76 Intake and management The XMD-250 is naturally aspirated with throttle-body injection, the fuel and air mixing conventionally as they travel down a tuned intake manifold about 9-12 in (22.8-30.5 cm) long, guiding them to the stator inlet and transfer port. Injection quantities and timing, along with the throttle servo angle, are mapped and governed using an off-the-shelf Walbro EFI/ECU combination unit. “Dynamometer testing uses a straight intake pipe on the engine. However, we’d naturally design something different for a field-use or flight-testing model for better integration, tuning and mixing,” Wilkinson says. “The engine-fuelling configuration can be customised to the design of the UAV platform. Typically, there’s a gravity flow of fuel to the EFI fuel pump, which then supplies fuel to the injector pump. The whole system runs on a 12 V power source. “And while the Walbro unit has been great for testing purposes, for UAV applications there are ECU and EFI solutions that Northwest UAV (NWUAV) can implement on our behalf – they’ve been of invaluable help in their advice and support throughout our r&d.” Avadi’s longer-term plan is to stay agnostic to ECUs and EFI controllers (porting the control software and mapping from one system to the next) to enable the adoption of new, more optimised ECUs or those preferred by a UAV developer. Combustion matters The combustion chamber has an unconventional shape due to the angled, tubular transfer port. “Normally, you want your spark plug right at the centre of the combustion chamber, so the flame can propagate out to all the edges of the combustion chamber as smoothly as possible, but because the transfer port inescapably forms part of our combustion chamber, we get a very asymmetric chamber geometry,” Dardalis says. “What we’ve found, however, is that as the piston closes on TDC, that squishing effect forces the gas to rapidly compress inside the transfer port – as it has nowhere else to go – and that creates high turbulence inside the port.” The spark plug is positioned centrally in the cylinder head. This substantially offsets it from the centre of the transfer port, but it cannot be offset from the centre of the cylinder head to position it more symmetrically (relative to the transfer port’s shape) due to how the cylinder head rotates, and the wiring and integration problems this would cause. Avadi, however, sees potential for CFD optimisation of the transfer port geometry for improved squish, tumble and combustion quality. “We might, for instance, add more volume in the head towards the rear of the port and still be able to hit our high compression ratio. The port design is only in its first draft now, so we anticipate many more versions of it, and the XMD-250 to follow,” Dardalis says. Wilkinson adds: “The piston, of course, could also play a part in that optimisation: a plug or other convex shape atop the crown could, for instance, protrude into the transfer port at TDC for higher compression. We had such a design feature in the MA-250’s piston; we just haven’t tried it in the XMD-250 yet.” Aside from all of this, the combustion and power stroke are conventional, with no direct efficiency relationship between the thermodynamics of the engine and its mechanical design. “Essentially, the combustion doesn’t know that an unusual crank system exists, but it could integrate efficiency improvements over time; for instance, the double connecting rod could one day use a very low connecting rod-to-stroke ratio, assisting with turbulence at high or low speeds,” Dardalis says. The intake and exhaust ports are offset at roughly 250° (crank angle degrees) from each other, with about 110° between the exhaust and intake, as the width of the ports (about 1 in, or 2.54 cm) means gases start flowing through long before the transfer port perfectly lines up with the inlet or outlet. Again, however, Wilkinson and Dardalis emphasise that this first design of the stator and transfer port was based on calculated assumptions of flow coefficients, with real-world CFD testing of the XMD-250 prototypes and benchmarking of the actual flow coefficients to inevitably refine the next versions’ geometries. “We don’t want to wait for the CFD to be over before we build prototypes. We want to develop as many of the concepts as we can, so the next version can have power efficiency and generation improvements across the board,” Dardalis adds. On power generation, Wilkinson notes that the flywheels (which primarily serve to smooth power and vibration from the piston, as conventional) present a strong opportunity not only for efficiency optimisations but also for capturing the engine’s power output as electricity by replacing the current steel flywheels with permanent magnet rotors running in copper-wire stators. He comments again that ePropelled’s motor-generators hold great interest for August/September 2024 | Uncrewed Systems Technology As the piston closes on TDC, that squishing effect forces the gas to rapidly compress inside the transfer port, and creates high turbulence
RkJQdWJsaXNoZXIy MjI2Mzk4