EDF theory and practice 1

Electric Ducted Fan – theory and practice

R.A.Sharman, B.Sc, Ph.D, Ceng

Table of Contents
1 Introduction......................................................................................................................................2
1.1 Objectives of EDF...................................................................................................................5
1.2 What do we need to know ?......................................................................................................6
1.3 What don't we need to know ?..................................................................................................7
1.4 A digression on Jet Engines (Gas Turbines).............................................................................8
1.5 The aim of this document.........................................................................................................8
2 A simple explanation of EDF propulsion for model aircraft............................................................9
2.1 The components of an EDF system..........................................................................................9
2.2 The operation of an EDF system............................................................................................10
3 A theory of Ducted Fan propulsion.................................................................................................11
3.1 Assumptions about the physics of airflow and motion...........................................................11
3.2 Defining the properties of an EDF system..............................................................................11
3.2.1 The geometry of a ducted fan system.............................................................................12
3.2.2 The electrical properties of a ducted fan system.............................................................12
3.2.3 The aerodynamic properties of a ducted fan system.......................................................12
3.3 Calculating the Thrust of an EDF system...............................................................................13
3.3.1 Conservation of mass......................................................................................................14
3.3.2 Mass and thrust...............................................................................................................14
3.3.3 The power needed for flight............................................................................................15
3.3.4 Balancing the forces acting on the EDF system..............................................................15
3.3.5 Finding the efficiency of the EDF system.......................................................................16
3.4 A summary of what has been achieved...................................................................................16
3.5 An Example – the West Wings Bae Hawk in level flight.......................................................17
3.6 The theory of Static thrust.......................................................................................................18
3.6.1 Assumptions and methodology.......................................................................................18
3.6.2 Assumptions made for the static analysis........................................................................19
3.6.3 Calculating the exhaust velocity from the system power................................................20
3.6.4 Calculating Static Thrust.................................................................................................20
3.7 An example – the BAe Hawk in static test.............................................................................20
3.7.1 Using the static case theory.............................................................................................20
3.7.2 Using the observed current consumption........................................................................21
3.7.3 Using the observed thrust measurement.........................................................................22
3.7.4 A possible adustment to the KS theory...........................................................................23
3.8 Conclusions on the utility of the static test method................................................................23
4 A Critique of the KS theory............................................................................................................24
5 References......................................................................................................................................24
6 Glossary..........................................................................................................................................24
 EDF theory and practice 2

Illustration Index
Illustration 1: My West Wings Hawker Hunter. .................................................................................3
Illustration 2: My West Wings Bae Hawk. .........................................................................................3
Illustration 3: The Max-Thrust GR4 Tornado......................................................................................4
Illustration 4: The E-flite Habu............................................................................................................4
Illustration 5: A wemotec 70mm 5-bladed fan with Mega-EDF motor fitted......................................5
Illustration 6: A pair of 60mm EDF 3-blade units................................................................................6
Illustration 7: Normal airstream flow through a Fan unit at flying speed..........................................13
Illustration 8: Airstream through an EDF unit when it is not moving (static)...................................19

1 Introduction
The most common type of model aircraft seen flying under, radio control at model flying
sites in the UK, is probably of the “spitfire” pattern – that is, a model aircraft with propeller
at the front driven by an electric or IC engine, a monoplane wing in the middle, and a rear
tailplane and fin. However, many modellers aspire to build and fly different types of
aeroplanes (of which there are a very large number) such as biplanes and canards.
Additionally, the motor and propeller can be mounted centrally, at the rear, on the wings and
so on. In fact, model designs appear in a greater degree of profusion than is ever seen on
full size aircraft in the modern era. This is in marked contrast to historical experience (in the
period up to the second world war) when full size designs proliferated but models were
rather simple and more uniform.
One type of model which has become very popular in recent decades is the “jet” style of
aircraft, often conceived as a scale, or semi-scale, model of a military jet fighter, or even a
model of a commercial jet. Tremendous strides were made with IC ducted fan powered
model jet planes in the 1980's and 1990's. The complications of the design, and the marginal
power of the motorised fans, resulted in these have been the province of a rather small set of
very dedicated, and highly skilled, model aircraft enthusiasts, rather than the general model
flying fraternity.
With the advent of electric powered model flight the possibility has arisen to drive ducted
fans by electric motors, giving rise to Electric Ducted Fan models (EDF). More or less at
the same time there has been a parallel development in model Jet turbines which develop
considerably more power (albeit with a lot of extra complexity and cost). Consequently the
jet model enthusiasts have more or less split into two camps:
• EDF powered small models, rather inexpensive, for general model enthusiasts
• turbine powered large models, rather expensive, for specialists.
In the following we shall consider only EDF as shown in the following examples:
 EDF theory and practice 3

Illustration 1: My West Wings Hawker Hunter.

This has a 32” wingspan with fixed undercarriage, powered by a Wemotec fan with Mega
16EDF motor and Overlander 2600mAh 40C Lipo battery. The model was built in January
2009 and first flown in May 2009.

Illustration 2: My West Wings Bae Hawk.

This has a 35” wingspan with electric retracting wheels, powered by a Wemotec fan with
Mega 16EDF motor and Overlander 3200mAh 60C Lipo battery. The model was built in
February 2011 and first flown in June 2011.
 EDF theory and practice 4

Illustration 3: The Max-Thrust GR4 Tornado

This has working swing wings, powered by a 50mm ducted fan unit. This model is all foam,
and has a low wing loading, hence its ability to take off from the ground (rather than bungee
launch).

