Heat Transfer Updated 31 Aug 20211

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 66

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/323144799

Heat transfer introduction

Preprint · February 2018

CITATIONS READS

0 89,745

1 author:

Kevin Graham Harding


University of the Witwatersrand
146 PUBLICATIONS   1,154 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Environmental Assessment of Microbial Bioprocesses through Life Cycle Assessment (LCA) View project

The fatty acid analysis of microalgae for biodiesel production View project

All content following this page was uploaded by Kevin Graham Harding on 14 September 2021.

The user has requested enhancement of the downloaded file.


Heat transfer – Introduction

1 Heat transfer
1.1 Introduction
1.1.1 What is heat transfer?
Heat transfer is the process of the movement of energy due to a temperature difference. The
calculations we are interested in typically include determining the final temperatures of
materials or how long it takes for these materials to reach these temperatures. This can help
inform the level of insulation required to ensure heat is not lost from a system. Typically, heat
loss is proportional to a temperature gradient (driving force or potential).

Heat transfer can be achieved by conduction, convection or radiation.

1.1.2 Conduction
Conduction is the form of heat that exists due to direct contact without movement. A
temperature gradient within a substance causes a flow of energy from a hotter to a colder
region. These gradients can exist in solids, liquids, and gases; provided there is no movement
in the fluid phases, i.e., fluids which are not well mixed. Over time the temperature difference
will reduce and approach thermal equilibrium (same temperature). Conduction occurs in a
solid, liquid or gas; provided there is no bulk movement.

Examples of conduction include the end of a metal rod placed in a fire heating up from one end
to the other, hot coffee heating through the mug or the cooling of ice cream in the bowl it is
placed in.

Figure 1.1: Conduction – heat moves from the hot end to the cold end of a metal pole

1.1.3 Convection
Convection is the transfer of heat due to the bulk movement of fluids. As such convection only
applies to heat transfer within a fluid or between a solid and fluid but not the heat transfer
within a solid. This heat transfer is achieved by the movement of molecules within the fluid.
The term convection can refer to either mass transfer and/or heat transfer. Typically, when
referred to as ‘convection’, heat transfer is meant.

Convection is the sum of advection and diffusion:


Advection is the heat transported by the large-scale movement of currents in the fluid; and
Kevin Harding (kevin.harding@wits.ac.za) 1
School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Introduction

Diffusion is the random Brownian motion of individual particles in the fluid.

Examples of convection include the effect of hot air rising and falling (convection currents) or
the large-scale convection currents of the atmosphere and oceans.

Figure 1.2: Convection – movement of water as it heats in a pot

1.1.4 Radiation
Radiation is the transfer of energy due to electromagnetic waves when thermal energy is
converted by the movement of the charges of electrons and protons in the material. When a
body radiates, the energy comes from the entire depth of the body, not just the surface.
Radiation does not require a temperature gradient. A person standing some distance from the
source will still feel the effects of the heat, e.g., a person near a fire is heated by the fire, not
by the air surrounding them.

Examples include infra-red radiation such as, an incandescent light bulb emitting visible light,
the infrared radiation emitted by a common household radiator or electric heater, as well as the
sun heating the earth.

Video 1.1: Heat transfer – Introduction (YouTube)

Example 1.1: Keeping a cup of coffee warm

The power is going to be down for the next 5 hours. You have just enough time to boil a cup of water for a cup
of coffee. However, you don’t really want a cup of coffee right now – but would rather have it in 30 min time.
Your colleagues come up with several suggestions on how to ensure the coffee is as warm as possible when
you want to drink it. Unfortunately, they are not sure if these will make the coffee hotter or colder. For each
suggestion indicate which would keep the coffee hotter for when you want to drink it. Explain why.

Kevin Harding (kevin.harding@wits.ac.za) 2


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Conductive heat transfer

a) Pouring the (cold) milk in when the coffee has been made OR pouring the milk in just before drinking.
Assume the milk stays at the same temperature even with the power out due to good insulation of the fridge.
b) Covering the top of the coffee cup with something OR leaving the top of the cup open to the atmosphere.
c) Using a steel mug OR a ceramic mug.
d) Leaving a spoon in the coffee cup OR taking the spoon out.
e) Leaving the coffee near an open window where there is a slight draft OR where there is no wind.

Solution:
a) Pouring the cold milk in as soon as the coffee has been made will keep it warmer.
Pouring cold milk into the coffee after the coffee has been left to cool for some time results in a larger
temperature gradient and the coffee will cool down faster at the start.
b) Covering the top of the coffee cup will keep it warmer.
This will reduce convective heat transfer between surface of the coffee and the surrounding air.
c) Using a ceramic mug will keep it warmer.
Ceramic is a good insulator, while metal is a very good conductor. Convective heat loss from the coffee
through the walls of the mug to the surrounding air will be faster through metal walls than ceramic walls.
d) Taking the spoon out of the coffee mug will keep it warmer.
The spoon (assuming it is metal) is a good conductor and acts as a type of fin and increases the surface
area for heat transfer. This would increase the convective heat loss.
e) Leaving the coffee cup where there is no wind will keep it warmer.
A draft around the outside of the coffee mug will increase the rate of convective heat loss from the walls of
the mug and from the liquid surface. The draft moves heat away from the outer edges of the mug and the
liquid surface faster; thereby creating a larger temperature gradient. Heat will then be transferred from
the system more quickly.

1.2 Conductive heat transfer


1.2.1 Fourier’s Law
Conduction is governed by Fourier’s Law: The energy flux (rate of energy transfer per unit
area; W/m) is proportional to the temperature gradient, and inversely proportional to the
thickness of the object.

∆T
q′ ∝ 1.1
∆x

where:
q’ = Heat flux (W/m2)
T = Temperature difference (K)
x = direction in which there is a temperature gradient (direction of heat transfer) (m)

Including a proportionality constant, this can be written as:


dT
q′ = −k dx 1.2

where:
k = thermal conductivity (W/(m.K))

Or more conveniently:

Kevin Harding (kevin.harding@wits.ac.za) 3


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Conductive heat transfer

dT
q̇ = −k. A dx 1.3

where:
q̇ = Heat (W)
A = Cross sectional area (m2)

Thot
Tcold

Assumptions:
- Thot is at the same temperature across entire area;
- Use area perpendicular to the direction of heat transfer; and
- x is the direction from hot to cold (direction of temperature gradient).

1.2.2 Thermal conductivity


The constant, k, is the thermal conductivity of the material through which heat is transferring.
Simply, thermal conductivity is the property of a material to conduct heat. Materials with a
good heat transfer ability, e.g., metals, have a high conductivity, while those with a poor
conductivity (gases) have a low k-value (Table 1.1).

Table 1.1: Typical thermal conductivities

Material Thermal conductivity (W/(m.K))


Air 0.025
Wood 0.04 – 0.4
Hollow fill fibre insulation 0.042
Alcohols and oils 0.1 – 0.21
Polypropylene 0.25
Mineral oil 0.138
Rubber 0.16
LPG 0.23 – 0.26
Cement, Portland 0.29
Water (liquid) 0.6
Thermal grease 0.7 – 3
Thermal epoxy 1–7
Glass 1.1
Soil 1.5
Kevin Harding (kevin.harding@wits.ac.za) 4
School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Conductive heat transfer

Concrete, stone 1.7


Ice 2
Stainless steel 12.11 ~ 45.0
Lead 35.3
Aluminium 237 (pure)
120 — 180 (alloys)
Gold 318
Copper 401
Silver 429

In reality, k is not a constant but dependent on temperature:

Figure 1.3: Thermal conductivities of various materials

Good conductors have high thermal conductivity, e.g., copper.


Poor conductors, i.e., good insulators, have low thermal conductivity, e.g., asbestos.

1.2.3 Conduction through a wall


Case 1: Conduction through a single slab
Consider the heat transfer across a single slab.

Kevin Harding (kevin.harding@wits.ac.za) 5


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Conductive heat transfer

Th
Tc
k

L
x=0 x=L
where:
Th: Hot temperature on one side of the slab (K)
Tc: Cooler temperature on other side of the slab (K)
q̇ : Heat Flow (W)
A: Area perpendicular to heat flow (m2)
L: Length of the slab (m)
x: Direction in which heat flows (m)
k: Thermal conductivity of the slab (W/(m.K))

From Fourier’s Law:


dT
q̇ = −k. A dx 1.3

Assumptions:
- Area through slab is constant
- k is constant
- Steady state conditions. “Energy in one face of the slab” = “Energy out other face of
the slab”

This is a separable differential equation:


Rearranging:

dx = dT 1.4
−kA

Therefore:
x=L q̇ T
∫x=0 dx = ∫T c dT 1.5
−kA h

Since none of q̇ , k or A are functions of temperature or x, we can remove these from the
integral:
q̇ x=L T
∫ dx = ∫T c dT 1.6
−kA x=0 h
Kevin Harding (kevin.harding@wits.ac.za) 6
School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Conductive heat transfer

And integrate:

(L − 0) = (Tc − Th ) 1.7
−kA

Therefore:
kA(Th −Tc )
q̇ = 1.8
L

Example 1.2: Conduction through a single slab

An aluminium plate (k = 215 W/m°C) is heated to 300°C. If the heat flux is 8.6 MW/m2, how hot is the other
face, if the metal is 5 mm thick?

Solution:
dT
𝑞 ′ = −k
dx
Assuming constant k and A:
k(Th − Tc )
𝑞′ =
L
𝑞′L
∴ = Th − Tc
k
𝑞′L (8600000)(0.005)
∴ Tc = Th − = 300 −
k 215
∴ Tc = 100

Case 2: Conduction through 2 slabs


Consider the heat transfer across two slabs.

T1

Th
Tc
k1 k2

L1 L2
x=0 x = L1 x = L 1 + L2
where:
Th: Hot temperature on one side of the slab (K)
T1: Temperature at the end of slab 1 (K)
Tc: Cold temperature on the other side of slabs (K)
q̇ : Heat Flow (W)
Kevin Harding (kevin.harding@wits.ac.za) 7
School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Conductive heat transfer

A: Area perpendicular to heat flow (m2)


L1: Length of slab 1 (m)
L2: Length of slab 2 (m)
x: Direction in which heat flows (m)
k1: Thermal conductivity of slab 1 (W/(m.K))
k2: Thermal conductivity of the slab 2 (W/(m.K))

Assumptions:
- Area through slab is constant
- k1 ≠ k 2
- Steady state conditions; therefore, “Energy in one face of the slab” = “Energy out other face of
the slab”

From single slab calculation:


kA(Th −Tc )
q̇ = 1.9
L

Therefore:
k1 A(Th −T1 )
Slab 1: q̇ = 1.10
L1
k2 A(T1 −Tc )
Slab 2: q̇ = 1.11
L2

Re-arranging, this gives:


q̇ L1
Th − T1 = 1.12
k1 A
q̇ L2
T1 − Tc = 1.13
k2 A

Adding these equations, we are left with an expression without T1 and can calculate q̇ :
q̇ L1 q̇ L2
Th − T1 + T1 − Tc = + 1.14
k1 A k2 A
q̇ L1 q̇ L2
Th − T1 + T1 − Tc = + 1.15
k1 A k2 A
q̇ L1 L2
Th − Tc = ( + ) 1.16
A k1 k2
q̇ Th −Tc
= L L 1.17
A ( 1+ 2)
k1 k2

To determine T1, either of the original heat transfer equations can then be used.

Example 1.3: Conduction through two slabs

The wall of a furnace is constructed from an inner steel layer of 0.5 cm (k = 40 W/m.K) and a brick outer layer of 10 cm (k
= 2.5 W/mk). The inner surface temperature is 900K and the outside surface temperature is 460K. What is the temperature
between the steel and the brick?

Solution:
Assumptions:

Kevin Harding (kevin.harding@wits.ac.za) 8


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Conductive heat transfer

- Steady State (energy in = energy out)


- Linear heat transfer (ignore the effects of edges)

𝑑𝑇
From Fourier’s Law: 𝑞 = −𝑘𝐴
𝑑𝑥
𝑞 𝑇𝑖𝑛 −𝑇𝑜𝑢𝑡
For 2 slabs this simplifies to: = 𝐿1 𝐿2
𝐴 +
𝑘1 𝑘2

Therefore:
𝑞 (900 − 460) 440
= = = 10965.732 … 𝑊/𝑚2
𝐴 0.005 0.1 (0.000125 + 0.04)
+
40 2.5

If we let 𝑇1 be the temperature between the steel and the brick and:
𝑄.𝐿1
𝑇1 = 𝑇𝑖𝑛 − = 900 − (10965)(0.000125) = 898.629 … 𝐾
𝐴.𝑘1

The temperature between the brick and steel is approximately 625°C

Case 3: Conduction through n-slabs


Consider the heat transfer across multiple slabs.

T1 T2 T3 Tn-1

Th

k1 k2 kk33 ... kknn Tc

L1 L2 L3 Ln
x=0 x = L1 x = L 1 + L2 + L3
x = L 1 + L2 x = L 1 + L2 + L 3 + … + L n

where:
Th: Hot temperature on one side of the slab (K)
Ti: Temperature at the end of slab i (K)
Tc: Cold temperature on the other side of slabs (K)
q̇ : Heat Flow (W)
A: Area perpendicular to heat flow (m2)
Li: Length of slab i (m)
x: Direction in which heat flows (m)
ki: Thermal conductivity of slab i (W/(m.K))
n: Number of slabs

Assumptions:
Kevin Harding (kevin.harding@wits.ac.za) 9
School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Conductive heat transfer

- Area through slab is constant


- ki ≠ ki +1
- Steady state conditions; therefore “Energy in one face of the slab” = “Energy out other face of
the slab”

As before we can solve for the heat flow through each slab:
k1 A(Th −T1 )
Slab 1: q̇ = 1.18
L1
k2 A(T1 −T2 )
Slab 2: q̇ = 1.19
L2
k3 A(T2 −T3 )
Slab 3: q̇ = 1.20
L3
kn A(Tn−1 −Tc )
Slab n: q̇ = 1.21
Ln

Rearranging as before:
q̇ L1
Th − T1 = 1.22
k1 A
q̇ L2
T1 − T2 = 1.23
k2 A
q̇ L3
T2 − T3 = 1.24
k3 A
q̇ Ln
Tn−1 − Tc = 1.25
kn A

Adding these equations, we are left with an expression without Ti and can calculate q̇ :
q̇ L1 L L Ln
Th − Tc = ( + k2 + k3 + ⋯ + ) 1.26
A k1 2 3 kn

This is more commonly written as:


L L L Ln
Th − Tc = q̇ (k 1A + k 2A + k 3A + ⋯ + ) 1.27
1 2 3 kn A

Which can also be expressed as:


∆T
q̇ = Li 1.28
∑n
i=1 k A
i

Video 1.2: Heat transfer – Conduction through slabs (YouTube)

Kevin Harding (kevin.harding@wits.ac.za) 10


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Conductive heat transfer

Example 1.4: Conduction through multiple layers

On a chilly winter day, you decide to add on a few layers of clothes to stay warm. What heat flux do you
experience for each successive layer added as below? Assume your skin is at 36°C and the temperature outside
is at 4°C:
1) T-Shirt (k = 0.05 W/mK; Thickness = 0.75 mm)
2) Jersey 1 (k = 0.06 W/mK; Thickness = 2 mm)
3) Jersey 2 (k = 0.05 W/mK; Thickness = 2 mm)
4) Jacket (k = 0.005 W/mK; Thickness = 3 mm)
5) Overcoat (k = 0.02 W/mK; Thickness = 4.5 mm)

It is estimated that the body released between 90-140 W of heat. Further, the body has an average surface
area of between 1.6 and 1.9 m2. Given this information, what would be a reasonable number of layers to wear?

Solution:
∆T ∆T
q̇ = Li OR 𝑞 ′ = Li
∑n
i=1 ∑n
i=1
ki A ki

′ ∆T 32
1) 𝑞 = L1 = 0.75×10−3
= 2133.3 𝑊/𝑚2
k1 0.05
∆T 32
2) 𝑞 ′ = L1 L = 0.75×10−3 2×10−3
= 662.1 𝑊/𝑚2
+ 2 +
k1 k2 0.05 0.06
∆T 32
3) 𝑞 ′ = L1 L L = 0.75×10−3 2×10−3 2×10−3
= 362.2 𝑊/𝑚2
+ 2+ 3 + +
k1 k2 k3 0.05 0.06 0.05
∆T 32
4) 𝑞 ′ = L1 L L L = 0.75×10−3 2×10−3 2×10−3 3×10−3
= 46.5 𝑊/𝑚2
+ 2+ 3+ 4 + + +
k1 k2 k3 k4 0.05 0.06 0.05 0.005
′ ∆T 32
5) 𝑞 = L1 L L L L = 0.75×10−3 2×10−3 2×10−3 3×10−3 4.5×10−3
= 35 𝑊/𝑚2
+ 2+ 3+ 4+ 5 + + + +
k1 k2 k3 k4 k 5 0.05 0.06 0.05 0.005 0.02

From the average body data, it can be shown that the body releases between 47 and 87 W/m2. Therefore, is
would be a reasonable assumption that the average person would be most comfortable with four layers as
calculated.

