Observational Methods and NATM
Because prediction of geotechnical behaviour is often difficult, it is sometimes appropriate
to adopt the approach known as the ‘observational method’, in which the design is
reviewed during construction.1
The principles of the Observational Method were first described in detail by R.B. Peck
(1969). Peck’s method provides a basis for understanding of the requirements, although
when reviewing the method later Peck felt that his efforts to formalize it were too
contrived and rigid. Observational methods have in fact always been used to a degree in
applied soil mechanics.
The Observational Method has a specific meaning. Peck set forward the following
(a) Exploration sufficient to establish at least the general nature, pattern and properties
of the deposits, but not necessarily in detail
(b) The assessment of the most probable conditions and the most unfavourable
conceivable deviations from these conditions, in this assessment geology often play
a major role
(c) The establishment of the design based on a working hypothesis of behaviour
anticipated under the most probably conditions
(d) The selection of quantities to be observed as construction proceeds and the
calculation of their anticipated values on the basis of the working hypothesis
(e) The calculation of values of the same quantities under the most unfavourable
conditions compatible with the available data concerning the subsurface conditions
(f) The selection in advance of a course of action or modification of design for every
foreseeable significant deviation of the observational findings from those predicted
on the basis of the working hypothesis
(g) The measurement of quantities to be observed and the evaluation of actual
(h) The modification of design to suit actual conditions
The method is inapplicable where there is no possibility to alter the design during
construction. The ability to modify the design is appropriate if the method is to be applied
only during construction and the focus is on the temporary conditions. However, there are
situations where the method could be applied after construction, e.g long-term monitoring
of dams and buildings.
Eurocode 7 – BSI 1995
Chapter 8 166
Peck emphasises the importance of asking the critical questions. These must ensure that the
observations are appropriate and meaningful. The key is to combine comprehensiveness
with reliability, repeatability and simplicity. Observations are often far more elaborate and
costly than necessary.
The Base Design
The base design developed in (c) will typically be based on analysis, such as finite
element. However, analysis cannot replace judgement. Possible modes of failure –
particularly those of a sudden or brittle nature, or those who could lead to progressive
collapse – must be assessed carefully. It is a fundamental element of the Observational
Method to overcome the limitations of analysis by addressing actual conditions. The
design in (c) may therefore present difficulties associated with the term ‘most probably’,
and in practice (c) has been interpreted as ‘unlikely to be exceeded’. Some margin of
conservatism is always necessary; it may therefore be more appropriate base the design on
a ‘moderately conservative’ approach. A moderately conservative design would be less
conservative than a conventional design, but more conservative than one based on Pecks
‘most probable’, so that modifications to the original design become exceptional, not the
Feedback from Observations
Feedback and assessment from observations must be timely in order to confirm predictions
or to provide adequate warning of any undue trends in ground movements or loadings.
There must be sufficient time to enable planned contingency measures to be implemented
effectively. This emphasises a further aspect of the Observational Method. Measurements
of quantities must occur at the required times during a construction sequence. It may be
necessary to interrupt construction progress and may even influence the way construction
Other Observational Approaches
As set out by Peck, the procedures (a) – (h) for the Observational Method may be
unnecessarily cumbersome and often impossible to achieve. Further, the ‘most probable’
condition in (c) is very difficult to find in a statistically reliable manner. Simpler versions
of an observational approach have been suggested, as e.g. by Muir Wood.
Management of observational approaches are often described in flowcharts, often
including risk levels and responses.
Chapter 8 167
Figure 8.1: System for Observational approach to tunnel design
Eurocode 7 (EC7) includes the following remarks concerning an observational method.
Four requirements shall all be made before construction is started:
1. The limits of behaviour, which are acceptable, shall be established.
2. The range of behaviour shall be assessed and it shall be shown that there is an
acceptable probability that the actual behaviour will be within the acceptable
3. A plan of monitoring shall be devised which will reveal whether the actual
behaviour lies within the acceptable limits. The monitoring shall make this clear
at a sufficient early stage; and with sufficiently short intervals to allow
contingency actions to be undertaken successfully. The response time on the
instruments and the procedures for analysing the results shall be sufficiently
rapid in relation to the possible evolution of the system.
