The Scaling of Loss Pathways and Heat Transfer in Small Scale Internal Combustion Engines

Understanding the performance parameters of small remotely piloted aircraft powerplants.

The rapid expansion of the remotely piloted aircraft market includes an interest in 10 kg to 25 kg vehicles (Group 2) for monitoring, surveillance, and reconnaissance. Power plant options for those aircraft are often 10 cm3 to 100 cm3 displacement internal combustion engines. Both power and fuel conversion efficiency decrease increasingly rapidly in the aforementioned size range, with fuel conversion efficiency falling from approximately 30% for automotive and larger scale engines (greater than 100 cm3 displacement) to less than 5% for micro glow fuel engines (less than 10 cm3 displacement).

Two-stroke engine cycle (reprinted from Heywood with permission from McGraw Hill Education)

Based on the literature, it was unclear which loss mechanisms were responsible for the increasing rate of decreasing performance. Moreover, predictive models for losses such as friction, heat transfer, and short-circuiting (scavenging) were unavailable for ICEs in the stated size range. Previous research also indicated that these losses could cause an inherent relaxation in an engine's fuel octane requirement, possibly allowing small ICEs below a certain size to be converted from gasoline to JP-8 or diesel with little to no modification. To investigate these issues, three research objectives were proposed addressing the scaling of loss pathways; the modelling of heat transfer, friction losses, and gas exchange; and the determination of fuel anti-knock requirements for engines in the 10 cm3 to 100 cm3 displacement size range.

Objective 1 addressed the relative importance of loss pathways. The Small Engine Research Bench was constructed to measure the energy pathways in a family of geometrically similar, commercially available engines with 28 cm3, 55 cm3, and 85 cm3 displacements. At peak power, the loss pathways ranked by relative magnitude were: short-circuiting (40%-50%), sensible exhaust enthalpy (14%-18%), incomplete combustion (11%-17%), brake power (14%-15%), and heat loss from the cylinder (10%-11%). The measured losses were compared to balances for large and smaller engines, showing that the primary losses driving efficiency in 10 cm3 to 100 cm3 displacement two-stroke engines is short-circuiting and incomplete combustion and that thermal losses do not begin to increase substantially until engine displacement decreases below 10 cm3. The objective concluded with a parametric study describing the interactions between engine operating parameters, engine performance metrics, and the loss pathways that identified ways to improve engine performance over current commercial-off-the-shelf configurations.

Objective 2 examined models for heat transfer, friction, and gas exchange. Data from the engines were fit to Taylor and Toong's spatially and temporally averaged heat flux model. Compared to fits developed for larger compression-ignition and spark-ignition engines, the engines tested herein had substantially lower values for the Reynold's number exponent, 0.3-0.4 instead of 0.7-0.75 as was common in the literature. The results showed that the Taylor and Toong model was not formulated to handle the observed variations in throttle, speed, and cooling flow rate and was highly sensitive to the trapped mass used to calculate in-cylinder temperature. A model for engine friction was developed that predicts friction for similar two-stroke engines less than 100 cm3 displacement using the surface area to volume ratio, engine rotational speed, and delivery ratio. Engine performance data was compared to the perfect isothermal mixing and perfect isothermal displacement models and indicates that the perfect isothermal mixing model using the trapped volume to calculate delivery ratio predicts engine scavenging performance within ±5% (absolute) at most points and ±10% (absolute) over the whole operational range.

Objective 3 investigated the conversion of 10 cm3 to 100 cm3 displacement two-stroke, SI engines to low anti-knock index fuel. The knock limit was defined as the more conservative of 5 bar maximum amplitude of pressure oscillations or 5 bar/deg peak pressure rise rate for 1% of 400 consecutive cycles. Knock was strongly dependent on rotational speed. Generally, once knock occurred, the knock-limited speed range increased by 500 rpm for every 10 octane number (ON) decrease. The results also showed a dependence on engine size, with the 55 cm3 and 85 cm3 engines being knock-limited about 500 rpm and 1000 rpm faster (or 10 ON and 20 ON higher) than the 28 cm3 engine, respectively. Switching from 98 ON fuel (manufacturer recommended) to 20 ON (JP-8, diesel equivalent) fuel led to a 2%-3% increase in power at non-knock-limited conditions. Changing from 98 ON fuel to 20 ON fuel improved fuel conversion efficiency by 0.5%-1%, which translates to an approximately 6% increase in range or endurance. Of the three parameters investigated, combustion phasing had the greatest potential for controlling/ increasing knock-limited power, and offered the most control authority relative to its impact on fuel conversion efficiency.

This work was done by Joseph K. Ausserer, Captain, USAF, of the Air Force Institute of Technology for the Air Force Research Laboratory. AFRL-0251

This Brief includes a Technical Support Package (TSP).
Document cover

(reference AFRL-0251) is currently available for download from the TSP library.

Don't have an account? Sign up here.