Open access peer-reviewed chapter

The Effects of Hall and Joule Currents and Variable Properties on an Unsteady MHD Laminar Convective Flow Over a Porous Rotating Disk with Viscous Dissipation

By Abdus Sattar and Mohammad Ferdows

Submitted: November 19th 2011Reviewed: July 30th 2012Published: October 24th 2012

DOI: 10.5772/51982

Downloaded: 1289

1. Introduction

Heat transfer from convection in a rotating body is of theoretical as well as practical importance in the thermal analysis of rotating components of various types of mechanical devises. The rotating disk is one of a number of such geometrical configurations of rotating bodies which is of primary interest. Many practical systems can be modeled in terms of disk rotating in an infinite environment or in a housing. The importance of heat transfer from a rotating body can thus be ascertained in cases of many types of machineries, for example computer disk drives(Herrero et al 1994), rotating disk reactors for bio-fuel production and gas or marine turbines(Owen and Rogers 1989).

Heat transfer from a rotating disk by convection has been investigated theoretically by Wagner(1948), Millsaps and Pohlhausen(1952), Kreith and Taylor(1956), Kreith, Taylor and Chong(1959) and Sparrow and Gregg(1959). The theory thus established predicts that in the laminar flow regime heat and mass transfer coefficients are uniform over the entire surface of a rotating disk. Following pioneer treatment of von Karman(1921) for a flow over a rotating disk, an exact solution of complete Navier-Stokes’ and energy equations was obtained by Millsaps and Pohlhausen for laminar convective flow. The rate of heat and mass transfer from a rotating disk at various speeds in an infinite environment in both laminar and turbulent flows were measured by Kreith, Taylor and Chong(1959). On the other hand Popiel and Boguslawski(1975) measured the heat transfer coefficient at a certain location over a disk rotating at different angular speeds.

The applied magnetic field effects on a steady flow due to the rotation of a disk of infinite or finite extend were studied by El-Mistikaway and Attia(1990) and El-Mistikaway et al. (1991). Aboul-Hassan and Attia(1997) also studied steady hydrodynamic flow due to an infinite disk rotating with uniform angular velocity in the presence of an axial magnetic field with Hall current. Attia(1998) separately studied the effects of suction as well as injection in the presence of a magnetic field on the unsteady flow past a rotating porous disk. It was observed by him that strong injection tend to destabilize the laminar boundary layer but when magnetic field works even with strong injection, it stabilizes the boundary layer.

The heat transfer phenomenon along with magneto-hydrodynamic effect on an unsteady incompressible flow due to an infinite rotating disk were studied by Maleque and Sattar(2003) using implicit finite difference scheme of Crank-Nicolson method. Later Maleque and Sattar(2005) investigated numerically the steady three-dimensional MHD free convective laminar incompressible boundary layer flow due to an infinite rotating disk in an axial uniform magnetic field taking into account the Hall current.

In classical treatment of thermal boundary layers, fluid properties such as density, viscosity, and thermal conductivity are assumed to be constant. But experiments indicate that the assumption of constant fluid property only makes sense if temperature does not change rapidly as far as application is concerned. To predict the flow behavior accurately, it may be necessary to take into account these properties as variables. It is of course known that these physical properties may change significantly with the change of temperature of the flow. Zakerullah and Ackroyd(1979) taking into account the variable properties analyzed the laminar natural convection boundary layer flow on a horizontal circular disk. Herwig(1985) analyzed the influence of variable properties on a laminar fully developed pipe flow with constant heat flux across the wall. He showed how the exponents in the property ratio method depend on the fluid properties. Herwig and Wikeren(1986) made a similar analysis in case of a wedge flow. In case of a fully developed flow in a concentric annuli, the effects of the variable properties have been investigated by Herwig and Klamp(1988). Maleque and Sattar(2002), however, studied the effects of variable viscosity and Hall current on an unsteady MHD laminar convective flow due to a rotating disk. Similar unsteady hydromagnetic flow due to an infinite rotating disk was studied by Attia(2006) takig into account the temperature dependent viscosity in a porous medium with Hall and ion-slip currents. The effects of variable properties(density, viscosity and thermal conductivity) on the steady laminar convective flow due to a rotating disk were shown by Maleque and Sattar(2005a) while Maleque and Sattar(2005b) further investigated the same problem in presence of Hall current. Osalusi and Sibanda(2006) revisited the problem of Maleque and Sattar(2005b), considering magnetic effect. Osalusi et al.(2008), however, considered an unsteady MHD flow over a porous rotating disk with variable properties in the presence of Hall and Ion-slip currents. Rahman(2010) recently made a similar study on the slip-flow with variable properties due to a porous rotating disk.

