Geoscience Reference
In-Depth Information
Eq. (2.49) may be derived by substituti ng the log distribution of flow velocity along
the flow depth and the relation l m
z
into Eq. (2.48). Therefore, model
(2.49) can be seen as a special case of model (2.48).
= κ
(
1
z
/
h
)
2.3.3 One-equation turbulence models
In the often used one-equation turbulence model, the e d dy viscosity is determined using
the Kolmogorov-Prandtl expression, which adopts k as the velocity scale and reads
kL m
c µ
ν
=
(2.50)
t
where c µ
is a coefficient of about 0.084.
Unlike the mixing length model, the one-equation turbulence model uses a trans-
port equation to determine the turbulent energy k and, in turn, the fluctuating
velocity scale. The transport equation of k can be derived in exact form from the
continuity and Navier-Stokes equations. For high Reynolds numbers, this equation
reads
u i
u j u j
2
u i
u i
k
+
) =−
p
ρ
u i u j ¯
u i
x j ν
x j
x i ( ¯
u i k
+
(2.51)
t
x i
x j
The three terms on the right-hand side of Eq. (2.51) represent the diffusion, produc-
tion, and dissipation of k , respectively. To close this equation, the diffusion term is
treated in analogy to Eq. (2.46), and the dissipation term is determined as c D k 3 / 2
/
L m ,
thus yielding the modeled k equation:
ν
c D k 3 / 2
L m
k
t +
) =
σ k
k
t
x i ( ¯
+
P k
u i k
(2.52)
x i
x i
u i u j ¯
where P k is the production of turbulence by shear, defined as P k =−
u i
/∂
x j ;
σ k
c µ
is a coefficient of about 1.0; and c D is a coefficient, usually set as c D
.
For the turbulence in a state of local equilibrium, its production is equal to dissipa-
tion, and then Eq. (2.52) can be simplified as
0.08
/
2
c D k 3 / 2
0 in the case
of shear flows. By using Eqs. (2.48) and (2.50), the following relation can be derived
(Rodi, 1993):
ν
(∂ ¯
u
/∂
z
)
/
L m
=
t
c 1 / 4
D
c 3 / 4
L m
l m =
(2.53)
µ
Therefore, L m can be determined using simple empirical formulas similar to those
for the mixing length l m .
 
Search WWH ::




Custom Search