Notes On Heat Transfer
Uploaded by
adsharma29042003Notes On Heat Transfer
Uploaded by
adsharma29042003SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
Heat Transfer
Class notes prepared for our students
This class note is prepared using the text and reference books on heat transfer
For more details please go through text books
Chapter Content Page no.
1 Introduction to Heat Transfer 01
2 One-Dimensional Heat Diffusion 10
3 Heat transfer from extended surfaces-Fin 24
4 Transient Heat Conduction 37
5 Two-Dimensional Heat Diffusion 43
6 Numerical Heat Transfer 57
7 Heat Exchangers 73
8 Introduction to Convection --
Chapter 8 to be send latter
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
0
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
CHAPTER - 1
Introduction to Heat Transfer
In this Chapter, the following terms have been discussed in brief.
Energy, Heat, Work, three modes of heat transfer (Conduction, Convection & Radiation), thermal
diffusivity, energy conservation expression (first law of thermodynamics) & Taylor’s series.
Energy
Internal Energy Energy in Transit
(Stored energy) (The energy that gets
E = mcpT transferred from one body to
(or energy possessed by an another body)
object due to its mass &
temperature
Heat energy (Q) Work energy (W)
1.1 Heat: Energy transfer due to temperature difference only is called heat energy.
It is denoted by Q (Joule).
A T1 > T2
B
m2
m1
cp2
cp1 Q=40J
T2
T1
The amount of energy transfer Q = 40 J (say) from A to B is due to temperature difference. This
energy is named as Heat or heat energy.
1.2 Work: It is the energy transfer due to any other reason other than temperature difference. It
may be due to electric p.d., Pressure difference, Gravitational P.d. etc.
1.2.1 Heat transfer: It is also called rate of heat transfer (q)
Q
q (J/s = Watt),
dt
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
1
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
where Q = Infinitesimally small amount of heat transfer during infinitesimally small time interval.
q W q
q (heat flux) 2 , q (W/ m 3 ) = volumetric heat generation rate.
A m vol
Q dq dq
Q = Joule and q = (Watt) , q (Watt / m 2 ) and q (Watt / m 3 )
dt dA dV
1.3 Three – modes of heat transfer: Conduction, Convection, Radiation
1.3.1 Conduction: Heat transfer in a medium where medium does not have motion as a bulk. But
heat transfer takes place due to molecular diffusion or lattice vibration. Metals have higher rate of
heat transfer due to availability of large number of free electrons. Conduction heat transfer is
mainly in a solid medium.
dT
Conduction: Fourier Law of Heat Transfer; q x kA c
dx
(Conductive heat transfer along x x (direction)
dT
– direction) and the heat flux qx k
dx qx
The negative sign represents that heat is flowing O (T1 T2 T3……)
along decreasing direction of temperature. T1 T2 T3 T4
k = thermal conductivity, kAg = 418.7 W/m.K, kcu = 387.6 W/mK, kiron (pure) = 62.3 W/mK.
The thermal conductivity of a material is a measure of the ability of the
material to conduct heat. It can be defined as the rate of heat transfer through a unit
thickness of the material per unit area per unit temperature difference.
In conduction, when temperature increases, molecules/lattice start to vibrate more due to
thermal energy. This increase in vibration decreases mean free path of molecules thereby its
free electrons. Hence the thermal conductivity of metallic conductors decreases with
temperature.
In semiconductors also the mean free path decreases with increases in temperature but
valence electrons are converted to free electrons. Hence thermal conductivity increases with
temperature.
In gases the thermal conductivity increases with increase in temperature due to increased
number of molecular collisions.
In case of liquids thermal conductivity is proportional to density as fewer particles leads to
fewer collisions. Since density decreases with increasing temperature the net effect is
decreasing thermal conductivity.
1.3.2 Convection: Heat transfer in a medium where there is a bulk movement of medium particles.
Convection may be further classified as free convection (or natural convection) and forced
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
2
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
convection. In free convection, the motion of the medium is due to buoyancy force which is due to
density difference caused by thermal expansion.
In force convection, the motion of the medium is caused by external agents like fan, blower &
pump etc.
Convection: Newton’s Law of Cooling; q = hAs[Ts - T], (Convective heat
transfer rate over surface area As which is at surface temperature (Ts) and surrounding
temperature (T)
W
q = h[Ts - T], where h = heat transfer coefficient 2
m K .
Convective heat transfer coefficient (h) is not a property of the material of
the medium. It depends on many factors like thermo-physical properties (, kfluid, cp, fluid)
flow property (v = velocity) and geometrical parameters (length or diameter).
The convective heat transfer coefficient for forced convection is more than that of free
convection as the fluid velocity and hence the rate of fluid flows is more in case of forced
convections. The convective heat transfer coefficient of liquids is more than that of gases
due to increase density. In phase change process convective heat transfer coefficient is
highest.
Type of Convection h, W/m2 K
Free Convection of gases 2-25
Force convection of liquids 10-1000
Forced convection of gases 25-250
Forced convection of liquids 50-20,000
Boiling and condensation 2500-100000
1.3.3 Radiation: It is the form of energy transfer due to electromagnetic radiation. No medium is
required for radiation energy exchange between two bodies at different temperature.
Radiation: Stefan Boltzmann Law of Radiation; Radiation heat exchange
between a small surface (A) kept in a large enclosure is
q A TS4 Tsur
4
h r A TS Tsur ,
where ,
W
hr = TS Tsur TS2 Tsur
2
4 TS3 = radiation heat transfer coefficient 2 ,
m K
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
3
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
W
= 5.67 x 10-8 = Stefan – Boltzmann Constant, and Ts & Tsur are in Kelvin, and
m2K 4
= emissivity of the surface. Emissivity is 1 for black body surface.
k
Thermal diffusivity () = (m 2 / sec)
cP
It measures the ability of a material to conduct thermal energy relative to its ability to store thermal
energy.
Conservation of Energy: Ein Egen Eout Estore [Watt] , also called energy conservation equation
in rate form. It is based on 1st law of thermodynamics.
J
where Egen q vol. & Estore mcP T (Watt),and c P
kg.K
1.3.4 Thermal Resistance: The resistance offered by the medium to heat transfer is called thermal
resistance (Rth). Analogous to electric resistance, thermal resistance is also defined as the ratio of
thermal potential difference (T) to heat current (q)
q = heat
T T1 T2 current T2
Thermal Resistance (Rth) = T1
q q
Rth
T L K L
For conduction Rth =
T kA W
kA
L
T 1 K
h c As T h c As W
For convection Rth =
T 1 K
For radiation Rth = W ,
h r As T h r As
hc = Convective heat transfer coefficient.
hr = Radiative heat transfer coefficient.
1.4 Review of Fundamentals
E = energy (Joule = kg.m2/sec2) - any form of energy
Energy = Force x displacement = mass x acceleration x displacement
= kg x m/sec2 x m = kg m2/sec2
Estore = Einternal energy = m x cP x T
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
4
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
Energy J
So, cP = (specific heat)
mass temperature kg.K
d
E store Estore J / sec Watt
dt
Q = heat energy (Joule)
Q
q= (J/s = Watt) : heat transfer rate
dt
q Watt
q = (heat flux)
dA m 2
q Watt
q (volumetric heat generation rate)
d m3
q q d q x (AC (x))
Watt
q q volume 3 x m3 Watt
m
density () = mass/volume mass (m) = density () x volume (v).
Density of water at 40C is 1000 kg/m3 or 1 gm/cm3.
k
Thermal diffusivity () = (m 2 / sec)
cP
1.5 Assignment 1.1
1. A brick wall 15 cm thick with thermal conductivity 1.2 W/mK is maintained at 300C at one
face and 2300C at the other face. Determine q across 4m2.
2. The heat flow rate across an insulating material of thickness 3 cm with thermal conductivity
0.1 W/mK is 250 W/m2. If the hot surface temperature is 1750C, what is the temperature of
the cold surface.
3. Water at a mean temperature of 200C flows over a flat plate at 800C. If the heat transfer
coefficient is 200 W/m2K, determine heat transfer rate per unit area over time of 5 hours.
4. A 25 cm diameter sphere at 1200C is suspended in air at 200C. If the natural convection heat
transfer between the sphere and the air is 15 W/m2. 0C, determine the rate of heat loss from
the sphere.
5. A small sphere of 10 cm diameter is suspended inside a large evacuated chamber whose
walls are kept at 300 K. If the surface of the sphere has emissivity = 0.8 and is maintained
at 500 K, determine q.
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
5
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
6. Two large parallel plates, one at a uniform temperature 500 K and the other at 1000 K, are
separated by a nonparticipating gas. Assuming that the surfaces of the plates are perfect
emitters and that the convection is negligible, determine the rate of heat exchange between
the surfaces per square meter.
7. Write the unit of thermal conductivity (k), heat transfer coefficient (h), thermal diffusivity
(), Specific heat (cp), Stafan-Boltzmann constant (), heat flux ( q ), volumetric heat
generation rate ( q ) and thermal resistance (Rth)
8. How heat and work are different.
9. The thermal diffusivity is the controlling transport property for transient conduction.
Using appropriate values of k, cP and for aluminum at 300 K & at 700 K, calculate .
Data: At 300 K At 700 K
= 2702 kg/m3 = 2702 kg/m3
cp = 903 J/kg.K cp = 1090 J/kg.K
k = 237 W/mK k = 225 W/mK
10. The temperature distribution across a wall 1 m thick at a certain instant of time is given by
T(x) = 900 – 300 x – 50x2. A uniform heat generation q 1000W / m is present in the wall
3
of area 10 m2 having the properties = 1600 kg/m3, k = 40 W/mK, & cp = 4000 J/kg.K.
(a) Determine the rate of heat transfer entering the wall (x = 0) and leaving the wall (x = 1 m)
(b) Determine the rate of change of energy store in wall.
Ans: (1) 6.4 kW, (2) 1000C (3) 216 MJ (4) 294.5 W (5) 77.52 W.
(6) 53.16 kW/m2 (9) 300 K = 97.1 x 10-6m2/sec. 700 K = 76 x 10-6m2/sec.
(10.a) 120 kW, 160 kW, (10.b) -30 kW
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
6
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
1.6 Taylor Series
dx d 2Tx ........
2
d
Taylor Series: Tx dx Tx (dx) (Tx )
dx 2! dx 2
dx d 2 q dx d3 q ........
2 3
d
q x dx q x (dx) (q x ) x x
dx 2! dx 2 3! dx 3
dx d 2 dx d3
2 3
d
q x dx q x (dx) (q x ) qx q x ........
dx 2! dx 2 3! dx 3
dx d 2Tx dx d3Tx
2 3
d
Tx dx Tx (dx) (Tx ) ........
dx 2! dx 2 3! dx 3
Tx-dx Tx Tx+dx Tx+2dx
(x-dx)
x=0
x
x +dx
x + 2dx
For small value of dx, the higher powers of dx may be neglected, therefore, (dx) 2 (dx)3 0,
(dx)4 0 ……
So, the Taylors series is approximated as
d
Tx dx Tx (dx) (Tx ) 0 (dx) 2
dx
d
Tx dx Tx (dx) (Tx ) 0 (dx) 2
dx
d
q x dx q x (dx) (q x ) 0 (dx) 2
dx
dT Tx x Tx
0 x
1
Solved example 1. Prove that
dx x x
Solution:
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
7
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
From Taylor’s Series:
x d 2 T x d3 T ........
2 3
d
Tx x Tx (x) (Tx ) x x
dx 2! dx 2 3! dx 3
Ignoring the terms starting with (x)2 and onwards.
d
(Tx ) 0 x
2
Tx x Tx (x)
dx
d
(Tx ) 0 x
2
Tx x Tx (x)
dx
Dividing the equation by (x),
Tx x Tx d
(Tx ) 0 x
1
(x) dx
d Tx Tx x Tx
0 x
1
Or Proved
dx (x)
The above equation is called 1st order accurate forward finite difference equation for first order
derivative.
The above equation may be written as,
d Tx Ti 1 Ti
0 x
1
dx i (x)
Ti-2 Ti-1 Ti Ti+1 Ti+2
x x
1.7Assignment- 1.2
Tx
Tx - x
dT T T
x x x 0 x
1
1. Prove that
dx x (x) i-1 i
x
(First order accurate backward finite
x=0 Ti - 1 Ti
difference equation of first order derivative)
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
8
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
d 2 d
Hints: Tx x Tx Tx (x) (Tx ) x (Tx )........
dx dx 2
d T Ti Ti 1
0 x
1
The required prove may be written as
dx i (x)
dT Txx Tx x
0 x
2
2. Prove that
dx x 2(x)
d T Ti1 Ti 1
0 x
2
Or
dx i 2(x)
(2nd order accurate central finite difference equation of first order derivative)
Tx - x Tx Tx + x
d 2 T Ti 1 Ti 1 2Ti
0 x x x
2
3. Prove that
dx i2
(x) 2
i-1 i i+1
x=0
Ti - 1 Ti Ti + 1
1.8 Conclusions
In this Chapter we have discussed difference between heat transfer and work transfer, different
modes of heat transfer and Taylor series expansion. Emphasis has been given more on
fundamentals as well as units of different terms. In the next chapter the heat diffusion equation in
Cartesian, cylindrical and spherical coordinate system will be discussed. Also the concept of critical
insulation and overall heat transfer coefficient will be discussed.
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
9
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
CHAPTER – 2
One Dimensional Heat Diffusion
In this Chapter, the following terms have been discussed in brief.
