Professional Documents
Culture Documents
1264
3
1
=
2
2 1 + ( r / z )
Q
z2
(1)
dx
dy
q
x
x
B
z
A
Figure 1. Uniformly loaded rectangular area
Equation of vertical stress at a depth below a corner of a uniformly loaded rectangular area
has been described by Newmark [9] by obtaining through integration of Boussinesqs [5]
point load expression is given as follows:
z =
q
4
2 BLz z 2 + B 2 + L2
2 BLz ( B 2 + L2 + 2 z 2 )
+ tan 1
B 2 L2 + z 2 ( B 2 + L2 ) + z 4
( B 2 + z 2 )( L2 + z 2 ) z 2 + B 2 + L2
(2)
1265
= q
1
cot
2
1 2
2 2
1
1 1 2
2 + 2 +
n 2 2
m
2
1
2 2
m n
(3)
2
3
5
6
7
Loading handle
Hydraulic jack
Proving ring
Loading tool
Model plate
0.7m
B=0.15m
Wooden plate
Model
ground
Measurement
system
Pressure transducer
0.7m
(a)
Glass plate
0.7m
B=0.15m
Square footing
Model
ground
Wooden plate
Steel profile
0.7m
(b)
Figure 3. Test set-up (a) Elevation (b) Plan
1267
100
90
80
Percent Geen
finer %
70
60
50
40
30
20
10
0
10.00
1.00
0.10
0.01
ap (mm)
Grain size
1268
150mm
z
Figure 5. Model plate
Vertical stresses that occur during loading beneath the centre of the square footings were
measured using a 10 bar capacity, diaphragm type, vertically mounted pressure transducer
(Figure 6). There are many methods to measure the stresses at any material due to the
loading. These methods are mechanical, optical, acoustic, pneoumatic and electrical based
experimental methods. Most of these methods can not measure the stresses directly.
Stresses are calculated using the measurable parameters like deformations [2]. The
advantage of the pressure transducer used in the study is that it measures the stresses
directly.
1269
Transducer Readings
120
100
y = 11.341x
R2 = 1.000
80
60
40
20
0
0
2
4
6
8
Applied Vertical Pressure (kN/m2)
10
(a)
(b)
Figure 8. (a) 2D model (b) 3D model
1271
Parameters
Value
Symbol
Unit
kN/m2
Cohesion
kN/m
()
41
Dilatation angle
()
11
Poissons ratio
0.20
Youngs modulus
28000
In the analyses, the dimensions of loaded area were taken similar to experimental study as
150150mm. In 2D analyses of vertical stresses in ANSYS the square footing was
transformed to circular footing (D=B170mm).
5. RESULTS AND DISCUSSIONS
In the study, vertical stress values occurring beneath the centre line of the square footings
were measured at three different depths (1.0B, 1.5B ve 2.0B) under different loadings. But,
due to space constraints and observing the similar behaviour for other loads only stress
values under the load of q=10kN/m2 were presented. Stress values obtained in the
numerical analyses under the load of 10kN/m2 were similar to the values obtained in the
experimental study. The vertical stress values obtained from experimental and numerical
analyses were also compared with the stress values computed using the Boussinesqs [5]
solution.
5.1. Comparison of Experimental, Numerical and Theoretical Results
Table 2 and Figure 9 show the vertical stress values obtained from experimental study,
numerical analyses (2D) and theoretical Boussinesq solution.
Table 2. Comparison of experimental, numerical (2D) and theoretical results
Vertical Stresses, z (kN/m2)
1272
Depth
Experimental
Linear-Elastic
Drucker-Prager
Boussinesq
1.0B
3.715
3.469
3.629
3.360
1.5B
2.067
1.928
1.971
1.770
2.0B
1.348
1.203
1.164
1.106
z (kN/m )
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
1.0
1.2
1.4
z/B 1.6
Deneysel
Experimental
1.8
Lineer Elastik
Linear
Elastic 2D
2D
Drucker-Prager 2D
2D
Drucker-Prager
2.0
Boussinesq
2.2
(a)
(b)
Experimentall
Linear-Elastic
Drucker-Prager
Boussinesq
1.0B
3.715
3.390
3.466
3.360
1.5B
2.067
1.893
1.836
1.770
2.0B
1.348
1.143
1.090
1.106
As seen from Table 2 and Table 3 the stress values obtained from the 3D Linear Elastic and
Drucker-Prager models give smaller stress values than those obtained from 2D analyses for
the same models.