Most of the above studies were in cases of steady flows accept few. The reason is that the theoretical treatment of unsteady problems is a difficult task. However, one can rely on the sophisticated numerical tools such as finite difference or finite element methods to solve the unsteady problems but the solutions such obtained are non-similar. Problem of course arises when one tries to obtain similarity solutions of an unsteady flow. A similarity technique for unsteady boundary layer problems was thus fathered by Sattar and Hossain(1992), which has been incorporated here to investigate the effects of Hall and Joule currents on an unsteady MHD laminar convective flow due to a porous rotating disk with viscous dissipation.

2. Nomenclature

a, b, c arbitrary exponents

B0magnetic flux density

Fdimensionless radial velocity

Gdimensionless vertical velocity

Hdimensionless tangential velocity

JhJoule heating parameter

kthermal conductivity

kuniform condition of thermal conductivity

mHall current

Mmagnetic parameter

ppressure of the fluid

puniform condition of pressure

PrPrandtl number

NuNusselt number

q velocity vector

qwrate of heat transfer

rradial axis

Rexmagnetic Reynolds number

Rerotational Reynolds number

Ttemperature of the fluid

Twuniform surface temperature

Tfree stream temperature

(u,v,w)velocity components along (r,φ,z)coordinates

U0mean velocity of the fluid

wwuniform suction/injection velocity

zvertical axis

2.1. Greek letters

δa time dependent length scale

φtangential axis

γrelative temperature difference parameter

λa constant

θdimensionless temperature

σelectrical conductivity

ηsimilarity variable

τr,τtradial and tangential stresses

υkinematic coefficient of viscosity

υuniform condition of kinematic coefficient of viscosity

μviscosity

μuniform condition of viscosity

ρdensity of the fluid

ρuniform condition of density

Ωangular velocity

3. Physical model

Let us consider a disk which is placed at z = 0 in a cylindrical polar coordinate system (r,φ,z)where zis the vertical axis and rand φare the radial and tangential axes respectively. The disk is assumed to rotate with an angular velocity Ωand the fluid occupies the regionz>0. Let the components of the flow velocity q=(u,v,w)be in the directions of increasing (r,φ,z)respectively. Let pbe the pressure, ρbe the density and Tbe the temperature of the fluid while the surface of the rotating disk is maintained at a uniform temperatureTw. For away from the wall, free stream is kept at a constant temperature Tand at a constant pressurep. The fluid is assumed to be Newtonian, viscous and electrically conducting. An external magnetic field is applied in the z-direction having a constant magnetic flux density B0which is assumed unchanged by taking small magnetic Reynolds number(Rex1). The electron-atom collision frequency is assumed to be relatively high, so that the Hall effect is assumed to exist. Geometry of the physical model is shown below.

The fluid properties viscosity(μ), thermal conductivity (k)and the density (ρ)are taken as functions of temperature alone and obey the following laws [Jayaraj (1995)]

μ=μ[TT]a,k=k[TT]b,ρ=ρ[TT]cE2

where a, b and c are arbitrary exponents and μ,k,ρare the uniform conditions of viscosity, thermal conductivity and the density. For the present analysis the fluid considered in the flue gas. For flue gases the values of the exponent are a = 0.7, b = 0.83 and c = -1.0 (ideal gas).

Based on the above features, the Navier Stokes equations and Energy equation, which are the governing equations of the problem, due to unsteady axially symmetric, compressible MHD laminar flow of a homogenous fluid take the following form:

t(ρr)+r(ρru)+z(ρrw)=0E3
ρ(ut+uurv2r+wuz)=pr+r(μur)+r(μur)+z(μuz)σB021+m2(umv)E4
ρ(vt+uvruvr+wvz)=r(μvr)+r(μvr)+z(μvz)σB021+m2(v+mu)E5
ρ(wt+uwr+wzz)=pz+r(μwr)+1rr(μw)+z(μwz)E6
Tt+uTr+wTz=r(kTr)+krTr+z(kTz)+σB021+m2(u2+v2)+E7
+μ[(uz)2+(vz)2]E8

Scheme 1.

