CHAPTER 2

LITERATURE REVIEW

2.1. General
Over the last decades considerable attention has been raised on the behavior of
engineering structures under blast or impact loading. The use of explosives by
terrorist groups around the world that target civilian buildings and other structures is
becoming a growing problem in modern societies. Explosive devices have become
smaller in size and more powerful than some years ago, leading to increased mobil–
ity of the explosive material and larger range effects. Usually the casualties from such
a detonation are not only related to instant fatalities as a consequence of the direct
release of energy, but mainly to structural failures that might occur and could result in
extensive life loss. After the events of the 11 th September 2001 that led to the collapse
of the World Trade Center in New York it was realized that civilian and government
buildings, as well as areas with high people concentration (metro and train stations,
means of mass transportation, stadiums etc.) are becoming potential bombing targets
of terrorist groups. Since most engineering structures are vulnerable to such type of
loading scenarios, a guide should be introduced to the designer in order to guarantee
structural integrity even under those extreme situations.
Estimation of loading and predicting the response of a structure is one of the
key responsibilities of a structural engineer. Loading conditions experienced by a
structure during its lifespan can be roughly divided into two major categories, static
and dynamic loading. Static loading is considered the everyday loading experienced
by the structure, including the dead and imposed loads. Estimation and prediction of
structural response against the static loads is straightforward, even with different
combinations of loads.
Explosive events mainly originate from terrorist attacks targeting civilian or
commercial structures. However, accidental events such as explosions in storage
facilities or gas explosions also occur from time to time. The severe nature of loading
results in catastrophic failures of structural elements and ultimately loss of life either
 5

direct or indirect effects of the explosion. Hence, it is essential to estimate and predict
the effects of explosions and provide designs to protect structures against the potential
explosive events.
A detailed review of explosive loading and structural response is presented in
this chapter. Different explosive types and their specific properties, comparison
methods of different explosives and prediction of blast pressure and impulse
parameters are discussed in terms of high explosives. These prediction methods are
also compared against the available code guidelines on blast pressure prediction.
Finally, structural response to blast loading and associated failure mechanisms is
assessed.
The development of a procedure that will give practical design solutions is
essential for the design of new or the retrofitting of existing structures so as to be able
to withstand the effects of explosive loads. The engineer needs to calculate the acting
forces according to a certain blast scenario, which includes the type and weight of the
used explosive, the distance from the structure and the geometry of the surrounding
area and the structure itself. These forces should then be applied on the structural
system at hand in order to design structural members, size and materials properties
that will ensure sufficient robustness of the building to survive the effects of the
computed actions.
In my thesis, an overview of a design procedure for structures under blast
loading is provided. The material presented has been collected from various sources
and mainly from UFC, DoD and FEMA. The analysis focuses on ways to estimate
blast loading on structures and, to a lesser extent, on their response under such types
of actions. Next, a short duration, lateral blast load was applied and the response of
the structure was calculated. The analysis and design of structures subjected to blast
loads require a detailed understanding of blast phenomena and the response of various
structural elements. This gives a comprehensive overview of the effects of explosion
on structures.

2.2. Blast Loading

The blast loading on structures can be divided into two main groups based
upon the confinement of the explosive charge. In this thesis, surface burst load of
unconfined explosion type are used because this type is most used by terrorist other
than explosion.
 6

2.2.1 Unconfined Explosions

This group can be divided into three types based upon the relative charge
location and the blast loading produced on the structures.
(a) Free Air Burst Loads
The blast wave propagates away from the center of the explosion
striking the structures without intermediate amplification of the initial shock
wave.
(b) Air Burst Loads
The explosion is located at a distance away and above the structure so
that ground reflections of the initial wave occur before the blast wave reaches
the structure.
(c) Surface Burst Loads
The explosion is located close to or on the ground so that the shock
wave is amplified at the point of detonation due to ground reflection. Only this
type of blast loads considered in this thesis.

Figure 2.1.Types of External Explosions and Blast Loadings; (a) Free-air Bursts, (b)
Air Bursts, and (c) Surface Bursts.
 7

2.2.2 Partially Confined Explosions

This can be divided into following.
(a) Exterior or Leakage Pressures
The detonation occurs near the ground surface and behind an
obstruction so that the shock wave is interfered with before reaching the
structure.
(b) Interior or High Pressure Loads
The detonation is located within or immediately adjacent to barrier
type structures and the blast pressure amplified due to their multiple reflection
by the structure as a result of the closeness of the structure to the explosion.

2.3. Blast Wave Phenomena

The violent release of energy from a detonation in a gaseous medium gives
rise to sudden pressure increase in that medium. The pressure disturbance, termed the
blast wave is characterized by an almost instantaneous rise from the ambient pressure
to a peak incident pressure (Pso). Incident pressure is the pressure on a surface
parallel to the direction of the blast wave. This pressure increase or shock travels
radially from the burst point with a diminishing velocity (U) which is always in
excess of the sonic velocity of the medium. Gas Molecules, making up the front move
at lower velocities (u). This latter particle velocity is associated with a dynamic
pressure or the pressure formed by the winds produced by the shock fronts and it’s a
function of air density and wind velocity. As the shock front expands into increasingly
larger volumes of the medium, the peak incident pressure at the front decreases and
the duration of the pressure increases.
At any point from the burst, the pressure disturbance has the shape shown in
Figure 2-2. The shock front arrives at time (tA) and, after the rise to the peak value,
the incident pressure decays to the ambient value in time (to) which is the positive
phase duration. This is followed by a negative phase with duration (to-) longer than
the positive phase and characterized by a pressure below the pre-shot ambient
pressure (maximum value of Pso -) and a reversal of the particle flow.
The incident impulse associated with the blast wave is the integrated area
under the pressure-time curve and is denoted as (Is) for the positive phase and (Is-) for
the negative phase as shown in Figure 2.2.
 8

Figure 2.2. Typical Blast-induced Pressure-time.

