Professional Documents
Culture Documents
Abstract: This study proposes a limit equilibrium approach to estimating the bearing capacity of strip footings placed on geosynthetic-
reinforced soil structures (GRSSs). To assess the multiple mechanisms that may govern the ultimate resistance sustained by GRSSs,
logarithmic-spiral, two-part wedge, two-sided general shear, one-sided general shear, and failure above the uppermost geosynthetic layer
are proposed. Each of these mechanisms is assessed considering geometry, geotechnical properties, and geosynthetic rupture, whereupon an
algorithm selects the minimum, critical bearing capacity and associated failure mechanism for design. Additionally, the effects of foundations
placed near the transition between reinforced and unreinforced soil are evaluated. Considering these factors, both the failure mechanism
and bearing capacity attained from this analysis are compared with rigorous numerical models, demonstrating agreement. The multimechan-
ism approach is then extended to assess bearing capacity considering various geometric configurations and material properties.
Finally, a set of dimensionless charts are presented for convenient assessment of the ultimate bearing capacity of strip footings placed
on GRSSs. DOI: 10.1061/(ASCE)GT.1943-5606.0002072. © 2019 American Society of Civil Engineers.
Introduction However, there are few analytical solutions for evaluating bearing
capacity on reinforced structures considering multiple failure
Geosynthetic-reinforced soil structures (GRSSs) are a proven, eco- mechanisms and the relationship between footing location and
nomical means of earth retention, slope stabilization, and, in recent reinforced structure geometry.
years, foundation support. Limited space may dictate the need Numerous laboratory-scale experimental studies have been per-
to place bridge decks or building foundations on geosynthetic- formed to evaluate the load-carrying behavior of reinforced soil.
reinforced soil. The design of GRSSs includes evaluation of the Selvadurai and Gnanendran (1989) conducted a test with a footing
service limit state (SLS), ultimate limit state (ULS), and extreme located at the crest of a 400-mm-tall slope reinforced with one layer
event limit state (EELS) (Nicks et al. 2016). The use of GRSSs of geogrid, demonstrating that reinforcement properties (stiffness,
with closely spaced reinforcements has shown good performance tensile tension) and location affect bearing capacity. Huang et al.
under serviceability conditions attributed to the composite behavior (1994) performed scaled tests where reinforced and unreinforced
of the reinforced mass (Ketchart 1997; Abu-Hejleh et al. 2000, slopes were loaded to failure with a strip footing, observing that
2002; Elias et al. 2001). However, evaluation of the ULS is conven- the bearing capacity was improved by reinforcement location.
tionally evaluated through limit equilibrium (LE) or limit analysis Lee and Manjunath (2000) demonstrated that bearing capacity
(LA) approaches. Evaluating the ULS requires the assessment of decreased with increasing slope angle and decreasing clearance be-
various potential failure mechanisms, particularly when consider- tween the crest and footing. Yoo (2001) performed laboratory-scale
ing reinforced soil or proximity to a slope (Leshchinsky 2015;
tests on a geogrid-reinforced slope to validate finite-element analy-
Leshchinsky and Xie 2017). When considering the ULS on a
ses and concluded that the zone of shear for reinforced slope was
GRSS, one must consider the contributions of structure geometry,
wider and deeper than the unreinforced case and that the footing
reinforcement spacing, soil–reinforcement interaction, mobiliza-
location and reinforced extents influence realized bearing capacity.
tion of reinforcement tension, footing location, and footing size.
Similar tests on marginal backfills, such as soft clay (El Sawwaf
To date, research that has focused on the ULS of bearing resistance
2007) and fly ash (Choudhary et al. 2010; Gill et al. 2013), have
of GRSSs includes semiempirical approaches (Adams et al. 2014),
demonstrated similar trends. Xiao et al. (2016) conducted a series
experimental testing (Selvadurai and Gnanendran 1989; Yoo 2001;
of tests on small-scale, vertical geogrid-reinforced walls with a seg-
Xiao et al. 2016), and numerical modeling (Yoo and Kim 2008).
mental block facing, investigating the relationship between bearing
1
Staff Geotechnical Engineer, Earth Engineers, Inc. 2411 SE 8th Ave., capacity, footing location, footing width, reinforcement length,
Camas, WA 98607. and connection type. This study observed a transition in failure
2
Assistant Professor, Dept. of Forest Engineering, Resources and mechanism as the footing location was placed further from the
Management, Oregon State Univ., 280 Peavy Hall, Corvallis, OR 97331 GRSS face, transitioning from a two-part wedge to a log-spiral and
(corresponding author). Email: ben.leshchinsky@oregonstate.edu eventually a general shear failure when the clearance was suffi-
3
Professor, Civil, Environmental, and Architectural Engineering Dept., ciently large.
