Development of a MEMS-Based In-Place Inclinometer-Accelerometer

Array for Monitoring and Evaluation of Geotechnical Systems

By
Victoria Gene Bennett
A Thesis Submitted to the Graduate
Faculty of Rensselaer Polytechnic Institute
in Partial Fulfillment of the
Requirements for the degree of
DOCTOR OF PHILOSOPHY
Major Subject: Civil Engineering

W
Approved by the
Examining Committee: IE
_________________________________________
Tarek Abdoun, Thesis Adviser
EV

_________________________________________
Ricardo Dobry, Member
PR

_________________________________________
Mourad Zeghal, Member

_________________________________________
Thomas Zimmie, Member

_________________________________________
Anand Puppala, Member

Rensselaer Polytechnic Institute

Troy, New York
May 2010
 UMI Number: 3420903

All rights reserved

INFORMATION TO ALL USERS

The quality of this reproduction is dependent upon the quality of the copy submitted.

In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if material had to be removed,

W
a note will indicate the deletion.

IE
EV

UMI 3420903
PR

Copyright 2010 by ProQuest LLC.

All rights reserved. This edition of the work is protected against
unauthorized copying under Title 17, United States Code.

ProQuest LLC
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, MI 48106-1346
 W
IE
EV
PR

© Copyright 2010
by
Victoria Gene Bennett
All Rights Reserved

ii
 CONTENTS
LIST OF TABLES............................................................................................................ vi
LIST OF FIGURES ......................................................................................................... vii
ACKNOWLEDGMENT ................................................................................................. xv
ABSTRACT .................................................................................................................. xvii
1. Introduction and Historical Review............................................................................. 1
1.1 Current Geotechnical Instrumentation................................................................ 3
1.2 Recognizing the Need for Geotechnical Instrumentation................................... 7
1.3 Scope of Present Research.................................................................................. 8
1.4 Thesis Organization .......................................................................................... 12

W
2. Sensor Design and Specifications.............................................................................. 13
2.1 Sensor Design ................................................................................................... 13
2.1.1
IE
Installation and Retrievability .............................................................. 17
2.2 Method of Displacement Calculation ............................................................... 22
EV

2.3 Sensor Stability with Time ............................................................................... 27

3. Small-Scale Laboratory Tests.................................................................................... 30
3.1
PR

Shaking Table Calibrations Tests ..................................................................... 30

3.1.1 Initial Testing and Troubleshooting ..................................................... 30
3.1.2 Final Test Series................................................................................... 33
3.1.3 Acceleration Results ............................................................................ 36
3.1.4 Permanent Displacement Results......................................................... 41
3.2 Tension Tests on Composite ShapeAccelArray Joints..................................... 42
4. Full-Scale Laminar Container Tests at University at Buffalo ................................... 44
4.1 ShapeAccelArray Installation in Laminar Container for Test LG-0 ................ 45
4.2 ShapeAccelArray Results from Test LG-0....................................................... 47
4.3 ShapeAccelArray Installation in Laminar Container for Test SG-1 ................ 65
4.4 ShapeAccelArray Results from Test SG-1 ....................................................... 69

iii
5. Initial Field Implementations of ShapeAccelArray System ...................................... 77
5.1 Unstable Slope Monitoring with ShapeAccelArray System ............................ 77
5.1.1 Field Installation of ShapeAccelArray in an Unstable Slope Site with
Caltrans Collaboration ......................................................................... 77
5.1.2 Geotechnical Analysis of Unstable Slope in California with SLOPE/W
.............................................................................................................. 87
5.1.2.1 ‘Grid and Radius’ Slip Surface Specification........................ 90
5.1.2.2 ‘Entry and Exit’ Slip Surface Specification........................... 90
5.1.2.3 SLOPE/W Results using Variations with ‘Grid and Radius’
Definition of Lower Slope Failure......................................... 91
5.1.2.4 SLOPE/W Results using Variations of Uniform Vertical

W
Surcharge Load Representing Highway Load of Lower Slope
Failure .................................................................................... 97
5.1.2.5 SLOPE/W Results using Variations of ‘Entry and Exit’
IE
Definition of Lower Slope Failure......................................... 99
5.1.2.6 SLOPE/W Results using Variations with ‘Grid and Radius’
Definition of Upper and Lower Slope Failure ..................... 100
EV
5.1.2.7 Discussion of Findings from SLOPE/W Simulations.......... 103
5.2 Field Installation of ShapeAccelArray in Caltrans Plantable Geosynthetic
Reinforced Retaining Wall, San Diego, California........................................ 104
PR

