Professional Documents
Culture Documents
Comparison Between 2d and 3d Behaviour of Sheet Piles by Finite Element Method
Comparison Between 2d and 3d Behaviour of Sheet Piles by Finite Element Method
ABSTRACT
In many cases, two dimensional (2D) idealizations of what are essentially three
dimensional (3D) problems may not produce reliable results. Whereas highly realistic 3D
analyses might not have been within the reach of the practicing geotechnical engineer in
the past, developments in numerical technology are changing this situation and bringing
realistic 3D analysis within the reach of the practicing engineer. The conventional
methods, based on the limit equilibrium approach, are commonly used in the design of
sheet pile walls. This paper presents a series of 3D numerical analysis performed by the
nite element method (FEM) on sheet piles in a loose sandy soil, and the eect of
dierent length to height ratio of sheet piles on maximum bending moment, lateral earth
pressure and sheet pile hinge point position was obtained. The purpose of 3D nite
element method simulations was to investigate the soil-sheet pile interaction mechanism
by comparing the nite element modeling comprising 2D plane strain analysis of sheet
pile walls. The results indicate that, in sheet piles with length to height (L/H) ratio less
than 1, 2D analysis is very conservative in comparison with 3D analysis. On the other
hand, for the L/H ratio more than 5, there seems to be a threshold value of 3D to 2D
nite element method results beyond which the ratio of 3D to 2D analysis became
insensitive to any increase in L/H ratio.
Sheet pile walls are widely used in excavation support systems, coerdams, cut-o
walls under dams, slope stabilization, waterfront structures, and oodwalls.
Although there are several other materials (such as timber, reinforced concrete,
and plastics) used for sheet piles, steel sheet piles are the most common in
retaining walls (Bilgin, 2010). Sheet pile walls consist of continuously interlocked
pile segments embedded in soils to resist horizontal pressures (DeNatale &
Ibarra-Encinas, 1992; King, 1995; Madabhushi & Chandrasekaran, 2005). The
sheet pile walls can be either cantilevered or anchored. The selection of the wall
type is based on the function of the wall, the characteristics of the foundation
soils, and the proximity of the wall to existing structures. While the cantilevered
walls are usually used for wall heights less than 3-5m, anchored walls are required
for higher walls or when the lateral wall deformations need to be restricted
(ASCE, 1996). Cantilevered sheet pile walls are frequently used for retaining in
low- height granular soil in river protection walls, excavation, and as temporary
supports in foundation construction.
Current and common design practice of the use of sheet pile walls is based
on the limit equilibrium approach. The design utilizes the active and passive
earth pressures, which are concerned with the failure condition based on the
Mohr-Coulomb failure criterion. Wall penetration depth, anchor force, and
pile section selection come from the force and moment equilibrium conditions
using the lateral earth pressures. A factor of safety is applied to the passive
pressures during the determination of lateral earth pressures (US Navy, 1986,
US Army Corps of Engineers, 1994). However, the actual earth pressure
distributions along sheet pile wall have not been very well understood.
Experimental and eld-measured data will be helpful to verify, and when
needed, to modify current design criteria. Substantial experimental research
and eld measurements (Terzaghi, 1934; Coyle & Bartoskewitz, 1976; Sherif
et al., 1982, Fang & Ishibashi, 1986; Bentler & Labuz, 2006) have been
carried out regarding the lateral earth pressures developing against rigid
retaining walls. By contrast, only a limited number of studies have been
conducted to examine current sheet pile design procedures via experimental or
eld-measured data (Peck, 1943; DiBiagio, 1977; Stille, 1979; Finno, 1989;
Endley et al., 2000). Furthermore, most of these studies were related to
anchored or strutted sheet piles in excavation, or only measured strut or
anchor loads, instead of measuring earth pressures directly. Even fewer
studies are known to have investigated passive earth pressures against the
supporting face of a sheet pile. Strom and Ebeling performed model tests on
free embedded cantilever walls in coarse sand in 2002. Lyndon and Pearson
performed centrifuge tests on free embedded cantilever walls in 1984. Bica
and Clayton suggested penetration depth ratios (D/H) for design based on
experimental data, and also suggested design guidelines in 1989. In 1990, Das
explained the basic principles of the estimation of net lateral pressure
distribution on a cantilevered sheet pile, which can be explained with the aid
of Figure 1.
