The Eighth Asia-Pacific Conference on Wind Engineering,

December 10–14, 2013, Chennai, India

STRUCTURAL DESIGN OF TALL NDCT BASED ON BOUNDARY

LAYER WIND TUNNEL EXPERIMENTAL STUDIES ON GROUP OF
TOWERS

Girish Patil1, M.K.Nema1, Arvind Shrivastava1

1
Civil Engineering Group, Nuclear Power Corporation of India Ltd., Mumbai.

ABSTRACT

Wind loading is extremely important in cooling tower design for several reasons. First of all, the amount of
reinforcement, beyond a prescribed minimum level, is often controlled by the net difference between the tension
due to wind loading and compression due to dead load. Second, the quasistatic velocity pressure on the shell wall
is sensitive to the vertical variation of the wind and circumferential variation of the wind around the tower. There
are also additional wind effects such as internal suction, dynamic amplification, and group configuration. The
wind load variation along height and around circumference is specified in the codes like IS 11504 and BS 4485
with a restriction on height of the tower not greater than 120m. Nuclear Power Corporation of India Limited has
proposed to set up NPP at Kakrapar, where sill diameter of NDCT is expected to be as large as 120 m and the
height is expected of the order of 165m. It is proposed to setup two numbers of NDCT for each 700 MWe
capacity. As the tower forms part of tower group and its height is greater than 120 m, aero-elastic model testing
in an atmospheric boundary layer wind tunnel studies has been carried out. It is found from the experimental
studies that there is change in the wind load distribution compared to given the codes. The paper discusses the
structural analysis and design of tall NDCT based on the boundary layer wind tunnel experimental studies. The
paper also discusses the difference in the outcome of structural analysis and design based on the aero-elastic
model testing and approach given in BS 4485 part-IV.

Keywords: Cooling tower, Wind loading, Reinforced concrete shells, Bucking, Amplification factors

Introduction
To meet the growing electricity demand, a number of nuclear power plant (NPP) of
high capacity have been planned and are under construction in India by Nuclear Power
Corporation of India Limited (NPCIL). With the increase in capacity of power plant it
becomes necessary to increase both the height and diameter of cooling towers. It is proposed
to setup two numbers of NDCT for each 700 MWe capacity NPP project, where sill diameter
of NDCT is expected to be as large as 120 m and the height is expected of the order of 165m.
In this, just a linear extrapolation from normal sizes of 120m to a height 165m is not possible,
because the overall dimension like height and diameter are increased non-proportionally in
comparison with shell thickness and reinforcement. Thus the dynamic behavior changes in an
unfavorable way, leading to more bending action in the shell. Cooling tower is largest thin
shell structures of NPP which is highly sensitive to dynamic wind actions and their response
varies randomly in time and space. At such large dimension, rational assessment of wind
loading condition & structural response along with shape of tower is of much importance for
the state of stress (structural safety), for the elastic stability (overall stiffness) and for the
vibration properties (dynamic load amplification) of the structural response, for the initiation
of concrete cracking (durability) compared to smaller towers [Dieter Busch et al., Reinhard
Harte et al., Wilfried Kratzig et al., Ulrich Montag et al. (2002)]. The optimization procedure
including the derived most optimum shape for the tall NDCT which is being proposed to be
constructed at Kakrapara is discussed in detail by Girish Patil, Nema, M.K., Arvind
Shrivastava, Verma U. S. P. (2007). This work has been carried out based on the guidelines

Proc. of the 8th Asia-Pacific Conference on Wind Engineering – Nagesh R. Iyer, Prem Krishna, S. Selvi Rajan and P. Harikrishna (eds)
Copyright c 2013 APCWE-VIII. All rights reserved. Published by Research Publishing, Singapore. ISBN: 978-981-07-8011-1
doi:10.3850/978-981-07-8012-8 P10 353
 Proc. of the 8th Asia-Pacific Conference on Wind Engineering (APCWE-VIII)

