Sie sind auf Seite 1von 12

Concrete Structures for Mitigation of Debris-Flow Hazard in

the Montoro Inferiore Area, Southern Italy




Cosenza, E.
Department of Structural Analysis and Design, University of Naples Federico II, Via Claudio, 21 80125 Naples, Italy

Cozzolino, L., Pianese, D.
Department of Hydraulic and Environmental Engineering, University of Naples Federico II, 80125 Naples, Italy

Fabbrocino, G.
Department SAVA, Engineering & Environment Division, University of Molise, Via De Sanctis 86100 Campobasso, Italy

Acanfora, M.
TRE Consortium - Technologies for the Building Recovery, Naples 80125 Naples, Italy



INTRODUCTION

The paper deals with some aspects related to the design of structures able to mitigate the debris-flow hazard
in the Montoro Inferiore area, Southern Italy. In particular, the attention has been focused on the integrated
design process used to define structural performances of different concrete components and structures in
the Di Mezzo and Casapepe hollows. Structural design represents one of the components of the process
that has been characterised by a multidisciplinary approach to the problem of management and protection of
large areas exposed to debris-flow hazard: geologists, hydraulic engineers and structural engineers worked
together to give a rational answer to the problem of the definition and estimation of design actions on
structures and consequently to assess relevant performances of components and systems from the
structural viewpoint. Obviously, the extension of the area of intervention required accurate and planned
geological surveys and tests; these investigations led to the estimation of volumes expected to be mobilised
during extreme events, to the related distribution in plan and elevation and to the mechanical
characterisation of the different soils concerned by the structures along the flow course. The study of the
propagation of a debris-flows of given volume, accomplished by means of hydraulic numerical models,
aiming at the determination of the loads acting on the mitigation structures, constituted the core of the debris-
flow hazard mitigation project, bridging the gap between geology and structural design. In the following,
attention will be focused mainly on one of the cited mitigation interventions, namely the protection system of
the Di Mezzo Hollow, which will be extensively discussed.

Keywords: Environmental design, debris flow hazard; debris-flow protection systems.


DI MEZZO HOLLOW PROTECTION SYSTEM

The Di Mezzo Hollow protection system, Fig. 1, consists of three components: a main closed check dam,
which can be filled with about 2000 m
3
of debris material; an adequate number of secondary closed check
dams, able to contain 100 m
3
of debris material; a valley stilling basin, with roof-shaped bottom, bounded by
earthen levees, where 11 debris-breakers have been placed. Depending on the debris-flow volume
mobilised from the basin slopes and from the hollow bed, the main upstream check dam behaves properly. If
the debris-flow volume is less than 2000-2200 m
3
, the closed check dam firstly accomplishes a stopping
function, and subsequently behaves as a sort of retaining wall, with loads that are greater than ordinary
geotechnical ones. Whenever the debris-flow volume produced during an event (or during a rapid
succession of events), exceeds the limit of 2200 m
3
, the following functions are accomplished:

2



Fig. 1. Di Mezzo hollow protection system overview. From the left: the main check dam, the secondary dams along the hollow, the stilling basin with earthen levees boundaries and
concrete flow-breakers.


