Está en la página 1de 12

Ground Improvement (2000) 4, 141±152 141

Simpli®ed models for soil-nail lateral interaction


S. A. TAN, S. Q. LUOy and K. Y. YONG
 Department of Civil Engineering, National University of Singapore; y Presscrete Engineer-
ing Pte Ltd, Singapore

A series of simpli®ed failure modes has been used in this Dans cette eÂtude, nous avons utilise une seÂrie de modes de
paper to describe the behaviour of soil-nail lateral inter- deÂfaillance simpli®eÂs pour deÂcrire le comportement de
action. The criteria for determination of the possible l'interaction lateÂrale sol-clou. Nous donnons les criteÁres de
failure modes for a soil-nail structure are set out on the deÂtermination des modes de deÂfaillance possibles pour
basis of the relative stiffness and strengths of the soil and une structure sol-clou en nous basant sur la rigidite et la
nail and the local relative lateral displacement across the reÂsistance relatives du sol et du clou et sur le deÂplacement
shear zone. This local relative displacement is produced lateÂral relatif local aÁ travers la zone de cisaillement. Ce
along a potential sliding surface of the reinforced soil-nail deÂplacement relatif local se produit le long d'une surface
retaining structure. Expressions for calculation of soil-nail de glissement potentielle dans la structure de souteÁnement
lateral resistance have been derived for each potential renforce sol-clou. Nous en avons deÂduit les expressions
failure mode. It is found that soil-nail lateral resistance is pour le calcul de la reÂsistance lateÂrale sol-clou pour chaque
dependent not only on the relative stiffness and yield mode de deÂfaillance potentiel. Nous avons trouve que la
strengths of the soil and nail but also on the local relative reÂsistance lateÂrale sol-clou deÂpend non seulement de la
lateral displacement across the shear zone. Good agree- rigidite relative et des limites eÂlastiques du sol et du clou
ment among the results of numerical simulation of soil- mais aussi du deÂplacement lateÂral relatif local aÁ travers la
nail lateral performance, published experimental data and zone de cisaillement. Nous deÂmontrons qu'il existe une
theoretical predictions have been demonstrated, to support bonne correÂlation entre les reÂsultats de la simulation
the validity of the proposed approach for estimating soil- numeÂrique de la performance lateÂrale sol-clou, les donneÂes
nail lateral resistance capacity in soil-nailing design. expeÂrimentales publieÂes et les preÂvisions theÂoriques, a®n
de soutenir la validite de l'approche proposeÂe pour estimer
Keywords : Soil-nails; soil-nail lateral interaction and la capacite de reÂsistance lateÂrale sol-clou lors de la concep-
lateral resistance tion d'un cloutage de sol.

Notation Tc soil-nail lateral resistance, shear force (kN)


Tp axial tensile capacity of a nail subjected to tension
d diameter of a soil-nail (mm or m) only (kN)
dcr critical soil-nail diameter (mm or m) ã9 effective unit weight of soil (kN=m3 )
E Young's modulus of steel reinforcement member ä relative soil/nail displacement (mm or m)
of a soil-nail (kPa) äe relative displacement corresponding to the soil
G soil shear modulus (kPa) limit bearing pressure (mm or m)
I moment of inertia of the steel reinforcement ämax maximum relative soil/nail displacement (mm
member of a soil-nail (m4 ) or m)
ID relative density of soils k constant
K bulk modulus (kPa) ì Poisson's ratio
Ka the lateral soil pressure coef®cient ó9 soil reaction pressure on a soil-nail (kPa)
Ks modulus of subgrade reaction (kN=m3 ) ó9b soil limit bearing pressure (kPa)
lb the minimum soil-nail length required beyond the ót normal tension stress on the cross-section of a soil-
point of maximum bending moment (m) nail due to axial tension force (N=mm2 or kN=m2 )
ls width of shear band (m) ó9v vertical effective soil pressure (kPa)
l0 transfer length of a soil-nail (m) óy steel yield strength (N=mm2 or kN=m2 )
Mlimit limit plastic bending moment (kNm) ö9 angle of internal friction (degree)
Mmax maximum bending moment (kNm)
Mp full plastic bending moment (kNm)
S minimum sliding displacement of soil in active
zone along the sliding surface (mm or m)
Introduction
T axial tension force in a nail (kN) Soil-nailing is an in situ soil reinforcement technique. It has
been widely used in stabilization of slopes and excavation
by retaining the native soils with relatively small fully
(GI 075) Paper received 25 March 1999; last revised 19 October 1999; bonded inclusions that are generally steel bars or other
accepted 1 March 2000 metallic elements. Along a potential sliding surface (as in

1365-781X # 2000 Thomas Telford Ltd


Downloaded by [ NEW YORK UNIVERSITY] on [15/09/16]. Copyright © ICE Publishing, all rights reserved.
S. A. Tan et al.

