Biplane Driven by Compressed-air Engine
The model aeroplane illustrated by Fig. 148 has been designed to
suit the compressed-air plant fully illustrated and described in the
preceding chapter. It is from the results obtained from the testing
of the plant that the dimensions of a suitable model for it are
determined; and while the design may suit the majority of the plants
constructed from the illustrations shown in pp. 95 to 101, it is
chiefly given to show the correct method of designing a “power-driven”
machine, since the power unit (unlike the elastic motor) cannot be
varied, and recourse to some established line of reasoning becomes
essential.
The first thing to do, then, once the plant has been “tuned up,” is to
ascertain the thrust obtainable from it. This is found by suspending
the plant by the valve on a balance, with a container fully inflated,
the weight registered being carefully noted. The container pressure
should now be released, and the weight registered when the motor is
running observed. By subtracting the former from the latter the thrust
is obtained.
Thus, assuming the plant, at rest, to weigh 8 oz., and when running 12
oz., it is clear that the thrust is equal to 4 oz. Now, it is necessary
to know the _average_ thrust developed, since, as hitherto explained,
the thrust is not constant, but gradually diminishes as the density of
the air in the container approaches normal atmospheric conditions; that
is, 14·6 lb. per square inch (known as an atmosphere). It is possible
to obtain some very interesting data by plotting a graph of the thrust
given off at various moments from the release of the pressure in the
container. Meanwhile it can be taken as a good rule that the thrust
registered after one-third of the effective run of the motor represents
approximately the average thrust; and the figure given above (4 oz.)
will serve for the purpose of illustration.
It is next necessary to know the weight of the model it will lift. It
is well established that a plant will fly a machine weighing from four
to six times the weight of the thrust it develops, although, of course,
much depends on the efficiency of the model; the greater the complexity
of frame members the lower the lift drag ratio, and consequently
the lower the ratio between the thrust and the weight of the model.
Compromising, and taking 5: 1 as the ratio, 20 oz. is obtained as the
total weight of the plant and model.
The next point to be decided on is the loading, and as the model is to
be a biplane a comparatively light loading can be used. In the case of
the machine shown in side elevation by Fig. 146 and in front elevation
by Fig. 147, 4 oz. has been taken as the loading per square foot. So
that the total area of the wings will be
20 = 5 sq. ft. = 720 sq. in.
----
4
A span of 54 in. for the top plane and 46 in. for the bottom one has
been decided on, and by using a chord of 8¾ in. the total area of the
wings vies very approximately with this figure, allowing a small margin
for excess weight. The area of the tail, which is non-lifting, need not
be taken into account. Although the “gap” is given as being equal to
the chord, it could be made, if anything, ½ in. greater.
Now with regard to actual materials. Birch is to be used for the
longitudinals, straight in the grain and of the cross-sections
illustrated. The lower member is bent under steam to the curvature
shown—of 6½-in. radius. Two vertical struts support the wings, and
these should be cut from hickory. A short tie-strut secures the bottom
longitudinal to the front inter-strut, the joint being made by means of
side angle-plates bound into place. It will be found good practice to
make a full-size drawing of the machine in side elevation, so that it
can be used as a template to fit up the cross members—particularly with
regard to the cutting of the angles.
The joint of the longeron to the cross member is shown separately at
=A= (Fig. 146). The usual fish-plates are employed, so made that a
small wiring plate is left protruding from the binding, to which the
cross-sectional and longitudinal-sectional wires are made off.
[Illustration: Fig. 148.—Plan View of Biplane Driven by Compressed Air]
The plant itself is slung into the framework by means of eight wires,
each being made off to the wiring plates. Each should also be provided
with a small ¾-in. wire strainer to enable the plant to be fixed quite
rigidly—albeit permitting of its being removed for inspection or
repairs. The wires from the engine itself are taken off from the four
small eyes soldered to the stationary portion of the crank-shaft. Great
care should be taken to ensure that the plane of rotation of the screw
is at right angles to the main planes. A 1¼-in. dihedral is given to
the bottom plane by means of the bracing wires passing between the
inter-struts, and shown on the preceding page.