Heat diffusion equation for Cartesian, cylindrical and spherical coordinates, different boundary and
initial conditions, more about thermal resistance, critical insulation & overall heat transfer
coefficient.
2. Heat diffusion equation or heat equation
2.1 Heat Diffusion Equation in Cartesian Coordinates
Statement: Write an energy balance equation for a differential element; derive the one-dimensional,
time dependent heat conduction equation with internal energy generation and having constant
thermal conductivity in the rectangular coordinate system, along x-direction.
2T q 1 T
i.e.,
x 2 k t
Sol: Consider a rod having uniform area of cross-section. Ends of the rod are maintained at
different temperatures. The thermal conductivity, density, specific heat & volumetric heat
generation rate are specified as k, , cp &. q
AC dx dV = AC.dx
Figure shows a differential volume element
having volume dV = Ac. dx and mass (m) = dV qx qx+dx
= Ac dx.
AC
x Differential
dTx
qx = kA C (heat conducted into the element volume element
dx (x+dx)
per sec.)
dq x
qx + dx = qx + (dx) 0(dx) 2 (heat conducted out of the element, by using Taylors series)
dx
Energy balance for the differential volume element:
Ein Egen Eout Estore
d d
q x (qdV) q x (dx) (q x ) mc P Tx
dx dt
d d
qA c dx dx (q x ) (A C dx c P Tx )
dx dt
Substituting qx in the above equation.
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
10
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
d dTx dT
qA c dx dx k Ac A C dx c P x
dx dx dt
d 2Tx dT
q k 2
c P x
dx dt
d 2Tx q cP dTx
Or
dx 2 k k dt
Since T = T(x,t), so its differential is partial. Therefore
2T q 1 T
x 2 k t
k
where T = T(x, t), thermal diffusivity = (m 2 / s)
c P
This is the one-dimensional, time-dependent heat conduction equation in rectangular
coordinates. It is a 2nd order partial non homogeneous differential equation. The solution of this
equation needs two boundary conditions and one initial condition.
Similarly, for 3-dimentional coordinate system,
2T 2T 2T q 1 T
, where, T = T(x, y, z, t)
x 2 y2 z 2 k t
If the problem is in steady state i.e., T T(t) and T T(x) only,
d 2 T q
0 , where T = T(x) only, hence ordinary differential equation.
dx 2 k
d 2T
If the problem has no internal energy generation term ( q = 0), then 0.
dx 2
The solution of this equation is: T = C1 x + C2. To solve for C1 and C2, two boundary conditions are
required.
2.2 Heat diffusion equation in cylindrical Coordinate
Statement: Write an energy balance equation for a differential volume element (Cylindrical
domain); derive the one-dimensional, time dependent heat conduction equation with internal energy
generation and having constant thermal conductivity in the cylindrical coordinate system along r-
direction.
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
11
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
1 T q 1 T
i.e., r
r r r k t
qr+dr
qr
dr
l area of this face
= (2r)dr
(Radial heat conduction)
AC = 2rl [This area is perpendicular to the heat transfer qr]
dV = (2r) (dr) l = Volume of the cylindrical differential element.
Energy conservation principle:
Ei E g E 0 E s
d
q r qdV q r dr mc p T
dt
d d
q r qdV q r (dr) q r mc p T ,
dr dt
qdV (dr)
d
dr
d
q r dVc p T ,
dt
dT
Substituting qr = kA c in the above equation,
dr
where, qr represents heat transfer by conduction in radial direction.
d dT dT
qdV (dr) kA c c p dV ,
dr dr dt
d dT dT
k(dr ) 2 rl q2 rl drl c p d2 rl drl ,
dr dr dt
T T
k r qr drl c p r , [ T = T(r,t)]
r r t
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
12
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
Dividing by (k.r)
1 T q 1 T
r* ,
r r r k t
* this r cannot be taken out as it is a variable.
2.3 Heat diffusion equation in spherical coordinates
Statement: Write an energy balance equation for a differential volume element (spherical domain),
derive the one-dimensional, time dependent heat conduction equation with internal energy
generation and having constant thermal conductivity in the spherical coordinate system along r-
direction.
1 2 T q 1 T
i.e., r
r 2 r r k t
Solution : Consider a thin spherical shell of radium r and thickness dr as show. The radial heat
transfer about radius r is given by
dT
qr = -kAc
dr
d
and qr + dr = qr + (dr) (qr) [Taylor’s series with 2nd order accurate]
dr
Applying energy conservation principle to the differential control volume
qr + dr
Ei E g E 0 E s dr
r
d
q r qdV q r dr mc p T
dt dV=4r2dr qr
a
dq dT Ac=4r2
q r qdV q r (dr) r mc p
dr dt
b
d
q r qdV q r dr mc p T
dt
qr + dr
dq dT
q r qdV q r (dr) r mc p
dr dt
qdV (dr)
d
dr
q r dVc p
dT
dt
qr
d dT dT
qdV (dr) kA c dVc p m = dV
dr dr dt
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
13
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
d 2 dT dT
q4 r 2dr (dr ) k4 r 4 r dr c p
2
dr dr dt
d 2 dT dT
q4 r 2dr (dr ) k4 r 4 r dr c p
2
dr dr dt
d 2 dT 2 dT
r 2 q k r c p r
dr dr dt
2 T 2 T
or r 2 q k r r c p [ T = T(r,t)]
r r t
Dividing the equation by (k.r2),
qr
r q k 2 T r cp T
2 2
r
2
Ac = 4r , it is perpendicular to heat
kr 2 kr 2 r r kr 2 t transfer (qr) direction
q 1 2 T cp T
r
k r 2 r r k t
1 2 T q 1 T k
or r , where (m2 / s)
r r r k t
2
cp
This equation is called 1-D, heat conduction for spherical geometry. It is a 2nd order partial non-
homogeneous and linear differential equation with having constant coefficient.
2.4 Boundary Conditions for Heat Diffusion Equations
For heat diffusion equation. Ts
i) Constant surface temperature (Dirichlet condition) T(x)
T x 0 Ts x
ii) Constant surface heat flux (Neumann condition)
(a) Finite heat Flux
q x
dT
k qx T(x)
dx x 0 x
(b) Adiabatic or insulated surface
dT
0 T(x)
dx x 0 x
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
14
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
iii) Convection surface condition (Robins)
T0
T h
dT
k h T T x 0 T(x)
dx x 0
x
2.5 Assignment 2.1
1. Consider a slab of thickness L as illustrated in figure. The boundary
surfaces at x = 0 and x = L are maintained at constant, but different,
temperatures T1 and T2, respectively. There is no energy generation
in the solid, and the thermal conductivity k is constant. Develop an
expression for the temperature distribution T(x) in the slab. Develop
an expression for heat flow Q through an area A of the slab.
2. Consider a slab of thickness L and constant thermal conductivity k in which energy is
generated at a constant rate of q W/m3. The boundary surface x = 0 is insulated (adiabatic)
and that at x = L dissipates heat by convection with a heat transfer coefficient h into a fluid
at a temperature T.
Develop expressions for the temperature T(x) and the heat flux q(x) in the slab.
Calculate the temperatures at the surfaces x = 0 and x = L under the following conditions: L
= 1 cm, k = 20 W/(m. oC), q = 8 x 107 W/m3, h = 4000 W/(m2.oC), and T = 100oC.
Ans: T(0) = 500oC, T(L) = 300oC
3. Consider a solid cylinder of radius r = b in which energy is generated at a constant rate of g0
W/m3, while the boundary surface at r = b is maintained at a constant temperature T2.
Develop an expression for the one-dimensional, radial, steady-state temperature distribution
T(r) and the heat flux q(r).
Calculate the center temperature T(0) and the heat flux at the boundary surface r = b for b =
1 cm, q = 2 x 108 W/m3, k = 20 W/m.oC), T2 = 100oC.
Ans.: T(0) = 350oC, q(b) = 106 W/m2.
4. The inner surface at r = a and the outer surface at r = b of a hollow cylinder are maintained
at uniform temperatures T1 and T2, respectively. The thermal conductivity k of the solid is
constant.
Develop an expression for the one-dimensional, steady-state temperature distribution T(r) in
the cylinder.
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
15
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
Develop an expression for the radial heat flow rate Q through the cylinder over a length H.
Develop an expression for the thermal resistance of a hollow cylinder of length H.
6. The inner surface at r = a and the outer surface at r = b of a hollow sphere are maintained at
uniform temperatures T1 and T2, respectively. The thermal conductivity k of the solid is
constant.
Develop an expression for the one-dimensional, steady-state temperature distribution T(r) in
the sphere.
Develop an expression for the radial heat flow rate q through the hollow sphere.
Thermal p.d. T Kelvin
2.6 Thermal Resistance (Rth) =
heat current (q) Watt
2.6.1 For Conduction T1 T2
T1 T2 kA T1 T2 Ac q
q= C
R th L
L L
So, R th,cond
kA C
q
T1 Rth T2
This formula for thermal resistance conduction is used when AC (area of cross section) is uniform
throughout the length (L) and also k is uniform.
2.6.2 For Convection
T T(Temp. of medium)
Rth conv.
Ts q
Ts (surface temperature)
q = hAs (Ts - T), As- surface area.
R th conv
Ts T Ts T 1
q hA s Ts T hA s
1 (K/w)
R th conv
hA s
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
16
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
2.6.3 For Radiation
A mall object placed in a large enclosure:
Ts q Tsur
Ts = Surface temperature of the small object in Kelvin.
Rth radiation
Ts, As
Tsur = Surroundings temperature in Kelvin
As = Surface area of small object
= emissivity of small object.
q = As Ts4 Tsur
4
Rth radiation =
Ts Tsur Ts Tsur
q As Ts4 Tsur
4
Tsur
Ts Tsur
As Ts Tsur Ts2 Tsur Ts Tsur
= 2
qconvection to medium at T
1
Rth rad. =
As Ts Tsur Ts2 Tsur
2
qrad
If Ts Tsur Tsurface then Ts + Tsur 2Ts and Ts2 Tsur
2
2Ts2
1 1 Ts
Rth rad. = ,
4 Ts As h r As
3
where hr = 4 Ts3 = radiation heat transfer coefficient
W
= 5.67 x 10-8
m2K 4
Combined (convection & radiation) heat transfer Mechanism
q = qconv + qradi
q = hAs[Ts - T] + As Ts4 Tsur
4
,
where T = ambient temperature of the convective medium.
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
17
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
2.6.4 Thermal Resistance for cases when Ac is variable
2.6.4.1 Thermal Resistance for spherical shell Ac=4r2
dr
Consider a thin elemental spherical shell of radius ‘r’ and
thickness dr. Thermal resistance of this thin shell is
r
dr dr a
d Rth =
kAc K4 r 2
1 1 1 b a
b
1 dr T1
R th
4k a r 4k a b 4abk
2
b
k
Rth =
b a b a where, Ai = 4a2, A0 = 4b2.
4 abk k Ai A o T2 T1 > T2
2.6.4.2 Thermal resistance for cylindrical shell
Resistance of the thin cylinder shell is, (T1 > T2)
dr dr
dRth = , Ac = surface area of cylinder shell = 2rl
K Ac 2 rlk
l
q (radial
dr heat flow)
r
a
T1
b
k
T2
b b
dr 1 dr 1
R th ln b / a
a
k2rl k2l a r 2l
ln b / a
R th
2lk
L
Note: Rth = , Ac = area of cross section which is uniform over length L.
kA c
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
18
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
b a
Rth = ; for spherical shell.
4abk
ln b / a
Rth = ; for cylindrical shell.
2lk
2.7 Critical Insulation
The radius of insulation at which the total thermal resistance (sum of conduction resistance and
convection resistance) is minimum and hence the rate of heat transfer becomes maximum is called
critical radius of insulation.
2.7.1 For cylinder:
Insulation
q (h, (kin) (h,
T) Ti T) r
ri
Ti
ri
kin r
l
Cable
Consider a cable of radius ri which is required to keep at surface temperature Ti. An insulation layer
having thermal conductivity kin is provided over the cable upto radius r as shown. It is required to
find the radius of insulation ® at which net thermal resistance is minimum and hence heat transfer
will be maximum.
Rcond q
Ti (Surface
temp. of cable) T
Rconv
ln r
i , R convection
r 1
R cond
2lk in h2rl
ln r
Rnet =Rcond + Rconv = i
r 1
2lk in h2rl
The heat transfer rate ‘q’ is maximum when Rnet is minimum
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
19
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
d
i.e.; R net 0
dr
r
d ri d 1
ln
0
dr 2lK in dr h2rl
1 1 1 1
0
2lk in r 2lh r 2
2lk in k in
r
2lh h
k in
r is the expression of critical radius of insulation.
h
k in (r ri )
So, rc Rcond=
h 4rri k in
Ti T
2.7.2 For Sphere: Critical radius of insulation over sphere. 1
Rconv=
h4 r 2
r ri
1
Rnet =
4rri k in h4r 2
d R net 2k
For rc, 0 rc in
dr h
2.8 Overall Heat Transfer Coefficient (U)
Consider a hollow cylinder of radii ri & r0 with thermal conductivity k and length l. The
temperatures and heat transfer coefficients for inner fluid are T, i and hi and for outer fluid are
T, 0 and h0 respectively.
r0
ln o
r
, R2 = i
1 1 r
R1 = = h0
hi Ai hi 2 ri l 2lk
ri T, 0
1 1
R3 = T, I, hi
h 0 A 0 h 0 2r0 l
K
1
R1 = Convection resistance of the inner fluid =
hi Ai
q T, 0
R2 = Conduction resistance of the wall of the cylinder. T, I
R1 R2 R3
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
20
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
R3 = Convection resistance of the outer fluid.
ln o
r
i
1 r 1
Rnet = R1 + R2 + R3 =
h o 2ri l 2lk h 0 2r0 l
The net resistance may arbitrarly be defined as only convection type based on either inner surface
or based on outer surface of the cylinder as explained:
1 1
i.e, Rnet =
Ui Ai U0 A0
where Ui = overall heat transfer coefficient based on inner surface area (Ai = 2ril)
U0 = Overall heat transfer coefficient based on outer surface area (Ai = 2r0l)
ln 0
r
i
1 1 r 1 1
U i A i 2ri lh i 2rlk 2r0lh 0 U 0 A o
2ri l ln 0
r
2ri l
1
ri 2ri l
U i 2ri lh i 2rlk 2r0lh 0
1 1 ri r0
ln ri 1
Ui h i k ri r0 h 0
Similarly,
r 1 r
o 0 ln 0 1
1 r
ri h i k i h0
U0 r
2.9 Assignment 2.2
1. Consider a slab of thickness L as illustrated in the figure. A
fluid at a temperature T1 with a heat transfer coefficient h1
flows over the surface at x = 0, and another fluid at a
temperature T2 with a heat transfer coefficient h2 flows over
the surface at x = L of the plate.