3
a. during the initial phase, the check dam exerts a stopping action on the flow, causing an active dissipation
of the flow total head;
b. the free volume at the back of the check dam is filled with the debris-flow material, which is subtracted to
the downstream flow along the hollow, thus preventing, or reducing, the scour of the hollow bed
downstream;
c. once the volume at the back of the check dam is totally filled, the exceeding volumes are discharged by
a properly designed weir onto the downstream flow. It is worth to observe that the fall of the discharge
onto the downstream flow actively dissipates part of the total head, and reduces the destructive flow
velocity and its scouring capability.
Secondary check dams, because of the small free volume at their back, are slowly filled during the operating
life of the protection system, thus reducing the slope of the hollow bed upstream: this fact contributes to the
reduction of the velocity of the flow discharged by the main dam. Moreover, the fall of the flow passing over
the secondary dams onto the downstream flow causes again the debris-flow energy dissipation, preventing
from the acceleration of the flow along the hollow downstream of the main dam.
Finally, the trapezoidal stilling basin, located at the end of the hollow in the valley plain area, accomplishes a
number of functions:
a. the debris-flow moving from the hollow end is forced to spread over a vast area, with consequent
reduction of flow height and velocity due to the roof-shape of the stilling basin bottom;
b. a number of flow-breakers is placed into the stilling basin, they subdivide the flow, contribute to the
dissipation of the debris-flow energy, and reduce the loads acting on the boundary levees;
c. a detention function is developed, in fact the basin is designed to contain a debris volume of 22000 m
3
,
so the overloading of the discharging channel is prevented;
d. the discharging channel is intended to be located at a lower level respect to the basin bottom, thus
causing again energy dissipation consequent to the fall.


DEBRIS-FLOW DESIGN ACTIONS ON PROTECTION STRUCTURES

In order to evaluate the design actions on the protection structures, a two-dimensional Finite Volume
numerical model aiming at the simulation of debris-flows has been implemented. The cited model
approximates the solution of the following set of partial differential equations, namely the Shallow-water
Equations
( )
( )
fy oy y
fx ox x
S S gh
y
V
h
gh
K hV
y x
V
y
U
h hUV
x t
hV
S S gh
x
V
y
U
h hUV
y x
U
h
gh
K hU
x t
hU
y
hV
x
hU
t
h
=
|
|
.
|

\
|

+
|
|
.
|

\
|
|
|
.
|

\
|

=
|
|
.
|

\
|
|
|
.
|

\
|

+
|
|
.
|

\
|



2
2
2
2
0
2
2
2
2
(1)
where h is the flow depth; U is the x-component of the vertically-averaged velocity W; V is the y-component
of the vertically averaged velocity W; g is the gravity acceleration; S
ox
and S
oy
are the x- and y-component,
respectively, of the bed-slope; S
fx
and S
fy
are the x- and y-component, respectively, of the friction slope S
f
;
K
x
, K
y
are the correction factors of the hydraulic thrust; is the total cinematic viscosity. The Shallow-water
Equations are completed by the following set of closure models, in order to take in account for the dissipation
of energy due to the debris-flow rheology:
f
ghS h .1 0 =
(2)
and

|
|
.
|

\
|

|
|
.
|

\
|

|
|
.
|

\
| +
= >
=
= +
f
c
f
c
f
f
c
f
c
hS g hS g
h
k
S g
W
sen g
h
W
sen g
h
W

1
2 1
1
1
0
1
1
(3)
In the friction-slope Eq. 3,
c
, and are the parameters of the Herschel-Buckley type model

4

|
.
|

\
|
+ =
dz
dw
k
c
(4)
where the following values have been assumed:
c
= 100 N/m
2
, = 1600 kg/m
3
, = 3 Pas, = 1.
It is worth that the Shallow-water Equations, are strongly non-linear, and exhibit discontinuous solutions,
such as trans-critical flow conditions, also in the cases of continuous initial conditions. Moreover, the exact
extension of the computational domain is not a-priori known, due to the presence of moving wetting-drying
fronts.

5437 5437



Fig. 2. Vallone di Mezzo stilling basin: example of the computational grid used for the debrisflow propagation model
nearby the flow-breakers.

