THE performance of airplanes, whether military or civil, is primarily a problem in distribution of weight. Speed depends on the weight that can be devoted to power-plant. Range depends on the weight of fuel that can be carried. The capacity to reach great altitudes depends, again, on the weight that is put into power-plant, and on the extension of wing surface. Manœuvrability calls for weight assigned to the same purposes as climbing ability. Fighting power, in the case of military airplanes, demands the expenditure of weight on guns, bombs, ammunition, sights and armor. Yet the demands of all these qualities except the last are arrayed against a simple increase in total weight -- a measure which would be self-defeating. Speed, for example, demands that weight be put into the power-plant without increasing the total weight of the airplane as it stands ready for flight. This means that the weight must be taken from somewhere else, and put into power-plant at the expense of fuel supply, or structure, or armament, or navigating equipment, or some other of the elements of which the total weight is compounded.

I do not want to give an oversimplified impression. Obviously the performance of aircraft cannot be reduced to a simple series of algebraic formulas. Obviously the skill and ingenuity of the designer, and the extent and rigor of the preliminary testing and developmental research applied to the particular design, do much to determine its precise qualities. The Spitfire and the Airacobra and other aircraft of deserved reputation owe their particular success to the ingenuity entering into the original concept or the choice of its details, and to the skill and care with which the engineering organization worked out the design. Still it remains true that the distribution of weight is the fundamental control which governs the limits of performance in any existing state of the art. Designers will approach those limits with varying degrees of closeness, but they may not hope to exceed them except through new research opening new vistas of possibility to all designers alike. Moreover, the skill of the designers is itself revealed largely through the measures adopted to conserve weight in one element of the aircraft in order that it may be more lavishly expended elsewhere in the interest of improving performance.

Laymen rather generally seem to picture the designing of airplanes as a field in which revolutionary inventions come fairly tumbling on one another's heels. This is strikingly inaccurate. The general history of the last dozen years has been one of gradual refinement in detail, and of successive steps in the struggle to reduce such basic weight factors as the amount of weight of the engine for each horsepower that it delivers and the amount of fuel consumed for each horsepower that the engine develops and each hour that it runs. A group of changes of revolutionary importance between 1930 and 1933 altered the whole form of the modern high-efficiency airplane. Another change that may be of almost equal importance (though much less spectacularly apparent to casual observation) is now in the stage of incubation in the laboratory, and hence is not yet properly subject to public and detailed discussion. But in general what I have referred to as the maximum limits of performance, or those relationships which, in the existing state of the art, govern performance in terms of weight distribution, have changed only relatively little over the past eight years. The improvement in performance has been spectacular within that time, especially in military aircraft, but it has been built mainly upon increase of engine power and changes in weight distribution and in relation of wing area to weight. Just how rapidly some of the other elements have actually improved will be described in later paragraphs.


Before considering where the weight is actually used, and why, we may consider the performance attainable by aircraft in terms of the relations of their principal characteristics. Total weight, engine power and wing area are the basic features in an airplane's design. They are commonly combined into the two ratios of power loading and wing loading, or total weight per horsepower of the engines and total weight per square foot of wing surface.

Engine power drives the airplane forward; the resistance of the air to the passage of the wings and body and other structural elements retards it; and the speed that it will attain is determined by the balance of the two. To increase speed, either the power must be increased or the resistance of wings, body and tail surfaces must be reduced. If the wings and body and other elements be given the best form that the accomplishments of research have up to this point put at the designer's disposal, their resistance can be further reduced only by a reduction of the body's size or by a clipping of the wing area. Assuming a fixed total weight for the airplane, then, we see that more speed demands either more power or less wing, or both. It demands either a lower power loading or a higher wing loading; and speed -- and most other elements of performance -- can be expressed reasonably well in the relationship between those two ratios.

The curves in Figure 1, on the next page, show how the speed of an airplane at sea-level may be expected to vary with power loading for various wing loadings, assuming that the designer has made the best possible choice of form, and that the particular military specifications for this particular task did not lay on him exceptional burdens in the form of protruding guns and other excrescences.