Develop an expression for the heat flow q through an area A
of the plate. Calculate the heat transfer rate through A = 1
m2 of the slab for T1 = 130oC, T2= 30oC, h1 = 250 W/(m2.oC), h2= 500 W/(m-2-. oC), L
= 4 cm, k = 20 W/(m.oC). Ans: q = 12.5
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
21
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
2. An iron plate of thickness L with thermal conductivity k is
subjected to a constant, uniform heat flux q” W/m2 at the
boundary surface at x = 0. From the other boundary surface at
x = L, heat is dissipated by convection into a fluid at
temperature T with heat transfer coefficient h. Figure shows
the geometry and the nomenclature.
Develop expressions for the determination of the surface
temperatures T1 and T2 at the surfaces x = 0 and x = L,
respectively.
Calculate the surface temperature T1 and T2 for L = 2 cm, k = 20 W/(m.0C), q” = 105 W/m2,
T = 50oC, and h = 500 W/(m2.oC). Ans.: T1 = 350oC, T2 = 250oC
3. A hollow cylinder with inner radius r = a and outer radius r = b is heated at the inner surface
at a rate of q” W/m2 and dissipates heat by convection from the outer surface into a fluid at
temperature T with a heat transfer coefficient h. There is no energy generation, and the
thermal conductivity of the solid is assumed to be constant.
Develop expressions for the determination of the temperatures T1 and T2 of the inner and
outer surfaces of the cylinder.
Calculate the surface temperatures T1 and T2 for a = 3 cm, b = 5 cm, h = 400 W/(m2.oC), T
= 100oC, k = 15 W/(m.oC), and q” = 105 W/m2.
Ans.: T1 = 352.2oC, T2 = 250oC.
4. A hollow sphere of inside radius r = a and outside radius r = b is electrically heated at the
inner surface at a constant rate of q” W/m2. At the outer surface it dissipates heat by
convection into a fluid at temperature T with a heat transfer coefficient h. the thermal
conductivity k of the solid is constant.
Develop expressions for the determination of the inner and outer surface temperatures T1
and T2 of the sphere.
Calculate the inner and outer surface temperatures for a = 3
cm, b = 5 cm, h = 400 W/(m2.0C), T = 100oC, k = 15
W/(m.0C), and q” = 105 W/m2. Ans.: T1=270oC, T2 = 190oC
5. Consider the composite of two materials combined in parallel
paths with the ends maintained at uniform temperatures as
illustrated in the figure. Various quantities are specified: A1 =
0.2 m2, k1 = 20 W/(m.oC), A2 = 0.4 m2, k2 = 15 W/(m.0C), L
= 0.5m, Ta = 1500C, and Tb = 300C, Calculate the rate of heat
transfer Q across the composite medium.
Ans.: q = 2400W
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
22
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
6. A steel tube with 5-cm ID, 7.6 cm OD, and k = 15 W/(moC) is covered with an insulative
covering of thickness t = 2 cm and k = 0.2 W/(m.oC). A hot gas at Ta = 3300C, ha = 400
W/(m2.0C) flows inside the tube. The outer surface of the insulation is exposed to cooler air
at Tb = 300C with hb = 60 W/(m2.oC).
Calculate the heat loss from the tube to the air of H = 10 m of the tube.
Calculate the temperature drops resulting from the thermal resistances of the hot gas flow,
the steel tube, the insulation layer, and the outside air.
Ans.: q = 7521W, Thot gas = 12oC, Ttube wall = 3.3oC,
Tinsulation = 2530C, Tout side air = 31.7oC
7. A tube with OD of D = 2 cm is maintained at a uniform temperature and is covered with an
insulative tube covering [k = 0.18 W/(m.oC)] in order to reduce the heat loss. Heat is
dissipated from the outer surface of the cover by natural convection with h0 = 12 W/(m2.oC)
into the ambient air at constant temperature. Determine the critical thickness of the
insulation. Calculate the ratio of the heat loss from the tube with insulation to that without
any insulation for (1) the thickness of insulation equal to that at the critical thickness and (2)
the thickness of insulation 2.5 cm thicker than the critical thickness.
q with
Ans.: roc = 1,5 cm, = 1.067.
q without
8. A small, thin metal plate of area A m2 is kept insulated on one side and exposed to the sun
on the other side. The plate absorbs solar energy at a rate of 500 W/m 2 and dissipates it by
convection into the ambient air at T = 300K with convection coefficient hc = 20 W/m2K
and by radiation into a surroundings area which may be assumed to be a blackbody at Tsur =
280 K. The emissivity of the surface is = 0.9. Determine the equilibrium temperature of
the plate. (Ans: 315.5 K)
2.10 Conclusion: In this chapter we have discussed on the derivation of 1-D steady state
conduction heat transfer equation in different coordinate system and determined temperature
distribution equation using appropriate boundary conditions. Electrical analogy of heat conduction
equations is applied for solving problems, using concept of thermal resistances. Also the concepts
of critical insulation and overall heat transfer coefficients have been discussed. One dimensional
heat conduction & the performance of fin will be discussed in the next chapter.
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
23
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
CHAPTER – 3
Heat transfer from extended surfaces-Fin
In this Chapter, the following terms have been discussed in brief.
The temperature distribution in a fin, fin heat transfer rate, fin effectiveness and fin efficiency.
3.1 Introduction of Fin (Heat transfer from extended surfaces)
Definition: A fin is a surface that extends from an object to increase the rate of heat transfer to or
from the environment by increasing convection.
Or combined conduction and convection effect in which extended surfaces are used to enhance heat
transfer between solid and adjoining fluid is known as fin.
Convective heat transfer from a surface is given by:
q = hAs[Ts - T] … … (1)
From Equation (1) it can be shown that q can be increased by increasing either h or As or (Ts - T).
The enhanced heat transfer by increasing As i.e., by attaching an extended surface is called heat
transfer through fins.
Tb h, T
Tb
Tip of the fin
Extended surface
Ac
(No fin) called fin
Ac = Base area of fin
(a) Bare Surface Tb = Base temp. of fin
Fig 1: (b) Schematic diagram of pin fin
In order to find the heat transfer from the fin
surface (qfin), the temperature distribution in the
x t
fin is to be determined. The temperature of the T
b
fin decreases from base of the fin to the tip of Tx Ttip
w
the fin.
L Ac= W x t
(Fin)
Fig-2: Nomenclature of a rectangular fin
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
24
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
Tx = temperature of the fin at a distance x from the base of the fin
Tb = base temperature
Ttip = temperature at the end point
w = width of the fin
L = length of the fin
t = thickness of the fin
Tb > Tx > Ttip
Ac = area of x-section of fin = Base area of fin.
3.2.1 Derivation of governing differential equation of fin (GDE)
Develop a differential equation governing the temperature distribution T(x) in a fin of uniform
cross section assuming the thermal conductivity ‘k’, heat transfer coefficient of ‘h’ and ambient
temperature T.
In order to determine the temperature distribution of a
fin, we have to solve the developed governing qx conv
differential energy equation by performing an energy Tb
balance on differential element.
Consider a fin of uniform x-sectional area Ac. Base
temperature of the fin is Tb. The fin is exposed to a qx qx + dx Ac x
medium having convective heat transfer coefficient h
Tx
and temperature T, thermal conductivity of the fin is k. h
dx
x
In steady state with no heat generation within the fin,
applying the energy balance within the element as Fig-3 Conduction and convection in a fin
shown in the figure (3) and (4).
qx conv
0(no gen) 0(steady)
Ei E g E o Es
Ac
qx cond
E in E o
qx + dx
(cond.)
qx, cond = qx + dx, cond + dq conv. P = 2r dx
(Perimeter)
qx, cond = heat entering by conduction to the element
Fig.-4 Nomenclature for the derivation
qx + dx, cond = heat leaving the element by conduction at 1-D governing fin equation
dq conv. = heat leaving through the surface by convection.
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
25
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
Expanding qx+dx by Taylor series expansion
dx
qx, cond = qx, cond +
d
1! dx
q x,cond hdAs Tx T
dx dxd kA c
dTx
dx
hp dx Tx T 0
dAs = P.dx where P = perimeter of x-section of the fin.
d 2Tx
kAc hD Tx T
dx 2
d 2Tx hp
or Tx T 0
dx 2 kAc
Generalized governing energy equation in terms of local temperature.
hP d 2x d 2Tx
Taking: m 2 & x Tx T &
kAc dx 2 dx 2
x = excess temperature at any location.
d 2 x
m 2 x 0 … … (2)
dx 2
Equation (2) represents the general differential (G.D.E.) of fin and it is a linear 2nd order
homogeneous ordinary one dimensional differential equation.
3.2.2 Solution of G.D.E. for fin
Solving the governing differential equation i.e, Equation (2) the temperature distribution can be
found out over the length of the fin.
d 2 x
m 2 x 0
dx 2
Solution of the above equation is given by c1emx + c2emx
Solving this above differential equation we get,
x= C1emx + C2e-mx … … (3)
C1 & C2 will be obtained from both end B.Cs of the fin.
Both C1 & C2 can be calculated by taking two B.C., i.e.
i) x = 0
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
26
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
ii) and other at x = L.
3.2.2.1 Case-I
Very Long Fin (or infinitely long fin) Tx Ttip = TL
Tb = T
b
d 2 x
Starting from GDE: 2
m2x 0 , in x > 0
dx x
x=0 x = T x - T x = L(Large)
B.C. 1: (x) = Tb - T = b at x = 0
b = T b - T L = 0
B.C. 2: (x) = 0, at x = (large) Fig. 5: Case-1: Infinite long fin
Putting the B.C’s in Equation 3 i.e. (x) = C1emx + C2e-mx
Putting B.C. 1: b = C1 + C2
Putting B.C.2: 0 = C1emL + C2e-mL, where L .
0 = C1emL + 0 C1emL = 0
C1 = 0 So, C2 = b
e- = 0
emL cannot be zero
hence C1 = 0
(x) = b e-mx … … (4)
Or be-mx
d
q = -KAc KAc me mx b
dx x 0
d
q = - KAc kAc .me mx b .
dx x 0
hP
q x 0 q fin KAc .mz.KAc
kAc
and qfin = b hpkA c … … (5)
Equation (4) represents the temperature distribution and equation (5) represents the heat transfer
through the fin in a very long fin.
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
27
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
3.1.2 Case-II
T, h
Similarly solving Equation (2) and putting b = T b - T L
the B.C. in equation (3). qtip 0
i.e. (x) = c1emx + c2e-mx x
Tx and
Putting the B.C.s as x = Tx - T
Fig. 6: Finite long fin with negligible heat
d 2 loss at the tip (insulated tip)
GDE: 2
m 2 0
dx
B.C.1: b = Tb - T at x = 0 kA
dT
0
dx L
d
B.C. 2 : 0 at x = L (tip of the fin).
dx x L dT
0
dx tip
kd
=0
dx x L d
0
dx L
Solution of the GDE (x) = C1emx + C2e-mx
Putting B.C(1) in Equation (1) b = C1 + C2
d
Putting B.C(2) i.e.; 0 mC1emL + C2(-m) e-mL = 0
dx x L
C2 e mL
C1 = and substituting in Equation (2)
emL
C2 e mL emL e mL
b = + C2 = C2
emL e
mL
b emL b e mL
C2 = and C and substituting in Equation (1),
emL e mL emL e mL
1
(x) e(m(L x) e m(L x) cosh m(L x)
… … (6)
(b) e mL e mL cosh mL
and qf b tanh(mL) hpkAc … … (7)
Equation (6) and (7) represents the temperature profile and heat transfer from a fin with insulated
tip.
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
28
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
Let us discuss, what should be the length of the fin so
that it will be called as very long fin and as a result of b
Ttip = TL T.
It has been obtained that b
x = be-mx
qfin = b hpkA c for very long fin and (for long fin)
qfin = b tanh(mL) hpkAc for negligible heat loss
from the tip.
This implies, if tanh(mL) 1, then it will be the
case of very long fin. x=0
Fig. 7: Variation of temperature with length
When mL 2.65, tanh(2.65) = 0.99 for long fin and insulated fin
Note: when mL ≥ 2.65, the fin is said to be very long fin.
hp
This graph quickly falls to zero for large values of m when m = . Then think about the
kAc
factors influencing m and x = f(x).