6. Construction Monitoring with NYSDOT ................................................................ 109

6.1 Bridge Replacement and Realignment Site – Fort Ann, NY.......................... 109
6.1.1 Vertical ShapeAccelArray Installation and Results........................... 111
6.1.2 Horizontal ShapeAccelArray Installation and Results....................... 118
6.2 Geotechnical Analysis of Clay Settlement ..................................................... 127
6.2.1 Prefabricated Vertical Drain (PVD) Consideration ........................... 131
6.2.1.1 Drain Spacing Factor ........................................................... 133
6.2.1.2 Soil Disturbance Factor........................................................ 134
6.2.1.3 Drain Resistance Factor ....................................................... 134
6.2.2 Discussion of Findings from PVD Software Consideration .............. 135
6.2.3 Summary of Observations.................................................................. 144

iv
 6.3 Active Landslide – Springville, NY .............................................................147
7. Discussion and Conclusions .................................................................................154
REFERENCES ..........................................................................................................160
Appendix 5.1: Methods of Analysis in Slope/W Software ..........................................166

Appendix 5.2: Complete Set of SLOPE/W Simulations..............................................168

Appendix 6.1: Method of Calculation in FoSSA Software..........................................175
Appendix 6.2: Complete Set of FoSSA Simulations ...................................................177

W
IE
EV
PR

v
 LIST OF TABLES
Table 3.1: Input frequency and amplitude combinations for shaking table calibration
tests. ......................................................................................................................... 35
Table 4.1: Detailed instrument coordinates for test LG-0 (Bethapudi, 2008). ................ 50
Table 5.1: Recommended soil strength parameters (courtesy Caltrans). ........................ 89
Table 5.2: Resultant factor of safety values and critical slip surface depths at the
ShapeAccelArray location for changes in grid mesh density definition. ................ 92
Table 5.3: Resultant factor of safety values and critical slip surface depths at the
ShapeAccelArray location for percent decreases in fill φ angle.............................. 94
Table 5.4: Resultant factor of safety values and critical slip surface depths at the
ShapeAccelArray location for percent decreases in fill cohesion. .......................... 95

W
Table 5.5: Resultant factor of safety values and critical slip surface depths at the
ShapeAccelArray location for percent decreases in fill φ angle and cohesion........ 97
IE
Table 5.6: Resultant factor of safety values and critical slip surface depths at the
ShapeAccelArray location for increases in vertical surcharge load representing
EV
highway load............................................................................................................ 98
Table 6.1: Typical soil profile for FoSSA simulations.................................................. 130
Table 6.2: Typical consolidation data for FoSSA simulations. ..................................... 130
Table 6.3: Typical prefabricated vertical drain (PVD) properties for FoSSA simulations.
PR

............................................................................................................................... 130

vi
 LIST OF FIGURES
Figure 1.1: Aerial view of the 1995 La Conchita, CA landslide (photograph by Mark
Reid, U.S. Geological Survey, 1995). ....................................................................... 1
Figure 1.2: San Fernando dam failure during the 1971 San Fernando earthquake
(photograph by R.E. Wallace, U.S. Geological Survey, 1971). ................................ 2
Figure 1.3: Slope inclinometer probe, cable and readout box (Slope Indicator, 2010)..... 3
Figure 1.4: Data collection with traditional probe inclinometer........................................ 4
Figure 1.5: Slope Inclinometer Array (Slope Indicator, 2010).......................................... 6
Figure 1.6: Downhole accelerometer arrays at Lotung experiment site (Elgamal et al.,
1995). ......................................................................................................................... 7
Figure 1.7: Sketch of the vision for utilizing ShapeAccelArrays to monitor active soil

W
and soil-structure systems.......................................................................................... 9
Figure 1.8: Wireless network for real-time monitoring of geotechnical systems.............. 9
IE
Figure 1.9: MEMS accelerometer.................................................................................... 10
Figure 1.10: ShapeAccelArray on reel for shipping to field sites and storage. ............... 11
EV
Figure 2.1: Photograph of the ShapeAccelArray and corresponding visualization of
sensor positions........................................................................................................ 13
Figure 2.2: Schematic drawing of vertical field version of the ShapeAccelArray.......... 15
Figure 2.3: Sand settlement within 51 mm (2 in) grouted-in-place casing for vertical
PR

array installation. ..................................................................................................... 19

