The Rotary damper has been an innovation that has recently come and gone in F1. Typically F1 cars use Lineartelescopic dampers. However, in 2003 Sachs Race Engineering developed a Rotary version of a classic monotube damper for Ferrari. This was a tidier solution for packaging the dampers at the rear of the car. Since 2003 several teams have used Rotary dampers, Toyota, Midland and most notably Brawn GP. Who won the 2009 championship with these fitted at the rear of the car. I have recently come across a Rotary damper as fitted to the Midland car in 2006 and thus we have the chance to look into the detail of this development. I’ve also been in touch with two of the designers who have raced this device and they have given us some unique insight into its use.
Firstly John McQuilliam, who is now Designer for Marussia, but back in 2006, was the designer for Midland F1 (previously Jordan, latterly Spyker and Force India). The Midland M16 was the first of the teams’ cars to exploit Rotary dampers, as McQuilliam explains “I remember we were working with Sachs, as we had their Clutches on the car at the time. The Ferrari had Rotary dampers and it looked to be such a neat installation, we wanted to do something similar “. But this was not a total success, as their use on the Midland was short lived as they were replaced mid season with Linear dampers.
Secondly we have insight from the ex-Brawn Designer Jorg Zander, now running his own consultancy (http://jz-engineering.com/). As Honda were leaving F1 and the Brawn GP team was created, due to the Honda link the team had been using Japanese Showa dampers, but Zander was persuaded to switch to a Rotary design. As Zander recalls “Showa did a brilliant job and they would have provided us with continued support, but my boss was keen to go down the Rotary damper path with Sachs”. He adds that “the car still won the championship!”. Along with the Ferrari wins during their use of the Sachs Rotary damper, it’s clear that the technology has had a lot of success in F1. But it is still blighted by a reputation of being a troublesome technology.
A conventional Linear damper is made up from a cylindrical body in which a damper rod slides in and out of (much like a bicycle pump). A valve at the end of the damper rod passes through the inside of the cylinder which is filled damper oil and the oil passing through the vales controls the rate of movement of the damper. Being cylindrical it is easy to machine in a lathe and oil is kept in place with simple circular seals.
The Dampers are installed on the car by one end being secured to the gearbox case, the other end to a rocker which re-orientates the movement from the pushrod into the movement of the damper. The rocker can be made up of several levers splined to a single shaft and requires several bearings to make the suspension movement friction free. This amasses into quite a lot of separate components which all need space to fit in.
Rotary Damper Anatomy
Essentially the Rotary damper rotates the axial movement of the Linear damper into a Rotary movement. All the components have the same function, they just move in a different orientation. Thus the damper body is deep “keyhole” shape and the damper rod (known as the vane) rotates in the body, the vanes arm sweeping through a small angular rotation in the keystone section of damper body. The damping valves are placed in the face of the damper vane and work in exactly the same manner as a Linear damper.
Obviously the shape is the biggest difference and manufacturing the damper is far harder due to the fact that the parts have to be CNC milled and not spun in a lathe.
We can see the damper body is a robust metal housing machined from solid. Open on one side to allow the damper vane to be installed, there is a closing plate sealed with an “O” ring and 13 small bolts. To allow the damper vane to pivot the body features two bearings and seal arrangements, one on the closing plate the other is the main housing.
As the housing also acts the suspension rocker, there are also elements machined into its outer face to accommodate this function. Firstly the housing acts as the rocker linkage, so a rocker arm is part of the machined shape. One of these eyes in the rocker will mount the heave damper and the other eye has bearing to accept the pushrod. Showing the packaging efficiency of the Rotary damper a further spherical bearing is fitted to a machined section of the cover plate. This accepts the Anti Roll Bar linkage. Thus for the Midland, no further suspension elements need to be fitted to anything but the damper housing. In the case of the all F1 Rotary dampers, the damper body rotates and the damper rod is fixed to the gearbox (via splines). As one part moves and the remains static, the vane moves through the oil filled cavity, the reaction force of these two parts creates the damping effect. To allow the damper body to rotate in the gearbox, two bearing surfaces are machined into the outer face of the housing and cover plate.
