Professional Documents
Culture Documents
In the
FACULTY OF ENGINEERING & TECHNOLOGY
TO
GANPAT UNIVERSITY, KHERVA
February 2011
Research Guide:
Dr. H. S. Patel
Research Guide
Ganpat University
Kherva-382711, Mehsana (N.G.), India.
Forwarded through:
Prof. S. M. Patel
Dean
Faculty of Engineering & Technology
I hereby declare that the topic entitled “COMPUTER AIDED
ANALYSIS & DESIGN OF STRUCTURES IN TSUNAMI PRONE AREA”
which is submitted herewith to Ganpat University, Kherva, for the
award of Doctor of Philosophy in the Faculty of Engineering &
Technology is the result of work done by me in Ganpat University,
Kherva under the guidance of Dr. H. S. Patel, Asst. Professor, L. D.
College of Engineering, Research Guide of Ganpat University, Kherva.
I further declare that the results of this work have not been
previously submitted for any degree or fellowship.
I hereby seize the opportunity to express my sincere indebtedness and deep sense of
gratitude to thank Dr. H. S. Patel, Associate Professor, Applied Mechanics Department,
L. D. College of Engineering, Ahmedabad, who took care of each and every facet of this
work in congregation with personal attention and proper orientation. I am grateful to him
for excellently tutoring and grooming me with respect to professional ethics and real life
normal values in order to face squarely this swift changing competitive world with
compound nerve and verve. His approach and insight to the problem has helped me in
presenting this thesis in its present form.
I am grateful to Dr. B. K. Rastogi (D.G., ISR, G‘Nagar) for giving permission to use ISR
library, simulation model & other resource materials. I also thank scientists and librarian
of ISR, G’Nagar for their kind support for our research. My special thanks are in
particular to Dr. A. P. Singh who was always available to answer my questions and to
provide me with the necessary information concerning the research work and also for his
critical reflections on the numerical modeling.
My sincere thanks go to The Director, Mehsana District Education Foundation and The
Vice Chancellor, Ganpat University, Mehsana for his kind permission to carry out this
study.
I would like to thank Dr. H. S. Patil and Dr. B. B. Mistry for their valuable suggestions to
enhance the present work.
Sincere thanks for the friendship, support and professionalism of members of the BSPP
& UVPCE namely – Vijendra, Maitry, Bhavsar, R J Patel, D N Sheth, H T Patel, Vitthal,
Qureshi, V P Patel, Chhatani, Sitaram, Jigar & Bhaskar. Thank you also to the always-
cheerful staff of the Department of Civil & Computer.
I would like to thank V N Modi, V K Patel & Ramanuj who eagerly shared their wealth
of knowledge and wisdom for Neural Network application. Lastly I can not forget to
thank my esteemed mentors and many people who provided assistance for my research
visits across India, in particular Dr. S.K.Jain, Dr. Pradeep Kumar & Dr. Dholakia.
The love and support of my Wife, Parents and Family made this journey both possible
and worthwhile.
Vikram M. Patel
Ganpat University
The Indian Ocean tsunami on December 26, 2004 resulted into massive destruction to
coastal communities with more than 3,00,000 fatalities, and severe damage to buildings,
bridges and other infrastructure causing serious socio-economic problems. Such deadliest
mega event made the coastal community realize the need for the preparedness against
initial ground shaking and subsequent effects followed by tsunami. The casualty can be
minimized with a proper Early Warning System and a network of solid refuge buildings
and roads. As most of the tsunamis are earthquake induced tsunamis, it is necessary that
coastal structures should be designed against both earthquake and tsunami loads.
The experience of the recent Indian Ocean Tsunami was new to the people of India. As
the western coast of Gujarat is also sensitive to tsunami attack, it is essential to take the
tsunami risk into consideration. This study will focus on the suitability of the tsunami
resistant structures for the Dwarka and western coast of Gujarat. Although this study is
aims at western coast of Gujarat, it can also be applied in a broader perspective.
Aim of this research is to help minimize vulnerability of Dwarka and western coast of
Gujarat against the devastating effect of earthquake induced tsunami and enable
sustainable and long-term economical development in these regions. For Dwarka and
western coast of Gujarat the feasibility of tsunami protection alternatives are reviewed in
this study.
The study consists of two parts: 1) the derivation of a design of tsunami wave height &
probable arrival time; and 2) the study of behavior of buildings under tsunami loads along
with other loadings and special design considerations for tsunami resistant structures.
Eastern and Western parts of the Makran Subduction Zone (MSZ) of southern Pakistan
are potential zones for great earthquakes that can generate tsunamis affecting west coast
of Gujarat. The eastern part of the MSZ has produced the 1945 Mw 8.1 earthquake that
generated the last major tsunami in the Arabian Sea. Some sectors of the MSZ are un-
ruptured for a long time and can produce large earthquakes in near future. In this study
investigation has been carried out for the far-field amplitude of tsunamis in the Arabian
Sea of mega-thrust earthquakes, ranging from probable to possible.
In this study available design codal provisions were reviewed to choose a comprehensive
set of loading equations that may be used to estimate the loads imposed by tsunami flow
on structural components. Historical tsunami events were reviewed with the intention of
evaluating experimental structural response.
In this research seven typical R.C. buildings were chosen for the analysis considering
different types of structural configurations and a comparative study is made on the
ACKNOWLEDGEMENT ........................................................................................ ii
ABSTRACT ........................................................................................................... iii
TABLE OF CONTENTS ........................................................................................ v
LIST OF TABLES ................................................................................................ xii
LIST OF FIGURES .............................................................................................. xiv
2.4 LANDSLIDES.............................................................................................. 11
2.4.1 Volcanic Eruptions................................................................................. 12
INTRODUCTION
The western coast of India and Vicinity has experienced six tsunami events since
326 B.C. Distant tsunamis have originated from Makran Subduction Zone
(MSZ) and Iranian Coast. Although, the historical records are incomplete, it is
believed that such Tsunamis were destructive on the western coasts of India. The
most destructive tsunami ever recorded in the Arabian Sea was by the 28th
November 1945 earthquake in Makran region. More than 4000 people lost their
life along the Makran coast of Pakistan by both the earthquake and the tsunami.
The tsunami was responsible for great loss of life and destruction along the
coasts of India, Pakistan and Iran. The earthquake’s Richter Magnitude (Ms) was
7.8 (Pendse, 1948) & the Moment Magnitude (Mw) was revaluated to be 8.1
(Byrne et al, 1992). The run-up height during the tsunami was of 13m at Makran
coast 1.5 m at Karachi and 2 m near Bombay (Jaiswal et al., 2008).
Seismic gap area along the subduction zone of Makran is possible site of future
great earthquake, which could generate tsunamigenic condition along western
Indian coast (George, 2006 & Jaiswal et. al., 2008). Apart from known Eastern
Makran Zone the recent discovery shows that Western Makran Zone is also
capable of producing major far field tsunami (Fig 1.1). A review of various
tsunami catalogs (Rastogi and Jaiswal, 2006; Jordan, 2008; Murty, 1977) reveals
that the western coast of Gujarat is at risk from a locally generated tsunami too.
Evaluate design loads for tsunami resistant structures based on run up height of
waves, arrival time and inundation depth.
Provide insight into the feasibility of tsunami resistant structures for Dwarka and
western coast of Gujarat and to set up a design procedure for tsunami resistant
structures.
This study focuses on the Western Coast of Gujarat (Fig 1.2). On western side of
this region there is Arabian Sea, where large tsunami capable of reaching up to
Gujarat was generated in the past.
From these activities, in combination with results from hazard mapping and a
risk assessment, it was concluded that for the Western coast of Gujarat, where
there is one of the highest possibilities of loss of human life and economic value
in future, a further study into the feasibility of a tsunami resistant structure is
required. Based on these initial findings, the focus has been set to Dwarka and
western coast of Gujarat. However, many questions remain unanswered.
Tsunami resistant structure is unknown in this area, probably because no tsunami
has been reported since 1945. This yields questions concerning the frequency
and heights of future tsunamis. Finally a question is raised whether tsunami
protection is required.
With reference to the above question, does the matter concern with the actual
load on a structure? Is it anyhow possible to protect against extreme tsunami
events and preceding earthquakes? To what extend is tsunami impact influenced
by structure type, structure location, bathymetry, etc.
This section provides an overview of the content of each chapter. The thesis is
structured into four core components over seven chapters comprising (refer to
Table 1.1): Introduction; Background; Modeling, Results and Analysis; and
Synthesis.
Various numerical models have been used in this study. To study the general
behavior of tsunamis for varying bathymetry, shape and height, simple
numerical models are used. These models are also used to get a general idea
about the effectiveness of several earthquake parameters (and combinations) in
tsunami waves. A description of a Numerical model is also presented.
Application of Neural Network is discussed in the field of tsunami warning
system (Chapter 3).
Past tsunami effects on various small scales coastal structures are investigated
for comparison between their structural performance and the different structural
parameters used. The available literature on tsunami load equations for tsunami
resistant structures from various building codes and published literature are
reviewed to develop one set of equations to calculate the tsunami forces (Chapter
4).
Structural behavior of selected buildings under earthquake load case, wind load
case and tsunami load case are studied and compared in this section. In the case
of tsunami, the buildings are analyzed subjected to tsunami flow at the depths of
2.464, 3, 5 and 10 meters. The resulting forces in structural elements are
compared with the seismic design forces. Also, the effects of different shapes of
building are studied under the tsunami loads (Chapter 6).
The term tsunami is derived from two Japanese words: ‘tsu’ , means harbor,
and ‘nami’, means wave. That is because these waves may create large surges
or oscillations in bays or harbors, which are not responsive to the action of normal
sea waves. In deep water, a tsunami is hardly noticeable, but near the coast
various mechanisms cause the wave to grow with sometimes devastating effects.
The term tsunami was created by fishermen who returned to their ports to
find the surrounding area devastated, although they had not been aware of any
wave in the open water.
Tsunamis are caused by rapid perturbations of the seabed or of the water column
above it, which either lift the sea surface up above its normal level (the usual
case) or depress it. This perturbation produces a series of waves, or wave train,
which propagates outwards from the source area until it either, dissipates or collides
with a coastline. The physics of this propagation process is considered later.
The earth’s crust can generally be divided into fifteen major rigid plates. These plates
move relative to each other causing them to collide at certain locations and drift apart
elsewhere. These boundaries are called fault lines. The contact between these massive
plates does not run smooth, resulting in a buildup of stresses along the fault. When the
stress exceeds the resistance of the rocks, an earthquake occurs that (partially)
releases the built-up stresses in a certain time span. This process can result in
displacements of the earth’s surface up to several meters and subsequent displacement
of the overlying water mass. This will sometimes result in a series of waves, called
tsunamis. When this happens, this earthquake is called tsunamigenic.
Three main fault types can be distinguished: a normal fault, thrust (reverse) fault and
a strike slip fault. Normal fault and thrust fault are also called dip-slip fault. A dip-slip
earthquake can be on a vertical or a dipping plane. Views of different fault types are
shown in Fig 2.2.
In each case rupturing can occur at any point along the fault line. This point is known
as the focal depth of the epicenter. The hypocenter (focus) is the center of the
The orientation of the rupture plane depends on two angles. The direction of the
fault plane relative to the Earth's local surface is defined by the dip angle δ.
The strike angle φ is the clockwise angle between the geographical north and the
strike. The classification of an earthquake is determined by the slip angle, which
indicates the direction in which the upper block moves with respect to the lower
block. The main types are the dip-slip (λ= ±90˚) and strike-slip (λ=0˚, 180˚). The
magnitude of an earthquake is associated with the amount of the movement along the
plane, the so-called slip.
