Propellers Design - LUKY Software - Moment of Inertia - Performances - Change of propeller?

Propeller testing is essential to validate calculations and models, and to compare performance with other propellers. Every E-Props model is tested on the ground to confirm its performance, mechanical strength and compatibility with the intended engines, before any flight testing begins.

Ground test facilities

Terminator — laser-scanning dimensional inspection

Every blade and component is checked with the Terminator laser-scanning system: high-precision laser sensors measure the actual surface geometry of the part and compare it, point by point, with its digital reference model. This 3D control is applied systematically, before and after mechanical testing, to confirm that no dimensional drift has occurred during fatigue, vibration or impact testing.

RESONATOR — torsional vibration test bench

For the development of propellers intended for direct-drive engines, E-Props built a dedicated vibration test bench able to reproduce, with high fidelity, the vibratory loads generated by aeronautical piston engines — in particular large 4- and 6-cylinder units. The bench combines a torsion bar, a flywheel, an electric servo-motor and a high-precision digital encoder, reproducing the vibratory component of engine torque up to three times the levels measured in real operation. Current objective for the ASCALON range: to demonstrate an unlimited service life, in line with the requirements of EASA CS-P §370.

Engine torque vibrations are high-frequency torque exchanges between the engine and the propeller, generally imperceptible to the pilot — yet they have a major impact on internal stresses and component ageing. This is precisely the phenomenon that explains the restricted (forbidden) RPM ranges found on certain engine/propeller combinations.

Chicken Gun — bird-strike impact test bench

Bird impacts are among the most demanding events a propeller can encounter in service. To assess blade robustness under such conditions, the aerospace industry uses a "chicken gun" to simulate a controlled impact with a bird-sized projectile at representative velocity. E-Props designed and built this bench in-house; it is used in the development and validation campaign of the ASCALON range, dedicated to direct-drive engines and currently undergoing EASA CS-P and FAA 14 CFR Part 35 certification.

SHEEVA — high-pressure hydraulic test bench

The SHEEVA bench tests the governor, regulator and calculator assemblies of the GLORIEUSE variable-pitch propeller range. It simulates real flight conditions hydraulically, so that every such component is individually validated before delivery.

Traction (thrust) test bench

An instrumented ground bench fitted with a traction (thrust) measurement system, built around a 40-tonne hydraulic cylinder. It is used to verify the static thrust delivered by a propeller/engine combination and to validate design calculations before flight testing.

Flight test facilities

E-Props' own aircraft fleet allows flight-test campaigns of every propeller model and prototype, flown from the Sisteron-Vaumeilh airfield (LFNS).

The SKYRANGER, Rotax 912S (100 hp), acquired in 2015, is E-Props' first flight-test aircraft: its simple tube-and-fabric structure makes it easy to fit measurement equipment (a boom under the wing for airspeed and angle-of-attack, a cockpit fitted with sensors and cameras...), and it covers a useful low-speed test range from 30 to 190 km/h.
The PENA Bilouis is an aerobatic aircraft. It was purchased by E-Props in 2024. It is fitted with a Lycoming O-360 engine (180 hp) and has a high VNE, which makes it very valuable for testing.
The JODEL DR1054 (red) joined the fleet in early 2026. It is powered by a Lycoming O-320 (150 hp), which allows every test to be carried out on this widely used engine. It is also E-Props' liaison aircraft, with 4 seats, an excellent cruise speed and reasonable fuel consumption.

Few propeller manufacturers have their own dedicated flight-test aircraft fleet — a capability the E-Props team is happy to highlight, as it remains rare in the industry.

Samy Dupland, test pilot

All flight and ground test campaigns are conducted by Samuel "Samy" Dupland, a professional pilot and E-Props' dedicated test pilot since 2020.
He is a passionate aviator with more than 5,000 flight hours on numerous aircraft types.

ELIAS — in-flight data-acquisition system

ELIAS is the onboard data-acquisition system designed and manufactured entirely by E-Props. Originally developed to measure propeller/engine vibration and blade strain, it now integrates a complete flight-performance measurement chain: propeller thrust and torque, temperatures, static and dynamic pressures, engine RPM, angle of attack and sideslip, vibratory loads and blade-root strain. Onboard the aircraft, the recording unit — nicknamed Bob — houses the control screen, the recorder and the battery. Thousands of measurements are transmitted in real time (Wi-Fi, and 3G for flight campaigns) to this onboard computer and to a cockpit display that lets the pilot refine the flight profile according to the required test parameters.

