Handout Jirka PDF

Environmental Fluid Mechanics Ι 1 Institute for Hydromechanics
MASS TRANSPORT PROCESSES
When a tracer cloud is introduced into a fluid flow (e.g. river) two processes are visible:
• The cloud is carried away from the point of discharge by the mean current.
• The cloud spreads in all directions due to irregular motions and grows in size.
Advection - mass transport by mean velocity field
Diffusion - mass transport by irregular fluctuations ("random movements")
mean t1 t2
flow
PASSIVE MIXING PROCESSES - advection + diffusion as they exist

in environment (passive source)
ACTIVE MIXING PROCESSES - generation of mean and random

velocity field
by source - momentum or buoyancy
passive mixing Active mixing

(weak source) (strong source e.g. high velocity)
MOLECULAR DIFFUSION
− molecular diffusion of dissolved or suspended matter is caused by random movement

through molecular motions of individual molecules
∂c
<0
∂x
window of unit area
mass ⎡ M ⎤
J x = diffusive mass flux = =⎢ 2 ⎥
area , time ⎣ L , T ⎦
Fick's Law of Diffusion:

∂c
− rate of transport is proportional to the spatial concentration gradient
∂x
∂c
Jx = − D AB Fick´s law (1-D)
∂x
⎡ L2 ⎤
D AB = coefficient of molecular diffusivity = ⎢ ⎥ = f (solvent, solute, temperature...)
⎣T⎦
Typical values for molecular diffusion:
Dissolved matter in water: D ∼ 2 × 10-5 cm2/s

Gases in air: D ∼ 2 × 10-1 cm2/s
Generalization:
r ⎛ ∂ ∂ ∂⎞
(
J = J x , J y , Jz ) ∇=⎜ , , ⎟ divergence vector
⎝ ∂ x ∂ y ∂z ⎠
r
J = −D∇C Fick´s law (3-D)
Example: Salt diffusion in a vessel
Water Water two hours later

initial sharp diffuse interface
interface
~ 1,0 cm
Water + NaCl+ Dye Water + NaCl+ Dye

ADVECTIVE MASS FLUX
− transport of matter due to mean motion (current)
r
q = (u, v, w ) = hydrodynamic velocity
r ⎡ M L⎤ ⎡ M ⎤
c q = ⎢ 3 ⎥ = ⎢ 2 ⎥ = advective mass flux
⎣ L T ⎦ ⎣L , T ⎦
window of unit area
Example:
Pure advection in x-direction (no diffusion no reaction)
t1 t2
no change of shape
“plug flow”
TOTAL MASS FLUX: Advection + diffusion

r r r
( )
r
N = N x , N y , Nz = c q + J = c q − D∇C
MASS CONSERVATION PRINCIPLE
R = rate of production or decay per unit

⎡ M ⎤
volume = ⎢ 3 ⎥
⎣L , T ⎦
⎡Rate of production / ⎤
⎢decay of mass ⎥ ⎡Rate of mass ⎤
⎡net flux of mass⎤ ⎢ ⎥ ⎢accumulation ⎥
⎢ (In − Out ) ⎥ + = ⎢ ⎥
⎣ ⎦ ⎢(chemical, physical, ⎥
⎢ ⎥ ⎢⎣ within element ⎥⎦
⎣bio log ical ... ) ⎦
∂ Nx ∂ Ny ∂ Nz ∂c
− ∆ x∆ y ∆ z − ∆ x∆ y ∆ z − ∆ x∆ y ∆ z + R ∆ x∆ y ∆ z = ∆ x∆ y ∆ z
∂x ∂y ∂z ∂t
∂ c ∂ N x ∂ N y ∂ Nz ∂ c r ∂c r
+ + + = + ∇⋅N = + ∇ ⋅ c q − ∇ ⋅ (D ∇ c) = R
∂t ∂x ∂y ∂z ∂t ∂t
∂c r
+ q ⋅ ∇ c = D ∇2 c + R Convective Diffusion Equation
∂t
∂c ∂c ∂c ∂c ⎛ ∂2 c ∂2 c ∂2 c ⎞
+ u + v + w = D⎜ 2 + + ⎟ + R
∂t ∂x ∂y ∂z ⎝∂x ∂ y2 ∂ z2 ⎠
↑ 1444442444443 14444244443 ↑
Temporal Advective transport Diffusive transport Reaction
change
Solution of the convective diffusion equation:
1. Basic solution: Instantaneous point source (IPtS) in uniform current and infinite
space
z
r
q = (U, 0,0) , (uniform velocity field)
U y
R = − k c (first order decay)
t=0 t 1
t 2
x
∂c ∂c ⎛ ∂2 c ∂2 c ∂2 c ⎞
+U =D⎜ 2 + + ⎟ − kc
∂t ∂x ⎝∂x ∂ y2 ∂ z2 ⎠
Initial Conditions (I.C.) t < 0, c = 0 t=0 Sudden mass input

