Compressible Flow - TME085

Lecture 5

Niklas Andersson

Chalmers University of Technology

Department of Mechanics and Maritime Sciences
Division of Fluid Mechanics
Gothenburg, Sweden

niklas.andersson@chalmers.se
Chapter 4
Oblique Shocks and Expansion
Waves

Niklas Andersson - Chalmers 2 / 50

Overview
expansion
fans no
shock
se
reflection
shock
expansion
theory
governing conser-
oblique equations vation
shocks form
2D Flow
friction nozzles

heat
addition Quasi
1D Flow
diffusers

1D Flow Conservation
normal
shocks laws
integral form

isentropic
flow
Learning Outcomes

4 Present at least two different formulations of the governing equations for

compressible flows and explain what basic conservation principles they are
based on
7 Explain why entropy is important for flow discontinuities
8 Derive (marked) and apply (all) of the presented mathematical formulae for
classical gas dynamics
b normal shocks*
e oblique shocks in 2D*
f shock reflection at solid walls*
g contact discontinuities
h Prandtl-Meyer expansion fans in 2D
i detached blunt body shocks, nozzle flows
9 Solve engineering problems involving the above-mentioned phenomena (8a-8k)

why do we get normal shocks in some cases and oblique shocks in other?
Niklas Andersson - Chalmers 4 / 50
Roadmap - Oblique Shocks and Expansion Waves
Shock-expansion theory

Oblique shocks Expansion waves

Shock intersection Prandtl-Meyer expansion

Detached shocks

Shock systems Pressure-deflection diagram

Solid boundary reflection Mach reflection

Oblique shock relations The θ − β − M relation

Niklas Andersson - Chalmers 5 / 50
Motivation

Come on, two-dimensional flow, really?! Why not three-dimensional?

the normal shocks studied in chapter 3 are a special casees of the more general
oblique shock waves that may be studied in two dimensions

in two dimensions, we can still analyze shock waves using a pen-and-paper

approach

many practical problems or subsets of problems may be analyzed in

two-dimensions

by going from one to two dimensions we will be able to introduce physical

processes important for compressible flows

Niklas Andersson - Chalmers 6 / 50

Oblique Shocks and Expansion Waves

Supersonic two-dimensional steady-state inviscid flow

(no wall friction)

In real flow, viscosity is non-zero ⇒ boundary layers

For high-Reynolds-number flows, boundary layers are thin ⇒ inviscid theory still
relevant!

Niklas Andersson - Chalmers 7 / 50

Mach Waves
A Mach wave is an infinitely weak oblique shock

subsonic sonic supersonic

V <a V =a V >a

at
at at

Vt Vt Vt
B A
B A B A

at a 1
sin µ = = =
Vt V M
Niklas Andersson - Chalmers 8 / 50
Mach Wave
A Mach wave is an infinitely weak oblique shock

Mach wave

M1 M2
µ

No substantial changes of flow properties over a single Mach wave

M1 > 1.0 and M1 ≈ M2
Isentropic

Niklas Andersson - Chalmers 9 / 50

Oblique Shocks

M>1

M>1

compression corner gradual compression

Niklas Andersson - Chalmers 10 / 50

Roadmap - Oblique Shocks and Expansion Waves
Shock-expansion theory

Oblique shocks Expansion waves

Shock intersection Prandtl-Meyer expansion

Detached shocks

Shock systems Pressure-deflection diagram

Solid boundary reflection Mach reflection

Oblique shock relations The θ − β − M relation

Niklas Andersson - Chalmers 11 / 50
Chapter 4.3
Oblique Shock Relations

Niklas Andersson - Chalmers 12 / 50

Oblique Shocks

Two-dimensional steady-state flow

y Stationary shock

Flow condition
1

M>1

Flow condition
2

β >µ

Niklas Andersson - Chalmers 13 / 50

Oblique Shocks
w1 w2

v2
β−θ

θ
v1
u2
Stationary oblique shock
u1

Niklas Andersson - Chalmers 14 / 50

Oblique Shock Relations
w1 w2
A
v2

A
v1
Ω u2

u1

Two-dimensional steady-state flow

Control volume aligned with flow stream lines

Niklas Andersson - Chalmers 15 / 50

Oblique Shock Relations
w1 w2
A
v2

A
v1
Ω u2

u1

x
Velocity notations:
u1 v1
M n1 = = M1 sin(β) M1 =
a1 a1
u2 v2
M n2 = = M2 sin(β − θ) M2 =
a2 a2
Niklas Andersson - Chalmers 16 / 50
Oblique Shock Relations

