HEALTH MONITORING AND MANAGEMENT OF INTEGRATED CIVIL INFRASTRUCTURE

1
HEALTH MONITORING AND MANAGEMENT OF INTEGRATED
CIVIL INFRASTRUCTURE

• Kaan Ozbay, Ph.D., Associate Professor
• Hani Nassif, Ph.D. , Associate Professor
• Department of Civil and Environmental Engineering

2
INTRODUCTION

• Structural health monitoring (SHM) and management
  of infrastructure facilities are becoming
  increasingly popular in the U.S. and Europe
• Remaining life of aging infrastructure.
• Performance of new and advanced materials, e.g.,
  fiber reinforced polymer, high performance and
  self-compacting concrete, advanced composites,
  etc.
• Calibration and validation of design codes.
• Monitoring the security and safety of various
  civil infrastructures.

3
SHM
Data Acquisition System
Modem
Threshold
4
NJ Turnpike Delaware River Bridge
Focus of Fatigue Study
5
Doremus Avenue Bridge

• Average Daily Traffic 8900 veh/d
• Truck Traffic 40
• Equiv. Truck Weight RMC, Weq 43.5k

6
OBJECTIVES

• Develop a truck live load model based on actual
  truck weight measurement using a weigh-in-motion
  (WIM) system.
• Develop a fatigue load model to estimate the
  remaining life of the bridge.
• Verify Bridge Design Specifications
  (LRFD-AASHTO)
• Girder distribution factors GDF
• Dynamic Load Amplification
• Multiple Presence events
• 5. Verify New Jerseys deflection limits and
  develop provisions for design and formulation for
  deflection as well as vibration serviceability
  limit states.
• 6. Evaluate the dynamic characteristics of
  bridge substructure.

7
Sensor Wiring Layout
8
(No Transcript)
9
VWSG
Sensors Verification
Running the Cables
Data Logger
10
Sensor Network
11
Sensor Network
Girders 4 8
Typical Girder
12
Vibrating Wire Strain Gages

• VWSG
• 48 Sensors in each Stage of Construction

13
Laser Doppler Vibrometer
14
Long-Term Truck Monitoring Program

• 60 permanent WIM sites throughout NJ
• 8 sites near Doremus Avenue selected for detailed
  analysis
• 5 years of data (typical), with 10 years at
  certain sites

Typical Site with Piezo WIM Sensors
WIM sites unmarked undetected by drivers
15
Weigh-in-Motion
16
Weigh-In-Motion
Saw-cut WIM location
Clean using air compressor
Place sensors with spacers
Seal using epoxy
17
Weigh-In-Motion (WIM) System
Data Logger
Bending Plate
Magnetic Loops
18
Live Load Statistics

19
TWO TRUCKS FOLLOWING
TWO TRUCKS SIDE-BY-SIDE
SINGLE TRUCK
20
MATERIAL PROPERTIES
21
BRIDGE FINITE ELEMENT MODELS
Shell-Beam (S-B) Model

• Shell-Shell(S-S)
• Model

Shell Element
Beam Element
22
Unloaded
Loaded (Span 1)
Loaded (Span 3)
Loaded (Span 2)
23
DYNAMIC LOAD TEST
24
3-D Dynamic Model
Vehicle Model
5-axle semi tractor-trailer dynamic model is
composed three components, (1) tire, (2)
suspension, (3) truck body.
25
Model Validation
Comparison for Most Loaded Girders
Span 1
Span 2
Span 3
26
Basis for Fatigue Design

• NCHRP Report 286 Evaluation of Fatigue Tests
  and Design Criteria for Welded Details. Keating
  and Fisher 1986.
• Constant variable amplitude tests on full scale
  beams.
• SN-Curves based on lower 95 CI for full scale
  lab tests.

27
The To-From Rainflow Matrix
Proposed Method Rainflow Extrapolation
Can be used for seasonal effects To predict
extreme events (magnitude).
(Johannesson 2005)

• The Min and Max for each cycle is itemized in the
  To-From rainflow matrix
• Diagonal is always zero
• Elements of each sub-diagonal have equal stress
  ranges.

(Johannesson 1999)
28
Illustration of RF ExtrapolationHybrid Approach
Few Measurements
Infinite Measurements
(Johannesson 2005)

• Two methods extreme value theory kernel
  smoothing.
• For the limiting RF Matrix, G, as N?8 the shape
  stabilizes.
• For any period the RF matrix is GN

29
Instrumentation
Strain Transducer
WIM System
Fatigue System
WIM Sensors
30
Simulation Flowchart
31

• Propagation Characteristics of Dynamic
  Information Collected by In-Vehicle Sensors in a
  Vehicular Network
• This study has two major goals
• to test the feasibility and accuracy of getting
  data from on-board and external sensors
• to test the feasibility of disseminating this
  information over a peer-to-peer network using a
  microscopic traffic simulator namely, Paramics.

32
(No Transcript)
33
Location of bumps on GPS Map
34
Microscopic traffic simulation model
35
Microscopic traffic simulation model
36
NJ Turnpike Simulation Model

• Transfer of geometric and operational data from
  ArcView GIS and TransCAD models
• TransCAD network ? ArcView shape file
• ArcView GIS (shape file) ? PARAMICS network
  (Tool S2P (Shape to Paramics) software)
• Size of the NJTPK simulation model network
• With alternate routes 4244 nodes, 8800 links
• Without alternate routes 2766 nodes, 6042 links

37
NJ Turnpike GIS Network
38
Input Data

• Vehicle transaction data of all vehicles between
  2004 and 2006.
• The transaction data include
• Vehicle origin (entrance toll plaza)
• Entrance time, date
• Entrance lane
• Vehicle type (passenger car or truck)
• Vehicle destination (exit toll plaza)
• Exit time, date
• Exit lane
• Travel time

39
Model Network

• In NJ Turnpike network, there is only a single
  route between two exits. Therefore, zones are
  placed outside each toll plaza in the model
• There are 26 entry and exit toll plazas
• Computation Time 5-6 hrs. for 4 hour simulation

40
(No Transcript)