given in BS-4485 Part IV for wind loading, interference effects, amplification factors etc. It is
also recommended in the work that as the tower forms part of tower group and its height is
greater than 120 m, aero-elastic model testing in an atmospheric boundary layer wind tunnel
should be studied in detail to determine the stresses in the shell and final design should be
based on the wind tunnel studies. For rational assessment of wind loads and evaluation of
structural responses codal analytical methods can be used which are based on more
sophisticated calculations using the theory of random vibration and are applicable in principle
to isolated towers. Other method based on wind tunnel studies [Cermak Jack E. et al. (2003)]
to find out the wind loading and its associated effects which are accepted internationally as
the most desirable design tool because it considers the project specific factors like the
aerodynamic effect of the actual shape of the structure, influence of adjacent building and
topography, detailed wind directionality effects, aero elastic interaction between structural
motion and airflow etc. Therefore the wind tunnel studies become a requirement before
finalizing the structural design of large NDCT. The paper discusses the structural analysis and
design of tall NDCT based on the boundary layer wind tunnel experimental studies conducted
using BLWT facility at CSIR-SERC [Selvi Rajan, et al. (2010 a&b)]. The paper also
discusses the difference in the outcome of structural analysis and design based on the aero-
elastic model testing and approach given in BS 4485 part-IV.
Wind Load
The natural wind induces fluctuating loads. Their effects may be split into static,
quasi-static and resonant components. For practical application it is common to introduce
equivalent static loads. Their intensity and distribution are determined in such a way that they
cover the peaks of all relevant stresses and strains in the structure with an adequate level of
reliability [Dieter Busch et al., Reinhard Harte et al., Hans- Jurgen Niemann et al. (1998)].
For cooling towers, BS code has adopted the concept of the gust wind pressure which is based
on theory of random vibration. Cooling towers for NPP in India up to height of 120m
designed with this concept, have not experience any noticeable damage in the past. It is clear,
however, that a certain amount of ambiguity is introduced when the complicated random
process is represented by a static equivalent load. Moreover, the concept applies in principal
to isolated towers. When the wind flow is altered by adjacent high building or tower, the
mean and fluctuating loads will be less severely affected. With new dimension of the present
project it is therefore necessary to inspect wind load in detail. With the above in view, series
of wind tunnel tests are performed and the recommended mean external and total pressure
coefficient for isolated cooling tower are given Table 1 [Selvi Rajan, et al. (2010 a)].
Table 1: Recommended external and total pressure Table 2: Fourier Coefficient corresponding to
coefficient for isolated cooling tower IS 11504/ BS4485 and wind tunnel studies
Total Pressure IS 11504 /BS Wind Tunnel
Angle Mean External
Coefficient including n 4485 Studies
(deg.) (Cp)
internal coefficient (Cpș) Fn Fn
0 1 1.4 0 -0.00071 -0.09842
15 0.921 1.321 1 0.24611 0.33656
30 0.313 0.713 2 0.62296 0.92032
45 -0.538 -0.138 3 0.48833 0.39278
60 -1.157 -0.757 4 0.10756 -0.06683
75 -1.437 -1.037 5 -0.09579 0.01449
90 -1.503 -1.103 6 -0.01142 0.01117
105 -1.227 -0.827 7 0.04551 -0.06577
120 -0.63 -0.23 8 -0.02283
135 -0.374 0.026 9 0.00622
150 -0.334 0.066 10 -0.01499
165 -0.345 0.055 11 -0.01429
180 -0.34 0.06

354
 Proc. of the 8th Asia-Pacific Conference on Wind Engineering (APCWE-VIII)

The wind pressure acting at any point on the shell surface is computed as
q (z, ș) = q (z) Cpș (1+g) (1)
Where, q (z) = effective velocity pressure at a height z above the ground level
Cpș = coefficient for circumferential distribution of wind pressure
1+g = gust response factor
g = peak factor
The obtained circumferential pressure coefficient from wind tunnel studies are
represented in a Fourier cosine series of the form
Cpș = ™ An cos (nș) (2)
The comparison of the Fourier coefficient for IS code, BS code along with
recommended circumferential coefficient of the wind tunnel studies are given Table 2.The
circumferential distribution of the wind pressure is shown in Fig. 1. The key regions which
can be compared between pressure distribution obtained from wind tunnel studies and IS/BS
code are as given below:
1. At windward meridian, ș=0o, pressure coefficient are same for both the case.
2. Side Suction: a) Starting of side suction at ș=42.5o is same for both the case.
b) Maximum side suction: at ș=70o , Cp (70)= -0.927 for IS/BS case
at ș=90o, Cp (90)=-1.103 from wind tunnel studies
c) End of side suction : at ș=112.5o for IS/BS case
at ș=132.5o from wind tunnel studies