Fig. 1. Pressure distribution on a cantilever sheet pile wall penetrating sandy soil (Das, 1990).
The objective of this study is to analyze the behavior of sheet pile walls under
2D and 3D analysis in a loose sandy soil. The purpose of 3D nite element
method simulations was to investigate the soil-sheet pile interaction mechanism
by comparing the nite element modeling comprised 2D plane strain analysis of
sheet pile walls. The objective was achieved by a series of comprehensive nite
element analyses for varying length-to-height ratios. The eect of dierent
length-to-height ratios of sheet piles on maximum bending moment, lateral
earth pressure and sheet pile hinge point position was obtained.
NUMERICAL MODELING PROCEDURE
Conventional design of sheet pile walls
Traditionally, design and analysis of sheet pile walls (DeNatale & IbarraEncinas, 1992; King,1995; Madabhushi & Chandrasekaran, 2005) are carried
out on the basis of force and/or moment equilibrium to determine the required
penetration depth of walls, regardless of the magnitude or direction of wall
movements. For the stability of the wall, the following two conditions must be
satised:
(I) P horizontal forces per unit length of wall = 0
(II) P moment of the forces per unit length of wall about the base of sheet pile wall = 0
Finite element analysis
wall. The bottom boundary of the mesh was xed in all directions and the
vertical boundaries were xed only in the normal direction. In the two
dimensional analysis, the soil-sheet pile system was considered under plane
strain condition, with the length of one meter. The purpose of 3D nite element
method simulations was to investigate the soil-sheet pile interaction mechanism
by comparing the nite element modeling comprised 2D plane strain analysis of
sheet pile walls, and hence to obtain helpful insight for designers when they
choose between a 2D or 3D analysis for sheet pile design and analysis.
In soil related problems, the rst step in simulating a real problem by a
numerical method is to simulate realistic soil parameters that correspond to the
eld condition. Therefore, a loose sandy soil was selected for this purpose. An
elastic-plastic model was used to describe the soil-structure interface behavior.
The interface strength, Rint, is determined by Equation 1.
Rint
tan
tan
where is the interface friction angle and is the soil friction angle. A ratio of
the interface friction angle to soil friction angle, , ranges from 0.5 to 0.9 for the
sand and steel interface (Kulhawy et al., 1983). The modulus of elasticity, for
example, is an average value selected from the typical range given by Kulhawy
& Mayne (Kulhawy et al., 1990) and it is representative of secant moduli within
common design stress levels. The Mohr-Coulomb constitutive model for soils
had been commonly used in nite element modeling of retaining wall behavior
(Fan & Luo, 2008; Grande et al., 2002; Krabbenhoft et al., 2005; Neher &
Lachler, 2006; Tan & Lu, 2009; Tan & Paikowsky, 2008). The Mohr-Coulomb
model has been successfully used for granular soils and therefore was also
employed in this study to model the stress-strain behavior of sands. The MohrCoulomb model is a linearly elastic and perfectly plastic constitutive model. The
parameters needed for the Mohr-Coulomb model are the Young's modulus, E,
and Poisson's ratio, , for the elastic strain component of the soil behavior. The
eective strength parameters cohesion, c, and friction angle, , are needed for
the plastic strain component of the soil behavior. The properties selected for the
soil and sheet pile are given in Table 1.
Value
10000 (kN/m2)
0.35
17.5 (kN/m3)
2.1x108 (kN/m2)
0.25
78 (kN/m3)
30 (degree)
0 (kN/m2)
0.6
The validity of the 2D nite element model employed in the program is veried
by the conventional methods, which are based on the limit equilibrium
approach. Calculations of earth pressures in sheet pile design are usually based
on classical soil mechanics. Figure 2 schematically indicates the variation of
earth pressures acting on both sides of the sheet pile wall (Babu & Basha, 2008).