4. A plan of contingency actions shall be devised which may be adopted if the
monitoring reveals behaviour outside acceptable limits.
During construction the monitoring shall be carried out as planned and additional or
replacement monitoring shall be undertaken if this becomes necessary. The results of the
monitoring shall be assessed at appropriate stages and the planned contingency actions
shall be put in operation if this becomes necessary.
Chapter 8 168
Figure 8.2: Management review process for in-tunnel monitoring (CIRIA 1997)
The organisational procedure for a project using an observational approach is illustrated in
Figure 8.2. There must always be emphasis on the time element to enable supplementary
measures (contingency actions) to be put in place, including administrative actions.
Although primarily a tool for the geotechnical area, the Observational Method is not
limited to tunnel construction. For example, during construction of the Copenhagen Metro,
the method was applied for a yes/no decision for constructing a temporary waler beam at a
metro station box. The base design was not to build the waler, and the observed parameters
were deflection of the station walls, settlement of above buildings, etc.
NATM – New Austrian Tunnelling Method
One of the most well known methods using some elements of an observational approach it
the New Austrian Tunnelling Method, or NATM. The method, which is in fact a broader
concept of geotechnical engineering than a single ‘method’, has often been mentioned as a
‘value engineered’ version of tunnelling due to its use of light, informal support. It has
long been understood that the ground, if allowed to deform slightly, is capable of
contributing to its own support. NATM, with its use of modern means of monitoring and
surface stabilisation, such as shotcrete and rock bolts, utilizes this effect systematically.
Chapter 8 169
Traditional tunnelling used first timber supports and later on steel arch supports in order to
stabilise a tunnel temporarily until the final support was installed. The final support was
masonry or a concrete arch. Rock loads developed due to disintegration and detrimental
loosening of the surrounding rock and loosened rock exerted loads onto the support due to
the weight of a loosened rock bulb (described by Komerell, Terzaghi and others).
Detrimental loosening was caused by the available excavation techniques, the support
means and the long period required to complete a tunnel section with many sequential
intermediate construction stages. The result was very irregular heavy loading resulting in
thick lining arches occupying a considerable percentage of the tunnel cross-section (in the
early trans-Alpine tunnels the permanent structure may occupy as much as 40% of the
In the first part of the 20th century tunnellers and scientists at that time understood the
necessity to reduce deformations in order to utilise the carrying capacity of the rock mass,
and the reciprocal relationship between support resistance and deformations.
The New Austrian Tunneling Method (NATM) grew out of experience with the old
methods. In his book "Gebirgsdruck und Tunnelbau" 1944 Prof. L.v. Rabcewicz published
a systematic of rock pressure phenomena and their interpretation. In his Patent of 1948 the
basic principles of the concept were formulated. The essence was as follows:
With a flexible primary support a new equilibrium shall be reached. This shall
be controlled by in-situ deformation measurements. After this new equilibrium is
reached an inner arch shall be built. In specific cases the inner arch can be
From 1956 to 1958 Rabcewicz built the first large size tunnels in Venezuela according to
these principles. In Austria the first attempts were made in the fifties with smaller hydro
tunnels. In 1963 the "New Austrian Tunneling Method" was introduced at the
Geomechanics Colloquy in Salzburg. The method has been further developed with regard
to sequences of excavation works and supports. Support means and auxiliary measures
have been improved, instrumentation, interpretations techniques and skills have extended
the applicability of NATM to heavily squeezing ground with extreme deformations, to soft
ground in built up areas and to large sections or complicated geometrical tunnel
The New Austrian Tunnelling Method constitutes a design where the surrounding rock- or
soil formations of a tunnel are integrated into an overall ring like support structure. Thus
the formations will themselves be part of this support structure. The definition together
with the main principles was published in 1980.
With the excavation of a tunnel the primary stress field in the rock mass is changed into a
more unfavourable secondary stress field. Under the rock arch we understand those zones
around a tunnel where most of the time dependent stress rearrangement processes takes
place. This includes the plastic as well as the elastic behaving zone.
Chapter 8 170
Under the activation of a rock arch we understand our activities to maintain or to improve
the carrying capacity of the rock mass, to utilise this carrying capacity and to influence a
favourable development of the secondary stress field.
