Formula Student is Europe’s most established educational engineering competition and presents some unique technical challenges. Student teams worldwide design, build, test, and race small-scale formula-style racing cars and compete in static and dynamic events. The static events include engineering design, cost & manufacturing, business presentation and technical inspection. The deployment of the Formula Student car is in dynamic events, including figure 8 skidpad, 1 km autocross/sprint, 75 m acceleration and a 22 km endurance where judges also consider energy consumption.
In the 2021 Formula Student event, the University of Glasgow Racing team, UGRacing, posted its best-ever result with an eighth-place overall finish while retook the Scottish Formula Student crown. It also won the Engineering Design event for 1st Year Vehicles and the Racecar Engineering Engagement, Outreach and Communications Award.
One of the most significant developments of the UGR20 was that to its Honda CBR600 inline four-cylinder engine. Neil Dawson, Powertrain Manager at UGRacing, explains how UGRacing improved the engine’s drivability, eliminating the severe torque dip between 4-6krpm seen in UGR19’s engine (see Figure 1), thus allowing the powertrain to produce a predictable torque curve for the amateur driver to utilise.
‘Analysis of data from the 2018 endurance run showed that we spend >80% of an endurance lap at 4000-8000rpm,’ says Dawson. ‘Therefore, UGR19s torque curve is poorly tuned for Formula Student events. This correlates with driver feedback complaining of the engine “bogging down” coming out of slow-speed corners.’ This can be seen in Figure 2.
Through 1D simulation and empirical investigation, the team found reverse flow through the inlet port during the valve overlap period at lower engine speeds was the cause of this torque dip. ‘As the engine is port injection, the fuel cannot make it to the cylinder, and the engine runs lean in this torque dip, effectively burning exhaust gasses in the cylinder,’ notes Dawson. ‘The 20mm restrictor imposed by the formula student rules likely caused this, which results in lower inlet port pressures than those seen on the standard Honda CBR600 engine.’
The stock Honda camshafts are tuned for engine speeds above 12,000rpm (peak power @ 14,000) and therefore have large lifts and overlap periods to promote power production at higher rev ranges. UGRacing established a custom camshaft profile to reduce the valve overlap and produce a torque curve more suited to Formula Student competition. This, combined with more conventional methods of formula student powertrain design (exhaust length tuning, plenum runner tuning etc.) and internal engine modifications (higher compression pistons, superfinishing of gears, fitting a more extended first gear), mean that the team could produce an engine with better power at lower rev ranges, without completely sacrificing peak power.
The team developed a 1D model of its UGR19 powertrain in GT Suite to inform design decisions taken to establish its UGR20 powertrain. Members took physical measurements from the UGR19 powertrain to build the 1D model. ‘We validated the model against the UGR19 dyno data to validate its accuracy,’ says Dawson. ‘The 1D model also incorporated the ECU’s fuel and spark timing maps to allow the team to investigate potential map changes ahead of time to save valuable dyno time.’
The engineering team then used the Design of Experiments (DOE) feature within GT Suite to optimise the lift profiles of the camshafts thermodynamically. For the UGR20 camshafts, the DOE solver was instructed to maximise the area under the torque curve between 2500-11,000 rpm. The DOE solver was allowed to modify the lift, timing and duration of both the inlet and exhaust valves, as shown in figure 3 below. Note that multipliers of the stock camshaft controlled the durations and lifts.
The DOE simulation took those parameters and cycled through a total of ~400 camshaft designs to provide the team with the thermodynamically optimised camshaft. Compared to stock, this produced an exhaust camshaft with reduced duration, reduced lift, and a 6° advance. The intake camshaft also had a reduced duration, but the maximum lift and timing were the same as the stock camshaft. The new “thermodynamically optimised” camshafts produced a higher torque across the entire rev range with a smooth torque curve.
However, Dawson highlights that the thermodynamically optimised shapes were very aggressive, so the team conducted a mechanical analysis of the profiles to ensure the mechanical reliability of the engine. For this mechanical analysis, VTDesign was used to modify the new valve lift profiles found by the thermodynamic studies. ‘For this, we used the measured engine data to build a kinematic model of the valvetrain and analyse the forces acting on the valves and valve to piston clearances,’ says Dawson. ‘For valve velocity, we compared acceleration and jerk to the stock camshaft—the lift profiles of the new camshaft modified to bring these in line with the stock engine. We also monitored the valves’ spring rates and closure rate to avoid valve bounce and float.’ The shape of the inlet and outlet camshafts can be seen in Figure 4.
These profiles were then machined from a set of stock CBR camshafts by Newman Cams and fitted to the UGR20 engine. In addition to the camshaft work, the team also reduced parasitic losses on the engine by “knife-edging” the crankshaft, superfinishing the gears in the gearbox and removing the stock mechanical water pump in favour of an electrical one. The 1D engine model also tuned the primary exhaust lengths, plenum volume, and intake runner lengths. All of this resulted in a torque curve that was much smoother than the one seen on UGR19. Driver feedback was that the new torque curve meant that the car was less “twitchy” coming out of slow-speed corners as the power was instant and predictable, rather than a delay before a lump of power came in (as it was on UGR19). A comparison of UGR19’s dyno data to UGR20’s is in figure 5 below.