The selection of a few illustrative figures from the chart may serve to make it more clear. It appears, to begin with, that in order to secure a speed at sea-level of 200 miles an hour the total weight of the airplane must be kept down to 14 pounds for each horsepower delivered by the engine if there is one square foot of wing area for every 20 pounds of weight; but the same speed could be secured with a power loading as high as 21 pounds for each horsepower if the wing area were reduced to one square foot for every 40 pounds of total weight. In the same way, to secure a speed of 300 miles an hour it is necessary to keep the power loading down to 4.5 pounds per horsepower if the wing loading is 20 pounds per square foot; while power loading may rise to 8.0 pounds per horsepower if the wing loading is brought up to 40 pounds per square foot.[i]

All this might indicate that it would be advantageous to increase the wing loading almost without limit. So it would, if nothing but maximum speed at low altitudes were to be considered. But other elements of performance, also important, suggest a different course. As between two airplanes described in the previous paragraph as both likely to have a top speed at sealevel altitude of 300 miles an hour -- one with a wing loading of 20 pounds per square foot and the other with double that figure -- the latter would need more than twice as large a field for a safe take-off as would suffice for the former, and its minimum flight speed for contact with the ground would be about 40 percent higher. To be specific on the latter point, it is the general rule in present design practice that an airplane with a 20-pound wing loading lands at 60 miles an hour, one with a 40-pound loading at 85 miles an hour. However, where a device is used for reefing the wing in flight and extending its area somewhat on landing -- a device which is not yet very common but which may become more so -- those figures can be reduced to about 50 and 70 miles an hour, respectively.

The use of very high wing loadings also diminishes an airplane's ability to climb very rapidly or to reach great altitudes. All these factors must be brought into a compromise decision. It is not permissible to cite the exact characteristics of the latest military aircraft. However, the general history of the past twenty years has been one of gradual increase in wing loading, from a typical 15 pounds per square foot in both military and civil aircraft a dozen years ago to something between 25 and 30 pounds for each square foot of wing area in present commercial aircraft; to about the same values in the case of the latest fighters for which the data are public property; and to even a little higher ratio of weight to wing area in the case of bombers. With a wing loading of 30 pounds per square foot, power loading would have to be held down to about 6.2 and 4.0 pounds per horsepower, respectively, to permit the attainment of speeds of 300 and 350 miles per hour, respectively, at sea-level.

Military aircraft do but little of their fighting near sea-level, however, and transports do not by choice cruise near the ground. Speed is attained with less effort in the more rarefied upper air, and it is there that both combat and transport aircraft spend most of their flying lives. If an airplane engine were like that of an automobile, the maximum speed would fall off as the altitude increased, in spite of the reduction in air resistance. The decrease in engine power as the air density went down would more than offset the decrease in resistance. But the airplane engine, at least as designed for military aircraft, follows a different pattern. It is designed never to be operated with fully open throttle at low altitudes, but to permit, by a gradual and progressive opening of the throttle as the altitude is increased, the maintenance of the same maximum power output under any conditions from those prevailing at sea-level up to a considerable height. Five years ago it was common for engines designed for military service to maintain their maximum rated power at all altitudes up to about 10,000 or 12,000 feet. Now, however, the highest level of full-power performance -- known to engineers and pilots as the critical altitude -- has been raised to 15,000 or 20,000 feet, or even substantially more. Whatever the critical altitude of the particular engine may be, at that altitude the airplane in which the engine is installed will have its maximum speed; and from sealevel up to the critical altitude, the speed of a high-performance military or transport aircraft will increase by just about one percent for each 1,000 feet of increase of height. Thus the conditions, in the way of wing loading and power loading, that were cited as sufficient to develop a speed of 350 miles an hour at sealevel would be equally good for a speed of about 420 miles an hour at an altitude of 20,000 feet, provided only that the engine used had a critical altitude as high as that. In the last ten years the speed of fighters has increased nearly 200 miles per hour. Nearly 40 miles of that increase has been due to increases in the critical altitude of the engines.