Figure 2.4: Small casing (25 mm; 1 in) ShapeAccelArray installation method with
anchored webbing.................................................................................................... 20
Figure 2.5: Profile of sensor array within small casing (25 mm; 1 in) ten days after
installation................................................................................................................ 21
Figure 2.6: Sketch of small casing method with sand backfill. ....................................... 21
Figure 2.7: Three-dimensional description of a segment of the ShapeAccelArray......... 23
Figure 2.8: Degrees of freedom rotations of one segment of the ShapeAccelArray....... 24
Figure 2.9: Transformation of segments into an arbitrary displacement profile by a series
of rotations and bends.............................................................................................. 25

vii
Figure 2.10: 3D polyline with the same displacement profile as a physical
ShapeAccelArray resulting from a series of orientations determined from virtually
aligned accelerometer tilt values. ............................................................................ 26
Figure 2.11: ShapeAccelArray system stability in stable soil layer over a three and a half
year monitoring period. ........................................................................................... 28
Figure 2.12: Stability of MEMS sensor installed in stable slope layer for 3.5 years (18 m
(59 ft) depth of profile in Figure 2.11). ................................................................... 29
Figure 3.1: Early prototype design of ShapeAccelArray hardware with field connectors.
................................................................................................................................. 31
Figure 3.2: Early prototype of ShapeAccelArray in aluminum setup with 1g, 2 Hz input
acceleration with 1g-acceleration cap...................................................................... 32

W
Figure 3.3: Improved ShapeAccelArray prototype with corresponding improved
response for aluminum setup with resultant accelerations greater than 1g. ............ 33
IE
Figure 3.4: Setup for testing a ShapeAccelArray on RPI’s 1g shaking table.................. 34
Figure 3.5: Schematic of ShapeAccelArray test setup on RPI’s 1g shaking table. ......... 35
Figure 3.6: Comparison between acceleration (g) measured using traditional
EV
accelerometers and the ShapeAccelArray on a flexible assembly, at 0.5 Hz
excitation frequency................................................................................................. 36
Figure 3.7: Comparison between acceleration (g) measured using traditional
PR

accelerometers and the ShapeAccelArray on a flexible assembly, at 2.0 Hz

excitation frequency................................................................................................. 37
Figure 3.8: Comparison between acceleration (g) measured using traditional
accelerometers and the ShapeAccelArray on a rigid assembly, at 0.5 Hz excitation
frequency. ................................................................................................................ 38
Figure 3.9: Comparison between acceleration (g) measured using traditional
accelerometers and the ShapeAccelArray on a rigid assembly, at 2.0 Hz excitation
frequency. ................................................................................................................ 39
Figure 3.10: Earthquake detector accelerometer characteristic curve from RPI 1g shaking
table tests. ................................................................................................................ 40
Figure 3.11: Comparison between static lateral displacements (mm) measured using
traditional position sensors and the ShapeAccelArray. ........................................... 41

viii
Figure 3.12: Composite material tension test on joints of ShapeAccelArray on Instron
8562 testing machine. .............................................................................................. 43
Figure 3.13: Average results of Instron 8562 tension tests on composite joints of
ShapeAccelArray..................................................................................................... 43
Figure 4.1: Assembly of full-scale laminar container at the University of Buffalo. ....... 44
Figure 4.2: Hydraulic filling technique for soil placement in laminar container. ........... 45
Figure 4.3: One-point clamp for reference end of ShapeAccelArray in laminar container
level ground test LG-0. ............................................................................................ 46
Figure 4.4: 2.65 m (8.7 ft) extra length of ShapeAccelArrays protruding above the soil
surface in test LG-0. ................................................................................................ 47
Figure 4.5: Test LG-0 instrumentation layout – plan view (modified from Bethapudi,

W
2008). ....................................................................................................................... 48
Figure 4.6: Test LG-0 soil model and instrumentation plan (Bethapudi, 2008).............. 48
IE
Figure 4.7: Cone penetration testing at University at Buffalo test LG-0......................... 51
Figure 4.8: Average normalized cone penetration resistance and interpreted relative
density profiles for test LG-0 (Bethapudi, 2008)..................................................... 51
EV
Figure 4.9: Test LG-0 input acceleration (g) motion....................................................... 52
Figure 4.10: Acceleration time history comparison of bottom-most vertex of SAA19 (0.9
m; 3 ft east of center along longitudinal centerline) versus base accelerometer
PR

(B1X). ...................................................................................................................... 53
Figure 4.11: Acceleration time history comparison of bottom-most vertex of SAA19 (0.9
m; 3 ft east of center along longitudinal centerline) without DC component versus
base accelerometer (B1X)........................................................................................ 54
Figure 4.12: Progression of liquefaction front as shown by ShapeAccelArray (located 0.9
m (3 ft) east of the center of the container along longitudinal centerline)
accelerations within the soil deposit. ....................................................................... 56
Figure 4.13: Progression of liquefaction front as shown by ShapeAccelArray (located 1.5
m (5 ft) east of the center of the container along longitudinal centerline)
accelerations within the soil deposit. ....................................................................... 57

ix
Figure 4.14: Progression of liquefaction front as shown by ShapeAccelArray (located 0.9
m (3 ft) south of the center of the container along transverse centerline)
accelerations within the soil deposit. ....................................................................... 58
Figure 4.15: Progression of liquefaction front as shown by ShapeAccelArray (located at
the center of the container) accelerations within the soil deposit............................ 59
Figure 4.16: Acceleration time history comparison between bottom-most vertex of
SAA19 and west accelerometers on ring 1 of laminar container. ........................... 61
Figure 4.17: Acceleration time history comparison between a mid-point vertex (7) of
SAA19 and west accelerometers on ring 10 of laminar container. ......................... 62
Figure 4.18: Profile of excess pore pressure during test LG-0 (Bethapudi, 2008).......... 63
Figure 4.19: Acceleration time history comparison between a vertex (14) of SAA19 near