If the shock wave impinges on a rigid surface oriented at an angle to the

direction propagation of the wave, a reflected pressure instantly developed on the
surface. This pressure is a function of the pressure in the incident wave and the angle
formed between the rigid surface and the plane of the shock front. The duration of the
reflected pressure is controlled by the size of the reflecting surface. The peak positive
reflected pressure is denoted as (Pr), the peak negative reflected pressure is (Pr -), and
the unit- impulses associated with a completely reflected incident wave are (Ir) for the
positive phase and (Ir -) for negative phase as shown in Figure 2-3. When the shock
wave impings on a surface that is perpendicular to the direction of travel of the shock
wave, then the point of the initial contact will be subjected to the maximum reflected
pressure and impulse.

Figure 2.3. Complete Overpressure - time Profile.

 9

2.4. Blast Wave Scaling Law

One of the most critical parameters for blast loading computations is the
distance of the detonation point from the structure of interest. The peak pressure value
and velocity of the blast wave, which were described earlier, decrease rapidly by
increasing the distance between the blast source and the target surface, as shown in
Figure 2. In the figure only the positive phases of the blast waves are depicted, whose
durations are longer whenever the distance from the detonation point increases.
A comparison of different explosive events for blast wave/shock front
parameters is required when different types of explosives and different charge
weight/standoff distance combinations are used. In order to compare different charge
weight/standoff distances, the cube-root scaling law was introduced and for the
comparison of different explosive types, the TNT equivalence was proposed.
The Hopkinson-Cranz scaling law, known as cube-root scaling law
(Hopkinson, 1915, Cited in Ohashi et al. 2001), compare the different explosive
events with different explosive charges and standoff distances. As quoted by Smith
and Hetherington (1994) similar blast waves are produced at identical scaled distances
(Z) when two explosive charges of similar geometry and of the same explosive, but of
different weights, are detonated in the same atmosphere. The scaling parameter, Z, is
defined as the scaled distance, using the cube-root scaling law as shown in Equation
2.1.
The effect of distance on the blast characteristics can be taken into account by
the introduction of scaling laws. These laws have the ability to scale parameters,
which were defined through experiments, in order to be used for varying values of
distance and charge energy release. The experimental results are, in this way,
generalized to include cases that are different from the initial experimental setup. The
most common blast scaling laws are the ones introduced by Hopkinson-Cranz and
Sachs. The idea behind both formulations is that during the detonation of two charges
of the same explosive that have similar geometry but different weight and are situated
at the same scaled distance from a target surface, similar blast waves are produced at
the point of interest as long as they are under the same atmospheric conditions. Sachs
scaling is also suitable in the case of different atmospheric conditions. According to
Hopkinson-Cranz law, a dimensional scaled distance is introduced as described by
 10

R
Equation (2.1), Z = 1
W3
(2.1)
Where, R= the radial distance between the centre of explosion (the distance from the
detonation source to the point of interest [m/ft] ).W= the explosive weight in TNT
(the weight (more precisely: the mass) of the explosive [lb].)
Thus, suppose that an explosive charge of weight W1 and characteristic size
d1, situated at distance R1 from the point of interest, produces at this point a blast
wave of peak overpressure P, impulse i1, duration to1, with arrival time ta1 and that

R1
3 = λ .Then, what this scaling law implies is that a blast wave with the same peak
√W 1
overpressure P and similar form would be produced at this point by another explosive
charge W2 of characteristic dimension D2= λ d1, situated at distance R2= λ R1.
Further, at the given point due to W2 we would have: impulse i2= ɻ i1, duration to2=
λ to1, and arrival time ta2= λ t al. It is essential to underline that under this formulation
all distance and time parameters of a blast wave are scaled by the same factor λ but
pressure and velocity values remain unchanged at similarly analogous times.
It is essential to underline that under this formulation all distance and time
parameters of a blast wave are scaled by the same factor λ but pressure and velocity
values remain unchanged at similarly analogous times.

Table 2.1. Damage to Structures Vs. Scaled Distance (Source: Department of the US
Navy, 1999)

Scaled Peak Description of damage to a

distance presssure PS
+ ¿¿
light, unreinforced structure
1
𝑍 = 𝑅/W 3

6 29 psi Complete destruction

9 12 psi Severe structural damage, global collapse

10.5 8.4 psi Major structural damage, local collapse

 11

3.6 psi Damage costing 50% of replacement cost

18

2.3 psi Damage costing 20% of replacement cost

25
Table 2.1. (Continued-)
10% of replacement cost; 100% glazing
30 1.7 psi
broken
Minor structural damage; 60% glazing
40 1.1 psi
broken
Negligible structural damage; 30% glazing
50 0.8 psi
broken

2.4.1. Explosive Type and Weight

In the present thesis, the focus will be on building structures, as these have
proven to be the most common targets of terrorist attacks with the use of explosive
devices. Nevertheless, the procedure that should be followed in the case of different
structural elements is practically the same. The first step in designing a building to
sustain blast loading is the definition of the type and weight of the explosive for
which the design will be performed. Several types of explosives are available
nowadays, any of which could be used for conducting an attack against a structure. In
the majority of the cases solid explosives will be used in improvised explosive
devices (IED), because of their transportability, relatively easy manufacturing and the
possibility of their placement in vehicles that could be moved in the vicinity, adjacent
or within (e.g. underground garages) a building. The wide variety of explosives has
led to the adoption of a universal quantity, which is used for all necessary
computations of blast parameters. TNT (Trinitrotoluene) was chosen as its blast
characteristics resemble those of most solid type explosives. An equivalent TNT
weight is computed according to Equation (2.2) that links the weight of the chosen
design explosive to the equivalent weight of TNT by utilizing the ratio of the heat
produced during detonation:
H expd
W e = W exp d (2.2)
H TNT
where, W e is the TNT equivalent weight [kg],
W exp is the weight of the actual explosive [kg],
 12