While the vertical dip-slip (with δ ≈ 90˚) mechanism seems to be a logical one for
tsunami generation because it abruptly displaces large sections of the seafloor
vertically, the area of uplift cancels out the area of subsidence, resulting in small or
non-existent tsunami. An impression of a dip-slip fault is presented in Fig 2.2. But
this mechanism is still better at producing tsunami than the strike-slip pattern. The
thrust dip-slip fault is the most preferred fault mechanism for tsunami generation.
Shallow subduction zones earthquakes are one of the most common sources of
destructive tsunamis in the world.
Normal
fault
Strike-slip fault
Fig 2.2: Dip-slip and strike-slip faults
Subduction zones typically have average dip values of δ ≈ 25˚ ± 9˚, with the largest
tsunami associated with higher dip values. As the dip-angle (δ) decreases, the
tsunami is more likely to have a leading trough.
As can be seen in Fig 2.3, the fault-type determines the characteristics of the wave. If
a coast-line is located on the sub-ducted plate a negative wave is expected, means an
initial drop-down of the water level. The coastline located on the sub-ducting plate (at
the left) will receive a positive tsunami. Negative tsunamis are generally more
powerful then positive tsunamis. Tsunami wave heights are highly variable. Many
parameters determine the final wave height at the coast. Besides that, earthquakes can
go together with landslides.
2.4 LANDSLIDES
Two mechanisms can be distinguished, submarine and sub-aerial landslides. See Fig
2.4. As explained in the figure, sub-aerial landslides are more effective in generating
waves, since it yields a net addition of volume to the sea floor. Wave generation by
landslides depends primarily upon the volume of the material moved and submerged,
the speed of the landslide and the mechanism of movement.
Volcanic eruptions can generate tsunami in many different ways. The majority of
eruptions are accompanied by seismic tremors, which can trigger tsunami if they are
big enough and lie in or near the ocean. Volcanic activity can also induce submarine
landslides and submarine eruptions/explosions. The latter can cause violent tsunami,
when ocean water comes in contact with the magma chamber. This water is converted
instantly into steam, causing an explosion, which can generate large ocean waves. It is
believed that the August 1883 Krakatau eruption produced a 40m tsunami by this
mechanism.
Unlike earthquakes, which cause most tsunami but have a well-defined upper limit,
the potential tsunami height caused by meteor-impact is almost unlimited. However,
most objects smaller than 100 - 200 m in diameter explode in the atmosphere and will
not produce significant waves.
Although a distinction has been made among different causes for tsunami, it is most
likely that a lot of tsunamis are caused by a combination of these mechanisms.
In Table 2.1 the percentage distribution of events and deaths are presented for
tsunami-events over the past 2000 year. Also joint occurrences are listed. From this
table it becomes clear that earthquake-induced tsunami form the biggest threat as they
are responsible for over 80% of all tsunami events and deaths.
Note: The listed events are only most-probable and definite tsunami events (Source:
Tsunami Event Database of the National Geophysical Data Centre)
2.5.1 Terminology
The terminology used to describe tsunami waves is shown schematically in Fig 2.5.
Much of this terminology is similar to the terminology used for wind waves.
However, there are some differences. The run-up height is referred to a reference
level, mostly mean sea level and not to the water level at the actual moment the
tsunami occurred. This is because of the run-up height is mostly measured some days
after the tsunami occurred. So one should be aware that the actual run-up height of the
tsunami is the reported run-up height Hr corrected with the tidal component. The
period of the wave is the time it takes for two successive peaks to pass a fixed point.
2.5.2 Propagation
The earthquake rupture triggers a series of fast-moving, long waves of initial low
amplitude that radiate outward in a manner resembling the waves radiating when a
pebble is dropped in a pond (Fig 2.6). Part of the tsunami travels into the deep ocean
(distant tsunami) and a part travels to the nearby coast (local and regional tsunami).
Most tsunamis generated by large earthquakes travel as wave trains. These wave
trains contain several long waves with wave lengths that often exceed 200 km in the
deep ocean. Usually one of the waves is more pronounced than the other. In deep
water even the highest wave seldom exceeds 0.5m. Their steepness is so small that a
ship out at sea does not feel a tsunami pass.
The maximum ocean depth lies between 8 to 10 km and tsunamis in the deep ocean
have typical wave lengths of hundreds of kilometers. Thus the wave length exceeds
many times the water depth (L>>d), therefore tsunami waves travel as shallow water
waves. The propagation speed of shallow water waves is solely a function of the
water depth:
= ( . ℎ) Eqn (2.1)
The wave length of a tsunami can be expressed as a simple function of the wave speed
(c) and the period of the wave (T), for a constant depth:
= . Eqn (2.2)
L = wave length (m); v = wave speed (m/s); T = time period of the wave (s)
Tsunamis typically have periods o f 100 - 2,000 seconds (1.6-33 min), referred to as
the tsunami window.
Fig 2.7: Shoaling process of a tsunami wave, clearly showing the steepening of
the wave
When water depths become shallower, the tsunami waves slows down Eqn (2.1) and
becomes compressed Eqn (2.2) causing them to grow in height (Fig 2.7). This is
called shoaling of the wave and it is a reversible process as long as the wave does not
break. This makes that the changes in the depth and in the seafloor cause the tsunami
waves to continuously evolve and change in shape as they propagate.
On approaching land the wave height will have increased dramatically, whereas the
wave length will have significantly decreased. Despite of the steepening of the
tsunami waves, most waves hit the coast as a fast rising or falling tide (Fig 2.8).
Occasionally when the wave steepness does cause the tsunami waves to break, they
hit the coast as a bore (Fig 2.9). It happens very rarely that tsunami waves break on
the coast.
When the tsunami acts as a rapidly rising tide, the resulting incident is that the current
velocities are relatively low and most initial damage will result from buoyant and
Tsunamis in the open ocean are approximately sinusoidal in shape (Fig 2.10). At this
stage the waves can be described by the linear wave theory. As they approach the
coast, the waves become more peaked: the wave peak sharpens and the wave through
becomes more flattened. Mathematically the wave in this stage can be described by
the 2nd order Stokes wave theory. As the tsunami waves approach the coast the wave
through disappears and only the wave peak remains. The wave train evolves to a
succession of solitary waves.
Until this was the common vision among scientists; most experiments and
calculations assumed that tsunamis at the shore manifested themselves as solitary
waves. However, in the nineties scientists began to doubt this approach. First of all
because almost every big tsunami investigated, was reported by people having a
dipole form: a large through accompanied the tsunami arrival. This kind of wave
One of the reasons it took so long before scientists started to accept the N-wave form,
is that the solitary wave was somewhat easier in describing mathematically and also
easier in generating during experiments. One must realize that before mid-eighties
computational modeling and the computational control of experiments was not as
obvious as it is now.
Nevertheless several aspects of solitary waves and N-waves correspond quite well;
the solitary wave and the N-wave both preserve their wave form, and the propagation
is comparable. But N-waves can become steeper than solitary waves before they
break. This makes that the older work based on solitary waves can still contain a lot of
valuable information.
Some older work can give quantitative insight in the behavior of tsunami waves and
can be used for a preliminary approach to calculate refraction (based on the linear
wave theory), behavior of tsunamis at abrupt depth transitions, wave reflections,
resonance phenomena and many more tsunami wave related phenomena.
2.5.4 Breaking
As stated above tsunami waves do rarely break, because of their very long wave
lengths and modest wave height. A breaking criterion is given by Bryant (2001):
²
= Eqn (2.3)
( ² )
Br: Breaker Parameter (Br>1, wave breaks); H: Wave Height (m); : Slope of the
coast (rad); ω: Radiation frequency of the waves (1/s)
However it is not stated, which wave form is considered. Until the solitary wave was
regarded as the tsunami wave form in coastal waters, it is most likely that the breaker
criterion is derived for solitary waves. At this moment the form of tsunami waves are
often regarded as N-waves, which can become steeper than solitary waves before they
break. So, conveniently, the presented breaker criterion Eqn (2.3) can be regarded as a
lower limit for breaking of tsunami waves, as N-waves break later than solitary
waves.
There are two distinct measures for describing tsunamis: intensity and magnitude.
The intensity gives an indication of the strength of the tsunami at a given location and
its magnitude is an indication for the total energy of a tsunami. The most common
tsunami scales, based on work of Hatori (1995) and Bryant (2001) listed below are
used to quantify tsunami events.
Tsunamis are not as common in the Indian Ocean as in the Pacific. As compared to
average eight tsunamis per year in the Pacific, Indian Ocean has one in three years or
so. Though rare, tsunamis have hit India earlier. The tsunamis in the Indian region and
vicinity are listed in Table 2.2. About eighty percent of the tsunamis of the Indian
Ocean originate in Sunda arc covering Java and Sumatra.
Table 2.2: List of Tsunamis that Affected Indian Region and Vicinity
Sr Date Location Lo Lat. Eq. Cause P Max Ref.
. ng. Ma r Run
N g. o Up
o. b (Run
Ups)
1 326 Indus delta 1 4 (Lisitzin,
B.C. 1974)
2 About Poompuhar, 79. 11.1 4 Wikipedia
500 AD Tamilnadu 52 2
(Probably due to
karkatau
cruption)
3 900AD Nagapattinam, 79. 10.4 4 (Kalaki
Tamilnadu (may 53 6 Krishnmurty)
be from Sunda –
Andaman arc)
4 1008 Iraninan Coast 60 25 1 4 (Murty et al.,
1999)
5 1524 Dabhol, 73. 17 (Bendik and
Maharashtra 2 Bilham, 1999)
6 May Samaji-Delta of 68 24 (Oldham,
1668 Indus 1883)
7 1762.04. Bay of Bengal 92 22 1 4 >2 (Mathur,
12 (Bangladesh) (1) 1998)
8 1819.06. Kutch 26. 71.9 Mw 1 3 (Macmurdo,
16 6 7.8 1821)
9 1842.11. N. Bay of 90 21.5 1 4 (3) (Oldham,
11 Bengal 1883)
10 1845.06. Kutch 23. 68.3 1 3 Nelson
The cause and probability of the tsunamis are shown by “Cause” and “Prob.”
respectively. The cause and probability of the tsunamis are given by following codes.
Cause Code:
Event Probability:
The western coast of India and Vicinity has experienced six tsunami events since 326
B.C. Distant tsunamis have originated from Makran Subduction Zone (MSZ) and
Iranian Coast (Jordan, 2008; Rastogi and Jaiswal, 2006; Rajendran et al., 2008). Local
tsunamis have originated form Kutch region and Maharashtra
The oldest record of tsunami is available from November 326 BC earthquake near the
Indus delta /Kutch region that set off massive sea waves in the Arabian Sea (Lisitzin,
1974). Tsunami has been observed in the North Indian Ocean on the Iranian coast
from a local earthquake between 1 st April and 9 th May 1008 (Murty, 1977).
An earthquake occurred during 1524 A.D. off the coast of Dabhol, Maharashtra and. a
resulting large tsunami caused considerable alarm to the Portuguese fleet that was
assembled in the area (Bendick and Bilham, 1999).
The town of Samawani (or Samaji) sunk in to the ground with 30,000 houses during
an earthquake of Samaji-Delta of Indus in May 1668.
