In the early stages of torpedo practice, in fact not infrequently in recent times, it often happened that erratic performances of this weapon were looked upon in mystery, and the only explanation offered was that the torpedo was a freak. Fortunately the recovery of most of these abnormal torpedoes enabled the personnel in charge to find out the definite cause of troubles and to devise some remedy against similar occurrences. It was also found that certain precautions had to be taken in the care, preservation and adjustment of the torpedo to insure against the wreckage of the entire apparatus and that if these precautions were observed the torpedo behaved properly; if not an erratic performance ensued.
The development of aeroplanes and the operation thereof has progressed along similar lines, with the disagreeable exceptions, however, in that the personal equation is introduced in the latter and that the wreckage of recovered machinery is usually too confused to make the solution of the trouble apparent. There is an authentic instance of a so called freak performance of an aeroplane after which the pilot, who had luckily succeeded in righting his machine, foreswore aviation on the ground that some machines were uncanny. As a matter of fact his machine had acted only naturally, just as the gyro acts, in "tumbling" sometimes, naturally obeying the laws of precession. The occasional escape of a pilot from aeroplane accidents affords some account of what took place in these accidents. The experiences of the actual-flyer together with a conservative application of the laws of natural philosophy, experiments upon strength of material, wind and gust stresses etc., have given a clear idea of the phenomenon of flight, so that the accidents nowadays which throw mystery into the popular mind, are not difficult to explain and are by no means unavoidable. As aviation stands now the precautions to be taken before the flight of an aeroplane are as definitely known as those prescribed for torpedoes, and were it not for errors in piloting, casualties due to material would be few indeed and fatalities to personnel even rarer. Obviously aeroplane accidents are attributable to three general causes, namely:
(1) Air conditions.
(2) Material.
(3) Personal error.
In the first stages of aeroplane construction when the factor of safety was small and the machines were slow and under powered, casualties from causes Nos. x and 2 were frequent. As frequent as they may have been, however, it is more than probable that the majority of casualities were due to personal errors rather than to weakness of material or atmospheric conditions. In assuming, however, that the majority of aeroplane accidents are due to general cause No. 3, it is not to be inferred that accidents are practically eliminated, for there is no factor more variable than the personal factor and none more vital in the operation of aircraft. Unfortunately for both military and naval aviation, the War and Navy Departments have been forced to purchase only such material as was obtainable in this country, and this too in limited quantities.
The science of aeroplane construction in the United States appears, however, to be improving, especially in one or two companies, but up to date it has been patent to all flyers, who have had the opportunity to observe, that the strength, design, workmanship and finish of foreign material are vastly superior to anything we have. This fact must therefore be taken into consideration in mentioning such accidents, due to material, that may be hereinafter enumerated.
ACCIDENTS DUE TO AIR CONDITIONS
In a machine of a given weight, lifting surface and coefficient of fineness, there is a definite minimum velocity relative to the air from which sustentation can be gained. Corresponding to this speed there is but one angle of incidence at which the machine can fly. This speed is known as the lower critical speed in order to distinguish it from the other critical speed existing at the high-speed end of the speed incidence curve. Now, up to 1913-14, most of the machines built in the United States, especially the Wright, were of slow speed. Some of these types were slow due to design and others to the fact that the motors were underpowered. The range of safe flying speeds was therefore much reduced in both instances. As a consequence it made little difference whether the machines were flying at their normal incidence or at the critical speed, the velocity of air under the lifting surface was so small that the decelerations encountered in disturbed air, however small they may be, were often sufficient to reduce the relative velocity below that necessary for sustentation and dynamic control; and the machine fell. The mere fact that the speed of old time machines was slow does not necessarily mean that gust actions in such cases would probably be fatal. The serious point with the old slow speed machine is that the range of safe speeds is small and that its normal speed is essentially close to its critical speed. Another serious difficulty is that in the slow speed machine the stabilizing surfaces lack the dynamic righting moment possessed by such surfaces in high speed machines. Hence it has frequently happened that machines have been wrecked by gusts, not because the gust deprived the machine of control but that its righting moments were too slow to enable the pilot to avoid obstacles which chanced to be near. In many of the worst aeroplane accidents, which were obviously due to gust action, the personal error was so great due to the fact that the pilot was ignorant of the machine's critical angle, that, had this error been eliminated, the gust force alone would have been unable to upset the machine's equilibrium. It is thus seen that causes of accidents which are often attributed to gusts, are so closely related to personal errors that it may be a mooted question where to place the blame. The motor may be gradually failing; the pilot may be ignorant of this fact when a gust suddenly robs the machine of control and sustentation; and wreckage ensues. It has sometimes happened that sentiment for the pilot has thrown the burden of the accident upon the precipitating cause; and gust-action is accordingly classified as one of the great causes of aeroplane accidents.
