This is a more detailed version of an earlier posting on this topic. That one, "Inlets Capable of Supersonic Flight Use" dated 21 August 2026 below, presumed the reader was an amateur. This one presumes the reader actually knows something about compressible flow with shock waves, and about how inlet performance data are usually book-kept and used.
-----
Supersonic Inlets for Ramjets
Supersonic inlets have spikes or ramps as external
compression features that act before the airstream is captured at the cowl
lip, going to internal flow from
there. Spikes go with axisymmetric
inlets that have circular cross sections,
ramps with two-dimensional (2-D) inlets that have a rectangular cross
section. The chin inlet is a variation
on the nose-mounted axisymmetric spike inlet,
being just a sector of the full circle,
and using part of the vehicle nose as its external compression spike
surface. And,
there are a variety of side-mounted inlets, both round or rectangular in cross
section, or even half-round.
The external compression surfaces create oblique shock waves
with a flow deflection off of the axial direction. The flow behind an oblique shock wave is
still supersonic, and the total pressure
loss across the wave is quite modest.
There is at least one, and most
often, two (or more), such oblique waves in a typical supersonic
inlet. The last (“terminal”) shock wave
is a normal shock, perpendicular to the
flow, creating subsonic flow behind the
wave, and with a much larger total
pressure loss across the wave, when compared
to an oblique shock at the same Mach number.
See Ref.1 for equations,
tables, and charts.
By choosing the deflection angles, their geometric
locations, and the number of waves
correctly, about the same total pressure
loss ratio across each wave can be had.
The sum of these losses is less than the loss across a single normal
shock at the full flight Mach number (or even more if supercritical), as in a pitot/normal shock inlet.
All of these inlets have a minimum cross section area called
the “inlet throat”, located at, or just downstream of, the inlet capture location into internal flow. They all have divergent diffuser passages
downstream of that throat, where
subsonic flow is finally achieved. With
the throat at the cowl lip, there is
no internal flow contraction, these
are called “all-external compression” inlets.
With the throat downstream of the cowl lip, there is internal flow contraction for
more compression, these are called
“mixed-compression” inlets. The local
Mach number at this inlet throat is always still somewhat supersonic. There are no feasible all-internal
compression inlets, not with any
technology known today! In-line diffusers, yes.
Inlets, no.
All of these inlets share similar-looking characteristic
curve shapes for what is called the critical total pressure recovery ratio PRCR, the critical captured streamtube area ratio
ARCR, and the additive drag coefficient
CDadd. The item of greatest
influence on the shapes of these curves is the shock-on-lip Mach number, that being the speed (measured as flight Mach
number) at which the oblique shock waves created by the external compression
surfaces are focused to fall exactly upon the cowl lip, by deliberate design.
There is a modest slope change from below shock-on-lip to
faster than shock-on-lip in the PRCR curve,
yet it still decreases from high values to low values as flight speed
increases. There is a sharp slope change
at shock-on-lip speed in the ARCR curve.
Below shock-on-lip, the recovered
streamtube is much smaller than what is theoretically swept out by the cowl
lip. Above shock on lip, the recovered streamtube is very close to
that theoretical swept-out value. What
that really means is that the inlet can ingest lots of air flow above
shock-on-lip speed, and very little air
flow below shock-on-lip speed. The
first priority by far for ramjet application is maximum possible
ingested air flow rate!
There is also a sharp slope change in the CDadd
curve at shock -on-lip speed. CDadd
is defined as the drag coefficient of “additive drag”, based on the swept-out inlet capture area AC. The higher values at the lower speeds are the
effect of minimally-decelerated still-supersonic air diverted past the cowl lip
by the streamtube geometry below shock-on-lip.
This is because the local streamlines behind the oblique shock waves must
run parallel to the external compression surfaces that created the waves. Above shock-on-lip speed, the CDadd values are low, and in the case of nose-mounted inlets, often zero.