3.2 Fin Performance
The performance of fin is determined by the following parameters
(1) Fin effectiveness (f)
(2) Fin efficiency (f)
Fin Effectiveness: Fin effectiveness is defined as the ratio of the fin heat transfer to that of which
would occur from the surface on which the fin is attached.
Or it can be defined as the heat transfer with fin that of heat transfer without fin.
heat transfer with fin q
Fin Effectiveness (f) = fin
heat transfer without fin q nofin
0 hpkAc pk
For very long fin: (f) = … … (8)
0 hAc hA c
Equation (8) represents the fin effectiveness at a very long fin.
Effectiveness greater than 2 is justified. Effectiveness will be more when
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
29
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
It can be observed from Equation (8) that f depends on
x
(i) thermal conductivity (k) of the fin is high
Ac = x2
(ii) use fin on the side where h is less. P = 4x
P 4
permiter(P) Ac x
(iii) The shape should be such that should be large. This x/2 x/2
Area (Ac) x/2 x/2 Ac = x
2
is possible by using many pin fins. P = 8x
P 8
P 12x 4 4 8 Ac x
[ 2 , so, . That means use many pin fins] instead of a
Ac 3x x x 3x
single thick fin.
For finite fin with adiabatic tip:
q fin tanh mL hpkAc
f = b … (9)
q nofin b hAc fin
pk
f = tan h(mL)
hAc
Tb T
Rcond & conv. Ttip R conv.
Equation (9) represents effectiveness of adiabatic fin.
Rt, f
Fin effectiveness may also be quantified in terms of thermal
resistance.
Rt, f = Rcond, conv + Rconv (thermal resistance of the fin)
Tb T Tb T
Rt,f =
q fin b tanh mL hpkA c,b
T
R t,b
Tb T 1 Tb
hA c,b Tb T hA c,b
1
Rt,b =
Thermal resistance due to convection at the exposed base) hAc,b
(No fin)
Ac,b = Cross sectional area of the base.
R t,b Thermal resis tan ce convection at the base (at no fin)
f
R t,f Thermal resis tan ce of the fin
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
30
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
1
= hAc1b
tanh mL
pK
1 hA c,b
tanh mL hpkA c,b
To enhance effectiveness of fin (i.e., to enhance heat transfer through fin) it is necessary to reduce
the thermal resistance of the fin (Rt,f).
3.3 Fin efficiency (f): It is defined as the ratio of actual heat transfer through the fin to ideal heat
transfer through the fin, if entire surfaces were at fin base temperature T0.
Let us consider a fin of length L (finite). The heat loss at the tip is negligible. The heat loss at the
tip is negligible. The temperature of the fin decreases from Tb (maximum at the base) to Ttip. As
discussed earlier,
Heat transfer through the fin with insulated tip.
qfin = b tanh(mL) hpkA c … … (10)
Now let us take an ideal fin of same length L and area of cross section Ac. For ideal fin thermal
conductivity k i.e, Rth 0. As there is negligible thermal resistance, the surface temperature of
the entire fin is equal to the base temperature (Tb) of the fin. Hence, the heat transfer by the fin is
maximum. So,
Qmax = hAf,s (Tb - T) = h(PL) (Tb - T) = h(pL)b
Af,s = entire surface area of the fin = (PL + Ac) PL (Neglecting Ac in comparision to surface area
of fin.
(since k i.e., Rth, fin 0, So temperature of the entire
L Tb
surface of fin is equal to Tb)
k
qunfine = q - qf Tb
Tb P(Perimeter)
qfin actual = qfin ideal
qtotal = qfin + qunfined
q fin b tan h(ml) hpkA c tanh mL
Fin efficiency (f) = … (11)
q max b hpL mL
tanh(mL)
f 0 f 1 … … (12)
mL
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
31
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
Very long fin Finite length but heat loss at the tip in negligible
x mx x em(L x) e m(L x) Cosh m(L x)
e
b b emL e mL Cosh mL
& qfin = b hpkA c … 1 Qfin = b tanh(mL) hpkAc … 2
Case Tip condition at Temperature distribution θ/ θ0 = Fin heat transfer Qfin =
x=L
A Infinite long fin e-mx M
B Insulated tip cosh m( L x ) M tanh mL
cosh mL
C Convection heat cosh m( L x) (h / mk ) sinh m( L x) sinh mL (h / mk ) cosh mL
M
transfer at tip cosh mL (h / mk ) sinh mL cosh mL (h / mk ) sinh mL
D Prescribed ( L / 0 ) sinh mx sinh m( L x) cosh m( L 1)
temperature M(θL/θ0 +1)
sinh mL sinh mL
Where, m2=hP/(kAc) θ0 = T0 - T∞, θ0 = T - T∞, M= hPkAc (T0 T )
Solved Example 3.1: A steel rod of diameter D = 2cm, length L = 25 cm, and thermal conductivity
k = 50 W/mK is exposed to ambient air at T = 200C with h = 64 W/m2K. If one end of the rod is
maintained at a temperature of 1200C, calculate temperature at x = 10 cm, heat loss by the rod and
fin effectiveness.
Sol.: Since the entire surface of the rod is exposed to convection, so it is a fin problem. Let us first
decide whether the fin is long or short.
hp hD 4h 4 64
m= 16 T
kAc 2 kD 50 2 10 2
k D
4 Tb = 1200C
T
16 25
mL = 4 (having no unit). T
100
Since mL > 2.65, so this is a case of a very long fin. For a very long fin;
x mx
e and qfin = b hpkA c
b
Tx T 4x
e
Tb T
T10cm 20 4 0.01
At x = 10 cm; e T10cm
120 20
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
32
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
And qfin = (120 – 20) hpkAc 100 64 D 50 D2
4
q fin pk D k 4k 4 50
f 12.5
q nofin hAc 2 hD 64 2
h D 100
4
Solved Example 3.2
A 1 m long, 5 cm diameter, cylinder placed in an atmosphere of 400C is provided with 12
longitudinal straight fins (k = 75 W/m.K), 0.75 mm thick. The fins protrudes 2.5 cm from the
cylinder surface. The heat transfer coefficient is 23.3 W/m2.K. Calculate the rate of heat transfer, if
the surface temperature of cylinder is at 1500C.
Solution
Given: A cylinder with longitudinal fins
= 1 m, d = 5 cm = 0.5 m,
T = 400C, Nfin = 12,
H = 23.3 W/m2. K,
T = 0.75 mm = 0.75 x 10-3m,
L = 2.5 cm = 2.5 x 10-2m,
T0= 1500C k = 75 W/m.K.
To find: the heat transfer rate from surface.
Analysis; The fin is of finite length, it will dissipate heat by convection from its tip. Therefore,
using corrected length.
t 0.75 x10 3
Lc = L + = 2.5 x 10-2 + = 0.025375
2 2
P = 2 = 2m
Ac = x t
= 1 x 0.75 x 10-3 = 0.75 x 10-3m2
hP 23.3x 2
mLc = Lc = 0.025375 x = 0.73
kA c 75 x 0.75 x103
The heat transfer rate from fins surfaces
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
33
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
Qfin = Nfin hPkAc (T0 - T) tanh(mLc) = 12 x 23.3 x 2 x 75 x 0.75 x103 x (150 – 40) x tanh (0.73)
= 379.5 W
Hence the total heat transfer
= Qunfin + Qfin = 379.5 + 1332 = 1711.5 W.
Some important fact of fins:
If heat is to be transfer between two medium one with high h and other low, so fins are
generally attached in the surface where the convective heat transfer coefficient is low. i.e. if
heat is transfers between water and air then fins will be provided in air side.
If mL 2.65 than fin may be assumed to be long fin.
Fin increase the surface area for heat transfer it also increases the thermal resistance over
the portion of the surface where fins are attached.
Pk
1 to justify use at fins.
Ah
Internal conductance should be much greater than convective heat transfer coefficient.
Fin effectiveness can be increased by higher k, higher P/Ac, low h.
Triangular and parabolic has less material so more efficient than rectangular.
The fin efficiency decreases with increase in the fin length because of decreasing in fin
temperature with length.
Longer the fin the larger is the heat transfer area and thus higher the rate of heat transfer
from the fin.
The larger is the fin, bigger the mass, the higher the price and larger fluid section.
3.4. Assignment 3.1
1. A fin has 5 mm diameter and 100 mm length. The thermal conductivity of fin material is
400 Wm-k. One end of the fin is maintained at 130oC and its remaining surface is exposed
to ambient iar at 30oC. If the convective heat transfer coefficient is 40 Wm-2K-1. Determine
the heat loss (in W) from the fin. Ans. -5W.
2. An aluminium alloy fin (k-200 W/mK) 3.5 mm thick, 1m width and 2.5 cm long protrudes
from a wall. The base is at 420oC and ambient air temperature is 30oC. The heat transfer
coefficient may be taken as 15 W/mk/m2K. The heat loss and fin efficiency, if the heat loss
from fin is negligible.
Ans.: 213 W, 99%
3. A rod 2 cm in diameter and 20 cm long protrudes from heat source at 300oC into air at
40oC. The heat transfer coefficient is 5 W/m2K on all exposed surface. Neglecting the radial
variation of temperature and heat loss from the tip, find the temperature of the fin tip.
Find the temperature at the fin tip. Also determine fin efficiency.
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
34
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
Ans. 41.23, 16.5, K = 1.09 W/mK
4. The heat loss from a fin is 6 W. The effectiveness and efficiency of the fin are 3 and 0.75,
respectively. The heat loss (in W) from the fin, keeping the entire fin surface at base
temperature, is __________
5. A long stainless steel rod (k = 16 W/m.K) has a square cross-section 12.5 cm x 12.5 cm and
has one end maintained at 2500C. It is exposed into a fluid at 900C with h = 40 W/m2.K.
Calculate the heat lost by the rod. Ans. 357.77 W.
6. A rectangular copper fin has one end maintained at 2000C, while remainder of the fin
surface is exposed to convective environment at 250C with h = 35 W/m2.K. If the thermal
conductivity of the copper is 386 W/m.K, determine the heat lost by fin per unit depth. The
length of the fin is 5 cm and thickness is 4 mm. Assume the fin tip to be insulated.
Ans. 612.41 W
7. A straight rectangular fin (k=55 W/m.K), 1.4 mm thick and 35 mm long is exposed to air at
200C with h = 50 W/m2.K. Calculate the maximum possible heat loss for a base temperature
of 1500C. What is actual heat loss for this base temperature? Ans. 487.7 W, 309.6 W
8. Consider a long, slender copper rod of diameter D = 1 cm and thermal conductivity k = 380
W/(m.oC) with one end thermally attached to a wall at 2000C. Heat is dissipated from the
rod by convection with a heat transfer coefficient h = 15 W/(m2.oC). Determine the heat
transfer rate from the rod into the surrounding air at T = 30oC. Ans.: 20.16 W.
9. Copper-plate fins of rectangular cross section having thickness t = 1 mm, height L = 10 mm,
and thermal conductivity k = 380 W/(m.oC) are attached to a plane wall maintained at a
temperature T0 = 230oC. Fins dissipate heat by convection into ambient air at T = 30oC with
a heat transfer coefficient h = 40 W/(m2.oC). Fins are spaced at 8 mm (that is, 125 fins per
meter). Assume negligible heat loss from the fin tip.
(a) Determine the fin efficiency.
(b) Determine the area-weighted fin efficiency.
(c) Determine the net rate of heat transfer per square meter of plane wall surface.
(d) What would be the heat transfer rate from the plane wall if there were no fins attached?
Ans.: (c) 26.87 kW/m2, (d) 8 kW/m2.
10. Consider a steel rod of length L = 50 cm, diameter D = 2 cm, and thermal conductivity k =
55 W/(m.oC). One end of this rod is thermally attached to a hot surface maintained at T1 =
150oC and the other end is attached to a cold surface maintained at T2 = 50oC. The rod
dissiplates heat by convection into the ambient air at temperature T = 20oC with a heat
transfer coefficient h = 15 W/(m.oC). Determine the heat loss from the rod into the ambient
air. What fraction of this heat loss is from the surface maintained at T1 = 150oC?
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
35
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
11. Aluminum fins of rectangular profile are attached on a plane wall with 5-mm spacing. The
fins have thickness t = 1 mm, length L = 10 mm, and thermal conductivity k = 200
W/(m.oC). the wall is maintained at a temperature T0 = 200oC, and the fins dissipate heat by
convection into the ambient air t T = 40oC with heat transfer coefficient h = 50
W/(m2.oC).
(a) Determine the fin efficiency.
(b) Determine the are-weighted fin efficiency.
(c) Determine the heat loss per square meter of wall surface.
Ans.: 37.8 kW/m2
3.5 Conclusions: In this chapter we learn the calculation of temperature distribution and heat
transfer through a fin also the calculation of the performance of the fin (, ). Here we got a basic
idea about the selection and design criteria of fin. In the next chapter transient heat conduction in
lump system will be discussed.
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
36
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
CHAPTER – 4
Transient Heat Conduction
This chapter discusses the transient analysis of heat conduction. Our objective, in this chapter, is to
develop a method to determine time dependent temperature distribution within a solid placed in a
convective medium where the conductive internal resistance is very small in comparison to the
convective external resistance. Lumped system of analysis is represented here where the
temperature of the body varies with time only and temperature gradients inside the body are
neglected.
4.1 Transient Conduction
Any phenomenon that changes with time is called transient or unsteady state phenomenon. Cooling
or heating {i.e., temperature changes with time} is an example transient conduction.