Geometry of the physical model

In the above equations (2)-(6), mrepresents the Hall current and in equation (6) the last two terms respectively represent magnetic and viscous dissipation terms.

The appropriate boundary conditions of the flow induced by the infinite disk(z=0)which is started impulsively into steady rotation with constant angular velocity Ωand a uniform suction/injection velocity wwthrough the disk are given by

u=0, v=Ωr, w=ww, T=Twat z=0}

u0, v0, TT, ppas

zE9
.

4. Similarity transformations

In order to tackle the unsteady character of the motion unlike other approaches for example that of Chamkha & Ahmed(2011), a new similarity parameter taken as a function of time is introduced asδ=δ(t). Here δis a time dependent length scale and is a new parameter that has been fathered by Sattar & Hossain(1992).

Hence to obtain similarity solutions of the above governing equations the following similarity transformations which are little deviated from the usual von-Karman transformations are introduced in terms of the similarity parameter δ:

η=zδ,
u=rΩF(η),v=rΩH(η),w=rΩG(η),T=T+ΔTθ(η),p=p+2μΩP(η)E10

where Ωis a constant angular velocity and ΔT=TwTand Twis the temperature of the disk wall.

Following the laws in (1) the unsteady governing partial differential equations (2)-(6) are then transformed respectively to the following set of dimensionless nonlinear ordinary differential equations through the introduction of the transformations in (8):

cγ(1+γθ)1[ReHθ'δυdδdtηθ']+2ReF+ReH'=0E11
(1+γθ)ca[Re(F2G2+HF')δυdδdtηF']=γa(1+γθ)1F'θ'+F''E12
M1+m2(FmG)(1+γθ)aE13
(1+γθ)ca[Re(2FG+HG')δυdδdtηG']=γa(1+γθ)1G'θ'+G''
E14
M1+m2(G+mF)(1+γθ)aE15
δυdδdt(1+γθ)ca[HH']+P'=γa(1+γθ)1H'θ'+H''E16
Pr(1+γθ)ca[ReHθ'δυdδdtηθ']=θ''+γb(1+γθ)1θ2+
M1+m2JhPr(G2+F2)+E17
JhPr(1+γθ)a(F'2+G'2)E18

where M=σB02δ2μis the magnetic parameter, Pr=ρυcpkis the Prandtl number, Re=Ωδ2υis the rotational Reynolds number, Jh=r2Ω2cpΔTis the Joule heating parameter and γ=ΔTTis the relative temperature difference parameter which is positive for heated surface and negative for cooled surface and zero for uniform properties.

The equations (10) to (14) are similar in time accept for the term (δυdδdt)where tappears explicitly. Thus the similarity conditions requires that(δυdδdt)must be a constant. Hence following the work of Sattar & Hossain(1992) one can try a class of solutions of equations (10)-(14) by assuming

δυdδdt=λ(a constant).

Thus introducing (14) with the conditions that δ=0whent=0, one obtain

δ=2λυtE19

It thus appears from (15) that the length scale δis consistent with the usual length scale considered for various non-steady flows(Schlichting,1958). Since δis a scaling factor as well as a similarity parameter, any value of λin equation (15) would not change the nature of the solutions except that the scale would be different.

Now making a realistic choice of λto be equal to 2 in equation (15), equations (9) to (13) finally become

cγ(1+γθ)1[ReHθ'2ηθ']+2ReF+ReH'=0E20
(1+γθ)ca[Re(F2G2+HF')2ηF']=γa(1+γθ)1F'θ'+F''M1+m2(FmG)(1+γθ)aE21
(1+γθ)ca[Re(2FG+HG')2ηG']=γa(1+γθ)1G'θ'+G''M1+m2(G+mF)(1+γθ)aE22
2(1+γθ)ca
[HH']+P='γa(1+γθ)1H'θ'+H''E23
Pr((1+γθ)cb[ReHθ'2ηθ']=θ''+γb(1+γθ)1θ'2+M1+m2JhPr(G2+F2)E24
++
JhPr(1+γθ)a(F'2+G'2)E25

With reference to the transformations (8), the boundary conditions (7) transform to

F(0)=0,G(0)=0,H(0)=WS,θ(0)=1F()=0,G()=0,p()=0,θ()=0E26

where, WS=wwδΩrepresents a uniform suction (WS0) or injection (WS0)at the surface.