A streamlined version of ELIAS is supplied with every ASCALON propeller: integrated into a spacer between the engine flange and the propeller hub, it lets owners perform, in a few minutes of static ground running, a personalized vibration survey of their own engine/propeller installation.

Case study: torsional vibration on a Lycoming O-360

To illustrate how these test means are combined in practice, E-Props instrumented a Lycoming O-360-A1D engine (hollow crankshaft, internal strain gauges) with the ELIAS system (16 channels, 6.5 kHz/channel, GPS-synchronized) and compared four propellers on ground and in flight:

The E-Props three-blade carbon propeller, designed with LUKY's dynamic engine/propeller model and evolutionary optimization, showed markedly lower torsional excitation than the other three propellers — demonstrating, on real flight and ground data, the value of designing the propeller and engine as a single dynamic system rather than tuning the propeller afterwards.
Reference: J. Buiatti, Experimental Torsional Vibration Analysis of Different Propellers on a Lycoming O-360 Engine, E-Props, January 2026.

These test means support compliance with the applicable propeller airworthiness requirements: EASA CS-P (including the fatigue/vibration requirements of §370), ASTM standards, and FAA requirements as applicable to each market. Links to the corresponding test reports are available in the comments section of each propeller in the E-Props catalog, and additional test videos can be found on the E-Props YouTube channel.

Comparing two propellers in flight is one of the most demanding exercises in propeller development: it calls for sophisticated measuring equipment, a strict protocol, an experienced pilot, stable weather, and a great deal of time. Skip any one of these, and the result is not a comparison but a coincidence.

What a comparison must establish

A propeller has no performance of its own: it only performs together with a specific engine and a specific airframe. Comparing propellers therefore only makes sense within that same combination. Each propeller is judged on three criteria: take-off roll distance, initial climb rate at the speed recommended by the aircraft manufacturer, and cruising speed (fast or economical) for a given fuel flow. The choice of propeller also affects: aircraft weight, engine and gearbox longevity through the propeller's moment of inertia, aircraft availability through robustness and TBO, piloting comfort through vibration, cockpit noise, and the pilot's budget.

Measuring equipment

In flight: engine speed, aircraft speed (IAS, CAS, TAS), intake pressure (MAP — without it, only full-throttle points can be legitimately compared), altitude, rate of climb, sideslip indicator. On the ground: precise blade pitch (digital protractor), precise propeller weight complete with spacer, spinner and all hardware, and take-off roll distance — this last measurement, often skipped in practice, is nonetheless a safety matter, particularly on short runways or in hot weather. For its flight tests, E-Props uses its onboard ELIAS system (see above), which records all these parameters, synchronized and time-stamped.

Ground-adjustable pitch propeller settings

Comparing ground-adjustable pitch propellers means matching not the pitch angle itself, but the aerodynamic operating point it produces — for a given intake pressure, the same engine speed. Two approaches are possible: set the pitch so both propellers deliver the same take-off performance and climb rate, then compare cruising speeds; or set the pitch so both propellers reach the same full-throttle level-flight speed, then compare take-off, climb rate and cruising speed. The soundest method is to set each propeller to the pitch that lets it meet the aircraft's non-negotiable safety margins (take-off and obstacle clearance in hot weather), and only then compare the cruising speeds achieved at that setting.

Test conditions

Four conditions must be met: same aircraft and same engine (only the propeller changes); same weight and same fuel mass (on an over-powered light aircraft, +1% of weight costs about 1% of climb rate); same pilot, ideally little influenced by preconceived opinions about the equipment under test, flying symmetrically and holding altitude precisely; same weather (in mountainous terrain, even a light breeze or thermal activity is enough to generate updrafts and downdrafts that mask the true difference between propellers).

Data analysis

Once flown, the data is processed and written up in a detailed report, with tables and graphs — graphs in particular make the scatter in the data, and therefore its confidence level, immediately visible. Each test campaign is documented on a standard field form, filled in flight by flight (date, aircraft, propeller, pitch, RPM, speed, intake pressure, fuel flow, take-off distance, climb rate...).

In order to understand propeller operation, it is simpler to perform the analysis at propeller level rather than at airfoil level.

Thrust and the reaction principle

Newton's third law states that if a part A applies a force F_A on a part B, part B applies a force F_B on part A, of the same value, the same line of action, but the opposite direction — the "action = reaction" principle. For a propeller to generate forward thrust, it must apply a backward force on the mass air flow passing through the disc swept by the blades: this mass air flow equals the disc surface multiplied by air speed and air density. The blades act like wings: their airfoils let the propeller apply lift forces on the airflow, modifying its speed.

