1 Conservation Laws
1.1 Continuity
Unsteady: Z
X X
dV = (A~v )out (A~v )in (1)
CV t
Steady: X X
(A~v )in = (A~v )out (2)
1.2 Linear momentum
Unsteady: Z
X X X
F~ = ~v dV + v )out
(m~ (m~
v )in (3)
t CV
Steady: X X X
F~ = v )out
(m~ (m~
v )in (4)
1.3 Energy
1.3.1 Steady flow energy equation (SFEE)
1 1
h1 + ~v12 + ~g z1 = h2 + ~v22 + ~g z2 Q + Ws (5)
2 2
1.3.2 Bernoullis equation
1
P + ~v 2 + ~g z = constant (6)
2
2 Compressible Flows
2.1 Ideal gas law
P = RT (7)
2.2 Specific heat capacity
Q = mCv T if V = 0
(8)
Q = mCp T if P = 0
Cp
= (9)
Cv
R = Cp Cv = (10)
M
2.3 Internal energy
u = Cv T (11)
2.4 Enthalpy
h = Cp T = u + P V (12)
2.5 Entropy
S 0 (13)
T2 P2
S = S2 S1 = Cp ln Rln
T1 P1
(14)
T2 2
= Cv ln Rln
T1 1
2.6 Mach number
~v
M= (15)
a
s
p P
a= RT = (16)
2.7 Mach number relations
Assume inviscid, no work, no potential change, ideal gas. Table lookup.
2.7.1 Adiabatic and non-isentropic
r
T0 1 2 a0 T0
=1+ M1 = (17)
T1 2 a1 T1
2.7.2 Adiabatic and isentropic
1 1
1
P0 T0 0 T0
= = (18)
P1 T1 1 T1
2.8 Theoretical maximum velocity
p p
~vmax = 2h0 = 2CP T0 (19)
2.9 Compressible duct flows
2.9.1 Converging nozzle
d~v d~v
For M < 1, >0 For M > 1, <0 (20)
dx dx
2.9.2 Diverging nozzle
d~v d~v
For M < 1, <0 For M > 1, >0 (21)
dx dx
2.9.3 Converging-diverging nozzle
Choked flow when M = 1 at throat.
" +1
# 2(1)
1 2
A1 1 1+ 2 M1
= +1 (22)
A M1 2
1
1 r
2 2
m
max = 0 A RT0 (23)
+1 +1
2.10 Fanno flow
Compressible duct flow with friction. Assume steady, 1D, adiabatic, ideal gas, constant area duct, no work, no potential
change, wall shear stress correlated by Darcy friction factor. Table lookup.
fL
fL fL
= (24)
D D 1 D 2
Table 1: Flow property changes in Fanno flow. : Increase. : Decrease.
Flow property Subsonic Supersonic
P
~v
P0 , 0
T
M
S
2.11 Rayleigh flow
Inviscid compressible duct flow with heat transfer. Assume steady, 1D, inviscid, ideal gas, constant area duct, no work,
no potential change. Table lookup.
Salient points:
Heat transfer cannot be due to wall heat exchange because the mechanism of convection is friction (not inviscid).
Heat transfer changes the stagnation enthalpy of the flow.
Because heat transfer allows for S < 0, no restrictions on property changes.
For 1/ k < M < 1, heating causes a decrease in T . Due to conversion of heat into kinetic energy.
Table 2: Rayleigh flow effects on acceleration. : Accelerate. : Decelerate.
Heating Cooling
M <1
M >1
Table 3: Flow property changes in Rayleigh flow. : Increase. : Decrease.
Heating Cooling
Property M < 1 M >1 M <1 M >1
T0
M
P
~v
P0
S
T 1 2
1: for 0 < M < 1/k, for 1/k < M < 1
2: for 0 < M < 1/ k, for 1/ k < M < 1
2.12 Normal shock waves
Convention is 1-upstream, 2-downstream. Table lookup for properties normal to the shockwave.
