To change gear as fast as the gearbox will allow, every time, without wearing the transmission or triggering false shifts.
A pneumatic system running CO2 at 100-110 psi from a 1000 g bottle through a pressure regulator. Three solenoids feed two actuators: a dual-acting combined up/down shift actuator and a single-acting clutch actuator. Control is handled through the MoTeC M150's GPRP feature set. From CR26I the system is closed loop. A Corsa-Technic performance linear potentiometer on the shift lever reports actuator position. On an upshift request the actuator is preloaded for 100 ms, which alone saves over 150 ms of shift time, then torque reduction begins; power is restored when the shift position target is reached or the shift time upper limit is exceeded. On a downshift the same preload is applied (saving over 100 ms) and drag torque reduction is used instead, which improves shift reliability and reduces harshness.
Time the power cut to the mechanism, not to a fixed number. Closed loop lets the timing vary with gearbox temperature, traction state and driveline load, so the torque interruption is minimised for every shift regardless of condition, and dog engagement is properly timed. That is what protects the transmission. Do not build a custom shift drum. The recorded recommendation is explicit: tolerances have to be incredibly precise and stay precise through service life, and anything out of tolerance will cause a catastrophic gearbox failure. Auburn's 7075 aluminium shift drum failure at Lincoln 2019 is the cited example. Electric shifting has been evaluated and rejected. Shift time fell from 300 ms to 55 ms on the pneumatic system, reliability is good, and the advantages once claimed for electric (no tank refills, no burst lines, variable force, better logic, encoder feedback) have all been addressed within the pneumatic architecture. 1. Route every line and fitting in SolidWorks. Lines cannot be within 1 in of the engine casing or 3 in of the exhaust; bend radius is 4 in. 2. Size the cylinder bore and line inner diameter with full calculations rather than by picking a catalogue part. 3. Fit the linear potentiometer and set the position bounds that define a completed shift. 4. Tune the preload duration and the torque reduction window on track. 5. Measure the result. CR26I averaged 65 ms cut time across endurance closed loop against 85 ms open loop; across 180 upshifts that is 9 seconds under power cut instead of 15.3 seconds.
- Improve overall system serviceability. - Completely remove the possibility of falsely triggered and failed shift events. - Reduce transmission wear. - Further reduce shift times. Target reliability is 99.9% shift-to-execution. - Reduce weight where possible without losing the closed-loop capability.
- Pneumatic lines bursting. This stopped once the CO2 tank was mounted in the only orientation allowed for CO2 bottles. - CO2 tank refills. Less of an ordeal since the team gained the ability to fill in house. - Gear position unavailable at a standstill. Gear is currently calculated from transmission output shaft speed and engine speed, so with the car stationary the driver has no gear reading. Candidate fixes recorded: resistors on the shift drum using the neutral switch as a rotary potentiometer, or incrementing gear from confirmed shift completions on the linear potentiometer with a reset to neutral from the neutral light. - 1-2 shifts are unlikely to go below 85 ms because of the extended lever travel between first and second.
SolidWorks for line and fitting routing. MoTeC M1 Build for the GPRP shift logic and M1 Tune for calibration. MoTeC i2 Pro for shift time, actuator position, cut percentage and engine speed traces. A test bench is called for before any architecture change.