Conservation of mass:

d y {
ρdV + ρv · ndS = 0
dt
Ω ∂Ω

Mass conservation for control volume Ω:

0 − ρ1 u1 A + ρ2 u2 A = 0 ⇒

ρ1 u1 = ρ2 u2

Niklas Andersson - Chalmers 17 / 50

Oblique Shock Relations

Conservation of momentum:

d y { y
ρvdV + [ρ(v · n)v + pn] dS = ρfdV
dt
Ω ∂Ω Ω

Momentum in shock-normal direction:

0 − (ρ1 u21 + p1 )A + (ρ2 u22 + p2 )A = 0 ⇒

ρ1 u21 + p1 = ρ2 u22 + p2

Niklas Andersson - Chalmers 18 / 50

Oblique Shock Relations

Momentum in shock-tangential direction:

0 − ρ1 u1 w1 A + ρ2 u2 w2 A = 0 ⇒

w1 = w2

Niklas Andersson - Chalmers 19 / 50

Oblique Shock Relations

Conservation of energy:

d y { y
ρeo dV + [ρho v · n] dS = ρf · vdV
dt
Ω ∂Ω Ω

Energy equation applied to the control volume Ω:

1 1
0 − ρ1 u1 [h1 + (u21 + w12 )]A + ρ2 u2 [h2 + (u22 + w22 )]A = 0 ⇒
2 2

1 1
h1 + u21 = h2 + u22
2 2

Niklas Andersson - Chalmers 20 / 50

Oblique Shock Relations

We can use the same equations as for normal shocks if we replace M1 with Mn1 and
M2 with Mn2

Mn21 + [2/(γ − 1)]

Mn22 =
[2γ/(γ − 1)]Mn21 − 1

Ratios such as ρ2 /ρ1 , p2 /p1 , and T2 /T1 can be calculated using the relations for
normal shocks with M1 replaced by Mn1

Niklas Andersson - Chalmers 21 / 50

Oblique Shock Relations

What about ratios involving stagnation flow properties, can we use the ones
previously derived for normal shocks?

Niklas Andersson - Chalmers 22 / 50

Oblique Shock Relations

What about ratios involving stagnation flow properties, can we use the ones
previously derived for normal shocks?

The answer is no, but why?

Niklas Andersson - Chalmers 22 / 50

Oblique Shock Relations

What about ratios involving stagnation flow properties, can we use the ones
previously derived for normal shocks?

The answer is no, but why?

Po1 , To1 , etc are calculated using M1 not Mn1 (the tangential velocity is included)

Niklas Andersson - Chalmers 22 / 50

Oblique Shock Relations

What about ratios involving stagnation flow properties, can we use the ones
previously derived for normal shocks?

The answer is no, but why?

Po1 , To1 , etc are calculated using M1 not Mn1 (the tangential velocity is included)

OBS! Do not not use ratios involving total quantities, e.g. po2 /po1 , To2 /To1 ,
obtained from formulas or tables for normal shock

Niklas Andersson - Chalmers 22 / 50

Deflection Angle (for the interested)
w1 w2

v2
β−θ

α1
θ
v1 α2
u2

u1

x
   
w w
θ = α2 − α1 = tan −1
− tan −1
u2 u1

∂θ u2 u1
= 2 −
∂w w + u22 w2 + u21
Niklas Andersson - Chalmers 23 / 50
Deflection Angle (for the interested)

∂θ u2 u1
= 2 − 2 =0⇒
∂w w + u2 w + u21
2

u2 (w2 + u21 ) − u1 (w2 + u22 ) (u2 − u1 )(w2 − u1 u2 )

2 2 = 0 ⇒ =0
(w2 + u2 )(w2 + u1 ) (w2 + u22 )(w2 + u21 )

Two solutions:
I u2 = u1 (no deflection)
I w2 = u1 u2 (max deflection)

Niklas Andersson - Chalmers 24 / 50

Shock Polar
Graphical representation of all possible deflection angles for a specific Mach number
1

4.0
0.5 3.0
2.5
2.0

1.5
Vy
No deflection cases: 0
a∗
I normal shock
(reduced shock-normal velocity)
−0.5
I Mach wave
(unchanged shock-normal velocity)
−1
0 0.5 1 1.5 2
Vx
a∗