On June 16 1819, a severe earthquake with Mw 7.8 was recorded in Kutch, India
resulting in large changes in the elevation of the land. The town of Sindri (26.6N
71.9E) and adjoining countries were inundated by a tremendous rush from the ocean,
and all submerged, the ground sinking apparently by about 5m (Macmurdo, 1821)
On June 19 1845, “The sea rolled up the Koree (Kori creek, 23.6N 68.37E) (the east)
mouth of the Indus overflowing the country as far westward as the Goongra river,
northward to the vicinity of Veyre, and eastward to the Sindree Lake in Kutch, India,”
(Nelson, 1846)
On Sunday, December 26, 2004 at 06:28:53 a.m. (IST), the earthquake that created
the tsunami is the world’s fifth largest with a magnitude of 9.3 on the Richter scale of
the west coast of Northern Sumatra, Indonesia (Borrero, 2006; EERI, 2005-1) (Fig
2.11). The epicenter located at (3.09◦N, 94.26◦ E), and aftershocks occurred along a
fault line stretching from Indonesia to the Andaman Islands in the north where the
Indo-Australian plate subducting Sunda and Burma sub-plates at a velocity of roughly
60mm/year to north east. The faulting released elastic strains accumulated for
centuries in a few minutes. The estimate of the seismic moment of the earthquake is
3.95x1029 dyne.cm (3 95 × 1022Nm). A maximum slip of 13.9 meters is estimated
from the data. This is a very large slip compared to great earthquakes. Like, the
Imperial Valley earthquake of M 7.1 showed 1.4 m maximum slip and for the Kobe
earthquake (M 7.3), the maximum slip is 2.6 m.
The aftershock activity concentrated more towards the north of the main shock and
very little or nil towards the south. The large offset in vertical component is dominant
as compared to the horizontal components. This is one of the possible reasons for
generation of a large tsunami in the Indian Ocean. The ground motion in the epicenter
zone continued for a period of about 2000 seconds (nearly 33 minutes).
For the Sumatra earthquake, currently available fault plane models differ in many
interesting details, but are consistent in implying that the fault rupture propagated to
the northwest from the epicenter and that substantial fault rupture occurred within
hundreds of kilometers northwest of the epicenter.
The fault plane model is an impulsive model, including three adjacent rectangular
fault planes covering the entire 150km × 1170km fault region (http://
www.pmel.noaa.gov /∼titov/Titov/).
Fault depth 5 km 5 km 5 km
Strike, Dip, Slip 300˚, 13˚, 90˚ 345˚, 13˚, 90˚ 365˚, 13˚, 90˚
Dislocation 15 m 15 m 15 m
Fig 2.12 shows the estimated tsunami travel time in the Indian Ocean. The tsunami
attacked the coasts of Indonesia in 75 minutes, Thailand and Sri Lanka in 2 hours, the
Maldives within 4 hours, and east Africa in 10 hours. The tsunamis were observed at
a great number of tidal stations in the Indian Ocean.
In areas where the beach is gently sloped, the tsunami waves were very powerful and
this is evident in places like Lhok Nga, Ulhue Lhe, Krueng Raya and Meulaboh. The
destruction by the tsunami in those areas was very severe (Table 2.3 & 2.4).
In areas close to the center of the city of Banda Aceh, damage by impact force could
be observed. The impact force is caused by floating of fishermen’s boats and also by
cascading debris in mud. In Lhok Nga and several villages on the west coast of
Sumatra Island near Banda Aceh, where tsunamis 15 to 30 meters high hit coastal
villages, no body survived.
The tsunami devastated the distant shores of Thailand and Malaysia to the east, Sri
Lanka and India to the west and Burma to the north. The tsunami arrived in the A&N
Islands at 7:15 a.m., and it caused extensive devastation of the built environment.
Tamil Nadu and Kerala were extensively damaged, while Andhra Pradesh sustained
moderate damage.
Survey reports states that the Nicobar Islands indicate the most severe losses: out of
the total population of 42,068, about 1,395 are reported dead, 5,764 missing, and
27,497 were in the relief camps (as of March 10, 2005) (EERI, 2005-2) (Table 2.4).
Damage of telecommunication
Iran
Pakistan
Oman
Makran
India
Srilanka
INDIAN OCEAN
Sumatra
Thrust type earthquakes along subduction zones that cause vertical movement of the
ocean floor are usually tsunamigenic (Jaiswal et al., 2008-2). Minor tsunamis can be
generated due to dip-slip faulting along oceanic ridges.
Far-field tsunami hazard in an ocean basin is a direct function of the seismic potential
for the extremely large events required to generate tsunamis capable of exporting
death and destruction to distant shores. Research indicates that a seismic moment in
excess of 7 x 10 28 dyn.cm is required in this respect (Okal and Synolakis, 2008).
The M 9.3 earthquake of the coast of Sumatra triggered the tsunami. The earthquake
occurred due to the thrusting of the Burmese plate over the Indian plate. The fracture
propagated unidirectionally from Sumatra, toward north along the plate boundary
paralleling the Andaman and Nicobar Islands, at a velocity of about 2.4 m/s for the
first 600 km and then it slowed down to about 1.5 km/s (Jaiswal et al., 2008-2). The
total length of the fracture as seen from the aftershocks distribution is about 1250 km.
The convergence of the Indian plate with the Arabian plate, with the Arabian and
Iranian microplates of the Eurasian tectonic blocks has created an active east-west
subduction zone along the Makran coast in southern Pakistan.
The Makran coast is rugged and tectonic in origin with uplifted terraces, cliffs and
headlands. The entire coastline is characterized by extreme sediment accretion
(Jaiswal et al., 2008-2; Byrne, 1992; Singh et al., 2008; Mohammad et al., 2009). It is
one of largest sediment accretionary wedges on earth, with up to 7 km of sediments
deposited in the Gulf of Oman to the west and major rivers contributing vast amount
of sediment to the offshore region in the east. The accretionary complex is more than
900 km long and there is no evidence of very active volcanism. The complex has an
east-west orientation and is bounded on both sides by large transform faults associated
with tectonic plate boundaries.
The Balochistan section of the Makran coast of Pakistan has several small river deltas.
In the eastern Sindh region of Pakistan the Indus River has formed one of the largest
deltas in the world. Past meandering of Indus has formed extensive deltas east of
Karachi. Extensive sedimentation from the erosion in the Himalayas has widened the
continental shelf of the Sindh coast to about 150 km. Along the Balochistan region
where there is less sedimentation; the continental shelf measures only 15-40 km (Fig
2.14) (George, 2006).
The MSZ extends east from the strait of Hormoz in Iran to near Karachi in Pakistan
with the length of about 900 km. Stonely (1974) first proposed that a subduction zone
off the Makran coast forms the boundary between the Arabian and Eurasian Plates.
Quittmeyer and Jacob (1979) stated that the seismicity in the Makran region is
consistent with the interpretation of this area as an active subduction zone. According
to Page et al. (1979), the raised beaches along the Makran coast confirm the tectonic
model of subduction zone along this region. Recently, a comprehensive study on the
tectonic and climatic evolution of the Makran region was conducted by Clift et al.
(2002).
Offshore, the active tectonic convergence of the India plate with the Arabian and
Iranian microplates of the Eurasian tectonic block has created a tectonic plate margin
- an active subduction zone along the boundary of the Arabian plate on the Makran
coast. The tectonic plates there converge at an estimated rate of about 30 to 50 mm/y
(George, 2006).
The recent Tsunami has made us to understand that the planning of coastal areas
involves more care than the normal land use planning since our coast is vulnerable to
natural disasters.
The Tsunami hazard map may be empirically defined using a deterministic approach,
based upon potential maximum wave heights for the scenario tsunamis. For the
terrestrial environment the hazard may be presented as inundation levels, in terms of
run-up heights at specified land contours. For the marine environment (“ON
WATER”) Harbour, Bay and Reefs – hazard may be given in terms of potential
maximum wave heights. The definition of the tsunami hazard zones, as preliminary
estimates, is given in Table 2.6.
The elevation data sets are the most important input for inundation mapping in
tsunami prone areas. Preparation of a vulnerability map could inform coastal
community and others about susceptibility to inundation corresponding to various
wave-heights.
By integrating the hazard and vulnerability assessments, the tsunami risk assessment
is to be developed in terms of zonation and inundation maps and associated affects.
In tsunami disasters, the biggest damages and losses occur in the urban areas.
Once the tsunami strikes, there will be tremendous losses and damages in the city.
Therefore, it is very important to prepare the city and its community with a disaster
mitigation plan in order to reduce or minimize potential loss or damage.
To support the mobility of evacuation and aid efforts, the planning for emergency
road network is strongly required. Emergency road network is organized into escape
road and relief road. Escape roads accommodate people in the hazard zone to escape
from disaster in a short time. Relief roads mainly function for immediate treatment
(first-aid), evacuating citizens, and supplying relief materials. The road network
ensures connection with safer area in surrounding region of tsunami hazard zone.
The following structures are required to build in coastal area in order to reduce
destructive energy of tsunami waves. Materialization of these structures should
consider local condition of tsunami-prone area in terms of geography, topography,
and financial resources.
Detached breakwaters are structures situated offshore and generally parallel to the
shore. Detached breakwaters protect the adjacent shoreline by reducing incoming
wave energy due to storm surge, mid-scale and small-scale tsunami. Sand transported
along the beach is then carried into the sheltered area behind the breakwater where it
is deposited in the lower wave energy portion.
Today, nations around the Ocean are trying to decide whether to allow rebuilding on
the coast, which structures to rebuild and which ones to relocate, and how to rebuild
to minimize losses due to future tsunamis. There are a wide range of technical and
management options for coastal protection, which include the sea wall construction
and off shore breakers as artificial barriers.
If Mangrove ecosystem or Wet lands had found in these areas the impact would have
reduced as mangrove ecosystem would have acted as buffer zones People living along
the coastal zones must be encouraged to organize afforest ration programme of
mangrove for preventing future impact of tsunami in the coastal zones. While there is
no guarantee that replanting mangroves will prevent another tsunami tragedy, coastal
communities need many more life-saving belts to help filter the energy of strong
winds and tidal waves. Until recently, mangrove forests have been recognized mainly
for their direct-use values (Fig 2.15).
Tidal gate located in the river mouth to prevent tsunami run-up and inundation
through the river channel. It also prevents the collapse of bridges due to hydraulic
bore of tsunami waves that travel to upstream. Lower priority is given to the
construction of tidal gate because of high construction cost. The tidal gate would be
required when development of facilities and infrastructures along river channel
is carried out.
Vertical evacuation structure (VES) can be intended for general use by the
surrounding population, or by the occupants of a specific building or group of
buildings. Choice between various options available for vertical evacuation structures
will depend on emergency response planning and needs of the community, the type of
construction and use of the buildings in the immediate vicinity, and the project-
specific financial situation of the state, municipality, local community, or private
owner considering such a structure.
The 2004 Sumatra Earthquake and Indian Ocean Tsunami vibrantly demonstrated the
menace of major tsunamis. Interestingly, the majority of reinforced-concrete
buildings, except those very close to the shoreline, survived with minor structural
damage, even though they were not designed for tsunamis or earthquakes
(Lukkunaprasit et al., 2008). This indicates that it is possible to design buildings to
withstand moderate tsunamis to serve as evacuation shelters. However, development
of design guidelines for safe and economical buildings against tsunami is still in its
early life.
Vertical Evacuation Structures (VES) should be located such that all persons
designated to take refuge can reach the structure within the time available between
tsunami warning and tsunami inundation.
Table 5-1 summarizes approximate warning times associated with the distance
between a tsunamigenic source and the site of interest.
The average, healthy person can walk at approximately 4-mph. Portions of the
population in a community, however, may have restricted ambulatory capability due
to age, health, or disability. The average pace of a mobility-impaired population can
be assumed to be about 2-mph (FEMA P646, 2008).
