Non Fouling Monitoring and Performance of Railway Track and Transitions

Transcription

1 1/50 Non Fouling Monitoring and Performance of Railway Track and Transitions Timothy D. Stark Stephen T. Wilk University of Illinois 34 th Annual Geotechnical Seminar GEO-Omaha 2017 Friday, February 10, 2017 Omaha, Nebraska

2 Outline 2/50 Introduction Non Fouling Instrumentation Examples Sites Site #1 Site #2 Wireless Instrumentation

3 Transition Zones Rapid change in track structure Differential movements Locations of increased track maintenance Root cause of differential movement Settlement of approach Impact loads in approach Reduced performance balance usually present or developing Need cost effective remedial solutions 3/50

4 Outline 4/50 Introduction Non Fouling Instrumentation Examples Sites Site #1 Site #2 Wireless Instrumentation

5 Non Fouling Instrumentation Objectives Collect data without fouling 10 to 30 minute setup/takedown Mobile 5/50

6 Video of Ties #2 and #3 6/23 Stark et al. (2015)

7 Non Fouling Instrumentation 7/50 Two systems 3 high speed video cameras 8 piezo electric accelerometers Tie displacements Insight into transient movements Tie accelerations Insight into loading and impacts

8 High Speed Video Cameras 8/50 Type: Casio EX FH100 Resolution: 448 x 336 pixels FPS: 240 fps (1000 fps maximum) Accuracy: 0.25 to 0.5 mm Process: Identify target, select color, binarize

9 High Speed Video Cameras 9/50 Process with MATLAB script Track centroid of 2 x2 orange target Remove vibrations with stake

10 Tie Accelerations Piezo electric accelerometers vibrations & impacts 500g, up to 20,000 Hz 4,000 Hz sampling rate Measure tie displacements vibration, and impacts v. DC = displacements 1. Wheel rail impacts 2. Wheel rail vibrations 3. Superstructure impacts 4. Tie loading Each factor produces unique signature Analyze in time and frequency domains Filter at 500 Hz 5. Track and tie vibrations 6. Tie ballast impacts 7. Tie displacement 10/50 Superstructure impacts (30 to 500g) Tie loading (~4g) Figure 4.7 Figure 4.9

11 Filtering Remove noise & isolate particular frequencies Multiple tie acceleration mechanisms High frequency vibrations (noise): > 500 Hz Impacts & low frequency vibrations: Hz Tie displacements: <30 Hz 8 th order Butterworth filter cutoff at 500 Hz 11/50

12 Tie Accelerations Double Integration 12/50 Estimate peak to peak displacements Compare/validate with cameras Must remove low frequency accelerometer noise

13 Tie Accelerations Procedure 13/50 1. Bandpass filter at 0.75 and 30 Hz 2. Integrate 3. Bandpass filter at 0.75 and 30 Hz 4. Integrate

14 Tie Accelerations Typical Results 14/50 Video camera v. Double integration Comparable peak to peak displacement 13 mm v. 15 mm (0.5 v. 0.6 ) Limitations Train speed (<25 mph) Video Camera Double Integration

15 Outline 15/50 Introduction Non Fouling Instrumentation Examples Sites Site #1 Site #2 Wireless Instrumentation

16 Site #1 Original Bridge Class 2 single mainline 30 MGT Open deck bridge Reoccurring track geometry defects Profile, crosslevel, and warp 16/50

17 Site #1 Remediation Selection Scheduled upgrade Concrete ballasted deck bridge Conversion of existing open deck Constructed new concrete ballasted deck Double mainline Approach remediation Approach #1: Geoweb Approach #2: Hot mixed asphalt (HMA) Approach #3: Geoweb Approach #4: Existing slurry grout Compare effects of remediation 17/50

18 Site #1 Remediation Selection 18/50 Geographical North Railroad EAST Approach #1: Geoweb Approach #4: Slurry/Grout Approach #2: HMA Approach #3: Geoweb

19 Site #1 Construction HMA (250 ft.): 6 of crushed aggregate for subgrade 8 of HMA 12 of AREMA #4A granite ballast Geoweb (50 ft.): 6 of crushed aggregate for subgrade Nonwoven geotexile 6 thick geoweb Fill geoweb with crushed aggregate 12 of AREMA #4A granite ballast 19/50

20 Site #1 Construction 20/50 Geoweb installation Approach #1 Packaged geoweb Approach #3

21 Site #1 Remediated Bridge 21/50 Geoweb Approach #1 Geoweb Approach #3

22 Site #1 Instrumented Approaches 4 Approaches monitored 3 Days of monitoring 16 November April July /50

23 Site #1 Approach #2 (HMA) Ballasted deck bridge Hot mixed asphalt (HMA) Accelerometers & video cameras 3 days of monitoring 23/50

24 Site #1 Typical Video Camera Results Approach #2 (HMA Varying displacements along track Hanging tie at end of structure C#1 (13 ft) C#6 (54 ft) 24/50

25 Site #1 Video Camera Results 25/50 A#1 (Geoweb) v. A#2 (HMA) v. A#3 (Geoweb) v. A#4 (Grout) A#2 (HMA) & A#3 (Geoweb): Hanging ties at edge of structure; less displacement away from abutment (approaches) A#1 (Geoweb) & A#4 (Grout) show consistent displacements A#1 & A#4 Existing line (prior loading) A#2 & A#3 New line (no prior loading)

