Part X

March 1893.

At the Paris Exposition of 1889, Commandant Renard, of the French Aeronautical Department, exhibited, in connection with the dirigible war balloon "La France," an apparatus which he had designed some years before (1873) as embodying his conception of a flying machine, and which he termed a "dirigible parachute."

This is shown in fig. 64, and consists in an oviform body, to which is pivoted a couple of standards carrying a series of narrow and long superposed flat blades, intended to sustain the machine when gliding downward through the air.

FIG. 64 -- RENARD -- 1889.

The dotted lines in the side view indicate the maximum angle of inclination which it was proposed to give to this similitude of a Venetian blind, and it is evident that by setting it at the proper angle, and dropping the apparatus from a balloon, it can be made to travel back against the wind a considerable distance, and also that it ma' be steered laterally by the addition of a rudder. Beneath the body a sort of skate will be noticed, probably intended to glide over the ground in alighting, or in obtaining initial velocity to rise should a motor be applied; but the French War Department is reticent concerning its experiments in aerial navigation, and the writer has been unable to gather any information concerning the working of this apparatus.

It will be noted that Commandant Renard proposed to equip this machine with flat blades, thus conforming to the predilection in favor of plane surfaces exhibited by most of the experimenters with aeroplanes already noticed except Captain Le Bris and M. Goupil who took a different view as to the best shapes to employ. In point of fact, as already intimated, those who have succeeded in the air, the true experts in gliding, the soaring birds, do not perform their evolutions with plane surfaces. Their wings are more or less convex on top and concave beneath, and are warped surfaces of complicated outlines. It is true that in many cases they do not differ greatly from planes, and the mind of man so strongly tends to the simplification of complicated shapes, that most inventors have assumed that the effect on the air will be practically the same.

Flight is possible with flat planes, as witness the butterfly, the dragon fly, and insects generally, but such creatures are endowed with greater relative power, as already explained; and, moreover, the elasticity of their wings produces change of shape under action. In the case of the birds, although the outer ends of the feathers are elastic, yet the wing is stiffer as a whole, and the curved surfaces may prove more efficient than planes in obtaining support from the air.

this view seems to have prevailed with Mr. H. F. Phillips, for he patented, in 1884 a whole series of curved shapes, intended to be used in conjunction with suitable propelling apparatus for raising and supporting an aerial machine in the air. These shapes were to be utilized in a set of narrow blades arranged at suitable distances apart; the idea being to defect upward the current of air coming into contact with their forward edges when under motion, so as to cause a partial vacuum over a portion of the upper surface of the blade, and thus to increase the supporting effect of the air pressure below the blade.

These shapes were the result of a series of experiments tried by Mr. Phillips in artificial currents of air, produced by induction from a steam jet in a wooden trunk or conduit, and described in London Engineering in its issue of August 14, 1885.

A cross-section of the shapes patented will be found on fig. 65 Nos. 1-8. The following table gives the results observed, the last column having been added by myself:


of Form.
Speed of Air
Current. Feet
per second.
of Forms --
Foot Pounds
Per Pound
16 x 5
2. 86.7
Shape 1
16 x 1.25
Shape 2
16 x 3
Shape 3
16 x 3
Shape 4
16 x 5
Shape 5
16 x 5
Shape 6
16 x 5
Rook's Wing
0.5 sq. ft.

The intent of these experiments seems to have been to ascertain the speed of current required to sustain various forms and areas of surfaces, carrying the same weight in a soaring attitude. For this purpose they were exposed to the varying current with their long edges transversely thereto, and they were loaded with a weight applied one third of the width back from the forward edge, which point was thought to be the center of pressure. These shapes were swung by two wires attached to their front edges, and when they assumed a soaring attitude in the velocity of current required to sustain the weight, the "thrust" or drift was then measured.

FIG. 65. -- PHILLIPS -- 1884-1891.

The most efficient shape is, of course, that which requires the least expenditure of power, or the smallest number of foot-pounds per pound of weight to keep it afloat and this is seen to be shape No 5, which soared with 3.77 foot-pounds per pound, or at the rate of 146 lbs. sustained per horse power, while the flat plane absorbed more than twice as much power.

