Three-Dimensional Effects for Supported

Excavations in Clay
Richard J. Finno, M.ASCE1; J. Tanner Blackburn, A.M.ASCE2; and Jill F. Roboski, A.M.ASCE3

Abstract: This paper presents the results of 150 finite-element simulations conducted to define the effects of excavation geometry, i.e.,
length, width, and depth of excavation, wall system stiffness, and factor of safety against basal heave on the three-dimensional ground
movements caused by excavation through clays. The results of the analyses are represented by the plane strain ratio 共PSR兲, defined as the
maximum movement in the center of an excavation wall computed by three-dimensional analyses normalized by that computed by a plane
strain simulation. A simple equation for PSR is presented based on excavation geometry, wall system stiffness, and factor of safety against
basal heave. This PSR equation reasonably represents trends in results of the 150 simulations as well as those simulations reported in
literature. When the excavated length normalized by the excavated depth of an excavation wall is greater than 6, results of plane strain
simulations yield the same displacements in the center of that wall as those computed by a three-dimensional simulation.
DOI: 10.1061/共ASCE兲1090-0241共2007兲133:1共30兲
CE Database subject headings: Three-dimensional analysis; Excavation; Clays; Geometry; Stiffness.

Introduction when one can reduce the maximum ground movement estimated
with a semiempirical method on the basis of a 3D effect. Further,
Evaluating the magnitude and distribution of ground movements as inverse analysis techniques become more common in the ap-
adjacent to an excavated wall is an important part of the analysis plication of the observational method to supported excavations
and design process when excavating in an urban environment. 共e.g., Finno and Calvello 2005兲, it is important to know when an
Three-dimensional 共3D兲 effects caused by the higher stiffness at excavation can be adequately modeled as plane strain so as to
the corners of an excavation lead to smaller ground movements distinguish between corner stiffening effects and constitutive re-
near the corners and larger ground movements toward the middle sponses of the soil without resorting to a full 3D simulation of the
of the excavation wall. Another, and not necessarily universally excavation.
recognized, consequence of the corner stiffening effects is that the This paper presents the results of a finite-element parametric
maximum movement near the center of an excavation wall may study conducted to define the effects of excavation geometry 共i.e.,
not correspond to that found from a conventional plane strain length, width, and depth of excavation兲, wall stiffness, and factor
simulation of the excavation, i.e., 3D and plane strain simulations of safety against basal heave on the 3D ground movements
of the excavation do not yield the same movement at the center caused by excavation through clays. The results of the analyses
portion of the excavation, even if the movements in the center are are represented by the plane strain ratio 共PSR兲, defined herein as
perpendicular to the wall. While the former 3D effect is clear in the maximum movement in the center of an excavation wall com-
all field data reported in literature, the latter effect cannot be puted by 3D analyses divided by that computed by a plane strain
evaluated solely by field data. simulation. It is common practice to consider plane strain results
While performance data reported in literature are inherently as representative of deformations near the center of an excavation
3D, common semiempirical methods to define wall movements wall. Ou et al. 共1996兲 originally defined PSR in terms of the
共e.g., Clough et al. 1989兲 based on wall stiffness and factor of width-to-length 共B / L兲 ratio of the wall, and the distance from the
safety against basal heave are based in part on plane strain finite- corner.
element simulations of excavations. It is important to recognize Results of parametric studies presented herein indicate that
the value of PSR is affected by 共1兲 the ratios of the length of wall
1
Professor, Dept. of Civil and Environmental Engineering,
to the excavation depth 共L / He兲; 共2兲 the plan dimensions of the
Northwestern Univ., Evanston, IL 60208. E-mail: r-finno@ excavation, L / B, with L being the side where movements are
northwestern.edu computed, 共3兲 the wall system stiffness 共EI / ␥h4兲 as defined by
2
Assistant Professor, Dept. of Civil Engineering, Texas A&M Univ., Clough et al. 共1989兲; and 共4兲 the factor of safety against basal
College Station, TX. heave. Of these factors, the L / He ratio was the most influential for
3
Engineer, GeoSyntec Consultants, Kennesaw, GA 30144. E-mail: the range of parameters considered herein.
[email protected]
Note. Discussion open until June 1, 2007. Separate discussions must
be submitted for individual papers. To extend the closing date by one
month, a written request must be filed with the ASCE Managing Editor.
Background
The manuscript for this paper was submitted for review and possible
publication on February 28, 2006; approved on July 28, 2006. This paper The geometry of an excavation is described by its plan view
is part of the Journal of Geotechnical and Geoenvironmental Engineer- dimensions, depth of excavation, and total height of wall. This
ing, Vol. 133, No. 1, January 1, 2007. ©ASCE, ISSN 1090-0241/2007/1- geometry has a significant effect on the ground response due to
30–36/$25.00. excavation. 3D responses of excavations were reported by Bono

