VŠB - Technical University of Ostrava

Faculty of Mechanical Engineering

MODELING OF HEAT, MASS AND

MOMENTUM
Tutorials

Milada Kozubková
Marian Bojko
Veronika Mořkovská
Patrik Marcalík

Ostrava 2020
Annotation
The aim is to briefly acquaint the student with the basic concepts of mass,
momentum and heat transfer in applications to heat exchangers. The following is an
illustrative example containing defining the problem, the physical properties of the
flowing media, and the boundary conditions. With the help of Ansys - Fluent software,
geometry preparation, calculation network creation, calculation and evaluation of
results and their comparison with analytical solution are realized. The following is a
series of examples to solve the procedure described above. After studying the module,
the student should be able to describe the problem of the exchanger, build a physical
and mathematical model, prepare the problem for numerical calculation, and perform
and calculate the calculation. Then evaluate the numerical calculation with the
analytical solution.

2
Table of contents
1 1 TRANSMISSION OF MASS, MOMENTUM AND HEAT ..................................4
2 2 HEAT EXCHANGERS ..........................................................................................6
2.1 2.1 Heat output ......................................................................................................6
The heat output can be determined as the value calculated in Fluent. ...........8
2.2 Pressure loss ..........................................................................................................8
2.3 Dimensionless criteria ...........................................................................................9
3 FOURIER'S LAW - HEATING IN THE BAR ...................................................... 13
3.1 ANSYS Workbench ............................................................................................ 14
3.2 ANSYS DesignModeler ....................................................................................... 14
3.3 ANSYS Meshing ..................................................................................................20
3.4 ANSYS Fluent .....................................................................................................28
3.5 Varianty výpočtů ................................................................................................. 46
4 LAMINARY FLOW - WATER FLOW BETWEEN PLATES ............................. 51
4.1 Creating geometry and mesh .............................................................................. 52
4.2 Calculation in Fluent ........................................................................................... 54
5 TURBULENT - WATER FLOW BETWEEN THE PLATES .............................. 65
5.1 5.1 Geometry and mesh ....................................................................................... 66
5.2 ANSYS Fluent .....................................................................................................67
6 SAMPLE EXAMPLE SOLUTION - CO-CURRENT EXCHANGER ................. 74
6.1 6.1 Mathematical model and theoretical-empirical estimation of the problem 75
6.2 6.2 Geometry creation ......................................................................................... 76
6.3 6.3 Creation of mesh ........................................................................................... 79
6.4 ANSYS FLUENT ................................................................................................ 81
7 HEAT DISTRIBUTION OF CONDUCTIONS AND AIR CONVECTIONS ..... 109
7.1 Mathematical model and theoretical-empirical estimation of the problem .... 110
7.2 Geometry and mesh creation. ........................................................................... 111
7.3 Calculating the Gravity Problem...................................................................... 112
7.4 Výsledky ............................................................................................................ 115

3
1 1 TRANSMISSION OF MASS, MOMENTUM AND HEAT

The basic laws of conservation of mass, momentum and energy are described
by integral or partial differential equations with boundary and initial conditions that
significantly influence the result of the solution. In a general conservative form, the form
of the equations is as follows:
  
S   dS
  u n

 dS

V t
dV  S  

  
V
S  dV
(1.1)
accumulation + convection = diffusion + source
where  is a general variable and the terms in the equation are progressively

convective (related to the velocity vector u ), diffusion and source term, therefore the
equation is also called convection - diffusion equation.
This equation can be expressed in differential form (more common in
hydromechanics and thermomechanics textbooks)):
  
    u    

       S
t   (1.2)
accumulation + convection = diffusion + source
If  represents temperature, admixture or other scalar quantity, then it is a second order
linear equation, if  represents the velocity component, it is a nonlinear equation.
The task of finding a solution to equation (1.2) satisfying both boundary and
initial conditions is called a mixed problem. If the boundary conditions are zero, they
are called homogeneous boundary conditions, similarly if the initial conditions are zero,
they are called homogeneous initial conditions. Instead of boundary conditions, other
types of conditions, also called boundary conditions, may be given. Consideration of
boundary and initial conditions for temperature is valid for the general variable. An
analytical solution of such systems is possible only in significantly simplified
applications. Therefore, the emphasis is currently on the numerical solution and in
order to specify its possibilities.
Numerical modeling allows to solve various problems, eg.:
 planar two-dimensional flow, axially symmetrical flow, general three-
dimensional flow
 stationary, non-stationary and transient flow
 laminar and turbulent flow in simple and complex geometries
 compressible and incompressible flow
 heat transfer, natural and mixed convection, radiation
 chemical transfer including chemical reactions, combustion
 multiphase flow, free-flow flow, solid-particle flow and bubbles
 flow through porous environment, etc.
Powerful CFD (Computational Fluid Dynamics) software systems, such as
Ansys-Fluent, Ansys-CFX, OpenFoam, Star CCM +, etc. are available for this purpose.

4
With the development of computer technology, the requirements for its users are
changing, especially in the field of designing. Recently, the knowledge leading to the
right choice of the computational model, computational methods and interpretation of
results has gained a significant advantage over the mathematical and programming
aspects of the problem. It remains dedicated to top mathematics and programming
specialists and problem-oriented software specialists.
The calculation method is based on the finite volume method. The user should
know their nature to the extent necessary for reliable use in standard cases. In Fluent
program it is necessary to know what shapes of finite volumes will be used, which
implies the choice of mesh density, which approximation schemes it will be suitable to
use, dynamics to have an idea of the time dependence of individual quantities and the
resulting time step size, etc. it is necessary to understand the general diction of the
manuals, because without this tool it is not possible to seriously process the
assignment. Equally important is the evaluation of results, which is particularly difficult
for three-dimensional tasks. It is optimal to have at least approximate values of
calculated quantities, ideal is to compare the results with the experiment. This textbook
should provide guidance on how to deal with the above problems.

5
2 HEAT EXCHANGERS

Heat exchangers are devices that ensure the transfer of internal thermal energy
(enthalpy) between two or more fluids, between a solid surface and a fluid, or between
particles and a fluid, in their interaction without the external work and heat supplied.

Temperatures Fluid flows

th,I - hot liquid inlet Qm,c - mass flow of heated fluid
th,O - cooled fluid outlet Qm,h - mass flow of cooled liquid
tc,I - cold fluid inlet
tc,O - heated fluid outlet S - heat exchange surface
th,s - fixed wall temperature, hot side q - heat flow
tc,s - fixed wall temperature, cold side d - thickness of solid wall
th - temperature curve in the cooled liquid  c ,h - heat transfer coefficient
tc - temperature curve in the heated liquid

Fig. 2.1 - Scheme of fluid and heat flows through the exchanger (counterflow
exchanger)

The fluids may generally be mono-constituent, or may be a mixture, both

monophasic and binary. Typical applications are two-fluid heaters and coolers, where
both fluids are separated by a solid wall, and evaporators in thermal and nuclear power
plants. Typical exchangers can be divided into several groups
 Tubular, tubular, spiral (co - current, counter - current and cross - flow
exchangers),
 Honeycomb exchangers,
 Plate heat exchangers.
The basic design parameters for heat exchanger description are heat output and
pressure drop, which will be defined for simplicity according to the diagram in Fig. 2.1.

2.1 2.1 Heat output

The energy analysis is based on a calorimetric equation that describes the heat
exchange of two or more objects. Thus, heat passes through the solid wall of the
exchanger and subsequently also through the fluid and is then affected by the flow.

6
 Heat conduction through a solid wall, ie heat output, is described by the following
equation
t t
P   h ,S c ,S S (2.1)
d
whre  is the thermal conductivity coefficient Wm-1K-1, th,s is the wall temperature,
hot side, tc,s is the fixed wall temperature, cold side, S is the heat exchange surface
m2, d is a characteristic dimension m. However, there is a velocity and temperature
boundary layer close to the wall. The temperature boundary layer is related to a heat
transfer coefficient that defines how intensely the heat is transferred from the fluid to
the solid wall or vice versa. The heat transfer equation for hot and cold walls is given
by the following equations
P   c tc ,s  tc S
P   h th ,s  th S
(2.2)

where  h is the heat transfer coefficient on the hot fluid side,  c is the heat transfer
coefficient on the cold fluid side, th is the temperature curve in the cooled liquid, tc is
the temperature curve in the heated liquid. Next, a quantity is called a heat transfer
coefficient
1
k 
1 d 1 (2.3)
 
h  c
After the introduction of the heat transfer, the power equation is transformed into:
P  k th  tc S (2.4)
By analyzing the previous relation, it is possible to determine parameters that
influence the heat exchanger performance. If the intention is to maximize performance,
then the following conditions must be considered
1. The wall thickness should be as small as possible (this is the reason for the thin
walls in the exchangers)
2. the thermal conductivity of the solid wall should be as large as possible (that is
why materials with high thermal conductivity, aluminum, copper etc. are used)
3. the heat exchange surface should be as large as possible (that is why there are
a large number of fins, honeycombs, small tubes below.)
4. the heat transfer coefficient should be as large as possible, its value can be
partially influenced by the fluid velocity, but with increasing velocity they increase with
the square of pressure loss.

When flowing through the pipe system, there is a significant change in temperature,
then the heat output would be greatly overestimated when using a temperature
difference Δ T  T s T ref . As the fluid moves through the pipe system, the wall
temperature decreases and thus the temperature difference. Therefore, the so-called
logarithmic temperature difference is used

7
 ΔT lm 
T s T I   T s T O 
 T T I   (2.5)
ln s 

T
 s  T 
O 

where T I ,T O are the inlet and outlet temperatures of the flowing medium. The outlet
temperature that is needed to determine Δ T lm can be estimated from the relationship
T s T O  dN      
 exp     T O   exp   dN  .T s T I  T s 
T s T I      vN S c  
  vNT S T c p
   T T p  
where N is the total number of tubes in the system and is the number of pipes in the
vertical plane, v is an estimate of the flow rate. Therefore Δ T lm is known. Of course,
when using numerical calculation, temperature values are determined as average
values at the inlet and outlet edges.
The heat output per unit length of the pipe can be calculated from the relation
P  N d  T lm  (2.6)

The heat output can be determined as the value calculated in Fluent.