Temperature profile through n-slabs


From the calculation of heat transfer in one slab it was show that:
kA(Th −Tc )
q̇ = 1.8
L

Rearranging:
q̇ L = kA(Th − Tc ) 1.29
q̇ L
= Th − Tc 1.30
kA
−q̇ L
Tc = + Th 1.31
kA

When 𝑦 = 𝑚𝑥 + 𝑐, plot of T vs. x gives a straight-line temperature profile (for a constant k,


at steady state). An insulator has a LOW thermal conductivity, therefore (-q̇ /k.A) is LARGE,
and the slope is great, i.e., not much temperature change. NOTE: Only applicable for a
constant heat transfer constant.

Kevin Harding (kevin.harding@wits.ac.za) 11


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Conductive heat transfer

k
Th

Tc

x
0 L

In the same way, this can be extended for n-slabs in series. The temperature profile in each
slab is linear (provided k is constant), with a different linear slope depending on the value of
k.

k1 k2 k3 ... kn
Th

Tc

x
x1 x2 x3 xn-1 xn
0

1.2.4 Conduction through a cylinder


Case 1: Conduction through a single cylinder
Consider the heat transfer across a single cylinder.

Kevin Harding (kevin.harding@wits.ac.za) 12


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Conductive heat transfer

Tin
ri
Tout
ro

where:
Tin: Temperature on the inside of the cylinder (K)
Tout: Temperature on the outside of the cylinder (K)
q̇ : Heat Flow (W)
ri: Inner radius of the cylinder (m)
ro: Outer radius of the cylinder (m)
k: Thermal conductivity of the slab (W/(m.K))

Assumptions:
- Cylinder thickness is constant throughout
- k is constant
- Steady state conditions; therefore, “Energy in one face of the slab” = “Energy out other face of
the slab”
- No axial heat flow
- Area through which heat flows is NOT constant!

From Fourier’s Law:


dT
q̇ = −k. A 1.3
dx

But we do not have an x-direction in a cylinder. Therefore, convert Fourier’s Law into
cylindrical equivalent:

dT
q̇ = −k. A dr 1.32

The area is that which is perpendicular to the direction of heat transfer. Therefore:
A = 2π.r.L (circumference of a cylinder × length) 1.33

And Fourier’s Law becomes:

Kevin Harding (kevin.harding@wits.ac.za) 13


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Conductive heat transfer

dT
q̇ = −k. (2πrL) dr 1.34

Assuming k is constant and solving the separable differential equation:


q̇ ro dr T
∫ = ∫T out dT 1.35
−k2πL ri r in

Integrating:
q̇ r
ln ( ro ) = Tout − Tin 1.36
−k2πL i

Or:
q̇ r
Tin − Tout = k2πL ln ( ro ) 1.37
i

Video 1.3: Heat transfer – Conduction through cylinders (YouTube)

Example 1.5: Conduction through a cylinder

Water enters a pipe at a temperature of 60°C. Assuming the pipe is made of stainless steel (k = 30 W/m 2) and
that the pipe has an inner and outer diameter of 20 and 25 cm respectively, what is the outer temperature of
the pipe if heat is lost at a rate of 15 kJ/m.s? Assume the temperature is constant down the length of the pipe.

Solution:
𝑞̇ 𝑟𝑜
𝑇𝑖𝑛 − 𝑇𝑜𝑢𝑡 = 𝑙𝑛 ( )
𝑘2𝜋𝐿 𝑟𝑖
𝑞̇ 𝑟𝑜 15000 25
𝑇𝑜𝑢𝑡 = 𝑇𝑖𝑛 − 𝑙𝑛 ( ) = 60 − 𝑙𝑛 ( )
𝑘2𝜋𝐿 𝑟𝑖 (30)2𝜋 20
𝑇𝑜𝑢𝑡 = 42.2°𝐶

Example 1.6: Conduction through a sphere

Liquid ammonia at 0.5 atm (-45°C) is stored in a spherical container of ID 1 foot and OD 18 inches. The walls
of the sphere are filled with insulation (k = 0.0692 W.m-1.K-1), and the outer surface is at room temperature of
20°C. What is the heat transfer through the sphere? Assuming the temperature in the sphere remains constant,
how much ammonia is lost per hour by evaporation? Hvap NH3: 1390 kJ/kg

Solution:
Fourier’s law:
𝑑𝑇
𝑞̇ = −𝑘𝐴
𝑑𝑟
𝐴 = 𝑠𝑢𝑟𝑓𝑎𝑐𝑒 𝑎𝑟𝑒𝑎 𝑜𝑓 𝑠𝑝ℎ𝑒𝑟𝑒 = 4𝜋𝑟 2
Kevin Harding (kevin.harding@wits.ac.za) 14
School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Conductive heat transfer

𝑞̇ 𝑑𝑟
∴ = 𝑑𝑇
−4𝜋𝑘 𝑟 2
𝑟𝑖 𝑇𝑖
𝑞̇ 𝑑𝑟
∫ = ∫ 𝑑𝑇
−4𝜋𝑘 𝑟 2
𝑟𝑜 𝑇𝑜
𝑞̇ −1 𝑟𝑖
∴ [ ] = 𝑇𝑖 − 𝑇𝑜
−4𝜋𝑘 𝑟 𝑟𝑜
𝑞̇ 1 1
( − ) = 𝑇𝑖 − 𝑇𝑜
4𝜋𝑘 𝑟𝑖 𝑟𝑜
4𝜋𝑘(𝑇𝑖 − 𝑇𝑜 )
𝑞̇ =
1 1
( − )
𝑟𝑖 𝑟𝑜
4𝜋(0.0692𝑊𝑚−1 𝐾 −1 )(228𝐾 − 293𝐾)
𝑞̇ = = −25.84 𝑊 = −25.84 𝐽𝑠 −1 = −93.024 𝑘𝐽ℎ−1
1 1
( − )
0.1524𝑚 0.2286𝑚
−1
𝑞̇ 93.024 𝑘𝐽ℎ
𝑚̇ = = = 0.06692 𝑘𝑔ℎ−1 = 66.92 𝑔ℎ−1
̂
∆𝐻𝑣𝑎𝑝,𝑁𝐻3 1390 𝑘𝐽𝑘𝑔−1

1.2.5 Conduction through an irregular shape


When the heat flows through a material which is not a uniform shape, the area term can no
longer be taken out of the integral term as was done previously.

From the Fourier form as before:


x=L q̇ T
∫x=0 dx = ∫T c dT 1.5
−kA h

Now:
q̇ x=L 1 T
∫ dx = ∫T c dT 1.38
−k x=0 A h

Typically, area will be given as a function of x, e.g., A = x + 3; or A = 2x2 – 4 or others and


needs to be integrated appropriately.

Example 1.7: Conduction when area is not constant

In a building in a hot climate (45°C), the inside is cooled to 15°C. The walls of the building include an irregular
shaped steel girder which forms a structural member of the wall. If no heat is lost or gained to or from the
insulation through the sides of the girder, calculate the heat flow from the room from a single girder.
Given: Thickness of the wall, 20 cm; k = 45 W.m-1.K-1; A = -50.x2 +3

Solution:
𝑑𝑇
𝑞̇ = −𝑘. 𝐴
𝑑𝑥
𝑑𝑇
𝑞̇ = −𝑘. (−50𝑥 2 + 3)
𝑑𝑥
𝑞̇
𝑑𝑥 = 𝑑𝑇
−𝑘(−50𝑥 2 + 3)
𝑥=𝐿 𝑇ℎ
𝑞̇
∫ 𝑑𝑥 = ∫ 𝑑𝑇
−𝑘(−50𝑥 2 + 3)
𝑥=0 𝑇𝑐

Kevin Harding (kevin.harding@wits.ac.za) 15


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Conductive heat transfer

𝑥=𝐿 𝑇ℎ
𝑞̇ 1
∫ 𝑑𝑥 = ∫ 𝑑𝑇
𝑘 50𝑥 2 − 3
𝑥=0 𝑇𝑐
𝐿
𝑞̇ 2 2 × 3𝑥
[ 𝑡𝑎𝑛−1 ] = 𝑇ℎ − 𝑇𝑐
𝑘 √4 × 50 × 3 √4 × 50 × 3 0
𝑞̇ 2 6𝑥 0.2
[ 𝑡𝑎𝑛−1 ] = 45 − 15 = 30
45 √600 √600 0
(30 × 45)
𝑞̇ =
2 6(0.2) 2 6(0)
( 𝑡𝑎𝑛−1 )− ( 𝑡𝑎𝑛−1 )
√600 √600 √600 √600
1350
𝑞̇ =
1.2
(0.081 𝑡𝑎𝑛−1 )− 0
24.5
1350
𝑞̇ = = 810.6 𝑊
1.665

Example 1.8: Conduction through irregular shape (1)

The figure shows the plan view of a steel girder forming a structural member in the wall of a cold room. The
thickness of the wall is 30 cm. The width of the girder is 10 cm on the inside and 20 cm on the outside. Assuming
no heat transfer across the insulation on the sides, calculate the heat flow from the cold room per metre of the
girder. k = 45W.m-1.K-1

To = 25C

Ti = 5C
Solution:
Drawing the system in 3-D, with heat flowing perpendicular to the diagram, the surface area perpendicular to
the flow of heat will be rectangular, with the area increasing as the heat flows from the cold room to the outside
because of the tapering sides of the girder (the girder boundaries become increasingly further apart).

20 cm To = 25C

30 cm
10 cm x

Ti = 5C
Area

We need an expression to describe how the rectangular surface area perpendicular to the direction of heat
flow changes with distance through the wall. The sides of the girder can be described using the equation for
a straight line. Defining an x-y coordinate system such that at x = 0, y = 10 cm , and at x = 30 cm, y = 20 cm.

Given two points and knowing that y = mx + c, we can find:


1
𝑏(𝑥) = 𝑥 + 0.10, where b is the breadth of the rectangle
3

Kevin Harding (kevin.harding@wits.ac.za) 16


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Conductive heat transfer

Now from Fourier’s law:


𝑑𝑇
𝑞̇ = −𝑘𝐴
𝑑𝑥
With: 𝐴 = 𝑙 × 𝑏(𝑥)
𝑥𝑜 𝑇𝑜
𝑞̇ 𝑑𝑥
∴ ∫ = ∫ 𝑑𝑇
−𝑘 1 𝑥 + 0.10
𝑥𝑖 3 𝑇𝑖
Let u = 1/3 x + 10
du = 1/3 dx
3du = dx
And at x = 0, u = 0.10
at x = 0.30, u = 0.20
0.20 𝑇𝑜
𝑞̇ 3
∴ ∫ 𝑑𝑢 = ∫ 𝑑𝑇
−𝑘 𝑢
0.10 𝑇𝑖
3𝑞̇ 0.20
. 𝑙𝑛 ( ) = 𝑇𝑂 − 𝑇𝑖
−𝑘 0.10
−𝑘(𝑇𝑂 − 𝑇𝑖 ) −45𝑊𝑚−1 𝐾 −1 (298𝐾 − 278𝐾)
𝑞̇ = = = −433 𝑊. 𝑚−1
3 × 𝑙𝑛 (2) 3 × 𝑙𝑛 (2)
The sign is negative because heat flow is conventionally down the temperature gradient (i.e. from hot to cold)

1.2.6 Conduction when thermal conductivity is not constant


Typically, thermal conductivity is not constant, but a function of temperature. As such:

From the Fourier form as before:


x=L q̇ T
∫x=0 dx = ∫T c dT 1.5
−kA h

Now:
q̇ x=L T
− A ∫x=0 dx = ∫T c kdT 1.39
h

Typically, k will be given as a function of temperature, e.g., k = T + 30; or k = 2T2 – 4 or others


and needs to be integrated appropriately.

Example 1.9: Conduction with variable conductive heat transfer coefficient (1)

A furnace wall has a thickness of 0.5 ft, with inner and outer temperatures of 200°F and 100°F respectively.
The thermal conductivity is given by: K(BTU/h.ft.°F) = 0.1 + 0.001T(°F).
Calculate the heat transfer per square foot through the furnace wall.

Solution:
𝑑𝑇
𝑞̇ = −𝑘𝐴
𝑑𝑥
𝑥 𝑇2
𝑞̇
∫ 𝑑𝑥 = − ∫ 𝑘𝑑𝑇
0 𝐴 𝑇1

Kevin Harding (kevin.harding@wits.ac.za) 17


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Conductive heat transfer

𝑞̇ 𝑥 𝑇2
∫ 𝑑𝑥 = − ∫ (0.1 + 0.001 𝑇)𝑑𝑇
𝐴 0 𝑇1

𝑞̇ 0.001𝑇2 2 0.001𝑇1 2
(𝑥 − 0) = − [(0.1𝑇2 + ) − (0.1𝑇1 + )]
𝐴 2 2
𝑞̇ 0.001(100)2 0.001(200)2
(0.5) = − [(0.1(100) + ) − (0.1(200) + )]
𝐴 2 2
𝑞̇ ((10) − (40))
=−
𝐴 0.5
𝑞̇
= 60 BTU/hr.ft2
𝐴

Example 1.10: Conduction with variable conductive heat transfer coefficient (2)

A house in a cold climate with a flat insulated rook is heated and maintained at 23°C. The roof consists of
three layers that are in contact with one another: ceiling board (1.50 cm thick), insulation (40.0 cm thick) and
a waterproof layer on the outside (0.5 cm thick). The thermal conductivities of the ceiling board, insulation
and waterproof material are k1 = 0.690 W.m-1.K-1, k2 = 85.0 W.m-1.K-1 and k3 = 377 W.m-1.K-1. Lying on the
waterproof material is a layer of snow, which gets thicker as the snow falls. Due to compression of the snow
in contact with the roof, the thermal conductivity is a function of the thickness of the snow according to:
1
𝑘𝑠𝑛𝑜𝑤 =
𝑎𝑙 + 𝑏
Where:
l is the thickness of the snow in m, with a = 15.6 K.W-1 and b = 1.24 m.K.W-1

The film heat transfer coefficient on the inside is 18 W.m-2.K-1 and on the outside it is 50 W.m-2.K-1. If heat is
lost at a rate of 50.0 W.m-2 and the outside temperature is -5°C, what is the thickness of the snow on the roof
and the overall heat transfer coefficient?
Ta
h0 T4
L4

L (m)

L3 = 42cm
3
L2 = 41.5cm
2

1 L1 = 1.5cm

0cm
hi

Solution:
𝑞̇⁄ −2
𝐴 = 50.0 𝑊. 𝑚
𝑇𝑎 = −5℃ = 268𝐾
For the inner film:
𝑞̇ 𝑞̇
= ℎ𝑖 (𝑇𝑖 − 𝑇0 ) ⇒ 𝑇0 = 𝑇𝑖 −
𝐴 𝐴ℎ𝑖
∴ 𝑇0 = 293.22𝐾
For the ceiling board:

Kevin Harding (kevin.harding@wits.ac.za) 18


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Conductive heat transfer

𝑞̇ 𝑑𝑇 𝑞̇
= −𝑘1 ⇒ (𝐿 ) = 𝑇1 − 𝑇0
𝐴 𝑑𝐿 −𝑘1 𝐴 1
∴ 𝑇1 = 292.13𝐾
For the insulation:
𝑞̇ 𝑑𝑇 𝑞̇
= −𝑘2 ⇒ (𝐿 − 𝐿1 ) = 𝑇2 − 𝑇1
𝐴 𝑑𝐿 −𝑘2 𝐴 2
∴ 𝑇2 = 291.89𝐾
For the waterproofing:
𝑞̇ 𝑑𝑇 𝑞̇
= −𝑘3 ⇒ (𝐿 − 𝐿2 ) = 𝑇3 − 𝑇2
𝐴 𝑑𝐿 −𝑘3 𝐴 3
∴ 𝑇3 = 291.89𝐾
For the snow:
𝑞̇ 1 𝑑𝑇
=−
𝐴 (𝑎𝐿 + 𝑏) 𝑑𝐿
𝑞̇
⇒ − (𝑎𝐿 + 𝑏)𝑑𝐿 = 𝑑𝑇
𝐴
Integrating, with L = thickness of the snow:
𝑞̇ 𝑎𝐿2
− ( + 𝑏𝐿) = 𝑇4 − 𝑇3 (1)
𝐴 2
For the outer film:
𝑞̇ 𝑞̇
= ℎ0 (𝑇4 − 𝑇𝑎 ) ⇒ 𝑇4 = 𝑇𝑎 +
𝐴 𝐴ℎ0
∴ 𝑇4 = 269𝐾
From (1):
15.6
−50𝑊. 𝑚−2 (( 𝐾𝑊 −1 ) 𝐿2 + (1.24 𝑚𝐾𝑊 −1 )𝐿) = 269𝐾 − 291.89𝐾
2
(7.8 𝐾𝑊 −1 )𝐿2 + (1.24 𝑚𝐾𝑊 −1 )𝐿 − 0.4578 𝑚𝐾. 𝑊 −1 = 0

Using the quadratic formula, and the positive result:


L =0 .1755m
The snow is 17.55 cm thick

And from 𝑞 = 𝑈∆𝑇


𝑞
𝑈=
∆𝑇
50
𝑈=
23 − (−5)
𝑈 = 1.78 𝑊/𝑚2 𝐾

1.2.7 Thermal Resistance


From the slab calculations, it was shown that:
L L L Ln
Th − Tc = q̇ (k 1A + k 2A + k 3A + ⋯ + ) 1.40
1 2 3 kn A

Or:
Th −Tc
q̇ = L1 L2 L L 1.41
( + + 3 +⋯+ n )
k1 A k2 A k3 A kn A
∆T
q̇ = Li 1.42
∑n
i=1 k A
i

Kevin Harding (kevin.harding@wits.ac.za) 19


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Conductive heat transfer

This can be simplified to:


∆T
q̇ = 1.43
R

where:
T: Temperature difference (K)
R: Thermal resistance (K/W)

Li
For slab calculations, R = ∑ni=1 , however, R can represent the resistance for any range of
ki A
shapes.