Fig. 1), soil-nails are normally subjected to tension, shear would not constitute a failure limit space. And the model
forces and bending moments. The soil-nail lateral interaction adopted was less possible. Schlosser (1991a) also expressed
mechanism is a very complex aspect of soil-nail behaviour. that the plastic analysis ignored two soil-nail interaction
There are mainly two theoretical approaches in the literature mechanisms. One was that a soil-nail structure generally
for evaluation of soil-nail lateral resistance. One is the elastic fails due to soil yielding and the other was that even if the
analysis approach proposed by Schlosser (1982) and the maximum bending moment in a reinforcement member
other is the plastic approach developed by Jewell (1990) and reached its plastic capacity, the soil still yields eventually
Jewell and Pedley (1990, 1992). because of the ductile characteristics of the soil-nail steel
Adopting the closed form solution obtained by Hetenyi reinforcement member.
(1946) for an in®nitely long laterally loaded pile or beam in The debate has since cautioned practitioners that both
a medium with a constant modulus of subgrade reaction Ks , approaches appear controversial. Clari®cation is therefore
the elastic approach using Winkler springs was derived. The needed for application of these theoretical approaches in
plastic approach developed by Jewell and Pedley (1990, soil-nailing design, especially as to the appropriate ®eld
1992) was based on the assumption that soil within the shear application conditions for either of these methods to be
band had yielded plastically, where plastic hinges are valid. The authors realized that both the elastic and plastic
formed in the soil-nail reinforcement member at the two approaches represented two different stages of the soil-nail
edges of the shear band. The width of the shear band can be lateral deformation process in a particular progressive fail-
determined by the soil limit bearing pressure, ó9b , the yield ure mechanism. The evolution from one stage to another
strength of the reinforcement member, ó y , and the diameter stage is essentially controlled by the magnitude of the soil-
of the soil-nail, without involving the modulus of subgrade nail relative lateral displacement. This paper describes the
reaction Ks . soil-nail lateral interaction behaviour according to these
There is a series of serious debates in the literature about progressive failure mechanisms. The corresponding possible
the role of soil-nail bending stiffness and the errors in these failure mode for each stage is also presented. Criteria for
two analytical approaches, as highlighted by Jewell and application of these modes are established on the basis of
Pedley (1990). Jewell and Pedley (1991) pointed out that the stiffness, yield strengths of soils and nails and the relative
combination of axial and shear forces represented by the lateral displacements across the shear band. The movement
elliptical strength criterion employed in the elastic approach of soil along a potential sliding surface of a soil-nail
was unlikely to become most critical, and the maximum reinforced structure (as shown in Fig. 1) produces the
shear force in a soil-nail should be determined by the relative lateral displacement of a soil-nail across the shear
moment capacity of the reinforcement member. In addition, band. Results from numerical simulation and published tests
the magnitude of shear force that may be mobilized in a are used to verify the developed soil-nail lateral interaction
soil-nail could be overestimated by a factor 10±20 with the theories.
elastic analysis approach.
However, Bridle and Barr (1990) indicated that the plastic
hinges in a nail assumed in the plastic analysis approach
Structural bending capacities of soil-
nails
A soil-nail is normally subjected to tension, shear forces
Active zone Resistance zone
and bending moments, along a potential soil sliding surface.
The soil-nail will yield structurally once any of these forces
or moments reaches its plastic limit. Jewell and Pedley's
(1990) study shows that the lateral structural failure of a
soil-nail is determined by the failure mode when the
maximum bending moment reaches its plastic moment
capacity. Therefore, it is acceptable in general that its plastic
bending moment capacity limits the structural lateral resis-
tance capacity of a soil-nail. Jewell and Pedley (1990)
describe this as:
Sliding surface ! !
T2 ó 2t
Mlimit ˆ Mp 1 ÿ 2 ˆ Mp 1 ÿ 2 (1)
Tp óy

where Mlimit is the limit bending moment, Mp is the plastic


moment of the nail subjected purely to bending, T and T p
are tension force and tension capacity of the nail subjected
solely to tension; ó t is the tension stress on a cross-section of
the nail due to the tension force T , and ó y is the yield
strength.
Two types of soil-nails are commonly used in practice.
The ®rst is grouted nails, installed by grouting steel bars in
predrilled boreholes in full length. The other is driven nails,
installed by driving steel inclusions into the ground to form
soil-nails. This paper deals only with the latter. The plastic
bending moment of a driven nail is simply determined by
Fig. 1. Soil±nail lateral interaction in a soil-nailed wall (Mitchell and Villet, the yield strength, ó y , cross-section area and shape of the
1987) nail bar. For a soil-nail consisting of a steel bar of circular

142
Downloaded by [ NEW YORK UNIVERSITY] on [15/09/16]. Copyright © ICE Publishing, all rights reserved.
Simpli®ed models for soil±nail lateral interaction

cross-section, the plastic bending moment capacity is given πl0 /4 L ≥ 3l0


by
d3 ó y
Mp ˆ (2) 4
6 l0 = √4EI /KsD

where d is the diameter of the steel bar. Tc

δ Mmax
Mmax x
0
Soil limit bearing pressure
Tc
A soil-nail structure will also fail if soils around a nail
yield. Soil yielding occurs when the pressure exerted on
soils by a nail exceeds the soil limit bearing pressure ó9b . An y
important condition for this is that the relative lateral
displacement of the soil-nail along the sliding surface must
reach a ®nite value, normally taken as 25 mm (Bowles, σb′
Lateral stress
1997). Similar to pile-soil lateral interaction mechanism, soil-
nail lateral interaction mechanism can also be described by σb ′
using the Winkler model (Schlosser 1982; Schlosser et al.
1983). The relationship between a soil reaction pressure ó,
and the relative lateral displacement (ä, see Fig. 2) can be
expressed as Tc

ó ˆ äKs (3) Shear force

where Ks is the soil modulus of subgrade reaction (kN=m3 ).


For a given limit bearing pressure ó9b , and the correspond-
ing relative displacement äe , the modulus of subgrade
reaction can be determined by
Ks ˆ kó9b (4) Mmax

where k is a constant, k ˆ 1=äe , in the range 40±167, for Bending moment


äe ˆ 6±25 mm. Generally, it is assumed that äe ˆ 25 mm, as
suggested by Bowles (1997). The lower safe limit of ó9b ,
corresponding with a punching shear failure around a nail,
was derived by Jewell and Pedley (1992) as
    
1 ‡ Ka ð ö9 ð lb ls lb
ó9b ˆ ó9v tan ‡ exp ‡ ö9 tan ö9 (5)
2 4 2 2
Fig. 2. Soil-nail lateral interaction proposed by Schlosser (1983) and quoted
where ó9v is the vertical effective stress in the soil and Ka is by Jewell and Pedley (1992)
the active earth pressure coef®cient.
The corresponding upper limit of ó9b is bounded by the
relationship ism is determined by the relative stiffness and strengths of
  the soil and nail. If the nail is much stiffer and stronger than
ð ö9
ó9b ˆ ó9v tan2 ‡ exp(ð tan ö9) ˆ ó9v Nq (6) the soil, soil will yield before the nail structural yielding
4 2
takes place, and vice versa. Type C failure mechanism only
where Nq is the bearing capacity factor de®ned by Vesic occurs in the case where the stiffness and strength of a nail
(1973). match those of its surrounding soils in terms of lateral
resistance capacities.
Considering another important factor affecting the soil-
nail lateral interaction behaviour, that is the local relative
Soil±nail lateral interaction displacement across the shear band (ä, see Fig. 2), the
mechanisms following four possible failure modes can be developed from
types A±C failure mechanisms.
According to the foregoing discussion, a soil-nail structure
fails either when the maximum bending moment exceeds its (a) Mode A1: Plastic soil±elastic nail failure mode. Here the soil
bending capacity Mlimit , or when the lateral pressure exerted yields at the point of the maximum reaction pressure
on the soil reaches the soil limit bearing pressure, ó9b . This when the nail has not yet yielded. The relative displace-
implies that three types of failure mechanism are possible ment corresponding to this mode is relatively small.
(b) Mode A2: Plastic soil±plastic nail failure mode. This mode
(a) Type A: soil-nail structure failure triggered by soil develops from mode A1 with a larger relative displace-
yielding ment when the maximum bending moment in the nail
(b) Type B: soil-nail structure failure triggered by nail also reaches its bending capacity.
yielding (c) Mode B: Plastic nail±elastic soil failure mode. Here a nail
(c) Type C: soil-nail structure failure due to soil and nail yields at the point of maximum bending moment when
reaching yield simultaneously. the soil has not yet yielded. This mode corresponds to
For a given soil-nail structure, the lateral failure mechan- failure mechanism type B.