It has been thought advisable to attach a small rear wheel, to enable
the model to rise off the ground with as little loss of power as
possible. Such a wheel, with attachments, need weigh no more than ¼
oz., and is a great improvement over the cane skid usually employed.
[Illustration: Fig. 146.—Side Elevation of Biplane Driven by Compressed
Air]
In bracing the outriggers, or longerons, some care will be required to
ensure their being quite true. It will be easier to finish each section
off first, so that they are quite parallel at the joints.
The part plan view of the model (Fig. 148) will make the relative
position of the various component parts quite plain. The two top tail
outriggers pass through the fabric at the point where the spar is
located, their front ends being pinned and cross bound to the wing
spar, which is made of greater cross-section in the centre, so that its
strength is not materially impaired through the piercing of it. Birch
is to be used for the wing spars and ribs of the sections indicated.
The planes are ribbed at periods of 6 in. and given a camber of ¾ in.,
the greatest depth of which is 2¼ in. from the leading edge. It is far
easier to impart the camber after the wing framework is made than to
camber each rib separately. Each rib should be cut 1 in. longer than
necessary, and pinned and glued to the spars, with ½ in. overlapping
each of these latter. When the glue is quite set, the pins may be
clinched over by supporting the wing on an iron weight and tapping them
back flush to the spars.
The full-size section of the camber should be drawn upon a board, with
which to check the accuracy of the first rib to be cambered (the end
rib).
The ribs are cambered in a jet of steam, the convex or top sides being
placed nearest to it. Having cambered the end rib carefully to agree
with the drawing, the others may be matched to it. It will thus be easy
to ensure that every rib is of the correct curvature, as any mistake in
the steaming of the rib will distort the wing spar at the point of its
attachment.
If, however, it is thought advisable to camber the ribs first, a
wooden bending jig should be made, to enable several ribs to be bent
at one operation. The ribs should be tied down to the jig with string,
and thus held under the steam jet, being well dried in front of the
fire before they are detached from the jig. All three spars pass
_underneath_ the ribs.
A very light fabric should be chosen, such as can be obtained from the
model-aero accessory warehouses, or an unproofed Japanese silk can be
used and varnished when on the wing. If this latter is used, it will
be found advantageous to use a yellow hue, as this colour is least
affected by the action of the varnish. But the covering of the wings
must be left for the time being, for the reason that the sockets to
which the inter-struts are made fast must first be attached. Further,
the top plane must be covered after the tail outriggers have been
assembled, as it is so much easier to make the joint between the wing
spars and these latter before the fabric is attached.
To render it unnecessary to refer to the point further, it may be
noted that the fabric is brought over the leading spar of each wing to
pocket it out. It is much neater to sew the fabric along on the leading
edge, as when glue is used an unsightly black smear shows through. The
fabric should be stretched from end to end first, the fabric overlaps
being glued on the bottom face of each end rib. Drawing-pins should be
partially pressed into the ribs to secure the fabric until the glue is
set.
At =B= in Fig. 148 is shown the method of securing the bottom plane to
the inter-struts. Convenient notches are cut in the struts into which
the plane is sprung. It will have been noticed from the side elevation
(Fig. 146) that the width of the inter-struts increases towards the
bottom or lower ends, and also that they incline slightly; this is
to provide for the entry of the lower plane, since the top plane is
attached _outside_ the struts, while the bottom is placed _inside_
them. At =C= is shown the method of attaching the inter-struts.
The tail is built up from split bamboo, ⅒ in. by ³/₃₂ in. in
cross-section, and the rudders are framed up from No. 20 gauge piano
wire. The ends of the rudder frames are forced through the longerons,
and the ends bent back in alignment with them; they are then bound to
the longerons with black three-cord carpet thread. The rudders are
covered after being fixed to the outriggers. When it is necessary to
adjust them, the piano wire will be found sufficiently ductile to admit
of a warp being placed thereon.