M on (x1, y1, z1)
Boundary Conditions (B.C.) c → 0 as (x, y, z) → ∞
Solution:
[( x − x ) − Ut] + ( y − y1) + ( z − z1)

2 2 2
1
−
M 4D t
c = e e− k t
(4 π D t) 3/ 2
123 14444424444443 123

cmax spatial distribution non-conservative
(reaction) effect
1. Position of cloud center:

( x1 + U t, y1, z1)
t1
2. Cloud size:
4 D t = 2 σ2 , σ = variance t2
σ = 2D t = standard deviation ∼ t 0,5

cmax
Distance relative to center

3. Maximum concentration
cmax ∼ t-3/2
2. Other solutions
Boundary conditions (finite space):
on ⎡: c = specified
Ex: c = 0 Absorption condition

(Dirichlet B.C.)
∂c
= specified
∂n
∂c
Ex: =0 Diffusion barrier (Reflection condition)
∂n
(Neumann B.C.)
Multiple sources → superposition, due to linearity of governing equation

→ image sources
z
k
c = ∑ ci i=1
i =1
x
i=2
Time dependence
Initial condition:
− Instantaneous sources
− Continuous sources
TURBULENCE - an instability of a base flow (shear flow)
Possibility 1:
- damping of disturbance
(viscosity)
damping - stable flow
LAMINAR FLOW
Possibility 2:
- insufficient damping
- large eddies
- secondary eddies
disturbance TURBULENT FLOW

grows
SPECTRUM OF EDDIES
Large eddies: Integral length scale lΙ Integral velocity scale uΙ

lΙ
=> Integral time scale tΙ =
uΙ
⎡ L2 ⎤
K.E. uΙ u
2 3
Extraction of energy from mean flow: ε~ =⎢ 3 ⎥
~ = Ι
time l Ι / uΙ l Ι
⎣T ⎦
Energy cascade (transfer of energy from larger to smaller scales)
Smallest eddies: at some scale (Kolmogorov length scale) the eddy size will be small
enough so that viscous damping will prevent further breakdown
⎡ L2 ⎤ ⎡ L2 ⎤
Dimensionality: viscosity ν = ⎢ ⎥ , energy dissipation rate ε = ⎢ 3 ⎥
⎣T⎦ ⎣T ⎦
ν3 / 4
Kolmogorov length scale η = 1/ 4
ε
ε
Turbulence spectrum: lΙ η
" equilibrium subrange"
UL
Overall criterion: Reynolds number Re = > 100 to 1000
ν
U ∼ base velocity, L ∼ base flow length
TURBULENT DIFFUSION
Random motion due to turbulent velocity fluctuations cause additional transport ("small
scale advection").
u ′ ( t) = fluctuating velocity
1 t −T 1 t −T
u (t) = ∫ u ( t ′) d t ′ = mean velocity u′ = ∫ u′ ( t ′ ) d t ′ = 0
T t T t
u = u + u′ w = w + w′ Averaging process:
Reynolds decomposition
v = v + v′ c = c + c′ uc = u c + uc ′ + u′ c + u′ c ′
1 t + T ⎡ ∂c ∂uc ∂ vc ∂ wc ⎛ ∂2 c ∂2 c ∂2 c ⎞ ⎤
∫ ⎢ + + + = D ⎜ 2 + 2 + 2 ⎟ + R⎥ d t
T t ⎢⎣ ∂ t ∂x ∂y ∂z ⎝ ∂x ∂y ∂z ⎠ ⎥⎦
∂c ∂c ∂c ∂c ∂ ∂ ∂ ⎛ ∂2 c ∂2 c ∂2 c ⎞
+u +v +w =− u′ c′ − v ′c ′ − w ′ c′ + D ⎜ 2 + 2 + 2 ⎟ + R ↓
∂t ∂x ∂y ∂z ∂x ∂y ∂z ⎝ ∂x ∂y ∂c ⎠
144424443 14444244443 14442443 ↓
time advective transport "small scale advection" molecular reac-
change by ⇓ diffusion tion
of mean mean velocity turbulent mass transport ≈0
conc.
r ⎡ M ⎤
( )
Jt = u′c ′, v ′c ′, w ′c ′ = turbulent diffusive mass flux = ⎢ 2 ⎥
⎣L , T ⎦
r
Jt needs parameterization : analogy to molecular diffusion (Fick's law) ?
Example: Tank with Grid Stirring
net transport w ′c′ > 0
DIFFUSION IN TURBULENT FLOW
→ fluctuating velocity field ⇒ multiple scale eddies (spectrum)!
Ex: Release from POINT SOURCE
0,5
σ ∼ t
σ ~t
a) Short time after release t<t Ι , σ < l Ι