143
Downloaded by [ NEW YORK UNIVERSITY] on [15/09/16]. Copyright © ICE Publishing, all rights reserved.
S. A. Tan et al.

(d) Mode C: Plastic soil±nail failure mode. This corresponds to


failure mechanism type C.
σb′
The form of soil reaction pressure distribution on the nail Tc
differentiates these failure modes. The details of the soil±nail
δe
lateral interaction behaviour in each failure mode are
Mmax
described in the following subsections. 0
Mmax

Plastic soil±elastic nail failure mode: mode A1 σ b′ Tc


lb
In a slope reinforced with soil-nails (Fig. 1), the lateral ls lb
resistance of a soil-nail is mobilized due to relative displace-
ments between the nail and the surrounding soils when the
active zone above the sliding surface settles down. The
sliding surface divides a soil-nail into two segments: one in
Fig. 3. Simpli®ed distribution of soil reaction pressure on a soil-nail for
the active zone and the other in the resistance zone. General mode A1
observations indicate that the relative lateral displacement
across the shear band has its maximum value at the sliding
surface, and decreases sharply with the distance away from
the sliding surface. The soil reaction pressure on the nail is
approximately proportional to the relative lateral displace- It may be noticed from equation (10) and Fig. 3 that the soil
ment before it reaches the limit bearing pressure, ó9b . reaction pressure is ó9 ˆ 0 at x ˆ ls , which is the same as
A soil-nail is analogous to an in®nite beam supported by Hetenyi's (1946) exact solution.
Winkler springs subjected to double shearing forces at the The maximum bending moment Mmax is obtained from
same point but in opposite directions. Hetenyi (1946) the moment equilibrium
obtained the mathematical description about the behaviour 1 2
of this beam; this is also adopted by Schlosser (1982) in his Mmax ˆ l dó9b (11)
12 s
elastic analyses of a soil-nail. The distributions of the soil
(Mmax , Mlimit )
reaction pressure, shear forces and bending moment about
the sliding surface were depicted as in Fig. 2 by Schlosser The soil-nail lateral resistance Tc is given by
(1983), and Jewell and Pedley (1992). The following form 3
according to Hetenyi's (1946) solution expresses the relative Tc ˆ ls dó9b (12)
8
lateral displacement
2Tc ÿx= l0 x Substituting equation (11) into equation (12) yields
äˆ e cos (7)
l0 Ks d l0 4:5Mmax
Tc ˆ (13)
ls
where, as denoted in Fig. 2, Tc is the shear force, x is the
distance from the sliding surface along the nail, and l0 is the which is quite close agreement to
transfer length that is de®ned by 4:9Mmax
s Tc ˆ
ls
4 4EI
l0 ˆ (8) derived from the elastic analysis approach of Schlosser
Ks d
(1982), and Jewell and Pedley (1992).
where E and I are the Young's modulus and moment of Generally, the corresponding relative displacement äe to
inertia of the nail. reach soil limit bearing pressure ó9b is about 25 mm. For a
It is noted from equation (7) that for x < (ð=2)l0 , the values soil-nail structure to reach this failure mode, the sliding
of the relative displacement ä are approximately linearly displacement of the active zone along the sliding surface
related to x. The distribution of the soil reaction pressure on must be equal to or larger than 50 mm (S > 50 mm). It is
the nail can be simpli®ed as that shown in Fig. 3. When the noted that the solutions from this mode closely agree with
maximum soil reaction pressure at the sliding surface reaches those for the elastic analysis approach by Schlosser (1982).
the limit bearing pressure, the soil yields. The state of soil±
nail interaction is then called plastic soil±elastic nail failure
mode as the soil attains plastic yielding, but the nail behaves Plastic soil±plastic nail failure mode: mode A2
elastically. In this mode, the shear band width ls , that is the This failure mode develops from mode A1. If a soil-nail
distance between the two points experiencing maximum structure has not reached an equilibrium state at mode A1,
moment at both sides of the sliding surface is given by soil in the active zone will continue to settle downwards,
s which drags the nail to displace further laterally. The soil
ð ð 4 4EI yielding propagates away from the sliding surface along the
ls ˆ l0 ˆ (9)
2 2 Ks d nail and the bending moment in the nail increases. When
the bending moment reaches its plastic capacity, the plastic
This is to ensure that this simpli®ed model can closely re¯ect soil±plastic nail failure mode (A2) is achieved. With the
the exact solutions obtained by Hetenyi (1946). assumption that the steel of soil±nail bars is perfect elastic±
According to the force equilibrium ls and the length lb plastic material, the soil reaction stress distribution is then in
(see Fig. 2), minimum length beyond the point of maximum the form as shown in Fig. 4. This mode is the same as that
bending moment has the following relationship adopted in the plastic analysis approach developed by
Jewell and Pedley (1990, 1992).
ls ˆ lb (10) The soil-nail lateral resistance is given by

144
Downloaded by [ NEW YORK UNIVERSITY] on [15/09/16]. Copyright © ICE Publishing, all rights reserved.
Simpli®ed models for soil±nail lateral interaction

the partially plastic sections are controlled by the deforma-


Tc
tion of the still-elastic interior ®bres. By ignoring the effect
of axial force on the lateral de¯ection, the de¯ection of the
δp segment of the soil-nail where the bending moment is in the
range My to Mp is described by (Hetenyi, 1946)
σb′ δe Mmax
 
