image

On the far right above we can see the assembled pinion and spool valve.
The three parts on the left above are the lower pinion shaft, the torsion bar, and the spool valve.
The part in the middle is the input - upper pinion shaft.

The upper and lower pinion shafts locate together axially by means of loosely fitting dog features. This feature can be seen on the lower end of the upper pinion shaft(middle of photo). These dogs never normally transmit any rotating motion from the upper pinion to lower pinion shaft and would have maybe 10 degrees of rotational play. They are there for when the car is turned off, when the hydraulic system fails, and when the steering is on full lock. In an ideal world, there would be no need for these loose fitting dogs and the torsion bar would suffice on its own.

(Worth noting, many tuners often convert a power assisted rack back to manual for preferred steering ratios by disconnecting the hoses and capping the ports. This is not how to do it correctly as the system will have play since the output pinion is being driven via the loose fitting dogs - the two shafts must be welded together)

Inside the upper and lower pinion shafts lies the torsion bar. This torsion bar is fixed at either end to both upper and lower shaft parts with small 3mm cross pins as can be seen. Once the upper and lower pinion shafts are assembled it should now become clear that the torsion bar provides torsional elasticity from the input, to output/pinion.
Since the pinion is in mesh with the rack - which requires a far bit of force to move linearly, any input to the upper shaft causes the pinion shaft assembly to 'wind up' a few degrees.
Given that the spool valve is attached to the lower pinion shaft with a cross pin, and the corresponding oil ways matching the internal features of spool are machined into the upper pinion shaft this torsional elasticity between both parts allows for oil to flow depending on rotational orientation with each other.

A plan view cross section below of upper pinion shaft (center) and outer spool valve part, and oil paths at various positions,

Image

And with wheel turned,

Image

Since the torsional windup of the pinion shaft assembly depends on the resistance the lower pinion shaft sees at rack interface, once the spool valve has directed the oil to either chamber at rack piston the rack then moves to position and equilibrium is thus reached between lower pinion shaft, and upper pinion shaft and the spool rests in its ''straight ahead'' position even if wheels are turned to left or right.

Having taken all of the above into account we can now progress onto the F1 Steering Rack Control Valve with a better understanding of features, and also on how it differs from a normal road car.

The part below is from the RedBull F1 RB4 (2008). It is very clear it features an axial spool valve,

Image

A bit of searching brought up a few pictures and from those the research below started which ended up being possibly the largest body of work ever done to date on an F1 part in order to create a technical post.
Obviously it is well known how a hydraulic power assisted steering rack works as outlined above but F1 Steering racks are very different. In the end, a rough cad model was generated from many pictures gathered online. Motion Simulations were added gradually to this rough cad model based on each part having one function/possible motion given its location - basic application of classical mechanics if you will. Following on from this it was clear for certain as to how the rack functioned then.

So first off some pictures of various racks in F1 cars. As you are probably aware it is extremely hard get photos of the car in this area since the nose cone has to be removed,

Image

A Lotus F1 car, nose off, a closer look you will notice end of axial spool valve is connected to a silver part,

Image

A later Redbull car, nose off,

Image

Finding this image was a huge breakthrough however it leaded to further confusion,

Image Image credit @KevTs

A great photo also, however again, lead to confusion,

Image Image credit @KevTs

The issue regards understanding function at first glance is to do with the control spool axial displacement, it is only +/- .75mm. Remember, the F1 spool which we will look at shortly has the same function of the basic rotary spool mentioned earlier in thread - it has to see the steering spindle input before the pinion does in order to provide fluid assistance to rack.

The area of concern below,

Image

The grey part looked to be fixed to end of pinion shaft - but it couldn't be - the spool would only allow grey part to rotate a few degrees either way - severely limiting steering lock! If the grey part sat on end of pinion shaft on a bearing as it looks like it does above, then the pinion end would just rotate freely inside it, while at the same time being constrained by the bearings in the main rack housing.

The grey part sitting on end of pinion shaft and controlling spool axial displacement was modeled in cad and motion applied to the system, below is the full cad model displayed,

Image

Image

Image

Image

The main casing was loosely modeled on a part found online made by an Aerospace Company in Italy - the part is 2014 series billet aluminium and looks to be a Redbull item from studying various spool valve mount locations,

Image

Image

Many simulations and constraints were ran with not much success,

Image

Finally a realization was made - it turns out that the pinion shaft is not actually running in bearings in the main housing above at all, and instead just sitting on the what is now known as two flexure mounts at the front and rear of the housing,

Image

A closer look at the flexure mounts which were modeled up from the pictures shows two important breaks either side of pinion bearing housing area. These gaps allow the pinion to float from side to side.