Gusts upon aeroplanes may be classified under four headings although they frequently attack a machine in a combination. The general effects upon an aeroplane, in so far as accidents are concerned, may be studied in any one of the four headings:
(1) Rear gusts.
(2) Downward gusts.
(3) Head -gusts.
(4) Upward gusts.
To avoid confusion in studying gust action and effect upon an aeroplane, it must be continually borne in mind that the stability of a machine depends upon gravity relative to the air rather than upon the absolute gravity. It must also be remembered that in the problem of stability, it is the machine's path relative to the air, not its absolute path, that must be taken into account.
The foregoing enumeration is in order of relative importance to the machine's stability. The downward gust, which is severer in effect than the rear gust, has not as many accidents to its charge as the rear gust for the reason that downward gusts are, as a rule, encountered in very disturbed air, under which conditions attempts at flight are infrequent, while the rear gust frequently occurs when air conditions are apparently normal.
ACCIDENTS DUE TO REAR GUSTS
The simplest explanation of this phenomenon is that the aeroplane flies suddenly into a region of air being accelerated in the direction of the machine's path. Were the air acceleration strong enough, the direction of air flow through the wings would be momentarily reversed, and the machine would drop precipitously. There is however, no record of air currents being accelerated to such an extent, no matter what velocity the wind may have. Experiments have found few air accelerations beyond fifteen miles per hour, though it is believed that in electric and cyclonic storms they are much greater. What ordinarily happens is that the velocity relative to the aeroplane is momentarily reduced and, the machine's inertia preventing an immediate absorption of the gust's speed, lift and control are lost. The old popular idea of an aeroplane "getting a puff under the tail" or of a "puff coming up from behind" is quite incorrect. Even with very slow aeroplanes it is more than improbable that a gust could overtake them. The term "rear gust" is therefore a misnomer and no doubt came into use during the early stages of aviation when gust action was not understood. The aeroplane in rear gusts noses downward, throwing the tail upward, thus giving the idea that a gust has come up under the tail. An upward gust striking the tail surface only would, to an observer on the ground, appear to effect the machine similarly to a rear gust, but the "feel" of the machine would be different in that control would not be lost. If the machine be close to the earth when a strong upward gust strikes the tail surfaces there would be danger of the nose striking, but with seaplanes, it would probably mean nothing more than porpoising and certainly not the incorrigible "nose-dive" resulting from rear gusts.
The effect of a rear gust upon an aeroplane may possibly be better understood when the consideration of air velocity is disregarded and the machine is assumed to be sustained in normal flight. With these premises, any variation in the conduct of the machine must be due to a variation of "gravity relative to the air."
It is very important that this latter expression be thoroughly understood. When a machine is being sustained in undisturbed air, that is, air whose velocity is undergoing no accelerations or decelerations—gravity relative to the air and absolute gravity are the same. It must be borne in mind that gravity is nothing more than a downward accelerating force. Now if the supporting medium be suddenly accelerated, the machine's equilibrium must be momentarily adjusted to the resultant of the two accelerations, gravity and the gust. This resultant is the new gravity relative to the air, the direction of which is perpendicular to the earth only when the air acceleration—gust—is directly upward or downward.