2.2 Pressure loss

The power to be supplied to the fluid to flow through the exchanger in a given
amount can be determined by the pressure drop from the following relationship:
Q p
P  m

1  4l
P  f Re  for laminar flow (2.7)
2  2 dh
0. 046  0. 2 4l Q m2. 8
P  for turbulent flow
2  2 d h S 01. 8d h0. 2
l is the length over which heat transfer occurs, d h is the hydraulic diameter and S 0
is the minimum flow area of the exchanger.
In general, the pressure drop of the exchanger depends on the following
parameters:
1. frictional losses associated with fluid flow around heat exchange surfaces and
thus frictional (viscous) forces
2. torque effect, which is related to the change in density in the exchanger flow
3. Compression and expansion of fluid by flowing bodies (heat transfer surfaces))
4. geometric parameters of the exchanger (for large vertical exchanger it is
necessary to include also the static pressure exerted by gravity, for gases this loss is
neglected.
When flowing through a pipe system, the pressure drop is dependent on the
loss coefficient of the pipe system and empirically determined.
 u max
2
  8Q m2 
p  N L  
  resp. p  N L   
2 4  (2.8)
 2     d 

8
The loss coefficient is specific to the different pipe arrangements. In the arrangement
of the tubes in succession, it is defined as follows:
 S 
    N L L A  B 
 ST 
2 2
(2.9)
S  S d S
Where A  0. 028  T  a  T B   T  1
 2a  2  2a 
In cross-sectional arrangement of tubes, it is similarly defined:
  S 
    0. 7  0. 8 NL L A  B  
  S T 
(2.10)
2 2
S  S d S
kde A  0. 028 T  a  T B   T  1
 2a  2  2a 
Coefficient  depends on Reynolds number. For values higher than 40000 it is equal
to one and for values lower than 40000 it is estimated from empirical measurements
and is shown in Fig. 2.2.

arrangement behind cross arrangement

Fig. 2.2 - Coefficient values  depending on Re number
As can be seen, the solution flow around such a pipe system is dependent on a
series of empirically determined coefficients, the specification of which is not the object
of this subject. In Fluent, the pressure drop is obtained directly using the average
pressure values at the inlet and outlet edges. It is also possible to determine the
backward coefficient, so it can be the result of the calculation.
p  p 2tot
  1tot (2.11)
p 2 dyn

2.3 Dimensionless criteria

When preparing the mathematical model it is necessary to decide on the type
of flow and to compare the numerical solution with the analytical solution, therefore it
is necessary to define dimensionless parameters such as:
Reynolds number (Re), which is determined from boundary and physical
conditions to specify laminar or turbulent flow. Its value characterizes the flow in the
transition region between laminar and turbulent flow [3].

9
 udh
Re  (2.12)

where the so-called hydraulic diameter represents the diameter of the pipe when
flowing in the pipe, while the pipe diameter also flows around the pipe, the mean
velocity of the flowing medium. When flowing in a pipe, if Re <2320 is a laminar flow
(particles move in layers). At higher Re> 2320 it is a turbulent flow (the particles swirl)
[4].
The Prandtl number is only dependent on the material properties of the fluid.
It refers to boundary layer thicknesses, reference speed and temperature.
 c p 
Pr   (2.13)
 a
For air, its value can be assumed to be constant 0.7.

The Fourier number is the ratio of heat conduction to its accumulation in a solid body

Fo  (2.14)
c p d h 2
 is the time constant.

The Nusselt number expresses the effect of flow on the heat flow through the
wall, and depends on the geometric reference parameter (which is well definable).
d h
Nu  (2.15)

Heat transfer coefficient  includes thermal conductivity  solid walls that separate
both fluids and a heat transfer coefficient  1, 2 for the interface between the solid wall
and the two fluids. However, this coefficient is dependent on both the material
properties of the flowing fluid and the flow pattern around the solid wall.
The second definition of the Nusselt number contains better measurable quantities
such as heat output P , characteristic dimension d h , surface S , to which the heat
transfer, the temperature gradient between the wall temperature and the reference
ambient temperature is determined T  T s -T ref . The temperature gradient can also
be specified as the mean log difference.
P dh
Nu  (2.16)
S T 
The heat transfer coefficient can be determined on the basis of a number of empirical
relationships, and in practice the similarity theory is most often used. If we know the
value of the Nusselt number we can determine the heat transfer coefficient. The
Nusselt number is generally a function of other similarity criteria
Nu  f Re, Pr, Fo  (2.17)
In the case of forced convection, the value of the Nusselt number is determined by the
value of the number.

10
 Tab. 2.1 Forced convection

laminar flow around the Nu  0,664 Re1L/ 2 Pr 1 / 3 0,6  Pr

plate, TS is constant
uL
Re L  , 10 4  Re L  5.10 5 , L plate length


laminar flow around the Nu  0,908 Re 1L/ 2 Pr 1 / 3 0,6  Pr

plate, q is constant
uL
Re L  , 10 4  Re L  5.10 5 , L plate length


turbulent flow around the Nu x  0.0405 Re 4L./ 5 Pr 1 / 3 0,6  Pr  60

plate, TS is constant
5.10 5  Re L  10 8

laminar flow in the tube Nu=4.36 pro q=const. for wall

Nu=3.66 pro T=const. for wall

turbulent flow in the pipe Nu  0,023 Re 0,8 Pr m , m=0.3 for cooling

3.10 4  Re L  10 6 m=0.4 for heating

laminar, transitional and Nu  C1 Re Cě Pr 0,38

turbulent transverse pipe
Re C1 C2
wrap
0,4 ÷ 4 0,989 0,330
4 ÷ 40 0,911 0,385
40 ÷ 4 000 0,683 0,466
4 000 ÷ 40 000 0,193 0,618
40 000 ÷ 400 000 0,0266 0,805

laminar, transition and Nu D  C1 Re mD ,max pro N L 10,

turbulent wrapping of the
2000  Re mD ,max  40000 Pr  0.7 , constants C1 a m
tube bundle, NL is the
number of tubes are given in the table
SL – horizontal pipe spacing, ST – vertical pipe
spacing

11
Tab. 2.2 Constants for determining the Nusselt number when wrapping a bundle of tubes

direct system ST/D= 1.25 ST/D= 1.50 ST/D= 2.00 ST/D= 3.00
SL/D C1 m C1 m C1 m C1 m
1.25 0.348 0.592 0.275 0.608 0.100 0.704 0.063 0.752
1.50 0.367 0.586 0.250 0.620 0.101 0.702 0.068 0.744
2.00 0.418 0.570 0.299 0.602 0.229 0.632 0.198 0.648
3.00 0.290 0.601 0.357 0.584 0.374 0.581 0.286 0.608

system cross ST/D= 1.25 ST/D= 1.50 ST/D= 2.00 ST/D= 3.00
SL/D C1 m C1 m C1 m C1 m
1.000 0.497 0.558
1.125 0.478 0.565 0.518 0.560
1.250 0.518 0.556 0.505 0.554 0.519 0.556 0.522 0.562
1.500 0.451 0.568 0.460 0.562 0.452 0.568 0.488 0.568
2.000 0.404 0.572 0.416 0.568 0.482 0.556 0.449 0.570
3.000 0.310 0.592 0.356 0.580 0.448 0.562 0.482 0.574

A number of relationships can be found in the literature to determine the value

of the Nusselt number. These equations are predominantly empirically determined and
have limited validity in certain specific cases. In the previous text, only a very brief
selection of the most commonly used relationships was given.

12
3 FOURIER'S LAW - HEATING IN THE BAR

Example
Solve the temperature distribution in the bar of given length (Fig. 3.1) in ANSYS
Fluent program. The task is to create geometry, computational mesh (mesh onward),
define physical model, physical properties of material, boundary and initial conditions,
mathematical model in the programs DesignModeler, ANSYS Meshing and ANSYS
Fluent. The next step is to realize the numerical calculation and evaluate the
calculated quantities.

Fig. 3.1 - Bar of defined length

The dimensions of the area are shown in Tab. 3.1 and physical properties of individual
materials in Tab. 3.2.
Tab. 3.1 Geometry area
area length l [m] 0,5
area diameter D [m] 0,08

Tab. 3.2 Physical properties of the material (steel, aluminum, copper, wood) at 300 K
material wood steel aluminium copper
density  [kg∙m-3] 700 8030 2719 8978
specific heat capacity c p [J∙kg-1∙K-1] 2310 502,48 871 381
thermal conductivity  [W∙m-1∙K-1] 0,173 16,27 202,4 387,6

The boundary conditions are defined on the left wall (see Figure 3.1) by temperature
𝑇0 and on the right wall („right wall“) temperature 𝑇𝑙 (Tab. 3.3). Outer wall („outer
wall“) or tube sheath is considered to be insulated 𝑞 = 0 𝑊/𝑚2 .

13
Tab. 3.3 Boundary conditions
left wall right wall outer wall
𝑇0 = 50°𝐶 𝑇𝑙 = −10°𝐶 𝑞 = 0 𝑊/𝑚 2

Mathematical model
There is no flow in this task, so the flow with zero velocity is fictitious, ie as a laminar
flow.

3.1 ANSYS Workbench

Start the program in Start / All Programs / ANSYS 2019 R3 / Workbench 2019
R3. After running the program in the menu toolbar on the left side of the window,
double-click Fluid Flow (Fluent), as shown in Figure 3.2. For example, name the
newly created panel as bar (never use diacritics and mathematical symbols). Now save
the entire project File / Save as to any directory under any name, again do not use
diacritics and mathematical symbols.

Fig. 3.2 - Working environment of ANSYS Workbench 2019 R3 with Fluid flow
block.

3.2 ANSYS DesignModeler

In the first stage, you must create the geometry in DesignModeler. Right-click
on "Geometry" and select "New DesignModeler Geometry" (see Figure 3.3). The
working environment of DesignModeler is shown in Fig. 3.4..

14
 Fig. 3.3 - Starting DesignModeller

Fig. 3.4 - Program DesignModeler

Creating geometry
After launching DesignModeler, set the appropriate units, in your case it is meter - the
“Units - Meter” pull-down menu. In this case, the model represents a simple cylinder
of defined dimensions. The procedure for creating a 3D model is to create a simple
cylinder shape using the “Create - Primitives - Cylinder” drop-down menu (Fig. 3.5).

15
Select the coordinate plane of the cylinder base (XYPlane), change the position of the
center of the base (Origin), the axis length (Axis) and the radius according to Fig. 3.6.

Fig. 3.5 - Creating cylinder geometry

Fig. 3.6 - Setting the cylinder dimensions

16
 Fig. 3.7 - Resulting geometry

Naming boundary conditions

Because the model is three-dimensional, the boundaries will be areas of the area
(cylinder). In the first stage, the selection mode is changed to Face (see Figure 3.8).

