Вы находитесь на странице: 1из 52

A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof.

 Pagliano
 

Lecture Course
Building Physics

Prof. Lorenzo Pagliano

Academic Year 2017/2018

Part A

 Transport Mechanisms
 Transport of Heat: Conduction
 Steady-state conduction

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 1  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 

Dispense del Corso di Fisica dell’Edificio.


Sommario Parte A

1 Mechanisms of transport ................................................................................ 3


2 Principle of local balance ................................................................................ 9

3 Flux and flux density of a quantity ................ Errore. Il segnalibro non è definito.
4 Transport of energy for moleular collision or conduction ............................................. 11
4.1 Heat-flow density (Fourier's postulate) in a solid or a fluid at rest .................................... 11
4.2 thermal conductivity ........................................................................................... 16
4.3 Values of conductivity for dry and wet materials .......................................................... 19
4.4 The insulating materials ....................................................................................... 23

5 Elements of vector calculus ........................................................................... 26


6 balance of an extensive quantity relative to a control volume ......................................... 28
7 Demonstration and physical meaning of Fourier equation ............................................. 29
8 Resolution of Fourier eq. in the particular case of flat wall under steady state conditions and without

internal heat generation, with imposed surface temperatures .............................................. 34


9 Definition of thermal resistance and unit thermal resistance .......................................... 38
10 Resolution of Fourier equation: the general case ................................................... 43
10.1 Using and classification of temporal and spatial boundary conditions ............................... 43

11 Special cases and simplifications of the equation ................................................... 45


12 expression in different coordinate systems ......................................................... 45
12.1 flat wall ................................................... Errore. Il segnalibro non è definito.
12.2 Solid or hollow cylinder ................................. Errore. Il segnalibro non è definito.

12.3 Solid or hollow sphere ................................ Errore. Il segnalibro non è definito.

13 Some solutions of the general equation of conduction .............................................. 47


13.1 Conduction through a flat wall with internal heat generation, under steady-state conditions and
imposed surface temperatures. Principle of superposition ......................................................... 48
13.2 Conduction through a cylindrical wall in steady state conditions, with no internal heat generation
and imposed surface temperatures ................................ Errore. Il segnalibro non è definito.

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 2  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 

1 TRANSPORT MECHANISMS

Thermodynamic deals with systems in equilibrium. It can be used to predict how much energy is
needed to move a system from one equilibrium state to another. It can not be used to predict how long it
takes to complete the transformation from one equilibrium state to another because the system is not in
equilibrium during the process and therefore the thermodynamic theory that we have treated so far (in the
previous chapters) is in principle not applicable.
Heat transfer theory adds, to the first and second law of thermodynamics, additional assumptions (from
which one can deduce results comparable with the experience) which can be used for example to determine
how much energy is transferred from one system to another in a unit of time .
We recall here that work and heat are defined in thermodynamics as exchanges of energy between two
systems. In particular heat is defined as a transfer of internal energy between two systems occurring via the
hidden coordinates. Hence we should talk of “energy transfer in the form of heat”, rather than use the
expression “heat transfer”. But this expression is so deeply rooted in practice that we will often use it.

In what follows we frame the energy transport phenomena within the broader context of transport
phenomen.

The transport phenomena are defined as phenomena in which there is a transfer of extensive physical
quantities (mass, energy, momentum) between different parts of a system (that is, between the subsystems
that compose it), or through its boundary (i.e. between the system and the environment).

We consider systems where generally it is possible the movement of macroscopic portions of matter (ie,
portions of material identifiable by means of macroscopic measurement (that is sets consisting of a very large
number of molecules) as opposed for example to the oscillatory motion of the molecules around their
position equilibrium in a solid, or movement of electrons in the conduction band of a solid metal). This is
possible for example to a fluid (in liquid or gas) or a fluid-like (e.g. fine sand). The case in which there is no
macroscopic movement (fluid at rest or solid) will be treated as a special case of the general case.

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 3  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 

../../../../../../../Movies/Wondershare Free YouTube Downloader Videos/YouTube - Laminar Flow then Turbulent


Flow.flv

The transport mechanisms can be classified into two groups:  


 transport by molecular collision (or interaction) exchange  of  extensive  physical 
quantities  (mass,  energy,  momentum)  is  due  to  collisions  or  interactions  between  molecules,  without  any 
displacement  of  macroscopic  portions  of  the  system  in  the  direction  of  displacement  of  the physical 
quantities  
 transport by turbulent collision: the  exchange  of  extensive  physical  quantities  is  due  to  the 
movement of macroscopic parts of matter in the direction of the displacement of the physical quantities 

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 4  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 5  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 6  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 

Example of transport by molecular collision: mass diffusion; a drop of ink falls


in a motionless fluid (e.g. water) and by molecular collision it is diffused into the entire
mass of the fluid 1

Figure A.1:transport by molecular collision.

Diffusione carta crespa.flv


YouTube - ink in water test.flv

Example of transport by molecular collision: diffusion of heat in a solid or a


motionless fluid: in a solid, the vibrations of the crystal lattice and the movement of electrons
in the conduction band (in metals) or collisions between individual molecules in their random
movement (in gases and fluids) enable the progressive transfer of kinetic energy from one area to
another without any movement of macroscopic portions of matter .

Example of transport by molecular collision: laminar flow in a fluid

Figure A.2: fluid with laminar flow

We call laminar flow a flow of a fluid where macroscopic portions of fluid are moving in the x direction,
and at the same time the macroscopic velocity in the y direction is zero; therefore, no net macroscopic
transport of mass occurs in the direction y.
The exchange of energy and momentum in the y direction hortogonally to the macroscopic motion that
occurs in the x direction is due only to molecular collisions

 examples of transport by turbulent collision:

 agitation by scoop,

 convective motion in the vicinity of a hot surface immersed in a fluid,

                                                                 

1 cfr. moto Browniano e sua discussione da parte di A. Einstein. 

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 7  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 
 turbulent flow in a duct.

In the case of turbulent flow in a duct, there are macroscopic movements (ie, movements of
sets consisting of a very large number of molecules) that are transverse to the motion, i.e. the
component wy of the macroscopic velocity in this case is different from zero, accompanied by

fluctuations of the component wy of the macroscopic velocity around the mean value over time. On a

sufficiently long time, from an average point of view, the mass moves only in the x direction and not
in the y direction because the movements up and down give on average a net result equal to zero in
terms of the macroscopic mass, but a ink drop (single molecules distinguishable from those of the
surrounding fluid) and the momentum are transferred quickly even in the y direction thanks to this
macroscopic mixing (vortices). In Reynold’s experience with sufficiently high velocity, we can see the
thread of ink spreads rapidly, highlighting the turbulent vortices (ie macroscopic parts of fluids, groups
of molecules moving "all together").

Figure A.3: at the beginning with low fluid velocities, the dye placed in the tube is configured as a thin thread
and well defined (the case of laminar flow) (a).
We can note that gradually opening the valve that regulates the flow of water, the speed increases and the
initial thread blends with the surrounding water and fills the rest of the duct with a mass of colored water (b).
Looking at the tube with the aid of a light source, we can note as the thread of the dye takes the form of
more or less distinct curls that demonstrate the presence of vortices (c).

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 8  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 

Figure A.4: schema experiment carried out by Reynolds.

2 PRINCIPLE OF  LOCAL  EQUILIBRIUM

The relaxation times (i.e., of redistribution of energy, pressure, temperature, etc. ... until you get a uniform
distribution of these quantities) of an infinitesimal volume dV can be much lower than the time during which
there is interaction with the outside of the volume.
If this condition is met, then the volume can be considered at all times to be in a condition of thermodynamic
equilibrium, and therefore if we isolate this from the rest of the system (with a contour restrictive in relation
to the exchanges of all the extensive variables), it will not change its state.

In the following we will make the assumptions :


 that the relaxation times of an infinitesimal element of volume dV is much shorter than the time
interval in which there is interaction with the outside of the same volume
 and that under these conditions the principle of local balance is valid:

"For each infinitesimal element the relations between u, s, v, t, p, ... of its


thermodynamic equilibrium are valid, the same relations that would apply in the
case the element is isolated and in a state of thermodynamic equilibrium ”

or using an alternative formulation:

“ Every infinitesimal element of a continuum performs a quasi-static


transformation (ie a succession of equilibrium states)." .

