Daniel Gregley 2016
Daniel Gregley 2016
Daniel Gregley 2016
functions
M. Holling, H. Herwig
1. Introduction
gx T h3 gx T L3
Gr = or GrL = . (1)
2 2
In the near wall region, the governing equations for turbulent natural convection reduce to
T
0 = a v T (2)
y y
u
0 = u v + gx (T T0 ) (3)
y y
Figure 1: Standard geometries for turbulent natural convection. Left: infinite vertical channel
[1, 2]; Middle: vertical plate [3, 4]; Right: cavity with differentially heated sidewalls [5]
1
viscous sublayer
outer layer (y )
overlap
wall layer
inner layer
(y y )
y
and Nieuwstadt [1] and Henkes and Hoogendorn [7] showed deviations between this theory and
their data. An alternative approach was proposed by Yuan et al. [8], with profiles obtained by
dimensional considerations and curve fitting.
Despite the intensive research in natural convection for the last decades there is no theory
available today that can accurately describe the near wall region of natural convection. There-
fore, we want to present new results, first shown in Holling and Herwig [9] together with their
implementation as wall functions in a CFD code. Results shown afterwards are valid for fluids
with Pr = 0.7.
2. Temperature Profiles
An expection of the energy equation, (2), shows that the heat flux consists of a molecular part
(a T /y) and a turbulent part (v T ) with its sum equal to the wall heat flux (a T /y|w ).
Thus, the wall heat flux can be identified as the crucial parameter in turbulent natural convection
and a characteristic temperature Tc should be based on it. Following dimensional considerations
we introduce
!1/4
a2 T 3 Tw T
Tc and (4)
g y w Tc
T 1
= Tc (5)
y w
and the non-dimensional wall distance for the inner layer is y y/. For the outer layer a
geometric length (channel width h or distance along the plate L) can be used as a length scale,
resulting in y/h (or y/L).
As shown in Figure 2 the two layers are not sharply separated but an overlap layer exists
in-between. In this overlap layer, the temperature field can be simultaneously described in terms
of the inner layer (using y ) and the outer layer (using ). An asymptotic matching of gradients
in this layer results in
lim y = lim . (6)
y y 0
2
12
10
8
+ . . .
2 = C ln(y ) + D
= y
0 1 0 1 2 3
10 10 10 10 10
y
Figure 3: Non-dimensional temperature profiles of Betts and Bokhari [2] for Gr = 1.43 106
(), from Ampofo and Karayiannis [5] for Gr = 1.59 109 (), of Tsuji and Nagano [3] for
GrL = 1.55 1010 (), for GrL = 3.62 1010 (), for GrL = 7.99 1010 (), for GrL = 8.44 1010 (),
for GrL = 8.99 1010 (), for GrL = 17.97 1010 (+) and of Cheesewright [10] for GrL = 8.65 1010
(). Profiles are shifted by one unit each for better clarity.
This equation can only be fulfilled if both sides are equal to a constant C. Integration of equation
(6) therefore leads to
= C ln y + D. (7)
The constants C = 0.427 and D = 1.93 were determined by comparison with experimental
and numerical data. Figure 3 shows temperature profiles from various studies in their non-
dimensional form. The agreement between the data and the logarithmic profile according to
equation (7) is quite good. Additionally the temperature profile for the viscous sublayer is
plotted. In this region adjacent to the wall heat transport is purely molecular (see Figure 2)
and the temperature profiles are = y , following from equation (2). This again is in good
agreement with the data.
Here, as well as for the velocity profiles in the next section, variable property effects had to
be taken into account when constants were determined from experimental data, see Holling and
Herwig [9] for details.
3. Velocity Profiles
In a straightforward approach, as applied by George and Capp [6], velocity gradients are matched
in the overlap layer. This, however, results in a poor representation of the flow field [1, 7]. In-
stead, in our study the velocity profile in the overlap layer is directly derived from the momentum
equation (3). With an eddy-viscosity approach, u v = t u/y, together with t equation
(3) in non-dimensional form reads
t U
0= + 0 (8)
y y
3
with U = u/uc as non-dimensional velocity based on
gx Tc3 T 2
uc = . (9)
y w
U can be obtained by integrating equation (8) with all terms rewritten as functions of y .