Fig. 3.8 - Selection of type of face selection mode (“Face”)

An example of naming the right_wall boundary condition as specified is shown in

Figure 3.9, Figure 3.10. The desired area is highlighted and the “Named selection”
menu is selected. In the second phase naming NamedSel2 is done as "right_wall"
and "Generate", see Fig. 3.10.

17
 Fig. 3.9 - Area selection for naming the boundary condition ("Named
Selection")

Fig. 3.10 - Naming the boundary condition

The boundary condition is newly displayed in the command tree under the newly
created item "right_wall", see Fig. 3.11.

18
 Fig. 3.11 - Representation of the boundary condition "right_wall"
If there are multiple areas of the same meaning in the area (eg inlets for the pipe
system in the exchanger), then all of them can be selected (using Ctrl) and named
with one name.
Follow the same procedure to define and name the remaining boundary conditions
(“left_wall, outer_wall”), which are shown in Figure 3.12.

Fig. 3.12 - Naming boundary conditions

Now the model geometry is complete and ready for the computer network creation in
ANSYS Meshing. You can save the entire project from DesignModeler using the "File
/ Save Project" command and close the program. Go back to Workbench. You can
save the entire project from the Workbench at any time with the "File / Save" command.
If the geometry is created without errors, then the Geometry item has a green tick (Fig.
3.13).

19
 Fig. 3.13 - Workbench environment after creating geometry without errors
You can then proceed to the creation of a mesh in the ANSYS Meshing program,
which runs from the Workbench environment similar to the DesignModeler program,
see Figure 3.13.

3.3 ANSYS Meshing

In the project, double-click on the “Mesh” item to start the ANSYS Meshing
program, which allows the generated components to be meshed (Figure 3.14). This
may take several minutes depending on the model's complexity.

Fig. 3.14 - Environment of the ANSYS Meshing program

After you start the program and load the components, you have several options to
create a mesh. From a simple scheme, basically just double-clicking "Mesh" (right click

20
and select the “Generate mesh” command (very simple automatic mesh according to
preset parameters and for most of the cases unsatisfactory, (Fig. 3.15) up to a user-
defined mesh shape

„Mesh“

„Generate Mesh“

Fig. 3.15 - Creating a simple automatic network

In this application, the computational area is cylindrical, so regular hexahedron
elements are used as the elements to create a so-called computational mesh
compaction towards the outer wall. Ie. a combination of an automatic mesh with a
user-defined mesh is used.

In this example, you will use three operations to create a mesh:

 • Automatic element size adjustment
 • Defining wall compaction parameters
 • Define the Sweep method

Automatic setting of mesh element type

Click on “Mesh” in the “Outline” panel of Figure 3.16 to get information about the
meshing parameters in the “Details of Mesh” panel. There are many items in this
panel. Click “Sizing” to get predefined element size information. You can change these
values as you like. The values are given in units (meter), if they are given in millimeters,
for example, it is necessary to change the units (in the “Units” drop-down menu).
Redefine the element size in “Element Size, Max Size, Defeature Size, Curvature
Min Size”.

21
 Fig. 3.16 - Details of Mesh

For cylindrical bodies resp. The Sweep method is used for crosslinking. Use the
“Method” function in the “Mesh / Insert / Method” menu and the “sweep” method,
which is suitable for the cylinder geometry, see Fig. 3.17.

22
 Giant. 3.17 - Inserting a method

Select the volume "Geometry". We must define "Source Face" in the

"Details of Sweep Method" table. In "Src / Trg Selection" select "Manual
Source". The source left_wall. The number of elements by length can be specified
in “Type” (“number of division” and “sweep num divs”). The number of elements
along the length of the region is inserted (eg number of division = 100). The
settings and the resulting area are shown in FIG. 3.18.

23
 Giant. 3.18 - Setting parameters for the Sweep method
Then we can generate the mesh by clicking on the "Generate" command. The
resulting mesh is shown in FIG. 3.19.

24
 Giant. 3.19 - The resulting mesh

It can be seen that the net is not densified by the wall, which is useful in the case of
turbulent flow. Therefore, the Inflation method is used to repair the mesh.
Defining Inflation parameters
Inflation is defined for the “Source Face” in the “Mesh / Insert / Inflation” menu. You
can access this menu with the right mouse button (Fig. 3.20).

Fig. 3.20 - Selecting the “Inflation“

25
In general, the following parameters must be specified to create Inflation:
 the geometry (2D or 3D areas) where Inflation will be generated
 boundary at which Inflation will be created (in 2D it is the edge (line), in
3D it is the surface)
 Inflation parameters, ie reduction of the first cell at the border, number
of Inflation layers (cells), growth factor characterizing the gradual
increase in cell size,

The characteristics of the parameters defining Inflation are shown in Figure

3.21. The yellow highlighted “No Selection” must be selected from the model
geometry. “Geometry / No Selection” represents the selection of the area (area or
volume) where the Inflation will be. By changing the “Geometry Selection” you can
select by region name. First, click on the “No Selection” window (go to “Apply”, see
Figure 3.22). Then select the area by clicking on the model (a green background will
appear). Click “Apply” to confirm the result.

Fig. 3.21 - Inflation characteristics

Fig. 3.22 - Selection of the area in which Inflation will be created

Follow the same procedure to define the boundary to which Inflation is defined. Define
the edge in the “Boundary” item (Fig. 3.23). First, click in the "No Selection" field.
Then select the edge of the model. Then click on the border and select "Apply". The
result is shown in Fig. 3.23.

26
 Fig. 3.23 - Edge selection to define Inflation
Then define the Inflation parameters (Fig. 3.23)
- factor characterizing gradual cell size reduction - 0,272
- number of layers (cells) of Inflation - 5
- growth factor - 1,2
Then we can generate the network by clicking on the "Generate" command. The
resulting mesh, including densification, is shown in FIG. 3.24.

Giant. 3.24 - The resulting computing network

Save the project in ANSYS Meshing with the command "File / Save Project".

27
3.4 ANSYS Fluent
Before starting ANSYS Fluent, it is necessary to check whether the
“Geometry” and “Mesh” items have a green tick. If this is not the case, then the
"Geometry" or "Mesh" must be updated using the "Update" command. In this case it
is necessary to perform the "Update" for the "Mesh" by the right mouse button (Fig.
3.25).

Fig. 3.25 - Illustration of the Update mesh check mark

The resulting project in the Workbench environment is shown in Figure 3.26.

28
 Fig. 3.26 - Resulting panel project
The ANSYS Fluent program is started by double-clicking “Setup”. When Fluent is
started, the area dimension (3D) is verified and the calculation will be performed with
the usual or double precision (“Double Precision”). Define “Double Precision” (Fig.
3.27). It is also advisable to set the parallel calculation in "Processing Options /
Parallel" for larger numbers of cells. The number of cores is eg 4.

Fig. 3.27 - Starting ANSYS Fluent 2019 R3

Then the ANSYS Fluent program opens (Fig. 3.28).

29
 Fig. 3.28 - Basic environment of the program ANSYS Fluent
In the first phase it is necessary to check the mesh by displaying all boundaries
(boundary conditions) and the whole area by the command “Domain / Mesh /
Display” (Fig. 3.29). Selecting all items in the “Surfaces” window will display the
boundary conditions.

Fig. 3.29 - Mesh and boundary conditions check

It is also necessary to check the mesh size units with the command “Domain / Mesh /
Scale” (Fig. 3.30). If the calculation area is created in other dimensions (mm, cm, ...),
you can use the “Scaling” and “Specify Scaling Factors” commands to convert the
dimensions into basic units of meters (m).

30
 Fig. 3.30 - Checking the unit of dimensions
The next check concerns the number of cells in the mesh using the "Domain / Mesh /
Info / Size" command. Subsequently, a row will appear in the text window (Console)
with information about the number of cells (Cells), areas (Faces) and nodes (Nodes)
of the mesh, see Figure 3.31.

Fig. 3.31 - Display of the number of cells, areas and nodes

The following is a check of the existence of negative volumes in the mesh with the
command “Domain / Mesh / Check / Perform Mesh Check” (Fig. 3.32), which can
occur in complicated geometries, in which case it is necessary to recreate the mesh.

31
 Fig. 3.32 - mesh check for negative volumes

If all data is correct, proceed from left to right and top to bottom in the menu. Many of
the essential commands that appear in the tab menu are also in the left drop-down
panel (Figure 3.33).

Fig. 3.33 - Menu of commands defining a mathematical model

The first commands from the "Solver" menu ("Physics / Solver") define the solver
type, "Time-Steady" for a time-independent solution. Next, define “Type-Pressure-
Based”, “Velocity Formulation-Absolute”. The setting of the “Solver” commands is
shown in Figure 3.34. It is also possible to define the external force (eg gravity) by
accelerating the Gravity in any direction and change the physical units from the SI
system to another set of units or only units of selected quantities.

32
 Fig. 3.34 - Commands from the “Solver“

Other commands are from the menu "Models" ("Physics / Models"), which defines
the physical nature of the task according to very illustrative menu, ie "Multiphase",
"Energy", "Species", "Discrete Phase" respectively. “Viscous”, where laminar flow,
turbulent flow can be defined using various turbulent models, and a special case of the
ideal liquid flow “Inviscid” can be solved (Fig. 3.35)).

Fig. 3.35 - Characteristics of the “Models“

33
This problem solves the problem of heat transfer, ie define only the energy equation
"Energy". There is no flow in the task, so there is a fictitious solution of zero velocity
flow as a laminar “Laminar” (Fig. 3.36).

Fig. 3.36 - Setting the mathematical model of the solved problem

Defining rod material
Fluid type defines flowing medium (water, air,…). The “Solid” type defines a solid
material (steel, copper,…). In this example, we define the heat transfer in a solid
(material "Solid"), the type of material is steel. Define the material using the “Physics
/ Materials / Create / Edit Materials” command, see Figure 3.37.

34
 Fig. 3.37 Characteristics of the “Materials“

Select "Solid" in the ANSYS Fluent database ("Fluent Database") and change the
"Material Type" menu to "Solid". Next, select “steel” in the “Fluent Solid Materials”
menu. Confirm the move to the mathematical model using the “Copy” command, see
Figure 3.38. Physical properties (Density, Specific Heat, Thermal Conductivity,…) are
visible at the bottom of the menu and may vary according to the solver's requirements.

35
 Fig. 3.38 - Steel selection from ANSYS Fluent database
As a result, the “steel” material is moved to the “Materials” item (Fig. 3.39)

Fig. 3.39 “Steel” material in the “Materials“

The final assignment of the steel material to the area is done by the command
"Physics / Zones / Cell Zone Conditions", see Figure 3.40. First select "Type"
"solid". Then select “Material Name” - “Steel” and confirm with OK (Fig. 3.40).