Now let it be considered that the aeroplane is sustained in flight at the point A, flying in the direction AB, when it is attacked by a rear gust of force CA.
Now it must be further borne in mind that the reason air accelerations are considered gusts is that the inertia of the aeroplane prevents an immediate assumption of the gust velocity. Therefore if the air is being accelerated in the direction CA relative to the aeroplane, then the aeroplane is being accelerated in the direction of AC and with the force AC relative to the air. Thus the machine is simultaneously subjected to two accelerations AC and AD, or the resultant AE. Therefore, the virtual horizontal is AF, at right angles to virtual gravity AE. Now it was assumed that the aeroplane had just sufficient power to sustain it in the horizontal with reference to AD. With the new gravity AE, the machine attempts to adjust itself to the virtual horizontal AF. But gravity has not only changed in direction; its force also is increased. The machine therefore drops along lines anywhere between AKK' and AHH', according to the duration of the gust, the inertia of the machine and the damping effects of its stabilizing surfaces. An inspection of the paths AHH' and AKK' will disclose the improbability of recovery by a machine (without reserve power) when attacked by a violent rear gust at a low altitude.
ACCIDENTS DUE TO DOWNWARD GUSTS
The fatalities due to downward gusts though not as numerous as those due to rear gusts, have been frequent, while injuries of a less serious nature, both to materiel and personnel, are common. As stated before the reason that casualties resulting from the downward gust have not assumed greater proportions is that, heretofore, flights were infrequently attempted when this type of gust was prevalent.
The downward gust is doubly dangerous, as compared with the rear gust, both on account of the dangerous absolute path of the machine and its critical incidence to the air when the gust is past. It has often happened in exhibition flying that the pilot, while attempting to trim a pylon or the corner of a building closely, has been thrown down upon the very object which he was sure of clearing. There have been several fatalities on hot aviation fields (due to turbulent air rising therefrom) when the aeroplane, at an altitude of a few feet after rising or within a few feet of landing, was smashed bodily to earth.
With reference to the aeroplane's conduct during a downward gust and its attitude relative to the air after the gust is past, it would appear at first glance that in these respects its condition would not be serious. An examination, however, of the relative gravity diagram will reveal the hazardous position of the machine (Fig. 2).
Assume A the position at a given instant flying horizontally in direction AB, when struck by a downward gust of force CA. The acceleration of A relative to the air is AC, opposite to the direction of the gust. Therefore A is subjected to two accelerations, opposite in direction AC and AG, which is absolute gravity. Thus gravity relative to the air is momentarily reduced, and, if the gust be strong enough, relative gravity is reversed. The path of the machine relative to the air is in the direction AF or AE according to its damping qualities. The absolute path however, is in the direction AL or Al, which illustrates the double danger due to attitude relative to the air and to the earth when the machine is at low altitudes. The downward gust accounts for the phenomenon observed in very disturbed air when the machine seems to be struck with such violence as to make a small wrench or pliers jump up from the pilot's seat.
To the downward gust may be attributed many stalls that are ordinarily assumed to have arisen from other causes and it has sometimes caused the smash of land machines when they have been thrown bodily to the earth just after making the "get-a-way."
ACCIDENTS DUE TO HEAD GUSTS
Assuming the machine to be in normal flight, the encounter of a head gust momentarily increases the velocity of air under the lifting surfaces, thereby increasing the lift under the machine. The latter immediately starts to climb. When the gust is past, the machine being pointed upward retakes the undisturbed air at a larger angle than it can fly. Hence a stall ensues. The dangers of head gusts are fewer and smaller than those attributed to the foregoing causes, however great may be the dangers resulting from a stall. Stalling may be brought about in so many ways—by personnel, material, or air conditions—that it will be treated later under a separate heading. Since the head gust may cause a stall, it is classified as an air condition that may cause aeroplane accidents. However, with an aeroplane possessing a normal reserve power a pilot of fair skill may employ the pure head gust quite advantageously in an effort to make a rapid climb. Sufficient practice will enable him reflexively to push down the elevator towards the end of the gust, decreasing the incidence as the deceleration of the gust increases it, thereby maintaining the aeroplane at an incidence for which it is designed.
