1 Field monitoring of a motorway viaduct moving on an extremely slow landslide Tombolato Sara, Pedrotti Matteo, Simeoni Lucia, Mongiovì Luigi University of Trento, Italy Introduction The A22 Modena-Brenner motorway runs in Northern Italy from the Alps to the Po plain, and connects Italy with Austria. In the mountainous area more than one third of the A22 motorway was built on viaducts, and their monitoring is of vital, economic and strategic importance for the whole network. Regularly, a team from the A22 company monitors the viaducts and carries out works for the maintenance of the structures and for preserving their functionality and safety. In particular, the relative pier-deck bridge displacements are checked by visual inspections and measurements to find out if shear deformations have occurred. At the late 1980s anomalous displacements were surveyed at the Micheletti viaduct (Figure 1), and since then the A22 office have been monitoring the piers and the slope. In 2007 the A22 board requested that the University of Trento study the stability of the viaduct by defining the causes of the displacements. This paper describes the instrumentation used for the monitoring and how the measurements were processed for assessing their reliability. In fact, dealing with an extremely slow movements, the changes of the measurements could be as small as their precision or accuracy. The reliability assessment is therefore fundamental to use the measurements for interpreting the failure mechanism. Different instruments for measuring the pier viaduct movements and the slope movements were used. Piers were monitored using a Total Station, pendula and clinometers; slope movements were measured with inclinometers. The reliability was assessed by analysing the redundancy of the measurements, and results were used for identifying the failure mechanism.
18
Embed
Field monitoring of a motorway viaduct moving on an ......Inclinometer casing Direct pendulum Total station prism Biaxial clinometer Cropping out rock Isarco river SS12 road pier 20
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
1
Field monitoring of a motorway viaduct moving on an extremely slow landslide Tombolato Sara, Pedrotti Matteo, Simeoni Lucia, Mongiovì Luigi
University of Trento, Italy
Introduction The A22 Modena-Brenner motorway runs in Northern Italy from the Alps to the
Po plain, and connects Italy with Austria. In the mountainous area more than
one third of the A22 motorway was built on viaducts, and their monitoring is of
vital, economic and strategic importance for the whole network.
Regularly, a team from the A22 company monitors the viaducts and carries out
works for the maintenance of the structures and for preserving their functionality
and safety. In particular, the relative pier-deck bridge displacements are
checked by visual inspections and measurements to find out if shear
deformations have occurred. At the late 1980s anomalous displacements were
surveyed at the Micheletti viaduct (Figure 1), and since then the A22 office have
been monitoring the piers and the slope. In 2007 the A22 board requested that
the University of Trento study the stability of the viaduct by defining the causes
of the displacements.
This paper describes the instrumentation used for the monitoring and how the
measurements were processed for assessing their reliability. In fact, dealing
with an extremely slow movements, the changes of the measurements could be
as small as their precision or accuracy. The reliability assessment is therefore
fundamental to use the measurements for interpreting the failure mechanism.
Different instruments for measuring the pier viaduct movements and the slope
movements were used. Piers were monitored using a Total Station, pendula
and clinometers; slope movements were measured with inclinometers. The
reliability was assessed by analysing the redundancy of the measurements, and
results were used for identifying the failure mechanism.
2
a)
b)
Figure 1. a) Piers 23 (right) through 25 (left) of the Micheletti viaduct; b) deformed
elastomeric pad revealing the relative displacement pier-deck.
Geology The viaduct Micheletti lies on the western slope of the Isarco Valley, about
10 km North of Bolzano in Northern Italy (Figure 2).
Figure 2. A22 Highway map and Micheletti Viaduct location
The Isarco Valley is a U-shaped glacial valley, with current terrace formations
ascribed to alternation of glacial and post-glacial (fluvial) quaternary deeping
processes.
Figure 3 shows a geological cross section of the western valley flank where the
viaduct is located.
PIER
DECK
3
The profile exhibits different gradients corresponding to different geological
deposits. The lower valley flank is 40° dipping and covered by debris. These
debris consists of soil of gravitational origin (scree slope, rockfall deposit) mixed
with glacial deposits and alluvional lens. Debris spread out to a quasi vertical
ignimbrite rock slope outcropping from 410 to 475 metres above sea level. Up
here the slope becomes smoothly as a typical post-glacial landscape. Till and
morenic deposits covers these area, which ends in vertical outcropping rock
(not in figure).