Velocity increment across the disc

The difference between upstream and downstream air speed is: ΔV = Thrust / mass air flow, which follows from Newton's second law (F = d(m·v)/dt). This speed variation induced by the thrust is applied half upstream and half downstream of the disc. The mass air flow equals air density multiplied by the swept disc surface and by flight speed plus half of ΔV.

Propulsive efficiency

From this, the useful power delivered to the aircraft (thrust × flight speed) and the absorbed power (thrust × [flight speed + ΔV/2]) can be derived, giving the propulsive efficiency factor r_p = P_u / P_a — an absolute limit that remains the designer's goal. Too small a propeller diameter leads to mediocre performance, a problem that worsens at low flight speed; increasing the number of blades can reduce this loss, but never replaces an adapted diameter. The designer cannot reduce these fundamental losses, but must avoid making them worse through a poor thrust distribution across the disc — this is the role of pitch, chord and airfoil selection.

Blade drag

Other energy losses come from blade drag, which splits into two parts. Friction drag is more complex to model on a blade than on a wing because speed varies from root to tip: at the root, low speed and small chord give a mediocre Reynolds number (poor airfoil performance); at the tip, high speed close to the speed of sound raises the Mach number, degrading airfoil characteristics, and with the slightest curvature or incidence defect, the airflow may turn supersonic — generating noise and losing performance. Lift-induced drag is likewise harder to calculate on a blade than on a wing because of this variable speed along the span: having found no suitable method in the specialized literature, E-Props engineers developed their own calculation method — 90% of aerodynamic modeling time is spent determining these induced effects.

Full technical note (PDF)

Inertia is the capacity of an object to resist a variation of speed. For an object in translation, this resistance depends only on its mass. For a rotating object, mass alone is not enough: its distance from the axis of rotation must also be known to characterize its resistance to a variation of angular speed. This resistance is the moment of inertia (MOI), expressed in kg·cm², and it is a critical parameter for a propeller: I = m · r² — for a given mass, the further it sits from the axis, the greater the moment of inertia. A heavy hub close to the axis therefore generates comparatively little moment of inertia, while the mass distribution along the blades (root, mid-span, tip) generates a great deal: two propellers of identical mass can have very different MOI depending on this distribution.

Why MOI matters for an engine

Most aeronautical piston engines rely on the inertia of their rotating assembly to regularize the rotation of the crankshaft, accelerated at each power stroke and slowed down during the rest of the cycle. The propeller is by far the largest rotating inertia in the powerplant. On a geared engine, the torque peaks linked to this inertia are absorbed by the gearbox; on a direct-drive engine, the crankshaft itself takes the load. In both cases, fitting a propeller whose MOI falls outside the range specified by the engine manufacturer reduces the service life of the affected components, and can, in extreme cases, lead to failure of the gearbox or of the mounting hardware.

The E-Props approach

The moment of inertia of every E-Props propeller is calculated during design, then verified and measured on each finished part. This value is published for every model in the E-Props catalogue, so it can be checked against the engine manufacturer's limits before installation. Thanks to their light weight — achieved through carbon construction and thorough aerodynamic and mechanical optimization — E-Props propellers respect, with margin, the minimum and maximum MOI limits recommended by engine manufacturers.

See also: Rotax Service Instruction — Propeller mass moment of inertia

Jérémie Buiatti, designer of E-PROPS propellers, has taken a close interest in what happens between the engine and the propeller, and wrote a reference article on the subject for the Colloque de Cachan 2016 (INTER ACTION association): "The pistons pound, the gearbox perishes".

Why a reduction gearbox?

On a direct-drive engine, the propeller turns at the same speed as the engine: either the engine is slow (~2600 RPM, large displacement, heavy, noisy), or it runs faster (3300-3600 RPM), but then the propeller diameter must be limited to avoid exceeding the speed of sound at the blade tip — at the cost of mediocre efficiency, especially at take-off, and a high noise level. A reduction gearbox allows engine speed and propeller speed to be optimized separately, with better efficiency for both, particularly at take-off. Developing one is nonetheless complex, because of engine torque ripple — caused by piston and rod acceleration, and by pressure differences between the piston faces.

The "backlash operation" problem

Engine torque is far from constant, while the propeller's aerodynamic torque stays essentially constant at a given RPM. It is the inertia of the rotating parts — crankshaft, flywheel, gearbox, propeller — that absorbs this torque ripple and supplies the energy shortfall during the non-driving phases. In a gear, toothed-belt or chain reduction gearbox, there is almost always a small amount of backlash between the teeth. As long as torque stays positive (energy flows from the engine to the propeller), the teeth stay engaged and this backlash has no effect. But if torque reverses on every engine cycle — the propeller then feeding energy back to the engine — the rotational speeds decouple for the time it takes to cross this backlash, and the gap must be absorbed by an impact each time the teeth re-engage. This "backlash operation" regime wears the gear teeth prematurely, and can become severe under resonance.