The main principles of NATM are:
The main load-bearing component of the tunnel is the surrounding rock mass.
Support is ‘informal’ i.e. it consists of earth/rock-anchors and shotcrete, but support
and final lining have confining function only.
Maintain strength of the rock mass and avoid detrimental loosening by careful
excavation and by immediate application of support and strengthening means.
Shotcrete and rock bolts applied close to the excavation face help to maintain the
integrity of the rock mass.
Rounded tunnel shape: avoid stress concentrations in corners where progressive
failure mechanisms start.
Flexible thin lining: The primary support shall be thin-walled in order to minimise
bending moments and to facilitate the stress rearrangement process without
exposing the lining to unfavourable sectional forces. Additional support
requirement shall not be added by increasing lining thickness but by bolting. The
lining shall be in full contact with the exposed rock. Shotcrete fulfils this
Statically the tunnel is considered as a thick-walled tube consisting of the rock and
lining. The closing of the ring is therefore important, i.e. the total periphery
including the invert must be applied with shotcrete.
In situ measurements: Observation of tunnel behaviour during construction is an
integral part of NATM. With the monitoring and interpretation of deformations,
strains and stresses it is possible to optimise working procedures and support
The concept of NATM is to control deformations and stress rearrangement process in order
to obtain a required safety level. Requirements differ depending on the type of project in a
subway project in built up areas stability and settlements may be decisive, in other tunnels
stability only may be observed. The NATM method is universal, but particularly suitable
for irregular shapes. It can therefore be applied for underground transitions where a TBM
tunnel must have another shape or diameter.
Observations of tunnel behaviour
One of the most important factors in the successful application of observational methods
like NATM is the observation of tunnel behaviour during construction. Monitoring and
interpretation of deformations, strains and stresses are important to optimise working
procedures and support requirements, which vary from one project to the other. In-situ
observation is therefore essential, in order to keep the possible failures under control.
Considerable information related to the use of instruments in monitoring soils and rocks
are available from instrument manufacturers. Figure 8.3 shows an example instrumentation
in a tunnel lined with shotcrete.
Chapter 8 171
Legend Measuring objective Instrument
1 Deformation of the Convergence tape
excavated tunnel surface Surveying marks
2 Deformation of the ground Extensometer
surrounding the tunnel
3 Monitoring of ground Total anchor force
support element ‘anchor’
4 Monitoring of ground Pressure cells
support element ‘shotcrete Embedments gauge
Figure 8.3: Examples of NATM tunnel measurement equipment
Measurements of tunnel behaviour can be automatic, but it is outside the scope of these
notes to go into detail. Reference is made to instrumentation manufacturers catalogues.
A variant of NATM using steel-fiber reinforced shotcrete instead of mesh-reinforced
concrete is referred to as the Norwegian Method of Tunnelling (NMT). The method is
most suitable in jointed rock which tends to overreach. By 1984 NMT had replaced NATM
in such ground conditions in Sweden and Norway. NMT is often used with drill and blast
tunnelling, but can be used in conjunction with TBMs in clay zones.
Chapter 8 172
NATM Process on site
The simplified steps of an underground transition created with NATM are shown below.
Cutting a length of tunnel,
here with a roadheader
Applying layer of shotcrete on
Primary lining applied to whole
cavity, which remains under
Chapter 8 173
Final lining applied. Running tunnels
Completed underground transition
Tunnel stresses and failure mechanisms
The following main failure mechanisms are observed:
These mechanisms develop from the roof in case of loosening of the rock mass or lack of
horizontal stress in the roof.
Split tensile and buckling failure - develops near the excavation line in straight side
walls or inverts
Shear failure - develops in high pressure exerting rock if lateral confinement is
These failures are generally of progressive nature and can be kept under control with the
information of careful in situ observation. Other problem areas may be rock burst in brittle
behaving rock under high uniaxial stress or swelling rock like clays or anhydrite. Rock
sample testing should be carried out under laboratory conditions as well as in situ
examinations. The gained values of the rock-mechanical properties, their variability in
particular long-term changes and also the effects of water inflow must be taken into
Chapter 8 174
Stage 1 Stage 2 Stage 3
Figure 8.4: Sketch of mechanical process and sequence of failure around a cavity by stress
Figure 8.5: Schematic representation of stresses around a circular cavity with hydrostatic
The Fenner-Pacher curve
The Fenner-Pacher curve shows the relationship between the deformation ΔR/R and
required support resistance Pi. Simplistically, the more deformation is allowed, the less
resistance is needed. In practice, the support resistance reaches a minimum at a certain
radial deformation, and support requirements increase if deformations become excessive.
