Experimental Vibration Analysis for Civil Engineering Structures

133

Advanced vibration measurement system using laser Doppler

vibrometers for structural monitoring
T. Miyashita
Nagaoka University of Technology, Nagaoka, Niigata, Japan

H. Ishii
Yokogawa Bridge Corp., Funabashi, Chiba, Japan

K. Kubota
Keisoku Research Consultant CO., Hiroshima, Hiroshima, Japan

Y. Fujino
University of Tokyo, Tokyo, Japan

ABSTRACT: This paper introduces three studies of advanced vibration measurement system for
bridges using laser Doppler vibrometers (LDV), which makes possible to conduct high accurate,
non-contact and long distance measurement. 1) At a railway steel box girder bridge, damage
was observed on the web of a girder at the bottom end of a welded vertical stiffener. The objective of this study is to clarify the cause of the damage using conventional sensors and LDVs. 2)
In order to measure three-dimensional behavior of an object high accurately as well as spatial
densely, vibration measurement system using three scanning type LDVs was developed. 3)
When a LDV is far away from measured points, it is especially difficult to confirm the location
of the laser point on a structural surface. Therefore, a non-contact measurement system combining LDV with Total Station for long distance measurements, which also has the ability of high
accurate positioning, was developed.
1 INTRODUCTION
In Japan, continuous investments in infrastructures have formed a huge stock. However, since
the infrastructures were constructed rapidly and emphatically in high economic growth period,
the number of the degraded ones comes to increase rapidly. Therefore, the development of
quantitative and efficient maintenance techniques is strongly required.
Laser Doppler Vibrometer (LDV) is an optical instrument employing laser technology to
measure velocity. The characteristics of LDV are the followings: first, in comparison with conventional transducers such as accelerometers, non-contact and long distance measurement is
possible without adding mass or stiffness to an object. Second, resolution of velocity is very
high, and frequency bandwidth is very wide. Third, by attaching a scanning unit of mirror in
front of the laser sensor head, measurement on multiple points is made possible.
This paper introduces three studies of advanced vibration measurement system using LDVs in
order to monitor bridges.
2 UNDERSTANDING OF HIGH-SPEED TRAIN INDUCED LOCAL VIBRATION OF A
RAILWAY STEEL BRIDGE (Miyashita et al. 2007)
2.1 Bridge for measurement
The investigated bridge is a pair of steel mono-box girders with 4 spans as shown in Figures 1. The
stiffeners were not welded to the lower flange in the sections where positive moment affects the
girder. In bridges which have similar detail, damage was observed on the web of the girder at the
bottom end of a welded vertical stiffener. The parts of the bridge of similar detail were retrofitted
such as in Figure 1(b) using a T-shape member installed between the web and the lower flange using high tension bolts.

134

EVACES07

4@40 000
M

Web

2600

Lower Flange

2000

E D E C E D E C E D E C E D
@1600

600

600

2000

2000

Center of span

Crack
not welded

Section C
(Section of Diaphragms)

Section D
(Transverse
Rib
Lower Chord
Member)

Section E
(Transverse
Rib
Vertical
Stiffener)

Damage

(a) Detail of bridge

Figure 1. Focused bridge for measurement

Retrofitting

(b) Detail of Retrofitting (Section E)

1000

AX2L

2000

1000

600

800 800

Vertical Stiffener

AX3C
S01R

1000 1000

2000

:: X:
Strain
Gauge

gage
(Displacement
Sensor)

:: Accelerometer

Figure 2. Measurement using conventional sensors

2.2 Measurement using conventional sensors

Accelerometers and strain gauges were installed in the bridge for measurement of train induced
vibration. Since it is suspected that the cause of high local stresses was local vibration of the
web or lower flange of the girder, accelerometers were installed on both the web and the lower
flange of the girder. The strain gauges were installed on the main girder web 20mm away from
the toe of welding. Furthermore, the strain gauges were installed on the lower flange on the
main girder as shown in Figure 2.
2.3 Measurement results
Figures 3 show examples of measurement. Figure 3(a) shows that there is one large waveform
caused by trains passing on the bridge and there are 17 cycles corresponding to each vehicle
passing the cross section. It is shown that the dominant frequency is 3.0 Hz. This phenomenon
is known as speed effect by cyclic loading (Matsuura 1976) which depends on the interval
that bogies or train vehicles pass at constant speed. Figure 3(b) shows the presence of high frequency local vibration, and the frequencies of these vibrations are 20Hz-30Hz.
2.4 The relation between vibration and train speed
Figure 4 shows the relation between the dominant frequency of acceleration of the girder and
train speed of 39 trains. This figure shows that the dominant frequency of vibration of the girder
depends on the train speed and it agrees with speed effect by cyclic loading. Figure 5 shows
the relation between the dominant frequency of acceleration of the lower flange and train speed.
This figure shows that the dominant frequency depends on train speed, similar to the vibration
of the girder and that the frequency is an integer multiple of the frequency of the girder.