Vertical stress distribution obtained from 3D finite element analysis is presented in Figure
12.
In the experimental study, while the model footing is reduced by a certain scale, sand soil
particles were used with true dimensions. Therefore, the model footing and soil may not
show the real behaviour and this might influence the experimental results. The other aim of
this study is to provide data for further experimental studies using centrifugal modeling or
in-situ testing techniques.
1274
z (kN/m )
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
1.0
1.2
1.4
z/B 1.6
Deneysel
Experimental
Lineer Elastik
Linear
Elastic 3D
3D
Drucker-Prager 3D
Drucker-Prager
Boussinesq
1.8
2.0
2.2
Figure 11. Experimental, numerical (3D) and theoretical vertical stress values
(a)
(b)
DP 2D
DP 3D
Boussinesq
0.1B
9.865
9.774
9.994
0.2B
9.550
9.365
9.604
0.3B
8.866
8.716
9.120
0.4B
8.004
7.847
7.976
0.5B
7.122
6.928
7.008
1.0B
3.629
3.466
3.360
1.5B
1.971
1.836
1.770
2.0B
1.164
1.090
1.106
3.0B
0.723
0.639
0.558
4.0B
0.627
0.521
0.320
From Table 4 and Figure 13 it can be seen that the stress values obtained from 2D axisymmetric analyses using Drucker-Prager model are greater than the stress values obtained
from 3D analyses at all depths. Boussinesqs results show very good agreement with the
results of the Drucker-Prager model from the footing base to the depth of 2.0B but it gives
smaller stress values at depths greater than 3.0B.
1276
z (kN/m2)
0
10
0,0
0,5
1,0
z/B
1,5
2,0
2,5
3,0
Drucker-Prager 2
3,5
Drucker-Prager 3
Figure 13. Drucker-prager models with 2D and 3D
6. CONCLUSIONS
In this study, the vertical stress values below the square footings resting on sandy soil were
investigated experimentally, numerically and theoretically. Based on the results obtained
from the study,
(1) The vertical stress distribution on horizontal planes at depth of z from the footing base
show the same characteristic form in all methods presented in this work and the
pressure beneath the centre line of the model plate decreases as the depth increases.
(2) Experimental results show general agreement with numerical and theoretical results.
However, it gives slightly greater values up to 14%, 19% and 18% in 2D and 3D
numerical and theoretical results, respectively.
(3) Numerical and theoretical results are in very good agreement up to especially depths of
two times of the footing width and it is observed that theoretical results give smaller
values at depths greater than three times of the footing width.
(4) It is not observed considerable differences between the results obtained from Linear
Elastic and non-linear Drucker-Prager models. But, it is suggested that the DruckerPrager model be used for soil stress problems. This is because it gives the failure load
and reflects the soil behaviour better than the Linear Elastic model does.
1277
: Point load
: Depth
: Horizontal distance
: Vertical stress
: Foundation length
: Foundation width
: Uniform load
: Poissons ratio
m, n
: Geometrical coefficients
Dr
: Relative density
dry
: Youngs modulus
: Cohesion
: Dilatation angle
: Diameter
References
[1] Uzuner, B.A., zml Problemlerle Temel Zemin Mekanii, Ankara, Teknik
Yaynevi, 1998 (in Turkish).
[2] Salamer, A., Kohezyonsuz Zeminlerde Skunetteki Toprak Basnc Katsaysnn
Zemin Parametreleri Cinsinden fadesi, Ph. D. Thesis in Civil Engineering, stanbul
Teknik niversitesi, stanbul, 1972 (in Turkish).
1278