Mmax d2 d2 y
EI x ‡ Ks yd ˆ 0 (18)
dx 2 dx 2
Tc
where Ix is the moment inertia of soil-nail cross-sections and
y is the soil-nail lateral de¯ection which is considered to be
lb ls lb equal to ä in this study. The moment inertia Ix varies with
the extent of the plasti®cation on the sections which in turn
is determined by the magnitude of the bending moment and
Fig. 4. Simpli®ed distribution of soil reaction pressure on a soil-nail for the cross-section shape factor.
mode A2 As structural steels can be normally idealized as perfect
elastic±plastic materials, for this soil-nail analysis equation
4Mlimit (18) can be reduced to
Tc ˆ (14)
ls d4 y
EI ‡ Ks yd ˆ 0 (19)
Tc ˆ 12 ls dó9b (15) dx 4
which is exactly the same as the governing equation for the
where ls is calculated by earlier elastic analysis (Hetenyi, 1946). This equation stands
v
! for all sections except the sections where plastic hinges
u
u4ó y ó 2
locate. This suggests that the plasti®cation only takes place
ls ˆ dt 1 ÿ 2t (16)
3ó9b óy at the plastic hinges; the soil-nail except at the locations of
the plastic hinges still behaves elastically when the maxi-
The minimum required displacement of the active zone mum bending moment reaches the plastic bending capacity.
along the sliding surface to achieve this failure mode is Therefore, the shear band width ls between the two plastic
given by hinges at the two sides of the sliding surface (Fig. 6) can also
S ˆ 4:3äe (17) be determined by equation (9). The pro®le of the soil
reaction pressure distribution is shown in Fig. 6 in dotted
As äe is normally 25 mm, this suggests that the active zone lines which is very similar to that for mode A1.
must slide downwards by at least 108 mm along the sliding The soil-nail lateral resistance here is then determined by
surface to reach this failure mode. the soil-nail bending capacity, instead of soil limit bearing
If the soil-nail structure is still not stable at this stage, the pressure
further nail lateral movement will result in the distribution 4:5Mlimit
of soil reaction pressure on the nail in a form as shown in Tc ˆ (20)
ls
Fig. 5. This was used for the plastic analysis of soil-pile
lateral interaction by Hansen and Lundgren (1960), and was where ls is calculated according to equation (9).
quoted by Jewell and Pedley (1992). As an excessive lateral The maximum soil reaction pressure ó9max at the sliding
displacement is required to achieve this failure mode, it will surface is given by
not have much practical value for soil-nailing design. Thus 12Mlimit
no further details of this failure mode are discussed here. ó9max ˆ (21)
l2s d
ó9max , ó9b
Plastic nail±elastic soil failure mode: mode B
In this case, the soil is relatively stiffer and stronger than
the nail in terms of lateral resistance capacity. The soil-nail
yielding starts from the section where the bending moment
exceeds its elastic proportional limit (My ). The curvatures of
σ b′
Tc
δe

0 Mmax
Mmax
Tc
σm
σb ′ Mmax
Mmax σ
Tc σb′ Tc
σb′

L = ls /2 + lb

lb ls lb lb ls lb

Fig. 5. Soil reaction pressure distribution on pile postulated by Hansen and Fig. 6. Simpli®ed soil reaction pressure distribution on a soil-nail for mode
Lundgren (1960) quoted by Jewell and Pedley (1992) B

145
Downloaded by [ NEW YORK UNIVERSITY] on [15/09/16]. Copyright © ICE Publishing, all rights reserved.
S. A. Tan et al.

and the corresponding displacement of the active zone along The critical steel bar diameter dcr for a soil-nail structure
the sliding surface is to fail in this mode can be obtained by substituting equation
24Mlimit (9) into equation (23)
Sˆ (22)
Eó 9b
2
Ks l2s d 1
dcr ˆ 0:3 !2 (24)
Ks ó 2y ó2
The theoretical value of S is smaller than 50 mm as 1 ÿ 2t
óy
ó9max , ó9b .
If the soil-nail structure has not achieved its equilibrium which sets out the soil-nail compatible condition for this
at the state of mode B, soil in the active zone will continue failure mode.
to move downwards which will cause further soil-nail
lateral displacement. Because of the ductile characteristics of
the nail, the plasti®cation will gradually develop from the
hinge towards the sliding surface until the maximum soil Criteria for determination of failure
reaction pressure at the sliding surface also attains the soil modes
limit bearing pressure. In this case, the soil-nail behaves like
a chain, largely in¯uenced by the characteristics of the soil- Equation (24) can now serve as the criterion for determin-
nail strength hardening. This will not be investigated further ing the possible failure mode for a given soil-nail structure.
here as normally a soil-nail structure is considered as failed This is elaborated in the following paragraphs.
if the bending moment in the nail has reached the plastic If d . dcr , the nail is relatively stronger than the soil, then
bending capacity. soil-nail lateral interaction behaves according to type A
mechanism. The soil-nail structure attains a failure state in
mode A1 after the active zone moves down about 50 mm
Plastic soil±nail failure mode: mode C along the sliding surface, then gradually develops into
The condition for a soil-nail to fail in this mode is that the failure mode A2 after a displacement of about 108 mm if the
soil stiffness and strength can match those of the nail for soil-nails are not suf®cient to resist the driving forces at the
resistance to lateral loads. The maximum pressure on soil state of mode A1. This may continue until the slope
reaches the soil limit bearing pressure while the maximum completely fails, if the soil-nail resistance at mode A2 is not
bending moment in the nail attains its plastic bending adequate.
capacity simultaneously. The distribution of the soil reaction If d , dcr , the soil is relatively stiffer and stronger than the
pressure is the same as that in mode A1 (Fig. 3) but in this nail, and the nail will laterally interact with soils according
case Mmax ˆ Mlimit . The soil-nail lateral resistance can be to mechanism type B. The complete collapse will occur if
calculated according to equation (12) or (20) with substitu- soil-nails cannot prevent the active zone from moving
tion of the following ls beyond the state of mode B.
v
! If d ˆ dcr , the soil strength and stiffness are comparable to
u
u2ó y ó 2
those of the nail, and the soil and nail will fail simulta-
ls ˆ dt 1 ÿ 2t (23)
ó9b óy neously. The required minimum lateral displacement of the
soil-nail to achieve this failure mode is about 50 mm.
The maximum bending moment can be obtained from The typical values of dcr are plotted in Fig. 7 against soil
equation (11). The required soil displacement in the active limit bearing pressure ó9b for a steel nail of E ˆ
zone along the sliding surface is S ˆ 2äe ˆ 50 mm. 205 kN=mm2 , ó y ˆ 460 N=mm2 , and soil modulus of sub-

1000 T/Tp
0·8
σy = 460 N/mm2
E = 205 kN/mm2
Ks = 40 σb′ 0·7

0·6
0·5
Critical soil-nail diameter: mm

100
0·4

0·3

0·2

10
0·1

1
100 1000 10 000
Soil limit bearing pressure: kPa

Fig. 7. Critical diameter of soil-nail plotted against soil limiting bearing pressure

146
Downloaded by [ NEW YORK UNIVERSITY] on [15/09/16]. Copyright © ICE Publishing, all rights reserved.
Simpli®ed models for soil±nail lateral interaction

grade reaction of Ks ˆ 40 ó9b . This chart can be very useful


as a design aid in choosing the appropriate soil-nail Example 1
diameter for a given design situation, when tension force In order to check the soil-nail lateral interaction beha-
ratio and soil limit bearing pressure are determined prior to viour, a large shear box test is simulated in this example.
a soil-nail reinforced structure. The shear box is 3 3 3 3 6 m long, split into two halves as
shown in Fig. 9. The half at the left-hand side is ®xed and
the half at the right-hand side can move along the y
direction. The box is ®lled with dense sand of density
Veri®cation with numerical 㠈 16 kN=m3 , friction angle ö9 ˆ 408, Young's modulus
simulation results E ˆ 60 MPa, and Poisson's ratio ì ˆ 0:4. A 3 m long steel
bar of diameter d ˆ 25 mm, Young's modulus E ˆ
Numerical approaches 205 kN=mm2 and yield strength ó y ˆ 460 N=mm2 is hori-
A ®nite difference programme, FLAC (Fast Lagrangian zontally placed in the middle of the box. A con®ned
Analysis of Continua, version 3.3), is used to simulate the compression pressure of 80 kN=m2 , equivalent to a 5 m high
soil±nail lateral interaction behaviour. earth pressure, is applied in the x, y and z (out-of-plane)
A soil-nail is represented by two-dimensional elements directions. The limit bearing pressure according to equation
with three degrees of freedom (x-translation, y-translation (5) is 700 kPa. The right-half box moves along the y direction
and one rotation) at each end node. These elements have the at a very slow speed of 2:5 3 10ÿ3 mm=s.
combined features of beam and cable elements. Beam Based on these data, the critical diameter dcr calculated
elements are used to represent a structural member that can according to equation (24) is equal to 5 mm (dcr ˆ 5 mm)
sustain bending and yield when the bending moment which is smaller than the actual steel bar size of 25 mm. This
reaches its plastic limit. Cable elements are one-dimensional means that the soil±nail lateral interaction will behave
axial elements that can yield in tension or compression. according to type A mechanism; that is, the soil-nail will fail
The nail interacts with the soil grid via shear and normal in mode A1 ®rst then in mode A2. The numerical simulation
coupling springs. The coupling springs are non-linear does show that the soil-nail behaves in this manner. The
connectors that transfer forces and motion between the nail details of the numerical simulation are further discussed.
elements and the grid at the nail element nodes. The shear With the downward movement of the half-box at the
behaviour of the soil±nail interface is represented as a right-hand side, the soil reaction pressure on the nail
spring-slider system at the nail node points shown in Fig. 8. increases. When the maximum pressure reaches 700 kPa
It is described numerically by the soil shear stiffness. The (soil limit bearing pressure), the soil-nail behaviour is as
normal behaviour of the nail±soil interface is represented by plotted in Fig. 10. This state corresponds to that as
a linear spring with a limit normal force that is calculated theoretically described in failure mode A1. The theoretically
from the soil limit bearing pressure. calculated soil reaction pressure, according to mode A1, is
To reduce the three-dimensional problem with soil-nails plotted in Fig. 10(a) in dotted lines. The other theoretical
to the equivalent two-dimensional problem, the properties predictions are compared with the results of the numerical
of soil-nails are scaled by the out-of-plane thickness of the simulation in Table 1. These demonstrate that the theoretical
slice taken from a long soil-nail structure (plane strain predictions agree well with the numerical simulation results.
conditions) for this study. This is according to the recom- With the further downward movement of the half-box at
mendation made in the FLAC manual, which is based on the right-hand side, the soil yielding zone spreads away
the suggestion by Donovan et al. (1984) that linear scaling of from the shear plane, and the bending moment in the nail
material properties is a simple and convenient way of further increases. When the maximum bending moment
distributing the discrete effect of elements over the distance attains 1´2 kNm (nail plastic bending capacity), the soil-nail
between elements in a regularly spaced pattern. The out-of- behaviour is as shown in Fig. 11. This is the state corre-
plane thickness of the slice taken for this study is equal to sponding to mode A2. The distribution of the soil reaction
one unit (1 m), therefore the relevant soil-nail properties are pressure calculated theoretically according to mode A2 is
scaled by 1 m when the three-dimensional problem is
simulated with the equivalent two-dimensional model.
1·5 m 3·0 m 1·5 m

Reinforcing
element (steel)
1·5 m

Excavation x
1·5 m

Axial stiffness
of steel
Vy
Slider 3·0 m 3·0 m
Reinforcement (cohesive strength)
nodal point

Shear stiffness y

Fig. 8. Conceptual mechanical representation of soil±nail shear interaction Fig. 9. Shear box for numerical simulation of shear tests (all dimensions in
for numerical analysis (Itasca, 1996) metres)

147
Downloaded by [ NEW YORK UNIVERSITY] on [15/09/16]. Copyright © ICE Publishing, all rights reserved.
S. A. Tan et al.

Shear plane
700 kPa
700 kPa Theoretical
profile

700 kPa

Theoretical
profile
700 kPa (a)

(a)

Shear plane
6·27 kN

(b)

3·45 kN 1·2 kNm

(b)

–1·2 kNm

0·5 kNm 750 mm


(c)

–0·5 kNm

174·5 mm
600 mm

(c)
(d)

Fig. 11. Numerical simulation of soil-nail performance when nail yields,


showing: (a) soil reaction pressure distribution; (b) shear force distribution;
65 mm

(c) bending moment distribution; and (d) soil-nail lateral displacement

(d)
Table 2. Comparison of theoretical predictions and numerical simulation
Fig. 10. Numerical simulation of soil-nail performance when soil starts to resultsÐmode A2
yield, showing: (a) soil reaction pressure distribution; (b) shear force
distribution; (c) bending moment distribution; and (d) soil-nail lateral Tc : Mmax : ls : ämax : S:
displacement Description kN kNm mm mm mm
Analytical prediction 6´48 1´2 740´0 54´0 108´0
based on mode A2
Numerical analysis 6´27 1´2 750´0 71´8 174´5
Table 1. Comparison of analytical prediction and results of numerical analysisÐ results
model A1
Tc : Mmax : ls : ämax : S:
Description kN kNm mm mm mm Shear plane
Analytical prediction 4´0 0´53 607´5 25´0 50´0
based on mode A1 700 kPa
Numerical simulation 3´45 0´5 600´0 25´0 65´0
results
700 kPa

plotted in Fig. 11(a). The other theoretical predictions are Fig. 12. Soil reaction pressure distribution when nail laterally deformed
compared with the results of the numerical simulation in 448 mm
Table 2, which shows good agreement.
After the soil-nail has laterally deformed 448 mm, the
distribution of the soil reaction pressure reaches a pro®le as It may be noted that the numerical simulation shows that
shown in Fig. 12, which is very similar to that postulated by the lateral de¯ection of the soil-nail is larger than the
Hansen and Lundgren (1960) for the analysis of soil±pile predicted value. This is apparently because the moving soil
interaction as shown in Fig. 5. in the half-box at the right-hand side has in¯uence on the

148
Downloaded by [ NEW YORK UNIVERSITY] on [15/09/16]. Copyright © ICE Publishing, all rights reserved.
Simpli®ed models for soil±nail lateral interaction

soil adjacent to the sliding surface in the half-box at the left- results are compared in Table 3. These comparisons again
hand side. A thin layer of soil in the left side of the sliding give good agreement.
surface is also dragged down by the moving soil in the
active zone due to the friction. This means that the soils in
the resistance zone near the sliding surface do not stay static
when the soils in the active zone are moving down. The Veri®cation with experimental data
relative displacement between the nail and soil in the
resistance zone is then less than the absolute nail lateral Large shear box test by Bridle (1990)
displacement. Since the Winkler springs are independent Bridle (1990) reported a large-scale shear box test of an
and are not able to represent this shear interaction between instrumented soil-nail. A steel bar 3 m long and 20 mm in
soil particles, the theoretical modes cannot re¯ect the in¯u- diameter with yield strength of 570 N=mm2 was placed in
ence of moving soils in the active zone on soil mass in the the middle of the shear box ®lled with sand as shown in Fig.
resistance zone. From this point of view, the theoretically 14. When one half of the shear box moves along the shearing
calculated soil-nail lateral displacement can only serve as the plane passing through the middle point of the soil-nail, the
lower limit, which is normally smaller than the actual nail nail is subjected to bending moment and shear forces only.
lateral resistance. The forces required to displace half of the box against nail
de¯ection and bending moments at different stages are
measured and shown in Fig. 15. The measured shear band
Example 2 width is ls ˆ 600 mm. The soil-nail lateral resistance in-
This example is set out to numerically check the nail-
failure-®rst mechanism (type B). Therefore, the soil used is
relatively stronger and the nail is relatively weaker in this Table 3. Comparison of theoretical predictions and numerical simulation
example. The set-up for the shear test is the same as for resultsÐmode B
example 1. The soil-nail is a 10 mm diameter steel bar with
Tc : ls : Mmax : ämax : S:
the other parameters being the same as used in example 1. Description kN mm kNm mm mm
Its full plastic bending capacity is 77 Nm. The soil is very
Analytical prediction 1´75 198´0 0´077 14´7 29´0
dense sand with bulk density 㠈 18 kN=m3 , friction angle based on mode B1
ö9 ˆ 458, shear modulus G ˆ 29 MPa and bulk modulus Numerical simulation 1´30 198´0 0´077 12´6 54´0
K ˆ 133 MPa. The soil limit bearing pressure calculated from results
equation (5) is ó9b ˆ 4000 kN=m2 and the modulus of sub-
grade reaction is obtained to be Ks ˆ 160 kN=m3 with
k ˆ 40. The calculated critical diameter dcr is equal to 29 mm
which is larger than the steel bar size of 10 mm. This
suggests that the soil-nail structure will fail in mode B.
The numerically simulated behaviour of soil±nail lateral
interaction at the state that the maximum bending moment
achieves the plastic bending capacity is graphically illus-
trated in Fig. 13, which corresponds to that described in Shear Sand

1·5 m
mode B. only
The theoretical predictions and the FLAC calculation

Shear plane

2008 kPa Jack

Theoretical
0·4 m 3m 0·4 m
profile

2008 kPa

(a)

Shear + Nail
1·5 m

77 Nm

–77 Nm Jack
198 mm
3m

(b)
Deflection
Fig. 13. Numerical simulation of soil-nail behaviour when nail yields,
showing: (a) soil reaction pressure distributions; and (b) bending moment
distribution Fig. 14. Shear box (Bridle, 1990)

149
Downloaded by [ NEW YORK UNIVERSITY] on [15/09/16]. Copyright © ICE Publishing, all rights reserved.
S. A. Tan et al.

35 Mean Table 4. Comparison of analytical results with the experimental test data by
30 Instrumented Bridle (1990)
25 Tc : Mmax : ämax : ls :
10 kN Uninstrumented
Description kN kNm mm mm
Load: kN

20
Analytical prediction 6´47 0´53 22´0 370´0
15 Sand only based on mode A1
φ peak
10 Experimental measurement 5´0 0´5 25´0 600´0
Mean of instrumented
5 and uninstrumented nails
0
0 20 40 60 80 100 120 140
Deflection: mm a rigid steel cube of internal size 1 3 1 3 1 m, split at mid-
(a) height. In the test, shearing took place along the horizontal
central plane.
1·2
Approximately 1 kNm The soil used was quartz Leighton Buzzard sands, with
0·9
particle size in the range 0:60±1:18 mm. The unit weight was
0·6 300 m 29·74 kN
17:7 kN=m3 and relative density ID ˆ 90%. The friction angle
Moment: kNm

0·3 22·79 kN
27·58 kN measured at the applied vertical stress ó9v ˆ 100 kN=m3 was
0 ö9ps ˆ 468. Steel bars used were 920 mm long with yield
–0·3
stress ó y ˆ 393 3 103 kN=m2 . Two types of bar diameters
(d ˆ 15:9 mm and d ˆ 25:4 mm) were used. The steel bars
–0·6 300 m
were placed in the soil in the shear box at inclination angles
–0·9
of 0, 15 and 258 with respect to the vertical central axis of the
–1·2
–1500 –1000 –500 0 500 1000 1500 shear plane respectively.
Length along nail: mm The shear resistance due to the steel bars reached the
(b) maximum values at shear displacements about 35 mm
(äe ˆ 17:5 mm). The suggested value of the soil limit bearing
Fig. 15. Graphs showing: (a) load±displacement curve; and (b) bending pressure was ó9b ˆ 2200 kN=m2 (Pedley et al., 1990). The
moment distribution (after Bridle, 1990)
modulus of subgrade reaction is then obtained as Ks ˆ
125 400 kN=m3 by substituting the aforementioned data into
equation (4). The calculated critical diameters from equation
creases with the increase of the lateral movement, and (24) for 15:9 and 25´4 mm diameter steel bars are 15´4 mm,
reaches its ultimate value of 14 kN at a lateral displacement with the assumption of Young's modulus of steel bars
of about 95 mm, while the maximum bending moment E ˆ 205 kN=mm2 . These suggest that the 15´9 mm steel bar
attains its ultimate bending capacity of 1 kNm. This is a would interact with the soil according to type C mechanism
typical case where the soil-nail behaved according to the where the steel bar and soil yield simultaneously, and the
interaction mechanism of type A. These data describe the 25´4 mm soil-nail would behave in mode A1. These predic-
behaviour of the soil-nail lateral interaction corresponding to tions agree well with observations in the test. The maximum
that in mode A1 and mode A2. To verify the foregoing bending moments measured in 15´9 mm steel bars are very
developed theories, these data at the state of mode A2 can close to their plastic bending capacity (Mp ) as shown in Fig.
be used as the input data to back-analyse the soil-nail 16, while the maximum bending moments in 25´4 mm bars
interaction behaviour at the state corresponding to mode A1. were still smaller than their plastic bending capacities, even
By comparing the analytical and the experimental results, at a shear displacement of 60 mm.
the validity of the developed theories can be proved again The theoretical predictions for the behaviour of 15´9 and
as follows. 25´4 mm steel bars, according to corresponding modes, are
Using the data for the mode A2 state, substituting compared with the measured values in Table 5. The agree-
Tc ˆ 14 kN and ls ˆ 0:6 m into equation (15) gives ó9b ˆ ment between the predicted values and measured data is
2333 kN=m2 . From equation (17), the relative lateral dis- quite good.
placement is calculated to be äe ˆ 22 mm. Then the modulus In the publication by Pedley et al. (1990), Jewell's mode
of subgrade reaction is estimated to be Ks ˆ 45ó9b ˆ (mode A2) was adopted to interpret these test results.
104 985 kN=m3 according to equation (4). By substituting However, the authors believe it would be more accurate if
these data into equation (24) the critical diameter is obtained mode C had been used for the interpretation of the results of
as dcr ˆ 10 mm with the assumption of Young's modulus of 15´9 mm bars and mode A1 for the test results of 25´4 mm
steel E ˆ 205 kN=mm2 . The critical diameter is smaller than bars. This is because the soil reaction pressure distribution
the actual steel bar diameter of 20 mm. This suggests that on 15´9 mm bars at the yielding is closer to the form
the soil-nail structure will fail progressively from mode A1 described by mode C than that by Jewell's mode. One of the
to A2, which is quite consistent with observations from the important conditions for a soil-nail structure to fail accord-
test (Bridle, 1990). The back-calculated results describe the ing to Jewell's mode is that the maximum bending moment
state of mode A1, and the measured data (scaled from Fig. in the nail must reach its plastic bending capacity, otherwise
15) are listed in Table 4 for comparison, which show equation (16) cannot be applied. The test results showed that
reasonably good agreement. the maximum bending moments in 25´4 mm bars were
smaller than the plastic bending capacity. In addition, the
lateral relative displacement measured is about 17´5±30 mm,
Large-scale shear box test by Pedley et al. which is less than the required 54 mm for Jewell's mode.
(1990) Therefore, the distribution of the soil reaction pressure on
Pedley et al. (1990) presented a series of direct shear box 25´4 mm bars is unlikely to be of the form de®ned by
tests for the investigation of the effect of reinforcement Jewell's mode. If Jewell's mode is used to calculate the soil-
bending stiffness. The direct shear test apparatus consists of nail lateral resistance for these cases, the values will be

150
Downloaded by [ NEW YORK UNIVERSITY] on [15/09/16]. Copyright © ICE Publishing, all rights reserved.
Simpli®ed models for soil±nail lateral interaction

shear band. The failure modes of a particular failure


mechanism represent the different stages of the whole
failure process. For a given soil-nail structure, its failure
mechanism can be determined by comparing the soil-nail
diameter with the calculated critical soil-nail diameter from
equation (24). Then the corresponding failure mode can be
determined according to the magnitude of the relative lateral
displacement of the soil-nail.
One important point that can be inferred from the study
of these theoretical modes is that the soil-nail lateral
resistance depends not only on the stiffness and yield
strengths of soil and nails, but also on the relative lateral
displacement. Soil-nail lateral resistance generally increases
with the relative lateral displacement before the soil yields.
In addition, the relative displacement between a nail and
soil is the most important factor that controls the soil±nail
lateral interaction behaviour. The correct failure modes can
be established based on the magnitude of the relative
displacement in the shear zone. The required minimum
displacement of the soil in the active zone along the sliding
surface for failure mode A1 and C is about 50 mm and that
for mode B is less than 50 mm. However, for mode A2
( Jewell's mode) it is about 108 mm, which is unlikely to be
acceptable for a soil-nailing design in practice. Therefore it
may not be prudent to use this mode to evaluate the soil-
nail lateral resistance for soil-nailing design.

