Bridge Instrumentation and Nondestructive Load Testing for Long Term Structural Health Monitoring

Transcription

1 Bridge Instrumentation and Nondestructive Load Testing for Long Term Structural Health Monitoring NSF-PFI Grant: Whatever Happen to Long-Term Bridge Design? A case study: Vernon Avenue Bridge Masoud Sanayei, Tufts University Erin Santini Bell, University of New Hampshire Brian Brenner, Fay, Spofford and Thorndike, INC. RICC 2010, Research to Reality Northeastern University October 19, 2010

2 Presentation Outline Background Instrumentation Load Test (Static, Imaging, and Dynamic) Modeling (ERM & EDM) Model Calibrations New Bridge Design Paradigm Conclusions 2

3 National Science Foundation Partnership for Innovations Stimulate the transformation of knowledge created by the research and education enterprise into innovations that create new wealth; build strong local, regional and national economies, and improve the national well-being. Broaden the participation of all types of academic institutions and all citizens in activities to meet the diverse workforce needs of the national innovation enterprise. Catalyze or enhance enabling infrastructure that is necessary to foster and sustain innovation in the long-term. This material is based upon work supported by the National Science Foundation under Grant No Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation. 3

4 Motivation 1 in 3 Bridges Near End of Design Life (FHWA, 2009) Entering Rebuilding Phase Increased Public Awareness Opportune Time to Consider Changes to Design Paradigm Tobin Memorial Bridge 4

5 Vernon Avenue over the Ware River Bridge Barre, MA WM Barre Landfill Image Courtesy of Google Maps 5

6 Vernon Avenue Bridge (VAB) 6 Steel Girders with Reinforce Concrete Deck Beams and Deck are Composite 3 Continuous Spans Field splice in center span 150 Feet Long with 75 ft Center Span Vernon Ave Bridge, Looking South 6

7 Instrumentation Plan Girder Section at Station Concrete Temperature 12 LEGEND Steel Temperature TEMPERATURE GAUGE STRAIN GAUGE NORTH SECTION GIRDER 2 TILTMETER ACCELEROMETER PRESSURE CELLS Strain Gauges 7

8 Instrumentation Strain Gauges 8

9 Instrumentation Tiltmeter Accelerometer 9

10 Instrumentation Pressure Cells Concrete Temp 10

11 Instrumentation Summary Quantity Instrument Type 100 Strain Gauges 36 Girder Temperature Sensors 30 Concrete Temperature Sensors 16 Accelerometers 16 Tiltmeters 3 Ambient Temperature Sensors 11

12 Instrumentation by Geocomp 12

13 Concrete Pour 13

14 Non-Destructive Load Tests 3 Types of Tests Stop Locations Tests Crawl Speed Tests Impact Tests 3 Lanes West Shoulder Center East Shoulder 3 Trials of each 27 total runs One 72 Kips Truck 14

15 Load Test Lane and Stop Location Plan LANE 1 LANE 2 LANE 3 SOUTH ABUT SOUTH PIER NORTH PIER NORTH ABUT 15

16 Truck Tracking Using AMTS (Automated Motorized Total Station) 16

17 Bridge Modeling ERM - Model Using SAP2000 Solid and Shell Elements with fine mesh For both models: Include Exact geometry Include System Behavior Updated to Reflect As-built Condition EDM - Model Using SAP2000 Bridge Information Modeler (BrIM) with shell and frame elements 17

18 Enhanced Researcher Model (ERM) Modeling Requirements Detailed finite element model that captures bridge performance Exact geometry Including bridge components such as diaphragms, safety cubs, and parapets Girders: Shell Elements Deck: Solid Elements Boundary Conditions: Neoprene Pads 18

19 ERM Calibrated Model Crawl Speed Load Test Strains 19

20 ERM Load Rating Factors, ASD (a) Inventory (b) Operating Vernon Avenue Bridge Rating Factors using ERM 20

21 Enhanced Designer Model (EDM) Modeling Requirements SAP2000 Bridge Information Modeler (BrIM) Girders: Frame Elements Deck: Shell Elements Boundary Conditions: Roller-Roller-Roller-Pin 21

22 ERM Calibrated Model Stop Location Load Test Strains Bottom Flange of Girder #2 at North Pier (Station 8) 22

23 EDM Modeling for Bridge Management Similar ratings for Exterior Girders Interior Girders gain extra capacity in EDM Rating System Behavior Stiffer Exterior Girder picks up more load Lowers rating of Exterior Girders Increases rating of Interior Girders Load Rating LRFR vs. Baseline EDM Rating by Girder # EDM Rating 2 LRFR Rating Girder Number 23

24 DIC Data for Independent Evaluation Deflection (in) Vertical Deflection of Vernon Ave Bridge at South Span Truck in West Lane Time (s) Vertical Deflection of Vernon Ave Bridge near Midspan Truck in West Lane DIC Original Updated Deflection (in) DIC Original Updated Time (s) 24

25 Dynamic Testing (ongoing) APS Dynamics Shaker 9 excitation locations 9 Wilcoxon Accelerometers Use of transfer functions for FEM updating 25

26 Advantages and Disadvantages of Two Different Types of Models: ERM & EDM ERM - Solid/Shell Model EDM Shell/Frame BrIM Model Advantages Disadvantages Advantages Disadvantages Detailed behavior Beam assumptions Direct access to element stresses for strain calculations Need to calculate stress / strain from forces / displacements Obtain stress directly Need to define effective width for bridge composite section to determine bridge section forces and moments for stress calculations Complex modeling/adjustments Higher computational time Need to create programs to obtain standardized outputs Ease and speed of modeling/adjustments Lower computational time Standardized outputs (including bridge influence lines) 26