Eective earth pressures acting against sheet pile walls can be calculated by
using the following equations:
2
P2
L1 k ;
where P2 is the eective active earth pressure acting against the sheet pile at the
dredge line level.
3
P3
d2 k k ;
Where P3 is the net eective earth pressure acting at the bottom of the sheet
pile acting towards the backll.
4
P4
L1 k
Dk k and
a
P5
L1 kp d1 kp ka
where P4 and P5 are the net eective earth pressure acting at the bottom of the
sheet pile
acting against the backll side of wall.
is the unit weight of the soil (in
3
kN/m ), L1 is the height of cantilever sheet pile wall above the dredge line (in m),
D = (d1 + d2) = total penetration depth below the dredge line (in m), and H =
(D + L1) is the total height of sheet pile (in m).
Fig. 2. Pressure distribution on a cantilever sheet pile wall penetrating sandy soil
The comparison between the 2D nite element predicted and the classic
passive Rankine earth pressures is shown in Table 2. It can be seen that the 2D
FEM calculated earth pressures match well with the classic passive Rankine
earth pressures in this study. It seems that the friction force between a loose
sandy soil and the sheet pile wall (interfacial friction) was not mobilized. Similar
results were also observed by Tan & Paikowsky, 2008, who noted that the
friction force between the peat and the sheet pile wall was not mobilized.
Table 2. Comparison of 2D nite element results with the classic passive
Type of analysis
2D nite element
results
Classic passive
Rankine results
P2
(kN/m)
P3
(kN/m)
P4
(kN/m)
P5
(kN/m)
29.1
161.2
452.7
291.5
29.5
163.2
453.8
293.2
A series of 2D and 3D nite element analyses have been carried out to study the
inuence of L/H ratio on the maximum bending moment, horizontal earth
pressure and hinge point position of sheet piles, and the threshold value of L/H
is presented. The results obtained from this investigation with reference to
various parameters are presented below.
Maximum bending moments
Table 3 shows a comparison between the maximum bending moments per meter
run of wall computed by the 3D nite element method at various planes in the
longitudinal direction, and the maximum bending moments computed by the
2D analysis. It can be seen from Table 3 that, for the L/H ratio less than 1.0, the
maximum bending moment found by the 2D analysis is greater than the
moment obtained from the 3D analysis. Therefore, in this condition, the 2D
approach led to an over-design analysis of sheet pile. For the L/H ratio more
than 1.0, the computed 3D bending moments of sheet pile were greater than
those computed by the 2D analysis. In other words, by increasing the L/H ratio
of sheet pile, maximum bending moment in 3D analysis gave bigger value and
thus more caution should be taken when using the 2D analysis.
Table 3. Bending moment results.
2D
3D
3D
3D
3D
3D
3D
3D
3D
3D
3D
3D
3D
3D
3D
3D
Type of analysis
L/H=0.1
L/H=0.4
L/H=0.8
L/H=1.0
L/H=1.2
L/H=1.6
L/H=1.8
L/H=2.0
L/H=2.2
L/H=2.4
L/H=2.6
L/H=3.0
L/H=4.0
L/H=5.0
L/H=6.0
Mmax (kN.m/m)
286
225
230
230
260
290
301
301
301
301
301
301
300
296
290
285
The ratio of Mmax (3D) to Mmax (2D) versus dierent values of L/H is
demonstrated in Figure 3. It can be seen that, for L/H ratio less than 1, data are
more scattered. On the other hand, for L/H ratio more than 5, the comparison
of results between the 2D and 3D nite element method indicates an overall
consistency and the 2D analysis has a good compatibility with real conditions.
In other words, the ratio of M (3D) to M (2D) increases with increasing of L/H
to a threshold value around 1, beyond which M (3D) / M (2D) becomes
insensitive to any increase in L/H.