Tsunami VES should be spaced such that people will have adequate time not only to
reach the structure, but to enter and move within the structure to areas of refuge that
are located above the anticipated tsunami inundation elevation. Stairs or elevators are
traditional methods of ingress and vertical circulation in buildings, especially when
designated users have impaired mobility. Ramps, such as the ones used in sporting
venues, however, can be more effective for moving large numbers of people into and
up to refuge areas in a structure.
Special hazards in the vicinity of each site should be considered in locating VES.
Potential site hazards include breaking waves, sources of large waterborne debris, and
Wave breaking takes place where the water depth is sufficiently finite. In the design
of usual coastal structures (e.g., breakwaters, seawalls, jetties), critical wave forces
often result from breaking waves. In general, tsunamis break offshore. In the case of
very steep terrain, however, they can break right at the shoreline, which is known as a
collapsing breaker.
Forces from collapsing breakers can be extremely high and very uncertain. Location
of vertical evacuation structures within the tsunami wave-breaking zone poses
unknown additional risk to the structure. While the possibility of tsunami wave
breaking at an on-shore location is not zero, it is considered to be very rare. For these
reasons, recommended sites for vertical evacuation structures are located inland of the
wave-breaking zone, and wave breaking forces are not considered in this research
work. Fig 2.16 shows location of VES at the coast of Dwarka just opposite to
Dwarkadhish temple.
In order to serve effectively as a VES, it is essential that the area of refuge be located
well above the maximum tsunami inundation level anticipated at the site.
To account for this uncertainty, the magnitude of tsunami force effects is determined
assuming a maximum tsunami run-up elevation that is 30% higher than values
predicted by numerical simulation modeling or obtained from tsunami inundation
maps.
The recommended minimum freeboard is one story height, or 10 feet (3 meters) above
the tsunami run-up elevation used in tsunami force calculations. The recommended
minimum elevation for a tsunami refuge area is, therefore, the maximum tsunami run-
up elevation anticipated at the site, plus 30%, plus 10 feet (3 meters) (FEMA P646,
2008).
This chapter describes the processes involved to create the files and steps
necessary to complete the numerical modeling of a tsunami for the Arabian Sea.
Numerical modeling of tsunamis is commonly carried out to better understand
events that have occurred either during or before historical times. Numerical
modeling can also help to predict the effects of a future tsunami.
The present study uses the finite difference code of TUNAMI – N2 to predict
wave propagation (Imamura, 2005). Static displacement on the surface of an
elastic half space due to elastic dislocation was computed on the basis of
equations provided by Mansinha and smylie (1971).
Model runs of eighty six earthquake events were simulated for a far source
generated tsunami caused by an offshore earthquake. The models show a tsunami
wave approaching and starting to hit the west coast of Gujarat 140 to 185 minutes
after the earthquake event (Table 6.1 to Table 6.8). Out of 86, one model is in
keeping with what has already been seen in Makran Earthquake on November 28,
1945; a tsunami struck the coast approximately 150 minutes after the earthquake
event (Jaiswal et al.-1, 2008). As it has been suggested, there may also be other
possible generation mechanisms (Rastogi and Jaiswal, 2006; Mohammad et al.,
2009).
High resolution bathymetric data from the whole Arabian Sea is not often
collected due to the time, money and area involved to get such data. An only
small portion of the Western coast of Gujarat is known in sufficient detail to
recognize and count all the effects.
Western coast of Gujarat has not been extensively surveyed before. The
bathymetric grid was built from GEBCO 30 second resolution database. But due
to the large spatial variation of these points they didn’t give high enough
resolution to allow for accurate hydrodynamic modeling. This was a problem as it
would have a major effect on the accuracy of the results. Latest hydrographic
charts of Gujarat Maritime Board (GMB) and latest topographical data available
The basic GEBCO grid is in the form of World Geographic System format, it has
been converted in the form of Cartesian System in meter form by latitude-
longitude conversion constants with the help of Global Mapper and Surfer.
MS Excel was used to manipulate the digitized data and merged the entire
individual files into one, which was in turn loaded into SURFER 8.0 to
create the bathymetric files and maps that would be used in the TUNAMI-N2
numerical modeling suite. Fig 3.1 shows bathymetry of Arabian Sea and Fig 3.2
shows closer contours of Gujarat Region.
Compared to seismic hazard analysis a few research works have been devoted to
tsunami hazard assessment. Different researchers employed three main methods
for tsunami hazard assessment as follows: (a) Direct Statistical Tsunami Hazard
Assessment (DSTHA), (b) DTHA, and (c) Probabilistic Tsunami Hazard
Assessment (PHTA).
In this study, the method of DTHA has been employed for Makran for the
following reasons: a) the historical data of tsunami in the Makran region are rather
poor. Therefore, the direct statistical method (DSTHA) cannot yield useful results.
b) Establishment of specific tsunamigenic zones offshore Makran coast is difficult
because only one event is instrumentally recorded. This makes the application of
PTHA difficult. Furthermore, as this is one of the initial studies on the Makran
tsunami hazard assessment; there is a preference for the deterministic method over
the probabilistic one. The present deterministic methodology includes the
following steps: (a) selecting the characteristic earthquake, (b) modeling initial
condition, and (c) numerical modeling of tsunami.
TUNAMI-N2 program uses the bathymetry of the area as input data. The
bathymetry of the area is usually stored as data files. This file consists of three
values; x coordinate, y coordinate and the depth values. However data files are
typically randomly spaced files and this data must be converted into an evenly
spaced grid before using as input file of the program. To convert into a grid file, a
program called Surfer is used.
Below is the procedure for converting the bathymetry data file to grid file by
SURFER.
1. Start Surfer.
3. Specify the name of the XYZ data file which is the bathymetry data of the
area, and then click OK.
4. In the Grid Data dialog, specify the parameters for the type of grid file
you want to produce.
When creating a grid file you can usually accept all of the default gridding
parameters.
Data Columns
Individually specify the columns for the X data, the Y data, and the Z data. Surfer
defaults to X: Column A, Y: Column B, and Z: Column C, which represents the x
coordinate, y coordinate and the depth respectively.
Gridding Method
The gridding method should be set to Kriging which is the recommended gridding
method with the default linear variogram. This is actually the selected default
gridding method because it gives good results for most XYZ data sets.
TUNAMI-N2 requires a specific file format for the output grid file which is
ASCII grid file format.
The tsunami generation location was chosen due to the history of the area and past
events. As seen in chapter 2, different researchers have plotted historical
earthquake events that have preceded tsunami waves that are known to have
caused damage to the Western Coast of Gujarat in Past.
Byrne et al. (1992) calculated the source parameters of the 1945 earthquake as
seismic moment of 1.8 x 1021 Nm, moment magnitude of Mw 8.1, rupture length
of 150-200 km, rupture width of about 100 km, slip angle of 89˚, and depth of 17-
30km. Quittmeyer and Jacob (1979) reported surface magnitude (Ms), and rupture
length of the 1945 event to be 8 and 100-200 km, respectively. Jaiswal et al.-1
(2009) has taken fault parameters of the earthquake as fault length 200km and
width 100km, angle of strike, dip and slip 270˚, 15˚ and 90˚, respectively, focal
depth 10km and slip 10m.
In present study, the earthquake’s epicenter was moved along the MSZ to define
different scenarios for tsunami modeling. Based on the seismic parameters of the
1945 event and some other future probabilistic events, Eighty Six earthquakes
spaced along the MSZ from Karachi to the Strait of Hormoz have been considered
because any similar earthquake is capable of rupturing about 200-250 km of the
plate boundary (Rastogi and Jaiswal, 2006).
Total eighty six different fault dislocation scenarios were used. In this study there
were total nine earthquakes’s epicenter along the MSZ have been taken. For each
epicenter different fault parameters have been taken with the variation of ± 20% in
each fault parameter with reference to 1945 Makran Earthquake. All the faults are
of dip slip type.
The displacement field theory has been generalized to real Earth models by
Mansinha and Smylie (1971). They included initial hydrostatic stress, a liquid
core, self-gravitation, and radial variation of elastic constants, density and gravity
(Mansinha and Smylie, 1971).
The generation and propagation of tsunami waves are modeled using TUNAMI-
N2 code with the kind permission of Institute of Seismological Research,
Gandhinagar. TUNAMI-N2 is authored by Fumihiko Imamura of Tahoku
University, Japan, and developed in Middle East Technical University (METU)
and in the University of Southern California. It is an outcome of UNESCO TIME
All numerical models are based on two equations, which are solved by time
stepping. These equations are the momentum and mass conservations equations.
The momentum equation is used to describe the accelerations and retardation of
the water body when forces are applied to it. The equation is based on Newton’s
2 nd law of motion, where the acceleration is directionally proportional to any force
acting upon it. The mass conservation equation determines the total mass entering
and leaving model cells and ensures that mass is conserved in these cells (Black
1995).
The tsunami modeling process can be divided into three parts: generation,
propagation, and run-up.
Tsunami generation and run up are more difficult to model with computer
simulations. Simulation of wave run-up is very important for assessing the
magnitude of flooding in coastal areas (Enet and Grilli, 2005). Wave run-up
depends on both the wave amplitude and the wave period (Singh et al., 2008). It is
the main objective of this research to determine run-up height of tsunami wave at
different locations and how it will affect the western coast of Gujarat.
One common characteristic of all run-up laws, is a square root dependence on the
slope, and an almost linear dependence of the run-up on the wave height = H.
TUNAMI-N2 is one of the key tools for developing studies for propagation and
coastal amplification of tsunamis in relation to different initial conditions. It
solves nonlinear shallow water equations in Cartesian coordinates using the leap-
frog scheme of finite differences (Yalciner et al. 2002). TUNAMI-N2 and MOST
are the only two existing nonlinear shallow water codes, validated with laboratory
and field data (Yeh et al., 1996).
Tsunamis which are mainly generated by the movement of sea bottom due to
earthquakes belong to long waves. In the theory of such waves, the vertical
acceleration of water particles are negligible compared to the gravitational
acceleration except for an oceanic propagation of tsunami [Kajiura, 1963].
Consequently, the vertical motion of water particles has no effect on the pressure
distribution. It is a good approximation that the pressure is hydrostatic.
Based upon these approximations and neglecting the vertical acceleration, the
equations of mass conservation and momentum in the three dimensional problem
(see Fig 3.5) are expressed by the following theory:
+ + + =0
1 1
+ + + + + + + =0
1 1
+ + + + + + + =0
g+ =0 Eqn (3.1)
where x and y are horizontal axis, z the vertical axis, t time, h the still water depth,
the vertical displacement of water surface above the still water surface, u, v and w
are water particle velocities in the x, y and z directions, g the gravitational
acceleration, and Tij the normal or tangential shear stress in the i direction on the j
normal plane.
Any wave propagation problem can be solved using the governing equations -
Eqn (3.1) with boundary conditions. The dynamic and kinetic conditions at
surface and bottom are given as follows:
w= +u +v at z= Eqn (3.3)
Now, let us integrate Eqn (3.1) from the bottom to the surface using by Liebnitz
rule. For example, the first term of the momentum equation in the x-direction is
rewritten as follows:
( )
∫ = ∫ − ⃒ + ⃒ Eqn (3.5)
With dynamic and kinetic condition - Eqs. (3.2)-(3.4), the following two
dimensional equations are obtained, this is called the shallow water theory.