36
 Fig. 3.40 - Characteristics of the Cell Zone Conditions command

Defining boundary conditions

We define the boundary conditions using the menu "Physics / Zones / Boundary
Conditions", see Figure 3.41.

37
 Fig. 3.41– Boundary Conditions and types of boundary conditions
The default type of boundary condition in ANSYS Fluent is wall. If we name a certain
boundary condition in the ANSYS Meshing program according to the ANSYS Fluent
conventions, a specific type will be assigned to this condition. The types of boundary
conditions can be defined according to the menu, see Figure 3.41. E.g. the axis
condition is assigned the axis condition type (“axis”). Furthermore, for naming inlet,
the type of velocity inlet of the flowing medium is assigned to the area, and for the
naming outlet is assigned the type of output (pressure outlet) of the flowing medium
from the area, etc..
Specification of boundary conditions
 left wall – type „wall“ - „Edit“ (𝑇0 = 50 °𝐶 = 323.15𝐾), see Figure 3.42.

38
 Fig. 3.42 - Defining boundary condition „left_wall“

 right_wall – type „wall“ - „Edit“ 𝑇𝑙 = −10°𝐶 = 263.15𝐾, see Figure 3.43.

Fig. 3.43 - Defining the boundary condition "right_wall"

 outer_wall – type „wall“ - „Edit“ 𝑞 = 0 𝑊/𝑚2 , see Figure 3.44.

39
 obr. 3.1 – Definování okrajové podmínky „outer_wall“

Initialization
Subsequently, the Standard Initialization of the computational area is performed;
Define initial conditions for the entire region using the "Solution / Initialization /
Method" command. Define by default "Method-Standard" initialization using the
"Options" menu, see Figure 3.45. In this task we define only temperature. Define the
mean temperature value T = 293.15K. Confirm initialization with the "Initialize" button
(Figure 3.45).

40
 Fig. 3.45 - Initialization of the computation area
Calculation
After initialization, the iteration calculation is started with the “Solution / Run
Calculation” command, see Figure 3.46. Number of Iterations must be specified. The
predefined value is 0. Enter a value quite high, eg 1000, when it is assumed that
convergence will be achieved.

Počet
iterací

Run

Figure 3.46 - Run „Canculation command“

41
The iteration calculation is then started with the “Calculate” button. Convergence can
be monitored both graphically and numerically (Fig. 3.47). Since we do not solve the
flow, the components of speed and continuity are not calculated. Only the temperature
(energy) is calculated and as soon as the desired accuracy (“Results / Residuals”) is
reached, the calculation is terminated by a note that the solution is converged, see
Figure 3.47

Fig. 3.47 - Progression of convergence

Evaluation of calculation
First we need to create a longitudinal section of the geometry. We display the geometry
using the command "Domain / Display" to find out in which axis we have to create the
section. To create a section, use the command "Domain / Surface / Create / Iso-
Surface". Select “Mesh” as the “Surface of Constant” and select the z-axis (see
Figure 3.48). Click Compute to find the coordinates in the z-axis. Since we want to
have a cross section in the middle of the cylinder, enter a value in Iso-Values that is in
the middle of these coordinates. Click Create to confirm the slice.

Giant. 3.48 - Creating a section (Iso-Surface)

42
For the sake of clarity, evaluation options are presented, ie filled temperature
isocurves, other variables are meaningless, even if they are offered, such as pressure,
speed, etc. The isocurves are evaluated in the created longitudinal section. The setting
of temperature isocurve plots is shown in Fig. 3.49.

Fig. 3.49 - Setting the drawing of temperature isolines by command

„Contours“
The result of the temperature isocurve evaluation is shown in Figure 3.50, where a
linear temperature drop from 323.15K to 263.15K can be seen. This is in line with the
analytical solution (a line connecting the temperature boundary values)).

Fig. 3.50 - Temperature distribution in the whole area [K]

43
In addition, the temperature distribution along the area can be evaluated using the
“Results / Plots / XY Plot” command, see Figure 3.51. Select “Temperature / Static
Temperature” in the “Y Axis Function” menu and select the longitudinal section of
the area in “Surfaces”. In the "Plot Direction", enter the correct direction in which the
cylinder geometry lies.

Fig. 3.51 - Characteristics of „XY Plot command“

A plot of the temperature distribution along the length of the area is shown in Figure
3.52.

Fig. 3.52 - Temperature distribution along the length of the areas

It is very interesting to evaluate the amount of heat passing through the left wall and
right wall. The evaluation is performed by the command “Results / Reports / Fluxes”,
see Figure 3.53. Select “Total Heat Transfer Rate” in the “Options” menu and select

44
left wall and right wall in the “Boundaries” menu. The resulting values are listed under
“Results” and Tab. 3.4.

Fig. 3.53 - “Fluxes“

Tab. 3.4 – Heat passing through the wall
Heat passing through the wall Q [W] steel
left wall 45.27
right wall -44.40

The heat transfer through the wall elements in units [𝑊 ∙ 𝑚−2 ] can also be evaluated
in detail at each wall location. In this simple case, it is constant because the
temperature distribution is linear in the z direction and the mesh is lengthwise with a
constant step, so there is a single slope (the temperature derivative is a flow), but in
general geometry it will not. To do this, use the “Results / XY Plot” command, see
Figure 3.54. In the "Plot Direction" menu, define X = 0, Y = 1, Z = 0, in the "Y Axis
Function" menu, select "Wall Fluxes / Total Surface Heat Flux" and in "Surfaces"
select left_wall and right_wall. The graph shows a value line inside the surface and
several values that are on the border and modified in connection with the outer wall
boundary condition“
.

45
 Fig. 3.54 - Distribution of heat flow through walls „left wall and right wall“

3.5 Varianty výpočtů

In other variants of numerical calculations, first define a different material (Table 3.5)
of the calculation area (bars). Perform numerical calculations and compare the results
as shown in the example.

Tab. 3.5 - Physical properties of material (steel, aluminum, copper, wood))

material Wood Steel Aluminium copper
density  [kg∙m-3] 700 8030 2719 8978
specific heat capacity c p [J∙kg-1∙K-1] 2310 502.48 871 381
thermal conductivity  [W∙m-1∙K-1] 0.173 16.27 202.4 387.6

Then define the variations of the different temperature boundary conditions on the left
wall and right wall as shown in Tab. 3.6.

Tab. 3.5 – Variations of boundary conditions on left wall and right wall
OKRAJOVÉ PODMÍNKY

Varianta left wall right wall right wall right wall

𝑇0 [°𝐶 ] 𝑇𝑙 [°𝐶 ] 𝑞𝑙 [𝑊. 𝑚−2 ] 𝛼 [𝑊. 𝑚−2 . 𝐾 −1 ] 𝑇∞ [°𝐶 ]
A 50 -10
B -20 100
C 50 162700
D 50 0
E 50 1000 -10
Where 𝑇0 is the temperature at „left wall“
𝑇𝑙 is the temperature at „right wall“

46
 𝑞𝑙 is the specific heat flux at „right wall“
𝑇∞ is the ambient temperature
𝛼 is the heat transfer coefficient to „right wall“

- Prepare a solution area consisting of three rods of different diameters, when

defining the geometry, always use the movement of the coordinate system to
the end of the rod (“Create / New Plane”). The boundary conditions are the
same. When sweeping, use the sweep method on the first and third rods, the
second rod relative to a different diameter is transmitted only by inflation
Geometry creation:

47
Boundary conditions:

48
Meshing:
Inflation to the right_wall, the border is a circle, sweep is applied only to the third
tube (the next tube changes the diameter, so it is not possible to continue with
the sweep)

Inflation on the left_wall, the border is a circle, the sweep is applied only to the first
pipe

Inflation on the volume of the second tube, the boundary is the area of the tube

49
The resulting mesh

The next calculation is based on the previous task.

50
4 LAMINARY FLOW - WATER FLOW BETWEEN PLATES

Example
Solve water flow between two infinitely large plates, see Figure 4.1. The physical model
is given by the shape of the area, the type of flow and the hydraulic flow parameters.
Define the numerical calculation in ANSYS Fluent. Use DesignModeler and ANSYS
Meshing to create the computational area (geometry) and computational mesh.

Fig. 4.1 - Area diagram

The water flows into the area at the speed of 0.05 𝑚. 𝑠 −1 and exits into the atmosphere
where the relative pressure is 0 Pa. The problem is given as a 3D model and represents
the flow in a cuboid area of given length, thickness and width, see Tab. 4.1. The
physical properties of the flowing medium are given in Tab. 4.2.

Tab. 4.1 – Geometry area

area length l [m] 0.5

height of the area s [m] 0.02

area width b [m] 0.1

Tab. 4.2 - Physical properties of water

density of water  [kg.m-3] 998

dynamic viscosity  [kg.(m.s)-1] 0.001003

51
 Boundary conditions
VELOCITY INLET is defined on the "inlet" and a static pressure condition
(PRESSURE-OUTLET) on the "outlet". On the walls (“top wall, bottom wall”) there
is a boundary condition of the type WALL where zero flow velocity is assumed (it is
predefined). A side condition is defined by a boundary condition of the SYMMETRY
type (infinitely large boards). The marginal conditions are given in Tab. 4.3.

Tab. 4.2 – Boundary conditions

inlet – medium speed u s [m.s-1] 0.05

outlet – static pressure p [Pa] 0

Mathematical model
The selection of the mathematical model will be solved in the next chapters, now
it will be left predefined (laminar flow model).
The laminarity criterion is Reynolds number:
u.d 0.05 . 0.02
Re    1000
 1.106
The flow is therefore laminar.

4.1 Creating geometry and mesh

In the Workbench, select "Fluid Flow / Fluent" and drag it to the work window.
Right-click on "Geometry" and select "New DesignModeler Geometry". Create the
cuboid geometry with given dimensions using the “Create / Primitives / Box”
command (Fig. 4.2). Confirm the box by clicking on "Generate“.

Giant. 4.2 - Creating cuboid geometry

52
In the next phase, name the boundary conditions as described in Figure 4.1. To
rename boundary conditions, use the “Named Selection” command. The resulting
marking and naming of all boundary conditions is shown in Fig. 4.3.