30 / JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / JANUARY 2007

Downloaded 11 Jun 2012 to 180.211.192.67. Redistribution subject to ASCE license or copyright. Visit http://www.ascelibrary.org
et al. 共1992兲, Wong and Patron 共1993兲, Ou et al. 共1993, 1996, Table 1. Summary of 3D Finite-Element Analyses for Parametric Study
2000a,b兲, Chew et al. 共1997兲, Lee et al. 共1998兲, Finno and Bryson Primary length Secondary length
共2002兲, and Finno and Roboski 共2005兲. The following observa- Stratigraphy Height of cut, of cut, of cut,
tions can be made from these data regarding the movements near 共see Fig. 2兲 He共m兲 / FSBH L 共m兲 B 共m兲
the corner of an excavation:
A 9.8/ 1.7, 13.4/ 1.68, 20 20, 40, 80
1. In all cases, the ratio of corner to center movements perpen- 16.3/ 1.8
dicular to a wall, ␦Corner / ␦Center, was less than 1.0 indicating 40 20, 40, 80
that movements decrease near the corners of the excavation
80 20, 40, 80, 160a
due to the stiffening effects of the corners. Note that this is
160 80a
not the plane strain ratio defined previously herein.
B 9.8/ 1.63, 13.4/ 1.42, 20 20, 40
2. In general, the shorter the plan dimension of the excavation
16.3/ 1.28
wall, the smaller the movement that will be measured near
40 20, 40, 80
the center of that excavation wall due to the stiffening effects
80 40, 80
of the corner.
3. Deeper excavations experience smaller ␦Corner / ␦Center ratios, Note: System stiffnesses of 32, 320, and 3,200 were considered for each
of the 50 cases.
or higher reductions in movements near the corners of the a
excavation as compared to shallower excavations in similar Analyzed for He equal to 9.8 m only.
soil conditions with similar support systems.
An empirical procedure that relates the geometry of the exca- Finite-Element Model and Procedures
vation to the distribution of ␦x / ␦Center where ␦x is the lateral
movement at any distance x along a wall, and, hence, the distri- The commercially available PLAXIS 3D FOUNDATION and 2D,
bution of ground movements parallel to an excavation wall, has version8.0, three-dimensional and plane strain geotechnical finite-
been proposed by Roboski and Finno 共2006兲. However neither the element software packages, respectively, were used to conduct the
field data nor the empirical procedure provides direct information parametric study. Structural elements were modeled with aniso-
concerning whether the maximum movements can be reliably es- tropic linear and nonlinear elastic elements. Soil elements are
timated on the basis of plane strain assumptions. To develop such 15-node wedge elements which are created by the projection of
guidelines, both 3D and plane strain analyses of the same exca- 2D, 6-node triangle elements. Variations in stratigraphy interfaces
vation must be conducted. By comparing the results of such 共i.e., nonhorizontal兲 were modeled by 13-node pyramid elements
analyses, one can define the conditions wherein the two- and 10-node tetrahedral elements. Support structure elements
dimensional 共2D兲 plane strain results are applicable to the actual consist of 3-node line elements for beams, and 6-node and 8-node
3D geometry and develop factors that define the stiffening effects plate elements for walls. Soil-structure interaction is simulated by
near the corners. 12-node and 16-node interface elements. Soil responses are de-
A number of finite-element studies have been performed to fined by the hardening soil 共HS兲 model 共Schanz et al. 1999兲. For
evaluate this question including those reported by Ou and Chiou more details concerning the finite-element representation, see
共1993兲, Ou et al. 共1996兲, Chew et al. 共1997兲, Lee et al. 共1998兲, and Blackburn 共2005兲.
Lin et al. 共2003兲. From these studies, the following observations
may be drawn concerning the difference between the plane strain
2D calculation of movements near the center of the excavation Parametric Variables
wall and the 3D calculation. One hundred and fifty 3D finite-element analyses were made to
1. As was seen in the field data, all studies show that smaller evaluate the influence of geometric and structural parameters on
movements develop at the corners as compared to the center horizontal soil deformation, as summarized in Table 1. One-
of the excavation wall. Further, movements near the center of quarter of an excavation was represented to take advantage of
the excavation wall calculated by finite-element approaches symmetry, as shown in Fig. 1. The primary length, L, represents
may be different in 2D analysis than in 3D. the side of the wall where the lateral movement is reported, and is
2. For excavations with large distances between a rigid stratum not necessarily the longer of the two sides 共see Table 1兲. L varied
and excavation bottom, 2D calculation of movements near from 20 to 160 m. The secondary length of the wall, B, varied
the center of the excavation wall generally overpredicted the from 10 to 160 m, such that the L / B ratio varied from 0.25 to 4.
measured field response. Results of 3D analysis more closely The smallest excavation modeled was 20 m by 20 m and the larg-
reflected the field response. est was 160 m by 80 m, such that the plan areas analyzed spanned
3. The 2D and 3D calculation of movements near the center of those of typically sized excavations in urban areas. The excava-
a “long wall” are similar for an excavation with a rigid layer
immediately below the excavation bottom.
4. For smaller ratios of length of wall to height of excavation
共L / He兲, the 2D analysis overestimated the amount of move-
ment which would occur near the center of the excavation
wall, whereas the results of the 3D analysis better agreed
with the measured movements.
In summary, both field and numerical studies show that the 3D
effects depend on the plan geometry and depth of excavation,
support system stiffness, and depth to a rigid stratum below the
excavation. However, no systematic evaluation of all these factors
has been made. The following parametric study addresses these
influences. Fig. 1. Finite-element mesh

JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / JANUARY 2007 / 31

Downloaded 11 Jun 2012 to 180.211.192.67. Redistribution subject to ASCE license or copyright. Visit http://www.ascelibrary.org
 Table 2. Hardening Soil Parameters Used in Parametric Study
Soft Medium Stiff
Parameter Sand clay clay clay
ref
E50 共kPa兲 7,185 421 1,284 17,723
ref
Eoed 共kPa兲 7,185 295 884 12,406
cref 共kPa兲 1 1 1 1
␾ 共°兲 37 24 26 32
⌿ 共°兲 5 0 0 0
M 0.5 0.8 0.85 0.85

plasticity. Plastic strains are calculated assuming multisurface

yield criteria. Isotropic hardening is assumed for both shear and
volumetric strains. The flow rule is nonassociative for frictional
shear hardening and associative for the volumetric cap. The pa-
Fig. 2. Subsurface conditions for parametric studies
rameters used for the HS soil model are representative of the
compressible glacial clays in the Chicago area, and were obtained
tion depth, He, varied from 9.8 to 16.3 m, such that L / He varied by comparing computed values of lateral movements with those
from 0.5 to 12. The side boundaries of the mesh are constrained measured at the Ford Design Center excavation in Evanston, Ill.
by “roller” fixities to prevent displacement in the perpendicular 共Blackburn 2005兲 using inverse analysis techniques and are listed
direction to the boundary and the bottom boundary prevents dis- in Table 2. These parameters include the friction angle, ␸, cohe-
placements both horizontally and vertically. In all cases, the mesh sion, c, dilation angle, ␺, the reference secant Young’s modulus at
boundaries were located at least 120 m from the excavation the 50% stress level, Eref
50 , the reference oedometer tangent modu-
boundary. This distance was approximately seven times the maxi- lus, Eref
oed, and the exponent m which relates reference moduli to
mum depth of excavation, He, which exceeds the minimum dis- the stress level dependent moduli 共E representing E50 and Eoed兲