The comparison would have been more satisfactory if the soaring angles of incidence had been stated. This is given for the plane only as having been 15° by measurement. This agrees fairly well with calculation; for if the "thrust" is to the "lift" as the tangent of the angle of incidence, then we have 2/9 = 0.222 = tang. 12° 32'. But all the results obtained were probably somewhat vitiated by assuming that the center of pressure was uniformly one-third of the distance back from the front edge, and therefore applying the load at that point.

We have already seen that this center of pressure varies with the angle of incidence in accordance with JoOEssel's law, and the load should have been attached accordingly. If, for instance, the possible soaring angle were 4°, we should have for the position of the center of pressure, back from the front edge, a distance of 0.2 + 0.3 sin 4° = 0.22 per cent, So that it seems probable that if its load had been applied at 22 per cent. instead of 33 per cent. back from the front edge the flat plane would have soared at a flatter angle than 15° and would have shown less "thrust," because the effect of placing the weight so far back was to tilt the plane unduly, and thus to increase both the angle of incidence and the thrust. It is not known whether JoOEssel's formula applies to curved surfaces; but be this as it may, it is reasonable to believe that it would be but little modified, so that perhaps the error in locating the center of pressure operated to the disadvantage of the curved forms nearly as much as to that of the plane. We may, therefore, accept the general statement that greater weights per horse power can be sustained in the air with concavo-convex surfaces than with flat planes; but it seems very desirable that further experiments should be made, for it is quite possible that, in consequence of the loading of the blades at a point differing from the center of pressure, the shapes patented by Mr. Phillips are not absolutely the most efficient forms.

It will be interesting, in this connection, to note how these various shapes behaved. It was found that in order to get the maximum efficiency from any given surface, the greatest depth of hollow should be one-third of the total width from the forward leading edge, and that the amount of concavity of the lower surface and the convexity of the upper surface should bear a relation to the speed of the air current. Thus in shapes 1 and 2 the under surface was nearly flat, and the upper curvature not great, while speeds of current of 60 ft. and 48 ft. per second were required respectively to produce a soaring attitude. In shape 3 the curvature was more marked, and the required speed fell to 44 ft. per second. Shapes 4 and 5 were made broader, with a moderate degree of curvature both above and below, and the speeds of current to produce soaring were 44 ft. and 39 ft. per second respectively. Shape 6 was an extreme case, in which the distinguishing features of the experiments were purposely carried to excess; for when impinged upon by a current of air of 27 ft. per second in the direction of the arrow a0, it was seen (by a fine attached ribbon) that there was an induced current flowing outward in the direction a1,

Shapes 7 and 8 were used to demonstrate that the impinging air is deflected upward by the forward part of the upper surface, and that a partial vacuum results in the after part; they were not loaded with weights, and when exposed to a current of air of sufficient velocity, coming in the direction of the arrow, they rose into the position shown in the figure,

In 1890 Mr. Phillips patented an aerial vehicle in which these curved surfaces were applied to an apparatus similar to the "dirigible parachute" of Commandant Renard, except that there were to be two (or more) series of curved blades behind each other at suitable distances apart, They were to be attached to an elongated body, which he indicated might be of fish shape, and, say, 30 ft. long, The cross-blades, which he termed "sustainers," might be 15 ft. long, 6 in. wide, and 2 in, apart, so many being superposed as to furnish the required supporting air surface. Each set of "sustainers" was to be held in place by a number of vertical bars of angular form! so as to offer the least resistance to the air.

The propelling power was not indicated specifically, save the general statement that it should be "suitable," but a rudder was located at the top of the front series of curved blades, being affixed to a spindle bar terminating below (at the body) with a lever arm. A shifting weight was also provided, capable of being moved across the body, transversely to its line of motion, in order, when moved to either side, not only to depress it, but, by the resistance of the air acting on the surface of that weight, to check forward motion on that side, and thus cause the machine to describe the curve required.