W
the top of the soil deposit and west accelerometers on ring 16 of laminar container.
................................................................................................................................. 64
IE
Figure 4.20: Test SG-1 setup and instrumentation layout (modified from Dobry et al.,
2010). ....................................................................................................................... 66
Figure 4.21: Base input motion for sloping ground laminar container test SG-1 (modified
EV
from Dobry et al., 2010). ......................................................................................... 67
Figure 4.22: Clamping of ShapeAccelArray at bottom of laminar container for sloping
ground test SG-1. ..................................................................................................... 68
PR

Figure 4.23: Acceleration (g) and lateral displacement comparison (mm) between
ShapeAccelArray and laminar ring accelerometers and potentiometers, respectively,
at the soil surface, mid-depth and bottom of soil deposit; ACC_RING =
Accelerometer on Laminar Ring, POT = Potentiometer on Laminar Ring, SAA =
ShapeAccelArray..................................................................................................... 70
Figure 4.24: Progression of liquefaction front as shown by ShapeAccelArray
accelerations within the soil deposit at 0.305 m (1 ft) interval at the center of the
laminar container. .................................................................................................... 71
Figure 4.25: Progression of liquefaction front as shown by ShapeAccelArray
accelerations within the soil deposit at 0.305 m (1 ft) interval at 0.9 m (3 ft) east of
the center of the laminar container along the longitudinal centerline. .................... 72

x
Figure 4.26: Profile of noncyclic (permanent) displacement during strong shaking in test
SG-1, after removing the cyclic component (filtered above 1 Hz) – best fitting
curve (solid lines), data from North video camera (circles), data from
potentiometers (diamonds), and data from ShapeAccelArray (squares) (Dobry et al.,
2010). ....................................................................................................................... 74
Figure 4.27: Comparison of noncyclic displacement between ShapeAccelArray and
laminar ring potentiometers at critical time step for lateral spreading initiation..... 75
Figure 4.28: Strain time histories from ShapeAccelArray displacements near the bottom
of the laminar container at the initiation of lateral spreading.................................. 76
Figure 5.1: Aerial view of active slope site in California (courtesy of Caltrans)............ 78
Figure 5.2: Photo of curvilinear cracks at active slope location...................................... 78

W
Figure 5.3: Profile of unstable slope site at ‘Site A’, California (courtesy of Caltrans). 80
Figure 5.4: ShapeAccelArray packaged for field installations........................................ 80
IE
Figure 5.5: ShapeAccelArray unrolled along highway shoulder prior to installation..... 80
Figure 5.6: Advancing PVC casing segments along the ShapeAccelArray. ................... 83
Figure 5.7: Earth station with buried batteries and supplemental solar panels. .............. 83
EV
Figure 5.8: Earth station antenna installed on nearby fence............................................ 83
Figure 5.9: Comparison of ShapeAccelArray (smooth lines) and traditional inclinometer
(lines with markers) data for a nineteen-month monitoring period with an updated
PR

SAA reading. ........................................................................................................... 84

Figure 5.10: ShapeAccelArray system stability in stable soil layer over a four-year
monitoring period at ‘Site A’, California. ............................................................... 86
Figure 5.11: Top view of ShapeAccelArray system stability in stable soil layer over a
four-year monitoring period at ‘Site A’, California................................................. 87
Figure 5.12: Idealized soil profile at ‘Site A,’ California (courtesy Caltrans). ............... 89
Figure 5.13: Stability analysis of lower slope with 15 x 15 grid definition. ................... 93
Figure 5.14: Stability analysis of lower slope with 50% reduction in fill φ angle. ......... 95
Figure 5.15: Stability analysis of lower slope with 18.75% reduction in fill cohesion... 96
Figure 5.16: Stability analysis of lower slope with vertical surcharge load increased to
238.65 kPa. .............................................................................................................. 98

xi
Figure 5.17: Stability analysis of lower slope with ‘Entry and Exit’ slip surface
definition.................................................................................................................. 99
Figure 5.18: Stability analysis of upper and lower slopes with ‘grid and radius’ slip
surface definition – Class-A Prediction................................................................. 101
Figure 5.19: Stability analysis of upper and lower slopes with ‘grid and radius’ slip
surface definition – Class-B Prediction. ................................................................ 102
Figure 5.20: Plantable geosynthetic reinforced (PGR) retaining wall in San Diego,
California. .............................................................................................................. 105
Figure 5.21: Placement of sand backfill to couple ShapeAccelArray to PVC casing... 105
Figure 5.22: Solar panel cabinet with the ShapeAccelArray earth station components.107
Figure 5.23: Fourteen months of displacement data from a 32 m (104 ft) long