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 9  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 
This model works very well for most systems of interest for engineering applications of heat transfer (in the
sense that this assumption, together with others which we will expose below, let us make predictions that
agree with “good” accuracy with the results of experiences).
The cases where this assumption is not applicable (ie, leads to predictions not in agreement with
observations) are studied by the irreversible or non-equilibrium thermodynamics .

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 10  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 

3 TRANSPORT OF ENERGY BY MOLECULAR COLLISION OR 

CONDUCTION 

3.1 Density of heat flow rate (Fourier's postulate) in a


solid or a fluid at rest
The gradient of a scalar quantity X is defined analytically in Cartesian coordinates as the vector :
 X ˆ X ˆ X ˆ
grad X  X  i  j  k
x y z

One can easily verify that in each point the gradient vector of a scalar quantity X points to the direction of
steepest growing of the quantity X in the neighborhood of the point .

Figure A.6: representation of the scalar field of a quantity X. In black are represented the highest values of
the scalar field while in white the lowest, blue arrows indicate the direction and orientation of the gradient of
X (from Wikipedia).

Verification can be done easily in a one‐dimensional case. The following figure shows a scalar quantity 
X  which  decreases  in  the  direction  of  increasing  x  coordinate,  ie,  dX  /  dx  <0. The  gradient  vector  of  X 
points therefore towards the opposite direction as regards the unit vector of x axis.  

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 11  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 
 dX ˆ
X  i where dX/dx < 0
dx

Consider now a SOLID or a fluids at rest (i.e. the velocity of macroscopic portions of fluid is zero, even if the
individual molecules move or vibrate very quickly).

Figure A.7: subsystems A and B at different temperatures.

If two subsystems have different temperatures (for example) and these two subsystems are not separated by
an adiabatic surface (Figure1 A.7), the overall system is not in thermodynamic equilibrium, and, in accordance
with the second law of thermodynamics, it will tend to a state of equilibrium through a process in which the
energy moves from the subsystem at higher T to one with lower T.

If we are in the presence of a temperature field in which T varies from point to point the same reasoning will
apply to each pair of volumes which are in contact and with different temperatures. In each point of the
field, we can calculate a temperature gradient vector (grad T) that has direction and orientation of the
maximum increase of temperature T in the neighborhood of that point .
 T ˆ T ˆ T ˆ
T  i  j  k
x y z
Since (the body being solid or a fluid at rest) the velocity of macroscopic portions of the body is equal to zero,
we’ll have a phenomenon of transport of energy through molecular collision. Since this energy is not
manifestin itself in the macroscopic mechanical form, according to the definitions given in the thermodynamics
theory, we are observing a transfer of energy in the form of heat, which we call from now on "heat transfer
by conduction."

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 12  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 
If we consider the phenomenon from the microscopic point of view, within the solid (or fluid at macroscopic
rest) the molecules have their own microscopic velocity, and in the zone with temperature T 1 the molecules
will have more kinetic energy which is transferred by molecular collisions to the area with lower temperature
T2 .

Consider a system consisting of a cylindrical rod of known length L and cross section area A and laterally
surrounded by a material completely restrictive to heat transfer (ie a perfect insulator, ie, the lateral surface
of the cylinder is subject to an adiabatic constraint Figure A.8).

T1 T1 > T2 T2
 Q D A

L
Figure A.8: cylinder with adiabatic lateral surface and T1> T2.

We keep the bases of the cylinder at different and constant temperatures T1 and T2, for example, with T1>
T2 (for example, we hold the ends in contact with two systems in a condition of two-phase equilibrium at
constant pressure, so that T1 and T2 do not change - see the chapter on the subject of phase transitions -)
for a time interval delta t. Under these conditions, due to the symmetry of the problem, we can expect that
the temperature is the same in every point of any circular section perpendicular to the axis, and it’s different
only in the direction of the length of the cylinder (which we denote by y-coordinate). Given the assumption
T1> T2, the temperature will have a decreasing trend with increasing y-coordinate and it will be independent
of the other coordinates. This is consistent with the fact that, again due to the symmetry of the problem, we
expect a shift of energy in the y direction only, and not radially.

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 13  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 

Figure A.9: hypothetical and plausible trend of the temperature along the bar.

T
We can assume that the density of heat flow [W/m2] is a function of and we know that it is equal to zero
y
T
for 0 (i.e. when there are no temperature differences from point to point). Also, since the heat flow
y
will move in the positive y-axis, i.e. from areas of higher temperature to the lower temperature, the density
of hear flow vector will have the same direction and orientation of the y-axis and will point in the direction of
increasing y, ie same direction but opposite orientation to the gradient of T.

If we stop at the first term of the Taylor series expansion, we can make the hypothesis, as Fourier did, that
the density of heat flow rate vector has the expression:

( Postulate or law of Fourier, 1815 )

J W
With  J Q   
m2 s m2

Where we have introduced the constant of proportionality ƛ, denoted by the name of thermal
conductivity. To indicate the conductivity in English literature the symbol k is used (e.g. ASHRAE), while
elsewhere we use ƛ (UNI for example, in Italy)

Overall it is a tensor of rank two     x, y, z,T,direction , since the conductivity is a physical property

that may depend


on the position for inhomogeneous media,
on the the temperature (and therefore also from the position, since T can depend on the position),
and on direction for anisotropic media (i.e. we have for each Cartesian axis a different value of thermal
conductivity) .

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 14  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 
For homogeneous and isotropic media and portions of matter where temperature variations are not too high,
you can make the approximation that  has uniform value in space (such as choosing as value the average
value of the values taken in the area concerned).

If we are in a case where we can assume that  is a scalar quantity and has uniform value in space, in the

case of the previous graph the flux density vector points in the positive direction of the y-axis as (with
 
the assumptions made on temperature distribution, to fix ideas) it is the case for the physical heat flow:
T T
 0 and   0 and then the product   will be > 0 . 
y y
The fact that the heat flow density points in the positive y-axis is in agreement with the fact that the heat
passes from the point at temperature T 1 (greater) to the one at temperature T 2 (lower).

In the more general case T can also vary in the x and z directions, and the formula will show the gradient
vector of the temperature with non-zero components along all three axes (instead of along a single axis):

 T ˆ  T ˆ T ˆ
T  i  j  k
x y z

Figure A.10: Graph of the temperature gradient. Here are emphasized some isotherms, with T1 <T2 <T3 <T4
<T5. The gradient vector points in the direction of increasing temperature at any point and it’ s perpendicular
in each point to the isothermal passing in that point.

In conclusion, the density of heat flow in the more general case will be written as:

 T ˆ T ˆ T ˆ
J Q  x ( x , y , z ,T ) i   y ( x , y , z ,T ) j   z ( x , y , z ,T ) k
x y z

Reduced, in the case of isotropic substances, to:

   T ˆ  T ˆ T ˆ
J Q    ( x , y , z , T ) T where T  i  j  k
x y z

And in the case where  is uniform in space and independent of temperature

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 15  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 
 
J Q    T

The law applies to all three states of aggregation (provided that the substance is at rest), despite the
diversity of mechanisms. The values of conductivity for a given substance are different depending on whether
it is in the state of aggregation solid, liquid or gaseous.

Since:

 J Q   W m 2
It follows that the units of conductivity are:

 J Q  W / m2 W
    
 dT / dx  K / m m K

3.2 The thermal conductivity


The kinetic theory of gases predicts and experiments confirm that the thermal conductivity of gases is
proportional to the square root of the absolute temperature T and it is inversely proportional to the square
root of the molar mass M .

T

M
Therefore, the thermal conductivity of a gas increases with increasing temperature and decreases with
increasing molar mass. If we try to reduce heat loss through glazed areas, it will be better to fill the chamber
with argon (M = 40) or Kripton ( M= 83) than with air (M = 29), and you certainly will not use helium, whose
thermal conductivity (M = 4) is much higher.
For this reason the use of Argon or Kripton in double glazing instead of air improves the thermal insulation.
The values of thermal conductivity of gases in practice can be considered independent of pressure.