First small eddies, then eddies of increasing size will cause spreading
σ x = u′ 2 t ~ t1 Rapid growth (Taylor theorem, 1921)
u′2 =rmsvelocity~uΙ
b) Long time after release t>t Ι : eddies are independent of each other
Large eddies control spreading => random process
( )
σ x = 2 uΙ t Ι t ~t1 / 2
2
"Fickian" behavior: corresponds to spreading with
gradient-type diffusion and a constant coefficient
⎡ L2 ⎤
∴ uΙ t Ι =uΙ l Ι =E x = turbulentdiffusivity = ⎢ ⎥
2
Analogy:
⎣T⎦
∂c ∂c ∂c
J tx = u′ c ′ = − E x , Jty = v ′ c ′ = − E y , J tz = w ′ c ′ = − E z
∂x ∂y ∂z
PRACTICAL TURBULENT DIFFUSION EQ.: c → c
∂c ∂c ∂c ∂c ∂ ⎛ ∂ c⎞ ∂ ⎛ ∂ c⎞ ∂ ⎛ ∂ c⎞
+u +v +w = ⎜ Ex ⎟+ ⎜ Ey ⎟+ ⎜E ⎟ + R + D ∇2 c
∂t ∂x ∂y ∂ z ∂ x ⎝ ∂ x⎠ ∂ y ⎝ ∂ y ⎠ ∂ z ⎝ z ∂ z⎠
14444444244444443
turbulent diffusive transport
Valid for t > tΙ ; σ > lΙ
Diffusivities can have dependence on spatial position and direction:
- E x , E y , E z = f ( x , y , z) non-homogeneous
∂2c ∂2c ∂2c
if E x , E y , E z = const. homogeneous → Ex + E + E
∂ x2 y
∂y 2 x
∂z2
- Ex ≠ Ey ≠ Ez anisotropic
⎛ ∂2c ∂2c ∂2c ⎞
if Ex = Ey = Ez = E isotropic → E⎜ 2 + + ⎟
⎝ ∂x ∂y 2 ∂z2 ⎠
Estimation of turbulent diffusivities:
E = f (base flow characteristics)
Order of magnitude:
E ~uΙ l Ι single scale only!
TURBULENT CHANNEL FLOW

Uh
Re = > 500 for turbulence
ν
τ 0 = ρ gh S = shear stress
τ0
= u∗ = friction (shear ) velocity
ρ
u* = g hS
Dominating scales:
uΙ ~u∗,l Ι ~h
f f
Also: τ 0 =ρ U 2 f = Darcy-Weisbach coefficient, 0.02 .. 0.1 u∗ = U²
8 8
1 gn 2
Or: U = h 2 / 3 S1/ 2 n = Mannings's n, 0.02 ... 0.05 u∗ = U²
n h 1/ 3
Rule of thumb: u ∗ ≈ ( 0.05 to 0.1) U

