A dual purpose feature to reduce both drag and downforce at high speed in the simplest, lightest way possible
Post-publishing note: This concept was eventually abandoned for the 2024 undertray in favour of instead adding additional structural stiffening to account for the increased downforce magnitudes. By having the flap mounted on a single pivot, torsional stiffness of the undertray was significantly compromised to the point where the endplate flexed enough to noticeably adjust the friction on the flap, making DRS activation imprecise. Shorter-than-planned testing time also meant the flap activation speed and overall performance benefits could not be confirmed prior to the competition. It still remains an interesting concept, but is perhaps better suited to a front wing due to the naturally increased stiffness and potentially iarger performance benefit. The ability to manually mount the flap in a single minimum-drag position was maintained for use in the acceleration event.
Following on from the previous post where I mentioned the high drag sensitivity, I started coming up with methods to reduce the drag of the floor design further. Despite the very draggy appearance of the large vertical wall in front of the rear wheel, the least efficient part of the whole floor is the diffuser flap, thanks to its large frontal area. The brow wing is also very inefficient (relatively speaking) given its high angle of attack, short span, and lack of endplates, however these features all serve a purpose and making any changes here would be structurally difficult. Given the diffuser flap is effectively a second element on a multi-element wing, the most obvious answer was to implement a typical drag reduction system approach of actively reducing its angle of attack. I didn't really want to do this with an actuator though; these are unavoidably quite heavy and would be difficult to mount in a way that wouldn't significantly disrupt otherwise clean flow. Enter the ongoing controversial "flexi" wings in Formula 1.
There's no rules about flexible wings in FSAE other than a maximum deflection allowance for static loading, where the intention is clearly stated as ensuring wing structures won't break or fly off at speed. I decided to look into it and see if something like this was feasible. Obviously at the speeds achieved by FSAE cars, the wing elements themselves and any decent mounting isn't going to be flexing. It probably wouldn't pass the rule I just mentioned anyway if it did. The alternative was to have a spring-loaded wing, where the mounting and aerofoil are rigid, but rotation is free within a limited angle range. All this theoretically required in addition to a typical mount design is a spring, and one additional bolt. Seems a lot lighter than a servo or actuator.
The downside of this method is you have much less control over when the DRS activates. If set up properly, it will activate above a certain speed, and deactivate below that speed, and no driver or computer input will be able to alter this behaviour. We can work with this though; we know approximately what the fastest grip-limited corner will be on the competition track, so the activation speed should simply be marginally faster than this. There are also the implications of failure of the system (what if the spring falls off, or a bolt gets pulled out?). Because it only works if the spring holds the DRS closed against the aerodynamic force, it will fail "unsafe". This is ok for a small wing element positioned between the axles like this; a small loss of downforce will simply push a well-balanced car towards the outside of the corner. This "passive" DRS is likely not suitable for a main rear wing element though, where in the event of failure, a large loss of downforce heavily biased towards one axle could create a scenario where the driver loses control. Given the system is driven by aerodynamic forces though, it follows that any loss in downforce upon activiation will likely be minimal (and in fact our design does not lose total force, but rather holds approximately constant as the flap opens and velocity increases). The benefit of having it fail unsafe is we can safely assume the wing will always be in the open position above a certain speed (plus a bit of tolerance, and with a pre-drive check for free rotation). That means our maximum downforce magnitude is limited, so we don't need as much stiffness in the floor or strength in its mounts, saving weight and material and giving a little peace of mind when hurtling down a straight. The difference for us at Vmax of 120 kph is 71 kg vs 53 kg, giving a 9 kg (25%) reduction per side.
The general idea of how the spring is set up is shown in Figure 1, which is viewed from the outside of the car looking in at the endplate, behind which sits the diffuser flap.
Figure 1: Spring setup for the diffuser flap passive DRS. The flap pivots around the pin mounted in the bearing, and the spring is mounted between a fixed point on the endplate and a fixed point on the flap, which pivots through a slot in the endplate to define maximum angle limits.
I made up a Matlab program to find the optimal mount position and spring preload for a range of available spring rates of tension springs we had lying around the office. The target was to achieve activation at the predetermined speed (mentioned above), and to have it activate as rapidly as possible to maximise the benefit. Figure 1 shows the geometry of the best result, utilising a decreasing moment advantage for the spring as the aerodynamic moment on the flap decreases with activation. The expected activation behaviour is shown in Figure 2, and an example force balance between the spring and aerodynamic moment at the identified critical speed is shown in Figure 3 (we decided to cap rotation at 40 degrees to reduce the maximum force applied to the spring mounting for part longevity and reduced mass, given the drag and downforce difference between 40 and 50 is negligible). The fully closed and open states are shown in Figure 4. In reality we expect the critical speed might be slightly lower based on drivability, driver confidence, and reduced aerodynamic forces compared to CFD estimates, but we have various springs we can change out with slightly different stiffness to fine-tune activation at track. The design will also have the ability to be locked in the fully-open position for the acceleration event by adjusting the spring mount position to a different lower bolt, reversing the force direction. Track testing to ensure the activation behaviour is correct will be performed using a potentiometer mounted to the endplate.
Figure 2: Diffuser flap passive DRS activation angle prediction using values output from CFD with the spring parameters given by Matlab
Figure 3: Spring and aerodynamic (fluid-structure interaction) forces at the identified critical speed of 23 ms-1
Figure 4: Fully (left) closed and (right) open (40 degrees) position of the diffuser flap
Track testing is also critical to ensure both flaps open at the same velocity. It's not critical that both open perfectly in synchronisation, as the idea of the activation speed is that the car never enters a grip-limited scenario above this speed, so the worst that would happen with asymmetric activation is a slight yaw moment is applied to the vehicle; an advantage of having the DRS system as close to the centreline of the car as possible.