Illustration 4: The E-flite Habu

This is powered by an 80mm 5-bladed fan. This model has retracts wheels, flaps all normal
functions. Its performance is such that it is considered a suitable model to train on for those
interested in later moving to turbine powered planes.
 EDF theory and practice 5

1.1 Objectives of EDF

EDF power units are defined as an electric motor driving a small, multi-bladed, impeller
housed within a cylindrical shroud.

Illustration 5: A wemotec 70mm 5-bladed fan with Mega-EDF

motor fitted

The advantageous properties of an EDF unit are:

• The ducted fan unit is compact and enclosed and can therefore be mounted inside a
fuselage, as well as externally. This makes it a suitable for powering a jet type of
aircraft.
• Being electric the unit has no need for fuel tanks, throttle servo, exhaust ducting,
starting equipment, ignition and do on, and is therefore easier to construct and
easier to operate than IC powered fans.
• The shrouded propeller should be more efficient that a propeller of equal size in free
air, as the losses due to tip vortices, pressure decrease at the blade tip, and so on are
reduced.
• The potential exists for vectored thrust by diverting the exhaust flow direction to a
small extent, which may be useful for some aircraft to perform certain unusual
manoeuvres.
• Multiple engined models are much easier to build than an equivalent IC engined
plane.

The disadvantageous properties of an EDF unit are:

• The ducted fan unit produces less thrust and is therefore less effective than a normal
propeller driven by an equal amount of power. It is generally of smaller diameter,
 EDF theory and practice 6

and has a higher propeller blade loading, than the two-bladed propeller for the same
model.
• In order to increase the thrust of a given EDF unit it is usual to increase the power
and speed of the motor, but this requires more Watts of energy, which in turn
requires a bigger battery capable of higher current draw rates.

An example of the sort of units suitable for a multi engined plane, such as a learjet, is shown
in the net picture.

Illustration 6: A pair of 60mm EDF 3-blade units

These fans are suitable for external attachment to wings or fuselage

1.2 What do we need to know ?

The sort of questions that arise when planning an EDF-powered model are:
• How to choose the best EDF unit for a given model
• How to design a new model around a particular EDF unit
• How to check the performance of the model and EDF unit is as it should be
• How to improve the performance of a given model set up.

What is needed is an understanding of how fans work (i.e. the physics of the ducted fan
propulsion process) and therefore a method of solving problems, making adjustments and so
on.
 EDF theory and practice 7

1.3 What don't we need to know ?

A considerable amount of “mis-information” or “dis-information” seems to exist about the

mechanics of ducted fans. Some of these misconceptions will be addressed here (and the
reasons why they are wrong).

Some common ideas (which are NOT true) are:

• “a ducted fan speeds up when the model gets in the air, so that the model flies
faster” (not true). It is often claimed that the “note” emitted (the whine of the motor)
increases in pitch because the fan moves into a more efficient regime (”gets on the
step”) and therefore produces more power. The note may in fact, rise, although this
is hard to prove, and is confused by the Doppler effect when the plane is approaching
the pilot (it seems to be accelerating, and the apparent pitch of the sound does
increase). But there is no regime change in EDF operation: in fact the power
produced by a fan unit in flight actually decreases with increasing speed – if it didn't
the plane would eventually reach the sound barrier).

• “ducted fan planes need to be bungee launched to get them flying fast quickly” (not
true). It is often observed that bungee-launched models seem to “sag” as they leave
the bungee, but quickly pick up speed. This is due to the fact that most EDF planes
under underpowered for their weight and thus need time to accelerate to flying
speed. The bungee gives good initial acceleration, but the low power unit is slow to
take over. The same applies to hand launching. As EDF plane designs and
equipment has improved planes are able to take off from a ground run as is normal
for other model aircraft.

• “a ducted fan is an electric jet and produces power in the same way a gas turbine
does” (not true). An EDF unit is similar in principle to the compressor stage of a gas
turbine, but only that. An EDF unit does not heat air, combust fuel or produce large
quantities of surplus thrust. For the sake of simplicity we call them “electric jets”
but this is a misnomer.

• “a ducted fan plane can be made to go better(faster) by substituting a more powerful

motor” (not always true). An EDF unit depends on applying some Watts of electrical
energy to push air backwards, and therefore push the model forwards. More watts
applied should make for more air pushed faster. However, a higher wattage motor
may be heavier, and will require a bigger battery, which will also be heavier. The
airframe may need to be strengthened to withstand that increased strains of the motor
and the increased aerodynamic forces on the flying surfaces, resulting in a heavier
airframe. The higher overall weight of the plane will result in a higher wing loading
making the plane difficult to take off, liable to stall in flight, and to have a higher
landing speed – all undesirable characteristics. Longer flight times, more
manoeuvrable flight, and better flying characteristics can often be better achieved by
decreasing the all up flying weight of the plane.
 EDF theory and practice 8

1.4 A digression on Jet Engines (Gas Turbines)

Gas turbine engines (jet engines) work by

1. using a compressor to convert incoming air to high pressure, then
2. combusting fuel to produce a hot jet exhaust, and finally
3. using a turbine to extract some of the exhaust energy to drive the compressor.