Figure 3. Geological cross section of the studied valley flank.
The viaduct structure The highway viaduct is supported by a row of 8 concrete piers 35 meters apart.
The piers are 35 meters height and are founded on footing foundations in the
steeply inclined slope (about 40°). The depth of foundations is similar in all piers
and it is about 10 m from the ground surface. The piers were isolated from the
surrounding soil by elliptical and empty caissons, that are structurally jointed to
the piers foundations. At the top, each pier has six elastomeric bearings
arranged in 2 arrows which allow the deck to rotate and translate by distorting
the elastomeric pads.
336
386
436
536
met
ers
abov
e se
a le
vel
viaduct pier n.24
SS 12 street
Isarco river bed
486
Till and morainicdeposit
40°
Tuff rock 70°
Ignimbriti rock
cropping outrock
SS12 way
? ??
?
?
??
?
30°
Gravitationalalluvionalglacial deposit
0 50 m
4
The monitoring The monitoring was planned to measure both piers and slope displacements.
Instrumentation for the piers monitoring consisted of 6 biaxial clinometers, 4
direct pendula and a Total Station with 37 prisms. 13 inclinometers were used
for the slope monitoring.
The monitoring covered an area of about 150 m long and includes 8 piers,
numbered from 20 to 27 as showed in Figure 4.
Inclinometer casingDirect pendulumTotal station prismBiaxial clinometer
Figure 9. Pier tilts measured by the biaxial clinometers
9
Slope monitoring Soil displacements were measured using a mobile biaxial SisGeo inclinometer
probe with servoaccelerometer sensors. Measurements were carried out four
times by lowering the probe in every groove of the casing. Consequently, the
measurements were redundant and could be processed in four different ways:
two ways by assuming the probe biaxial (elaboration 1-3 and elaboration 2-4),
and two ways by assuming the probe monoaxial (elaboration A-A and
elaboration B-B). Figure 10 shows the cumulative displacements calculated at
the inclinometer T5 with the four elaborations. It is worth noticing that the
displacements differ considerably.
Figure 10. Inclinometer T5. Cumulative displacements (October 2009 – April 2010) By observing Figure 10, it is clear that difference increases topwards from the
bottom of the casing as it happens for the propagation of systematic errors.
None of the errors studied by Mikkelsen, 2003 was recognized and corrected.
Hence, attempts were made to find any other type of systematic errors, such as
10
those due to no-perpendicular grooves, or empirical relationships with
measurements, but none of them resulted successful.
In order to reduce the effects of the propagation of the systematic errors, soil
displacements were estimated by integrating the displacements just in the
intervals of depth where the local displacements exhibited both a magnitude
significantly larger than the precision (Simeoni et al., 2007) and an azimuth
coherent with the slope. Figure 11 shows two examples of local displacements
in inclinometer I3 and inclinometer T5. In the latter it could be seen that there is
not any significant local displacements. Therefore, in this inclinometer the local
displacements were not integrated and the inclinometer was assumed fixed,
despite the conventional cumulative displacement calculation had provided
displacements different form zero at the ground surface (Figure 10).
a) Inclinometer I3 b) Inclinometer T5
Figure 11. Inclinometers I3 and T5: local displacements (October 2009 – April 2010).
11
Reliability analysis In order to assess the reliability of the measurements different sets of data were
compared. The redundancy of the horizontal relative displacements was studied
by comparing the data from the total station, the biaxial clinometers and the
direct pendula.
The redundancy of the absolute horizontal displacements was studied by
comparing the structure displacements with the soil displacements at the pier
foundation level.
Redundancy of total station and pendulum measurements
Both Total Station and direct pendulum measurements were performed in the
same period, from November 2004 to April 2010. It was therefore easy to study
the redundancy by comparing the two type of measurements in terms of
magnitude and orientation. The agreement between the two different
instruments was good for each pier. As an example, Figure 12 compares the
pier relative horizontal displacements top-foundation calculated at the pier 25.
Even though the simplicity of the instrument, it may be concluded that the direct
pendulum provided very reliable measurements.