H expd is the heat of detonation of the actual explosive [MJ/kg], and

H TNT d is the heat of detonation of the TNT [MJ/kg].
It is worth mentioning that approximately one third of the total chemical
energy of the explosive is released by detonation. The rest is released at a slower rate
as heat of combustion through burning of the explosive products mix with the
surrounding air. These values can be used for the calculation of the equivalent TNT
weight with the use of Equation (2.2).
In order to compare the blast wave parameters generated by different
explosive materials, a non-dimensional parameter is defined. TNT equivalence factor
is based on the specific energy of the explosive material. TNT itself has a specific
energy of 4520 kJ/kg (Smith and Hetherington, 1994) and a ratio of specific energy of
the explosive material to specific energy of TNT is defined as the TNT equivalency
factor. TNT equivalency factors and specific energies for several explosives materials
are shown in Table 2.2.
Table 2.2 shows some predetermined TNT equivalent weight factors. These
factors can be used to determine the weight of TNT that produces the same blast wave
parameters as the ones from another explosive of certain weight. The comparison of
these blast wave parameters can be done either for pressure or impulse values, so the
following table contains two factors depending on the used method.

Table 2.2. TNT Equivalent Factors and Specific Energies for Several Explosive
Materials (Source: United States Army Corps of Engineers, 1986)

Mass specific
Explosives TNT equivalent (CF)
energy (kJ/kg)

TNT(Trinitrotoluene) 4520 1

GDN 7232 7232/4520=1.6

Pyroxilene 4746 1.05

Pentrinite 6689.6 1.48

Dynamite 5876 1.30

Schneiderite 3164 0.7

Dinitrotoluene (DNT) 3164 0.7

 13

Ethylenedinitramine 5650 1.25

Compound B[0.6RDX+0.4TNT] 5190 1.148

RDX(cyclonite) 5360 1.185

Table 2.2. (Continued-)

HMX 5650 1.256

5660 1..25
Semtex
5103.08 1.129
Dentolite 50/50

DENT 5795 1.282

Amatol 80/20
2650 0.586
Composition B (60/40)
5190 1.148
HMX
5680 1.256
Lead Azide 1540 0.340
Mercury Fulminate
1790 0.395
Nitroglycerine (Liquid)
6700 1.481
PETN
5800 1.282
Pentolite 50/50 5110 1.129
Tetryl 4520 1.000
Blasting Gelatin 4520 1.000
Nitroglycerine Dynamite
2710 0.600

The weight of an explosive is usually estimated by taking into account a

relevant attack scenario,which would involve a vehicle-borne or a personnel-borne
improvised explosive device. Clearly, the larger the used vehicle that could be
directed towards a structure, the larger the weight of the explosives it could carry
leading to higher equivalent TNT weight values. In Table 2.3 an estimate of the
quantity of explosives that could be transported by various vehicle types is presented.
The engineer, following the relevant regulations and in consultation with the building
owner, should decide on the type of explosive and size of vehicle that could be used
 14

for transportation, so as to be able to compute the equivalent weight of TNT for which
the structure should be designed. Due to a variety of such uncertainties, it is
recommended to apply a safety factor to the charge weights and augment them by
approximately 20%.

Table 2.3. Indicative TNT Equivalent Mass Factors.

Both explosive sand bombs are categorized as small, medium and height or
large.
(a) Small explosives devices --------up to 5kg TNT
(b) Medium explosives devices-------up to 20 kg TNT
(c) Large explosives devices---------up to 100kg TNT
(d) Very large explosives devices and bombs----up to 2500 kg TNT
These explosives or bombs are sometimes given in lb TNT or ton TNT
designated as yield.
When the high explosive is other than TNT , the equivalent energy is obtained

R
by using the charge factor. This will form the value of the scaled distance Z = 1 .
W3
The charge factor is equal to

actual mass of the charge

(2.3)
mass of TNT equivalent

2.4.2. Blast Wave Reflection

 15

Once the blast wave is generated by the detonation of an explosive, it

propagates away from the source at supersonic speeds, until it meets obstacles. When
the blast wave encounters an object in its path, then reflection, refraction and
diffraction occurs similar to a sound wave travelling through an air medium.
The occurrence of reflection, diffraction or refraction depends on the physical
and geometrical properties of the obstruction. When the blast wave strikes
perpendicular to the object, reflection occurs with increased pressure, density and
temperature. The maximum increase of these parameters is observed when the
obstruction has infinite dimensions perpendicular to the blast wave travel direction.
These increases in physical parameters are directly due to the interaction of the
original wave and the reflected wave. The overpressure measured after the reflection
is known as the reflected overpressure (Pr), which is several times higher than the
incident overpressure.
When the blast wave encounters a finite object in its path, diffraction may
occur at the edges of the obstruction. In particular when a 3-dimensional object is
exposed to blast waves, diffraction may reduce the effects of blast pressure on the side
walls as shown in Figure 2.4. The front of the rectangular structure will encounter the
maximum blast pressure while side walls and back wall will encounter much lower
incident pressure.

