Sponsored by

William W. Hay Railroad Engineering Seminar

“Non-destructive in-place condition assessment technologies for deterioration in railroad ties”

John PopovicsAssociate Professor University of Illinois at Urbana-Champaign

Date: Friday, February 27, 2015 Time: Seminar Begins 12:20

Location: Newmark Lab, Yeh Center, Room 2311University of Illinois at Urbana-Champaign

John S. PopovicsThe University of Illinois at Urbana-Champaign

RailTEC Hay SeminarFebruary 27, 2015

NON-DESTRUCTIVE IN-PLACE CONDITION ASSESSMENT TECHNOLOGIES FOR DETERIORATION IN RAILROAD TIES

Presentation outline

1) Objective, motivation, review of previous

work

2) Development of data analysis/evaluation

schemes

3) Improvement of ultrasonic testing set-up

4) Experimental results

5) Future Work

3

Motivation

Concrete and timber crossties are important components in the rail bed structure:

* Distribute wheel loads (Support)* Maintain track geometry (Stability)* Electrically isolate rails (Isolation)

<J.R. Edwards>

Material integrity of ties especially important for high speed rail structures

4

Vision for ultrasonic evaluation of rail ties

Concept of implementation for performing cost effective inspection: one-sided, air-coupled ultrasound techniques

carried out from a moving platform

5

Timber rail ties have long been used in rail structures in the US. Structural deficiency arises from natural degradation mechanisms

Deteriorated wood Sound wood

6

A variety of damage

mechanisms can

compromise concrete ties

Microcracks

Delamination

Rail seat deterioration(RSD)

<Walker et al. 2006>

<S. Naar>

<J.R. Edwards>

Concrete rail ties are becoming more prevalent, but can suffer deterioration

Mechanical body waves in solids: P-waves

Direction of Travel

Excitation

Direction of Particle Motion

also: Longitudinal (L-) Waves, Compression Waves

( )( )( )2ν1ν1ρ

ν1EvP −+−

=

Wave Velocity: Governing ParametersYoung’s Modulus EPoisson’s Ratio vDensity ρ

<T. Voigt >

Mechanical body waves in solids: S-waves

Direction of Travel Direction of Particle Motion

Wave Velocity in solids: Governing Parameters

Shear Modulus GDensity ρ

no propagation in liquids or gases !

also: Transverse (T-) Waves, Shear Waves

Excitation

ρGvS =

<T. Voigt >

VP > VS in all known solids

Guided waves: Surface waves

<www.lamit.ro/earthquake-early-warning-system.htm>

Rayleigh surface wave travels along free surface but do not propagate far into the body of the material. Rayleigh waves travel slightly slower than shear waves, and

show coupled longitudinal and shear motion

Guided waves: Lamb waves in plates

<N. Ryden>

Lamb wave are set up in large plates

Multiple (infinite) modes of propagation, with varying motion character and propagation velocity

Can be visualized as a propagating resonance

Increasing frequency or plate thickness

Surface guided wave

Contactless MEMSReceiver

Contactless electrostaticSender

Fully contactless ultrasonic technique

AirConcrete

12

2 3 4

x 10-4

-0.01

-0.005

0

0.005

0.01A

mpl

itude

(V)

Time(sec)

MEMS 1 at same locationMEMS 2 at same location

Key feature of contactless sensing: 1) Signal consistency

1.5 2 2.5 3 3.5 4 4.5 5

x 10-4

-0.04

-0.03

-0.02

-0.01

0

0.01

0.02

0.03

0.04

Am

plitu

de(V

)

Time(sec)

1st Try with same Accelerometer2nd Try with same Accelerometer3rd Try with same Accelerometer

0 1 2 3 4 5 6

x 10-4

-0.06

-0.04

-0.02

0

0.02

0.04

0.06

Am

plitu

de(V

)

Time(sec)

AccelerometerMEMS SensorCondensor MicrophoneDynamic Microphone

Accelerometer(contact)

Smooth

Medium

Extremely Rough

Distance

0 1 2 3 4 5 6 7

x 10-4

-0.4

-0.2

0

0.2

0.4

0.6

Ampl

itude

(V)

Time(sec)

0 1 2 3 4 5 6 7

x 10-4

-0.4

-0.2

0

0.2

0.4

0.6

Am

plitu

de(V

)

Time(sec)

0 1 2 3 4 5 6 7

x 10-4

-0.4

-0.2

0

0.2

0.4

0.6

Ampli

tude

(V)

Time(sec)

Key feature of contactless sensing: 2) Application to rough surfaces

Development of data analysis/evaluation

schemes

Ultrasonic surface wave data scans reveal that signal velocity and

attenuation indicate presence of distributed damage

Amplitude (v)

Wave arrivals

Forward propagating

energy attenuation

0 % 0.6 %0.3 %100 150 200 250 300

0.5

1

1.5

2

x 10-12

Location or distance (mm)

Area

of a

mpl

itude

2 (on

ly th

ree

peak

s)

zero0.30.6

Most attenuated

Characterization of forward propagating signal energy through short-time-interval computation

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

x 10-4

-3

-2

-1

0

1

2

3

x 10-3

Time(sec)

ampl

itude

(V)

A short-time-interval average signal a window width equal to the duration of the excitation pulse, of the square of the filtered signal was then constructed

(Weaver & Sachse, 1995).