An iron block initially was inside a furnace at temperature
say 2000C and is now placed inside a room at temperature
300C. The temperature of the block will change (decrease) T0
T
with time. This phenomenon of heat transfer from block to r
the surroundings is an example of transient conduction. So, Ri Ro
in this case T = f(space, time). If internal resistance of the T(r,t)
block is very small or negligible, then temperature of the
block becomes a function of time only i.e.; T = f (time). The Lc
Ri = Conduction Resistance=
analysis of heat transfer under such an assumption (where, kA
T = f(t)) is called the lumped system analysis. 1
Ro = Convection resistance =
hA s
With reference to the Figure-1, due to the resistance of the
Figure:1 Schematic diagram showing
solid sphere (i.e; conduction resistance or internal
internal and external resistance
resistance) the temperature of the sphere is a function of its
position i.e., T = f(r) and due to cooling, T = f(t) also. So T = T(r,t). If R1(conduction resistance) is
negligible then temperature of the sphere from center to surface will be same and this temperature
will decrease with time only. So, T = f(time) only. This is called Lumped System analysis.
4.2. Develop a governing differential equation to determine temperature distribution in
Lumped system analysis:
Consider a hot metal block, initially at a uniform temperature Ti, is suddenly immersed in a well-
stirred, cold liquid bath which is maintained at a uniform temperature T. The heat transfer
W
coefficient between the block and the liquid is h 2 . The metal block has a volume V(m3),
m K
J
surface area AS(m2), and specific heat cp . Assuming that the lumped system analysis is
kg.K
applicable, develop an expression for the determination of temperature T(t) of the block as a
function of time.
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
37
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
Let Tt be the temperature of the hot metal after ‘t’ second. Applying energy conservation principle
to the hot metal,
Eio E go E 0 E st where, As
(solid)
E 0 E st … … (1) T
V h, T
E 0 = qconv = hAs(T(t) - T∞) (ambient
Cp fluid)
d
E s = ( mc pTt )
dt
Figure 2: Schematic diagram of sudden cooling of hot
m = V, where = density of metal block metal block surrounded by an ambient fluid
Substituting the expressions of E 0 and E s in Equation (1)
d
- hAs[Tt - T] = mc p Tt
dt
d
VCp Tt hA s Tt T 0
dt
dTt
hAs
Tt T 0
dt c pV
It is a 1st order ordinary non homogeneous differential equation having constant coefficients. To
convert into homogeneous forms let us consider:
t = Tt - T
d t dTt
or
dt dt
d dT
or
dt dt
dt hAs
Hence, t = 0 (2)
dt c pV
Equation (2) represents the transient heat conduction equation for lumped system analysis.
Separating the variables in the Equation (2)
dt hAs
Or dt
t cp V
Integrating Equation (2) within the time limit of 0 to t
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
38
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
At t = 0, i = Ti - T
At t = t sec, t = Tt - T
1
t
d t
t
hAs Decreasing order
c pV 0
dt
i
t of (Bi x F0)
t
hAs i
ln t t
i C p V
t Tt T hAs
Or exp . t (3)
i Ti T c V
p
o t
Equation (3) represents the temperature
distribution in a lumped system wrt time. Figure 3: Variation of non dimensional
temperature with time.
hL c t ht
Further, Bi Fo = .
K L c 2 kL c
hkt hA t
= s
V c pV
c p k
As
hAs t
Bi Fo = . cp V
,
t Tt T
so. e Bi Fo (4)
i Ti T
Where, Bi = Biot number = Rinternal / Rexternal ; Biot number is
Figure 4: Temperature
defined as the ratio of resistance to conduction to that of
distribution wrt Biot number
resistance to convection.
t
Fo = Fourier Number = : it is the ratio of rate of heat conducted to the rate of heat stored. It is
L2c
also known as non-dimensional time.
4.3Validity of the lumped capacitance Method
As discussed earlier, lumped system analysis is applicable if Ri = negligible [internal or conduction
resistance] is very small in comparison to Ro (external or convective resistance). In fact, it is the
R
ease when Ri<<Ro . In other words, we can fix a criterion that is i 0.1 for lumped system
Ro
analysis.
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
39
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
Ri Rcond Rint ernal
This ratio Bi (Biot number) It is a non-dimensional quantity
Ro Rconv Rexternal
Lc 1
R cond , R conv
kA hA
Lc
kA hL
R cond
c
So, Bi = ,
R conv 1 k
hA
where, Lc =
V volume
characteristic length
A s surface area
Figure 5: Temperature distribution wrt Biot
k = Thermal conductivity of the solid. Number at different time interval
Note: Bi < 0.1 [T = f(t)] only, Lumped system
analysis is applicable. The temperature profile is shown in Figure 4 and Figure 5.
Bi > 0.1 [T = f(space, time)], Lumped system analysis is not applicable, here T = f(Space, time)
4 3
r
V 3
For sphere: LC = r
AS 4r 2 3
V r 2l r
For long cylinder: LC =
AS 2rl 2
V L
For a thin rectangular block: Lc = , where L = thickness of the block.
As 2
Solved example 4.1
W kg kJ
An aluminium plate [k = 160 , = 2790 3 , cp = 0.88 ] of thickness L = 3 cm and at a
mK m kg.K
uniform temperature of Ti = 2250C is suddenly immersed at time t = 0 in a well-stirred fluid
maintained at a constant temperature. T = 250C. The heat transfer coefficient between the plate
W
and the fluid is h = 320 2 . Determine the time required to reach 500C. Refer Figure 6: Solved
m K
Example 4.1.
Sol.
kJ J
Given: k = 160 Watt/m K, = 2790 kg/m3, cp = 0.88 = 880
kg.K kg.K
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
40
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
W
h = 320 , T= 250C, Ti = 2250C, Tt = 500C, L
m2K
t = ?, L = 3 cm. Y
Let us check for lumped system Analysis. Y
Bi =
hLc h V h
XYL L
k k As k 2 XY XL YL X
Figure 6: Solved example 4.1
h L 1 hL 320 310 2
= 0.03
k 2 1 L L 2k 2 160
Y X
Since X & Y are very large in comparison to L. Hence L/Y and L/X may be neglected.
Bi = 0.03 < 0.1, so lumped system is applicable. That means T = f(t) only
T T 50 25
e Bi Fo e 0.03F0
i Ti T 225 25
t
Where, Fo , Solving t = 239 sec.
L2c
4.2 Assignments
1. The temperature of a gas stream is measured with a thermocouple. The junction may be
approximated as a sphere of diameter D = 1 mm, k = 25 W/(m.0C), ρ = 8400 kg/m3, and cp
= 0.4 kJ/(kg.0C). The heat transfer coefficient between the junction and the gas stream is h =
560 W/(m2.0C). How long will it take for the thermocouple to record 99 percent of the
applied temperature difference? Ans.: time = 4.6 sec
2. A solid copper sphere of 10-cm diameter [ρ = 8954 kg/m3cp = 383J/(kg.0C), k = 386
W/m.0C)], initially at a uniform temperature Ti = 2500C, is suddenly immersed in a well-
stirred fluid which is maintained at a uniform temperature T∞ = 500C. The heat transfer
coefficient between the sphere and the fluid is h = 200 W/(m2.0C).
(a) Check whether the lumped system analysis is suitable.
(b) If it is suitable, determine the temperature of the copper block at t = 5, 10, and 20 min
after the immersion. Ans.: (b) 1200C, 74.50C, 530C
3. Using the lumped system analysis, determine the time required for a solid steel ball of
diameter D = 5 cm [ρ = 7833 kg/m3, cp = 0.465 kJ/(kg.0C), and k = 54 W/(m.0C)] to cool
from 600 to 2000C if it is exposed to an airstream at 500C having a heat transfer coefficient
h = 100 W/(m2.0C). Ans.: 6 min. 34s
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
41
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
4. 2-cm-diameter, stainless-steel ball [ρ = 7865 kg/m3, cp = 0.46 kJ/(kg.0C), and k = 61
W/(m.0C)] is uniformly heated to Ti = 8000C. It is to be hardened by suddenly dropping it
into an oil bath at T∞ = 500C. If the quenching occurs when the ball reaches 1000C and the
heat transfer coefficient between the oil and the sphere is 300 W/(m2.0C), how long should
the ball be kept in the oil bath?Ans:. 1 min 6 sec
5. A 3-cm-diameter aluminium sphere [k = 204 W/(m.0C), ρ = 2700 kg/m3, and cp = 0.896
kJ/(kg.0C)] is initially at T0 = 1750C. It is suddenly immersed in a well-stirred fluid at T∞ =
250C. The temperature of the sphere is lowered to T(f) = 1000C in t = 42 s. Calculate the
heat transfer coefficient. Ans.: h = 200.1 W/(m2.0C)
4.3 Conclusions
This chapter discusses the transient heat conduction analysis. The lumped capacitance method is
discussed assuming that the temperature gradient does not exist within the body and the expression
of variation of temperature with time is obtained. The expression is also used in different practical
problem solving. The next chapter discusses the two dimensional heat conduction analyses.
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
42
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
CHAPTER - 5
Two-Dimensional Heat Diffusion
2T 2T
Differential Equation (D.E): =0
x 2 y2
2T 2T
Non-homogeneous D.E: = 50
x 2 y2
Homogeneous direction:
X-direction is said to be homogeneous direction if, both left
boundary and right boundary indicate homogeneous boundary
condition i.e., T=0 T=0
T = 0 at both left boundary and right boundary. With
homogeneous direction, the temperature field is harmonic. Here x-
direction is homogeneous, so the differential for T with x is :
x
2 2
dT dT
2
- T + 2T = 0.
dx dx 2
The solution of this differential equation gives a harmonic solution
i.e., T = C1Cosx + C2sinx
T=500C
Non-Homogeneous direction:
Y-direction is said to be non-homogeneous if, bottom boundary is at 00C
and top boundary is at any temperature. Other than zero, (say = 500C). y
With non-homogeneous direction, the temperature field (or temperature
distribution) is not harmonic. In this case y-dirction is non-homogeneous
so, the differential equation for temperature with y-variable will be,
T= 00C
d 2T d 2T 2
+ T or 2 - T = 0
2
dy2 dy
The solution for this differential equation will be,
T = C3exy + C4 e-y
Conditions for Method of Separation of variables:
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
43
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
1. Differential Equation must be homogeneous.
2. One of the directions must be homogeneous.
3. While the other direction is non-homogeneous.
Note: Homogeneous direction leads to harmonic solution Non-homogeneous direction leads to
non-harmonic solution.
Fourier Series:
a0
f(x) = b n cos nx c n sin nx , where f(x) is continuous over ≤ x ≤ + 2
2 n 1 n 1
Fourier sine series:
f(x) = c
n 1
n sin nx ,
The coefficient can be obtained as follows:
L
f (x) sin nx dx
x 0
Cn = L
sin
2
nx dx
0
Example-1: Using the method of separation of variables find T(x,y) in a two-dimensional
conduction domain as shown in the figure. Given, there is no
generation and is in steady state.
Y
Sol: T = 500C L, W
0, W
T T2 2
GDE: 0 … (1)
x 2 y2 T=0 T=0
This problem can be solved using method of separation of
variable. Because the G.D.E is homogeneous and x-direction is
X
homogeneous & y- direction is non-homogeneous. O T=0 L, 0
Let T = XY … (2)
We have X = f(x) & Y = f(y)
2T d2X T dY 2 2
Y and X 2
x 2
dx 2
y 2
dy
Substituting in equation (1),
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
44
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
2
d2X dY
Y 2
+ X 2 =0
dx dy
2
1 d2X 1dY
Or 2
= 2
= 2 where f(x,y)
X dx Y dy
Since, x-direction is homogeneous which leads to harmonic solution so (-2) is acceptable.
2
1 d2X 1dY
Therefore, 2
= -2 and 2
= - 2
X dx Y dy
d2X
2
+ 2 X = 0 … (3)
dx
d2Y
and - 2Y = 0 … (4)
dy2
d2X
Solving equation (3) i.e., 2
+ 2X = 0
dx
X = C1 Cosx + C2 Sinx … (5)
d2Y
Solving equation 4 i.e., 2
- 2 Y = 0
dy
Y = C3 ey + C4e-y … (6)
T = (C1 Cosx + C2 Sinx) (C3 ey + C4e-y) … (7)
The coefficients C1 , C2, C3 & C4 will be obtained from four boundary conditions.
B.C. 1 at x = 0, T = 0 [left face], putting in equation (7)
0 = c1(c3ey + c4 e-y) c1 = 0
T = c2Sinx(c3ey + c4e-y) … (8)
B.C. 2: at y = 0, T = 0 [Bottom surface]
0 = c2Sinx (c3+ c4) c3+ c4 = 0
c4 = -c3 and putting in equation (8)
T = c2c3 Sinx(ey – e-y) = 2c2c3Sinx . Sinhy
T = cnSinx . Siny. … (9)
where cn = 2c2c3
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
45
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
B.C. 3: at x = L, T = 0 [right surface]
0 = cn SinL Sinhy SinL = 0 = Sin n
L = n
n
Or = , where n = 1, 2, 3, ……
L
n n
T = cn Sin x .Sinh y … (10)
L L
The value of cn will be obtained from the fourth B.C. i.e.,
at y = W, T = 500C (top surface)
n n
50 = cn Sin x Sinh w
L L
n n
{cnSinh w } Sin x which is a Fourier sine series.
L L
n
So, cn Sinh w will be obtained as,
L
n
L
n
50Sin L x dx 2 1 1
n 1
Cn Sinh w= 0
50
L
L
n n
0 Sin L x dx
2
2 1 1
n 1
50
cn = . & substituting in equation (10)
nw
Sinh
L
nx n
100 1n 1 1 Sin L .Sinh L y
So, T(x,y) =
n 1
nw
Sinh
L
Note: Since, the solution is linear, the general solution is the superposition of all possible solution
i.e., for n to n = .
