Properties of Dynamically Loaded Soils 135

Field Test Measurements

4.10 Reflection and Refraction of Elastic Body

Waves —Fundamental Concepts
When an elastic stress wave impinges on the boundary of two layers, the wave is
reflected and refracted. As has already been discussed in Chapter 3, there are two
types of body waves — that is, compression waves (or P–waves) and shear
waves (or S–waves). In the case of P–waves, the direction of the movements of
the particles coincides with the direction of propagation. This is shown by the
arrows in Figure 4.37a. The shear waves can be separated into two components:

a. SV-waves, in which the motion of the particles is in the plane of

propagation as shown by the arrows in Figure 4.37b
b. SH-waves, in which the motion of the particles is perpendicular to the plane
of propagation, as shown by a dark dot in Figure 4.37c

If a P-wave impinges on the boundary between two layers, as shown in

Figure 4.38a, there will be two reflected waves and two refracted waves. The
reflected waves consist of (1) a P-wave shown as P1 in layer 1 and (2) an SV-
wave shown as SV1 in layer 1.

Figure 4.37 P-wave, SV-wave, and SH-wave

136 Chapter 4

Figure 4.38 Reflection and refraction for (a) an incident P-ray; (b) an incident SH-ray;
and (c) an incident SV-ray

The refracted waves will consist of (1) a P-wave, shown as P2 in layer 2,

and (2) an SV-wave, shown as SV2 in layer 2. Referring to the angles in Figure
4.38a, it can be shown that

α1 = α2 (4.37)
and
sin D1 sin D 2 sin E 2 sin D 3 sin E3
(4.38)
X p1 X p1 X s1 X p2 X s2

where
 Properties of Dynamically Loaded Soils 137

X p1 and X p2 = the velocities of the P-wave front in layers 1 and 2, respectively

X s1 and X s2 = the velocities of the S-wave front in layers 1 and 2, respectively

If an SH-wave impinges the boundary between two layers, as shown in

Figure 4.38b, there will be one reflected SH-wave (shown as SH1) and one
refracted SH-wave (shown as SH2). For this case

β1 = β2 (4.39)

and

sin E1 sin E3
(4.40)
X s1 X s2

Lastly, if an SV-wave impinges the boundary between two layers, as shown

in Figure 4.38c, there will be two reflected waves and two refracted waves. The
reflected waves are (1) a P-wave, shown as P1 in layer 1 and (2) an SV-wave,
shown as SV1 in layer 1. The refracted waves are (a) a P-wave, shown as P2 in
layer 2, and (b) an SV-wave, shown as SV2 in layer 2. For this case, β1 = β2:

sin E1 sin D 2 sin E 2 sin E3 sin D 3

(4.41)
X s1 X p1 X s1 X s2 X p2

The mathematical derivations of these facts will not be shown here. For
further details the reader is referred to Kolsky (1963, pp. 24–38).

4.11 Seismic Refraction Survey (Horizontal

Layering)
Seismic refraction surveys are sometimes used to determine the wave
propagation velocities through various soil layers in the field and to obtain
thicknesses of each layer. Consider the case where there are two layers of soil, as
shown in Figure 4.39a. Let the velocities of P-waves in layers 1 and 2 be X p1
and X p2 , respectively, and let X p1 < X p2 . A is a source of impulsive energy. If
seismic waves are generated at A, the energy from the point will travel in
hemispherical wave fronts. Consider the case of P-waves, since they are the
fastest. If a detecting device is placed at point B, which is located at a small
distance x from A, the P-wave that travels through the upper medium will reach it
first before any other wave. The travel time for this first arrival may be given as
138 Chapter 4

Figure 4.39 Seismic refraction survey — horizontal layering

x
t= (4.42)
X p1

where AB = x.
Again, consider the first arrival time of a P-wave at a point G, which is
located at a greater distance from A. In order to understand this, one considers a
spherical P-wave front that originates at A striking the interface of the two layers.
At some point C, the refracted P-wave front in the lower medium will be such
that the tangent to the sphere will be perpendicular to the interface. In that case,
the refracted P-ray (shown as P2 in Figure 4.39a) will be parallel to the boundary
 Properties of Dynamically Loaded Soils 139

and will travel with a velocity X p2 . Note that because X p1 < X p2 , this wave front
will travel faster than those described previously. From Eq. (4.38)

sin D1 sin D 3
X p1 X p2

Since D 3 = 90°, sin D 3 = 1, and

È X p1 Ø
α1 = sin–1 É Ù = αc (4.43)
Ê X p2 Ú

where αc = critical angle of incidence.