W
ShapeAccelArray at the San Diego PGR site. ....................................................... 108
Figure 6.1: Instrumentation layout (plan view) for bridge replacement and realignment
IE
over the Champlain Canal (courtesy NYSDOT; Barendse, 2008). ....................... 110
Figure 6.2: Soil profile and location of the vertical ShapeAccelArray at Champlain Canal
(courtesy NYSDOT; modified from Barendse, 2008)........................................... 110
EV
Figure 6.3: Pre-installation preparation of the vertical ShapeAccelArray at the Fort Ann
site.......................................................................................................................... 111
Figure 6.4: Retrieval of ShapeAccelArray from dummy installation with fine sand
PR

backfill. .................................................................................................................. 112

Figure 6.5: Sand backfill placement for the vertical ShapeAccelArray at Champlain
Canal site. .............................................................................................................. 113
Figure 6.6: Comparison of observed displacements; Settlement of sand and sensor array
within PVC casing, Oct 2006 – April 2007, and measured lateral displacements,
April 2007 – July 2007. ......................................................................................... 114
Figure 6.7: Comparison of lateral displacement measurements between the
ShapeAccelArray system (line without markers) and traditional inclinometer (line
with markers); April 2007 – August 2007............................................................. 116
Figure 6.8: Settling of the vertical ShapeAccelArray within 50 mm (2 in) PVC casing
due to sand backfill voids and self-weight of sensor array.................................... 117

xii
Figure 6.9: Final profile of vertical ShapeAccelArray at Fort Ann, NY site; maximum
measured displacement of 15 mm (0.6 in) at 4 m (13 ft) depth............................. 118
Figure 6.10: Soil profile and instrument locations at Champlain Canal (courtesy
NYSDOT; modified from Barendse, 2008)........................................................... 119
Figure 6.11: Settlement gage installation. ..................................................................... 120
Figure 6.12: Relative locations of horizontal ShapeAccelArray and settlement gage
installation.............................................................................................................. 120
Figure 6.13: Horizontal ShapeAccelArray installation. ................................................ 121
Figure 6.14: Settlement profile from horizontal ShapeAccelArray (contour plot) and
nearest set of surface settlement plates (discrete lines). ........................................ 121
Figure 6.15: Time history of displacement from three ShapeAccelArrays and surface

W
settlement plate locations (consolidation data)...................................................... 122
Figure 6.16: Settlement profile from horizontal ShapeAccelArray (contour plot) and
IE
settlement plates located 13 m (42 ft) east (discrete lines).................................... 124
Figure 6.17: Survey targets on settlement plate pipes at Champlain Canal site............ 125
Figure 6.18: Settlement profile from the horizontal ShapeAccelArray (contour plot) and
EV
settlement plates located 13 m (42 ft) east without centerline readings (discrete
lines). ..................................................................................................................... 126
Figure 6.19: NYSDOT Geotechnical Engineering Bureau spreadsheet for settlement
PR

calculation.............................................................................................................. 128
Figure 6.20: Schematic of primary soil layer setup in FoSSA software simulations. ... 129
Figure 6.21: Settlement versus time curves at embankment centerline comparing results
from changes in drain resistance factor (Fr). ......................................................... 138
Figure 6.22: Settlement versus time curves at embankment centerline comparing results
from changes in soil disturbance factor (Fs). ......................................................... 139
Figure 6.23: Settlement versus time curves at embankment centerline comparing results
from combined changes in drain resistance factor (Fr) and soil disturbance factor
(Fs). ........................................................................................................................ 140
Figure 6.24: Settlement versus time curves at embankment centerline comparing results
from changes in horizontal coefficient of consolidation (ch)................................. 141

xiii
Figure 6.25: Settlement versus time curves at embankment centerline comparing results
from changes in horizontal coefficient of consolidation (ch) and increased
effectiveness of PVDs............................................................................................ 142
Figure 6.26: Settlement profile from the horizontal ShapeAccelArray (contour plot) and
‘Class-A’ FoSSA simulation results (discrete lines). ............................................ 144
Figure 6.27: Recommendation for increase in drain resistance factor (decrease in
drainage capacity) with time within FoSSA software. .......................................... 146
Figure 6.28: Resultant cracks from active landslide in Springville, NY. ...................... 148
Figure 6.29: Vertical ShapeAccelArray installation at Springville, NY landslide site.
Array housed in 25 mm (1 in) gray PVC conduit.................................................. 149
Figure 6.30: Three tremie hoses utilized at Springville, NY ShapeAccelArray to ease

