This is only a preview of the March 1998 issue of Silicon Chip. You can view 43 of the 96 pages in the full issue, including the advertisments. For full access, purchase the issue for $10.00 or subscribe for access to the latest issues. Articles in this series:
Items relevant to "Sustain Unit For Electric Guitars":
Items relevant to "Multi-Purpose Fast Battery Charger; Pt.2":
Items relevant to "Command Control For Model Railways; Pt.3":
Items relevant to "PC-Controlled Liquid Crystal Display Board":
Articles in this series:
Articles in this series:
|
RADIO CONTROL
BY BOB YOUNG
Jet engines in model aircraft; Pt.3
Last month we looked at the input of air to the
jet engine via the compressor and we established
that the radial or centrifugal compressor was
the best choice for model jet engines. This month
we look at the diffuser, combustion chamber and
turbine.
Having accelerated the air to the
outside edge of the com
p ressor
wheel, it is now time to collect the
air, achieve the full measure of com
pression possible and remove as much
turbulence in preparation for the entry
of the air to the combustion chamber.
To do this we use a diffuser (stator or
collector ring). Here we are faced with
some difficult choices.
Essentially, diffusers may be divid
ed into two categories, bladeless and
bladed diffusers.
Professional engine men refer to
the bladeless diffuser as a bladeless
annular space. They are easy to man
ufacture and can be quite efficient if
designed correctly. As there are no
blades there is no flow breakaway.
The main advantage of the bladeless
diffuser is that the compressor as a
whole has excellent self-regulatory
characteristics.
The disadvantages are that they
must be of a larger diameter than the
bladed diffuser and that they cannot
smooth out the twisting motion im
parted to the airflow by the compres
sor. As the overall diameter of the
finished engine is very important to
the modeller, the increased diameter
is a serious drawback.
Therefore, the best solution for a
model turbine is the bladed diffuser
with the blades set back from the com
pressor, leaving a clear annular space
where the airspeeds are the highest
84 Silicon Chip
and most unevenly distributed. The
wedge-shaped diffuser blade can be
useful for mounting the diffuser to
the outer casing. Screws can be fair
ed into the wedge with a minimum
disturbance to the airflow.
The action whereby the air is
slowed and compressed in the diffus
er is a complex mix of subtle factors.
If you are interested in model aircraft
jet engines, this book by Thomas
Kamps and entitled Model Jet Engines
is a good one to have. It’s published
by Traplet Publications UK (ISBN 0
9510589 9 1).
The spiral law governing action of
a fluid in a centrifugal compressor
states that the product of the radius
(r) in the diffuser system and the
speed (cu) in the peripheral direction
is constant (spiral law: r x cu = con
stant). Thus, as the radius increases,
the speed is reduced.
This basic law plays an important
role in model jet engines and an
interesting analog is found in the
common tea cup. As we stir a cup
of tea we speed up the centre of the
mixture but at the edge the speed is
the slowest. This causes an increase
in the fluid pressure and the level of
the fluid at the wall of the cup rises,
leaving a dish or well at the centre;
the faster the rotational speed, the
deeper the well.
According to Bernoulli’s theorem,
the total energy in the flow must al
ways remain constant. Therefore, if
the speed decreases as the air moves
out, away from the compressor into
the larger diameter diffuser, the pres
sure goes up, as in the tea cup. So
contrary to what one might expect,
compression is due largely to the
centrifugal force applied to the air
leaving the compressor and moving
outwards into a space of a larger di
ameter, and not due to the change in
volume between the compressor and
the diffuser.
True, this change in volume will
also cause a slowing of the airflow
with the subsequent increase in pres
sure but not of the order required.
In fact, if the size of the annular
non-bladed duct is increased, there
is a danger of the airflow breaking
up into turbulence with a severe loss
of efficiency. For this reason, some
bladeless diffusers have a cross sec
tion which tapers to a more narrow
section as the diameter increases,
This close-up view shows the JPX-T-240 turbine engine fitted to Kevin Dodds’
(Tingalpa, Qld) A-10 “wart hog”. The maximum engine speed is 122,000rpm!
Note the discoloration on the rear of the fuselage from the exhaust.
forcing the air to move more quickly
to the outer edge.
