J Ijheatmasstransfer 2013 12 042
J Ijheatmasstransfer 2013 12 042
J Ijheatmasstransfer 2013 12 042
a r t i c l e i n f o a b s t r a c t
Article history: In this study, numerical analysis of the heat transfer enhancement and fluid flow characteristics of a
Received 4 October 2013 rotating cylinder under the influence of magnetic dipole in the backward facing step geometry is con-
Received in revised form 9 December 2013 ducted. The governing equations are solved with a finite element based commercial solver. The effects
Accepted 13 December 2013
of Reynolds number (10 6 Re 6 200), cylinder rotation angle (75 6 X 6 75) and strength of the mag-
netic dipole (0 6 c 6 16) on the heat transfer characteristics are studied for backward facing step flow.
It is observed that the length and size of the recirculation zones can be controlled with magnetic dipole
Keywords:
strength and cylinder rotation angles. As the Reynolds number increases, local Nusselt number increases
Magnetic dipole
Rotating cylinder
and number of peaks in the presence of the magnetic field decreases. The effect of cylinder rotation on the
Backward facing step local Nusselt number distribution is more pronounced at low Reynolds number.
CFD Ó 2013 Elsevier Ltd. All rights reserved.
0017-9310/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.ijheatmasstransfer.2013.12.042
F. Selimefendigil, H.F. Öztop / International Journal of Heat and Mass Transfer 71 (2014) 142–148 143
Nomenclature
v0
vm ¼ ð7Þ
1 þ aT þ T 0
Finally, using the constitutive equation in (4), the body force in
the momentum equation can be stated as
direction.
On the cylinder surface, specified velocity components with adi- 2.1. Validation study
abatic wall boundary condition is used, u ¼ xðy y0 Þ;
v ¼ xðx x0 Þ; @T@n
¼ 0. The numerical code is first checked against the benchmarked
results of backward facing step reported in the literature [28–
The relevant physical nondimensional numbers are Reynolds 32]. Table 2 shows the results of the reattachment length divided
number (Re), magnetic number (Mn) and cylinder rotation angle by step height at Reynolds number of 100 for expansion ratio of
(X). Eqs. (1)–(8) along with the boundary conditions are solved 2. Minimum deviation for the percentage in the error is obtained
with COMSOL Multiphysics simulation package (a general purpose for the results of [28] which is 0.2%. The agreement between
finite element solver [27]). P2 – P1 Lagrange finite elements are the other sources is less than 5 percent, only 6.8 percent error
used to discretize velocity components and pressure, and La- is obtained for the results of [32].
grange-quadratic finite elements are chosen for temperature.
COMSOL solver adds artificial diffusion with the streamline up- 3. Results and discussion
wind Petrov–Galerkin method (SUPG) to handle local numerical
instabilities. Segregated parametric solvers are used for fluid flow As stated earlier, the overall purpose of this study is to investi-
and heat transfer variables. Biconjugate gradient stabilized itera- gate the effects of rotating cylinder in the presence of a magnetic
tive method solver (BICGStab) is used for fluid flow and heat trans- dipole located below the channel downstream of the step on the
fer modules of software. The computational domain is divided into heat transfer and fluid flow characteristics.The main parameters
65,362 triangular elements. The mesh is finer near the walls of the that effect the fluid flow and thermal characteristics are Reynolds
channel and cylinder to resolve the high gradients in the thermal number (Re), magnetic number (Mn), cylinder rotation angle (X),
and velocity boundary layer and in the vicinity of the step for the cylinder position and diameter, location and strength of the mag-
recirculation region downstream of the step. Mesh independence netic dipole, expansion ratio and step height. In this study, the
study is also assured to obtain an optimal grid distribution with influence of the Reynolds number (10 6 Re 6 200), cylinder rota-
accurate results and minimal computational time. Five different tion angle (75 6 X 6 75) and strength of the magnetic dipole
grid sizes are tested and the convergence in the length-averaged (0 6 c 6 16) are examined for convective heat transfer enhance-
Nusselt number (along the bottom wall downstream of the step) ment over a backward facing step flow.