Since the temperature profile is already known in terms of y , see (7), only t / = f (y ) has
to be found. First, this expression is rewritten with the turbulent Prandtl number t as
t t at
= . (10)
Pr a
From the DNS data of Versteegh and Nieuwstadt [1] it is found that t is nearly constant in
the overlap layer with t = 0.9, like for forced convection. The ratio of the turbulent thermal
diffusivity at and the molecular thermal diffusivity a can be obtained by differentiating the
temperature profile (7) and comparing it to the non-dimensional form of equation (2). From
this
at y t t y
= = . (11)
a C Pr C
Now, all terms in equation (8) depend on y only and it can be integrated, from which
CPr
U =
y C ln(y ) 2 + D 0 + E ln(y ) + F (12)
t
results. The constants of integration, E and F , are constant with respect to y but they
depend on the non-dimensional wall shear stress U /y |w :
U U
E = 0.49 2.27 and F = 1.28 + 1.28 (13)
y w y w
This was found by comparison with numerical and experimental data.
Figure 4 shows non-dimensional velocity profiles from various studies. The velocity profiles
in the overlap layer, equation (12), are a good representation (better than the profiles from
George and Capp [6]) of the experimental data as long as the Grashof number is high enough.
Additionally, the velocity profiles in the viscous sublayer are shown, obtained by integrating
equation (3) with u v = 0. In its non-dimensional form they read
1 3 1 2 U
U = y 0 y + y . (14)
6 2 y w
Equation (14) again is a good representation of the experimental data.
The temperature and velocity profiles, (7) and (12), respectively, are valid for equilibrium
boundary layers and for wall angles (angle between the wall and a horizontal line) with
30 . < 90 . For 6= 90 only the component of the gravitational acceleration parallel to
the wall has to be taken into account (gx = g sin ). For 0 (Rayleigh-Benard convection)
alternative temperature and velocity profiles can be derived, see e.g. Holling and Herwig [11, 12].
The temperature and velocity profiles derived above can be used as wall functions for natural
convection in CFD codes, so that coarse grids can be used near the wall. Then, the flow and
temperature fields are not resolved up to the wall, but the center of the first control volume
lies within the overlap layer. The region close to the wall is brigded by the wall functions,
i.e. gradients at the wall are determined using the universal profiles (7) and (12).
First, results for natural convection from a commercial CFD-code are shown to emphasize
the need for a new wall treatment. In the subsequent section results are given that were obtained
with the new wall functions.
4
80
(13)
70
(15)
60
U + . . . 50
40
30
20
10
0 1 0 1 2 3
10 10 10 10 10
y
Figure 4: Non-dimensional velocity profiles of Betts and Bokhari [2] for Gr = 1.43 106 (),
from Ampofo and Karayiannis [5] for Gr = 1.59 109 (), of Tsuji and Nagano [3] for GrL =
1.55 1010 (), for GrL = 3.62 1010 (), for GrL = 7.99 1010 (), for GrL = 8.44 1010 (), for
GrL = 8.99 1010 (), for GrL = 17.97 1010 (+) and of Cheesewright [10] for GrL = 8.65 1010
(). Profiles are shifted by five units each for better clarity.
5
360 0.3
350 0.2
340 0.1
u in m/s
T in K
330 0
320 0.1
310 0.2
300 0.3
0 0.05 0.1 0 0.05 0.1
y in m y in m
Figure 5: FLUENT results ( : fine grid; : coarse grid) with the k- turbulence model
compared to the DNS data of Versteegh and Nieuwstadt [1] for Ra = Gr Pr = 5.0 106 ().
360 0.3
350 0.2
340 0.1
u in m/s
T in K
330 0
320 0.1
310 0.2
300 0.3
0 0.05 0.1 0 0.05 0.1
y in m y in m
Figure 6: CAFFA results (: eight cells; : four cells) using new wall functions compared to
the DNS data of Versteegh and Nieuwstadt [1] for Ra = Gr Pr = 5.0 106 ().
6
0.02 60
50
0.015
40
k in m2 /s2
at /a
0.01 30
20
0.005
10
0 0
0 0.02 0.04 0.06 0.08 0.1 0 0.2 0.4 0.6 0.8 1
y in m y/h
Figure 7: CAFFA results for the turbulent kinetic energy k (, left) and the ratio of turbulent
and molecular thermal diffusivity, at /a, (, right) compared to the DNS data of Versteegh and
Nieuwstadt [1] for Ra = 5.0 106 ().
The proposed set of wall functions / boundary conditions can be used when the center of the
first control volume is within the overlap layer (y & 5).
