CFD SIMULATION USING

ANSYS FLUENT
Fluid Mechanics MECH 340

Dr. Fadi Alnaimat

UAEU Faculty of engineering Mechanical engineering Department

GROUP MEMBERS:

ROLL # NAME

Group 4
www.uaeu.com
TABLE OF CONTENTS
1 Abstract .............................................................................................................................................. 2
2 Introduction........................................................................................................................................ 2
3 Mathematical Formulation................................................................................................................. 4
4 Results and discussion ....................................................................................................................... 6
4.1 problem with boundary condition .............................................................................................. 6
5 Summary and Conclusions .............................................................................................................. 15
6 References........................................................................................................................................ 16
1 ABSTRACT
Computational Fluid Dynamics (CFD) is a very broad approach of estimating the fluid behavior around
any physical boundary. The boundary can be of any shape either circular like in pipes or rectangular
like in ducts. These ducts usually found in vast engineering applications like refrigeration and heat
pump system, power plants, biomedical instruments, aircrafts, automobile and heat ventilation and air
conditioning etc. Analytically analyzing CFD means through equations like PDEs and ODEs
commonly shows exact results but analytical solutions only apply to known shapes, but for irregular
shaped body analytical method is unable to give accurate and instant results. On the other hand, by use
of numerical technique means through computers we can solve various simultaneous equations through
algorithms and loops in a software, there are tens of commercial software are available in the market to
compute CFD in any irregularly shaped body but they give approximate solutions like ANSYS,
SOLIDWORKS, SAP2000, PRO-MECHANICA, NISA, CATIA, CREO, LS-DYNA, MSC-
NASTRAN etc. By using these software, we can use CFD for many findings like optimization of
design, heat transfer analysis, head loss prediction, shear forces calculation, fluid uniformity, multi-
physics simulation and transient analysis etc. But there are some assumptions in both analytical and
numerical methods like supposed boundary conditions, laminar flow consideration, incompressible
flow, no slip boundary condition, uniform walls, negligible wall thickness, one dimensional steady
state flow low Reynold’s number and hundred percent uniform mesh etc. Main approach and methods
used in CFD are Finite volume method (FVM), Finite element method (FEA) for analytical approach,
turbulence modelling, meshing, multiphase flow, heat transfer etc. In this report we calculated average
shear force at wall and pressure drop at inlet and outlet of duct, while our main findings present that
software solutions are quite accurate than solutions from governing equations that is due to main
difference in boundary conditions and assumptions. But on industrial scale, using ANSYS fluent for
CFD simulation is not recommended because for complex problems their way of finding solution
ignores lots of physics behind it and gives absolutely inaccurate results.
Keywords:
ANSYS fluent, FVM, FEA

2 INTRODUCTION
In this project we are conducting a CFD simulation using ANSYS fluent software of two circular ducts
of vary diameter attached to each other at center line. Diameter of pipe one is 0.05m and diameter for
second pipe is 0.025m, similarly length for former is 0.5m and for latter is 0.75m. Fluid passes from
inlet, wall and outlet is taken as water with following properties.