It is possible to relate thermal resistance to the resistance in an electrical circuit. In this way
the resistance through solid slabs can be represented as below.

R1 R2 R3

Example 1.11: Conduction through an irregular shape (2)

A circular pipe of 20 cm is enclosed centrally in a square section insulator of side 36 cm. The thermal
conductivity of the material is given as 8.5W/mK. The inside surface is at 200°C, while the outside is at 30°C.
Determine the heat flow for a length of 5 m.

1 1 1.08𝑎
Given: 𝑅 = [ 𝑙𝑛 ]
2𝜋𝑙 𝑘 2𝑟

where:
a = side length of length of square section
l = length
r = radius of inner portion

k = 8.5 W/m.K
30°C

0.2 m
b=5m
200°C

a = 0.36m

Solution:
∆𝑇
𝑞̇ =
𝑅
1 1 1.08 × 0.36
𝑅= [ 𝑙𝑛 ] = 0.002489362 °C/W
2𝜋 × 5 8.5 0.2
2.
2
∆𝑇 200 − 30
∴ 𝑞̇ = = = 68290.59 = 68 000 W
𝑅 0.002489 …

Kevin Harding (kevin.harding@wits.ac.za) 20


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Convective heat transfer

Example 1.12: Conduction through an irregular shape (3)

A pipe of 30 cm diameter is insulated by a material of thermal conductivity 0.45 W/mK. Due to space
restrictions, the pipe is placed slightly off-centre (5cm), resulting in a portion of insulation thicker than the rest
of the pipe (see diagram). The inner surface is at 250°C and the outer surface is at 60°C. Determine the heat
loss for a 5 m length of pipe.

1 √[(𝑟2 +𝑟1 )2 −𝑒 2 ]+√[(𝑟2 −𝑟1 )2 −𝑒 2 ]


Given: 𝑅 = 𝑙𝑛
2𝜋.𝑘.𝑙 √[(𝑟2 +𝑟1 )2 −𝑒 2 ]−√[(𝑟2 −𝑟1 )2 −𝑒 2 ]
where:
e = eccentricity, m

60°C

0.15 m

0.05 m
0.3 m

250°C

k = 0.45 W/m.K

Solution:
∆𝑇
𝑞̇ =
𝑅
1 √[(0.15 + 0.3)2 − 0.052 ] + √[(0.15 − 0.3)2 − 0.052 ]
𝑅= 𝑙𝑛
2𝜋 × 0.45 × 5 √[(0.15 + 0.3)2 − 0.052 ] − √[(0.15 − 0.3)2 − 0.052 ]
1 0.4472 + 0.14142
𝑅= 𝑙𝑛 = 0.046325 ℃/W
2𝜋 × 0.45 × 5 0.4472 − 0.14142
250 − 60
∴ 𝑞̇ = = 4101 W
0.046325

1.3 Convective heat transfer


1.3.1 Newton’s Law of Cooling
Convection (film) heat transfer occurs through the bulk movement of fluid within a fluid or
between a solid and a fluid.

Consider the convective heat from surface.

Kevin Harding (kevin.harding@wits.ac.za) 21


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Convective heat transfer

Ts

Tb

Solid surface
Approximation: Linear temperature flow
through boundary layer

Film of fluid; stationary


(boundary layer)

where:
δ: thickness of stationary boundary layer (m)
Ts: Surface temperature (K)
Tb: Bulk temperature (K)

If we assume that there is no movement in a boundary layer between the slab and bulk fluid (δ
– Greek delta), we can approximate the heat transfer to Fourier’s Law:
kA(Ts −Tb )
q̇ = 1.44
δ

Defining a heat transfer coefficient as h:


k
h= 1.45
δ

The equation for heat transfer becomes:


q̇ = hA(Ts − Tb ) 1.46
q̇ = h. A. ∆T 1.47

where:
h: Heat Transfer Coefficient (W/m2.K)

This is known as Newton’s Law of Cooling:

Note:
It is harder to measure wall temperatures compared to the bulk fluid temperatures. Therefore,
the equations are developed to use the bulk fluid temperatures and not intermediate wall
temperatures.

Kevin Harding (kevin.harding@wits.ac.za) 22


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Convective heat transfer

1.3.2 Convection from a wall


Case 1: Convection from a single slab:
Consider the convective heat from a single slab (ignoring the heat transfer through the wall).

T1

Tc

hc

where:
T1: Temperature at the end of slab 1 (K)
T2: Temperature at the end of slab 2 (K)
Tc: Cold temperature (K)
q̇ : Heat Flow (W)
A: Area perpendicular to heat flow (m2)
hc: Heat Transfer coefficient on the cold side (W/m2.K)

Assumptions:
- The fluid has a heat transfer coefficient (h)
- Area through convective area is constant
- Steady state conditions, therefore, “Energy in” = “Energy out”

Convective heat transfer is given as:


q̇ = hA(Ts − Tb ) 1.46
Using the nomenclature here:
q̇ = h𝑐 A(T1 − Tc ) 1.48

Example 1.13: Convective heat transfer

A stainless-steel plate is maintained at 150°C, while the surrounding temperature is 20°C. The heat transfer
coefficient is given as 11.2 W/m2.K. What is the convective heat loss per square meter?

Solution:
𝑞̇ = ℎ𝐴(𝑇2 − 𝑇1 )
𝑞̇
= ℎ(𝑇2 − 𝑇1 )
𝐴

Kevin Harding (kevin.harding@wits.ac.za) 23


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Convective heat transfer

𝑞̇
= 11.2(150 − 20) = 1456 𝑊/𝑚2 = 1.456 𝑘𝑊/𝑚2
𝐴

Case 2: Convection on either side of 2 slabs:


Consider the heat across two slabs, considering both conduction and convection.

T0 T1 T2

Th
Tc

hh k1 k2 hc

L1 L2
x=0 x = L1 x = L1 + L2

where:
Th: Hot temperature on one side of the slab (K)
T0: Temperature at the start of slab 1 (K)
T1: Temperature at the end of slab 1 (K)
T2: Temperature at the end of slab 2 (K)
Tc: Cold temperature on the other side of slabs (K)
q̇ : Heat Flow (W)
A: Area perpendicular to heat flow (m2)
L1: Length of slab 1 (m)
L2: Length of slab 2 (m)
x: Direction in which heat flows (m)
k1: Thermal conductivity of slab 1 (W/(m.K))
k2: Thermal conductivity of the slab 2 (W/(m.K))
hh: Heat Transfer coefficient on the hot side (W/m2.K)
hc: Heat Transfer coefficient on the cold side (W/m2.K)

Assumptions:
- The slabs have thermal conductivities (k); the fluids have heat transfer coefficients (h)
- Area through the slabs is constant
- k1 ≠ k2 (k1 and k2 are constant)
- Steady state conditions, therefore, “Energy in one face of the slab” = “Energy out other
face of the slab”

Kevin Harding (kevin.harding@wits.ac.za) 24


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Convective heat transfer

Heat transfer equations from inside to outside:


For inside convective film: q̇ = hh A(Th − T0 ) 1.49
k1 A(T0 −T1 )
Slab 1 (conduction): q̇ = 1.50
L1
k2 A(T1 −T2 )
Slab 2 (conduction): q̇ = 1.51
L2
For outside film: q̇ = hc A(T2 − Tc ) 1.52

Rearranging for each in terms of temperature:



Th − T0 = 1.53
hh A
q̇ L1
T0 − T1 = 1.54
k1 A
q̇ L2
T1 − T2 = 1.55
k2 A

T2 −Tc = 1.56
hc A

Then adding the last four equations, we are left with an expression without Tc and are able to
calculate q̇ :
q̇ q̇ L1 q̇ L2 q̇
Th − T0 + T0 − T1 + T1 − T2 + T2 −Tc = +k +k +h 1.57
hh A 1A 2A cA
q̇ q̇ L1 q̇ L2 q̇
Th − T0 + T0 − T1 + T1 − T2 + T2 −Tc = +k +k +h 1.58
hh A 1A 2A cA
q̇ 1 L L 1
Th −Tc = ( + k1 + k 2 + h ) 1.59
A hh 1 2 c
q̇ Th −Tc
= 1 L1 L2 1
1.60
A ( + + + )
hh k1 k2 hc

Given the area, heat transfer coefficients ad thermal conductivities, only need to know inner
and outer temperatures (NOT any of the inside temperatures) to solve for q.

Example 1.14: Conduction and convection through a furnace wall

The inner surface of a 23 cm brick furnace wall is kept at 820°C, while the temperature of the outer surface is
180°C.

a) Calculate the heat loss from the furnace, given a thermal conductivity of the brick of 0.865 W.m-1.K-1.

b) A further 23 cm is added to the furnace wall, consisting of insulating brick with a thermal conductivity of
0.26W.m-1.K-1. Calculate the reduction in the heat loss, and the maximum temperature to which the
insulating brick is exposed. Assume that the inner temperature remains at 820°C. The temperature of the
surrounding air is 20°C.

c) Re-calculate the heat loss and the maximum temperature to which the insulating is exposed, but allow for
a convective heat transfer coefficient on the outer surface of 11.9 W.m-2.K-1. Assume that the inner
temperature remains at 820°C, and the temperature of the surrounding air is 20°C.
d) How do the answers compare when convection is included?

Solution:

Kevin Harding (kevin.harding@wits.ac.za) 25


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Convective heat transfer

𝑑𝑇
a) 𝑞̇ = −𝑘𝐴
𝑑𝑥

𝐿 𝑇𝑜
𝑞̇
∴ ∫ 𝑑𝑥 = ∫ 𝑑𝑇
−𝑘𝐴
0 𝑇𝑖
𝑞̇ 𝑘 0.865 𝑊𝑚−1 𝐾 −1
= (𝑇𝑖 − 𝑇0 ) = (1093𝐾 − 453𝐾) = 2407 𝑊. 𝑚−2
𝐴 𝐿 0.23𝑚
b) For wall 1:

𝑑𝑇
𝑞̇ = −𝑘1 𝐴
𝑑𝑥
𝐿1 𝑇𝑤
𝑞̇
∴ ∫ 𝑑𝑥 = ∫ 𝑑𝑇
−𝑘1 𝐴
0 𝑇𝑖
𝑞̇ 𝐿1
⇒ = 𝑇𝑖 − 𝑇𝑤
𝑘1 𝐴
𝑞̇ 𝑘1
= (𝑇𝑖 − 𝑇𝑤 ) (1)
𝐴 𝐿1
For wall 2:
𝑑𝑇
𝑞̇ = −𝑘2 𝐴
𝑑𝑥
𝑞̇ 𝑘
= 2 (𝑇𝑤 − 𝑇0 ) (2)
𝐴 𝐿2
Equating (1) and (2)
𝑘1 𝑘2
(𝑇 − 𝑇𝑤 ) = (𝑇𝑤 − 𝑇0 )
𝐿1 𝑖 𝐿2
𝑘 𝑇 𝑘 𝑇 0.865 𝑊. 𝑚−1 𝐾 −1 . 1093 𝐾 0.26 𝑊𝑚−1 . 𝐾 −1 . 293 𝐾
( 1 𝑖 + 2 𝑜) ( + )
𝐿1 𝐿2 0.23𝑚 0.23𝑚
𝑇𝑤 = =
𝑘 𝑘 0.865 𝑊. 𝑚−1 . 𝐾 −1 0.26 𝑊. 𝑚−1 . 𝐾 −1
( 1 + 2) ( + )
𝐿1 𝐿2 0.23𝑚 0.23𝑚
𝑇𝑤 = 908 𝐾 = 635℃ is the maximum temperature to which the insulating brick is exposed

From (2):
𝑞̇ 0.26 𝑊. 𝑚−1 . 𝐾 −1
= (908 𝐾 − 293𝐾) = 695.2 𝑊. 𝑚−2
𝐴 0.23 𝑚
∆𝑞 = 2407 𝑊. 𝑚−2 − 695 𝑊. 𝑚−2 = 1712 𝑊. 𝑚−2 reduction in heat loss
c) For wall 1:
𝑑𝑇
𝑞̇ = −𝑘1 𝐴
𝑑𝑥
𝑞̇ 𝑘1
= (𝑇𝑖 − 𝑇𝑤 ) (1)
𝐴 𝐿1
For wall 2:
𝑞̇ 𝑘2
= (𝑇𝑤 − 𝑇𝑓 ) (2)
𝐴 𝐿2
For the film:
𝑞̇ = ℎ𝐴(𝑇𝑓 − 𝑇0 )
𝑞̇
𝑇𝑓 − 𝑇0 = (3)
ℎ𝐴
Add (1) + (2):
𝑞̇ 𝐿1 𝐿2
𝑇𝑖 − 𝑇𝑓 = ( + ) (4)
𝐴 𝑘1 𝑘2
Add (3) + (4)
𝑞̇ 𝐿1 𝐿2 1
𝑇𝑖 − 𝑇0 = ( + + )
𝐴 𝑘1 𝑘2 ℎ

Kevin Harding (kevin.harding@wits.ac.za) 26


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Convective heat transfer

𝑞̇ 𝑇𝑖 − 𝑇0
∴ =
𝐴 (𝐿1 + 𝐿2 + 1)
𝑘1 𝑘2 ℎ
𝑞̇ 1093𝐾 − 293𝐾
= = 648 𝑊𝑚−2
𝐴 ( 0.23𝑚 0.23𝑚 1
+ + )
0.896 𝑊𝑚−1 𝐾 −1 0.26 𝑊𝑚−1 𝐾 −1 11.9 𝑊𝑚−2 𝐾 −1
From (2):
𝑞̇ 𝐿2
𝑇𝑤 = + 𝑇𝑓
𝑘2 𝐴
From (3):
𝑞̇
𝑇𝑓 = + 𝑇0
ℎ𝐴
𝑞̇ 𝐿2 𝑞̇ 𝑞̇ 𝐿2 1
∴ 𝑇𝑤 = + = 𝑇0 + ( + )
𝑘2 𝐴 ℎ𝐴 𝐴 𝑘2 ℎ
0.23𝑚 1
∴ 𝑇𝑤 = 293𝐾 + 648 𝑊𝑚−2 × ( −1 −1
+ ) = 920.68𝐾 = 648℃
0.26 𝑊𝑚 𝐾 11.9 𝑊𝑚−2 𝐾 −1
𝑞̇ 648 𝑊𝑚−2
𝑇𝑓 = + 𝑇0 = + 293𝐾 = 347.45 = 74.5℃
ℎ𝐴 11.9 𝑊𝑚−2 𝐾 −1
The maximum temperature is 648°C.

d) The heat loss was 695.2 W/m2 neglecting the convective film heat transfer on the outer surface of the
furnace wall. Taking the convective heat transfer into account (as calculated above), there is a 47.2 W/m 2
reduction in heat loss. The film acts as an additional insulating layer.

Reminder:
∆T
q̇ = 1.61
R

For the above equation:


∆T
q̇ = 1 L L 1 1.62
( + 1+ 2+ )
hh A k1A k2A hc A
∆T
q̇ = ( R1 +R2 +R3 +R4 )
1.63

where:
Ri : individual thermal resistances for different layers (K/W)

Compare this to U as defined later.

