NGE4_ROCK ANCHORS

Anchor Materials and Anchorage Methods

Permanent rock anchors performance shall meet the following criteria:

a. High reliability of materials and installation of both the tieback and head
components;
b. Static or cyclic structural loads as high as 5MN;
c. Predictable low deformation tolerances;
d. Minimum service life of 50 years.

Allowable working loads and safety factors

The allowable working load of an anchor is the design load that the anchor is required
to sustain under normal service conditions.

The allowable working load is expressed as a percentage of the characteristic strength

of the steel. The characteristic strength of the steel may be either the guaranteed
ultimate tensile stress (GUTS) or the yield stress.

The allowable working load is generally taken to be between 50% and 62.5% of the
ultimate tensile strength, i.e. the factor of safety against failure of the anchor material is
between 2 and 1.6, but usually a safety factor of 2 is considered for most permanent
applications.

Steel relaxation

Tensioned anchors may lose load with time as a result of both steel relaxation and
creep of the anchorage.

The factors that influence steel relaxation are the stress level, the service temperature,
time after stressing, and in the case of a strand, the tendency to unwind.

At stress levels up to 50% of the ultimate strength, relaxation is negligible. For stress
levels of 75% of the ultimate strength and temperatures of 20ºC, a load loss of 5-10%
of the applied stress occurs in ordinary stress relieved steel.

The equation defining the loss of stress due to relaxation at normal ambient
temperatures is as follows (Libby, 1977):

log 10 (t) σ i
Δσr = σi . . ( −0.55)
10 σy

Δσr - relaxation stress loss at time t hours after stressing;

σi – initial stress;
σy – 0.1% offset stress.

This equation is only applicable when initial stress is higher than 0.55 of the 0.1% offset
stress. Below this limit, relaxation is negligible.

In situations where the levels of relaxation are unacceptable, restressing at time of

1000 hours will reduce the further relaxation to about one quarter of its normal value at
an initial stress of 70% of GUTS.
Another method of reducing relaxation is to overload the anchor at the time of initial
stressing and hold this stress for a period of up to 10 minutes which disposes of the
rapid initial relaxation (Littlejohn and Bruce, 1976).

It is also found that the relaxation rate increases rapidly at temperatures over 20ºC
which may be of significance in some applications.

Strength properties of steel bar and strand

Properties of steel bar and strand anchors that are required for design are the yield
stress, the ultimate tensile stress, the elastic modulus and the relaxation
characteristics. Manufacturer´s specifications will provide this technical information for
design purposes (refer Dywidag).

Applications of rigid bar anchors

Types of steel bars used as rock anchors include deformed reinforcing steel and
continuously threaded bar such as Dywidag Threadbar or Williams all-thread bar.

Deformed or threaded bars are normally used because of the improved steel-grout
bond strength in comparison with smooth bar.

Reinforcing steel is used where the primary function of the anchors is to secure a
footing to a rock surface and the loading conditions consist of purely compressive
loads, or uplift and/or shear loads only occur infrequently. The installation procedure
would be to drill a pattern of holes in the rock foundation, anchor the reinforcing bar
with cement grout, and then cast the footing with the exposed part of the anchors
embedded in the concrete to form a passive anchor; or they could be sleeved trough
the concrete and the prestressed against the top surface of the concrete footing.

Rigid bar anchors are commonly installed where the design working load is in the
range of about 100-600 kN and where the required length is less than about 8 m.

The advantage of rigid bars is the ease of handling short lengths which can be coupled
together as required. The disadvantages of bar anchors are their limited load capacity
and the difficulty to handling long, continuous lengths.

Applications of strand anchors

The required load capacity of strand anchors is obtained by assembling a bundle of

strands up to 25 strands which has an ultimate load capacity in excess of 4000 kN and
requires a drill hole with a diameter of at least 0.20 m.

Because it is not possible to join lengths of strand, the entire anchor assembly, with the
corrosion protective sheaths, has to be fabricated in one piece. As the weight of the
anchor can be considerable, the installation method is an important consideration
during the design phase.