[Illustration: Fig. 147.—Front Elevation of Compressed-air-driven Model]
Eleven ribs connect the spars of the top plane and nine those of the
lower, the camber of each being the same; that is, the same depth of
camber is maintained throughout. Before the wings are covered, the
angle-plates to which the inter-struts are fixed must be bound on; and
these are cut from No. 30 gauge sheet tin. They should be cut less in
width than the spar to which they are attached, in order that their
sharp edges shall not cut through the binding. To prevent the plates
from moving, they should be lightly sunk into the wing spar with two
centre-punch dots, and a film of glue should also be spread over the
face of the plate coming in contact with the spar.
The inter-struts are streamlined in cross-section (see Fig. 149); but
they are to be left rectangular in section at their ends, to provide a
flat surface for the plates to bed home on. The ends of the plates are
turned back over the binding, which may be of the light machine variety.
The lower ends of the inter-struts are cut off to the same angle as the
dihedral on the lower plane, to avoid distortion of the plane. Spruce
or American whitewood may be used for them, the greatest cross section
being ½ in. by ⅛ in. The greatest cross-section is situated at the
middle of the strut, whence it tapers to ³/₁₆ in. by ⅛ in. Fig. 149
shows the attachment of the inner strut to the wing spar. In Fig. 150
are shown the brackets forming the guides for the axle, and also the
supports for the rubber shock-absorbers. Piano wire is used for them.
The width of the guide should be such that the umbrella-ribbing, which
constitutes the inner portion of the axle, rides freely within it. The
wheel axles are cut from No. 16 gauge piano wire, and they are soldered
to the umbrella-ribbing, being sunk into the channel of the latter,
bound with No. 30 gauge tinned iron wire, and then soldered.
Ordinary elastic, as used for a rubber-driven model, can be used as the
shock-absorber, and it should be neatly and fairly loosely bound to the
vertical guide, the axle of course being first seated therein. In order
that the absorber brackets may maintain a vertical position, their ends
are shaped to a form similar to the letter =U=. They are wire-bound to
the skids and lightly soldered.
[Illustration: Fig. 149.—Details of Inter-struts.]
[Illustration: Fig. 150.—Detail of Axle Guides.]
[Illustration: Fig. 151.—Outrigger and Wing-spar Joint.]
[Illustration: Fig. 152.—Tail Crossbar Joint.]
[Illustration: Fig. 153.—Detail of Front Skid Angle-plates.]
[Illustration: Fig. 154.—Tail Incidence-quadrant.]
[Illustration: Fig. 155.—Details of Axle.]
[Illustration: Fig. 156.—Rear Chassis Details.]
[Illustration: Fig. 157.—Inter-strut and Longeron Joint.]
The bottom plane only must be covered; it will be easier to cover the
top plane when the machine is assembled, for it would be a difficult
matter to secure the top outriggers to the spars were the fabric
attached.
Having assembled the outriggers and completed the bracing of it, it
will be possible to attach it to the wings.
Small elliptical holes are cut in the fabric of the lower wing, through
which the central supports or stanchions pass, and the bottom plane
is seated home in the notches alluded to in Fig. 148. Next, the top
outrigger ends are fitted up, being cut off to correct length and
halved on to the wing spars, as shown in Fig. 151. The vertical support
is then glued, pinned, and cross-bound to the outrigger.
Great care will be necessary to ensure that the outriggers are quite
central with the planes. A point to be made clear is that if in the
fitting of the top outriggers one is cut even ¹/₃₂ in. short, the
tail end of the machine may be ¾ in. out of centre. In order to check
inaccuracy in this direction it would be advisable to mark the centre
of the horizontal tail member, insert a drawing-pin, and take the
measurement to the corner of the wing tip, on both sides of the model;
the outriggers should be temporarily lashed to the wing spars, and
gradually adjusted until they are located centrally with the planes.
Perhaps it may be interesting to here mention that this is the method
employed in locating the fuselage of full-size machines.
The bracing of the planes should now be undertaken. All lift wires
should first be fixed, beginning from the wing tips. Just sufficient
tension should be placed on each wire to ensure rigidity. A wooden
straightedge should be used to reveal any distortion of the spar. The
top plane must be given a slight dihedral, so that when the anti-lift
wires are inserted it assumes a perfectly parallel position.