Very few accidents, if any, are due to the upward gust. Discussion of this feature of air conditions is therefore unnecessary.
ACCIDENTS DUE TO RUPTURES OF MATERIAL
Fatalities due to material unquestionably have been frequent. This is true in the development of any new method of locomotion, as well as any particular design of vehicle new to its own specie. Many ruptures have regrettably been due to the willingness of the pilot to take a chance when he knew of the inherent weakness of the structure upon which he had to depend or when he was culpably ignorant of faults in the machine which would have been patent to a careful observer. There are instances on record where certain exhibition flyers, limited in finance and inspired by competition, substituted an old fence picket for a cracked strut or renewed a stranded control wire with a piece from around a bale of hay. It is needless to mention the resultant casualties. A notable example of folly in disposition of aeroplane material recently occurred when the drift wires were reduced to such an extent that it was impossible for the wing to withstand the stresses due to high speed; and total collapse followed.
There has been, however, a steady and rapid improvement in material, and fatalities, marked against material as a primary cause, are quite frequently due to the locality in which the aeroplane is forced to land. At the Aeronautic Station casualties in material are still numerous, principally on account of the power plants. Among the incidents mentioned below escapes of the pilot have been due to only good luck. The following are a few instances:
(a) Crank Case Explosions.—This has occurred several times. Debris of the motor shot through the planes at several points, cracked. struts and narrowly missed the pilot's head. These explosions were due to poor material in cylinders, in pistons and piston rings, which allowed an explosive mixture to accumulate in the crank case and finally to ignite from the regular explosion within the cylinder. Remedies for such faults suggest themselves.
(b) Breakage of Piston Heads and Walls.—This has happened a few times without disabling but one cylinder. At other times it has resulted similarly to crank case explosions. The pistons appeared to be cast from poor material and to have been poorly and unevenly finished. In fine there appeared to be no necessity for having such a piston although the motor containing it was the best that could be obtained on the American market. Government inspection at aeroplane factories promises to relieve this difficulty.
(c) Ruptures of Chain Drive.—Very few aeroplanes, other than the Wright, employ chain driven propellers. The Wright Company employs two chain driven propellers, turning in opposite directions. This latter feature requires one of the chains to be crossed. Frequent casualties both to material and personnel have been caused either by the chain drive or by the cross in the chain. In some instances the accident has resulted from the chain jumping the sprocket: in others, from the propeller supports weakening, with similar results. A few of these accidents have occurred so close to the ground that by cutting the spark the machine could not be righted from the sudden skewing effect of one propeller. In one instance it was just as well that the machine was close to the water, since the motor's rolling up the cross chain put the warp out of commission, and equilibrium was impossible. The location of the motor forward of the pilot's seat, in the last named accident, was the only thing that saved the personnel from being pulped. Upon impact with the water the motor was thrown out through the bow and was recovered by one wire that did not break.
The undesirable feature of chain drive is as patent in the design of other vehicles as in aeroplanes and will no doubt be eliminated in the future. Closely allied with chain drive in disadvantages is the poorly stranded control wire with its numerous sheaves and fair leads. Many accidents in the early stages of aviation were due to these latter causes, and one may be cited recently in which the parting of an aileron wire caused the wreckage of a machine. Weakly stranded control wires, sheaves and fair leads appear about as antiquated as chain drives. It is expected therefore that either bell-crank sectors with heavy cables or control rods and levers will soon relieve the devices less safe.