Univ. of Kansas, 2150 Learned Hall, 15th St., Lawrence, KS 66045. Owing to the constraints in labor, material needs, and loading
ORCID: https://orcid.org/0000-0003-3137-733X
capacity, few large-scale loading tests have been performed on re-
Note. This manuscript was submitted on March 12, 2018; approved on
January 16, 2019; published online on June 26, 2019. Discussion period inforced slope or wall-supported foundations. Bathurst et al. (2003)
open until November 26, 2019; separate discussions must be submitted for performed an experimental investigation evaluating bearing capac-
individual papers. This paper is part of the Journal of Geotechnical and ity of a strip footing placed on unreinforced and geosynthetic-
Geoenvironmental Engineering, © ASCE, ISSN 1090-0241. reinforced slopes with a height of 3.4 m. The reinforced slope
Due to the complexity associated with the ULS of reinforced log-spiral mechanism has been observed under various numerical
soil, few analytical methods evaluate the bearing capacity of shal- and physical experiments on bearing capacity, particularly when
low foundations on reinforced soil structures. Zhao (1996a) devel- the footing is close to the slope face or the slope inclination is
oped slip-line solutions for reinforced soils to calculate the limit steep (e.g., Selvadurai and Gnanendran 1989; Huang et al. 1994;
loads on geosynthetic-reinforced soil slopes, treating the reinforced Sommers and Viswanadham 2009). However, a two-part wedge
mass as a composite. Within this study, stress characteristic fields failure surface has been observed in other relevant studies, particu-
were investigated to better understand developed regions of plas- larly apparent when the footing is close to the slope and the incli-
ticity within the reinforced mass. Zhao (1996b) built on this work, nation of the GRSS would classify it as a slope (i.e., inclinations
presenting a kinematically admissible limit analysis approach that less than 70°) according to AASHTO (2016) guidelines (Gourc
assessed the internal stability of geosynthetic-reinforced soil slopes et al. 1995; Bathurst et al. 2003; Xiao et al. 2016). Xie and
considering rotational and translational mechanisms. Jahanandish Leshchinsky (2015) observed a transition from a log-spiral to
and Keshavarz (2005) extended the slip-line method to analyze asymmetrical two-sided, general shear mechanism with increasing
the bearing capacity of spread footings on geosynthetic-reinforced clearance from a GRSS face. Failure above the uppermost reinforce-
soil structures under seismic loading. Wu and Pham (2013) devel- ment has also been recognized as a common failure mechanism of
oped an analytical solution to evaluate the load-carrying capacity
strip footing located at the reinforced soil (Binquet and Lee 1975;
of closely spaced geosynthetic-reinforced soil, treating the rein-
Wayne et al. 1998). In this study, five different failure mechanisms—
forced structure as a composite. The composite behavior of the re-
two-part wedge, log-spiral, two-sided, one-sided general shear
inforced soil was evaluated with a modified cohesive shear strength
mechanisms, and failure above the uppermost geosynthetic layer—
representative of the effects of closely spaced reinforcements. Sim-
were considered simultaneously to evaluate the bearing capacity on
ilar to the work of Wu and Pham (2013), studies by Zhao (1996a, b)
a GRSS. The results of this analysis are presented in a series of nor-
and Jahanandish and Keshavarz (2005) treated reinforced soil as
malized design charts with varying reinforcement tensile tension,
a single, composite material. However, few analytical methods
footing location, footing width, GRSS height, and GRSS slope.
can be used to evaluate the multiple failure mechanisms associated
By accounting for these parameters, the transition in failure mecha-
with the interaction of GRSS geometry, soil properties, and discrete
reinforcements. nism and bearing capacity can be defined for various design scenar-
The bearing capacity of foundations placed upon GRSSs has ios. The observed transition in mechanism is important because it
conventionally been assessed using slope stability analyses. Using captures the influence of the proximity of shallow foundations near
Janbu’s generalized slope stability analysis, Huang and Tatsuoka a GRSS face or near the retained soil.
(1994) developed an analytical method that considers composite
failure surfaces to evaluate the stability of reinforced sand slopes.