Figure 2.4. Interaction of Blast Wave with an Object with Finite Dimensions

The magnification factor for the peak reflected pressure (Peak reflected
pressure (Pr) / Peak positive incident pressure (Pso+)) depends on the rigidity of the
 16

obstacle and incidence angle (α a). The incident angle is defined as the angle between
the blast wave propagation direction and the normal of the facing element of
obstruction. When the obstruction is rigid and perpendicular to the blast wave, (α a=0o)
(i.e. Front wall, Figure 2.9) maximum magnification will occur and when the
obstruction is parallel to blast wave (αa=90o) (i.e. Side walls, Figure 2.9)
magnification will be equal to unity.
Reflection of blast waves can be classed either normal or Mach reflections
depending on the incidence angle. Mach reflection is a supersonic shockwave effect
observed when the shock wave propagates over a solid edge, which involves the
formation of a triple point reflection. It has been found that Mach reflections can
occur only in the case of incident angle greater than 40 o and when the incident angle
is less than 40o, normal reflection will occur.
The blast winds generated by the dynamic pressure also interact with the
obstructions in its path in the same way as normal winds on structures, resulting in
drag pressures. Drag pressures on a surface of an object depend on the dynamic
pressure and the drag coefficient for that specific surface. Typical drag coefficients
used for a rectangular structure are shown in Figure 2.5.

Figure 2.5. Drag Coefficients for a Rectangular Building Subjected to Explosive

Loading

The interaction between an object and a blast wave generates a pressure

pattern which is different than the idealized time history presented in Figure 2.7. As
the blast wave travels through space, decreasing in speed and peak pressure value, it
encircles every object/structure that lies within its range. The load that has to be
withstood by a structure depends on various parameters, such as the type and weight
of the explosive charge, the distance of the detonation point, the structure’s geometry
and type, the interaction of the wave with the environment and the ground, etc. When
 17

the blast wave comes to contact with a rigid surface the pressure that is reflected is
larger than the incident peak pressure Pso . The reason for this rise is attributed to the
nature of the propagation of the blast wave through the air. While the wave travels, it
moves along air particles that collide with the surface upon arrival. In an ideal linear-
elastic case the particles should be able to bounce back freely leading to a reflected
pressure equal to the incident pressure, and thus the surface would experience a
doubling of the acting pressure. In a strong blast wave, which as a shock wave is a
non-linear phenomenon, the reflection of these particles is obstructed by subsequent
air particles that are transferred there, thus leading to much higher reflected pressure
values. In this case the surface would experience an acting pressure much higher than
the incident one.
Clearly it is this reflected pressure to be used for design. Figure 2.11 shows the
difference between the incident and the reflected pressure in an infinite surface. As
just noted, the reflected pressure can be several times larger than the incident
pressure, depending on the geometry of the structure, the type, size, weight and
distance of the explosive as well as the interference of other obstacles between the
detonation point and the structure.

Figure 2.6. Incident, Reflected and Dynamic Pressure Time Histories.

Three types of reflection can take place depending on the angle of the
reflecting surface with the propagation direction of the blast wave. The most severe
 18

case, in terms of loading values, is when a surface is perpendicular to the direction of

the wave, during which normal reflection occurs. When the propagation direction of
the wave intersects at a small oblique angle with the surface it causes the creation of
an oblique reflection, whereas the third case is linked to a phenomenon known as
Mach stem creation, which occurs whenever the wave impinges on a surface at a
specific angle, as will be explained below.
Clearly, in all cases the reflected pressure is always greater than the incident
pressure. As mentioned earlier, the peak incident pressure is also referred to as the
peak side-on overpressure because it is equal to the reflected pressure on a surface
that is parallel to the direction of the blast wave. Figure 5 illustrates the difference
between the reflected and the side-on overpressures in the case of a free air explosion
without wave amplification. The value of the reflected pressure from the surface
becomes maximum at the point of normal distance RA between the detonation source
and the surface. As expected, the reflected pressure’s value decreases (however not
monotonically!) as the angle of incidence α increases. Its minimum value is equal to
the incident pressure and is created on surfaces perpendicular to the shock front
(α=90o). Equation (2.4) can give the value of the peak reflected overpressure in the
case of reflections at zero angle ,

4 Pso +7 Po
Pr =2 PSO
P so+ 7 P o
(2.4)

where, Pso= incident peak overpressure and

Po= ambient pressure.
For this equation it was assumed that the explosion takes place at standard sea
level conditions and that the ratio of specific heats of gases is equal to 1.4. According
to Equation (2.4) the maximum and minimum reflected pressures could be obtained
by letting Pso→∞ and Pso→0, respectively. This substitution leads to reflected
pressures values ranging from 2 to 8 times the incident pressure for small and large
shocks, respectively. Nevertheless, experimental investigations have concluded that
the reflected pressure can be several times higher than 8 times the incident pressure.
This is attributed to the fact that under severe blast loads the air does not behave in an
ideal manner, which is an underlying assumption of Equation (2.4). It also indicates
 19

that the ratio of the normal maximum reflected pressure and the incident pressure is
not constant, but depends on the value of the incident pressure.

Figure.2.7. Definition of Reflected Pressures for a Cube Type Structure.

