ASE 435 Aerodynamics –

Supersonic and hypersonic flows

by

Balaji
Assistant Professor
Division of Aerospace Engineering
School of Mechanical and Aerospace Engineering
Scope
• Linearized theory
• Second-order theory
• Shock-expansion theory
• Comparisons
• Supersonic aerofoils
• Supersonic wings
• Swept wings
• Hypersonic flows
• Viscous effects at hypersonic speeds
• Heating effects at hypersonic speeds
• Newtonian theory
• Lift and drag characteristics
• Mach number independence
• Waveriders
Introduction
• Supersonic flows over an aerofoil can be analyzedusing
– Linearized theory
– Second-order theory
– Shock-expansion theory
• Linearized theory – quite similar to linearized theory learnt for subsonic
compressible flows earlier on (with some appropriatemodifications)
• Second-order theory – based on foundations of linearized theory but more
accurate, due to the inclusion of higher-order terms
• Shock-expansion theory – based on the principles ofoblique shocks and
expansion fans learnt earlier in compressible flowtheory
• It has been shown that all three provide reasonable estimations which are
close to oneanother
Linearized theory
• Similar to subsonic compressible flow, the perturbation velocity potential
function of a supersonic flow can be linearized foraerodynamic predictions
 2ˆ  2ˆ
• Recall for subsonicflow: 1 M x2  y 2  0
2


2  2
ˆ  ˆ
2
• For supersonic flow, let uswrite:  2  y2  0 where   M2 1
x

• Without going through the mathematics, it can be shown that the solution
to the above equationis
ˆ  f x  y 
̂ ̂
  uˆ f '   vˆ f '
x y

• Combining them will give: vˆ

uˆ 


• Recall from earlier: ˆ ˆ

 v  Vtan  V
y
Linearized theory
V
 uˆ 

2uˆ
• Recall from earlier again: cp  
V
2 cp,0
• Substituting: cp  Compare to: cp 
M 2  1 1  M 2
(Subsonic case)
• Supersonic flow  cp a function of surface inclination relative to the
incoming flow, 

• For an interesting comparison:

Linearized theory
• Consider points A, B, Cand D on
an aerofoil in supersonicflow:
2 A 2
cp,A  cp,B  B

M 2 1 M 2 1

Note: A and B arepositive

relative to incomingflow
2 2
cp,C   C cp,D   D

M 2 1 M  1
2

Note: Cand D are negative

relative to incomingflow

• To assess the lift/drag coefficient,

consider the situation of a flat-
plate (the simplest aerofoil
possible) at angle-of-attack of 
Second-order theory
• Linearized theory is based on first-order Taylor series expansion of pressure
field in terms of   but Taylor series expansion can go up to several orders
• To improve accuracy in predicted performance for aerofoils in supersonic
flows, second-order term of theTaylor series expansion is incorporated
• In this case, pressure coefficient can be rewritten as

   1M 4  4M2  4  2 Recall for

2 2
cp    linearized c p
M 2 1  2M  1  M  1
2

2 2

theory:
1st-order term 2nd-order term
 c  c   c 2
p 1 2
Shock-expansion theory
• Since the application of shock-expansion theory here is similar to that
considered in oblique shocks and expansion fans  will notgo through the
details again
Oblique shock
Expansionfan
Oblique shock
Supersonic 1
freestream 2
3 Slip line

6
4 5 Oblique shock
Oblique shock Expansionfan

• Treat each surface separately

– concave or convexturn
– oblique shock or expansionfan
– Determine Mach number, pressure, temperature etc
– Lift and drag forces calculated by using pressure, geometryand
resolving in the rightdirections
Comparisons
• Consider supersonic flow pasta
triangular wedge shown in the
figure:
– Good agreement between
theory and experimental
data for lift anddrag
– Slightly better agreement
between shock-expansion
theory and experiments
– But significant deviations in
quarter-chord moments
• Hence, theories are not perfect
and have limitations in some
aspects
Supersonic aerofoils
• Research has shown that good supersonic aerofoils should possess
– Sharp leading-edges and trailing-edges
– Relatively thin sections