Chapter 8 175
Fenner-Pacher-type diagrams can be generated to help evaluate the support methods best
suited to the conditions. In terms of analysis, it is convenient to carry out quick and simple
convergence-confinement calculations with and without support. Resistances of shotcrete,
steel reinforcement, and anchors/bolts can be calculated. Functions of support elements,
radial stresses, support resistances of inner and outer arches, deformations, and failures can
be analysed with respect to time.
Figure 8.6: The Fenner-Pacher Curve.
A classical approach is that of a hole in a homogenous uniformly stressed solid which
behaves linearly elastically up to a certain stage of stress and perfectly plastically
thereafter. The radial and tangential stresses assumptions around the opened cavity, and the
plastic zone can be seen in figure 8.7.
r=radius of cavity
R=radius of plastic zone
0=primary stress condition
r0 and t0=radial and tangential stresses
respectively with r=0
r1 and t1=radial and tangential stresses
respectively for r-r.
Figure 8.7: Schematic representation of stresses around a cavity.
Chapter 8 176
Skin resistance which counteracts the radial stresses forming around the cavity, becomes
smaller in time, and the radius of the cavity decreases simultaneously. These relations are
given by the equations of Fenner-Talobre and Kastner.
P = -c Cotg + [c Cotg + P (1 - Sin ) ]( r / R) 1 Sin
Pi = skin resistance
C = cohesion
= angle of internal friction
R = radius of the protective zone
r = radius of the cavity
P0 = H; overburden
Following the main principle of NATM, the protective ring around the cavity (R-r), is a
load carrying part of the structure. The carrying capacity of the rock arch is formulated as;
S R Cos S nR Sin
Pi = R
PiR = resistance of rock arch (t/m2)
S = length of shear plane (m)
R = shear strength of rock (t/m2)
= angle of internal friction (0)
b = height of shear zone (m)
nR = normal stress on shear plane (t/m2)
, R, nR, can be measured in laboratories, where as S can be measured in meters , on a
drawing made to scale.
Figure 8.8: Skin resistance Pi required to establish equilibrium of a cavity as a function of
Ø angle of internal friction and P0 = H.
Chapter 8 177
Principles of dimensioning the supporting system
Generally two separate supports are carried out. The first is a flexible outer arch or
protective support designed to stabilize the structure accordingly. It consists of a
systematically anchored rock arch with surface protection, possibly reinforced by ribs and
closed by an invert.
The behaviour of the protective support and the surrounding rock during the readjustment
process can be monitored by a measuring system.
The second means of support is an inner concrete arch, generally not carried out before the
outer arch has reached equilibrium. In addition to acting as a final, functional lining (for
installation of tunnel equipment etc.) its aim is to establish or increase the safety factors as
In order to plan a project and design standard sections for the documents it is necessary to
establish the required carrying capacity of the support for different types of rock.
The carrying capacity of the outer arch can be decided by the r/r curve, which is
characteristic for any type of rock and primary stress condition.
Figure 8.9: Schematic representation of ther, r1, r/R1 and T1 for supports of different
yield 1 and 2, and time of application. Key: r0=radial stress for r/R=1;
r=required radial stress as a function of r; pia and pi1=support resistances of
outer and inner arches respectively; s=safety factor; and C and C’=loaded and
unloaded condition of the inner arches respectively.
Chapter 8 178
The required radial stress pia to obtain equilibrium decreases if the border zone is allowed
to yield and a plastic zone develops simultaneously. The rate of the decrease being mainly
a function of the primary stress condition 0 and the angle of the internal friction of the
rock as a rule diminishes rapidly. At any intersection between pi and the r curve,
equilibrium is reached for the respective support resistance.
It is a particular feature of NATM that the intersections always take place at the
descending branch of the curve. For instance, should the support partially fail for any
reason, a new equilibrium comes into being without any additional strengthening at a lower
point of intersection, as long as this lower point does not fall below the minimum of the r
curve (marked B on figure 9) where the detrimental loosening starts.
With conventional methods on the other hand, the intersection point is usually situated at
the ascending branch of the r curve. With any failure, the intersection point moves to the
right and the supporting structure has to be strengthened above its former carrying
Loosening is considered detrimental, with open cracks and fissures appear in such as way
that the rock is no longer capable of conveying shear and compressive stresses. The weight
of the loosened masses is added to the lining, actually causing the free area of the cavity to
To be able to plot the r/r curve, the following parameters have to be established: the
primary stress condition 0 with the direction of principal stresses, the angle of internal
friction , the uniaxial compressive strength gd parallel and normal to stratification, and
the corresponding modulus of deformation and elasticity.
These parameters can be determined by measuring and the course of the curve computed
by the finite-element method, taking into consideration the method of excavation (full-face
driving or subdivision of the section).
While, after Kastner, the r0 for r = 0 is theoretically given by the equations:
2 0 gd 1 sin 2c
1 1 sin tan(45 / 2)
(here c = cohesion and φ = angle of internal friction)
Establishing of rmin is influenced by the magnitude of r0 on one hand and geological
conditions on the other. This can be explained by the following example.
With a road tunnel situated in fairly compact rock with a small overburden, the tangential
border stresses only slightly exceeding its uniaxial compressive strength, Pi min will be very
small, particularly if the rock has in addition a high standing capacity.
The same type of rock under a large overburden is bound to develop a fairly large plastic
zone causing significant deformations. The rock in this case becoming fractures to a depth
of several meters, requires a far greater Pi min, the more so it should it be crossed by a
Chapter 8 179
system of even joints instead of being interlocked, due to e.g. well-defined interlocking of
The value of the required carrying capacity of the outer arch Pia must be chosen as to
combine maximum economy with and acceptable degree of safety, and Pia should therefore
be as close as practically possible to Pi min in order to obtain a sufficient factor of safety
from the additional lining resistance Pi1 of the inner arch.
Should a stiffer type of support be chosen for the outer arch (as marked ‘2’ in figure 8.9 for
example), the intersection with the r curve is bound to rise, while the safety factor
The minimum carrying capacity of the inner arch is decided by the smallest lining
thickness that will allow suitable compaction of the concrete. Should a greater Pi1 be
required, the thickness can be chosen according to Pia and the required safety factor s.
Once the carrying capacity of the outer arch has been established for certain standard
sections, the means of strengthening can be chosen and computed accordingly. The
computation has been slightly altered in the mean time and it is therefore shown again in
figure 8.11. The resistance of the lining material (shotcrete) is:
sin s (b / 2)
An additional reinforcement (steel ribs, etc.) gives a resistance of:
F st st
sin s (b / 2)
s E st
15 s (for concrete)
The lining resistance is:
PiL = Pis + Pist
The anchors are acting with a radial pressure:
f st p
With the lateral pressure given by:
3 = pis + pist + piA
and with Mohr’s envelope, the shear resistance of the rock mass R and the shear angle is
determined, assuming that the principal stresses are parallel and at right angles to the
The carrying capacity of the rock arch is given by:
Chapter 8 180
S R cos S n sin
The resistance of the anchors against the movement of the shear body towards the cavity
af st p cos
et (b / 2)
The total carrying capacity of the outer arch is then:
Pi w Pi L Pi R Pi A Pi m in
Figure 8.10: Shear forces
Figure 8.11: Design scheme of arch for a given carrying capacity
Chapter 8 181
Symbol Unit Description
Pi Mp/m² Resistance of shotcrete
Pi Mp/m² Resistance of steel
Pi Mp/m² Resistance of rock arch
Pi Mp/m² Resistance of anchors
Pi Mp/m² Total support capacity
b M Height of shear zone
d M Thickness of lining
e,t M Distance between rock bolts
s M Length of shear plane
w M Width of carrying ring
gs Mp/m² Uniaxial compressive strength of rock
º Angle of internal friction
c Mp/m² Cohesion of rock
Mp/m² Shear resistance of lining material
Mp/m² Proportaion of shear resistance of reinforcement
E ,E Mp/m² Modulus of elasticity of reinforcement and lining material,
s º Shear angle of lining material
Fst cm² Area of rock bolts
pst Mp/m² Proportional limit of anchor steel
R Mp/m² Shear strength of rock
nR Mp/m² Normal strength of rock
º Shear angle of rock
º Average inclination of shear plane
º Inclination of anchors
As to the reciprocal mode of action of the basic supporting members of the NATM,
shotcrete and the anchored rock-arch, experience shows the following:
1. With the same type of rock and overburden the relationship between the size of the
joint-bodies and the excavation area is decisive for the mobility of the material;
2. With small sections (ie -10-16 cm²) and joint bodies of a few dm³, a simple
shotcrete sealing with d = 3 cm = 0.017 R usually stabilizes the tunnel;
3. With an underground power station of 400-600 m2 on the other hand, a rock with
joint-bodies of this size behaves like a cohesionless mass, and a simple shotcrete
lining of 0.017 R = 19-24 cm would never do. A systematically anchored rock arch
is imperative in this case.
The surrounding ground acts as the main carrying member, the shotcrete lining merely
having the function of stabilizing the surface between the anchorage points.
Chapter 8 182
The greater is the r and the section of the cavity and, the smaller is , the more important
is the system anchoring in comparison with the shotcrete.
With conventional wedge or expansion bolts the plate exerts the supporting action, and the
anchor is always stressed equally over its whole length. Grouted anchors have their main
carrying effect from the bond between the grout and rock. The bond consists mostly of
friction caused by the tangential stresses in the surrounding rock (besides a minor share of
The tensile stress of the anchor increases from zero at the end to a maximum at the plate,
and any radial border stresses possibly remaining are additionally conveyed to the anchor
by the plate Fig. 8.12
The movement of the rock towards the cavity is inhibited in this way, and an arch effect is
creates between neighbouring anchors, as is shown in the drawing.
The carrying capacity piB can be described analytically in simplified form by the equation:
piB = ld(a+tan mt) +Frfee
The term Fr can possibly rise to r.
Although the carrying capacity of both the expansion and the grouted type anchors is the
same (limited by the tensile strength of the steel), the stabilizing effect of grouted anchors
is much greater.
As a further reinforcing measure in NATM, light steel ribs of the channel-section type are
used, connected by overlapping joints and fastened to the rock by the anchors.
Chapter 8 183
The ribs serve primarily as a
protection for the tunnelling crew
against rock fall and as local
reinforcement to bridge across
zones of geological weakness.
The static share of the ribs in the
lining resistance is relatively low.
The stiffness of the ribs contrasts
with the relatively high yielding
capacity of the shotcrete, and
with large sections and
deformations minor cracks in the
shotcrete along the ribs must be
Final dimensioning is based on
Inseparably connected with the
NATM, and a basic feature of the
method, is a sophisticated
Deformations and stresses are
allowing determination of
whether the chosen support-
resistance corresponds with the
type of rock in question, and
what kind of additional
reinforcing measures are needed,
In a case of the lining being over-
dimensioned, the reinforcing
measures can straight away be
reduced accordingly when the
same or similar mechanical
conditions of the rock are
encountered during further tunnel
Figure 8.12: Schematic showing the mode of action of the grouted anchors.Key: ri =
radial border stress, ti= tangential border stress, t = tangential stress, r =
radial stress, 0 = H, tm = average tangential stress on l, a = adhesion of
grout/rock, PiB= lining resistance of anchor, fe = area of anchor steel, e =
tensional strength of steel, F = area of plate, d = grouted diameter hole.
An empirical dimensioning is carried out in this way, based on the principles explained
Chapter 8 184
During the execution of a series of NATM tunnelling works during the last many years
satisfactory measuring systems have been developed.
In order to control the behaviour of the outer arch and surrounding the different
construction stages in practice, main measuring sections are chosen at distances determined
by the significant geological points.
These will be equipped with double extensometers and convergence measuring devices to
measure deformations and pressure pads to measure radial and tangential stresses. In
addition, roof and floor points can be monitored geodetically.
In-between the main measuring sections, secondary points are selected at suitable distances
where only convergence are made. Readings are made every other day at the beginning,
decreasing to once a month according to the velocity of deformation and change of
stresses. The measurement results are plotted in graphs as a function of time, which
enables the changes in the rock caused by mechanical
R1-R8 : Radial pressure pads
T1-T8 : Tangential pressure
H1, H2, H3: Convergence
E1-E6 : Long extensometers
Ea1-Ea6: Short extensometers
VF, Vs: Geodesic control points
Figure 8.13: Standard main measuring section
This method of establishing stress-time graphs gives a high degree of safety, allowing any
situation to be recognized long before it becomes critical.
Since the readjustment process takes a very long time, being possibly influenced locally by
subsequent alterations of the geological conditions (e.g. increase in the water content of the
surrounding rock), it is essential from both the practical and the theoretical point of view to
measure also the stresses and deformations of the inner lining
This is done by placing a series of tangential pressure pads or strain gauges, both in pairs
outside and inside the lining, and also by using convergence measuring devices.
Chapter 8 185
Figure 8.14: A drawing showing some of the results of measurement made at the Austrian
Tauern Tunnel North, showing the conspicuous decrease of deformation with
respect to time and the excavation and outer lining section. Key: H1-H3 =
convergence measurement readings, E4 and E5 = long extensometer readings.
Numerical example for NATM
Tunnel size: 12.10 x 12.00 m (fig.10)
H = 15.0 m overburden according to tests on samples found: = 27º, c = 100 t/m² (three
axial tests). For establishing a new equilibrium condition in the opened cavity, following
supporting is going to be used:
1. Use the supporting ring which develops around the cavity after excavation as a self-
supporting device and select a type of supporting which can bear the developed
rock loads and deformable when necessary.
2. Design the inner lining under final loads
Figure 8.15: Numerical Example Tunnel Section
Chapter 8 186
The (1) supporting system is capable of carrying safely the loads, the (2) lining is for safety
and to bear the additional loads which are probable to develop after the supports are
Supporting will consists of:
a: Shotcrete (15+10) cm in layers by two shots
b: Bolts spaced 2.00 x 2.00 m in rings with diameter Ø 26 mm.
c: Rib steel channel supports (2 x 14)
d: by ground supporting ring
To find the radius of the disturbed zone R:
Pi c cot g c cot g P0 (1 sin ) ( ) 1sin
Values entered into the formula:
= 2.5 t/m³
H = 15.0 m
P0 = H = 2.5 x 15.0 = 37.5 t/m²
C = 100 t/m²
2 sin 27
6.05 1sin 27
Pi 100 cot g 27 100 cot 27 37 .5 (1 sin 27 ) (
2 x 0.450
Pi 100 1.96 100 1.96 37 .5 (1 sin 27 ) (
196.3 217.08 0
R = 6.45 m
d = 25 cm
c28 = 160 kg/cm² compressive strength shear in concrete (assume 20% of c28)
the capacity carrying load
sh = 0.20 · 160 = 32 kg/cm² = 320 t/m²
d = thickness of shotcrete in (cm)
= /4 - /2 angle of shear plane with vertical
Chapter 8 187
b = shear failure height of the cavity (see Fig.10)
pic ; sin 31.5 = 0.520; b/2= r cos = 6.05xcos31.5=5.15
pic 29.9 t / m 2
pic 29 .9 t / m 2
Bolt Ø 26 mm
St III, sh = 4000 kg/cm²
f = 5.3cm²
pib 0.53 kg / cm 2
pib 5.3 t / m 2
c) steel ribs
t = spacing 2.00 m
F = 2 · 20.4 cm² = 40.8 cm² = 0.00408 m²
st = c · 15 = 320 · 15 = 4800 t/m²
F · st 0.00408 4800
Pi st 36.56 t / m²
b 0.52 5.15 2.00
Pc 36 .56 t / m²
N = normal stress = 142
read on X
Z = minor stress = 37.2
R = shear stress = 170 t/m² read on Y
Chapter 8 188
Figure 8.16: Finding stresses geometrically
Connect AB and find the centre (W) draw the circle. Tangent at the point B (BB’) so
OB’ = Cohesion = 100 t/m²
= internal friction angle (27º)
R calculated (Talobre formula) as 6.45 m. Width of the protective ring 6.45-6.05 = 0.4m
drawn through A, B and the intersection bisecting with the middle ring (C); ABC shear
failure line drawn and thus (S) measured.
Bolt length l = 4.00 m is taken and inclination measured
Pi = pic + pib + pist = 29.9 + 5.3 + 36.56 = 70.26 t/m²
P 70.26 t / m²
d) bearing capacity of the supporting ring
Sin = sin 27 = 0.450
Cos = cos 27 = 0.891
S R cos S N sin
thus S = 4.54 m (from figure 17)
b/2 = 5.15 m
R = 170 t/m²
N = 142 t/m²
Chapter 8 189
enter the formula
4.6 170 0.891 4.6 142 0.450
pC 78 .22 t / m 2
e) The resistance of the bolts (anchors) against the movement of the shear body towards
the cavity is:
a f sh cos
et b/ 2
(b/2, a = 4.20 m, = 35.5º from Fig 17)
4.20 5.20 4000 0.81
pib 3.43t / m 2 < 5.3 t/m²
2.00 2.00 5.15
shear = 4000 kg/cm² ; e.t1 = bolts arrangement = 2.00 · 2.00 m
So the total bearing capacity of supporting will be
Pi = Pic + Pib + Pist + PiR = 29.9 + 3.43 + 36.56 + 78.22 = 160.1 t/m²
Chapter 8 190
P 148.11 t/m²
Failed implementation and criticism of NATM
It may be useful for the student to be aware of some of the controversy that has surrounded
The name itself has been used to cover a variety of different approaches to tunnelling, at
times surrounded with too much ‘philosophical baggage’. The method is quite practical,
however. It is a universal principle of good tunnelling to adapt the support to the expected
behaviour of the ground. As regards the ‘ring-like support structures’, any opening in the
ground causes at least a degree of ground support to be conferred by circumferential stress
on the ground.
More importantly, since the beginning of NATM engineers and geologists doubted the
existence of sufficient bearing capacity of the ground. In principle at least, the criticism
was unfounded, but there were indeed a dozen accidents within the first thirty years use of
the method. Also in more recent years, a number of NATM tunnels collapsed, and debates
on the method are still continuing. A well-known NATM tunnel collapse was the
Heathrow Express Tunnel in October 1994, causing extensive media coverage. There were
allegations that the method was unsuitable for tunnelling in the London Clay; allegations
that was deemed groundless by NATM proponents the success of the method in many
other clay conditions all over the world. Two collapses in Germany, the Munich Metro and
Krieberg Tunnels received similar media attention, and in Turkey the Bolu tunnel
experienced massive problems.
The collapse of the Heathrow Express tunnel in 1994 caused a symposium by the Health
and Safety Executive (HSE) in the UK. The HSE was at that time aware of totally 116
NATM related collapses.
When analysing the background of these collapses, the most important fact is that the large
majority of NATM tunnel collapses have occurred during construction. Moreover, the
principle collapse in NATM is the face failure, i.e. collapses have occurred only at the face
where the lining is still weak and cantilevered. Completed, correctly constructed NATM
linings have almost never failed.
In literature, and in the NATM debate in the tunnelling society, one particular item is
addressed as very important: skill and experience of the site-engineering team. This view is
defended by the fact, that the majority of incidents have been associated with night shifts.
For instance, the 1994 Heathrow Express Tunnel collapse took place at 01.00h.
A contractor, however, need not be an expert in NATM tunnelling provided the correct
business relations are established to obtain expertise elsewhere. The design of NATM
tunnels relies heavily on the available data of the soil conditions. During the discussions it
has been claimed, that designers of NATM tunnels have often had insufficient data to work
with, and in fact most NATM collapses are connected with 'unexpected ground conditions'.
NATM is a safe method if properly applied. Proper application requires:
Theory and calculations
Experience of engineers
Good labour skills
Chapter 8 191
Chapter 8 192