Table 4: Property changes over a shockwave
Increase Decrease Constant
Property P , , T P0 , 0 , M T0
Moving shock:
~v1 = ~v + ~vshock (25)
2.13 Oblique shocks
Use the normal shock tables with the normal Mach numbers.
M1N = M1 sin()
(26)
M2N = M2 sin( )
2 cot M12 sin 1
tan = (27)
M12 ( + cos(2)) + 2
2.14 Weak shock approximations
Approximations for 5 only.
P2 P1 M 2
= p 21 tan (28)
P1 M1 1
S2 S1 ( 2 1)M16
= 3 tan (29)
Cp 12(M12 1) 2
k+1
sin = sin + tan (30)
4 cos
2.15 Prandtl-Meyer expansions
1
Assume = = arcsin M1 . For expansion fans and compression waves. Table lookup.
12 = (M2 ) (M1 ) (31)
3 Dimensional Analysis and Similitude
3.1 Buckingham Pi Theorem
Method of reducing number of dimensional variables in a function. Objective is to develop a dimensionless term. Work
in {M LT } or {F LT }.
1. List all variables in problem, n
2. List all dimensions in problem, k
3. Number of terms, j = n k
4. Select k repeating variables. Cannot form a term on their own.
5. Form each of the j terms, each for one dependent variable
(a) Raise each repeating variable to unknown exponents and take the product.
(b) Multiply by dependent variable.
(c) Express each variable in dimensional form. Set equal to M 0 L0 T 0 0 = .
(d) Form system of equations with unknown exponents as independent variables, and solve.
(e) Re-write term with the solved exponents.
3.2 Similitude
Geometric similarity: Model scaled equally in length (L) dimensions. m = p if = f (L).
Kinematic similarity: Model scaled equally in length (L) and time (T ) dimensions. m = p if = f (L, T ).
Dynamic similarity: Model scaled equally in length (L), time (T ), and force (F ) dimensions. m = p if = f (L, T, F ).
Table 5: terms, and when they should be used for scaling
term When to use
Re = ~v L/ Always
F r = ~v 2 /~g L Frictionless free-surface flow
W e = ~v 2 L/ Y Free-surface flow
St = f L/~v Unsteady oscillating flow
M = ~v /a Compressible flow
4 Viscous Flow Theory
4.1 Shear stress between fluid layers
du
= (32)
dy
4.2 Laminar and turbulent boundary layers (BL)
Laminar BL for Re < 5 105 . Turbulent BL for Re > 5 105 . Turbulent BL faster at wall, and thicker.
4.3 BL thickness parameters
Disturbance thickness:
= y(u = 0.99u ) (33)
Displacement thickness: Z
u
= 1 dy (34)
0 u
Momentum thickness: Z
u u
= 1 dy (35)
0 u u
4.4 Momentum integral equation
Shear stress at the wall:
2 d
0 = U (36)
dx
Skin friction coefficient:
0 d
CF = 1 2
=2 (37)
2 u dx
Skin friction drag: Z
1 2
F~DF = 0 dA = CDF u bL (38)
A 2
Skin friction drag coefficient:
~FDF
CDF = 1 2
(39)
2 u bL
4.5 Blasius solution
For laminar BL on flat plates. Exact solution.
5.0x 1.721x 0.664x 0.664
= = = Cf = (40)
Rex Rex Rex Rex
4.6 Von Karmans solution
For laminar BL on flat plates. Approximate solution. Assume parabolic BL velocity profile:
2y y 2
u
= 2 (41)
u
5.48x 1.83x 0.730x 0.730
= = = Cf = (42)
Rex Rex Rex Rex
4.7 Prandtls solution
For turbulent BL on flat plates. Assumes:
1)
14
0 = 0.0225u2 (43)
u
2)
u y 71
= (44)
u
0.162x 0.0203x 0.0158x 0.0271
= 7
=
7
=
7
Cf = 7
(45)
Rex Rex Rex Rex
Good luck! ,