Bi-convex aerofoil Double wedge aerofoil

• Reasons being
– Blunt leading-edges produce strong, detached leading-edge shocks
large wave drag
Oblique shock
Expansionfan
Oblique shock

Oblique shock
Oblique shock Expansionfan
Supersonic wings
• At supersonic flow regime, the overall drag experienced byan aircraft
comes from
– Skin-friction drag
– Induced drag
– Wave drag
• Recall from earlier: CD  CD,0  kCL2

Skin friction Induced

& wavedrag drag

• Hence: CD  CD,skinfriction  CD,wave drag  kCL2

Swept wings
Supersonic flow
Subsonicflow

• Swept wings allows aircrafts to fly in the supersonic flow regime easier than
rectangular wings
• In supersonic flow regime, the swept wing leading-edge can be treatedas a
concave turn
• Recall: supersonic flow over concave turn  oblique shock  subsonic flow
regime downstream of shock

• Swept wing leading-edge is downstream of oblique shock  experiences

subsonic flow rather than supersonicflow
Swept wings
• However, swept wings do not workas well
in actual practice
– viscous effects
– complex flow separationbehaviour
• Types of flow separations
– Leading-edge flow separation
– Flow separation due to spanwiseflow
– Inboard shockseparation
– Trailing-edge shockseparation
• Recall from previous slides that theoretical
considerations do not include effects
coming from flow separations  highly
complex  not possible to incorporate
everything analytically
Swept wings
• Leading-edge flow separation
– Similar to counter-rotating vortex systems formed above delta wings
– Induce low-pressure above swept wings aswell
• Flow separation due to spanwiseflow
– Spanwise flows thicken boundary layers nearwing-tips
– Leads to flow separations since thicker boundary layers areless
resistant to adverse pressuregradients
• Inboard shockseparation
– Occurs at wing/body interface alongleading-edge
– Sufficiently strong shocks may lead to flowseparations
• Trailing-edge flow separation
– Happens when shocks exist along trailing-edges
– Again, if strong enough, these shocks may induce flowseparations
there
Swept wings

• Compared to subsonic conditions, swept wings insupersonic conditions

produce
– Lower lift at thesame angle-of-attack
– Lower lift-to-drag ratio
Hypersonic flows
• If we exceed M = 5 flight speed, technically we are going into hypersonic
flight regime (i.e. because “hyper” >“super”!)

• Hypersonic flight regime

– Extreme temperatures
– Extreme shock wave systems
– Extreme chemical behaviour
* NASA
– Extreme material requirements
– Extremely difficult to analyze theoreticallyas well

• For engineers
– Difficult to understand due to interactions between heat, pressure,
chemical processes etc
– Most knowledge comes from experimentation  even then, scarce
– Difficult to compute flow fields (many coupled phenomena)
Hypersonic flows
• Despite its difficulties, hypersonic flight remains very attractivegoal
– Potentially cutting down air travel time significantly (no, not from your
hall/home to LPU)
– Ability to send in weapons faster than ICBMs for rapid response (at such
speeds, you don’t need explosives  kinetic energy will besufficient)
– Just because we can…

• We are still many years away from a realistic and useful hypersonicflight
vehicle
– Current material limitations (either too weak, heavy or expensive)
– Very expensive undertaking (with the global economy like this…)
– Safety considerations
– Immature SCRAMJETtechnology at such highspeeds
Hypersonic flows

*Imperial War Museum

*NASA
• The first man-made hypersonic vehicle is
probably the German “V-2”  up to
1500m/s flight speed
• It is also the forefather of the ICBM 
sub-orbital trajectories