26 Site #1 Accelerometers Approach #1 (Geoweb) v. Approach #2 (HMA) Similar results (under 5g) 26/50

27 Site #1 Summary Existing v. New line Existing line consistent displacements New line hanging ties at bridge end Potential subgrade settlement on new line Future monitoring required Every 6 months to year 27/50

28 Outline 28/50 Introduction Non Fouling Instrumentation Examples Sites Site #1 Site #2 Wireless Instrumentation

29 Under Tie Pad Installation Reoccurring track geometry problems Amtrak high speed line (110 mph) Tangent track, multiple closely spaced bridges 29 UTPs installed on Upland Street Southern Track #2 Upland Street Northern Track #3 previously instrumented 29/50 Google Earth

30 Track #3 Instrumentation Upland Street Northern Track #3 Permanent & transient displacements Strings of LVDTs Identify location of movement with depth Results 14 mm/yr at Upland (15 ft.) Detrimental movement at tie ballast gap 60 ft. 15 ft. 30/50 Track #3 Permanent Displacements Track #3 Transient Upland (60 ft.) Track #3 Transient Upland (15 ft.) 1 mm/yr 14 mm/yr

31 Tie acceleration 10g v. 30g Impact from closure of tie ballast gap Upland Avenue 31/50 LVDT #1 & Tie Accelerations Figure 4.27(a)

32 Under Tie Pad Installation Reoccurring track geometry problems Amtrak high speed line (110 mph) Tangent track, multiple closely spaced bridges 29 UTPs installed on Upland Street Southern Track #2 Upland Street Northern Track #3 previously instrumented 32/50 Google Earth

33 Under Tie Pads (UTPs) Potential Benefits Improved distribution of tie load Reduced tie ballast contact stress Increased vibration damping Installation Relatively inexpensive 33/50

34 Non Fouling Instrumentation Track #2 UTPs 12 August 2015 Transient displacement and acceleration Measure multiple locations along track UIUC & Amtrak instrumentation 346 days after UTP installation Three displacement measurements Direct: video cameras (UIUC) & bending beams (Amtrak) Indirect: accelerometers (UIUC & Amtrak) Compare results of each instrument with distance from bridge Accelerometer (UIUC) Bending Beam (Amtrak) 34/50

35 Instrumentation Layout 35/50

36 Bending Beams 36/50 Amtrak 20 ft 8 accelerometers & Beam 22 ft 1 Bending Beam

37 Instrumentation Comparison Rough agreement (<2 Hz & >30 Hz) Tie #4: 4.5 mm v. 4.1 mm (0.18 v in) Camera Tie #10: 0.8 mm v. 1.4 mm (0.03 v in) Bending Beam Differences in analysis methods 37/50 Video Camera Tie #4 (8 ft.) Accelerometer Tie #4 (8 ft.) Bending Beam Tie #10 (20 ft.) Accelerometer Tie #10 (20 ft.)

38 Displacements with Distance Instrument comparison Differences in measurement techniques With distance Greatest displacements near approach 38/50

39 Accelerations with Distance UTP Response Tie #1 (2 ft.) > 5g Multiple tie acceleration mechanisms High frequency vibrations (noise): > 500 Hz Impacts & low frequency vibrations: Hz Tie displacements: <30 Hz Raw Data No Filter 500 Hz Filter 30 Hz Filter 39/50 ~30g ~30g ~2.5g Frequency Domain

40 Accelerations with Distance Tie #1 (2 ft.) tie ballast impacts: 25 to 30g Loading & vibrations: 10 to 15g Tie displacement: <3g Similar response to displacements No geometry since Track #2 UTPs 40/50

41 Comparison with Track #3 Similar open track Better approach support Track #3 Comparison 41/50

42 Outline 42/50 Introduction Non Fouling Instrumentation Examples Sites Site #1 Site #2 Wireless Instrumentation Cameras Accelerometers

43 Non Fouling Track Assessments Prototype system Non Fouling Mark rail and/or tie Set up cameras 20 feet away from track Monitor wirelessly 43/50

44 Wired v. Wireless 44/50

45 Wired v. Wireless 45/50 Wireless (30 fps)

46 Wired v. Wireless 46/50 High speed (240 fps)

47 Non Fouling Assessments Alternative method Attach camera to van (if access is possible) Monitor wired or wirelessly 47/50

48 Non Fouling Assessments Remove subjectivity Developing software red, yellow, green displacements 48/50

49 Non Fouling Assessment 49/50

50 Summary Non Invasive/Non Fouling Monitoring Assess Transitions, Fouled Ballast Areas, Remedial Measures. Upgrading to wireless Long Term Monitoring of Track assess impact of environmental conditions loading and speed Balance Displacements between Approach & Transition Bridge Ballasted Deck, Mats, MSE Approach HMA, Geoweb, Geogrid, Wing Walls 50/50

51 51/50 Non Fouling Monitoring and Performance of Railway Track and Transitions Timothy D. Stark Stephen T. Wilk University of Illinois 34 th Annual Geotechnical Seminar GEO-Omaha 2017 Friday, February 10, 2017 Omaha, Nebraska