CHAPTER-3

METHODS FOR DESIGN AND ANALYSIS TO

DEVELOP THE WORK

3.1 Procedure for CFD solution

CFD codes are structured around the numerical algorithms that can be tackle fluid
problems. In order to provide easy access to their solving power all commercial CFD
packages include sophisticated user interfaces input problem parameters and to
examine the results. It includes mainly three steps which has been shown in fig 3.1
below

 Pre-Processing (Gambit) FLUENT

 Solver (Fluent) Solver
 Post-Processing (Fluent) Transport Equation Physical Models
Mass
 Turbulence
GAMBIT -Species mass fraction  Combustion
 Radiation
Pre-Processing -Phase volume fraction  Multiphase
 Phase
Momentum
Change
Solid Mesh Energy  Moving
modele Generator Zone
Equation of state  Moving
r Mesh
Supporting physical
models

FLUENT  Material Properties

 Boundary Conditions
Post Processing  Initial Conditions

Fig 3.1: CFD frame work

14
3.1.1 Pre-Processing

This is the first step in building and analyzing a flow model. Preprocessor consist of
input of a flow problem by means of an operator –friendly interface and subsequent
transformation of this input into form of suitable for the use by the solver. The user
activities at the Preprocessing stage involve:

 Definition of the geometry of the region: The computational domain.

 Grid generation the subdivision of the domain into a number of smaller, non
overlapping sub domains (or control volumes or elements Selection of
physical or chemical phenomena that need to be modeled).
 Definition of fluid properties
 Specification of appropriate boundary conditions at cells, which coincide with
or touch the boundary. The solution of a flow problem (velocity, pressure,
temperature etc.) is defined at nodes inside each cell. The accuracy of CFD
solutions is governed by number of cells in the grid. In general, the larger
numbers of cells better the solution accuracy. Both the accuracy of the
solution and its cost in terms of necessary computer hardware & calculation
time are dependent on the fineness of the grid. Efforts are underway to
develop CFD codes with a (self) adaptive meshing capability. Ultimately such
programs will automatically refine the grid in areas of rapid variation.

3.1.2 Gambit (CFD Preprocessor)

GAMBIT is a state-of-the-art preprocessor for engineering analysis. With advanced

geometry and meshing tools in a powerful, flexible, tightly-integrated, and easy-to
use interface, GAMBIT can dramatically reduce preprocessing times for many
applications. Complex models can be built directly within GAMBIT’s solid geometry
modeler, or imported from any major CAD/CAE system. Using a virtual geometry
overlay and advanced cleanup tools, imported geometries are quickly converted into
suitable flow domains. A comprehensive set of highly-automated and size function

15
driven meshing tools ensures that the best mesh can be generated, whether structured,
multi block, unstructured, or hybrid.

3.1.3 Solver (Fluent)

The CFD solver does the flow calculations and produces the results. FLUENT,
FloWizard, FIDAP, CFX and POLYFLOW are some of the types of solvers.
FLUENT is used in most industries. FloWizard is the first general-purpose rapid flow
modeling tool for design and process engineers built by Fluent. POLYFLOW (and
FIDAP) are also used in a wide range of fields, with emphasis on the materials
processing industries. FLUENT and CFX two solvers were developed independently
by ANSYS and have a number of things in common, but they also have some
significant differences. Both are control-volume based for high accuracy and rely
heavily on a pressure-based solution technique for broad applicability. They differ
mainly in the way they integrate the fluid flow equations and in their equation
solution strategies. The CFX solver uses finite elements, similar to those used in
mechanical analysis, to discretize the domain. In contrast, the FLUENT solver uses
finite volumes. CFX software focuses on one approach to solve the governing
equations of motion (coupled algebraic multigrid), while the FLUENT product offers
several solution approaches (density-, segregated- and coupled-pressure-based
methods).

FLUENT, is used because of its adaptability to a wide variety of grid structures.