+ + =0 (2 − )
² ² ²
+ + +g + = +
² ²
² ² ²
+ + +g + = +
² ²
where D is the total water depth given by h+η, τx and τy the bottom frictions in the
x and y directions, A the horizontal eddy viscosity which is assumed to be
constant in space, the shear stress on a surface wave is neglected. M and N are the
discharge fluxes in the x- and y- directions which are given by,
= = (ℎ + ) =
= ∫ = (ℎ + ) = Eqn (3.7)
²
= / √ + Eqn (3.8)
Table 3.1: Values of Coefficient of Bottom Friction (Linsley and Franzini, 1979)
For the propagation of tsunami in the shallow water, the horizontal eddy
turbulence could be negligible compared to the bottom friction except for run-up
on the land. The following equations are therefore given as the fundamental
equations in the present model.
+ + =0
² gn²
+ + +g + /
+ =0
² ²
+ + +gD + /
N√M + N = 0 Eqn (3.9)
For the first step to describe the numerical scheme for the tsunami model, the
linearized long wave equation without bottom friction in one dimensional
propagation, Eqn (3.10), is introduced.
The above finite methods provide stable result as long as the C.F.L condition is
satisfied:
C (celerity) < ∆ /∆
Imamura (1996) investigated the truncation errors in three kinds of typical scheme
for long wave’s simulations and showed that in term of numerical accuracy the
staggered leap-frog scheme is the best among them.
The present program is only for tsunamis. No wind waves and tides are included.
The still water level is given by tides and is assumed constant during
tsunamis are computed. Accordingly, no motion is assumed up to the time n-1.
It means, therefore, in sea,
, , , =0 Eqn (3.11)
, ,
For run-up computation on land, the initial water level is equal to the ground
height h.
, = −ℎ , Eqn (3.12)
This yield
The wave front is located between the dry and submerged cells. Discharge across
the boundary between the two cells is carried out if the ground height in the dry
cell is lower than the water level in the submerged cell. In other cases, discharge
is considered zero.
The equations belong to the wave equation, for which the CFL condition should
be satisfied for stability of numerical computation.
Eqn (3.15)
Where t and x are the temporal and spatial grid lengths and hmax is the
maximum still water depth in a computation region.
In this study four nested domains, namely A, B, C, D are used (Fig-3.6). The
increased resolution is essential in order to simulate as best as possible the travel
time and tsunami amplitude of the waves. The intermediated grid (B) allows for a
better resolution all around the Arabian Sea. For grids A and B the model runs in
the linear mode which, although is not good enough for run-up estimates, is good
enough for travel time estimates. Another reason for increasing resolution as we
go into shallower water (grid C & D) is the fact that (Imamura, 1996; Murty,
1977) each tsunami wavelength should be covered by at least 20 grid points in
order to diminish numerical dispersion. Ramming and Kowalik (1980) found that
10 grid points per wave length is sufficient if we are willing to accept 2% error in
the phase velocity. Still another reason is that numerical stability considerations
require that the finite differences time step be such that Δt ≤ [Δx / √(2gh)], where
Δx is the space discretization size, g the gravitational acceleration, and h is the
maximum depth in the given grid. As the wave propagates into shallower waters h
decreases and by decreasing Δx we can maintain a constant Δt.
The data of Makran 1945 tsunami wave heights on the coastlines are rather poor,
and no tide gauge data is available. However, we were able to collect some data
about tsunami wave heights on various Makran coasts. Based on Page et al.
(1979), the 1945 tsunami reached a height of 7-10 m in Pansi. According to
Ambraseys and Melville (1982), tsunami wave height in Pansi was about 4-5 m
and was about 1.5 m high in Karachi. According to Pendse (1946), tsunami wave
height in Pansi was about 12-15 m and was about 1.37 m high in Karachi.
Numerical model adopted here is successfully reproducing the wave height of
between 1 to 2 m at Karachi as well as 7-12 m in Pansi.
TUNAMI does not use the fault data as it is. There is programme called
FAULTWAVE.f which runs the fault data and produces the initial wave by the
fault. TUNAMI uses this initial wave data to start with the modeling.
HH : Depth (m)
Other than the input file, the dimensional parameters in the FORTRAN code are
needed to be modified according to the number of grids of the bathymetry file.
The output file of the program is the initial wave in the sea which contains the
water surface elevations at each grid point of the bathymetry data. This output file
which is usually named as fault***.grd is also one of the input files of the
TUNAMI program.
Start
Ye
Output
End
Tsunami-N2 model is used for the propagation of tsunami waves for the event of
November 28, 1945, that happened in the Makran region of the Arabian Sea. The
tsunami propagation states at every 1-min interval are simulated. In this study,
model outer domain has a horizontal resolution of 2700 m over the Arabian Sea
including the Indian sub-continent (7–26N and 62–80E). The simulation is carried
out for duration of 360 min and the Sea states at 0, 30, 60, 90, 120, 150 and 180
min in the Arabian Sea are presented in (Fig-3.9.1 to 3.9.7). Because of the
variability in the bathymetry of the Arabian Sea and the earthquake that triggered
the tsunami waves, the wave amplitude varies with the propagation of waves. At t
= 0 min, the wave amplitude that is shown in the side bar with different colors,
next to the simulation figure, shows red at the point of epicenter. This indicates
the wave height is in the range of 5–6 m on the land-ocean boundary. At t = 30
min, the wave starts propagating toward the Makran and the western coast of
India. The wave amplitude varies with the forward motion of the tsunami waves.
Boundary conditions play a significant role in the separation of the land and ocean
boundary. Further, the red color indicates in (Fig-3.9.1) that the water surface is
higher than normal, while the blue means lower. Because of the fault geometry,
the waves propagating to the Makran Coast begin with a receding wave, which
explain why the Sea started to recede minutes before flooding the coast.
Directivity of earthquake fault in Arabian Sea and run-up height at Gujarat coast
is also evaluated with the help of software (Fig. 3.10 &3.11).
Photographs show the maximum amplitude simulated across the Indian Ocean
Basin. In order to validate quantitatively the simulation, it is compared to the only
available measurements of the 1945 Makran tsunami. The results obtained are
configured with the available reports of 1945 Makran Tsunami (Pendse, 1948;
Jaiswal et al.-1, 2008).
Table -3.2 to 3.9 shows the time of arrival of maximum amplitude of water waves
and wave run up height at Okha, Mandvi, Dwarka and Porbandar for different
fault models of MSZ. These give clear idea about how tsunami wave is
propagating in Arabian sea and what are the effects of shoreline bathymetry and
topography on it. The model also gives data about how far waves enter in the
landside. For exact inundation condition, a rigorous analysis with actual buildings
and other obstructions in coastal area is required with the help of this model data
as input.
EPI - 2
EPI - 3
EPI - 4
EPI - 5
EPI - 6
EPI - 7
EPI - 8
EPI - 9
Fig 3.12: Comparision of Run-Up Height with diff. Epicenters for Mandvi
0.3
0.2
0.1
0.0
EPI - 1
EPI - 2
EPI - 3
EPI - 4
EPI - 5
EPI - 6
EPI - 7
EPI - 8
EPI - 9
Fig 3.13: Comparision of Run-Up Height with diff. Epicenters for Okha
1.0
0.8
0.64
0.6
0.4
0.2
0.0
EPI - 1
EPI - 2
EPI - 3
EPI - 4
EPI - 5
EPI - 6
EPI - 7
EPI - 8
EPI - 9
Fig 3.14: Comparision of Run-Up Height with diff. Epicenters for Dwarka
1.0 0.86
0.8
0.58
0.6
0.4
0.2
0.0
EPI - 1
EPI - 2
EPI - 3
EPI - 4
EPI - 5
EPI - 6
EPI - 7
EPI - 8
EPI - 9
Fig 3.15: Comparision of Run-Up Height with diff. Epicenters for Porbandar
1.0
0.8
0.6 0.54
0.4 0.34
0.2
0.0
MANDVI OKHA DWARKA PORBANDAR
Fig 3.16: Comparision of Run-Up Height for diff. location for EPI-1
The elevation data sets are the most important input for inundation mapping in
tsunami prone areas. Preparation of a vulnerability map could inform coastal
community and others about susceptibility to inundation corresponding to various
wave-heights. The Gujarat state has important installations like ports, jetties,
industries along the coast, and also other socio-economical perspective, which can
be affected by tsunami trigged due to such events, and hence the determination of
possible inundation areas is important (Singh et al., 2008).
Inundation model is prepared for coastal parts of Gujarat belt on the basis of
existing topographic and bathymetric (water depth) data sets. For preparation of
the inundation map, high resolution of SRTM data set has been used. Surfer 8.0
software is used to plot elevation data. The different colors in Fig. 3.22 represent
various elevations from 0 to 10 m, over the parts of Gujarat state. From this Fig it
can be inferred that the Rann of Kutch region has very low elevations, and
therefore inundation would be more due to tsunamigenic conditions than other
costal parts of Gujarat.
Determination of tsunami wave height and arrival time is complex and time
consuming procedure and dependent on number of parameters. For such type of
complex problem artificial neural network architecture can be useful tool to
predict wave height and arrival time for future tsunami using available data sets of
tsunami propagation modeling. This cost effective tool may be useful as tsunami
warning system.
This section describes preprocessing, model design & training, model simulation
and post processing for generation of ANN prediction models.
Before applying inputs and outputs for ANN training, data have to be converted in
to range of 0 to 1 or -1 to 1 i.e. data should be normalized for ANN training. Eqn
3.16 shows an equation used for data normalization which ranges the data to [0,
1]. Normalized result table is shown in Appendix - 4. All 86 tsunami data sets are
divided for training, validation and testing. Different Neural Network models
were trained by Levenberg-Marquardt (LM) algorithm which used 76 data sets for
training and 10 data sets for testing. Use of more data sets in training reduces
processing time in ANN learning and improves generalization capability of
models, therefore maximum number of data sets available were used to train the
models.
Eqn (3.16)
Where,
X = Value of variable X
It is desirable to discard abnormal data sets for ANN model generation which
creates very large errors in prediction results of models.
The neural network design and training was done using MATLAB R2009a and its
associated GUI for Neural Network Toolbox.
For feed-forward back propagation networks, the default training algorithm is the
Levenberg-Marquardt (LM) algorithm. According to literature review this
algorithm is well suited to functionalize approximation or prediction problems
with networks of moderate size and number of parameters (Beale et al., 2010). It
is also well suited to problems that require the approximation to be very accurate.
As the problem under study fits that description, these recommendations were
considered and the conclusions on the actual data for this problem were validated.
Fig. 3.23 suggests how this model is designated. This designation covers various
properties of created ANN model. It covers type of training algorithm used,
number of neurons in hidden layer, transfer function used in between input and
hidden layer, and in between hidden and output layer.
The simulated model was trained for 76 test data and tested for 10 data sets. Fig.
3.24 shows MATLAB window of predicted output result for supplied test data
after simulation of LM14TT_76 model.
ANN model is trained by changing and storing proper weights in inter connection
links between neurons lying in various layers. These weight values are the
responsible parameters which gives prediction capability to trained ANN models.
It in connection links among input and hidden neurons, and neurons in hidden and
output layer for LM14TT_76 are shown in Table 3.10.
Before the Indian Ocean Tsunami, it was assumed that there was no need for the design
of structures against tsunami-induced forces. This situation was due to the assumption
that tsunamis are “rare” events, with significantly large return periods. The devastation
that may be caused by a tsunami of a large magnitude can be catastrophic as
demonstrated by the 2004 Indian Ocean event which induced significant structural
damage on infrastructure, killing over 3,00,000 people and leaving an estimated 1.5
million homeless (Ghobarah et al., 2006; EEFIT, 2006). It is shown that engineered
reinforced concrete buildings generally survived the tsunami without much damage (M.
Saatcioglu, 2009). Non-engineered concrete frame and confined masonry buildings
suffered different degrees of structural and non-structural damage, depending on the
topography and their proximity to the shoreline.