冉 冊
tance to the mesh boundary of 5He, recommended by Roboski
c cot ␾ − ␴⬘3 m
共2004兲. The excavation was supported by three or four levels of E = Eref 共2兲
lateral support. The walls were “wished” into place for all analy- c cot ␾ + pref
ses 共i.e., installation of the wall caused no stress changes or dis- where pref = reference pressure equal to 100 stress units and
placements in the surrounding soil兲. Soil was excavated uniformly ␴⬘3 = minor principal effective stress. Note that finite-element re-
1 m below each support level prior to adding the support. sults previously reported in literature will be compared later in
As shown in Fig. 2, two soil stratigraphies were considered to this paper with results based on the HS model and the given set of
evaluate the influences of distance to a stiff layer and basal sta-
bility on 3D restraining effects. The base case soil stratigraphy
and support system geometry used in this analysis corresponds to Table 3. Wall Stiffness Parameters
the Lurie Excavation 共Finno and Roboski 2005兲 in Chicago, and
Wall
is shown in Fig. 2共a兲. The water table is located at an elevation of
0 m Chicago city datum 共CCD兲. The major difference between Parameter Flexible Medium Stiff
the stratigraphies is the depth between the excavation bottom and
an underlying stiffer clay layer. Fig. 2共a兲 represents a shallower Plane strain FE parameters
depth, resulting in factors of safety against basal heave 共Terzaghi System stiffness, S 32 320 3,200
1943兲 of 1.6–1.8, which are larger than the values of factor of Bending stiffness, 50,400 504,000 5,040,000
safety of 1.28 to 1.42 computed based on the stratigraphy with a EI 共kN m2 / m兲
greater depth to a stiff layer shown in Fig. 2共b兲. Axial stiffness, 3,427,000 34,270,000 342,700,000
The wall system stiffness, S 共Clough et al. 1989兲 is EA 共kN/m兲
Element thickness 共m兲 0.42 0.42 0.42
EI Poisson’s ratio 0 0 0
S= 共1兲
␥ wh 4 Three-dimensional FE parameters
where EI bending stiffness of the wall; h = average vertical spac- Young’s modulus, 8,160,000 81,600,000 816,000,000
E1 共kPa兲
ing of lateral support elements; and ␥w = unit weight of water.
Young’s modulus, 408,000 4,080,000 40,800,000
Values of 32, 320, and 3,200 were used to represent flexible,
E2 共kPa兲
medium, and stiff walls, respectively. The 3,200 value is higher
Young’s modulus, 200,000,000 2,000,000,000 20,000,000,000
than most “stiff” wall systems and represents an upper bound on
E3 共kPa兲
the system stiffness. The depth of embedment of the wall was at
Shear modulus, 408,000 4,080,000 40,800,000
least 20% of the exposed height of the wall in all cases to prevent
G12 共kPa兲
the toe of the wall from excessively deforming toward the
Shear modulus, 400,000 4,000,000 40,000,000
excavation.
G13 共kPa兲
Shear modulus, 1,330,000 13,300,000 133,000,000
Soil and Structural Parameters G23 共kPa兲
Poisson’s ratio 0 0 0
All soil layers were modeled using the HS model. This effective
Element thickness 共m兲 0.42 0.42 0.42
stress model is formulated within the framework of elasto-

32 / JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / JANUARY 2007

Downloaded 11 Jun 2012 to 180.211.192.67. Redistribution subject to ASCE license or copyright. Visit http://www.ascelibrary.org
 Fig. 4. Effect of plan dimensions and excavation depth on PSR

ments. The movements computed by the 3D analysis are less than

those computed by plane strain simulations for the smaller exca-
vations but are almost the same for the larger excavations.