All of these behaviors and characteristics are illustrated
in Figure 1.
Figure 1 – Basic Characteristics of Supersonic Inlets with
External Compression
These components must be used differently with turbojets
versus ramjets, because of the
quite-different characteristics of those engines. The turbojet is almost like a
positive-displacement pump, in that, at any given rotation speed and air
density, there is only one engine air
massflow! Rotation speed is determined
by the throttle setting. The turbojet
inlet must deliver the exact air flow rate that the engine demands, regardless of the flight speed and the
inlet ARCR recovery characteristic curve!
To do that, it must
operate “subcritically”, with the
terminal shock “unswallowed”, and thus located
on the external compression surface in front of the cowl lip, as shown in Figure 2. There is a sometimes-large spillage of
subsonic air around the cowl lip, in
order to deliver the “right” ingested air massflow rate to the turbojet engine.
In contrast, the
ramjet acts like a simple obstruction,
producing a pressure at the entrance to the min obstruction area that
depends only on the total air and fuel massflow rates fed to the engine, and the combusted temperature and properties
corresponding to them. The thrust of the
ramjet depends most strongly on the amount of air flow ingested! That means the inlet must be operated
supercritically (with a swallowed terminal shock), for ingesting the maximum airflow! The more air,
the higher the combustor pressure,
and the higher the combustor pressure,
the greater that nozzle thrust force!
This is also shown in Figure 2.
Figure 2 – The Same Components Must Be Used Differently for
Turbojets and Ramjets
Exactly how the ramjet engine-inlet balance works is shown
in Figure 3, for which the most
important behavior is supercritical versus subcritical operation. The inlet characteristics are represented by
the cross plot of PRCR versus ARCR at any one flight Mach number, very similar to the one shown in Figure 1, just with some extra details added here. The corresponding depictions of the inlet
flow field and the distribution of conditions through the engine are shown in
the figure, one each for subcritical
operation, operation exactly at the
critical inlet point, and for supercritical
operation. As already discussed, best ramjet performance obtains when ingested
air flow rate is maximized (supercritical inlet).
Note that the ramjet engine determines the total pressure
at the nozzle entrance Pt4 from the flow rates, and it determines the pressure drops across
hardware items and flame zones, so that
it determines the actual operating total pressure Pt2 at the inlet
diffuser outlet! The inlet determines
only the maximum-possible diffuser total pressure Pt2CR!
Only at critical are these equal! Supercritically, Pt2CR > Pt2, the difference measured by the supercritical pressure margin PM. If the engine-demanded Pt2 is too high, the inlet will instead operate subcritically, and spill just enough air to make Pt2 = Pt2CR. The amount of this air spillage is measured by the subcritical spillage margin SM.
Figure 3 – Exactly How the Engine-Inlet Balance Works with
Ramjet
So, for a supersonic inlet, there are basically 3 modes related to shock-on-lip speed, and 3 modes related to inlet criticality. That’s a total of 9 possible modes and flow patterns, as depicted in Figure 4. The critical-inlet condition right at shock-on-lip speed is the design point for the inlet (but not necessarily for the engine-inlet system in the ramjet). All supersonic inlets, regardless of the application, share these 9 behaviors. Only some of them are appropriate for ramjet, and only some of them for turbojet. In the upper left illustration, both the streamtube geometry is reduced, and there is subcritical spillage.
Figure 4 – The Nine Behaviors Common to All Supersonic
Inlets
The differences between all-external compression inlets and
mixed-compression inlets are shown in Figure 5. The mixed-compression inlet has both external
compression by oblique shock waves, and
internal compression by both passage contraction and another oblique shock
wave. The all-external compression inlet
has no internal compression, only an
always-divergent internal passage! It
only has the external oblique shock waves,
and its min-area throat is right at the cowl lip station. At inlet design Mach, both types of inlet have a terminal normal
shock very close to the min-area throat, wherever it is.
