see also: evolution propeller design methods

1.8 m aircraft propeller CFD: cruise and static A first aircraft propeller designed in Heliciel Propeller at take-off: off-design calculation

See Tutorials Heliciel

History of the aircraft propeller, 1920 to today: materials, variable pitch and blade shapes

The history of the aircraft propeller, from 1920 to today, reads as a series of answers to two constraints: blade tip speed and the gap between climb speed and cruise speed. Wood gave way to metal, the flat twisted blade to the cut profile, direct drive to the reduction gear, and fixed pitch to variable pitch. This page follows those steps up to about 1940 and draws out what they teach designers of propellers today. For a wider view, we also have a page on the history of the propeller in general, and another on the aircraft propeller and how it is calculated.

This part draws on Jean Grampaix's book L'Hélice, listed in the bibliography of the site. We keep its facts, dates and names, but we rephrase and comment on them in our own words. The dates and figures come from that book; we could not cross-check all of them, and we write "according to Grampaix" wherever it matters.

On this page:

1: Before 1920: the wooden aircraft propeller

Before and during the 1914-1918 war, the technique was the same everywhere: the propeller was made of glued wooden laminations. The shape was first cut by hand, then, at some propeller makers, with a copying machine that reproduced a model. The propeller was then varnished or lacquered.

Workshop making wooden aircraft propellers in the 1920s, with workers shaping blades on benches

A propeller maker's workshop (the print carries a stamp dated 1922): the wooden blades are still shaped by hand. In the foreground a blade rests on a bench; other blades wait at the back of the workshop.

Plan shapes varied widely. The leading edge was sometimes curved, more or less strongly, and sometimes straight, which gave the same variety in the neutral fibre. The blade tip could be rounded, square or very tapered. As for blade profiles, Grampaix gives orders of magnitude at mid-radius: a relative thickness of 10 to 30% and a blade width of 9 to 11% of the diameter. The pressure side is flat in the working part, and the pitch is constant along the blade or increases from hub to tip.

To gain finesse (lift-to-drag ratio), and so efficiency, some makers tried symmetrical profiles. The idea was soon dropped: it widened the gap between the rotation speed in level flight and in climb, which made the propeller less flexible in use.

Manufacture was as much a matter of craft skill as of technique, and it stayed that way until metal propellers appeared. The propellers of the time drove engines of about fifty horsepower, measured 2 to 2.50 m in diameter and turned at 1,200 rpm. According to Grampaix, the tip speed stayed below 150 m/s. A quick calculation gives the same order of magnitude: at 1,200 rpm, that is 20 turns per second, the tip of a 2 m diameter blade moves at about 126 m/s and that of a 2.50 m blade at about 157 m/s. At such low speeds, the propellers held up easily, even though they rested on rough calculations and imprecise techniques.

Matching was not a problem either. Aircraft had little excess power: their maximum level speed stayed close to their climb speed. The advance per turn (the distance travelled during one propeller revolution) therefore varied little, and the propeller kept practically its maximum efficiency in climb.

Between 1920 and 1925 the wooden propeller improved more in how it was made than in its aerodynamic quality, and it proved insufficient for the new engines. Attention turned to the metal propeller, which first appeared as a twisted plank and then as a cut profile.

2: 1920 to 1940: why wood gave way to metal

In this period aircraft performance grew a great deal. For propellers, two difficulties followed:

  1. tip speed approached 300 m/s, roughly double that of the previous period. Centrifugal force grows with the square of the speed, so the factors tied to strength were multiplied by 4;
  2. aircraft had more excess power, so the gap widened between maximum level speed and climb speed. A real matching problem followed: in climb, a propeller could lose up to 25% of its maximum efficiency.

Wood was no longer enough, either aerodynamically or in strength of materials. Grampaix credits metal with several advantages over wood:

The first metal propellers weighed 30 to 40% more than their wooden counterparts. Later, the weights became practically equivalent.

3: The twisted metal "plank" propeller (Reed, Levasseur)

According to Grampaix, the American Reed filed the first patent for a metal propeller in March 1921. A 30-hour endurance test on the test bench followed, then a first flight test on 30 August 1921. These were twisted metal "planks" of uniform relative thickness, about 5%, over the whole blade radius.

Reed asked the flexibility of his blades to balance the aerodynamic moments partly against the centrifugal moments. Stiffness then came from centrifugal force rather than from the blade's own strength, so the bending loads were "passed through". This concept is found today on helicopter rotors. A certain balance was indeed obtained, but the great flexibility of the blades had two serious drawbacks:

Pierre Levasseur brought these propellers to Europe in 1924, under Reed's licence. They were made in one piece, machined from a forged duralumin blank. They were used for a long time, until about 1930, and allowed Bonnet to fly at more than 400 km/h in a Bernard-Hispano.