The quantities which are of physical interest relevant to our problem are the local skin-friction coefficients (radial and tangential) and the local Nusselt number.

Now since the radial (surface) and tangential stresses are respectively given by

τr=[μ(uz+wr)]z=0and

τt=[μ(vz+1rwφ)]z=0E27
,

the dimensionless radial and tangential skin-friction coefficients are respectively obtained as

U02δυγΩ(1+γθ)a=F(0)E28
U02δυγΩ(1+γθ)ac=G(0)E29

where U0is taken to be a mean velocity of the flow.

Again the rate of heat transfer from the disk surface to the fluid is given by

qw=(kTz)Z=0E30
.

Hence the Nusselt number defined by

Nu=δqwkΔTE31

is obtained as

Nu=θ(0)E32
.

5. Numerical method

The nonlinear coupled ordinary differential equations (16) to (20) with the boundary conditions (21) have been solved numerically applying Natchtsheim-Swigert(1965) iteration technique(for detailed discussion of the method see Maleque and Sattar(2002)) along with sixth-order Runge-Kutta integration scheme. A step size of Δη=.01was selected to be satisfactory for a convergence criteria of107. The value of ηwas found to each iteration loop by the statementη=η+Δη. The maximum value of ηwas determined when the value of the unknown boundary conditions at η=0does not change to successful loop with an error less than107.

6. Steady case

When the flow is steady, δis no longer a function of time rather can be considered to be a characteristic length scale such asL. Thus in equations (9) to (13) we can take

dδdt=dLdt=0.E33

Thus putting dδdt=0in equations (9) to (13) we obtain the following equations:

H'+2F+cγ(1+γθ)1Hθ=0E34
F''+aγ(1+γθ)1F'θ'[RL(F2G2+HF')](1+γθ)caM1+m2(FmG)(1+γθ)a=0E35
G''+aγ(1+γθ)1G'θ'[RL(2FG+HG')](1+γθ)caM1+m2(G+mF)(1+γθ)a=0E36
H'+γa(1+γθ)1H'θ'P'=0E37
θ+''bγ(1+γθ)1θ2PrRL(1+γθ)ca+M1+m2JhPr(G2+F2)+JhPr(1+γθ)a(F'2+G'2)=0E38
.

In the above equations

RL=ΩL2υE39
.

The above equations exactly correspond to those of Maleque and Sattar(2005a), therefore the solutions to the above equations have not been explored here for brevity. However, numerical values of the radial, tangential and rate of heat transfer coefficients for three different values of the relative temperature difference parameter γis presented in Table-1 and compared with those of Maleque and Sattar(2005a).

γF(0)Maleque and Sattar(2005a)G(0)Maleque and Sattar(2005a)θ'(0)Maleque and Sattar(2005a)
-0.50.4570.4682.0842.0860.8680.867
0.00.3710.3721.2341.2330.7210.720
0.50.1670.1680.6240.6220.5590.559

Table 1.

Values ofF(0),G(0)and θ'(0)for various values of γwhen RL=1, Jh=0, λ=0,m=M=0.1.

Although the comparison should show exact values, due to the differences in the present code and that of Maleque and Sattar(2005a) there are differences in the calculated values ofF'(0),

G'(0), andθ'(0). Percentage wise differences have therefore been calculated and found to be maximum 2.35% and minimum 0.0% w.r.t three decimal places of the calculated values which shows a good agreement between our calculated results and that of Maleque and Sattar(2005a).

7. Unsteady solutions

As a result of the numerical calculations the radial, tangential and axial velocity profiles and temperature profiles are displayed in Figures 1-24 for various values of the governing parameters. In the analysis the fluid considered is flue gas for which Pr=0.64and the values of the exponents a,band care taken to be asa=0.7, b=0.83andc=1.0.