The problems of weight distribution as applied to some particular types of aircraft will be discussed in due course. I pause first to dwell briefly upon the extraordinary economy of weight that must be achieved in the design and building of an airplane which is to attain present-day fighter speeds. To make 400 miles an hour at an altitude of 20,000 feet requires, if the wing loading is not to be allowed to go above 30 pounds per square foot, that the total weight of the aircraft, ready for flight, should not exceed 4.5 pounds for each horsepower of the engine. For a powerplant designed for any purpose except that of airplane propulsion, 4.5 pounds per horsepower would be an extraordinarily light weight for the engine alone.

It must again be emphasized that the speeds indicated here are not by any means attained by every machine with the indicated combination of power loading and wing loading. These speeds are presented, rather, as being about the best that can be attained in the present state of development. In practice, the speeds of actual aircraft fall a little below the curves in Figure 1. Even aircraft which would be regarded as competently designed, but whose designer worked under some handicap imposed by specification or otherwise, actually attain speeds as much as 15 percent below those indicated by the curves. Using the same methods of calculation that were employed to construct the curves, we find that to reach a speed of 500 miles per hour at 20,000 feet, even if the wing loading were allowed to rise to 50 pounds for each square foot, the power loading would need to be kept down to 3.7 pounds per horsepower.

But at such speeds as this the relatively simple system of computation that serves at lower speeds begins to break down. Above 450 miles an hour new complications appear. From time to time a newspaper announces that some scientist has discovered that it is impossible to fly at as much as 600 miles an hour. A few days after a recent appearance of this story came the announcement that a new fighter had actually attained a speed above that figure while being tested in a dive. Actually there is no fixed barrier beyond which speed cannot increase. However, the various parts of an airplane, like a projectile, ultimately attain a speed at which the air flow changes fundamentally, and above which the resistance begins to increase at much more than the normal rate. The speed at which this change takes place is not the same for the wings as for the nose of the body or the windshield form; nor is it the same for wings or bodies which differ among themselves in form. At best, the onset of the régime of rapidly rising resistance begins to be troublesome at speeds in the 450-550 mile range. Theoretically it would be possible to attain still higher speeds, just as it is possible to put enough power behind a projectile to make it travel through the air at speeds enormously higher than those at which the rapid increase of resistance takes place. But the price paid in power for each additional mile of speed in the high-resistance régime would be such that in the present state of our knowledge we may consider it at least four times as difficult to increase the speed of an airplane from 500 miles an hour to 550 miles an hour as it was to make the increase from 350 to 400.


When a pilot is concerned exclusively with making the longest possible flight with the least possible consumption of fuel, he flies at a much lower speed than he ordinarily uses for cruising where fuel economy is not the paramount objective. The actual value of the most economical speed depends primarily on the wing loading. The following course is typical:


Wing Loading Speed of Flight for
  Maximum Economy
(pounds per square foot) (miles per hour)
20 110
30 135
40 155

The foregoing are the speeds for greatest economy in flying at sea-level. At higher altitudes they increase at the rate of about 1½ percent per 1,000 feet of altitude. To tell the story in terms of the aircraft with which travelers in the United States are most familiar, it appears that the most economical speed for the present model of twin-engine Douglas transport, at an altitude of 10,000 feet, is about 140 miles an hour. Actually it cruises at just over 180 miles per hour at that altitude, and pays for the higher speed by consuming about 15 percent more fuel per mile flown than it would if the pilot at all times held strictly to the most economical speed. On a really long bombing raid (where weight of fuel is a paramount consideration, as the airplane may be barely able to carry enough fuel to make the flight), and on transoceanic flights of commercial aircraft, pilots seek to keep closely to the most economical speed, even though the penalty for small variations from it is slight. An increase in speed by 10 percent from the most economical value will increase the fuel consumption per mile by only about 2 percent. However, a 30 percent increase in speed causes a 15 percent increase in use of fuel.