is checked. The averaged Nusselt number results at
(X ¼ 75; c ¼ 16) for Reynolds numbers of 10 and 100 are tabu-
3.1. Effects of magnetic dipole strength
lated in Table 1. From this table, grid size of 65,362 is decided to
be fine enough to resolve the flow and thermal field for the given
Figs. 3 and 4 demonstrate the effects of magnetic dipole
flow parameters. Distribution of the grid in the vicinity of the step
strength on the structure of flow and thermal patterns for fixed
and cylinder is depicted in Fig. 2.
values of Re = 10, X ¼ 75. In this case, cylinder rotation is in the
Local Nusselt number is defined as
counterclockwise direction. For this position of the cylinder, the
hx S @h top half of the cylinder is located above the step and the flow over
Nux ¼ ¼ : ð9Þ
k @n S the cylinder is effected. This result is supported by Ref. [15]. It is
Table 2
Reported values for the reattachment lengths X R at Reynolds number 100 (Expansion
Table 1 ratio of 2).
Variation of average Nusselt number for different grid densities.
X R /S Error (%)
Grid size Re = 10, Numean Re = 100, Numean
Present 4.98 0
8644 0.981 2.052 Acharya et al. [15] 4.97 0.2
18,582 0.956 2.032 Lin et al. [16] 4.91 1.4
31,323 0.954 2.023 Dyne et al. [17] 4.89 1.8
65,362 0.927 1.954 El-Refaee et al. [18] 4.77 4.21
104,424 0.925 1.945 Cochran et al. [19] 5.32 6.82
F. Selimefendigil, H.F. Öztop / International Journal of Heat and Mass Transfer 71 (2014) 142–148 145
behavior [24,25]. Isotherm plots show that the effect of the convec-
tion currents due to magnetic force is seen as the fluctuating of the
isotherms behind the step and on the bottom in the vicinity of the
cylinder. Fig. 5 shows the x-component of the velocity at channel
(a) γ = 0
section x ¼ 2S for different values of magnetic dipole strength. As
it is shown in Fig. 5, negative values of the x-component of the
velocity corresponding to a recirculation zone first forms at
c ¼ 8, then decrease in size but increases in strength for c ¼ 16.
The length and size of the recirculation zone can be controlled with
(b) γ = 8 magnetic dipole strength. Lengthwise local Nusselt number distri-
butions along the bottom wall for varying Reynolds number and
magnetic dipole strength are depicted in Fig. 6. x-axis denotes
the normalized (by step height) distance from the step. The peak
in the Nusselt number corresponds to a point in the vicinity of
the reattachment point. At Re = 10, with increasing the strength
(c) γ = 16 of magnetic dipole, several peaks which indicate the reattachment
points of the additional reattachment points of the separation bub-
Fig. 3. Effect of magnetic dipole strength on the streamlines for fixed values of
Re = 10, X ¼ 75.
bles are seen in the plot. With further increment of the magnetic
dipole strength, the peak values in the Nusselt number increases.
As the Reynolds number increases, local Nusselt number increases
and number of peaks in the presence of the magnetic field de-
creases. Table 3 shows the maximum, average (length averaged
along the bottom wall downstream of the step) of the Nusselt
number along with the location of the maximum Nusselt number
(a) γ = 0 for varying Reynolds numbers and magnetic dipole strength val-
ues. At Re = 10, although the peak value in the Nusselt number in-
creases at c ¼ 16, average Nusselt number decreases in the
presence of the magnetic field. This is due to the fact that external
magnetic field acts in a way to decrease the local heat transfer in
some locations and increase it in some others (see Fig. 6). At
(b) γ = 8
Re = 100 and Re = 200, maximum and average Nusselt number in-
creases as the magnetic dipole strength increases.