The mechanism of heat conduction in a liquid is complicated by the fact that the molecules are arranged
closer together, creating a force field more intense. The thermal conductivities of liquids have values between
those of solids an gases. The thermal conductivity of a substance is normally higher in the solid phase and
lower in the gas phase.
Unlike gas, the thermal conductivities of most liquids decrease with increasing temperature, except for water
which is an important exception.
Like gas, the thermal conductivity of liquids decreases with increasing molar mass.
Liquid metals such as mercury and sodium have high thermal conductivity and their use is very convenient in
applications where you want to transfer a high heat output to a liquid, such as nuclear power plants. But
there are problems of toxicity and explosion hazards associated with these fluids.

The thermal conductivity in solids is obtained by summing the effects of heat conduction due to the waves of
vibration of the lattice, produced by the vibratory motion of the molecules that occupy relatively fixed
positions (component of lattice), with the energy carried by the free flow of electrons (electronic
component ). The thermal conductivities of pure metals depend mainly on the part of the lattice, in turn
strongly dependent on the arrangement of moleculas: for example, the diamond, highly ordered crystalline
solid, has the highest thermal conductivity, at ambient temperature.

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 16  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 
Unlike metal, good conductors of electricity and heat, the crystalline solids such as diamond and
semiconductors such as silicon are good conductors of heat but bad electrical conductors and for this reason
they are widely used in the electronics industry: for the excellent thermal conductivity of the diamond,
despite the high price, the diamond heat sinks are used in cooling of electronic components especially
sensitive; silicon oils and seals are commonly used in the assembly of electronic components, because they
provide good heat conductivity and good electrical insulation.
Since the pure metals have high thermal conductivity, we expect that even the metal alloys have high
conductivity between the conductivities of the metal constituents. It may be noted that the thermal
conductivity of an alloy of two metals is usually much lower than that of each metal , because in a pure metal
even small amounts of “extraneous” molecules, even if belonging to good conductors materials,
disrupt severely the flow of heat, for example, the thermal conductivity of steel containing just 1% of
chromium is 62 W / (m K), while the thermal conductivity of iron and chromium, are respectively, 83 and 95
W / (m K).
As mentioned the thermal conductivity of materials varies with temperature; this variation in certain
temperature ranges is negligible for some materials, but significant for others (Figure A.11).

Figure A.11: variations of thermal conductivity of some solids, liquids and gases with temperature.

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 17  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 

The thermal conductivities of some solids show a significant increase at temperatures close to absolute zero,
where these solids become superconductors, for example, the conductivity of copper reaches a maximum
value of 20 000 W / (m ° C) at 20 K, corresponding to approximately 50 times the conductivity at room
temperature.

Since the dependence of thermal conductivity on the temperature complicates considerably the analysis of
the conduction , in practice on ranges of temperature not too high, is adopted a constant value of thermal
conductivity , calculated at an average temperature.
In the study of heat transfer usually we consider the material as isotropic, ie with constant properties in all
directions. This assumption is realistic for most materials except for those that show different structural
characteristics in different directions, such as laminated composite materials and wood: for example, the
thermal conductivity of the wood perpendicularly to the fibers is different than the one parallel the fibers.

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 18  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 
3.3 Values of conductivity for dry and wet materials

Figure A.12: Values of thermal conductivity (real or apparent) for various materials. .

The thermal conductivity is a property of the material. It does not depend on the size of the material but on :
 temperature
 density
 chemical composition of the material .

The thermal conductivity values reported in the following tables are valid for values of temperature normally
occurring in buildings. There are no significant variations between 273-343K (0-70 ° C). When we consider
situations with very high temperatures, such as in ovens, we must take into account the influence of
temperature.
Generally, lightweight materials are better insulators than heavy materials since lightweight materials often
contain significant amounts of air. The still and dry air has very low conductivity. Please note however, that
thermal energy can be transferred through a layer or air bubble by radiation and convection (see for example
discussion on double glazing).
The apparent thermal conductivity is defined as the value of conductivity which should have a homogeneous
material in order to transfer the same flow that is actually transmitted by the inhomogeneous material under
the same conditions of thickness, frontal area, temperature difference applied to faces.

It’s prevalently the low-conductivity of the still gas which keep down the value of the apparent conductivity

Figure .... Structure of Rockwool, from wikipedia

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 19  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 
The apparent thermal conductivity is in fact defined as such that its use in conjunction with the postulate of
Fourier provides a value of the flow equal to the one which would be measured in practice with the same
delta T, despite the presence of convection and radiation besides conduction: i.e. we define:

T
 misurato   apparente A
x
And the explicit definition of the apparent thermal conductivity is:

 misurato
 apparente 
T
A
x
In Table A.1 all the values of conductivity of the materials containing air spaces, are apparent conductivities.

When a material becomes wet, the air pockets within the material are filled with water.Since water is a better
conductor than the air, as a result also the conductivity of the material increases. From here it follows that it
is very important to install insulating materials when they are dry and to operate so that they remain so.

Table A.1: The table shows a list of building materials and their relative conductivity when they
are dry and wet.

actual or apparent thermal


Density - ρ
conductivity [W/mK]

[kg/m 3 ] dry wet

Metals

Aluminium 2800 204 204

Copper 9000 372 372

Lead 12250 35 35

Steel, Iron 7800 52 52

Zinc 7200 110 110

Natural stone

Basalt, Granite 3000 3,5 3,5

Bluestone, Marble 2700 2,5 2,5

Sandstone 2600 1,6 1,6

Masonry

Brick 1600-1900 0,6-0,7 0,9-1,2

Sand-lime brick 1900 0,9 1,4

1000-1400 0,5-0,7

Concrete

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 20  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 

Gravel concrete 2300-2500 2,0 2,0

Light concrete 1600-1900 0,7-0,9 1,2-1,4

1000-1300 0,35-0,5 0,5-0,8

300-700 0,12-0,23

Pumice powder concrete 1000-1400 0,35-0,5 0,5-0,95

700-1000 0,23-0,35

Isolation concrete 300-700 0,12-0,23

Cellular concrete 1000-1300 0,35-0,5 0,7-1,2

400-700 0,17-0,23

Slag concrete 1600-1900 0,45-0,70 0,7-1,0

1000-1300 0,23-0,30 0,35-0,5

Inorganic

Asbestos cement 1600-1900 0,35-0,7 0,9-1,2

Gypsum board 800-1400 0,23-0,45

Gypsum cardboard 900 0,20

Glass 2500 0,8 0,8

Foam glass 150 0,04

Rock wool 35-200 0,04

Tiles 2000 1,2 1,2

Plasters

Cement 1900 0,9 1,5

Lime 1600 0,7 0,8

Gypsum 1300 0,5 0,8

Organic

Cork (expanded) 100-200 0,04-0,0045

Linoleum 1200 0,17

Rubber 1200-1500 0,17-0,3

Fibre board 200-400 0,08-0,12 0,09-0,17

Wood

Hardwood 800 0,17 0,23

Softwood 550 0,14 0,17

Plywood 700 0,17 0,23

Hard-board 1000 0,3

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 21  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 

Soft-board 300 0,08

Chipboard 500-1000 0,1-0,3

Wood chipboard 350-700 0,1-0,2

Synthetics

Polyester (GPV) 1200 0,17

Polyethene, Polypropene 930 0,17

Polyvinyl chloride 1400 0,17

Synthetic foam

Polystyrene foam, exp. (PS) 10-40 0,035

Ditto, extruded 30-40 0,03

Polyurethane foam (PUR) 30-150 0,025-0,035

Phenol acid hard foam 25-200 0,035

PVC-foam 20-50 0,035

Cavity wall filling

Cavity wall isolation 20-100 0,05

Bituminous materials

Asphalt 2100 0,7

Bitumen 1050 0,2

Water

Water 1000 0,58

Ice 900 2,2

Snow, fresh 80-200 0,1-0,2

Snow, old 200-800 0,5-1,8

Air

Air 1,2 0,023

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 22  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 

