You are on page 1of 14

Introduction

In most modern lightweight hovercraIt, the thrust required Ior generating Iorward motion
comes Irom the use oI either a propeller or a commercially available Ian. This document
discusses the methods oI calculating the thrust obtained Irom the latter, and selection issues.
The Iormulae available Ior calculating thrust Irom propellers give misleading results when
they are applied to commercially available Ians. Most suitable Ians are intended to run
within a duct and have blades mounted onto a hub. The hub is usually oI much larger
diameter than a typical propeller boss.
These Ians are normally manuIactured Ior industrial purposes such as air handling, smoke
and Iume extraction etc. ManuIacturers do not expect them to be used as propulsive units Ior
air cushioned vehicles and consequently, do not publish thrust Iigures Ior their Ians. They do
however, provide the volume oI air that the Ian can provide at diIIerent speeds and
pressures.
What is thrust?
Thrust is the Iorce applied by the volume (mass Ilow) oI air passed at the discharge oI the
Ian. The basic equation Ior Gross thrust is given by:-
Where:-
T
g
Gross thrust (exclusive oI drag or losses)
Q
d
Quantity oI air at the discharge
V
d
Discharge velocity (EIIlux velocity) and
7 (Rho) density oI air
Momentum drag
A thrust Ian works by taking still air Irom in Iront oI it and using the Ian blades to increase
its pressure and velocity. II the air at the inlet already has some momentum, the Ian is unable
to increase its velocity by the same amount, this diIIerence is reIerred to as momentum
drag` (D
m
)
Where:-
V
o
the Eree stream Velocity`
T
g
Q
d
x V
d
x 7 (Equation 1)
D
m
Q
d
x V
o
x 7 (Equation 2)
Calculation of thrust in a ducted fan assembly for hovercraft
P. EitzPatrick - HovercraIt Club oI Great Britain (S.E. Branch)
Net thrust
The Net thrust (T
n
) is the gross thrust less the momentum drag (D
m
)
ThereIore Net thrust is given by:- T
n
(Q
d
x V
d
x 7) - (Q
d
x V
o
x 7)
Which can be rewritten as:-
The quantity oI air (Q
d
) can be given by the Ian Area x discharge Velocity (A x V
d
)
ThereIore Equation 1 can be modiIied to give T
g
A x V
d
x V
d
x 7 or:-
And equation 3 can be modiIied to give:-
Note: Convert Newtons (N) to Pound-Eorce (lbI) by dividing by 4.44822
The data required to use these Iormulae is speciIic to the choice oI Ian
Example:-
Eor a 1000mm Multi-Wing 6-12/5Z 40Deg Blade angle and 2200 rpm.
Duct Dia 1000mm (Area 0.785m
2
)
Discharge Velocity 30.04 m/sec
Density oI air 1.22 kg/m
3
Using Equation 4, Gross Thrust (30.04
2
* 0.785 * 1.22)/4.4482 194.29 lbI
Fan selection
This data would need to be acquired Irom the manuIacturer or Ian supplier. In the event that
data is only available at certain rotational speeds (e.g. 1440 rpm or 2990 rpm) it would be
necessary to extrapolate the data required using the standard Ian laws (see 'Understanding
and Selecting LiIt Eans)
Once the horsepower available Ior thrust has been established, the approximate Ian
discharge velocity needs to be chosen to suit the perIormance requirement oI the craIt. This
T
g
V
d
2
x A x 7 |m/sec
2
x m
2
x kg/m
3
| Newtons (Equation 4)
T
n
Q
d
x 7 (V
d
V
o
) (Equation 3)
T
n
V
d
x A x 7 (V
d
V
o
) |m/sec x m
2
x kg/m
3
x (m/sec)| Newtons (Equation 5)
inIormation can then be used to select one or more Ians with the correct duty. II the Ian is to
be used Ior thrust only, no static pressure development is required other than that needed to
overcome the resistance oI the short duct and the obstructions to the Ian (guard, Irames,
rudders etc.). This will typically be in the order oI 50 Pascals (Pa).