One additional external feature is machined into the damper body, a single damper valve. I believe this is a valve to compensate for the effect of heat on the volume of oil inside the casing. The valve offsets the thermal expansion of the oil to ensure there is a constant volume of oil within the damper cavity.
Inside the damper body sits the damper vane. This is a highly finished and possibly DLC coated component. Even when degreased this part had the feel of a lightly oiled component. The friction reducing coating being there to reduce the friction created by the vane moving inside the housing. Again machined (as most F1 parts are) from solid, the damper vane is formed of two shapes.
Firstly, the spindle that sits in the bearings that allows the arm to the rotate. One end of this spindle has splines machined into it.
Then secondly a flat plate shape is formed into the vane, this is the equivalent of the Linear dampers damping rod. This arm needs to be a close fit to the damper cavity in order to accurately control oil for the damping effect. It’s the edges of this arm that need to be sealed against the housing. A single square edged seal is fitted into a machined groove around the periphery of the vane. There are also four friction reducing pads (two on each side) to aid the movement of the vane against the body.
Providing the damping effect the damper valve is a simply circular shim stack arrangement fitted to hole machined into the vanes face. This valve set up is almost identical to the set up used on the end of a Linear damper. This is perhaps the only aspect of the Rotary damper that directly echoes a Linear set up.
As already alluded to, the Rotary damper is fitted to the gearbox casing and forms both the damper and the rocker. The damper vane slides into splines machined into the gearbox casing and a bearing locates on the bearing surface of the cover plate. Then another bracket fits to the rear of the gearbox to locate the rear bearing and secure the damper in place. The torsion bar passes through the damper engaging in splined in the spindle of the damper vane and also on the front face of the gearbox. Splines on the protruding section of the damper body are probably for the preload adjuster arm.
The pushrod passing up from the lower wishbone fits to the rocker, as does the heave springdamper which sits over the top of the damper to attach to the other side damper. One end of the anti roll bar attaches to each spherical bearing and then the installation is complete.
Pros and Cons of Rotary Dampers
As explained the main benefit is the packaging of these units. Typically Linear dampers are operated by rockers and the dampers are then either laid across the top of the gearbox or sit vertically (requiring a recess in the top of the gearbox case). Either option carries complications in the cars structure or aero. With a Rotary damper the unit forms both the rocker and the damper and takes up far less volume.
As the damper vane permanently sits inside the casing, as the vane sweeps through the radial movement, no oil displaced. This is described a constant volume system. Unlike a monotube damper where the damper rod displaces fluid inside the damper body. This requires a method to offset the movement of the excess fluid. Typically separate nitrogen charged cylinder is used, the gas is compressed by the displaced fluid. But this in itself creates a small spring effect inside the damper. Other methods include the though rod damper, whereby the damper rod passes through both ends of the damper body, thus displacing no fluid. However through rod dampers do require additional seals and this creates some additional friction in the design.
Other key benefits of the simpler Rotary design are weight reduction, with fewer parts the 950g damper with its integrated rocker is lighter than a conventional damper and separate rocker set up. John McQuilliam confirms “We did achieve a weight saving over the conventional layout when you consider the damper, its drive arm and reaction bracket.” McQuilliam goes to on to highlight its packaging and resultant aero benefits “also the packaging is easier, without finding a volume for the damper. Akio Haga who is now alternating chief designer at Force India was laying out the rear suspension back then and we had a few different lay outs, mainly to try and keep the mechanicals out of which ever area the Aero was telling us was most important”.
With there being less parts to be splined together and less bearings in the Rotary design is also stiffer and suffers much less from slop. Jorg Zander explains “the good side of the Rotary damper is that because of the rocker integration, the system is very stiff and direct, so there is little losses due to backlash in linkages, ball joints, etc. which meant it had a good high frequency response”.