Soil

Woodland soil 1450 0,8

Clay with sand 1780 0,9

Damp sandy soil 1700 2,0

Soil (dry) 1600 0,3

Floor covering

Floor tiles 2000 1,5

Parquet 800 0,17-0,27

Nylon felt carpet 0,05

Carpet (with foam rubber) 0,09

Cork 200 0,06-0,07

Wool 400 0,07

3.4 Insulating materials


Thermal insulation products are used primarily for resistance to heat flow. Most insulation is made of
heterogeneous materials of low thermal conductivity and air cavities, the latter characterized by one of the
lowest thermal conductivity: examples are a bale of straw, fur animals and by derivation a wool sweater ,
polystyrene, the rock wool, glass wool, fleece or polar fabric ....
The flow of heat between two bodies at different temperatures can be slowed by placing "barriers" along its
path, consisting of thermal insulation, that are particularly important in the design and manufacture of all
devices and "energy-efficient" systems. The energy savings achievable with the isolation is not limited to hot
surfaces. You can save energy and money, in fact, isolating cold surfaces (surfaces whose temperature is
lower than the ambient temperature), such as chilled water piping, cryogenic tanks, reserve, refrigerated
trucks and air-conditioning channels, since the "cold" is produced by the cooling machine that takes energy to
work, provided generally in the form of electricity. In the absence of thermal insulation, the heat moves from
the external environment to the refrigerated environment with the result that the refrigerated chiller must
work more strenuous and longer to compensate for this heat gain, consuming more electricity or power: For
example, a refrigerator with well insulated walls, will consume much less power than another similar one,
which has lower or null thermal insulation . (see for example www.topten.info, www.topten.ch, www.eerg.it
Micene project report) ,

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 23  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 

www.topten.ch
The thermal insulation in the form of mud, clay, straw, rags and strips of wood was used for the first time in
the eighteenth century on steam engines, in order to avoid that the workers were burned by hot surfaces. As
a result, temperatures in the engine room dropped and there was evidence of a reduction in fuel consumption,
which stimulated the search for materials with improved thermal efficiency. One of the first materials used
for the thermal insulation was the mineral wool which was accidentally discovered in 1840, when a producer
of iron in Wales directed a jet of steam at high pressure at the slag of a blast furnace. Subsequently, this rock
wool byproduct of the manufacture of cannons for the Civil War, made its way rapidly in various industrial
uses. Since 1880, manufacturers began to install mineral wool even in homes.

The energy crisis of 1970 greatly increased the public aware about the importance of energy and on
thelimitedness of reserves, promoting energy saving. Save energy by reducing the heat flow is the main
reason for the use of thermal insulation, so you have materials of satisfactory performance in the temperature
range from -268 ° C to 1000 ° C. Most insulation is accomplished by blocking air or other gas into a matrix of
fibers, or flakes, or foam. These insulators, since the propagation of heat occurs by conduction through the
solid material, and by conduction, convection and radiation through the air, are described by means of the
apparent thermal conductivity, which is often much less than the solid material into homogeneous
configuration (see cellular cement in comparison with dense cement, a mineral - rock or glass - in comparison
with mineral wool- rock wool, glass wool-

Insulation elements with extremely low apparent thermal conductivity (about one millionth of that of air),
called superinsulating, are achieved by using highly reflective layers separated by glass fibers in vacuum.

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 24  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 

Figure A.13: Values of thermal conductivity (actual or apparent) for various materials

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 25  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 

4 REFERENCES FOR CALCULATION  V ECTOR

Vectors as directional derivatives

A vector may also be defined as a directional derivative: consider a function f(xα) and a
curve xα(τ). Then the directional derivative of f is a scalar defined as

where the index α is summed over the appropriate number of dimensions (e.g. from 1 to 3 in 3-
dimensional Euclidian space, from 0 to 3 in 4-dimensional spacetime, etc.). Then consider a
vector tangent to xα(τ):

We can rewrite the directional derivative in differential form (without a given function f) as

Therefore any directional derivative can be identified with a corresponding vector, and any
vector can be identified with a corresponding directional derivative. We can therefore define a
vector precisely:

Interpretations of the gradient

Consider a room in which the temperature is given by a scalar field T, so at each point (x,y,z)
the temperature is T(x,y,z) (we will assume that the temperature does not change in time).
Then, at each point in the room, the gradient of T at that point will show the direction in which
the temperature rises most quickly. The magnitude of the gradient will determine how fast the
temperature rises in that direction.
Consider a hill whose height above sea level at a point (x,y) is H(x,y). The gradient of H at a
point is a vector pointing in the direction of the steepest slope or grade at that point. The
steepness of the slope at that point is given by the magnitude of the gradient vector.
The gradient can also be used to measure how a scalar field changes in other directions, rather
than just the direction of greatest change, by taking a dot product. Consider again the example
with the hill and suppose that the steepest slope on the hill is 40%. If a road goes directly up
the hill, then the steepest slope on the road will also be 40%. If instead, the road goes around
the hill at an angle with the uphill direction (the gradient vector), then it will have a shallower
slope. For example, if the angle between the road and the uphill direction, projected onto the
horizontal plane, is 60°, then the steepest slope along the road will be 20% which is 40% times
the cosine of 60°.
This observation can be mathematically stated as follows. If the hill height function H is
differentiable, then the gradient of H dotted with a unit vector gives the slope of the hill in the
direction of the vector. More precisely, when H is differentiable the dot product of the gradient

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 26  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 
of H with a given unit vector is equal to the directional derivative of H in the direction of that
unit vector.

Formal definition

The gradient (or gradient vector field) of a scalar function f(x) with respect to a vector variable
is denoted by or where (the nabla symbol) denotes the vector
differential operator, del. The notation is also used for the gradient.
By definition, the gradient is a vector field whose components are the partial derivatives of f.
That is:

(Here the gradient is written as a row vector, but it is often taken to be a column vector; note
also that when a function has a time component, the gradient often refers simply to the vector
of its spatial derivatives only.)
The dot product of the gradient at a point x with a vector v gives the directional
derivative of f at x in the direction v. It follows that the gradient of f is orthogonal to the level
sets of f.

Vector operations

Vector calculus operations are functions between either scalar and vector fields or two vector
fields . They are typically expressed in terms of the del operator ( ). The four most important
operations in vector calculus are:

Operation Notation Description

Gradient Measures the rate and direction of change in a scalar field.

Curl Measures the tendency to rotate about a point in a vector


field.

Divergence Measures the magnitude of a source or sink at a given point


in a vector field.

Laplacian A composition of the divergence and gradient operations.

Divergence The integral of the divergence of a vector field


theorem over some solid equals the integral of the flux
through the surface bounding the solid.

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 27  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 

5 BALANCE OF AN EXTENSIVE QUANTITY WITH  RESPECT   TO   A  

CONTROL  VOLUME  

The balance equations of an extensive quantity with respect to a well-defined control volume (C.V.) have the
generic form of the type:

Time‐varying  size  X  "contained"  in  the  C.V. 


=  Net  flow  of  X  incoming  to  C.V.  through  the  boundary 
+ Generation rate of the quantity X within the C.V.

Consider an example. To calculate with respect to time, the variation of the amount of bicycles in Italy (we
assume, therefore, as control volume the national territory and as its borders its political boundaries), you will
have to take account of imports and exports (the latter can be considered as flows with negative value) and
at the same time of bicycles produced in Italy and of those demolished, all related to a certain interval of
time, eg a year.We can therefore write:

Which is a balance done on the entire control volume. This equation will have as unit the same of the quantity
X divided by the unit of time i.e.  X  t 1 , in our example: number of bicycles (dimensionless) / years
In the relation:

dX / dt is the variation of the X quantity in control volume per unit of time chosen (for example, the change
in the number of bicycles in Italy in one year)

 in
X is the sum of the flows of X through the boundary of the C.V. with the convention to be represented

by positive relative numbers if entering the C.V. and negative if outgoing, i.e.  in
X represents the net

balance between import and export of bicycles in a year or "net inflow" in the C.V.

is the sum of "generations" of bicycles in the C.V. (Represented by positive relative numbers if

productions and negative if demolitions) that is the difference between the bicycles produced and destroyed
within the boundary of the control volume (or, in other words, the net production) per unit of time chosen.