MIXING IN RIVERS
Eddy diffusivity Ez
We expect: E z ~ u ∗ h
Velocity profile
= f (u ∗ , z)
du
dz
Dimensional analysis:
du u ∗ du u ∗
~ or =
dz z dz κ z
Upon integration:
u 1
= lnz + C 1 log-law κ = von Karman constant = 0.4
u∗ κ
C1 = f (Re, roughness)
⎛ z⎞ du
τ = τ 0 ⎜ 1− ⎟ = ρ ε z model for momentum exchange
⎝ h⎠ dz
(Boussinesq)
⎛ z⎞
τ0 ⎜ 1− ⎟
⎝ h⎠ ⎛ z⎞ ⎡ L2 ⎤
εz = = κ u∗ ⎜ 1 − ⎟ εz = eddy viscosity = ⎢ ⎥
du ⎝ h⎠ ⎣T⎦
ρ
dz
εz z⎛ z ⎞ Ez
= κ ⎜ 1− ⎟ ≡ Reynolds analogy for turbulent flows
u∗ h h ⎝ h ⎠ u∗ h
Mass exchange ≈ momentum exchange
Depth averaged:
Ez κ
= = 0.067 ≈ 0.1
u∗ h 6
Rules of thumb:
River U, h u ∗ ≈ 0.1U
E 2 ≈ 0.1u ∗ h ≈ 0.01Uh
Ex:
U = 2 m/s, h = 2m
Ez = 0.04 m²/s = 400 cm2/s >> D = 2 × 10-5 cm/s !!
VERTICAL MIXING
e.g. source at surface
x
σ z = 2E z t = 2E z
U
Complete vertical mixing: 2,15 σz = h 10% criterion with Gaussian profile
2
h xm ⎛ h ⎞ U
= 2 Ez ⇒ xm = ⎜ ⎟
2.15 U ⎝ 2.15 ⎠ 2E z
2
⎛ 2 ⎞ 2
Ex: xm = ⎜ ⎟ = 25 m Very fast
⎝ 2.15 ⎠ 2 × 0.04
Typically: Distance to complete vertical mixing 10 to 20 water depth!
LATERAL MIXING Wide Channel, Depth h
Conditions:
x > xm
σy > h
Assume isotropy: E y ≅ Ez = 0.067 u ∗ h
Ey
Data: ≅ 0.15 to 0.20 straight, uniform channels
u∗ h
Evaluation of field / laboratory tests:
1 d σy 1 σ y ( t 2 ) − σ y ( t1 )
2 2 2
= Ey ≅
2 dt 2 t 2 − t1
1 σ y ( x 2 ) − σ y ( x1 )
2 2
or
2 ( x 2 − x1 ) / U
Outlines of tracer cloud
Data from Mississippi River study on

longitudinal mixing (after McQuivey and
Keefer 1976b)
NATURAL CHANNELS; IRREGULARITIES
Ey
Data: = 0.5 to 10
. =α α = 0.6 ± 50% (Fischer (1972)
u∗h
2
⎛ UB ⎞
α = 0.4 ⎜ ⎟ Yotsukura and Sayre (1976)
⎝ u∗ Rc ⎠
Cumulative Discharge Method Yotsukura and Cobb (1976)
dq = hudy = local discharge → from

measurement or numerical model
y
q ( y) = ∫ hudy = cumulative discharge

0
B
Q = ∫ hud y = total discharge
0
Approximate local balance:

∂c ∂ ⎛ ∂c⎞
u ( y) h ( y) = ⎜ E y ( y) h ( y) ⎟
∂x ∂y ⎝ ∂y⎠
↓ ↓
depth integrated depth integrated
advection lateral diffusion
∂c ∂ ⎛ ∂c ⎞ ∂c ∂ ⎛ 2 ∂c⎞
uh = uh ⎜ h 2 E y u ⎟ ⇒ = ⎜ h Ey u ⎟
∂x ∂q ⎝ ∂q ⎠ ∂x ∂q ⎝ ∂ q⎠
14243
⎡ L5 ⎤
"Diffusitivity" = ⎢ 2 ⎥ ≈ constant
⎣T ⎦
Understanding of advective velocity

field is key to many environmental
diffusion problems!
y/B q/Q
Transverse distributions of dye observed in the Missouri River near Blair, Nebraska, by
Yotsukura et al. /1970), plotted (a) versus actual distance across the stream and (b)
versus relative cumulative discharge [Yotsukura and Sayre (1976)].
LONGITUDINAL DISPERSION (SHEAR FLOW DISPERSION)