The duty oI the Ians selected can then be examined to select the most eIIicient at the
proposed Ian speed. The more eIIicient Ian will provide greater thrust per horsepower.
In general terms, the larger and slower the Ian the more eIIicient it will be. II the Ian is too
large the discharge velocity (V
d
) will be too low and limit the potential speed. Eor this
reason, it is wise to select the minimum V
d
acceptable early in the process.
Static and Dynamic thrust
Static thrust is the measured thrust with the craIt stationary and is equal to the gross thrust.
To Iorm any useIul reIerence it must be measured in still air conditions, i.e. with the Iree
stream velocity Zero.
Eor Dynamic thrust, the losses must be taken into the equation.
Assuming a still day, the air some distance in Iront oI the Ian will have zero velocity, (Eree
stream velocity). Erom equation 2 the momentum drag will also be zero. So the static thrust
Iigure will not alter (except Irom the eIIect oI losses due to any additional obstructions etc.)
II the Iree stream velocity is greater e.g. 10 m/sec the losses must be calculated and
subtracted Irom the static thrust. Using Equation 5, The example above would give:-
30.04 x 0.785 x 1.22 x (30.04 10) /4.4482 129.61 lbI
So in a 22 mph headwind (10 m/sec) the Ian thrust will be reduced to approx. 130 lbI. The
same is true iI the Iree stream velocity is zero but the Iorward speed oI the craIt is 22 mph.
As the craIt picks up Iorward speed the Iree stream velocity eIIectively equals the craIt
speed and the value oI dynamic thrust drops with increasing Iorward speed.
Erom equation 3, T
n
Q
d
x 7 (V
d
V
o
) it can be seen that when the value oI V
o
V
d
the net
velocity will be zero and thereIore the net thrust will also be zero.
Ignoring other losses, such as skirt drag and aerodynamic hull drag, we can see that this
equation gives a maximum theoretical speed Ior the craIt in the example above, oI 30 m/sec
(67 mph). In practice the maximum speed will be lower as these other losses increase
exponentially with speed and Ior a small craIt could equate to 50-100 lbI.
The main beneIit oI this, to demonstrate the need to choose the thrust Ian to match the
purpose to which the craIt is put. A cruising craIt may require a maximum speed oI 44 mph.
with the ability to operate at this speed, in a 22 mph headwind at Iull throttle. The Ian
chosen in the example will allow this perIormance (ignoring the other losses), whereas the
same Ian would not allow a racing craIt to achieve 60 mph in even a light wind. Such a craIt
would require a smaller Ian with greater discharge velocity, which Ior the same power
absorption would need to run at a higher rotational speed to develop a similar mass Ilow.
To achieve the thrust required, it is necessary Ior the Ian to use or absorb power. Eig. 1
shows the chart Ior a common Multi-Wing Ian used on small hovercraIt.
Figure 1. Fan Performance Chart - Multi-Wing 900/6-12/5Z/PAG
Fan 900 /6-12/5Z Area = 0.636 Sqr.Mtr
Pitch SPEED
2300 2400 2500 2600 2700 2800 2900 3000 3100 3200 3300
VoIume (Cu/M/sec)
25 12.77 13.32 13.88 14.43 14.99 15.54 16.10 16.66 17.21 17.77 18.32
30 15.56 16.24 16.91 17.59 18.27 18.94 19.62 20.30 20.97 21.65 22.33
32.5 17.13 17.88 18.62 19.37 20.11 20.86 21.60 22.34 23.09 23.83 24.58
35 18.88 19.70 20.52 21.34 22.16 22.98 23.80 24.62 25.44 26.26 27.08
37.5 19.83 20.69 21.55 22.41 23.28 24.14 25.00 25.86 26.72 27.59 28.45