Here we come across another dis
advantage of the centrifugal or radial
compressor. To get a worthwhile
increase in diameter in the diffuser
we need to have a diffuser of ap
proximately twice the diameter of
the compressor. Practical experience
has established that this figure may be
cribbed somewhat but a minimum ra
tio is around 1.6 times the diameter of
the compressor wheel if we introduce
vanes or guides into the diffuser. Thus
it is the diffuser that most contributes
to the dumpy appearance of the model
jet engine.
With blades in the diffuser, we now
have real cause for concern because
the path of the air leaving the com
pressor rim is a very complex func
tion. It requires careful calculation
to get the diffuser blades set at the
correct angle so as to minimise flow
breakaway on the guide vanes.
What must be kept in mind with
these engines are the very high air
speeds involved. In an engine using
a shaft speed of 100,000 rpm and a
66mm diameter compressor, the rim
speed (Rs) of the compressor is given
by the formula Rs = n x d x π/60, where
n is the shaft speed and d is the diam
eter. This works out to 345.5 metres
per second or 1243.44km/h. But wait,
I hear you cry, that is in excess of the
speed of sound!
Not so, for we are working with air
at higher temperatures and pressures,
so the speed of sound in the medium
is much higher. Even at rim speeds as
high as 450m/s, the sound barrier can
not be exceeded inside these engines.
However these are phenomenally
high airspeeds and if the diffuser
blades are set incorrectly then there
are serious rami
fications. Unfortu
nately these are the sorts of speeds
required if the Reynolds numbers
are to be moved up into a reasonably
efficient range.
Yet model jet engines can be throt
tled down successfully to much lower
speeds, such is the amount develop
ment work that has been poured into
this the most difficult of all modelling
dreams. If any reader is interested in
a full mathematical analysis of the
model jet engine then there is a very
good book on the subject written by
Thomas Kamps and entitled Model
Jet Engines, available from Traplet
Publications UK (ISBN 0 9510589 9 1).
If we can slow the air by 50% in the
diffuser, we will convert about 75% of
the speed energy to pressure energy, as
the energy in the gas is proportional
to the square of its speed. Thus, re
turning to our motor using a 66mm
compressor at 100,000 rpm, we find
that typical throughput of air will be
about 1.35 - 1.75kg. As the thrust of
the motor rises in proportion to the
throughput, the higher figure is the
more acceptable. At this point the
compressor will deliver a compres
sion ratio of about 1.9:1.
Another important factor in the
compressor/diffuser design is the
expansion angle of the diffuser blades.
The blades start off more close
ly
spaced and gradually move apart as
they move out to the rim of the dif
fuser. This divergence angle is known
as the expansion angle and it plays a
large part in the compression of the
incoming air. Too shallow an angle
will mean more losses as the air will
stay in the duct longer and boundary
layer losses will rise.
Typical expansion angles are
around 15 degrees which calls for 24
blades in the diffuser. Smaller angles
will call for more blades and more
friction losses. For this reason it is
better to use blades which are curved
forward slightly, forming gently wid
ening ducts.
This type of diffuser and a radial
compressor with retro curved blades
will result in an engine capable of
rapid throttle response and which will
be quite resistant to surging.
The compressed and stabilised air
now passes to the combustion cham
ber and we haven’t even got to the
March 1998 85
Table 1: Model Jet Engine Fuels
Densi ty (kg/l )
H0u (MJ/kg)
Boi l i ng Range (oC)
Diesel
Petrol
JP1/Jet A
JP4
Propane
Methanol
0.85
0.76
0.804
0.76
0.5(1)
0.79
42.8
42.5
43.3
>42
46.3
19.5
190-334
80-130
160-260
60-240
-42
65
Fuel tank Capaci ty (ml )
880
990
920
990
1380
(5 mi nutes, 30N thrust)
Fl ammabi l i ty/Fi re
Low
H i gh
Low
H i gh
Very Hi gh
Hazard
(1) Li qui d under pressure; (2) Suffi ci ent for 5 mi nutes of powered fl i ght at a thrust
of 30 Newtons (speci fi c consumpti on = 0.3kg/N/h)
2080
H i gh
Source: Model Jet Engi nes, by Thomas Kamps
hard part yet. Is it any wonder that
the model jet engine took so long to
develop?
The combustion chamber
Single stage turbines and compres
sors take up little space but not so the
combustion chamber. This is why
model jet engines do not look at all
like their full-size cousins from the
outside. Actually, the proportions are
almost reversed. In the full-size motor,
the combustion chamber is a short
section between the compressor and
turbine, whereas in the model engine
the combustion chamber is the largest
component.