The all-external compression inlet has a higher cowl lip
drag because of its strongly-curved shape toward the spike or ramp. But it has no internal contraction, and so is easy to start (swallow the shock
system), because that geometry
entirely avoids the Kantrowitz-Donaldson conundrum (see Ref. 2) peculiar
to internal supersonic contractions! It helps if the
demanded backpressure at start is also low,
at values more like large supercritical margin.
Without any throat bleed means, the initially-subcritical mixed compression
inlet throat area simply cannot subsonically pass the full swept-out supersonic
massflow rate (at its design speed),
thus requiring very serious overspeed above design Mach, in order to start! Even so,
in order to start at overspeed, the
demanded backpressure must also be low,
corresponding to levels seen at high supercritical pressure margin.
This difficulty with this approach is that once the inlet is
started, if you slow down to, or below, design speed,
and also happen to experience inlet unstart for any reason
whatsoever, you cannot re-start the
inlet! There is no restart, unless you have some other means to
re-accelerate the vehicle back up to the feasible starting overspeed.
Throat bleeds can be through the lateral cowl cheekwalls, as cutback slots in the 2-D variants, or by slots cut in the axisymmetric cowl wall tube from ahead of, to just past, the throat. It is also possible to put a throat bleed slot in the centerbody spike or ramp, leading overboard directly in the 2-D variant, and leading outward through the centerbody support struts, for the axisymmetric variant.
Figure 5 – Characteristics for All-External Vs Mixed
Compression Inlets
The supersonic inlet has a swept-out capture area defined by
the cowl lip and the bottom or centerline of the ramp or spike external
compression surface. When viewed from
ahead, it is obvious to even-the-untrained-eye
that the ramp or spike features present an obstruction to the oncoming flow. If the oblique shock waves are formed
properly by the external compression surfaces,
this blockage is not a problem,
as those waves divert the flow parallel to the external compression
surface, and compress it to a smaller
size that actually will fit through the available passage area at the inlet
throat station! This is illustrated in
the right-hand image of Figure 6.
This is still true, even if there
is some air not ingestible due to the streamtube and shockwave geometry, while traveling at somewhat lower than the
supersonic shock-on-lip speed.
However, as speed
decreases further, there comes a point
where the external compression surfaces cannot maintain the existence of
oblique shock waves that are attached to the surface. The shock system detaches, and becomes a curved bow wave, out in front of the external compression
features at some standoff distance, with
locally-normal shock characteristics.
This is shown in the middle illustration in Figure 6. A nominal speed at which this occurs is shown
in the figure, but the actual value
depends upon the angles off axial of the compression surfaces. The reader is referred to the ramp and cone
charts for oblique shock waves in NACA 1135 for further details.
This behavior persists at some bow shock standoff distance that varies, until flight speed is subsonic, and the bow shock disappears entirely. Flow ingestion is still very poor due to the presence of the obstruction that the subsonic approaching flow can “sense”. This is shown in the left illustration in Figure 6.
Figure 6 – Low-Speed Behaviors Seen Only in Turbojet
Installations
The only “fix” by means of which to increase air ingestion
is the spring-loaded blow-in door, shown
in the bottom illustration of Figure 6.
This only works if a suction can be pulled in the inlet duct, and only a gas turbine can do that. Ramjet cannot.
Various inlet integrations are possible. For modern high-speed combustor layouts, some options are shown in Figure 7. The old stovepipe-engine designs with
V-gutter flameholders, used in the
1950’s, were suitable only as engine
nacelles added to the basic vehicle, as
in the IM-99 Bomarc interceptor missiles.
In the stovepipe designs, there
is no place inside the engine structure,
in which to package anything,
even the fuel. And these were
limited by materials survivability issues,
to flight speeds under about Mach 3.
The nose inlet offers the lowest vehicle drag, with the air transfer duct either coaxial
along the centerline as illustrated, or
alternatively, as an annular passage
about a centerbody containing the equipment and fuel.