Wind conditions also play a part in these calculations. The economical speeds so far described have been those for flights in still air, or at right angles to the wind. When flying with the wind, it is most economical to fly still more slowly; when flying against it, to increase the speed somewhat. Take the Douglas transport again as an example. I gave the most economical speed for it, at 10,000 feet in still air, as being 140 miles an hour. With a 30-mile headwind, the value should be raised to 149; with a 30-mile tailwind, it should be lowered to 133. All these factors are elaborately tabulated and charted, so that the pilot engaged on a long flight can read off immediately during the flight, without pausing for any computations of his own, the rate at which fuel will be consumed at any particular speed under any possible combination of wind and other conditions.

Under conditions of maximum economy, an airplane that represents the best possible practice in the existing state of the art -- i.e., has the best possible form to create minimum air resistance, the highest possible propeller efficiency, and the lowest possible unit fuel consumption by the engine, and has the engine and carburetor controls carefully adjusted to give that minimum consumption -- will consume, for each mile flown, a weight of fuel and oil equal to just a little less than 1/10,000 of its own weight. That is to say, an airplane weighing 10,000 pounds, and otherwise meeting the specifications laid down in the previous sentence, will consume just a little less than one pound of fuel during its first mile of flight after having climbed to the intended operating altitude and settled down to steady cruising at the economical speed. As time goes on, however, the amount of fuel consumed for each mile of flight will be reduced, since the weight of the airplane is itself constantly being reduced by the amount of the fuel consumption. It therefore does not need twice as much fuel to fly 4,000 miles as 2,000, but only about 1.8 times as much.

The results of actual calculations on this point are shown in Figure 2, where the total weight of fuel and oil required for a flight, stated as a fraction of the total weight of the aircraft at the time of starting, is plotted against the length of flight. The chart contains two curves. The lower of the two, showing minimum consumption, represents the very best that may be considered attainable in the present state of development of aeronautical engineering, assuming the conditions presented by the design specification to be the most favorable possible. This lower curve thus represents the apparent limit of present attainment. The upper curve represents the actual current probabilities for actual military aircraft, which must have turrets and other irregularities of form, with engines and propellers chosen for best performance at maximum power output as well as for consideration of extreme economy. Transports are better off than bombers, not having projecting armament and turrets. The fuel consumption of a transport plane designed today might be expected to lie near to the lower of the two curves.

These figures are so significant that I think I should supplement the curves by a table setting forth a few selected values.


Length of Flight in Total Weight of Fuel and Oil Consumed
Still Air (as percentage of total weight of the aircraft at time of take-off)
  Best possible Most probable for
(miles) at present bombing aircraft
1,000 8.3 10.9
2,000 15.9 20.6
3,000 22.7 29.2
4,000 29.1 36.9
5,000 34.9 43.8

In planning a 3,000-mile flight for a specific military or commercial purpose, it is not enough to provide the amounts of fuel and oil that would be consumed in the course of a cruise of 3,000 miles in still air. Even though it is anticipated that, on the average, the wind will have no effect, weather forecasts are never perfectly accurate and unforeseen headwinds may develop. In the case of long commercial flights over water, where the very highest degree of safety must be maintained, it is customary to allow a reserve sufficient to take care of an unforeseen headwind of at least 30 miles an hour. Reserves must also be allowed for errors in navigation, and for the possible necessity of deviating from the course to fly around storms. Again, bombers have to forsake their economical speed, and fly at close to their highest possible speed, when passing through areas strongly defended by anti-aircraft fire or fighting planes; and they have to spend some time, and a corresponding amount of fuel, in manœuvring around their target when they reach it. Then there is the fuel consumed during the take-off and climbing to cruising height. This is roughly equal to the amount required to fly an extra 50 miles at cruising speed.

Taking all this together, we reach the rough general rule that a bomber should carry enough fuel and oil to fly at least 21/2 times the distance from its base to the given target in still air and at the most economical speed. A fighter which has to make a trip to enemy territory and possess enough reserve fuel to perform any useful service on arrival would certainly need a total supply (figuring on a still-air flight basis) for at least three times the distance from its base to the point over which it was expected to engage in combat. The total reserve of fuel carried in transport planes on long non-stop flights is from 20 to 50 percent of the amount of fuel required for covering the total distance of the flight in still air, the exact reserve depending on the wind conditions existing and forecast at the time of take-off. The distance from California to Honolulu, for example, is almost exactly 2,400 miles. The upper curve in Figure 2 would indicate that a conservative expectation of the fuel requirement for a flight of that length would be for 24 percent of the initial total weight of the aircraft. Actually, the Pan American Clippers which fly that route take off with a total maximum weight of 86,000 pounds, and carry a load of fuel and oil ranging from 24,000 to 29,000, or from 27.9 to 33.7 percent of the total take-off weight.