Fig. 1 Circumferential pressure distribution (including internal suction) around shell cross section
It appears from Fig.2 that, n=0 mode represent uniform expansion and contraction of
the circumference; n=1 corresponds to beam like bending and higher mode i.e. n > 1 are
related to shell undulating deformations. The relatively large fourier coefficient associated
with n=2, 3, 4 indicate that a significant portion of the loading is causing shell deformation. It
appears from Table 2 and Fig. 2 for n=0 that the contribution of expansion and contraction of
the circumference is more corresponding to wind tunnel studies. It can also observed that the
contribution of beam like bending will be more corresponding to wind tunnel studies. From
the above, it can be concluded that wind tunnel pressure distribution would cause more axial
tension or compression and bending of cooling tower shell compared to codal approach. It is
also concluded from the Fig. 2 that fourier coefficient are more or less same for higher modes,
hence it is expected that the factor of safety against bucking will be same for both the case.

Dynamic Amplification
The dynamic amplification of the effective velocity pressure is represented by the
parameter g in equation (1) which is corresponding to resonant part of the response of the
structure. Value of this response component is dependent upon the dynamic characteristics of
the tower. The variation of the gust factor w.r.t height corresponding to recommendation
given in IS 875, BS: 4485 and wind tunnel studies are shown in Fig.3. BS-4485(1996) has

355
 Proc. of the 8th Asia-Pacific Conference on Wind Engineering (APCWE-VIII)

introduced a methodology to account for fluctuating stresses in the shell, which are caused
due to dynamic response of cooling tower. Such dynamic behaviour is taken into account
indirectly by determining the wind load enhancement factor (ijG). This factor is used to
enhance the mean hourly wind stress resultants. Based on wind tunnel studies, it is suggested
that tower design stress resultant for isolated tower can be obtained from by multiplying a
gust factor of 2 to analytical stress resultant. As per wind tunnel studies, a gust factor of 2 is
suggested to be applied uniformly throughout the height of the tower whereas based on the
procedure given in BS: 4485, gust factor is varying throughout the height of the tower. From
Fig.3, it appears that the gust factor as per BS: 4485 is around 2.2 for height in between 15 m
to 130 m when the coefficient CD and CG related to dynamic correction factor and group
correction factor is equal to 1.There are local regions at the bottom and top of the cooling
tower, where gust factor as per BS code is comparatively high due to low mean stress
resultant. As per IS: 875 (Part 3), gust factor is coming out be 1.45.

Fig. 2 Harmonic Component of Pressure coefficient Fig. 3 Gust factor variation as per IS 875, BS4485
and wind tunnel studies

Fig. 4 Gust factor variation as per BS4485 and Fig. 5 Wind Tunnel Test for interference studies[Selvi
wind tunnel studies Rajan, et al. (2010 a)]

Interference Effect
In realistic situation, the tower is part of a dense arrangement of large surrounding
power plant buildings some of them are comparable about half the size with the cooling
tower. At kakrapara NPP project, it is proposed to construct two tall cooling towers for each
700 MWe unit and therefore total four for the twin unit plant; hence it is essential to consider