The expansion of the burning fuel creates enormous energy in the form a high pressure
exhaust stream which can be used in various ways:
• as direct thrust, similar to rocket propulsion in a supersonic jet (as used on many
military fighter aircraft)
• to make the turbine power a propeller – the turbo-prop engine (as used on many
small and intermediate commercial planes)
• to make the turbine power a fan – the turbo-fan engine (as used on most larger
commercial airliners).

An EDF unit is similar in principle only to the compressor stage of a jet engine. It is a very
inefficient compressor relative to the axial compressor of a gas turbine engine, or even the
centrifugal compressor of a non-axial jet engine. In all other respect an EDF unit is totally
different to a gas turbine as it does not greatly compress incoming air or combust fuel. As
a result it cannot produce large quantities of surplus thrust.

The corollary is that, since an EDF unit is a low-power thrust device, it will only have
application in certain areas, such as:
1. small planes carrying small batteries, but with potentially powerful motors
2. very light planes carrying small batteries
3. multi-engined planes, where motors can potentially share batteries

It is notable that to date (2012) the most successful EDF model planes use military fighter
scale appearances, are of foam construction, are of the order of 1 metre wingspan and weigh
about 1.5Kg. Below these limits most of the weight of the plane is equipment (fans, motors,
batteries, radios, etc.) and the airframe tends to be rather fragile. Above these limits and the
total all up weight is rather high, requiring more thrust to generate sufficient lift.

1.5 The aim of this document

Surprisingly there is little scientific literature written on the theory of how EDF units work.
There is, of course, a great deal of un-scientific literature on the subject, most of it wrong,
misleading or incomplete. Information on how to make better units, and make model planes
fly better is also scarce. The most important omission is a readily available means to
 EDF theory and practice 9

calculate the thrust produced by a fan unit for help in selecting the right unit to use,
matching it to a speed controller (ESC) and battery appropriate for the type of aircraft and
flying style desired. This document seeks to fill some of these gaps.

Some articles which appeared in various model magazines, starting with an MFI magazine
article of 1984 in which Klaus Scharnhorst derived an equation for the velocity of air exiting
from a static fan (i.e. one strapped to the bench). This value can then be checked against
observation. This work culminated in a theory of EDF operation published in 2005[1] (in
English). The velocities of air into and out of, an EDF system are related to the power
required to drive the fan in both static and dynamic cases. Given the geometry of the model
(ducts, fans, etc.) and other information (such as the speed of the model) we can work out
how much power is needed, and therefore what motor, battery is required.

2 A simple explanation of EDF propulsion for model aircraft

The components of an electric ducted fan are exemplified in the following diagram:

THRUST

Air flow

Free air Entry duct Fan exhaust duct exhaust

A simple depiction of an EDF system in which air is allowed to flow in (at the left) and is
pushed out (at the right), resulting in a force on the airframe (to the left).

2.1 The components of an EDF system

A Ducted Fan system can be thought of as the following components, arranged in order:
• an entry duct, or forward duct, which channels air from an entry aperture to the front
of a fan unit. In some cases there may be two entry apertures, one on each side of
the fuselage. These usually join at a confluence ahead of the fan unit. The forward
duct is not usually cylindrical (although it could be). More commonly the forward
duct is tailored to the shape of the fuselage it is mounted in.
• A fan unit which comprises a short cylindrical tube, in which is mounted a motor
driving a fan (or impeller). The impeller may have 2, 3, 4, 5, 6 or more blades
according to its design. The entry and exit diameter of the fan unit is almost always
the same. Note that the presence of the motor in the centre of the fan unit
automatically takes a small amount of the fan area away. The remainder of the space
 EDF theory and practice 10

through which air flows is called the fan-swept-area (FSA).

• An exhaust duct, or rear duct which connects the fan unit to the outside. The rear
duct, or exhaust, is most often cylindrical or nearly so. The rear duct usually tapers
to the exit by a small amount. A little tapering causes a beneficial speeding up of the
exhaust flow and generates more thrust. Too much tapering introduces resistance
into the system which causes a lack of thrust.

2.2 The operation of an EDF system

The following section attempts to give a simple, non-technical description of how and why a
ducted fan unit produces sufficient propulsion to power a model aircraft.
Evidently, the EDF unit is connected to a battery, via an ESC (electronic speed controller)
under the command of a radio receiver supporting a throttle control servo channel. The pilot
sets the throttle to a given level, the fan produces an airflow through the ducts resulting in a
reactive force on the airframe and the plane is pushed forward.

The various components of this picture are described in a little more detail:

1. In front of the entry duct (the area into which the plane is flying) is a free air-stream
This air may be moving (according to the wind at the time) and may be flowing
towards the plane (the plane is flying upwind), or away from the plane (the plane is
flying downwind), or in some cross-wind direction. To simplify the discussion we
assume here that there is no external wind and the plane is flying in still air. This is
unrealistic as in natural conditions there is always some wind, but it will do for a
theoretical abstraction, since in practice the wind velocity as low relative to the
model velocity (the plane is flying fast). Remember, too, that a plane flying in a
moving air-stream is also moving with the air-stream Relative to the pilot on the
ground it may appear to fly slowly in one direction and fast in another, but of course
relative to the air it is flying at the same speed most of the time.
2. In the entry duct air is moving along the duct, essentially at the same speed that the
plane is flying. There will be some resistance to flow caused by friction on the walls
of the duct, but this can be considered to be small (air is not very viscous, and the
duct is not so small that the resistance would be significant).
3. As the air passes through the fan unit it is slightly compressed, and therefore its
temperature will in fact be raised by a small amount (probably too small to measure).
This is the result of the rotating fan blades compressing the air. Considerable energy
is exerted by the motor to turn the fan at a relatively high speed. In a typical example
350 watts of electrical energy (11.4 volts at 30 amps) is used to turn a fan at 45,000
rpm generating around 9 Newtons of force.
4. In the exhaust duct the slightly compressed air escapes to the outside atmosphere,
and reverts to normal atmospheric pressure. The escaping air is in the form of a
stream of faster moving air. This creates an equal and opposite force on the airframe
which pushes the plane forward.
 EDF theory and practice 11

Some observations can be made immediately: Air must pass easily through the ducts and
fan without restrictions, otherwise the EDF unit will not operate at its best. We know
already that it is not likely to have power to spare, so none must be wasted by tortuous or
long ducts, narrow openings, and so on.

3 A theory of Ducted Fan propulsion

This treatment follows the presentation of Klaus Scharnhorst. It is suggested where it might
be incomplete. A more accurate treatment is proposed later.

3.1 Assumptions about the physics of airflow and motion

The following assumptions are made:
• the airflow is incompressible.
• The effective size of the EDF unit is determined by the Fan Swept Area.
• The ducts are frictionless.
• Ducts can be assumed to be circular in cross section.
• The bifurcation and joining of ducts can be ignored.

Actually, air as a fluid is compressible, as will quickly be noticed when you pump up a
bicycle tyre, for instance. The pump increases the pressure of some air which is then stored
inside the tyre. However, heat is also produced by this process (the pump warms up).
Similarly, when air is compressed in the first stage of a gas turbine, it heats up, to
approximately the temperature at which the second stage combustion process works.

Temperature rise is not significant in the case of the flow of air through an EDF unit.
The increase in air pressure caused by the fan does not cause the exhaust duct to bulge out
because of high pressure, nor does the entry duct collapse because of low pressure. Actually
this can happen if the ducts are not properly built, but it is rare. The unit does not get hot, or
freeze (although the electric motor may generate heat because of electrical inefficiency).
FSA is defined as the area of the impeller minus the area masked by the motor. If the
diameter of the EDF unit is d u and the diameter of the motor is d m, then the FSA is defined
as:
FSA = π (du/2)2 – π (dm/2)2 = π (du2 – dm2) / 4

3.2 Defining the properties of an EDF system

One of the problems in calculating useful properties of EDF units, and the resulting
performance of the plane for which this supplies power, is having some measured values to
start from. It is no use having a method of calculating the exhaust velocity from the mass
 EDF theory and practice 12

flow induced by the fan if we have no way of measuring that in the first place.

3.2.1 The geometry of a ducted fan system

The geometrical properties we can measure on the plane are:
the entry duct diameter is d1, and therefore the entry duct area is A1 = π (d1/2)2. Note that in
the case of an entry duct which is not circular, which may be common for many scale-
appearance planes, the area can be measured in other ways, and if needed, the effective
diameter calculated.
The entry duct length is l1. This may be important if duct losses are expected to be
significant,
The fan diameter is df, and therefore the fan area is Af = π (df/2)2
The motor diameter is dm and the motor cross-sectional area is A m.. This is often given by
the manufacturer. Remember to include any thickness added by the motor mounting
cylinder,
The fan-swept-area is Asw = Af – Am.
The exit duct length is l2.
The exit duct final diameter is d2 , and therefore the exit duct area is A2 = π (d2/2)2

3.2.2 The electrical properties of a ducted fan system

The electrical properties we can measure on the bench in static testing are:
the voltage, v, at which the motor operates. e.g. v = 11.4 for a 3s Lipo battery.
the amps, a, drawn by the motor at various throttle settings, including the maximum throttle
position. e.g. 35 amps
The watts, w, being consumed by the motor and fan running at maximum speed. e.g. power
consumed equals 11.4 volts * 35 amps = 399 watts. Note that this is not the same as the
power produced (thrust generated) by the EDF unit because of inefficiencies in the
conversion of electrical power into useful work. We shall return to the question of the
efficiency of the EDF system later.

3.2.3 The aerodynamic properties of a ducted fan system

The aerodynamic properties we can usefully measure in the air, when the model is in flight,
are:
the airspeed of the plane, u, in metres per sec, m/sec. This can be measured by a pitot tube
airspeed indicator (as on full size planes) and recorded by an onboard logger.
The voltage, amperage and wattage of the system if we have an on-board logging system
with appropriate sensors.
 EDF theory and practice 13

The airflow through and over a ducted fan system is shown in the following diagram. The
streamlines represent lines of equal pressure and denote the path taken by a individual air
particle.

Illustration 7: Normal airstream flow through a Fan unit at flying speed

3.3 Calculating the Thrust of an EDF system

This section is a brief summary of the Klaus Scharnhorst paper. The EDF system is
considered (for the moment) to be a single “back box” system. Looked at from the outside
it has an entry and an exit, through which air passes, and which producs thrust. We don't
need to consider the internals in too much detail at the moment.