Haza et al. (2000) used a LE-based double-wedge method to evalu- Theoretical Framework
ate the stability of a GRSS subject to localized loading from a strip
footing. Blatz and Bathurst (2003) adopted the conventional two- In this study, a rigid strip footing of width B is placed at a distance
part wedge LE approach to predict the bearing capacity of a strip of Sb from the top of a slope with a clearance of St from the
footing placed on the crest of geosynthetic-reinforced slopes based transition between reinforced and retained backfill (Fig. 1). The
on comparisons with large-scale experimental results (Bathurst GRSS has a height of H and an inclination of β. Geosynthetic rein-
et al. 2003). Luo and Bathurst (2018) used finite-element modeling forcements with tensile strength T i are evenly placed throughout
to consider the effects of compaction and spatial variability on bear- the height of the GRSS at a vertical spacing of Sv and length
ing capacity on reinforced slopes. Ausilio (2014) used upper-bound Lreinf . The distance from footing to the top layer of reinforcement
limit analysis to develop the seismic bearing capacity of strip is u. For a given failure mechanism, the convention for reinforce-
footings placed close to the crest of geosynthetic-reinforced soil ment numbering in the analysis begins with 1 at the top of the wall
structures using a log-spiral mechanism. Within AASHTO (2016) and increases with distance from the footing. The footing supports
guidelines, a bilinear failure mechanism is proposed, consisting of a surcharge pressure of q (or Q as a resultant force). Both the
vertical block and underlying active wedge underneath a footing reinforced and retained soils are cohesionless materials with an in-
placed on reinforced soil (Berg et al. 2009), exhibiting a slip surface ternal angle of friction ϕ 0 and unit weight γ. The effects of footing
that, in an average sense, is similar to that of a log-spiral. Xiao embedment, footing shape, and cohesion are ignored in the pre-
et al. (2016) compared rotational and two-part wedge stability sented analysis, which generally provides more conservative re-
analyses, finding that the latter often governed. They noted that sults. Furthermore, the influence of facing elements, often used
bearing capacity increased with footing setback, but dropped as in steeper GRSSs (i.e., walls), is ignored in this study because it
the footing reached the transitional zone between the reinforced fill is a conservative and common practice in design. Consistent with
and retained soil. Xie and Leshchinsky (2015) used computational current practice, the mechanical effects of facing elements are
upper-bound limit analysis to evaluate the bearing capacity of ignored for conservatism but would influence observed bearing
spread footings on soil with varying reinforcement spacing and capacity and its corresponding mechanism. The influence of
1L L where L ¼ Sb þ B. The moment induced by the surcharge load q is
f2 ¼ 2 cos θo − sin θo ð3bÞ
6 ro ro
B
M Q ¼ qB ro cos θo − ð4Þ
2
1 L
f 3 ¼ exp½ðθh − θo Þ tan ϕ 0 sinðθh − θo Þ − sin θh The moment due to reinforcement is calculated as
6 ro
X
N
L zi
× cos θo − þ cos θh exp½ðθh − θo Þ tan ϕ 0 ð3cÞ M D ¼ ro T i sin θo þ ð5Þ
ro i¼1
ro
sv ; Lreinf Þ, the backfill ðγ; ϕ 0 Þ, the footing ðB; Sb ; St Þ, and the mo- 2003). Future work may better consider the stress-strain response
bilized reinforcement tensions (T i ), the bearing capacity consider- and the role of soil deformation in the calculation of these limit state
ing a log-spiral, two-part wedge, asymmetric two-sided general conditions. Another approach is that of Sharma et al. (2009), which
shear mechanism, one-sided general shear mechanisms, and failure considers elastic settlement and corresponding reinforcement mo-
above the uppermost geosynthetic layer can be estimated. Of the bilization under loading through modification of the Schmertmann
bearing capacities determined from the log-spiral, two-part wedge, et al. (1978) approach. However, this approach necessitates knowl-
asymmetric two-sided general shear and failure above the upper- edge of the soil stiffness and correction factors for materials and
most geosynthetic mechanisms, the lowest estimated bearing geometry. Alternatively, Liu and Won (2014) developed an analyti-
capacity, and its associated mode of failure will govern in consid- cal approach to evaluating horizontal strains and forces in rein-
eration of bearing failure influenced by the slope face. The bearing forcements assuming full strain compatibility at the location of
capacity derived from the one-sided general shear mechanism will the maximum reinforcement tensile tension. This was done through
govern when in the range of influence of the retained soil. The consideration of a hyperbolic stress-strain relationship that consid-
lower value of these two cases yields the ultimate bearing capacity ered the dilative behavior of the soil but does not explicitly consider
for a given configuration. However, there are various assumptions the effects of a discrete surcharge. These methods demonstrate util-
associated with the required distribution and magnitude of mobi- ity that could be incorporated into the proposed method. However,
lized reinforcement tensions, described in further detail in the a decision was made here to use the approach proposed by Binquet
following section. and Lee (1975) for one- and two-sided shear mechanisms because it
is simple to implement and demonstrates reasonable agreement
with field studies and numerical models (e.g., Yoo and Kim 2008).
Mobilized Reinforcement Tensile Resistance For log-spiral and two-part wedge mechanisms, the reinforcement
rupture capacity (T ult ) is assigned to each reinforcement layer. This
The described approach presents a means of evaluating the bearing assumption is conservative but consistent with the current state-of-
capacity of strip footings on a GRSS; however, implicit in the practice for assessing the ULS for the internal stability of reinforced
appropriate application of this method is the use of a reasonable soil structures.
estimation of the mobilized reinforcement tensile tension, T i .
Assessing the distribution of mobilized reinforcement resistance
is a function of surcharge load, material extensibility, interaction Verification of Analytical Solutions
with surrounding soil, and soil properties (Hatami and Bathurst
2006). For limit state analyses, it is often held that the reinforcement The presented mechanisms demonstrate a rational means of assess-
rupture capacity (T ult ) may be assumed for each reinforcement layer ing the bearing capacity of GRSSs under reasonable combinations
(Leshchinsky and Boedeker 1989; Blatz and Bathurst 2003). of geometry, reinforcement configurations, and material properties.
