On the Use of Ground Penetrating Radar to Detect Rebar Corrosion in Concrete Structures

Transcription

1 On the Use of Ground Penetrating Radar to Detect Rebar Corrosion in Concrete Structures David Eisenmann, CNDE, ISU Frank J. Margetan, CNDE, ISU Shelby Ellis, ISU This work is supported by the Iowa DOT Midcon 2017 Conference August 2017

2 Outline 1. Objective. 2. GPR Barrier Rail Inspections Old and New. 3. Laboratory measurements using a concrete phantom : Effect of rebar orientation and metal loss on GPR signals. 4. Summary

3 Objective The goal of this project is to develop a bridge inspection technique that will increase safety while reducing overall costs. The inspection technique is aimed at identifying the unsafe condition where corrosion has thinned the reinforcing steel that attaches the side barrier rail to the bridge deck to the point where it can no longer withstand the necessary impact. By quantifying corrosion, the test will enable reinforcement of only those barrier rails that are no longer safe.

4 10 nanoseconds 6 inches Inspection Basics Raw GPR Data Horz. Line Scan of Wall 4 feet FW+DC FW+DC Vert. Rebar Vert. Rebar Wave length in concrete ~ 2.5 inches 1600 MHz

5 Outline 1) Objective. 2) GPR Barrier Rail Inspections Old and New. 3) Laboratory measurements using a concrete phantom : Effect of rebar orientation and metal loss on GPR signals. 4) Summary.

6 East 202 West Field Trial at a Local Highway Bridge Separate magnetic testing equipment Separate magnetic testing equipment Assumptions: (1) Corrosive thinning will reduce the amplitude of the rebar signal. (2) Larger signal reduction if antenna is closer to the thinned region. 2 x 4 spacer US 35 GPR equipment cart & antenna 1 2

7 Field Trial at a Local Highway Bridge Time (nanoseconds) Scan Distance, X (feet) Antenna on roadbed (Antenna closer to thinned region) Time (nanoseconds) Antenna elevated (Antenna further from thinned region)

8 Field Trial at a Local Highway Bridge File 14 Above Deck File 10 On Deck 2012 Data East end 45,000 40,000 35,000 30,000 25,000 20,000 15,000 File 10 on deck File 14 above deck 10,000 5, Distance (ft) 3

9 Field Trial at a Local Highway Bridge A dozen thinned rebar candidates were identified. These have unexpectedly low amplitude relative to signal arrival time (rebar depth in concrete). 18

10 New Data Collected June 2017 Three new sets of data Removal of the top layer of the road bed exposing the cold joint of the barrier rail New analysis with SAFT Position of Antenna Above Deck Normal Bridge Deck Top Coat Removed to this Level On Deck On Cold Joint

11 New Data Collected June 2017 Above Deck On Deck On Cold Joint

12 New Data Collected June 2017 SAFT Processed Above Deck On Deck On Cold Joint

13 V pk to pk New Data Collected June 2017 SAFT Processed 35, Data East end 30,000 25,000 20,000 Cold Joint 15,000 On Deck Above Deck 10,000 5, Distance (ft)

14 Response Combined Data Rebar Response Comparison of Rebar Signal (East End) File 10 On Deck File 14 Above Deck 8000 File 5 Cold Joint File 7 On Deck 6000 File 9 Above Deck Rebar

15 Response (V) Combined Data SAFT Processed Rebar Response (East End) File 10 On Deck File 14 Above Deck File 5 Cold Joint File 7 On Deck File 9 Above Deck Depth (ns)

16 Outline 1) Objective. 2) GPR Barrier Rail Inspections Old and New. 3) Laboratory measurements using a concrete phantom : Effect of rebar orientation and metal loss on GPR signals. 4) Summary.

17 Factors Affecting Peak GPR Echo From Rebar Rebar Z direction into wall Y Metal loss region Rebar depth in concrete. Rebar length. Rebar rotation angle (XY plane). X (scan) Rebar tilt angle (in YZ plane). Size of metal loss region. Location of metal loss region (in Y direction). Last year we began an investigation of each factor using rebar in air

18 Antenna Antenna suspended from ceiling. Wooden side supports being used for tiltedrebar measurements All measurements use standard ½-inch rebar Lab Setups for Earlier Rebar-in-Air Measurements Styrofoam table being used for rebar-gap measurements

19 Lab Setups for Rebar-in-Concrete-Phantom 2 concrete block Antenna 2 moist sand Before sand is added 2 concrete block Metal plate Moisture content is 5.4% by weight Rebar which can slide thru holes in wooden frame Wooden frame After sand & top block added

20 GPR Measurements in Air: Setups Used Last Year X into page Air Antenna Vary Depth X into page Air Antenna Length Vary Top view Rotation X into page Antenna Tilt Air Antenna X θ Vary New phantom data Air φ Vary X into page Air Antenna Vary Metal loss length X into page Air Antenna Vary Metal loss location New phantom data