The patent drawings show the vertical standards carrying the blades as being rigidly attached to the body instead of being pivoted thereto, as in the case of Commandant Renard's device, and hence the angle of incidence of the machine could not be conveniently varied in order to rise or to descend; but it is probable that Mr. Phillips has long since remedied this defect, for he is understood to have been continuously experimenting, although the results attained have not as yet been published.

He apparently concluded that he had not developed the best shape in 1884, for he patented, in 1891, the form shown at the bottom of fig. 65. In this, the upper side A of the blade was made convex, as formerly, but the after portion of the lower side of the blade was made concave, as shown at B, while the curvature of the forward portion of this lower side was in the form of a reverse curve consisting of a convex curve, C, at the forward edge, followed by a concave curve, D He states in his patent:

"The particles of air struck by the convex upper surface 4 at the point E are deflected upward, as indicated by the dotted lines, thereby causing a partial vacuum over the greater portion of the upper surface. The particles of air under the point E follow the lower convex and concave surface C D until they arrive at about the point G, where they are brought to rest. From this point G the particles of air are gradually put into motion in a downward direction, the motion being an accelerating one until the after edge Fof the blade is passed. In this way a greater pressure than the atmospheric pressure is produced on the under surface of the blade."

Mr. Phillips indicates that such blades may be of wood, 12 ft. in length and 6 in. in width, from the leading edge E to the rearward edge F, but further experiment led him to make these blades still narrower, and he finally constructed an experimental machine which was tested in the early part of 1893, and has been described in various English journals, notably in Engineering of March 10 and May 5, 1893 the latter issue containing four illustrations, which were reproduced in the AMERICAN ENGINEER for June, 1893. From these various publications the following description of the Phillips experimental machine is compiled.

Instead of providing two series of curved blades, one behind the other, there was but one set, approximately as shown in fig. 64. The apparatus looks like a huge Venetian blind with the slats open. There are 50 of these slats or "sustainers" I 1/2 in. wide and 22 ft. long, fitted 2 in. apart in a frame 22 ft. broad and 9 1/2 ft. high. The sustainers have a combined area of 136 sq. ft.; they are Convex on the upper surface and concave below, the hollow being about 1/16 in. deep. The frame holding the sustainers is set up on a light canoe-shaped carriage, composed principally of two bent planks like the two top streaks of a whale boat, and being 25 ft. long and 18 in. wide, mounted on three wheels I ft. in diameter, one in front and two at the rear. This vehicle carries a small boiler with compound engine, which works a two-bladed aerial screw propeller revolving about 400 times per minute. The fuel used is Welsh coal. There is said to have been no attempt to provide exceptionally light machinery. The weights of the various parts of the machine are, approximately: carriage and wheels, 60 lbs.; machinery with water in boiler and fire on grate, 200 lbs.; sustainers, 70 lbs.; total weight, 330 lbs. The machine was run on a circular path of wood with a circumference of 628 ft. (zoo ft. diameter), and to keep it in position (preventing erratic flight) wires were carried from various parts of the machine to a central pole, as in the Tatin experiments heretofore described. Still further to control the flight, which there is no means of guiding, as the machine is not of sufficient size to carry a man, the forward wheel is so balanced that it never leaves the track, and therefore serves as a guide, carrying some 17 lbs. of the weight, the remainder being on the hind wheels.

On the first run 72 lbs. dead weight were added, making the total lift 402 lbs. As soon as speed was got up, and when the machine faced the wind, the hind wheels rose some 2 or 3 ft. clear of the track, thus showing that the weight was carried by the air upon the Venetian blind sustainers. A second trial was made with the dead weight reduced to 16 lbs. and the circuit was made at a speed of about 28 miles per hour (2,464 ft. per minute), and with the wheels clear of the ground for about three-fourths of the distance That the machine can not only sustain itself, but an added weight, was demonstrated beyond all doubt, even under the disadvantages of proceeding in a circle, with the wind blowing pretty stiffly.

It is stated in the journal Iron that the boiler is a cylindrical phosphor-bronze vessel 12 in. in diameter and 16 in. long. The fire grate area is 70 sq. in., and the fuel Welsh coal. The engine is compound, having cylinders 1 3/4 in. X 3 5/8 in. X 6 in. stroke, fitted with ordinary slide valves. The working pressure of steam is 180 lbs. per square inch, The propeller is 6 ft. in diameter and 8 ft. pitch, with a projected blade surface of 4 sq. ft. The machine was also moored by a stern rope in which a dynamometer was inserted, and on the engine being run at full speed the dead pull was 75 lbs.