40 20.94 21.85 22.76 23.67 24.58 25.49 26.40 27.31 28.22 29.13 30.04
45 23.16 24.17 25.17 26.18 27.19 28.19 29.20 30.21 31.21 32.22 33.23
50 25.06 26.15 27.24 28.33 29.42 30.51 31.60 32.69 33.78 34.87 35.96
VeIocity (M/sec)
25 20.07 20.94 21.82 22.69 23.56 24.43 25.31 26.18 27.05 27.93 28.80
30 24.46 25.52 26.59 27.65 28.71 29.78 30.84 31.90 32.97 34.03 35.09
32.5 26.93 28.10 29.27 30.44 31.61 32.78 33.95 35.12 36.29 37.47 38.64
35 29.67 30.96 32.25 33.54 34.83 36.12 37.41 38.70 39.99 41.28 42.57
37.5 31.17 32.52 33.88 35.23 36.59 37.94 39.30 40.65 42.01 43.36 44.72
40 32.91 34.34 35.77 37.21 38.64 40.07 41.50 42.93 44.36 45.79 47.22
45 36.40 37.99 39.57 41.15 42.73 44.32 45.90 47.48 49.06 50.65 52.23
50 39.40 41.11 42.82 44.53 46.25 47.96 49.67 51.38 53.10 54.81 56.52
Thrust (Lbf)
25 69 75 82 88 95 102 110 118 126 134 142
30 103 112 121 131 141 152 163 175 187 199 211
32.5 124 136 147 159 171 184 198 212 226 241 256
35 151 165 179 193 208 224 240 257 274 292 311
37.5 167 182 197 213 230 247 265 284 303 323 343
40 186 202 220 238 256 276 296 316 338 360 383
45 227 248 269 291 313 337 362 387 413 440 468
50 266 290 315 340 367 395 423 453 484 516 548
Power Consumed at Speed/Pitch (H.P.)
25 10.90 12.38 13.99 15.74 17.63 19.66 21.84 24.18 26.68 29.34 32.18
30 16.36 18.59 21.01 23.64 26.47 29.52 32.80 36.31 40.06 44.07 48.33
32.5 21.15 24.03 27.16 30.56 34.22 38.16 42.40 46.94 51.79 56.97 62.48
35 25.78 29.29 33.11 37.24 41.71 46.52 51.68 57.21 63.13 69.44 76.15
37.5 29.69 33.74 38.13 42.89 48.04 53.57 59.52 65.89 72.70 79.97 87.70
40 33.52 38.09 43.05 48.43 54.23 60.49 67.20 74.39 82.08 90.29 99.02
45 47.09 53.51 60.48 68.03 76.18 84.97 94.40 104.51 115.31 126.83 139.10
50 60.66 68.92 77.90 87.63 98.14 109.45 121.60 134.62 148.53 163.38 179.18
Modified Thrust (aIIowing 4 cu/M/sec for Iift)
25 47 53 58 64 70 76 83 89 96 104 111
30 76 84 93 101 111 120 130 140 151 162 174
32.5 95 105 115 126 137 149 161 174 187 200 214
35 119 131 144 157 171 185 200 215 231 248 265
37.5 133 146 160 175 190 206 223 240 258 276 295
40 150 165 181 197 214 232 251 270 290 310 332
45 188 207 226 246 267 289 312 336 360 386 412
50 224 246 268 292 317 343 370 398 427 456 487
Aerodynamic, Skirt and Wave Drag Losses
Drag is the term used to describe the Iorces acting against the craIt, which tend to slow it
down. The largest oI these Iorces are :-
Aerodynamic drag:
This is the resistance the craIt hull and superstructure presents to its progress through the air.
Skirt drag:
Similarly, is the resistance the skirt presents to the surIace over which it is travelling. It is
caused by the physical impingement oI the surIace (water, grass etc) on the skirt material.
Wave drag:
This is the Iorce absorbed in the creation oI waves when running on water. Below hump
speed, a bow-wave is Iormed in Iront oI the skirt and the losses are quite high. Once over
hump the Iorces are reduced, but the eIIect oI the air moving through the cushion causes
some localised disturbance to the surIace, as the craIt moves Iorward. This loss is also
Figure 2. Drag forces and Thrust Loss
encountered over other surIaces where the cushion can disturb the media over which, it is
Ilying including grass, sand and snow. On a smooth hard surIace such as ice, the wave drag
and skirt drag will be reduced considerably, allowing the craIt to achieve a higher maximum
speed.