As specified, Equation (2.4) is valid in the case of normal reflection. If there is

an angle α (angle of incidence) between the wave propagation direction and the
affected surface, the process of reflection may be quite different. This angle affects
the resulting reflection and, consequently, the blast loading on a structure. Figure 2.7
shows the influence of the peak incident overpressure Pso on the peak reflected
pressure Prα versus the angle of incidence α, expressed in terms of the reflection
coefficient C rα, which is defined as

Prα
C rα=
P SO
(2.5)

The effect of the angle of incidence may be neglected and the structure is
studied under the normal reflected pressure, which is on the safe side for most cases,
especially if the incident overpressure is large. More specifically, from the curves of
the diagram it can be concluded that whenever the angle of incidence is
approximately less than 40’, the use of the normal reflected pressure leads to
conservative design as the normal reflected pressure is bigger (up to 25%). For small
 20

to moderate values of peak incident pressures (0.01-3.50 MPa) the reflected pressure
at angles of incidence between 40o-55o can be underestimated if the assumption of a
normal reflection has been made.The reason for this reflected wave behaviour for
angles of incidence between 40o-55o is attributed to the creation of a Mach stem. As
already discussed, the blast wave due to an air detonation above the ground is made
up of an incident wave, emanating from the explosive charge, and of a reflected wave,
which is produced from the reversal due to the impingement of the wave to the
ground. For shallow incident angles with the ground (up to 40 o) the incident wave is
ahead of the reflected wave produced by the surface and typical reflection occurs. For
larger angles though, coalescence between the incident and the reflected wave takes
place, creating a Mach stem. This coalescent wave can in some cases be substantially
larger than the normal reflection pressure values. The point of intersection of the
incident, reflected and Mach waves is known as the triple point. The Mach stem is
considered to have a constant value throughout its height, even though some small
variations may exist.
The case of an air burst is schematically examined where the blast wave is
reflected from the ground before it hits the structure. The distance of the detonation
point from the ground is much smaller than the distance from the obstacle (structure),
thus leading to the creation of a Mach stem on the ground. The Mach front pressure-
time history is similar to the incident pressure, but its value is larger. This results in
situations where structures:
(a) can either be loaded by uniform pressure, if the Mach front height is larger
than the height of the structure, or
(b) can be loaded uniformly by the Mach front over their lower part (bellow
triple point) and by the separate action of the incident and reflected pressures
over their upper part (above triple point).
The corresponding incident overpressure profiles are schematically shown in
the middle part of Figure 2.8. Of course, for design purposes this second loading case
can be substituted by applying the uniform pressure of the Mach front (whose value is
larger) all-over the structure’s height. Design manuals such as contain diagrams for
the calculation of the Mach pressure, impulse and height according to the distance and
weight of the blast charge from the structure under study. At the bottom part of Figure
2.8.diagram for the calculation of the triple point’s height is presented.
 21

Similar to Figure 2.8 for Prα, the peak reflected impulse values i ra, when the
angle of incidence is greater than 90’, can be conveniently calculated and used. The
relevant curves, which are strictly valid for free-air bursts and spherical waves, are
shown in Figure 2.8(a) and 2.8(b).
(a)

(b)

(c)
 22

Figure 2.8. (a) Mach Stem Creation over a Horizontal (Ground) Surface and Loading
of a Vertical Surface; (b) Profile of Incident Overpressure within the Mach Front
Height, and of that above the Triple Point, Respectively; (c) Estimation of Mach Front

HC HG
Height H T from the Scaled Charge Height 1 and Scaled Horizontal Distance 1 .
3 3
W W
2.4.3. Surface Burst and Loading
A common case for a terrorist attack bombing scenario is the use of a vehicle
loaded with explosives that is detonated by using a remote control or time trigger.
What is special about this type of explosion, which is of practical importance for
designers, is the fact that due to the closeness to the ground there is an immediate
interaction between the ground and the blast wave. This is the so-called hemispherical
surface burst. Instead of the creation of a Mach front at a certain distance from the
point of blast initiation, the incident wave is reflected immediately from the ground
leading to higher pressure values.
If the ground were a rigid surface, the generated pressure would be twice that
produced by the same charge under free-air burst conditions. This observation allows
the use of the pressure relationships derived for free-air blasts also in the case of
surface bombings, by doubling the original charge weight. In reality, depending on
the soil and explosive type, some kind of energy absorption takes place from the
ground by the creation of a crater, and thus the multiplicative factor applied to the
charge is approximately 1.7 to 1.8. The generated wave has a hemispherical shape,
and its characteristics resemble those of the Mach front wave that was described
earlier. For a sufficient large stand-off distance this wave can be considered locally as
plane, and the pressure loading of a structure can, for practical purposes, be
considered as uniform.
In appendix apart from the pressure on the front side, the pressures applied at
the top of the structure are also schematically shown. When the blast wave comes to
contact with an obstacle it eventually engulfs it, thus exerting pressures to all of its
exposed sides. In the case of a stand-alone building the blast wave first hits its front
face and subsequently it diffracts around the rest of the structure, loading the roof, the
side and the rear walls. As was pointed out earlier, the roof and the side walls are
parallel to the blast wave propagation direction, which means that they are loaded
according to the incident pressure, (with a reduction due to the negative drag
 23

pressure). In order to calculate the pressure to be applied at the side walls and the top
of the structure a procedure is proposed in which is based on an equivalent load factor
C E and a drag coefficient C D . These factors are computed as functions of the
wavelength and essentially they average the applied non-uniform pressure over the
length of the surface. In case that the side walls are not parallel but at an angle to the
blast wave propagation direction, the procedure proposed for the front walls should be
adopted.
Concerning the rear wall, as the pressure wave passes the edges of the side
walls and the roof, secondary waves are created which are responsible for loading the
back of the structure. For design purposes the calculation of the pressures acting at the
rear wall is essential for the determination of the overall drag force acting on the
structure. Usually the same formulation used for the side walls and the roof is applied,
and involves the peak pressure value acting at the back edge of the roof slab, as will
be shown below.