In order to overcome these difficulties, a specific software program based on the Finite Volume numerical
solution procedure for Shallow-water Equations has been implemented. The domain has been subdivided N
in triangular non-overlapping finite volumes (see Fig. 2), where the Shallow-water Equations have been
spatially integrated:


= +
i i i
Sd d F n Ud
dt
d

(5)
The original system of partial differential equations finally reduces to a set of ordinary differential equations


+ =
i i
Sd d F n
dt
U d i
i

(6)
where

=
i
Ud U
i
i
1
(7)
is the vector of the mean value of the conserved physical quantities in each control volume. In order to
correctly manage the discontinuities at the interface between the finite volumes, an approximate solver of the
Riemann problem has been applied for the computation of the advective fluxes, namely HLL [1], while the
diffusive fluxes have been evaluated by means of the approach by Anastasiou and Chan [2]. The robust
Euler algorithm has been chosen for the solution of Equations 6.
Further details about the numerical procedure are reported elsewhere, i.e. [3, 4]. A comparison between
different approaches used in technical field has been issued. To this end, a one-dimensional Finite Volume
model, able to solve the conservative form of the de Saint-Venant Equations, has been also implemented [5].
The debris-flow design hydrograph used as boundary condition for the Shallow-water Equations is shown in
Fig. 3.a. It has been evaluated according to the data and specific recommendations given by the geological
consultants.
Fig. 3.b reports the computational triangular grid generated for the numerical solution of the Shallow-water
Equations in the area upstream of the main check dam.

5
0
10
20
30
40
50
60
70
80
90
100
0 100 200 300 400 500 600
t (s)
Q

(
m
3
/
s
)
a) b)

Fig. 3. Vallone di Mezzo: debris-flow design hydrograph (a); computational grid of the terrain upstream the main check
dam (b).

a) b)
Fig. 4. Vallone di Mezzo: main check dam layout a); digital elevation model of the terrain upstream the main check
dam.

300
302
304
306
308
0 5 10 15 20
Abscissa (m)
E
l
e
v
a
t
i
o
n

(
m
)
Hollow transect
1D model
2D model
a)
294
298
302
306
310
0 10 20 30 40 50
Abscissa (m)
E
l
e
v
a
t
i
o
n

(
m
)
Hollow transect
Check dam crest
1D model
2D model
b)

Fig. 5. Vallone di Mezzo: flow depth upstream the main check dam at time t=60 s (a); flow depth at the main check dam at
time t=60 s (b).

The layout of the main check dam and the topography of the surrounding area are shown in Fig. 4.a. A bird-
fly view of the same area is proposed (from upstream to downstream) and reported in Fig. 4.b. Fig. 5.a
compares the flow depth, as supplied by the 1D and the 2D models, upstream the main check dam at the

6
Transect 2 abscissa (compare with Fig. 4.a), at time t=60 s. It is easy to recognise that the average flow
depth for the two numerical models is quite similar, but the one-dimensional model cannot take in account for
the strong curvature and the complicate topography of the channel. This circumstance has relevant
consequences for the design of the main check dam. In fact, in Fig. 5.b the flow depths, as supplied by the
two models immediately upstream of the main check dam, are compared: the differences of the flow depths
introduce a large scatter in the value and the distribution of loads along the main check dam. Moreover, the
volume of debris material discharging from the upper weir increases passing from the two-dimensional to the
one-dimensional model.

0
100,000
200,000
300,000
400,000
500,000
0 20 40 60 80
t (s)

Q
V

(
N
)
1D model
2D model
a)
0.E+00
3.E+06
5.E+06
8.E+06
0 20 40 60 80
t (s)


(
N
)
1D model
2D model
b)

Fig. 6. Vallone di Mezzo: dynamic load on the main check dam (a); static load on the main check dam (b).

Fig. 6 shows the dynamic (a) and static loads (b) on the main check dam, as supplied by the two models,
and a comparison between the two models is reported. It is worth noting that static loads are of an order of
magnitude greater than the dynamic loads. The above plots point out that the loads increase of about 65%
passing from the two- to the one-dimensional model. Again, similarly to the case of the volumes discharged
by the upper weir, the differences can be related to the difference of the flow depth distribution supplied by
the two models. In Fig. 7 the trapezoidal stilling basin with the design debris-flow hydrograph entering the
system is represented, the velocity vectors at different times are represented as arrows. The two-
dimensional model can be used to optimise the position of the flow-breakers, in order to exploit the energy
dissipation, and to spread the flow on a vast area.