Image

Taking it into motion study and running a deformation simulation on the assembly, it is now clear on how the spool valve is axially displaced. When the pinion is rotated by the steering wheel, the resistance provided by the rack/contact patch of the tyres causes the pinion to move to the left or right as it tries to travel along the rack. The flexure keeps the pinion firmly engaged with the rack in the z direction but allows pinion float in the x direction, in turn actuating the spool valve and in turn supplying fluid to whatever side of the rack chamber the pressure is required to move the wheels and put the system back into equilibrium again.

A large amount of displacement first for display purposes,

Image

Image

Then a more real amount of deflection - ignore values, they are meaningless,

Image

Image

So with all that understood, lets take a look at the spool valve itself, it is a pretty compact part - approx 100mm wide,

Image

Machined from 2 series aluminium billet,

Image

Image

The two small lower ports are oil entry/exit into main rack housing,

Image

Image

Image

Image

Image

The end of axial spool where it connects to flexure, there are two ball joints here given the motion translation required from the flexure,

Image

The steel barrel part appears to be a tuned mass damper - its solid steel. Possibly needed to tune out any vibrations that maybe present in the flexure during high frequency bump steer events or such other vibration/harmonic/hysteresis type events across the mechanical to fluid link.

Image

The male dry break connections,

Image

The pressure sensors (broken) - two of, each monitoring fluid in both rack compartments,

Image

Image

Image

Removing the spool valve end retention plate, titanium part,

Image

Image

Image

Removed other end plate, allows the spool center shaft to be removed,

Image

Image

Image


Interestingly, on inspecting the housing chamfer detail it was not fully deburred and had damaged the seal at time of assembly, pretty surprising on a part like this,

Image

A close up of the center shaft,

Image

The end cap bung seal detail - these seal against main spool assembly, still inside housing,

Image

Image

Image

Removed, take note of fluid passage window ports, axial displacement of center spool shaft covers or uncovers these windows,

Image

Image

Image

Moving onto the multi link control bar input from flexure,

Image

Image

Image

As you can see, the balljoint clearance can be dialled in and the castle nut locked off with outer ring,

Image

Image

The inboard joint is same,

Image

The filter chamber - all hydraulic parts on F1 cars contain their own filter, guessing its to catch any dirt entry from dry break connections, the exit from filter entering center of spool chamber,

Image

Image

Image

Dry break connectors removed - male and female types,

Image

Image

Image

The exploded assembly,

Image

Image

Image

Further research showed that the spool assembly is an off the shelf item from Moog,

Image

Image

So that's that. A much larger body of research work than intended as mentioned, was really flying blind on this one simply because I rarely branch away from Engine technology, that and the fact that actual F1 steering rack information is extremely thin on the ground; until now...

Please feel free to share all info and pictures, no need for permission/credit of any images.

Brian,

ps - I'm not sharing cad of rack as it is useless to anyone, and was just roughed together for sim purposes.
If you think you cant, you wont, If you think you can, you will

Greg Locock
Greg Locock
215
Joined: Fri Jun 29, 2012 11:48 pm

Re: F1 Steering Rack, a closer look

Post

Wow. It might be worth pointing out that at least in a conventional rack the hydraulic fluid flows continuously at straight ahead, and the valve restricts the flow on the outlet side as the wheel turns.

This is why HPAS uses a fair amount of power continuously.

The other option is probably how people think it works, like a hydraulic ram, with spool allowing pressurised oil IN on side of the piston. I have driven a tractor with that system, but it is no good for precise control.

Brian.G
Brian.G
311
Joined: Fri Dec 10, 2010 10:52 pm
Location: Ireland

Re: F1 Steering Rack, a closer look

Post

Greg, Yes! That is worth pointing out...this ''effect'' has killed many Bmw power steering pumps on me..mainly because the insides of hydraulic lines start shedding pretty quick and end up in the non serviceable tank filter in the form of black goop. This then leads to pump cavitation, worn vanes, and much noise.

Gladly no such power drain in an F1 engine though since if there is no demand for flow the pump swash plate ''goes into neutral'' Link on pump for other readers - viewtopic.php?t=28398

Hope you're keeping well,

Brian,
If you think you cant, you wont, If you think you can, you will

User avatar
flynfrog
Moderator
Joined: Thu Mar 23, 2006 9:31 pm

Re: F1 Steering Rack, a closer look

Post

excellent work as always Brian.

Brian.G
Brian.G
311
Joined: Fri Dec 10, 2010 10:52 pm
Location: Ireland

Re: F1 Steering Rack, a closer look

Post

flynfrog wrote:
Wed Nov 17, 2021 2:55 am
excellent work as always Brian.
Thank you!

