2-D channel Flow with/without thermal field

Muzammal Arif
Graduate Student, Department of Mechanical Engineering
King Fahd University of Petroleum and Minerals, (KFUPM)
Dhahran, Saudi Arabia
Email: [email protected]

Dr. Shahzada Zaman Shuja

Professor Mechanical Engineering Department
KFUPM, Dhahran, Saudi Arabia

Department of Mechanical Engineering

King Fahd University of Petroleum and Minerals

G201409640

MUZAMMAL ARIF

AFM-I TERM PAPER (ME-532)

Table of Contents
Introduction----------------------------------------------------------------------------------(3)
Relevance to Engineering and Industry--------------------------------------------------(3)
Literature Review --------------------------------------------------------------------------(4)
Viscous open-channel flow in laminar regime------------------------------------------(5)
Method for implementing the proposed solution along with improvements. -------(7)
Velocity Distribution in 2D channel with varying angle () -------------------------(10)
Velocity Distribution in pipe with varying angle () ---------------------------------(11)
References/Bibliography-----------------------------------------------------------------(15)

G201409640

MUZAMMAL ARIF

AFM-I TERM PAPER (ME-532)

1. Introduction
Channel flow is defined as flow in one direction within a conduit. It may be closed channel
flow or open channel flow depending upon whether the fluid flowing is entirely in contact with
rigid boundaries or it has one of its surface free from solid boundary i.e. open to atmosphere.
Pipe flow is an example of closed channel flow while flow in rivers or canals are examples of
open channel flow. In both closed and open channel flow, the flow can be either uniform or
non-uniform, steady or unsteady flow. In this article we will focus our attention on two
dimensional flow in a closed conduit. The flow will be laminar, uniform, steady and fully
developed and our objective will be to find distribution of different variable like velocity,
temperature and shear stress with or without the application of thermal field. We will utilize
generalized form of Navier-Stokes equations.

2. Relevance to Engineering and Industry

Channel flow is of great importance in engineering and Industry specially flow in a pipe
which is a form of closed channel flow is used extensively in Boilers, heat exchangers
and condensers etc. Closed channel flow in rectangular duct that we are going to deal
with hold its application in different field of engineering like flow in ducts of heating,
ventilation and air-conditioning systems. Flow in Plate type heat exchangers etc.

G201409640

MUZAMMAL ARIF

AFM-I TERM PAPER (ME-532)

3. Literature Review
In paper [1] Simulating 2D open-channel flows through an SPH model by I. Federico, S. Marrone,
A. Colagrossi, F. Aristodemo, M. Antuono, two dimensional model of open channel flow is given.
They developed an appropriate algorithm to enforce different upstream and downstream flow
conditions and simulate uniform, non-uniform and unsteady flows. The main focus was in velocity
field, pressure forces, water depths etc. Comparisons between numerical results, theory and
experimental data is provide.
The governing equations were

Boundary Conditions:
Free surface
Solid boundary
In/Out flow boundary
The free surface boundary conditions can be easily handled by the SPH model. The implicit
enforcement of the free surface dynamic boundary condition is one of the main advantages of the
SPH method in comparison to other solvers where this boundary condition has to be forced
directly.

G201409640

MUZAMMAL ARIF

AFM-I TERM PAPER (ME-532)

The channel bottom is modelled through the fixed ghost particle technique. In contrast to the
classical ghost particles that are instantaneous mirrors of the fluid particles with respect to the body
surface, the fixed ghost particles are associated to interpolation points internal to the fluid. The
main advantage of using the fixed ghost particles instead of the classical ghost technique is that
their distribution is always uniform and does not depend on the fluid particle positions.