Fig. 7. The trapezoidal stilling basin with the design debris-flow hydrograph entering the system.


The dynamic, static and total loads per linear meter on the first flow-breaker (see Fig. 2) are represented in
Fig. 8. The large variability in time of the loads is not due to numerical instabilities, but instead is related to

7
the pulsating behaviour of the vertical-axis eddies that are present in the stilling basin.

0
100
200
300
400
500
600
700
0 50 100 150 200 250 300 350 400
time (s)
Dynamic
cell load
(kN)

a)
0
50
100
150
200
250
0 50 100 150 200 250 300 350 400
time (s)
Static cell
load (kN)

b)
0
100
200
300
400
500
600
700
800
0 50 100 150 200 250 300 350 400
time (s)
Total cell
load (kN)
c)
Fig. 8. Time plots of the load (per linear meter) acting on one of the flow breakers (see also Fig. 2): a) static load, b)
dynamic load, c) total load.


STRUCTURAL DESIGN OF CONCRETE COMPONENTS OF PROTECTION SYSTEM

Assumptions made for estimation of dynamic and static loads due to debris flow on the protection systems
led to select concrete structures and pointed out the role of components and sub-structures to ensure the
required energy dissipation. In fact, Figg. 2 and 7 show that protection systems are basically rigid, so that
interaction between hydraulic flow and the constructions can be neglected and the changes of the finite
volume meshes during the loading process are avoided. This is one of the basic requirements of the
constructions that oriented the conceptual design of components and even the relative verifications
according existing constructions codes [7,8]. Time plots of the loads calculated according to the hydraulic
mode show that eddies generated by the impact of debris flow on the concrete surface are characterised by
a dominating frequency around 1.7 Hz, so that the dynamic properties of breakers and dams was analysed
to keep the first natural frequency of the system far for this value and around 10 Hz. Load combinations took
account of gravity loads due to structural and non structural components; soil thrusts in the case of retaining
walls to limit areas at different elevations [9]; seismic actions, since Montoro area is assumed to be subjected
to medium to high seismicity (Zone 2 according to the Italian Seismic Classification) [10]; debris flow impact
loads. Among the reported load cases, obviously gravity loads, soil thrusts and seismic action refer to
consolidated procedures, while specific review of hydraulic load data has been required in order to define the
load combinations and local values on each structure. Debris flow loads and seismic actions in some cases

8
were quite different from a quantitative standpoint, so that the latter were generally neglected. This
circumstance was related to another relevant design option related to the maintenance program of the
system that is aimed to avoid any filling of the basins in order to ensure a continuous and full service of the
system. Structural design has been carried out assuming a linear elastic response of structures and limiting
stresses acting on materials within allowable values. This approach reflects again the need of full
consistency between models used to estimate debris flow equivalent loads and the expected structural
response of structures. In fact, an ultimate limit state approach leads to yielding and plastic deformations that
decrease the stiffness of constructions and lead to change the boundary conditions to flows. On the other
hand, the capability of the hydraulic model to differentiate applied loads on the structure makes possible to
estimate even distortions and tangential stresses due to debris flow impact.
In the following a brief overview on the most relevant components of the protection system is reported.

Main check dam at the di Mezzo Hollow
The main component of the hydraulic protection system is made of a large dike that plays the relevant role of
bearing the first impact of the flow and dissipating energy by means of a Cipolletti weir, see Fig. 9. The width
of the weir ranges between 15.0 and 19.5 m and is 7.5 m high on the bottom of the channel. Design
requirements of the structures led to simplify the structural shape and ensure a sort of structural regularity to
the construction in order to control the stiffness and optimise the structural dimensions. The structural model
of the check dam is presented in Fig. 10.a [11]; it is easy to recognise that it is shaped to fit the complex
shape of the selected narrow gorge and is made of a concrete wall stiffened by orthogonal concrete walls.
The wall thickness ranges between 1.0 m at the base and 0.6 m at the top of the dike; the maximum
transverse dimension is about 10 m. Stiffeners have been located so that there is no interaction with debris
flow falling from the weir. Outflanking is prevented using lateral jet-grouting all around the construction.