W
retrieval. ................................................................................................................. 149
Figure 6.31: Displacement data profile from the ShapeAccelArray system at the
IE
Springville, NY landslide showing two shear zones. ............................................ 150
Figure 6.32: Approximate locations of slope indicator casing and ShapeAccelArray
shear depths. .......................................................................................................... 151
EV
Figure 6.33: Time history of 32 m (104 ft) ShapeAccelArray at Springville, NY
landslide (Barendse, 2009). ................................................................................... 152
PR

xiv
 ACKNOWLEDGMENT
I would like to start with a considerable thank you to my thesis advisor, Professor Tarek
Abdoun, for his advice and support throughout this project. Completing this thesis has
been a trying experience and I especially thank Professor Abdoun for seemingly
believing in me when I no longer believed in myself. There were many times that I was
sure I was lost but if I had not been here doing this work, I would have never tried to
find myself. I am thankful for the support of my thesis committee members, Professors
Ricardo Dobry, Mourad Zeghal, Thomas Zimmie and Anand Puppala. If it had not been
for Professor Zimmie’s commitment to encouraging women to pursue graduate degrees
in civil engineering, I would have never started on this journey. I am grateful to Dr.
Anand Puppala for his encouragement of my involvement with the Transportation

W
Research Board Committee on Soils and Rock Instrumentation. This has been an
extremely valuable experience and I am particularly thankful to George Machan for all
IE
the valuable knowledge he has shared with me. This research was supported by the
National Science Foundation, Grant No. CMS-0330043 and is gratefully acknowledged.
EV
I am forever grateful to my parents and my brother for a lifetime of love and
encouragement. I am extremely fortunate for everything they have given me and can
never thank them enough for the support they have offered me, particularly in the last
two years. I would also like to thank my friends, Jillian Blake, Elizabeth Chrestler,
PR

Lindsay Bassett, Katie Muller, Joanna Mason, Michael Triller and Diane Saunders, for
their never-ending morale support and for being such an important and wonderful part of
my life. I am extraordinarily grateful to Mrs. Janet Pertierra and Mrs. Marcia Hartnett for
giving me something to look forward to during the workday and for listening to the same
frustrations over and over and over again. I truly would not have survived my time in
graduate school without them. I owe a great deal of gratitude to Mrs. Tasha McDonough
and Mrs. Kimberly Boyce for always lending me a sympathetic ear and sharing in a
healthy dose of sarcasm. I would also like to thank Dr. Pickett Simpson for being an
inspiration to me and for his friendship. I greatly appreciate the support and
collaboration I received from my former graduate student colleagues: Drs. Claudia
Medina, Javier Ubilla, Da Ha, Marcelo Gonzalez and Ahmed Elmekati. Dr. Inthuorn
Sasanakul and Anthony Tessari offered me a great deal of support through our shared

xv
work with the Center for Earthquake Engineering Simulation at RPI and I am especially
grateful for this. During his graduate studies at RPI and as the teaching assistant for my
soils courses, Dr. Usama El Shamy was a great example to me as an undergraduate
student and I am particularly grateful for all the things he taught me.
Dr. Lee Danisch and Mr. Terry Patterson of Measurand Inc. were especially
supportive of my participation on this project. It was a great pleasure to work alongside
them at all the field installations. Their knowledge and commitment to their work is
unique and special. Mr. Matthew Barendse of NYSDOT was an integral link to
practitioners interested in this instrumentation. I especially thank him for all the hard
work and time he devoted to the integration of this instrumentation to NYSDOT
projects. Throughout my work on several different projects with many different

W
organizations, I have collaborated with so many engaging and knowledgeable
individuals without whom much of this work would not have been possible. In particular
IE
I would like to acknowledge: Mark Pitman, Tom Shantz, Professor Michael Symans, Dr.
Damon Reigles, Dr. Sabanayagam Thevanayagam, Dr. Nurhan Ecemis, Deh-Jeng Jang,
Ronnie Gu, Gem-Yeu Ma, Raghudeep Bethapudi and all the Caltrans and NYSDOT
EV
drillers and maintenance staff who participated in these field installations.
PR

xvi
 ABSTRACT
Real-time monitoring of civil infrastructure provides valuable information to assess the
health and condition of specific associated systems. An important effort for the future of
civil engineering is thus the development of instrumentation capable of accurate real-
time monitoring of geotechnical systems. The use of Micro-Electro-Mechanical Systems
(MEMS) accelerometers in geotechnical instrumentation is relatively new but on the
rise. This thesis describes a new MEMS-based system for in situ deformation and
vibration monitoring. The system has been developed in an effort to combine recent
advances in the miniaturization of sensors and electronics with an established wireless
infrastructure for on-line geotechnical monitoring. The concept is based on triaxial
MEMS accelerometer measurements of static acceleration (angles relative to gravity)