However, it is apparent that geosynthetic reinforcements, par- To verify their applicability, a comparison of the proposed approach
ticularly when placed in multiple layers, do not always reach rup- to both experimental and numerical results was performed and
ture as the surrounding soil reaches the critical state at smaller presented, found in the corresponding supplementary files docu-
strains in comparison to those required to rupture geosynthetic rein- ment. From these comparisons it is seen that the critical mecha-
forcements (Sommers and Viswanadham 2009; Allen and Bathurst nism (i.e., the failure geometry that demonstrates the lowest calcu-
2013). This is particularly relevant for localized, general shear fail- lated bearing capacity) transitions with changing input properties.
ures. One conservative means of estimating the distribution of ten- Hence, the proposed approach is compared with results from
sion within reinforcements is to assume that mobilized tension has both available experimental data and a finite-element package
an inverse relationship with increasing quantities of reinforcement that employs upper-bound computational limit analysis (UB-LA,
layers with depth. Thus, the approach of Binquet and Lee (1975) is OptumG2) to determine bearing capacity and its associated
adopted for the assessment of one- and two-sided shear mecha- collapse mechanism. In the numerical comparisons, a strip footing
nisms owing to its simplicity and its agreement with physical tests (B ¼ 1 m) is placed on the reinforced backfill (ϕ 0 ¼ 34°, γ ¼
evaluating bearing capacity on reinforced soil (e.g., Yoo and Kim 20 kN=m3 , Sv ¼ 0.5 m) of a GRSS with a height of H ¼ 6 m
2008). General agreement with numerical models presented in the and a slope inclination (β) ranging from 70° to 90°. The primary
following section supports this assumption. This approach states variable in this comparison was footing setback (Sb ), which was
that for N layers of reinforcement, the mobilized tension in a varied from 0.5 to 6 B. The location of the foundation in comparison
reinforcement is defined as to the retained soil was also assessed. The reinforcement strength
T1 was also varied from 10 to 100 kN=m, demonstrating a notable
Ti ¼ ð19Þ effect on the realized bearing capacity and failure mechanism.
i
Pullout was not observed to be critical in these numerical models,
where i = number of layers within failure zone; and T 1 = mobilized likely owing to the relatively long reinforcement layers. Experimen-
reinforcement force in top reinforcement layer. Because this study tal comparisons are done with tests from Khing et al. (1993),
In general, the proposed failure mechanisms for bearing capac- Zornberg et al. 1998; Yang et al. 2016). Limited sensitivity in re-
ity on a GRSS sufficiently captures relevant failure modes for sults was observed for footing widths (B) ranging from 0.5 to 3 m,
different reinforcement properties, slope geometries, and footing which covers a range of footing widths commonly observed in
setbacks when failure is governed by the presence of both a slope design (Adams et al. 2011). For fixed kt =γB and the geometries
and retained soil. The proposed solutions are comparable to those described, limited sensitivity was observed when varying Sv be-
derived from robust numerical models that do not assume failure tween 0.2 and 0.5B (<15%, limited to only the general shear
mechanisms a priori. The maximum difference between the pre- mechanism, <5% for other mechanisms); thus, the results provide
sented numerical models and proposed solutions for different cases reasonable estimates for the range of dimensions proposed. When
are 10% or less. applicable, retained soil is considered to have the same material
properties as reinforced backfill.
Fig. 7. Normalized bearing capacity for a range of soil friction angles and footing setbacks of (a) Sb =B ¼ 1; and (b) Sb =B ¼ 4.
Fig. 8. Normalized bearing capacity for varying slope inclinations and footing setbacks of (a) Sb =B ¼ 1; and (b) Sb =B ¼ 4.
exponentially with backfill strength, demonstrating the influence of When the footing setback is large (Sb ¼ 4B), the bearing capacity
current, conservative design standards, where backfill strength is and failure mechanism are not affected by the slope face, demon-
assumed to have a ϕ 0 value of 34° in the absence of strength testing strated by the critical mechanism being above the uppermost
(AASHTO 2016). When using more closely spaced reinforcements geosynthetic for all slope inclinations because reinforcement is
(or stronger reinforcements), bearing capacity may increase by strong [i.e., kt =γB ¼ 5, Fig. 8(b)], applicable when the clearance
more than 50% if a more realistic ϕ 0 of 45° is considered. However, between the footing and the uppermost reinforcement is greater
these benefits are less pronounced when weaker overall reinforce- than 0.5 m.
ment is present (Fig. 7).
Influence of GRSS Height
Influence of GRSS Inclination
The height of a GRSS affects the mechanism and load associated
The GRSS facing inclination directly influences the mechanism and with bearing capacity, particularly when a compound slope
load associated with bearing capacity. For a slope of height 6 B, stability–bearing capacity mechanism, such as a log-spiral, is ex-
footing setback of Sb ¼ B and Sb ¼ 4B, and a backfill strength of pected. For a vertical facing (β ¼ 90°), a footing setback of Sb ¼ B,
ϕ 0 ¼ 34° (c 0 ¼ 0), changing the inclination from 90° to 50° results and a backfill strength of ϕ 0 ¼ 34° (c 0 ¼ 0), changing the GRSS
in a slight increase in bearing capacity (Fig. 8). This change—or height from H=B ¼ 1 to 10 results in a decrease in bearing capacity
lack thereof—is governed primarily by a transition in mechanism. that is governed by footing location (Fig. 9). For shorter structures
Relative changes are smaller when the mechanism is constrained to (e.g., H < 2B), the two-part wedge tends to govern for both overall
a localized failure in the upper portion of the GRSS. When the over- reinforcement strengths. However, as the GRSS height increases,
all reinforcement in the GRSS is relatively weak, the bearing capac- the bearing capacity decreases, eventually leveling off because
ity is greatly influenced by the slope, demonstrated by critical the influence of the footing load is minimal compared with the
mechanisms consisting of either a log-spiral or two-part wedge relative stability of the slope [i.e., when i.e., kt =ðγBÞ ¼ 2] or
for steep facing inclinations, respectively [Fig. 8(a)]. However, when it is constrained to the upper reaches of the GRSS
when the overall reinforcement is strong (i.e., strong reinforcements [i.e., kt =ðγBÞ ¼ 5]. For we aker overall reinforcement and a small
or closely spaced, weaker reinforcements) and the strip footing footing setback, the decrease in bearing capacity becomes small
is near the facing, the failure is confined to the upper reaches of when H=B > 4 since the log-spiral dominates bearing capacity
the GRSS. In this scenario, bearing capacity failure is dominated and its critical slip surface geometry is constrained to upper
by a two-sided general shear and two-part wedge mechanism portions of the GRSS. On the other hand, the bearing capacity for
for steep and vertical slope inclinations, respectively [Fig. 8(a)]. weaker overall reinforcement and large footing setback (Sb =B ¼ 4)
Fig. 9. Normalized bearing capacity for varying wall heights and footing setbacks of (a) Sb =B ¼ 1; and (b) Sb =B ¼ 4.
Fig. 10. Charts for bearing capacity of GRSS considering influence of slope face.
decreases when H=B > 4 since a failure mass develops in the from H=B ¼ 1 to 10 for facing inclinations of 90°, 80°, 70°, 60°,
whole structure with increasing structure height. For the given and 50°. Normalized reinforcement values of kt =γB ¼ 1, 2, and 5
geometry with small footing setback and stronger reinforcements, are presented to capture a realistic range in reinforcement strengths
a minimum is reached when H=B > 3, limited to the relative depth and backfill densities. The shear strength of the reinforced and re-
of influence of a two-sided general shear mechanism [Fig. 9(a)]. tained backfill in the absence of laboratory testing is considered to
When footing setback is large and the overall reinforcement is be a conservative ϕ 0 ¼ 34°, compatible with conventional design
strong, failure above the uppermost geosynthetic tends to govern values used in AASHTO (2016).
for structure heights changing from 1 to 10B [Fig. 9(b)]. These charts demonstrate that increasing reinforcement
strength, increasing footing setback, decreasing facing inclination,
and decreasing GRSS height will result in higher bearing capacity
Charts for Bearing Capacity on GRSS values (Fig. 10). In particular, increasing setback and decreasing
facing inclination result in a relatively constant bearing capacity
For convenient evaluation of the bearing capacity on a GRSS and with height as the mechanism is constrained to the upper reaches
its associated failure mechanisms, a series of dimensionless solu- of the GRSS. This benefit can also be achieved by using stronger
tions was generated in chart form. Two sets of charts are presented: reinforcements (or using closely spaced weak reinforcements). In
the first for determining the bearing capacity considering the influ- the absence of such an approach, compound slope stability–bearing
ence of the slope face (Fig. 10) and the second considering the dis- capacity failures manifesting as a log-spiral result in lower bearing
tance between the footing and retained soil (Fig. 11). The minimum capacity with increasing height because the overall structure is less
value determined from these charts defines the critical, ultimate stable. As the foundation nears the transitional zone between rein-
bearing capacity for a given configuration of geometry, soil proper- forced and retained soil, bearing capacity is generally constant, the
ties, and reinforcement layout. The GRSS heights evaluated ranged exception being when St is between 0.25 B and −0.5B (where a
Fig. 11. Charts for bearing capacity of GRSS considering influence of retained soil.
positive St implies that the footing is placed entirely on reinforced charts, three stages of linear interpolation are suggested using
soil and negative implies overlap with retained fill). When a footing bounding values of Sb =B, kt =γB, and β: (1) interpolate the eight
approaches the retained soil, the governing mechanism tends to bounding values of N γ based on kt =γB, (2) interpolate the four
be a one-sided failure toward the unreinforced soil. Increasing modified N γ values based on kt =γB, and (3) interpolate the two
reinforcement strength results in higher bearing capacity when modified N γ values based on β to arrive at an estimated N γ
the footing is placed upon reinforced soil, but it also demonstrates representative of the proposed conditions.
a greater drop in bearing capacity in proximity to unreinforced,
retained fill (Fig. 11).
To find the critical bearing capacity for a given GRSS geometry Example
using the proposed charts, one must know the footing location and To illustrate the application of the proposed solutions, a practical
width, GRSS height and inclination, reinforcement strength, length example is presented. Due to limited right-of-way, a GRSS is to be
and spacing, and backfill properties. One determines the appropri- constructed with an inclination of 90°, a height of 5 m, and a
ate combination of H=B, Sb =B, kt =γB, and β and subsequently cohesionless backfill with ϕ 0 ¼ 34° and γ ¼ 20 kN=m3 . It will
determines a modified bearing capacity factor, N γ . Upon determi- be reinforced with geogrid layers that are 4 m in length and
nation of the modified bearing capacity factor, the bearing pressure have a tensile strength of T ult ¼ 20 kN=m, a vertical spacing of
can be determined by multiplying N γ by 0.5γB. For a given bearing Sv ¼ 0.5 m, and an interface friction of 2/3 the friction angle
capacity, insight into the corresponding failure mechanism can be (δ ¼ 22.7°). The GRSS will support an nonembedded footing that
gained from the presented marker shape—in Fig. 10, the circle, is 1 m wide (B ¼ 1 m) placed 2 m (Sb ¼ 2 m) from the facing
upper triangle, lower triangle, and diamond represent a log-spiral, and 1 m from the retained soil (St ¼ 1 m). Based on these values,
two-part wedge, two-sided general shear failures, and failure above the normalized inputs are Sb =B ¼ 2, St =B ¼ 1, kt =ðγBÞ ¼ 2, and
the uppermost geosynthetic, respectively. For geometries and H=B ¼ 5 for a vertical GRSS. Using Figs. 10 and 11, the modified
reinforcement strengths not explicitly represented in the proposed bearing capacity factors are N γ ¼ 21 and N γ ¼ 47, respectively.
height is indicative that the failure is confined to the upper reaches particularly when a two-sided mechanism is critical. In this case,
of the GRSS. This implies that closely spaced reinforcements may closely spaced reinforcements can be applied to a selected por-
be most advantageous at the crest of the GRSS, a finding consistent tion of the GRSS for improved bearing capacity—namely, an
with observations in the practice and literature (e.g., Xie and upper portion of the GRSS corresponding to the minimum
Leshchinsky 2015; Nicks et al. 2016). This observation shows that H=B that yields a constant bearing capacity and mechanism.
close spacing need not be applied to the entire GRSS in the context Characterizing a margin of safety in design is commonly per-
of bearing capacity but only until a H=B of 3.5 is reached, the depth formed through (1) a reduction of soil shear strength in the case
corresponding to the inflection point representative of a two-sided of slope stability, (2) through a reduction in T ult in the case of in-
shear mechanism governing regardless of depth. This implies that ternal stability of GRSS, and (3) through a comparison of expected
one can attain increased cost efficiency using stronger or more footing loads versus capacity in the case of the bearing capacity of
closely spaced reinforcements in an appropriate portion of a GRSS. shallow foundations. Considering the numerous mechanisms dis-
cussed in this study, it is difficult to propose assessing safe design
using the first two criteria as the governing mechanism may change.
Conclusions However, since the proposed results represent the bearing capacity at
a state of equilibrium, a comparison of demand (i.e., expected ver-
The ultimate bearing capacity of strip footings placed on a GRSS tical footing load) versus capacity (i.e., bearing capacity at equilib-
was evaluated using an analytical approach where a minimum rium) presents a rational means of assessing a margin of safety that is
failure load was determined from five simultaneous failure mech- compatible with the current design of shallow foundations.
anisms: a two-part wedge, log-spiral, two-sided general shear, fail- The load-bearing behavior of a GRSS is complex and governed
ure above the uppermost geosynthetic layer, and one-sided general by potential strain incompatibilities between reinforcement, soil,
shear failure. This approach accounts for GRSS geometry, footing and foundation that are difficult to capture with limit equilibrium
location, backfill properties, and reinforcement spacing, length, and methods. Also omitted from this approach are pressure-dependent
strength. The proposed multimechanism method demonstrated parameters, such as stiffness, dilation, and anisotropic behavior
good agreement with rigorous numerical approaches, capturing commonly associated with structural backfills. While these param-
the influences of both the GRSS slope face and the transition from eters, particularly stiffness, may affect the load-displacement re-
reinforced backfill to retained soil. A sensitivity study was per- sponse of a footing, they are not considered in this study, which
formed and a comprehensive set of solutions presented in a series focused on assessing the ultimate limit state in the context of fully
of dimensionless charts that account for reinforcement spacing/ plastic conditions, i.e., equilibrium. The assessment of the service
strength, footing location, and GRSS inclination. The main conclu- limit state is a critical consideration in the design of GRSSs. Com-
sions of the study are as follows: plex design parameters, such as settlement and deformation, may
• Increasing backfill and reinforcement strength or decreasing be best assessed using calibrated numerical techniques, such as fi-
reinforcement spacing and the ratio of GRSS height to footing nite elements or finite difference modeling. The proposed approach
width will result in higher bearing capacity. Notably, the use extends conventional ULS design approaches for shallow founda-
of conservative backfill strength values in design results in tion design and slope stability to reinforced soil to present a com-
potentially significant underestimation of bearing capacity. prehensive framework for evaluating the bearing capacity on
• The log-spiral mechanism typically governs when the GRSS GRSS. However, as for any complex GRSS, one should account for
slope is steep and overall reinforcement is weak. This mechan- compound stability using LE or LA packages that ensure that a
ism is more likely to occur when the footing is located closer to system will be globally stable.
the facing or the GRSS is tall.