The result is a set of algebraic equations through which mass, momentum, and energy
transport are predicted at discrete points in the domain. In the freeboard model that is
being described, the segregated solver has been chosen so the governing equations are
solved sequentially. Because the governing equations are non-linear and coupled,
several iterations of the solution loop must be performed before a converged solution
is obtained and each of the iteration is carried out as follows:

16
  Fluid properties are updated in relation to the current solution, if the
calculation is at the first iteration, the fluid properties are updated consistent
with the initialized solution.
 The three momentum equations are solved consecutively using the current
value for pressure so as to update the velocity field.
 Since the velocities obtained in the previous step may not satisfy the
continuity equation, one more equation for the pressure correction is derived
from the continuity equation and the linearized momentum equations once
solved, it gives the correct pressure so that continuity is satisfied. The
pressure–velocity coupling is made by the SIMPLE algorithm, as in FLUENT
default options.
 Other equations for scalar quantities such as turbulence, chemical species and
radiation are solved using the previously updated value of the other variables;
when interphase coupling is to be considered, the source terms in the
appropriate continuous phase equations have to be updated with a discrete
phase trajectory calculation.
 Finally, the convergence of the equations set is checked and all the procedure
is repeated until convergence criteria are met.

3.1.4 Post-Processing (Fluent)

This is the final step in CFD analysis, and it involves the organization and
interpretation of the predicted flow data and the production of CFD images and
animations. Fluent's software includes full post processing capabilities. In addition,
FLUENT CFD solutions are easily coupled with structural codes such as ABAQUS,
MSC and ANSYS, as well as to other engineering process simulation tools. Thus
FLUENT is general-purpose computational fluid dynamics (CFD) software ideally
suited for incompressible and mildly compressible flows. Utilizing a pressure-based
segregated finite-volume method solver, FLUENT contains physical models for a
wide range of applications including turbulent flows, heat transfer, reacting flows,

17
chemical mixing, combustion, and multiphase flows. FLUENT provides physical
models on unstructured meshes, brings the benefits of easier problem setup and
greater accuracy using solution-adaptation of the mesh. FLUENT is a computational
fluid dynamics (CFD) software package to simulate fluid flow problems. It uses the
finite-volume method to solve the governing equations for a fluid. It provides the
capability to use different physical models such as incompressible or compressible,
inviscid or viscous, laminar or turbulent, etc. Geometry and grid generation is done
using GAMBIT which is the preprocessor bundled with FLUENT. Owing to
increased popularity of engineering work stations, many of which has outstanding
graphics capabilities, the leading CFD are now equipped with versatile data
visualization tools. These include

 Domain geometry & Grid display.

 Vector plots.
 Line & shaded contour plots.
 2D & 3D surface plots.
 Particle tracking.
 View manipulation. (Translation, rotation, scaling etc.)

3.2 Mathematical Equations in CFD

In CFD mathematical equations are solved which govern fluid flow, heat transfer, and
related phenomena for a given physical problem. Navier-Stokes equations are most
general equations which can handle wide range of physics. Generalized Navier-stroke
equations are as below:

3.2.1 Conservation of Mass

𝜕𝜌
⃑ =0
+ ∇. 𝜌𝑉
𝜕𝑡

18
3.2.2 Conservation of Momentum
𝜕𝑢 𝜕𝑢 𝜕𝑢 𝜕𝑢 𝜕𝑝 𝜕 𝜕𝑢 2 𝜕 𝜕𝑢 𝜕𝑣 𝜕 𝜕𝑢 𝜕𝑤
𝜌( +𝑢 +𝑣 +𝑤 )=− + [𝜇 (2 ⃑ )] +
− ∇. 𝑉 [𝜇 ( + )] + [𝜇 ( + )] + 𝐵𝑥
𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧 𝜕𝑥 𝜕𝑥 𝜕𝑥 3 𝜕𝑦 𝜕𝑦 𝜕𝑥 𝜕𝑧 𝜕𝑧 𝜕𝑥

𝜕𝑣 𝜕𝑣 𝜕𝑣 𝜕𝑣 𝜕𝑝 𝜕 𝜕𝑣 2 𝜕 𝜕𝑢 𝜕𝑣 𝜕 𝜕𝑣 𝜕𝑤
𝜌( +𝑢 +𝑣 +𝑤 )= − + [𝜇 (2 ⃑ )] +
− ∇. 𝑉 [𝜇 ( + )] + [𝜇 ( + )] + 𝐵𝑦
𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧 𝜕𝑦 𝜕𝑦 𝜕𝑦 3 𝜕𝑥 𝜕𝑦 𝜕𝑥 𝜕𝑧 𝜕𝑧 𝜕𝑦