Example 1.15:

Solution

Case 3: Convection on either side of n-slabs:


Consider the heat transfer across multiple slabs.

Kevin Harding (kevin.harding@wits.ac.za) 27


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Convective heat transfer

T0 T1 T2 T3 Tn-1 Tn

Th
Tc

hh k1 k2 k3 kn hc

L1 L2 L3 L3
x=0 x = L1 x = L 1 + L2 + L3
x = L1 + L2 x = L 1 + L2 + L 3 + … + Ln

Using similar derivations as above:


q̇ Th −Tc
= 1 Li 1 1.64
A ( +∑n
hh i=1k +hc )
i

Example 1.16:

Solution

1.3.3 Convection (and conduction) through a cylinder


Case 1: Convection (and conduction) through concentric cylinders
Consider the heat transfer across multiple layers of a cylinder.

Kevin Harding (kevin.harding@wits.ac.za) 28


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Convective heat transfer

Insulation
Pipe wall

Tout T2 T1 T0 Tin

r2
hin
r0
k1
k2 r1
hout

where:
Tin: Temperature on the inside of the cylinder (K)
Tout: Temperature on the outside of the cylinder (K)
q̇ : Heat Flow (W)
r0: Inner radius of the cylinder (m)
r1: Outer radius of the 1st cylinder (pipe) (m)
r2: Outer radius of the 2nd cylinder (insulation) (m)
k1: Thermal conductivity of cylinder 1 (W/(m.K))
k2: Thermal conductivity of cylinder 2 (W/(m.K))
hin: Heat Transfer coefficient on the inside (inside) (W/m2.K)
hout: Heat Transfer coefficient on the outside (outside) (W/m2.K)

Assumptions:
- Cylinder thickness is constant throughout

Kevin Harding (kevin.harding@wits.ac.za) 29


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Convective heat transfer

- Steady state conditions, therefore, “Energy in one face of the slab” = “Energy out other
face of the slab”
- Area through which heat flows is NOT constant!
- k1 ≠ k2
- No axial heat transfer

Convection on inner surface:


q̇ = hin A0 (Tin − T0 ) 1.65
A0 = 2πr0 L 1.66

Area is the area in contact with the film (we don’t know how thick the film is so can’t use any
other area)
∴ q̇ = hin 2πr0 L(Tin − T0 ) 1.67

Tin − T0 = 1.68
hi 2πr0 L

Conduction across pipe:


dT
q̇ = −k1 . (2πrL) dr 1.69

Assuming k is constant and solving the separable differential equation:


q̇ r1 dr T
∴ ∫ = ∫T 1 dT 1.70
−k1 2πL r 0 r 0

Integrating:
q̇ r
T0 − T1 = k ln (r1 ) 1.71
1 2πL 0

Conduction across insulation:


q̇ r
As above for conduction through pipe: T1 − T2 = k ln (r2 ) 1.72
2 2πL 1

Convection on outer surface:


As for convection on inner surface:
q̇ = hout 2πr2 L(T2 − Tout ) 1.73
Therefore:

T2 − Tout = 1.74
ho 2πr2 L

Consolidating for each layer:



Tin − T0 = 1.75
hin 2πr0 L
q̇ r1
T0 − T1 = k ln (r ) 1.76
1 2πL 0
q̇ r
T1 − T2 = k ln (r2 ) 1.77
2 2πL 1

Kevin Harding (kevin.harding@wits.ac.za) 30


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Convective heat transfer


T2 − Tout = 1.78
hout 2πr2 L

Adding equations these four equations:


q̇ q̇ r q̇ r
Tin − T0 + T0 − T1 + T1 − T2 + T2 − Tout = +k ln (r1 ) + k ln (r2 ) +
hin 2πr0 L 1 2πL 0 2 2πL 1

1.79
hout 2πr2 L
1 1 r 1 r
Tin − T0 + T0 − T1 + T1 − T2 + T2 − Tout = q̇ (h +k ln (r1 ) + k ln (r2 ) +
in 2πr0 L 1 2πL 0 2 2πL 1
1
) 1.80
hout 2πr2 L
1 1 r 1 r 1
Tin − Tout = q̇ (h +k ln (r1 ) + k ln (r2 ) + ) 1.81
in 2πr0 L 1 2πL 0 2 2πL 1 hout 2πr2 L
Tin −Tout
q̇ = 1 1 r 1 r 1
1.82
( + ln( 1 )+ ln( 2 )+ )
hin 2πr0 L k12πL r0 k22πL r1 hout 2πr2 L

Example 1.17: Conduction and convection through a cylinder

Saturated steam at 14 bar (195°C) condenses inside a nickel tube 10 m long with a 2 cm ID and 4 cm OD. The
temperature of the surrounding air is 25°C. The thermal conductivity of nickel is 140 W.m-1.K-1. The pipe is
covered with a 0.5 cm layer of insulation (thermal conductivity 0.21 W.m-1.K-1).
a) Calculate the conductive energy across the pipe and insulation.
b) Given a convective heat transfer coefficient on the outside of the insulation (10 W.m -2.K-1), determine the
overall heat transfer coefficient, Uo
c) Determine the mass of steam which condenses per hour.
d) How do answers compare when convection is included vs when it is not?

Solution:

Ta T3 T2 T1 r3
r1 r2

h
𝑑𝑇
a) 𝑞̇ = −𝑘𝐴
𝑑𝑟
𝐴 = 2𝜋𝑟𝐿
𝑞̇ 𝑑𝑟
∴ = 𝑑𝑇
−2𝜋𝑘𝐿 𝑟

Across the pipe:


𝑟2 𝑇2
𝑞̇ 𝑑𝑟
∫ = ∫ 𝑑𝑇
−2𝜋𝑘𝑁 𝐿 𝑟
𝑟1 𝑇1

Kevin Harding (kevin.harding@wits.ac.za) 31


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Convective heat transfer

𝑞̇ 𝑟
∴ . 𝑙𝑛 ( 2) = 𝑇2 − 𝑇1 (1)
−2𝜋𝑘𝑁 𝐿 𝑟1

Across the insulation:


𝑞̇ 𝑟
. 𝑙𝑛 ( 3 ) = 𝑇3 − 𝑇2 (2)
−2𝜋𝑘𝐼 𝐿 𝑟2

Adding (1) + (2):


2𝜋𝐿(𝑇1 − 𝑇3 ) 2𝜋 × 10𝑚 × (468𝐾 − 298𝐾)
𝑞̇ = 𝑟2 𝑟3 = = 10 006𝑊 ≅ 10 𝑘𝑊
𝑙𝑛( ⁄𝑟1 ) 𝑛( ⁄𝑟2 ) 𝑙𝑛( 2𝑐𝑚⁄ ) 𝑛(2.5𝑐𝑚⁄2𝑐𝑚)
( + ) ( 1𝑐𝑚 + )
𝑘𝑁 𝑘𝐼 140 𝑊𝑚−1 𝐾 −1 0.21 𝑊𝑚−1 𝐾 −1

The conductive energy across the pipe and insulation is approx. 10 kW.

b) For concentric cylinders with film heat transfer:


𝑇𝑖 − 𝑇𝑎
𝑞̇ =
1 𝑙𝑛 (𝑟2 ⁄𝑟1 ) 𝑙𝑛 (𝑟3 ⁄𝑟2 ) 1
( + + + )
2𝜋ℎ𝑖 𝑟0 𝐿 2𝜋𝑘1 𝐿 2𝜋𝑘2 𝐿 2𝜋ℎ0 𝑟3 𝐿

For this system, the overall heat transfer co-efficient (U0) is given by:
1
𝑈0 =
𝑙𝑛 (𝑟2 ⁄𝑟1 ) 𝑙𝑛 (𝑟3 ⁄𝑟2 ) 1
( + + )
2𝜋𝑘𝑁 𝐿 2𝜋𝑘𝐼 𝐿 2𝜋ℎ𝑟3 𝐿
𝑈0
1
=
𝑙𝑛 (2𝑐𝑚⁄1𝑐𝑚) 𝑙𝑛 (2.5𝑐𝑚⁄2𝑐𝑚) 1
( + + )
2𝜋 × 140 𝑊𝑚−1 𝐾 −1 × 10𝑚 2𝜋 × 0.21 𝑊𝑚−1 𝐾 −1 × 10𝑚 2𝜋 × 10𝑊𝑚−2 𝐾 −1 × 0.025𝑚 × 10𝑚
1
𝑈0 = = 12.4 𝑊𝐾 −1
0.08065 𝑊 −1 𝐾

c) 𝑞̇ = 𝑈0 (𝑇𝑖 − 𝑇𝑎 ) = 12.4 𝑊𝐾 −1 (468𝐾 − 298𝐾) = 2108 𝑊 = 2.108 𝑘𝐽𝑠 −1

Energy balance for a flow process:


∆𝑣 2
∆𝐻̂+ ̂𝑠
+ 𝑔∆𝑧 = 𝑞̂ + 𝑊
2
𝑞̇
̂ = 𝑞̂ =
∴ ∆𝐻
𝑚̇
𝑞̇
⇒ 𝑚̇ =
∆𝐻 ̂

For saturated steam at 14bar, Ts = 195°C, Hf = 830 kJ/kg and Hg = 2790 kJ/kg
∴ ∆𝐻𝑣𝑎𝑝 = 1960 𝑘𝐽/𝑘𝑔
The enthalpy change is the heat of vaporization since the steam is condensing.
2.108 𝑘𝐽𝑠 −1
𝑚̇ = = 1.076 × 10−3 𝑘𝑔𝑠 −1 = 3.872 𝑘𝑔ℎ−1
1960 𝑘𝐽𝑘𝑔−1

d) The heat transfer is smaller when convection is included (2.108 kW compared to 10 kW)

1.3.4 Log Mean Radius and Log Mean Area


We define the log mean radius as follows:
r −r
rLM = 2 r21 1.83
ln
r1

Kevin Harding (kevin.harding@wits.ac.za) 32


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Overall heat transfer coefficient

where:
rLM = Log mean radius
r2 = outer radius
r1 = inner radius

Note:
r2 − r1 r2 − r1
r ≠ 1.84
ln 2 2
r1

From this we can define the log mean area:


ALM = 2πrLM L 1.85

where:
ALM = Log mean area

From the heat flow through two cylinders, (conduction and convection, it was shown that:
Tin −Tout
q̇ = 1 1 r 1 r 1
1.86
( + ln( 1 )+ ln( 2 )+ )
hin 2πr0 L k12πL r0 k22πL r1 hout 2πr2 L

Which we can re-write as:


Tin −Tout
q̇ = 1 1 r1 (r1−r0 ) 1 r (r −r ) 1
1.87
( + ln( ).(r + ln( 2 ).(r2 1 )+ )
hin 2πr0 L k12πL r0 1−r0 ) k22πL r1 2−r1 hout 2πr2 L

From the definition of RLM, and converting wherever possible to surface area instead of radius,
this becomes:
Tin −Tout
q̇ = 1 1 (r −r ) 1 (r −r ) 1
1.88
( + . 1 0+ . 2 1+ )
hin A0 k1 2πL rLM,10 k22πL rLM,21 hout A2
Tin −Tout
q̇ = 1 (r1−r0 ) (r −r ) 1
1.89
( + + 2 1 + )
hin A0 𝑘1 ALM,10 𝑘2ALM,21 hout A2

For n concentric circles:


Tin − Tout
q̇ = 1 n ∆Ri 1
1.90
( +∑i=1 + )
hin A0 ki ALM,i hout An

1.4 Overall heat transfer coefficient


Previously, heat transfer was expressed in terms of thermal resistance as:
∆T
q̇ = 1.91
R

where:
T: Temperature difference (K)
R: Thermal resistance (K/W)

Kevin Harding (kevin.harding@wits.ac.za) 33


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Overall heat transfer coefficient

For multiple slabs in series the thermal resistance was shown to be:
Li
R = ∑ni=1 1.92
ki A

Which can be re-written to include the convection as:


1 Li 1
R=h + ∑ni=1 +h 1.93
h A kA
i cA

And for cylindrical systems:


1 ∆Ri 1
R= + ∑ni=1 k A +h 1.94
hin A0 i LM,i out An

A new term is now introduced for the Overall Heat Transfer Coefficient (U):
1
U = A.R 1.95
∴ q̇ = UA(Tin − Tout ) 1.96

Note: Thermal resistance (R) includes the term for area (A), while the overall heat transfer
coefficient (A) does not. Further, R and U are inverse relationships.

Defining the overall heat transfer coefficient for cylinders:


q̇ = UA(Tin − Tout ) 1.97

We can measure heat and temperatures independently; UA can be calculated as a combined


measured value.

1
UA = 1 ∆ri 1
1.98
( +∑ni=1 k + )
hin i hout

While we typically use the ALM for these calculations, or assume the thickness is thin enough
not to worry about which radius, and thus aera, to use, we can then define different U values
for different areas:

UA = U0A0 = UiAi 1.99

UA
U0 = For outer cylinder 1.100
A0
UA
Ui = For inner cylinder 1.101
Ai

Example 1.18: Convective heat transfer, with conduction, across multiple layers (overall heat transfer
coefficient)

A furnace is constructed of a 130 mm steel layer and an outer surface of bricks which is 170 mm thick. The
thermal conductivities of the steel and brick layers are 144 W/m.K and 4.38 W/m.K respectively, whilst the
heat transfer of the inner and outer surface are 18W/𝑚2 . 𝐾 and 73W/𝑚2 . 𝐾. Given that the temperature inside

Kevin Harding (kevin.harding@wits.ac.za) 34


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Heat transfer through composite walls

the furnace is kept at 650°C and the temperature of the surroundings is 45°C, calculate the overall heat flow
and overall heat transfer coefficient.

Solution:
Because the system is at steady state, with areas and thermal conductivities, it was shown earlier that:
𝑞 𝑇𝑖 − 𝑇𝑎
=
𝐴 1 𝐿 𝐿 1
+ 1+ 2+
ℎ𝑖 𝑘1 𝑘2 ℎ0

Therefore:
𝑞
= 5552𝑊/𝑚. 𝐾
𝐴
Now: 𝑞 = 𝑈𝐴(𝑇𝑖 − 𝑇𝑎 )
𝑞
𝑈= = 9.18 W/𝑚2 . 𝐾
𝐴(𝑇𝑖 −𝑇𝑎 )

1.5 Heat transfer through composite walls


1.5.1 Heat transfer through series configuration
What we have looked at thus far has been heat transfer through a series configuration. This
can be represented in a similar way to an electrical circuit.

Th
Tc

k1 k2 k3

q̇ hh hc

L1 L2 L3

Rhh R1 R2 R3 Rhc

where:
Ri = Thermal resistance through each layer (including convection and conduction)

In the same way as in an electrical circuit, to get the total resistance in series, add the resistance.

For the slab above:


R = R hh + R1 + R 2 + R 3 +R hc 1.102

where:
Kevin Harding (kevin.harding@wits.ac.za) 35
School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Heat transfer through composite walls

R = Total thermal resistance

Therefore, from:
∆T
q̇ = 1.91
R
∆T
q̇ = R 1.103
hh +R1 +R2 +R3 +Rhc
∆T
q̇ = 1 L L L 1 1.104
+ 1+ 2+ 3+
hhh. A k1 A k2 A k3 A hhc. A

1.5.2 Heat transfer through parallel configuration


Similarly, heat transfer through a parallel configuration can be equated to an electrical circuit.

The sum of resistance in a parallel configuration is equal to the sum of the inverse resistances,
all to the power of negative one.

For the diagram below:


R1

R2

1 1 −1
R = (R + R ) 1.105
1 2

where:
R = Total thermal resistance

Given the configuration below, heat will travel from the left to the right through the parallel
configuration. In the calculations we have performed so far, we have assumed 1-dimendionsal
heat transfer. However, with the parallel configuration now in place, when k2 and k3 are not
equal, heat will move between these sections. To ensure the 1-dimensional assumptions in the
calculations, two scenarios are presented with simplifying assumptions.