Fig. 9. Di Mezzo hollow main check dam elevation (a) and down-hill view (b).

9
a)
b)
Fig. 10. Structural model of the main check dam (a); distribution of the debris flow loads (b).

Deep foundations have been adopted in order to ensure a reliable restraint at the base and resist to the large
horizontal shear and overturing moments applied at the base of the walls and primarily at the base of the
concrete stiffeners.
Structural design under debris flow loads has been carried out taking into consideration a number of complex
load conditions; the latter were related to the development of the dynamic impact of the debris flow. To this
end a correspondence between finite volume cells used for the dynamic simulation were assumed as the
base for the structural discretization. In this way a load versus time relation for each element was available.
In this way envelops of bending moments, torsional moments, shear forces were defined for structural
verifications of cross sections and reinforcement; a typical distribution of loads is reported in Fig. 10.b.
Debris flow breakers in the still basin

The debris flow breaker plays a relevant role in the energy dissipation mechanisms that governs the
expansion of debris flow in the still basin. It is characterised by a prismatic shape with a trapezoidal cross
section. It is six meters tall from the foundation, which is made of a square concrete block 6.5 m wide and
1.0 m thick. It is designed to bear the dynamic and static loads due to the debris flow. Depending on the
location of the breaker within the still basin, base details can change so that the run-off of rainfall can take
place freely precipitations.
A coupled foundation system has been provided in order to ensure that the stability of the structure during
the debris flow impact. In fact, primarily the foundation has been designed in order to directly transfer the
load due to the debris flow impact; however, in order to prevent the loss of stability due to the erosion of the
emerging soil layers, a number of piles (diameter 600 mm, 10.0 m long) have been provided. The latter have
been designed to bear the debris flow impact even under the unfavourable assumption that a heavy erosion
takes place. To prevent this effect of the debris flow, a concrete slab 150 mm thick has been provided. Use
of high ductility steel reinforcement in this slab increases the structural robustness of the system.

10
3| 3|
!a ta
t
!a
3a
2a 2
!| !|
!a
!
t
3
!a
/
3
2
ta
9
9a 9a
9
7
t
9
8
t ! t !


Fig. 11. Main debris breaker: geometry and reinforcement detailing.



Fig. 12. Deformation of the main debris breaker under the debris flow design loads.

11
Secondary check dams at Casapepe hollow

A different example of check dam is reported in Fig. 13, that refers to the secondary check dams used for the
energy dissipation in the Casapepe Hollow protection system. The role of such secondary check dams is to
control the velocity of the debris flow and dissipate kinetic energy during the event.
The constructions are obviously shaped depending on the configuration of the natural channel and are
characterised by the presence again of a weir that enable the flow to go over the obstacle.
In this case, optimisation of the structural performance and dimensions has been pursued adopting a
concrete wall and shallow foundations. It is worth noting that this kind of solution is sustainable since
overturning moment is balanced by a stabilising action due to the weight of the debris flow itself. Shear
resistance and robustness against erosion are ensured by proper preparation of the channel bottom and by
the passive soil resistance.





Fig.13. Overview of the secondary check dam of the Casapepe hollow protection system.