Variation of the radial, tangential and axial velocity profiles and temperature profiles under the influence of γare shown in Figures 1-4. From Fig.1 it can be seen that due to the existence of the centrifugal force the radial velocity attains maximum values0.23γ=0.5, 0.13γ=0.0and 0.10γ=0.5close to the surface of the disk(approximately atη=0.75). Thus at η=0.75the boundary layer thickness of the surface of the disk is reduced due to the increase inγ. From Fig-2, it is seen that the tangential velocity profile decreases in the intervalη[0,0.75], but for η0.75this situation breaks down and the consequence is that the tangential velocity increases with the increase ofγ. From Fig-3, it is seen that close to the disk surfaceγhas a tendency to reduce the motion and induce more flow far from the boundary indicating that there is a separation flow and is detected at η=0.75(approximately). On the other hand, from Fig-4 it is observed that there is a small rate of decrease of the temperature close to the surface and then the temperature distributions starts increasing with increasingγ. This means that the thermal boundary layer induces more flow far from the surface of the disk.

Figure 2.

The dimensional radial velocity profiles against η for different values of γ

Figure 3.

The dimensional tangential velocity profiles against η for different values of γ

Figure 4.

The dimensional axial velocity profiles against η for different values of γ

Figure 5.

The dimensional temperature profiles against η for different values of γ

Figures 5-8, present the effects of uniform suction as well as injection(Ws) on the flow, which characterizes the flow behavior. It is evident from these figures that the boundary layer is increasingly blown away from the disk to form an interlayer between the injection and the outer flow regions. Also all flow profiles increase monotonically with increasing Ws. From Fig.5 it is apparent that radial velocity in this case attains maximum values approximately0.19|Ws=4,η=1,0.11|Ws=2,η=0.75,0.05|Ws=0,η=0.5,0.03|Ws=2,η=0.25. This implies that the momentum boundary layer thickness decreases due to an increase in the values of Ws in different regions likeη=1,0.75,0.5,0.25. Thus reduced flows are observed for increase in injection (Ws<0) and induced flows are observed for increase in suction (Ws>0) in the total flow behavior.

Figure 6.

Fig-5: The dimensional radial velocity profiles against η for different values ofWs.

The effects of the magnetic parameter Mon the radial, tangential and axial velocities and temperature profiles are depicted in Figures 9-12. We see that the radial velocity increases with the increase inM. It can also be seen that at each value of M there exists local maxima

in radial velocity distributions. The maximum values of velocities are approximately0.13|M=1,η=1,0.11|M=0.5,η=1,0.08|M=0,η=1. Thus we can say that at η=1the boundary layer thickness increases. From Fig.-10, we see that tangential velocity decreases with increasingM. On the other hand, axial velocity decreases with increasing Mand seen that at each value of Mthere exists local minima in the profiles. The effect of Mis found to be almost not significant for temperature distributions.

Figure 7.

The dimensional tangential velocity profiles against η for different values ofWs.

Figure 8.

The dimensional axial velocity profiles against η for different values ofWs.

Figure 9.

The dimensional temperature profiles against η for different values ofWs.

Figure 10.

The dimensional radial velocity profiles against η for different values of M.

Figure 11.

The dimensional tangential velocity profiles against η for different values of M.

Figure 12.

The dimensional axial velocity profiles against η for different values of M.

Figure 13.

The dimensional temperature profiles against η for different values ofM.

The effects of varying the Hall current parameter mon the flow distributions are shown in Figures 13-16. It can be seen that the radial velocity distributions increases with increasingm. This is due to the fact that for large values ofm, the term 11+m2is very small and hence the resistive effect of the magnetic field is diminished. This phenomenon for small and large values of mhas been effectively explained by Hassan and Attia(1997). The maximum velocities are approximately0.12|m=50,η=1,0.10|m=5,η=1,0.9|m=2,η=1,0.05|m=0,η=1. Thus at η=1the boundary layer thickness increases due to the increase inm. Tangential distribution increases with the increase ofm. On the other hand, the axial velocity decrease with increasing mand shows local minima indicating that the boundary layer thickness decreases. Like the magnetic parameter Mthe effect of mis also found to be not much significant in case of temperature distributions.

Figure 14.

The dimensional radial velocity profiles against η for different values ofm.

Figure 15.

The dimensional tangential velocity profiles against η for different values ofm.

Figure 16.