The design of structures in flood-prone areas has previously been investigated, but only
few existing codes specifically address the design of onshore civil structures built in
tsunami-prone areas. However, tsunami-induced forces and the impact of debris are not
properly accounted for in the existing codes and significant improvement is needed. A
comprehensive review of tsunami forces presented by Kason Hoku Pacheco (2005), M.
Saatcioglu (2009), Palermo and Nistor (2008), Nouri et al. (2008) and Okada et al. (2006)
are referred here. Recommendation by existing building codes and some other references
are depicted in this chapter. These include: the City and County of Honolulu Building
Code (CCH, 2000); the minimum Design Loads for Buildings and Other Structures;
SEI/ASCE 7-02 (ASCE 7, 2002); the Federal Emergency Management Agency- “Coastal
Construction Manual” (FEMA CCM, 2000); and the Federal Emergency Management
Agency- “Guidelines for Design of Structures for Vertical Evacuation from Tsunamis”
(FEMA P646, 2008); Okada et al., (2005); M. Saatcioglu, 2009.
The City and County of Honolulu Building Code Chapter 16 Article 11 authored by the
Department of Planning and Permitting of Honolulu, Hawaii, contains definitions, flood
proofing requirements in certain areas, flood proofing methods, structural requirements
and coastal flood water design to resist tsunami effects. Floodwater equations are
provided for the design of elevated structures which are based from the Dames & Moore,
1980 report.
The ASCE 7-02 "Minimum Design Loads for Buildings and other Structures" is authored
by the American Society of Civil Engineers, Committee 7. This model code provides the
various forces involved with flood and wave loads on specific types of structural
elements. The code also covers important definitions that relate to flooding and coastal
high hazard areas.
The Federal Emergency Management Agency Guidelines for Design of Structures for
Vertical Evacuation from Tsunamis chapter 6 of the code provides expressions for flood
loads and wave loads on specific types of structural members. The code also recommends
certain loading combinations for specific types of structural elements. FEMA P646
briefly introduces latest aspects of tsunami loading.
There are three parameters essential for defining the magnitude and application of
designing forces on structures: a) flow direction; b) inundation depth; and c) flow
velocity. These parameters mainly depend on: 1) run-up height of tsunami and arrival
time; 2) coastal topography; and 3) roughness of the coastal inland. The flow velocity and
direction is generally more difficult to estimate.
A set of generalized equations were selected from currently available building codes and
published literatures, which contain information and recommended equations on tsunami
forces on structures.
The buoyant force is the vertical force acting through the center of mass of a submerged
body. Buoyant forces are a concern for basements, empty above-ground and below-
ground tanks, and for swimming pools. Any buoyant force on an object must be resisted
by the weight of the object and any opposing force resisting flotation. If there is
insufficient building weight to resist buoyant forces, tension piles may be used to increase
the resistance to flotation, but reduction in pile side friction due to anticipated scour
around the tops of the piles must be considered.
When water flows around a building with moderate to high velocity hydrodynamic loads
caused by drag are applied to the building. They are usually called the drag forces, which
are combination of the lateral loads caused by the impact of the moving mass of water
and the friction forces as the water flows around the obstruction. These loads are a
function of flow velocity and structure geometry, and include frontal impact on the
upstream face, drag along the sides, and suction on the downstream side. The flow is
assumed to be uniform, and therefore, the resultant force will act at the centroid of the
projected area or in the form of uniformly distributed load along the inundation depth.
The hydrodynamic force is directly proportional to the square of the tsunami bore
velocity as indicated in Eqn (4.4). Therefore, the estimation of the bore velocity remains
to be one of the critical elements on which there is significant disagreement in literature.
Fd = Cd u2 ρ A Eqn (4.4)
Where :
Cd = drag coefficient (recommended values are 2.0 for square or rectangular
piles/columns and 1.2 for round piles/columns).
The combination h u 2 represents the momentum flux per unit mass. Note that h u 2
max
2
does not equal hmax umax . The maximum flow depth, hmax and maximum flow velocity,
hmax at a particular site may not occur at the same time. The hydrodynamic forces must be
based on the parameter h u 2 , which is the maximum momentum flux per unit mass
max
2
z z
hu
2
max
2
gR 0.125 0.235 0.11
R R
Eqn (4.7)
Where g is the acceleration due to gravity, R is the design run-up elevation, and z is the
ground elevation at the base of the structure. The design run-up elevation, R, is taken as
1.3 time the maximum run-up elevation, R*, which is the maximum inundation elevation
at the structure from a detailed numerical simulation model, or the ground elevation at
maximum penetration of the tsunami from available tsunami inundation maps. To use this
formula, the sea level datum must be consistent with that used in the inundation maps.
The surge force is generated by the impingement of the advancing water front of a
tsunami bore on a structure. The surge force is computed as a force per unit width on a
vertical wall subjected to a surge from the leading edge of a tsunami. Due to lack of
detailed experimental research specifically applicable to tsunami bores running up the
shoreline, the calculation of the surge force exerted on a structure is subjected to
substantial uncertainty.
The surge-force computation by Dames and Moore (1980), CCH (2000) may result in
excessively overestimated values. The estimation made by CCH (2000) implies that the
surging force would be 9 times the hydrostatic force alone. Such an excessively large
surging force is contradictory to the laboratory results by Ramsden (1993) and Arnason
(2005) referenced by (Yeh et al., 2005). Lukkunaprasit et. al. (2010) confirms that the
predicted forces by FEMA P646, 2008 provide a reasonable upper bound for the
measured forces.
The equivalent static pressure resulting from the tsunami impact is associated with a
triangular distribution where water depth equals three times the tsunami inundation depth
(Asakura et. al, 2000)
qx = ρ g (3h – z) Without soliton (solitary wave) breakup
qx = max (5.4h – 4z, 3h –z) ρ g with soliton (solitary wave) break up Eqn (4.10)
where qx is the tsunami wave pressure for structure design, z is the height of the relevant
portion from ground level (o ≤ z ≤ 3h).
FEMA P646 recommended that the impulsive forces or surge forces be taken as 1.5 times
the hydrodynamic force.
Fs = 1.5 Fd Eqn (4.11)
A high-speed tsunami bore traveling inland carries floating debris such as driftwood,
small boats, floating pieces of houses, cars, automobiles, etc., or any object transported by
floodwaters, striking against buildings and structures or parts thereof. The impact of
floating debris can induce significant forces on a building, but it is very difficult to predict
the magnitude of these loads. Yet some reasonable allowance must be made for them.
The object is assumed to be at or near the water surface level when it strikes the building.
Therefore, the object is assumed to strike the building at the water level. Uncertainty
about the duration of the impact time is the most likely cause of error in the calculation of
debris impact loads. The duration of impact is influenced primarily by the natural
frequency of the building, which is a function of the building’s stiffness. This stiffness is
determined by the properties of the material being struck by the object, the number of
supporting members (columns/piles), the height of the building above the ground, and the
height at which the building is struck.
Fi = m * a Eqn (4.12)
Where : Fi = impact force in pounds (kN)
V
a
t
wV
Fi
gt Eqn (4.13)
Where :
1
V = velocity of water in m/sec or approximated by gd ,
2
t = duration of impact in seconds (1.0 seconds for wood construction, 0.5 seconds for
steel construction and 0.1 seconds for reinforced concrete).
z
umax 2 gR 1
R Eqn (4.15)
where g is the acceleration due to gravity, R is the design run-up height that is 1.3 times
the ground elevation R at the maximum tsunami penetration, and z is the ground elevation
at the structure (the datum must be at the sea level).
For a shipping container or other similar large debris with draft d, the ratio of the draft d
to the maximum run-up height R can be computed, and Fig 4.5 can be used to estimate
the maximum flow velocity, Draft d can be estimated using Eqn (4.15):
d = W / (ρ g Af) Eqn (4.15)
where W is the weight of the debris, ρ is the fluid density including sediment, g is the
acceleration due to gravity, and Af is the cross-sectional area parallel to the water surface
such that the product d x Af represents the volume of water displaced by the debris.
Fig 4.5: Maximum Flow Velocity of Depth, d, at the Ground Elevation, Z, and
Maximum Run-up Elevation, R.
Based on the appropriate curve for η (d/R), and ration between the elevations of the
structure relative to the design run-up elevation z/R, Fig 4.5 will provide an estimate of
the maximum flow velocity.
1 up 2
FR pg h
2 2g
Eqn (4.16)
Where FR = hydrostatic force (kN/m) on a wall, per unit width of wall
p = density of water
u p = component of velocity of flood flow perpendicular to the wall (m/s)
The resultant force will act horizontally at a distance of hg above the base of the wall,
1 up 2
hg h
3 2g
Eqn (4.17)
1 2
fm rd s
2 Eqn (4.18)
Where : f m = hydrostatic force per unit width (kN/m) resulting from flooding against
vertical element
Past research suggest that scour depth depends on soil type and that scour depths in areas
up to 91 meter and above 91 meter from the shoreline can be determined as a percentage
of the Stillwater depth h, as shown in Table 4.1.
Above stated values may be reduced by 40%, if a substantial dune or beam higher
than the regulatory flood elevation protects the building site. In case of the entire
region is essentially flat, values may be reduced by 50% .
This code is also recommending the same scour depth as recommended by CCH (2000).
Subgrade failure resulted from scouring Loss of support under shallow foundations
(Kamala, Phuket) (Thailand)
Fig 4.8: Damage of Building Due to Scour
Tsunami waves tend to break offshore and approach shoreline as a broken hydraulic bore
or a soliton, depending on wave characteristics and coastal bathymetry. Buildings and
shelters are usually constructed on land some distance from the shoreline. Hence, wave-
breaking force exerting directly onto the building would be a rare case and need not to
consider while designing a onshore building.
Where Bd is the breadth of the debris dam. A minimum debris dam width of Bd = 12m,
representing a sideways shipping container or a mass of floating lumber, is recommended.
The estimation of design flood velocities in coastal flood hazard areas is subject to
considerable uncertainty. There is little reliable historical information concerning the
velocity of floodwaters during coastal flooding. The direction and velocity of floodwaters
can vary significantly throughout a coastal flood event. Flood velocities should be
estimated conservatively by assuming floodwaters would approach from the most critical
direction and by assuming high flow velocities. Prevailing codes do not mention the site-
specific tsunami velocity.
V=h
Tsunami
V 2 gd , Eqn (4.22)
Upper Bound
V gd ,
z
umax 2 gR 1
R Eqn (4.23)
Where g is the acceleration due to gravity, R is the design run-up height that is 1.3 times
the ground elevation R at the maximum tsunami penetration, and z is the ground elevation
at the structure (the datum must be at the sea level).
Fig 4.9: A Definition Sketch for Upward Buoyant Force Exerted on an Elevated
Floor.
The total uplift force on the floor system can be estimated using Eqn (4.24)
Fu = ½ Cu ρ Af uv2 Eqn (4.24)
Where Cu is a coefficient (taken as 3.0), Af is the area of the floor panel or floor framing
component, and uv is the estimated vertical velocity or water rise rate.
Unless a detailed hydrodynamic study is performed, the value of uv for the condition of
sloping terrain below the building can be estimated using Eqn (4.25):
The maximum potential downward load per unit area, fr can be estimated using Eqn
(4.26):
fr = ρ g hr Eqn (4.26)
where hr is the maximum potential depth of water retained on the elevated floor
determined using Eqn (4.27):
Where hmax is the maximum inundation level predicted at the site, h1 is the floor
elevation above grade, and hbw is the maximum water depth that can be retained before
failure of the wall due internal hydrostatic pressure.