(c) γ = 16 Figs. 7 and 8 show the effect of varying cylinder rotation angle
(X) on the streamlines and isotherms for fixed values of Re = 10
Fig. 4. Effect of magnetic dipole strength on the isotherms for fixed values of and c ¼ 12. The case X ¼ 0 corresponds to a motionless cylinder,
Re = 10, X ¼ 75.
which is shown in Fig. 6b. In this case, several vortices appear be-
hind the step and in the vicinity of the cylinder close to the bottom
wall. A recirculation zone is also seen on the upper wall of the
1 channel. A negative value of the rotation (X) indicates clockwise
0.8 rotation of the cylinder. When the cylinder rotates in the clockwise
0.6 γ=0 γ=16
direction, more flow is entrained into the wake of the cylinder. The
0.4
size and extent of the recirculation bubble appearing on the upper
wall increase compared to motionless cylinder case. some portion
0.2
of the flow is directed towards the bottom of the cylinder and re-
u
0
lated to this effect, the vortex appear on the bottom of the cylinder
−0.2 for motionless cylinder case disappears. When the cylinder rotates
−0.4 γ=8
in the counterclockwise direction X ¼ 75, the motion of the fluid
−0.6 flowing through the top of the cylinder and the size of the recircu-
−0.8 lation zone on the upper channel wall are not affected compared to
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
y/S
motionless cylinder case. The flow structure near the bottom and
right of the cylinder are affected to some extent. Fig. 8 shows the
Fig. 5. x-component velocity distribution for different values of magnetic dipole effect of cylinder rotation on the isotherms. For clockwise rotation
strength at x ¼ 2S, Re = 10, X ¼ 75. direction, the isotherms fluctuates more on the upper wall and
right and bottom parts of the cylinder due to the formation of
the vortices compared to motionless cylinder case. Fig. 9 demon-
observed that with increasing values of magnetic dipole strength, strates the effect of cylinder rotation angles on the x-component
the flow behind the step and below the cylinder are affected. A of the velocity at channel section x ¼ 2S at Re = 10, c ¼ 12. A neg-
recirculation region is seen and increases in size and strength with ative values of the x-component of the velocity corresponds to a
increasing values of magnetic dipole strength. The flow is seen to recirculation zone. As it can be seen in the figure, the extent and
be directed towards the bottom wall downstream of the step and strength of the recirculation bubble can be controlled with cylinder
cylinder. This effect is due to the spatial variation in the magneti- rotation angle. The effect of varying cylinder rotation angles on the
zation which is induced through temperature gradient. The inho- local Nusselt number distributions are depicted in Fig. 10 for
mogeneous magnetic body force is responsible for this flow Re = 10, 100 and 200 at c ¼ 12. The peaks in the Nusselt number
146 F. Selimefendigil, H.F. Öztop / International Journal of Heat and Mass Transfer 71 (2014) 142–148
20
15
5
10
5
0 −1 0 1 0
10 10 10 −1
10
0
10 10
1
(c) Re=200,Ω=75
25
20 γ=0
γ=8
15 γ=16
Nux
10
0 −1 0 1
10 10 10
x/S
Fig. 6. Effect of magnetic dipole strength on the local Nusselt number distribution for different Reynolds number, X ¼ 75.
Table 3
Variation of average, maximum and location of maximum Nusselt number for
different Reynolds number and strength of magnetic dipole at X ¼ 75.
(c) Ω = 75
Fig. 8. Effect of cylinder rotation angles on the isotherms for fixed values of Re = 10,
(a) Ω = −75 c ¼ 12.
(b) Ω = 0
0.5 Ω=−75
Ω=0
U
Ω=75
(c) Ω = 75
Fig. 7. Effect of cylinder rotation angle on the streamlines for fixed values of −0.5
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
Re = 10, c ¼ 12.
y/S
15
10
10
5 5
0 −2 0 −2 −1 0 1
10 10
−1 0
10 10
1 10 10 10 10
x
Nu
10
0 −2 −1 0 1
10 10 x/S 10 10
Fig. 10. Effect of cylinder rotation angle on the local Nusselt number distribution for different Reynolds number, c ¼ 12.
Table 4
Variation of average, maximum and location of maximum Nusselt number for
different Reynolds numbers and cylinder rotation angles at c ¼ 12.
(c) Re=200
(a) Re=10 Fig. 12. Effect of Reynolds number on the isotherms for fixed values of
X ¼ 75; c ¼ 8.