Niklas Andersson - Chalmers 25 / 50

Shock Polar
Graphical representation of all possible deflection angles for a specific Mach number
1

0.5

Vy
0
a∗
Solutions to the left of the sonic line
are subsonic −0.5
∗
M = 1.0

−1
0 0.5 1 1.5 2
Vx
a∗

Niklas Andersson - Chalmers 26 / 50

Shock Polar
Graphical representation of all possible deflection angles for a specific Mach number
1

0.5
θmax

Vy
0
a∗
It is not possible to deflect the flow
more than θmax −0.5
∗
M = 1.0

−1
0 0.5 1 1.5 2
Vx
a∗

Niklas Andersson - Chalmers 27 / 50

Shock Polar
Graphical representation of all possible deflection angles for a specific Mach number
1

0.5

For each deflection angle θ < θmax , Vy

θ

0
there are two solutions a∗
I strong shock solution
I weak shock solution −0.5
∗
M = 1.0
Weak shocks give lower losses and
therefore the preferred solution
−1
0 0.5 1 1.5 2
Vx
a∗

Niklas Andersson - Chalmers 28 / 50

Shock Polar
Graphical representation of all possible deflection angles for a specific Mach number
1

0.5 β

θ
Vy
0
The shock polar can be used to a∗

calculate the shock angle β for a

given deflection angle θ −0.5
∗
M = 1.0

−1
0 0.5 1 1.5 2
Vx
a∗

Niklas Andersson - Chalmers 29 / 50

Flow Deflection

M>1
weak shock family
sonic line

strong shock family

θ > θmax
M<1
M2 θ < θmax

M1 > 1 M1 > 1

M2
M<1

strong shock family

sonic line
weak shock family

M>1

Niklas Andersson - Chalmers 30 / 50

Roadmap - Oblique Shocks and Expansion Waves
Shock-expansion theory

Oblique shocks Expansion waves

Shock intersection Prandtl-Meyer expansion

Detached shocks

Shock systems Pressure-deflection diagram

Solid boundary reflection Mach reflection

Oblique shock relations The θ − β − M relation

Niklas Andersson - Chalmers 31 / 50
The θ-β-M Relation

It can be shown that

M12 sin2 β − 1
 
tan θ = 2 cot β
M12 (γ + cos 2β) + 2

which is the θ-β-M relation

Does this give a complete specification of flow state 2 as function of flow state 1?

Niklas Andersson - Chalmers 32 / 50

The θ-β-Mach Relation

90
80
A relation between: 70 M2 < 1.0
θmax
I flow deflection angle θ
60
I shock angle β
I upstream flow Mach number M1 50 M2 > 1.0
β
40
30
M12 sin2 (β) − 1
 
tan(θ) = 2 cot(β) 20
M1 (γ + cos(2β)) + 2
2
10
0
Note! in general there are two solutions 0 10 20 30 40 50
for a given M1 (or none) θ

Niklas Andersson - Chalmers 33 / 50

The θ-β-Mach Relation

90
I There is a small region where we may find
weak shock solutions for which M2 < 1 80
70 M2 < 1.0
θmax
I In most cases weak shock solutions have
60
M2 > 1
50 M2 > 1.0
β
I Strong shock solutions always have M2 < 1 40
I In practical situations, weak shock solutions 30
are most common 20
10
I Strong shock solution may appear in special
situations due to high back pressure, which 0
0 10 20 30 40 50
forces M2 < 1 θ

Niklas Andersson - Chalmers 34 / 50

The θ-β-M Relation M12
sin2 β − 1
!
tan θ = 2 cot β
M1 (γ + cos 2β) + 2
2

Example: Wedge flow

β
M1 > 1 θ

Weak solution:
smaller β, M2 > 1 (except in some cases)
Strong solution:
larger β, M2 < 1

Niklas Andersson - Chalmers 35 / 50

The θ-β-M Relation

Note! In Chapter 3 we learned that the mach number always reduces to subsonic
values behind a shock. This is true for normal shocks (shocks that are normal to the
flow direction). It is also true for oblique shocks if looking in the shock-normal
direction.