W
and dynamic accelerations. The dynamic acceleration sensitivity range provides signals
proportional to vibration during earthquakes or construction activities. This MEMS-
IE
based in-place inclinometer system utilizes the measurements to obtain three-
dimensional (3D) ground acceleration and permanent deformation profiles up to a depth
EV
of one hundred meters. Each sensor array, which can be used vertically or horizontally,
or group of arrays can be connected to a wireless earth station to enable real-time
monitoring as well as remote configuration. This thesis provides a technical assessment
of MEMS-based in-place inclinometer systems for geotechnical instrumentation
PR

applications by reviewing the sensor characteristics and providing small- and full-scale
laboratory calibration tests.
Descriptions and validations of recorded field data from a bridge replacement site
and two unstable slopes are included. This new instrumentation system was also
included in full-scale laminar box tests of level and sloping saturated fine sand deposits.
These full-scale tests provided a means of evaluating measured acceleration data. In all
cases, data recorded with the developed in-place inclinometer system is compared to
data measured with state-of-the-practice instrumentation. These comparisons were
extremely favorable and justified the future use of this instrumentation for many
geotechnical applications. This study also includes a practical evaluation of
commercially available geotechnical software predictions as compared to measured site
data.

xvii
 1. Introduction and Historical Review
The health and state of the aging and overburdened civil infrastructure in the United
States has been subjected to renewed scrutiny over the last few years. The American
Society of Civil Engineers reports that this situation threatens the economy and quality
of life in every state, city and town in the nation. As one example, the United States
Army Corps of Engineers noted in early 2007 that nearly 150 United States levees pose
an unacceptable risk of failure during a major flood. Landslides during the rainy season
occur in nearly every state. The U.S. Geologic Survey (USGS) estimates that landslides
cause more than $2 billion in damages and about 25 to 50 deaths each year in the United
States (Figure 1.1). Potentially unstable slopes that may affect critical infrastructure
require monitoring systems that can provide an alert if movement occurs (Kane and

W
Beck, 1999). Such real-time geotechnical monitoring systems to detect unstable slope
conditions have been implemented by national, state and local agencies (Baum et al.,
IE
2005). Due to implementation costs, data transmission methods and power requirements,
the extent and reliability of current landslide monitoring systems are often limited.
EV
PR

Figure 1.1: Aerial view of the 1995 La Conchita, CA landslide (photograph by

Mark Reid, U.S. Geological Survey, 1995).

1
 Earthquakes are another significant source of geotechnical problems. In addition to
triggering landslides and failures of earth embankments, the phenomenon of liquefaction
of loose, water-saturated sands and other granular soils is another major hazard
specifically related to earthquakes. This liquefaction can lead to ground failure,
significant permanent vertical deformations and lateral spreading. These ground failures
are usually associated with very costly damage to port facilities, bridges, dams, buried
pipes, houses and buildings of all types. For instance, the 1971 San Fernando, California
earthquake caused more than five hundred million dollars in damage (NRC, 1982),
which included a failure of the San Fernando dam (Figure 1.2). The 1995 HyogoKen
Nanbu earthquake in Kobe, Japan caused more than one hundred billion dollars in total
damage, with about ten billion dollars of that total attributed to soil liquefaction and

W
ground deformation. Similar losses associated with failures of soil systems continue to
grow in the United States and elsewhere because of increased urban development in
IE
hazard-prone areas. The control and mitigation of the effects of these failures requires a
better understanding of the field response of soil systems. The use of advanced in situ
real-time monitoring devices is a key to this understanding.
EV
PR

Figure 1.2: San Fernando dam failure during the 1971 San Fernando earthquake
(photograph by R.E. Wallace, U.S. Geological Survey, 1971).

2
1.1 Current Geotechnical Instrumentation
Real-time monitoring of active soil systems can provide early indications of a
catastrophic movement. It can also enable a better understanding of the underlying
geotechnical phenomena that can be used in modeling and predicting similar failures.
Many of the United States’ Departments of Transportation and federal agencies spend
significant resources monitoring possible slope failures using manual slope
inclinometers or slope inclinometer arrays. Vertical slope inclinometers are instruments
for measuring relative horizontal displacements affecting the position of a flexible guide
casing installed in the ground or along a structure (AASHTO, 1980). This instrument has
been in use for geotechnical monitoring for more than fifty years. The theoretical
background of inclinometer probe data is described in ASTM (2005), Dunnicliff (1993),

W
Mikkelsen (1996) and Wilson and Mikkelsen (1977).
Inclinometer probes (see Figure 1.3) typically contain sensors that measure the
IE
amount of tilt of the probe in a guide casing in two perpendicular planes; with this
information and the known length of the probe, displacement magnitudes and directions
EV
can be calculated. The bottom end of the guide casing should be a stable reference and
must be installed beyond the displacement zone. The position of the casing is determined
by taking 50 cm (2 ft) increment measurements, which is the distance between the
wheels, over the length of the guide casing. Relative displacement over time is
PR

determined by repeating measurements at the same depths and comparing data sets
(Figure 1.4). Cornforth (2005) provides detailed guidance on collecting initial data sets
and subsequent monitoring measurements.