Kevin Harding (kevin.harding@wits.ac.za) 36


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Heat transfer through composite walls

L2

Th
Tc

k1 k2 k4

q̇ hh hc

k3

L1 L3 L4

Scenario 1: Assuming the surfaces normal to the direction of heat flow are isothermal, i.e., the
entire length of slabs 1 and 4 have the same temperature, this gives:

Rhh R1 R2 R4 Rhc

R3

where:
Ri = Thermal resistance through each layer (including convection and conduction)

∆T
q̇ = 1.91
R

In a similar manner to calculating electrical resistance, the resistance through the parallel
section can be written as:
1 1 −1
R 2/3 = ( + ) 1.106
R2 R3
k A k3 A −1
R 2/3 = ( L2 + ) 1.107
2 L3

Therefore:
∆T
q̇ = 1 L k A k A −1 L 1
1.108
+ 1 +( 2 + 3 ) + 4 +
hhh. A k1 A L2 L3 k4A hhc. A

Scenario 2: However, if we assume that the surfaces parallel to heat flow are adiabatic, i.e., no
heat flows from slab 2 to 3, a better approximation is given by:

Kevin Harding (kevin.harding@wits.ac.za) 37


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Heat transfer through composite walls

Rhh R1 R2 R4 Rhc

Rhh R1 R3 R4 Rhc

Now:
∆T
q̇ = 1.91
R

Adding all the sections which are in series, inverting to account for addition of multiple parallel
sections and again inverting the sum of this:

∆T
q̇ = −1 1.109
1 1
( 1 L L L 1 + 1 L L L 1 )
+ 1 + 2 + 4 + + 1 + 3 + 4 +
hhh. A k1 A k2 A k4 A hhc. A hhh. A k1 A k3 A k4 A hhc. A

Video 1.4: Heat transfer – Parallel configurations (YouTube)

1.5.3 Contact resistance


The diagrams shown thus far indicate that the materials in successive layers are in perfect
contact with each other. However, if we look at these more closely, at a microscopic level in
some instances, we will find there are spaces between each layer. This is often most clear in
rough materials. Depending on the space between the materials, and the type of materials
involved, this can act as a further level of insulation.

Kevin Harding (kevin.harding@wits.ac.za) 38


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Heat transfer through composite walls

Th
Tc

k1 k2

L1 L2

R1 RC R2

where:
RC = Contact resistance (m2.K/W)

Table 1.2: Examples of contact resistance values

Interface RC × 104 (m2.K/W)


Silicon chip/lapped aluminium in air (27-500 kn/m2) 0.3 – 0.6
Aluminium/aluminium with indium foil filler (~100 kN/m2) ~0.07
Stainless/stainless with indium foil filler (~3500 kN/m2) ~0.04
Aluminium/aluminium with metallic Pb coating 0.01 – 0.1
Aluminium/aluminium with Dow Corning 340 grease (~100 kN/m2) ~0.07
Aluminium/stainless with Dow Corning 340 grease (~3500 kN/m2) ~0.04
Silicon chip/aluminium with 0.02 mm epoxy 0.2 – 0.9
Brass/brass with 15 m tin solder 0.025 – 0.14

1.5.4 Porous materials


Not all materials are homogenous, and this will impact the heat transfer through them. Once
example are porous materials. The porosity of a material will influence the heat transfer
through the material.

Kevin Harding (kevin.harding@wits.ac.za) 39


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Heat Transfer from Extended Surfaces

Th
Tc

Various estimations of the heat transfer through the material are available, including ways such
as dividing the material into parallel and series systems (not presented here), but a more
comprehensive method is given as:

From Fourier’s Law:


dT
q̇ = −k. A dx 1.3

Simplifying for a porous material, with constant area, at steady state:


keff A(Th −Tc )
q̇ = 1.110
L

where:
keff: Effective heat transfer coefficient across slid and porous section of slab (W/(m.K))

For a porous material:


k + 2k𝑠 − 2𝜀(k𝑠 − k𝑓 )
k eff = [ kf ] 1.111
f + 2k𝑠 + 𝜀(k𝑠 − k𝑓 )

where:
kf: Heat transfer coefficient of the fluid in the void (W/(m.K))
ks: Heat transfer coefficient of the solid (W/(m.K))
: Void fraction (-)

1.6 Heat Transfer from Extended Surfaces


1.6.1 Introduction
To cool something down, we may have a cooling fin. A hot object may have a long flat (or
thin) piece protruding from it to increase surface area and the conductive and convective heat
transfer, thereby cooling the object down faster.

Kevin Harding (kevin.harding@wits.ac.za) 40


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Heat Transfer from Extended Surfaces

Fins may come in various forms. These could be as straight, annular or pins.

Figure 1.4: Some examples of fins: Rectangular, Pins, Triangular, Parabolic, Annular.

1.6.2 Fins of Uniform Cross-Sectional Area


Fins of a uniform cross-sectional area could be rectangular (Figure 1.5) or pins (Figure 1.6).

TW TA

Convection

y
Conduction
TS
w
x x + x
Convection

Figure 1.5: Graphical representation of a rectangular fin protruding from a wall

Kevin Harding (kevin.harding@wits.ac.za) 41


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Heat Transfer from Extended Surfaces

TW TA

Convection

Conduction
r
TS

x x + x

Figure 1.6: Graphical representation of a circular pin protruding from a wall

Simplifying assumption (for rectangular fins of uniform cross-sectional area): Assume thin
slice of fin that has a temperature profile in x-direction only, NOT in y-direction.

By an energy balance:
Energy in (left – conduction) = energy out (right - conduction) + energy out (top + bottom -
convection) 1.112
q̇ |x = q̇ |x+∆x + h(2w∆x)(Ts − TA ) 1.113

(Multiplied by 2 because there are 2 surfaces)

where:
q̇ : Conductive heat transfer
h: convective heat transfer coefficient
x: element through which heat flows
TS surface temperature
TA: Ambient/air temperature

Replacing q̇ terms with conductive heat flow terms:


dT dT
−kA dx | = −kA dx | + h(2w∆x)(Ts − TA ) 1.114
x x+∆x
dT dT
−k(w. y) dx | = −k(w. y) dx | + h(2w∆x)(Ts − TA ) 1.115
x x+∆x

Rearrange and divide by x

Kevin Harding (kevin.harding@wits.ac.za) 42


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Heat Transfer from Extended Surfaces

dT dT
−k(w.y) | +k(w.y) |
dx x+∆x dx x
= −h(2w)(Ts − TA ) 1.116
∆x

Taking limits as x → 0
dT dT
−k(w.y) | +k(w.y) |
dx x+∆x dx x
lim = −h(2w)(Ts − TA ) 1.117
∆x→0 ∆x
d dT
[−k(w. y) dx ] = −h(2w)(Ts − TA ) 1.118
dx

Assuming k is constant:
d2 T
−k(w. y) = −h(2w)(Ts − TA ) 1.119
dx2

Simplifying:
d2 T
ky dx2 = 2h(Ts − TA ) 1.120

And since there is no temperature profile in the y-direction, Ts = T at all x


Therefore Ts = T

And:
d2 T
ky dx2 = 2h(T − TA ) 1.121
d2 T 2h
= ky (T − TA ) 1.122
dx2

Let:
 = T – TA 1.123
And:
m = √2h/ky 1.124

d2 θ
= m2 θ 1.125
dx2

D2 − m2 θ = 0 1.126
(D − m)(D + m)θ = 0 1.127
θ = C1 emx + C2 e−mx 1.128

We have two constants, therefore need two boundary conditions (BC):


BC1: At x = 0,  = 1 Known temp at wall
BC2: At x = L, d/dx = 0 Temperature has stationary point at x = L

From BC1:
1 = C1 + C2 1.129

Kevin Harding (kevin.harding@wits.ac.za) 43


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Heat Transfer from Extended Surfaces

From BC2:

= C1 memx − C2 me−mx 1.130
dx
0 = C1memL – C2me-mL 1.131
C2 = C1e2mL 1.132

θ1
C1 = 1.133
1+ e2ml
θ1 .e2ml
C2 = 1.134
1+ e2ml

θ1 .emx θ1 .e2ml .e−mx


θ= + 1.135
1+ e2ml 1+ e2ml
θ eml (e−ml .emx +eml .e−mx )
θ = 1+ e12ml [emx + e2ml . e−mx ] = θ1 1.136
eml (e−ml +eml )
θ e−ml .emx +eml .e−mx
= 1.137
θ1 e−ml +eml
θ cosh m.(L−x)
= 1.138
θ1 cosh mL

Re-introducing the temperature and h, k and y terms:


Reminder:
 = T – TA 1.123
m = √2h/ky 1.124

T−TA cosh √2h/ky.(L−x)


= 1.139
TW −TA cosh √2h/ky.L

where:
TW: Wall temperature (K)

cosh √2h/ky. (L − x)
∴ T = (TW − TA ) ( ) + TA
cosh √2h/ky. L

Video 1.5: Heat transfer – Conduction through slabs (YouTube)

Kevin Harding (kevin.harding@wits.ac.za) 44


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Radiative Heat Transfer

1.7 Radiative Heat Transfer


Black Body: A perfect idealized physical body which absorbs all incident electromagnetic
radiation and is also the best possible emitter of thermal radiation.

q̇ = AσT 4 Stefan Boltzmann Law 1.140

where:
q̇ : Heat Flux (W)
A: Cross Sectional Area (m2)
σ: Stefan Boltzmann Constant = 5.67 × 10-8 W.m-2K-4
2π5 k 4B
=
15h3 c 2
kB = Boltzmann constant – 1.38 × 10-23 J/K;
h = Plank constant – 6.626× 10-34 J.s
c = speed of light m/s
T = Temperature (K)

Example 1.19: Radiative heat transfer

What is the radiative heat transfer per square meter for a body at the following temperatures?
a) 100°C
b) 400°C
c) 800°C

Solution:
𝑞̇ /𝐴 = 𝐴𝜎𝑇 4
𝑞̇ /𝐴 = 𝜎𝑇 4
𝑞̇ /𝐴 = 5.76 × 10−8 . 𝑇 4

a) 𝑞̇ /𝐴 = 1.12 kW/m2
b) 𝑞̇ /𝐴 = 11.8 kW/m2
c) 𝑞̇ /𝐴 = 76.4 kW/m2

Given that the Stefan Boltzmann constant is very small, radiative heat transfer only plays a part
at high temperatures (approx. greater than 400°C), while conduction and convection describe
heat transfer at lower temperatures.

1.8 Heat Exchangers


1.8.1 Introduction
To transfer heat efficiently, both for energy and cost savings, industrial processes require heat
transfer equipment. Hot fluids may also need to be cooled before being discarded to the
environment – either for environmental and/or personal safety reasons.

Various heat transfer equipment exists. These include:


- Heat Exchanger; including shell and tube-, plate and frame-, radiators, and others

Kevin Harding (kevin.harding@wits.ac.za) 45


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Heat Exchangers

- Air cooling; including fin fan coolers, fans, and others


- Water cooling; including cooling towers, water jackets, and others
- Radiators
- Heat pipes

A heat exchanger is a piece of equipment built for efficient heat transfer from one medium to
another.

Figure 1.7: Plate and frame heat exchanger (metal plates transfer heat between two fluids)
(https://commons.wikimedia.org/wiki/File:Plate_frame_1.svg)

Figure 1.8: Passive heat sink on a motherboard (metal pins act as fins to allow for cooling)
(https://commons.wikimedia.org/wiki/File:Harumphy.dg965.heatsink.jpg)

Kevin Harding (kevin.harding@wits.ac.za) 46


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Heat Exchangers

Figure 1.9: Typical radiator found in a car (liquid inside the mechanism allows for conductive cooling
despite the contradictory naming) (https://commons.wikimedia.org/wiki/File:Automobile_radiator.jpg)

Figure 1.10: Various cooling tower designs (https://en.wikipedia.org/wiki/Cooling_tower)

A shell and tube heat exchanger is a class of heat exchanger most commonly used in oil
refineries and other large chemical processes. It consists of a shell with a bundle of tubes
inside. One fluid runs through the tubes, and another fluid flows over the tubes (through the
shell) to transfer heat between the two fluids.

1.8.2 Types of Shell and Tube Heat Exchangers


Shell and tube heat exchangers can be as simple as having a single pipe enclosed in a second
pipe (double pipe heat exchanger), or can become more elaborate to include baffles, bends or
more elaborate designs.

Kevin Harding (kevin.harding@wits.ac.za) 47


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Heat Exchangers

Figure 1.11: Double pipe heat exchanger (one Figure 1.13: Shell and tube heat exchanger (two
pass tube side; one pass shell side; no baffles) pass tube side; one pass shell side; with baffles)
(https://commons.wikimedia.org/wiki/File:Double- (https://commons.wikimedia.org/wiki/File:Straight-
Pipe_Heat_Exchanger.png) tube_heat_exchanger_2-pass.png)

Figure 1.12: Shell and tube heat exchanger (one Figure 1.14: U-tube heat exchanger (two pass
pass tube side; one pass shell side; with baffles) tube side; one pass shell side; with baffles)
(https://commons.wikimedia.org/wiki/File:Straight- (https://commons.wikimedia.org/wiki/File:U-
tube_heat_exchanger_1-pass.png) tube_heat_exchanger.png)

It is common to put the dirtier fluid (fluid which would foul) inside the tubes as the tubes are
easier to clean or replace. The easiest way to increase the area of a heat exchanger could simply
be to in increase the number of tubes.

1.8.3 Co-current Heat Exchangers


Two streams of fluid enter at the same end of the heat exchanger – one hot and one cold. The
two fluids exchange energy with the hotter getting colder and the colder getting hotter. The
temperature of the cold stream can never be greater than the hot stream and vice versa.

Kevin Harding (kevin.harding@wits.ac.za) 48


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Heat Exchangers

Temperature

Length

Figure 1.15: Temperature profile through a co-current heat exchanger

1.8.4 Counter Current Heat Exchangers


Two streams of fluid enter on opposite ends heat exchanger – one hot and one cold. The two
fluids exchange energy with the hotter getting colder and the colder getting hotter. The
temperature of the cold stream exiting can be hotter than the temperature of the hot stream
exiting.

Typically, the maximum temperature difference at any point can be no lower than 10°C. Note
in the counter current setup that the outlet temperature difference can be less than this.

Temperature

Length

Figure 1.16: Temperature profile through a co-current heat exchanger

Video 1.6: Heat transfer – Heat exchangers introduction (YouTube)

1.8.5 Derivation of Heat Exchanger Equation


Consider two fluids flowing in a counter current fashion, one hot and one cold.
Kevin Harding (kevin.harding@wits.ac.za) 49
School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Heat Exchangers

Side 1 Th Th + dTh Side 2


Hot Th

Tc Cold

Tc + dTc Tc

dA

- Consider a temperature, Th on the hot stream and a temperature Tc on the cold stream
- Consider a small area dA, over which it can be assumed the temperature is constant

Assumptions:
- No phase change
- No reaction
- Constant pressure
- Constant specific heat (Cp)

From overall heat transfer equations:


q̇ = UA∆T (1) 1.141

For the portion over dA (assuming the hot and cold fluids don’t change over the length of the
heat exchanger – because the length is so small):
dq̇ = UdA(Th − Tc ) 1.142

For the hot stream: Energy balance


dq̇ = ṁh Ĉph [Th − (Th + dTh )] 1.143
dq̇ = −ṁh Ĉph . dTh (2) 1.144

For the cold stream: Energy balance


dq̇ = ṁc Ĉpc [Tc − (Tc + dTc )] 1.145
dq̇ = −ṁc Ĉpc . dTc (3) 1.146

(2) x ṁc Ĉpc


dq̇ (ṁc Ĉpc ) = −ṁh Ĉph . dTh (ṁc Ĉpc ) (4) 1.147

(3) x ṁh Ĉph


dq̇ (ṁh Ĉph ) = −ṁc Ĉpc . dTc (ṁh Ĉph ) (5) 1.148

(4) – (5)
dq̇ (ṁc Ĉpc ) − dq̇ (ṁh Ĉph ) = −ṁh Ĉph . dTh (ṁc Ĉpc ) + ṁc Ĉpc . dTc (ṁh Ĉph ) 1.149
Kevin Harding (kevin.harding@wits.ac.za) 50
School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Heat Exchangers

dq̇ (ṁc Ĉpc − ṁh Ĉph ) = −ṁh Ĉph . ṁc Ĉpc (dTh − dTc ) 1.150

Substituting (1)
−UdA(Th − Tc )(ṁc Ĉpc − ṁh Ĉph ) = ṁh Ĉph . ṁc Ĉpc (dTh − dTc ) 1.151

Rearranging:
̂ pc −ṁh C
ṁc C ̂ ph dTh −dTc
−UdA ̂ ph .ṁc C
̂ pc
= 1.152
ṁh C Th −Tc

Now:
̂ pc −ṁh C
ṁc C ̂ ph
Let ̂ ̂
=R (6) 1.153
ṁh Cph .ṁc Cpc

Or
̂ pc −ṁh C
ṁc C ̂ ph 1 1
R= ̂ ph .ṁc C
̂ pc
= (ṁ ̂
− ṁ ̂ pc
) 1.154
ṁh C h Cph cC
dTh −dTc
∴ −RUdA = 1.155
Th −Tc
d(Th −Tc )
∴ = −RUdA 1.156
Th −Tc

Integrate:
∆T
ln(Th − Tc )∆T21 = −RUA 1.157
∆T
ln ∆T2 = −RUA 1.158
1

Or:
∆T
ln ∆T1 = RUA 1.159
2
∆T1
ln
∆T2
R= (7) 1.160
UA

Overall energy balance for each fluid:


ṁh Ĉph [Th,in − Th,out ] = ṁc Ĉpc [Tc,out − Tc,in ] = Q
Q
̂ ph
= [Th,in − Th,out ] (8) 1.161
ṁh C
Q
̂ pc
= [Tc,out − Tc,in ] (9) 1.162
ṁc C

(8) – (9)
1 1
Q (ṁ ̂
− ṁ ̂ pc
) = [Th,in − Th,out ] − [Tc,out − Tc,in ] 1.163
h Cph cC
1 1
Q (ṁ ̂
− ṁ ̂ pc
) = [Th,in − Tc,out ] − [Th,out − Tc,in ] 1.164
h Cph cC

Substituting (6):
QR = ∆T1 − ∆T2 1.165
Kevin Harding (kevin.harding@wits.ac.za) 51
School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Heat Exchangers

∆T1 −∆T2
Q= 1.166
R

Substituting (7)
∆T1 −∆T2
Q = UA ( ∆T1 ) 1.167
ln
∆T2

where:
∆T1 −∆T2
( ∆T1 ) = ∆TLM 1.168
ln
∆T2

∴ Q = UA∆TLM 1.169

Assumptions:
i) Overall heat transfer co-efficient (U) is constant
ii) Heat exchanger is well insulated
iii) Physical properties of each fluid are constant
iv) Steady state
v) No phase change, no reaction, no mixing, constant pressure

If there is a phase change, deal with each phase separately


The TLM calculation can only be done on a fluid which does not have a phase change OR if
there is a phase change ONLY.