Fig. 3. The ratio of Mmax (3D) to Mmax (2D) versus L/H ratio.
The lateral earth pressure analysis details of the 2D and 3D analysis are
presented in Table 4. It can be seen from Table 4 that, for the L/H ratios less
than 1.0, the lateral earth pressures (P2, P3, P4 and P5) increase in the 2D
analysis of sheet piles, and for P4 load this eect is more. Similar to maximum
bending moment results, for the L/H ratio less than 1.0, 2D analysis is very
conservative in comparison with 3D analysis. For the L/H ratio of sheet pile
between 1 and 5, lateral earth pressure distribution in 3D analysis gives bigger
value and thus more caution should be taken when using the 2D analysis. On
the other hand, for the L/H ratio less than 5, there is no dierence in the lateral
earth pressure distribution between the 2D and 3D analysis.
10
2D
3D
3D
3D
3D
3D
3D
3D
3D
3D
3D
3D
3D
3D
3D
3D
L/H=0.1
L/H=0.4
L/H=0.8
L/H=1.0
L/H=1.2
L/H=1.6
L/H=1.8
L/H=2.0
L/H=2.2
L/H=2.4
L/H=2.6
L/H=3.0
L/H=4.0
L/H=5.0
L/H=6.0
P2
(kN/m)
29.1
22.1
21.1
22.7
26.7
31.3
30.2
30.2
30.4
30.4
30.4
30.4
30.3
30.2
30.0
29.4
P3
(kN/m)
161.2
117.7
112.1
120.9
145.1
166.2
166.8
168.2
167.6
167.3
166.0
164.4
163.6
163.2
163.0
162.5
P4
(kN/m)
452.7
348.5
331.8
362.1
411.9
473.2
472.6
471.7
470.8
470.3
468.5
469.9
467.6
469.4
465.0
462.0
P5
(kN/m)
291.5
227.3
215.4
230.3
268.1
302.2
306.0
310.2
307.5
306.0
306.6
305.5
304.6
305.6
303.0
300.0
In Figures 4a, 4b, 4c and 4d, the ratio between the 3D and 2D lateral
pressures versus L/H ratio are plotted. The Figures show that the ratio of
3D to 2D lateral pressures increase with the increase in L/H ratio less than
2. Regression analysis using between the ratio of P (3D) / P (2D) and the
L/H ratio shows that there is a threshold P (3D) / P (2D) around 1, beyond
which the ratio of 3D to 2D lateral pressures became insensitive to any
increase in L/H.
11
The exact positions of hinge point in 2D and 3D analyses are presented in Table
5. It can be seen from Table 5 that for L/H ratio less than 1.0, the hinge point
position moves downward in the 2D analysis. From Table 5 and Table 4 it may
also be concluded that for the L/H ratio less than 1.0, the lateral earth pressure
distribution in 3D analysis increases with the decrease in hinge point distance
from the bottom of sheet pile.
12
2D
3D
3D
3D
3D
3D
3D
3D
3D
3D
3D
3D
3D
3D
3D
3D
Type of analysis
L/H=0.1
L/H=0.4
L/H=0.8
L/H=1.0
L/H=1.2
L/H=1.6
L/H=1.8
L/H=2.0
L/H=2.2
L/H=2.4
L/H=2.6
L/H=3.0
L/H=4.0
L/H=5.0
L/H=6.0
d (m)
0.64
0.93
0.89
0.82
0.72
0.71
0.70
0.69
0.68
0.68
0.67
0.67
0.66
0.66
0.65
0.64
13
CONCLUSION
14
- Based on the results of this study, for the L/H ratio more than 5, the
comparison of results between the 2D and 3D Finite Element Method
indicated an overall consistency.
ACKNOWLEDGEMENT
The authors are thankful for the editorial board and reviewers' precision and
their valuable comments. They have made for a more clear and detailed paper.
REFERENCES
ASCE. 1996. Design of sheet pile walls. Technical engineering and design guides as adapted from
Babu, G.L.S & Basha, B.M. 2008. Optimum design of cantilever sheet pile walls in sandy soils
using inverse reliability approach. Journal of Computers and Geotechnics 35: 134-143.