(d) Casualties Due to Propellers.—With military aeroplanes propeller ruptures have been few, except when struck by a bullet or shell. The seaplane, on the other hand, has had and will have a harder proposition. Water being 800 times denser than air, the air propeller fares badly even when struck by spray. Up to the present time aeroplane propellers have not been designed, on account of weight and inertia, with a view of withstanding the shock of a wave or heavy spray. As a consequence the propellers furnished our seaplanes have to be resheathed. Before the necessity of such reinforcement was realized some very serious casualties occurred—and this quite recently. In one instance the sheathing of a blade, shaken loose by impact with spray, flew off while the machine was in the air. The high rotational velocity of the propeller threw the sheathing out with great force and cut one of the tail bamboos cleanly out. At the same time the sudden unbalancing of the propeller brought such violent stresses upon the power plant that the sheathing of the other blade was also thrown off, the engine beds cracked and the motor put generally out of commission. Fortunately for the occupants, one wire from the cut bamboo escaped, and this wire carried the tail load during the glide to a safe landing.
For military aeroplanes the tractor type appears to have eliminated dangers from propeller ruptures. For seaplanes the tractor type will relieve the pilot of dangers incident to propeller ruptures, but whether tractor of pusher, in seaplanes, the propeller must be mounted with a good clearance from the water and must also be strongly sheathed in order to withstand the spray that will be encountered now and then. On account of its offensive and propelling efficiency the pusher type of aeroplane is preferred to the tractor. It may be, therefore, that other precautions against the dangers of propeller ruptures will be sought in the former type rather than upon the assumption that the propeller will have to smash sooner or later and should accordingly be upon a tractor mount. In all events it is evident that casualities due to propeller rupture will be of a serious nature, whatever the type of the machine: and thus far only theory, not practice, has found a remedy.
(e) Accidents Due to .Ruptures of Gasoline Tanks or Leads Therefrom.—There are many cases on record where ruptures of the gasoline tank, exhausting the fuel, have forced landings in undesirable places. There are several instances where back fires from the carburetor have caught in the drip from a gasoline leak and burnt the machine completely. At the Aeronautic Station there have been two notable instances of casualties resulting from gas lead ruptures. In one of these the machine caught fire in the air. The pilot succeeded in reaching the water before the machine was consumed. The exact cause of this accident—what part broke first is not known—but it is more than probable that the rupture was due to the excessive vibration of the Curtiss power plant. In the second of the two accidents mentioned above, the after part of the gasoline tank broke off bodily and struck the propeller. Both blades were broken off close to the hub and the debris of blades and copper were thrown through the right upper wing (Dunne biplane), cutting a strut, cracking the rear spar and tearing out five square feet of surface. The breakage of the propeller unbalanced the motor so suddenly that, before the spark could be cut, the engine beds were badly broken. Luckily the design of the machine (Burgess-Dunne) was such that the debris hurled by the propeller did not cut the controls, and a glide from 1000 feet altitude was made to a safe landing. Several small slivers had, however, punctured the central float, which began filling rapidly, and the machine was saved only by the prompt assistance of the rescue boat.
There are numerous other possibilities of casualties to an aeroplane which, in times of peace, might not prove serious either to materiel or personnel, but which in war would frustrate the machine's mission as effectively as a crank case explosion. Use and vibration effect details of a machine's structure where a defect is least expected, and not infrequently does it happen that some small piece of the power plant or machine's structure, such as a lock washer or the head of a small bolt, is thus broken, which, "per se" is of little importance but when struck by the propeller may cause irreparable damage. As previously pointed out, however, the tractor eliminates the immediate danger of such trouble.
When the subject of material is treated in detail, it becomes an easy matter to enumerate the many things that can go wrong. But taken as a whole, the obtainable aeroplane material is good and is improving daily. A rigid, systematic inspection of adjustments appears to be the greatest requisite. With this requisite fulfilled the percentage of successful flights, despite the counterweight in the risk and loss of human life, is much higher than the percentage of successful torpedo runs.