Cement grout anchorage

Cement grout is the most used in anchors for permanent installations when compared
with epoxy resin and mechanical anchors.
The advantages of this type of anchorage are the availability and low cost of the
materials, simple installations procedures and its suitability for a wide range of soil and
rock conditions. In addition, cement provides an environment that protects the steel bar
or strand from corrosion, and when properly installed the strength of the grout will
improve rather than deteriorate with time.

The disadvantage is that if the rock is fractured, grout may flow into even fine fractures
(width greater than about 0.25 mm), resulting in an incompletely filled hole with a brittle
grout after setting that can be damaged by movement during installation and stressing.

For the hole diameter of a grout anchorage, consideration shall be given to economy
(as small as possible) and as to have sufficient annulus thickness so as to transmit the
shear stresses from the steel to the surrounding rock. Usual practice in the selection of
the drill hole diameter is to ensure a diameter of the hole (dh) to diameter of the
anchorage (da) ratio that falls within:

dh
= 1.5 to 2.5
da

The low end of this range would be used in strong massive rock while the high end
would be used in fractured rock.

The most important factors for the selection of the bond length are the strength and
fracture characteristics of the rock in the bond zone. The working bond stress has
values which range from 350 kPa for weak rock to a high of 1400 kPa for strong rock.

The grout mix used in the anchorages must be strong enough to withstand the high
stresses that develop around the steel anchor, must not degrade with time and must be
non-corrosive so as not to affect the properties of the steel anchor.

The factors to consider in grout mix design are the water-cement ratio, setting time,
and use of additives to reduce shrinkage and improve workability. Use of grout mixes
containing sand or fine aggregate is usually not recommended as these granular
materials tend to block grout tubes.

A water cement ratio between 0.4 and 0.45 will produce a grout that can be readily
pumped down small diameter grout tubes and will produce a strong, continuous grout
column.

While grout mixes comprising only cement and water are generally satisfactory,
additives can be used to improved grout performance, namely, control bleed, improve
flowability, reduce water content and retard set. A common non-shrink agent used is
Intraplas N (Sika Products) that is added to the mix at about 1% of the cement weight.

Rock anchors are usually grouted at atmospheric pressure. Pressure grouting is only
used where the rock is sufficiently loose and fractured that the grout will be forced into
the rock mass to consolidate and strengthen it, forming a mass of grouted rock integral
with the anchor zone. Depending on the degree to which the rock is fractured, pressure
grouting can increase capacity of anchorage up to 100%.

The bar or strand must be fully embedded in a continuous and uniform grout column to:
develop a high strength bond in the anchor zone and be protected against corrosion.
This can be achieved by centralizer sleeves placed at intervals along the bond zone
(usually between 0.5 and 3m).
Design Procedure for Tensioned Anchors

Load transfer mechanism between anchor and rock

When a tensile load is applied to a steel bar or cable that is anchored in rock with a
column of grout (either cement or epoxy resin), shear stresses are developed at both
the steel-grout and grout-rock interfaces. Under elastic conditions, the shear stress
distribution is non-linear with high stresses concentrated at the top of the bond length
which diminish rapidly down the hole.

As the applied stress is increased, a greater portion of the tensile load is carried at the
distal end of the anchor. As the load is further increased, the bond at the proximal end
of the bond length will start to fail. Once the bond has been broken, the shear strength
will be equal to the friction of the surface. General design practice is to select a
combination of applied load and anchorage dimensions such that there is no slippage,
this is, no bond failure.

Allowable bond stresses and anchor design

The non-linear shear stress distribution along the anchor length is difficult to predict
due to a wide range of conditions that may exist within a tensioned anchor. For this
reason and for design purposes, shear stress is taken as uniformly distributed along
the bond length (simplified assumption).

The magnitude of the average shear stress for both rock-grout and grout-steel
interfaces has been established empirically from the results of tests on full-scale and
laboratory anchors.

Calculation of the bond length is a two stage process that ensures that the working
bond stresses are not exceeded at either rock-grout or grout-steel interfaces. First, the
bond length and drill hole diameter are selected such that the average shear stress at
the rock-grout interface is less than or equal to the working bond strength. Second, the
length of the anchor is checked against the required design development length of the
steel which is the length of embedment required to support the applied tensile load.