Bica, A.V.D. & Clayton, C.R.I. 1989. Limit equilibrium design methods for free embedded
cantilever walls in granular materials. Proceedings of Institutions of Civil Engineers 1: 879-98.
Bilgin, O. 2010. Numerical studies of anchored sheet pile wall behavior constructed in cut and ll
conditions. Journal of Computers and Geotechnics 37: 399-407.
Bjerrum, L., Clausen, C.J.F. & Duncan, J.M. 1972. Earth pressures on exible structures: A state-
of-the-art report. 5th European Conference on Soil Mechanics and Foundation Engineering
(Madrid) 2: 169-96.
Chiralaksanakul, A. & Mahadevan, S. 2005. First order approximation methods in reliabilitybased design optimization. ASCE Journal of Mechanical Design 127: 851-7.
Das, B.M. 1990. Principles of Foundation Engineering. PWS - KENT Publishing Company, USA.
DeNatale, J.S. & Ibarra-Encinas, G.A. 1992. Total stress analysis of cantilever sheetpiling in
layered clay. Journal of Geotechnical Engineering 118(7): 1064-82.
Duncan, J.M., Byrne, P., Mabry, P. & Wong, K. 1980. Strength, Stress-Strain, and Bulk Modulus
Parameters for Finite-Element Analyses of Stresses and Movements in Soil Masses. University
of California, Berkeley, Calif.
Fan, C.C. & Luo, J.H. 2008. Numerical study on the optimum layout of soil-nailed slopes.
Computer Geotech 35(4): 585-99.
Grande, L., Soreide, OK. & Tefera, T.H. 2002. Large scale model testing on the moment
distribution and deformation behaviour of a sheet pile wall. In: Proceedings, 2nd International
Conference on Soil Structure Interaction in Urban Civil Engineering, Zurich, Switzerland, pp.
389-94.
King, G.W.J. 1995. Analysis of cantilever sheet-pile walls in cohesionless soils. Journal of
Geotechnical Engineering 121(9): 629-35.
Krabbenhoft, K., Damkilde, L. & Krabbenhoft, S. 2005. Ultimate limit state design of sheet pile
walls by nite elements and nonlinear programming. Computer Structures 83(4 and 5): 383-93.
Kulhawy, FH. & Mayne, PW. 1990. Manual on estimating soil properties for foundation design.
Electric Power Research Institute EL-6800.
Kulhawy, FH., Trautmann, CH., Beech, JF., O'Rourke, TD. & McGuire, W. 1983. Transmission
line structure foundations for uplift-compression loading. Electric Power Research Institute EL2870.
Lee, J.O., Yang, Y.S. & Ruy, W.S. 2002. A comparative study on reliability-index and targetperformance-based probabilistic structural design optimization. Computer Structures 80(3-4):
15
257-69.
Lyndon, A. & Pearson, RA. 1984. Pressure distribution on a rigid retaining wall in cohesionless
material. Proceedings of Symposium Application of Centrifuge Modeling to Geotechnical
Design, Manchester, pp. 271-80.
Madabhushi, S.P.G. & Chandrasekaran, V.S. 2005. Rotation of cantilever sheet pile walls. Journal
of Geotechnical Geoenvironmental Engineering 131(2): 202-12.
Neher, H.P. & Lachler, A. 2006. Numerical modeling of a diaphragm wall production process in
Rotterdam compared to monitoring data. Sixth European Conference on Numerical Methods
in Geotechnical Engineering (Austria): pp. 417-22.
Ou, C.Y. & Chiou, D.C. 1993. Three-dimensional nite element analysis of deep excavation.
Proceedings, 11th Southeast Asian Geotechnical Conference, Singapore.
Strom, R.W. & Ebeling, R.M. 2002. Simplied procedures for the design of tall, sti tieback walls.
Technical Report ERDC/ITL TR-02-10, U.S Army Engineer Research and Development
Center, Vicksburg, MS, USA.