Δt

1 1.5 2 2.5 3 3.5 4 4.5

x 10-4

-11

-10

-9

-8

-7

-6

-5

Time(sec)ln

(E)

x 160mm

1 1.5 2 2.5 3 3.5 4 4.5

x 10-4

-11

-10

-9

-8

-7

-6

-5

Time(sec)ln

(E)

x 160mm

Shifting of energy envelope indicates energy dissipation from wave scattering

1.2 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 3 3.2

x 10-4

-10

-9.5

-9

-8.5

-8

-7.5

-7

-6.5

-6

Time(sec)

ln(E

)

x 160mm

0 %0.3 %0.6 %

Wave scattering

The reflection of ultrasonic energy away from the original direction of propagation; caused by reflection, refraction and mode conversion from internal

inclusions. Causes signal loss, signal dispersion and scattering “noise”

Detected back-scattered signal

<Oelze 2007>

Can we make use of ultrasonic backscatter measurements to characterize distributed damage in

concrete using surface waves?

10mm

MEMS array

Transmitter

baffle

coherent

12

3

45

6

7

backscattering

coherent

Incoherent (backscattering )

Concrete samples subjected to sets of repeated hot-cold and wet-dry cycles to impart distributed damage

Ultrasonic backscatter and resonant frequency data for concrete samples subjected to many damage cycles

Standard resonance frequencyUT Backscattering variance

68o to ice 80o to ice

120o (dry) to 23o (wet)

Symmetric Lamb

Anti-symmetric Lamb

A0

A1

S0

S1 A2S2

Delamination(defect)

Concrete

Fast MASW

Air-coupled ultrasonic approach to detectdelamination in concrete ties using Lamb waves

<S. Naar>

20 - 100 mm

Improvement of ultrasonic testing set-up

Improved ultrasonic pulse control enables empowered test schemes and data analysis approaches

We have improved output amplitude, voltage biasing control, signal to noise ratio and frequency and bandwidth control of input signal for the transmitting transducer

o Center frequency control between 10-90 kHz

o Improved pulse duration and shape control: chirp and tone burst signals now possible

26

Ultrasonic hardware developments

Impr

oved

sen

sor

desi

gn

Crosshair Laser targeting

Contactless capacitance transducer Micro processer

Signal process

Gyro tilt sensor

Contactless sensor bracket

Adjustable height

Contactless MEMS sensor

Scan

ning

sys

tem

s

Increased offset distance between sensors and rail tie is critical for practical application: at least 20 cm (8 inch) offset needed

35mm 10mm

81mm0 0.5 1 1.5 2 2.5

x 10-3

-3

-2

-1

0

1

2

3x 10-3

Time(sec)

Am

plitu

de(V

)

4 5 6 7 8 9 10 11 12

x 10-4

-1

0

1

x 10-4

Time(sec)

Am

plitu

de(V

)

Direct acoustic wave

Surface wave

Surface wave

28

Increased transmitter and receiver offsets yield good surface wave signal. However large lateral spacing may require

modification of signal analysis schemes

1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 2.1 2.2

x 10-3

-5

0

5

x 10-5

Time(sec)

Am

plitu

de(V

)

200mm

200

mm

200

mm

29

Experimental Results

Optimal (for concrete) air-coupled configuration applied to samples from timber ties. Contact sensor used for comparison

31

Ultrasonic signals from sound timber across varying distance. 100 times averaging used. Both P-waves and surface waves readily generated in timber.

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

x 10-4

-5

0

5x 10

-3

ampl

itude

(V)

time(sec)

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

x 10-4

-5

0

5x 10

-3

ampl

itude

(V)

time(sec)

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

x 10-4

-5

0

5x 10

-3

ampl

itude

(V)

time(sec)

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

x 10-4

-5

0

5x 10

-3

ampl

itude

(V)

time(sec)

MicrophoneAcelerometer

4166.6 m/s: P-wave

1388.8 m/s:Surface wave

32

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

x 10-4

-2

0

2x 10

-3

ampl

itude

(V)

50mm

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

x 10-4

-4

-2

0

2x 10

-3

ampl

itude

(V)

100mm

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

x 10-4

-4

-2

0

2x 10

-3

ampl

itude

(V)

150mm

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

x 10-4

-4

-2

0

2x 10

-3

ampl

itude

(V)

200mm

microphoneacellerometer

4054.1 m/s

1369.8 m/s

Ultrasonic signals from deteriorated timber across varying distance. 100 times averaging used. Both P-waves and surface waves show distinction of material quality.