So with the damper being a constant volume design, structurally stiff, lightweight and easy to package why has it not become the norm for F1 suspensions?
There is a simplistic argument bandied about that the damping effect is the reason they do not work so well. Neither designer suggested this was the case to me, although at the time McQuilliam did suggest to me it was a factor, but corrected himself with hindsight “I think the damping was actually reasonable”. McQuilliam continues to highlight a more specific deficiency in the Rotary design, “there is a much more complicated sealing arrangement in the damper, adding stiction”. Stiction is one of the enemies of the suspension designer, a mix of terms meaning “sticky” “friction”. This is seen as initial friction, then smoother running. This non Linearity of response is hard to design out. Where as the inherent spring rate added by a gas charged damper can be considered in the overall suspension spring rate, stiction cannot. Zander also echoes this issue “However, the downside is that due to the high internal pressure, the hydraulic seals had to have reasonable preload, which induced a large amount of friction and hysteresis”. This internal pressure also lead to structural issues, the Ex Brawn designer tells me “the high internal pressures caused local deformation of the housing, as well as the vane, this lead to increased friction and pretty inconsistent damping characteristics. Initial developments started with Aluminium, then Ti and steel options to gain stiffness to reduce the issues caused due to deformation. The Midland Damper that I have appears to be titanium. This was an earlier design compared to the Sachs dampers run on the Brawn in 2009. It is possible to see the ribs machined in to the casing to reinforce the device from deformation.
Another argument quoted in the press is the small size of the damper, which results in a lack of oil or radial movement. I removed some 75cc of fluid from the damper and the damping chamber was quite large. The limited radial movement was not seen to be a major problem, being taken into account in the rocker sizing, although both designers point out it still had to be factored in. Zander explains “The angular displacement wasn’t so much of an issue in 2009 and in previous years, but with the heavier fuel loads from 2010 that should be something to be considered.” Seeing as the 2009 Brawn BGP001 was succeeded by the Mercedes W01, it’s interesting to note the latter went back to Linear dampers.
Away from the technical argument of Rotary over Linear there was one other factor which perhaps underlines the less publicised aspect of the designers’ role, budgets! But when we recall that Midland and Brawn were both teams managing a tight budget, this last issue makes sense. Both designers highlighted price as one of the major issues of the Rotary damper. McQuilliam starting by saying “They were mighty expensive, so not good value for money for the weight saving”. Zanders more recent recollection of the Brawn days provides this insight “I think a bigger argument against it for some teams, were the cost of such a Rotary damper. Depending on the specification, it was in the region of about €15.000 per piece”. With these being sealed dampers for each set up, a pair of dampers (€30,000) would be on the car and a multitude of other dampers pairs in the pits, each set up for different damping characteristics.
Clearly the stiction and internal stiffness issues need to be addressed with the design. Evolution via detail design has overcome similar issues with Linear dampers, so presumably the same could be resolved for Rotary dampers too.
The cost issue still remains; inherently they are a complex and high precision part. Where as turning on a lathe produces the correct finish for a Linear damper, careful milling operations are required for the damper body of the Rotary damper, which will inevitably make this an expensive part to produce. No doubt more teams using the damper would drive the price down.
It’s interesting to note that the Sachs Race Engineering website no longer details Rotary dampers as part of their range. Instead conventional Linear dampers and a Through-Rod version of the same, form the basis of their product range and their current Formula1 teams use these Linear format dampers.
Given the choice between Rotary and Linear, Zander sums up the decision well “I would at the current stage of damper technology development prefer Linear dampers over Rotary ones. The friction can be controlled in a better way and problems like cavitation are well understood and do not cause any issues in contemporary Linear damper designs. Also the flexibility with Linear damper designs is much wider, considering systems like: frequency depending damping, high & low speed damping, drop off characteristics (blow by valves, pressure relief valves)”.
Perhaps the Rotary damper will be explored in F1 again at some point, but for now the Linear damper is the sole solution in F1.