To ensure that the values we are calculating are consistent and meningful, we need to keep
track of the dimensions of the various quantities using dimensional analysis. In this analysis
the following terms are used:

• M for a quantity of mass. The unit of mass is the kilogram,

 EDF theory and practice 14

• L for a quantity of length. The unit of length is the metre,

• T for a quantity of time. The uit of tim is the second.

Thus a quantity of area is L2, and quantity of volume is L3. Velocity has dimensions LT-1.
Mass flow has dimensions MT-1, and force MLT-1 which we call Newtons. Power has
dimensions of ML2T-2 which we call watts.

3.3.1 Conservation of mass

When the plane is in straight and level flight it is not accelerating or decelerating,:

The inlet area is Ai and air enters this with velocity v i (the speed of the model). The volume
of air passing through this in unit time is

Qi = Ai vi.

With dimension L3T-1. Note that area is measured in square metres, or m 2, and velocity in
m/sec.

The air entering must exit somewhere. The volume of air exiting in unit time is

Qe = Ae ve

where Ae is the exit area and ve is the exit velocity. Since there is no combustion, or
temperature change involved, we can assume that air is incompressible. Now, as a result of
the principle of continuity for fluid flow:

Qe = Qi .

Therefore, re-arranging this equation, we can arrive at an equation for the velocity of the
exhaust, which is:

ve = (Ai /Ae) * vi

3.3.2 Mass and thrust

The mass of air being moved is

M = ρ Qi
 EDF theory and practice 15

where ρ is the density of air ( = 1.224 kg/m3 at Standard Sea Level). This has dimensions of
MT-1. Mass flow is measured in kg/sec.

The speed up of the air produced by the fan is the difference between the speed of the
incoming air and the speed of the exiting air, or

dv = ve – vi .

The thrust is the force produced by the extra moving air and is given by the momentum
increase of the moving air mass, which is the mass being moved multiplied by the increase
in speed, or

T = M dv.

Thrust is a force which has dimensions of MLT-1 and is measured in Newtons. One Newton
is the force of an apple falling under gravity, or about 3.5 ounces = 102 grams.

3.3.3 The power needed for flight

We can denote the power needed for flight as P flight which is a force applied at a given speed.
It therefore has the dimensions of MLT-1 * LT-1 or in total ML2T-2 which is called watts. The
power needed for flight, measured in watts, at this speed is:

Pflight = T vi

3.3.4 Balancing the forces acting on the EDF system

Another way of looking at the EDF system is to calculate the force produced by the fan in
terms of all the forces acting on the system. By the principle of conservation of energy
(energy may be neither created nor destroyed) the forces must sum to zero when all are
considered.

The power gained by the system because of the input of flowing air is P gain, is given by the
kinetic energy of the mass of air entering the system. The kinetic energy of an object (here,
the moving mass of air) is given by Newton's laws of motion as a mass for a distance, times
an acceleration. It therefore has dimensions of M * L * LT-2 = ML2T-2, which is in watts . It
is defined as:

Pgain = 1/2 M vi2

The power lost from the system, Ploss, caused by ejecting a mass of flowing air is, similarly,
 EDF theory and practice 16

given by the kinetic energy of the exhaust, and is:

Ploss = 1/2 M ve2

So, using the conservation principle, we have

Pfan - Ploss + Pgain = 0

The power produced by the system, Pfan, which is due to the fan working away inside is
again measured in watts, and is, therefore:
Pfan = Ploss - Pgain

3.3.5 Finding the efficiency of the EDF system

The "efficiency" of the fan is the nearness to which the energy required is matched by the
energy provided, or
Efficiency = Pflight / Pfan.

As a percentage this is usually around 80%.

The Power required from the motor, Pmotor, also depends on the losses in the system due to
friction in the duct, restrictions in the geometry of the duct and so on. KS assumes (from an
analysis of pipe flow theory, etc..) that P motot = Pfan/0.85 is a reasonable approximation. The
efficiency of the motor is therefore:

efficiency(motor) = Pfan / Pmotor = 0.85 approx.

3.4 A summary of what has been achieved

The theory outlined above is remarkable in that it has created a logically reasoned argument
which connects the static properties of the EDF system to the power budget of the system in
operation. This is by no means a small achievement.

In summary we can describe this theory as a method which requires ONLY the size of the
entry and exit ducts, AND the speed of the model in level flight, to determine the power
required for flight. From a consideration of other physical properties we can also find the
efficiency of the motor.
 EDF theory and practice 17

In terms of the flying performance of the EDF powered plane we can say that:

IF the plane is of the given physical specification, AND if flies at the given speed,
THEN the thrust required for flight is the one calculated.

To use this theory to calculate a useful design property such as the power required from the
motor (ie which motor to choose to install):

Calculate the power required for flight from the geometry of the ducts and the
assumed flight speed. Then calculate the required power of the fan,and finally the
required power of the motor.

Note that these considerations only apply to the plane in straight and level flight at its
cruising speed. If the plane is to take off from the ground and climb to a reasonable flying
height, and if it is expected to do some interesting aerobatics, then of course more power
will be required. So,

The power calculated is the minimum required for flight.

3.5 An Example – the West Wings Bae Hawk in level flight

This example uses observed values from my West Wings BAE Hawk in actual flight.