Thus the airplane that sets out from England to bomb Berlin must start its flight with about 18 percent of its total weight in the form of fuel and oil. A fighter accompanying it as a defensive escort would need to carry at least 25 percent of its own initial weight in consumable supplies. To turn from military to civil operations, an airplane that is to fly non-stop from England to the United States with passengers or cargo, with a safe reserve for coping with headwinds, should have a little over one-third of its total weight in the form of fuel and oil at the time of take-off.


In climbing, two factors are important -- the ceiling, or maximum height to which the climb can be carried, and the length of time taken to reach various altitudes along the way. A fighting plane's ability to get above the enemy fighters gives an enormous advantage in high-altitude combat, as the pilot who has felt the gnawing anxiety of being outclassed in that particular very well knows. Only a little less important is the ability to climb more rapidly than the enemy, and so to gain superiority of altitude after having started on the same level.

The main features governing ceiling are critical altitude of the engine, power loading and wing loading -- in that order of importance. Modern fighters have an absolute ceiling of between 35,000 and 40,000 feet. For every 1,000 feet of increase in the critical altitude, or altitude up to which the engine is able to produce its full rated power, the ceiling is raised, other things remaining equal, by about 700 feet. An increase of wing loading from 30 pounds per square foot to 40 (other things again remaining unchanged), cuts the ceiling by approximately 2,500 feet -- part of the price that has to be paid for the clipping of wing surface in the interest of high speed. A 20 percent reduction in power loading, by an increase of rated power output or a reduction of total weight, is good for nearly 4,000 feet of additional ceiling. Thus an increase of 5,000 feet in critical altitude of the engine is worth nearly as much, in terms of ceiling performance, as the combined effect of a 20 percent increase in total power and a 20 percent increase in wing area, both secured without adding to the total weight of the airplane.

Whether or not future aerial combat will be carried on at levels still higher than any yet attained depends mainly upon the efforts and the successes of the supercharger designers. One of the greatest services of American research and engineering in the preparation for victory in the air may be found to lie in the introduction and improvement of the type of engine supercharger which is attached to the engine as an accessory. It is driven by the action of the exhaust gas upon a turbine instead of being operated directly from the crankshaft through a train of gears in accordance with the practice universally followed in European design.

The influence of power loading is paramount in determining the rate of climb of fighters at altitudes below the critical. This is a little less true in the case of bombers, for there the wing loading appears as a factor of some significance, a high wing loading being unfavorable to a rapid climb. A fighter with a wing loading of 30 pounds per square foot and a power loading of 6.2 pounds per horsepower, a combination which has already been presented as offering a maximum probable speed of 300 miles an hour at sea-level and 360 at 20,000 feet, may at best be able to climb to 10,000 feet in about five minutes. A reduction of power loading to 4.6 pounds per horsepower, keeping to the same wing loading, would increase the speed by only about 10 percent, but reduce the time needed for a climb to a given altitude by more than a quarter.


The facility with which an airplane leaves the ground, though apparent at only one instant of each flight, yet may operate indirectly to control both range and speed. For it determines the size of the field needed for take-off, and so determines the readiness with which new fields of suitable extent may be found or hastily constructed in mountainous or otherwise difficult terrain. It also influences range by limiting the maximum weight that can be carried from available fields. A machine capable of carrying enough fuel for a 4,000-mile flight, but needing a 5,000-foot field from which to make a take-off with that amount of weight, may have its extreme range reduced to 2,000 miles if only a 3,500-foot field is available. It influences speed by imposing limitations on the increase of wing loading. Finally, enemy action may have a more disruptive effect if the take-off area is large and smooth, for then the enemy has a better chance of rendering some part of it unusable by bombing attacks.