• A two-part wedge mechanism tends to occur when the footing is
located closer to the GRSS face and overall reinforcement is Acknowledgments
strong.
• The two-sided general shear mechanism governs when a footing The primary author would like to acknowledge the Geosynthetics
is located sufficiently far from the slope facing and overall re- Institute (GSI) for partial support of this work. Furthermore, he
inforcement is strong. Use of stronger or more closely spaced would like to honor Jerry Yamumuro, who kindly bequeathed a
reinforcements increases the likelihood of this mechanism fellowship program that partially supported the first author
governing instead of a log-spiral. during this work.
• Failure above the uppermost geosynthetic governs when the
distance from footing to the top layer of reinforcement is large,
the footing is far from the slope facing, and overall reinforce- Supplemental Data
ment is strong.
• The one-sided general shear mechanism is appropriate when the Figs. S1–S9 and Table S1 are available online in the ASCE Library
footing is located within 0.5B of the transitional zone between (www.ascelibrary.org).
Adams, M., J. Nicks, T. Stabile, J. Wu, W. Schlatter, and J. Hartmann. 2011. Jahanandish, M., and A. Keshavarz. 2005. “Seismic bearing capacity of
Geosynthetic reinforced soil integrated bridge system:interim imple- foundations on reinforced soil slopes.” Geotext. Geomembr. 23 (1):
mentation guide. Rep. No. FHWA-HRT-11-026. McLean, VA: Federal 1–25. https://doi.org/10.1016/j.geotexmem.2004.09.001.
Highway Administration. Ketchart, K. 1997. “Performance of geosynthetic-reinforced soil bridge pier
Adams, M. T., P. S. Ooi, and J. E. Nicks. 2014. “Mini-pier testing to es- and abutment, Denver, Colorado, USA.” In Mechanically stabilized
timate performance of full-scale geosynthetic reinforced soil bridge backfill, 101–116. Rotterdam, Netherlands: A.A. Balkema.
abutments.” Geotech. Test. J. 37 (5): 20140007. https://doi.org/10 Khing, K. H., B. M. Das, V. K. Puri, E. E. Cook, and S. C. Yen. 1993. “The
.1520/GTJ20140007. bearing-capacity of a strip foundation on geogrid-reinforced sand.”
Allen, T. M., and R. J. Bathurst. 2013. “Comparison of working stress and Geotext. Geomembr. 12 (4): 351–361. https://doi.org/10.1016/0266
limit equilibrium behavior of reinforced soil walls.” In Sound geotech- -1144(93)90009-D.
nical research to practice, edited by A. W. Stuedlein and B. R. Lee, K. M., and V. R. Manjunath. 2000. “Experimental and numerical
Christopher, 499–513. Reston, VA: ASCE. studies of geosynthetic-reinforced sand slopes loaded with a footing.”
Ausilio, E. 2014. “Seismic bearing capacity of strip footings located close Can. Geotech. J. 37 (4): 828–842. https://doi.org/10.1139/t00-016.
to the crest of geosynthetic reinforced soil structures.” Geotech. Geol. Leshchinsky, B. 2015. “Bearing capacity of footings placed adjacent to c′-ϕ′
Eng. 32 (4): 885–899. https://doi.org/10.1007/s10706-014-9765-4. slopes.” J. Geotech. Geoenviron. Eng. 143 (1): 04015022. https://doi
Bathurst, R. J., J. A. Blatz, and M. H. Burger. 2003. “Performance of .org/10.1061/(ASCE)GT.1943-5606.0001306.
instrumented large-scale unreinforced and reinforced embankments Leshchinsky, B., and Y. Xie. 2017. “Bearing capacity for spread footings
loaded by a strip footing to failure.” Can. Geotech. J. 40 (6): 1067– placed near c 0 -ϕ 0 slopes.” J. Geotech. Geoenviron. Eng. 143 (1):
1083. https://doi.org/10.1139/t03-052. 06016020. https://doi.org/10.1061/(ASCE)GT.1943-5606.0001578.
Berg, R. R., B. R. Christopher, and N. C. Samtani. 2009. Design of me- Leshchinsky, D. 1997. Design procedure for geosynthetic reinforced steep
chanically stabilized earth walls and reinforced soil slopes—Volume slopes. Newark, DE: Leshchinsky, Inc.