Effects of Excavation Size and Depth

Fig. 3. Lateral soil movements behind the wall: plane strain versus
3D The influence of excavation geometry on lateral soil displacement
is evaluated by comparing the PSR values for several normalized
geometric parameters. Fig. 4共a兲 shows the relationship between
parameters to show the relative insensitivity of the PSR to the PSR and the ratio of primary wall length to elevation depth, L / HE
assumed soil model. A detailed description of the parameters used based on all cases shown in Table 2. While a variety of wall
to model the internal bracing in 3D is presented by Blackburn and stiffness, soil stratigraphy, and soil models were employed to de-
Finno 共2006兲. Table 3 summarizes the wall stiffness parameters. velop these results, the general trends in the PSR are similar. The
The horizontal bending stiffness is computed assuming that the trends indicate that L / HE ratios greater than 6 result in an exca-
wall is 20 times more flexible in the horizontal direction 共the vation response which has a PSR approximately equal to 1, thus
two-direction in Table 3 indicating the direction along the length suggesting that results of plane strain and 3D analyses will yield
of the wall or the horizontal direction兲 to account for the rotations the same maximum wall displacement in the center of the ex-
in the connections of a sheet pile wall and the lack of continuity cavation. Large differences between plane strain and 3D re-
in stiffer wall systems in this direction. sponses are apparent when L / He is less than 2, implying that as
the excavation gets deeper relative to its length, more restraint is
provided by the sides of the excavation. Fig. 4共b兲 shows the same
Geometry, Stiffness, and Base Stability Effects results plotted versus L / B ratio. When this ratio is less than or
equal to 2, L / HE must be taken into consideration for determining
the PSR. Smaller values are apparent for L / B values less than 1,
General Trends
indicative of movement on the shorter side of the excavation.
To illustrate the pattern of lateral deformations, ␦h, Fig. 3 shows Note that there is less scatter in the PSR-L / He plot in Fig. 4共a兲
the results of 2D and 3D calculations for both the 20 m by 20 m than in the PSR-L / B plot in Fig. 4共b兲 suggesting that of these two
and 80 m by 80 m excavations. The lateral deformations repre- geometric parameters, L / He is more influential in defining the
sent those at the end of the excavation for a vertical line located PSR. The scatter in Fig. 4共a兲 can be attributed to the L / B ratio,
3 m behind the center of the wall. Results are presented for system stiffness, and FSBH.
excavation depths of 9.8, 13.4, and 16.3 m. The maximum move-
ments occur slightly below the bottom of the excavated surface.
Effects of Wall Stiffness
Note that very little cantilever movements occur, and thus the
results presented hereafter are applicable to excavations where Fig. 5 shows the PSR-L / HE relationship for each excavation
this type of movement is minimized by installing the first level of depth and wall stiffness. For L / HE values less than 2, whereas
support prior to the development of significant cantilever move- the PSR values of the medium and flexible walls increase as

JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / JANUARY 2007 / 33

Downloaded 11 Jun 2012 to 180.211.192.67. Redistribution subject to ASCE license or copyright. Visit http://www.ascelibrary.org
 Fig. 7. Parameter k as a function of support system stiffness

Guidelines for Estimating Plane Strain Ratio

While Figs. 4–6 provide guidelines for estimating PSR for the
conditions shown, to quantify more generally the influences of
L / He, L / B, S, and FSBH on PSR, the following equation was
Fig. 5. Effects of support system stiffness developed from the results of the finite-element parametric study:

PSR = 共1 − e−kC共L/He兲兲 + 0.05共L/B − 1兲 共3兲

where C = factor that depends on the FSBH and k = factor that
the excavation becomes deeper, the stiff wall PSR values remain depends upon system stiffness. Eq. 共3兲 was found by separately
low 共approximately 0.6兲, indicating additional corner restraint due evaluating the changes in PSR from a base case excavation for
to the higher wall stiffness. each main parameter. The base case represents a square excava-
tion with a FSBH approximately equal to 1.8 supported by a flex-
Effects of Factor of Safety against Basal Heave ible wall with S equal to 32. In the base case, both k and C are
equal to 1.
Fig. 6 shows the relationship between PSR and L / HE for flexible The value of k in Eq. 共3兲 depends on the support system stiff-
and stiff walls and different factors of safety. The trend lines ness, S, and is taken as
plotted through the data show that for an L / HE ratio of six or
greater, the PSR value is one, regardless of stratigraphy and sta- k = 1 − 0.0001共S兲 共4兲
bility. For all wall stiffnesses, the PSR decreases with smaller This relation is illustrated in the results shown in Fig. 7 where k
L / HE values. For flexible walls, the FSBH has relatively little ef- values based in Eq. 共2兲 for the computed PSR for all L / B = 1 data
fect on the PSR values. In contrast, stiff walls have a greater are plotted for the three system stiffness values. The trends in the
restraining effect, and thus a smaller value of PSR when the FSBH computed responses are reasonably represented by Eq. 共4兲.
is lower. The effects of wall stiffness on PSR as predicted via Eq. 共3兲
are shown in Fig. 8. An excavation depth of 13.4 m with
L / B = 1 is shown to illustrate the following trends. Stiff walls
result in lower values of PSR for a given geometry with the

Fig. 6. Effects of basal stability Fig. 8. Effects of support system stiffness for L / B = 1

34 / JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / JANUARY 2007

Downloaded 11 Jun 2012 to 180.211.192.67. Redistribution subject to ASCE license or copyright. Visit http://www.ascelibrary.org
 Fig. 9. Parameter C as a function of basal stability

largest effects at small values of L / He. As L / He gets larger, the

effects of wall stiffness become less pronounced.
The value of C in Eq. 共3兲 depends on the factor of safety
against basal heave, FSBH, and is taken as
Fig. 11. Computed and empirically derived lateral movements along
C = 1 − 兵0.5共1.8 − FSBH兲其 共5兲 wall
This relation is illustrated in the results shown in Fig. 9 where C
values are based on Eq. 共2兲 for all cases where the PSR was less
than 0.9. The trends in the computed responses are reasonably with a combined kC constant in Eq. 共2兲 equal to 0.5. A kC value
represented by Eq. 共5兲. equal to 0.5 reflects a range of extreme conditions that would
A comparison of the predictions made via Eq. 共2兲 and results influence the PSR, including a very stiff wall 共S = 5,000兲 in stable
of all parametric studies presented herein and those presented in material 共FSBH = 1.8兲 to a flexible wall 共S ⬃ 30兲 in an unstable
literature is made in Fig. 10. In the figure, the solid line represents material 共FSBH = 0.8兲. Also, the second term in Eq. 共3兲 is assigned
a base case with the k and C constants equal to 1. This corre- a value of 0.0375 for calculation of the lower bound curve. This
sponds to a flexible support system 关where the 0.0001S term is corresponds to a low L / B ratio 共0.25兲, which represents the case
negligible in Eq. 共3兲兴 and a factor of safety against basal heave where the length of the primary wall is much shorter than the
共FSBH兲 greater than or equal to 1.8. secondary wall. Large corner restraining effects would be ex-
The dashed lines in Fig. 10 represent upper and lower bounds pected for this case, thus the PSR value is reduced for all L / He
of Eq. 共3兲. The upper bound shown in Fig. 10 corresponds to a values. In spite of different soil models and assumptions used in
base curve 共k and C equal to 1兲 with an excavation geometry term making the finite-element analyses, the trends in the finite-
共L / B兲 greater than 4. This geometry approximates plane strain element results are reasonably represented by limits computed
conditions and, therefore, a PSR value close to unity for all L / He from Eq. 共3兲.
values would be expected. The lower bound curve is calculated In summary, the magnitude of the corner effects depends on
the geometry of the excavation, the stiffness of the lateral support
system, and the factor of safety against basal heave. In general,
greater corner effects are observed for relatively deep excava-
tions, as evidenced by small L / He values, on the shorter of the
two walls, as evidenced by L / B values less than 1, for stiff walls
and for lower factors of safety against basal heave. Conversely,
when L / He is larger than 6, plane strain and 3D analyses yield the
same maximum movements in the center of the excavation for the
range of conditions analyzed herein.