𝜕𝑤 𝜕𝑤 𝜕𝑤 𝜕𝑤 𝜕𝑝 𝜕 𝜕𝑤 2 𝜕 𝜕𝑤 𝜕𝑢 𝜕 𝜕𝑣 𝜕𝑤
𝜌( +𝑢 +𝑣 +𝑤 )=− + [𝜇 (2 ⃑ )] + [𝜇 (
− ∇. 𝑉 + )] + [𝜇 ( + )] + 𝐵𝑧
𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧 𝜕𝑧 𝜕𝑧 𝜕𝑧 3 𝜕𝑥 𝜕𝑥 𝜕𝑧 𝜕𝑦 𝜕𝑧 𝜕𝑦

3.2.3 Conservation of Energy

𝜕𝐸 𝜕𝐸 𝜕𝐸 𝜕𝐸
𝜌( +𝑢 +𝑣 ⃑ + 𝑄̇𝑣 + 𝑄̇𝑔
+ 𝑤 ) = ∇. (𝑘∇𝑇) − ∇. 𝑝𝑉
𝜕𝑡 𝜕𝑥 𝜕𝑦 𝜕𝑧

19
3.3 Design of Shock Tube
3.3.1 Steps involved in Pre-processing
Step1: Define Problem
One of the design models of this project is shock tube. Laboratory shock tube is
shown in fig3.2. The cross sectional model is shown in fig3.3.

D=Ø56mm

L=1.63 m

Fig3.3: Shock Tube Model

Fig3.2: Laboratory Shock Tube

Step2: Geometry Creation

As in fig3.3, model is circular in cross section so Gambit model to be designed as a
2d axis-symmetric problem. It is as follows

D C

𝐷
= 28mm
2

O L=1.63 m
A

Fig3.4: Axis Symmetric Model of Shock Tube

20
 Fig3.5: Basic Geometry of Axis Symmetric Model of Shock Tube Created in
GAMBIT

Step3: Apply Mesh

In this step, mesh is applied to the above created geometry. After creation of mesh,
the model will look like as below.

Fig3.6: Basic Geometry meshed in GAMBIT

Step4: Apply Boundary Conditions

In this step boundary conditions are applied on basic geometry as below:

Table3.1: Boundary Conditions for Shock Tube Model

Boundary Name of the Boundary Boundary Condition

OA Axis Axis

AC End wall Wall

CD Side wall Wall

OD Diaphragm Wall

21
Step5: Export the Meshed File

After applying boundary conditions, save the meshed geometry file and export it as a
2d mesh file (as model is two dimensional).

3.3.2 Steps Involved in Solver Settings

Step1: Import the Meshed File

The mesh file which is exported from Gambit is imported into 2ddp (2-Dimensional
double precision) version of Fluent.

Step2: Grid Check

Check the grid, and see the minimum volume on screen, it should not be negative. If
grid check fails exit from the Fluent and remesh the geometry after deleting old mesh
using Gambit.

Step3: Solver Settings

In this step, Solver settings are changed according to need of the problem (i.e.
Pressure based, Axis-symmetry, implicit and unsteady model).

Step4: Material Properties

In this step, air is taken as working fluid under ideal gas conditions because it has
considered as compressible fluid, not only air but other fluids can also be considered
as working fluid.

Step5: Enabling Dynamic Mesh Parameters

In this problem fluid is being compressed by giving motion to the one of the
boundary (i.e. Piston). Since, one of the boundaries is moving in this problem relative
to other boundaries, it is considered as a dynamic mesh problem, so enabled the
dynamic mesh parameters.