The area of the inlet is 3667 mm2 ( = 0.003667 m2 ) and the area of the outlet is 2463 mm 2
( = 0.003667 m2 )

But what is the speed of flight ? It is much faster than my scale Tucano (which flies around
60mph as measured by on-board EagleTree telemetry) , but not as fast as a pylon racer
(known by timing trials to be of the order of 180mph). This creates a difficulty (which we
could resolve by fitting an [Eagle Tree] logger measuring airspeed). For the moment let's
assume it could be 37m/sec = 81 mph as a reasonable guess.

Then, the volume of air being consumed at the inlet, per second, is Q = 0.14 m 3 with a
corresponding mass flow of M = 0.17 kg/sec.

So, calculating the speed of the exhaust we obtain ve = 55.1m/sec (about 110 mph)

The force acting on the plane is the thrust T = 3.0 Newtons (=10.8 oz = .307 grams).

The power needed for flight is then Pflight = 111.2 watts.

From the conservation of energy, the power of the fan is P fan = 138.4 watts, assuming a fan
system efficiency of 80% .

Assuming a motor efficiency of 85% we estimate that the required power of the motor is
Pmotor = 163 watts.
 EDF theory and practice 18

The actual power consumption in flight, as measured by total battery depletion over flight
time, is obtained as follows: a 3s 2200mAh battery was exhausted in a 5 minute flight
consisting of a take off run at full throttle, some level flight at two thirds throttle, as a
landing approach and touchdown at one third throttle. So,

2200mAh = 2.2Ah = 132 Amp minutes.

The average amperage over the flight is therefore 132 / 5 amps = 26.4 amps.
A 3s battery at 11.4 volts, so average consumption is 26.4 *11.4 = 301 watts.

The efficiency of the motor therefore seems to be 45% rather than 85%, why is this ?
Clearly, some loss of power goes to making the ESC hot, the motor hot, the battery hot, etc.
There is some measure of agreement between theory and practice, but perhaps the difference
is due to over use of maximum throttle in the take-off and climb phase of flight before the
straight and level test began. Consequently, the duration based calculation of power
required is an over-estimate. However, every flight requires a take-off phase, so the 45%
average efficiency may be a more realistic practical guide.

3.6 The theory of Static thrust

Note that the figures given in the last example are NOT the figures for static thrust and
power consumption as obtained on the bench. And yet the static thrust on the bench is the
only thing we can measure prior to the maiden flight. If we want to be sure that the plane
will fly on its maiden flight we need to know the relationship between the static thrust
measurement and the dynamic thrust expected in flight.

The whole problem of understanding the STATIC thrust is also analysed by KS.

The important point is that we can EASILY measure static thrust with a dynamometer on the
bench, and we can easily measure static power consumption with an ammeter on the bench,
but we cannot easily measure either of these properties in the dynamic flight environment.
Yet dynamic thrust is what we want to know, since that determines model performance.
What we need is to calculate dynamic thrust from static thrust.

3.6.1 Assumptions and methodology

The EDF system is assumed to be held in a static position for testing. For example the
complete plane, ducts and fan, can be anchored to a workbench. When the system is
powered up and the fan made to pump air out of the exhaust duct the airflow through and
around it will be as shown in the following diagram:
 EDF theory and practice 19

Illustration 8: Airstream through an EDF unit when it is not moving (static)

Note, especially, that while the exhaust airflow looks much as it did in the dynamic case, the
inlet flow looks very different. Air is being drawn from a large volume around the inlet
opening, which implies that the effective inlet area is much larger, and the velocity of the
incoming air is correspondingly much slower, than in the dynamic case.
The methodology of this section is to work backwards through the equations we developed
above in order to find the value of v e from the observation of the power consumed in the
bench test. Once we have an estimate of the exhaust velocity we can calculate the thrust in
the same way as before.

3.6.2 Assumptions made for the static analysis

Of course the continuity requirement means that an amount of air must be drawn in at the
front and the same quantity expelled at the back. However, the difference between the inlet
velocity and the exhaust velocity is now much more extreme. We will therefore make the
approximation that dv → v e (read this as “dv tends to the value of v e) and in the limit we
can assume that dv = ve.
Also, we will assume that since v i is very low, the kinetic energy gained by the system as
result of the incoming air, P gain, is very low. In fact, since the air entering the system is
doing so as a result of work done by the fan (sucking air in from a region of still air around)
that it is reasonable to assume that Pgain = 0.
Finally, we will assume that the electrical power consumed by the motor is the same as that
consumed at the same throttle setting in flight. But note that this is a quantity which we can
also measure on the bench since we can connect an Ammeter / Volt meter in between the
battery and the ESC.
 EDF theory and practice 20

3.6.3 Calculating the exhaust velocity from the system power

From the energy balance equation we can substitute Pgain = 0 and obtain
Pfan = Ploss – Pgain = Ploss
In the assumptions above we decided that it was reasonable to make P fan(dynamic) =
Pfan(static) so that now we know the value of Ploss(static).
From the kinetic energy definition we know that
Ploss = ½ M ve2 from the kinetic energy equation
= ½ Qe ρ ve2 from M = Qe ρ
= ½ Ae ve ρ ve2 from Qe = Ae ve
= ½ Ae ρ ve3 by collecting terms

Therefore, by re-arranging this equation, and solving for ve we obtain:

ve3 = (2 Ploss) / ( Ae ρ ) by re-arranging, and then
ve = { (2 Ploss) / ( Ae ρ ) }-3 by taking the cube root of the expression.