Assuming that the shear stress is uniformly distributed on the surface of the drill hole,
the bond length is calculated from:

Q
Lb =
π . d . τa

Lb – bond length
Q – applied load at the head of the anchor
d – diameter of drill hole
τa – working bond strength of rock-grout interface
In design, the average bond stress can be matched to the working bond stress by
increasing the bond length or the hole diameter as required. However, a practical limit
on the bond length is in the range 8-10 m, with usual rock drilling equipment limiting the
drill hole diameter to about 150 mm. The reason for this restriction is that the peak
stress is developed at the proximal end of the bond and if this stress is greater than the
ultimate bond strength, failure of the grout in the proximal end of the bond will occur
regardless of the bond length.
An approximate relationship between the rock-grout bond strength and the uniaxial
compressive strength of the rock has been developed from the results of load tests on
anchors installed with cement grout anchorages in a wide range of rock types and
strengths. The following equations provide design values for working and ultimate rock-
grout bond strengths for cement grout anchorages:

σ u (r )
τ a =
30
up to a maximum value of 1.4 MPa;

σ u (r )
τ u =
10
up to a maximum value of 4.2 MPa;

τ – rock-grout allowable (working) strength,

a

τ – rock-grout ultimate strength,

u

σ – ultimate rock uniaxial compressive strength of weakest rock in the bond zone.
u(r)

Values of τa are obtained assuming a safety factor of 3 applied to τu. Refer table below
for reference values.

ROCK TYPE τ a (MPa)

Granite, basalt 0.55 – 1.0
Dolomitic limestone 0.45 – 0.70
Soft limestone 0.35 – 0.50
Slates, strong shales 0.30 – 0.45
Weak shales 0.05 – 0.30
Sandstone 0.30 – 0.60
Concrete 0.45 – 0.90
Weak rock 0.35 – 0.70
Medium rock 0.70 – 1.05
Strong rock 1.05 - 1.40

Judgement should be used in the application of equations and table above for rock-
grout strengths. Unfavourable conditions requiring a low value of τa include a smooth
hole surface produced by rotary drilling compared with percussion drilling, a zone of
loose fractured rock in the bond length, drill cuttings smeared on the walls of the hole,
holes from that cannot be completely cleaned and flowing water.

Actual conditions in the hole are likely to be unknown, usual practice is to conduct
performance tests on selected anchors to ensure that anchor meets specified
acceptance criteria.

Second step in anchor design is to check that shear stress developed at the steel-grout
bond interface does not exceed working bond stress. Working development lengths
required to develop the full strength of the bar are given by the following equations
(Canadian Portland Cement Association, 1984):
 0.019 . A b . σ y
Ld = for 35 mm diameter bars and smaller, but not less than Ld = 0.058.db.σ y;
√ σ uc
26 . σ y
Ld = for 45 mm diameter bars;
√ σ uc
34 . σ y
Ld = for 55 mm diameter bars;
√ σ uc
0.36 . d b . σ y
Ld = for prestressing strand;
√ σ uc
Ld – development length (mm)
Ab – cross sectional area of bar (mm2)
σy – yield strength of non-prestressed reinforcement (MPa)
σuc – compressive strength of grout (MPa)
db – nominal diameter of bar or strand (mm)

Equations above determine working development lengths which should be suitable for
most anchoring designs, although a safety factor up to about 1.5 may be used in poor
anchoring conditions.

Prestressed and passive anchors

Prestressed and passive anchors can be used to support tension loads. The
advantages of using prestressed anchors instead of passive anchors are: smaller
deflection on anchor head and anchorage strength increase due to better confinement
of rock around anchorage.

Rock cone failure

Other failure mode to consider in design is the pull out of the rock mass around the
anchorage in the form of a cone with its apex near the midpoint of the bond length
(consideration based on uplift test results).

After establishing the bond length of the anchorage, the next step is to check that the
anchor will mobilize a sufficient volume of rock to support the applied load.

A simplifying assumption can be made that the apex angle is 90º (sides of cone at 45º
in regard to vertical axis of the cone) and that the position of the apex is at the midpoint
of the bond length.

Resistance against the pull out failure mode of the rock cone is provided by the cone
weight and the strength of the rock on the surface of the cone.

A narrow apex angle should be used for weak rock (60º). In strong rock, the apex angle
may be as great as 120º. The rock cone shape is also strongly influenced by rock
fractures, with the most favourable case being that of fractures perpendicular to the
anchorage, and the least favourable being that of fractures parallel to the anchorage.

The strength of the rock cone surface can be estimated from the following equation:
 σ u (r) 1
σ t= . [ m−(m2+ 4 . s )1/ 2 ] .
2 FS

σt – working tensile strength on surface of cone (negative number)

σu(r) – ultimate unconfined compressive strength of rock
m, s – rock mass constants (refer table below)
FS – factor of safety applied to rock strength

FS may vary from 2 for massive rock with predominant discontinuity set at right angles
to the anchor, to 4 for closely fractured rock, or where the discontinuities are parallel to
the anchor.

The buoyant weight of the cone Wc (refer figure on next page) is:
 π θ
W c = . tan 2
3 2()
. [ D 3 . γ r−( D−D w )3 . γ w ]

The resisting force f(r) developed on the curved surface area of the cone is:

σ t . π . D2 . tan (θ /2 )
f (r )=
cos ( θ/2 )

The uplift capacity Q of the rock cone is:

f (r )+ W c . cos ( Ψ c )
Q=
FS

θ – apex angle of rock cone (degrees)

D – depth of apex below ground surface
Dw – depth of water table below ground surface
r – rock unit weight
w – water unit weight
c – angle between upward vertical and load direction (degrees)
FS – factor of safety

Group action

Where a number of anchors are required to support the structural load, the combined
effect of the group of anchors must be evaluated.

In order to prevent excessive stress concentrations being developed around the

anchors that could fracture the rock, and minimize the risk of drill holes intersecting, it
is usual practice to specify both a minimum spacing and a stagger between the bond
zones. While there are no codes defining spacing and stagger, one commonly used
criterion for the minimum spacing is that it should be the lesser value of four times the
diameter of the bond zone or 1.2 m.

Also, the South African Code of Practice (1972) recommends that for anchors spaced
at less than 0.5 times the bond length, the stagger between alternate anchors should
be 0.5 times the anchor length.

Cyclic loading of anchors

Where the anchorage is in closely fractured rock, the cyclic loading may cause
loosening and dilation of the rock mass and eventual reduction in the capacity of the
anchor.

The installation of prestressed anchors under these conditions will maintain the
interlock between the blocks of rock in the anchor zone and minimize the risk of
movement of the anchorage. In addition, placement of the top of the bond zone at
some depth below the rock surface will provide confinement to the rock in the most
highly stressed area of the rock and minimize the risk of loosening of the rock mass.

Time dependent behavior and creep

Time dependent behavior of rock anchors will result from both relaxation of the steel
bar or strand, and creep of the grout and rock in the bond length.

Corrosion Protection

This is one of the most important aspects of design and construction of rock anchors.

Permanent anchors are defined as those with a service life exceeding 2 years.
Corrosion protection measures should always be provided for permanent anchors.

The general requirements of a corrosion protection system are as follows:

a. There will be no break down, cracking or dissolution of the protection system

during the service life of the anchor.
b. The fabrication of the protective system can be carried either in a plant or on
site in such a manner that the quality of the system can be verified.
c. The installation and stressing of the anchor can be carried out without damage
to the protection system.
d. The materials used in the protection system must be inert with respect to both
the steel anchor and the surrounding environment.

The material most commonly used for corrosion protection is cement grout, primarily
because it creates a corrosion free environment, is inexpensive, simple to install, has
sufficient strength for most applications and has a long service life. The downside to
this material is its tendency to crack. To minimize shrinkage cracks, common practice
is to incorporate additives in the grout mix to reduce shrinkage and reduce bleed.

Testing

Water tightness test

Cement grout may flow into fractures with apertures greater than 0.25 mm and it will
be necessary to seal such fractures prior to installing the anchor.

Water tightness of the drill hole can be tested by filling the hole with water and
subjecting it to a pressure of 35 kPa in excess of the hydrostatic head as measured at
the top of the hole. The rate at which the water level falls in the hole is observed and
the hole is acceptable if the seepage rate does not exceed 9.5 liters over a 10 minute
period (time should be allowed for the rock in the walls of the hole to be saturated). If
this rate is exceeded, then it is necessary to grout the hole using a low water-cement
ratio grout or a sanded grout, let the grout set for a period of about 8-24 hours, and
then redrill the hole.
The water tightness test would then be repeated to ensure that the fractures had been
sealed, and if necessary, the hole is regrouted and redrilled.

Alternatively to filling the entire hole with water, packers could be used to isolate the
bond zone to determine seepage conditions only in this area.

Load testing

The stressing procedure for all permanent anchors incorporates tests to evaluate their
performance and ensure that they meet specified acceptance criteria.

The tests comprise applying a load to the bar or strand with a hydraulic jack, and
monitoring the extension of the anchor by measuring the movement of the head.

The purpose of making load-extension measurements is to ensure that the anchor is

behaving elastically, that there is no loss of load with time, and that the required
volume of rock to support the applied load is being mobilized by the anchor. This
behavior is evaluated by carrying out four types of tests:

a. Performance tests,
b. Proof tests,
c. Creep tests,
d. Lift-off tests.

Performance tests are a detailed examination of the load-extension behavior. They are
carried out on the first two to five anchors, and on a minimum 2% of the remaining
anchors

Proof and creep tests are carried out on all remaining anchors.

For any anchor that fails either test, an additional two performance tests are performed,
and the failed anchor is either improved to meet the required load or is replaced.

A lift-off test is performed on every anchor.

Performance tests

The performance test comprises a cyclic loading procedure in the following sequence
with an extension measurement being made at each increment:

AL, 0.25P, AL;

0.25P, 0.5P, AL;
0.25P, 0.5P, 0.75P, AL;
0.25P, 0.5P, 0.75P, P, AL;
0.25P, 0.5P, 0.75P, P, 1.2P, AL;
0.25P, 0.5P, 0.75P, P, 1.2P, 1.33P – hold for creep test, AL;
P – lock off

AL – alignment load required to take out slack out of the system,

P – design working load (lock-off)

Application of an overload is an important part of checking that the anchor has capacity
in excess of the design load. However, the value of the maximum load must be
compared with the yield strength of the steel to ensure that it does not exceed 80% of
the elastic limit of the steel.
Proof tests

The loading sequence for a proof test is as follows:

0.25P, 0.5P, 0.75P, P, 1.2P, 1.33P – hold for creep test, AL (optional), P – lock off
The loading sequence is equivalent to the last cycle of the performance test.

Creep tests

At the applied maximum load of 1.33P in both the performance and proof tests, creep
extension readings are taken at intervals of 1,2 ,3 4, 5, 6 and 10 minutes. If the total
creep movement between 1 and 10 minutes exceeds 1 mm, the test should be
maintained for an additional 50 minutes and the extension recorded at 20, 30, 40, 50
and 60 minutes.

Lift-off tests

When the anchor has been locked off at the design load, the jack pressure is reduced
to zero and then reapplied to determine the load at which the nut or wedges are lifted
off the bearing plate.

Acceptance criteria

An anchor is acceptable if the results of the performance, proof, creep and lift-off tests
meet the following criteria for permanent anchors:

1. Total elastic extension measured in performance or proof tests should exceed

80% of theoretical elongation of the free stressing length, and be less than the
theoretical elastic elongation of the free stressing length plus 50% of the bond
length.
2. Creep extension should not exceed 1 mm during the period of 1-10 minutes. If
this value is exceeded, then the total creep extension within the period 6-60
minutes should not exceed 2 mm.
3. The lift-off load should be within 5% of the specified transfer load.

First criterion ensures that the rock mass between the head and the bond length is
mobilized by the applied load, and that the major portion of the bond stress is
developed in the top half of the anchorage.

In general, creep in rock anchors is small. If creep exceeds the limit in criterion 2, the
anchor is inacceptable because this is an indication of failure of the bond rather than
the creep of the surrounding rock.

Refences

“Foundations on Rock”, Duncan C. Wyllie, 2nd ed.