The stretching of a tracer cloud due to vertical and horizontal velocity shear is called
longitudinal dispersion. In this case, the velocity differences interact with the transverse
diffusion effect and produce an additional transport mechanism
Observations: t > t mixing or σ z > h
For long time:

1. Gaussian distribution
2. σ L − t
Therefore: behavior is analogous to

Fickian diffusion
Analysis:
∂c ∂c ∂ ⎛ ∂c ⎞ ∂ ⎛ ∂c ⎞
+ u ( z, t) = ⎜E x ⎟ + ⎜Ez ⎟ 2-D
∂t ∂x ∂x ⎝ ∂z⎠ ∂z ⎝ ∂z⎠
Spatial decomposition: u = U + u ′′ , c = C + c ′′ , u ′′ , c ′′ = spatial deviations
1h
Substitute and average
h ∫0
( ) dz U, C = spatial averages
h
∂C ∂C ∂u ′′ c ′′ ∂ ⎛ ∂C ⎞ 1 ⎛ ∂ c ⎞
+U + = ⎜E ⎟ + ⎜E ⎟
∂t ∂x ∂x ∂ x ⎝ x ∂ x ⎠ h ⎝ z ∂z ⎠
0
14243
h
1
h ∫0
u ′′ c ′′ = u ′′ c ′′ dz = 0 no flux
⎡ M ⎤
JL = u ′′ c ′′ = dispersive mass flux = ⎢ 2 ⎥
⎣L , T ⎦
⎡ M ⎤
JL = u ′′ c ′′ = dispersive mass flux = ⎢ 2 ⎥
⎣L , T ⎦
∂C ⎡L2 ⎤
Analogy: JL = − EL EL = coeff. of longitudinal dispersion = ⎢ ⎥
∂x ⎣T⎦
− neglect: long. diffusion, Ex << EL
∂C ∂C ∂2 C
+ U = EL 1-D
∂t ∂x ∂ x2
↓ ↓ ↓
change of advection spreading by
mean conc. mean velocity deviations
velocity + transverse diffusion
Local balance as seen by moving observer (U):
∂C ∂ ⎛ ∂ c ′′ ⎞
u ′′ = ⎜E z ⎟ Taylor (1953, 1954)
∂x ∂z ⎝ ∂z ⎠
↓ ↓
transport in x ⇔ diffusion in
direction by vertical direction
differential by turbulence
velocities
"Stretching" ⇔ "Homogenization"
∂C
With given properties: u ′′ ( z), E z ( z) one can compute c ′′ ~ ,
∂x
and evaluate u ′′ c ′′
This leads to:
h z z
1 1
EL = − ∫ u ′′ ∫
h 0 0 E z ∫o
u ′′ dz dz dz
Applications: 2-D Flows
1. Channel flows:
u ′′ ~ log− profile
EL = 5.9 u∗ h Elder (1959)
EZ ∼ parabolic
Compare: E y ≈ E x = 0.2 u∗ h
EL >> Ex
Plan view of a drop of dye diffusing in the turbulent flow in an open channel.
Distribution of concentration C, normalized to have a maximum of 10. The flow is
to the left; h = 1.43 cm, x/h = 90 ( Elder, 1959)
2. Turbulent pipe flow:
EL = 10.1 u∗ ro Taylor (1954)

ro = pipe radius
3. Laminar pipe flow:
r 2 o U2
EL mol
= Taylor (1953)
48 D
General rule for applicability: t > tM = transverse mixing time (vertical)
h2
= 0.4 Et = transverse diffusivity
Et
EL
Natural Rivers Data: = 100 to 1000 3-D non-uniformities
u∗ h
Lateral velocity
deviations dominate
→ Stretching
+
Lateral (transverse) diffusion
Typical cross section of a natural stream

(Fischer 1968)
Fischer (1965) u ( y), E y ( y) , A = cross-section
B y y
EL = − ∫ ( u − U) h ∫ ∫ ( u − U) h dydydy
1 1
u(y): measurements
A0 0
Ey h 0
↓ ↓ or numerical model
velocity deviation in lateral

direction y
- approximate formula (typical conditions):
U2 B 2
EL = 0.011 good within factor 4 (σL ∼ within 2)
u∗ h
Applicability: t > tm = transverse (lateral mixing time