This equation for the value of the exhaust velocity in the static case was first developed by
Klaus Scharnhorst in the 1980's and published in various magazine articles. It has probably
not got the attention it deserves as a truly remarkable step forward, maybe due to the length
of development which necessarily proceeds it, and which has been reproduced here.

3.6.4 Calculating Static Thrust

Now that we have an expression for the exhaust velocity in the static case it is simple to find
the thrust produced in the static case, since T = M * dv as before, and we have made the
assumption that dv = ve at least approximately. So we now have:
T(static) = M ve

3.7 An example – the BAe Hawk in static test

The Bae Hawk was anchored to the bench and the motor run to perform the static test. In
this configuration the battery is connected to a meter which us then connected to the ESC.
The meter displays instantaneous volts, amps and watts on the main battery circuit.

3.7.1 Using the static case theory

First calculate the expected (dynamic) thrust from the results of the dynamic case as
calculated above. We accept that the motor should use the same power in the static case as
 EDF theory and practice 21

in the dynamic case, that is 163 watts. It has the same efficiency as before, 85%, and so the
power required by the system is Psys = 138 watts.
We have assumed that Pgain = 0, and so Ploss = Psys – Pgain = 138 watts.
Using the equation developed above we can make the following steps:
ve = { (2 Ploss) / ( Ae ρ ) }-3 = 45.1 m/sec
The first thing to note is that the (static) exit velocity calculated is lower in this case than in
the dynamic case – why is this ? Remember that in the dynamic case air is entering the EDF
at flying speed and the EDF is accelerating the air a bit more, whereas in the static case air is
entering the EDF at a very low speed. This slow air is then being accelerated greatly, but in
total not to the same speed as is achieved in the dynamic case.
We can also calculate the mass flow, M(static), and the static thrust, T(static) using the
standard equations:
M(static) = Ae ve ρ = 0.14 k/sec
T(static) = M dv = 6.139 Newtons = 22 oz = 626 grams
Notice that the mass flow is slightly less in the static case than in the dynamic case (as a
result of the lower exit velocity) but the surprise is that the static thrust is greater than the
dynamic thrust. In fact it is double. What this seems to be saying is that the EDF unit that
powers the plane to fly straight and level at the specified speed with a thrust of about 3
Newtons should produce a static thrust of over 6 Newtons on the bench, if we did that test.
An alternative way of expressing it would be to say that if this EDF system is observed to
produce 6 Newtons of thrust on the bench then, when flying at the given speed, it will be
producing only 3 Newtons of thrust in the air. It is another set of considerations now to
decide whether the 3 Newtons is sufficient to fly the aircraft.

3.7.2 Using the observed current consumption

Since we can run an EDF system on the bench we can easily measure the current it actually
consumes, and therefore gain an insight into whether some of our assumptions are correct.
In the case of the Bae Hawk, runnning it up to maximum throttle gives a set of values for the
current used as follows:
Battery type Volts Amps Watts
3S 2200mAh 40C 10.3 27 283
3S 2600mAh 40C 10.3 30 317
3S 4000mAh 30C 10.2 32 353
4S 2200mAh 60C 13.2 39 514
4S 3200mAh 25C 12.8 38 488

These results clearly show that the bigger the battery capacity, and the higher the battery
voltage, the more watts are consumed (and therefore hopefully the more thrust is produced).
Additionally the C-rating of the battery has an influence on how much current the battery
 EDF theory and practice 22

can deliver in a given time.

The target battery for the plane is the 3S 2600mAh size of battery. Larger batteries carry an
additional weaight penalty, so the choice of battery is a compromise between the weight and
the thrust produced.
If we take the 317 watts as a typical bench mark it would seem that the efficiency of the
EDF unit (motor, fan and duct system) is really 138 / 317 = 44%, not the 85% previously
assumed. There are a couple of caveats around this conclusion, however:
1. the bench test was full throttle, whereas the plane flying straight and level will
(hopefully) not be at full throttle. Until we fly it we don't actually know what throttle
level it will require. It will probably be about 2/3rds throttle, and the power
consumed with therefore be more like 200 – 250 watts, indicating an efficiency of
about 60%.
2. the motor and fan unit have unknown inefficiencies. While the (brushless) motor
may be 90% efficient the fan blades themselves may be inefficient in turning this
into thrust, there may be leakages around the edges of the blades, and so on.
3. The ducts are of unknown efficiency. Clearly the rather long, and curved intake
ducts are responsible for some losses (over the straight nacelle shape shown in the
diagrams. The question of duct efficiency needs to be dealt with separately.