If the latter figures be correct, then the power developed was 75 X 2,464 ÷ 33,000 = 5.6 horse power, and the weight carried per horse power was 402 ÷ 5.6, or 72 lbs. per horse power, which is inferior to the 110 lbs. per horse power carried by M. Tatin's apparatus, and probably due to the increased resistance produced by the frame which holds the sustainers.

Mr. Phillips's experimental machine neglects any provisions for maintaining equilibrium in full flight, or for rising and alighting safely. Those he may add later; but whether he does or not, he is entitled to great credit for having been among the first experimenters who have tested concavo-convex surfaces instead of adhering to plane surfaces, and who have thus drawn attention to what may prove to be a very important line of inquiry.

Almost all scientific experiments in air have hitherto been tried with planes, and such few formula as have been proposed are based upon the effect on flat surfaces. It is probable that such formulae--those of Smeaton, Duchemin, Joessel and others-will be found to need modification, either in form or in constants, when applied to curved surfaces. In such case the tables of "lift" and "drift" heretofore given herein will either need recalculation for each specific curved shape, or require the application of a variable coefficient, as exemplified in the calculations of the power expended by the pigeon as heretofore given. In any case it seems very desirable that further scientific experiments be made on concavo-convex surfaces of varying shapes, for it is not impossible that the difference between success and failure of a proposed flying machine will depend upon the sustaining effect (with a given motor) between a plane surface and one properly curved to get a maximum of "lift."

Fig. 66 represents a kite-like aeroplane proposed by M. de Graffigny a French aeronaut, and the author of several works upon aerial navigation, This apparatus was to consist of a kite 46 ft. across, with its fabric surface capable of bagging to a certain extent, and attached to a longitudinal frame, as shown, which was to be trussed both above and below. In front, a stiff triangular head was to be affixed, and an adjustable horizontal tail was to be placed in the rear. Between these, a boat-shaped body containing the machinery and aviators was to be swung on trunnions and attached to the frame. In front of this car a two-armed screw was to rotate, and behind the car a vertical steering rudder was to be placed, above the surface of the kite.

FIG. 66 -- GRAFFIGNY -- 1890.

M. de Graffigny estimated that the power required to drive the apparatus we. in the proportion of one horse power for every 110 lbs., and he proposed the use of liquefied carbonic acid gas, which he states to weigh but 55 lbs. per horse power, including the motor, the recipient and a supply for several hours. This, of course, was a mere makeshift, a reservoir of power for experiment, and not a prime mover; inasmuch as the whole apparatus was to weigh but 396 lbs. and to have sufficient sustaining surface (some 1,300 sq. ft.) to come down like a parachute, should the motor break down while in the air. The screw was to be 6 ft. in diameter and 10 ft. pitch, and its shaft was to remain constantly horizontal (this being the object of hanging the car on trunnions), so that the position of the propeller should be independent of the angle of incidence of the sustaining surface, in accordance with the theory of the designer.

M. de Graffigny states that he experimented with a model of this apparatus in 1890. The screw was rotated some 300 turns per minute by a skein of twisted rubber threads weighing, in the aggregate, I.l lbs., and producing 1,085 foot-pounds in 2 1/2 minutes, or at the rate of 7.23 foot-pounds per second, which proved quite insufficient to give to the apparatus (mounted on three wheels, the foremost of which was adjustable) the velocity necessary to cause it to rise upon the air. The designer expresses himself as unable to state what would be the result with a full-sized apparatus.

It will be noted that this proposal resembles a number of others which have already been described. It is probable enough that the best form for sustaining a given weight and for propelling it with a minimum of surface and of power, or for maintaining equilibrium, were not selected. but M. de Graffigny in the book31 in which this design is incidentally described, strongly advocates the kite principle generally, as the one most likely to lead to success in devising a flying machine, and in learning how to manage it in the air.

This will have occurred to many readers, and it may be interesting to them to inquire as to what has been published upon past experiments with kites, a subject upon which the writer has found distressingly little on record.

Among the first, if not the very first, to call attention to the fact that the study of the kite as a means of obtaining unlimited lifting and tractive power had been unduly neglected was Mr. Wenham, who, in his celebrated paper on "Aerial Locomotion," published in 1866, described briefly some very interesting experiments with kites, and who has kindly furnished the writer with some additional particulars. Mr. Wenham states that his principal summary of facts was taken from a little book, styled the "History of the Charvolant, or kite Carriage," by Mr. George Pocock, of Bristol, England, who also published a small work on "Aeroplastics," both of them, unfortunately, now having become very rare.

The experiments described took place more than half a century ago, and the purpose of the inventor was not to evolve a flying machine, but to provide a floating observatory to serve in warfare, or to drag wheeled vehicles over land.

The apparatus was, in fact, a huge kite, of suitable size to carry the intended weight, with a chair swung just below and so rigged that by tightening or slackening the different cords which held it, the wind would meet it at an>, angle desired, and the apparatus would rise or fall, or could be made to swing a considerable distance to one side or the other. It was so arranged that in case the cords broke, it would act like a parachute, and thus insure safety.

The following quotation, descriptive of the experiments, was given by Mr. Wenham in his paper:

"While on this subject we must not omit to observe that the first person who soared aloft in the air by this invention was a lady, whose courage would not be denied this test of its strength. An arm-chair was brought on the ground; then, lowering the cordage of the kite by slackening the lower brace, the chair was firmly lashed to the main line, and the lady took her seat. The main brace being hauled taut, the huge buoyant sail rose aloft with its fair burden, continuing to ascend to the height of 1000 yards. On descending she expressed herself much pleased with the easy motion of the kite and the delightful prospect she had enjoyed. Soon after this another experiment of a similar nature took place, when the inventor's son successfully carried out a design not less safe than bold--that of scaling, by this powerful aerial machine, the brow of a cliff coo ft. in perpendicular height. Here, after safely landing, he again took his seat in a chair expressly prepared for the purpose, and, detaching the swivel line, which kept it at its elevation, glided gently down the cordage Lo the hand of the director. The buoyant sail employed on this occasion was 30 ft. in height, with a proportional spread of canvas. The rise of the machine was most majestic, and nothing could surpass the steadiness with which it was manoeuvred; the certainty with which it answered the action of the braces, and the ease with which its power was lessened or increased.... Subsequently to this an experiment of a very bold and novel character was made upon an extensive down, where a wagon with a considerable load was drawn along, while this huge machine, at the same time, carried an observer aloft in the air, realizing almost the romance of flying.

"It may be remarked (continues Mr. Wenham) that the brace lines here referred to were conveyed down the main line and managed below; but it is evident that the same lines could be managed with equal facility by the person seated in the car above; and if the main line were attached to a water-drag instead of a wheeled car, the adventurer could cross rivers, lakes, or bays with considerable latitude for steering and selecting the point of landing, by hauling on the port or starboard bracelines as required. And from the uniformity of the resistance offered by the water-drag, this experiment could not be attended with any greater amount of risk than a land flight by the same means.

The reader may perhaps inquire whether there was not some risk that the kite should run away with the wagon when the wind freshened; but Mr. Wenham further explains that the kite attached to the "charvolant" or chariot was provided with a smaller "pilot," or upper kite, which was sufficient to support the "draft," or lower kite, when it was relaxed or allowed to float edgewise, on the wind. The "draft" kite had two cords, one attached well forward, and the other attached well aft, running through rings to keep the cords together. I! the aft cord was slacked off by the driver of the chariot, the "draft" kite floated edgewise on the wind, and the wagon stopped; but by pulling on the aft cord the kite could be made to face the wind absolutely, and to produce the maximum of draft.