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 28  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 

6 DEMONSTRATION AND PHISICAL MEANING OF  THE 
EQUATION OF  FOURIER

Since here we’re dealing with conduction, we only deal with cases in which the energy is propagated by
molecular collisions, and therefore there is no macroscopic movement of matter (we are so in a solid or in a

fluid at rest). This means that we assume w  0 (the assumption that the macroscopic velocity is zero at
every point implies that the transport of energy occurs only at the level of molecular collisions, does not
appear as macroscopic mechanical work, so for our definitions, it is heat transfer through the conduction
mechanism).
Consider a surface S, and an infinitesimal element of area dA on this surface. The orientation in the space of
this element can be described by the unit vector (vector whose lenght is 1) n̂ with direction perpendicular to
the element and orientation towards the interior of the control volume.

If at that point the flux density vector JQ is known, then you can calculate the infinitesimal flow through the

element with area dA as d   J Q  n̂ dA .


Figure A.14: graphical representation of the flux density vector J Q and the unit vector n


We note that the scalar product J Q  n̂ (according to the definition of scalar product of two vectors

a b  a  b cos where  is the angle formed by the two vectors) will be:

 A positive number if the flux density vector forms an angle less than 90 ° with the unit vector n̂ and
that this situation is physically an input of energy in the control volume
 A negative number if the flux density vector forms an angle more than 90° with the unit vector n̂
(this physically coincides with an output in the control volume)
 a zero value if  = 0.

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 29  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 

Figure A.15: the dot product between the flux density vector and the unit vector normal to the surface is
positive when  <90 ° (a) and negative when > 90 ° (b), having chosen the unit vector entering the CV.


So d   J Q  nˆ dA it is a positive number when the energy is physically entering the CV, i.e. coincides with

a quantity that we denote with the superscript "IN":



d   J Q  nˆ dA  d  IN [W]

The sum of d IN along all the closed surface S takes into account both the areas where the flow is
incoming and the ones where it is outgoing; it is therefore the net heat flux entering. Being a sum of infinite
infinitesimal terms, it can be written as an integral:


in   ( JQ  n̂) dA [W]
S


Where (we recall) J Q [W/m 2 ] is the vector of heat flux density and n̂ is the unit vector normal to the

surface dS pointing to the inner part of the CV.

In addition we have seen previously that the density of heat flow by conduction is expressed as
 
J Q    T [ W/m 2 ], according to Fourier's postulate.
We can then write:


 IN   (  T )  n̂ dS [W]
S

Applying the divergence theorem, with n̂ as the normal unit vector directed towards the interior of the

volume V (note the minus sign, which depends on the choice of the direction of the unit vector n̂ 2):

 
  A  nˆ  dS       A  dV

S V

where the divergence of a vector A is defined as:

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 30  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 
  A A A
 A  x  y  z
x y z

We get in our case:


    
 IN  
S
J  ˆ
n dS     J dV     (
V
 T ) dV
V

• If we introduce the hypothesis that conductivity does not depend on the position, the direction and

T then conductivity can filter out of the operator  ,and the integral since they operate both on
spatial coordinates. Note that  must be independent of T. In general T can vary from point to point
and in time, so if  depends on T, it would also depend on the spatial coordinates and the time

With this assumption we obtain that the expression of the net heat flux entering the finite volume V is

   
 
inQ        T dV         T dV      2T dV
V V
  V
  [W]

Where we introduced the Laplacian operator as application of the divergence operation to a vector obtained
through an operation of gradient (which can be indicated symbolically as the scalar product of the divergence
operator and the gradient operator - remember the expression of scalar product of two vectors a and b in
terms of their Cartesian components:
 
a  b  a x bx  a y by  a z bz - ) :

  2 2T 2T 2T


  (T )   T   
x 2 y 2 z 2
Notice that this operation depends on position, as the derivatives do.

Since the volume V on which we carry out the balance can be chosen arbitrarily (for all volumes V must apply
energy conservation) the expression of the net flow entering any infinitesimal volume dV is therefore:

d  in Q    2T dV

The energy balance equation must be valid for any volume, so even for the single infinitesimal volume dV, for
which we can write:

dE
  inQ  

Q [W]
dt  

B C
A

Ove:
A is the energy change per unit time in the infinitesimal volume dV in question ;
B is the net heat flux entering the infinitesimal volume dV ;
C is the power generated in the infinitesimal volume dV .

                                                                                                                                                                                                               

2 Si noti che nella formula del teorema della divergenza presentata nel capitolo sui vettori in vettore unitario  n̂ è 
stato scelto con direzione verso l’esterno  
Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 31  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 
The equation in this case then becomes :

dE
dt
   dV
  2T dV    [W]

where is the power generated per unit of volume , in W/m 3

To make explicit an expression for dE / dt we have to introduce the following assumptions :

 macroscopic velocity constantly equal to zero and fixed control volume, so there is no change in
kinetic or potential energy

 in the system considered will not occur phase change phenomena (otherwise we should introduce a
term for the heat of phase change ,...)

 the volume is constant (we consider only ideal solids and liquids and gas in rigid containers), then,
since V = const, the deformation work is zero (it will be realistically negligible compared to other
energy exchanges onside

The variation of the total energy E, therefore, coincides with the change in internal energy U of the volume
dV. Since the exchange of work is zero, and the system is closed (there is no movement, so there is no mass
exchange), the first principle for closed systems tells us :

dU   Q in   W in   Q in [J]

 Q IN
Reminding that csp   dU   Q IN  c M dT  c dV dT
dT M
then we get :

dU   c dV dT [J]

where "c" is the specific heat at constant volume; in the case of incompressible fluids and solids it varies little
with the nature of the transformations and then we can use Cp. The local balance equation (ie for the volume
dV) becomes:

c
T
t

dV   2T dV   
  dV [W]

keeping into consideration the fact that the derivative of T with respect to time is a partial derivative, because
it is calculated in the point on which is centered the point V, ie x, y, z are fixed.

By simplifying the common factor dV, we obtain the Fourier equation (equation of heat conduction).

T  
c   2T   [W / m3 ]
t
Which can also be rewritten in the form:

T  2  

 T  [K/s]
t  c c

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 32  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 
In summary, the Fourier equation is applicable when the following assumptions are satisfied:
1. local equilibrium
2. conductivity independent of position, direction and temperature
3. zero macroscopic velocity (solid or fluid at rest from the macroscopic point of view)
4. there are no ongoing phase changes
5. the volume is constant (we consider only ideal solids and liquids and gases at rest and which do
not change volume significantly with T)

We define :


DEF:  [m2 / s ] thermal diffusivity of the medium
c

which is also called the coefficient of molecular diffusion of heat.


In the absence of heat generation and one-dimensional conduction (ie, T depends only on x) we obtain
the equation:

T  2T
 2 [ K/s]
t x

formally analogous to the equation of mass diffusion or Fick equation that will be discussed in the chapters
that discuss transport of mass.
The thermal diffusivity  of the medium has dimensions:

 T  K
  2

    2tT 
K
s  m  [ D]
  s
 x2  m 2

That, we will see are the same dimensions of the mass diffusion coefficient D.
Not to be confused with:

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 33  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 

7 RESOLUTION  OF  FOURIER EQ . IN THE PARTICULAR CASE OF  


FLAT WALL  UNDER STEADY‐ STATE CONDITIONS ,  WITHOUT 
INTERNAL HEAT GENERATION,  WITH IMPOSED SURFACE 
TEMPERATURES .

Piano  x = s  
Piano  x = 0  
Supponiamo  che  T  abbia 
Supponiamo  che  T  abbia  lo 
lo  stesso  valore  T 2   in  
stesso  valore  T 1   in    tutti  i 
tutti  i  punti  di  questo 
punti  di  questo  piano  e  che 
piano  e  che  sia  costante 
sia costante nel tempo. 
nel tempo. 


0  s 

Figure A.16: schematization of undefined flat wall.

Consider a wall consisting of a homogeneous solid, with constant conductivity λ, thickness s in one direction (
assume this as x direction ) and infinite-dimensione in the two perpendicular directions (in practice we might
think to have this situation when the sizes in the other two directions are much larger than s).

There is no heat generation inside the wall.