There are other difference between
model and full-size turbines in terms
of specific power. Model size com
pressors and turbines are less efficient
than industrial aircraft engines. If the
engine is to run at all, the turbine must
extract most of the available energy
from the exhaust flow at the turbine.
As a result, there is little left in the
residual exhaust flow to produce
thrust. For this reason, the shape of
the tail-cone is vitally important; a
correctly shaped tail-cone can increase
the thrust dramatically.
The low residual thrust combined
with the low compression ratios avail
able in the model engine means that
only 3 - 8% of the energy contained in
the fuel is turned into thrust. Howev
er, due to the low mass of the model
engine, thrust to mass ratios are much
the same as full-size engines. The
drawback in the model engine is fuel
consumption. Modellers wishing to
use a jet in their new model should
leave plenty of space for the fuel tank.
Table 1 shows the most common fuels
suitable for use in model turbines.
Early model jets used propane gas
but there was some risk with this fuel.
Theoretically, the jet engine is not re
86 Silicon Chip
stricted to one type of fuel, the main
requirement being that the maximum
energy is released during combustion.
In practice, most jets are designed to
run on one of the many mineral oil
products commerc ially available.
Alcohol fuels such as methanol are
not suitable due to their low energy
densities. These days, most model en
gine manufacturers choose kerosene.
The design of the combustion cham
ber is critical. If this component falls
short in any way, there are serious
consequences, the most drastic being
the destruction of the motor.
If combustion is uneven, then the
incoming air will not be heated to full
temperature in parts of the combustion
chamber. The enthalpy of this portion
of the air rises only slightly and con
sequently does little work on its way
through the engine. Worse still, to
compensate, the rest of the air must
become that much hotter to keep the
engine running. The result is uneven
speed distribution in the turbine and
lower overall efficiency. In the worst
case, the engine will not run at all.
The purpose of the combustion
chamber is to heat the air in order that
it can do more work when it is decom
pressed than was required to heat it. If
the air is heated during decompression
then this effect is largely nullified.
For this reason, combus
tion must
be contained inside the combustion
chamber as much as possible. If the
flames are too long, they will extend
into the turbine area and the turbine
will overheat. The clue for this prob
lem is high exhaust gas temperature.
Mixture considerations
Stable combustion can only be
achieved if a stoichiometric mixture
is present. This is referred to as an air
surplus of one. A mixture is said to
be rich if the air surplus is less than
one and lean if it is greater than one.
A lean mixture can result in the flame
being blown out if the throttle is closed
suddenly because the compressor is
still delivering a large quantity of air
to a weak flame.
If the mixture is too rich, the flame
burns yellow due to glowing carbon
particles. These cannot be fully burnt
because the necessary oxygen is ab
sent. The result is a layer of soot depos
ited on the combustion chamber walls.
When using kerosene or diesel, sto
ichiometric combustion occurs with a
fuel/air ratio of 14.7:1 and results in
burn temperatures of about 2000°C,
even in model engines. To reduce
this temperature to the desirable 650850°C, cool air must be introduced by
dividing the combustion chamber into
two parts. These are called the primary
and secondary zones.
In full-size engines, high combus
tion chamber temperatures (900°C)
are a real problem and nickel based
alloys (Nimonic or Inconel) are the
usual solution. In a model engine, this
heating is not so severe due to the low
er compression ratios. This means the
air temperature is not as high from the
compressor and therefore the cooling
effect is much greater.
A neat trick is to drill small holes
(1-1.5mm dia.) in the primary zone to
introduce cooling air. This air forms a
thin, cool boundary layer and protects
the combustion chamber primary zone
from the stoichiometric temperatures.
As a result, V2A sheet steel is OK for
the combustion chambers.
On the other hand, it is desirable
that the cooling air in the secondary
zone penetrates deeper into the com
bustion chamber and this calls for
larger holes. The number, size and
location of the holes has a large effect
on the overall exhaust gas temperature
and temperature distribution and is a
key factor in the engine design.
The aim is to obtain perfectly even
heating with as short a flame length as
possible and with all fuel completely
burnt. The exhaust gas heat must be
directed away from the root of the tur
bine blades, as the stresses are highest
at this point. Most heat is directed
towards the centre of the turbine disc,
towards the shaft and bearings.
By now, the reader should be aware
of why it took so long to make one of
these engines run at all. Did Mr Ball
ever have those engines running in
SC
1947? I really doubt it.
|