Water
Density 1000 kg/m3
Coefficient of viscosity 10-3 Pa.s
Simulation in ANSYS consist of various steps that includes sketching of model in geometry section of
ANSYS-workbench, extrusion and apply boundary conditions by defining wall, inlet, outlet and fluid,
apply material of model, development of very fine mesh by discretization of whole body into small
matrices, in setup section drawing contours of velocity, pressure and streamlines, plotting a graph
between pressure drop vs different Reynold’s numbers, calculating average shear force at inlet and exit
through fluent solution section and view results in report form. On the other hand, writing governing
equations for given problem and theoretically find pressure drop and average shear force by applying
some assumptions like constant temperature, no slip condition, zero wall thickness and steady state
condition etc. Finally compare the results with theoretical solution and calculation percentage error.
Give reasons and conclusion of why there is error in results comparison. Now, let’s comes towards
literature review of CFD in ANSYS fluent.
Continuous equations, Navier-Stokes equations, and energy equations are the primary governing
equations that regulate the physics of fluid mechanics and environmental sciences and are used in
computational fluid dynamics (CFD) investigations [1]. The governing equations are derived from
Newton's Laws and Reynold's Transport Theorem, that might be represented in a general version of
integral equations. Yet, such a broad form makes exact analysis right down to the scale of fluid element
packets difficult [2]. McFarland and Landy (1980) conducted water testing with three distinct circular
pipe layouts. They collected transient and steady-state data for fluid pressure and temperature and
compared it to ocular observations of mixing processes in various pipe layouts. One of their most
important findings was that adequate mixing between similar fluids may be achieved within a modest
width (L/D 10) without significant pressure drops [3]. A. Began (2000) focused on the geometric
optimization of circular pipes in order to reduce turbulence and pressure errors, and hence the energy
wasted during fluid flows. They primarily computed the ratios of main and branch pipe diameters and
length ratios to minimize damages [4]. Vaasa (2007) made essential contributions to the study of head
loss caused by fluid flow through a circular pipe and pipe angles. He used computations to get the
results for all types of flows at various velocity ratios and compared them to the empirical formulas of
Vazsonyi and Gardel. He also studied the relationship between angle and radius and the difference in
head loss as computed by formula and numerical approaches [5]. Stigler (2011) investigated a more
ideal version of a circular pipe, assuming that the pressures at the ends of pipes are the same from input
to output. They compared PIV measurement to numerical modeling for pipe flow. There were two
items that were compared. The first one was a streamlining comparison. It depicts an overall
perspective as well as an assessment of fluid flow in a pipe. The second is a comparison of velocity
profiles at each pipe junction branch [6]. Nimbalkar (2010) was especially interested with reducing
thermal fatigue in pipe-joints caused by the mixing of two fluids or the same fluid at different
temperatures. Their primary objective was to relate branch and main pipe velocity ratios to velocity and
temperature changes in the main pipe. They likewise established the minimal length necessary for exit
and entrance effects by simulating different lengths of the pipe [7].
Continuity equation of Newtonian Fluid flowing over the boundary in three dimensional Cartesian
coordinate is given below:
𝜕𝑢 𝜕𝑣 𝜕𝑤
+ 𝜕𝑦 + = 0 [8]
𝜕𝑥 𝜕𝑧
Similarly we have DARCY–WEISBACH equation for calculation of pressure drop across ducts as
shown below:
𝐿𝑉 2
∆𝑃 = 𝑓 [9]
2𝐷
3 MATHEMATICAL FORMULATION
In this project we have two circular pipes that are inter connected to each other at common centerline,
fluid water flows from left (inlet) to right (outlet) and after some length there is sudden contraction
occur, the main factor for pressure losses and shear force increment is friction. We are assuming
laminar flow, steady state condition and ignoring frictional heating of fluid etc. Shear force created at
wall due to this friction irrespective off that we are considering no-slip condition as shown in figure
below:

if we know the mass flow rate, we can easily calculate the average velocity and similarly velocity
streamlines (represented in a figure) by using following equation:
𝑚̇ = 𝜌𝑉𝑎𝑣𝑔 𝐴𝑐
𝑚̇
𝑉𝑎𝑣𝑔 =
𝜌𝐴𝑐

Where, Ac is a cross-sectional area of pipe and 𝜌 is a density of flowing fluid.

Figure 3.1

Either the fluid in pipe is laminar or turbulent can be estimated by using the equation of Reynold’s
number that is:

𝜌𝑉𝑎𝑣𝑔 𝐷
𝑅𝑒 =
𝜇

Where D is the diameter of pipe and 𝜇 is a coefficient of viscosity of flowing fluid.

Pressure drop at two points in a channel can calculated from the following relation:

32𝜇𝐿𝑉𝑎𝑣𝑔
∆𝑃 = 𝑃1 − 𝑃2 =
𝐷2
This is a total pressure drop from inlet to outlet of pipe, where D and L are diameter and length of pipe.
There are other pressure losses as well that is due to viscous effects like friction with the wall and
sudden contraction (vacuum in streamlines generated) due to change in cross-section of pipe. The
expression is as follows:
2
𝐿 𝜌𝑉𝑎𝑣𝑔
∆𝑃𝐿 = 𝑓
𝐷 2
Where f is a Darcy friction factor,

8𝜏𝑤
𝑓= 2
𝜌𝑉𝑎𝑣𝑔

For circular, horizontal and laminar flow pipe f is given by,

64
𝑓=
𝑅𝑒
Where Re is a Reynolds number, that shows the nature of flow either laminar, transitional or turbulent.
Similarly, for average shear stress, we calculate average shear force for laminar flow and given
assumption.

𝜏𝑎𝑣𝑔 = 𝜇𝑉𝑎𝑣𝑔
𝑠ℎ𝑒𝑎𝑟 𝑓𝑟𝑜𝑐𝑒 = 𝜏𝑎𝑣𝑔 (𝐴) = 𝜇𝑉𝑎𝑣𝑔 (𝐴)

Effect of surface roughness and friction between fluid layers with itself is ignored.
Here are some equations of motion of fluid flowing in computational fluid dynamics domain that
shows fluid behavior with three equations in Cartesian coordinate system and four unknowns.