A face is kept at temperature T 1 uniform on the whole face and constant over time, and the opposite is kept
at temperature T 2 uniform on the whole face and constant over time (to fix ideas suppose that T 1 > T 2 ).
We want to calculate:

1. the  temperature  field  in  each  point  x   of  the  solid  and  in  every  instant  of  time  t,  ie  the 

function T  T (x ,t )

2. the  flux  density  in  each  point x   of  the  solid  and  in  every  instant  of  time  t,  ie  the  function
  
   (x ,t )

To do this we can integrate the Fourier’s equation to find T (x, y, z, t) and then apply the Fourier’s postulate
to calculate the flux density.
We start by integrating Fourier’s eq.:

T  
c  2T  
t
Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 34  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 
Inside the wall there is no production of heat, therefore we can set . Due to the symmetry of
geometry and boundary conditions, all dependent on the variable x alone, the solution can only depend on x
alone, ie T (x, y, z, t) = T (x, t).
In addition, the boundary conditions are constant over time, so even if at the beginning the wall is likely to be
place of time-dependent phenomena, we can assume that after a certain time, the temperature distribution is
going to stabilize on a given function T (x) , the flux stabilizes on constant values over time, that the
internal energy stabilizes on a constant value, ... In short, we can assume that after an initial transitory all
the quantities will lead to constant values, ie, no more variable over time, i.e. that all the partial derivatives
with respect to time are zero. We can say that we are in steady state conditions, and thus in particular we
have:

T
 0
t

The Fourier equation is therefore reduced to the expression,  2T  0 , and considering that the problem is

one-dimensional, everything is reduced to the form: :

 2T
0
x 2

An ordinary differential equation of second order in x alone, 2° , whose solution will be of the
 
form T ( x, y , z , t )  T ( x ) ,consequently, the functions T  T ( x, y, z , t ) and    ( x , y , z , t ) , which we are

looking for will depend on the variable x alone, i.e. we are now looking for:
1. the temperature field in each section where x = const of the solid, ie the function T  T ( x) that is
independent of time
 
2. the flux density in each section where x = const of the solid, ie the function   (x ) that is
independent of time

The equation can then be rewritten (being T function of x alone, to derivate with respect to x means to
calculate the total derivative) as:

d  dT 
dx  dx   0
 

and solved with two successive integrations:


d  dT  dT
0   A  cos t  T  Ax  B
dx  dx  dx
Thus the functional form of T (x) is linear in x; on a graph (T, x) will be represented by a straight line.
But this line is not determined as the constants of integration can take any values (in the sense that the
function T = Ax + B satisfies the equation for any choice of the cost values of A and B, (you can test it
directly by calculating the second derivative with respect to x).
We must then determine the values of A and B specific for our specific problem.
This is possible through the use of what are called boundary conditions as they describe the physical situation
actually present on the boundary of the volume in question (in this case, the fact that the temperatures on
both sides are uniform and constant in time):

T ( x)  T1 per x  0

 T ( x)  T2 per x  s

From which follows:

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 35  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 
 B  T1
 B  T1 
   T2  T1
 A  s  B  T2  A  s

T2  T1 T1  T2
T  Ax  B  x  T1  T1  x
s s
The temperature is then a linear function of x, which, in the range (0, s) its maximum value (T1) at x = 0,
and decreases linearly as x increases, reaching a value T2 at x = s.
This temperature profile is a function of x alone is not time dependent, as we expected since the boundary
conditions are independent of time and we have chosen to neglect the initial transient and steady focus on
the conditions that are established after the transient.

The density of heat flow:

  dT ˆ  T1  T2  T1  T2 ˆ
J Q    T    i   ˆ
dx   s  i   s
i [W/m 2 ]
 
0

is a vector pointing in the direction of x in the direction of increasing x, ie, the decreasing T, in agreement
with the second principle.

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 36  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 

Figure A.17: diagram, temperature and direction and orientation of the flux density. .

We calculate the heat flux and heat flux through the x-axis in the positive direction a vertical section of the
solid, for example at x = 0, chosing n as having the same direction and orientation of i, we obtain:

T T T T T T
 

 Q  Q   J Q  n̂ dA    1 2 iˆ  n̂ dA    1 2 dA   1 2 A [W]
A A
s A
s s

where A is the area of the wall in m 2 .

We note that the flow

It is not time dependent (obviously since we have imposed at the outset that T and all other variables are not
dependent on time, as a consequence of the fact that the boundary conditions do not depend on time),

and depends on space, that is, by the coordinate x. The flow has exactly the same value in all sections of the
wall (each section is a plane corresponding to a certain value x = const),

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 37  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 
This latter finding is consistent also with the fact that the hypothesis of the absence of generation and
stationary balance of energy for a solid layer contained between any the two sections x = x 1 and x=x 2 is:

dU
   IN    x  x1   x  x2  0
dt

8 DEFINITION OF  THERMAL  RESISTANCE AND UNIT 


THERMAL  RESISTANCE 

If we rewrite the flow or thermal power transmitted through the wall (flat indefinite, and steady-state with no
internal heat generation) in the form:

T T
Q    1 2 (W)
s
A

we may be tempted to write the heat flux in the form:

T T
Q    1 2 ( Equation A)

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 38  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 
defining the “thermal resistance” as follows:

T1  T2
DEF:   [K/W] Thermal Resistance

It follows tha the expression of the conductive thermal resistance in the particular case of plane geometry is:

s
 [K / W ]
 A

NOTE: pay attention to the distinction between the definition of thermal resistance and its expression in
specific cases. The definition of resistance is generally not dependent on the particular mechanism of
transport of energy. Applying the definition in various geometry configurations we get different
s
expressions. In the case of conductive flat wall the expression of the unit resistance is  but will be
 A
different depending on the geometry (eg cylindrical or flat wall conduction in applying the definition above we
get different expressions, as we shall see shortly).

We can compare the expression (A) with the expression that links electrical current (I in amps), potential
difference (V in volts) and electric resistance ( elett , ohms) in a linear conductor (Ohm's law):

V1  V2
I
elett
The formulas of the electric current (flow of electric charge) and heat flux are then presented showing a
similarity: in both the flow is proportional to a "driving force" (respectively electric potential difference
between two points or temperature) and inversely proportional the resistance between those two points
(electrical or thermal). We will see below that the properties of the thermal resistances are shown entirely in
the field of study of heat transfer.

NOT been shown since by analogy with the formulas of Ohm's electricity and ARE TOTALLY INDEPENDENT.

In addition, we will see that you can define the thermal resistance for convective and radiation phenomena.
There is in fact a theory of thermal networks, just as there is a theory of electrical networks.

The two theories are completely independent, while presenting some formal analogies.

Returning to the phenomenon of thermal conduction only, the formulas of written result for a flat wall thermal
resistance is indefinite (in accordance with physical intuition):
• s is directly proportional to the thickness of the wall;
• inversely proportional to the area in which it passes through the wall of the heat;
• inversely proportional to the conductivity λ.
The thermal resistance is measured in K / W as:

T1  T2 K 
 W 
  

T T
NOTE: The relationship Q    1 2 between flow, delta T and resistance, of course, is valid ONLY when

the assumptions are valid according to which we have obtained:
 stationary conditions (which occur if the boundary conditions are stationary, and when the initial
transient is exhausted);

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 39  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 
 absence of heat generation in the wall.

 one-dimensional flow

 assumptions and validity of the Fourier equation

s
As already noted the particular expression of the resistance  depends on the fact that we have
 A
considered a case of plane geometry. We will see that for one-dimensional flows in different geometries you
can still write heat flow = (temperature difference / resistance) but the expression of the resistance as a
function of conductivity and geometrical parameters will be different.

Furthermore, if and only if the assumptions


 plane geometry
 layers in series, and all having the same area A
the expression of the flow can be rewritten by relating it to a unit area, as follows, and be the same in each
section.

 Q T T
  JQ    1 2 [W/m2]
A A s
This suggests that we can rewrite the density of heat flow in the form:

T1  T2

R

defining the “unit thermal resistance” R as:

T1  T 2  m 2K 
R   
  W 

which brings, only in this case of a flat wall, to the expression:

s  m2 K  s
R   ( Compared with   K/W)
  W   A

Since:

 T1  T2
 
A R
T T T T
 A 1 2 A  1 2
R 

it follows:

R

A

that is "thermal resistance (  )" = "unit thermal resistance ( R )” DIVIDED by area.

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 40  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 
We see then that the unit thermal resistance is NOT the thermal resistance per unit area. The name
probably derives from the unit thermal resistance being defined starting from the heart flow density
(ie, flow per unit area), while the resistance is defined starting from the heat flow.