Mr. Wenham also mentions in his paper Captain Dansey's kite, for communicating with a lee shore, as described in Vol. XLI. of the "Transactions of the Society of Arts." This was made of a sheet of holland fabric exactly 9 ft. square, and, as stretched by two spars placed diagonally, spread a surface of 55 sq. ft., the remarkable fact about its performance being that in the experiment about to be quoted this surface of 55 sq. ft. sustained no less than 92 1/4 lbs. The quotation is as follows:

"The kite, in a strong breeze, extended I,100 yards of line in. in circumference, and would have extended more had it been at hand. It also extended 360 yards of line I) in. in circumference, weighing 60 lbs. The holland weighed 3 1/2 lbs., the spars, one of which was armed at the head with iron spikes for the purpose of mooring it, weighed 6 3/4 lbs., and the tail was five times its length, composed of 8 lbs. of rope and 14 lbs. of elm plank, weighing together 22 lbs."

This latter kite seems to have been provided with a tail to steady it in the air, and in considering the bearing of such experiments upon possible flying machines, it is preferable to select those upon tailless kites, sailed with one single line, for it is easy to maintain the stability if several restraining cords be used. Mr. Wenham has kindly furnished to the writer the particulars concerning a tailless kite, or, rather, series of superposed kites, patented in Great Britain in 1859 by E. J. Cordner, an Irish Catholic priest, who designed the apparatus to save life in shipwrecks, and who preferred to arrange hexagonal disks of fabric (stretched upon three sticks), above each other on the same line, so that they would all pull together. The operation was to be as follows:

When a sailing-vessel had struck, which almost in every case occurs by the ship being blown on a lee shore, a common kite was to be elevated in the usual way from on board the vessel. When enough cord had been paid out to keep the kite well suspended, the end of the cord on board was to be attached in a peculiar manner to the back of another and larger kite (without tail), and the second kite was then to be suffered to ascend. The end of the suspending rope was to be attached in a similar manner to the back of another and still larger kite, and the process to be repeated until enough elevating and tractive power was obtained, when a light boat or basket with one occupant was to be fastened to the kite line, the latter being paid out until the occupant reached the shore and alighted, when by means of a light running line, extending from the ship to the person ashore, it was deemed easy to haul the basket back and forth as many times as necessary to rescue the passengers and crew.

It is not known whether this ingenious method of saving life without extraneous aid was ever used in a case of actual shipwreck, but it was tested by transporting a number of persons purposely assembled on a rock off the Irish coast, one at a time, through the air to the main land, quite above the waves, and it was claimed that the invention of thus superposing kites so as to obtain great tractive power was applicable to various other purposes, such as towing vessels, etc.

Many proposals have been made at various times and in various countries to utilize kites in life saving, but none seem to have come into practical use. Such attempts may have suggested to Mr. Simmons (the English aeronaut) the experiments which he is said to have tried, in 1876, of gliding downward under such buoyant sails.

The only accounts which the writer has found of these experiments are given in the Aéronaute for April, and for November, 1876 The apparatus of Mr. Simmons is described as consisting of a huge "pilot" kite 49 ft. high and 49 ft. wide, with another kite below, still larger. The pilot kite was first to be raised, and to carry up the second the two were to be adjusted to the breeze, and the aeronaut was to be suspended in a car, and allowed to ascend 200 or 300 yards. Then by adjusting his weight by means of guy lines, so as to obtain a proper angle of incidence, the apparatus was said to glide downward to the ground, being slightly dirigible through the guy lines, and to be arrested by the bystanders seizing a dragging guide rope.

Mr. Simmons is said to have been fairly successful with his experiments in England, but to have failed to repeat the feat at Brussels, Belgium. In the latter case it was claimed that there was not sufficient wind, but steadiness of breeze would be more important. The surfaces operated with seem to have been very large--some two to three square feet per pound in order to alight gently; but such extent of surface is so unmanageable in a gusty wind as probably to have led to the abandonment of the experiments.

The exploit is feasible, and would prove useful in-experimenting with various shapes and extent of surfaces, but such experiments should be tried with areas more nearly corresponding to the proportions which exist in soaring birds, and the operator should invariably alight in water until he has learned how to manage his apparatus.


31 "Traite d'Aerostation." H. de Graffigny 1891, p. 189.
List of Illustrations Table of Contents Index