[Illustration: Fig. 158.—Biplane Driven by Compressed Air]
The upper and lower longerons are spanned at the tail end with light
spruce cross-bars, ⅛-in. by ¼-in. section, which are let into mortises
cut in the longerons; and two vertical posts are halved on to these
cross members (see Fig. 152), to provide the fulcrum about which the
tail swings in the quadrant, to be referred to presently. They are
spaced 4 in. apart, which is equivalent to the distance between two
ribs; and on the outside of them a groove is cut in the centre of each
to provide a seating for the two central tail ribs. These grooves must
be cut =V=-shaped, the apex of the =V= facing the trailing edge of the
post. The object of the groove is to form a guide for the tail when it
is desired to alter the angle of it. A pin should be driven through
the rib and into the groove to constitute the pivot on which the tail
swings; and the ribs must be bound with fine thread on each side of
the pin to prevent the rib from splitting. It will be found that it is
better to bind the ribs before inserting the pin.
The central inter-struts are attached to the skids by angle-plates,
and in Fig. 153 the form of these is given. It must be understood that
there are two plates to each joint, one on each side of it, and for
neatness and simplicity they can be cut from one piece of tin, both
plates being thus formed in the one. No. 30 gauge tinplate is suitable.
The plates are pinned, clinched, and bound into place, and constitute
an exceedingly rigid piece of construction, which is needed in this
portion of the machine, bearing as it does the impact of landing. Glue
should be neatly brushed into all the joints. The tie-strut is to be
streamlined as far as practicable without materially impairing the
strength of it.
A very neat finish can be given to the binding if it is just brushed
round with japan black, which shows up in pleasing contrast to the
light brown varnish with which the framework is coated.
Fig. 154 gives the shape of the quadrant, which makes possible the
variation to the angle on the tail. It is cut to a radius of 4 in.,
and is pinned into position. The pitch of the teeth is ⅛ in., and this
facilitates a very fine adjustment. It should be so fixed that the
tail springs tightly into notches, but not so tightly as to render
adjustment difficult. Trial and error will be found the best method of
locating its position.
[Illustration: Fig. 159.—Compressed-air-driven Monoplane]
It was mentioned in the preceding chapter (see pp. 94 to 103) that the
axle is composed of two portions, umbrella-ribbing and piano wire,
and Fig. 155 shows the construction. It will be seen that the piano
wire beds into the channel (which is fixed in a trailing position),
wherein it is bound and soldered. The wheels are spaced apart by means
of small brass-tube collars, soldered to the piano-wire axles in their
respective positions. The axle itself, as mentioned earlier in the
chapter, is attached between the shock-absorber brackets, being held
there by means of suitable radius wires secured to any convenient part,
the rubber binding forming the absorber. The radius wires are essential
in order to maintain the lateral position of the axle relatively to the
planes. Sufficient rubber binding is to be used to absorb the shocks
the model is bound to receive, the exact quantity, of course, being
impossible to define.
The rear wheel members are fixed to the longerons in the following
manner. The ends are bent back parallel to align with the frame member.
The apices of the =V= chassis members are soldered to the short axle
carrying the back wheel, the axle being cut a length suitable to the
hub of the wheel. No. 20 gauge wire is used for all portions of the
rear chassis. Fig. 156 makes this clear.
Fig. 157 shows the joint of the trailing central inter-struts to the
top longeron. It will be seen that the joint is a halved one, pinning
and binding forming the security.
All woodwork may be polished by filling the grain with gold size, and
finishing with a good varnish.
In flying the model the writer would point out that full pressure
should not be given to the plant until adjustment has been completed,
also the importance of tuning the machine by starting it from the
ground, thus obviating many vexing smashes. Further, the rudder must
be set to counteract torque; if the screw is of left-hand pitch, then
torque will tend to bank the machine to the right, and the rudder must
therefore be set to the left, and vice versa.
A sketch of the finished machine is given by Fig. 158, and a design for
a monoplane driven by the same plant in Fig. 159.