Twisted sheet-metal aircraft propeller of the Levasseur-Reed type fitted to a biplane, front view

Twisted sheet propeller (Reed type built by Levasseur) on a biplane: the blade is thin, almost without thickness, very different from a wooden blade. (fig. 37 of Grampaix's book)

4: The cut-profile propeller and the Junkers "cheek" propeller

After the era of the flat twisted metal blade came that of the cut-profile propeller, a stage we have not left. One special case exists: hollow blades of mild steel sheet, gas-welded. They already existed in 1930 as an exception (Leitner propellers). Their bending strength was sufficient, but they were criticised for holding centrifugal loads poorly, which made the metal slip along the weld lines.

From the start of cut propellers, two light metals were considered, duralumin and magnesium. Duralumin was chosen. Magnesium was criticised for its low elastic limit, hence permanent deformations, and for the difficulty of always obtaining homogeneous magnesium.

The general layout of the 1930 propeller looked like this:

One particular layout is worth mentioning: the Junkers "cheek" propeller. The blade widens at the root to cool the engine. The system was good, because the cooling adapted itself automatically to the various flight conditions. However, such a layout was poorly suited to variable pitch because of the loads it created when the pitch was changed.

Drawing of a Junkers cheek propeller: two blades widened at the root near the hub

Junkers cheek propeller: the two blades widen at the root, near the hub, to cool the engine better.

5: Two, three and four blades: the number of blades

The propeller of these decades was generally two-bladed, but the first three-bladed propellers appeared. Aircraft became more manoeuvrable, finer and faster. This turning rate (the aircraft's turn), combined with the rotation of the blades, creates gyroscopic effects. Analysis shows it is worth fitting three blades instead of two. With three blades, the vibrations due to gyroscopic effects disappear, and these are all the more harmful when aircraft have reduction gears.

Grampaix explains why. A two-bladed propeller has a moment of inertia (about an axis across the disc) that is zero, or nearly so, when it is horizontal and maximal when it is vertical. At each turn the gyroscopic couple therefore goes through two zero values and two maximum values. On a three-bladed propeller the moment of inertia is constant, and so is the couple.

The advantage of the three-bladed propeller is all the clearer because no weight increase is to be feared: a well-designed three-blader weighs as much as its two-bladed counterpart, sometimes less. Another advantage: when compressibility at the tip became a concern, the three-blader allowed the diameter, hence the tip speed, to be reduced by about 7%.

From about 1930, compressibility effects at the tip were a worry when tip speed was close to the speed of sound. A sudden drop in efficiency under these critical conditions had already been recorded. It was also found that very thin, very narrow profiles with a sharp leading edge reduced these effects at the tip, with the maximum thickness located near mid-chord. These are the features found today on the wings of supersonic aircraft. Propeller experience thus served aircraft makers when they faced very high speeds.

In these respects the four-bladed propeller brings the same improvements as the three-blader, but its efficiency is lower, because blade interaction reduces it. We come back to this compromise at more length in our page on the number of blades of a propeller.

6: Pitch ratio: growth and limit

The gradual improvement of maximum propeller efficiency, from the early days of aviation to 1930, corresponds to a continuous increase in pitch ratio. The pitch ratio is the ratio of the propeller's pitch to its diameter (H/D). A high pitch ratio means the blade is set at a large angle to the plane of rotation.

According to Grampaix, the pitch ratio went from 0.4 to 1.2. It reached 1.4 only on speed-record aircraft, whose power per square metre of swept area exceeded 350 hp. Beyond that, for the same tip speed, maximum efficiency falls as the pitch ratio rises.

There is more. Efficiency tends to decrease as tip speed grows. The efficiency curves as a function of advance per turn spread out less and less: the propeller's matching is less flexible. As the propeller is a compromise, take-off and climb suffer from this lower flexibility. These curves also show that, for the same tip speed, the thrust per square metre of swept area begins to fall when the pitch ratio approaches 1.8.

The case of the Schneider Cup seaplanes is an instructive aside. These aircraft had propellers with a large pitch ratio, and almost the whole propeller surface contributed to propulsion.

Schneider Cup racing seaplane seen from the side, with its two-bladed large pitch ratio propeller

A Schneider Cup seaplane: here the same propeller has to give good efficiency at very high speed, at the price of a difficult take-off.

What solution, then, improves the performance of the propulsor and so of the aircraft? The reduction gear became common on aircraft of the time. By lowering the rotation speed, it allows larger diameters, hence a larger pitch ratio, and it spreads the efficiency curve as a function of advance per turn. But Grampaix does not think it brings much of a solution well beyond the usual pitch ratios. It was then accepted that the only solution was automatic pitch variation, continuously following changes in forward speed. The variable pitch propeller was close.