I. Rep. No. FHWA-NHI-10-024. Washington, DC: Federal Highway Leshchinsky, D., and R. H. Boedeker. 1989. “Geosynthetic reinforced soil
Administration. structures.” J. Geotech. Eng. 115 (10): 1459–1478. https://doi.org/10
Binquet, J., and K. L. Lee. 1975. “Bearing capacity tests on reinforced earth .1061/(ASCE)0733-9410(1989)115:10(1459).
slabs.” J. Geotech. Geoenviron. Eng. 101 (12): 1241–1255. Liu, H., and M. S. Won. 2014. “Stress dilatancy and reinforcement load of
Blatz, J. A., and R. J. Bathurst. 2003. “Limit equilibrium analysis of vertical-reinforced soil composite: Analytical method.” J. Eng. Mech.
large-scale reinforced and unreinforced embankments loaded by a strip 140 (3): 630–639. https://doi.org/10.1061/(ASCE)EM.1943-7889
footing.” Can. Geotech. J. 40 (6): 1084–1092. https://doi.org/10.1139 .0000686.
/t03-053. Luo, N., and R. J. Bathurst. 2018. “Deterministic and random FEM analysis
Choudhary, A. K., J. N. Jha, and K. S. Gill. 2010. “Laboratory investigation of full-scale unreinforced and reinforced embankments.” Geosynthetics
of bearing capacity behaviour of strip footing on reinforced flyash Int. 25 (2): 164–179. https://doi.org/10.1680/jgein.17.00040.
slope.” Geotext. Geomembr. 28 (4): 393–402. https://doi.org/10.1016/j Mandel, J., and J. Salencon. 1972. “Force portante d’un sol sur une assise
.geotexmem.2009.09.007. rigide (étude théorique).” Geotechnique 22 (1): 79–93. https://doi.org
Elias, V., B. R. Christopher, and R. R. Berg. 2001. Mechanically stabilized /10.1680/geot.1972.22.1.79.
earth walls and reinforced soil slopes-design and construction guide- Meyerhof, G. G. 1963. “Some recent research on the bearing capacity
lines. Publication No. FHWA-NHI-00-043. Washington, DC: Federal of foundations.” Can. Geotech. J. 1 (1): 16–26. https://doi.org/10
Highway Administration. .1139/t63-003.
El Sawwaf, M. A. 2007. “Behavior of strip footing on geogrid-reinforced Michalowski, R. 1997. “An estimate of the influence of soil weight on bear-
sand over a soft clay slope.” Geotext. Geomembr. 25 (1): 50–60. https:// ing capacity using limit analysis.” Soils Found. 37 (4): 57–64. https://
doi.org/10.1016/j.geotexmem.2006.06.001. doi.org/10.3208/sandf.37.4_57.
Gill, K. S., A. K. Choudhary, J. N. Jha, and S. K. Shukla. 2013. “Exper- Nicks, J. E., D. Esmaili, and M. T. Adams. 2016. “Deformations of
imental and numerical studies of loaded strip footing resting on rein- geosynthetic reinforced soil under bridge service loads.” Geotext.
forced fly ash slope.” Geosynthetics Int. 20 (1): 13–25. https://doi Geomembr. 44 (4): 641–653. https://doi.org/10.1016/j.geotexmem
.org/10.1680/gein.12.00036. .2016.03.005.
Gourc, J. P., P. Gotteland, E. Haza, H. Perrier, and E. Baraize. 1995. “Geo- Schmertmann, J. H., J. P. Hartman, and P. R. Brown. 1978. “Improved
textile reinforced structures as bridge abutments: Full-scale experimen- strain influence factor diagrams.” J. Geotech. Geoenviron. Eng.
tation.” In Proc., 5th Int. Conf. on Geosynthetics, 79–92. Jupiter, FL: 104 (8): 1131–1135.
International Geosynthetics Society. Selvadurai, P. S., and C. T. Gnanendran. 1989. “An experimental study of a
Guido, V. A., D. K. Chang, and M. A. Sweeney. 1986. “Comparison of footing located on a sloped fill: Influence of a soil reinforcement layer.”
geogrid and geotextile reinforced earth slabs.” Can. Geotech. J. Can. Geotech. J. 26 (3): 467–473. https://doi.org/10.1139/t89-059.
23 (4): 435–440. https://doi.org/10.1139/t86-073. Sharma, R., Q. Chen, M. Abu-Farsakh, and S. Yoon. 2009. “Analytical
Hatami, K., and R. J. Bathurst. 2006. “Numerical model for reinforced soil modeling of geogrid reinforced soil foundation.” Geotext. Geomembr.
segmental walls under surcharge loading.” J. Geotech. Geoenviron. 27 (1): 63–72. https://doi.org/10.1016/j.geotexmem.2008.07.002.
Eng. 132 (6): 673–684. https://doi.org/10.1061/(ASCE)1090-0241 Sommers, A. N., and B. V. S. Viswanadham. 2009. “Centrifuge model tests
(2006)132:6(673). on the behavior of strip footing on geotextile-reinforced slopes.” Geo-
Haza, E., P. Gotteland, and J. P. Gourc. 2000. “Design method for local load text. Geomembr. 27 (6): 497–505. https://doi.org/10.1016/j.geotexmem
on a geosynthetic reinforced soil structure.” Geotech. Geol. Eng. 18 (4): .2009.05.002.
243–267. https://doi.org/10.1023/A:1016623619511. Terzaghi, K. 1943. Theoretical soil mechanics. New York: Wiley.