The dimensional axial velocity profiles against η for different values ofm.

Figure 17.

The dimensional temperature profiles against η for different values ofm.

Figures 17-20 present the effects of the rotational Reynolds number, Re, on the flow behavior. It is seen that Reynolds number accelerates the fluid motion in radial and tangential velocity profiles and temperature profiles. However the behavior of the axial velocity profiles decreases with increasingRe.

Figure 18.

The dimensional radial velocity profiles against η for different values ofRe.

Figure 19.

The dimensional tangential velocity profiles against η for different values ofRe.

Figure 20.

The dimensional axial velocity profiles against η for different values ofRe.

Figure 21.

The dimensional temperature profiles against η for different values ofRe.

Figures 21-24 present the effects of the Joule heating parameter, Jh, on the flow behavior. It is seen that the axial velocity profiles and temperature profiles increase with increasing Jhwhile velocity profiles is generally much smaller between 0.1 and 0.5 except for large values ofJh, when it increases above 0.5, which is expectable on physical basis. The radial and tangential velocity profiles have no significant impact.

Figure 22.

The dimensional radial velocity profiles against η for different values ofJh.

Figure 23.

The dimensional tangential velocity profiles against η for different values ofJh.

Figure 24.

The dimensionless axial velocity profiles against η for different values ofJh.

Figure 25.

The dimensional temperature profiles against η for different values ofJh.

8. Concluding remarks

In this study, we have investigated numerically the heat transfer phenomenon along with the effects of variable properties for a 2-D unsteady hydrodynamic flow past a rotating disk taking into account viscous dissipation, Joule and hall currents. Using a new class of similarity transformation close to von-Karman, the governing equations have been transformed into non-linear ordinary differential equations that are locally similar. These equations have been solved using the Nachtsheim-Swigert shooting iteration technique along with a sixth-order Runge-Kutta integration scheme. Based on the resulting solutions the following conclusions can be drawn:

  1. Similarity approach adopted in the analysis has the advantage that one can separately obtain the steady and unsteady solutions.

  2. A comparison of the steady results for the radial and tangential stresses and the rate of heat transfer with those from the available literature leads credence to the numerical code used and hence to the results obtained in the unsteady case.

  3. The relative temperature difference parameter γtaken as the variable properties parameter has marked effects on the radial and axial velocity profiles. Close to the surface of the disk tangential velocities and temperature slow down but shortly after they increase with the increasing values ofγ.

  4. As an influence of the relative temperature difference parameterγ, the thermal boundary layer induces more flow far from the surface of the disk.

  5. Separation of flow was detected in different regimes of the momentum and thermal boundary layers.

  6. Reduced flows have been observed for increase in injection(Ws<0)while induced Flows were observed for increase in suction(Ws>0).

  7. Local maxima and local minima have been observed in the cases of radial and axial velocities forM, mandJh.

  8. Re

© 2012 The Author(s). Licensee IntechOpen. This chapter is distributed under the terms of the Creative Commons Attribution 3.0 License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

How to cite and reference

Link to this chapter Copy to clipboard

Cite this chapter Copy to clipboard

Abdus Sattar and Mohammad Ferdows (October 24th 2012). The Effects of Hall and Joule Currents and Variable Properties on an Unsteady MHD Laminar Convective Flow Over a Porous Rotating Disk with Viscous Dissipation, Heat Transfer Phenomena and Applications, Salim N. Kazi, IntechOpen, DOI: 10.5772/51982. Available from:

chapter statistics

1289total chapter downloads

More statistics for editors and authors

Login to your personal dashboard for more detailed statistics on your publications.

Access personal reporting

Related Content

This Book

Next chapter

Thermal Interaction Between a Human Body and a Vehicle Cabin

By Dragan Ružić and Ferenc Časnji

Related Book

First chapter

Measurements of Local Heat Flux and Water-Side Heat Transfer Coefficient in Water Wall Tubes

By Jan Taler and Dawid Taler

We are IntechOpen, the world's leading publisher of Open Access books. Built by scientists, for scientists. Our readership spans scientists, professors, researchers, librarians, and students, as well as business professionals. We share our knowledge and peer-reveiwed research papers with libraries, scientific and engineering societies, and also work with corporate R&D departments and government entities.

More About Us