The chin inlet is sort-of like a circumferential segment of
a nose inlet, using part of the vehicle
nose as its external compression spike.
But, to achieve shock-on-lip at a
reasonable flight Mach number, the nose
half-angle may have to be steeper (more drag) than would otherwise be desired
for low drag but high-volume packaging.
Even so, it still offers a very
low-drag vehicle design, plus it is well-adaptable
to noncircular vehicle cross sections that have the flat bottoms by which to be
“wave riders” for higher lift. ASALM-PTV
was one of these, and reached Mach
6, although that was by accident. It was designed for cruise at Mach 4, limited more by materials survival than by thrust-over-drag.
The side entry inlets offer packaging in the front of the
vehicle without any geometric interference from the air transfer ducting. They do increase the vehicle drag, so that top speeds usually fall in the Mach 3
to 4 range, limited by thrust-over-drag. These can take a variety of forms, leading to a variety of different sudden-dump
flameholding geometries.
The single side inlet offers no added lift capacity, only just the inlet fairing drags, which are cowl lip, skin drag,
aft boat-tail pressure drag, and
aft face base drag. All of these are
traditionally charged against airframe drag,
no matter the thrust-drag accounting scheme. The inlet needs to go on the belly, so that at lifting angles of attack, the inlet flow is not blocked by the vehicle
body. Remember, ramjets cannot suck air like a pump, only gas turbines can!
The two-inlet configurations are either symmetrically placed
on the two sides, or asymmetrically
placed on those two sides, but at or below
the horizontal. Both of these geometries
offer added lift from the inlet fairings,
which can replace the presence of the forward wing-fins of
missiles. You really do not want
inlets anywhere around the dorsal side,
as they have a hard time ingesting air:
the missile body blocks the oncoming stream to one extent or another.
Three inlets are possible,
but that configuration has not been used. It always suffers from one or two inlets
suffering the “dorsal inlet” problem, to
one extent or another.
Four inlets might do better in the X-configuration, but in some cases having the fins in the
X-configuration might be more important to the vehicle designer. There are missiles out there with
X-configuration inlets, like the
“Sunburn” anti-ship missile. There was
another with the fins in X, and the
inlets in cross-configuration, as shown
in Figure 7. That was the SA-6
“Gainful” surface to air missile. Both
work.
Any of the illustrated geometries are able to serve as
flameholders for LFRJ engines, but not
the GG-fed RJ’s that do not use hypergolic magnesium fuels! The asymmetric two-inlet configuration is the
most successful for GG-fed RJ using fuels that do require flameholding.
There is a preferred location zone for the capture features of any of the side inlet configurations. This is about 4 to 6 diameters aft of the nose. Too far forward, and the flow about the body has not yet re-stabilized from the disturbances presented by that nose. That puts unrepeatability into the inlet capture characteristics. Too far aft, and you get into inlet and fin interference, plus increased additive drag along the greater wetted length.
Figure 7 – Various Inlet Integrations Are Possible
The SFRJ to date has shown a preference for nose or chin
inlet designs, although side inlets
could also feed the same air plenum just ahead of the fuel chamber. The real design restriction would be, not being forced to put the side inlet capture
features too far forward.
A very important aspect of the design of ramjets or gas
turbine-powered vehicles is the flight envelope. This is a region in altitude-speed space
in which the inlet/engine combination is expected to operate, and is not expected to operate outside
of it. This kind of thing is illustrated
in Figure 8. Bear in mind that
the minimum ramjet flight speed is an appropriately-high ramjet takeover
speed, while the minimum gas turbine
speed is literally zero on the surface up to some elevation, and below vehicle stall speed above it.
As already discussed above,
the minimum takeover speed of a ramjet is usually where you set the
design shock-on-lip Mach number for its inlet.