10-N
ov-0
414
-Dec
-04
6-A
pr-0
5
6-Se
p-05
18-O
ct-0
5
1-M
ar-0
612
-Jun
-06
2-A
ug-0
622
-Nov
-06
22-M
ay-0
7
2-O
ct-0
715
-Nov
-07
14-M
ar-0
83-
Jun-
08
10-M
ar-0
9
23-S
ep-0
9
12-A
pr-1
0
0
1
2
3
4
5
mag
nitu
de (m
m)
Direct pendulumTotal station
10-N
ov-0
414
-Dec
-04
6-A
pr-0
5
6-Se
p-05
18-O
ct-0
5
1-M
ar-0
612
-Jun
-06
2-A
ug-0
622
-Nov
-06
22-M
ay-0
7
2-O
ct-0
715
-Nov
-07
14-M
ar-0
83-
Jun-
08
10-M
ar-0
9
23-S
ep-0
9
12-A
pr-1
0
0
60
120
180
240
300
360
orie
ntat
ion-
north
dire
ctio
n (°
)
Direct PendulumTotal station
(a) magnitude (mm) (b) orientation-north direction (°)
Figure 12. Redundancy of total station and pendulum measurements (pier 25)
12
Redundancy of biaxial clinometer and pendulum
As showed in Figure 13 the pier inclinations calculated from the clinometers
data resulted completely different from the ones calculated from the direct
pendula data. Therefore the clinometer measurements were rejected.
28-J
un-9
514
-Nov
-95
18-J
un-9
6
10-J
ul-9
79-
Apr
-98
14-S
ep-9
81-
Apr
-99
15-F
eb-0
09-
Aug
-00
18-D
ec-0
027
-Jun
-01
25-M
ay-0
25-
Nov
-02
18-D
ec-0
314
-Sep
-04
7-A
pr-0
519
-Oct
-05
13-J
un-0
623
-Nov
-06
23-M
ay-0
7-0.005
0
0.005
0.01
0.015
0.02
ango
lo α
risp
etto
alla
ver
tical
e (°
)
Pila 22Pila 23Pila 24Pila 25
clinometri biassiali
filo a piombo
Pier
incl
inat
ion
(°)
Biaxial clinometers
Direct pendula
Pier 22Pier 23Pier 24Pier 25
Figure 13. Redundancy of the biaxial clinometer and pendulum measurements.
13
Redundancy of total station and inclinometers
For studying the redundancy between the inclinometer and the total station data, the
absolute soil displacements at the foundation (inclinometer cumulative displacement)
were compared to the absolute pier displacements (Total Station displacement).
Because the measurements were neither contemporary nor carried out on the same
date, the comparison was made in term of displacement rates. Soil displacements
were measured from October 2009 to July 2010, whereas pier displacements were
measured from November 2004 to April 2010.
All the inclinometers, except inclinometer I2, were drilled next to a pier (Figure 4).
Therefore the comparison was made between each inclinometer and the pier next to
it (Table 1). The soil displacements were estimated by integrating the local
displacements in the way explained before and only below the foundation level.
In Table 1 It is worth noticing that the soil displacements rates were very similar to the
pier ones, and ranged between 7 and 10 mm/year. Accordingly to Cruden and
Varnes, 1994 the slope movement was classified as extremely slow.
The orientation of the displacements calculated from either the inclinometers or the
Total Station data were very similar and close to 110°clockwise from the North
direction). This value was coherent with the slope morphology.
14
Mechanism of failure Sliding surface limits Despite the presence of discontinuities in the outcropping ignimbrite, their orientation
did not identify any sliding surface. Moreover, the inclinometers located at the base of
the slope, close to the road SS12 (inclinometers T2 and T5), did not detect significant
local displacements. Accordingly, it seemed reasonable to assume that the sliding
surface developed in the gravitational-alluvional-glacial deposit immediately above the
road SS12.
Kinematic of the soil-viaduct structure The soil displacement analysis identified two different sliding surfaces (Figure 14).
One developed in the northern zone (pier 22 trough pier 24) and it was located at the
level of the foundations. It was called upper sliding surface. The other developed in
the southern zone (pier 24 trough pier 26) and it was located 3-4 meters below the
foundations. It was called deeer sliding surface. In the middle zone (pier 24) both
surfaces were identified; each of them exhibited a smaller displacement rate than on
the sides, but the sum was similar (Table 1).