Can use TLM calculation

Phase change; no
temperature change

Kevin Harding (kevin.harding@wits.ac.za) 52


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Heat Exchangers

Can use TLM calculation

Steam; phase change,


water
Steam; phase change,
water

TLM,a
TLM,b
TLM,c

CAN NOT use TLM calculation BUT: If we break the problem into 3, can
use TLM calculation

Example 1.20: Heat exchanger – log mean temperature vs. temperature difference

Calculate the log mean temperature difference given the following information for a counter current system:
Temperature hot in: 90C
Temperature hot out: 65C
Temperature cold in: 40C
Temperature cold out: 70C

Solution:

90C

Temperature 65C
T2
70C T1

40C

Length

T1 = 25C; Thot,out – Tcold,in = 65 – 40C = 25C


T2 = 20C; Thot,in – Tcold,out = 90 – 70C = 20C

∆𝑇1 − ∆𝑇2 25 − 20
∆𝑇𝐿𝑀 = ( )= ( ) = 22.4℃
∆𝑇1 25
𝑙𝑛 𝑙𝑛 ( )
∆𝑇2 20

25+20
This is not the same as average temperature: ∆𝑇𝐴𝑉𝐸 = ( ) = 22.5℃
2

Example 1.21: Heat exchanger – log mean temperature undefined

Calculate the log mean temperature difference given the following information for a counter current system
Temperature hot in: 90C

Kevin Harding (kevin.harding@wits.ac.za) 53


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Heat Generation

Temperature hot out: 60C


Temperature cold in: 40C
Temperature cold out: 70C

Solution:

90C

Temperature
T2 60C
70C
T1
40C

Length

T1 = 20C; Thot,out – Tcold,in = 60 – 40C


T2 = 20C; Thot,in – Tcold,out = 90 – 70C

∆𝑇1 − ∆𝑇2 20 − 20
∆𝑇𝐿𝑀 = ( )= ( ) = 𝑢𝑛𝑑𝑒𝑓𝑖𝑛𝑒𝑑
∆𝑇1 20
𝑙𝑛 𝑙𝑛 ( )
∆𝑇2 20

Since TLM is undefined, can use TAVE. Also: since T is constant across the length of heat exchanger (and also equal to
TAVE), TAVE can be used.

20 + 20
∆𝑇𝐴𝑉𝐸 = ( ) = 20℃
2

1.9 Heat Generation


Heat generation in terms of heat transfer:

Energy can be generated in a solid element by:


- Nuclear reaction;
- Catalyst and exothermic reactions; and
- Electrical current

1.9.1 Nuclear Reactions


Given in terms of G (per volume of radioactive material)

1.9.2 Catalysts
Metals are expensive. To maximize the surface area to volume ratio of a catalyst, we typically
use metal oxide supports which have a large surface area, e.g., activated carbon. Reactants
adsorb and react in the pores; products desorb and are released.

Kevin Harding (kevin.harding@wits.ac.za) 54


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Heat Generation

1.9.3 Electric Current

Example 1.22: Heat generation due to an electrical current

Consider an electric wire of circular cross section with radius R and electrical conductivity k e ohm-1cm-1.
Through this wire an electric current is passed with a current density of I amp/cm2. This process is an
irreversible process, converting some electrical energy into heat (thermal energy). The rate of heat production
per unit volume is given by the expression:
𝐼2
𝑆𝑒 =
𝑘𝑒

The quantity Se is the heat source resulting from electrical dissipation. It is assumed that the temperature rise
in the wire is not so large that the temperature dependence of either the electrical or thermal conductivity need
be considered. The surface of the wire is maintained at temperature T0.

Find the radial temperature profile within the wire.

Solution:

R
L

r

For the system we take a cylindrical shell of thickness r and length L.

The contributors to the energy balance across this shell are:

Kevin Harding (kevin.harding@wits.ac.za) 55


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Heat Generation

Rate of heat across the cylinder at r (2𝜋𝑟𝐿)(𝑞𝑟 |𝑟 ) = (2𝜋𝑟𝐿𝑞𝑟 ) |𝑟


Rate of heat out across cylindrical surface at r + r (2𝜋(𝑟 + ∆𝑟)𝐿)(𝑞𝑟 |𝑟+∆𝑟 ) = (2𝜋𝑟𝐿𝑞𝑟 ) |𝑟+∆𝑟
Rate of thermal production by electrical dissipation (2𝜋𝑟∆𝑟𝐿)𝑆𝑒

By an energy balance: Energy in + Energy generated = Energy out


(2𝜋𝑟𝐿𝑞𝑟 ) |𝑟 + (2𝜋𝑟∆𝑟𝐿)𝑆𝑒 = (2𝜋𝑟𝐿𝑞𝑟 ) |𝑟+∆𝑟
Re-arranging:
(2𝜋𝑟𝐿𝑞𝑟 ) |𝑟+∆𝑟 − (2𝜋𝑟𝐿𝑞𝑟 ) |𝑟 = (2𝜋𝑟∆𝑟𝐿)𝑆𝑒
÷ 2Lr
(𝑟𝑞𝑟 ) |𝑟+∆𝑟 − (𝑟𝑞𝑟 ) |𝑟
= 𝑆𝑒 𝑟
∆𝑟

Taking limits as r → 0
(𝑟𝑞𝑟 ) |𝑟+∆𝑟 − (𝑟𝑞𝑟 ) |𝑟
𝑙𝑖𝑚 = 𝑆𝑒 𝑟
∆𝑟→0 ∆𝑟
𝑑
(𝑟𝑞 ) = 𝑆𝑒 𝑟
𝑑𝑟 𝑟

This is a first order DE and can be integrated to:


𝑆𝑒 𝑟 𝐶1
𝑞𝑟 = +
2 𝑟

The integration constant C1 must be zero, since at r = 0, qr is not infinite


𝑆𝑒 𝑟
𝑞𝑟 =
2

Now, from Fourier’s Law for heat transfer:


𝑞𝑟 = −𝑘. (𝑑𝑇/𝑑𝑟)

Equating:
𝑑𝑇 𝑆𝑒 𝑟
−𝑘. ( ) =
𝑑𝑟 2

Assuming k is constant, integrating:


𝑆𝑒 𝑟 2
𝑇= + 𝐶2
4𝑘

At r = R; T = T0
𝑆𝑒 𝑅2
𝐶2 = + 𝑇0
4𝑘
𝑆𝑒 𝑅2 𝑟 2
𝑇 − 𝑇0 = [1 − ( ) ]
4𝑘 𝑅

Maximum temperature is at r = 0
𝑆𝑒 𝑅 2 𝑆𝑒 𝑅 2
𝑇𝑚𝑎𝑥 − 𝑇0 = OR 𝑇𝑚𝑎𝑥 = + 𝑇0
4𝑘 4𝑘

Average temperature:
2𝜋 𝑅
∫0 ∫0 (𝑇(𝑟) − 𝑇0 )𝑟𝑑𝑟. 𝑑𝜃 𝑆𝑒 𝑅2
〈𝑇〉 − 𝑇0 = 2𝜋 𝑅 =
∫0 ∫0 𝑟𝑑𝑟. 𝑑𝜃 8𝑘

Kevin Harding (kevin.harding@wits.ac.za) 56


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Unsteady State Heat Transfer

1.10 Unsteady State Heat Transfer


A non-steady state system is on that has properties (state variables – independent of path) that
are changing with time.

Non-steady state conduction problem:

q̇ 1,in q̇ 2,out

q̇ 1,out q̇ 2,in

1 2

- Some of q used to heat up temperature


- q1 ≠ q2
- At the point where q no longer increases, then all of q1,in = q1,out = q2,in = q2,out

Example 1.23: Non-steady state heat transfer through sphere

A well-stirred spherical vessel of diameter 6.5 cm containing water at 50C is placed in a constant temperature
bath at 20C. The overall heat transfer coefficient is 625 W/m2.K.

a) Assuming a constant heat loss over time, how much energy is lost from the sphere in 2 minutes?
b) What is the temperature of the water in the vessel after 2 minutes?

Solution:
a) 𝑞 = 𝑈0 𝐴(𝑇𝑖 − 𝑇0 )∆𝑡
𝑞 = 625 𝐽𝑠 −1 𝑚−2 𝐾 −1 × 4𝜋(0.0325)2 𝑚2 × (323𝐾 − 293𝐾) × 120𝑠
𝑞 = 625 𝐽𝑠 −1 𝑚−2 𝐾 −1 × 4𝜋(0.0325)2 𝑚2 × (323𝐾 − 293𝐾) × 120𝑠
𝑞 = 29865 𝐽 = 29.865 𝑘𝐽

b) Set up a differential balance describing how the temperature of the water within the vessel (the system)
varies with time. This is a constant pressure system => ∆H = q

𝐻|𝑡 = 𝐻|𝑡+∆𝑡 + 𝑞̇ ∆𝑡

𝐻 = 𝑚𝐶̂𝑝 ∆𝑇 = 𝑚𝐶̂𝑝 (𝑇 − 𝑇𝑏 )
where Tb is the basis temperature relative to which the Cp is calculated

Also note that q is a function of Ti


𝑚𝐶̂𝑝 (𝑇𝑖 − 𝑇𝑏 )|𝑡 = 𝑚𝐶̂𝑝 (𝑇𝑖 − 𝑇𝑏 )|𝑡+∆𝑡 + 𝑈0 𝐴(𝑇𝑖 − 𝑇0 )∆𝑡
∴ 𝑚𝐶̂𝑝 (𝑇𝑖 )|𝑡+∆𝑡 − 𝑚𝐶̂𝑝 (𝑇𝑖 )|𝑡 = −𝑈0 𝐴(𝑇𝑖 − 𝑇0 )∆𝑡
𝑚𝐶̂𝑝 (𝑇𝑖 )|𝑡+∆𝑡 − 𝑚𝐶̂𝑝 (𝑇𝑖 )|𝑡
= 𝑈0 𝐴(𝑇0 − 𝑇𝑖 )
∆𝑡

Kevin Harding (kevin.harding@wits.ac.za) 57


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Additional Reading

𝑑𝑚𝐶̂𝑝 𝑇𝑖
⇒ = 𝑈0 𝐴(𝑇0 − 𝑇𝑖 )
𝑑𝑡
𝑚𝐶̂𝑝
𝑑𝑇 = 𝑈0 𝐴𝑑𝑡
(𝑇0 − 𝑇𝑖 ) 𝑖
𝑇𝑓 𝑡
1
𝑚𝐶̂𝑝 ∫ 𝑑𝑇 = 𝑈0 𝐴 ∫ 𝑑𝑡
𝑇0 − 𝑇𝑖 𝑖
323 0
where Tf is the final temperature of the water inside the sphere

𝑥 1 𝑥
Note: ∫𝑥 2 𝑑𝑥 = [− 𝑙𝑛(𝑎 − 𝑥)]𝑥12
1 𝑎−𝑥

𝑇
∴ 𝑚𝐶̂𝑝 [−𝑙𝑛 (𝑇0 − 𝑇𝑖 )]50𝑓 = 𝑈0 𝐴𝑡
𝑙𝑛 (𝑇0 − 323)
𝑚𝐶̂𝑝 = 𝑈0 𝐴𝑡
𝑙𝑛 (𝑇0 − 𝑇𝑓 )
𝑇0 − 323 𝑈0 𝐴𝑡
𝑙𝑛 ( )=
𝑇0 − 𝑇𝑓 𝑚𝐶̂𝑝
𝑈0 𝐴𝑡
𝑇0 − 323 ⁄𝑚𝐶̂
( )=𝑒 𝑝
𝑇0 − 𝑇𝑓
𝑈0 𝐴𝑡
⁄𝑚𝐶̂
𝑇0 − 323 = 𝑒 𝑝 × (𝑇0 − 𝑇𝑓 )
𝑈0 𝐴𝑡 𝑈0 𝐴𝑡
⁄𝑚𝐶̂ ⁄𝑚𝐶̂
𝑇0 − 323 = 𝑇0 . 𝑒 𝑝 − 𝑇𝑓 . 𝑒 𝑝

𝑈0 𝐴𝑡
⁄𝑚𝐶̂
𝑇0 (𝑒 𝑝 − 1) + 323
∴ 𝑇𝑓 = 𝑈0 𝐴𝑡
⁄𝑚𝐶̂
𝑒 𝑝

𝐴 = 𝑠𝑢𝑟𝑓𝑎𝑐𝑒 𝑎𝑟𝑒𝑎 𝑜𝑓 𝑠𝑝ℎ𝑒𝑟𝑒 = 4𝜋(0.0325)2 𝑚2


𝑡 = 2𝑚𝑖𝑛 = 120𝑠
𝑣𝑜𝑙𝑢𝑚𝑒 𝑜𝑓 𝑠𝑝ℎ𝑒𝑟𝑖𝑐𝑎𝑙 𝑐𝑜𝑛𝑡𝑎𝑖𝑛𝑒𝑟 4⁄ 𝜋(0.0325)3 𝑚3
𝑚= = 3 = 0.14 𝑘𝑔
𝑠𝑝𝑒𝑐𝑖𝑓𝑖𝑐 𝑣𝑜𝑙𝑢𝑚𝑒 𝑜𝑓 𝑤𝑎𝑡𝑒𝑟 𝑎𝑡 50℃ 0.001 𝑚3 𝑘𝑔−1
𝐶̂𝑝 = 4.182 𝑘𝐽𝑘𝑔−1 𝐾 −1 = 4182𝑘 𝐽𝑘𝑔−1 𝐾 −1 𝑎𝑡 323𝐾 (50℃)

625×4𝜋(0.0325)2×120⁄
( 0.14×4182)
293 × (𝑒 − 1) + 323
𝑇𝑓 = 625×4𝜋(0.0325)2 ×120⁄
( 0.14×4182)
𝑒
𝑇𝑓 = 298.5𝐾 = 25.5℃

Additional Reading
Çengel, YA, Ghajar, AJ, 2011. Heat and Mass Transfer – Fundamentals and Applications,
McGraw Hill.
Coulson, JM, Richardson, JF, Backhurst, JR, Harker, JH, 1999. Coulson & Richardson’s
Chemical Engineering, Volume 1, Elsevier.
Incropera, FP, DeWitt, DP, Bergman and Lavine, 2013. Fundamentals of Mass and Heat
Transfer, 7th Edition, John Wiley & Sons.
Perry, RH, and Green, DW, 1984. Perry's Chemical Engineers Handbook.
Rathakrishnan, E, 2012. Elements of Heat Transfer, CRC Press, Taylor & Francis Group.
Kevin Harding (kevin.harding@wits.ac.za) 58
School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Additional Reading

Skogestad, S, 2009. Chemical and Energy Process Engineering, CRC Press.


Welty, JR, Wicks, CE, Wilson, RE, Rorrer, GL. 2008. Fundamentals of Momentum, Heat,
and Mass Transfer, Wiley.