Fig. 4.3 - Marking of boundary conditions

Networking is done in ANSYS Meshing. Since it is a flow between the boards,

it is necessary to insert on the inlet Inflation surface to both walls (top wall, bottom
wall). Create Inflation for both edges at the same time as in chap. 3.3 using the
“Meshing / Inflation” command (the parameters are shown in Figure 4.4). Then use
the "Sweep" method according to chap. 3.4. The parameters are shown in Fig. 4.4.
Then generate a new mesh with the "Generate Mesh" command. The resulting form
of the mesh including meshing parameters is shown in Fig. 4.5.

Fig. 4.4 - Inflation parameters and Sweep method

53
 Fig. 4.5 - The resulting form of a computer network for flow between boards

4.2 Calculation in Fluent

After creating the mesh, go back to the Workbench, Fig. 4.6. Before running
the ANSYS Fluent program, it is necessary to update the mesh by the command
“Update” for the item “Mesh” with the right mouse button (a green check should have
appeared). The ANSYS Fluent program is started by double-clicking “Setup”. Do not
forget to set the calculation with a higher order of precision "Double precision" and
parallel calculation using "Processing Options / Parallel".

54
 Fig. 4.6 - ANSYS Workbench after update

After running ANSYS Fluent, check the calculation area dimensions and boundary
conditions as in the previous task (Chapter 3.4)

If all data is correct, proceed in the job setup in ANSYS Fluent:

 Physics Solver - General / Solver-Type (Pressure-Based) Solver Command
 Command to set time-independent Physics - General / Solver -Time (Steady)
solution
 Physics - General / Solver Gravity (no) command
 Command to set Physics - General / Solver-Units - SI
 Command to set the Physics - Models - Viscous Model - Laminar laminar model
Definition of physical properties of fluid
 Command to copy water from Physics database - Materials-Create / Edit
Materials - Fluent Database Materials (select “Material Type” water-liquid and
copy with Copy command)
 Physics - Zones-Cell Zones Conditions command (define Zone (solid) and
select material water-liquid). The zone must be of Fluid type
Defining boundary conditions
 Defining boundary conditions • bottom wall - type wall (define fixed stationary
wall, default setting)
 inlet - type of velocity inlet (define speed according to Table 4.3)
 outlet - type of pressure outlet (define the static pressure according to Table
4.3)
 top wall - type wall (define fixed fixed wall, default setting)
 symmetry1, symmetry2 ¬– type of symmetry

55
Initialization
Subsequently, the current field is initialized; Define initial conditions for the entire
area using the “Solution-Initialization-Method (Standard / Options)” command. The
values are defined based on the input boundary condition “Compute from Inlet”, see
Figure 4.7.

Fig. 4.7 - Initialization based on input boundary condition

Before starting the calculation, set the stabilization diagrams for the calculation of the
individual variables with the “Solution / Methods” command, see Fig. 4.8 with regard
to the stability of the numerical calculation.

56
 Fig. 4.8 - Setting of stability schemes
Then you run the iterative calculation “Solution-Run Calculation”. You must
specify the number of "Number of Iterations" iterations. The predefined value is 0.
Enter a fairly high value, eg 1000, when it is assumed that convergence will be
achieved, see Figure 4.9. Convergence can be monitored both graphically and
numerically.

Fig. 4.9 - Setting the number of iterations and starting the calculation

57
 The residual list is activated from the menu by the “Results / Residuals /
Residuals Monitors” commands. The course of the residuals is shown in Fig. 4.10.
Residual values (relative error) for each calculated variable (pressure - continuity,
velocity in x - x - velocity direction, velocity in y - y - velocity direction and velocity in z
- z - velocity direction) must be less than 0.001. When this accuracy is reached, the
calculation is terminated by itself.

Fig. 4.10 - The course of residuals

The next steps will be to evaluate this calculation variant. For the sake of clarity,
it is possible to create auxiliary sections with given coordinates, in which eg speed
vectors are displayed. This will be followed by the creation of transverse planes at
distances x = 0.1m, 0.2m, 0.3m and 0.4m and a longitudinal section through the center
of the z-axis (Fig. 4.11).

Fig. 4.11 - Created planes for evaluation

58
 To create transverse planes at distances x = 0.1m, 0.2m, 0.3m and 0.4m, use
the command “Results / Surface / Create / Iso-suface”. The creation of the transverse
plane at the distance x = 0.1m is shown in Fig. 4.12. In the “Surface of Constant”
menu, select Mesh / X-Coordinate. Next, enter 0.1 for Iso-Values. 0.3m and 0.4m To
make a longitudinal section, select Z-Coordinate.

Fig. 4.12 - Creating an auxiliary plane at a distance x = 0.1 m

Then, velocity vectors, velocity profiles, and filled isolines can be evaluated to
illustrate. The velocity vectors are defined in each cell of the computational domain by
the command "Results / Graphics / Vectors /", where it is possible to define the
coloring of the vectors by another variable (eg temperature). In addition, “Scale” allows
you to reduce the size of the vector and “Skip” to skip a certain number of vectors to
make the vectors less dense.

Fig. 4.13 - Speed vectors for Scale = 0.3 and Skip =0 (𝑢[𝑚. 𝑠 −1 ])

59
Next, we draw the speed profiles with the command "Results / Graphics / Vectors /"
in each cross-section. 4.15 Then the vectors are plotted using the “Display” command,
see Figure 4.16.

Fig. 4.14 - Definition of velocity vectors in cross-sections

Fig. 4.15 - Menu for defining the front view.

60
 Fig. 4.16 - Speed vectors in individual sections (𝑢[𝑚. 𝑠 −1 ])

The evaluation shows that the parabolic velocity profile is gradually formed
along the length of the computational area. In order to achieve the desired shape of
the velocity profile (from the previous solution), the calculation area is short. The
velocity size contours are drawn using the command “Results / Graphics / Contours”,
Fig. 4.15. It is further specified whether the velocity or velocity component or other
variable in the longitudinal section is plotted. Levels define the number of isosurfaces,
check Filled in Options to display filled isocurves, otherwise they are contour lines, the
result is shown in Fig. 4.18.

Fig. 4.17 - Menu for Creating Filled Isoplases of Velocity

61
 Fig. 4.18 - Contour velocity magnitude in the computational area ( )

Similarly, the plot of static pressure isocurves of static pressure is set in Fig. 4.19

Fig. 4.19 - Filled isocurves of static pressure in the computational area

[Pa]

The next evaluation presents speed profiles in individual sections from the inlet
input to the outlet output with a step of 0.05 m along the length of the computational
area. This illustration is very illustrative if it is necessary to compare the profiles of
quantities at the input, output, or in other sections of the area. Rendering is performed
using the command "Results / Plots / XY Plot / Solution XY Plot". In the Y Axis
Function menu, select Velocity - Velocity Magnitude, and in the X Axis Function
menu, select Direction Vector. Next, in the Plot Direction menu, edit X = 0 and Y = 1
and Z = 0. you can plot dependence on Y, see Figure 4.20 and select the appropriate
slices in the Surfaces menu.

62
 Fig. 4.20 - Menu for creating velocity profiles

The results show the formation of a velocity profile from a constant velocity at
the inlet input to a parabolic velocity profile at the outlet exit from the area (Fig. 4.21).
Another way to get data is to use the Options-Write to File menu, which exports data
to an external text file. This file is then read and edited in Excel.

Fig. 4.21 - Formation of speed profile

Another evaluation is the course of static pressure along the length of the
computational area. The static pressure is evaluated in the longitudinal section of the
calculation area, see Figure 4.22.

63
Fig. 4.22 - Static pressure profile along the length evaluated in the pipe axis (
[Pa])

64
5 TURBULENT FLOW - WATER FLOW BETWEEN THE PLATES

Example
Solve the flow of water between two infinitely large plates (Fig. 5.1). The physical model
is given by the shape of the area, the type of flow and the hydraulic flow parameters.
Define the numerical calculation in ANSYS Fluent. Use DesignModeler and ANSYS
Meshing to create the computational area (geometry) and mesh.

Fig. 5.1 - Area diagram

Water flows into the area at a speed of 1 m.s-1 and exits into the atmosphere, where
the relative pressure is 0 Pa. Area dimensions shown in Tab. 5.1. The problem is given
as a 3D model and represents the flow in a rectangular gap of given length and gap
thickness. The physical properties of the flowing medium are given in Tab. 5.2.

Tab. 5.1 – Geometry area

area length l [m] 0.5

height of the area s [m] 0.02

area width b [m] 0.1

Tab. 5.2 – Physical properties of water

density of water  [kg.m-3] 998

dynamic viscosity  [kg.(m.s)-1] 0.001003

Boundary conditions

65
 VELOCITY INLET is defined on the inlet and the static pressure condition
(PRESSURE-OUTLET) is specified at the outlet. On the walls (top wall, bottom wall)
there is a boundary condition of the WALL type, where zero flow velocity (predefined)
is assumed. Boundary conditions including turbulent conditions are given in Tab. 5.3.

Tab. 5.3 Boundary conditions

Inlet Medium speed u s [m.s-1] 1

Turbulent intensity [%] 1

Hydraulic diameter [m] 0.02

Outlet Static pressure p [Pa] 0

Turbulent backflow intensity [%] 1

Hydraulic diameter [m] 0.02

Mathematical model
The choice of mathematical model depends on Reynolds number.
The laminarity criterion is Reynolds number:
u.d 1 . 0.02
Re    20000
 1.106
The flow is therefore turbulent, but with a low Reynolds number, so RNG k-ε turbulent
mathematical model will be used.

5.1 5.1 Geometry and mesh

The geometry and mesh will be used from the previous example (laminar flow) by
copying the entire panel in the Workbench environment. Copying is done by the
command "Duplicate", which is invoked by the right mouse button, see Fig. 5.2.

66
 Fig. 5.2 - Copying of panel by command „Duplicate“
Then rename the panel to, for example, "turbulent flow between plates" and run
ANSYS Fluent with the "Setup" command to modify the task to turbulent flow between
plates. The other setting remained from the laminar flow task, given only for repeat.
Only the boundary conditions change.