4. Conclusions
(b) Re=100
In this study, a numerical study for the analysis of the heat
transfer enhancement and fluid flow characteristics of a rotating
cylinder in the backward facing step geometry under the influence
of magnetic dipole is performed. The effects of Reynolds number
(10 6 Re 6 200), cylinder rotation angle (75 6 X 6 75) and
(c) Re=200 strength of the magnetic dipole (0 6 c 6 16) on the heat transfer
characteristics are studied for backward facing step flow. Following
Fig. 11. Effect of Reynolds number on the streamlines for fixed values of
conclusions can be drawn:
X ¼ 75; c ¼ 8.
At Re = 10, the external magnetic field acts in a way to decrease [10] H. Stuer, A. Gyr, W. Kinzelbach, Laminar separation on a forward facing step,
Eur. J. Mech. B/Fluids 18 (1999) 675–692.
the local heat transfer in some locations and increase it in some
[11] M. Sherry, D. LoJacono, J. Sheridan, An experimental investigation of the
others and at Re = 100 and Re = 200, maximum and average recirculation zone formed downstream of a forward facing step, J. Wind Eng.
Nusselt number increases as the magnetic dipole strength Ind. Aerodyn. 98 (2010) 888–894.
increases. [12] H. Abu-Mulaweh, Turbulent mixed convection flow over a forward-facing step
– the effect of step heights, Int. J. Therm. Sci. 44 (2005) 155–162.
When the cylinder rotates in the clockwise direction, more flow [13] H. Abu-Mulaweh, A review of research on laminar mixed convection flow over
is entrained into the wake of the cylinder and some portion of backward- and forward facing steps, Int. J. Therm. Sci. 42 (2003) 897–909.
the flow is directed towards the bottom of the cylinder and [14] I. Yilmaz, H.F. Oztop, Turbulence forced convection heat transfer over double
forward facing step flow, Int. Commun. Heat Mass Transfer 33 (2006) 508–
related to this effects the size and extent of the recirculation 517.
bubble appearing on the upper wall and on the bottom of the [15] A. Kumar, A.K. Dhiman, Effect of a circular cylinder on separated forced
cylinder are affected. convection at a backward-facing step, Int. J. Therm. Sci. 52 (2012) 176–185.
[16] J. Nie, Y. Chen, H. Hsieh, Effects of a baffle on separated convection flow
The effect of rotation on the local Nusselt number distribution is adjacent to backward-facing step, Int. J. Therm. Sci. 48 (2009) 618–625.
more pronounced at low Reynolds number. [17] M. Rahman, M. Alim, M. Sarker, Numerical study on the conjugate effect of
At Re = 10, the maximum and average heat transfer increase Joule heating and magnato-hydrodynamics mixed convection in an obstructed
lid-driven square cavity, Int. Commun. Heat Mass Transfer 37 37 (2010) 524–
with cylinder rotation but for higher Reynolds numbers, there 534.
is slightly change of peak value of the Nusselt number and its [18] R. Nasrin, Laminar combined magnetoconvection in a wavy enclosure with the
location with cylinder rotation compared to motionless cylinder effect of heat conducting cylinder, Int. Commun. Heat Mass Transfer 38 (2011)
1269–1278.
case.
[19] A. Ishak, R. Nazar, I. Pop, Mhd convective flow adjacent to a vertical surface
with prescribed wall heat flux, Int. Commun. Heat Mass Transfer 36 (2009)
554–557.
References [20] O. Aydin, A. Kaya, Mhd-mixed convection from a vertical slender cylinder,
Commun. Nonlinear Sci. Numer. Simul. 16 (4) (2011) 1863–1873.
[21] M. Pirmohammadi, M. Ghassemi, Effect of magnetic field on convection heat
[1] D. Barkley, M. Gabriela, M. Gomes, R.D. Henderson, Three-dimensional transfer inside a tilted square enclosure, Int. Commun. Heat Mass Transfer 36
instability in flow over a backward-facing step, J. Fluid Mech. 473 (2002) (2009) 776–780.