T1  T 2  m 2K 
R   
  W  is usually referred to as "R value" or "unitary thermal
In the U.S. the quantity
resistance"

In this course we will use


T1  T2 K 
the "thermal resistance" , defined as   W 
  
T1  T2  m2 K 
and " unit thermal resistance " defined as R  
  W 

Figure A.18: Graph comparing the unit thermal Figure A.19: Graph comparing the unit thermal

resistance (s / λ )  m K  of different materials resistance (s / λ )  m K  of different materials in a


2 2

   
 W   W 
with the same thickness (10mm) . building typical thickness (note that the scale is
different compared to Figure 1.22).

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 41  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 

Figure A.20: Comparison between the unit thermal resistance (s / λ )  m2 K  for thicknesses of different
 
 W 
materials in a typical building.

LP to do: enter here a matrix relationship between flow and temperature

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 42  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 

9 RESOLUTION OF  THE EQUATION OF  FOURIER:  GENERAL 


CASE

The equation of Fourier is a partial differential equation T  T  x, y , z , t  in the second order with respect to

spatial variables (x, y and z in Cartesian coordinates) and the first order with respect to the variable t. Its

integration enables the determination of the temperature range but does not

determine the value of the arbitrary constants of integration. To determine the constants of integration is
necessary to assign boundary conditions :

 spatial boundary conditions: for example, the temperature distribution on the boundary at all times
(Dirichelet problem) or heat flux at each point of the boundary and that the temperature gradient at
all times (Neumann problem), or mixed conditions

 boundary conditions of time or the initial condition: temperature distribution throughout the body at
time t = 0  

9.1 Using the boundary conditions and classification of


temporal and spatial boundary conditions
As we have seen, in solving a problem of heat transfer by conduction is usually necessary to calculate the
temperature distribution by integrating the Fourier equation, which is a quadratic equation in space and first
order in time.

In the case of non-stationary conditions there should be an integration with respect to time, which will give
rise to an arbitrary constant of integration. To calculate the value must be used as an initial condition that is

familiar with and use the temperature distribution at time t = 0 at any point x belonging to the solid area

for which you are integrating the Fourier equation, analytically T  T ( x , t  0) .

Suppose you have a separate part of the solution as a factor that depends on time, or to be in stable
condition, you must perform the integration over the space. Integrating twice with respect to space is a
general solution of the equation contains two arbitrary constants of integration (ie two constants whose value
is not determined, since the equation is satisfied for any value they assume). The general solution thus
describes a family of solutions, and must determine the particular solution (ie, characterized by well-defined
values of the constants) that satisfies the equation is given to the specific conditions of the problem.

These are indicated by the term boundary conditions, and are usually divided into four categories. The use of
two boundary conditions on a particular problem to determine the value of the two constants of integration of
space in that particular situation.

ENGLISH TO BE IMPROVED BELOW HERE!!!!

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 43  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 
 The conditions of the first kind or Dirichlet, are all conditions where the temperature is known at all
points of a given part of the surface contour of the volume that integrates the Fourier equation for
each time time, these conditions are represented mathematically by equations of the form, where
ranges over all points belonging to the area allocated. We have already encountered this kind of talking
about the issue of conditions conductive wall, and did we mention that it is normally difficult to find a
practical case where there is known a priori, the temperature along a given area.

 The conditions of the second kind, or Newman, if you have, along a given part of the surface contour
of the volume that integrates the Fourier equation, are known values of the gradient of temperature
(and therefore the heat flow conductive) and for each instant of time, a condition of the second kind is

therefore an equation of type T  f x ,t  . A special case of boundary condition of the second kind
is related to an adiabatic surface: the flow along that wall is zero and the condition reduces to

T  0 .

 The conditions of the third kind or Robin, when you have known the relationship between the gradient
of temperature (and therefore the flow of conductive) and the temperature itself along a certain part of
the surface contour of the volume that integrates the 'Fourier's equation and for each instant of time,
 
equations, then type  
f T , T , x , t  0 . Because they contain information relating to the combination

of T and gradT, the conditions of the third kind are also called mixed conditions; meet for example in
the case of heat exchange between a solid and a fluid at a known temperature undisturbed: then the
condition to boundary lies in the fact that the heat flux due to conduction and that due to convection
along the surface S are the same solid-liquid separation, and is described by equations of the type
T
k  h T  x   T  , x  S (for simplicity we have assumed that the problem is one-
x S
dimensional). Note that if h is very high T  x on the surface (i.e xS ) and T infinito are almost

equal. For example, for water condensation (or boiling) h is very high and we can say that the wall
temperature coincides with the condensation temperature of steam. In this case the condition of the
third kind is reduced to a condition of the first kind, in fact, this physical situation is one of the few
ways to create a boundary condition of the first kind. Another possible case of the third kind of
condition concerning the conduction of heat through a solid, when the heat is transferred out of the
solid by radiation exchange by radiation with a wall temperature Tp: across the surface of the solid is,

as always, the continuity of the flow ( Q  Q conduz ), and therefore the boundary condition becomes
irragg


T
x S
 4

  Tp4  T  x   h T  x   Tf  , x  S (for simplicity we have assumed that the problem

is one-dimensional).The equations of heat flow exchanged by radiation will be discussed later. NOTE:
to know the relationship between T and gradT on the boundary does NOT mean to know the values of
T and gradT separately, but only the relationship between them, as illustrated by the examples .

 The conditions of the fourth species is a condition that occurs between two solid bodies in contact
surface. The condition in question is expressed by saying that the flow exiting from the first
conductive body must be equal to that entrant in the second body, which is also :

T1 T2
1  2 for each instant of time
n S n S

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 44  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 

10 SPECIAL CASES OF  THE EQUATION AND SIMPLIFICATIONS

The equation can be simplified in special cases :

 If there is no power generation, ie   = 0 :



 c T
 2T 
 t

 On steady state (  T/  t = 0 ):
 

2T  0 ( Poisson equation )

 In steady state with no power generation :

 2T  0 ( Laplace equation )

In the last two cases are not required boundary conditions in time.

11 EXPRESSION AND DIFFERENT COORDINATE SYSTEMS

The Fourier equation, and in particular the Laplacian operator can be expressed in Cartesian
coordinates (x, y,z) or cylindrical coordinates (r,  , z) or spherical coordinates (r,  ,  ).
These are convenient to use than when the boundary conditions and power generation (and hence
the resultingtemperature distribution) have a symmetry plane, respectively, cylindrical or spherical.

Consider the simple case where the temperature, as well as a function of time is a function of one spatial
coordinate, and we write the Fourier equation for such cases.
In particular :

 infinite flat wall in the y and z directions, T function only of x:

2T 

 
  x , t

c T
x 2
  t

 solid or hollow cylinder of infinite height, T function only of r:

2T 1 T 
 
 
  x , t

c T
r 2
r r   t

 solid or hollow sphere, T function only of r:

2T 2 T 
 
 
  x , t

c T
r 2
r r   t

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 45  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 
The above analytical expressions are obtained by inserting the structure of the Laplace, expressed as a
function of the type of symmetry, the general Fourier equation and performing the appropriate partial
derivatives.

The Laplace operator or Laplacian, denoted by the vector differential operator nabla, can be connected to the
differential operators of divergence and gradient :

 
     
div grad f x    f x  2 f x  

This assumes the following structures in relation to the type of symmetry of the system in question, especially
following applies :

 Three-dimensional symmetry plane (Cartesian coordinates x, y, z):

2 f 2 f 2 f
 2 f  x, y , z     
x 2 y 2 z 2

 cylindrical symmetry (cylindrical coordinates r, z ):

1   f  1  2 f  2 f  2 f 1 f 1  2 f  2 f
 2 f  r , , z      2   
r r  r  r 2  2 z 2
r
r r r r 2  2 z 2

 spherical symmetry (spherical coordinates r,  ,  ):

1   2 f  1   f  1 2 f
 2 f  r , ,     r    sin    
r 2 r  r  r 2 sin      r 2 sin 2   2
1   rf 
2
1   f  1 2 f
  2  sin   2 2
r r 2
r sin      r sin   2

Figure A.21: Cartesian reference system (a) cylindrical frame of reference (b), spherical frame of reference
(c).