Niklas Andersson - Chalmers 36 / 50

The θ-β-M Relation M12
sin2 β − 1
!
tan θ = 2 cot β
M1 (γ + cos 2β) + 2
2

No solution case: Detached curved shock

M1 > 1 θ

Niklas Andersson - Chalmers 37 / 50

The θ-β-M Relation - Wedge Flow
Wedge flow oblique shock analysis:

1. θ-β-M relation ⇒ β for given M1 and θ

2. β gives Mn1 according to: Mn1 = M1 sin(β)

3. normal shock formula with Mn1 instead of M1 ⇒

Mn2 (instead of M2 )

4. M2 given by M2 = Mn2 / sin(β − θ)

5. normal shock formula with Mn1 instead of M1 ⇒

ρ2 /ρ1 , p2 /p1 , etc

6. upstream conditions + ρ2 /ρ1 , p2 /p1 , etc ⇒

downstream conditions
Niklas Andersson - Chalmers 38 / 50
Chapter 4.4
Supersonic Flow over Wedges and
Cones

Niklas Andersson - Chalmers 39 / 50

Supersonic Flow over Wedges and Cones

What about cone flows?

M>1

I Similar to wedge flow, we do get a constant-strength shock wave, attached at

the cone tip (or else a detached curved shock)

I The attached shock is also cone-shaped

Niklas Andersson - Chalmers 40 / 50

Supersonic Flow over Wedges and Cones

What about cone flows?

M>1

I The flow condition immediately downstream of the shock is uniform

I However, downstream of the shock the streamlines are curved and the flow
varies in a more complex manner (3D relieving effect - as R increases there is
more and more space around cone for the flow)

I β for cone shock is always smaller than that for wedge shock, if M1 is the same
Niklas Andersson - Chalmers 41 / 50
Roadmap - Oblique Shocks and Expansion Waves
Shock-expansion theory

Oblique shocks Expansion waves

Shock intersection Prandtl-Meyer expansion

Detached shocks

Shock systems Pressure-deflection diagram

Solid boundary reflection Mach reflection


Oblique shock relations
Niklas Andersson - Chalmers

The θ − β − M relation
42 / 50
Chapter 4.6
Regular Reflection from a Solid
Boundary

Niklas Andersson - Chalmers 43 / 50

Shock Reflection

Regular reflection of oblique shock at solid wall

(see example 4.10)

θ
θ β2
M3 > 1
M1 > 1 M2 > 1

y β1
θ

Assumptions:
I steady-state inviscid flow
I weak shocks
Niklas Andersson - Chalmers 44 / 50
Shock Reflection

first shock:
I upstream condition:
M1 > 1, flow in x-direction
I downstream condition:
weak shock ⇒ M2 > 1
deflection angle θ
shock angle β1

second shock:
I upstream condition:
same as downstream condition of first shock
I downstream condition:
weak shock ⇒ M3 > 1
deflection angle θ
shock angle β2

Niklas Andersson - Chalmers 45 / 50

Shock Reflection

Solution:
first shock:
I β1 calculated from θ-β-M relation for specified θ and M1 (weak solution)
I flow condition 2 according to formulas for normal shocks (Mn1 = M1 sin(β1 ) and
Mn2 = M2 sin(β1 − θ))

second shock:
I β2 calculated from θ-β-M relation for specified θ and M2 (weak solution)
I flow condition 3 according to formulas for normal shocks (Mn2 = M2 sin(β2 ) and
Mn3 = M3 sin(β2 − θ))

⇒ complete description of flow and shock waves

(angle between upper wall and second shock: Φ = β2 − θ)

Niklas Andersson - Chalmers 46 / 50

Roadmap - Oblique Shocks and Expansion Waves
Shock-expansion theory

Oblique shocks Expansion waves

Shock intersection Prandtl-Meyer expansion

Detached shocks

Shock systems Pressure-deflection diagram

Solid boundary reflection Mach reflection


Oblique shock relations
Niklas Andersson - Chalmers

The θ − β − M relation
47 / 50
Chapter 4.11
Mach Reflection

Niklas Andersson - Chalmers 48 / 50

Regular Shock Reflection

Regular reflection possible if both primary and reflected shocks are weak (see θ-β-M
relation)

θ
θ β2
M3 > 1
M1 > 1 M2 > 1

y β1
θ

Niklas Andersson - Chalmers 49 / 50

Mach Reflection

normal shock

slip line

reflected oblique shock

incident oblique shock

y
θ

Mach reflection:
I appears when regular reflection is not possible
I more complex flow than for a regular reflection
I no analytic solution - numerical solution necessary
Niklas Andersson - Chalmers 50 / 50