Mercedes: Innovative Linked Rear Suspension

As we get towards the end of the season we often see teams start to get relaxed with the usual secrecy in the pitlane. This weekend in Korea was no different with several technical details being bared to the cameras for the first time.
In particular was the first picture I’ve seen of the Mercedes rear suspension, http://www.f1talks.pl/?p=11598&pid=6274  (Credit to F1Talks.pl and SuttonImages.com for the picture)
The surprise is that Mercedes appear to have adopted a hydraulic solution for managing rear roll and\or heave stiffness. Nothing is new in F1, this solution closely matches the aims of the 1995 Tyrrell Hydrolink system, which I hope to cover in detail in a future blog post. Indeed this is not even new in current F1, as several other teams already run similar and perhaps even more developed systems. But this is the first evidence I’ve had of teams interconnecting the suspension with hydraulics.
I spoke to renowned race car designer and suspension expert Andy Thorby about the use of just such a system, “I think most or all the teams are using linked hydraulic actuators on the corners.” adding “they allow you to tune the attitude change of the car under aero load, independently of corner spring rates” by altering both heave and roll stiffness.

Background
Mercedes were one of the many teams to switch to pull rod suspension for 2011, to gain the aero benefits and a lower CofG. With space at a premium at the back of an F1 car, compromises in packaging the various suspension elements need to be made. At its launch it was clear the Mercedes pull rod arrangement placed the rocker quite rearward, in comparison to other pull rod arrangements which place the rocker towards the front of the gear case. The conventional forward rocker placement, puts the heave spring and antiroll in the space at the front of the gear case, packaged around the clutch.
In Mercedes case the rocker is packaged the other side of the gear cluster, just under the gearboxes cross shaft. This leaves little room for the antiroll bar and heave spring. It does however place the rocker and torsion bars very low for the benefit of packaging, aero and CofG. Albeit these are small benefits, perhaps Mercedes choice of a short wheel base did not leave space for the suspension to be packaged around the clutch, as the gearbox length is a critical factor in wheelbase length.
So left with the lack of space to place either a mechanical heave element or a antiroll bar, it appears that Mercedes have opted to create a passive hydraulic system. This is not to be confused with any form of active suspension or the cars high pressure hydraulic systems, this system will be entirely self contained to remain within the rules on suspension design. As the system reacts only to suspension loads, it is clearly legal and there is no question of interpretation in its acceptance by the FIA.

What Mercedes have in place of a conventional anti roll bar and heave spring are hydraulics units (yellow), which probably also act as the dampers. These are connected via fluid lines (blue) to the central valve block and reservoir (red). Springing for the rear wheels is managed by the torsion bars. One end of which is conventionally located within the rocker pivot, the torsion bar then leading forward and connecting to the front of the gear case.
There is still some hardware at the top of the gearbox, which looks like it might be the mounting for an anti roll bar (ARB). But in this set up, its hard to see how the suspension rocker will act on the ARB. So Its not clear if the car started the season with a mechanical system, or whether it was designed purely with this solution in mind.
The teams early season struggle with rear tyre wear, may or may not be attributable to this system. My feeling is that other rear suspension and car layout factors have influenced the tyre problem, to a greater degree than this hydraulic solution. Although in a car that had a difficult pre-season and fundamental design problems. Getting the hydraulic suspension to work as well, may have been just another drain on resources for a team trying to recover its pace.

How it works
A cars individual wheel dampers displace hydraulic fluid as the suspension moves, creating higher pressure in one end of the damper and lower pressure in the other. To act as a damper, valves in the damper control the rate in which the fluid moves between the two chambers to create the damping effect.

In the passive hydraulic unit, the fluid is displaced not from one chamber to another, but via pipes through a valve block and into the opposite hydraulic unit. How the upper and lower chambers are interconnected left to right make the system react differently to inputs from the suspension. These being a resistance to roll or heave.

In a simplified view we can see the system working in two modes, with the fluid lines in ‘Parallel‘, where one units upper chamber connected to the opposite units upper chamber. Or, in ‘Crossover‘, where the upper chamber in one unit is connected to the lower chamber in the opposite unit.
In each mode we can see the effect of the car in roll (tilting from cornering loads) or heave (going down from aero or braking loads).