Computed Horizontal Displacements

Fig. 11 compares the computed horizontal soil deformation dis-

tributions for the 20 m by 20 m and 80 m by 80 m excavations at
an excavation depth of 13.4 m. The horizontal distributions cor-
respond to the displacement at an elevation of −7.0 m, which is
the approximate elevation of the maximum horizontal displace-
ment. For each excavation geometry, the maximum displacement
Fig. 10. Comparison between published data and results of decreases with increasing wall stiffness, however the distribution
parametric study shapes do not differ significantly for each wall stiffness. The

JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / JANUARY 2007 / 35

Downloaded 11 Jun 2012 to 180.211.192.67. Redistribution subject to ASCE license or copyright. Visit http://www.ascelibrary.org
shape of the distributions differs between the excavation geom- References
etries, as the zone of “plane-strain” deformation is much greater
for the 80-m-wide excavation. Also shown is the distribution of Blackburn, J. T. 共2005兲. “Automated remote sensing and three-
the lateral movements as computed by a complementary error dimensional analysis of internally braced excavations.” Ph.D. disser-
function defined as 共Finno and Roboski 2005; Roboski and Finno tation, Northwestern Univ., Evanston, Ill.
2006兲 Blackburn, J. T., and Finno, R. J. 共2006兲. “Representing internal bracing
systems in 3D models of deep excavations.” Proc., GeoCongress

冉 冋 册冊
冢 冢 冣冣
He 2006, Geotechnical Engineering in the Information Technology Age,
2.8 x + L 0.015 + 0.035 ln ASCE, Atlanta, in press.
1 L