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
46
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
T = 500C
W
T = 40
T = 30
T=0 T=0
T = 20
T = 10
T=0
0 L
T=0
Figure shows, isotherms for two-dimensional conduction in a rectangular plate.
Two Dimensional Heat Conduction Analysis (Steady state)
Problem – 1
x homogeneous direction (both boundaries are
homogeneous) Y x
T = Tm Sin
L
y Non-homogeneous direction (bottom face is
homogeneous where as, top is non-homogeneous)
Since x – direction is homogeneous, it will gives W T=0 T=0
solutions that are periodic in nature.
The governing differential equation is: X
O T=0
2T 2T
0 … (1) L
x 2 y2
This equation can be solved by using method of separation of variables.
Conditions Required for Method of separation of Variables
1. Both the directions of the problem can be expressed by homogeneous differential equations
and
2. One of the directions of the problem is subjected to homogeneous boundary conditions
called homogeneous direction.
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
47
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
3. While the other direction is subjected to one homogeneous and one non-homogeneous
boundary condition i.e., non-homogeneous boundary direction.
Using separation variable method, in which the solution to the differential equation is assumed to
be a product solution.
Let T = XY … (2)
Where X = X(x)
& Y = Y(y)
Substituting (2) in 1
2
1 d2X 1dY
- = = 2 … (3)
X dx 2
Y dy 2
where = separation constant.
Now we have to choose either +2 or -2 and the condition for this is,
The sign of 2 is chosen in such a way that, the homogeneous direction leds to a solution which is
periodic.
d2 y
Example: 4y = 0
dx 2
Here x-direction is homogeneous, so the differential equation along x-direction should be
1 d2X
2
2 = 0
X dx
d2X 2
Or X= 0 … (4)
dx 2
That means the sign of 2 is chosen as + 2.
d2Y 2
So, Y = 0 … (5)
dy2
Solving (4) X = c1 Cosx + c2 sinx
Solving (5) Y = c3e-y + c4 ey, (= c3 coshy + c4 Sinhy)
T = XY = (c1 Cosx + c2 sinx) (c3e-y + c4 ey)
Or T = (c1 Cosx + c2 sinx) (c3e-y + c4 ey) … (6)
The constants c1, c2, c3 & c4 will be obtained by using four boundary conditions.
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
48
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
B.C. 1: At x = 0, T = 0
B.C. 2: At x = L, T = 0
B.C. 3: At y = 0, T = 0
x
B.C.4: At y = W, T = Tx Sin
L
Putting Boundary Conditions in equation (6)
B.C. 1 0 = C1 (c3e-y + c4 ey)
c1 = 0
T = c2 Sinx(c3e-y + c4 ey) … (7)
B.C.3 0 = c2Sinx(c3 + c4)
c3 = - c4
T = c4c2 Sinx(ey - c4 e-y) … (8)
B.C. 2 0 = c4c2 SinL(ey - c4 e-y)
SinL = 0 = Sin n
x
=
L
x nL y nL y
T = c4c2 Sin x e e
L
Tm
c2c4 =
W
2sinh
L
nx ny
or T = 2c2c4 Sin Sinh
L L
So in the above form we got an infinite no. of solutions that satisfy the differential equation and
boundary conditions. Since the differential equation is linear, the general solution is the
superposition of individual solutions.
nx ny
Hence, T = c
n 1
n Sin
L
Sin h
L
… (9)
To evaluate cn let us apply the fourth B.C.
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
49
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
x
i.e., at y = W, T = Tm Sin
L
x
nx nw
Tm Sin =
L
c Sin
n 1
n
L
Sinh
L
(It is Fourier Sine Series)
x nx
L
nw
T m Sin
L L
dx
0
So, cnSinh = Tm (for n = 1
L 2 nx
L
0 Sin L dx
= 0, for n 1
Tm
Cn =
w
Sinh
L
ny
Sinh
nx L
T = Tm Sin
n 1 L Sinh w
L
Problem-2:
2T
Here, we assume a negligible heat transfer through the top and bottom face i.e; 0 and
z 2
conduction heat transfer is primarily in the x and y direction.
2T 2T
=0 … (1)
x 2 y2
Boundary Conditions:
At x = 0, T = TL, At x = L, T = T1
At y = 0, T = T1 & at y = w, T = T2
To simplify the solution, Let us assume,
T T1
= . So the modified equation is,
T2 T1
2 2
=0 … (2)
x 2 y2
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
50
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
B.Cs: At x = 0, = 0
At x = L, = 0 … (3)
At y = 0, = 0
At y = w, = 1
We now apply the separation of variables technique by assuming that the desired solution can be
expressed as the product of two functions, one of which depends only on x while the other depends
only on y. that is, we assume the existence of a solution of the form
(x,y) = X(x) Y(y) … (4)
Substituting (4) in equation (2)
1 d2X 1 d2Y
- 2
2
= X2 … (5)
x dx Y dy
Since x direction is homogeneous so
d2X
2
+ 2X = 0 … (6)
dx
d2Y
- 2Y = 0 … (7)
dy2
So, the general solutions to equations (6) & (7) are,
X = C1Cosx + C2 Sinx
Y = C3 e-y + C4 ey
= (C1Cosx + C2 Sinx) (C3 e-y + C4 ey) … (8)
Using Boundary Conditions:
At x = 0, = 0
0 = C1(C3 e-y + C4 ey) C1 = 0
= C2 Sinx(C3 e-y + C4 ey) … (9)
At y = 0, = 0 and equation (9) gives
0 = C2 Sinx (C3 + C4)
= C2C4 Sinx(ey – e-y) … (10)
At x = L, = 0 & equation (10) gives
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
51
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
0 = C2C4 SinL(ey – e-y) SinL = 0 = Sin(n)
n
= n = 1, 2, 3, 4, ………….
L
nx nLy nLy
= C2C4 Sin e e … (11)
L
nx ny
Or = Cn Sin Sinh
L L
Where Cn = 2C2C4
Since the problem is linear, a more general solution may be obtained from a superposition of the
form
nx ny
(x, y) = C Sin
n 1
n
L
Sinh
L
… (12)
To determine Cn, Let us apply the fourth B.C. i.e., At y = W, = 1 and equation (12) gives
nx nw
1= C Sin
n 1
n
L
Sinh
L
nw nx
1= C
n 1
n Sinh
L
Sin
L
The value of Cn can be obtained from an analogous infinite series expansion interms of orthogonal
functions, i.e.; Fourier sine series.
An infinite set of functions
g1(x), g2(x), g3(x) ……… gn(x) are said to be orthogonal in the domain a≤ x ≤ b if
b
g
a
n (x) g m (x) dx 0 , m ≠ n
nx nx
We know Sin and cos in the domain 0 ≤ x ≤ L exhibit orthogonal functions. This
L L
ideal we will use here to get Cn.
Let f(x) may be expressed interms of an infinite series of orthogonal functions as
f(x) = C1g1(x) + C2g2(x) + C3g3(x) + ….. + Cngn(x) + …. + Cmgm(x) + …..
Multiplying the above equation by gn(x) and by integrating between the limits a & b.
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
52
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
b b b b b
f (x) gn (x) dx C1g1 (x) g n (x)dx C2g 2 (x) g n (x)dx .... Cng n (x)dx .... Cmgm (x)dx .....
a a a a a
b b
f (x) g n (x) dx Cn g n (x) dx
a a
f (x) g
a
n (x) dx
Or Cn = b
g
2
n (x) dx
a
Our problem is
nw nx
1= C n Sinh
L
Sin
L
n 1
L
nx L nx
L
nw 0
1.Sin
L
dx n cos L
0
Cn Sinh = L L
L 2 nx 1 2nx
0 Sin L dx 2 0 1 cos L dx
L
n
Cosn Cos 0
=
1 x L Sin 2nx
L
2 2n L
L
1 1 2
n
n
1 1
n 1
=
1
(L 0) n
2
2 1 1
n 1
=
n
nw 2 1 1
n 1
Cn Sinh =
L n
2 1 1
n 1
1
Or, Cn =
nw n
Sinh
L
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
53
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
nx ny
(x,y) = C
n 1
n Sin
L
Sinh
L
2 1 1
n 1
1 nx ny
= Sin Sinh
n 1 nw n L L
Sinh
L
ny
Sinh
1 1 Sin nx
n 1
2
(x,y) = L is a covergent series.
n 1
n L Sinh nw
L
Figure shows, isotherms for two-dimensional conduction in a rectangular plate.
Conclusion:
2T 2T
0 … (1)
x 2 y2
T T1
= … (2)
T2 T1
2 2
0 … (3)
x 2 y2
B.C.s:
At x = 0, T = T1, = 0
At x = L, T = T1, = 0 … (4)
At y = 0, T = T1, = 0
At y = w, T = T2, = 0
The solution for is
ny
Sinh
1 1 Sin nx
n 1
2
(x,y) = L …
(5)
n 1
n L Sinh nw
L
ny
Sinh
1 1 Sin nx
n 1
T(x, y) T1 2
L …
(6)
T2 T1 n 1
n L Sinh nw
L
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
54
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
Heat flux may now be calculated as
T
qx k … (7)
x
T
and qy k … (8)
y
However, note that the discontinuities prescribed at the upper corners of the plate are physically
unterable. In reality, large temperature gradients could be maintained is proximity to the corners,
but discontinuities could not exist.
For Problem-1: Note:
1 d2X 1 d2Y
2
2
2
X dx Y dy
For the case 2 = 0
d2X
0 X = C1 + C2x
dx 2
d2Y
& 0 Y = C3 + C 4 y
dy 2
= (C1 + C2x) (C3 + C4y)
At x = 0, T = 0 0 = C2(C3 + C4y)L C2 = 0
Since C1 and C2 both are zero, so T(x,y) = 0, which is not possible. So 2 ≠ 0.
For the Case (-2).
d2X
- 2X = 0 X = C1e-X + C2eX
dx 2
d2Y
2
+ 2Y = 0 Y = C3 Cosy+ C4 Siny
dy
T = (C1e-X + C2eX) (C3 Cosy+ C4 Siny)
At x = 0, T = 0 0 (C1 + C2) (C3 Cosy+ C4 Siny)
So, C1 + C2 = 0 C1 = -C2 T = C2 Sinhx (C3Cosy + C4 Siny)
At x = L, T = 0 0 C2SinhL (C3Cosy + C4 Siny) C2 ≠ 0 So SinhL = 0,
eL – e-L = 0 e2L = 1
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
55
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
=0
At y = 0, T = 0 0 = C2(Sinhx) X C3 C3 = 0.
T = C2 C4Sinhx Siny = 0 for = 0
So (-2) is not possible.
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
56
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
CHAPTER – 6
Numerical Heat Transfer
In this Chapter, the following terms have been discussed in brief.
6. Numerical Heat Transfer(Finite Difference Solution)
dT Ti 1 Ti
0 x
1
dx i x
dT Ti Ti 1
0 x
1
dx i x
dT Ti 1 Ti 1
0 x
2
dx i 2 x
d 2 T Ti 1 Ti 1 2Ti
0 x
2
x
2 2
dx i
Example – 1.
Find the nodal temperatures using FDM.
Take: Gauss – seidel interative technique and 200C as Guess value for all 3 nodes.
Solution: Governing differential equation:
X X X X
T q 1 T
2
200C
100
x 2 k t
T1 T2 T3
1-D, Steady state heat
d 2T
= 0 ( steady & having no generation) conduction without generation
dx 2 (no generation)
d 2T
Discretizing about the node i,
dx 2
d 2 T Ti 1 Ti 1 2Ti
0 x 0
2
x
2 2
dx i
2Ti = Ti+1 + Ti-1, where i = 1, 2 & 3.
1
T1k 1 T2k 100
2
1
T2k 1 T1k 1 T3k
2
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
57
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
1
T3k 1 T2k 1 20
2
(by Gauss Seidel interative technique.)
Iteration (k) T1 T2 T3
Guess Value 20 20 20
1st 60 40 30
2nd 70 50 35
3rd 75 55 37.5
4th 77.5 57.5 38.75
5th 78.75 58.75 39.375
Exact Solution 80 60 40
80
Texact = x 100
L
MATLAB CODE: Please write the code to solve the problem
Heat Transfer : FDM
Example – 2
W
4
Find the nodal temperatures using FDM and G-S iteration 10 m 2
and taking 800 as guess temperature. 400C
Solution: T1 T2 T3
Governing differential equation: k = 20 W/mK
X = 5 cm
2T q 1 T q = 0
One-D, Steady State conduction
x 2 k t
d 2T
=0
dx 2
Discretizing using FDM about node i where I = 2 & 3 [for interior nodes]
1
Ti = [Ti+1 + Ti-1] This equation is applicable for only nodes 2 & 3.