𝜕𝑢 𝜕𝑣 𝜕𝑤 1 𝜕𝑃′ 𝜕 2𝑢 𝜕 2𝑣 𝜕 2𝑤
𝑥 − 𝑚𝑜𝑚𝑒𝑛𝑡𝑢𝑚: 𝑢 +𝑣 +𝑤 =− +𝑣( 2 + 2 + 2 )
𝜕𝑥 𝜕𝑦 𝜕𝑧 𝜌 𝜕𝑥 𝜕 𝑥 𝜕 𝑦 𝜕 𝑧

𝜕𝑢 𝜕𝑣 𝜕𝑤 1 𝜕𝑃′ 𝜕 2𝑢 𝜕 2𝑣 𝜕 2𝑤
𝑦 − 𝑚𝑜𝑚𝑒𝑛𝑡𝑢𝑚: 𝑢 +𝑣 +𝑤 =− +𝑣( 2 + 2 + 2 )
𝜕𝑥 𝜕𝑦 𝜕𝑧 𝜌 𝜕𝑦 𝜕 𝑥 𝜕 𝑦 𝜕 𝑧

𝜕𝑢 𝜕𝑣 𝜕𝑤 1 𝜕𝑃′ 𝜕 2𝑢 𝜕 2𝑣 𝜕 2𝑤
𝑧 − 𝑚𝑜𝑚𝑒𝑛𝑡𝑢𝑚: 𝑢 +𝑣 +𝑤 =− +𝑣( 2 + 2 + 2 )
𝜕𝑥 𝜕𝑦 𝜕𝑧 𝜌 𝜕𝑧 𝜕 𝑥 𝜕 𝑦 𝜕 𝑧

The assumption for above equation is that fluid is Newtonian, flow is steady, incompressible, laminar
with constant properties with free surface effect.
In the analysis section we will put values to these equations, and compare the results from numerical
solution given by software, we will check the trend of pressure drop from uniform section to lower area
section of pipe, similarly, shear force comparison with velocity profile across length is going to be
discussed in detail. Other equations of motion are transport equation or Navier-Stokes equation,

1 ′
(𝑉 ⃗ )𝑉
⃗ .∆ ⃗ =− ∆⃗ 𝑃 + 𝑣∆2 𝑉
⃗
𝜌
4 RESULTS AND DISCUSSION
Theoretically we’ll calculate pressure drop and shear force from inlet to outlet, from equations,

32𝜇𝐿𝑉𝑎𝑣𝑔
∆𝑃 =
𝐷2
32(1 × 10−3 )(1.25)(6.6667 × 10−3 )
∆𝑃 =
(0.05 + 0.025)2

∆𝑃 = 0.047 𝑃𝑎

64 8𝜏𝑤
= 2
𝑅𝑒 𝜌𝑉𝑎𝑣𝑔

𝑠ℎ𝑒𝑎𝑟 𝑠𝑡𝑟𝑒𝑠𝑠 = 𝜏𝑤 = 7.11 × 10−4 𝑁/𝑚2

𝑠ℎ𝑒𝑎𝑟 𝑓𝑟𝑜𝑐𝑒 = 𝜏𝑎𝑣𝑔 (𝐴) = 1.745 × 10−6 𝑁

4.1 PROBLEM WITH BOUNDARY CONDITION

𝑫𝟏 0.05m
𝑫𝟐 0.025m
𝑳𝟏 0.5m
𝑳𝟐 0.75m
Errors in Velocity:
Errors in velocity profiles between analytical and numerical results could be caused by turbulence
effects that were overlooked in the analytical model. Turbulence is inherently captured by numerical
simulations, contributing to discrepancies in velocity forecasts.
Errors in Pressure Drop:
The discrepancy in pressure drop can be related to the analytical model's presumptions, such as the
absence of turbulence effects and simplified geometry considerations. Pressure drop forecasts are
influenced by numerical simulations, offering a more thorough depiction of flow events.
Differences in Shear Force:
Differences in average shear force at the wall might be caused by different modeling methodologies.
Certain fluid characteristics that numerical simulations capture more thoroughly may be missed by the
analytical solution.
Treasures of Number:
Differences between analytical and Fluent findings might be attributed to numerical aberrations such as
mesh-induced errors or solver parameters. Mesh refinement and sensitivity assessments can help to
reduce these inconsistencies. Finally, variations between analytical and numerical results are to be
predicted due to analytical model constraints and the additional intricacies incorporated in numerical
simulations. Identifying these variances is critical for appropriately interpreting simulation results and
enhancing simulation integrity in future research.
Screen shot of every step in simulation is available in the report.