When the total drag is plotted along with the dynamic thrust, the true maximum speed oI the
craIt can be seen where the value oI dynamic thrust is equal to the total drag. Eigure 2 shows
a typical Eormula 2 type racing craIt with a 65Hp engine and 900mm thrust duct used in still
air conditions i.e. zero Iree stream velocity.
Fan Efficiency
The Ian is adjustable Ior pitch. By examining the section labelled Power Consumed at
Speed/Pitch`, (and reading between the lines) it will be seen that Ior a given horsepower, the
user has a choice oI which blade pitch to use at diIIerent Ian speeds. The chart partially
demonstrates the previous statement made (i.e. 'In general terms, the larger and slower the
fan the more efficient it will be). It can be seen in the highlighted example that this Ian will
give more thrust Ior 33 HP with a slower Ian and steeper pitch.
Eor this particular Ian, the eIIiciency at the highlighted settings is as Iollows:
Many Multi-Wing thrust Ians are working in
the mid-to-low 50 range oI eIIiciencies.
Slower Ians work more eIIiciently, but using
a steeper pitch is not always the most eIIicient
solution. UnIortunately there is no easy way
to assess the eIIiciency oI a range oI Ians.
Because these Ians were originally designed
Ior industrial use, they are rarely working within the most eIIicient part oI the Ian curve. The
overall eIIiciency is dependent on numerous Iactors including the speed, tip clearance, blade
shape and pitch, all oI which eIIect the power absorption. This would be ably demonstrated
iI the Ian blades above were replaced Ior 4Z types
Although the advantage might appear
signiIicant, it would only be so iI the integrity
oI the blade could be relied upon. In practice
the thinner lighter construction oI the 4Z
blade is more likely to distort at high speed
and load than the sturdier 5Z type.
Using the example speeds above, Ior a single speed, the 5Z blade peak eIIiciency is at 40
whereas the 4Z blade peaks at 35
Eans having Iewer blades are also more eIIicient, but a 2 blade Ian Ior example will need to
rotate Iaster to achieve the same mass Ilow and thrust as an equivalent Ian oI the same
diameter. This leads to higher tip speeds and the potential Ior greater noise pollution, This
increased noise is partially oIIset by the reduction in blade numbers which itselI reduces
noise levels.
One oI the characteristics oI the axial Ian, is the manner in which the air is discharged Irom
the blade. The air leaves the Ian blade at an angle, resulting in a rotating column oI air. Only
the axial component oI this column will provide thrust, the rotating component will be lost
and may cause unwanted torque reactions. The use oI Ilow straighteners (or stator blades)
will allow most oI this wasted energy to be reclaimed.
Integrated Lift Systems
On an integrated liIt design, the discharge Irom the Ian is divided. Part oI the lower portion
oI the Ian is used to provide liIt and the rest provides thrust. To Iind the percentage used Ior
thrust, the liIt area could be calculated and subtracted Irom the Ian area, but this rarely gives
accurate results. The reason Ior this is shown in Eig 3. The example shown demonstrates that
there is a diIIerence between the percentage area oI the Ian, separated by the splitter plate
2300
rpm
2400
rpm
2500
rpm
4Z 32.5 57% 57% 57%
4Z 35 61% 61% 61%
4Z 37.5 59% 59% 59%
4Z 40 58% 58% 58%
2300
rpm
2400
rpm
2500
rpm
5Z 35 53% 53% 52%
5Z 37.5 56% 54% 54%
5Z 40 56% 56% 56%
5Z 45 53% 53% 53%
and the percentage oI the volume oI air that will be used. This example shows only a small
diIIerence but this diIIerence can be greater depending on the size oI the Ian and the Iinal
position oI the splitter plate.
