The influence of bond stress distribution on ground anchor fixed length design.

Field
trial results and proposal for design methodology
L'influence de la rpartition des contraintes sur les tirants d'ancrage de longueur fixe. Rsultats de
planche d'essais et proposition de mthodologie de conception

Vukoti G.
Kellerterra, S.L.
Gonzlez Galindo J., Soriano A.
Universidad Politcnica de Madrid

ABSTRACT: This paper presents a brief analysis and comparison of different recommendations for a ground anchor fixed length
design and a load transfer capacity at grout-ground interface, comparing it with the full scale test results recently carried out in Spain.
Simple methodology for ground anchor routine design is proposed, incorporating efficiency factor as a conceptual control of anchor
capacity, and fixed length criteria to determine range of application of conventional anchors, with single fixed length unit, or Single
Bore Multiple Anchors (SBMA). Ground anchors discussed in this paper are cement pressure grouted, formed by pre-stressed strand
tendons that are installed in soil or rock.

RSUM: Cet article prsente une brve analyse et la comparaison des diffrentes recommandations pour la conception de tirant
dancrage de longueur fixe et la capacit de transfert de charge linterface coulis-sol, en le comparant avec les rsultats dessais
chelle relle obtenus en Espagne au cours des dernires annes. Une mthodologie simple pour la conception de tirant d'ancrage est
propose; elle intgre le facteur d'efficacit comme un contrle conceptuel de la capacit de lancrage et les critres de longueur fixe
pour dterminer le champ d'application des tirant dancrage conventionnels, dunique longueur fixe, ou dancrages multiples en un
unique forage (SBMA pour ses sigles en anglais). Les tirants d'ancrage dcrits dans ce document sont injects pression au moyen
dun coulis de ciment et sont forms par un faisceau de cbles dacier prcontraints qui sont installs dans le sol ou la roche.
KEYWORDS: anchor, fixed length, bond stress, efficiency factor.

1 INTRODUCTION This formula differs from experimental and theoretical

evidence that corroborate that there is no linear dependency of
Nowadays anchors represent a key medium to sustain and ultimate capacity on fixed anchor length.
strengthen slopes formed by instable soils and fractured rocks,
and to ensure the stability of various types of gravity structures. 2.2 Non-uniform bond stress distribution
Regarding bond stress and load transfer capacity at the
grout-ground interface, most procedures for anchor design are It is fully acknowledged by numerous researchers that the
empirical values or formulas derived by local experiences, very distribution of stress along the fixed anchor length is non-
difficult to extrapolate for different locations or execution uniform, both at low stress levels and at failure. This
systems. The design procedure is often simplified, considering phenomenon results from the general incompatibility between
direct proportionality between fixed anchor length and its load elastic modulus and corresponding deformation of the anchor
capacity, as it is prevailing practice in Spain and South strands, cement grout and ground.
America. Field tests on instrumented conventional anchors, reported
However, since late 1960s numerous authors demonstrate by Muller (1966), Berrardi (1967), Ostermayer (1974),
bond stress or skin friction distribution to be highly non- Ostermayer and Scheele (1977), Mastrantuono and Tomillo
uniform at all stages of loading, with high bond stress (1977), Barley (1995) and Briaud et al. (1998), showed that
mobilization along reduced fixed length. In the following when applying the initial load the bond stress is concentrated
chapters, based on presented references and analysis of the field over the proximal length of the fixed anchor, leaving a
trial results recently performed in Spain, the methodology for significant part of the fixed length towards distal end unstressed.
anchor fixed length design is proposed. By the evolution of the load the bond stress concentration zone
is transferred along the fixed anchor as the bond stress along
either the tendon/grout or grout/ground interface is exceeded.
2 DESIGN OF THE FIXED ANCHOR LENGTH

2.1 Current practice

Design assumption of uniform load distribution along the fixed
anchor length is not only limited to usual methodology and
standards in Spain and South America but is internationally
Figure 1. Development of bond stress distribution along a fully bonded
generally adopted. Considering this hypothesis the ultimate or fixed anchor length (Barley 1995).
capacity of the anchor is commonly expressed as follows:
Simultaneously, due to progressive debonding, stress at the
Tult d L fix ult (1) proximal end reduces to residual values. The bond stress
concentration zone reaches distal end of the anchor just before
where: d = anchor diameter, Lfix = anchor fixed length and the failure, as it can be seen in Figure 1.
ult = ultimate bond stress.

2119
 Proceedings of the 18th International Conference on Soil Mechanics and Geotechnical Engineering, Paris 2013

The same mechanism of a non-linear load and bond One of the most extensive attempts to model construction
distribution was confirmed by laboratory full-scale test technique, characteristics and behaviour of anchors have been
accomplished by Weerasinghe (1993). It is also important to accomplished by Mecsi (1995), based on analysis of results
mention investigation done by Coates and Yu (1970), which from numerous installed and monitored anchors.
studied stress distribution around a cylindrical anchorage in Analytical solution and simple graphical method based on
triaxial stress field using finite element methods. The results the theory of expanded cylindrical cavity provide the possibility
emphasize the non-uniform bond stress distribution for the ratio to define the approximate pull-out capacity. The Analysis of
of the elastic modulus of the anchor material (EA) and the rock load distribution for the known anchor geometry and rigidity
(ER) less than 10 (EA/ER<10) , which is very common for wide permits determination of the specific pull-out resistance of a 1
range of rocks and soils in which anchors are usually m anchor length (tult) and the length of the fully mobilised bond
constructed. stress (Lb). Considering that only reduced percentage of
maximum bond stress can be mobilised over the remaining
2.3 Efficiency factor fixed anchor length (Lfix Lb), the ultimate anchor capacity can
be expressed by the following expression:
There have been a number of attempts to quantify the non- 1
uniform load distribution and to introduce effects of progressive Tult ult L 0 thk(L fix L 0 )
k (7)
debonding into Formula 1. Casanovas (1989) recommended
design based on definition of apparent fixed length (Lve) over ult
which the ultimate bond stress can be mobilized: k
Esteel Asteel ult
1 (8)
ult
L fix log( 0.1 ) where: k = rigidity index, Esteel = steel deformation modulus,
0
L ve ( ) L0 (2) Asteel = steel tendon area, ult = (2)
elongation of the shear strength
L0 length (Lfix L0).
where: Lve = apparent fixed length over which ult (kN/m2)
operates, L0 = reference length of 1 m, 0 = reference value of 1 Based on data from Ostermayer and Scheele (1997), Woods
kN/m2. and Barkhordari (1997) proposed efficiency factor for its
incorporation in the expression for ultimate capacity of low-
To understand better efficiency factor concept it is possible pressured anchors in sand (Formula 10), recommended in BS
to analyze Figure 1 and to compare area A, that corresponds to 8081 (1989), which is a function both of fixed anchor length
the final and maximum load stage, with the total area below ult. and friction angle:
Area. A L fix (9)
f eff f eff exp(0.05 tan )
Area.below. ult L0
(3)
Then, ultimate anchor capacity can be expressed as follows: where: L0 = reference length of 1 m.
Tult d L fix fix f eff
(4) Tult f eff L fix ntan( ) (10)
Research based on over 60 full scale tests performed on
different anchor fixed lengths, installed in wide range of soil
(clays, silty clays, sandy clays, boulder clay and glacial till), 2.4 Single Bore Multiple Anchors - SBMA
permitted development of the concept of the efficiency factor This system involves the installation of a multiple unit anchors
(Barley 1995 and 1997, Barley and Windsor 2000). Figure 2 into a single borehole, with enough short unit lengths to reduce
presents the distribution of the values of the efficiency factor or even to avoid the progressive debonding. Each unit is formed
(feff) against anchor fixed length, and the best fit curve can be by individual tendon and is loaded with the corresponding unit
expressed by following expression: stressing jack, mobilizing its own capacity independently of
0 , 57
L fix other unit anchors.
f eff 1,6 Application of this system permits the unlimited theoretical
L0 (5) total fixed length, while conventional anchors formed by only
It is important to emphasize that Barleys efficiency factor is one unit do not provide beneficial effects in load capacity for
quite consistent with Ostermayer (1974) diagrammatic fixed length superior to 10 m as is stated by numerous authors
presentation of the ultimate medium skin friction against fixed and design guidelines or codes.
length for similar soil characteristics and anchor construction
process, as it can be seen in Figure 2.

Figure 3. Development of bond stress along a four unit single bore

multiple anchor (Barley, 1997).
Total fixed lengths of 10 to 20 m are frequently used in all
types of soils, achieving high anchor capacities (2000 to 5000
Figure 2. Ostermayers (1974) boundary lines vs. Barleys (1995) kN) that are almost three times greater than the normal
efficiency factor.
conventional anchors capacity values (Barley 1997, Barley and
Barley (1995) also suggested the efficiency factor for sands, Windsor 2000). A comparison of load distribution between
correlating efficiency with the fixed length and the friction conventional and SBMA anchor is presented in Figure 3.
angle:
Lfix
tan
f eff (0,91) L0 (6)

2120
 Technical Committee 207 / Comit technique 207

3 GENERAL FIELD TEST DETAILS 3.2 Test results in silty clays

The results of full-scale field tests, undertaken to verify the Table 2 summarizes results for trial anchors, 2 SMBA with two
influence of the fixed anchor length on the ultimate capacity units (in total 4 units) and two conventional anchors, performed
(Tult) and the average bond stresses (ault) obtained at failure in in stiff silty clay (CL) at depth of 15 to 25 m. SBMA anchors
gravelly sands, silty clays and clayey marl are presented in this were formed by two units of 2.5 m fixed length, and
chapter. conventional anchors had fixed length of 8.0 m. Diameter of all
Investigation test were done according to the Test Method 1 test anchors was of 160 mm.
(EN 1537), with incremental load cycles from a datum to
ultimate load. The test involved very strict measurement of Table 2. Test results for pressure grouted 160 mm diameter anchors in
tendon displacement versus applied load and, at the peak of silty clays.
each cycle, measurement of tendon displacement versus time. N/Type of Anc. Lfix Tult ault feff
Anchors subjected to investigation were high-pressure grouted,
m kN kN/m2
designed to fail at grout/ground interface. Field tests were
carried out near previously performed boreholes to establish SBMA (4 units) 2.5 400440 318-350 1
better correlation between obtained data and soil characteristics.
Values of efficiency factor presented in Table 1, 2, 3 and 4 2 Conventional 8.0 915-950 227-236 0.65-0.74
are calculated comparing average bond stresses (ault) of SBMA
units and conventional anchors, considering that units that
formed SBMA anchors with fixed length of 2.5 3.0 m were 3.3 Test results in clayey marl
short enough to avoid progressive debonding effects. Due to The trial anchors were carried out in two different sites with
that hypothesis it can be assumed that efficiency factor for similar soil characteristics, stiff to very stiff clayey marl (CH-
individual units of SBMA anchor have the value of 1. CI), with sand proportion less than 10%.
To avoid possible influences and collaboration of the free
anchor length on pull-out capacity and bond stress distribution, 3.3.1 Test site A
fixed units of all test anchors were separated by specially The trial anchors were constructed at depth of 20 m, with
designed compressible joints that prevent load transmission diameter of 150 mm. SBMA anchors were formed by two units
from fixed to free anchor length. of 3 m fixed length, and conventional anchors had fixed length
of 11 m.
3.1 Test results in gravelly sand
Table 3. Test results for pressure grouted 150 mm diameter anchors in
The trial anchors, six SMBA with two units (in total 12 units) clayey marl Test site A.
and two conventional anchors, were executed in dense to very
dense gravelly sand, with silt content that varied from 5 to 15%, N/Type of Anc. Lfix Tult ault feff
at depth of 20 to 30 m. 2
SBMA anchors were formed by two units of 2.5 m fixed m kN kN/m
length, while conventional anchors had fixed length of 7.5 m. 2 SBMA (4 units) 3.0 450 318 1
Diameter of all test anchors was of 178 mm. The following
table shows ranges of obtained results: 2 Conventional 11.0 785 110-152 0.35-0.48

Table 1. Test results for pressure grouted 178 mm diameter anchors in

gravelly sands.
Results presented in Table 3 demonstrate high inefficacy of
fixed anchor lengths longer than 8 or 10 m.
N/Type of Anc. Lfix Tult ault feff Efficiency factor for 11 m long fixed length obtained
m kN kN/m2 comparing average bond stresses varies from 0.35 to 0.48. It is
important to emphasize that results feet well with efficiency
6 SBMA (12 units) 2.5 780960 558-687 1 factor values proposed by Barley (1995), as it can be seen in
Figure 5.
2 Conventional 7.5 1880-1920 445-455 0.65-0.8

It is worthy to mention that values of the obtained ultimate

load capacities are very consistent with Ostermayer and Scheele
(1978) diagrammatic presentations of the load capacity against
fixed length (see Figure 4).

Figure 5. Efficiency factor obtained in field tests in silty clays and

clayey marlvs. values proposed by Barley (1995).

3.3.2 Test site B

The trial anchors were executed at depth of 30 to 35 m, with
diameter of 178 mm. SBMA anchors were formed by two and
three units of 2.5 and 3 m, while conventional anchors had fixed
length of 7.5 m.
Figure 4. Ultimate load capacity obtained in field test vs. values
proposed by Ostermayer and Scheele for sand and gravel (1977).
Table 4. Test results for pressure grouted 178 mm diameter anchors in
clayey marl Test site B.

2121
 Proceedings of the 18th International Conference on Soil Mechanics and Geotechnical Engineering, Paris 2013

is the application of SBMA. In this case fixed length of each

N/Type of Anc. Lfix Tult ault feff unit that forms SBMA is calculated considering corresponding
m kN kN/m2 efficiency factor (feff) Equation 3.
If the fixed length, obtained considering non-uniformity is
8 SBMA (18 units) 2.5 450540 322-386 1 greater than 10 m it is recommended to apply SBMA anchors.
2 Conventional 7.5 960-1080 230-254 0.59-0.79

3.4 Summary
An extensive series of field anchor tests performed in different
soils showed that:
- there is no direct proportionality between fixed anchor
length and its ultimate load capacity;
- obtained values of the average ultimate bond stresses
in cohesive soils fit well with Ostermayers (1974)
diagrammatic presentation of skin friction against fixed
length, while results obtained in gravelly sands fit well
with Ostermayers and Scheele (1978) presentation of
ultimate load capacity vs. anchor length;
- ranges of obtained efficiency factors are consistent
with tendency of values proposed by Barley (1995);
- efficiency factor can be considered as a conceptual
control of anchor ultimate capacity;
- fixed anchor lengths longer than 10 m do not Figure 6. Flow chart for design of fixed anchor length.
contribute significant beneficial effects on capacity.
- SBMA anchors permits construction of high anchor
capacities that approach more than two times that of 5 REFERENCES
the conventional anchors which utilize long inefficient
fixed length. Barley A.D. (1995). Theory and Practice of the Single Bore Multiple
Anchor System. Proc. Int. Symposium Salzburg. Balkema
Rotterdam, pp 293-301.
4 PROPOSAL FOR DESIGN METHODOLOGY Barley A.D. (1997). The Single Bore Multiple Anchor System. Proc.
Int. Conf.: Ground anchorages and anchored structures. London,
Based on the information presented in previous chapters, pp. 65-75.
proposal for design methodology for cement grouted anchors Barley A.D. and Windsor, C.R. (2000). Recent advances in ground
anchor and grout reinforcement technology with reference to the
formed by steel tendons is presented below, considering most development of the art. GeoEng 2000, Int. Conf. on Geotechnical
important parameters that define its capacity, like: soil and Geological Engineering. Melbourne, pp. 1083-1094.
characteristics, execution process, ultimate and average bond Berardi G. (1967). Sul Comportamento Deglic Ancoraggi Immersi in
capacity, fixed length, type of anchors (conventional or SBMA), Terreni Diversi. Universidad Genoa, Inst. Contr. Sc. Serie III (60),
stress distribution and efficiency factor. pp 18-19.
Emphasis is placed on the effects of progressive debonding Briaud J.L., Powers W.F. and Weatherby D.E. (1998). Should Grouted
that cause the non-uniform stress distribution along the fixed Anchors Have Short Tendon Bond Length?. Journal of Getechnical
length, with efficiency factor as a conceptual control of anchor and Geoenvironmental Engineering ASCE, pp. 110-119.
BrS 8081 (1989). British Standard Code of Practice for Ground
capacity. Due to the number of parameters that enter the Anchorages. BSI, London.
analysis, recommended methodology has an iterative character, Casanovas (1989). Anchoring in rock. Elsevier Scientific Publishing
as it can be seen in Figure 6. Some of the most important steps Company, pp 0-33.
of the flow chart are commented below. Coates D.F. and Yu Y.S. (1970). Three dimensional stress distribution
Phase I: Evaluation of the site subsoil conditions and around a cylindrical hole and anchor. Proc. of 2 Int. Conf. on Rock
relevant properties of in situ soil and rock, as a factor that Mechanics. Belgrade, pp. 175-182.
EN 1537 European Standrad (2010). Execution of special geotechnical
directly influence steps in the Phase II (construction system and
work Ground Anchors.
skin friction estimation). Mesci J. (1997). Some Practical and Theoretical Aspects of Grouted
Phase II: For the skin friction estimation it is recommended Soil Anchors. Proc. of Int. Conf. Ground Anchorages and
to use at least two sources, taking into account the concept or Anchorages Structures. London, pp. 119-130.
formula that will be applied for the anchor design. If pre-design Muller H. (1966). Erfhnungenmit Verakurungen System BBRV in Fels-
load tests are performed to evaluate ultimate anchor load und Lock-ergensteinen. Schweisezerische Bauzeitung, 84 (4), pp.
capacity, construction process has to be exactly the same as 77-82.
planned for production anchors, and fixed lengths should be Ostermayer H. (1974). Construction carrying behavior and creep
similar with test anchors.For the first iteration anchor length is characterics of ground anchors. Int. Conf. On Diaphragm Walls and
Anchorages. I.C.E. London, Septiembre 18-20, pp. 141-151.
calculated considering uniform bond stress distribution Ostermayer H. and Scheele F. (1977). Research and Ground Anchors in
(Equation 1). If calculated fixed length is larger than 5 m, Non-Cohesive Soils. Gotechnique 3, pp. 92-97.
construction process can be reconsidered (Alternative A), Weerasinghe R.B. (1993). The Behaviour of Anchorages in Weak
varying anchor diameter or type of grouting, with objective to Mudstone. PhD Tesis, University of Bradford.
reduce fixed length up to 5 m. Other option (Alternative B) is to Woods R.I. and Barkhordari K. (1997). The Influence of Bond Stress
introduce directly the efficiency factor. Distribution on Ground Anchor Design. Proc. of Int. Conf. Ground
Phase III: If the fixed length obtained considering non- Anchorages and Anchorages Structures. London, pp. 55-64.
uniformity is in the range between 5 and 10 m, two alternatives
are proposed. First alternative considers conventional type of
anchor, with unique fixed length unit calculated taking into
account efficiency factor (feff) Equation 2. Another alternative
2122