Professional Documents
Culture Documents
Merifield 2010 Ultimate Uplift Capacity of Multiplate Helical Type Anchors in Clay
Merifield 2010 Ultimate Uplift Capacity of Multiplate Helical Type Anchors in Clay
Abstract: In recent years, the use of helical anchors has expanded beyond their traditional use in the electrical power industry. The
advantages of rapid installation and immediate loading capability have resulted in their being used in more traditional civil engineering
Downloaded from ascelibrary.org by LEHIGH UNIVERSITY on 10/26/23. Copyright ASCE. For personal use only; all rights reserved.
infrastructure applications. Unfortunately, our current understanding of these anchors is unsatisfactory, and the underlying theoretical frame-
work adopted by engineers has proven to be largely inappropriate and inadequate. A better understanding of helical anchor behavior will lead
to increased confidence in design, a wider acceptance as a foundation alternative, and more economic and safer designs. The primary aim of
this research is to use numerical modeling techniques to better understand multiplate circular anchor foundation behavior in clay soils. A
practical design framework for multiplate anchor foundations will be established to replace existing semiempirical design methods that are
inadequate and have been found to be excessively under- or overconservative. This framework can then be used by design engineers to
confidently estimate the pullout capacity of multiplate anchors under tension loading. DOI: 10.1061/(ASCE)GT.1943-5606.0000478.
© 2011 American Society of Civil Engineers.
CE Database subject headings: Anchors; Numerical models; Numerical analysis; Clays; Plasticity; Uplifting.
Author keywords: Anchors; Numerical modeling and analysis; Theoretical analysis; Clays; Plasticity.
1
Numerous investigators have performed model tests on single helix
Senior Lecturer, Centre for Geotechnical and Materials Modeling, anchors in an attempt to develop semiempirical theories that can be
School of Engineering, Univ. of Newcastle, Newcastle, NSW 2308, used to estimate the capacity of anchors in soil. For anchors in clay,
Australia. E-mail: richard.merifield@newcastle.edu.au
results can be found in the works of Vesic (1971), Meyerhof and
Note. This manuscript was submitted on November 4, 2009; approved
on October 27, 2010; published online on November 4, 2010. Discussion Adams (1968), Meyerhof (1973), and Das (1978; 1980).
period open until December 1, 2011; separate discussions must be sub- One thing in common in all the aforementioned studies is that
mitted for individual papers. This paper is part of the Journal of Geotech- they are limited to anchors with a single helix. The proposed semi-
nical and Geoenvironmental Engineering, Vol. 137, No. 7, July 1, 2011. empirical theories cannot be easily applied to the problem of
©ASCE, ISSN 1090-0241/2011/7-704–716/$25.00. multihelix anchors in which a complication arises owing to the
Helix
Diameter
Transmission tower Submerged Pipeline
Pitch
Top Helix (b) Uplift
Downloaded from ascelibrary.org by LEHIGH UNIVERSITY on 10/26/23. Copyright ASCE. For personal use only; all rights reserved.
Helix
Diameter
Lower Pitch
Helix
Helix
Thickness
Fig. 1. Use of helical anchors—Some practical examples: (a) helical anchor definitions; (b) uplift; (c) new construction; (d) tiebacks
interaction between adjacent helices. This interaction can produce with a scope that is sufficient to overcome the problems of existing
overlapping stress zones that affect the failure mode and ultimate semiempirical theories discussed in this section.
capacity, as highlighted by Merifield and Smith (2010).
Experimental and field studies on multihelix anchors are Previous Theoretical Analyses
much more scarce and are limited to the works of Weikart and
There have been numerous theoretical studies that address the uplift
Clemence (1987), Mitsch and Clemence (1985), Mooney et al.
of single horizontal anchors as summarized by Merifield et al.
(1985), Lutenegger et al. (1988), Hoyt and Clemence (1991),
(2006, 2003). The majority of these studies, however, assume a
and Narasimha Rao et al. (1991, 1993).