The condition to respect

To stay in backlash-free operation, the aerodynamic torque must be able to decelerate the propeller more strongly than the compression torque decelerates the crankshaft — which, simplified, requires a propeller moment of inertia low enough relative to the engine's and to the square of the reduction ratio. A worked example from the article, for a Rotax 912S with a 2.43 reduction ratio: in the worst case (an instant throttle-up at 1700 RPM), negative torque reaches -80 N·m on the crankshaft side, against an aerodynamic propeller torque of only 12 N·m. The condition then requires a propeller moment of inertia below about 0.09 kg·m², while the lightest propellers for this engine are around 0.25 kg·m² — so backlash operation does occur during an abrupt throttle-up. Fortunately, the Rotax 912 gearbox includes a dog-clutch system that absorbs these torque spikes.

Available levers

The article lists several ways to favor backlash-free operation: a low propeller moment of inertia; a high engine moment of inertia (flywheel — but that is heavy); a high reduction ratio; avoiding prolonged low-RPM operation (for example at idle while warming up, a case flagged as dangerous); increasing the number of cylinders to reduce the amplitude of torque ripple; or using a centrifugal clutch to disengage the propeller at low RPM.

And without a reduction gearbox?

On a direct-drive engine, the backlash between propeller and crankshaft is zero: the two inertias then share the torque ripple in proportion to their value — the oscillating torque seen by the propeller equals the engine's oscillating torque multiplied by I_propeller / (I_propeller + I_engine). A low propeller moment of inertia therefore directly reduces the torque peaks experienced by the propeller and its engine mount — which is why direct-drive engine manufacturers impose a maximum propeller moment of inertia (for example 0.3 kg·m² for the Jabiru 2200), at the cost of a slightly less stable idle, corrected in practice by a small increase in idle RPM.

A propeller must be perfectly matched to the engine, to its moment of inertia and to its rotational speed, deliver the strongest possible thrust, be ultra-light for the best reactivity and for engine and gearbox longevity, and be as quiet as possible. E-PROPS propellers have very particular airfoils, patented designs (SCIMITAR, for example) and specific blade positioning to reduce drag and achieve the best possible thrust.

Efficiency

Engine power, diameter and number of blades set the absolute upper limit of possible propulsive efficiency; it is then up to the designer to get as close to it as possible. E-Props propellers all have a thin chord, which generates less drag than a wide chord: depending on blade geometry, this saves between 6 and 9% of fuel at the same engine speed. Better efficiency also means a lower noise level.

Number of blades

At equal diameter, more blades mean better efficiency — provided each blade's geometry is specifically adapted to its configuration: a universal blade simply fitted to a 3- or 4-blade hub will not deliver ideal performance.

Measurements on 170 cm diameter propellers, Rotax 912S engine, 2.43 reduction ratio.

Diameter

Increasing diameter improves propulsive efficiency: at equal blade count, a 180 cm propeller delivers more thrust than a 155 cm one (235 kg versus 264 kg, measured on a Rotax 912S, 2.43 reduction ratio, 3-blade). But not every engine-gearbox combination can accept a large diameter: peripheral speed at the blade tip must not exceed 0.75 Mach (about 900 km/h), beyond which noise becomes excessive.

Two-blade or three-blade: a myth to correct

It is sometimes said that the wake of the 3rd blade of a 3-blade propeller crosses the wake of the other two, reducing efficiency. This is physically false: blades never pass through each other's wake, because that wake is carried rearward by the propeller's own slipstream — the condensation trails visible on some aircraft propellers demonstrate this clearly. The only real case of wake crossing occurs during pitch reversal ("reverse") on a variable-pitch propeller, used to shorten the braking distance on landing — hence the very characteristic noise of that manoeuvre.

E-Props propellers are among the lightest in the world — not for communication or regulatory reasons, but to respect the laws of physics. Example: a 3-blade DURANDAL-3, 170 cm diameter, with titanium leading-edge armor and all hardware, weighs 2.1 kg.