5.2 ANSYS Fluent

Settings in ANSYS Fluent
• Solver Setting Command Setting Up Physics - General / Solver-Type (Pressure-
Based)
• Time-dependent Solution Setting Up Physics - General / Solver-Time (Steady)
Command
• Setting Up Physics Command - General / Solver Gravity (no)
• Setting Up Physics - General / Solver-Units - SI command
Setting Up Physics - Models - Viscous Model - RNG k-epsilon, Scable Wall Functions

Definition of physical properties of fluid

• Setting Up Physics - Materials-Create / Edit Materials - Fluent Database Materials
(select “Material Type” water-liquid and copy with Copy command)
• Setting Up Physics - Zones-Cell Zone Conditions (select Zone (surface_body) and
select material water-liquid)

67
Defining boundary conditions
• bottom wall - type wall (Setting Up Physics-Boundaries - define fixed stationary wall,
default setting)
• inlet - type of velocity inlet (define speed according to Table 5.3)
• outlet - type of pressure outlet (define the static pressure according to Table 5.3)
• top wall - type of wall (Setting Up Physics-Boundaries - define fixed stationary wall,
default setting)
Initialization
Subsequently, the current field is initialized; Define initial conditions for the entire
area using the “Solving Initialization-Method (Standard / Options)” command. The
values are defined based on the input boundary condition. Then adjust the stabilization
diagrams according to Fig. 4.8. The iteration calculation is then started. The resulting
residues can be seen in Figure 5.3.

Fig. 5.3 - The course of residuals

In the next steps the evaluation of this calculation variant will follow in the
transverse planes at distances x = 0.1m, 0.2m, 0.3m and 0.4m and in longitudinal
section through the center of the z-axis area (Fig. 5.4).

68
 Fig. 5.4 - Created planes for evaluation

Creating transverse planes is described in the previous chapter. Subsequently,

velocity vectors, velocity profiles and filled isocurves can be evaluated. The velocity
vectors are defined in each cell of the computational domain by the command
"Postprocessing / Graphics / Vectors". We use a longitudinal section to evaluate the
vectors. Adjust the settings to Scale = 1 and Skip = 1..

Fig. 5.5 - Speed vectors for Scale = 1 and Skip = 1 (𝑢[𝑚. 𝑠 −1 ])

69
 Fig. 5.6 - Menu for evaluation of velocity vectors in individual cross-sections

obr. 5.1 – Vektory rychlosti v jednotlivých řezech (𝑢[𝑚. 𝑠 −1 ])

The evaluation shows that the turbulent velocity profile is gradually formed along
the length of the computational area. The longitudinal section of the velocity magnitude
is shown in Fig. 5.8 and is drawn with the command „Postprocessing / Graphics /
Contours“.

70
 Fig. 5.8 - Contour velocities in the computational area (𝑢[𝑚. 𝑠 −1 ])

Similarly, the isocurves of static pressure in Figure 5.9 and the effective viscosity of
Figure 5.10 are plotted.

Fig. 5.9 - Filled static pressure lines in the computational area [Pa]

Fig. 5.10 - Filled isolines of effective viscosity

Further evaluation presents velocity profiles in individual cross-sections, see

Fig. 5.11 by graph. Rendering is done using the command "Postprocessing / Plots /
XY Plot / Solution XY Plot". Select Velocity- Velocity Magnitude in the Y Axis
Function menu and Direction Vector in the X Axis Function menu. Next, in the Plot
Direction menu, edit X = 0 and Y = 1. we want to plot the dependence on Y and in the
Surfaces menu select the appropriate slices.

71
 Fig. 5.11 - Formation of the speed profile
Another evaluation is the course of static pressure along the length of the
computational area. The static pressure is evaluated along the axis of the calculation
area, see Figure 5.12.

72
Fig. 5.12 The course of static pressure along the length evaluated in the pipe
axis ([Pa])

73
6 SAMPLE EXAMPLE SOLUTION - CO-CURRENT EXCHANGER

Create a mathematical model of the co-current exchanger and perform a three-

dimensional (3D) numerical simulation. The flowing fluids in the exchanger are water-
air combinations. The co-current exchanger model is shown in Fig. 6.1. Define
individual areas and parameters according to the specified boundary conditions and
graphically evaluate the results.

Fig. 6.1 - Co-current exchanger in 3D design.

Tab. 6.1 – Area dimensions

H1 0.5 m
D1 0.04 m
D2 0.08 m

In a given area, which is a co-current cooler, liquid-water flows in the center and
air flows around. The walls are made of steel pipes of different diameters.
Tab. 6.2 – Physical properties of material (steel, water, air) at 300 K
Material Steel Water Air
density  8030 998.2 1.225 [kg.m-3]
specific heat capacity c p 502.48 4182 1006.43 [J.kg-1K-1]
thermal conductivity  16.27 0.6 0.0242 [W.m-1K-1]
viscosity  0.001003 0.000017894 [kg.m-1s-1]

74
Tab. 6.3 – Boundary conditions
Inlet Outlet Outlet Wall Wall
Inlet air
water air water inner outer
temperature T 300 363.15 300 [K]
velocity u 3 0.3 [m.s-1]
pressure p 0 0 [Pa]
intensity of
1 1 1 1 [%]
turbulence I
hydraulic
0.02 0.04 0.02 0.04 [m]
diameter d h

Next, consider the thickness of the inner wall and the outer wall 003m. Wall material
consider steel.

6.1 6.1 Mathematical model and theoretical-empirical estimation of the

problem
In this task turbulent flow occurs, so the mathematical model of RNG is used k-
ε. The criterion of turbulence is the Reynolds number.
Re for water flow:
𝑣⋅𝑑ℎ 0.3⋅0.04
𝑅𝑒𝑣𝑜𝑑𝑎 = = 1.01𝑒−06 = 12000 (6.1)
𝜈

Re for air flow:

𝑣⋅𝑑ℎ 3⋅0.02
𝑅𝑒𝑣𝑧𝑑𝑢𝑐ℎ = = = 4323 (6.2)
𝜈 1.46𝑒−05

The calculation of the Nusselt number and the heat transfer coefficient is based
on empirical relations, which are described in detail in the literature [2]. In the next step,
only the analytical calculation is performed, which will be compared with the numerical
calculation. From the given parameters it is possible to calculate the above parameters
of flow and heat transfer (Reynolds number is calculated from the maximum speed).
The estimation of Nusselt's number is problematic and is only indicative. This estimate
is followed by the calculation of the wall heat transfer coefficient determined from the
 Nu.
Nusselt number   [2].
d
Calculation of the Nusselt number for the area of water flow in a pipe:
  c p  998.2  4182  1.01e  6
Pr    6.99 (6.3)
 0.6
𝑁𝑢 = 0,023 ⋅ 𝑅𝑒 0,8 ⋅ 𝑃𝑟 0,3

75
𝑁𝑢 = 0.023 ⋅ 120000,8 ⋅ 6.990.3 = 75.5 (6.4)

Then the heat transfer coefficient is

𝑁𝑢 75.5
𝛼= ⋅ 𝜆 = 0.04 ⋅ 0.6 = 1132.6 𝑊. 𝑚−2 . 𝐾 −1 (6.5)
𝑑ℎ

Calculation of the Nusselt number for the area of air flow around a pipe:
𝜌⋅𝑐𝑝 ⋅𝜈 1.225⋅1006.43⋅1.46𝑒−5
𝑃𝑟 = = = 0.707 (6.6)
𝜆 0.0242

𝑁𝑢 = 0.023 ⋅ 𝑅𝑒 0,8 ⋅ 𝑃𝑟 0,4

𝑁𝑢 = 0.023 ⋅ 43230,8 ⋅ 0. 7070,4 = 16.79 (6.7)

Then the heat transfer coefficient is

𝑁𝑢 16.79
𝛼= ⋅𝜆 = ⋅ 0.0242 = 20.3 𝑊. 𝑚−2 . 𝐾 −1 (6.8)
𝑑ℎ 0.02

6.2 6.2 Geometry creation

Run the ANSYS 2019 R3 program as per chap. 3.1. Name the newly created
panel, for example, heat exchanger. Then save the entire project under any name and
run the DesignModeler geometry program.
To create the geometry use the detailed instructions in chap. 3.2, because the
resulting co-current exchanger model is a 3D model similar to a 3D rod model. The co-
flow exchanger model represents two areas (interior water, interior air). Fig. 6.1.
These are therefore two cylinders that we must subtract from each other. You create
the regions identically using Create / Primitives / Cylinder as in the example of the
conduction heat in a rod. The final appearance of the interior water area created by the
Cylinder including the dimensions is shown in Figure 6.2.

Fig. 6.2 - Creating an interior water area ("Cylinder“)

The final appearance of the interior air created by the cylinder (cylinder) including
the dimensions is shown in Fig. 6.3.

76
 Fig. 6.3 - Creating an interior air area ("Cylinder“)
In the case of two areas to be separate volumes, you must define the Add Frozen item
in the Operation tools. This will not merge faces.
Now you need to subtract the cylinders from each other by Boolean operations using
the “Create / Boolean / Operation-Subtract” command. Select air area as target
points and water area as tool points. Select Preserve Tool Body to preserve the water
area. Clicking Generate creates two separate volumes for the water area and the air
area.
The last operation is to merge volumes into one unit, ie. New part. By merging the
volumes into one unit, the continuity of the computer network between individual areas
will be preserved. You can get the command by selecting both volumes in the 2 Parts,
2 Bodies tab and right-clicking on the “Form New Part” menu, see Figure 6.4.

Fig. 6.4 - Merging volumes into one unit ("Form New Part“)
The final form of the “Form New Part” command is shown in Figure 6.5.

77
 Fig. 6.5 - The final form of the “Form New Part“

In the next phase, name the boundary conditions as described in Figure 6.1 (inlet air,
inlet water, outlet air, outlet water, wall inner, wall outer). The naming of the
boundary conditions is done using the “Named Selection” command with Face mode
(3.2). The resulting marking and naming of all boundary conditions can be seen in
Figure 6.6. In addition to the boundary conditions on the walls, the model contains two
areas of interiors (volumes) that need to be defined (interior_air and interior_water).

Fig. 6.6 - Marking of boundary conditions

This completes the coil flow model in DesignModeller.

78
6.3 6.3 Creation of mesh
Now you can switch to the mesh in ANSYS Meshing. The procedure for starting the
program is described in chap. 3.3. Use the same tools as shown in Chap. 3.3. Create
a mesh with interior water and interior air boundary layers towards the wall inner and
wall outer.
To create a computational network in this form, you will use network compaction
(create only at the front of both cylinders) and sweep. These are therefore the same
operations as used to create the computer network in the example of heat conduction
in a rod. In the Details of Mesh panel, redefine the element size for Element Size to
4 mm, Max Size to 10 mm.