(a)

(b)
 24

Figure 2.9. Pressure Acting on a Building due a Surface Burst (a) at a Close Distance,
(b) at a Distance Sufficient for the Impinging Wave to be Consider as Plane.
2.4.4. Effect of Finite Reflecting Surface
The reflected curve shape shown in Figure 2.10 is valid in the case of a blast
wave acting on an infinite surface. For a single building though, the reflected pressure
from the front of the structure is eventually relieved by the leakage of the wave
around its sides and roof regardless of the type of blast (surface burst or free-air). This
phenomenon allows the reflected pressure to be reduced at a fast rate and finally to
become equal to the sum of the side-on and the dynamic pressure. The time required
for the reflected pressure to drop to this lower level is known as clearing time t C .
Figure 2.10 depicts the difference between the reflected pressure curves in the
case of a finite and an infinite surface. The peak reflected pressure is identical for both
cases, but the curves follow a different path during their decay part. The clearing time
is also shown according to the definition presented above. The positive phase duration
in the case of a finite surface is generally smaller than an infinite one, as in the former
the pressure values drop faster due to the wave’s propagation over the sides/roof of
the surface. The immediate impact of this effect is the reduced total force at the
surface of the obstacle since the impulse, which is computed as the integral of the
pressure-time diagram, is considerably smaller. As will be seen below, the clearing
time is not only dependent on the geometry of the obstacle, but also on the charge
type and weight and the detonation point’s distance from the building (stand-off
distance).
 25

Figure 2.10. Influence of Surface’s Finite Dimensions on the Reflected Pressure

Time-History.
2.5. Dynamic Pressure
The blast wave propagates through the air with a continuously decreasing
speed which is larger than the speed of sound. The air behind the front of the blast
wave moves also along the same direction as a wind but with smaller velocity. These
winds behind the blast wave front are responsible for loading a surface for the whole
duration of the positive phase and also for a small time afterwards. The pressure that
is produced is known as dynamic or drag pressure q(t), it has an initial peak value qo,
which is less than that of the incident or reflected pressures for medium and small
overpressure values, and it eventually dies out. The difference is that both incident
and reflected pressures last only for a very short time (usually far less than one
second), while the dynamic pressure may last for longer periods (up to 2-3 seconds).
Figure 2.11. shows the variation of the peak dynamic pressure as a function of
peak incident pressure values. The dynamic pressure depends on the density of the air
and the wind velocity behind the front of the blast wave, all of which are influenced
by the peak incident overpressure. Its effective value further depends, through a drag
coefficient, on the orientation of the surface with respect to the blast wind direction.
 26

Figure 2.11. Variation of Peak Dynamic Pressure qo Versus Peak Incident

Pressure.
2.6. Calculation of Blast Parameters
Blast parameter calculation are based on TNT equivalent and standoff
distance.

2.6.1. Blast Pressure Determination

There are various relationships and approaches for determining the incident
pressure value at a specific distance from an explosion. All the proposed relationships
entail computation of the scaled distance, which depends on the explosive mass and
the actual distance from the center of the spherical explosion.
Kinney presents a formulation that is based on chemical type explosions. It is
described by Equation (2.6) and has been used extensively for computer calculation
purposes,
 27

Z 2
808 [1+(1+ )]
4.5
PSO = PO 2 2
Z Z Z 2 0.5
(
{ 1+
0.08 )( 1+
0.32 )( 1+
1.35)}

(2.6)

where Z (m/kg1/3) is the scaled distance,

Po is the ambient pressure.
Other relationships for the peak overpressure for spherical blast include those
of Brode . They depend on the magnitude of the explosion, Equation (2.7) is valid
where the peak overpressure is over 10bar (=1MPa) (near field explosions) and
Equation (2.8) for pressure values between 0.1 bar and 10 bar (0.01MPa-1MPa)
(medium and far-field explosions). The scaled distance is measured in m/kg1/3 and
the pressure Pso in bars,

6.7
PSO = +1 for PSO > 10 bar
Z3
(2.7)

0.975 1.455 5.85

PSO = + + 3 - 0.019 for 0.1 < PSO < 10 bar
Z Z2 Z
(2.8)

Another formulation, that is widely used for computing peak overpressure

values for ground surface blast has been proposed by Newmark and does not contain
categorization according to severity of the detonation:
W W
PSO = 6784
R 3 + 93
R3 √
(2.9)
where, Pso is in bars,
W is the charge mass in metric tons of TNT
R is the distance of the surface from the center of a spherical explosion in m.
 28

Mills have also introduced an expression of the peak overpressure in kPa, in

which W is expressed in kg of TNT and the scaled distance Z is in m/kg1/3, which
reads:
1772 114 108
PSO = - +
Z3 Z2 Z
(2.10)
The most widely used and accepted approach for the determination of blast
parameters is the one proposed by Kingery-Bulmash. Their paper includes
formulations for both spherical (free air bursts) and hemispherical pressure waves
(surface bursts) and provide the values of incident and reflected pressures as well as
of all other parameters. The proposed blast parameters are valid for distances from
0.05 m to 40 m as the diagrams included in are referred to 1kg of TNT.
The most widely used and accepted approach for the determination of blast
parameters is the one proposed by Kingery-Bulmash. Their paper includes
formulations for both spherical (free air bursts) and hemispherical pressure waves
(surface bursts) and provide the values of incident and reflected pressures as well as
of all other parameters. The proposed blast parameters are valid for distances from
0.05 m to 40 m as the diagrams included in are referred to 1kg of TNT.