Why weight matters so much in aviation

Any extra weight degrades climb rate (more lift needed, hence more drag), increases fuel consumption, reduces cruise speed, increases stall speed, lengthens landing distance, and raises the risk of structural damage during manoeuvres. A light propeller also improves powerplant operation across the whole RPM range, reduces jolts and stress on the gearbox (by respecting the engine's maximum moment of inertia), and reduces vibration. A low moment of inertia also makes the propeller more responsive to control input, and stops it faster in the event of an impact — less kinetic energy stored, therefore less damage. Finally, every kilogram saved translates directly into fuel and range: a 4 kg weight saving (common with E-Props) represents about 5.6 litres of fuel, or 50 km of additional range on a Rotax 912-powered aircraft.

Why E-Props propellers are so light

Through their design, engineered to save weight from the outset. Through their materials: 100% carbon braid with continuous fibres between the upper and lower surfaces, a very high fibre content (63%), epoxy resin exclusively — no fibreglass, no mixing of materials — and titanium leading-edge armor, twice as light as steel or Inconel. Through their manufacturing process, RTM (Resin Transfer Molding, also used by Airbus or Boeing to lighten structural parts), which saves resin while staying very strong. And through their finish: the blades are simply polished, with no gel-coat or paint — a gel-coat can add more than 400 g to a 170 cm 3-blade propeller, and cannot be repaired once damaged.

Can a propeller be too light?

No. It is sometimes said that a heavy propeller makes an engine "run smoother" at idle. But while speed variations are then smaller, load variations between the propeller and the pistons are larger across the whole mechanical chain — and it is load variations that cause damage, not speed variations. Engine manufacturers, moreover, almost never specify a minimum moment of inertia: nothing prevents fitting an ultra-light propeller to an engine, at most the idle setting needs adjusting.

Mass comparison — 3-blade ground-adjustable propellers (Rotax 912S, ~170 cm, hardware included, no spacer)

Mass comparison — 3-blade constant-speed propellers (Rotax engines, with spinner, hardware included, no spacer)

Since 2008, the E-PROPS team has carried out in-depth research into propeller noise and how to reduce it as much as possible.

Propeller noise

Propeller noise is dominated by "tonal noise", linked to blade-passing frequency — hence to rotational speed and blade count — and split into thickness noise, loading noise, and (negligible) shear noise. On a tractor aircraft with a geared engine whose peripheral tip speed stays below 0.65 Mach, a modern, efficient propeller generates very little nuisance: perceived noise then comes 75% from the engine. The issue is mostly relevant to pusher-configuration aircraft (weight-shift trikes, gyrocopters, drones, paramotors), where 80% of the noise comes from the interaction between the propeller and the aircraft's wake — the same propeller, fitted to different paramotor frames, can therefore have very different noise signatures. Reducing pusher-propeller noise therefore starts with refining the airframe's aerodynamics, so the propeller operates in the least disturbed air possible.

E-Props concepts

With standard propeller concepts, noise reduction remains locked into a difficult trade-off with aerodynamic performance. E-Props has therefore developed its own concepts:

• QD2 (Dual-Phase 4-blade): a 4-blade propeller made of 2×2 blades at different angles. Noise reduction versus a standard 3-blade propeller: about 6 dB(A).
• H2D (Dual-Phase 6-blade): the same principle as some helicopter anti-torque rotors. Reduction: about 8 dB(A), with a different tone (closer to a turbine).
• SCIMITAR design: a special airfoil generating less noise than a standard straight blade. Reduction: about 5 dB(A).

Number of blades

Contrary to popular belief, a 3-blade propeller does not generate more noise than a 2-blade one: comparative test campaigns, run for months by the E-Props team with certified sound-level meters, show an identical or lower noise level — only the tone changes, sometimes higher-pitched. In flight, a 3-blade propeller's noise exposure cone is even narrower than a 2-blade one's, so it is heard on the ground for less time. On 4-blade propellers, the QD2 design (blades not crossed at 90°) brings a further gain of 3 to 4 dB(A) over a standard 4-blade propeller.

Spacers

Moving the propeller away from the engine also reduces noise: SCIMITAR propellers integrate a 10 mm spacer at the hub, for a gain of 4 to 5 dB(A). Standard aluminium spacers are heavy (20 g/mm, i.e. 2.4 kg for 120 mm), complex to fit and expensive. E-Props developed its own system, the ESU carbon spacers: 13 times lighter (1.5 g/mm), free of charge regardless of length, integrated into the hub (no superfluous hardware, hence less vibration), with interactive mounting drawings.

Blade tips

The design office tested more than ten blade-tip shapes (45° cut forward or backward, rounded, winglet, anti-vortex tube...). Technical verdict, despite the commercial appeal a more exotic design would have had: a clean 90° cut relative to the blade axis remains the most effective, both for noise reduction and for efficiency — the two being closely linked. Work continues, with several prototypes under development.