Fig. 6.7 - Defining element size

Defining Inflation Parameters

Then define the compaction mesh parameters. The number of densified layers (cells),
the growth factor characterizing the gradual reduction of the cell size towards the
boundary, the reduction ratio of the last cell of the densified area. Define two areas of
densification into each area (interior water, interior air) towards the wall inner.
Number of boundary layer layers (cells)
- Growth factor - 1,2
- Factor characterizing gradual cell size reduction - 0.272

79
Inflation parameters towards wall inner for interior water and interior air are shown in
Figure 6.8 and Figure 6.9.

Fig. 6.8 - Compression parameters for the interior water area

Fig. 6.9 - Compression parameters for the interior air area

Now insert the “Sweep” method (for settings, see chapter 3.3) and thereby pull the
surface mesh into volume.
To generate a computer network, use the Generate Mesh command. The resulting
form of the mesh is shown in Figure 6.10.

80
 Fig. 6.10 - The resulting form of the mesh

6.4 ANSYS FLUENT

You run the ANSYS FLUENT 2019 R3 in a similar way as in the example of heat
conduction in a rod.
After successfully loading the computer network into ANSYS Fluent 2019 R3, check:
• units of mesh size with the command "Domain / Mesh / Scale"
• number of network cells with the command “Domain / Mesh / Info / Size”
• the existence of negative volumes in the mesh using the "Domain / Mesh /
Check" command
• Mesh by displaying all boundaries (boundary conditions) and all areas with the
command "Domain / Mesh / Display"

When checking the mesh with the "Domain / Mesh / Display" command, all boundary
conditions are named as defined in ANSYSMeshing. Except for one newly created
boundary condition wall_inner-shadow. Which represents the same boundary
condition as wall_inner. A new boundary condition wall_inner-shadow was created
(Fig. 6.11), which together with the wall_inner condition defines a so-called two-layer
wall, where one is part of the interior water area and the other is part of the interior air
area. This type of boundary condition offers the definition of additional options at the
transition between the two regions.

81
 Fig. 6.11 - Checking boundary conditions

Use the following mathematical model settings:

• Time-stable flow
• Turbulent k-ε RNG flow model for water and air
• Without considering gravity acceleration
• Consider heat transfer (energy equation)
• Define constant physical properties of water and air (copy materials from Fluent
library)
Within the "General" command, define "Solver" of the "Pressure-Based" type. Steady
flow. Do not consider gravity acceleration. The setting of the “General” command is
shown in Fig. 6.12.

Fig. 6.12 - Command „General“

82
Another command is Models (“Physics / Define / Models”), where the physical nature
of the task is defined, ie Energy flow and the turbulent k-ε RNG model of Viscous flow
together with the Scable Wall Functions, see Figure 6.13.

Fig. 6.13 - Setting the mathematical model of the solved problem

To define the material, use the Materials command ("Physics / Materials / Create /
Edit Materials"). Analogous to the heat conduction in a rod, select the materials:
water, air, steel to be copied from the ANSYS Fluent database. These are fluid and
solid materials, and define a constant physical property for all materials. The resulting
form of material supply is shown in Figure 6.14.

83
 Fig. 6.14 - Required materials for the mathematical model
To define the flowing fluid into the area, use the "Physics / Cell Zone Conditions"
command. In this case we have two areas (interior water, interior air). Define water
in the interior water area, and define air in the interior air area, as shown in Figure
6.15.

Fig. 6.15 - Defining flow media to given areas

Use the "Physics / Boundary Conditions" command to define boundary conditions.
The conditions may be of different types depending on the characteristics of the
physical model. The list of conditions is apparent from Tab. 6.4.

84
 Tab. 6.4 – Types of individual boundary conditions
Okrajová podmínka Typ okrajové podmínky
inlet water VELOCITY INLET
inlet air VELOCITY INLET
outlet water PRESSURE OUTLET
outlet air PRESSURE OUTLET
wall inner WALL
wall outer WALL
interior water INTERIOR
interior air INTERIOR
Parameters on individual boundary conditions correspond to the input according to
Tab. 6.3. The setting of boundary conditions is shown in the following figures.

Fig. 6.16 - Parameters of boundary condition inlet water

Fig. 6.17 - Parameters of boundary condition inlet air

85
Fig. 6.18 - Parameters of boundary condition outlet water

Fig. 6.19 - Parameters of boundary condition outlet air

86
 Fig. 6.20 - Parameters of boundary condition wall inner

Fig. 6.21 - Parameters of boundary condition wall inner-shadow

87
 Fig. 6.22 - Parameters of boundary condition wall outer

The calculation area is then initialized ("Solving-Initialization-Method (Standard /

Options)"). defining initial conditions for the whole area. In the first step, define the
initial conditions (zero values, minimum temperature) based on the parameters in
the inlet_air boundary condition.

88
 Fig. 6.23 - Initialization of the computation area („Solution Initialization“)
In the second step, define the initial temperature value T = 363K for the entire
interior_water using the Patch command in the same window (Fig. 6.24) to speed up
the numerical calculation.

Fig. 6.24 - Initialization of water flow area by command Patch

Then run the numerical calculation using the “Run Calculation” command. The first
check of the calculation is the monitoring of residuals (relative errors). Once the

89
residual values are below 0.001 for all variables and 0.000001 for temperature, it is
guaranteed that the calculation has converged numerically. How real the results are,
ie whether the result is not deformed by random errors in the selection of materials or
boundary conditions, is a question of evaluating all calculated quantities. The course
of residuals is shown in Fig. 6.25 - The course of residuals.

Fig. 6.25 - The course of residuals

For the evaluation it is necessary to create a longitudinal section of the area using the
command “Results / Surface / Create / Iso-Surface”. In this section, subsequently
evaluate the graphic outputs. The setting of the longitudinal section through the water
flow area is shown in Fig. 6.26. Similarly, we create a section of the interior-air area,
with the difference that in the From Zones item select interior_air.

Fig. 6.26 - Creating a longitudinal section in the interior water area

90
The resulting section is shown in Figure 6.27.

Fig. 6.27– Longitudinal section through the center of the computing area
To evaluate the velocity vectors defined in each cell of the computation area, use the
"Results / Graphics / Vectors /" command to adjust the value of the "Scale"
parameter. Define a new value for the "Scale = 0.5" parameter, see Figure 6.28.

Fig. 6.28 - Velocity vectors (𝑢[𝑚. 𝑠 −1 ] )

The course of static pressure in the longitudinal section in the solved areas (interior
water, interior air) can be displayed using the filled out “Results / Graphics /
Contours” contours, see Fig. 6.29.

91
 Fig. 6.29 - Static pressure contours (Pa)

The static pressure curve can also be displayed using a 2D graph with the command
“Results / Plots / XY Plot / Solution XY Plot” in individual areas (interior water,
interior air). The subsequent plot of the pressure curve with the command “Results /
Plots / XY Plot / Solution XY Plot” in the individual areas is shown in Figure 6.30.
The results can be distorted into one graph.

Fig. 6.30 - Progression of static pressure along length in area interior water and
interior air

Further evaluations are the longitudinal section contour velocities using the filled in
“Results / Graphics / Contours” contours, see Figure 6.31.

92
 Fig. 6.31 - Velocity contours [m/s]
The graph of effective viscosity using filled out “Results / Graphics / Contours” contours
is shown in Fig. 6.32.

Fig. 6.32 - Effective viscosity [kg.m-1.s-1]

The evaluation of the temperature field using the filled in “Results / Graphics /
Contours” contours is shown in Figure 6.33.

93
 Fig. 6.33 - Temperature field [K]
The heat flow through the wall inner (wall inner-shadow) can be evaluated by the
command “Results / Plots / XY Plot / Solution XY Plot”, see Figure 6.35. The
interface wall is divided into two walls (wall inner and wall inner-shadow), one acting
as an interface for water and the other as an interface for air. Their exact designation
is related to the subsequent evaluation of the heat transfer coefficient and the Nusselt
number. To determine exactly which wall is part of a given flow area, use the "Physics
/ Zones / Boundary Conditions" command. Subsequent editing eg wall inner it is
stated that the wall is adjacent to the surrounding area (Adjacent Cell Zone) - interior
_water, ie. with water, see Fig. 6.34. In the case of wall_inner-shadow it will be the
opposite (the wall is adjacent to the air area)).

Fig. 6.34 - Identification of the wall inner adjacent to the surrounding water

94
The evaluation of the Total Surface Heat Flux through the wall inside and inside wall
with the “Results / Plots / XY Plot / Solution XY Plot” command is shown in Figure
6.35..

Fig. 6.35 - Heat flow (W / m2) through the wall inner (wall inner-shadow)

By analogy, the heat flow on the walls can be evaluated using the filled out “Results /
Graphics / Contours” contours“ (Chyba! Nenalezen zdroj odkazů.).

95
 Fig. 6.36 - Heat flow (W / m2) through the wall inner)

Fig. 6.37 - Heat flow (W / m2) through the wall inner-shadow)

96
Furthermore, the evaluation focuses on the heat transfer coefficients α and the
Nusselt number Nu into water and air, while it is necessary to define reference values
Evaluation for water
First, define the reference values according to the inlet water with the command
“Results / Reporst / Reference Values”. Under "Compute from", select inlet water.
In the “Reference Values” menu, specify the “Temperature” and the “Lenght” - (Tref
= 363.15 K, dh = 0.04 m), see Figure 6.38.

Fig. 6.38 - Reference values for evaluation into the water for the wall wall inner

Evaluation of the surface heat transfer coefficient α from the water side to the wall
inner is carried out with the command “Results / Plots / XY Plot / Solution XY Plot”,
see Figure 6.39.

97
Fig. 6.39 - Heat transfer coefficient through the interface wall (wall inner) [W.m-2.K-1]
Analogously, the heat transfer coefficients on the wall can be evaluated using the filled
out “Results / Graphics / Contours” contours“ (obr. 6.1).

obr. 6.1 – Heat transfer coefficient through interface wall (wall inner) [W.m-2.K-1] by
contours
Subsequently, the Nusselt number on the wall inner can be evaluated. First check
the reference values with the command "Results / Reporst / Reference Values"
(Temperature -Tref = 363.15 K and Lenght - dh = 0.04 m. Then draw the Nusselt
number with the command "Results / Plots / XY Plot“ (Chyba! Nenalezen zdroj o
dkazů.).

98
 Fig. 6.41 - Nusselt number evaluated on the interface wall (wall inner)

The Nusselt number can be evaluated using the filled out “Results / Graphics /
Contours” contours“ (Chyba! Nenalezen zdroj odkazů.).

Fig. 6.42 - Nusselt number evaluated on the interface wall (wall inner)

99
Evaluation for air
Next we evaluate the surface heat transfer coefficient a and the wall inner-shadow
Nusselt number Nu by the command “Results / Plots / XY Plot / Solution XY Plot”.
First, define the reference values according to the inlet air with the command “Results
/ Reporst / Reference Values”. Under "Compute from", select inlet air. In the
“Reference Values” menu, specify the “Temperature” and the “Length” (Tref = 300
K, dh = 0.02 m), as shown in Figure 6.43.