Wing loading has more effect in determining the length of take-off than it has on any other characteristic except landing speed; and the general tendency to increase of wing loading in recent years has brought with it a substantial change in the established idea of what constitutes a proper field on which to base either military or civil air operations. Ten years ago, a field 3,000 feet long seemed generously proportioned. Now, with wing loadings up to double the figures then prevailing, 3,500 feet is regarded as an absolute minimum for any serious operations, and paved runways 4,000 or even 5,000 feet long are considered desirable.

What we call "take-off distance" is affected both by wing loading and power loading. A 10 percent increase in power loading normally increases by about 12 percent the total distance required to clear the ground and reach a height of 50 feet. A similar increase in wing loading typically increases that distance by about 8 percent. A simple increase of 10 percent in the weight of an airplane, by overloading, without change of engine or wing surface, increases the length of the field needed for take-off by some 20 percent.

Take-off distance is no great problem in the case of the modern fighter plane, which needs hardly more than 1,000 feet in order to be able to reach a 50-foot altitude. But the distance which it requires to approach the ground and come to rest after landing is much longer; and it is this requirement which determines the size of field adapted to fighter operations.

For the bomber the reverse is true. Both its power loading and its wing loading are higher than the fighter's, and the take-off is correspondingly longer. On the other hand, when it comes to land it generally is light, having finished a long flight and dropped its bombs, so that the landing run required is correspondingly shorter. The distance which it needs in order to reach a 50-foot altitude under test conditions is likely to be between 2,000 to 3,000 feet; and in order to provide a proper margin of reserve the length of the field from which it operates, taking account of the effects of high temperature and other unfavorable weather conditions, and of variations in piloting technique and in the mechanical condition of particular airplanes, ought to be from 3,000 to 5,000 feet.

There has been much discussion, over many years, of the possibility of solving all such problems at a single stroke by catapulting the airplane into flight, as is done on shipboard. To catapult a large machine is difficult mechanically. More important, it would be of practical use only where flights of great length were to be made and where it was desired that a very high wing loading be employed; or in cases (analogous to that of shipboard operation) where it was impossible to find or prepare a large enough and smooth enough field to permit normal take-offs, but where the means exist for setting up a catapult. So far, no sufficiently portable catapult has been developed; but this may conceivably be done.


The designer's problem of where to allocate weight begins with the empty airplane. It begins with the airplane structure, the engine and propeller, and the indispensable accessories and equipment.

The weight of the airplane structure -- wings and body and tail surfaces and landing-gear -- will be determined largely by the use to which the machine is to be put. The structure of a bomber or long range transport which is not intended to operate at very high speed or to manœuvre at all violently may weigh as little as 25 percent of the maximum total weight carried in actual flight. From 28 to 32 percent, however, is the more typical proportion. Fighters, which must be strong enough to stand the manœuvres as gruelling as long dives followed by quick pull-ups, devote as much as 35 percent of their weight to structure.

The relative importance of the engine and propeller, as a fraction of the total weight carried, depends largely on power loading. The weight of the engine is fixed; and if the total weight of the airplane and all its added load is to be 5 pounds per horsepower, the engine weight will obviously be a much more considerable proportion of the total than it would have been if the total weight had been twice as great and the horsepower had remained the same.

The bare engine may weigh hardly 1.0 pound per horsepower; but the propeller, and starter and other engine accessories, and the cooling liquid in the cases where a liquid-cooled engine is used, increase the total weight of a present-day power-plant to a minimum of from 2.0 to 2.2 pounds per horsepower. The instruments and controls, and radio and navigating equipment, commonly take up about another 4 percent of the total weight. The weight of the crew averages about 3 percent. In a military plane, furnishings such as seats, and mechanical systems for heating, ventilating, and sound-proofing, and miscellaneous equipment such as fire extinguishers, take up about 3 percent of the weight. In a transport, which has elaborate sound-proofing arrangements and luxurious seating equipment for passengers, this figure must be increased by more than one-third the weight of the maximum number of passengers for whom provision is made. The amount left after all these items have been taken care of can be divided between armor and armament or commercial load, on the one hand, and fuel and oil on the other. The maximum percentage of the initial weight that can be expected to be available for these purposes may be tabulated as follows for three different classes of aircraft:


