Professional Documents
Culture Documents
Articulo Estructura Priscila Tituaña 9 B
Articulo Estructura Priscila Tituaña 9 B
net/publication/344024991
CITATIONS READS
0 77
3 authors:
Y. Tian
University of Western Australia
1 PUBLICATION 0 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Driven single piles and pile groups: numerical modelling View project
All content following this page was uploaded by Q. B. Liu on 16 September 2020.
Abstract: This technical note presents results from a series of undrained load tests performed at model scale on circular embedded footings
founded in a two-layered stratigraphy, comprising uniform stiff clay overlying soft clay. The observed foundation stiffness and bearing
capacities are compared with available empirical, analytical, and numerical solutions. The dependence of the equivalent linear stiffness
at a settlement ratio of 1% (E0.01 ) and the bearing capacity factor, N c , on the normalized depth to the soft clay (H=B) and strength ratio
of two clay layers (sut =sub ) are presented. A simple expression is developed for N c, which provides a good match to experimental data and is
consistent with numerical predictions. It is also shown that the critical H=B ratio, above which the underlying soft layer does not affect the
footing response, is significantly smaller for E0.01 than for N c at any given sut =sub value. This finding has important practical design
implications for footings on stiff-over-soft clay deposits. DOI: 10.1061/(ASCE)GT.1943-5606.0002393. © 2020 American Society of Civil
Engineers.
Author keywords: Footings and foundations; Model test; Clays.
S2-6 21.5 4.75 30 40 80 0.75 2.0 82 103 using an actuator to a settlement-to-footing-diameter ratio (s=B) of
Note: H = distance from the footing base to the top of the underlying soft between 0.15 and 0.25. Experience with the actuator has shown that
clay layer; and D = embedment of the footing below the surface. the displacement measurement has an accuracy of approximately
0.2 mm, which equates to s=B ¼ 0.01 for the smallest footing em-
ployed (B ¼ 20 mm).
tests and the best estimations of su for the upper and lower layer
of S1 are 19 and 9.5 kPa (sut =sub ¼ 2.0), while those for S2 are 21.5
Soft clay
and 4.5 kPa (sut =sub ¼ 4.75). These strength values are consistent
with maximum consolidation pressures experienced by each sam-
Geofabric ple layer.
Fig. 1. Elevation of chamber footing test setup (not to scale). To assist data interpretation, numerical analyses were performed
using the Plaxis FE program with a two-dimensional (2D) axisym-
metric mesh. The suitability of the FE model was first evaluated
by comparing its predictions with the experimental results. The
mesh was constructed using 15-noded triangular elements with a
stiff-over-soft S1 and S2 profiles were then reconsolidated to concentration of elements within the footing’s zone of influence;
stresses of 32 and 20 kPa, respectively, for a period of 1 day prior a preliminary iterative study of mesh refinement requirements was
to foundation tests. A full description of experimental work is pro- undertaken to enable accurate calculations of foundation bearing
vided in Tian (2018). capacity. The footing was represented as a discrete rigid body with
a fully rough base and the soil was modeled as an isotropic linear
elastic-perfectly plastic Tresca continuum with a friction angle of
Test Program
zero, a Poisson’s ratio (υ) of 0.49, and a cohesion for the upper and
Full details of the footing tests undertaken in profiles S1 and S2 are lower clay layers equal to the mean undrained shear strengths mea-
provided in Table 1 and a schematic of the test setup employed is sured in T-bar tests (sut and sub ). A constant ratio of the elastic
shown in Fig. 1. The model duralumin footings had roughened modulus to undrained strength (E=su ) was set for both soil layers.
(shot-blasted) circular bases with diameters (B) varying from 20 This soil specification is sufficient to represent a fine-grained mate-
to 80 mm and a central connection stem with a 10-mm-diameter rial subjected to a period of loading sufficiently short that no drain-
threaded hole at the top end where vertical load was applied. age will take place.
Special dibbers, illustrated in Fig. 1, were produced by a three- Preliminary trial analyses indicated that an E=su value of 110
dimensional (3D) printer to assist with excavation. The helical part gave reasonable predictions for the initial footing stiffness and that
of the dibber was designed to rotate around its axis to remove soil the selected E=su ratio had little influence on bearing stresses cal-
and form a cylindrical excavation with a flat base. The diameter culated at s=B > 5%. This E=su ratio falls at the lower end of the
of each helix (or cavity) was 5 mm wider than that of the tested range indicated by Atkinson (2000) and Vardanega and Bolton
footing to facilitate ease of placement and to avoid side friction. (2013) and is consistent with the normally consolidated and recon-
The footing tests were located in a manner that ensured that their stituted unaged nature of the kaolin layers. The calculated q versus
su =qc /12 20
20 su =qT-bar /10.5
su =qc /12
40
40 60
Downloaded from ascelibrary.org by China University Of Geosciences, Wuhan on 08/31/20. Copyright ASCE. For personal use only; all rights reserved.
80
60
Depth (mm)
Depth (mm)
interface, h=74mm
interface, h=110mm
100
80
120
100 140
160
CPT (before footing tests)
120 CPT_1 (before footing tests)
T-bar_1 (before footing tests)
180 CPT_2 (after footing tests)
T-bar_2 (after footing tests)
Best estimate
s of su profile Best estimate
Best estimate of
of Su
su profile
profile
140 200
(a) (b)
Fig. 2. Best estimate of su profile from CPTs and T-bar tests for (a) S1 ; and (b) S2 .
200 did not permit examination of the highly nonlinear response ex-
sut/sub=4.75
H/B=3.85 pected at s=B < 0.01.
H/B=1.92 The undrained bearing capacity of a footing on clay is calculated
160
routinely using the following equation:
qf ¼ N c dc sc su þ σv0 ¼ N c;c dc su þ σv0 ð1Þ
120 where N c = bearing capacity factor for a surface strip footing
q (kPa)
3
H s
2
Experiment N c;c ¼ 3.0 þ 6.05 ub ≤ 6.05 ð3Þ
Merifield et al. (1999) B sut
Meyerhof & Hanna (1978)
1 Brown & Meyerhof (1969)
Eq. (3) predicts N c;c values that are in relatively good agree-
Nc=1.5(H/B)+5.14(sub /sut)
ment with experimental and numerical results for the cases
0 of sut =sub ¼ 4.75 and 10, but overestimates N c;c when sut =sub ¼
0 1 2 3 4 5 6
1.25 and 2.
H/B The best-fit equation to the FE calculations, which is consistent
with the experimental data, is as follows:
Fig. 4. Experimentally derived N c values compared with proposals for
surface strip footing. The Merifield et al. (1999) trend line shown for
sut =sub ¼ 4.75 corresponds to their reported trend for sut =sub ¼ 5. sub s H
N c;c ¼ 1.18N c ¼ 6.05 þ 1 − ub tanh 0.85
sut sut B
ð4Þ
in reasonable agreement with the classical formula proposed by Eq. (4) provides a very good match to the numerical and exper-
Brinch Hansen (1970) when D=B < 2.5. imental results over a wide range of sut =sub and H=B ratios. This
equation assumes implicitly that the operational undrained shear
strength is that measured in a T-bar test. The underprediction of
Comparison of Test Results with Other N c;c for the 20-mm-diameter footing with D=B ¼ 4 (Fig. 5) arises
Experimental Data because of the overestimation by Eq. (2), compared with Brinch
Hansen (1970), of the dc value at a very large D=B value. On this
Values of N c;c were calculated for all footings tested in the labo- basis, it is recommended that dc calculated using Eq. (2) is limited
ratory chamber using Eqs. (1) and (2) and employing the undrained to a maximum value of 1.5 (i.e., corresponding to a D=B value of
strength of the upper layer (sut ) with capacity defined at s=B ¼ 0.2. 2.5). The suitability of Eq. (4) is further investigated in Fig. 6 em-
The N c;c value obtained for the footing with the large H=B ratio ploying the average of upper- and lower-bound N c values reported
of 3.85 (and D=B ¼ 1.65) is 6.04 and in excellent agreement with by Merifield et al. (1999). It is evident that Eq. (4) can provide
the theoretical N c;c value for a surface circular footing on homo- predictions that are in reasonably good agreement with all quoted
geneous clay, indicating that suT-bar (which was assumed for sut ) data and that the empirical formula for N c derived by Brown and
provides a good measure of the operational undrained shear Meyerhof (1969) from load tests on surface strip footing is too
strength. To determine values of N c , the shape correction factor, sc , conservative over a wide range of sut =sub ratios.
for a circular footing in Eq. (1) is taken to be 1.18 (¼ 6.05=5.14),
given that the exact solution for N c for a surface strip footing equals
5.14 and that an N c;c value of ∼6.05 was indicated in the load tests Equivalent Linear Stiffness
on an embedded circular footing with large H=B ratios as well as in
the FE analyses of surface circular footings. Fig. 4 compares the Because a high factor of safety against ultimate failure of shallow
experimental N c values (measured N c;c =1.18) with those reported footings is commonly applied, an important focus of designers is
in previous studies for surface strip footings. It is clear that the the assessment of footing settlement under service loads. A simple
largely empirical trends proposed by Brown and Meyerhof (1969) approach often used to assess footing settlement (s) under a bearing
and Meyerhof and Hanna (1978) are conservative, particularly for stress of q employs the following equation for a rigid punch in an
6 6
5 5
4 4
Nc,c
Nc,c
3 3
sut /sub=1.25 sut /sub=2.0
2 2
Downloaded from ascelibrary.org by China University Of Geosciences, Wuhan on 08/31/20. Copyright ASCE. For personal use only; all rights reserved.