ACCIDENTS DUES TO PERSONNEL
Accidents due to personnel, while varied in cause and in form, may be generally classified as follows:
1. Bad getaways.
2. Bad landings.
3. Side-slipping.
4. Working into the vortex of a spiral.
5. Stalling.
The above enumeration is made in the inverse order of importance. Bad getaways and bad landings are by far the most frequent, and, while hard on the aeroplane, they are less frequently fatal to the pilot. Except where forced landings bring the aeroplane down into exceedingly rough country or into local conditions which a pilot would avoid, bad getaways and bad landings are the beginner's faults, which a reasonable amount of practice should overcome. The commonest feature of these two faults is known as "porpoising." Whether in getaways or in landings it is usually caused by holding the boat too much down by the nose so that the reaction between the water and the curvature of the bow forces the machine into the air, when the speed of the air relative to the wings is insufficient to give sustentation. The machine being in a stalled condition drops to the water, nosing down again according to the phenomenon of stalling. This latter conduct causes a repetition of the performance, two or three times if the motor is dead, and almost a continuous performance if the reserve power is poor and the water rough, and the elevator badly handled. When the water is rough the porpoising is more violent and the machine's headway more reduced by impact with the slope of waves. It is for this reason, when the reserve power is weak, that it requires a judicious use of the elevator to recover from the porpoised condition; otherwise it would be better for the beginner to throttle down and abide the consequences of the second or third porpoise, rather than hold the throttle open and continue, by improper use of elevator, to pound the machine into the waves. When the aeroplane is equipped with a motor of good reserve power, the effects of porpoising are almost nil, since the large thrust of the propeller is sufficient to recover flying speed before the second impact with the water. It is probable that the most serious effects of porpoising are found during subsequent flights when the machine's strained condition permits of ruptures which would otherwise be inexplicable.
Side-Slipping.—When an aeroplane is stable on its flight path, the air streamline should attack at right angles to the span. If the air attack is at an angle other than go° to the span, the machine side-slips or skids. Like many other dangerous phenomena of flight, side-slipping is particularly dangerous when the aeroplane is so close to some fixed object as to make improbable the regaining of stability before collision. Side-slipping toward the center while the aeroplane is banked on a turn has caused a greater number of accidents than skidding outboard. A number of pilots have been badly hurt thereby, a few have been killed, and a great number of machines demolished. The cause of side-slipping will be more easily understood from the accompanying diagram (Fig. 3).
Let AB represent the surface of an aeroplane flying straightaway from the observer. P is the weight of the machine and L is the lift, which under the assumption of horizontal flight is equal and opposite to P. Assume now that the pilot wishes to turn to the right. He puts the rudder to the right and at the same time banks the machine along A'B' to prevent skidding outboard. The reaction to the lifting surface is now along OR. Therefore if the pilot wishes to maintain his altitude, he must either open the throttle a little more or increase the angle of incidence of the lifting surface, otherwise OL, the vertical component of OR will be less than the weight OP. Assuming that the motor is running "all out" or that the reserve does not permit of an increase of incidence, banking the machine will unbalance OL and OP, thus leaving the aeroplane to the resultant of the downward force (OP-OL) and the centripetal force LR. The result is side-slipping. Now side-slipping obliques the air flow to the wing span, a condition which further decreases OL. Additional decrease of OL tends to increase the side slip, so that the two accentuate each other. It is therefore apparent, when the aeroplane is close to the earth, that unless side-slipping is quickly corrected, a smash will quickly follow. A careful pilot will accordingly avoid sharp turns—that is, steep banks—at low altitudes.
Working into the Vortex of a Spiral.—Casualties both to materiel and personnel have not infrequently resulted from this cause. The condition of a spiral, in its ordinary acceptation, means that the aeroplane is making a turn while gliding or descending under reduced power. In spiral attitude the machine is therefore headed down, banked and turning. Now when a machine is turning, the outboard wing, having the greater velocity, has the greater lift, which tends to increase the bank. Now it may be seen from the last diagram that the lateral component of OR is a centripetal effect which produces turning. Turning produces banking and banking produces turning, so that each phenomenon increases the other. The radius of a spiral will therefore continuously decrease unless the pilot operates his controls to prevent such action. Meanwhile the machine is headed down; and it sometimes happens that the machine works into the vortex and approaches to a condition of spinning about its own longitudinal axis. Such a predicament would naturally be confusing, especially at low altitudes, where the pilot would probably start working all controls at once. Consequently if he pulled up the elevator, as his reflexes might prompt, the banked position of the machine would give the elevator such a rudder component as to make the spiral sharper without decreasing the slope of the flight path. As a consequence the uncontrolled machine smashes into the earth. It is hard to get safely out of a bad spiral. The safe policy, therefore, is to make them wide.
Stalling.—Although stalling is a common reference in popular accounts of aeroplane accidents, there appears to be a vague idea of this phenomenon even amongst flyers. The ordinary occurrence of the stalling of an automobile has associated the idea of engine stall so closely with the stall of an aerofoil that it is necessary to explain how the latter is a distinct phenomenon, which may or may not be caused by the motor, and which may occur in climbing flight, horizontal flight or in gliding flight. The clearest idea of the "aeroplane stall" is obtained by keeping in mind that it is not the aeroplane's attitude to the horizontal plane that determines its stability, but its velocity relative to the air. The minimum speed at which aeroplanes may fly varies with the design of the machine, but for every aerofoil there is a definite minimum air speed at which it will respond to the movements of the controls. When the aeroplane, at this reduced air speed, fails to answer control, it is said to be "stalled." The angle of incidence corresponding to the speed at which control is lost is called the "stalling angle" (Fig. 4).
In the accompanying diagram the graph AB represents the power required to sustain a given aeroplane in horizontal flight at corresponding speeds between E and F. The portion of the curve beyond D corresponds to high speeds at incidences so small as to incur the danger of getting a sudden pressure on top of the lifting surfaces. This portion need not therefore be considered with reference to stalling. The point C on the curve AB marks the speed and incidence, of this particular aeroplane, at which control is lost. In this connection it must be remembered that the speed of an aeroplane is a function of the angle of incidence, while the lift and the slope of the flight path are functions of the propeller thrust. In other words an increase of thrust, for a given angle of incidence, only increases the steepness of the climbing path without increasing the speed of the aeroplane thereon. Therefore for a given aeroplane there is a limit to the increase of the angle of incidence regardless of the power applied. Accordingly to an aeroplane of the characteristic curve AB, it is advisable to adopt a propelling plant, which, for the speeds at which the aeroplane is designed to fly, will deliver, at full throttle, the powers indicated along the curve CD. Examination of the superposed graphs shows that between the points C and D this aeroplane is flying at both safe speeds and safe incidences, and that, using full throttle there is a reserve power represented by the difference between ordinates of the curves CD and AB. Now it sometimes happens that the pilot may not realize what may be a safe flying speed or he may not realize what his best climbing angle may be. Assume that the pilot is using full throttle at speed-incidence corresponding to HKL, which gives him the greatest reserVe and the steepest climbing slope. Now a novice may not realize when this point is reached, erroneously thinking that, because the machine still answers the controls, additional up-elevator (increase of incidence) will steepen the climb. He thus gradually increases the incidence and decreases the speed until the point C is passed. The aeroplane stalls and drops without control.
Suppose, however, that the aeroplane is fitted with a power plant delivering at full throttle as represented by C'D'. Or, suppose that the original power plant is failing, and, at full throttle, can deliver only as represented by C'D'. Now if the pilot repeat his performance, the aeroplane may not stall so abruptly after passing the point C' as it did upon passing C, for the aeroplane, though lacking the lift necessary for sustentation in horizontal flight, still has a speed corresponding to C', at which speed the aeroplane will answer the controls. What takes place is as follows:
At C' sustentation for horizontal flight being lost, the aeroplane starts to descend. The pilot, still feeling that he has control, does not allow the machine to lie upon the flight path, but continues to pull the elevator up. The path of the aeroplane then becomes PS the resultant of two motions PM and PG (Fig. 5). Before the aeroplane began to lose altitude the angle of incidence was NPM. With the machine descending however, and held in its former attitude by the pilot, the angle-of incidence jumps to NPS. Then there is a quick change of speed-incidence from C' to C; and the aeroplane stalls.