Brian,
If you think you cant, you wont, If you think you can, you will

NathanE
NathanE
3
Joined: Fri Mar 31, 2017 6:49 am

Re: F1 Steering Rack, a closer look

Post

flynfrog wrote:
Wed Nov 17, 2021 2:55 am
excellent work as always Brian.
+1, thank you, I always find your posts detailed, comprehensive and thoroughly insightful. Much appreciated.

ojlopez
ojlopez
5
Joined: Fri Oct 24, 2014 9:33 pm
Location: Guatemala

Re: F1 Steering Rack, a closer look

Post

Beautiful insight, thanks!

Billzilla
Billzilla
11
Joined: Tue May 24, 2011 12:28 am

Re: F1 Steering Rack, a closer look

Post

Superb detail, as always.
FWIW I'm not sure if it's still the case but certainly a decade or so back an fellow Aussie mate of mine used to make steering racks for half the F1 field.

Greg Locock
Greg Locock
215
Joined: Fri Jun 29, 2012 11:48 pm

Re: F1 Steering Rack, a closer look

Post

Bishop's?

AngusF1
AngusF1
5
Joined: Sun Aug 13, 2017 9:54 am

Re: F1 Steering Rack, a closer look

Post

Incredible. Thank Brian.

It would be interesting to understand the benefits of this mechanism in F1 compared to the standard road-going design.

If I understand it correctly, in this method:
  • The steering shaft is unbroken and does not have a torsional spring in it.
  • Consequently rotation of the steering wheel is transmitted faithfully to the pinion gear.
  • The shaft is supported by two bearings each mounted onto a device which is sprung laterally (the "X" direction).
  • So when the steering wheel is turned the shaft and pinion rotate faithfully, causing a "normal" amount of relative movement between the pinion and rack.
  • However due to the laterally compliant mounts, the shaft, pinion *and rack* move a tiny amount laterally, causing less initial movement of the rack than if the mounts were stiff.
  • This lateral movement of the shaft end is picked up by the spool valve ie the hydraulic amplifier.

Is that the gist of it? I wonder why this method is preferable. Some possibilities:
  • This puts the mechanism in a location where there's space, and it's lower down.
  • Rotational faithfullness between the steering wheel, rack and pinion is maintained which ensures consistency of resistance in the rack and pinion mechanism and gets the stiction out of the way immediately. By comparison the traditional design initially results in less rack-pinion movement than would otherwise occur.

gruntguru
gruntguru
534
Joined: Sat Feb 21, 2009 6:43 am

Re: F1 Steering Rack, a closer look

Post

The system still results in motion loss between steering wheel and rack. How much motion loss will always be a function of spool valve amplification factor.
je suis charlie

Tommy Cookers
Tommy Cookers
578
Joined: Fri Feb 17, 2012 3:55 pm

Re: F1 Steering Rack, a closer look

Post

gruntguru wrote:
Mon Dec 06, 2021 11:19 pm
... motion loss between steering wheel and rack ... will always be a function of spool valve amplification factor.
yes a very interesting thread

afaik .....
we could see the above as a deadband that's the inverse of the forward path gain
eg Detroit and big aviation liked some deadband - as protection against the so-called 'sneeze factor'
using a servo type valve etc could eliminate the deadband
eg https://www.moog.com/products/servovalv ... alves.html

imo and fwiw
deadband is apparent in new (small) cars - but seems to reduce with wearing-in of tyres, toe, and seals
also .... my (Peugeot 206) pump drew 9 amps just maintaining pressure and 13 amps when moving fluid

AngusF1
AngusF1
5
Joined: Sun Aug 13, 2017 9:54 am

Re: F1 Steering Rack, a closer look

Post

gruntguru wrote:
Mon Dec 06, 2021 11:19 pm
The system still results in motion loss between steering wheel and rack. How much motion loss will always be a function of spool valve amplification factor.
How? Not sure I follow, could you explain?

To me it seems like for any given angular turn of the wheel, the pinion will turn by the same angle as the wheel and therefore the rack must move by the "normal" amount relative to the centre of the pinion. Both the pinion and rack will move slightly laterally due to the spring.

gruntguru
gruntguru
534
Joined: Sat Feb 21, 2009 6:43 am

Re: F1 Steering Rack, a closer look

Post

Imagine locking the steering rack and applying torque to the steering column. Yes the pinion will rotate but it is translating along the rack and deflecting the flexure - not moving the rack itself.
je suis charlie

Brian.G
Brian.G
311
Joined: Fri Dec 10, 2010 10:52 pm
Location: Ireland

Re: F1 Steering Rack, a closer look

Post

Angus, you have the description correct yes.

As mentioned in the posts following regards motion loss, I suspect as described the weight at end of spool where it attached to flexure is a tuned mass damper, and there to cut down on any axial frequency vibrations/hysteresis that may occur between the two parts of the system since it is the only link.

The response of the system as noted will be almost instantaneous, and in reality the pinion wont move in the x direct ''much at all'' before the spool follows it up with oil delivery.

Brian,
If you think you cant, you wont, If you think you can, you will