356
 Proc. of the 8th Asia-Pacific Conference on Wind Engineering (APCWE-VIII)

group effect in the design. The field of the flow-induced pressures changes more or less
considerably according to the specific arrangement, with regard both to the averaged static
and the buffeting forces. It has become usual to term this influence as interference effect. The
available cooling tower codes contain simplified rules to account for interference. A rule of
distance is given in most of the codes if the distance to adjacent building or tower is smaller
than some specified minimum value. The IS code recommends allowance to be provided for
load intensification due to turbulence induced in the incident wind by adjacent cooling towers
in a group or of the structures of the significant dimension in the vicinity but value to be
adopted are not specified. French code [H. J Niemann et al., H.D.Kopper et al. (1998)] include
magnification factor to be applied to the wind pressure. For a centre to center distance of three
times the tower base diameter, the factor is 1, for a distance of 2 it is 1.15.For closer spacing,
special advice is required. The British code applies interference factors to the wind induced
stresses. An interesting feature is that the factors are different over the shell height and
individual values are given for the quasi-static and the resonant contribution to the stresses.
The factors are specified for the certain building arrangement considering the effect of the
distance. As per BS: 4485 classification for correction to consider effect of adjacent structure,
cooling towers of kakrapara NPP falls under class 3. From Fig.4, it appears that the wind
amplification factor taking into account correction for adjacent structures as per BS: 4485 is
around 2.8 for height in between 15 m to 130 m.
To assess the group specific interference effect wind tunnel test were conducted to
investigate the effect of interference, for different angles of wind incidence ranging from 00 to
3600 in intervals of 300.The testing arrangement of the model with other interfering structures
for a typical wind angle is shown in Fig.5.The analysis clearly indicates that the effect of
interference manifest itself in discrete changes in the values of pressure at certain azimuth
angles corresponding to a given angle of wind incidence. This, in effect leads to a change in
the magnitude of force coefficient at various level. In the wind tunnel studies, interference
factor is defined as the ratio of the measured peak resultant force or mean resultant force on
the test cooling tower model under interference condition to the corresponding peak resultant
force or mean resultant force on the same tower under isolated condition. Precisely speaking,
the internal towers were found to have more effect of interference than the outer tower which
appears to be logical. Based on the wind tunnel studies for kakrapara NPP site interference
factor of 1.5 is judiciously evolved. Hence gust factor as per wind tunnel studies after
considering the effect of interference factor is coming out to be 3 (wind amplification factor
= 2 (gust factor) × 1.5 (interference factor)) compared to 2.8 obtained from the BS: 4485.

Structural Analysis and design

General
On the basis of thermal duty condition, thermal demand, tower characteristics,
following pre-designed overall dimension are considered.
Height of shell (hs) : 154.15 m
Base diameter of shell (2rL) : 116 m
Throat diameter of shell (2rT) : 70 m
Column height : 12m (10m clear Air opening height=hc)
Following possible variables are used to obtain the optimum shape of the cooling tower.
1. Height of throat (hT)
2. Column configuration
3. Angle made by hyperbola at the base and top (ȕL)
4. Thickness profile
The opening height of 10 m is selected considering the friction loss and consequent
increase in tower height and it is basically a thermal design parameter. It has been observed

357
 Proc. of the 8th Asia-Pacific Conference on Wind Engineering (APCWE-VIII)

that with the increase in the height of throat the fundamental frequency of cooling tower
reduces; hence calculation has been done for minimum required throat height. In the present
study, 44 numbers of “V” columns are provided. Different models based on angle made by
bottom hyperbola are considered, typically NDCT_14, NDCT_16, NDCT_18, NDCT_20 and
NDCT_21.76. The numeric nomenclature of the model name gives angle made by hyperbola
at the base. Thickness profile along the shell height is finalized based on the buckling criteria
as shown in Fig.8. Based on the earlier studies, Model NDCT_18 is found out to be most
optimum shape for a given preliminary dimension on the basis of structural safety and
economy and the same is adopted for the present study.
As discussed in the previous section, a wind tunnel study has been carried out for
KAPP site to determine the pressure distribution along the height and around the
circumference on the cooling tower. In this, aero-elastic model study is also carried out to
determine the dynamic behavior of the tower in the presence of wind. Based on the
experimental studies three well defined procedure for design are outlined which are as
discussed below:
Procedure I: Determine the velocity profile corresponding to Table 33 of IS 875 Part III
considering basic wind speed for kakrapara site at 10 m height as 44 m/sec with probability
factor of 1.07 and pressure coefficients as given in Table 1. It is recommended to use gust
factor of 2 and interference factor of 1.5 to be multiplied to analytical stress resultant
corresponding to 0.6 Vz2 × Cp.
Procedure II: In this procedure, it is recommended to use velocity profile corresponding to
category II class A structure with probability factor of 1.07 and pressure coefficient as given
in Table 1. As the velocity profile is determined based on the peak gust velocity averaged
over short time interval of 3 sec, no additional enhancement to gust factor is required. In this
procedure, interference factor of 1.5 is recommend to obtain design wind load stress resultant.
Procedure III: This procedure recommends design of tower based on the experimental design
stress resultant. Design based on this procedure is not discussed in this paper.