(B / 2) 2
= 0.4
Ey
Other physical non-uniformities:
Overall effects:
- stretching of "cloud"
- increased longitudinal
dispersion EL ↑
Practical longitudinal dispersion equation
∂C ∂C ∂2 C
+U = EL − kC t > tM
∂t ∂x ∂x2
(often neglected; but careful!)
m ′′ ⎡ ( x − Ut) 2 ⎤
Instantaneous Area Source: C= exp − ⎢ ⎥ exp [ − kt]
4 π EL t ⎢⎣ 4 EL t ⎥⎦
q′′ ⎡ xU ⎛ 4 kEL ⎞⎤
Continuous Area source: C= exp ⎢ ⎜1 m 1 + ⎟⎥
4kEL ⎢⎣ 2 EL ⎝ U2 ⎠ ⎥⎦
u 1+ 2
u
4kEL
In continuous problems Plug flow: →0
U2
Æ long. Dispersion usually negligible
q′′ ⎛ kx ⎞
C= exp ⎜ − ⎟
U ⎝ U⎠
OCEANIC DIFFUSION (coastal zone, estuaries, large lakes)
- eddy structure
RELATIVE DIFFUSION"
diffusion relative to center of mass, independent of actual location in fixed space

(x, y)
- mass will follow larger eddies, but will be diffused by action of smaller eddies (of
increasing size)
→ accelerating growth rate
Er = C ε1/3 σr4/3 „Richardson´s (1921) 4/3 Law“

↑
cloud size
d σ 2r
= 4 Er cylindrical coordinate r, standard deviation σr
dt
⇒ σ 2r ~ ε t 3 rapid growth
Er ~ ε t 2 ≠ const.!, diffusivity increases with time

Data: σ 2 r ~ t 2.3
Er ~ σ 11. r
Okubo (1971) → Oceanic Diffusion Diagrams
ocean, coastal zone
large lake, coastal zone
mid-size
small lake
A. Okubo, "Oceanic Diffusion Diagrams", Deep Sea Research, Vol. 18, 1971
Typical growth rates:
1 hr ∼ σr = 50 m
1 day ∼ 1 km
1 week ∼ 10 km
1 month ∼ 100 km
Applications:
Instantaneous Point Source (Accident): t2

t1
1 σr
cmax ~ 2-D growth σr
σ 2r
c max
(vertically mixed)
c max
⎛ r2 ⎞
− ⎜⎜ ⎟
2⎟
M ⎝ 2σ r ⎠
c= e
h 2 π σr 2
Continuous release (Routine problems):
∂c ∂ ⎛ ∂ c⎞
U = ⎜E ⎟ − kc Brooks (1960)
∂x ∂y ⎝ y ∂y ⎠
4/3
⎛L⎞
E y = E y0 ⎜ ⎟ 4/3 Law L = 2 3 σy
⎝ b0 ⎠
22% width
3/2
L ⎡ 8 E y0 x ⎤
= ⎢1 + ⎥ ∼ x3/2 !
b ⎣ Ub 2 0 ⎦
c max 3/2
= erf 3
c0 ⎛ 8 E y0 x ⎞
⎜1 + ⎟ −1
⎝ Ub 2 0 ⎠
References: Diffusion and Dispersion Processes
Bird, R.B., Stewart, W.E., and Lightfoot, E.N., 1960, "Transport Phenomena", John Wiley
& Sons, New York
Carslaw, H.S. and Jaeger, J.C., 1959, "Conduction of Heat in Solids", Oxford Press, 2nd
ed.
Elder, J.W., 1959, "The Dispersion of Marked Fluid in Turbulent Shear Flow", Fluid
Mechanics, Vol. 5, Part 4, pp 544-560
Fischer, H.B., 1973, "Longitudinal Dispersion and Turbulent Mixing in Open-Channel