Tan, Y. & Lu, Y. 2009. Parametric studies of DDC-induced deections of sheet pile walls in soft
soils. Journal of Comput Geotech; 36(5): 902-10.
Tan, Y. & Paikowsky, SG. 2008. Performance of sheet pile wall in peat. ASCE Journal of
Geotechnical and Geoenvironmental Engineering 134(4): 445-58.
US Army Corps of Engineers. 1994. Design of sheet pile walls, EM 1110-2-2504, Washington (DC).
US Navy. Naval Facilities Engineering Command (NAVFAC). 1986. Design manual, 7.02.
Foundations and Earth Structures. Virginia, USA.
Wissmann, K.J., Filz, G.M. & Martin, J.R. 1995. New design methods for sheet pile cellular
structures.
Submitted : ??/??/????
Revised : ??/??/????
Accepted : ??/??/????
16
,}Gf gOGRB
***
**
,!9}TF HGQ%|
?Y;L
k
O9g<}G ?+C;C {v9W| 9S9SC ,$ ,AyG O9g<}G ,F9"C - ?+y9D| J:9JyG #| Q+Dv ,p
O9g<}G ?+C;C {+zJ@ fC ?*9jzy ,gsGh &!C #+I ,p .9%+zf O9}Af:G #w}* HF9A! #f QqT@
9+F(y("wAyG d9G| ,p JGQ(aAyG fE .,\9}yG ,p ?+"t@(+GyG ,SO"%| dh9"A| ,p #w* ~y
,p &zgFh ,C;DyG Og=zy ,gsGh {+zJ@ ~*Ot@h h\(yG GP$ Q++j@ )zf IQO9s ?+}sQyG
OJyG fRG(@ H%! )zf e(t@ ,AyGh ?*O+ztAyG bQayG fE .#+SQ9}}yG #+SO"%}yG dh9"A|
#| ?zTzS EJ=yG GP$ VQgAT* .?sQ(yG eG(vC fGQOF ~+}Z@ ,p eOMAT@ 9| IO9f
eG(vC )zf )FEM( IOOJ}yG QY9"gyG ?t*Q_ 9%< e(t@ ,AyG O9g<}G ?+C;C ?+}sQyG {+y9JAyG
/d(ayG #| ?qzAM| >T! Q+Cz@ )zf d(ZJyG ~@ 9}v .?wwq}yG ?+z|QyG ?<QAyG ,p ?sQ(yG
?at! hs(|h ,tp}G VQ}G bj\ ,)Zs}G )"J"}yG ?dJy )zf ?sQ(yG eG(v} Y9q@Q:G
O9g<}G ?+C;C IOOJ}yG QY9"gyG ?t*Q_ I9v9J| #| VQjyG fE .?sQ(yG eG(v} {Zq}yG
IOOJ}yG QY9"gyG LP(}! ?!Q9t}< xyPh ?sQ(yG eG(v} ?<QAyG {f9q@ ?+yB ?pQg}y f9v
&!C )zf HF9A"yG Q+W@ .?sQ(yG eG(vC fGQOGy O9g<}G ,F9"C i(AT| bj\ {+zJA< ?!Q9t|
eqJA| 2D {+zJ@ f(w* ,OIGh #| {sC Y9q@QG /d(_ ?=T! f(w@ ,AyG ?sQ(yG eG(vC ,p
k
Y9q@Q:G /d(ayG ?=T! f(w@ 9|O"f ,iQL}G ?+I9"yG #| .3D {+zJ@ h| ?!Q9t| GOF
QY9"gyG ?t*Q_ HF9A"y 2D )yE 3Dvy )!O}G OJzy ?}+s c9"$ fC hO=* ,5 #| QDvC
?=T! ,p IO9*R jC )yE ?S9TI Q+i 2D )yE 3D {+zJ@ ?=T! f(w@ 9$Og< ,AyG ,IOOJ}yG
.Y9q@Q:G /d(ayG
: Os