Finite Element Model

Finite Element Model of tall NDCT is shown in Fig. 6. Shell behavior is modeled with
the help of four nodded plate elements while 3 dimensional beam elements are used for
modeling the circular diagonal supporting columns in STAAD-Pro software. At the top of
shell “U” shape ring beam is provided to stiffen the shell. The orientation of diagonal columns
are calculated such that at the junction of shell and diagonal column the slope of shell and
column remain same to avoid undue moment induced due to kink at the junction. Bottom
support of column is considered to be fixed on the top of annular raft.

Design loads
Self weight and wind pressure are the primary loadings considered for designing the
cooling tower. The density of Reinforced Concrete Shell is assumed to be 25kN/m3 for the
purpose of calculation of self weight. Basic wind velocity is considered as 44 m/sec at 10 m
height and mean hourly wind velocity at 10m height above the ground is calculated as per
table 33 of IS 875-Part 3.
Vm10 =1.07 x 0.67 x 1 x 44=31.54m/sec
The mean wind velocity (Vmz) and gust wind velocity (Vgz) at height z are calculated
as per BS 6399 (1997). The wind pressure is calculated as p=0.613* Cpș *(Vmz)2 Where, Cpș is
the pressure coefficient distribution around the shell circumference. The velocity profile for
mean wind and gust wind speed as per BS: 4485 are shown in Fig.7. Fig.7 also shows velocity
profile as per procedure I and procedure II of wind tunnel studies. It appears from Fig.7 that
velocity profile of mean wind almost matches with the velocity profile corresponding

358
 Proc. of the 8th Asia-Pacific Conference on Wind Engineering (APCWE-VIII)

procedure I. Velocity profile of gust wind does not matches with the gust velocity profile of
BS-4485 corresponding to procedure II.

Fig. 6 Finite Element Model of Tall NDCT Fig. 7 Velocity Profile for Mean and Gust Wind Speed

Buckling Resistance
Stress state of the cooling tower shell is non-axis symmetric under wind loading, [refer
Fig.1] therefore, the combined stress state under dead load and wind load is also non- axis
symmetric with biaxial stress state changing along the height and circumference. Hence it is
required to vary wall thickness over the height, taking the buckling resistance into account to
developing meridian shape of the tower. The final buckling configuration depends mainly on
the restraints of cooling tower shell and on the distribution of wind loading along the
circumference. To derive optimum shape for safety and economy, it is prerequisite that the
resistance of cooling tower against buckling shall be provided uniformly over the height of
the tower. With the increasing size of the tower, the natural frequency drops more rapidly,
therefore, it has become necessary to investigate method to improve the buckling stability and
vibration properties of cooling tower.

Fig. 8 Thickness Profile along the height of Fig. 9 Local buckling pattern as per BS code and wind tunnel
tower and details of U ring beam at the top studies
Design codes differs considerably in criteria for buckling safety, requiring either the
“snap through approach”, the local or “buckling stress states” (BSS) approach, or global
approach. The snap through criteria was first proposed by Der and Fidler at al. in 1968. Based
on result of wind tunnel tests, this approach prescribes design for a critical wind buckling
pressure, which is a function of shell geometry and material strength. The Der and Fidler
formula is incorporated in the British, Indian and German codes. The BSS approach
developed by Mungan et al. [2005] is based on the hypothesis that buckling starts locally and
is primarily dependant on the strength of shell material as well as the presence of geometric
imperfections. It is also incorporated in German and British code. On the other hand, the
global buckling criterion requires a full nonlinear analysis of the shell, and is the preferred
approach in USA.

359
 Proc. of the 8th Asia-Pacific Conference on Wind Engineering (APCWE-VIII)

Buckling safety of cooling tower is investigated by local and snap through buckling
criteria. In the present study, for the given basic tower dimension, 236 mm (considering
modulus of elasticity equal to 31.62×1006 KN/mm2) is the minimum thickness to obtain factor
of safety against snap through buckling equal to 5. Local buckling pattern as per BS
methodology for BS 4485 and wind tunnel based wind load is shown in Fig. 9 wherein
thickness of 275 mm is required for buckling safety factor equal to 5.