27 Conclusions Detailed 3D FEM of typical highway bridges is feasible to a high degree of accuracy Truck load testing performed on a newly constructed bridge can provide highly reliable strain data for calibrating baseline FEM s. DIC data verified the FEM independently Calibrated FEM and load test strain data matched closely There are pros and cons for using ERM and EDM. A calibrated FEM can be used as an effective tool for bridge management 27

28 The article, Whatever Happened to Long-Term Bridge Design? From the abstract: "The Chairman of the Board of one of the leading engineering firms in the nation ponders the future of American bridges. In today s designing of bridges, Thomas R. Kuesel points out that light, thin, elastic and graceful are the adjectives come to mind. He worries that old fashioned concepts of stiff, rugged and durable construction are not actively pursued by present day designers. He states the case for a long useful life for bridges and stresses the need for endurability. Ref:

29 Current AASHTO Bridge Design Process Ends at commission of bridge. Design calculations recorded and submitted to owner Generally no consideration for SHM and long-term management Structural modeling not generally part of submission 29

30 Typical AASHTO Bridge Inspection Every 2 years, Visual inspection of each member Condition Ratings (AASHTO Manual for Bridge Evaluation, 2008) Subjective Process Although Inspectors are well trained, results can vary MHD Bridge Inspection Report, Vernon Ave Bridge, Jan

31 Central Thesis How is long term design is addressed in the design process? Leverage advancing technology (instrumentation, analysis, data management, remote sensing) to improve the bridge design process that currently focuses on opening day, but not the 75 years that follows opening day.

32 New Bridge Design Paradigm Go beyond opening day design Creation of baseline model Integrate baseline modeling and model updating through bridge service life Modify reactive mode of bridge management Develop useful and cost-effective bridge instrumentation plan Deploy practical nondestructive tests Have continuous feedback from bridge about current structural health 32

33 Baseline Model Propose the development of a "baseline" model of the bridge during design and construction Verify initial design assumptions Live and Evolve with the bridge as a data management tool.

34 Research Motivation Leverage current technologies Bridge design today is elemental Bridge design is complete on opening day Design intelligence is not readily available during life of bridge Address long term behavior of bridges during initial design

35 Our Focus We are focusing on: measuring objective, verifiable, and useful data for bridge SHM providing practical data useful to bridge owners providing data that will be easy to use by state agencies for bridge management

36 Thank You for Listening Acknowledgements NSF-PFI Grant No Whatever Happened to Long Term Bridge Design? Program Director: Dr. Sara Nerlove MassDOT Bridge Construction Town of Barre, MA Bridge Management and Owner Fay, Spofford & Thorndike, Inc. Bridge Design Geocomp Corporation Instrumentation E. T. & L. Corp. Bridge Contractor High Steel Structures, Inc. Steel Fabricator Atlantic Bridge and Engineering, Inc. Steel Erector Bridge Diagnostics, Inc. Bridge Testing 36

37 Backup Slides 37

38 Objective Function for Manual Model Updating Scalar Objective Function for Crawl Speed Test 38

39 Material Property Model Updating Original Updated Units Density, w c kg/m 3 Unconfined Compressive Modulus of Elasticity, E MPa MPa 1. AS-BUILT MATERIAL PROPERTIES 2. CONCRETE SAFETY CURB Degree of Freedom Original Updated Units Axial, U z Fixed kn/mm Shear, U y Free kn/mm Shear, U x Free kn/mm Rotation, R x Free 1.77E+06 kn-mm/rad Rotation, R y Free 1.77E+06 kn-mm/rad Torsion, R z Free 1.77E+05 kn-mm/rad 3. BOUNDARY CONDITIONS 4. DECK REINFORCEMENT 39

40 Current AASHTO Bridge Design Process Typical Design Assumptions Elemental Design Effective width of slab, b e Composite Behavior Distribution Factor, mg, and Impact Factor, IM PP uu,tttttttttt = ηγγ(mmmm)(1 + IIII)PP tttttttttt (Adapted from AASHTO LRFD,2008) 40

41 Typical AASHTO Bridge Inspection Load Ratings: 1. Measured Section Loss - From Inspection 1. Calculate New Section Capacity: - Based on Reduced Section Properties 3. Calculate Load Rating Factor: - Ratio of LL Capacity to LL Applied Design Calculation Methodology MHD Bridge Inspection Report, Vernon Ave Bridge, Jan

42 Modeling for Bridge Management: AASHTO Load and Resistance Factor Rating (LRFR) RRRR LLLLLLLL = CC γγ DDDDDDDD γγ LLLL LLLL(1 + IIII) (Adapted from AASHTO Manual for Bridge Evaluation, 2008) RRFF MMooddeell = CC DDLL MMooddeell LLLL MMooddeell(1 + IIII) 42

43 Modeling for Bridge Management: AASHTO Load and Resistance Factor Rating (LRFR) RRRR LLLLLLLL = CC γγ DDDDDDDD γγ LLLL LLLL(1 + IIII) (Adapted from AASHTO Manual for Bridge Evaluation, 2008) γ DC =1.25; γ LL = dependent on rating type 2 Ratings: o Inventory Rating: Based on LL that can safely utilize bridge indefinitely. RF INV => γ LL =1.75 o Operational Rating : Based on maximum permissible LL. RF OP => γ LL =1.25 Elemental Approach LL and DL Distribution Factors 43

44 Modeling for Bridge Management: EDM Rating RRFF MMooddeell = CC DDLL MMooddeell LLLL MMooddeell(1 + IIII) γ DC =1.25 and γ LL =1.75 factored into applied loads DL from self weight of bridge LL applied to mimic 2008 LRFD Bridge Design Specs for Worst-Case Traffic Loading Scenario Accounts for System Behavior No Distribution Factors 44