FEM FEM
Equation (4) Experiment
1 1 Equation (4)
Equation (3)
Equation (3)
0 0
0 1 2 3 4 5 6 0 1 2 3 4 5
H/B H/B
7 7
6 6
5 5
D/B=4
4 4
Nc,c
Nc,c
3 3
sut /sub=4.75 sut /sub=10
2 FEM 2 FEM
Experiment Equation (4)
1 Equation (4)
1 Equation (3)
Equation (3)
0 0
0 1 2 3 4 5 6 0 1 2 3 4 5 6
H/B H/B
Fig. 5. Comparisons of experimental and numerical N c;c values with Eqs. (3) and (4).
elastic medium, using a single equivalent linear elastic modulus a large reduction in the bearing capacity of a footing, the impact
(Eeq ) to represent the stiffness of the medium (where υ ¼ 0.5 of this layer on the footing settlement at service loads is appreciably
for undrained loading): less. The E0.01 =su ratios of clays deposits, in general, vary with a
range of factors including overconsolidation ratio. However, Fig. 7
s π qð1 − ν 2 Þ may be used to assess the proportional reduction of E0.01 =su and
≈ ð5Þ
B 4 Eeq hence the relative influence on footing immediate settlement of an
underlying soft clay layer.
A typical design undrained s=B ratio is 0.01 and therefore Eeq
values at this ratio were back-figured from the chamber tests using
Eq. (5) and referred to here as E0.01 values. The variations with Conclusions
H=B of E0.01 normalized by sut for the two sut =sub values inves-
tigated experimentally are presented in Fig. 7(a), which also shows This technical note has extended the sparse database of footing tests
the results of FE computations at four sut =sub values. It is seen that on two-layered clay profiles. The model tests were conducted in
the footing stiffness is controlled by the stiffness of the upper clay two separate soil profiles comprising a stiff clay layer over a soft
layer when H=B is greater than a critical value. Above this critical clay layer and involved tests with a variety of distances from the
value, E0.01 =sut is equal to the best-fit E=su ratio of 110 (as em- footing bases to the underlying soft clay (H), footing diameters (B),
ployed for the FE calculations shown on Fig. 3). The general varia- and footing embedment (D). The experimental observations were
tion of E0.01 =sut with H=B is similar to that observed for N c;c in supported by FE analyses and enabled derivation of an improved
Fig. 5. However, there is a clear difference between the respective formulation for the undrained bearing capacity factor in a two-
critical H=B values, designated ðH=BÞcr . This difference is illus- layered clay profile that is generally consistent with existing
trated on Fig. 7(b), where it is evident that ðH=BÞcr for E0.01 =sut is experimental data and additional FE calculations as well as with
significantly lower than that of N c;c (or N c ) at any given sut =sub the average of upper- and lower-bound N c values presented by
ratio. Therefore, while the underlying soft clay layer can lead to Merifield et al. (1999). A simple method to estimate the undrained
5 5
Nc 4 4
Nc
3 3
sut /sub=1.25 sut /sub=2.0
2 2
Merifield et al. (1999) Merifield et al. (1999)
Downloaded from ascelibrary.org by China University Of Geosciences, Wuhan on 08/31/20. Copyright ASCE. For personal use only; all rights reserved.
0 0
0 1 2 3 4 5 0 1 2 3 4 5
H/B H/B
6 6
5 5
4 4
Nc
3 Nc 3
sut /sub=3.0 sut /sub=4.0
2 Merifield et al. (1999) 2 Merifield et al. (1999)
Equation (4) Equation (4)
1 Brown & Meyerhof (1969) 1 Brown & Meyerhof (1969)
Nc=1.5(H/B)+5.14(sub/sut ) Nc=1.5(H/B)+5.14(sub/sut )
0 0
0 1 2 3 4 5 0 1 2 3 4 5
H/B H/B
Fig. 6. N c values by Merifield et al. (1999) compared with Eq. (4) and empirical formula proposed by Brown and Meyerhof (1969).
140 3
120 2.5 Nc
80
(H/B)cr
E0.01 /sut
1.5
sut /sub =1.25, FEM
Sut/Sub=1.25, FEM
60
sut /s ub =2, Experiment
Sut/Sub=2, Experiment
sut /s ub =2, FEM
Sut/Sub=2, FEM 1
40
s ut /sub =4.75, Experiment
Sut/Sub=4.75, Experiment
sut /s ub =4.75, FEM
Sut/Sub=4.75, FEM
20 0.5
sut /s ub =10, FEM
Sut/Sub=10, FEM
0 0
0 1 2 3 4 5 6 0 2 4 6 8 10 12
(a) H/B (b) sut /sub
Fig. 7. (a) Variation of E0.01 =sut with H=B in two-layered clay profile with different sut =sub ratios; and (b) dependence of ðH=BÞcr for capacity and
stiffness on sut =sub .