This diIIerence is due in part to the Iact that commercial Ian blades are intended to operate at
lower speeds. An ideal Ian blade would produce the same pressure and volume oI air Irom
inIinite points along its length, but in
practice this does not happen because the
outer portion oI the blade is moving at a
higher velocity. Twisting the blade along
its length corrects this to some degree by
allowing the end oI the blade to run at a
shallower pitch, which reduces the
volume, pressure and load produced at
the tip. This technique works well Ior a
particular amount oI twist in a blade
rotating at one speciIic speed. Because
hovercraIt Ians are working outside the
normal envelope that the manuIacturer
intended, the problem remains.
Fig 3. Volume under a splitter Plate
Eor this reason it is better to estimate the volume used by the liIt system (see 'The Principles
oI HovercraIt Design) and then deduct this volume Irom the initial calculation. This was
the method used to provide the ModiIied Thrust` values in Eig 1.
Eor more accurate results the area oI the discharge can also be measured or calculated and
this value used to calculate the discharge velocity. This approach is necessary iI there is any
degree oI diIIusion in the duct, as would occur with a variable splitter plate. i.e. the area
below the splitter may alter, but the thrust discharge area remains constant.
Method of Calculating Duct Areas
When the duct is used only Ior thrust, its area is
simply that Ior a circle, given by the Iormula
tr
2
. When a splitter plate is installed, it might
be required to work out the area oI one or both
sections oI the duct. The representation here in
Eigure 4. Shows the dimensions which are
required to achieve this. h1 can be measured
easily on an existing duct, h2 calculated Irom
h1 and the diameter.
To obtain the area under the splitter, calculate
the area oI the quadrant AOBCA and subtract
the area oI the triangle AOBA.
Fig 4 Splitter Plate Dimensions
The angle O can be calculated Irom the Iollowing Iormula:
Example: A 900mm diameter duct has a splitter oI
748mm long. 0.748/0.900 0.831 the Inverse Sine
oI this number is 56.2 degrees. x 2 112.4 Degrees.
The area oI the quadrant AOBCA is given by the Iormula:
Example: 112.4/360 0.312 x t x 0.450
2
0.198 m
2
(Note: the radius is expressed in metres)
The area oI the triangle AOBA is given by the Iormula:
Example: The distance Irom the duct to the splitter is 200mm (h1)
h2 is thereIore 250mm (0.25 m). 0.25 x 0.748 0.187/2 0.094 m
2
The Area under the Splitter is AOBCA AOBA 0.198 0.094 0.104 m
2
II the thrust area is required, the Iigure obtained above can be deducted Irom the area oI the
entire duct.
Selection of fans in integrated systems
When discussing thrust Ians it was noted that the static pressure development required was
very low (typically in the order oI 50 Pascals about 1 lbI/It
2
).
A small single seat hovercraIt will typically require a cushion pressure between 300600 Pa
(6-12 lbI/It
2
), dependent on weight and payload. Eor an integrated system, the Ian chosen
must be capable oI producing this pressure.
Much conIusion reigns about the relationship between pressure and volume in an integrated
system. Taking as an example the 900/6-12/5Z/2300 rpm Ian selected Ior thrust. Erom the
table shown in Eig.1. it can be seen a 40 pitch will provide a volume oI 20.94 cu.m/sec. But
as was stated earlier, this was against a low resistance oI just 50 Pascals Static pressure.
This same Ian could be Iitted into a 900mm diameter, air conditioning system, where instead
oI a short duct, the air was supplied to a long convoluted pipe. This would oIIer more
resistance to the Ilow oI air, represented by a higher static pressure.
Eigure 5 shows just such an installation, where the system resistance is 750 Pascals and the
Ian is supplying a volume oI 16.6 cu.m/sec. II the system resistance were to change the
working point oI the Ian (shown by the red dot) would move up or down the line depicting
the airIlow. The system resistance could be altered, by introducing a volume control damper
into the ductwork. This would when altered, present a greater or lesser resistance to airIlow
by altering the static pressure oI the system.
The total pressure (Pt) within system is composed oI both, static pressure (Ps) and velocity
pressure (Pv). The static pressure is the pressure, which pushes against the inside oI the duct
Angle O =
1
2 xsin
Splitter
Diameter

| |
|
\ .
2
360
Angle o
r 5
| |
|
\ .