As mentioned above one of the boundaries is moving so the motion of that boundary
can be achieved by three ways. Those are:

22
  In cylinder model: In this model the velocity profile of piston will be
sinusoidal the magnitudes of these can be changed by setting the
appropriate RPM, Crank Length and Connecting rod length in In cylinder
panel under dynamic mesh window.
 UDF: In this method appropriate function is written to get the required
motion with the help of C code and it is Compiled and loaded to Fluent.
Only this method has been used in present problem to achieve the required
motion of piston.
 User Defined Profiles: In this method appropriate profiles are given to
achieve the motion of piston.

Step6: Solution Controls Settings

Set the controls parameters in control panel. All energy equation are used under
relaxation factors of ; Pressure= 0.3, Density= 1, Body Force= 1, Momentum= 0.7

Step7: Monitors Setting

In this step, all monitors are set according to the required results, which will save the
results after simulation at given location in the Notepad format.

Step8: Initialization and Iteration

Initialize the problem. Set the number of iterations and time step for each iteration in
iteration panel and then iterate it.

Step9: Saving the Case and Data File

On the completion of iteration write the case and data file at the required location,
which can be used to analyze the results of the problem. Exit the Fluent.

3.3.3 Steps Involved in Post-processing in Fluent

Step1: Reopen Case & Data File

Re-open the Case and Data file which has been saved in previous step.

23
 Step2: Display

Display Grid and Contours of CFD model. In display panel contours of velocity,
pressure, temperature etc can be displayed and analyzed.

Step3: Exit

Exit from Fluent.

Step4: Drawing Graphs and Analyzing Results

Data obtained by fluent monitors (which has been saved in Note pad format at the
location specified), graphs can be plotted and results can be analyzed.

3.3.4 Modified Model of Shock Tube

As results obtained by above model was not matching with the practical results
because of various reasons so to overcome those problems, the model has been
changed (fig 3.7), whose laboratory model is as shown in fig 3.8 (on next page).

L0 = 0.44m

L=3.5m

𝐷0 = Ø0.44m D=Ø 0.05m

Fig3.7: Model of Shock Tube

24
 Fig3.8: Laboratory Model of Shock Tube

Since, cross-section of the model is circular; the CFD model is designed as axis-
symmetric model which has shown below (Fig3.9).

L0
H G

𝐷0 /2 E
F

O
A B C D

Fig 3.9: Axis Symmetric Model of Shock Tube

25
Meshed CFD model is as shown in fig 3.10.

Fig 3.10: Axis Symmetric Model of Shock

The boundary conditions applied for this model is as follows:

Table3.2: Boundary Conditions for Shock Tube Model

Boundary Name of the boundary Boundary condition

OD Axis Axis

DE End wall Wall

GH Side wall of driver section Wall

OH Diaphragm Wall

EF Side wall of driven section Wall

FG Interface wall Wall

After this step remaining steps are similar as explained for previous model of shock
tube like importing the above meshed file into 2ddp version of Fluent, setting
dynamic mesh parameters, setting fluid properties, setting solution control panel,
initializing and iterating to solve the generated model.

At last in post-processing step, displaying the contours of pressure and Mach number
(as shown below in fig 3.11, 3.12, 3.13) to analyze the results obtained by simulation.

26
Fig 3.11: Contours of Static Pressure in Driver Section

Fig 3.12: Contours of Static Pressure in Driven Section

Fig 3.13: Contours of Static Temperature

27
3.4 Problems Faced during Convergence of Solution

1. This entire project is based on modeling of Shock tubes and Micro jets, where
all models are in cylindrical in shape, so initially models are created in three
dimensions (3D). Even though simulations converged, convergence time was
more, which increases the cost of simulations.
2. To overcome the above problem, the model is considered as 2D axis
symmetric problem with tri-pave mesh but there is a problem of negative cell
volume detection with finer mesh, so the model has remeshed with
quadrilateral mesh, which gave better results with less simulation cost.
3. When this model is operated with higher moving wall/boundary velocity,
there is sudden increment and/or decrement in either one or more of the
following factors like, temperature, pressure, viscosity. Whose values are
beyond integer value or limiting value, so there is an error displayed by fluent
as ‘floating point error’.
4. Because of the above problem, velocity of moving wall/boundary is optimized
to a particular value, depending on dimensions of the model to nullify error.

28