Parallel

Heave

When the car is in heave, both upper chambers create high pressure. This creates resistance between the two systems wanting to displace their fluid. This has the effect of increasing the cars heave stiffness.

Roll


When the car is rolling, the upper chamber on one side and the lower chamber on the other side create high pressure. As these chambers are now connected to the lower pressure chambers on their opposite side, the fluid displaced with little resistance. This presents no increase in the cars roll stiffness.

Crossover

Heave


When the car is in heave, both upper chambers create high pressure. As these chambers are now cross connected to the lower pressure chambers on their opposite side, the fluid is displaced with little resistance. This presents no increase in the cars heave stiffness.

Roll


When the car is rolling, the upper chamber on one side and the lower chamber on the other side create high pressure. As these chambers are cross connected to the high pressure chambers on their opposite side. This creates resistance between the two systems wanting to displace their fluid. This has the effect of increasing in the cars roll stiffness.

If a team simply want a hydraulic system to create one suspension effect, then they can rig up a basic system based on one of these patterns. However, with a valve system connecting in the centre of the pipes, then a single pair of hydraulic units and would be able to control both heave and roll stiffness. Such a system would not need external pressurisation or any control software to operate the valve block.

Development issues
However these systems are still present handicaps to development. Friction in the valve seals and the valve block, will create heat and variances in the systems response. This heat will be an enemy of the system, as it effect on the volume of fluid in the system, thus the stiffness the system provides to the suspension will alter. As a result the system will need to be a ‘constant volume’ system. Where the volume of fluid is managed depending on its rate of thermal expansion. This is probably part of the function of the small reservoir mounted to the valve block.
Equally important is the ‘installation stiffness’ of the system, that is the flexibility of any components, especially the flexible fluid lines, as this will alter the systems response.
But these and other issues related to hydraulic systems is already well understood by the teams with similar hydraulics being used both for the braking system and the high pressure electro-hydraulic control systems.

One area which presents trouble to the teams is the modelling of these systems. The design and simulation of the hydraulic element is not necessarily covered by conventional suspension and ride simulation software. I asked , Peter Harman, Technical Director of Deltatheta Ltd (http://www.deltatheta.com) about these issues. “I have advised teams on how best to simulate them“ adding “it sounds like it is a common development“. The problem is the hydraulic elements don’t fit in with conventional suspension design software. As Peter explains “Traditionally car companies have used MSC Adams for suspension modelling, and this has been adopted for ride simulation by most F1 teams, however Adams is really just a mechanical tool and doesn’t do hydraulics well“. Thus teams need to alter their approach, needing specialist add-ons and code to augment the already well established suspension development solutions.
Of course the systems will also be physically rig tested in back to back comparisons with their mechanical counterparts on the teams multi-post rigs.

Overcoming these issues with good approach to the detail design work, a hydraulic system should be able to get very close to the response of a Mechanical system. However the potential of the Hydraulic solution does offer some other benefits over purely mechanical systems.

Other possibilities
Once you have the ability to independently tailor the damping and stiffness of the heave and roll functions. The next obvious step is to control the pitch of the car. Pitch is when the car brakes or accelerates, one end of the car moves down and the other moves up. Braking creates a forward pitch, with reduced front ride height and greater rear ride height. Acceleration is the opposite situation.
As we’ve seen for the past few years controlling pitch is critical to maintaining a low front wing ride height, with out sacrificing splitter wear or excessive rear ride height (thus rear downforce).
Linking the hydraulic units\valve blocks between both front and rear axles, will allow the same resistance to pitch, as it does to heave on just one axle. This will increase the front heave stiffness, reducing forward pitch and preventing the splitter grounding excessively. This effect under braking could be further augmented with either gravitationally load sensitive valves, altering the displacement of fluid front to rear. Or similarly, a valve directly controlled by brake pressure. The former G-load system already in legal use on the individual wheel dampers and the latter solution a common fitment to motorbikes in the eighties, often termed Anti-Dive.

Summary
With Rake being ever important to the cars aero set up, such linked systems are increasingly being investigated by the teams. Indeed one team has run such a solution since mid 2009 and at least two other teams (one at each end of the grid) ran them last year.