2x
2
h Splitter
(Eor a hovercraIt, the cushion, the ground and all internal Iaces oI the liIt system.) The
velocity pressure is that pressure exerted by the motion oI the air within the system.
Fig. 5 Fan in a Ducted system
Resistance and volume
As the resistance (static pressure) increases, it should be observed that there will be a
corresponding drop in the volume oI air moved by the Ian and vice versa. The shape and
slope oI the Ian curve determines the rate oI change. The Ian is still producing the same total
pressure (Pt), but the static pressure (Ps) and velocity pressures (Pv) alter. The latter in
response to the change in volume and velocity in a Iixed diameter system | Pt Ps Pv |.
This Ian can work eIIectively at pressures between 0 and 980 Pascals, above 980 Pa the Ian
eIIiciency drop markedly, as shown by the eIIiciency curve. The Ian is working at its peak
eIIiciency close to the working point shown.
Although we have already seen that the air flow from the fan is not evenly distributed,
for the sake of convenience, we shall assume that the pressure is.
As the total pressure generated is evenly distributed across the disc oI the Ian, it does not
matter iI we take air Irom the entire disc or just a part oI it. The total pressure supplied will
remain constant, only the volume will change as we use a smaller percentage area oI the
disc.
II we use the same Ian in a hovercraIt, with the Ian still operating at 2300 rpm. We can
install a splitter plate to separate the Ilow. Above the splitter plate the static resistance oI the
short duct will be about 50 Pa and there will be a high air velocity and velocity pressure.
Below the splitter plate there is eIIectively a ducted system with a static pressure equal to
the cushion pressure any system losses. There will consequently be a high static pressure,
low velocity pressure and lower air velocity. The cushion pressure is determined by the mass
to be liIted (i.e. the weight oI the hovercraIt payload) and will not alter signiIicantly
except Ior the small adjustments, which occur due to gravity (see 'Understanding and
Selecting LiIt Eans).
It is important to note: that increasing or decreasing the area under the splitter plate
does not significantly alter the cushion pressure (see below), it will only alter the
volume of air fed to the cushion.
The total pressure within the system will not change irrespective oI the splitter plate
position. II the volume is altered, static pressure will change inversely to the velocity
pressure. However, the velocity pressure within the system is tiny in comparison with the
static pressure, so even large changes oI velocity will only cause very small changes in static
pressure. It is Ior this reason that we can, Ior all practical purposes consider the cushion
pressure to be Iixed.
Fan Solidity
The solidity oI the Ian is the percentage oI the disc space, which is Iilled with blades, but the
term is generally used to reIer to the number oI blades Iitted as a Iraction oI the total
possible. E.g. 3-12 / solidity 12-12 Eull solidity. The greater the solidity oI the Ian the
more power it will absorb and hence Ior a given amount oI power, the pitch must be oI a
shallower angle or the rotational speed reduced.
Eigure 6 shows the Ian curves Ior two similar Ians with diIIerent solidity. Eor each Ian the
curves are given at three diIIerent speeds.
Taking as an example, a small single seat hovercraIt with a cushion pressure oI 500 Pa (10.5
lbI approx.). Eitting an 800mm pure thrust duct would provide approx. 22.5 cu.m/sec oI air
Irom both, the 9-9/4Z Ian or the 6-9/4Z Ian (A1 or B1) when running at 3300 rpm. It can be
seen Irom the chart that the 6-blade Ian will produce a slightly higher volume (and thereIore
thrust) than the 9-blade unit (Shown ringed Yellow). Rising vertically Irom this position, to
the power curves, we can see that power consumption Ior both curves is virtually identical.
The same Ians when Iitted with a splitter plate set to give 20 volume, would produce about
17 cu.m/sec oI air Ior thrust and also be capable oI producing adequate pressure to meet the
500 Pa pressure required Ior liIt (Shown ringed Red). Rising vertically Irom this position, to
the power curves, we can see that the power consumption is now higher Ior the 9-blade Ian
than Ior the 6-blade. This value will be the maximum power drawn Irom the engine.