Power Loading at Maximum Percentage of Maximum Initial Weight Likely to be Available
Weight for Fuel and Oil and Military or Commercial Load
  Fighter Bomber Transport
(pounds per horsepower) (liquid-cooled engine) (air-cooled engine) (air-cooled engine)
4.5 6 -- --
5   11 -- --
6   18 26 20
8   27 36 28
10   -- 42 32
12   -- 46 35
15   -- 50 38

These figures (and those on following pages) differ from previous tabulations in representing typical, or probable, rather than best possible performances. Thus there have been few, if any, instances of aircraft attaining speeds in excess of those plotted in Figure 1, or using less fuel for a given length of flight than the amount indicated by the lower curve in Figure 2; but there have been numerous instances of improvement over the percentages of useful load just tabulated. The figures given, however, are representative of good practice, and are not likely to be improved upon materially.

An airplane with a power loading of 4.5 pounds per horsepower and a wing loading of 50 pounds per square foot, designed with the closest possible attention to aerodynamic refinement, might be able to make 380 miles an hour at sea-level and 460 at 20,000 feet; but it appears that such a machine would be able to carry fuel for only an hour of flight, and no guns, ammunition or armor at all. With the power loading increased to 5 pounds per horsepower, and with the wing loading reduced to 40 pounds per square foot in order that the ceiling and the take-off qualities may not be too adversely affected by the increase in power loading, the anticipated speed would be reduced to about 350 miles an hour at sea-level and to 420 at 20,000 feet. The residual weight available for military items and fuel would be 11 percent of the initial total weight. Now in order to meet present ideas regarding protection for the pilot and certain vital parts of the mechanism, the weight of armor in a fighter ought to be about 5 percent of the total weight of the airplane. The weight of guns and ammunition on such a machine, for really adequate fire-power, ought to run to 8 or 10 percent of the total weight, although of course a usable military airplane can be equipped with much less armament. For a bomber, the total weight of armor, guns, turrets, ammunition and bomb sights and racks may be expected to average about 12 percent.

Correcting for all these items, and assuming a reasonable variation of wing loading with power loading in the interest of maintaining adequate ceiling and take-off performance, the range and bomb-load (or, in the case of a transport machine, the payload of passengers and cargo) can be determined in terms of speed. Let us take first the case of a fighter, fully armed and armored according to the standards set forth in preceding paragraphs. Here the relation between the maximum speed and the range (at most economical speed) for which it would appear practicable to design an airplane at the present time would be:


Maximum Speed Maximum Range for which
at 20,000 Feet Fuel Could be Carried
  (in miles)
300 1,700
350 1,000
400 400

By reducing the weight of armor and armament by one-half, the airplane's range in any of these cases could be extended by some 600-800 miles, and its effective operating radius when used as a defensive escort could be increased by about 250 miles. But this obviously would seriously diminish its usefulness as a fighting machine.

There has been a great deal of talk in the last two years about the desirability of developing long-range fighters. A fighter adapted to long-range operations is inherently handicapped in combat with airplanes operating close to home, and therefore able to put into additional armament or additional power-plant the portion of the total weight which the long-range machine must put into fuel. There is no way of overcoming this handicap completely. However, it can be kept to a minimum by heavily overloading the machine on the take-off, which can be done safely if the pilot is sure of not having to undergo the strains of combat at that stage and of not having to climb to a very high altitude until the flight has progressed far enough to have substantially reduced the fuel load.

There follows a similar computation for bombers. It shows the distances to which bombers designed for various maximum speeds may be expected to carry out effective raids carrying bomb weights equal to 10 percent and 20 percent, respectively, of the initial total weight of the aircraft.