CONCLUSIONS

In this paper, some aspects of the design of the concrete structures aiming at the mitigation of the debris-
flow hazard have been considered. The estimation of the design actions on this type of structures requires

12
an integrated approach, where geologists and hydraulic engineers collaborate with structural engineers, in
order to determine the volumes of the probable debris events, together with the expected debris depth and
velocity. A finite volume numerical model aiming at the solution of the two-dimensional Shallow-water
Equations has been presented, which is able to simulate the special features of a debris-flow, such as
propagation over a dry bed, and the contemporary presence of sub-critical and super-critical flow: the cited
model, making use of unstructured grids, can manage complex topographies and computational domains
boundaries, such as those represented by the protection structures above described. For comparison
purposes, a one-dimensional model which preserves the capability of manage discontinuities of the flow
filed, has been implemented: the differences between the results of the two models, which can be attributed
to the complicate geometry of the computational domains, show that the two-dimensional model is preferable
in any circumstance.
From a structural perspective, it is worth noting that implementation of structural verification is made complex
by the nature of the loads and their distributions on the impacted surface. As a consequence, input of load
data becomes a critical issue of the design procedure and needs careful consideration. Obviously softwares
commonly used for structural design sometimes cannot fit the requirements of this type of calculations.
In the present example, elastic analysis of structures has been used to make structural models compliant
with hydraulic ones, but it is clear that further research is needed to implement optimised design procedures
that enable to explicitly the interaction between debris flow and structural components. In this way, design
can take advantage of ductility of members and reduce structural dimensions and impact of constructions.


REFERENCES

1. Fraccarollo L, and Toro E F. Experimental and numerical assessment of the shallow water model for
two-dimensional dam-break type problems, IAHR Journal of. Hydraulic Research, 1995; 33(6): 843-864.
2. Anastasiou H, Chan C T. Solution of the 2D Shallow water equations using the finite volume method on
unstructured triangular meshes, International Journal for Numerical Methods in Fluids, 1997; 24: 1225-
1245.
3. Pianese, D., e Barbiero, L.. Formulation of a two-dimensional unsteady debris-flow model for the
analysis of debris-flow hazards and countermeasures thereof. Proc. of 3rd International Conference on
Debris-Flow Hazards and Mitigation: Mechanics, Predictions and Assessment, Congress Center,
Davos (CH), September 2003, Vol. 1, 705-716.
4. Cozzolino L, Barbiero L, Della Morte R. and Pianese D. Criteri di dimensionamento di un particolare tipo
di dispositivo per la dissipazione dellenergia posseduta dalle colate, Proc. of Giornate di Studio su La
difesa Idraulica del Territorio, September 2003, Trieste, 203-216 (in italian).
5. Barbiero L, Cozzolino L, Iavarone V and Pianese D. Valutazione, attraverso modellazione uni- e bi-
dimensionale dei fenomeni di moto vario, delle interazioni tra correnti detritiche e briglie frangi-colata,
Proc of the workshop Modelli matematici per la simulazione di catastrofi idrogeologiche, march 2004,
Arcavacata di Rende (CS), 273-284 (in italian).
6. D.M. 16 Gennaio 1996 Criteri generali per la verifica di sicurezza delle costruzioni e dei carichi e
sovraccarichi. G.U. 5-2-1996, n.29. Ministry of Public Works. (in italian).
7. D.M. 6 Gennaio 1996 "Norme tecniche per il calcolo, lesecuzione ed il collaudo delle strutture in
cemento armato, normale e precompresso e per le strutture metalliche". G.U. 5-2-1996, n.29. Ministry of
Public Works. (in italian).
8. CEB Bulletin No. 213/214 (1993). CEB-FIP Model Code 90.
9. D.M. 11 Marzo 1988. Norme tecniche riguardanti le indagini sui terreni e sulle rocce, la stabilit dei
pendii naturali e delle scarpate, i criteri generali e le prescrizioni per la progettazione, lesecuzione e il
collaudo delle opere di sostegno delle terre e delle opere di fondazione. (G.U. 1.6.1988, n.167). Ministry
of Public Works. (in italian).
10. D.M. 16 Gennaio 1996 "Norme tecniche per le costruzioni in zone sismiche". G.U. 5-2-1996, n.29.
Ministry of Public Works. (in italian).
11. Computer & Structures, Inc. SAP2000, Structural Analysis Program. 1994, California.

Das könnte Ihnen auch gefallen