Variation of Design Stress Resultant

The dead load analysis results in Fig. 10 indicate that the shell is under compression in
both directions, except for a small circumferential tension near the top. This is a very
desirable feature of this geometrical form. To compare the results based on the Wind tunnel
studies and as per BS-4485, design stress resultant are obtained considering the effect of wind
amplification and interference effects.

Fig.10 Circumferential and meridional forces under Fig.11 Design Circumferential forces along the height
dead weight of the tower at windward direction
In Fig. 11 through Fig. 17, the results of an analysis for quasi-static wind load are
shown for circumferential and meridional direction. It appears that that large tensions in both
the meridional and circumferential direction are present for both the case i.e. as per codal
approach & wind tunnel studies. In Fig.12 through Fig. 13, results of variation of
circumferential forces are shown around the circumference of the cooling tower at height
equal to 30m and near to the top of the shell. In Fig. 15 through Fig. 17, results of variation of
meridional forces are shown around the circumference of the cooling tower at near to the
bottom & top of shell and at throat level. The regions of tension extent a considerable distance
along the circumferential from the windward meridian and the magnitude of the forces is
strongly dependent on the distribution selected.

Fig.12 Design Circumferential forces around the Fig.13 Design Circumferential forces around the
circumference (at height 30) circumference (near to the top of shell)

360
 Proc. of the 8th Asia-Pacific Conference on Wind Engineering (APCWE-VIII)

It is found that for region of tension near the windward direction & compression near
the side suction portion corresponding to wind tunnel studies is extended to a considerable
distance compared to BS4485.It is also found that the design stress resultant corresponding to
design process based on the wind tunnel studies i.e. Procedure I & II are more compared to
design stress resultant obtained based on the codal procedure of BS: 4485.

Fig.14 Design meridional forces along the height of Fig.15 Design Meridional forces around the
the tower at windward direction circumference (at height 30)
Table 3: Comparison of Global Reaction Forces and Moments
Fx (KN) Fy (KN) Fz (KN)

WT WT WT WT WT WT
Description
BS:4485 Studies Studies BS:4485 Studies Studies BS:4485 Studies Studies
( I) (II) (I) (II) ( I) (II)

Summation 14810 20371 19009 -11 -1602 -1573 0 0 0

Maximum
2224 2767 2636 5854 7128 6783 2257 2749 2626
Tension
Maximum
-733 -1089 -1042 -5988 -7282 -6908 -2257 -2749 -2626
Compression

Mx (KN-m) My (KN-m) Mz (KN-m)

WT WT WT WT WT WT
Description
BS:4485 Studies Studies BS:4485 Studies Studies BS:4485 Studies Studies
( I) (II) (I) (II) ( I) (II)
Summation 0 0 0 0 0 0 -9749 -13591 -13026
Maximum
525 699 730 187 257 265 223 308 294
Tension
Maximum
-525 -699 -730 -187 -257 -265 -692 -916 -955
Compression
Note: For global direction refer to Fig.6

Fig.16 Design Meridional forces around the Fig.17 Design Circumferential forces around the
circumference (at throat level) circumference (near to the top of shell)

361
 Proc. of the 8th Asia-Pacific Conference on Wind Engineering (APCWE-VIII)

Table 4 Loads Combination for Shell Support Structure

Limit state of strength
1.0 DL + 1.5 ijF Wk
1.4 DL + 1.5 ijF Wk
Where, DL-Characteristic Dead Load, Wk-Hourly Mean Wind Load
ĭF – Load Amplification factor evaluated at a height above the base shell equal
to the distance of between adjacent shell support nodes
Note: Description Amplification Factor
1. As per BS:4485 ijF =2.75
2. WT (Procedure I) ijF =g × IF = 2 × 1.5 = 3.0
3. WT (Procedure II) ijF =IF = 1.5 = 1.5

Table 3 gives the comparison of the support global reaction forces and moments for both the
cases i.e. as per BS 4485 and procedure I & II of wind tunnel studies. It is found out that there
is noticeable change in the reaction forces and moments as per wind tunnel devised procedure
compared to codal approach. With this view, design of columns has been carried out
considering the load combination given in the Table 4. It is found that maximum
reinforcement requirement has been increased from 2.78 % as per BS:4485 to 3.82 % as per
wind tunnel studies.

Conclusion
From the present study following conclusions can be made,

• In general, codal methods are helpful for preliminary design and for simple situation, but
provide conservative wind loads in most cases, but it is not always true. In fact, the code
recommends wind tunnel testing for complex problems. If the tower forms the part of the
tower group and its height is greater than 120,code recommends aero-elastic model testing
in an atmospheric boundary layer wind tunnel to determine the stresses in the shell.
• In the present case, it is found that codal analytical provisions gives lesser wind load
effects on tall cooling tower compared to wind tunnel studies. It is found that regions of
tension extent a considerable distance along the circumferential from the windward
meridian and the magnitude of the forces is strongly dependent on the distribution selected.
• There is difference in the support reaction and moment corresponding to wind tunnel
testing compared to codal provision. This may call for change in the size of the column or
increase in the reinforcement.
• Wind tunnel tests account for structure geometry ,surrounding details and this leads to
accurate estimation of wind load. In addition to accurate results for wind amplification,
interference etc. wind tunnel studies confirm the structural safety, economy and durability.
• Need is felt to revise the structural design based on IS:11504 so that the latest development
around the globe can be adequately represented in the code.

References
BS 4485 (1996), “British standard for water cooling towers, Part 4, Code of practice for structural design and
construction” ,British Standard Institution.
BS 6399 (1997), “British standard for Loading for Building, Part 2, Code of practice for wind loads”, British
Standard Institution.

Cermak Jack E., (2003), “Wind-tunnel development and trend in application to civil engineering,” Journal of
Wind Engineering and Industrial Aerodynamics, 91, 355-370, 2003.

362
 Proc. of the 8th Asia-Pacific Conference on Wind Engineering (APCWE-VIII)

Der T.J., Fidler R., (1968), “Model study of the buckling behavior of hyperbolic shells”, Proceeding of Institute
of civil Engineering, 41:105-18,
Dieter Busch, Reinhard Harte, Wilfried Kratzig, Ulrich Montag (2002), “ New natural draft cooling tower of 200
m of height,” Engineering Structures, 24, 1509-1521, 2002.
Dieter Busch, Reinhard Harte, Hans- Jurgen Niemann (1998), “ Study of a proposed 200 m natural draught
cooling tower at power plant Frimmersdorf/Germany,” Engineering Stuructures, Vol. 20, pp. 920- 927, 1998.
Girish Patil, Nema, M.K., Arvind Shrivastava, Verma U. S. P. (2007) , "Assessment of dynamic wind effect and
optimization of shell for tall NDCT using BS-4485 code approach," Proceedings of the Fourth National
Conference on Wind Engineering , SERC Chennai, India, October 30th to November 1st, 2007.

H. J Niemann, H.D.Kopper (1998), “Influence of adjacent building on wind effects on cooling towers,”
Engineering structures, Vol. 20, pp. 874-880, 1998.
IS 875 (1987), “Indian standard code of practice for design loads (other than earthquake) for building and
structures, Part 3, Wind loads”, Indian Standard Institution.
IS 11504 (1985), “Indian standard-Criteria for structural design of reinforcement concrete natural draught
cooling towers”, Indian Standard Institution.
Mohan Babu, Khasim Khan, Neme M.K, N.Raghavan (2004), “A comparative study of wind dynamic effects on
NDCTs as per various international standards”, International Symposium on Natural Draught Cooling Towers,
2004.

Mungan I., (2005), “Buckling stress state of hyperboloidal shells”, ASCE Journal of structural Engineering, 102
(ST10), 2005.
Selvi Rajan, S., Arunachalam, S., Ramesh Babu, G., Harikrishna, P., Abraham, A., Chitra Ganapathi, S., Sankar,
S. and Nagesh R. Iyer, (2010), “Wind tunnel experimental investigations on natural draught cooling towers
including interference effects – Pressure measurements”, Project No. S-1, SSP 075 41, March 2010.
Selvi Rajan, S., Arunachalam, S., Ramesh Babu, G., Harikrishna, P., Abraham, A., Chitra Ganapathi, S., Sankar,
S. and Nagesh R. Iyer, (2010), “Aero-elastic test on model of a natural draught cooling tower using boundary
layer wind tunnel”, Project No. S-2, SSP 075 41, June 2010.

363