21 Rebar Response (a.u.) Rebar Response (a.u.) For all measurements: Effect of Rebar Rotation (as measured) Use either 1.6 GHz or 2.6 GHz antenna. Measure GPR signal from rebar target + support structure. Measure GPR signal from support structure with rebar removed (background). Subtract to get GPR signal from rebar target alone. 0 degrees 25 degrees Time (nsec) Effect of Rebar Rotation (BKG removed) degrees 25 degrees 2.6 GHz 2.6 GHz Time (nsec) Front-wall + direct-coupled signal Rebar response Metal plate echo Rebar response after background subtraction

22 Response Rebar Response (a.u.) Measurement #1: Effect of varying the rebar rotation angle θ with a fixed separation between the rebar and antenna Top view Air Antenna θ Rotation Vary Rebar centered below antenna at 3-inch depth in concrete phantom Rebar length = 48 Effect of Rebar Rotation X (BKG removed) foot length Time (nsec) Time (nsec) 0 degrees 25 degrees Angle θ varies from 0 o to +/-25 o (θ = 0 o when rebar is perp. to antenna scan direction X) In bridge walls, any rotation angle is expected to be small (< 10 o ). Even the 25 o tilt at left only lowers rebar response by 15%

23 Norm'ed Rebar Response Norm'ed Rebar Response Measurement #1: Effect of varying the rebar rotation angle θ with a fixed separation between the rebar and antenna Rebar in Concrete Phantom Comparison to Rebar in Air Peak-to-Peak Rebar Response As measured BKG subtracted Rotation Angle (Degrees) Peak-to-Peak Rebar Response inch (air) inch (air) 3-inch (air) inch (concrete) GHz 1.6 GHz in air Rotation Angle (degrees)

24 Response Measurement #2: Simulates a 4-foot long rebar section having a 1 gap. The gap location varies w.r.t. the antenna center. Concrete Antenna Vary Z C Metal loss location Rebar centered below antenna at fixed Z = 3 in concrete Distance C between center of gap and center of antenna varies from 0 to 8 inches GHz C = 7" C = 0" Background subtraction has been used Rebar Metal plate GPR signal stable in time and shape. Large variation in amplitude seen when gap is centered under antenna. Time (nsec)

25 Norm'ed Peak-to-Peak Response Measurement #2: Simulates a 4-foot long rebar section having a 1 gap. The gap location varies w.r.t. the antenna center inch-long gap in rebar Normalized w.r.t. offset = 7 (Gap far from antenna center) GHz 1.6 GHz Offset Distance From Antenna Center (inches) Results shown for two antennas (1.6 and 2.6 GHz) Min rebar signal seen when offset is zero.

26 Norm'ed Peak-to-Peak Response Measurement #2: Simulates a 4-foot long rebar section having a 1 gap. The gap location varies w.r.t. the antenna center inch-long gap in rebar. 1.6 GHz Offset Distance C in Air (inches) GHz 1.6 GHz in concrete 1.6 GHz in air Offset Distance C in Concrete (inches) Comparing earlier measurements in air to new measurements in concrete. Antenna-to-rebar distance is 23.4 in air, 3 in concrete. For in-air data offset scale is three times larger. In-air offset extends to 19.

27 Response Measurement #3: Simulates a 4-foot long rebar section having a thinned region. Region location varies w.r.t. the antenna center. Concrete Antenna Vary C 2.6 GHz Z Rebar C = 7" C = 1" Time (nsec) Metal loss location Metal plate Rebar centered below antenna at fixed Z = 3 in concrete Thinned region is 1 long, 50% diameter loss. Distance C between centers of region and antenna varies from 0 to 7 inches. GPR signals stable in time and shape. Modest variations in amplitude seen. Minimum rebar signal seen when offset = C = 1 inch

28 Norm'ed Peak-to-Peak Response Measurement #3: Simulates a 4-foot long rebar section having a thinned region. Region location varies w.r.t. the antenna center inch-long 50%-thinned rebar section Normalized w.r.t. offset = 7 (i.e., thinning far from antenna center) GHz Results shown for two antennas (1.6 and 2.6 GHz) GHz Offset Distance From Antenna Center (inches) Min rebar signal seen when offset is between 1 and 2 inches.

29 Outline 1) Objective. 2) GPR Barrier Rail Inspections Old and New. 3) Laboratory measurements using a concrete phantom : Effect of rebar orientation and metal loss on GPR signals. 4) Summary

30 Summary We have embarked on Phase II of a project investigating the use of GPR to detect and quantify rebar thinning. This will combine laboratory measurements and measurements on a local bridge. Lab measurements indicate that GPR signal amplitude is sensitive to material loss of rebar due to corrosion. But care must be taken to compensate for effects due to rebar depth and orientation. Bridge measurements were made June 2017