Eor a practical application on a small hovercraIt the maximum volume required is likely to
be a little less than this example. (see Understanding and Selecting LiIt Eans)
The other two sets oI curves show the same two Ians at mid range speed and idle speed.
They highlight a proIound diIIerence between the Ians, which may eIIect the choice oI Ian.
The curve Ior the 6-blade 1800 rpm Ian (B3), shows that at this speed, it cannot provide the
500 Pa pressure required to liIt the craIt. The 9-blade Ian (A3) however does produce
suIIicient pressure and a volume oI approx. 1.7 cu.m/sec. The Ian will liIt the craIt and
payload at just 1800 rpm although it may not provide suIIicient air to overcome losses, so it
will drag the skirt.
Fig 6. Comparison of fans
Figure 6 compares two similar fans of different solidity, used in an integrated system.
A1 800/9-9/4Z/37.5 deg/ 3300 rpm B1 800/6-9/4Z/40.5 deg/ 3300 rpm
A2 800/9-9/4Z/37.5 deg/ 2500 rpm B2 800/6-9/4Z/40.5 deg/ 2500 rpm
A3 800/9-9/4Z/37.5 deg/ 1800 rpm B3 800/6-9/4Z/40.5 deg/ 1800 rpm
490. The THRUST figure is for the 800mm duct used for thrust onlv. When used as an
integrated duct the thrust figure must have the lift quantitv deducted.
The pitch of each fan is chosen to absorb the same power at maximum thrust
The other curves which show the mid range speed, demonstrates that Ior the same speed, the
6-blade Ian (B2) will use less power but produce less volume at the cushion pressure oI 500
Pa, than will the 9-blade (A2).
Conclusion and choice:
II the craIt were to be used Ior racing or Ior ultimate perIormance the 6-blade Ian would be
the obvious choice. It produces more thrust and almost the same liIt volume as the other Ian.
The low speed perIormance would be immaterial.
II the craIt were to be used Ior cruising and generally at lower speeds, the 9-blade Ian might
be the better choice. It produces the cushion much earlier than the 6-blade Ian and in the mid
range Ian speeds produces a greater volume oI air which may be useIul when negotiating
undulating surIaces at lower speeds.
Dual fan systems
A brieI comment should be made regarding the use oI dual thrust or integrated Ians. II two
Ians are used Ior thrust only, the resultant thrust will be twice that oI a single Ian oI the same
speciIication. The velocity will remain the same as will the rate oI thrust loss with increasing
speed. As it is likely that the craIt weight will increase there will be an increase in cushion
pressure and increased power requirement Ior the cushion. II designed careIully, a small
advantage in increased thrust may be evident in a twin Ian arrangement, compared to a
single Ian oI similar overall discharge area, but at a cost oI increased weight and complexity.
II twin integrated Ians are used with both Ians used to supply cushion air, it is important that
the Ians are selected to work in an eIIicient part oI the Ian curve. II the Ians work outside
this area, the perIormance may become unstable to the point where Ian stall can occur and
possibly cushion collapse. This is particularly likely as the Ian speed Ialls and the Ian
struggles to produce the required pressure (curves A3 and B3 in Eig 6 shows just such a
situation developing). Regardless oI how careIully the Ians are adjusted Ior speed and pitch,
diIIerences in pressure development will occur between the Ians, causing variations in Ilow.
This situation is exacerbated by the turbulent Ilow, which usually occurs in the splitter
opening. Although the air may be Ied into a common plenum, it is quite likely in normal use,
that each Ian provides air mainly to one side oI the craIt (particularly with contra-rotating
Ians). As the Ilow rate becomes unstable this situation may change and volume to one side
may rapidly decrease.
Matching fans to engines
II we look at a graphical representation oI a Ian power absorption curve, something like that
in Iigure 7 will be seen. The curves shown are Ior a Kawasaki KR1s motorcycle engine and
a 900mm 6-12/5Z Multi-Wing Ian.
The power requirement Ior an axial Ian increases with the cube oI the change in speed. e.g.
Example: Eig 7 shows the Ian power
required at 6000 rpm is 11Hp. What
power is required at 9000 rpm?