3.7.3 Using the observed thrust measurement

Using the selected battery (at the selected current consumption) the static thrust can be
measured by placing the plane on a weight balance (digital kitchen scales are appropriate
for this test). Note that the plane should be mounted vertically on the scales, the scale's zero
reset, and then power applied. The thrust produced is then observed as additional weight
which is read off from the display panel of the scales.
In the case of the Bae Hawk the measured (static) thrust was 440 grams = 4.313 Newtons =
15.3 oz at 317 watts for full throttle.
This value is a little less than the calculated value of 6.139 Newtons which suggests that the
theory is only approximately correct. No doubt some additional considerations need to be
taken into account (and will be discussed below).
Now working backwards through the methodology we can see that the assumption about the
flying speed may have been wrong. We have (as yet) no actual observation of this value
which was assumed (guessed) in the first instance. If the flying speed had instead been
estimated at 31 m/sec = 68 mph then the static thrust calculated would be 4.309 Newtons,
and the corresponding dynamic thrust would have been 2.11 Newtons.
In order to verify this we need to be able to do one of two things: either:
• calculate (by airfoil lift/drag considerations) whether enough lift would be produced
at the lower flying speed, or
• measure the actual speed in flight.
 EDF theory and practice 23

3.7.4 A possible adustment to the KS theory

One of the well-voiced objections to the KS theory is that it seems to imply that in the static
case the entry velocity of incoming air is zero, or effectively so. Clearly this cannot be
literally true, although we know the value of vi is small.
In concept the EDF unit in the static case is operating a little like a vacuum cleaner, sucking
air in at the front and expelling it at the back. The airstream diagram above shows that air is
drawn from all around the EDF inlet, and the rounded lips of the intake duct are clearly
essential in guiding the incoming in the right way. The result of this is that the effective
intake area is much larger than the actual intake size (which is the controlling dimension in
the dynamic case). Vacuum cleaners make use of this fact by reducing the inlet nozzle size
to a value as small as is practicable.
If we make the reasonable assumption that the effective diameter of the intake is about
double the actual size in the static case, we can then make an estimate of the average intake
velocity, since we have
Ai * vi = Ae * ve
and so by re-arranging,
vi = (Ae / Ai ) * ve
The new value of the inlet area can be calculated from the area of a circle with a diameter of
twice the original.
In the case of the Bae Hawk the inlet diameter is now 2 * 68.3 mm = 136.7 mm and the
corresponding inlet area is 14668 sq mm = 0.014668 sq m. Using the formula above for the
case of flight speed = 37 m/sec and dynamic thrust = 3.007 Newtons we obtain the exit
velocity of 45.1 m/sec as before. But now if we take account of dv = v e – vi we obtain a
value of vi = 7.6 m/sec and therefore dv = 45.1 – 7.6 = 37.5 m/sec. The revised value of
static thrust is now T(static) = M * dv = 0.14 * 37.5 = 5.108 Newtons.
As before if we adjust the estimate of the flying speed until we get agreement we arrive at a
figure of flight speed = 34 m/sec giving a T(static) of 4.313 Newtons corresponding to a
dynamic thrust of 2.539 Newtons.

3.8 Conclusions on the utility of the static test method

The following conclusions on the utility of the static analysis can be made:
1. The dynamic calculations, and the associated static calculation, can be conisdered a
generally good guide to basics of EDF operation. We have a logical understanding
of the mechanism of the EDF unit and its context in an actual plane. We have a
systematic was of estimating the power needed to make a plane fly at a given speed,
and we have a way of relating the measurement on the bench (the static case) to what
might happen in flight.
2. On the other hand, the calculated figures are of the right order of magnitude but
some significant discrepancies with respect to observations. Some of this is due to
the difficulty of making accurate measurements on a real plane, and some of it is
due to assumptions that have been made during the calculations. Clearly both of
these factors need to be refined.
 EDF theory and practice 24

3. The surprising result that static thrust > dynamic thrust means that simply taking
the result of a bench test and using it to decide on motor selection, battery selection
and so on may be misleading. What has been missing all along is an aerodynamic
calculation of how much lift is required to fly the plane in the first place, since
lift = weight in straight and level flight.
The lift generated by the airfoil will depend on the profile of the airfoil (its
coefficient of lift) and the size of the airfoil (its planform area). These are
calculations that should be done initially. Also, the speed of the plane will depend on
the drag induced by the airfoil, since
thrust = drag in straight and level flight
The total drag of the airframe will depend largely on the induced drag of the airfoil,
but also on the profile drag of the fuselage and other factors.

4 A Critique of the KS theory

While there is much to commend the KS theory there are a number of valid objections or
criticisms that can be raised. Among these are:
1. the thrust values calculated seem to be rather low. Is this because EDF is an
inefficient way of pwoering a plane, or because the estimates are incorrect ?
2. The power required for flight, as calculated, seems to be on the low side. General
observation of models of all types flying in normal conditions seem to suggest that
more power should be needed. It is obvious that applying more power will result in
faster flight, and it may simply be that the model aircraft fraternity prefers to fly
over-powered, rather than under=powered planes, and thus has a high expectation for
thrust which is needed.
3. A serious compaint is, however, the following: If in the geometry of the plane it is
designed that Ai = Ae (and this happens at least when the EDF fan unit is used
without ducts, as is often the case on simple models) then calculation would show
that ve = vi so dv= 0. However, in that case it would result that T = 0 and the plane
would NOT fly. However, common observation is that such planes DO fly, and often
very successfully. This is an important problem. One resolution of this is that the
theory ignores (or under rates) power applied at propellor.

5 References
Klaus Scharnhorst, 2005 which is available from SMAC
Propeller thrust calculation http://www.grc.nasa.gov/WWW/k-12/airplane/propth.html
Duct design http://electricjet.blogspot.co.uk/2009/04/electric-ducted-fans-duct-design.html

6 Glossary
Brushless an alternating current, synchronous, electric motor
duct the tubes leading air to the EDF and from it
EDF Electric Ducted Fan
IC Internal combustion [engine, usually methanol powered, also using nitromethane]