Figure 1.3: Slope inclinometer probe, cable and readout box (Slope Indicator,
2010).

3
 W
IE
Figure 1.4: Data collection with traditional probe inclinometer.

Slope inclinometers are the most common method of long-term slope monitoring,
EV

however surveys of the slope surface, tiltmeters, extensometers, and Time Domain
Reflectometry (TDR; Dowding and O’Connor, 2000) systems are also used to determine
direction, rate of displacement, and location of the failure plane (Dunnicliff, 1993).
PR

Monitoring with traversing probe inclinometers provides limited datasets and has proven
to be quite expensive in the long run. Another limitation of the traversing inclinometer
probe where multiple shear zones exist is that an upper deformation zone could cause the
guide casing to bend excessively and obstruct the probe from being lowered to measure
deeper shear zones.
The guide casing is typically installed vertically to measure horizontal ground
movements. Horizontal installations are also possible and use a specialized probe to
monitor settlement (Machan and Bennett, 2008). Horizontally installed inclinometer
casing can be used to measure differential settlement under embankments. This type of
installation prevents interference with embankment fill construction and compaction.
Interference with these processes is common with traditional settlement plates that use
vertical riser pipes for measurement. Puppala et al. (2008) describe the installation of an

4
inclinometer system with a closed end and a pulley system underneath a pavement.
Representative diagrams of this type of installation are presented in Slope Indicator
Company (2006) and FHWA (1998).
A slope inclinometer array has an inclination sensor attached to each segment of the
borehole that monitors the inclination at depths continuously with time. While slope
inclinometer arrays (see Figure 1.5) can provide real-time monitoring with good angular
resolution capability, they are very expensive and inflexible. This high cost makes it
impractical or impossible to install multiple arrays on a single slope, thus severely
limiting the ability of conducting continuous real-time monitoring. Myers et al. (2000)
present a landslide case history that describes an automated monitoring system, which
utilizes in-place inclinometers and tiltmeters to track ongoing creep movements and

W
provide early warning of the need to make adjustments to a deep elevator shaft
constructed through a landslide basal shear zone.
IE
The extent of damage due to seismic excitation is often directly correlated to local
site conditions. This correlation was evident during recent earthquakes in the form of
motion amplification and liquefaction-induced ground deformations, for example,
EV
Mexico City earthquake (Seed et al., 1987), San Francisco during the 1989 Loma Prieta
earthquake (Seed et al., 1990), and 1995 Kobe earthquake (Bardet et al., 1995; Comartin
et al., 1995; Sitar, 1995). The associated mechanisms of ground response are being
PR

monitored through a worldwide network of sites instrumented with accelerometer

downhole arrays, often supplemented at liquefiable sites with pore pressure sensors and
in some cases also with slope inclinometer arrays (Elgamal et al., 2001). Downhole
acceleration records provide direct insight into the response mechanisms of instrumented
layers within the ground, see Figure 1.6.
In the United States, early downhole data sets were recorded at the San Francisco
Bay area (Aisiks and Tarshansky, 1969; Joyner et al., 1976; Johnson and Silva, 1981),
and at Union Bay in Seattle, Washington (Seed and Idriss, 1970; Dobry et al., 1971).
During the 1980s, data from downhole seismic arrays that include pore-pressure
piezometers became available, for example, in Owi Island, Japan (Ishihara et al., 1987),
Wildlife Refuge, CA (Holzer et al., 1989) and Lotung, Taiwan (Tang, 1987). In spite of
the growing awareness of the importance of accelerometer downhole arrays (Elgamal et

5
al., 1995; Zeghal et al., 1995; Elgamal et al., 1996), particularly when combined with
pore pressure sensors to capture information on the poorly understood phenomena of
liquefaction and lateral spreading (Youd and Holzer, 1994; Zeghal and Elgamal, 1994),
these arrays still remain scarce due to their high cost. To date, geotechnical site response
prediction in the United States does not usually incorporate the most advanced soil
models due to a lack of confidence in these models for soft soils in the nonlinear range.
Model validation through earthquake-induced stress and strain measurements from
densely instrumented sites remains an important goal. Such dense arrays, with
measurements at one meter or sub-meter vertical spacing, would require new MEMS-
based instrument designs (de Alba et al., 2006).

W
IE
EV
PR

Figure 1.5: Slope Inclinometer Array (Slope Indicator, 2010).