Based on field and laboratory observations, Mitsch and condition of plane strain for the case of a continuous strip anchor
Clemence (1985), and Mooney et al. (1985) proposed semianalyt- or axi-symmetry for the case of circular anchors. The works of
ical expressions to estimate the uplift capacity of horizontal [see Merifield et al. (2006, 2003), and more recently, Wang et al.
Fig. 1(a)] multihelix anchors in sand and clay, respectively. These (2009), appear to be the only three-dimensional numerical analyses
two studies represent the most notable attempts at presenting a to ascertain the effect of anchor shape on the uplift capacity. How-
generic set of design guidelines for horizontal multihelix anchors. ever, these studies do not consider multihelix anchors.
Important to note is that these theories are simple limit equilibrium The common approach to modeling anchors is by using the limit
approaches based on a number of significant assumptions. This in- equilibrium method (LEM), even though most failures involve
cludes assuming a failure mechanism consisting of general bearing- complex deformations that bear little resemblance to the rigid block
capacity failure above the top helix and cylindrical failure between assumptions required by most limit equilibrium methods.
the lower helices. An approximate semiempirical theory for the uplift capacity of
A measure of the accuracy of the methods proposed by Mitsch single circular anchors has been proposed by Meyerhof and Adams
and Clemence (1985), and Mooney et al. (1985) can be found in the (1968). The analysis of strip anchors was extended by Meyerhof
paper by Lutenegger et al. (1988). Lutenegger et al. (1988) found and Adams (1968) to include circular anchors by using a semiem-
that when estimating the capacity of multihelix anchors in sand pirical shape factor to modify the passive earth pressure obtained
based on cone penetration test (CPT) data, the proposed procedures for the plane strain case. The paper by Meyerhof and Adams (1968)
of Mitsch and Clemence (1985) are overly conservative, particu- is widely referenced when considering the capacity of anchors.
larly for deep anchors. The theory proposed by Mooney et al. Kulhawy (1985) presents what appears to be one of the only
(1985) for anchors in clay was found to overpredict the ulti- methods of analysis for the uplift capacity of multihelix helical an-
mate capacity significantly; the calculated load was up to four chors. In his model, the upper helix develops a cylindrical shear
times the measured field load. Given the significant level of site surface that controls its behavior. The soil between the helices
investigation, Lutenegger and colleagues concluded that the dis- becomes an effective cylinder if the helices are sufficiently close
crepancies could not be explained through errors in strength together. The shearing resistance along the interface is said to
determination, and existing semiempirical methods need to be be controlled by the friction angle and state of stress in the dis-
modified. turbed cylinder of soil above the anchor. Key terms in the general
In summary, it is apparent that existing theories are unable to model are the modified bearing-capacity factors, the in situ stresses
produce estimates of the collapse loads in either clay or sand that and how they are modified by installation effects, and the soil
are consistent with measured field loads. Ultimately, what is needed strength along the shear surfaces. The overall design procedure
is a much more thorough experimental study of multihelix anchors is, again, semiempirical in nature.
stead have chosen to use one of three methods for predicting helical blade displacing the overburden soil in a logarithmic spiral
pullout-capacity, namely, “cylindrical shear,” “individual bearing,” mode. Thus, the capacity of each blade can be estimated using the
and an empirical method based on “installation torque.” In general, well-known Terzaghi bearing-capacity equation. The total ultimate
manufacturers suggest that the cylindrical shear and individual pullout-capacity is the sum over n blades, as given by
bearing methods be used to determine capacity and the installation
torque method be used as a field verification tool only. X
1
1
In the “cylindrical shear” method, the entire volume of soil be- Qu ¼ An 1:3cN c þ σvn N q þ γBN γ ð4Þ
tween the helical blades is assumed to be mobilized. This type of n
2
failure mechanism is shown in Fig. 2(a). Due to the rigid nature of
this mechanism, the anchor shaft is unlikely to contribute to the where An = blade area. All other parameters have been defined pre-
overall capacity. Ultimate pullout-capacity of a multiblade anchor viously. Contributions of soil unit weight in the third term in Eq. (4)
is a combination of shear along the cylinder of soil between the are generally ignored by manufacturers. Therefore, for anchors in
blades and bearing capacity of the top blade, given by purely cohesive clay, the total ultimate pullout-capacity is the sum
over n blades in clay, as given by
Qu ¼ πDðn 1ÞSðc þ σv tan ϕÞ þ 0:25πD2 ð1:3cN c þ σv N q Þ
X
1
ð1Þ Qu ¼ An ð1:3cu N c þ σvn Þ ð5Þ
n
where D = average blade diameter; S = spacing between each
blade; n = number of helical blades; σv = vertical soil pressure where An = plate area.
at the helices, c = soil cohesion; ϕ = angle of internal friction Both the “cylindrical shear” and “individual bearing” methods
of the soil; and N c and N q = bearing capacity factors for general have similarities to the semiempirical methods presented by Mitsch
shear (Terzaghi 1943). and Clemence (1985) and Mooney et al. (1985) with one significant
For a circular plate anchor in clay (ϕu ¼ 0), the ultimate pullout- exception, namely, they do not use adjusted bearing-capacity fac-
capacity of a multiblade anchor using the “cylindrical shear” tors or earth pressure coefficients to reflect the difference between
analogy can be estimated in two ways. First, the capacity can be soil uplift and footing bearing capacity. The writer believes that the
assumed to be the result of shear along the cylinder of soil between concept of applying existing bearing-capacity solutions derived ini-
all the blades; therefore Eq. (1) will become tially for surface footings to the problem of anchor uplift is not
appropriate and is a poor representation of the real problem. A more
Qu ¼ nπDScu ð2Þ compelling argument could be made to use Terzaghi’s bearing-
capacity theory if the anchor was loaded in compression as opposed
to tension.
Qu = π DnScu Qu = π D ( n − 1) Scu + 0.25π D 2 (1.3) cu N c The final theory currently used when estimating the capacity of
multihelix anchors is based on installation torque. Nowadays, this
empirical method based on installation torque is often used by an-
chor manufacturers who have access to vast quantities of anchor
field data. The details regarding these correlations and the data
S on which they are based are usually considered proprietary by
the manufacturers. The required torque is monitored during instal-
lation until the anchor reaches its design depth. The torque readings
are then averaged over the distance of the final three helix diameters
of installation. This averaged torque value is then multiplied by the
n=2 torque factor to obtain the ultimate capacity.
n=2
S ( Blades ) S In light of these discussions, an obvious question asked by de-
sign engineers would be: which one of the three available methods
for estimating the capacity of multihelix anchors is the most appro-
priate to use and under what conditions? An insight into the accu-
D D racy of these existing methods can be found in the work of Hoyt
(a) (b) and Clemence (1991).
Hoyt and Clemence (1991) analyzed 91 load tests at 24 different
Fig. 2. Helical anchors: (a) cylindrical shear failure; (b) cylindrical
sites, with sand, silt, and clay soils all represented. All of the tests
shear and bearing failure
used in the study were short-term; most were strain controlled and
This final load was taken as the ultimate capacity Qact. The capacity Qu ¼ Qui ≤ Qu ð6Þ
i¼1
ratio Qact =Qcal was obtained for each test by dividing the actual
capacity (Qact ) by the calculated capacity (Qcalc ). Qcalc was calcu- where Qui = ultimate pullout-capacity of anchor i; n = total number
lated by using the three different load capacity models discussed of anchor plates; and Qu = maximum/limiting capacity of a multi-
previously, namely the cylindrical shear, individual bearing, and plate anchor. The average ultimate bearing pressure for each anchor
torque correlation methods. These data were then compared and is
plotted on separate histograms. All three capacity theories exhib- Q
ited the capability of both overpredicting and underpredicting mul- qui ¼ ui ð7Þ
A
tihelix anchor capacity. On average, the methods overpredicted the
capacity by around 50%, but this could be as much as 600%, using where A = area of an individual anchor. After Rowe and Davis
the individual bearing theory [Eq. (4)]. The cylindrical shear (1982), the analysis of anchor behavior can be divided into two
method was also found to underestimate the capacity more often distinct categories, namely, those of “immediate breakaway” and
Downloaded from ascelibrary.org by LEHIGH UNIVERSITY on 10/26/23. Copyright ASCE. For personal use only; all rights reserved.
than other methods. “no breakaway.” In the immediate breakaway case, it is assumed
The observations of Hoyt and Clemence (1991) help to confirm that the soil/anchor interface cannot sustain tension, so that, upon
the argument that neither “individual bearing” capacity theory nor loading, the vertical stress immediately below the anchor reduces to
“cylindrical shear” theory are appropriate for the problem of buried zero, and the anchor is no longer in contact with the underlying soil.
This represents the case in which there is no adhesion or suction
anchors under uplift loads. They are unable to reliably predict the
between the soil and anchor. In the no breakaway case, the opposite
likely ultimate capacity of multihelix anchors in a wide range of
is assumed, with the soil/anchor interface sustaining adequate ten-
soil types.
sion to ensure the anchor remains in contact with the soil at all
times. In reality, it is likely that the true breakaway state will fall
somewhere between the extremities of the “immediate breakaway”
Problem of Anchor Capacity
and “no breakaway” cases. The anchor analyses presented in this
paper are performed for the case in which the soil/anchor interface
Problem Definition is unable to sustain any tension. The interface between the soil and
A general layout of the problem to be analyzed is shown in Fig. 3. anchor is free to separate under tension, but contact can reestablish
The anchor has a total of n individual plates of diameter D, spaced itself if the analysis desires.
After allowing for immediate and no breakaway behavior, an-
evenly along the shaft at a distance of S. The anchor spacing ratio is
chors can be further classified as shallow or deep, depending on
defined as S=D and the anchor embedment depth is H=D. Referring
their mode of failure. This point is illustrated in Fig. 4. An anchor
to Fig. 3, the total pullout-capacity for the anchor system can be
is classified as shallow if, at ultimate collapse, the observed failure
expressed as a summation of the individual plate capacities, namely
mechanism reaches the surface [Figs. 4(a) and ]. In contrast, a deep
as anchor is one whose failure mode is characterized by localized
shear around the anchor(s) and is not affected by the location of
n
Qu = ∑ Qui
the soil surface [Figs. 4(c) and 4(d)]. For an individual anchor plate
i =1 of diameter, D, and soil type, (γ, cu ), there exists a critical embed-
ment depth, H cr (Fig. 4), at which the failure mechanism no longer
extends to the soil surface and becomes fully localized around the
anchor(s). This is true for both single and multiplate anchors. When
cu , γ this type of behavior occurs, the total ultimate capacity of the an-
Anchor shaft
Q Dshaft = shaft diameter chor will have reached a maximum limiting value Qu (the asterisk
H qu1 = u1
A
on Qu denotes deep failure).
In the context of a multiplate anchor arrangement (Fig. 3), the
type of deep failure mechanism will be a function of the anchor
spacing ratio S=D. If the anchor spacing S=D ratio is large enough,
i = 1 to n then each plate anchor will act independently of each other and
n = number of anchor plates
produce a collapse mechanism like that shown in Fig. 4(d). This
transition will occur at a critical value of S=D ≥ ðS=DÞcr . However,
'
H S Qu 2
qu 2 = π D2
A A= if the anchor spacing ratio S=D is small enough, such that each
4
plate anchor is influenced by the location of adjacent anchors, then
the collapse mechanism is likely to be global, like that shown in
Fig. 4(c). From a design perspective, the maximum capacity of
a multiplate anchor Qu will be achieved if each of the anchor plates
acts independently of each other.
S Qun
qun =
A Capacity of a Single Circular Anchor Plate in
Undrained Clay
In contrast to formulating a methodology for multiplate anchors
based on bearing-capacity theory, it is proposed to extend the plate
anchor methodology of Merifield et al. (2003). Therefore, the ul-
D timate anchor bearing capacity for each anchor plate qui will be
expressed as a function of the undrained shear strength in the
Fig. 3. Problem definition
following form:
(a)
(b)
(c) (d)
Fig. 4. Shallow and deep anchor behavior: (a) and (b) shallow failure mechanism; (c) global deep failure mechanism; (d) local deep failure mechanism
Qui
¼ qui ¼ cu N ci ð8Þ 14
A
12.6
where, for a homogeneous soil profile, 12
qui γH
N ci ¼ ¼ N coi þ ≤ N ci ð9Þ 10
cu γ≠0 cu
Finite-Element Analysis
Qui*
= ( 3 + 3π ) = 12.42
Downloaded from ascelibrary.org by LEHIGH UNIVERSITY on 10/26/23. Copyright ASCE. For personal use only; all rights reserved.
anchor break-out factor N coi from Eq. (10) for a range of embed- D S
2
ment depths and anchor geometries. Analyses were conducted on n = 1, 2,3
two- and three-plate anchor geometries (n ¼ 2; 3) at anchor spacing SMOOTH D
Anchor i
The displacement contours at collapse for anchors in weightless capacity of each anchor along the shaft increases linearly with over-
soil are shown in Fig. 9 for anchors at H=D ¼ 2 with two and burden pressure up to a limiting value. This limiting value reflects
three plates (n ¼ 2; 3). This figure illustrates that the displacement the transition from shallow to deep anchor behavior, in which the
pattern for the shallowest anchor (Anchor 1) remains largely un- mode of failure becomes fully contained around the anchor. At a
changed, regardless of the proximity of the underlying anchors. given embedment depth, an anchor may behave as either shallow or
This agrees with the results shown in Fig 8(a). deep, depending on the dimensionless overburden ratio γH=cu . The
For anchor spacing ratios of S=D ≤ 2, the mode of failure is very type of deep failure mechanism will also be a function of the anchor
similar to the cylindrical shear mode of failure shown in Figs. 2(a) spacing ratio S=D.
and 2(b). As the anchor spacing increases, the failure mode tends
to more localized above the anchor, and the displacement pattern Deep Multiplate Anchors
begins to extend laterally outward. In addition, the displacement A summary of the ABAQUS results for the limiting ultimate capac-
pattern for the lower anchors appears to be independent of the ity Qu [Eq. (6)] of a multiplate anchor system can be seen in Fig. 10.
overall embedment ratio H 0 =D but is related to the anchor spacing Results for two, three, and four anchor plates (n ¼ 2; 3; 4) are
S=D. This is regardless of the number of anchor plates and agrees shown in this figure.
with the previous discussions on an equivalent embedment ratio Fig. 10 illustrates the transition from a global deep failure
ðH=DÞeq [Eq. (15)]. mechanism, encompassing all anchor plates, to an individual deep
The numerical results already discussed have been limited to failure mechanism, in which a local failure mechanism exists
soil with no unit weight, and, therefore, the effect of soil weight around each anchor plate. The transition between the two cases
(overburden) needs to be investigated. Implicit in Eq. (9) is the occurs when the anchor spacing ratio reaches a critical value, when
assumption of superposition, in which the ultimate anchor capacity S=D ≥ ðS=DÞcr .
increases linearly with the dimensionless overburden pressure The FE displacement patterns for a number of problems in
γH=cu . This assumption has been proven to be valid by Rowe which the overburden and/or anchor spacing are sufficient to lead
(1978) and Merifield et al. (2003, 2001). Not surprisingly, further to a deep mode of failure are shown in Fig. 11. These figures il-
FE analyses that include cohesion and soil weight confirm that lustrate the transition between the two types of deep anchor failure
this is also the case for multiplate circular anchors. The ultimate mechanism previously shown in Figs. 4(c) and 4(d). A global deep
Eq. (19), in conjunction with Eqs. (16) and (18) are shown in
Fig. 10. These expressions can be used to estimate the limiting
ultimate capacity of a multiplate anchor system with sufficient
accuracy.
Effect of Anchor Shaft
To determine the effect of the anchor shaft, a number of single an-
chor analyses were run with the anchor shaft included. Both deep
and shallow anchor failure mechanisms were investigated, and
the results are shown in Fig. 12. This figure indicates the reduction
Downloaded from ascelibrary.org by LEHIGH UNIVERSITY on 10/26/23. Copyright ASCE. For personal use only; all rights reserved.
Table 1. Comparison of Laboratory Test Results (Narasimha Rao et al.1991) and Numerical Results
Narasimha Rao et al. 1991 This study
Measured experimental Calculated load
Test no. Screw pile no. n S=D H=D load Qu exp (kN) Qu calc =Qu exp Qu calc (kN) Qu calc =Qu exp
2.0 70%, illustrating the “cylindrical shear” method must be used with
1.8 caution and, in many cases, is inadequate.
1.6
1.4
Qu _ calc 1.2 Conclusions
1.0
Qu _ exp Current understanding of the behavior of helical anchors is some-
0.8
0.6 what unsatisfactory and has essentially remained unchanged for
0.4 20 years. With the use of complex numerical modeling software,
Current Study
0.2 Narasimha Rao et al 1991 this paper has thoroughly investigated the multihelix anchor prob-
0.0 lem with the aim of improving our knowledge of such foundations.
0 1 2 3 4 5 6 7 A better understanding of helical anchor behavior will hopefully
H lead to a wider acceptance as a foundation alternative and lead
(a) D to more economic and safer designs.
A practical design framework for multiplate helical anchor
1.2
foundations has been presented to replace existing semiempirical
1.0 design methods that are inadequate and have been found to be ex-
cessively under- or overconservative. The proposed method avoids
0.8 the need to apply bearing capacity theory, that was derived initially
Qu _ calc for surface footings, to the problem of anchor uplift. Importantly, it
0.6
Qu _ exp encompasses the complex relationship between the various modes
0.4 of failure, the anchor geometry, and the soil properties.
This new framework can be used by design engineers to con-
0.2 fidently estimate the pullout-capacity of multiplate helical anchors
Current Study
under tension loading. The method is based on rigorous FE analy-
0.0 ses on a wide range of problems.
0 1 2 3 4 5 6 7
H
(b) D References
Fig. 13. Comparison between numerical and laboratory results for Das, B. M. (1978). “Model tests for uplift capacity of foundations in clay.”
multihelical plate anchors: (a) laboratory results of Narasimha Rao et al. Soils Found., 18(2), 17–24.
(1991); (b) laboratory results of Narasimha Rao et al. (1993) Das, B. M. (1980). “A procedure for estimation of ultimate uplift capacity
of foundations in clay.” Soils Found., 20(1), 77–82.
Hoyt, R. M., and Clemence, S. P. (1991). “Uplift capacity of helical anchors
in soil.” Proc., 12th Int. Conf. on Soil Mechanics and Foundation
by the ratio of Qu calc =Qu exp , also shown in Tables 1 and 2 and Engineering [Comptes Rendus du Congres International de Mecanique
shown graphically in Fig. 13. des Sols et des Travaux de Fondations], Vol. 2, A. A. Balkema
Figs. 13(a) and 13(b) indicate a good agreement between the Publishers, Rotterdam, Netherlands, 1019–1022.
experimentally obtained capacities and the capacities calculated Hubbell, Inc. (2003). “Helical screw foundation system design manual for
using the proposed method. For the majority of cases, the calcu- new construction.” A. B. Chance Div., Hubbell, Inc., Centralia, MO.
lated capacities are within 10% of the measured values, which 〈http://www.vickars.com/screwpile_manual/index.html〉 (May 25, 2011).
Kulhawy, F. H. (1985). “Uplift behavior of shallow soil anchors—An over-
is adequate for design purposes. Also shown in Fig. 13(a) is a com-
view,” Proc., ASCE Convention, Uplift Behavior of Anchor Founda-
parison between the experimental capacities and the capacities cal- tions in Soil, ASCE, New York, 1–25.
culated by Narasimha Rao et al. (1991, 1993) using the “cylindrical Lutenegger, A. J., Smith, B. L., and Kabir, M. G. (1988). “Use of in situ
shear” method as given in Eq. (3). These estimates do not compare tests to predict uplift performance of multihelix anchors.” Special topics
as favorably and tend to overestimate the capacity by as much as in foundations (GSP 16), ASCE, New York, 93–110.