Conclusions
The conclusions of this study can be summarized as
follows.

Fig. 16. Distribution of: (a) bending moment; and (b) shear force at a shear (a) A series of simpli®ed soil-nail lateral failure modes have
displacement of 60 mm (after Pedley et al., 1990) been adopted to describe the soil-nail lateral interaction.
The criteria for application of these modes have been
established.
overestimated by 8´5±21%. In addition, the measured ls =d (b) Expressions for the evaluation of soil-nail lateral resis-
for 15´9 and 25´4 mm bars would be quite different from the tance have been derived. The soil-nail lateral resistance
observed results. The calculated ls =d according to Jewell's is dependent on the relative stiffness and strengths of
mode is 15´4 regardless of the steel bar sizes, whereas both the soil and nail and the relative lateral displacement.
the experimental results and the proposed analytical calcula- (c) Numerical simulation results and published experimen-
tions clearly showed the in¯uence of bar diameter on this tal test data are used to verify these developed theories.
ratio. This illustrates that it may be inappropriate to use one Good agreements among the numerical simulation
mode to describe soil-nail interaction behaviour in different results, theoretical predictions and the published experi-
stages of a failure mechanism. mental data have been demonstrated.
(d) This study indicates that the states of soil-nail lateral
interaction behaviour described by the elastic (Schlosser,
Discussion 1982) and plastic ( Jewell and Pedley, 1990, 1992) ap-
proaches represent the two different stages of a particu-
Soil-nail lateral failure takes place in a progressive lar failure mechanism at different lateral displacements.
manner. There are three possible failure mechanisms, based It also illustrates that the required minimum settlement
on soil-nail relative stiffness and yield strengths. Each failure of the soil, along the sliding surface, in active zone
mechanism consists of one or two failure modes according relative to the stable soil is 108 mm for the mode
to the relative lateral displacement of the soil-nail across a proposed by Jewell and Pedley (1990, 1992). This

Table 5. Comparison of analytical predictions with measured data from tests by Pedley et al. (1990)
Tc : kN ls =d
Inclination angle è:
Steel bar size d: mm degrees Analytical prediction Measured Analytical prediction Measured
15´9 0 4´00 4´0 18´80 17´4
15 4´00 4´13 18´80 16´9
25 4´00 4´0 18´80 17´2
25´4 0 8´86 9´47 16´65 13´7
15 8´86 8´5 16´65 13´4
25 8´86 9´0 16´65 13´2

151
Downloaded by [ NEW YORK UNIVERSITY] on [15/09/16]. Copyright © ICE Publishing, all rights reserved.
S. A. Tan et al.

amount of settlement may not be practically allowed for Jewell R. A. and Pedley M. J. (1991) DiscussionÐclosure. Ground
in a soil-nail structure. Engineering, Nov., 34±39.
Jewell R. A. and Pedley M. J. (1992) Analysis for soil reinforcement
with bending stiffness. Journal of Geotechnical Engineering, 118,
No. 10, 1505±1529.
Acknowledgement Mitchell J. K. and Villet W. C. B. (1987) Reinforcement of Earth
Slopes and Embankments. Transportation Research Board, Na-
The authors wish to express their thanks to Mr B. T. Khoo tional Research Council, Washington DC, National Co-operative
of Presscrete Engineering Pte Ltd for his support for this Highway Research Program Report 290.
publication. Pedley M. J., Jewell R. A. and Milligan G. E. W. (1990) A large
scale experimental study of soil-reinforced interaction. Ground
Engineering, July/Aug., 44±50 and Ground Engineering, Sept.,
46±49.
References Schlosser F. (1982) Behaviour and design of soil nailing. Proceedings
of an International Symposium, Asian Institute of Technology,
Bowles J. E. (1997) Foundation Analysis and Design. McGraw-Hill, ?, Bangkok, 29 Nov.±3 Dec., pp. 399±419.
5th edn. Schlosser F. (1983) Analogies et diffeÂrences dan le component et le
Bridle R. J. and Barr B. I. (1990) DiscussionÐsoil nailing. Ground calcul des ouvrages de souteÁnement en terre armeÂe et par
Engineering, July/Aug., 30±31. clouage du sol. Sols et Fondations, 184, Annales ITBTP, No. 418,
Bridle R. J. (1990) Discussion of papers in Session 3. Performance of Oct., 8±23 (in French).
reinforced soil structures. Proceedings of an International Rein- Schlosser F. (1991a) Behaviour and design of soil-nailing. Sympo-
forced Soil Conference, BGS, Glasgow, pp. 284±286. sium on Recent Developments in Ground Improvement Techniques,
Donovan K., Pariseau W. G. and Cepak, M. (1984) Finite element Bangkok, pp. 399±413.
approach to cable bolting in steeply dipping VCR stopes. In Schlosser F. (1991b) Discussion: The multicriteria theory in soil
Geomechanics Application in Underground Hardrock Mining. Soc- nailing. Ground Engineering, 24, Nov., 30±33.
iety of Mining Engineers, New York, pp. 65±90. Schlosser F., Jacobsen H. M. and Juran I. (1983) Soil reinforce-
Hansen J. B. and Lundgren H. (1960) Hautprobleme der bodenmecha- ment. Proceedings of the 8th European Conference on Soil Mechanics
nick. Springer-Verlag, Berlin. and Foundation Engineering, Helsinki, 3, 1158±1180.
Hetenyi M. (1946) Beams on Elastic Foundations. University of Vesic A. S. (1973) Analysis of ultimate loads of shallow foundations.
Michigan Press. Journal of the Soil Mechanics and Foundations Division, ASCE, 99,
Itasca Consulting Group, Inc (1996) FLAC Manual, Version 3.3. SM2, 45±73.
Itasca Consulting Group, Inc, USA.
Jewell R. A. (1990) Review of theoretical models for soil-nailing.
Proceedings of an International Reinforced Soil Conference, 10±12
Sept., Glasgow.
Jewell R. A. and Pedley M. J. (1990) Soil-nailing design: the role of Discussion contributions on this paper should reach the
bending stiffness. Ground Engineering, Mar., 30±34. secretary by 10 January 2001

152
Downloaded by [ NEW YORK UNIVERSITY] on [15/09/16]. Copyright © ICE Publishing, all rights reserved.

También podría gustarte