33

The goal is to develop understanding of inter-relation between damage and surface wave behavior

Detection of rail seat damage (RSD) in concrete ties

Rail seat deterioration (RSD)

Concrete crossties are important

components

• Distribute wheel loads (Support)

• Maintain track geometry (Stability)

• Electrically isolate rails (Isolation)

Zeman,2010

degradation at contact interface between the concrete rail seat and the rail pad that can result in track geometry problems

Currently, freeze-thaw cracking, crushing, hydro-abrasive erosion, and hydraulic pressure cracking may contribute to RSD

Zeman,2010

Preliminary RSD testing configuration: small offset

Receiver array

Sender

R1 : no damage R2 : moderate RSD R3 : serious RSD

Test results with and without a bearing pad

2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5

x 10-4

-8

-6

-4

-2

0

2

4

6

8x 10-3

Time(sec)

Ampl

itude

(V)

NoPadSerious RSD damage

2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5

x 10-4

-8

-6

-4

-2

0

2

4

6

8x 10-3

Time(sec)

Ampl

itude

(V)

NoPad

Moderate RSD damage

2.8 3 3.2 3.4 3.6 3.8 4 4.2

x 10-4

-4

-3

-2

-1

0

1

2

3

4x 10-4

Time(sec)

Am

plitu

de(V

)

r0r1r2

r1r2r3

R1 R2 R3

Are individual signal data reliable?

r1r2r3

Integrated signal energy analysis

Spatially averaged signals from multi-sensor array

Savr t =1

𝑁𝑁𝑝𝑝𝑝𝑝𝑝𝑝𝑝�y=1

𝑁𝑁𝑝𝑝𝑝𝑝𝑝𝑝𝑝

Tavr_y t

Statistical analysis for inhomogeneous material

Array averaged = Group = position, p1, p2….

In total seventy signal of each damage region

2 3 4

x 10-4

-16

-15

-14

-13

-12

-11

Time(sec)

ln(E

)

region =3, position =10

Quantification and statistical interpretation

averaged data on position-1 has 7 signals

So, a Box plot hasseven array signals

Area

of l

n(E)

2 3 4

x 10-4

-16

-15.5

-15

-14.5

-14

-13.5

-13

-12.5

-12

-11.5

-11

Time(sec)

ln(E

)2

region =3, position =1

12345678910

10 different positions

-2400

-2300

-2200

-2100

-2000

-1900

1 2 3 4 5 6 7 8 9 10Position

area

of ln

(E)2

range from =0.002s, to =0.0035s

Area

of l

n(E)

ln(E

)

Statistical interpretation of test data shows distinction only of most severe RSD damage

-2400

-2300

-2200

-2100

-2000

-1900

1 2 3 4 5 6 7 8 9 10Position

area

of l

n(E

)2

range from =0.002s, to =0.0035s

-2400

-2300

-2200

-2100

-2000

-1900

1 2 3 4 5 6 7 8 9 10Position

area

of l

n(E

)2

range from =0.002s, to =0.0035s

-2400

-2300

-2200

-2100

-2000

-1900

1 2 3 4 5 6 7 8 9 10Position

area

of l

n(E

)2

range from =0.002s, to =0.0035s

R1 R2 R3

Solid

Rough 1

Rough 2

Follow on RSD testing configuration: large offset

Testing configuration

250 mm250 mm

Ultrasonic ray paths and sensor configuration

Path 1

Path 2

Path 3

Seven different MEMS receivers

1

2

3

4

5

6

7

Energy envelope data collected along path 1 shows clear distinction between damage extent levels

1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9

x 10-3

-10.5

-10

-9.5

-9

-8.5

Time(sec)

ln(E

)

solidrough1rough2

1.2 1.3 1.4 1.5 1.6 1.7 1.8

x 10-3

-11.5

-11

-10.5

-10

-9.5

-9

-8.5

-8

Time(sec)

ln(E

)

solidrough1rough2

Raw energy envelope data Envelope data with diffusion fits

Future work

Optimize data analysis/evaluation schemes

Test prototype development and evaluation

Develop schemes to target tie regions rather than specific defects, providing overall tie heath index

Incorporate hardware in moving test platform

Evaluate on in-place ties

46

Acknowledgments

This research is carried out with the help of support from the Association of American Railroads (AAR), Technology Scanning Program

47