7: The powerplant: reduction gear, supercharger and variable pitch

The evolution of the propeller, that of the engine and even that of the airframe went on in parallel. In the technical writings of the last years of the period, the initials GMP (groupe motopropulseur, powerplant) appeared, to signify that the propeller, the engine, its accessories, the reduction gear, the supercharger and the pitch-change system formed a whole. We must therefore consider here the reduction gear, the supercharger and pitch variation, whose introduction spans the two decades 1920-1930 and 1930-1940.

The reduction gear

A "reduction gear" existed on quite a number of aircraft of the early days of aviation. It was simply two sprockets joined by a chain drive. It was asked to raise maximum efficiency by increasing the propeller's rotation speed, because of the low speeds of aircraft at the time. It was therefore not a reducer but the opposite: a system that speeds the propeller up, which explains the quotation marks.

When aircraft speed increased, this "reduction gear" was abandoned and, for many years, propellers were fitted in direct drive. Aircraft speed was enough to give good maximum efficiency in such conditions. Engines turning at 1,500 or 1,800 rpm were accepted for a while. They were heavy per horsepower, but that speed was known to give good propeller efficiency, and everything was sacrificed to it.

Yet aircraft speed kept rising. For a reason diametrically opposed to that of the early days, reduction gears were fitted again, real reducers this time, around 1930. The aim was to reconcile high engine speeds with sufficiently slow propeller speeds. Grampaix credits them with several advantages:

The balance sheet has a counterpart. The reduction gear raises the pitch ratio, so the propeller turns more slowly as a receiver and the braking effect per unit of disc area decreases. Grampaix nevertheless concludes that the balance is beneficial: the reduction gear improves braking by the propeller.

More generally, by raising propeller efficiency, the reduction gear improved all the aircraft's performance, often by a great deal: level speed, ceiling, and above all range. The same improvements were then asked of variable pitch. Reduction gear and variable pitch are converging elements of one problem, propeller efficiency.

The supercharger

Take an aircraft fitted with a supercharged engine, at constant engine torque and with a fixed-pitch propeller. According to Grampaix, the rotation and translation speeds in level flight then both grow by 5% per 1,000 m. A propeller designed for a recovery altitude of 6,000 or 7,000 m will therefore work very badly on the ground and in climb.

With an engine without a supercharger, power falls together with air density, and so with the resistance to blade rotation. At a pinch, a fixed-pitch propulsor is enough: the pressure ratio is very close to the density ratio. The propulsor is set for maximum efficiency at the operating altitude, and take-off and climb characteristics are sacrificed. Aircraft of this category did not exceed 4,000 or 5,000 m in normal flight, and the efficiency gain that variable pitch would have brought would not have exceeded 15%. That 15% was given up to avoid the complex variable-pitch mechanism.

With a supercharged engine, it is different. Its power at altitude stays equal to that at ground level, despite the lower air density. Yet the propeller works in thin air, which resists it less. It races and imposes on the engine a speed incompatible with its mechanical strength.

The engine can be fitted with a device that holds its speed, and so the propeller's, constant. But the propeller then no longer throws back a sufficient mass of air. The flow falls, and the propulsive effect, which depends on it, falls too. This balance leads Grampaix to conclude that the supercharged engine can only be conceived if it drives a variable pitch propeller: constant-power engine and variable pitch propeller are inseparable.

Variable pitch

From the 1920s the variable pitch propeller appeared to be a necessity, but building it became complicated for engines of 100 or 150 hp. At that power, centrifugal force is not enormous, and the blade shanks can be pivoted in smooth metal sleeves without the friction due to centrifugal force becoming excessive. With 300 hp engines, according to Grampaix, this force already reached some twenty tonnes. Friction became considerable, which forced the use of special steels and led to larger hub dimensions.

The problem could be left in the drawer for a few years after 1920. It had to be tackled when devices designed to restore power at altitude appeared, such as the Rateau turbo-supercharger, which raised the ceiling of aircraft and their speed at high altitude. Engineers approached it from various angles. They even tried the variable-diameter propeller, whose blades, on the pilot's command, were to screw in or out on a mandrel to reduce then increase the "useful" diameter. That formula never found a workable solution.

The variable pitch propeller deserves a page to itself, and we have written it: see the variable pitch propeller, its mechanisms and its governing.

8: What this history teaches modern propellers

This history holds design lessons that still apply today, for an aircraft propeller as well as for a drone, boat or wind turbine propeller. We summarise them here without adding any figure to those of the book cited.

In Heliciel
Heliciel lets you revisit these compromises on your own propeller. It calculates efficiency and thrust as a function of rotation speed and, with the multiple analysis, compares several speeds or several pitch settings. See the optimisation of rotation speed and number of blades.

To go further, we can move on to the variable pitch propeller to see how pitch adapts to flight, or return to the choice of the number of blades, which extends part 5 of this page.