20 thoughts on “Mercedes: Innovative Linked Rear Suspension

  1. This looks really interesting. In effect it would create something of an “intelligently” working suspension without touching anywhere near the active suspension line of thought.

    This feels like just the thing a Brawn and Schumacher might be perfecting!

  2. Nice to see a bit more of the cars now.

    I know that Prost GP had a 4 damper system created for them around 1999/2000. All 4 dampers were interconnected and it was really quite similar to this. I believe there was only one set made and delivered and was used at a test, but never raced. I certainly never saw another set pass through the factory.

  3. Perhaps Mercedes buy their kit from Penske? Penske still sell a similar system.(It’s on Penske’s website under “Special Products”.

    Excellent blog, by the. way, Mr Scarborough. The last 2 entries have been superb.

  4. Pingback: Whitmarsh says teams will decide future of FOTA | F1 Fanatic – The Formula 1 Blog

  5. Doesn’t the suspension on the MacLaren MP4-12C operate in much the same way, using hydraulics to control pitch fore and aft and roll?

  6. Reminds me of the hydropneumatic suspension of the Citroen of the 50’s. Also very similar system to the self leveling suspension of the Mercedes W123-W140 series. Maybe a clever adaptation of it?

  7. There are better ( simpler ) ways to manage the roll and ride level at the same time. Simple solutions are always the best but most difficult to crack…
    Compliments to Mr. Scarborough for giving us so much insight on complexed F1 technical issues.

  8. Scarbs,

    I did not quite understand this :- “However, with a valve system connecting in the centre of the pipes, then a single pair of hydraulic units and would be able to control both heave and roll stiffness.”

    Can you shed some light on what you think the valving would be inside to make this ‘red’ reservoir? How I understand this is if there is no external tuning and no active system then the valving will be such that it is an intermediate of the ‘cross’ and ‘parallel’ configuration. In this case it will be a trade-off between the the heave and roll stiffness and the roll and heave stiffness will not be (almost) decoupled as in the case of the ARBs. Packaging seems to be the only real advantage. I am sure if all teams are doing this, there is more to this.

  9. The first time I saw a similar system on a race car was back in 2008, used by the University of Western Australia FSAE team. In fact it was almost identical to the one described, using pullrod suspension with rocker actuated interconnected Kinetic dampers and torsion bars. It even utilizes single-piece carbon fiber lower a-arms (acting as an additional springing element on pitch/heave and thus increasing pitch/heave effective spring rate) and smart driver adjustable anti-roll bar. More info on the system can be found on ATZ extra 2009 or on UWA’s website (http://motorsport.mech.uwa.edu.au/news.html)

  10. I remember Ron Dennis having a fit over John Barnards fixed A- frame suspension on the Ferrari, boy how things have changed. Nice work Mr Scarbs, and as you point out the trick of the system is how to manage the massive change in suspension frequency, lateral gs, tons of down force and maintain a eighth of a inch of ride height,plus keep it cool. And I’m sure it doesn’t change rake right? must have one super hydro pump!
    Ever get any pictures of the Red bull coming out of a corner at plus 100mph about 30 to 50 feet after the apex. I really wonder if that rear goes down. That seems to be the big difference in Vettle over Webber and Hamilton over Button.We don’t get much info in the states, not like Europe and England.The Red Bulls rears coming into a corner is very soft, even from side to side. I believe that rear goes down under hard acceleration, But can’t find the Pics!
    Thanks for your work, I don’t drink before I write now!

  11. Pingback: Lotus Renault GP: Fluid Inerter | Scarbsf1's Blog

  12. Mr Scarborough,
    I am fascinated by the tech and dynamics involved in the F1 car. Thanks for your insight and blog work! HOWEVER…
    I believe your illustration at the top of this page is incorrect. If “Heave” is the upward movement of the chassis as the arrow indicates, then the pullrods will move in the OTHER direction as the wheels move downward relative to the chassis.

    Also, Do any of the teams use Delco’s Magnetic reacting Fluid tech in their various damping systems?

    • I see what you mean, but Heave is merely a vertical direction, bump or droop are the specific directions of heave.

      Magna-rheological dampers are effectively prohibited in the rules.

      • Due to active dampers being illegal?

        Having a magnet moving around mechanically on the outside of the damper tube should be doable.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s