That said, you can set the appropriate mathematical descriptions that allow you to derive the formulas
presented above.

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 46  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 

12 SOME SOLUTIONS OF  THE EQUATION OF  GENERAL 


CONDUCT


To get the temperature range (ie the function T ( x , t ) which is the description of the manner in which the
temperature varies in space occupied by the solid and time in the studio) that is generated in the body, the
equations proposed in the previous paragraph must be integrated. To this end it is necessary to define the
initial conditions and some boundary conditions such as temperature and / or heat fluxes on the surfaces of
the body (more generally, the system boundary is that the differential equation).

Assuming that you are steady, why  T t = 0  , and there is a constant heat generation and

uniform    x , t      const . Fourier equation is reduced to 


2
T 0 and can be easily integrated,

resulting in:

  2

 for a flat wall simple :  
T x 
2
x  Ax  B ;

 for a solid or hollow cylinder : ;

  2 A

 For a spherical solid or hollow : 
T r 
6
r  B;
r

where A and B are constants of integration which are determined by imposing boundary conditions (two
boundary conditions for two constants of integration). Let's see in detail the procedure of integration and
calculation of the constants of integration in some special cases.

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 47  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 

12.1 Conduction through a flat wall with internal heat


generation, and constant and uniform surface
temperatures. Principle of superposition
Consider a monolayer of flat wall thickness s with uniform temperatures on the inner and outer surface and
 ''' (x, y,
constant in time and internal heat generation (power per unit volume W/m3) uniform, ie such that 
z) dV results the same for each point x, y, z of the wall, and constant in time.
We want to compute the evolution of temperature and thermal power (flow) through the wall. To do this we
must integrate the Fourier equation to find T (x, y, z, t) and then apply the Fourier assumption to calculate
the flow .

c T  

 2T 
 t 
 ''' does
Due to the symmetry of geometry and boundary conditions, all dependent on one variable x (in fact 
not change even as x changes), the solution can not only depend on x, ie T (x, y, z, t) = T (x, t), that is, our
problem is reduced to a one-dimensional problem.
In addition, the boundary conditions (temperatures and heat to the surface) are constant over time, so even if
the wall is likely to be home at the beginning of time-dependent phenomena, it is conceivable that after a
certain time interval the temperature distribution is stabilizes on a certain function T (x), which stabilizes the
flow of values also fluctuate over time, that the internal energy is stabilized to a constant value, ... In short,
we can assume that all quantities are leading to constant values, ie, no more variables over time, that all the
partial derivatives with respect to time are zero. We say that we are in steady state, and thus in particular we
have
T
 0
t
3
And T will therefore depend only on the variable x.
It follows that the partial derivatives in x are actually derived total (indicated
by straight instead of rounded  ).So the general equation of Fourier, in one-dimensional case,
with constant and uniform heat generation, and instationary conditions, this takes the form of Poisson:

d 2T   
 0
dx 2 

where x varies in the range. 0  xs.


Rearranging this equation we have :

                                                                 
3  WARNING: the fact that in steady state does not mean that the flows are non-existent, but that are
independent of time, you might be led to make this confusion by writing, but remember that in this
case the symbols mean heat transfer unit time derivative of the heat and NOT with respect to
time. There is the derivative of the heat because there is no change in the heat, the heat not being a
state function. There is, forexample, the time derivative of the internal energy U, since u is a state
function and therefore it makes sense to talk about the physical change in U when the system
switches from one state to another.. 
Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 48  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 
d 2T  d  dT   

     0   
dx 2 dx  dx  

That can be integrated for the first time :

d  dT   
 dx  dx  dx     dx thus providing:

dT  
 xA
dx 

She obtained her with a second addition :

  2

 
T x 
2
x  Ax  B

This is the general solution of the Poisson equation and allows, under suitable boundary conditions, to
determine the temperature field inside the wall .

The temperatures can be achieved by following a parabolic well and, as is the general term , are

constant and the conductivity   is positive, it appears that the coefficient of the term of the second

degree is a constant, negative value.

This means that the curve of temperature inside the wall, in case of generation,
has the trend of a parabola is concave downward .

All that remains is to impose boundary conditions to determine the extent of the two constants of integration
A and B. Selecting suitable reference system, ie as x = 0 by choosing one of the surfaces of the plate , :

T x
   x 0
 T1  B  T1

    ''' 2

T x   x s
 T2  2

T  
2
s  As  B

We solve this system of two equations in two unknowns by substitution :


'''  
''' 1
T2   s 2  As  T1  A   T2  T1  s2
2  2  s

and insert the two values found for the constants in the general solution obtained in this way and you get: :


'''  
''' 1
 
T x 
2
x 2   T2  T1 

s 2  x  T1
2  s

What can be rewritten as :

T T  '''

 
T x  T1  1 2 x 
2
x sx  
 s    
I II

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 49  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 
 Having written in this form can be seen that:  
• The term I represents the temperature trend (IT (x)) which is the solution to the problem of
conduction in infinite flat plate in steady state with no internal heat generation and temperatures
imposed on the two different faces;
• while the term is the solution II (TII (x)) the problem of steady run in infinite flat plate with
internal heat generation and temperature equal taxes on both sides,.

In the event that the two sides are kept at the same temperature, in fact the first term disappears and the
solution of the problem is :


'''
 
T II x 
2

x sx 
That means even a concave parabolic shape to the bottom of T (x), missing the deadline and having known
constant conductivity and the generic term, this parable is symmetrical to the median plane of the wall (x = s
/ 2) and has the vertex .

In conclusion, in case of heat generation, the temperature trend is parabolic with concave down and at any
point within the wall temperature reached will be the sum of a term due to the difference of surface
temperature and imposed a term linked to one generation of heat.
This is an example of a general rule, called the principle of superposition (which we discuss in more detail
below): given two functions T '(x, t) and T "(x, t) are solutions of Fourier, although their sum T '(x, t) = T' (x,
t) + T "(x, t) turns out to be solutions of the equation .

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 50  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 

Figure A.23b: solving the problem of steady-state conduction in the wall indefinitely with no internal
generation and different taxes and different temperatures on both sides (a), with internal heat generation and
temperature equal on both sides (b), internal heat generation and different temperatures on both sides (c).

Depending on the prevailing contribution of the first or second term of the vertex of the parabola will be close
to one of the two sides (one at a higher temperature) or to the median plane of the wall (x = s / 2).
In fact, the vertex of the parabola of the temperatures can fall outside the wall and in this circumstance,
however, the value assumed by the temperature passes through the values T1 and T2 on the faces .

In the case of indefinite flat wall without internal sources the density of heat flow is also uniform in space (ie
independent of the x-coordinate):
dT T T T T
 I    T      2 1   2 1  costante
dx s R [W/m 2 ]
While for the case with internal generation flux density depends on the coordinate x.
we have :

T T  '''

T  x   T1  1 2 x  x  s  x
  2
s  
I II

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 51  
A1_171117_CONDUCTION_v33_pp 1‐51.doc     DRAFT Course of Advanced Building Physics ‐ Prof. Pagliano
 
dT   '''
 T T
    T      ''' x  s  2 1
dx  2  s
II I

Recalling that the flux density vector we can write :

 T1  T 2 ˆ  s
  i  '''   x   iˆ W/m2
 s   2
I
 
II

As the temperature of the flux density can also be considered as the superposition of two parts :
 Part I is the flow that would occur if different from T1 and T2 there was no generation
 Part II is the flow that would occur in the presence of uniform generation in space and T1 = T2
This is also an example of a general rule, that the principle of superposition .

Part one is uniform in space and directed in the positive x axis if it is T1> T2. [Figure A.25 a]

Part II (the flow that would occur in the presence of uniform generation in space and T1 = T2) depends on
the position, and in particular :
 vanishes at x = s / 2 ,
 is directed in the positive direction of x for x> s / 2,
 is directed towards the negative of x for x <s / 2.
This is consistent with physical intuition: the thermal power generated by the wall can not build up inside
(since we are in steady state), therefore, will flow outward. If T1 = T2 will flow to the right and left
symmetrical to the median plane x = s / 2 [Fig A.25 b]

Faculty of Architectural Engineering ‐ Politecnico di Milano   page A / 52  

Вам также может понравиться