Flow", Annual Review of Fluid Mechanics, Vol. 5, pp 59-78, Annual Review, Inc., Paolo
Alto, Calif.
Fischer, H.B., et al., 1979, "Mixing in Inland and Coastal Waters", Academic Press, New
York
Holly, E.R. and Jirka, G.H., 1986, "Mixing and Solute Transport in Rivers", Field Manual,
U.S. Army Corps of Engineers, Waterways Experiment Station, Tech. Report E-86-11
(NTIS No. 1229110, AD-A174931/6/XAB)
Jirka, G.H. and Lee, J.H.-W., 1994, "Waste Disposal in the Ocean", in "Water Quality and
its Control", Vol. 5 of Hydraulic Structures Design Manual, M. Hino (Ed.), Balkema,
Rotterdam
Rutherford, J.C., 1994, "River Mixing", John Wiley & Sons, Chichester
Taylor, G.I., 1921, "Diffusion by Continuous Movements", Proc., London Math. Soc., Ser.
A, Vol. 20, pp 196-211
Taylor, G.I., 1953, "Dispersion of Soluble Matter in Solvent Flowing Slowly Through a
Tube", Proc. Royal Soc. of London, Ser. A, No. 219, pp 186-203
Taylor, G.I., 1954, "The Dispersion of Matter in Turbulent Flow Through a Pipe", Proc.
Royal Soc. of London, Ser. A, Vol. 223, pp 446-468
Yotsukura, N. and Sayre, W.W., 1976, "Transverse Mixing in Natural Channels", Water
Resources Research, Vol. 12, No. 4, pp 695-704

Handout Jirka PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Handout Jirka PDF

Uploaded by

Copyright:

Available Formats

Environmental Fluid Mechanics Ι 1 Institute for Hydromechanics

MASS TRANSPORT PROCESSES

Advection - mass transport by mean velocity field

Diffusion - mass transport by irregular fluctuations ("random movements")

PASSIVE MIXING PROCESSES - advection + diffusion as they exist

ACTIVE MIXING PROCESSES - generation of mean and random

passive mixing Active mixing

− molecular diffusion of dissolved or suspended matter is caused by random movement

window of unit area

Fick's Law of Diffusion:

Dissolved matter in water: D ∼ 2 × 10-5 cm2/s

Example: Salt diffusion in a vessel

Water Water two hours later

Water + NaCl+ Dye Water + NaCl+ Dye

ADVECTIVE MASS FLUX

− transport of matter due to mean motion (current)

window of unit area

Pure advection in x-direction (no diffusion no reaction)

TOTAL MASS FLUX: Advection + diffusion

MASS CONSERVATION PRINCIPLE

R = rate of production or decay per unit

Solution of the convective diffusion equation:

Initial Conditions (I.C.) t < 0, c = 0 t=0 Sudden mass input

Boundary Conditions (B.C.) c → 0 as (x, y, z) → ∞

[( x − x ) − Ut] + ( y − y1) + ( z − z1)

123 14444424444443 123

1. Position of cloud center:

σ = 2D t = standard deviation ∼ t 0,5

Distance relative to center

Boundary conditions (finite space):

Ex: c = 0 Absorption condition

Multiple sources → superposition, due to linearity of governing equation

TURBULENCE - an instability of a base flow (shear flow)

damping - stable flow

disturbance TURBULENT FLOW

Large eddies: Integral length scale lΙ Integral velocity scale uΙ

Example: Tank with Grid Stirring

net transport w ′c′ > 0

DIFFUSION IN TURBULENT FLOW

→ fluctuating velocity field ⇒ multiple scale eddies (spectrum)!

Ex: Release from POINT SOURCE

a) Short time after release t<t Ι , σ < l Ι

PRACTICAL TURBULENT DIFFUSION EQ.: c → c

Valid for t > tΙ ; σ > lΙ

Diffusivities can have dependence on spatial position and direction:

E = f (base flow characteristics)

TURBULENT CHANNEL FLOW

Rule of thumb: u ∗ ≈ ( 0.05 to 0.1) U

Complete vertical mixing: 2,15 σz = h 10% criterion with Gaussian profile

Typically: Distance to complete vertical mixing 10 to 20 water depth!

LATERAL MIXING Wide Channel, Depth h

Assume isotropy: E y ≅ Ez = 0.067 u ∗ h

Outlines of tracer cloud

Data from Mississippi River study on

NATURAL CHANNELS; IRREGULARITIES

Cumulative Discharge Method Yotsukura and Cobb (1976)

dq = hudy = local discharge → from

q ( y) = ∫ hudy = cumulative discharge

Approximate local balance:

Understanding of advective velocity

LONGITUDINAL DISPERSION (SHEAR FLOW DISPERSION)

Observations: t > t mixing or σ z > h

For long time: