Structural Design of Reinforced Concrete Deep Shafts

Inserted Into Fractured Rock Masses

Miguel Luzia Parrinha

Extended Abstract

October 2015
 Structural Design of Reinforced Concrete Deep Shafts Inserted Into
Fractured Rock Masses

Miguel Luzia Parrinha

Instituto Superior Tcnico, University of Lisbon, Portugal

Abstract: In the design of deep shafts inserted into fractured rock masses it is important to know the
behavior and distribution of the vertical loads between the concrete structure and the rock mass. With
the purpose to know this load distribution, this work develops an analytic model for the load distribution
between the concrete and the rock mass, and the respective study of this analytic model, comparing it to
a finite element method (FEM) model. The results reported by the analytic model provide a tool for
estimating the forces in concrete for the structural design of a deep shaft.

Key-Words: Vertical load distribution, structural design of deep shafts, fractured rock masses, reinforced
concrete.

1. Introduction There are various types of lining for shaft

structure, such as, reinforced concrete lining
The main goal of this work is to develop a
and steel lining. The use of reinforced concrete
theoretical/analytical model to obtain the load
lining can bring several advantages such as the
distribution between the rock mass and the
possibility of reinforcing week rock regions,
reinforced concrete structure, in the design of
water leakage control along the shaft,
deep shafts of circular cross-section, with a
prevention of rock erosion, prevention of
reinforced concrete lining, inserted in rocky
instability of the rock mass and the economic
environment (fractured rock mass), such as
advantage of being less expensive than a steel
shafts created in some hydropower plants
lining (ASCE, 1989).
projects. It is important to identify the load
distribution between the two materials given 2. Deep Shaft Excavation

that the depth of these shafts can reach 300

The excavation of deep shafts with large
meters and higher, and taking into account also
diameters, embedded in rock masses is usually
the irregular contact surfaces between the
performed through one of following methods:
reinforced concrete structure and the surface of
drilling and blasting method, Alimak method,
the rock mass.
and rotary drilling with Raise Boring Machine or

Shafts are structures that can be used for Shaft Boring Machine.

multiple purposes, such as access to high depth

The drilling and blasting method consists of two
levels, collect underground resources or
main phases: first, the excavation by use of
samples, conduct surveys, serve to support
percussion or rotation drilling methods,
other large underground works and also have
followed by blasting with explosives arranged in
ventilation functions.
the excavation already performed (Grieves,

1
1996). This second phase cause an irregular
surface on the rock wall (presenting wedges
throughout the depth of the shaft) and
fracturing of the rock in the shaft surroundings
by the violent vibrations and gas releases
provoked by the blasting (Fig. 1) (Hartmann,
1992).

Figure 2 - Excavation with Raise Boring Machine (adapted

from Terratech Copmanys Brochure 2011)

3. Analytical Model to Obtain the Load

Distribution

The load sharing is possible by means of wedges

Figure 1 - Illustration of the explosives effects on the shaft
surroundings (adapted from Silva, 2005)
Raise Boring Machine (Fig. 2) and Shaft Boring
Machine are automated machines
performing excavation up to 1260 meters depth
and a range of diameters from 0.7 to 8.5 meters.
The first one bores a pilot hole, breaking through
the rock from the surface to the level below, the
pilot bit is then replaced by a reamer head, with
the diameter planned for the shaft, then the
machine pulls the reamer head upward, while
rotating, to break a circular hole in rock
(Ozdemir ,1986). The second one is for shafts
excavation exclusive, and unlike the previous,
performs the excavation from the top down, but
also with the use of rotary drilling.
existing in the rock wall (Fig. 3), that result from
use of blasting excavation method, which have
the capacity to bear part of the loads acting on
the concrete. Together with the rock mass, the
wedges show a spring type of behavior and have
a certain stiffness, and it is through this stiffness
that is to be determined which load absorbed
for
either by the rock mass either by concrete. In
order to calculate this stiffness is first necessary
to calculate all the associated displacements.
As simplifying assumptions it is assumed that the
wedges are similar to each other along the shaft,
with a meters width and a angle with the
vertical axis, and that the rock mass
characteristics do not vary with depth.

2
 Supposing that the inner radium of the shaft is
and the concrete lining width is , the vertical
force is given by (4).

= (4)
2 ( + + )

From equation (2) it is possible to obtain the
perpendicular force to the surface of the wedge
(5).

Figure 3 - Contact surface of the representation of rock = (5)
mass and concrete ring where the wedges (sin + cos )

Replacing (5) in equation (1) it is possible to

3.1 Equilibrium at Slide Situation
achieve the horizontal force on the wedge (6).
To determine the displacements is first
(cos sin )
necessary to know the stresses and forces = (6)
(sin + cos )
actuating in a single isolated wedge, when it is
At last from (3) it is possible to obtain the
submitted to a force (Fig. 4).
distance between the application point of
and point A (7).

sin + cos cos2

=( +
2 sin
(7)
cos
)
2
It is also important to define the horizontal
tension on the wedge surface (8).

cos sin tan

= ( ) (8)
sin + cos

Figure 4 Equilibrium of an isolated wedge 3.2 - Relative Displacement Between

Concrete and Rock
Using to the three fundamental equilibrium
equations:
In order to determine the vertical relative
= 0 <=> cos
(1) displacement between the concrete lining and
sin = 0
the rock mass is first needed to determine the
= 0 <=> sen + horizontal displacements of both.
(2)
cos = 0
In the case of the concrete horizontal

= 0 <=> + + displacement, it has the radial pressure

(3)
=0 actuating on all the perimeter of the lining (Fig.
2 tan

3
5), caused by the rock. So admitting thin-wall
hypothesis the displacement is given by (9).

( + )
= ( + ) (9)

Figure 6 Representation of the fractured rock ring and

actuating pressures

The non-fractured rock zone is actuated only by

the pressure ( ) exerted by the fractured rock
Figure 5 Representation of the concrete under pressure
ring (Fig. 7).
In the case of fractured rock region, the action
of an inner pressure ( ) applied by concrete,
and outside pressure ( ) applied by the non-
fractured rock zone (Fig. 6). Both these
pressures cause displacements on the inner and
outer faces of the fractured rock ring,
respectively, that are given by (10) and (11)
(Timoshenko, 1940).

=
1 ( + )2 2
=
2 ( + )2
1 + (10) Figure 7 - Representation of the non-fractured rock ring and
( + ) + actuating pressures

( + )2 2 ( ) Considering Timoshenko (1940) it is possible to

(2 ( + )2 ) ( + )
determine the displacement of the non-
fractured rock ring inner face (12).
1 =
101
1 ( + )2 2 1 = [ + ] (12)
= 99
2 ( + )2
To set the pressure ( ) (13) it is necessary to
1 + (11)
+ equalize (11) and (12).

( + )2 2 ( )
= (13)
(2 ( + )2 )

4
In which is given by (14). For the zone 1 the diameter of the tension
dispersion and the area that receives tensions,
2 ( + )2
= (14)
+ are given by (18) and (19), respectively.
In which and are given by (15) and (16),
() = 2 ( cot + + + ) (18)
respectively.

() = ( cot + + + )2
= (2 ( + )2 ) (19)
( + )2
101 (15) This makes it easy to determine the vertical
[ + ]
99
displacement of the zone 1 (20).

= [(1 ) 3 + (1 + ) 1 =
( + )2 ] (16) ln( cot + )
= [
2 cot ( + )

Finally the vertical relative displacement is given ln( cot + 2( + ) + ) (20)

2 cot ( + )
by (17).
ln ln(2( + ) + )
]
, = ( + ) cot 2 cot ( + )
(17)
Analyzing the tension dispersion in the zone 2,
with a depth of meters below the bottom of
3.3 Vertical Deformation of the Rock
the shaft, it is necessary to define the upper and
Mass
lower receiving tensions areas 1 and 2 ,

For the determination of the rock wall vertical respectively.

displacement it is considered that there is a

1 = ( cot + + + )2
(21)
conical dispersion of the actuating tensions (Fig. ( + )2
8). [() + 2 cot ]2
2 = (22)
4
So the vertical displacement of the zone 2 is
given by (23).

2
2 = ln (23)
(2 1 ) 1

3.4 Stiffness of the Wedge-Rock Set

Considering Hookes law (Timoshenko &

Goodier, 1951), the stiffness of one isolated
Figure 8 Representation of the tensions dispersion
wedge is given by (24).
It is necessary to distinguish two zones of the

rock in depth: the zone which lies between the = (24)
,
top and bottom of the shaft (zone 1); and the
zone immediately below the first, and that
supports the shaft (zone 2).

5
In which , is the addition of the
displacements previously determined (25).

, = ( + ) cot + 1 + 2 (25)

It is possible to eliminate the force out of the

equation (24) through the adoption of the
following simplifications of the displacements.

1 = (26)


2 = (27)

1
1 = (28)

Figure 9 Displacements field of the shaft and free body
2 diagram of a portion of the shaft
2 = (29)

Considering also the number of wedges by It is now conceivable the expression for the

meter vertically () it is obtain the stiffness of distributed force on the rock (32).

the wedge-rock set of the shaft (30).

() = ( ) (32)


= (30) Integrating the equation above, we get the
(1 + 2 ) cot + 1 + 2
expression for the total force on the rock for a
determined depth (33).
3.5 Axial Load Distribution

() =
At this point are defined the terms needed to 0
(33)
obtain the axial forces both in solid rock and in 2
( )
2
concrete ring. Assuming that on the structure
Replacing (33) in (31) it is obtained the
acts only a distributed load () and the
expression for the total force on the concrete
possibility that the displacement field varies
lining for a determined depth (34).
linearly along depth (Fig. 9), it is given the
equation (31) by equilibrium.
() = +

(34)
2
() + () = (31) ( )
0 2
It is also important to define the displacement
on the top of the shaft (35).

()
=
0
(35)
3 2
=
2 ( 2 + 3 )

6
In short, it is thus possible to predict what the
N Concrete (kN)
amount of load charges transferred to the rock N Rock (kN)
mass (33), and also the axial force in the N total (kN)
350
concrete ring (34), depending on the depth.
300
4. Analytic Model Sensitivity Analysis 250
200

z(m)
At this point are tested the results of the analytic
150
model through the application to a shaft with
the physic and geometrical characteristics 100

represented in tables 1, 2, 3 and 4. 50

0
Table 1 Characteristics of the concrete lining -10000 0 10000 20000 30000 40000 50000
L (m) 300 N (kN)
Rc (m) 2,95
e (m) 0,3
Figure 10 Load distribution in function of z
Ec (Mpa) 20000
q (KN/m) 10 It is observed that there is tension in the initial
part of the shaft and compression closer to the
bottom, this due to the fact of the consideration
Table 2 Characteristics of the rock mass
of the spring-type effect of wedges, while the
Rd (m) 15
rock mass is always under compression.
Er (Mpa) 10000
Erf (Mpa) 10000
In figure 11 are shown the displacements in
Erv (Mpa) 10000
function of z, which are liner as expected.
0,7
r 0,3
rf 0,25 Total Displacement
350
300
Table 3 Characteristics of the wedge 250
() 30 200
z (m)

t (m) 0,2 150

n (wedge/m) 0,1 100
av 3
50
0
Table 4 Values for the tension dispersion
0,00 0,20 0,40 0,60 0,80
Vertical Displacements (mm)
() 45
Lm (m) 100
Figure 11 Vertical displacements in function of z

Now using the values of the tables above and

equations (33) and (34) it is obtain the load
distribution (Fig. 10).

7
 5. Comparison of the Analytic Model shaft compared to the results supplied by the
with a Computational Model calculation program.

At this point, it proceeds to a shaft modeling

FEM Analytic Model
through the SAP2000 finite elements method 350

(FEM) software, using shell elements (Fig. 12) so 300

as to approximate the model to the geometric 250

characteristics of the shaft of reinforced 200

z (m)
concrete circular cross-section presented 150
before.
100

50

0
0 20 40 60 80
Vertical Displacements (mm)

Figure 13 - Vertical displacements in function of z (without

springs)

5.2 Modeling With Springs

In order to simulate now a shaft in which there

Figure 12 Circular cross-section modeled through SAP2000
are wedges on the rock mass along its length,
5.1 Modeling Without Springs
were added springs on the FEM model,
It begins then to analyze the model consists of simulating the stiffness of =
shell elements without any spring along its 385011,4 //, given by equation (30).
length, which corresponds to the situation that Here, the total load imposed in the shaft, shall
there are no wedges throughout the depth of be notionally allocated between the rock mass
the shaft ( = 0) in which total load is and concrete.
supported only by reinforced concrete.
It is possible to observe the load distribution
Regarding the displacements (Fig. 13) note that, comparison of the analytical model and the
in fact, the admitted hypothesis for the SAP2000 model in the figures 14 and 15.
analytical model in which there is a linear
variation of the displacement field is not exact,
since the variation in displacement obtained by
FEM appears to have a parabolic form. Still, the
admitted hypothesis is a reasonable
approximation, notwithstanding the existence
of smaller displacements along the depth but
with the same displacement at the top of the

8
 FEM Analytic Model FEM Analytic Model
350 350

300 300

250 250

z (m)
200 200
z (m)

150
150
100
100
50
50
0
0 0 0,2 0,4 0,6 0,8
0 20 000 40 000 60 000 Vertical Displacement (mm)
Axial Force on the Rock(KN)

Figure 14 Axial force on the rock in function of z Figure 16 - Vertical displacements in function of z (with
springs)

Analytic Model FEM 6. Discussion of the Displacement Field

350
300 Due to the non-accuracy of the hypothesis
250 adopted, in which the displacement field varies
200 linearly in depth, it is now considered a new
z (m)

150 hypothesis, in which the displacement field

100 varies parabolically in depth (Fig. 17), in order to
50 make an improved approximation of the actual
0 result of forces and displacements of the shaft in
-5 000 0 5 000 10 000 15 000
Axial Force on the Concrete (KN) relation to the first one.

Figure 15- Axial force on the reinforced concrete in function

of z

It is noted that the results given by the analytic

model are not exact, this is due to the inaccuracy
of the hypothesis for the displacement field. Still
these results are a good approximation of the
real ones.

As anticipated, the displacements (Fig. 16) given

by the analytic model are different from the
displacements given by the SAP2000 model.

Figure 17 - Displacements field of the shaft and free body

diagram of a portion of the shaft

Having said that it is necessary to redefine the

equations corresponding to the force
distributed along the shaft (32), the axial force

9
on the rock mass (33), the axial force in the FEM
concrete (34) and displacement at the top of the
Analytic Model (Linear Hypothesis)
shaft (35).
350

() = 2 (2 2 ) (36) 300

250

() = 2 200

z (m)
0
3 (37) 150
(2 )
3 100

50
() = +
2 0
3 (38) -5 000 0 5 000 10 000 15 000
( 2 )
3 Axial Force on the Concrete (KN)

() Figure 19 - Axial force on the reinforced concrete in function
= of z
0
2 (39) The displacements approximation is also
=6
12 + 5 2 improved with the adoption of the new

Comparing the results of this new hypothesis hypothesis relatively to the previous one (Fig.

with the results from the previous one and with 20).

the results of the FEM model (Figs. 18 and 19) it

FEM
is viable to say that although these last results
Analytic Model (Linear Hypothesis)
are also not exact, they represent a better
approximation to the SAP2000 results than the 350
first hypothesis. 300
250
FEM
200
z (m)

Analytic Model (Linear Hypothesis) 150

350 100

300 50
0
250
0,00 0,20 0,40 0,60 0,80
200 Vertical Displacements (mm)
z (m)

150
Figure 20 - Vertical displacements in function of z
100

50

0 7. Structural Design of the Shaft

0 20 000 40 000 60 000
Axial Force on the Rock (KN) At this moment it is proceeded to the structural
design of the shaft, which includes checking the
Figure 18 - Axial force on the rock in function of z
concrete compression on the bottom of the

10
shaft, because that is where there is located the
highest compressive strength, and also the steel
reinforcement needed to accommodate the
tensile strength.
. So the needed reinforcement for
each model is presented in table 6 ( =
435 ).
Table 6 - Needed Reinforcement for each model

In order to ensure that the concrete does not Analytic Analytic

Model - Model -
come into compression failure, it is needed to FEM
Linear Parabolic
check condition (40) (Eurocode 2), which Hypothesis Hypothesis

() 780,76 3706,67 2376,45
provides the maximum limit for the compressive
() 1171,1 5560,0 3564,7
stress acting on the concrete. ( ) 26,9 127,8 81,9

( /) 1,32 6,26 4,01
= = (40)
Reinforcemen
As the concrete considered is a C30/37 (a value t on each side 6//0,30 10//0,20 8//0,20
of the ring
of = 30 ), the design value of the
concrete compression failure is given by (41).
It is also essential to determine the minimum
0,85 30 reinforcement for the shaft according to
= = 17
1,5 (41)
Eurocode 2 (43).

Comparing table 5 and equation (41) it is

, = (43)
possible to see that the concrete compression is

verified for all the three models ( = 1,5).

On table 7 is given the minimum reinforcement
Table 5- Values of the compression stress
adopted ( = 2,9 ).

() ( ) ()
FEM 2829,28 5,843 0,726 Table 7 Minimum Reinforcement
Analytic
, ( ) 338,9
Model -
12058,08 5,843 3,096
Linear
Hypothesis , 16,60
( /)
Analytic
Model - Reinforcement on each
9665,83 5,843 2,481 16//0,20
Parabolic side of the ring
Hypothesis

Regarding the longitudinal reinforcement 8. Conclusions

design, since it is a circular section vertical shaft, The primary goal of this work, that was obtaining
it can be calculated as if it were a reinforced an analytic model for the to obtain the load
concrete structural wall according to Eurocode 2 distribution between the rock mass and the
(42). reinforced concrete structure, was fulfilled,

even though the results given by this model are

= (42)
an approximation to the real ones, because of
Once there are no moments applied on the
the fact that the admitted hypothesis regarding
structure the tensile force it is given by =

11
the displacement field is not accurate;
nevertheless it is a plausible approximation.
Concerning the result of the load distribution
obtained by the analytical model developed, it is
noted that for the axial force in the concrete,
that traction is observed in large part of the shaft
length, and compression near the bottom of the
shaft, recognizing that this result is due to the
spring-type effect conferred by the wedges of
the rock mass, since the axial force in the rock
Hartman, H., (1992) S.M.E. Mining Engineering
Handbook, 3rd edition, Society for Mining,
Metallurgy, and Exploration, Colorado, U.S.A.
Ozdemir, L., (1986) Improving the performance
of tunnel, raise, and shaft boring machines for
coal mine applications, Technical Report, Bureau
of Mines, U.S. Department of the Interior,
Colorado, U.S.A.
Timoshenko, S., (1940) Strength of materials:

increases greatly in the first meters of depth and Part II advanced theory and problems, 2nd

tends to stabilize in the last meters. edition, D. Van Nostrand Company, Inc., New
York, U.S.A.
Regarding the structural design of the shaft is
concluded that either for analytical model, Timoshenko, S. & Goodier, J., (1951) Theory of
either to the SAP2000 model are verified the Elasticity, 2nd edition, McGraw-Hill, New York,
safety of concrete in compression and the U.S.A.
adopted steel reinforcement to correspond to
the minimum reinforcement imposed by
regulations.

9. References

American Society of Civil Engineers (1989), Civil

Engineering Guidelines for Planning and
Designing Hydroelectric Developments, Volume
2 (Waterways), New York, U.S.A.

CEN - NP EN 1992-1-1 (2010). Eurocdigo 2:

Projeto de estruturas de beto Parte 1-1:
Regras gerais e regras para edifcios. Instituto
Portugus da Qualidade (IPQ), Lisbon.

Grieves, M. (1996), Deep Shafts, in J. Bickel, T.

Kuesel, & E. King, Tunnel Engineering Handbook,
2nd edition, Kluwer Academic Publishers,
Norwell, Massachusetts, U.S.A.

12