Maximum Speed at Approximate Maximum Distance (in miles) at which Effective
20,000 Feet Bombing Raids Can be Carried Out
  Bomb Weight of 10% of Total Bomb Weight of 20% of Total
200 1,500 1,100
250 1,100 800
300 900 550
350 600 200

This calculation has been made for a bomber of approximately 40,000 pounds weight. The relative importance of the weights of armor and armament, of furnishings, equipment and radio, all decrease gradually as the size of the airplane increases; and the larger bombers also show somewhat better relative fuel economy than the smaller ones of equal power loading and wing loading. For a 25,000-pound bomber the smaller figures in the tabulation should be reduced by about 150 miles; the larger figures, at the top of the columns, by about 250. Increases in weight to above 40,000 pounds, on the other hand, probably would not increase the tabulated values very appreciably; although there might continue to be some slight gain up to a weight of about 100,000 pounds.

As persistent as the dream of a long-range fighter which will secure range without paying the price of sacrificing other elements of performance is the dream of a bomber which will out-speed the defending fighters, or out-gun them in direct combat, or serenely soar above them at altitudes which they are unable to reach. Barring the possibility of some new invention or great improvement in design which the enemy for the moment cannot match, this will remain a dream. Other things being equal, an airplane will always be able to obtain greater altitudes in a light condition than when it is heavy; and after the ideal bomber has been imagined, it is only necessary to imagine it modified by leaving off its weight of bombs and replacing them with half their weight of additional guns and ammunition, in order that there may result a machine with more fire-power than the bomber, greater climbing capacity, and at least as much speed. The airplane that accepts the handicap of carrying bombs will always be at a disadvantage as against the machine that carries only offensive armament and the crew to operate it. The airplane specialized for a single function will continue to maintain superiority in that function over the machine designed with some other objective, or with a multiplicity of objectives.


In reckoning future performances, one can say that the largest single promise of gain lies in the possibility of fundamentally modifying the section or form of the wings. This might so change the flow of air about the wing as to reduce its resistance by as much as a half. By that single change, and the others that would naturally accompany it without requiring any new invention on their own account, speed might be increased by as much as 15 or even 20 percent. The principal barriers to further improvement of performance, on the other hand, lie in the general increase in air resistance with increases of speed above 450 miles an hour or thereabouts, and in the rapid decrease of efficiency of airplane propellers as the speed of the aircraft increases above about 400 miles an hour. Propeller design for extremely high speeds presents one of the problems on which research is most needed.

Limitations of propeller efficiency, in fact, threaten rapidly diminishing returns as we seek to develop further the process which has brought a large part of the gains in fighter performance in the last 20 years -- namely, steady increase in power. In 1918 the largest engine used on any single-seater fighting airplane developed about 300 horsepower. By 1930 that figure had been doubled. Now it has been doubled again; and it may be very nearly doubled once more in the comparatively near future, with engines of 2,000 horsepower or more coming into use. The problem remains, as I have emphasized at such length, one of weight distribution, and of relative power rather than absolute power. But when the engine power of a fighter is doubled, the weight of the power-plant does not increase in quite so large a ratio; neither does the weight of the structure; and the weight of some of the furnishings and equipment items, and of the pilot, does not increase at all. It therefore becomes possible to reduce the power loading as the engine power is increased, and so to increase the speed. However, in increasing the power of a fighter from 1,000 horsepower to 2,000 the total weight of the machine is likely to be increased by at least 60 percent; so that the resulting increase in speed, even if the limitations of propeller efficiency were less controlling than they now are coming to be, would be only about 10 percent.

[i] Mathematical expressions and engineering formulas are avoided in this discussion. Those who are interested in the methods by which the present conclusions have been developed, and in the discussion of the ways in which performances are affected by variations in a considerable number of design factors, will find those matters treated in the writer's "Airplane Design-Performance."

You are reading a free article.

Subscribe to Foreign Affairs to get unlimited access.

  • Paywall-free reading of new articles and a century of archives
  • Unlock access to iOS/Android apps to save editions for offline reading
  • Six issues a year in print, online, and audio editions
Subscribe Now
  • EDWARD WARNER, Vice Chairman of the Civil Aeronautics Board; Assistant Secretary of the Navy for Aeronautics, 1926-1929; Editor of Aviation, 1929-1935; former Professor of Aeronautical Engineering at the Massachusetts Institute of Technology
  • More By Edward P. Warner