Variable Thermal Conductivity

Dr. Dushyant Singh

Associate Professor
Department of Mechanical Engineering
NIT Jalandhar, Punjab, India – 144027
Email: [email protected]
 Variable Thermal Conductivity, k(T)
Pervious lectures that thermal conductivity of a material, in general,
varies with temperature
When the variation of thermal conductivity with
temperature k(T) is known, the average value of
thermal conductivity in the temperature range
between T1 and T2 can be determined from:
𝑇2
‫𝑘 𝑇׬‬ 𝑇 𝑑𝑇
1
𝑘𝑎𝑣𝑒 =
𝑇2 − 𝑇1
(𝑇1 −𝑇2 )
𝑄𝑃𝑙𝑎𝑛 𝑤𝑎𝑙𝑙 = 𝑘𝑎𝑣𝑒 𝐴
𝐿
(𝑇1 −𝑇2 )
𝑄𝐶𝑦𝑙𝑖𝑛𝑑𝑒𝑟 = 2𝜋𝑘𝑎𝑣𝑒 𝐿
ln 𝑟2 /𝑟1
(𝑇1 −𝑇2 )
𝑄𝑆𝑝ℎ𝑒𝑟𝑒 = 4𝜋𝑟1 𝑟2 𝑘𝑎𝑣𝑒
(𝑟2 − 𝑟1 )
Dr. Dushyant Singh, Department of ME 01 Dr B R Ambedkar NIT Jalandhar
 Temperature Coefficient of Thermal Conductivity
𝑘 𝑇 = 𝑘0 1 ± 𝛽𝑇
 : temperature coefficient of thermal conductivity
𝑑𝑇
Substituting this in the Fourier’s Equation 𝑄 = −𝑘𝐴
𝑑𝑥
𝑑𝑇
𝑄 = −𝑘0 1 ± 𝛽𝑇 𝐴
𝑑𝑥
2 2
Integrating  න 𝑄𝑑𝑥 = 𝐴𝑘0 න 1 + 𝛽𝑇 𝑑𝑇
1 1

2 2
𝛽𝑇 𝛽𝑇22 𝛽𝑇12
𝑄 𝑥2 − 𝑥1 = −𝐴𝑘0 𝑇+ = −𝐴𝑘0 𝑇2 + − 𝑇1 −
2 1
2 2

(𝑇1 +𝑇2 )
= 𝐴(𝑇1 −𝑇2 ) 𝑘0 1+𝛽
2
Dr. Dushyant Singh, Department of ME 02 Dr B R Ambedkar NIT Jalandhar
 Conduction in a plane wall with
(𝑇1 +𝑇2 )
𝑄 𝑥2 − 𝑥1 = 𝐴(𝑇1 −𝑇2 ) 𝑘0 1+𝛽 where
2
(𝑇1 +𝑇2 )
𝑄 𝑥2 − 𝑥1 = 𝐴(𝑇1 −𝑇2 )𝑘𝑎𝑣𝑒 𝑘𝑎𝑣𝑒 = 𝑘0 1+𝛽
2

For Plane wall Case  𝑄𝐿 = 𝐴(𝑇1 −𝑇2 )𝑘𝑎𝑣𝑒

(𝑇1 −𝑇2 )
𝑄𝑃𝑙𝑎𝑛 𝑤𝑎𝑙𝑙 = 𝑘𝑎𝑣𝑒 𝐴
𝐿

For Cylinder Case  𝑄𝐶𝑦𝑙𝑖𝑛𝑑𝑒𝑟 = 2𝜋𝑘𝑎𝑣𝑒 𝐿

(𝑇1 −𝑇2 )
ln 𝑟2 /𝑟1

(𝑇1 −𝑇2 )
For Sphere Case  𝑄𝑆𝑝ℎ𝑒𝑟𝑒 = 4𝜋𝑟1 𝑟2 𝑘𝑎𝑣𝑒
(𝑟2 − 𝑟1 )
Dr. Dushyant Singh, Department of ME 03 Dr B R Ambedkar NIT Jalandhar
 Example 1:
An insulating wall 16 cm thick has one face at 600 oC while the other is at 100 oC .
The thermal conductivity of material is given by k = 0.078 (1+17.9510-4T) W/moC
and T is in oC. Determine the heat loss per unit area.

(𝑇1 −𝑇2 ) k = 0.078 (1+17.9510-4T)

𝑄𝑃𝑙𝑎𝑛 𝑤𝑎𝑙𝑙 = 𝑘𝑎𝑣𝑒 𝐴
𝐿
(𝑇1 +𝑇2 )
𝑘𝑎𝑣𝑒 = 𝑘0 1+𝛽 600oC
2
𝑞𝑃𝑙𝑎𝑛 𝑤𝑎𝑙𝑙
(600 + 100)
𝑘𝑎𝑣𝑒 = 0.078 1 + 17.95 × 10−4 = 0.127 W/moC
2 100oC

(600 − 100)
𝑞𝑃𝑙𝑎𝑛 𝑤𝑎𝑙𝑙 = 0.127 = 396.9 W/m2 0.16 m
0.16

Dr. Dushyant Singh, Department of ME 04 Dr B R Ambedkar NIT Jalandhar

 Example 1
To find the mid plane temperature, the heat flow for thickness up to mid plane is
equated to the heat flow Eq.
W (𝑇 − 𝑇 ) 600 oC
1 𝑀𝑖𝑑
𝑞𝑃𝑙𝑎𝑛 𝑤𝑎𝑙𝑙 = 396.9 = 𝑘𝑚𝑖𝑑
m2 0.08
396.9 𝑊/𝑚2 100oC

(600 + 𝑇𝑀𝑖𝑑)
𝑘𝑀𝑖𝑑 = 0.078 1 + 17.95 × 10−4
2
𝑇𝑀𝑖𝑑
396.9
−4
(600 + 𝑇𝑀𝑖𝑑) (600 − 𝑇𝑀𝑖𝑑 )
= 0.078 1 + 17.95 × 10
2 0.08

𝑇𝑀𝑖𝑑 = 383.8 𝑜𝐶
Dr. Dushyant Singh, Department of ME 05 Dr B R Ambedkar NIT Jalandhar
 Example 2
The thermal conductivity of an insulating material used over a hot pipe varies as
k = 0.0545 (1 + 28.4 × 10–4 T) where T is in °C and k is in W/mK. This insulation is
used for a thickness of 12 cm over a pipe of diameter 0.6 m. The pipe surface is
at 300°C and the outside insulation temperature is 60°C. Determine the heat flow
for a length of 5 m. Also find the mid layer temperature.
(𝑇1 −𝑇2 )
𝑄𝐶𝑦𝑙𝑖𝑛𝑑𝑒𝑟 = 2𝜋𝑘𝑎𝑣𝑒 𝐿 ????
ln 𝑟2 /𝑟1
(𝑇1 +𝑇2 )
𝑘𝑎𝑣𝑒 = 𝑘0 1+𝛽
2
−4 (300+60)
𝑘𝑎𝑣𝑒 = 0.0545 1 + 28.4 × 10 = 0.08236 W/moC
2

(300 − 60)
𝑄𝐶𝑦𝑙𝑖𝑛𝑑𝑒𝑟 = 2𝜋 × 0.08236 × 5 × = 1845.6 W
ln 0.42/0.3

Dr. Dushyant Singh, Department of ME 06 Dr B R Ambedkar NIT Jalandhar

 Example 2
To find the mid plane temperature, equate this
heat flow for half thickness of the cylinder
(𝑇1 −𝑇𝑀𝑖𝑑 )
𝑄𝐶𝑦𝑙𝑖𝑛𝑑𝑒𝑟 = 2𝜋𝑘𝑀𝑖𝑑 𝐿
ln 𝑟2 /𝑟1

−4
(300 + 𝑇𝑀𝑖𝑑)
𝑘𝑀𝑖𝑑 = 0.0545 1 + 28.4 × 10
2

(𝑇1 −𝑇𝑀𝑖𝑑 )
1845.6 = 2𝜋𝑘𝑀𝑖𝑑 𝐿
ln 𝑟2 /𝑟1

−4
(300 + 𝑇𝑀𝑖𝑑) (300 − 𝑇𝑀𝑖𝑑 )
1845.6 = 2𝜋 × 0.0545 1 + 28.4 × 10 ×5×
2 ln 0.36/0.3

𝑇𝑀𝑖𝑑 = 183.47 𝑜𝐶
Dr. Dushyant Singh, Department of ME 07 Dr B R Ambedkar NIT Jalandhar
 Alternative Conduction Analysis
Considering conduction in the system of below figure Assumptions:


𝑸𝒙 Steady-state
No heat generation
No heat loss from the sides

𝑸𝒙+𝒅𝒙

𝑸𝒙
Heat transfer rate (Qx)
System with a constant conduction heat transfer must be a constant
independent of x
For any differential element dx, 𝑸𝒙 = 𝑸𝒙+𝒅𝒙
Dr. Dushyant Singh, Department of ME 08 Dr B R Ambedkar NIT Jalandhar
 Conduction Analysis
Energy conservation requirement, and it must apply even if the area varies with
position A(x) and the thermal conductivity varies with temperature k(T).

 Moreover, even though the temperature distribution may be 2D, varying with x
and y, it is often reasonable to neglect the y-variation and to assume a 1 D
distribution in x.

Fourier’s law in the integral form:

𝑥 𝑇
𝑑𝑥
𝑄𝑥 න = − න 𝑘 𝑇 𝑑𝑇
𝑥0 𝐴(𝑥) 𝑇0

The cross-sectional area may be a known function of x, and the material

thermal conductivity may vary with temperature in a known manner.

Moreover, if the temperature T = T1 at some x = x1 is also known, integration

between x0 and x1 provides an expression from which 𝑄𝑥 may be computed.

Dr. Dushyant Singh, Department of ME 09 Dr B R Ambedkar NIT Jalandhar

 Conduction Analysis
Note that, if the area A is uniform and k is independent of temperature,
pervious Eq reduce as :
𝑄𝑥 ∆𝑥
= −𝑘∆𝑇
𝐴
Where, x = x1 − x0 and T = T1 − T0

We frequently elect to solve diffusion problems by working with integrated

forms of the diffusion rate equations.

However, the limiting conditions for which this may be done should be
firmly fixed in our minds: steady-state and one-dimensional transfer with
no heat generation.

Dr. Dushyant Singh, Department of ME 10 Dr B R Ambedkar NIT Jalandhar

 Example 3
The diagram shows a conical section fabricated from pyroceram. The thermal
conductivity of an insulating material is k = 3.46 W/mK. It is of circular cross
section with the diameter D = ax, where a = 0.25. The small end is at x1 = 50 mm
and the large end at x2 = 250 mm. The end temperatures are T1 = 400 K and T2 =
600 K, while the lateral surface is well insulated.
Derive an expression for the temperature
distribution T(x) in symbolic form, assuming
1D conditions. Calculate the heat rate Qx
through the cone.
𝑸𝒙
Assumptions:
1. Steady-state conditions.
2. 1D conduction in the x-direction.
3. No internal heat generation.
4. Constant properties.

Dr. Dushyant Singh, Department of ME 11 Dr B R Ambedkar NIT Jalandhar

 Example 3
𝑑𝑇 𝜋𝐷2 𝜋𝑎2 𝑥 2
𝑄𝑥 = −𝑘𝐴 Where, 𝐴 = =
𝑑𝑥 4 4
separating variables,
4𝑄𝑥 𝑑𝑥 𝑸𝒙
2 2
= −𝑘𝑑𝑇
𝜋𝑎 𝑥
Integrating from x1 to any x within the cone, and
recalling that 𝑄𝑥 and k are constants, it follows that

4𝑄𝑥 𝑥 𝑑𝑥 𝑇

2
න 2 = −𝑘 න 𝑑𝑇
𝜋𝑎 𝑥1 𝑥 𝑇1

4𝑄𝑥 1 1
𝜋𝑎 2
− +
𝑥 𝑥1
= −𝑘 𝑇 − 𝑇1  4𝑄𝑥 1 1
𝑇 𝑥 = 𝑇1 − 2 − +
𝜋𝑎 𝑘 𝑥 𝑥1

Dr. Dushyant Singh, Department of ME 12 Dr B R Ambedkar NIT Jalandhar

 Example 3
4𝑄𝑥 1 1
𝑇 𝑥 = 𝑇1 − 2 − +
𝜋𝑎 𝑘 𝑥 𝑥1
Although 𝑸𝒙 is a constant, it is as yet an unknown. However, it may be
determined by evaluating the above expression at x = x2, where T(x2) = T2 , Hence

4𝑄𝑥 1 1 𝜋𝑎2 𝑘(𝑇1 − 𝑇2)

𝑇2 = 𝑇1 − 2 −  𝑄𝑥 =
𝜋𝑎 𝑘 𝑥1 𝑥2 4 1ൗ𝑥1 − 1ൗ𝑥2

Substituting for 𝑸𝒙 into the expression for T(x), the temperature distribution
becomes
1 1
−
𝑥 𝑥1
𝑇 𝑥 = 𝑇1 + (𝑇1 − 𝑇2 )
1 1
−
𝑥1 𝑥2

Dr. Dushyant Singh, Department of ME 13 Dr B R Ambedkar NIT Jalandhar

 Example 3
Substituting numerical values into the foregoing result for the heat transfer rate

𝜋𝑎2 𝑘(𝑇1 − 𝑇2)

𝑄𝑥 =
4 1ൗ𝑥1 − 1ൗ𝑥2


𝜋(0.25)2 × 3.46 × (400 − 600)

𝑄𝑥 = = −2.12 W
4 1ൗ0.05 − 1ൗ0.25

When the parameter a increases, the cross-sectional area changes more rapidly
with distance, causing the one-dimensional assumption to become less
appropriate.

Dr. Dushyant Singh, Department of ME 14 Dr B R Ambedkar NIT Jalandhar