The wave front just described traveling with a velocity X p2 will create
vibrating stresses at the interface, and this will generate wave fronts that will
spread out into the upper medium. These P-waves will spread with a velocity
of X p1 . The spherical wave front traveling downward from D in layer 2 will have
a radius equal to DE after a time Δt. At the same time Δt, the spherical wave
front traveling upward from point D will have a radius equal to DF. The resultant
wave front in the upper layer will follow a line EF. It can be seen from the
diagram that

X p1 't DF
sin ic (4.44)
X p2 't DE

So ray DFG will make an angle ic with the vertical. It can be mathematically
shown that for x greater than a critical value xc, the P-wave that travels the path
ACDG will be the first to arrive at point G. Let the time of travel for the P-wave
along the path ACDG be equal to t. Thus, t = tAC + tCD + tDG, or

È z Ø È 1 Ø x  2 z tan ic È z Ø È 1 Ø
T= É É Ù É É Ù
Ê cos ic ÙÚ Ê X p1 Ú X p2 Ê cos ic ÙÚ Ê X p1 Ú
x 2 z sin ic 2z
=  
X p2 X p2 cos ic X p1 cos ic

where x = AG . But X p2 = X p1 /sin ic [from Eq. (4.44)]; thus

140 Chapter 4

x 2 z sin 2 ic 2z x 2 z È 1  sin 2 ic Ø
t=   
X p2 X p1 cos ic X p1 cos ic X p2 X p1 ÉÊ cos ic ÙÚ
x 2z
=  cos ic (4.45)
X p2 X p1

Since sin ic = X p1 / X p2

2
È Xp Ø
cos ic = 1  sin 2 ic 1 É 1 Ù (4.46)
Ê X p2 Ú

Substituting Eq. (4.46) into Eq. (4.45), one obtains

x 2 z X 2p2  X 2p1
t  (4.47)
X p2 (X p1 ) (X p2 )

If detecting instruments are placed at various distances from the source of

disturbance to obtain first arrival times and the results are plotted in graphical
form, the graph will be like that shown in Figure 4.39b. The line Oa represents
the data that follow Eq.(4.42). The slope of this line will give 1 X p1 . The line ab
represents the data that follow Eq. (4.47). The slope of this line is 1 X p2 . Thus
the velocities of X p1 and X p2 can now be obtained.
If line ab is projected back to x = 0, one obtains

2 z X 2p2  X 2p1
t = ti =
(X p1 )(X p2 )
or
(ti )(X p1 )(X p2 ) tiX p1
z (4.48)
2 X 2p2  X 2p1 2cos ic

where ti is the intercept time. Hence, the thickness of layer 1 can be

easily obtained.

The critical distance xc (Figure 4.39b) beyond which the wave refracted at
the interface arrives at the detector before the direct wave can be obtained by
equating the right-hand sides of Equations (4.42) and (4.47):
 Properties of Dynamically Loaded Soils 141

xc xc 2 z X 2p2  X 2p1
= 
X p1 X p2 X p1 X p2
or
X 2p2  X 2p1 X p1 X p2 X p2  X p1
xc = 2 z 2z (4.49)
X p1 X p2 X p2  X p1 X p2  X p1

The depth of the first layer can be calculated from Eq. (4.49) as

xc X p2  X p1
z (4.50)
2 X p2  X p1

A. Refraction Survey in a Three-Layered Soil Medium

Figure 4.40 considers the case of a refraction survey through a three-layered soil
medium. Let X p1 , X p2 and X p3 be the P-wave velocities in layers 1, 2 and 3,
respectively, as shown in Figure 4.40a ( X p1 < X p2 < X p3 ). If A in Figure 4.40a is a
source of disturbance, the P-wave traveling through layer 1 will arrive first at B,
which is located a small distance away from A. The travel time for this can be
given by Eq. (4.42) as t = x X p1 . At a greater distance x, the first arrival will
correspond to the wave taking the path ACDE. The travel time for this case be
given by Eq. (4.47) as

x 2 z1 X 2p2  X 2p1
t= 
X p2 (X p1 ) (X p2 )

where z1 = thickness of top layer.

At a still larger distance, the first arrival will correspond to the path
AGHIJK. Note that the refracted ray H-I will travel with a velocity of X p3 . The
angle ic2 is the critical angel for layer 3.

ÈXp Ø
ic2 = sin É X ÙÙ–1 É
2
(4.51)
Ê p3 Ú

For this path (AGHIJK) the total travel time can be derived as
142 Chapter 4

Figure 4.40 Refraction survey in a three-layer soil

x 2 z1 X 2p3  X 2p1 2 z2 X 2p3  X 2p2

t   (4.52)
X p3 (X p3 ) (X p1 ) (X p3 ) (X p2 )

where z2 = thickness of layer 2.

So, if detecting instruments are placed at various distances from the source
of disturbance to obtain first arrival times, they can be plotted in a t-versus-x
graph. This graph will appear as shown in Figure 4.40b. The line Oa corresponds
to Eq. (4.42), ab corresponds to Eq. (4.47), and bc corresponds to Eq. (4.52). The
slopes of Oa, ab, and bc will be 1 X p1 , 1 X p2 and 1 X p3 , respectively. The
thickness of the first layer z1 can be determined from the intercept time ti1 in a
similar manner, as shown in Eq. (4.48), or
(ti1 ) (X p1 ) (X p2 )
z1 =
2 X 2p2  X 2p1
 Properties of Dynamically Loaded Soils 143

The thickness of the second layer can be obtained from Eq. (4.52).
Referring to Figure 4.40b, the expression for the intercept time ti2 can be
evaluated by substituting x = 0 into Eq. (4.52):

2 z1 X 2p3  X 2p1 2 z2 X 2p3  X 2p2

t = ti2 = 
(X p3 )(X p1 ) (X p3 )(X p2 )
or
Ë 2 z1 X 2p3  X 2p1 Û (X )(X )
1Ì Ü p3 p2
z2 ti 2  (4.53)
2Ì (X p3 )(X p1 ) Ü X2  X2
ÌÍ ÜÝ p3 p2

B. Refraction Survey for Multilayer Soil

In general, if there are n layers, the first arrival time at various distances from the
source of disturbance will plot as shown in Figure 4.41. There will be n segments
in the t-versus-x plot. The slope of the nth segment will give the value 1 X pn (n =
1, 2,…). More details on advanced test methods (detecting inclination of the
bedrock is briefly described later) and interpretation could be found in several
geophysics books.
The value of P-wave velocity in a natural deposit of soil will depend on
several factors, such as confining pressure, moisture content, and void ratio.
Some typical values of X p are given in Table 3.1. It is worth noting that P-wave
velocity through saturated soils will be approximately 1500 m/sec. However, as
P-wave velocity value could reach the order of few kilometres per second in case
of rocks.

Figure 4.41 Refraction survey for multilayer soil

144 Chapter 4

Example 4.1

Following are the results of a refraction survey (horizontal layering of soil).

Determine the P-wave velocities of the soil layers and their thicknesses.
Distance (m) Time of first arrival (ms)
2.5 5.5
5.0 11.1
7.5 16.1
15.0 24.0
25.0 30.8
35.0 38.2
45.0 46.1
55.0 51.3
60.0 52.8

Solution

The time-distance plot is given in Figure 4.42. From the plot,

5
Xp = = 472 m/s
1
10.6 – 103
10
X p2 = = 1389 m/s
7.2 – 103
10
X p3 = = 3333 m/s
3 – 103

Figure 4.42
 Properties of Dynamically Loaded Soils 145

ti1 = 13.3 × 10–3 s; ti2 = 35.6 × 10–3 s. From Eq. (4.48)

(ti1 ) (X p1 ) (X p2 ) (13.3 – 10 3 ) (472) (1389)
z1 =
2 X 2p2  X 2p1 2 (1389) 2  (472) 2
= 3.39 m
From Eq. (4.53)
Ë 2 z1 X 2p3  X 2p1 Û (X ) (X )
1Ì Ü p3 p2
z2 = ti 2 
2Ì (X p3 ) (X p1 ) Ü X2  X2
ÌÍ ÝÜ p3 p2

1Ë 3 (2)(3.39) (3333) 2  (472) 2 Û

= Ì 35.6 – 10  Ü
2Ì (3333) (472) Ü
Í Ý
(3333)(1389)
–
(3333) 2  (1389) 2
1
= (0.02138) (1528) = 16.33 m
2

4.12 Refraction Survey in Soils with Inclined

Layering
Figure 4.43a shows two soils layers. The interface of soil layers 1 and 2 is
inclined at an angle β with respect to the horizontal. Let the P-wave velocities in
layers 1 and 2 be X p1 and X p2 , respectively ( X p1 < X p2 ).
If a disturbance is created at A and a detector is placed at B, which is small
distance away from A, the detector will first receive the P-wave traveling through
layer 1. The time for its arrival may be given by

x
td =
X p1
However, at a larger distance the first arrival will be for the P-wave following
the path ACDE — which consists of three parts. The time taken can be written
as
td = tAC + tCD + tDE (4.54)

Referring to Figure 4.43a,

z„
tAC = (4.55)
X p1 cos ic