2
1
T2 = (T1 + T3) (1)
2
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
58
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
1
T3 = (T2 + 40) (2)
2
One more equation is required for 3 unknowns and it can be obtained from B.C. at left end.
dT T T
q K 104 20 2 1
dx i 1 X
20
104 T2 T1
5 102
T1 = T2 + 25 (3)
dT Ti 1 Ti
dx i x
Using Gauss-seidel Iterative technique,
T1k 1 T2k 25
1
T2k 1 T1k 1 T3k
2
1
T3k 1 T2k 1 40
2
Iteration (k) T1 T2 T3
Guess Value 80 80 80
Ist Itertion 105 92.5 66.25
2nd 117.5 91.875 65.938
3rd 116.875 91.406 65.70
4th 116.40 91 65.5
5th
Exact 115 90 65
Texact = - 500x + 115
Example – 3:
i=1 i=2 i=3
Find the nodal temperatures using FDM and G-S-
Iteration and taking 900 as Guess temperature. 1200C T = 200C
W
T1 T2
T3 h = 200 2
Sol.: m K
k = 20 W/mK, X = 5 cm
One-D, heat conduction without
generation
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
59
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
d 2T
GDE =0
dx 2
Finite Difference equation: 2Ti = [Ti+1 + Ti-1] (for i = 1 & 2)
1
For i = 1; T1 = T2 120 (1)
2
1
i = 2; T2 = T1 T3 (2)
2
Using B.C. at the right end
dT
k h T3 T
dx 3
T T
k 3 2 h T3 T
x
20
T3 T2 200 T3 20
5 102
1
T3 = 2T2 20 (3)
3
Using G-S iteration:
1
T1k 1 T2k 120
2
1
T2k 1 T1k 1 T3k
2
1
T3k 1 2T2k 1 20
3
Iteration (k) T1 T2 T3
Guess Value 90 90 90
1st 105 97.5 71.6772
2nd 109 90.0 67
3rd 105 86 64
4th 103 83 62
5th
Exact Solution 100 80 60
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
60
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
120
Texact = 120 – 400x, where x is in meter.
Example-4: T1 T2
Find the nodal temperature using FDM. Take: Gauss-Seidel
60 40
iteration technique & guess value as 00C for all 4-nodes. T3 T4
Solution:
2T 2T q 1 T
GDE:
x 2 y2 k t 00C
X = Y
2T 2T q = 0 (no generation)
0
x 2 y2 Steady state model
Discretizing ther Differential equation about the node (i, j).
2 T 2 T Ti 1j Ti 1j 2Tij Tij1 Tij1 2Tij
2 2 0 x Y
2 2
x y ij x Y
2 2
4Tij = Ti+1j + Ti-1j + Tij+1 + Tij-1
1
Or Tij = [Tij = Ti+1j + Ti-1j + Tij+1 + Tij-1]
4
1
Or TP = [TE + TW + TN + TS]
4
For Node at (T1)
1
T1 = [T2 + 60 + 120 + T3]
4
1
Or T1 = [T2 + T3 + 180 ]
4
For Node at (T2)
1 1
T2 = [40+ T1 + 120 + T4] = [T1 + T4+ 60]
4 4
For Node at (T3)
1 1
T3 = [ T4 + 60+ T1 + 0] = [T1 + T4+ 60]
4 4
For node at (T4)
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
61
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
1 1
T4 = [ 40 + T3 + T2 + 0] = [T2 + T3+ 40]
4 4
Using G-S- Iteration,
1
T1k 1 T2k T3k 180
4
1
T2k 1 T2k 1 T4k 160
4
1
T3k 1 T1k T4k 60
4
1
T4k 1 T2k 1 T3k 1 40
4
Iteration (k) T1 T2 T3 T4
Guess Value 0 0 0 0
1st Iter 45 51.25 26.25 29.4
2nd Iter 64.4 63.4 38.5 35.5
3rd Iter
4th ITER
5th Iter.
Program – 2
Clc;
Close all;
Clear all;
W
k = 20 = Thermal conductivity
mK
L = 15 x 10-2m = length of the rod
N = No. of segments = 3
L
X = X DX
N
W
q HF 104 heat flux
m2
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
62
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
TN+1 = 400C
For i = 1 : (N + 1)
If i = 1
ai = 1;
bi = 1;
ci = 0;
di =
HF DX ;
K
else if i = (N + 1)
ai = 1;
bi = 0;
ci = 0;
di = TN+1;
else
ai = 2;
bi = 1;
ci = 1;
di = 0;
end
end
For I = 1 : (N+1)
If i = 1
Pi = 1;
HF DX
Qi = ;
K
else
bi
Pi = ;
a i ci Pi 1
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
63
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
d i Ci Qi 1
Qi = ;
a i ci Pi 1
end
end
For i = (N+1) : -1 : 1
if I = (N+1)
Ti = Qi
Else
Ti = Pi Ti+1 + Q;
T1 T2 T3 T4 = 400C
end
L
end.
N=3
segments
Problem – 2
q = 104
W
K = 20
mK
L = 15 cm
L
X = = 5 cm
N
W
q = 104
m2
TN+1 = 400C
dT
q K
dx 1
T T
q K 1 2
X
q X
T1 = T2 +
K
2T2 = T3 + T1
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
64
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
2T3 = T4 + T2
T4 = 40 = TN+1
CT1 = 115
T2 = 90
T3 = 65
T4 = 40
FDM-Fin
Example – 5: An iron rod L = 5 cm long of diameter D = 2 cm with thermal conductivity K =
W W
50 protrudes froma wall and is exposed to an ambient at T = 200C and h = 100 2 . The
mK m K
0
base of the rod is at Tb = 320 C, and its tip is insulated. Assuming one-dimensional steady-state
heat flow, calculate the temperature distribution along the rod and the rate of heat loss into the
ambient by using finite differences. Compare the finite difference solution with the exact analytical
solution of this problem. Devide the domain into 5 equal segments.
Sol:
x=0
x = Tx - T 1 2 3 4 5
Tb = 3200C
d dTx
or b = 3000C
dx dx x x
GDE: b x
d 2 W
x=L
m 2 0 h = 100
dx 2 m2K
Insulated
B.C. T = 200C
d
At x = 0, b = 3000C = Tb - T W i.e; 0
K = 50 dx L
mK
At x = L, qtip = 0
L
x = = 1 cm = 0.01 meter
Dt d 5
-K 0 0
dx L dx L
x cosh m(L x)
Exact solution:
b cosh mL
hP
Where m = = 20
KAc
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
65
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
2
P = D, AC = D
4
D = 2 cm = 0.02 meter.
qfin = b hpkAc tanh(mL)
Dsicretizing the above equation about node I, using FDM,
d 2 2
m2i 0 i 1 i 1 2 i m 2 i 0
2
dx i x
(2 + m2(x)2)I = i+1 + i-1, for i = 1, 2, 3 & 4 (all interior nodes)
2.04 i = i+1 + i-1; m2(x)2 = 400 x (0.01)2 = 0.04
1
i = 1, 1 = [2 + 300]
2.04
1
2 = [3 +1]
2.04
1
3 = [4 +2]
2.04
1
4 = [5 +3]
2.04
For boundary node i = 5
2.04 5 = 6 + 4 where 6 will be obtained as,
d
0 6 4 0 (Central differencing scheme)
dx 5 2 x
6 = 4 and substituting in above equation,
We have; 2.045 = 24
2
So, 5 = 4
2.04
Using Gauss-seidel Iterative technique, the equations are:
1
1k 1 k2 300
2.04
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
66
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
1
k2 1 3k 1k 1
2.04
1
3k 1 k4 k2 1
2.04
1
k4 1 5k 3k 1
2.04
2 k 1
And 5k 1 4
2.04
FDM
Iteration 1 2 3 4 5
Guess Value 100 100 100 100 100
1st Iteration 196 145 120 108 105
nd
2 218 166 134 117 115
3rd 228 177 144 127 125
th
4 234 185 153 136 133
5th
Exact 260 230.5 210.2 198.3 194.4
Calculation of heat loss:
qfin (exact) = b hpkAc tanh(mL) = 71.76 Watt. (Exact)
Numerical solution by FDM:
x
Let us apply energy conservation principle to element at b as shown
2
(steady)
Ei E g Eo E s
Ei Eo
b
d hp x
q f KAc b
dx x 2 x
2
2
hp x
(x)
= KAc 1 b 2 b
x Ei qf
KA c hp x
= b 1 b E o q cond q conv
x 2
d x
Qf = 72.1 Watt (Finite Difference) = KA c hP Tb T
dx x 2
2
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
67
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
Unsteady Conduction – Explicit Scheme:
1-D Conduction (Unsteady) without having generation.
0
2T q 1 T
GDE:
x 2
k t
2T T
(1)
x 2 t
Discretizing the GDE about node i and time ‘t’ with time step t:
t
2 T T
t
2
x i t i
With explicit Scheme:
T t T t 2T t Tit t t it
i 1 i 1 2 i
x t
t
Tit t Tit Tit1 Tit1 2Tit
x
2
t t 2t t
Or Tit t Tit1 Tit1 1 T
x
2
x
2
x 2 i
The solution will be stable, if all the coefficients in the above equation are positive, i.e.;
2t x
2
1 0 t
x 2 2
This is the condition for time step size (t) in explicit scheme to get stable solution.
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
68
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
X
2
Finite Difference Solution when (t) >
2
T(x,t)
Exact Solution
X
2
Finite Difference Solution when (t) <
2
O x
Example – 6.
W
A marble slab (k = 2 , = 1 x 10-6 m2/s) that is L = 2 cm thick is initially at a uniform
m.k
temperature. Tinitial = 200 C. Suddenly one of its surface is lowered to 00C and is maintained at that
0
temperature., while the other surface is kept insulated. Develop an explicit finite-difference scheme
for the determination of the temperature distribution in the slab as a function of position and time as
well as the heat flux at the boundary surface after 40 sec. divide the domain into 5 equal segments.
Solution:
2T T
GDE: =
x 2 t
T
B.C: (1): 0 at x = 0 (left surface)
x
B.C. (2): T = 0 at x = L (right surface)
Initial condition: T = 200 for t = 0, at all x.
Discretizing the GDE,
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
69
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
t t 1 2 t t
Tit t Tit1 Tit1 Ti
x x x
2 2 2
2 cm
Here, x = = 0.4 cm = 4 x 10-3 meter
5
4 10
2
x 3
2
For stable solution, (t) ≤ = 8 sec.
2 2 106
t 1
So, let us take (t) = 8 sec.
x 2
2
Setting t = 8 sec, the Finite difference equation are
Tit t
2
1 t
Ti 1 Tit1 for I = 2, 3, 4 & 5 (All interior nodes).
T
0
x 1
1 2 3 4 5 6
T3 T1t
1 t T6=0 (B.C.)
T2t t
2
T3t t
2
T4 T2t
1 t
T4t t
1 t
2
T5 T3t
T5t t
1 t
2
T6 T4t
T6t t 0
T1t t will be obtained from B.C. at left surface.
t t
dT T2t t T1t t
0 0
dx 1 x
T1t t T2t t
The heat flux at the right boundary at t + t = 40 sec.
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
70
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
t t
dT T t t T6t t
q K K 5
dx 6
Finite-Difference calculations for Tit t
(t) t, sec T1 T2 T3 T4 T5 T6
Time step x=0 x=0.4 cm x=0.8 x=1.2 x=1.6 x=2 cm
0 0 200 200 200 200 200 200
(Initially)0
1 8 sec. 200 200 200 200 200 0
2
2 16 200 200 200 200 100 0
3 24 200 200 200 150 100 0
4 32 200 200 175 150 75 0
5 40 187.5 187.5 175 125 75 0
6 48 181.2 181.2 156.2 125 62.5 0
7 56 168.7 168.7 153.1 109.4 62.5 0
The Finite Difference Equations are:
T1t t T2t t
1
T2t t T3t T1t
2
1
T3t t T4t T2t
2
1
T4t t T5t T3t
2
1
T5t t T6t T4t
2
T6t t 0
T5t t T6t t K
and qboundary K T540 T640
X x
2
= 75 0
4x10 3
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
71
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
= 37.5 x 103 W/m2
= 37500 W/m2
Unsteady Conduction – Implicit Scheme
2T T
GDE: [no generation]
x 2 t
Discretizing the GDE about any node i and time step t, the nodal temperature at any new time
(t + t) is:
Tit t t
1 Ti 1 2Ti
t t t
(T t t T t )
i i
x
2
t
t
Tit t t
1 Ti 1 2Ti
t t t
Tit t Tit
x
2
2 t t t t t t t t t
1 T T t
T
2 i 1
T
2 i 1
x x x
2 i i
Here, all the coefficients are positive. So, the solution is unconditionally stable.
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
72
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
CHAPTER – 7
Heat Exchangers
Heat exchangers are devices that facilitate heat transfer between two or more fluids at different
temperatures. Many types of heat exchangers have been developed for use at such varies levesl of
technological sophistication and sizes as steam power plants, chemical processing plants, building
heating and air conditioning, household refrigerators, car radiators, radiators for space vehicles, and
so on. In the common types, such as shell-and-tube heat exchangers and car radiators, heat transfer
is primarily by conduction and convection from a hot to a cold fluid, which are separated by a metal
wall. In boilers and condensers, heat transfer by boiling and condensation is of primary importance.
In certain types of heat exchangers, such as cooling towers, hot fluid (i.e., water) is cooled by direct
mixing with the cold fluid (i.e., air); that is, the water sprayed or falling down into in induced air
draft is cooled by both convection and vaporization. In radiators for space applications, the waste
heat carried by the coolant fluid is transported by convection and conduction to the fin surface and
from there by thermal radiation into the atmosphere-free space.
Classification of Heat Exchangers
Classification of Heat exchangers are mainly based on
(1) The transfer process, (2) Compactness (3) construction type (4) flow arrangement, and (5) heat
transfer mechanism.
Classification based on Transfer Process:
Direct Contact and Indirect Contact types
Classification according to Compactness:
The compactness of heat exchanger is defined as the ratio of heat transfer surface area on one side
of the heat exchanger to the volume of the heat exchanger. The above ratio which is more than 700
m2/m3 is referred to as compact heat exchanger.