In the deep cinematism close to pier 26 and 25 the magnitude of displacement rate is
included between 7 and 10 millimetres per year. Whereas the magnitude of
displacement rate of superficial cinematism, close to pier 22 and 23, is included
between 6 and 8 millimetres per year. Close to the pier 24 both the superficial
cinematism and the deep one have a magnitude of displacement rate approximately
3-4 millimetres per year.
Due to the redundancy between soil and structure displacement rates (Table 1) it
seemed reasonable to assume that any other deep surface was not relevant to the
movement of the viaduct. In fact, if another deeper sliding surface, developing under
the bottom of inclinometers T2 and T5, existed it would move extremely slowly and
could not affect the stability of the viaduct. Furthermore, if the sliding surface is
considered circular, the small vertical displacements, measured from total station
analysis (Figure 5), show that there is no deeper sliding surface which includes the
total station base or not.
15
Figure 14. Location of the two sliding surfaces.
Causes of the viaduct movement Because of the extremely slow movements, there was no a clear evidence on the
ground of the existence of one or more landslides. Therefore the viaduct piers had
moved either for a problem of bearing capacity of their foundations or for a problem of
global stability or both. The answer came from the analysis of the inclinometer
measurements. In fact, given that the inclinometer measurements were reliable
because redundant with the Total Station measurements, there were firm evidences
confirming that the failure mechanism was due to the slope instability. The first
evidence was the extraordinary repeatability of the deeper sliding surface location
beneath the foundations (3-4 meters). The second one was the uniformity of the
displacement rates and their directions in the whole area examined. Furthermore, the
inclinometer T6 was drilled 16 meters upslope pier 26 and detected the sliding
surface at the same elevation of the pier’s foundation (Figure 15). If the mechanism
was a bearing capacity problem, the sliding surface should intersect the inclinometer
at an higher elevation.
16
PIER 26
T6I
T10I
ISARCO RIVER
10 m
LOCALDISPLACEMENTS
Figure 15. Cross section at Pier 26.
10 m
LOCALDISPLACEMENTS
T4I
T5I
PIER 25
ISARCO RIVER
Figure 16. Cross section at Pier 25.
Likewise, the two inclinometers T3 and T4 installed upslope and downslope the pier
25, respectively, detected the sliding surface at approximately the same elevation
(Figure 16). This showed that the sliding surface was sub-horizontal.
Hence, a bearing capacity mechanism seemed to be incompatible with the shape of
the sliding surfaces derived from the inclinometer measurements.
17
Conclusions
The system composed by a Total Station and inclinometers has proved to be suitable
for monitoring extremely slow movements and for defining the failure mechanism
causing the viaduct to move. The reliability of measurements was assessed by
analysing the redundancy between the Total Station displacement rates and the
inclinometer displacement rates integrated only where the major displacements were
identified. The comparison between the Total Station and the direct pendula
measurements revealed that the pendulum can be assumed a reliable instrument
despite its simplicity. On the contrary, the clinometer measurements were not
redundant with those of the direct pendula and were rejected.
Given that the inclinometer measurements resulted reliable, they were used for
defining the shape of the sliding surfaces and to identify the failure mechanism. It was
proved that the viaduct movement was due to a slope instability instead of a bearing
capacity problem of the pier foundations.
References Cruden D.M. and Varnes D.J.; Landslides Types and Processes, Transportation
Research Board. National Academy of Sciences. "Landslides: Investigation and
Mitigation", pp. 36-75, 1994.
Mikkelsen P.E.; Advances in inclinometer data analysis. Proc., 6th International
Symposium on Field Measurements in Geomechanics, Oslo, Norway, September, 15-
18 2003, 555-567, 2003.
Simeoni L. and Mongiovì L.; Inclinometer monitoring of the Castelrotto landslide in
Italy. Journal of Geotechnical and Geoenvironmental Engineering, Vol. 133, No. 6,
June 2007, pp. 653-666, 2007.
18
Authors
Dr Sara Tombolato, Ph.D.* [email protected] Dr Matteo Pedrotti** [email protected] Dr Lucia Simeoni, Ph.D.* [email protected] Prof. Luigi Mongiovì* [email protected] *Department of Mechanical and Structural Engineering University of Trento Via Mesiano 77, 38123 Trento (Italy) www.unitn.it/ingegneria **Strathclyde University, Glasgow