167–190. [22] I. Nkurikiyimfura, Y. Wang, Z. Pan, Heat transfer enhancement by magnetic
[2] H. Iwai, K. Nakabe, K. Suzuki, Flow and heat transfer characteristics of nanofluids – a review, Renew. Sustain. Energ. Rev. 21 (2013) 548–561.
backward-facing step laminar flow in a rectangular duct, Int. J. Heat Mass [23] H.F. Oztop, K.A.-S.I. Pop, Mhd mixed convection in a lid-driven cavity with
Transfer 43 (2000) 457–471. corner heater, Int. J. Heat Mass Transfer 54 (2011) 494–3504.
[3] J.G. Barbosa-Saldana, N.K. Anand, Flow over a three-dimensional horizontal [24] T. Strek, H. Jopek, Computer simulation of heat transfer through a ferrofluid,
forward-facing step, Numer. Heat Transfer, Part A 53 (2008) 1–17. Phys. Stat. Sol. (b) 244 (3) (2007) 1027–1037.
[4] E. Erturk, Numerical solutions of 2-d steady incompressible flow over a [25] B. Finlayson, Convective instability of ferromagnetic fluids, J. Fluid Mech. 40
backward-facing step, part I: high reynolds number solutions, Comput. Fluids (1970) 753–767.
37 (2008) 633–655. [26] A. Jafari, T. Tynjala, S. Mousavi, P. Sarkoma, Simulation of heat transfer in a
[5] F. Selimefendigil, H.F. Oztop, Identification of forced convection in pulsating ferrofluid using computational fluid dynamics technique, Int. J. Heat Fluid
flow at a backward facing step with a stationary cylinder subjected to Flow 29 (2008) 1197–1202.
nanofluid, Int. Commun. Heat Mass Transfer 45 (2013) 111–121. [27] COMSOL FEMLAB 3.2 Users Guide, Comsol AB, 2005.
[6] F. Selimefendigil, H.F. Oztop, Numerical analysis of laminar pulsating flow at a [28] S. Acharya, G. Dixit, Q. Hou, Laminar mixed convection in a vertical channel
backward facing step with an upper wall mounted adiabatic thin fin, Comput. with a backstep: a benchmark study, ASME HTD 258 (1993) 11–20.
Fluids 88 (2013) 93–107. [29] J. Lin, B. Armaly, T. Chen, Mixed convection in buoyancy-assisted vertical
[7] Y.C. Shih, J. Khodadadi, K. Weng, H. Oztop, Transient leading to periodic fluid backward-facing step flows, Int. J. Heat Mass Transfer 33 (1990) 2121–2132.
flow and heat transfer in a differentially-heated cavity due to an insulated [30] B. Dyne, D. Pepper, F. Brueckner, Mixed convection in a vertical channel with a
rotating object, in: Proceedings of HT2007 2007 ASME-JSME Thermal backward-facing step, ASME HTD 258 (1993).
Engineering Summer Heat Transfer Conference, Vancouver, British Columbia, [31] M. El-Refaee, M. El-Sayed, N. Al-Najem, I. Megahid, Steady-state solutions of
Canada, 2007. buoyancy-assisted internal flows using a fast false implicit transient scheme
[8] B.F. Armaly, F. Durst, J.C.F. Pereier, B. Schonung, Experimental and theoretical (fits), Int. J. Numer. Methods Heat Fluid Flow 6 (1996) 3–23.
investigation of backward-facing step flow, J. Fluid Mech. 127 (1983) 473–496. [32] R. Cochran, R. Horstman, Y. Sun, A. Emery, Benchmark solution for a vertical
[9] S. Terhaar, A. Velazquez, J. Arias, M. Sanchez-Sanz, Experimental study on the buoyancy-assisted laminar backward-facing step flow using finite element,
unsteady laminar heat transfer downstream of a backwards facing step, Int. finite volume and finite difference methods, ASME HTD 258 (1993) 37–47.
Commun. Heat Mass Transfer 37 (2010) 457–462.
Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.
Alternative Proxies: