Computers and Geotechnics 86 (2017) 209–218

Contents lists available at ScienceDirect

Computers and Geotechnics

journal homepage: www.elsevier.com/locate/compgeo

Technical Communication

Bearing capacity of rectangular footings in uniform clay with deep

embedment
Jun Liu a,⇑, Mingzhi Li a, Yuxia Hu b, Congcong Han a
a
State Key Laboratory of Coastal and Offshore Engineering, Dalian University of Technology, Dalian 116024, China
b
School of Civil and Resource Engineering, University of Western Australia, 35 Stirling Highway, Crawley, WA 6009, Australia

a r t i c l e i n f o a b s t r a c t

Article history: This letter is concerned with the undrained bearing capacity of rectangular footings with various aspect
Received 18 July 2016 ratios and embedment ratios in uniform clay. It covers thin plate foundations with low aspect ratios and
Received in revised form 11 January 2017 high embedment depth with embedment ratio up to 150. The work is based on small strain finite element
Accepted 31 January 2017
analysis (FEA). After verification of the FEA model against existing solutions of the bearing capacity fac-
tors of rectangular footings, a series of FEA results are obtained. Based on the FEA results, a simple for-
mulation is proposed to calculate the bearing capacity factor for rectangle footing with different aspect
Keywords:
ratio in any embedment depth, extending the existing solutions to cover a wider ranges of footing aspect
Bearing capacity factor
Clay
ratios and embedment ratios.
Depth factor Ó 2017 Elsevier Ltd. All rights reserved.
Footing
Shape factor

1. Introduction
Gravity installed plate anchor (GIPLA) is a newly developed
anchor for offshore mooring floating facilities. Its mooring capacity
(i.e. uplift resistance) largely depends on its penetration depth dur-
ing gravity installation [12]. In order to assess the final penetration
depth of the anchor after installation, soil resistance needs to be
predicted accurately. As part of the soil resistance, the end bearing
of the anchor fins (see Fig. 1) is normally calculated based on the
bearing capacity formulation of deeply embedded strip founda-
tions [14,10,11].
Eq. (1) shows the general form of footing capacity in undrained
uniform clay:
qu ¼ q0 þ sc dc Nc0 su ¼ q0 þ N c su
where qu is the ultimate bearing pressure at the footing base;
q0 = cD is the surcharge loading at the footing base; D is the embed-
ment of the footing base from the soil surface; c is the unit weight
of soil; Nc0 = p + 2 [15] is the bearing capacity factor of surface strip
footing; su is the representative undrained shear strength of clay; dc
is embedment depth factor, which is the ratio of the strip footing
net bearing at depth D to that for an identical strip footing at the soil
surface; sc is footing shape factor, which is the ratio of the net bear-
⇑ Corresponding author.
E-mail addresses: junliu@dlut.edu.cn (J. Liu), 731426627@qq.com (M. Li), yuxia.
hu@uwa.edu.au (Y. Hu), 1059281811@qq.com (C. Han).
ing of a surface footing to that for a surface strip footing; Nc = qnet/su
(qnet = qu - q0) is the net bearing capacity factor of a footing with any
shape and embedded in any depth. Eq. (1) implies that the depth
and shape factors are independent [16]. However, the bearing
capacity factor Nc has taken into account the combined effects of
footing aspect ratio and embedment ratio.
Many scholars have studied the ultimate bearing capacity of
rectangular footings with different aspect ratios and embedment
depths. Table 1 summaries the previous studies on shape and
embedment factors (sc and dc). It can be observed that the formu-
lations of sc and dc are empirical before 1970s, as they are derived
from approximate analysis and prototype and model experiments
[18,13,20,1]. Due to the uncertainties involved in the empirical for-
mulations, the bearing capacity of footings have been studied
ð1Þ recently employing numerical methods, such as method of charac-
teristics (MoC) [19], upper bound analysis (UB) [22], finite element
limit analysis (FELA) [16], and finite element analysis (FEA)
[7,8,22]. Most of the studies are limited to a relatively shallow
embedment (i.e. D/B up to 2.5, where D is embedment depth and
B is the diameter of a circular footing or the width of a rectangular
footing). Among these research work, Salgado et al. [16], using
FELA, conducted a more comprehensive work, which covered the
bearing capacity factors for strip, circular and rectangular footings
with different aspect ratios and with embedment ratio extended to
D/B = 5. The footing-soil interface was considered rough and the
soil was homogenous clay. Edwards et al. [6], using the Imperial
College Finite Element Program, investigated the bearing capacity

http://dx.doi.org/10.1016/j.compgeo.2017.01.019
0266-352X/Ó 2017 Elsevier Ltd. All rights reserved.
210 J. Liu et al. / Computers and Geotechnics 86 (2017) 209–218

Nomenclature

A effective area of footing base L length of footing base

B width of footing base F bearing force without the soil buoyancy acting on the
D depth from the base of footing to the ground surface footing base
qu ultimate unit base resistance Nc bearing capacity factor of footing with different aspect
c unit weight of clay ratio in any embedment depth, calculated by F/suA
c0 effective unit weight of clay d displacement of footing
q0 surcharge at the base level, calculated by cD l Poisson0 s ratio of clay
qnet net unit base resistance E Young0 s modulus of clay
Nc0 bearing capacity factor of strip footing resting on the Ir rigidity index of clay, calculated by E/su
surface of clay M coefficient of bearing capacity factor, calculated by Nc/
su representative undrained shear strength of clay Nc0
dc depth factor that is the ratio of the net limit unit base c1, c2 parameters of M
resistance for a strip footing at depth D to that for an
identical strip footing at the soil surface
sc shape factor that is the ratio of the net limit unit base
resistance of an any shape footing resting on the soil
surface to that for a strip footing on the soil surface

factors of strip and circular footings in homogenous clay with
embedment ratio up to 4 and their results agree well with that
by Salgado et al. [16].
However, the formulas of the bearing capacity of rectangular
footing, including the effects of footing aspect ratio and embed-
ment ratio, might not be applicable to the end bearing assessment
for the anchor fins (see Fig. 1). In general, the penetration depth of
GIPLA is around 10.7–20.1 m, the thickness of the fin is about 0.1–
0.2 m and the width of the fin is about 2–3 m [24,17]. This means
that the embedment ratio can be as high as about 50–200 and foot-
ing aspect ratio can be as low as about 0.03–0.1.
In the present study, the bearing capacity factors (Nc) of rectan-
gular footing are studied with wide ranges of aspect ratios and
embedment ratios to extend the existing solutions to cover higher
embedment ratios and lower aspect ratios. The commercially avail-
able finite-element software ABAQUS [2] is used. The finite ele-
ment model is validated first with existing bearing capacity
results for footings in clay, followed by a series of FE analysis.
The FE results obtained can be used to establish a design formula
to calculate the bearing capacity factor of a rectangle footing with
any aspect ratios (i.e. B/L = 0–1.0) and its embedment ratio up to D/
B = 50.

2. Finite element model

The commercially available FEA software, ABAQUS, was chosen

as the computation platform, since it has been used successfully in
the computation of footing bearing capacity [7,23]. Small strain
Fig. 1. Typical dimensions of torpedo anchor and GIPLA. finite element models in plane strain and three-dimensional space

Table 1
Summary of published studies about sc and dc.

D/Ba kB/sub Methodc dc scd Authors

65 0 SE dc = 1 + 0.2D/B (D/B 6 2.5) sc = 1 + 0.2B/L Skempton [18]
dc = 1.5 (D/B > 2.5)
62.5 0 SE dc = 1 + 0.2D/B sc = 1 + 0.2B/L Meyerhof [13]
65 0 SE After Skempton [18] 1.2 for square and circular Terzaghi and Peck [20]
65 0 SE dc = 1 + 0.4D/B (D/B 6 1.0) After Skempton [18] Brinch Hansen [1]
dc = 1 + 0.4atan(D/B) (D/B > 2.5)
60.3 630 MoC dc = 1 + nD/B (n varies with kB/sum) sc is related to kB/sum and D/B (for circular footings) Tani and Craig [19]
65 0 FELA dc = 1 + 0.27(D/B)0.5 sc = 1 + c1B/L + c2(D/B)0.5 Salgado et al. [16]
(c1 and c2 vary with B/L)
61.2 65 UB&FEA dc = 1 + 0.25(D/B)0.4 – Yun and Bransby [22]
61 0 FEA dc = 1 + 0.86D/B  0.16(D/B)2 sc = 1 + 0.214B/L  0.067(B/L)2 Gourvenec et al. [7]
Gourvenec [8]
a
B represents the diameter of circular footing and the width of rectangle footing.
b
k is the gradient of shear strength with depth in linearly increasing shear strength profile. sum is the shear strength of clay at the mudline.
c
SE, semi-empirical; MoC, method of characteristics; UB, upper bound; FEA, finite element analysis; FELA, finite element limit analysis.
d
L represents the length of rectangle footing.
 J. Liu et al. / Computers and Geotechnics 86 (2017) 209–218 211

Refine zone

B/2

D
footing B/2

clay L/2 B

(a) model sketch (b) 3D FEA mesh

Fig. 2. Schematic diagram of a quarter of rectangular footing in clay.

were created in ABAQUS to represent rectangular footing with var-
ious aspect ratios and embedment ratios. Due to the symmetry
inherent in all of these analyses, and to save computation time,
only a quarter of the FE model was discretized for rectangular foot-
ings in 3D space and a half of the FE model for strip footings in
plain strain model. A schematic diagram of a quarter of a rectangu-
lar footing in clay is shown in Fig. 2(a) and a typical 3D mesh is
shown in Fig. 2(b).
In all analysis, the clay was modeled as an elastic-perfectly plas-
tic Tresca material with a constant undrained shear strength (i.e.
uniform clay). Poisson0 s ratio of clay, l, is 0.49 to simulate
undrained conditions. The Young0 s modulus of clay, E, does not
influence the ultimate capacity [6,23]. This is also confirmed by
the results shown in Fig. 3 where the ultimate bearing capacity fac-
tors are constant when soil rigidity index, E/su, varies from a mod-
erate to a very high values (E/su = 500–106). The soil unit weight, c,
has also no effect on the ultimate bearing capacity factor, Nc, as in
Fig. 4 for c0 = 16 and 0 kN/m3. The footing bearing capacity factor,
Nc, is calculated using Eq. (1).
The size of soil domain was adjusted with footing embedment
depth to maximize calculation efficiency. Since the plastic zone in
soil increases with increasing footing embedment depth, the soil
domain was also increased with increasing footing embedment
depth ratio to avoid domain boundary effect and minimize calcu-
lation time as well. The vertical boundaries of the soil domain
were set up as rollers and the base of soil domain was fixed.
The first order fully integrated hexahedral hybrid element
(C3D8H) and 4-node plane strain quadrilateral hybrid element
(CPE4H) were adopted in three-dimensional model and plain
strain model respectively. This was proved to be appropriate as
shown in Fig. 5(a) where identical results were obtained by using
higher order element C3D20 (i.e. 20-node quadratic hexahedral
element). Due to the higher plastic strain of soil near the corners
and edges at the footing base, a refined mesh near the footing
base was used. To achieve model efficiency, the size of refinement
zone near the footing base was tested. With the minimum size of
element hmin = B/50 (as shown in Fig. 5(b)), mesh a was selected
for all following analysis, since the two finer meshes (mesh b

16

14

12

10

8
Nc

3
6 square footing
3
(D/B=10)
4
3
strip footing
2 3
(D/B=10)
0
0 20000 40000 60000 80000
/suB
Fig. 3. Effect of soil rigidity index on footing bearing capacity factor, Nc. Fig. 4. Effect of soil weight on footing bearing capacity factor, Nc.
212 J. Liu et al. / Computers and Geotechnics 86 (2017) 209–218

9
Footing Clay
8
7
6
5
Nc

4
Contact
points
3
2 C3D20
1 C3D8H
0 The effective width of footing Half of the
0 100 200 300 400 500 is B/2 – B/100 (element size at element size
/suB the footing edge) B/100
(a) Effect of element type (B/L=0.1, /B=3)
B/2
9
8
7 Fig. 6. Numerical treatment of elements at footing edge.

6
5 The footing was modeled as a rigid body with same mesh as the
Nc

surrounding soil. The interface between the footing and the adja-
4
cent soil was simulated through contact pairs. Both smooth and
3 hmin=B/50
rough contacts were simulated at the footing base while the foot-
2 ing side walls were set up as smooth. The contacts at the edges of
hmin=B/100 the footing base were allowed to move along the footing sidewalls
1
as shown in Fig. 6. Thus the effective width and length of the foot-
0 ing base are B/2-B/100 and L/2-B/100, respectively in the quarter
0 1000
2000 3000 4000 model [9].
/suB
(b) Effect of minimum element size (B/L=0.1, /B=2.5)
3. Verification
18
16 In order to verify the numerical model built in ABAQUS, several
cases were simulated and the results were compared with existing
14
analytical, numerical and experimental solutions. The aspect
12 ratios, B/L, of the footing were 0 (strip footing), 0.2, 0.5 and 1.0
10 (square footing). The embedment ratios, D/B, of the footing base
Nc

were from 0 (surface footing) to 5. The soil was uniform clay with
8
mesh a undrained shear strength of su = 12 kPa. The contacts on the footing
6 base were rough to be consistent with the published results.
mesh b
4 Fig. 7 shows the bearing capacity factors predicted by the pre-
mesh c sent study, along with other published solutions from the analyses
2
of SE, UB, FELA and FEA (see Table 1). Fig. 7(a) shows the compar-
0
ison of the shape effect for rectangular surface footings. With the
0 10000 20000 30000
value of B/L approaching to unity, the discrepancy between the
/suB current results and the empirical formulation (SE) established by
(c) Effect of refined zone (B/L=1, /B=10) Skempton [18] is increasing. The current results agree quite well
with those (FEA) from Gourvenec et al. [7], with the maximum dif-
Fig. 5. Calibration of FEA.
ference is less than 5%.
Fig. 7(b) displays the effect of embedment depth on the bearing
and mesh c in Fig. 5(c)) provide the same footing capacity as capacity factor of a strip footing. From the comparison, the current
mesh a. The details of all three meshes were summarized in results show close agreement with the published results by Yun
Table 2. and Bransby [22] and Gourvenec [8] with the max difference of

Table 2
Summary of the information of the mesh.

Element type Minimum mesh size Numbers of elements Refined zonea

Mesh a C3D8H B/50 34,351 B/2  B/2  B
Mesh b 153,234 3B/4  3B/4  2B
Mesh c 319,302 5B/4  5B/4  2B
a
The three quantities represent the refined zone length of the footing’s width direction, length direction and embedment direction
from the footing’s outer surface, respectively.
 J. Liu et al. / Computers and Geotechnics 86 (2017) 209–218 213

6.2
(1951)
FEA, Gourvenecet al. (2006)
6.0
FEA,This study

5.8

Nc
5.6

5.4

5.2

5.0
0.0 0.2 0.4 0.6 0.8 1.0
B/L
(a) Effect of aspect ratio on surface footing capacity

7.1

6.8 UB & FEA, Yun & Bransby (2007)

FEA, Gourvenec (2008)
6.5 FEA, This study

6.2
Nc

5.9

5.6

5.3

5.0
0.0 0.2 0.4 0.6 0.8 1.0

. D/B
(b) Effect of embedment ration on strip footing capacity

15 FELA-UB, Salgado et al.(2004)

FEA, This study B/L=1
FELA-LB, Salgado et al.(2004)
14 FELA-UB, Salgado et al.(2004)
FEA, This study B/L=0.5
FELA-LB, Salgado et al.(2004)
13 FELA-UB, Salgado et al.(2004)
FEA, This study B/L=0.2
FELA-LB, Salgado et al.(2004)
12 FELA-UB, Salgado et al.(2004)
FEA, This study
Strip
FEA, Edwards et al.(2005)
11
FELA-LB, Salgado et al.(2004)
Nc

10

5
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
D/B
(c) Rectangular footing capacity factors with /B up to 5
Fig. 7. Verification of FEA.

4.66%. All the FEA results (current and previous) are higher than
the semi-empirical solutions by Skempton [18]. It should be noted
that, although a bonding interface was set up in the FEA analysis by
Gourvenec [8], the contribution to the vertical capacity from the
shear resistance mobilized along the side wall of the footing was
deducted in the paper. Thus the reported bearing capacities are
for the footing base only.
The bearing capacity factors of rectangular footings with deeper
embedment are displayed in Fig. 7(c). It covers a range of footing
aspect ratios from strip footing (B/L = 0) to square footing
214 J. Liu et al. / Computers and Geotechnics 86 (2017) 209–218

(B/L = 1). By comparing with the FELA solutions by Salgado et al. The bearing capacity factors of rectangular footings with vari-
[16], it is clear that the current FEA results locate right between ous aspect ratios and embedment ratios are depicted in Fig. 8.
the lower and upper bound (LB and UB) solutions of FELA. The The bearing capacity factors increase with increasing embedment
difference between UB and LB solutions in Salgado et al. [16] is ratios and there are no critical embedment depths observed where
quite large for all the non-strip cases, which indicates the rigorous the bearing capacity factor reached its limit. As the convergence of
solutions for rectangle footings are unavailable. Thus the the square footing0 s bearing capacity is poor with the increasing
verification may not be very strict for non-strip cases. For the strip embedment ratio, the bearing capacity is only computed with
footing, the current results agree well with the results from the embedment ratio up to 10.
Edwards et al. [6] as well. Based on the results and comparisons, Moreover, the bearing capacities of circular footings with both
the numerical model mentioned employed here is acceptable to flat and conical bases (the cone apex angle is 60°) and with embed-
calculate the bearing capacity factors of rectangular footing with ment ratio up to 20 are also calculated. The bearing capacity results
various aspect and embedment ratios. are depicted in Fig. 9(a) and the soil flow mechanism (accumulate
displacement) and the deformed mesh near the conical footing are
4. Results and discussions shown in Figs. 9(b) and (c). This problem was simulated as axisym-
metric, and 4-node bilinear axisymmetric quadrilateral, hybrid,
In the present study, eight aspect ratios were considered, i.e., constant pressure element (CAX4H) is employed. There are no crit-
B/L = 0, 0.025, 0.05, 0.06, 0.1, 0.2, 0.5, 1.0. The corresponding ical embedment depths observed where the bearing capacity factor
maximum embedment ratios were 150, 50, 50, 40, 50, 40, 30, 10 reached its limit. This is because that the undrained clay is incom-
for the eight aspect ratios respectively. This embedment range pressible. With deeper embedment, there is more soil around the
has been extended far beyond the exiting one (i.e. (D/B)max = 5.0 footing to be mobilized, similar to cavity expansion (see Fig. 10).
in Fig. 7(c)), which should cover the range for torpedo anchor or The displacement magnitude of the surface soil reaches at 1.0E
GIPLA fin design. All the cases studied in the paper are summarized 5–1.0E4 m, even the footing is embedded very deeply at D/
in Table 3. Generally speaking, the bearing capacity factors of B = 10. Although the plastic zones of clay with different embed-
footing with a rough interface are higher than those with a smooth ment ratios are almost the same, the maximum equivalent plastic
interface. Thus, for conservative design, all contacts were set up as strain near the footing edge is increasing with the increasing
smooth interface in this study. embedment ratio (see Fig. 11). This is why the footing bearing
capacity increases with its increasing embedment ratio. The foot-
ing aspect ratio also affects its bearing capacity where the bearing
capacity factor increases with increasing footing aspect ratio, B/L.
Table 3
This is because the mobilized soil from two-dimensional expansion
Summary of analysis.
(i.e. strip footing with B/L = 0) to three-dimensional expansion (i. e.
B/L D/B square footing with B/L = 1.0).
0 0, 0.5, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 12, 16, 20, 24, 32, 50, 64, 100, 150 The problem with large embedment ratio discussed in the pre-
0.025 0, 0.5, 1, 2, 3, 4, 5, 6, 8, 10, 20, 30, 40, 50 sent study is similar to the bearing capacity of pile or deep founda-
0.05 0, 0.5, 1, 2, 3, 4, 5, 6, 8, 10, 20, 30, 40, 50 tion. For deep foundations, Meyerhof [13] developed a theory of
0.06 0, 0.5, 1, 2, 3, 4, 5, 6, 8, 10, 20, 30, 40
0.1 0, 0.5, 1, 2, 3, 4, 5, 6, 8, 10, 20, 30, 40, 50
bearing capacity on the basis of plastic theory. Based on his theory
0.2 0, 0.5, 1, 2, 3, 4, 5, 6, 8, 10, 20, 30, 40 the base resistance of foundations in purely cohesive material is
0.5 0, 0.5, 1, 2, 3, 4, 5, 6, 8, 10, 20, 30 found to increase with foundation depth as the increase of shear
1 0, 0.5, 1, 2, 3, 4, 5, 6, 8, 10 band. True [21] extended the general solution of bearing capacity
0.075 0, 0.5, 1, 2, 3, 4, 5, 6, 8, 10, 20, 30
for wedges in frictional materials developed by Durgunoglu and

19 B/L
1
0.5
0.2
17 0.1
0.075
0.06
0.05
15 0.025
0(strip)

13
Nc

11 19

17

15
9
13
Nc

11

9
7
7

5
50 100 150
D/B
5
0 10 20 30 40 50
D/B
Fig. 8. Bearing capacity factors of rectangular footings in uniform clay.
 J. Liu et al. / Computers and Geotechnics 86 (2017) 209–218 215

21

18

15

12

Nc
9
Flat footing-FEA, This study
6 conical footing (60°)-FEA, This study
flat footing, Salgado et al. (2004)-FELA-LB
3 flat footing, Edwards et al. (2005)-FEA
flat footing, Salgado et al. (2004)-FELA-UB
0
0 5 10 15 20
D/B
(a) Bearing capacity factors of circular flat and conical footings in uniform clay

Accumulate Disp. 1e-4 m

(b) Soil flow mechanism near the conical footing base ( /B=5)

(c) Deformed mesh near the conical footing base ( /B=5)

Fig. 9. FEA results of circular flat and conical footings.

Mitchell [3–5] to the cases of axisymmetric penetrators with vari-
ous nose sharps and roughnesses penetrating non-frictional tar-
gets. The proposed depth factor Ncz is equal to 4 log2(D/D0 + 1),
where D0 is the lateral extent of the deformation zone adjacent to
the base of the penetrometer nose, which indicates the bearing
capacity factor increases with depth. In order to accommodate
experimental tests, both Meyerhof [13] and True [21] simplified
their equations on account for local shear failure and considered
the existing of a critical depth for the bearing capacity factor.
It is apparent that both the footing aspect ratio and embedment
ratio have effects on the footing bearing capacity factor. However,
the shape factor (aspect ratio B/L) and the depth factor (embed-
216 J. Liu et al. / Computers and Geotechnics 86 (2017) 209–218

-4 -4
magnitude magnitude

(a) B/L = 1 /B = 8 (b) B/L = 1 /B = 10

-5 -4
magnitude magnitude

(c) B/L = 0.06 /B = 8 (d) B/L = 0.06 /B = 10

Fig. 10. Soil flow mechanisms of rectangle footings (accumulate displacement/B = 0.025).


ment ratio D/B) are not independent as the conventional formulas B=L 6 0:064 c1 ¼ 5:599B=L þ 0:337 c2 ¼ 0:940  8:904B=L
suggested (see Table 1). B=L > 0:064 c1 ¼ 0:697  0:022B=L c2 ¼ 0:284 þ 1:339B=L
To incorporate the effects of footing aspect ratio and embed- ð4Þ
ment ratio on the bearing capacity calculation, a single factor M
is proposed as shown in Eq. (2). where the parameters c1 and c2 are also functions of the footing
aspect ratio (B/L). Fig. 12 displays the bi-linear function of c1 and c2
Nc F
M¼ ¼ ð2Þ against the footing aspect ratio (B/L). It can be seen that for a thin
Nc0 su AN c0 plate foundation (0 < B/L < 0.2), the functions change dramatically
The factor M is a function of footing aspect ratio (B/L) and at B/L = 0.064. The occurrence of the inflexion may be owing to
embedment ratio (D/B). By curve fitting based on the data in the data fitting only. This is an important finding, which would
Fig. 8, the M factor can be calculated using Eq. (3). not be shown without investigating the thin plate foundations.
 2    Eqs. (3) and (5) can be used for calculating footing capacity fac-
B B D tors. Although Eq. (5) indicates that the bearing capacity factor Nc
M ¼ 1 þ 0:189  0:108 þ c1 ln 1 þ c2 ð3Þ
L L B is a linear function of the bearing capacity factor of a surface strip
 J. Liu et al. / Computers and Geotechnics 86 (2017) 209–218 217

peeq 0.2521
peeq 0.275

(a) B/L = 1 /B = 8 (b) B/L = 1 /B = 10

peeq 0.1906
peeq 0.1586

(c) B/L = 0.06 /B = 8 (d) B/L = 0.06 /B = 10

Fig. 11. Soil equivalent plastic strain near rectangular footing base (accumulate displacement/B = 0.025).

footing (Nc0), the factor M changes with changing footing shape 1.8
and embedment depth non-linearly. The bearing capacity factors c1
(Nc) calculated using Eqs. (3)–(5) are plotted together with the cur- 1.5 c2
rent FEA as shown in Fig. 13. Close fit can be observed with all cor- Fitted curves (Eq. 4)
relation coefficients R2 > 0.99. 1.2
c1, c2

Nc ¼ M  Nc0 ð5Þ 0.9

In order to verify the reliability of the proposed Eqs. (3)–(5), two 0.6
more FEA cases were analyzed with one case of B/L = 0.075, D/
B = 10, and the other case of B/L = 0.35, D/B = 8. The bearing capac- 0.3
ity factors are 10.76 and 12.59 for these two cases respectively.
According to Eqs. (3)–(5), the bearing capacity factors of these 0.0
two cases are 10.85 and 12.32, respectively. The errors of the 0.0 0.2 0.4 0.6 0.8 1.0
two cases are 0.83% and 2.14%. And the results are shown in B/L
Fig. 13. On the whole, the proposed equations are reliable and
accurate. Fig. 12. Fitted functions of c1 and c2 in Eq. (4).
218 J. Liu et al. / Computers and Geotechnics 86 (2017) 209–218
Fig. 13. Comparison between FEA results and the predictions by Eqs. (3)–(5).
5. Conclusions
FEA has been employed to study the undrained bearing capacity
factors for rectangular footing with various aspect ratios and
embedment ratios in uniform clay. The current study has extended
the existing solutions to cover lower aspect ratios and higher
embedment ratios. The extension can be more beneficial to off-
shore anchor fin design. Based on the obtained bearing capacity
factors, a unique formulation was proposed to replace the previous
shape and depth factors. The proposed formulation can also be
used in other rectangular footing applications due to the large
ranges of aspect ratios and embedment ratios investigated.
Acknowledgements
This research was supported by the National Natural Science
Foundation of China (51479027, 51539008). These supports are
gratefully acknowledged.
References
[1] Brinch Hansen J. A revised and extended formula for bearing capacity. Danish
Geotech Inst Bull No 1970;28:5–11.
[2] Dassault Systèmes. Abaqus analysis users’ manual, Version 6.14; 2014.
[3] H.T.Durgunoglu, J.K.Mitchell . Static penetration resistance of soils. Space
sciences laboratory report series 14 issue 24. Geotechnical Engineering,
University of California, Berkeley; April 1973.
[4] H.T.Durgunoglu , J.K.Mitchell . Static penetration resistance of soils: analysis.
In: Proceedings of the specially conference on in situ measurement of soil
properties, Geotechnical Division, ASCE, North Carolina State University,
Raleigh; June, 1–4, 1975.
[5] H.T.Durgunoglu , J.K.Mitchell . Static penetration resistance of soils: evaluation
of theory and implication for practice. In: Proceedings of the specially
conference on in situ measurement of soil properties. Geotechnical Division,
ASCE, North Carolina State University, Raleigh; June, 1–4, 1975.
[6] Edwards DH, Zdravkovic L, Potts DM. Depth factors for undrained bearing
capacity. Géotechnique 2005;55(10):755–8.
[7] Gourvenec S, Randolph M, Kingsnorth O. Undrained bearing capacity of square
and rectangular footings. Int J Geomech 2006;6(3):147–57.
[8] Gourvenec S. Effect of embedment on the undrained capacity of shallow
foundations under general loading. Géotechnique 2008;58(3):177–85.
[9] Hu Y, Randolph MF. A practical numerical approach for large deformation
problems in soil. Int J Numer Anal Methods Geomech 1998;22(5):327–50.
[10] Kim YH, Hossain MS. Dynamic installation of OMNI-Max anchors in clay:
numerical analysis. Géotechnique 2015;65(12):1029–37.
[11] Kim YH, Hossain MS, Wang D. Numerical investigation of dynamic installation
of torpedo anchors in clay. Ocean Eng 2015;108(11):820–30.
[12] Liu J, Lu L, Hu Y. Keying behavior of gravity installed plate anchor in clay.
Ocean Eng 2016;114(3):10–24.
[13] Meyerhof GG. The ultimate bearing capacity of foundations. Géotechnique
1951;2(4):301–32.
[14] O0 Loughlin CD, Richardson MD, Randolph MF, Gandin C. Penetration of
dynamically installed anchors in clay. Géotechnique 2013;63(11):909–19.
[15] Prandtl L. Eindringungsfestigkeit und festigkeit von schneiden. Angew Math U
Mech 1921;1(15):15–20.
[16] Salgado R, Lyamin AV, Sloan SW, Yu HS. Two and three-dimensional bearing
capacity of foundations in clay. Géotechnique 2004;54(5):297–306.
[17] Shelton JT, Nie C, Shuler D. Installation penetration of gravity installed plate
anchors-laboratory study results and field history data. In: OTC Brasil. Offshore
Technology Conference; 2011.
[18] Skempton AW. The bearing capacity of clays. Proc Build Res Cong London
1951;1:180–9.
[19] Tani K, Craig WH. Bearing capacity of circular foundations on soft clay of
strength increasing with depth. Soils Foundations 1995;35(4):21–35.
[20] Terzaghi K, Peck RB. Soil mechanics in engineering practice. 2nd ed. John Wiley
& Sons; 1967. p. 223–4.
[21] True DG. Undrained vertical penetration into ocean bottom soils. Berkeley,
California: University of California; 1979.
[22] Yun G, Bransby MF. The undrained vertical bearing capacity of skirted
foundations. Soils Foundations 2007;47(3):493–505.
[23] Zhang Y, Bienen B, Cassidy MJ, Gourvenec S. Undrained bearing capacity of
deeply buried flat circular footings under general loading. J Geotech
Geoenviron Eng 2011;138(3):385–97.
[24] Zimmerman EH, Smith MW, Shelton JT. Efficient gravity installed anchor for
deep water mooring. In: Proceedings of the 41th Offshore Technology
conference, 20117, Houston; 2009.