Fig. 6.43 - Reference values for the wall wall inner-shadow

Fig. 6.44 - Coefficient of heat transfer to air for the interface (wall inner-shadow)
[W.m-2.K-1]

100
The evaluation of the heat transfer coefficient using the filled in “Results / Graphics /
Contours” contours is shown in Figure 6.45.

Fig. 6.45 - Heat transfer coefficient for air for interface (wall inner-shadow) [W.m-2.K-
1
]
Similarly, we evaluate the Nusselt number on the wall inner-shadow interface. Check
the reference values with the command "Results / Reporst / Reference Values"
(Temperature -Tref = 300 K and Lenght - dh = 0.02 m). Then draw the Nusselt number
with the command “Results / Plots / XY Plot / Solution XY Plot”, see Figure 6.46.

Fig. 6.46 - Nusselt number evaluated on the interface wall (wall inner-
shadow)

101
By analogy, the Nusselt number can be evaluated using the filled out “Results /
Graphics / Contours” contours“ (Chyba! Nenalezen zdroj odkazů.).

Fig. 6.47 - Nusselt number evaluated on the interface wall (wall inner-shadow)

Evaluation of average values

To evaluate the average Nusselt value on the wall inner for the water flow area, use
the "Results / Report / Surface Integral" command. First, define the reference values
as shown in Figure 6.38. Select “Area-Weighted Average” in the “Report Type”
menu. Next, in the “Field Variable” menu, select “Wall Fluxes-Surface Nusselt
Number” and in “Surface” select “wall inner” (Figure 6.48). Write the resulting value
in Tab. 6.5.

102
 Fig. 6.48 - Evaluation of the average value of the Nusselt number on the wall
inner for the water flow area
Follow the same procedure for evaluating the average value of the Nusselt number
on the wall inner-shadow for the air flow area, see Figure 6.49. Define the reference
values as shown in Figure 6.43.

Fig. 6.49 - Evaluation of the average value of the Nusselt number on the wall
inner-shadow for the area of air flow

103
In the same way, evaluate the mean value of the surface heat transfer coefficient
na on the wall inner for the water flow area using the "Postprocessing / Report /
Surface Integral" command (Figure 6.50).

Fig. 6.50 - Evaluation of the mean value of the heat transfer coefficient  on
the wall inner for the water flow area
Fig. 6.50 - Evaluation of the average value of the coefficient The same procedure
applies to the evaluation of the average value of the heat transfer coefficient  on
the wall inner-shadow for the air flow area, see Figure 6.51. Define the reference
values according to Fig. 6.43 water flow.

Fig. 6.51 - Evaluation of the mean value of the heat transfer coefficient  on
the wall inner-shadow for the air flow area

104
To evaluate the heat output P, select Total Heat Transfer Rate in the “Results /
Report / Fluxes” command in Options and select the wall inner and wall inner-
shadow in the Boundaries menu, see Figure 6.52.

Fig. 6.52 - Heat flow evaluation P [W]

Determine the loss coefficient  based on the respective pressures defined in the
equation below.
p1tot  p 2tot
  (6.9)
p 2 dyn

Always use the “Results / Report / Surface Integrals” command to evaluate the
pressures at the inlet and outlet of the flowing water and air. An example of evaluating
the p1tot of the total inlet air pressure is shown in Figure 6.53. Then enter the value in
Tab. 6.5. Identify the remaining pressure values identically (p2tot, p2dyn).

105
 obr. 6.2 – Evaluation of p1tot on inlet air
Calculation of the loss coefficient  for the airflow area:
p1tot  p2tot 9,16  5,8
    0,58 (6.10)
p2 dyn 5,78

Calculation of the loss coefficient  for the waterflow area:

p1tot  p2tot 71,48  47,08
    0,52 (6.11)
p2 dyn 47,003

106
Tab. 6.5 – Final comparison of average results

Estimate for Estimate CFD air CFD water

Units
air for water solutions solutions

u 3 0.3 3 0.3 [m.s-1]

Re 4108 11943 4108 11943 [1]

Nu 15,9 91,62 12.05 82,18 [1]

 19,2 1374,3 14.58 1232,7 [W.m-2.K-1]

P 85,77 85,77 [W]

p1tot 7,53 61,85 [Pa]

p2tot 5,59 45,73 [Pa]

p2dyn 5,59 45,67 [Pa]

 0,35 0,35 [1]

u velocity
Re Reynolds number
Nu Nusselt number
 Heat transfer coefficient
P Heat output
p1tot total pressure at the inlet
p2tot total pressure at the outlet
p2dyn dynamic pressure at outlet
 loss coefficient

Conclusion
Deviations in the solution are caused both by estimating the Nusselt number
analytically and by the numerical solution, where it is possible to test the influence of
mesh quality, models and physical properties. In particular, the analytical relations of
the Nusselt number estimate do not fully correspond to the characteristics of the given
co-current exchanger task. They are intended to provide basic information about the
Nusselt number estimate. Nusselt numbers obtained from analytical relations and
numerical calculation are in order, which can be considered satisfactory.

107
The accuracy of the numerical calculation depends on the quality of the mesh, which
can be subsequently compressed. There are several possible adaptations, for
example, a mesh can be prepared and a comparison of results can be made.

108
7 HEAT DISTRIBUTION OF CONDUCTIONS AND AIR
CONVECTIONS

Create a mathematical model theoretically analogous to a co-current exchanger

with the difference that the ambient air will be defined instead of the outer tube. Perform
a 3D simulation. The fluids are combined air-to-air. The model can be seen in Fig. 6.1.
Define individual areas and parameters according to the specified boundary conditions
and graphically evaluate the results.

Fig. 7.1 - Geometry and boundary conditions.

Tab. 7.1 – Area dimensions

Tube length H1 0.5 m
Pipe diameter D1 0.04 m

Block using two (0.0 -0.1 -0.1)

m
pointson the diagonal
(0.5 0.5 0.1)

The pipe flows in the middle of the liquid - water, the wall is formed by a steel pipe of
a given diameter. Also consider a wall water thickness of 003m. Wall material consider
steel.

109
Surrounding is air bounded by atmospheric pressure, a condition of pressure outlet.
The outlet bottom is an insulated wall.
Tab. 7.2 – Physical properties of material (steel, water, air) at 300K
Material Steel Water Air
density  8030 998.2 1.225 [kg.m-3]
specific heat capacity c p 502.48 4182 1006.43 [J.kg-1K-1]
thermal conductivity  16.27 0.6 0.0242 [W.m-1K-1]
viscosity  0.001003 0.000017894 [kg.m-1s-1]

Tab. 7.3 – Boundary conditions

Inlet Outlet Wall Outlet Wall
water water water air bottom
temperature
363.15 coupled q=0 [K]
T
velocity u 0.3 [m.s-1]
pressure p 0 0 [Pa]
intensity of
1 1 1 [%]
turbulence I
hydraulic
0.04 0.04 0.5 [m]
diameter d h

7.1 Mathematical model and theoretical-empirical estimation of the

problem
In this task turbulent flow occurs, so the mathematical model RNG k-ε is used.
The criterion of turbulence is the Reynolds number. There is almost no air flow, eg
velocity is 0.001 m / s.
Re for water flow:
𝑣⋅𝑑ℎ 0.3⋅0.04
𝑅𝑒𝑣𝑜𝑑𝑎 = = 1.01𝑒−06 = 12000 (7.1)
𝜈

The calculation of the Nusselt number and the heat transfer coefficient is based
on empirical relations, which are described in detail in the literature [2]. In the next step,
only the analytical calculation is performed, which will be compared with the numerical
calculation. From the given parameters it is possible to calculate the above parameters

110
of flow and heat transfer (Reynolds number is calculated from the maximum speed).
The estimation of the Nusselt number is problematic and is only indicative. This
estimate is followed by the calculation of the wall heat transfer coefficient determined
 Nu.
from the Nusselt number   [2].
d
Calculation of the Nusselt number for the area of water flow in a pipe:
  c p  998.2  4182  1.01e  6
Pr    6.99 (7.2)
 0.6
𝑁𝑢 = 0,023 ⋅ 𝑅𝑒 0,8 ⋅ 𝑃𝑟 0,3
𝑁𝑢 = 0.023 ⋅ 120000,8 ⋅ 6.990.3 = 75.5 (7.3)

Then the heat transfer coefficient is

𝑁𝑢 75.5
𝛼= ⋅ 𝜆 = 0.04 ⋅ 0.6 = 1132.6 𝑊. 𝑚−2 . 𝐾 −1 (7.4)
𝑑ℎ

7.2 Geometry and mesh creation.

Geometry is given by two entities, ie a cylinder and a box, using a Boolean subtraction
to create a water area and an air area. The methodology of networking is identical with the
methodology described in chap. 6, ie the inflation and sweep method for the pipe and the
inflation method for the air volume. The mesh has the following shape.

Fig. 7.2 - Surface mesh and detail with inflation.

111
7.3 Calculating the Gravity Problem.
Adjustments to deal with ambient heat transfer will be carried out in Fluent as follows:
The heat transmitted by the air flow to the surroundings is significantly influenced by gravity.
It is entered eg in menu „Physics / Operating Conditions / Gravity“ and density is specified
in „Physics / Operating Conditions / Operating Density“, whose value is 0. Then the
stratification of pressure in the result can be observed.

Fig. 7.3 Operating Conditions

Physical properties of air will depend on temperature or pressure, so density is given by state
equation and other physical properties by so-called kinetic theory.

112
 Fig. 7.4 Properties of Air
Too “loose” boundary condition of atmospheric pressure causes significant backflow and then
divergence. Therefore, it is advantageous to use a velocity condition with a very small value,
e.g. 0.001 m / s.

113
 Fig. 7.5 Velocity inlet for air
When gravitational acceleration is entered, hydrostatic pressure is generated automatically in
the air and tube area. Therefore, the pressure condition at the water outlet from the pipe will be
replaced by hydrostatic pressure.

Fig. 7.6 Pressure outlet for water

114
Then the task converges well and the results will be real.

7.4 Výsledky
Initialization is realized mainly by real temperature values (300 K). PATCH at 363 K is used
for the water area.

Fig. 7.8 - Contours of the hydrostatic pressure at the air / pipe boundary and
pressure drop in the pipe (PLOT XY)).

115
Fig. 7.9 - Temperature distribution in axial section of the area with air heating
above the pipe.

116