In Figure 6 results for the infinite vertical channel (Ra = 5.0 106 ) obtained with the new
wall functions and modified turbulence boundary conditions are shown. Deviations of the wall
gradients with respect to DNS results are within 3 % compared to 20 70 % of the FLUENT
results. Also, the computation time can be reduced considerably due to the coarse grids that can
now be used. Figure 7 shows the profile of turbulent kinetic energy k and the ratio of turbulent
and molecular thermal diffusivities, at /a, compared to the DNS data [1]. Though k deviates
from the DNS data, at /a is quite close to the data in [1]. This is due to the modified boundary
condition for , equation (16), by which the correct value of t (and therefore also at ) in the
control volume adjacent to the wall is enforced.
The good agreement of the mean flow field and the rather poor representation of k and
are not unique for natural convection but can also be observed in forced convection. Numerical
solutions for complex interior forced convection flows, as measured by Mocikat et al. [14], for
example, show good agreement for the mean flow profiles based on standard CFD code results.
However, e.g. the turbulent kinetic energy k then is wrong by more than 100 %.
The proposed set of wall functions / boundary conditions together with a standard turbulence
model may be a reasonable approach as long as there is no better turbulence model for natural
convection.
5. Conclusions
The temperature and velocity profiles for turbulent natural convection at vertical walls were
analysed asymptotically. A two-layer structure can be identified with a viscosity influenced
inner layer and a turbulent outer layer. The temperature profile is obtained by asymptotic
matching of temperature gradients in the overlap layer. The momentum equation is rewritten in
terms of the wall distance and can then be integrated to get the velocity profile for the overlap
layer. These profiles, (7) and (12), are in good agreement with numerical and experimental data
sets.
These profiles can be used in CFD-codes as improved wall functions for natural convection
since commercial codes with standard wall treatment give only poor results. The standard
k- turbulence model was chosen with a modified boundary condition for . Results for the
7
temperature and velocity profiles are in good agreement with DNS data, though the profile for
the turbulent kinetic energy k show considerable deviations.
An alternative turbulence model adapted to natural convection is needed to improve the
agreement for k (and ). Also, an extension to mixed convection at vertical walls is planned,
see for example Balaji et al. [15]. Additionally, an extension to non-equilibrium boundary layers
would be an option to increase the range of validity.
References
[1] T.A.M. Versteegh and F.T.M. Nieuwstadt, A direct numerical simulation of natural con-
vection between two infinite vertical differentially heated walls: scaling laws and wall func-
tions, Int. J. Heat Mass Transfer 42, 36733693 (1999).
[2] P.L. Betts and I.H. Bokhari, Experiments on turbulent natural convection in an enclosed
tall cavity, Int. J. Heat Fluid Flow 21, 675683 (2000).
[3] T. Tsuji and Y. Nagano, Characteristics of a turbulent natural convection boundary layer
along a vertical flat plate, Int. J. Heat Mass Transfer 31, 17231734 (1988).
[4] T. Tsuji and Y. Nagano, Turbulence measurements in a natural convection boundary layer
along a vertical flat plate, Int. J. Heat Mass Transfer 31, 21012111 (1988).
[5] F. Ampofo and T.G. Karayiannis, Experimental benchmark data for turbulent natural
convection in an air filled square cavity, Int. J. Heat Mass Transfer 46, 35513572 (2003).
[6] W.K. George and S.P. Capp, A theory for natural convection turbulent boundary layers
next to heated vertical surfaces, Int. J. Heat Mass Transfer 22, 813826 (1979).
[7] R.A.W.M. Henkes and C.J. Hoogendoorn, Numerical determination of wall functions for
the turbulent natural convection boundary layer, Int. J. Heat Mass Transfer 33, 1087
1097 (1990).
[8] X. Yuan, A. Moser and P. Suter, Wall functions for numerical simulation of turbulent
natural convection along vertical plates, Int. J. Heat Mass Transfer 36, 44774486 (1993).
[9] M. Holling and H. Herwig, Asymptotic analysis of the near wall region of turbulent natural
convection flows, J. Fluid Mech. 541, 383397 (2005).
[10] R. Cheesewright, Turbulent natural convection from a vertical plane surface, J. Heat
Transfer 90, 18 (1968).
[11] M. Holling and H. Herwig, A new Nusselt/Rayleigh number correlation and wall functions
for turbulent Rayleigh-Benard convection, Proc. IHTC13 Sydney, 13.-18. August, (2006).
[12] M. Holling and H. Herwig, Asymptotic analysis of heat transfer in turbulent Rayleigh-
Benard convection, Int. J. Heat Mass Transfer , 49, 11291136 (2006).
[13] J.H. Ferziger and M. Peric, Computational methods for fluid dynamics, 2nd ed., Springer,
Berlin (1999).
[15] C. Balaji, M. Holling and H. Herwig, New approaches to treatment of turbulent mixed
convection flows using asymptotics, submitted for publication (2006).