IRJET- Tall-Building Structure Shape Optimization using “Computational Fluid Dynamic (CFD)”

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 07 Issue: 01 | Jan 2020 www.irjet.net p-ISSN: 2395-0072
© 2020, IRJET | Impact Factor value: 7.34 | ISO 9001:2008 Certified Journal | Page 6
Tall-Building Structure Shape Optimization using “Computational
Fluid Dynamic (CFD)”
Samir Kumar Singh1
1M.Tech in Structural Engineering, Dept. of Civil Engineering, Indira Gandhi Institute of Technology, Odisha, India.
---------------------------------------------------------------------***----------------------------------------------------------------------
Abstract - Wind plays an important role in tall buildings.
So it is necessary to analyze wind behavior around and over
these tall buildings. In tall buildings, wind pressure can
reduce by adopting the suitable shape of buildings. CFD
(computational fluid dynamics) simulation is a suitable
option for study the sensitivity of wind pressure over tall
buildings. Analyzing the different building shape with
constant air flow around the buildings; after that, we can
select one of the building shapes. In this paper, we are using
ANSYS 17.2 for CFD simulation. Here we analyze turbulent
flow; so we are taking a turbulent model which is
“realizable k-ε turbulent model”. We are taking realizable k-
ε turbulent model because it is more accurate than SST k- ε,
standard k- ε, RNG k- ε models. A series of wind test have
been carried out to determine the wind characteristics of
many buildings with various configurations.
Key Words: Ansys, CFD, Tall Buildings, Wind analysis,
Turbulent flow.
1. INTRODUCTION
A tall building cannot be defined in specific terms
related just too height or the number of the floors. A
measurable definition of a tall building cannot be
universally applied because the tallness of a building is a
matter of community’s circumstance and their consequent
perception [1]. From the structural engineer point of view,
a tall building can be defined when the structure is
affected by the lateral loads. Due to lateral forces building
sway or drift at the top of the building relative to its base.
Drift is the magnitude of lateral displacement. The lateral
forces on buildings are increase with the height of the
building. Wind and earthquake are the important lateral
forces which play an important role in structural design.
Therefore these forces considered from the very beginning
of the design process. In tall buildings, wind loads are
one of the most symbolic forces which are responsible for
natural disasters.
Wind forthcoming a building is a complex issue; where
the flow pattern generated around the building is
complicated. Large wind pressure fluctuations develop on
the surface of a building due to the mean flow, the flow
separation, the vortex formation, wake development
characteristics of wind. As a result, large aerodynamic
loads are imposed on the structural system and the
fluctuation of force also large on the elevation of the
building. Due to these fluctuating forces; a building will
vibrate in linear as well as a torsional mode. This is called
vortex shedding. The amplitude such of such oscillations is
dependent on the nature of aerodynamic forces and the
stiffness characteristic and dynamic characteristic of the
buildings [2]. Therefore, the building shape is also the
most important aspect of a tall building design. Effects of
wind on tall buildings mainly depend on the magnitude of
wind speeds and directions of wind. The response of
buildings to wind loads basically 3 types as illustrated in
Figure 1; these are:-
1. Along with wind response
2. Crosswind response
3. Torsional effects
1.1 Along with wind response
Due to wind flow on the face of building the building
displace horizontally in the direction of wind flow. This
displacement is occurring due to drag forces. Pressure
fluctuations on the windward face and leeward face (back
face of the building) because of these drag forces.
Fig -1: Wind response directions
Some techniques such as the “gust loading factor”
approach are developed to predict forces as well as a
response in the along wind direction. The gust factor
approach is used in international standards codes to
calculate the dynamic forces along wind direction and
their effects on high rise buildings [3].

1.2 Cross wind response
The crosswind response means deformation
perpendicular to the wind flow direction due to wind
forces. Tall buildings have opposed the streamline of wind
that causes the flow to separate from the surface of the
structure, rather than follow the body contour. The wake
flow thus created behind the building produce various
degrees of periodicity, ranging from virtually periodic with
a single frequency to fully random. In each of the cases, at
any given instant, the wake flow is asymmetrical. The
lateral turbulent fluctuations in the oncoming flow may
also contribute to the crosswind forces. The complex
nature of the crosswind loading which results from an
interaction of incident turbulence, unsteady wake effects,
and building motion has inhibited reliable theoretical
predictions.
1.3 Torsional effects
Torsional or twisting moments are induced in the
buildings when the wind flows normal to the face of
prismatic type buildings by variations in the fluctuating
wind velocity around the face of the buildings. The larger
building has larger width also, so the greater will be the
fluctuating torsional moment. For the turbulence,
buffeting and vortex shedding these fluctuating torsional
moments are generate. Because of the eccentricity
between the instantaneous aerodynamic center and
building’s center of rigidity; torsional motions of a
building occur.
Most of the modern buildings and structures usually
result in a natural frequency in a torsional mode that is
greater than the lowest translational natural frequency.
Hence, torsional motions will develop only after lateral
motions are induced in the structure. By structural
modifications; the effects of the torsional moment can be
controlled. Due to the lack of firm guidance from design
codes to estimate the torsional effects accurately,
structural engineers depend on “wind tunnel tests” to
evaluate the torsional behavior of tall buildings.
2. COMPUTATIONAL FLUID DYNAMICS (CFD)
One of the main end products of this analysis is to find
the pressure variations on the surface of the building using
a rigid model approach. However, wind tunnel testing has
its own limitations, including the disguise of Reynolds
number due to the use of scaled models. Mr. E Achenbach
has been investigated the effect of Reynolds number on
the pressure coefficients of circular cylinders [4]. The
output showed that pressure values changed according to
the different Reynolds numbers. The scaled down models
used in wind tunnel and full-scale wind flow have
significantly different values of Reynolds numbers. Wind
analyzer generally adopts rough surfaces to simulate
turbulent flow which is comparable to real life condition;
however, this is a rudimentary method as the degree of
roughness required to simulation the real wind behavior is
unknown. Therefore, significant data studies are required
before the simulation for selecting the roughness of
elements. And another additional limitation of wind tunnel
test is time requirement, cost and need for technical
expertise. Computational Fluid Dynamics is a good
solution for the above difficulty.
In the modern world, Computational Fluid Dynamic
(CFD) is progressively used in wind engineering to analyze
wind loads, pedestrian level wind comfort and wind energy
harvesting, pollutant dispersions. The main advantage of
CFD is the capability to simulate full-scale models within
full-scale atmospheric boundary layer (ABL) flows by
creating a virtual domain. In wind tunnel test parameters
are only calculated at selected points, but CFD can provide
detailed information about wind pressure, velocity,
streamline, drag force, lift force, streamline of flow, etc.
CFD is more flexibility than conventional wind tunnel
method because wind tunnel test requires material, skilled
labor, time resources[5].
Computational Fluid Dynamic (CFD) is a branch of fluid
mechanics that use to analyze and solve problems that
involve fluid flows by using numerical analysis and data
structures. To simulate the free stream flow of the fluid,
computers are used to perform the required calculations.
High-speed super-computers are required when the
simulation is a complex problem. Some time high-speed
super-computers are used to calculate an accurate
solution. In CFD analysis, the examination of fluid flow in
accordance with its physical properties such as velocity,
pressure, temperature, density, and viscosity is conducted.
3. TURBULENT FLOW
In fluid dynamic when the flow motion is disorderly
changes in pressure and flow velocity; then the flow is
called as turbulence or turbulent flow. Turbulence is
commonly observed in our daily life such as fast flowing
rivers, billowing in storm clouds, smoke form by a chimney
and most fluids flows occurring in nature and created in
engineering application are turbulent. When the excess
kinetic energy generated in part of the fluid flow; then
turbulence is generated in fluid flow. Because of the kinetic
energy of fluid overcomes the damping effect of the fluid’s
viscosity. Turbulence is easier to create in low viscosity;
but more difficult in a highly viscous fluid.
The Reynolds number (Re) is an important dimensionless
quantity in fluid mechanics; which is used to describing the
flow characteristic whether it is laminar flow or turbulent
flow. If the Reynolds number is low then the flows are
laminar flow and if the Reynolds number is high then the
flow is turbulent flow. In case of flow through pipes of
diameter ‘D’ in a fully developed region, laminar flows
occur when Reynolds number (Re) is less than 2300 and

turbulent flow occur when Reynolds number (Re) is greater
than 2900. In case of flow over a Flat plate; if the Reynolds
number (Re) is above the value of (2*105) to (3*105) then
the flow is called as turbulent flow. Normally Reynolds
number (Re) is above (5*105) is taken as turbulent flow.
Reynolds number can be defined as;
Where,
ρ = Density of the fluid (SI units: kg/m3)
= Characteristic velocity of the fluid with respect to the
object (m/s)
L = Characteristic linear dimension (m)
µ = Dynamic viscosity of the fluid (Pa·s or N·s/m2 or
kg/(m·s)).
Fig -2: Velocity versus time
Refer to Fig -2,
Where,
t = time
= velocity
̅ = Reynolds average velocity
= fluctuation velocity
As shown in Fig -2; this is a graph of velocity versus
time. As it is a turbulent flow; so the velocity is fluctuating
at a high rate. We consider an average velocity ‘̅’ and
fluctuation about the average velocity is “ ”; which is
called as Reynolds average velocity and fluctuation
velocity. I can decompose the velocity ‘ ’ into the average
velocity ‘̅’ and fluction velocity ‘ ’.
We can write,
̅
The above equation is called as “Reynolds
decomposition”. In most engineering problems we need to
calculate only in “Reynolds average velocity” and
“Reynolds average pressure”.
3.1 Reynolds Average Navier- Stokes Equation
In Physics, the Navier-Stokes Equation named by
‘Claude-Louis Navier’ and ‘George Gabriel Stokes’. This
equation describes the motion of viscous fluid substances.
Navier-Stokes Equation arises by applying ‘Newton’s
second law’ to ‘fluid flow motion’, and then together with
the assumption that the stress in the fluid is the sum of the
fluid is the sum of ‘diffusing viscous’ term and a pressure
term which can describe viscous flow. Navier-Stokes
Equations are very useful for describing and analyzing the
physical phenomena of scientific and engineering. This
equation can model the ocean currents, water flow in a
pipe and air flow around a wing can help to design of
aircraft and cars, the study of blood, etc.
Navier-Stokes Equations for x-direction moment of
conservation is written below,
( )
Here, is an extra term comes dure to unsteadiness in
the flow: the flow is not really steady here because we
have these fluctuations. So we have to extend the Navier-
Stokes Equation to ‘unsteady case’ and we will get this
extra term .
4. TURBULENCE MODELS
In CFD simulations ‘turbulence modeling’ is a key issue.
All most all engineering applications are turbulent and
hence turbulence model is very important in CFD
simulations. Turbulent models are constructed for use of
the mathematical model to predict the effects of
turbulence. In case of simple flow only, the equations
governing turbulent flow can be solved directly. In most of
the real-life turbulent flows; CFD simulations predict the
evolution of turbulence by using the turbulence model.
Constitutive relations predict the statistical evaluation of
the turbulence model, and these constitutive equations are
simplified by turbulence models.
4.1 Two equation turbulence models
Two equation turbulent models are one of the most
common types of turbulence models. Models like the k-
epsilon model and k-omega model are commonly used for
most types of engineering problems; hence these models

have become industry standard models. Two equation
turbulence models are also very much use in research
work and new refined two-equation models are still in
underdeveloped.
According to the definition, two equation models include
two extra transport equations to represent the turbulent
properties of the flow. These extra transport equations
allow a two-equation turbulence model to account for
history effects like convection and diffusion of turbulent
energy. Mostly one of the transported variables is
‘turbulent kinetic energy (k)’. The second transported
variable varies depending on what type of two-equation
model it is. Commonly used are ‘turbulent dissipation ( )’
or the specific turbulence dissipation rate (ω).
The first variable kinetic energy (k) determines the
energy in the turbulence and second variable can be
thought of as the variable that determines the scale of
turbulence (i.e; length-scale or time-scale). In complex
flows, it is impossible to guess the distribution of length
scales. It may be difficult to define the distance to the wall
and the local level of ‘ ’ is affected by convection and
diffusion processes. One approach to avoid having to
prescribe ‘l’ building on the modeling, by adopted for ‘k’ is
to obtain ‘l’ from its own separate equation. Thus, we
devise a transport equation for some length scale
containing a variable of the form “ ”. The most popular
choice of the second variable has been esplion (ε) which is
the dissipation rate of ‘k’.
We know that
Here, a= 3/2 and b= -1
Other choices of the variable have been proposed which is
length scale (l) or frequency (f).one alternative that has
been widely tested is ω = ε/k
Generally, two types of two-equation models are mostly
used. Which are,
1. k-epsilon model (k-ε)
2. k-omega model (k-ω)
5. METHODS
5.1 Outline of the model
This simulation was conducted in “computer lab; civil
engineering department” of Indira Gandhi Institute of
Technology, Sarang. I am using ANSYS 17.2 software for
my simulation work. For CFD simulation I am using ANSYS
FLUENT tool.
In this simulation my interest to find the best shape for
the tall building when turbulent flow around the tall
building using aerodynamic. In other words, this
simulation is conducted for tall buildings shape
optimization using aerodynamic. Here I am taking six
different shapes of the building to find which shape is
better. Different shapes of the building are a circular
shape, square shape, rectangular shape, square with four
edge chamfer shape, Y shape, and oval shape. Again I take
six different dimensions of oval shape building for better
analysis. In my simulation, I am taking a total of eleven
number of the different shape of buildings. Floor area of
each building model is 100 m2 and the height of the
building model is 50 m. Flow domain shape varies
according to the shape of the building, in some case it is
square in shape and in other it is rectangular. In our case, I
consider floor area of the flow domain is 2500 m2 and
height is 70 m.
Wind specification:-
Wind velocity = 50 m/s
Viscosity of wind = 1.7894*10-5 kg/ms
Density of wind = 1.225 kg/m3
Let’s take a building model of square shape.
Fig -3: flow configuration and boundary locations for the
square building
The flow configuration and boundary locations for the
square building are shown in Fig -3. In this simulation
turbulent wind flow from the inlet to the outlet within the
flow domain. We consider turbulent wind flow velocity is
50 m/s. I want to find the pressure around the building
wall after turbulent air flow around the building within
the flow domain. On which building shape the pressure on
the building wall is small, that one is the best aerodynamic
shape of the tall building.
In my case there are three boundary conditions are
present in my model.

1. Inlet
2. Outlet
3. Wall
Here I am taking six different shapes of the building
shown in Fig -4. All building has the same floor area which
is 100 m2 and the same height also which is 50m. One by
one I will check which building shape reduce the wall
pressure. In which building small pressure acting on the
wall will be the best shape for our tall buildings.
The input parameters I will use are:-
The cross-sectional area of the building is 100m2.
Free stream velocity is 50 m/s.
The density of air is 1.225 kg/m3
The viscosity of air is 1.7894*10-5 kg/ms
Fig -4: Different shapes of the building
6. RESULT AND DISCUSSION
6.1 Test result for square building
Fig -5: Square shape building model
As shown in Fig -5; I take a square shape building and
start my simulation. After applying wind load our interest
to find out the net pressure acting on the surface of the
building. I found that the maximum pressure on the
surface of the building is 3071.43 Pa.
The velocity of wind also changes in different place
within flow domain. The velocity of the wind varies from 0
m/s to 72.8261m/s.
Fig -6: Velocity streamline
Fig -7: Velocity contour
2D represented of velocity contour of wind flow is
shown in Fig -7. From the above figure, we found that
velocity is decreased near the wall. Velocity is almost zero
at the wall of the building. The velocity of wind gradually
increases from the wall and after some distance from the
wall; the velocity is at the pick. After approaching the
maximum velocity the wind speed again decreases.
Fig -8: Surface pressure

Chart -1: Surface pressure
By pressure coefficients, the surface pressure on a
building is generally described. A pressure coefficient
describes the relative pressure throughout a flowing fluid
in fluid dynamics and it is a dimensionless number. In
aerodynamics and hydrodynamics, the pressure
coefficient is used. Pressure coefficient has its own unique
pressure coefficient at every point in a fluid field. Pressure
coefficient calculated through normalization with velocity
pressure in the undisturbed approaching wind at
reference height; which is often chosen as the height of the
building.
Pressure coefficient can represent as “Cp”.
Where
P = static pressure at the point at which pressure
coefficient is being evaluated.
P∞ = static pressure in the free stream.
v∞= free stream velocity
Chart -2: Pressure coefficient
Chart -2 graphically represents the pressure coefficient
against the building height. In this simulation maximum
pressure coefficient for square shape building is ‘1.97’.
6.2 Comparison of all shapes building
I have already conducted simulation against the different
shapes of the building, and those are a square shape, a
rectangular shape, square with chamfer shape, a circular
shape, Y shape. Now I am going to compare and study
which is the best shape for the tall building.
Chart -3: Surface Pressure of different shapes building
Chart -4: Surface Pressure of different shapes building
In Chart -4, I represent the surface pressure against the
different shape of the buildings. Here also clearly shown
that surface pressure acting very little on the circular
shape of the building, and the maximum amount of surface
pressure acting on the rectangular shape building. On
square with chamfer shape building; surface pressure is
little larger then the circular shape and Y shape building’s
surface pressure is larger than square with chamfer shape
building. On square shape, surface pressure is less than
the rectangular shape. Maximum surface pressure on the
different buildings is given in Table -1.
Table -1: Maximum Surface Pressure
Building Shape Maximum Surface
Pressure (in Pa)
Square 3071.95
Square with Chamfer 2202.78

Rectangular 4096.06
Circular 1737.46
Y 2995.53
6.3 Test result for oval shape building
We already see that the circular shape of the building is
the best shape because of the pressure coefficient very low
in the circular shape building. Now I am going to modify
the circular shape building to increase the floor area. So I
manipulate the dimension of circular shape building to
convert to oval shape building as shown in Fig -9.
Fig -9: Oval shape building
Here I am going to change the dimension of the length
(L) but the width will be constant at 10m. With different
dimension, I will conduct the simulation to check the
pressure parameter. I take a Y shape building and start my
simulation. After applying wind load our interest to find
out the net pressure acting on the surface of the building.
Here I will conduct four different dimensions of oval shape
building. The dimensions of the length (L) of five oval
shape model are 0.5W, 1.0W, 1.5W, 3W. And the width (W)
of five different models will be the same which is 10m.
Height remains the same for all buildings which are 50m.
Table-2: Maximum Surface Pressure of oval building
Oval dimension (L)
MAXIMUM SURFACE
PRESSURE IN (Pa)
0.5W 2299.05
1.0W 2321
1.5W 2345.05
3.0W 2390.22
After simulation all the four building I found maximum
pressures as shown in Table-2. Here I found that if I
increase the length the building then surface pressures are
also increasing accordingly.
Fig -10: maximum surface pressure of all buildings
7. CONCLUSIONS
First I consider five different building shape which is a
circular shape, square shape, rectangular shape, square
with chamfer shape, and Y shape. After simulation, these
five shape building I get minimum surface pressure on the
circular building which is 1737.46(Pa), and maximum
surface pressure on the rectangular shaped building which
is 4096.06(Pa). And the floor area of both circular and the
rectangular building is 100m2. After I conduct the same
simulation on the oval shaped building. I take four oval-
shaped building with different length, i.e; 0.5W, 1.0W,
1.5w, 3.0W. Height remains the same for all buildings
which are 50m. Here I get maximum surface pressure on
the oval building which length is 3.0W, and surface
pressure value is 2390.22(Pa).
On rectangular shaped building I get 4096.06(Pa) of
surface pressure which floor area is 100m2 only. But in the
case of the oval-shaped building, I get 2390.22(Pa) of
surface pressure which floor area is 378.5m2. So oval
shape is more suitable for tall building than the other four
models (i.e; square shape, rectangular shape, square with
chamfer shape, and Y shape) because surface pressure is
small even the floor area is larger than another model. So I
take oval shape as the best shape for a tall building which
is the best shape to resisting the turbulent air flow.
REFERENCES
[1] Structural analysis design of tall buildings by Bungale
S. Taranath.
[2] wind analysis and design of tall buildings, the state of
the art (ICSECM2017 – 463)
[3] Li, Q.S., J.R. Wu, S.G. Liang, Y.Q. Xiao, C.K. Wong, 2004.
Full-Scale Measurements and Numerical Evaluation of
Wind-Induced Vibration of 1 63-Story
Reinforcedconcrete Tall Building. Engineering
Structures.

[4] Distribution of local pressure and skin friction around
a circular cylinder in cross flow up to Re = 5*106 By E.
Achenbach
[5] Comparison of Reynolds Averaging Navier-Stokes
(RANS) turbulent models in predicting wind pressure
on tall building ; By D. Mohotti , K. wijesooriya
AUTHOR
Name :- Samir Kumar Singh
BE (Civil Engineering)
ME (Structural Engineering)

IRJET- Tall-Building Structure Shape Optimization using “Computational Fluid Dynamic (CFD)”

More Related Content

What's hot (20)

Similar to IRJET- Tall-Building Structure Shape Optimization using “Computational Fluid Dynamic (CFD)” (20)

More from IRJET Journal (20)

Recently uploaded (20)

IRJET- Tall-Building Structure Shape Optimization using “Computational Fluid Dynamic (CFD)”