Tsunami forces area combined on individual structural components (e.g., columns, walls,
and beams), as follows:
Impulsive force, Fs due to the leading edge of the tsunami bore, for maximum
hu 2
Hydrodynamic drag force, Fd plus debris impact, Fb at the most critical location
2
on the member, for maximum h u .
Debris damming, Fdm due to a minimum 12.2 m wide debris dam causing the
2
worst possible loading on the member, for maximum h u .
Hydrostatic pressure, Fd on walls enclosing watertight areas of a structure, for
maximum h.
For uplift on floor framing components, the following combinations of tsunami forces
should be considered:
Downward load due to water retained by exterior walls, f n combined with 100% dead
load.
The pacific basin has a long history of devastation by tsunami, because it is the most
active seismic region on earth. Tsunamis worldwide have killed hundreds of thousands of
people. On 26th December, 2004, an earthquake of moment magnitude 9.3 off the coast
of Sumatra in the Indian ocean, triggered a series of lethal tsunami that killed over
3,00,000 people, making this the deadliest tsunami in recorded history. It has been
concluded that any community near a coastline is vulnerable to a tsunami, which indicates
a need for a global tsunami warning system. Coastlines have always been a popular
location for people to settle, because of the natural resources the ocean provides and its
beauty.
Since tsunami is relatively infrequent, it is difficult to measure tsunami flow velocities,
and the resulting loads on structures. The maximum vertical height and horizontal
distance to which the water is observed with reference to sea level is referred to as run-up
and inundation, respectively.
Performance of buildings during the 2004 Indian Ocean tsunami indicate that tsunami
generated hydraulic pressures and debris impact forces can have significant effects on
structural behavior (M. Saatcioglu, 2009). In past tsunamis, the majority of non-
engineered structures are completely destroyed. However, much can be learned from
structures that survived the tsunami, or were only partially damaged. Field surveys and
photos obtained from a number of past tsunamis were used for further evaluation of the
characteristics of tsunami and structural response.
Following three structures are studied here; a) In 1992, an earthquake with the magnitude
of 7.0 generated a tsunami in Nicaragua which devastated communities along the
shoreline and caused 116 deaths. b) In 1993, a powerful earthquake west of Hokkaido in
the Sea of Japan caused a tsunami that demolished areas on nearby Okushiri Island. c) On
Sunday 26 December, 2004, at 7:58 am local time, a megathrust earthquake of moment
magnitude 9.3 occurred off the west coast of the Indonesian island of Sumatra (EERI-1;
Ghobarah et al., 2006).
This structure is located in Nicaragua near the shoreline approximately 100 feet from the
ocean. The structure consists of concrete slab foundation with a timber structure sitting 9
feet 6 inches high on timber columns (Fig 4.10 and 4.11). The 1992 tsunami inundated
approximately 6 feet 7 inches (2 meters) at this location. The estimated tsunami flood
velocity was approximately 4 meters per second. The structure survived since the tsunami
flood was able to flow freely under the structure without impacting the structural walls.
The “knee-bracing” at the top of the columns proved adequate to resist loads developed
on the columns during the tsunami. Scouring is evident around the edge of the foundation,
This Structure is located in an area that experienced almost devastation (Figure 4.12). The
structure consists of reinforced concrete walls and columns (Fig 4.13 to Fig 4.15). Only
non-structural damage was noted to the top level balcony railings and the front glass
windows. The tsunami flood reached 2.7 meters at the structure, with debris deposited on
the second floor level. The estimated tsunami flood velocity was approximately 5 meters
per second. The reinforced concrete framing in this structure was able to resist the
tsunami loads even though much of the first floor walls did not breakaway.
This Structure is a weather monitoring building of the Meteorological station at Takua Pa,
Phang Nga, Thailand. The building is about 200m from the shore line, relatively without
obstruction from the incoming waves except for a few lines of pine trees. The single story
reinforced concrete (RC) structure, whose plan and elevation are shown in Fig 4.17, is
supported by shallow foundations. The columns, 200mm × 200mm in cross section, are
reinforced with 4–12mm diameter longitudinal reinforcing bars. The building has
survived the 4.4 m wave height with minor structural damage (Fig 4.16 and 4.17). The
covers of the columns did not even spall off.
The ability of a structure near the shoreline in a tsunami inundation zone to survive a
tsunami is important in order to save lives. If the structure is capable of resisting the
tsunami loads, it can be used for vertical evacuation. The recent Indian Ocean tsunami
event illustrated the effectiveness of vertical evacuation particularly for a locally
generated tsunami. With limited warning time, it is often more appropriate for people to
seek refuge in the upper floors of a multistory building than to attempt horizontal
evacuation to higher grounds. It is important, therefore, that such buildings resist not only
the initial ground shaking but also the subsequent tsunami loads.
Several model studies have focused on tsunami loading on structures (M. Saatcioglu,
2009; Palermo and Nistor, 2008; Nouri et al., 2008; Okada et al., 2006; Lukkunaprasit et.
al., 2009; Yeh et al.,2005). Yeh. et. al. (2005) developed design guidelines for buildings
subjected to tsunami loading by analyzing tsunami force in detail and compiling
equations currently addressing loads under flooding. A recent design guidelines
document – FEMA P646, 2008 published by the US Federal Emergency Management
Agency (FEMA) – proposes a practical method to estimate the tsunami design forces at a
given locality with a known maximum tsunami run-up height. Lukkunaprasit et al. (2009)
confirms that the predicted forces by FEMA P646, 2008 provide a reasonable upper
bound for the measured forces.
For generalization of structural behavior under tsunami forces seven typical structures
having different configuration are taken for investigation. It is presumed that structures
are located on western coast of Gujarat for seismic and wind analysis. Analysis for
tsunami has been carried out for 3m, 5m and 10m design run-up height. The building
suggested for VES (Rectangular Building) is analyzed for seismic zone III, IV & V, wind
and tsunami having design run-up height 2.264m, 3m, 5m & 10m.
Among seven typical buildings a G + 11 Building is taken from “Evaluation of Tsunami
Loads and Their Effect on Reinforced Concrete Buildings” by Kason Hoku Pacheco,
2006 for validation of structural behavior under the tsunami effects using STAAD-Pro
v8i (Appendix – B). In this reference design of five reinforced concrete buildings for
tsunami effects are presented.
Six Storey Building is one another typical building which was taken from IITK-GSDMA
CODES (EQ-26) for validation of earthquake effects simulated using STAAD-Pro v8i
(Appendix – A). In this reference authors have provided analysis and design of six storey
reinforced concrete building for seismic effects under IS Codal provisions.
All seven reinforced concrete buildings incorporate different structural configurations.
Buildings are analyzed and designed according to the Indian Standard (IS) Codal
Provisions. Particularly this study utilizes IS 456, IS 1893-I, IS 875-I, IS 875-II, IS 875-
Different tsunami loading combinations were applied to the respective floor columns.
Structural members all floor levels were investigated. Lower storey columns are having
danger of impact loads. Failure of any of these members could lead to progressive
collapse of a significant portion of the building.
The lateral tsunami loads were applied to the structural members using STAAD-Pro
Version v8i, a structural analysis computer program. The maximum bending moment,
shear force and axial force obtained in tsunami load case were compared with the seismic
and wind load case for critically located members.
Buoynt Force per Unit area 31537.19 0.59 kN/m2 ON FLOOR AREA IN UPWARD DIR.
STEP-
4 Hydrodynamic Force - Fd
(hu2)max
(hu2)max = [g*R2(0.125 - 0.235*(Z/R) + m3 /
0.11(Z/R)2)] 125.54 sec2
u2 12.56
ζ = Z/R 0.17
flow depth ratio = d/R 0.02
umax / √(2*g*R) from fig - 6.7 - FEMA - P
646 0.14
umax 2.15 m/sec
AT LEADING EDGE OF STR AND AT
Fiw (realistic) = Cm * umax * √(k*m) 141186.10 141.19 kN WATER LEVEL (POINT LOAD)
STEP-
7 Impact Force (shipping container) - Fi - 3
draft d = m*g / (ρ*g*A box) 0.84 m
d/R 0.07
ζ = Z/R 0.17
umax / √(2*g*R) from fig - 6.7 - FEMA - P
646 0.14
umax 2.15 m/sec
AT LEADING EDGE OF STR AND AT
Fis (realistic) = Cm * umax * √(k*m) 1152779.67 1152.78 kN WATER LEVEL (POINT LOAD)
STEP-
8 Damming Effect of Water borne Debris - Fdm
Fdm = Fd * (width of ship / width of wall) 1807821.82 1807.82 kN
Damming Effect of Water borne Debris per AT LEADING EDGE OF STR AND AT
Unit length = Fdm 903910.91 903.91 kN WATER LEVEL (TWO POINT LOAD)
STEP-
9 Hydrodynamic Uplift Forces - F uplift
ζ = Z/R 0.17
Figs 5.1 to 5.10 depict line diagram of plan and elevation of seven typical buildings. The
structural element at critical locations indicated in the plan. The flow of tsunami is also
shown in plan. All dimensions of the buildings are in meter.
Fig 5.1: Typical Floor Plan of G+11 Building & Uneven G+11 Building
Fig 5.2: Elevation View of G+11 Building & Uneven G+11 Building
Fig 5.5: Typical Floor Plan of Rectangle Building & Uneven Rectangle Building
Fig 5.6: Elevation View of Rectangle Building & Uneven Rectangle Building
Earthquakes generating tsunami have occurred in the past; it has occurred now and
will continue to occur. As the population of coastal areas increases, the need for, and
value of, scientific understanding of earthquake and tsunami hazards also increases.
To mitigate this hazard, efforts in four directions are needed. Firstly, work has to be
done in the direction of developing Tsunami Early Warning System. Secondly,
tsunami hazard maps have to be prepared showing the possible inundation areas in
case of a tsunami attack. Thirdly, educating people and disseminating information
about the people likely to be affected. Lastly, planning, designing and construction or
retrofitting of structures in coastal areas considering tsunami loads.
Eastern and Western parts of the MSZ of southern Pakistan are potential zones for
great earthquakes that can generate tsunamis affecting west coast of Gujarat. The
eastern part of the MSZ has produced the 1945 Mw 8.1 earthquake that generated the
last major tsunami in the Arabian Sea. Some sectors of the MSZ are un-ruptured for a
long time and can produce large earthquakes in near future.
In this study investigation has been carried out for the far-field amplitude of tsunamis
in the Arabian Sea for 86 numbers of scenarios of mega-thrust earthquakes, ranging
from probable to possible. The most important lesson from the scenarios investigated
in this study is that the patterns of far-field maximum amplitudes predicted by these
simulations are matching with the observed in 1945 Makran Earthquake.
Each model is having 900 X 1360 nodes. Each model is simulated for 6 hours time,
which will take almost 15 hours time on Core 2 duo processor. Each model will
occupy 3.6 to 3.9 GB hard disk space.
Particularly in western coast of Gujarat, Dwarka is also one of the most important
cities. Dwarka, a religious and historical place of Gujarat and India, may have more
severe hazard against tsunami from the Arabian Sea. So Dwarka can be protected
with proper methods of mitigation and disaster management. Vertical Evacuation
Structure has been proposed for the Dwarka city.
Seven typical R.C. buildings are analyzed for wind, earthquake and forces subjected
to 2.264, 3, 5 and 10 meter high tsunami flow. Important conclusions have been
drawn from the analysis results.
Ph.D. Thesis – V M Patel / Chapter- 7 / Summary, Conclusion & Recommendations 181| P age
7.2 CONCLUSIONS
In some locations, high ground may not exist, or tsunamis triggered by local
events may not allow sufficient warning time for communities to evacuate
low-lying areas or densely populated areas like Dwarka Temple. Where
horizontal evacuation out of the tsunami inundation zone is neither possible
nor practical, a potential solution is vertical evacuation structure, which is
designed and detailed to resist the tsunami forces.