冋 册
␦共x兲 = ␦max 1 − ⴱ erfc Bono, N. A., Liu, T. K., and Soydemir, C. 共1992兲. “Performance of an
2 He internally braced slurry-diaphragm wall for excavation support.”
0.5L − L 0.015 + 0.035 ln Slurry walls: Design, construction, and quality control, ASTM STP
L
1129, D. B. Paul, R. G. Davidson, and N. J. Cavalli, eds., ASTM,
共6兲 Philadelphia, 169–190.
Chew, S. H., Yong, K. Y., and Lim, A. Y. K. 共1997兲. “Three-dimensional
Use of this equation requires only geometry and an estimate finite-element analysis of a strutted excavation.” Computer methods
of the maximum movement. As can be seen in Fig. 11, the and advances in geomechanics, Yuan, ed., Balkema, Rotterdam, The
complementary error function better matches the distributions Netherlands.
for the flexible wall than for the stiff wall, although reasonable Clough, G. W., Smith, E. M., and Sweeney, B. P. 共1989兲. “Movement
agreement is seen in all cases. Eq. 共6兲 was derived based on the control of excavation support systems by iterative design.” Current
observations at an excavation with a flexible wall. It has been principles and practices, Foundation Engineering Congress, Vol. 2,
shown to provide reasonable agreement for movements reported ASCE, Reston, Va., 869–884.
in literature for stiffer walls 共Roboski and Finno 2006兲, however, Finno, R. J., and Bryson, L. S. 共2002兲. “Response of a building adjacent
it is not altogether unexpected that better agreement is found for to stiff excavation support system in soft clay.” J. Perform. Constr.
flexible walls. Facil., 16共1兲, 10–20.
Finno, R. J., and Calvello, M. 共2005兲. “Supported excavations: The ob-
servational method and inverse modeling.” J. Geotech. Geoenviron.
Eng., 131共7兲, 826–836.
Conclusions Finno, R. J., and Roboski, J. F. 共2005兲. “Three-dimensional responses of
a tied-back excavation through clay.” J. Geotech. Geoenviron. Eng.,
Based on the results of the finite-element parametric studies pre- 131共3兲, 273–282.
sented herein, the following conclusions can be drawn concerning Lee, F. H., Yong, K. Y., Quan, K. C. N., and Chee, K. T. 共1998兲. “Effect
differences between displacements computed by plane strain and of corners in strutted excavations: Field monitoring and case histo-
3D analyses of supported excavations in clays: ries.” J. Geotech. Geoenviron. Eng., 124共4兲, 339–349.
1. The PSR, defined herein as the maximum lateral movement Lin, D. G., Chung, T. C., and Phien-wej, N. 共2003兲. “Quantitative evalu-
behind a wall found from the results of a 3D simulation ation of corner effect on deformation behavior of multi-strutted deep
excavation in Bangkok subsoil.” Geotech. Eng., 34共1兲, 41–57.
normalized by that from a plane strain simulation, depends
Ou, C. Y., and Chiou, D. C. 共1993兲. “Three-dimensional finite-element
on geometry expressed as L / He and L / B ratios, wall system
analysis of deep excavation.” Proc., 11th Southeast Asian Geotechni-
stiffness, and factor of safety against basal heave, as ex-
cal Conf., The Institute of Engineers, Malaysia, Kuala Lumpur,
pressed in Eq. 共3兲. Malaysia, 769–774.
2. When L / He is greater than 6, the PSR is equal to one and Ou, C. Y., Chiou, D. C., and Wu, T. S. 共1996兲. “Three-dimensional finite-
results of plane strain simulations yield the same displace- element analysis of deep excavations.” J. Geotech. Engrg., 122共5兲,
ments in the center of an excavation as those computed by a 337–345.
3D simulation. Ou, C. Y., Liao, J. T., and Cheng, W. L. 共2000a兲. “Building response and
3. All else being equal, smaller values of L / B produce lower ground movements induced by a deep excavation.” Geotechnique,
PSR than higher values. 50共3兲, 209–220.
4. All else being equal, stiff wall systems produce lower PSR Ou, C. Y., Shiau, B. Y., and Wang, I. W. 共2000b兲. “Three-dimensional
than flexible wall systems. deformation behavior of the Taipei National Enterprise Center
共TNEC兲 excavation case history.” Can. Geotech. J., 37, 438–448.
5. All else being equal, excavations with lower factors of safety
Roboski, J. F. 共2004兲. “Three-dimensional performance and analyses of
against basal heave, FSBH, produce lower PSR than excava- deep excavations.” Ph.D. dissertation, Northwestern Univ., Evanston,
tions with higher FSBH. Ill.
Roboski, J. F., and Finno, R. J. 共2006兲. “Distributions of ground move-
ments parallel to deep excavations.” Can. Geotech. J., 43共1兲, 43–58.
Schanz, T., Vermeer, P. A., and Bonnier, P. G. 共1999兲. “The hardening
Acknowledgments soil model—Formulation and verification.” Proc., Plaxis Symp.
“Beyond 2000 in Computational Geotechnics,” Amsterdam,
Financial support for this work was provided by National Science Balkema, The Netherlands, 281–296.
Foundation Grant No. CMS-0219123 and the Infrastructure Tech- Terzaghi, K. 共1943兲. Theoretical soil mechanics, Wiley, New York.
nology Institute 共ITI兲 of Northwestern University. The support of Wong, L. W., and Patron, B. C. 共1993兲. “Settlements induced by deep
Dr. Richard Fragaszy, program director at NSF, and Mr. David excavations in Taipei.” 11th Southeast Asian Geotechnical Conf.,
Schulz, ITI’s director, is greatly appreciated. Singapore.

36 / JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / JANUARY 2007

Downloaded 11 Jun 2012 to 180.211.192.67. Redistribution subject to ASCE license or copyright. Visit http://www.ascelibrary.org