Figure 2.12. Comparison of Curves of Peak Incident Overpressure Versus Scaled

Distance for both Free-air Bursts (Spherical Waves) and Surface Bursts
(Hemispherical Waves)
Techanical report ARBRL-TR-02555 also includes a full set of analytical
relationships providing the above blast parameters in terms of polynomial functions of
 29

the logarithm of the scaled distance. These relationships can be readily programmed
in appendix, show the diagrams of blast parameters for the positive phase of the blast
wave for both free-air and surface bursts. They are overall more comprehensive and
the curves have been drawn with respect to scaled distances from Z=0.05 m/kg1/3 to

m
Z=40 . From these diagrams in order to obtain the absolute value of each
kg1 /3
1
parameter, its scaled value has to be multiplied by a factor W 3 so as to take into

account the actual size of the charge. Clearly, as mentioned above for the Hopkinson-
Cranz scaling law, pressure and velocity quantities are not scaled.
The additional symbols stand for: U = shock wave speed (m/ms) and Lw =
blast wavelength (m). This latter can be defined, for a point at a given standoff
+¿¿
distance at a particular time instant, for LW as the length which experiences positive
−¿ ¿
pressure (or, negative pressure for LW ).
Figure 2.13 shows the diagrams for blast parameters of the negative phase of
the blast wave for both free-air and surface bursts for TNT charges at sea level. These
parameters are important if the behaviour of the structure to the whole blast
phenomenon is necessary.

Figure 2.13. Reflected Wave Pressure-Time History (Adapted from UFC 3-340-02,
2008)
 30

2.6.2. Blast Load and Modelling

These calculations are described follows.
(a) External Blast and Modelling
(b) Duration
The overpressure Pso in ots peak form acting dynamically on walls or front
faces of a building can be written as (In 1961, Newmark and Hansen introduced a
relationship to calculate the maximum blast pressure(Pso), in bars, for a high
explosive charge detonates at the ground surface as;

1
W W 2
Pso=6784 + 93 ( ¿ ¿
R3 R3
(2.11)

Where, W= total charge weight inTNT (tonne)

R= stand-off distance to the center of the detonation in m

For example;
W= 1.1 ton
R = 90 m
Pso = 0.1244 bar
2.5C D Pso
q o= peak drag =
7 Pso+ P o
P R= peak pressure of a reflected shock
6
= 2 Po + when Pso ≥ 10 bar
7 Pso+ Po
Where , Po = ambient atmospheric pressure ≈ 1 bar
C D = drag coefficient = 1.o in general
In detail , C Dr = drag coefficient for the rear face = .25 to 0.5
C Df = drag coefficient for the front surface = 0.8 to 1.6

Pressure-Time Relation
Notation
 31

3S B
T C= , S= H s ≤ T C= reflected pulsetime
U 2
(2.12)
S= H s when B≥ H
Where, Pf =pressure at front face
Pr =pressure at rear face
B= building width
H s =building height
2i r
T B=time for reflected wave =
PR
i r=total reflected pressure impulse
Duration
1
3
A=t B=shock pulse duration = 10.23 W PSO <¿ 70 bar (2.13)
√ P SO
1
3
= 20.77 W PSO ≥ 70 bar (2.14)
√ P SO
W 13
A=t D =time of drag = 20.77 ( ) for PSO <¿ 2 bar (2.15)
PSO
2.6.3. Basic Parameters of a Bomb Blast
The analysis for blast effects include the following:
(a)Computation of transient overpressure from the blast wave front
(b)Reflected shocks to evaluate peak pressure loading and total impulse
In case of buildings, the following parameters need urgent attention:
(a) The magnitude of the explosion in TNT
(b) The stand-off distance and altitude
(c) The geometry of a building structural system
(d) The structure orientation with respect to the explosive and the ground.
Generally, for the design purposes a simplified blast-structure interaction shall
be developed depending upon the building location and orientation (front, rear, sides,
and roof) relative to the point of burst.
In terms of incident pressures and phase durations the following two cases
define the blast loadings:
 32

+¿
(a) Peak positive incident (side-on) pressure PSO ¿ positive phase duration
t +¿ +¿
O ¿nd the corresponding positive incident impulse i S ¿ .
−¿
(b) Peak negative incident pressure PSO ¿(suction), negative phase duration
t −¿ −¿
O ¿, and the corresponding negative incident impulse i S ¿ .

2.7. Calculation of Pressure Loads on Building Surfaces

Blast pressure loads are acting on the building according to the type of
blasting. Front wall are much acted by blast pressure other than surface.
2.7.1. Blast Wave-Structure Interaction
A structure subjected to an external explosion will eventually be engulfed by
the blast wave and normal forces (pressures) will be applied to its exposed surfaces.
The above sections described the basic characteristics of a blast wave and the way to
calculate important values for design, such as its peak incident and reflected
pressures. Apart from these values, the response of a structure under explosion
loading depends also on various other factors such as the shape of the structure, the
relative location of the blast from the structure, the geometry of the area between the
structure and the detonation point, the natural period of the structure, its openings etc.
The fundamental (or equivalent, if a single-degree-of-freedom is considered)
natural period T n of the structure and its relation to the positive phase duration t O of
an excitation play an important role on its behavior. If the duration of the excitation is
longer than the natural period T n of the structure, then its maximum deflection takes
place before the excitation ceases, and dynamic analysis should be used, as often is
the case for earthquake loading. If the duration of the positive phase of the excitation
is a lot shorter than the natural period (< T n/4) of the structure, then one deals with an
impulsive loading. For a particular structure and load type (e.g. pressure), these
considerations can be schematically depicted, by an iso-response curve in a so-called
pressure-impulse (P-I) diagram. Thus, for very short duration loads (relative to T n) the
structural response is sensitive to the associated impulse and not to the maximum
pressure (vertical asymptote). However, as the load duration increases to quasi-static,
the response starts becoming sensitive to the maximum value of pressure and not to
impulse (horizontal asymptote).
The majority of the structures has a natural period that is far greater than the
duration of the positive phase of the blast wave, which is of the order of milliseconds.
 33