Somewhere along that low-speed boundary in the flight envelope, is your ramjet design point! At design,
and at the max mixture ratio expected,
you need enough supercritical pressure margin PM at engine design, to maintain a PM > 0 all along and near
that inner boundary. PM will increase as
you speed up across the envelope toward its outer boundary, and it will also increase as you lean back
mixture ratio in the ramjet engine. Ambient
air temperature also affects it.
Most gas turbine installations do not use actual supersonic
inlets, they use the simpler
pitot/normal shock inlets, because they
never fly more than about Mach 2.5.
Pitot/normal shock inlets are a separate topic, not discussed here. They operate down to very subsonic speeds.
But for a really high-speed gas turbine installation, not only are the same inlet components used
differently, but the design approach
itself is entirely different! Right up
to the maximum possible speed, you want
the oblique shocks to fall ahead of, or
at worst on, the cowl lip. (That is utterly different than what you
do designing for ramjet!)
That high-speed-side shock-on-lip gives you full freedom to
spill air with subcritical spillage margin all across the flight envelope, in response to changes in engine throttle
setting (and engine core rotation speed).
So, your inlet design
shock-on-lip Mach number is the highest speed you ever expect to fly with gas
turbine. To change that, requires a translating-spike variable inlet
geometry technology, something done with the SR-71.
With gas turbine installations, both the nozzle throat and nozzle exit areas
are variable. Their maxima get set by
the max thrust available at max throttle setting at low altitudes.
The selection of the inlet capture area AC is a
compromise between the large values needed at low altitude (and ground running), versus all the rest of the flight
envelope. This gets compromised further
by how many blow-in doors are included.
If you really have a supersonic inlet with a gas turbine installation, you may have to seriously consider including variable inlet geometry, in the form of relaxing the compression surfaces back to “flat” relative to the oncoming flow. That is how you get the removal of the obstruction effect, that you otherwise inherently have with supersonic inlet technology, at low or subsonic speeds, or even up to around Mach 1.5.
Figure 8 – Ramjet and Gas Turbine Use Different Design
Approaches
There are a lot of internal flow problems in all of
these different inlets, which can be
“fatal” to a given design. That
is also beyond scope here, being a
separate topic entirely, as are the
pitot/normal shock inlets themselves.
There is also a variation on the mixed-compression inlet
that starts easily without overspeed or bleed slots. It is called the “shock-follower inlet”, and it compromises its pressure
recovery with suboptimal internal shock strengths and locations to
accomplish this! The air ingestion was
not compromised, as this was for a
ramjet engine application on the old Russian surface-to-air missile known to
NATO as the SA-6 “Gainful”, which had 4
axisymmetric side-mounted mixed-compression side inlets.
The means to accomplish this was undersizing the
centerbody, in order to oversize the
inlet throat area enough to pass the swept-out airflow subsonically. This is often done in supersonic diffuser
work, per Ref. 2, although many diffusers are adjustable to a
smaller, more-efficient throat after the
shock system is swallowed. In the
SA-6, throat oversize was done, and left that way at fixed geometry, in order to swallow the shock system easily
in a very simple design. That pushed the
terminal shock at critical away from the throat, and closer to the full duct area, at the rear of the centerbody. Interference with the waves shed by the centerbody
struts does complicate this shock field,
but we will ignore that here.
Subcritically, all of this unswallows out in front of the cowl lip as with all the other supersonic inlets. The SA-6 being a ramjet, operating subcritically was avoided by design, even at the takeover points. Supercritically, there is a complicated train of oblique shocks trailing off the centerbody down into the constant-area ducts. The terminus of this was complicated, but effectively it is like a weak normal shock somewhere down in that duct. Having long ducts to accommodate large supercritical margins is essential. See Figure 9.
References:
#1. NACA Technical
Report 1135 “Equations, Tables, and Charts for Compressible Flow”, Ames Research Staff, 1953.
#2. Kantrowitz, A.,
and Donalson, C., “Preliminary Investigation of Supersonic Diffusers”, NACA TR L5D20, 1945.
No comments:
Post a Comment