All commands in design modeler in geometry section of simulation is given below:

Fine mesh is generated in mechanical design modeler,

All boundary conditions are applied like inlet, wall and outlet in meshing section as shown below,

We have applied steady, absolute, laminar and pressure based conditions in ANSYS Fluent as shown
below:
Similarly, we have applied boundary conditions in ANSYS Fluent
and given a proposed average velocity of 50 m/s at the inlet of pipe
shown in figure,

In the solution section we have

calculated the continuity, x, y,
z velocity by applying 75
iterations, we have observed a
sinusoidal behavior of velocity
profile in Cartesian
coordinates as shown in
screenshot of ANSYS-
FLUENT.
Velocity streamline across inlet and outlet is created,

Velocity contour is also created as shown below,

Pressure contour is also created as shown below,

Pressure drop across the length of pipe is as follows at Re = 500,

Figure no. 4.1

Velocity change at inlet and outlet is given by following figure,

Figure 4.2

Average shear stress is calculated from following result:

Similarly, friction coefficient at the wall is calculated below,

Average shear force from inlet to outlet is given be figure:

Pressure drop from inlet to outlet in Fluent is as follows,

There is a contraction in pipe, that’s why force and velocity first decrease then increase, reverse case
for pressure, as elaborate from above plots and figures.

Pressure Drop vs. Reynolds Number: A plot is generated displaying the pressure drop across the ducts
with regard to Reynolds numbers. This graph illustrates the relationship between pressure drop and
flow conditions and aids in understanding the hydraulic operation of the system.

Evaluation of Average Shear Force and Pressure Drop: Real-time results are compared to analytical
solutions for average shear force at the wall and pressure drop between the intake and outflow.
Analytical and numerical findings may differ due to biases in the mathematical framework, turbulence
effects not considered in the analytical solution, and numerical renditions in the modeling process.
5 SUMMARY AND CONCLUSIONS
Analytical results are better than software results because the ratio of assumption is lower in analytical
solution. There are lots of variation in actual verses animation cases exits like:

 Changes in friction
 Flaws in mesh
 Material used is aluminum, material properties are not included.
 Flow turbulence
 Wall thickness and weight

The gas flow properties within the interconnected square ducts were successfully captured by the CFD
simulation in ANSYS Fluent, offering useful insights into the system's behavior.
There were discrepancies between numerical and analytical conclusions, highlighting the need of
taking turbulence effects, geometric complexities, and numerical artifacts into account.
Differences in velocity profiles, pressure drop, and shear forces revealed the drawbacks of
mathematical equations in capturing precise fluid dynamics, demonstrating numerical simulations'
advantage in delivering more precise results.

Recommendations:

 Perform further sensitivity assessments on mesh resolution to verify the robustness of

simulation findings and to reduce mesh-induced errors.
 Investigate the effect of various turbulence models and boundary layer treatments on the
accuracy of CFD simulations.
 Look into including additional physics, such as heat transmission, to broaden the simulation's
scope and provide a more complete picture of the system.
 Coordinate with experimental investigations to validate numerical findings and improve
simulation overall dependability.
 Consistently modify and validate the numerical model in light of forthcoming studies and
advances in CFD methodology.

Finally, the research used ANSYS Fluent to successfully simulate fluid flow in interconnected square
ducts. Disagreements with analytical solutions highlighted the need of taking into account the
intricacies inherent in everyday situations, underlining the need for more advancement and innovation.
6 REFERENCES

[1] Tey-Wah-yen, "equations in CFD". china Patent http://www.akademiabaru.com/ProgEE.html,

2022.

[2] N. Azwadi, "Simulation in CFD technique". China Patent 10455, 2015.

[3] A. chaturvedi, "CFD SIMULATIONS OF FLOW AND PRESSURE DROP IN PIPE". India Patent
FM-059, 2011.

[4] A. Rocha, "Thermodynamic optimization of geometry". Japan Patent Int. J. Therm. Sci. 39, 949–
960, 2000. , 2000.

[5] V. S. Naik-Nimbalkar, "Thermal mixing in T-junctions". USA Patent 534, 2009.

[6] Stigler, "The fluid flow in the t-junction". Florida Patent 5-6, 2012.

[7] Nimbalkar, "Thermal mixing in T-junctions," Oxford University Press, Manchester, 2023.

[8] Cengel, Fluid mechanics a modern approach, 2020.

[9] KOTHAND, Fluid Mechanics Mathematical solution, 2017.