These structures barely experience a deflection before the complete passage of the
blast wave, so they are loaded according to the impulse value of the wave’s time
history. For this reason these structures are usually considered as rigid throughout the
positive and negative phase of a blast wave and the analysis can be performed as
decoupled. This means that the structural response is calculated without considering
the interaction between the blast pressure loading and the deforming structure. In this
study, surface blasting type are considered and pressure acting on the front wall are
calculated because front wall of this building are faced with blasting point.
2.7.2. Pressure on the Front Wall
The first surface that will be loaded from the blast wave is the building’s front
façade. In UFC 340-02 a procedure is proposed for the calculation of the load that has
to be sustained by the front face, but an assumption is made that the loaded face is
within the region of the Mach stem whose height exceeds the total height of the
building. For surface bursts this means that the detonation point is situated at such a
distance from the structure that the blast wave front can be considered as plane. The
first step for the definition of the total impulse applied to the front face of the building
is the calculation of the peak reflected pressure Pr at time t = t A when the blast wave
arrives at the surface. The pressure at the front face starts decreasing as the wave
moves past the surface surrounding the rest of the structure within the clearing time t C
. At time t A+ t C the pressure at the front surface will be substantially dropped (but still
remaining higher than the initial ambient pressure). After time t A+ t C the pressure is
the result of the combined effect of the incident and the dynamic pressure and is
provided by Equation (2.16),

P = PS + C D q (2.16)
where, Ps is the incident pressure,
C D is the drag coefficient, taken equal to C D =1 for the front wall, and
q is the dynamic pressure.

As mentioned above, t Cis known as the clearing time and various formulations
exist for its calculation. According to [8] it can be estimated by Equation (2.17)
 34

4S
tC =
( 1+ R ) C R
(2.17)
where, S is the smallest of the surface’s height H or half width W/2 ,
Cr is the sound velocity in the reflected medium, shown in Figure 17, and
R is the ratio of S/G, where G is the largest of the surface’s height H or
the half width W/2.
After the time t A+ t C the pressure values at the front face of the structure
continue to decrease to finally reach the value of the ambient pressure. In reality the
form of this decreasing curve is practically exponential, as has already been discussed
during the description of the ideal blast wave characteristics. In order to simplify the
analysis the decrease of the incident and the dynamic pressure can be assumed to be
linear and of triangular shape.
According to this approach the actual duration to of the positive phase of the
incident wave is replaced by a fictitious time t of (t of < to) that depends on the positive
impulse and peak pressure, as described by Equation (2.18).

2 is
t of = (2.18)
P so

where, 2 i s is the impulse value of the positive phase of the blast wave and Pso is the
peak incident pressure.
A similar procedure can be applied for the negative phase of the blast wave
−¿¿
thus defining a fictitious time t of by employing the corresponding impulse and peak
pressure values. The introduction of the fictitious time to produce simpler triangular
pulses results in a time gap between the end of the assumed positive phase and the
beginning of the negative.
 35

Figure 2.14. Substitution of Actual Incident Pressure Curve by Triangular Pulses and
Definition of Relevant Fictitious Times
The triangular shape assumption of the blast wave time history is not always
accurate, especially whenever high pressure values are involved. From equation
(2.18) it is deduced that computation of the clearing time depends on the geometry of
the impinged surface and the sound velocity in the reflected region, which in turn
depends on the amplitude of the incident wave and not on the intensity of the
explosion. Experiments have shown that whenever high pressure values are present,
the positive phase duration of the blast wave is extremely short, resulting in a
fictitious time t of that is shorter than the clearing time t C. In this case the
aforementioned procedure will result in a triangular diagram that will not represent
the true reflected pressure time history. The fact that the clearing time will not occur
means that the reflected wave pressure time history should correspond to that applied
on an infinite surface. For this reason an additional curve is usually constructed,
which uses the total reflected pressure impulse ir assuming a normal reflection. For a
given Pso the reflected impulse i r can be derived from the diagrams from appendix
and the fictitious duration t rf for the reflected wave is computed through Equation
(2.19).

2i r
t rf = (2.19)
Pr

where, i r is the total reflected impulse

Pr is the peak reflected pressure.
 36

Figure.2.15. Triangular Assumption of Pressure Time History on the Front

Face of a Structure
Thus for the positive phase of the reflected pressure two curves Pr -t are
constructed and compared: one corresponding to infinite surface conditions, and
another derived using the assumption that the finite surface geometry influences the
value of the reflected pressure. The curve to be used for loading the structure is the
one that produces the smallest impulse value.
2.8. Architectural, Landscaping and Blast Mitigation in Structure

Unlike other mitigation measures that improve the resistance of buildings and
infrastructure to disasters, protective and control measures focus on protecting
structures by deflecting the destructive forces from vulnerable structures and people.
Ideally, a potential terrorist attack is prevented or pre-empted through intelligence
measures. If the attack does occur, physical security measures combine with
operational forces (e.g., surveillance, guards, and sensors) to provide layers of defense
that delay and/or thwart the attack. Deception may be used to make the facility appear
to be a more protected or lower-risk facility than it actually is, thereby making it a less
attractive target. Deception can also be used to misdirect the attacker to a portion of
the facility that is non-critical. As a last resort, structural hardening is provided to save
lives and facilitate evacuation and rescue by preventing building collapse and limiting
flying debris.

Because of the interrelationship between physical and operational security

measures, it is imperative for the owner and security professional to define, early in
 37

the design process, what extent of operational security is planned for various threat
levels. If properly implemented, physical security measures will contribute toward the
goals listed below in prioritized order. In this thesis, third layer of defense called
structural hardening is emphasized to protect terrorist attack.