Classification by Construction type:
Tubular heat exchanger (shell-and-tube type heat exchanger, Plate heat exchanger, Plate-fin heat
exchanger, Tube-fin heat exchanger, Regenerative heat exchanger)
Classification by Flow Arrangement:
Parallel flow, Counter flow, Cross flow, Multipass flow
Classification by Heat Transfer Mechanism:
The possibilities for the heat transfer mechanism included a combination of any two of the
following:
1. Single-phase forced or free convection
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
73
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
2. Phase change (boiling or condensation)
3. Radiation or combined convection and transfer.
The examples for the above three categories of heat exchangers are: Condensers, Boilers, Radiators
for space power plants.
Analysis of heat exchanger is carried out by two methods
(1) LMTD method for heat exchanger analysis
(2) Effectiveness (ϵ) – NTU method for heat exchanger analysis
Before going for heat exchanger analysis it is required to know the overall heat transfer coefficient
as described below.
Overall Heat Transfer Coefficient (U)
Accumulation of deposits
1
R1 = = Inside flow convection causing Fouling conditions.
h i Ai
resistance
Rf ,i
R2 = = Inside flow fouling
Ai ri
resistance.
TI, hi
ln 0
r ro
R3 = ri = Conduction resistance
2lk
k
(wall thickness resistance)
Rf,o
R4 = = outside flow fouling
Ao
resistance. R4 T0
TI
1 R1 R3 R5
R5 = = Outside fluid convection
h 0A0
resistance.
Net resistance, R = R1 + R2 + R3 + R4 + R5 … (1)
1 1
where, R = = net resistance which is convective type based on inner or outer surface
Ui Ai U0 A0
of the tube. Substituting the values of R, R1, R2 …..R5 in the above Equation (1)
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
74
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
ro
1 1 1 1 Rf ,i ln ri R 1
= + + f ,o + … (2)
UA U i Ai U o Ao hi Ai Ai 2lk Ao h0 A0
W
Ui = Overall heat transfer coefficient 2 based on inner surface of the tube
m K
U0 = Overall heat transfer coefficient based on outer surface of the tube.
m2 K
R f ,i = Fouling factor at the inner surface of the tube
W
m2 K
Rf ,o = Fouling factor at the outer surface of the tube
W
The overall heat transfer coefficient (U) as described in Equation (2) depends on the heat transfer
coefficient (hi) of the fluid inside the tube and the heat transfer coefficient (ho) of the fluid outside
the tube. The value of hi and ho will be calculated from Nusselt number (NuD) for flow in tubes.
hDh 4 Ac
NuD = , where hydraulic diameter Dh = , Ac = flow cross section and Pw = wetted
k fluid Pw
perimeter of the flow domain)
kfluid = thermal conductivity of fluid which is flowing.
So, the value of hi and ho can be determined from Nusselt no. as,
kfi kf0
hi = NuDi x & h0 = NuD0 x ,
Dh i Dh 0
Now, we need to know the value of Nusselt number (NuD). The Nusselt no. again depends on flow
nature, i.e., whether Laminar flow or turbulent flow in a pipe.
For Laminar flow
48
NuD = (constant wall heat flux)
11
NuD = 3.66 (constant wall temperature)
For Turbulent flow
NuD = 0.023 Re 0D.8 Prn (Dittus-Boelter equation)
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
75
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
u m Dh
Where, Reynolds number for pipe flow ReD = ≤ 2300 (Laminar) else transition to turbulent
and n = 0.3 for the fluid in the pipe gets cooled or n = 0.4 for the fluid in the pipe gets heated.
depends on the mean flow across the cross section(Ac).
and mass flow rate m
m
= ρumAc or um =
m
Ac
Di = D
For Circular cross section pipe
4 D2
4 Ac 4 u m Dh um D 4m
Dh = = = D and ReD = = = Do= Di+ thickness
Pw D D Tube cross section
For annular cross section pipe
Ds (Diameter of
4 Ac
4
4
D s
2
Do
2
the Shell)
Dh = =
Pw Ds Do Shell
Do
= (Ds – Do) Di
d
Tube showing its
m
m ρumAc um = wall thickness
Ac
Shell & Tube cross section
u m Dh m 4 Ac 4m
RcD = = =
Ac Pw ( Ds Do )
Solved Examples for determination of overall heat transfer coefficient.
Solved example 5.1
Determine the overall heat transfer coefficient U0 based on the outer surface of a steel pipe with Di
W
= 2.5 cm and D0 = 3.34 cm, Thermal conductivity of the pipe wall k = 54 for the following
mK
W W m2K
flow and fouling conditions: hi = 1800 , h 0 = 1250 , R
f ,i = R
f ,o = 0.00018 ,
m2K m2K W
Solution:
W
[Ans: U0 = 487.1 ]
m2K
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
76
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
Heat transfer analysis in a heat exchanger can be done in two methods such as (1) LMTD Method
& (2) Effectiveness – NTU Method.
The LMTD Method for Heat Exchanger Analysis:
Heat transfer in a heat exchanger is defined as q = UAΔTmean, where U = Overall heat transfer
coefficient over surface area A and ΔTmean = mean temperature difference between hot fluid and
cold fluid over the entire length (L) of the heat exchanger.
For parallel flow and counter flow type shell and tube type heat exchanger,
ΔTmean = ΔTlogmean = LMTD (i.e. log mean temperature difference).
So, q = UA ΔTmean = UA ΔTln (for parallel & counter flow heat exchanger)
q = F UAΔTlm,count where F = Correction factor against ΔTlm,count for the ease where the flow
domain is neither parallel nor counter flow.
Here ΔTmean = F x ΔTlm,count. the value of F will be obtained from different heat exchanger charts.
Problem 1:
Problem. 2: Water at a mean temperature of Tm= 800C and a mean velocity of um = 0.15 m/s flows
inside a 2.5 cm. ID, thin-walled copper tube. Atmospheric air at T = 200C and a velocity of u =
10 m/s flows across the tube. Neglecting the tube wall resistance, calculate the overall heat transfer
coefficient and the rate of heat loss per 1-m length of the tube.
Given:
Properties of water at 800C
Water
flow
Tm = 800C, um = 0.15 m/s Di Water
0.364 x106 m2 / s flow
W
K = 0.668 , Pr. = 2.22
mK
umD Air flow: u= 10 m/s
ReD = ,
T = 200C
hD
Nu D 0.023Re 0.8 Pr 0.3
K fluid
(Dittus – Boelter equation) for turbulent flow i.e.; ReD > 2000
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
77
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
80 20
50 C (mean film temperature).
0
Properties of air at
2
18.22 x106 m2 / s
W hD
K = 0.0281 , Pr = 0.703, NuD = (0.4 Re0.5 + 0.06 Re2/3) Pr0.4 =
mK K fluid
W
Ans: Water side: Re = 10,300, Nu = 47.4, hi = 1267
m2K
W
Air Side: Re = 13,721, Nu = 70.52, h0 = 79.3
m2K
W W
U = 74.63 2
& q = UA [Tm - T] = 351.7
m K m
Problem – 3: An oil cooler for a large diesel engine
is to cool engine oil from 600 to 450C, using sea 600
water at an inlet temperature of 200C with a oil
temperature rise of 150C. the design heat load is q = 450
140 kW, and the mean overall heat transfer T1= 40 C
0
350
coefficient based on the outer surface area is 70
T2= 100C
W water
. Calculate heat transfer area (A) based on (a) 200
m2K
parallel flow (b) counter flow.
Solution of problem-3. (a) Parallel flow
q = UA Tlm
200C Sea water
= UA(LMTD)
T1 T2 450C
LMTD = 0
60 C
T
ln 1 Hot oil
T2 Oil in
Cold sea water
40 10 30
= = 21.640C
40 ln 4
ln
10 350C
Given: q = 140 kW = 140,000 W
W
U = 70
m2K
LMTD = 21.640C (Calculated)
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
78
SIKSHA ‘O’ ANUSANDHAN DEEMED TO BE UNIVERSITY
So, q = UA(LMTD)
q 140, 000
A= = 92.42 m 2
U(LMTD) 70 x 21.64
For parallel flow arrangement, area required
is 92.42 m2. 350C water
(b) For counter flow
450C
T1 = 250C = T2 600C
Hot oil
So, LMTD = 250C
q 140000
A= = 80m 2
U(LMTD) 70 x 25
200C
For counter flow arrangement, area
required is 80 m2. Where as for parallel flow it is found to be 92.42 m2. So, we note that less area is
required with the counter flow arrangement. It is because the counter flow arrangement is more
balanced type of heat exchanger.
600
T1= 250C oil
0 450
35
T2= 250C
water
200
The overall heat transfer coefficient can be calculated as,
ln 0
r
Rf i Rf 1
i 0
1 1 r
UA hi Ai Ai 2lk wall A0 h0 A0
m2k
where, Rf i = fouling factor of inner surface of the tube. .
w
m2k
Rf 0 = foulns factor of outer surface of the tube
w
D.N. Thatoi, S. Samantaray, D. Mohapatra, B. Pattnaik, S.K.Acharya
79
You might also like
- Course Syllabus: Bes - Emech Engineering Mechanics First Semester, AY 2020-2021No ratings yetCourse Syllabus: Bes - Emech Engineering Mechanics First Semester, AY 2020-20218 pages
- Chapter One: - Methods of Heat TransferNo ratings yetChapter One: - Methods of Heat Transfer24 pages
- Introduction and Basic Concepts: ObjectivesNo ratings yetIntroduction and Basic Concepts: Objectives35 pages
- Lectures Presented by B.K.Roy Assistant Professor Mechanical Engineering DepartmentNo ratings yetLectures Presented by B.K.Roy Assistant Professor Mechanical Engineering Department40 pages
- Unit 2.: Unit 1 3-30 Unit 1 Interview Questions 31,32 Unit 1 Numerical Questions First Mid Term Paper 33 Unit 2No ratings yetUnit 2.: Unit 1 3-30 Unit 1 Interview Questions 31,32 Unit 1 Numerical Questions First Mid Term Paper 33 Unit 212 pages
- Chapter 1: Fourier Equation and Thermal ConductivityNo ratings yetChapter 1: Fourier Equation and Thermal Conductivity19 pages
- Conduction - Part 1 - Heat and Mass TransferNo ratings yetConduction - Part 1 - Heat and Mass Transfer39 pages
- CHAPTER I: Introduction To Heat Transfer Methods of Heat Transfer, Conduction, Convection and RadiationNo ratings yetCHAPTER I: Introduction To Heat Transfer Methods of Heat Transfer, Conduction, Convection and Radiation83 pages
- Ch.1 Essential Concepts: 1.1 What and How? What Is Heat Transfer?No ratings yetCh.1 Essential Concepts: 1.1 What and How? What Is Heat Transfer?151 pages
- Feynman Lectures Simplified 2C: Electromagnetism: in Relativity & in Dense MatterFrom EverandFeynman Lectures Simplified 2C: Electromagnetism: in Relativity & in Dense MatterNo ratings yet
- Feynman Lectures Simplified 2B: Magnetism & ElectrodynamicsFrom EverandFeynman Lectures Simplified 2B: Magnetism & ElectrodynamicsNo ratings yet
- Centrifugal Compressor Parts & Their Function - 2020No ratings yetCentrifugal Compressor Parts & Their Function - 202013 pages
- Advances in Circular Dichroism Spectroscopy and Its Applications100% (2)Advances in Circular Dichroism Spectroscopy and Its Applications4 pages
- Recent Progress in Functional Analysis Proceedings of the International Functional Analysis Meeting on the Occasion of the 70th Birthday of Professor Manuel Valdivia 1st Edition Klaus D. Bierstedt - Instantly access the full ebook content in just a few seconds100% (1)Recent Progress in Functional Analysis Proceedings of the International Functional Analysis Meeting on the Occasion of the 70th Birthday of Professor Manuel Valdivia 1st Edition Klaus D. Bierstedt - Instantly access the full ebook content in just a few seconds54 pages
- ME-406 Software LAB Manual - 1710572459No ratings yetME-406 Software LAB Manual - 171057245947 pages
- Consumers' Purchase Intentions For Foreign Products: An Empirical Research Study in Istanbul, TurkeyNo ratings yetConsumers' Purchase Intentions For Foreign Products: An Empirical Research Study in Istanbul, Turkey8 pages
- Section: A (Very Short Answer Type Questions) : Cycle Test - 1 Physics STD: XII TIME: 1.5 HrsNo ratings yetSection: A (Very Short Answer Type Questions) : Cycle Test - 1 Physics STD: XII TIME: 1.5 Hrs3 pages
- JIPMAT 2024 Maths Syllabus Topic Analysis 2021 2022 2023No ratings yetJIPMAT 2024 Maths Syllabus Topic Analysis 2021 2022 20239 pages
- UT Dallas Syllabus For cs2305.081 06u Taught by Sergey Bereg (sxb027100)No ratings yetUT Dallas Syllabus For cs2305.081 06u Taught by Sergey Bereg (sxb027100)2 pages
- PC Card Standard 8.0 Volume 5 Card Services SpecificationNo ratings yetPC Card Standard 8.0 Volume 5 Card Services Specification240 pages
- How To Connect To Wifi On Windows 10 PDFNo ratings yetHow To Connect To Wifi On Windows 10 PDF3 pages
- Christensen 等 - 2023 - A Machine Learning Approach to Volatility ForecastNo ratings yetChristensen 等 - 2023 - A Machine Learning Approach to Volatility Forecast48 pages