Power 11Hp x (9000/6000)
3
37.12 Hp (as the graph will conIirm)
Power
New
Power
Original
x
3
New Speed
Original Speed
| |
|
\ .
Figure 7. Engine/Fan power curve
This peaky power output curves oI the
modern 2-stroke engine, are ideally
suited Ior matching with this type oI Ian
curve. As Eig 7 demonstrates there is an
almost constant surplus oI power
available right up to peak engine power
where the curves cross.
Where the curves cross, the Ian is
absorbing all oI the available power and
there is no power available to accelerate
the Ian to higher speeds. It is important
to ensure that this does not happen
Iurther down the curve.
Some older two strokes suIIer Irom
'Torque Trough, which is a dip in
power output mid range. The engine
power curve shown has a small trough
but not suIIicient to cause problems
with this Ian. II the Ian curve were to
intersect the engine power curve the Ian
would hang` at that speed and not pull
through` to the maximum speed.
This particular engine Ian combination has a 3.75:1 drive ratio. II the engine has a Ilatter
power curve (as would a 4-stroke) the drive ratio can be changed to adjust the relationship
between the two curves. II it is desired that the engine does not over-rev, then the Ian curve
must pass through the engine curve at peak rpm.
This may be considered important Ior a racing craIt where peak perIormance is required
Irom the Ian, but may not be desired on a small cruising craIt where the engine will
normally operate at 60-80 oI peak engine rpm.
Notes
The symbol E is used in some texts to denote Axial Thrust.
'Multi-Wing is a trademark oI E.S Anderson ApS oI Denmark.
USEFUL INFORMATION
Conversion Data
Pressure Velocity
1 Pascal
1 Pascal
1 Pascal

0.0208854 lbI/It
2
0.004014 in H
2
O
0.101972 mm H
2
O
1 m/sec
1 m/sec
1 m/sec

3.28084 It/sec
196.85 It/min
2.236 mph
1 lbI/It
2
47.8803 Pascals 1 It/sec 0.304 m/sec
1 millibar
1 in H
2
O
1 mm H
2
O

100 Pascals
249.089 Pa
9.80665 Pa
1 It/min
1 mph
1 mph

5.08 m/sec
0.447 m/sec
1.60934 Kph
Volume Force
1 m
3
/Sec 2118.88 It
3
min (cIm) 1 Newton 0.224 809 lbI
1 m
3
/Sec 35.3147 It
3
sec (cIs) 1 Newton 0.101972 kgI
1 cIm 1.699 m
3
hr 1 lbI 4.44822 N
1 lbI 0.453 592 kgI
Area
Power
1 m
2
10.763 It
2
1 It
2
0.0929 m
2
1 Kilowatt 1.3410 Hp (UK)
1 Kilowatt 1.359 Hp (Metric)
Density
1 Hp (UK) 0.74570 kw
1 Hp (Metric) 0.735499 kw
1 lb/It
3
16.0185 kg/m
3
1 kg/m
3
0.06242 lb/It
3
Useful Formulae
Tip Speed (V
t
)
60
t
n d
J
5
=
Axial Thrust (E)
2 2
2
4
4
s
q d
F P
d
7 5
5
= +
EIIiciency ()
t
p q
P
=
Area under a chord (splitter plate)
A
2
2 1 2 2
2 2
2
tan 1
R
R h R h
h


| |


|

\ .
]
Terms:
Tip Speed (V
t
) m/sec
Rotational Speed (n) RPM
Diameter (d) m
Density oI air ( 7 ) kg/m
3
AirIlow (volume) (q) m
3
/sec
Power (P) Watts
When h
2
R, A0 and when h
2
0, A tR
2
Disclaimer
Readers are Iree to use this document as they wish and pass it to others in its entirety. The authors` permission should be
sought iI it is intended to re-publish the document or any part thereoI. This document represents the views and experience
oI the author only. No warranty is given or implied as to the accuracy oI the content or liability accepted Ior diIIiculties
encountered by it use. Compiled April 2003.

You might also like