Kevin Harding (kevin.harding@wits.ac.za) 59


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Problems

Problems
Example 1.28: [MCQ] Units of k e.) Use the average (NOT log mean) temperature across the whole
Example 1.24: [MCQ] Conduction: Definition of A exchanger.
What are the units of k (thermal conductivity coefficient):
Conduction is given by the equation: q = -k.A.dt/dx, where x is the direction of heat flow. a.) J/(m.K.s) Solution:
A is the: b.) W/(kg.K) C
a.) Average surface area of the object, in m2 c.) J/(m.K)
b.) Area parallel to heat flow, in m2 d.) W/K Example 1.32: [MCQ] Overall heat transfer coefficient
c.) Area perpendicular to heat flow, in m2 e.) J/(m2.K)
d.) Amount of material through which heat flows, in kg UA is given as 100 W/K. Ao = 20 m2 and Ai = 10m2. The correct Uo and Ui respectively
e.) Total amount of material through which heat flows PLUS insulating Solution: are (in W/m2.K):
material which does not allow heat flow, in kg C a.) 2&5
b.) 2 & 10
Solution: Example 1.29: [MCQ] Conduction through the walls of a cylinder c.) 5 & 10
C d.) 5 & 20
Heat flows through a cylinder (from inside to outside). Which of the following equations e.) 10 & 20
Example 1.25: [MCQ] Definition of thermal insulator would you use to solve heat flow:
a.) 𝑞 = −𝑘2𝜋𝑟𝐿(𝑇ℎ − 𝑇𝑐 ) Solution:
A material that is considered a thermal insulator will have a: −𝑘2𝜋𝐿(𝑇ℎ −𝑇𝑐 ) C
b.) 𝑞= 𝑟
𝑙𝑛 𝑖𝑛𝑠𝑖𝑑𝑒
a.) High thermal conductivity 𝑟𝑜𝑢𝑡𝑠𝑖𝑑𝑒

b.) Low thermal conductivity c.) 𝑞 = (𝑟


−𝑘2𝜋𝐿(𝑇ℎ −𝑇𝑐 )
Example 1.33: Heat transfer around a pipe
2 2
𝑜𝑢𝑡𝑠𝑖𝑑𝑒 ) −(𝑟𝑖𝑛𝑠𝑖𝑑𝑒 )
c.) Large cross-sectional area −𝑘2𝜋𝐿(𝑇ℎ −𝑇𝑐 )
d.) Large surface area d.) 𝑞=𝑟
𝑜𝑢𝑡𝑠𝑖𝑑𝑒 −𝑟𝑖𝑛𝑠𝑖𝑑𝑒 Water flows through a 2” steel pipe (2.067” ID & 2.375” OD). The pipe is heated on
e.) None of the above e.) None of the above the outside by steam at 230°F. The overall heat transfer coefficient is 250 Btu/hr ft 2 °F
based on the outside area of the pipe. If the inlet and outlet temperatures for a 1 ft length
Solution: Solution: of pipe are 80°F and 180°F, what is the water flow rate?
B B
Solution:
Example 1.26: [MCQ] Definition of 𝒒̇ Example 1.30: [MCQ] Heat transfer equipment 𝑄 = 𝑈𝐴∆𝑇𝐿𝑀
𝐴 = 2𝜋𝑟0 𝐿
𝑞̇ has the units of: Which of the following is least likely to be classified as heat transfer equipment:
a.) kJ a.) Fin fan cooler ∆𝑇1 − ∆𝑇2 (𝑇𝑠𝑡𝑒𝑎𝑚,𝑖𝑛 − 𝑇𝐻20,𝑜𝑢𝑡 ) − (𝑇𝑠𝑡𝑒𝑎𝑚,𝑜𝑢𝑡 − 𝑇𝐻2𝑂,𝑖𝑛 )
∆𝑇𝐿𝑀 = =
b.) kJ/kg b.) Distillation column ∆𝑇 (𝑇 − 𝑇𝐻2𝑂,𝑜𝑢𝑡 )
𝑙𝑛 ∆𝑇1 𝑙𝑛 𝑠𝑡𝑒𝑎𝑚,𝑖𝑛
c.) kJ/m3 c.) Radiator 2 (𝑇𝑠𝑡𝑒𝑎𝑚,𝑜𝑢𝑡 − 𝑇𝐻2𝑂,𝑖𝑛 )
d.) kJ/s d.) Heat Exchanger
e.) kJ/m2 e.) Furnace Assume that Tsteam,in = Tsteam,out = 230°F
(230℉ − 180℉) − (230℉ − 80℉)
Solution: Solution: ∴ ∆𝑇𝐿𝑀 = = 91℉
(230℉ − 180℉)
D B 𝑙𝑛 (230℉
− 80℉)

Example 1.27: [MCQ] Conduction: Proportionality Example 1.31: [MCQ] Heat exchanger: Phase change 𝑄 = 𝑈𝐴∆𝑇𝐿𝑀
1𝑓𝑡
𝑄 = 250 𝐵𝑡𝑢 ℎ−1 𝑓𝑡 −2 ℉−1 × 2𝜋 × 1.1875𝑖𝑛 × × 91℉
Conductive heat transfer is proportional to: How is a phase change dealt with in a heat exchanger? In order to calculate the heat 12𝑖𝑛
a.) x/T exchanger area needed: 1 𝑘𝐽𝑠 −1
𝑄 = 14 145.35 𝐵𝑡𝑢 ℎ−1 × = 4.146 𝑘𝐽𝑠 −1
b.) T/x a.) Calculate the log mean temperature for three sections (liquid, phase 3412 𝐵𝑡𝑢 ℎ−1
c.) A/T change and gas) and use the average of these temperatures in a
At T = 27°C, hf = 113.1 kJ/kg
d.) T/A single calculation.
e.) None of the above b.) Use the log mean temperature across the whole exchanger.
At T = 80°C, hf = 334.9 kJ/kg
c.) Calculate the log mean temperature for three sections (liquid, phase
At T = 85°C, hf = 355.9 kJ/kg
Solution: change and gas) and calculate 3 individual areas.
By interpolation, hf = 343.3 kJ/kg at T = 82°C
B d.) Divide the calculation into one part for the liquid and one for the
gas.
Energy balance:

Kevin Harding (kevin.harding@wits.ac.za) 60


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Problems

∆𝐻 ̂𝐻2𝑂 = 𝑞̂ Assuming a section of composite material below, what is the inside wall temperature if
∴ 𝑞̇ = 𝑚̇𝐻2𝑂 ∆𝐻 ̂𝐻2𝑂 the outside is maintained at a temperature of 1000°C and the heat transfer is 4 kW/m. Which can also be simplified to (because portions of the parallel section are the same
𝑞̇ = 𝑚̇𝐻2𝑂 (ℎ̂𝑓,82℃ − ℎ̂𝑓,27℃ ) Give your answer to the nearest degree Fahrenheit. material, i.e., same k, AND same area):
k-values (W/mK): ∆𝑇
𝑞̇ 4.146 𝑘𝐽𝑠 −1 𝑞̇ =
⇒ 𝑚̇𝐻2𝑂 = = = 0.01801 𝑘𝑔𝑠 −1 Aluminium: 200 Ceramic: 0.52 Air: 0.0234 −1
(ℎ̂𝑓,82℃ − ℎ̂𝑓,27℃ ) (343.3 𝑘𝐽𝑘𝑔−1 − 113.1 𝑘𝐽𝑘𝑔−1 ) 𝐿1 𝐿2 𝑘3 𝐴3 𝑘4 𝐴4 𝐿8
+ + (3 (
𝑘1 𝐴1 𝑘2 𝐴2 𝐿3 ) + 2 ( 𝐿4 )) + 𝑘8 𝐴
𝑚̇𝐻2𝑂 = 0.01801 𝑘𝑔𝑠 −1 × 2.205 𝑙𝑏𝑘𝑔−1 × 3600 𝑠 ℎ−1 = 143 𝑙𝑏 ℎ−1
Assumptions to be used:
- Ignore convection Substituting numbers from above:
- Assume the surfaces normal to the direction of heat flow are isothermal ∆𝑇
𝑞̇ = −1
𝐿1 𝐿2 𝑘3 𝐴3 𝑘4 𝐴4 𝐿8
𝑘1 𝐴1 + 𝑘2 𝐴2 + (3 ( 𝐿3 ) + 2 ( 𝐿4 )) + 𝑘8 𝐴

Tinside = 24.84397…°C
Converting to Fahrenheit:
Video 1.7: Heat transfer – Heat transfer around a pipe (YouTube) Tinside = 85.71915…°F
Therefore, to the nearest degree:
Tinside = 86°F
Example 1.34: Heat loss and viscosity effects
Example 1.36: Heat transfer equipment

As hot oil flows through a pipe it loses heat, thereby reducing its viscosity. For the oil to
List as many types of heat transfer equipment as possible. How does these work?
remain flowing, the viscosity needs to remain above 6 508 cP, at which point it enters a
heat exchanger to increase the temperature to 80°C. Solution:
Heat exchangers A process stream is heated or cooled via exchange of heat
a) How much water (at 95°C) is required in the heat exchanger if 100 kg/hr of oil is with another process stream
flowing through the pipe? Assume the temperature of the steam drops by 10°C. Heaters A process stream is heated via exchange of heat with a hot
b) How many heat exchangers will be needed for a length of pipe 3 716 m long? service/utility stream (e.g. steam)
Solution:
Assume the oil needs to be at 80°C at the start and end of the pipe, i.e., include a Coolers A process stream is cooled via exchange of heat with a cold
It can be assumed that all three “bars” are 5 mm thick. Therefore, the length of the air
heat exchanger at each end of the pipe. service/utility stream (e.g. cooling water)
sections are:
Reboilers Typically associated with distillation columns. The liquid
(1m – 3(5mm)/2 = 0.4925 mm thick. Further assume that the section is 1m deep. NOTE:
Given: bottoms product is heated and partially or fully vaporized
Units DO NOT balance with this assumption and as such, the heat transfer given needs
𝑇 = 25 + 55𝑒 −0.02𝑥 Units: °C via exchange of heat with a hot service/utility stream (e.g.
to be written as 4kW as we are taking a basis of 1 m.
𝑉𝑖𝑠𝑐𝑜𝑠𝑖𝑡𝑦 = 5000 + 5000𝑒 −0.01𝑥 Units: cP steam)
Where x is the length down the pipe (in metres) Condensers Typically associated with distillation columns. The vapour
Numbering the sections from 1 – 8 from left to right (top to bottom):
Cp, oil = 1.6 kJ/kg.K top product is heated and partially or fully condensed via
Cp, steam = 2 kJ/kg.K exchange of heat with a cold service/utility stream (e.g.
Li Ki Ai Resistance per
cooling water/refrigerant)
section
Solution: 2 Agitated vessels (jacketed heaters) Process fluid is heated or cooled in an
(mm) (m) (w/mK) (m ) Ri = Li/KiAi
a) At a viscosity of 6508 cP; x = 119.8653 m [2nd equation] 1 Ceramic 40 0.04 0.52 1 0.076923077 agitated tank via exchange of heat with a hot or cold jacket
This also corresponds to a temperature of 30°C [1st equation] 2 Al left 20 0.02 200 1 0.0001 (steam or cooling water circulated through the outer
Calculating the heat needed to raise the oil from 30°C to 80°C: 3 Al top 500 0.5 200 0.005 0.5 jacket of the vessel)
4 Air top 500 0.5 0.0234 0.4925 43.38583019
𝑄 = 𝑚𝐶𝑝∆𝑇 = 100 𝑘𝑔 ∗ 1.6 𝑘𝐽/𝑘𝑔. 𝐾 ∗ (80 − 30)𝐾 = 7999.529 𝑘𝐽/ℎ𝑟
5 Al middle 500 0.5 200 0.005 0.5 Example 1.37: Shell and tube heat exchanger
Since the heat gained by the oil must be equal to the heat lost by the hot water: 6 Air bottom 500 0.5 0.0234 0.4925 43.38583019
𝑄𝑜𝑖𝑙 = 𝑄𝑤𝑎𝑡𝑒𝑟 = 7999.529 𝑘𝐽/ℎ𝑟 = 𝑚 ∗ 4.1868 ∗ (10) 7 Al bottom 500 0.5 200 0.005 0.5
∴ 𝑚𝑤𝑎𝑡𝑒𝑟 = 399.976 ≈ 400 𝑘𝑔/ℎ𝑟 8 Al right 24 0.024 200 1 0.00012
In a shell and tube heat exchanger:
b) Length at which viscosity is at the limit = 119.86m (from above). Therefore,
need a heat exchanger every 119 m Now: a) What are the role of baffles?
Therefore (3716/119) = 31 ∆𝑇 b) How does fouling influence heat transfer?
Plus adding the initial heat exchanger to the start of the line, a total of 32 heat 𝑞̇ = c) Can the outlet temperature of a hot stream be less than the outlet temperature of
𝑅
exchangers are needed. the cold stream? Explain and include a sketch to show this.
Or for the system above:
Example 1.35: Heat transfer through composite slab ∆𝑇 Solution:
𝑞̇ = a) Baffles define the flow path of fluid through the heat exchanger, increase the
𝐿1 𝐿 𝑘3 𝐴3 𝑘4 𝐴4 𝑘5 𝐴5 𝑘6 𝐴6 𝑘7 𝐴7 −1 𝐿
+ 2 +( 𝐿 + 𝐿 + 𝐿 + 𝐿 + 𝐿 ) + 8 surface area for heat exchange and support the tubes.
𝑘1 𝐴1 𝑘2 𝐴2 3 4 5 6 7 𝑘8 𝐴

Kevin Harding (kevin.harding@wits.ac.za) 61


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Problems

A stream (oil) of 9.2 kg/s is to be heated from 65°C to 95°C in a heat exchanger, using b) What is the energy input required at the ends of the rod to maintain the
b) Fouling refers to the deposition of solid material on the heat exchange surfaces. 16.7 kg/s of water entering at 105°C. If the overall heat-transfer coefficient is specified temperatures T1 and T2?
All process and utility streams will cause fouling and the extent depends on the 1300 J/m2s°C, calculate the area of heat exchanger required if the flows are counter
composition of the fluid. The fouled material acts as a resistance to heat transfer current. Solution:
and, therefore, reduce the rate of heat transfer. a) Energy in|x = Energy out|x+∆x – energy loss due to convection over ∆x
Solution:
c) In a co-current heat exchanger m this will never be possible since the hot is always 𝑞̇ |𝑥 = 𝑞|̇𝑥+∆𝑥 + ℎ(2𝜋𝑅∆𝑥)(𝑇 − 𝑇0 )
hotter than the cold when flowing in the same direction (below – left). However,
theoretically, in a counter-current configuration, the hot (coming in from the left) Applying Fourier’s Law:
could still ALWAYS be hotter than the cold (coming on the right) along the heat 𝑑𝑇 𝑑𝑇
−𝑘(𝜋𝑅2 ) | = −𝑘(𝜋𝑅2 ) | + 2𝜋ℎ𝑅∆𝑥(𝑇)
exchanger from left to right, but because the outlets are on different sides, by the 𝑑𝑥 𝑥 𝑑𝑥 𝑥+∆𝑥
end of the heat transfer process, these hot/cold temperatures could have ‘switched’ For the oil stream: Re-arranging and dividing throughout by ∆x:
(below – right). 𝑄 = 𝑚𝐶̂𝑝 ∆𝑇 𝑑𝑇 𝑑𝑇
𝑘𝜋𝑅2 (𝑑𝑥 | − 𝑑𝑥 | )
Assume vegetable oil or mineral oil (both have the same heat capacity) 𝑥+∆𝑥 𝑥
= 2𝜋ℎ𝑅𝑇
Cp,oil = 1.67 kJ/kg.K ∆𝑥
Taking lim
𝑄 = 9.2 𝑘𝑔𝑠 −1 × 1.67 𝑘𝐽𝑘𝑔−1 𝐾 −1 × (368𝐾 − 338𝐾) ∆𝑥→0