Figure 2: Initial sketch of the computational domain

In total, four sets of particles are used: fluid (f), fixed ghost (s), inflow (i) and outflow (o) particles.
Similarly to the fixed ghost particles, the in/out-flow particles affect the fluid particles but not vice
versa. At the inflow, the desired velocity and pressure conditions are assigned to the inflow
particles. When inflow particles cross the inflow threshold, they become fluid particles and they
evolve in accordance with the SPH equations. As concerns the outflow particles, it is possible to
impose specific outflow conditions (similarly to the inflow case) or open boundary conditions. In
the latter case, the fluid particles that cross the outflow threshold become outflow particles. Their
physical variables are frozen. Three different test cases were considered while I am going to
discuss only the first one i.e. viscous open-channel flow in laminar regime.
Viscous open-channel flow in laminar regime:
A uniform, steady and laminar flow in a free-surface channel is a special case of the Poiseuille
flow. The distribution of velocity u (z) for two-dimensional channel flow is given by a secondorder equation given by

G201409640

MUZAMMAL ARIF

AFM-I TERM PAPER (ME-532)

Where g is the gravity acceleration, h is the total water depth, S0 is the bottom slope and z is the
vertical abscissa whose origin is located at the channel bottom. The Reynold number Re =

().

is evaluated using the average horizontal velocity U =

The fluid domain has length L
= 2h and slope S0 = 0.001. The initial and inflow boundary conditions are imposed and elementary
fluid domain is sketched as follows

Figure 4: Particle distribution at Re=10

Figure 5: Particle distribution at Re=100

Figure 6: Particle distribution at Re=200

G201409640

MUZAMMAL ARIF

AFM-I TERM PAPER (ME-532)

4. Method for implementing the proposed solution along with improvements.

We will make analytical solution of the two dimensional laminar fluid flowing due to gravity
as a function of its slant height i.e. slope and visualize the distribution of velocities with the
changing of slope as well as Reynold number. Additionally we are going to replace the 2 D
channel with a pipe and visualize the distribution of parameters like velocity as a function of
slope as well as a function of Reynolds number.

= g sin ()

= g cos ()

v=w=0 which gives

=0

Navier Stokes equation in x-direction is

= 0 also
= 0 (no pressure gradient along x) and

variation in u along the z- direction) will give

Putting t= v = w =

= 0 (no

G201409640

MUZAMMAL ARIF

AFM-I TERM PAPER (ME-532)

2
=
2
2
= sin ()
2
2

sin ()
2

Integration gives

=
sin () y + A

Again Integration gives

sin () y2 + Ay +

Applying the BC
1. @ y = 0
u = 0 gives B = 0
2. @ y = h

= 0 i.e.

=0
0=

or

sin () h + A

sin ()h = A

Put the values of A and B in equation (1) gives

sin ()y2 +
sin ()hy
2

sin ()(2 2 )

(1)

G201409640

MUZAMMAL ARIF

AFM-I TERM PAPER (ME-532)

First we will reproduce the results shown in figure 3.

We will takesin () = 0.001, = 1000/3 and g = 9.81m2/s, =8.9x10-4 kg/(ms)

Graph at slope =0.001

1

Height of liquid column (m)

0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0

0.002

0.004

velocity (m/s)

0.006

0.008

0.01

G201409640

MUZAMMAL ARIF

AFM-I TERM PAPER (ME-532)

Velocity Distribution as a function of angle in 2D channel:

Now we will vary the angle theta of the 2D plane from 0 degree to 60 with an interval of 15 degree
and we will measure the distribution of velocities. Plot of velocities as a function of height as well
as a function of angle theta is given in the following diagram.

Velocities at different values of angle ()

1
0.9

15 deg

30 deg

45 deg

60 deg

Height of liquid column (m)

0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0.00E+00

1.00E+00

2.00E+00

3.00E+00

4.00E+00

velocity (m/s)

10

5.00E+00

6.00E+00

7.00E+00

G201409640

MUZAMMAL ARIF

AFM-I TERM PAPER (ME-532)

5. Velocity Distribution in pipe with varying angle ():

In our next step we will find velocity distribution of fluid flowing in a pipe having a gradient.

gz =g sin ()

gr = gcos ()

With u = ur = 0 and

=0

Continuity equation gives us

=0

Navier-Stokes r-direction:

+gr
N-S equation in r-direction for steady flow gives
0 =
11

G201409640

MUZAMMAL ARIF

AFM-I TERM PAPER (ME-532)

0 = ()
Navier-Stokes -direction:

+g

N-S equation in -direction for steady flow gives

0 =
Navier-Stokes z-direction:

N-S equation in z-direction for steady flow gives

0=

0=

12

1
(
(
)) +

1
(
(
)) + ()

()
(
)) =

G201409640

MUZAMMAL ARIF

AFM-I TERM PAPER (ME-532)

()
(
)) =

Integration w.r.t r gives

(

2
() +
)=

=
() +

2
Integrating again w.r.t r gives

2
=
() + +
4
Constants A and B can be found by the B.C which are

@r=0

=0

which gives A=0

=
@ r =R

uz =0

2
4

() +

which gives

2
()
4

Put the value of B in equation (2) gives

=

2
2
() +
()
4
4

13

() 2
( 2 )
4

(2)

G201409640

MUZAMMAL ARIF

AFM-I TERM PAPER (ME-532)

Distribution of velocities for the flow of a liquid in a pipe for the various angle is shown in the
figure. Angle is varied from. The various parameters taken are given below
= 1000/3 , g = 9.81m2/s, =8.9x10-4 kg/(ms), R=1 m

Velocity distribution as a function of angle ()

1
0.8

pipe radius (m)

0.6
0.4
0.2

15 deg
0
0.00E+00

5.00E-01

30 deg
1.00E+00

45 deg
1.50E+00

-0.2
-0.4
-0.6
-0.8
-1

velocity(m/2)

14

60 deg
2.00E+00

2.50E+00

3.00E+00

G201409640

MUZAMMAL ARIF

AFM-I TERM PAPER (ME-532)

References/Bibliography
1. Recurrence in 2D inviscid channel flow
(Y. Charles Li)

2. Simulating 2D open-channel flows through an SPH model

(I. Federico, S. Marrone, A. Colagrossi, F. Aristodemo, M. Antuono)

3. Understanding heat transfer in 2D channel flows including recirculation (M.H.

Dirkse, W.K.P. van Loon, J.D. Stigter, G.P.A. Bot)

4. CFD 2D simulation of viscous flow during ECAE through a rectangular die with
parallel slants
(Alexander V. Perig & Nikolai N. Golodenko)

5. Second law analysis of the 2D laminar flow of two-immiscible, incompressible

viscous fluids in a channel
(Fethi Kamish, Hakan F. oztop)

6. Guidelines for Modeling a 2D Rough Wall Channel Flow

(Stefano Leonardi Paolo Orlandi Lyazid Djenidi Robert A. Antonia)

7. Direct numerical simulation of a turbulent open channel flow with passive heat
transfer
(D. M. LU and G. HETSRONIt)

8. DNS of turbulent heat transfer in a channel flow with a high spatial resolution
(Makoto Kozuka, Yohji Seki, Hiroshi Kawamura)

9. DNS of velocity and thermal fields in turbulent channel flow with transverse-rib
roughness
(Yasutaka Nagano, Hirofumi Hattori a, Tomoya Houra)

10. Passive heat transfer in a turbulent channel flow simulation using large eddy
simulation based on the lattice Boltzmann method framework (Hong Wu, Jiao Wang,
Zhi Tao)

11. Use of direct numerical simulation to study the effect of Prandtl number on
temperature fields
(Yang Na a, Dimitrios V. Papavassiliou b, Thomas J. Hanratty)

12. Simulating two-dimensional thermal channel flows by means of a lattice Boltzmann

method with new boundary conditions
(Annunziata DOrazio, Sauro Succi)

13. An investigation of turbulent open channel flow with heat transfer by large eddy
simulation
(Lei Wang, Yu-Hong Dong, Xi-Yun Lu)

14. Flow distribution and pressure drop in 2D meshed channel circuits

(Daniel Tondeur , YilinFan b, LingaiLuo)
15. Large eddy simulation of stably stratified turbulent open channel flows with low- to
high-Prandtl number
(Lei Wang, Xi-Yun Lu)

16. Heat transfer in fully developed turbulent channel flow: comparison between
experiment and direct numerical simulations
(M. TEITEL and R. A. ANTONIA)
15