Historical records state that only structurally engineered buildings, and
structures raised above the tsunami flow, were able to survive the tsunami
forces without collapse. In some cases, the failure of breakaway walls at the
lower levels or large openings in walls at lower storey reduced the tsunami
loadings, hence the building survived.
Higher design forces in columns and beams below tsunami flow level were
observed in the typical buildings with solid walls.
Reduction in design forces were observed when longer side of the building is
in perpendicular direction to shoreline.
Ph.D. Thesis – V M Patel / Chapter- 7 / Summary, Conclusion & Recommendations 182| P age
With an additional cost of about 5%, a building with 50% opening can be
tsunami resistant for tsunami flow up to 3 meter.
With an additional cost of about 28%, a building with 50% opening can be
tsunami resistant for tsunami flow up to 5 meter.
Damming effect of water born debris and impact loads are two critical loads
for failure of buildings up to 5 meter height of tsunami flow, and these loads
can be eliminated with proper vegetation along the coast like mangrove trees.
Irregularity in plan & elevation of building will increase shear force and
bending moment of beam and column.
Overall axial force is lower in tsunami load case except 10 meter high tsunami
flow with solid wall case compared to earthquake load case.
A solid masonry wall cannot resist the tsunami flow of height 3 meter or
more, but the failure of solid wall will minimize the tsunami load and framed
structure may survive.
For Gujarat coastal area, there is a risk of 3.28 meter high tsunami flow.
Therefore, ground floor columns and foundations should be strengthened
according to tsunami requirements to make the whole building safe against
probable tsunamis.
In wind case, all member forces are observed lower, but non-structural
elements in the building should be designed considering local wind pressure
co-efficient in coastal area.
In case of more than 3 meter tsunami height, beams near corner which are
parallel to tsunami flow observed considerably higher shear force and bending
moment below tsunami height.
7.3 RECOMMENDATIONS
Large scale wave tank studies are recommended to verify tsunami velocity
and corresponding effect on acting loads and impact loads due to water borne
debris.
Effects of vegetation need to be checked experimentally.
Based on experimental studies, the tsunami loading equations can be revised.
It is recommended that breaking wave forces should not be considered.
Progressive collapse analysis may be recommended in case of higher impact
forces.
It is recommended to avoid use of solid walls perpendicular to the anticipated
tsunami flow in the buildings.
Reinforced concrete or steel frame structures are recommended in tsunami
prone areas.
Ph.D. Thesis – V M Patel / Chapter- 7 / Summary, Conclusion & Recommendations 183| P age
Further research is needed to improve design and detailing requirements for
structural and non-structural components.
Ph.D. Thesis – V M Patel / Chapter- 7 / Summary, Conclusion & Recommendations 184| P age
APPENDIX - A
The general layout of the building is shown in Figure 1. At ground level, the floor beams
FB are not provided, since the floor directly rests on ground (earth filling and 1:4:8 c.c. at
Storey numbers are given to the portion of the building between two successive grids of
beams. For the example building, the storey numbers are defined as follows:
Max: -251.829 kN
88.103 kN
Max: -268.251 kN
71.680 kN
Fig A1.3: Shear Force Diagram for Loadcase: [1.5 (DL + EXTP)]
-71.680 kN
Max: 268.251 kN
-88.103 kN
Max: 251.829 kN
Fig A1.4: Shear Force Diagram for Loadcase: [1.5 (DL - EXTP)]
-85.595 kNm
-19.893 kNm
-15.211 kNm
Max: -216.773 kNm
-180.251 kNm
Fig A1.5: Bending Moment Diagram for Loadcase: [1.5 (DL + EXTP)]
-180.251 kNm
Max: 575.427 kNm
-15.211 kNm
Fig A1.6: Bending Moment Diagram for Loadcase: [1.5 (DL - EXTP)]
-121.582 kN
Max: -121.582 kN
Rect 0.71x0.71
Rect 0.71x0.71
Max: -67.279 kNm
clear all;
close all;
load grid_a.grd; % input bathymetry
ind=find(grid_a < 0); % looking for dry land updated on 24.05.2005
[M,N]=size(grid_a);
a1=[];
a2=[];
a3=[];
a4=[];
for k = 150 : 156
filename = ['z' num2str(1000+k-1)];
fid=fopen(filename,'r');
z=fscanf(fid,'%f',[N,M]); z=z';
fclose(fid);
z(ind)=nan*z(ind);
text(100,500,[num2str(k-1), 'min']);
a1=[a1 z(615,842)];
a2=[a2 z(590,867)];
a3=[a3 z(594,876)];
a4=[a4 z(607,890)];
end
[max_val1,index1]=max(a1);
disp('max value PORBANDAR is: ');
disp(max_val1);
disp('frame for PORBANDAR is: ');
disp(['z' num2str(1150+index1-1)]);
[max_val2,index2]=max(a2);
disp('max value DWARKA is: ');
disp(max_val2);
disp('frame for DWARKA is: ');
disp(['z' num2str(1150+index2-1)]);
[max_val3,index3]=max(a3);
disp('max value OKHA is: ');
disp(max_val3);
disp('frame for OKHA is: ');
disp(['z' num2str(1150+index3-1)]);
[max_val4,index4]=max(a4);
disp('max value MANDVI is: ');
disp(max_val4);
Following Data Taken For Shear Strength Capacity Check of T-2.264 Model –
Column C1:
D = D/2 + Dr/∏
= 500/2 + 360/3.14
= 364.65 mm
Vs = [(Av x Fsh x D) / S]
[AASHTO, 2002 referred from Jae-Hoon Lee]
= [(6 x 78.54 x 415 x 364.65) / 140]
= 509.37 kN
Vn = Vc + Vs
= 208.37 + 509.37
= 717.75 kN > 612.00 kN (Shear Force acting on Column C1 at GF of T-2.264 of
Tsunami Model)
Bendick, R, and Bilham R, 1999, Search for buckling of the southwest Indian coast
related to Himalayan collision, In Macfarlane, A., Sorkhabi, R., and Quade, J. (Eds.),
1999, Himalaya and Tibet: mountain roots to mountain tops, Geological Society of
America Special Paper 328, p. 313-323.
Borrero J C, Synolakis C E, Fritz H, 2006, Northern Sumatra Field Survey after the
December 2004 Great Sumatra Earthquake and Indian Ocean Tsunami, Earthquake
Spectra, Vol.22, Nos.S3, pp. S93–S104.
Clift P D, Kroon D, Gaedicke C, Craig J, (2002), The Techtonic and Climatic Evolution
of the Arabian Sea Region, Geological Society, London, Special Publication, 195, pp.1-6.
Earthquake Engineering Field Investigation Team, (2006), The Indian Ocean Tsunami of
26 December 2004: Mission Findings in Sri Lanka and Thailand, Institution of Structural
Engineers, London, UK.
EERI -1, The Great Sumatra Earthquake and Indian Ocean Tsunami of December 26,
2004, EERI Special Earthquake Report — March 2005.
EERI -2, The Great Sumatra Earthquake and Indian Ocean Tsunami of December 26,
2004, EERI Special Earthquake Report — April 2005..
Enet F and Grilli, S, (2005), Tsunami Landslide Generation: Modeling and Experiments,
Ocean Waves Measurement and analysis, 5th International Symposium WAVES 2005,
Madrid, Spain, pp. 88-97.
HarryYeh, (2006), Maximum Fluid Forces in the Tsunami Run-up Zone, Journal of
Water Way, Port,Coastal, and Ocean Engineering, pp. 496-500.
Hatori, T., (1995), Magnitude Scales for the Central American Tsunamis, (1990-1993),
Pure and Applied Geophysics, Vol.144, Nos.3/4, pp.471-479.
Imamura, F., (1996). Review of Tsunami simulation with a finite – difference method in
Long –wave Run-up Models, World Scientific, Singapore, pp. 25-42.
Kajiura K, (1963), The Leading Edge of a Tsunami, Bulletin of the Earthquake Research
Institute, Uni. Tokyo, Vol. 41, pp. 535-571.
Kanamori H and Given W, (1983), Use of Long-Period Seismic Waves for Rapid
Evaluation of Tsunami Potential of Large Earthquakes, Tsunamis – Their Science and
Engineering, pp. 37-49.
Macmurdo C, (1821), Account of the earthquake which occurred in India in June 1819,
Edinburgh Phil.J. 4, 106-109.
Mansinha L, Smylie D (1971), The Displacement fields Of Inclined Faults, Bull Seismol
Soc Am 61: 1433–1440.
Murty T S, (1977), Seismic Sea Waves – Tsunamis, D. W. Frieses & Sons Ltd., Altona,
Canada, ISBN – 0-660-00565-4.
Nelson C, (1846). Notice of an earthquake and a probable subsidence of the land in the
district of Cutch, near the mouth of Koree, or the eastern branch of the Indus in June
1845, Geol. Soc. London, Quart. J., 2, 103.
Okada T, Sugano, Ishikawa, Takai & Tateno, (2006), Tsunami Loads and Structural
Design of Tsunami Refuge Building, www.bcj.or.jp / src / c05_research / results /
tsunami2006_1.pdf.
Rastogi B K and Jaiswal R K, (2006). A Catlog of Tsunamis in the Indian Ocean, Science
of Tsunami Hazards, Vol-25, pp 128-142.
Wang X and Liu P F, (2006), Ananalysis of 2004 Sumatra Earthquake Fault Plane
Mechanisms and Indian Ocean Tsunami, Journal of Hydraulic Research, Vol.00, No.0,
pp.1–8.
Yeh H I, Robertson, Pruess J., (2005), Development of Design Guidelines for Structures
that Serve as Tsunami Vertical Evacuation Sites, Washington State Department of
Natural Resources, Olympia, WA.
Codes:
FEMA P646, (2008), Federal Emergency Management Agency - Guidelines for Design
of Structures for Vertical Evacuation from Tsunamis.
IS-1905, (1987), Indian Standard Code of Practice for Structural Use of Unreinforced
Masonry.
IS-456, (2000), Indian Standard - Plain and Reinforced Concrete - Code of Practice
IS-875, PART-I, (1997), Indian Standard Code of Practice for Design Loads (Other Than
Earthquake) for Buildings and Structures - Dead Loads.
IS-875, PART-II, (1997), Indian Standard Code of Practice for Design Loads (Other
Than Earthquake) for Buildings and Structures, Imposed Loads.
IS-875, PART-III, (1997), Indian Standard Code of Practice for Design Loads (Other
Than Earthquake) for Buildings and Structures, Wind Loads.
IS-875, PART-V, (1997), Indian Standard Code of Practice for Design Loads (Other
Than Earthquake) for Buildings and Structures, Special Loads and Combinations
NTHMP – www.nthmp.tsunami.gov/publications.html
Saatcioglu, M., (2009), Performance of Structures During The 2004 Indian Ocean
Tsunami and Tsunami Induced Forces For Structural Design, “Earthquakes and
Tsunamis”, A.T. Tankut (ed): http://books.google.co.in/books?id=u-
BrJdXr70MC&pg=PA170&lpg=PA170&dq=impact+force+cch-2000&source= bl&ots=
RxdFdQRcL4&sig=suSqdJCdekK0vZIX77GkAwhToiQ&hl=en&ei=mwMQTd6OMcjm
rAfv66XECw&sa=X&oi=book_result&ct=result&resnum=6&ved=0CDcQ6AEwBQ#v=
onepage&q=impact%20force%20cch-2000&f=false