𝑄 = 460.92 𝑘𝐽𝑠 −1 𝑑 𝑑𝑇 2ℎ𝑇


( )=
𝑑𝑥 𝑑𝑥 𝑘𝑅
2
For the water stream: 𝑑 𝑇 2ℎ𝑇 𝑑 2 𝑇 2ℎ
∴ 2= 𝑜𝑟 − 𝑇=0
𝑄 = 𝑚𝐶̂𝑝 ∆𝑇 𝑑𝑥 𝑘𝑅 𝑑𝑥 2 𝑘𝑅
If D = d/dx and we let θ = (2h/kR)½ , then:
𝑄
∴ ∆𝑇 = (𝐷2 − 𝜃 2 )𝑇 = 0
𝑚𝐶̂𝑝 (𝐷 + 𝜃)(𝐷 − 𝜃)𝑇 = 0
⇒ 𝑇 = 𝐴𝑒 −𝜃𝑥 + 𝐵𝑒 𝜃𝑥
Assume Cp,water = 4.19 kJ/kg.K
Example 1.38: Counter current heat exchanger 460.92 𝑘𝐽𝑠 −1
∆𝑇 = Boundary condition 1: At x = 0, T = T1 = 150°C
16.7 𝑘𝑔𝑠 −1 × 4.19 𝑘𝐽𝑘𝑔−1 𝐾 −1 150 = 𝐴𝑒 −𝜃(0) + 𝐵𝑒 𝜃(0) = 𝐴 + 𝐵
A small counter current heat exchanger operates with the following stream temperatures: ∆𝑇 = 6.6𝐾 ∴ 𝐵 = 150 − 𝐴
∆𝑇 = 𝑇𝐻,𝑜𝑢𝑡 − 𝑇𝐻,𝑖𝑛 Boundary condition 2: At x = L, T = T2 = 60°C
∴ 𝑇𝐻,𝑜𝑢𝑡 = 371.4𝐾 = 98.4℃ 60 = 𝐴𝑒 −𝜃𝐿 + (150 − 𝐴)𝑒 𝜃𝐿
Cold stream in: 20°C
Cold stream out: 100°C 𝐴(𝑒 𝜃𝐿 − 𝑒 −𝜃𝐿 ) = 150𝑒 𝜃𝐿 − 60
𝑄 = 𝑈0 𝐴∆𝑇𝐿𝑀 150𝑒 𝜃𝐿 − 60
Hot stream in: 120°C
𝑄 𝐴 = 𝜃𝐿
Hot stream out 70°C ∴𝐴= 𝑒 − 𝑒 −𝜃𝐿
𝑈0 ∆𝑇𝐿𝑀
(𝑇𝐻,𝑖𝑛 − 𝑇𝐶,𝑜𝑢𝑡 ) − (𝑇𝐻,𝑜𝑢𝑡 − 𝑇𝐶,𝑖𝑛 ) (105℃ − 95℃) − (98.4℃ − 65℃) 150𝑒 𝜃𝐿 − 60
The unit has a total area for heat transfer of 1 m² and an overall heat transfer coefficient ∆𝑇𝐿𝑀 = =
(𝑇 − 𝑇𝐶,𝑜𝑢𝑡 ) (105℃ − 95℃) ∴ 𝐵 = 150 −
of 500W/m².K. What is the rate of energy transfer? ln ((98.4℃ ) (𝑒 𝜃𝐿 − 𝑒 −𝜃𝐿 )
ln ( 𝐻,𝑖𝑛 ) − 65℃)
(𝑇𝐻,𝑜𝑢𝑡 − 𝑇𝐶,𝑖𝑛 ) 150(𝑒 𝜃𝐿 − 𝑒 −𝜃𝐿 ) − (150𝑒 𝜃𝐿 − 60)
Solution: 𝐵=
= 19.40℃ (𝑒 𝜃𝐿 − 𝑒 −𝜃𝐿 )
460.92 𝑘𝐽𝑠 −1 60 − 150𝑒 −𝜃𝐿
𝐴= −1 −2 −1
= 18.28 𝑚2 𝐵=
1.300 𝑘𝐽𝑠 𝑚 ℃ × 19.40℃ 𝑒 𝜃𝐿 − 𝑒 −𝜃𝐿

Example 1.40: Heat transfer down a length of rod 150𝑒 𝜃𝐿 − 60 −𝜃𝑥 60 − 150𝑒 −𝜃𝐿 𝜃𝑥
𝑇= 𝑒 + 𝜃𝐿 𝑒
𝑒 𝜃𝐿 − 𝑒 −𝜃𝐿 𝑒 − 𝑒 −𝜃𝐿
𝜃𝑥
A metal rod of length L and radius R has its two ends maintained at temperatures T1 = (150𝑒 𝜃𝐿 − 60)𝑒 −𝜃𝑥 + (60 − 150𝑒 −𝜃𝐿 )𝑒
𝑄 = 𝑈0 𝐴∆𝑇𝐿𝑀 150°C and T2 = 60°C respectively. The length of the rod is exposed to air at constant 𝑇=
𝑒 𝜃𝐿 − 𝑒 −𝜃𝐿
(𝑇𝐻,𝑖𝑛 − 𝑇𝐶,𝑜𝑢𝑡 ) − (𝑇𝐻,𝑜𝑢𝑡 − 𝑇𝐶,𝑖𝑛 ) (393𝐾 − 373𝐾) − (343𝐾 − 293𝐾) temperature To = 0°C, and hence will be cooled by the surrounding air. The film
∆𝑇𝐿𝑀 = =
(𝑇 − 𝑇𝐶,𝑜𝑢𝑡 ) (393𝐾 − 373𝐾) coefficient of heat transfer between the curved surface of the rod and the air is h. Assume At x = L/2
𝑙𝑛 ( 𝐻,𝑖𝑛 ) 𝑙𝑛 ((343𝐾 )
(𝑇𝐻,𝑜𝑢𝑡 − 𝑇𝐶,𝑖𝑛 ) − 293𝐾) that at any point x along the length of the rod the metal temperature is constant over the 𝜃𝐿 𝜃𝐿⁄
(150𝑒 𝜃𝐿 − 60)𝑒 − ⁄2 + (60 − 150𝑒 −𝜃𝐿 )𝑒 2
= 32.74𝐾 circular cross-section of the rod – although this ‘constant’ temperature varies, along the 𝑇=
𝑒 𝜃𝐿 − 𝑒 −𝜃𝐿
length of the rod. Regard the thermal conductivity k of the metal as constant.
2ℎ
𝑄 = 500 𝑊𝑚−2 𝐾 −1 × 1 𝑚2 × 32.74𝐾 = 16.37 𝑘𝑊 𝑜𝑟 𝑘𝐽𝑠 −1 where 𝜃 = √𝑘𝑅
a) Obtain an expression for the temperature at the midpoint (x = L/2) of the
Example 1.39: Counter-current heat exchanger rod. b) For conduction down the length of the rod: 𝑞̇ = ℎ(2𝜋𝑅𝐿)(𝑇2 − 𝑇1 )

Kevin Harding (kevin.harding@wits.ac.za) 62


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Problems

𝑏⁄ 𝑏⁄
Example 1.41: 𝑊𝑖𝑡ℎ 𝑎 = 2 𝑊𝑖𝑡ℎ 𝑎 = 2 [ 1 − 1 ]
(2𝜋ℎ2 𝑟2 )2 (2𝜋)2 (ℎ2 𝑟2 )2 (ℎ1 𝑟1 )2
0.004 𝑊𝑚−1 𝐾 −2⁄ 0.004 𝑊𝑚−1 𝐾 −2⁄ 1 1
A thick-walled pipe of inner radius r1 and outer radius r2 is constructed of material whose 𝑎= 2 = 2.0264 × 10−6 𝑊 −1 𝑚 𝑎= 2[ − ]
thermal conductivity is a function of the temperature, i.e. k = a + bT, where a and b are (2𝜋 × 25 𝑊𝑚−2 𝐾 −1 × 0.20 𝑚)2 (2𝜋)2 (25 𝑊𝑚−2 𝐾 −1 × 0.20 𝑚)2 (150 𝑊𝑚−2 𝐾 −1 × 0.05 𝑚)2
constants. Inside the pipe, there is a liquid at temperature To. Outside, the pipe is 𝑎 = 1.1258 × 10−6 𝑊 −1 𝑚
𝑟
surrounded by air at a temperature TA. The film coefficients on the inside and outside of ln( 2⁄𝑟1 ) 𝑎 + 𝑏𝑇𝐴
𝑊𝑖𝑡ℎ 𝑏 = + 𝑟
ln( 2⁄𝑟1 ) 𝑎+𝑏𝑇0 𝑎+𝑏𝑇𝐴
the pipe are h1 and h2 respectively. 2𝜋 2𝜋ℎ2 𝑟2 With 𝑏 = + 2𝜋ℎ + 2𝜋ℎ
2𝜋 1 𝑟1 2𝑟2
1 (0.1 + 0.004(10℃))𝑊𝑚−1 𝐾 −1
𝑏= (ln(0.20 𝑚⁄0.05 𝑚) + ) = 0.22509 1 (0.1 + 0.004(100℃))𝑊𝑚−1 𝐾 −1
a.) If the coefficient h1 is very large, so that the temperature of the inner surface may 2𝜋 (25 𝑊𝑚−2 𝐾 −1 × 0.20 𝑚) 𝑏= (ln(0.20 𝑚⁄0.05 𝑚) +
2𝜋 (150 𝑊𝑚−2 𝐾 −1 × 0.05 𝑚)
be assumed to be at To, calculate the steady-state heat flow through the pipe wall
(0.1 + 0.004(10℃))𝑊𝑚−1 𝐾 −1
per unit length. 𝑏 + )
𝑊𝑖𝑡ℎ 𝑐 = 𝑎(𝑇0 − 𝑇𝐴 ) + (𝑇0 2 − 𝑇𝐴 2 ) (25 𝑊𝑚−2 𝐾 −1 × 0.20 𝑚)
b.) Recalculate 𝑞̇ /𝐿 allowing for h1 given in the data below. 2
0.004 𝑏 = 0.23570
𝑐 = 0.1 𝑊𝑚−1 𝐾 −1 (100℃ − 10℃) + 𝑊𝑚−1 𝐾 −2 (100℃2 − 10℃2 )
2
Data: −1
= 28.8 𝑊𝑚 𝑏
r1 = 5 cm 𝑊𝑖𝑡ℎ 𝑐 = 𝑎(𝑇0 − 𝑇𝐴 ) + (𝑇0 2 − 𝑇𝐴 2 )
2
r2 = 20 cm 0.004
𝑞̇ −0.22509±√0.225092 −4(2.0264×10−6 𝑊 −1 𝑚)(−28.8 𝑊𝑚 −1)
h1 = 150 W/m2.K = = 128 𝑊𝑚−1 𝑐 = 0.1 𝑊𝑚−1 𝐾 −1 (100℃ − 10℃) + 𝑊𝑚−1 𝐾 −2 (100℃2 − 10℃2 )
𝐿 2(2.0264×10−6 𝑊 −1𝑚) 2
h2 = 25 W/m2.K = 28.8 𝑊𝑚−1
To = 100°C b) Film heat transfer from fluid to inner pipe wall:
TA = 10°C 𝑞̇ −0.23570 ± √0.235702 − 4(1.1258 × 10−6 𝑊 −1 𝑚)(−28.8 𝑊𝑚−1 )
k = 0.1 + 0.004T in W/mK for T in °C =
𝑞̇ = ℎ1 (2𝜋𝑟1 𝐿)(𝑇0 − 𝑇1 ) 𝐿 2(1.1258 × 10−6 𝑊 −1 𝑚)
𝑇1 = 𝑇0 − 2𝜋ℎ
𝑞̇
(1) 𝑞̇
Solution: 1 𝑟1 𝐿
= 122 𝑊𝑚 −1

a) If h1 is very large, it offers minimal resistance to heat transfer 𝐿

Conduction through the pipe material: Example 1.42: Temperature distribution through a solid conical section
Conduction through the pipe material: 𝑑𝑇
𝑞̇ = −𝑘𝐴
𝑑𝑇 𝑑𝑟
𝑞̇ = −𝑘𝐴 𝑟2 𝑇2 A conical section fabricated in a wall (as below) has a diameter of D = ax. Assuming
𝑑𝑟 1 the lateral surface is well insulated, derive an expression for the temperature distribution
𝑟2 𝑇2
∴ 𝑞̇ ∫ 𝑑𝑟 = − ∫ (𝑎 + 𝑏𝑇)𝑑𝑇 assuming one dimensional heat flow.
1 2𝜋𝑟𝐿
∴ 𝑞̇ ∫ 𝑑𝑟 = − ∫ (𝑎 + 𝑏𝑇)𝑑𝑇 𝑟1 𝑇1
2𝜋𝑟𝐿
𝑟1 𝑇0
𝑞̇ 𝑟 𝑏
ln 𝑟2 = 𝑎𝑇1 + 2 𝑇1 2 − 𝑎𝑇2 − 2 𝑇2 2
𝑏
(2) a) Show an expression for T(x) in terms of T, q, k, a, and x values AND An
2𝜋𝐿
𝑞̇ 𝑟 𝑏 𝑏
ln 𝑟2 = 𝑎𝑇0 + 2 𝑇0 2 − 𝑎𝑇2 − 2 𝑇2 2 (1)
1
expression for T(x) in terms of T and x values ONLY
2𝜋𝐿 1
b) Given the following information, calculate the heat transfer rate (in Watts to 3
Film heat transfer from outer surface of pipe:
significant figures) through the cone.
Film heat transfer from outer surface of pipe: 𝑞̇ = ℎ2 (2𝜋𝑟2 𝐿)(𝑇2 − 𝑇𝐴 ) (3)
𝑞̇ = ℎ2 (2𝜋𝑟2 𝐿)(𝑇2 − 𝑇𝐴 ) a = 0.25
𝑇2 = 𝑇𝐴 + 2𝜋ℎ
𝑞̇
(2) Substitute (1) and (3) into (2): x1 = 50 mm
2𝑟2 𝐿 2
𝑞̇ 𝑟2 𝑞̇ 𝑏 𝑞̇ 𝑞̇ x2 = 250 mm
ln = 𝑎 [𝑇0 − ] + [𝑇0 − ] − 𝑎 [𝑇𝐴 + ] T1 = 600K
2𝜋𝐿 𝑟1 2𝜋ℎ1 𝑟1 𝐿 2 2𝜋ℎ1 𝑟1 𝐿 2𝜋ℎ2 𝑟2 𝐿
Substitute (2) into (1): 2 T2 = 400K
2 𝑏 𝑞̇
𝑞̇ 𝑟2 𝑏 𝑞̇ 𝑏 𝑞̇ − [𝑇𝐴 + ] k = 3.46 W/m.K
ln = 𝑎𝑇0 + 𝑇0 2 − 𝑎 [𝑇𝐴 + ] − [𝑇𝐴 + ] 2 2𝜋ℎ2 𝑟2 𝐿
2𝜋𝐿 𝑟1 2 2𝜋ℎ2 𝑟2 𝐿 2 2𝜋ℎ2 𝑟2 𝐿
Re-arranging to solve for q/L:
Re-arranging to solve for q/L: 𝑟2
𝑞̇ 2 𝑏⁄ 𝑏⁄
𝑞̇ 2 𝑏⁄ 𝑟2 2 2 ] + 𝑞̇ (ln( ⁄𝑟1 ) + 𝑎 + 𝑏𝑇0 + 𝑎 + 𝑏𝑇𝐴 )
( ) [ 2 ] + 𝑞̇ (ln( ⁄𝑟1 ) + 𝑎 + 𝑏𝑇𝐴 ) − [𝑎(𝑇 − 𝑇 ) + 𝑏 (𝑇 2 − 𝑇 2 )] ( ) [ −
(2𝜋ℎ2 𝑟2 )2 (2𝜋ℎ1 𝑟1 )2
0 𝐴 𝐿 𝐿 2𝜋 2𝜋ℎ1 𝑟1 2𝜋ℎ2 𝑟2
𝐿 (2𝜋ℎ2 𝑟2 )2 𝐿 2𝜋 2𝜋ℎ2 𝑟2 2 0 𝐴

𝑏
=0 − [𝑎(𝑇0 − 𝑇𝐴 ) + (𝑇0 2 − 𝑇𝐴 2 )] = 0 q
2
Using the quadratic formula:
Using the quadratic formula:
𝑞̇ −𝑏 ± √𝑏2 − 4𝑎𝑐
= 𝑞̇ −𝑏 ± √𝑏2 − 4𝑎𝑐
𝐿 2𝑎 =
𝐿 2𝑎
Note:
Note:
k (W/mK) = a (W/mK) + b (W/mK2) * T (°C)
k (W/mK) = a (W/mK) + b (W/mK2) * T (°C)
Solution:

Kevin Harding (kevin.harding@wits.ac.za) 63


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Problems

a) Assuming, steady-state, heat transfer only in x direction and well insulated lateral
surface – no heat transferred to wall.

q = -kAdT/dx

With: A = π D2/(4)
But D = ax
⸫ A = (π/4) × a2.x2

Substituting: q = -k [(π/4).a2x2]dT/dx

Rearranging:
4q/[-π.k.a2x2]dx = dT

Integrating:
[4q/- π.k.a2] × (-1/x + 1/x1) = T(x) – T1
T(x) = T1 – (4q/π.k.a2) × (1/x1 – 1/x)

At time t2: q = (T1 – T2)(π.k.a2)/[4 × (1/x1 – 1/x2)]

Substituting q in into T(x): T(x) = T1 + (T1 – T2) × [(1/x – 1/x1)/(1/x1 – 1/x2)]

b) Substituting all the constants to solve for q: q = 2.12 W

Kevin Harding (kevin.harding@wits.ac.za) 64


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos
Heat transfer – Problems

Kevin Harding (kevin.harding@wits.ac.za) 65


School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg (South Africa)
ORCID - research profile ResearchGate - publications & lecture notes YouTube - lecture videos

View publication stats

You might also like