In gliding flight there is no propeller thrust. Therefore the equivalent of the propeller thrust necessary to preserve the equilibrium of the aeroplane must be found in the component of the force of gravity along the flight path. Pursuit of the analogy makes it obvious that the angle of incidence may cause stalling on a glide in the same way as in horizontal or climbing flight.
As mentioned in the foregoing, one of the peculiar features of stalling is the "nose-dive." This phenomenon is easy to explain. Within the ordinary limits of incidence the center of pressure is situated at distances from the wing's leading edge, varying from .25 to .4 of the wing's depth (chord). At increasing the incidence beyond the critical angle, the center of pressure begins travelling aft. When the aeroplane is badly stalled, the center of pressure travels aft rapidly, reaching a position of about one-half chord. Now when an aeroplane is in equilibrium, the resultant of all pressures passes through the center of gravity, which, of course is fixed. When the center of pressure jumps abaft the center of gravity, the forces are unbalanced, the pressure has a big moment about the center of gravity, and the machine goes into the nose dive. Obviously if the aeroplane goes squarely into the nose-dive, it will, provided the altitude be sufficient, regain sufficient speed for control. Unfortunately in the most stalls on record the altitude was not sufficient. Each week there are numerous cases of intentional stalls during exhibition flights at altitudes of moo feet or more, where the aeroplane comes out of the nose-dive in good order. There are other instances of stalls at safe altitudes where the pilot failed, through ignorance, to extricate the machine. There is one instance, mentioned in the foregoing, when an aeroplane stalled at a safe altitude, nose-dived for several hundred feet and was finally righted by a chance maneuver close to the earth. Up to this maneuver, despite the fact that the regained speed was ample, the aeroplane refused to answer the controls. This phenomenon may be explained as follows (Fig. 6):
When, during a nose-dive, the aeroplane has gathered excessive speed, as will be the case in 400 feet or more, the incidence of the main surfaces become very small. Now at very small incidences of cambered surfaces, the center of pressure travels aft to positions varying between .4 and .55 of the chord. Thus the travel of the center of pressure on cambered surfaces tends to accentuate a decrease of incidence. It is seen therefore that both for excessively large and excessively small incidences the center of pressure moves aft. As said before the position of the center of gravity remains a fixed point about which the moments of the lifting surface and control surface pressures must be taken. From Fig. 6 it is seen that, for very small incidences, the pressure on the main surfaces has a moment about C (center of gravity) equal to LxCP. The pilot would naturally have the elevator pulled up to the position DH, in which position the moment of the elevator pressure about C is equal to TxDC. Now if the aeroplane refuse to answer the control, it follows that LxCP is equal to TxDC, a condition which may be called the "balance of forces." Since, while the moments are equal, the total pressure on the main surfaces, L, is much larger than the total pressure on the elevator, T, it follows that the aeroplane, thus caught in irons, will receive a translating force, without showing a tendency to come out of the nose-dive. Now what the pilot should do in the above condition is to throw the elevator to the position DK. The moments LxCP and T'xDC would then be in the same direction. The motion of the aeroplane would relieve the pressure on the main surfaces. Then a quick reversal of the elevator back to the position HD would give an unbalanced moment to TxDC, which would swing the aeroplane to a positive incidence sufficiently large to reduce the moment of L, and to allow the elevator to have control. This latter maneuver was the one which saved the pilot mentioned in the foregoing as having foresworn aviation.
With reference to the large and peculiar travel of the center of pressure upon cambered surfaces, it may be said that many of the latest wing sections have been designed to reduce this travel, without greatly impairing the lifting efficiency. As a result of such design the weight of an aeroplane may be so distributed as to bring the center of gravity into approximate coincidence with the center of pressure of the main surfaces, so that regardless of the incidence, L has practically no moment, and the elevator has full charge.
The developments of the last twelve months have left no doubt as to the naval advantage to be gained from the use of aircraft. The question now confronting us is not whether aircraft can accomplish their mission, but how soon and how well can the navy acquire their use. The seaplane is but a weapon, and, following the rule of weapons, it is dangerous to him who handles it without interest or without care. Ignorance, not unbelief, is our trouble.