• The first hypersonic research vehicle

would be the X-15. It was used to
explore problems and issues surrounding
high-speed, high altitude flights
Viscous effects at hypersonicspeeds

M =2 M =20
20 20

Shockwave 25
53.5

Shockwave

• For moderate supersonic flow (i.e. M = 2) over a concave turn,oblique

shocks are formed with moderate turningangle
• When the Mach number over the same turn increases, turning angle
decreases
Viscous effects at hypersonicspeeds

• So, for hypersonic flow over awedge:

– Oblique shock with very small turning angle,  , will be formed
– At high altitudes  low air density  lowReynolds number  thick
boundary layer
– High Mach number also leads to thicker boundary layer ( M2)

• Oblique shock interacts with viscous boundary layer  viscous interaction

phenomena
Viscous effects at hypersonicspeeds

• Inviscid scenario (clearly impossible)

– No boundary layer  flow not deflected
– Straight Mach wave extends from leading-edge to downstream 
constant pressure
• Viscous scenario
– Thick boundary layer  strong curved shock
– Much higher pressure near leading-edge  aerodynamic
heating
 Heating effects at hypersonic speeds
• Referring tothe figure, a blunt object traveling
at hypersonic speed will encounter 11,000 K
environment
–  is no longer 1.4 nor aconstant
Partially ionized
plasma
• Earlier equations cannot be used(!)
• Problems solved numerically
N  N+ + e- Causes communications
“blackout” duringre-entry – Gas becomes chemicallyreacting
O  O++ e-
• Oxygen molecules will disassociate at
2000 K< T< 4000K (O2  2O)
• Nitrogen molecules will disassociateat
4000 K< T< 9000K (N2  2N)
• Ionization will occur when T> 9000K
– Large heat transfer to objectsurface
Newtonian theory
• Unlike incompressible flow, where cp =
1 at stagnation point  cp  2 at
stagnation point when M  
• Realistically speaking, cp cannot reach 2
since M has to be finite  has to use
actual stagnation pressure, cp,max,
rather than 2

• Nonetheless, Newtonian theory

remains one of themost robust way to
estimate hypersonic flow performance
Mach number independence
• cp behind the hypersonic shock is pretty
much a constant at very high M  Mach
number independence
• Comparison shows that:
– Accuracy of Newtonian theory improves
as M increases
More accurate – Newtonian theory more accurate for3D
than 2Dbodies

• Another comparison showsthat:

– Sphere reaches Mach number
independence faster than cone-cylinder
– Regardless of geometry, there boundto
be a point where independence is
achieved
Waveriders
• Principles desired for a high L/Dratio
hypersonic vehicle
– All wing, delta shapedesign
– Shock wave attached to leading-edge
at designspeed
– Design “captures” or “traps” theshock
wave
– Vehicle “riding” upon the shockwave
 waverider

• Problems
– Lower than expected L/Dratio
– High skin friction drag
Waveriders
• For thin delta wings, as Mach number
increases
– Expansion fan along leading-edge  low
suction pressure along uppersurface
– Shock becomes increasingly attached to
lower surface
– No “leakage’ of air from lower to upper
surface
• Compare to a generic vehicle,where
upstream shock is detached
– Non-zero leakage ofair from lower to
upper surface
– Needs to fly at larger angle-of-attack for
similar lift as awaverider
Waveriders
• Waveriders have significantly higher lift than
generic hypersonic vehicles
• But L/D ratio is only comparatively higher  due
tohigh skin friction drag of waveriders
• Experiments carried out on waveridercandidates
(see figure) show
– Waverider type exhibits better lift than flat-
top type
– Waverider “inverted” type has lower lift
– Indicates the need to capture the lower
surface shock reliably for goodhypersonic
performance
Waveriders

• Base drag takes up 25-30% of total drag, skin friction drag another 25%
• Note that waveriders may not actually be the best hypersonic vehicle
– Stability and control
– Airframe-engine integration
– Heating effects
– Volumetric efficiency