135

0
-10
-20

10
15
Time [s]

20

1
3.00Hz
3.00 Hz
0.8
0.6
0.4
0.2
0
0 10 20 30 40 50 60
Frequency [Hz]

60
40
20
0
-20
-40
-60
0

10
15
Time [s]

20

Acceleration [m/s2]

10

Stress [MPa]

Stress [MPa]

20

Acceleration [m/s2]

Experimental Vibration Analysis for Civil Engineering Structures

6
5 3.00Hz
32.80Hz
4
3
2
1
0
0 10 20 30 40 50 60
Frequency [Hz]

(a) Stress response of lower flange

(b) Acceleration response of lower flange
Figure 3. The results of measurement by conventional sensors

Frequency [Hz]

3.2

Measurement
fb=V/90

3
2.8
2.6
2.4
230

240

250

260

270

280

Train Speed [km/h]

Train speed [km/h]

Figure 4. Relation between train speed and dominant frequencies of girder
1st Peak
2nd Peak
3rd Peak

300
280
260
240
220

10

11

12

13

200
25

30

35

Frequency [Hz]

40

Acceleration [m/s ]

Figure 5. Relation between train speed and dominant frequencies of lower flange
6.0
5.0
4.0
3.0
2.0
1.0
0.0

Lower flange
Web

0.0

1.0

2.0

3.0

Stress [MPa]
Figure 6. Correlation between stress and acceleration

2.5 The relation between local vibration and train speed

Figure 6 shows the relation between local vibration and local stresses. This figure shows that
there is a strong correlation between local stresses and the lower flange vibration Thus, it was
confirmed that the cause of high local stresses is local vibration of the lower flange. The vibration-mode inducing local stresses can be seen in Figure 7 based on the investigation of the phase
between measurements.
2.6 The relation between local vibration and local stresses
Figure 8 shows the relation between the Fourier spectral amplitude and the frequency of the
lower flange vibration. This figure shows that there are 2 peaks of different frequencies. A peak

136

EVACES07

occurs at 28.65Hz and another one is at 32.80Hz. These 2 peaks identify the natural frequencies
of local vibration. Figure 9 shows the relation between the Fourier spectral amplitude of the
lower flange vibration and the train speed in each mode. It is shown that each mode has each
peak at a certain speed and that the frequency at each peak is an integer multiple of the frequency of the girder. Thus, it was confirmed that the vibration is larger when the natural frequency of local vibration is an integer multiple of the frequency of the girder induced by cyclic
loading.
2.7 Measurement using LDVs
The measurement system consists of three scanning type LDVs at three sections and one single
point type LDV as shown in Figure 10. The single point type LDV always measures a reference
point and is used to calculate the phase between measurement points for the identification of
mode shapes. Using LDV, measurements of ambient vibration was conducted. During ambient
vibration measurement, measurements were conducted before and after retrofitting the bottom
part of the vertical stiffeners as shown in Figure 1(b). The objective of the measurement is to
identify natural frequencies and mode shapes at the sections before and after retrofitting.

10.0MPa
--9.3MPa
9. 3N/ mm 10.
0N/ m
m

5.4MPa
5.
4N/ m
m

0.0.12mm
12m
m

0.23mm
0.
23m
m

--6.2MPa
6. 2N/ m
m

0.40mm
0. 40m
m

0.11mm
0. 11m
m

0.08mm
0.
08m
m
0.66mm
0.
66m
m

Figure 7. Local mode shape causing local stress

32.80Hz
28.65Hz

Acceleration [m/s ]

6
4

1st Peak
2nd Peak
3rd Peak

2
0
25

30

35

40

Frequency [Hz]

Figure 8. Fourier spectral amplitude and peak frequencies of lower flange

28.65Hz
32.80Hz

Acceleration [m/s ]

8
6
4

32.80/(268.4/90)=11

28.65/(258.0/90)=10
32.80/(246.2/90)=12

2
0
220

230

240

250

260

Train Speed [km/h]

270

280

Figure 9. Train speed and Fourier spectral amplitude of lower flange

Experimental Vibration Analysis for Civil Engineering Structures

137

Osaka
L13-14

L11-12
L9-10

Tokyo

SLDV1
SLDV2

RLDV

SLDV3

Figure 10. Vibration measurement of a steel railway bridge using LDVs

Vertical
Lateral

Longitudinal

(a) Before retrofitting (29.35 Hz)

(b) After retrofitting (29.35 Hz)

(c) Before retrofitting (32.75 Hz)

(d) After retrofitting (33.35 Hz)
Figure 11. Identified modes shape from ambient vibration data using LDV

2.8 Identification of mode shapes using LDVs

Figures 11 show identified mode shapes based on the ambient vibration measurement using
peak-picking method. Figures 11(a), (c) and (b), (d) show the mode shapes before and after attaching the stiffener for retrofitting respectively. Notice that although natural frequencies of
both modes of Figures 11(a), (b) are the same 29.35 Hz, the mode shapes are greatly different.
In Figures 11(c), (d), the natural frequency after retrofitting increases a little.

138

EVACES07

3 DEVELOPMENT OF THREE-DIMENSIONAL VIBRATION MEASUREMENT SYSTEM

USING LDVS (Miyashita 2005)
3.1 Introduction
The objective of this study is to develop vibration measurement system using three LDVs,
which makes possible to measure three-dimensional behavior of an object high accurately as
well as spatial densely. Since developed system can clarify the three-dimensional local deformation of structures caused by train/traffic-induced vibration, it can contribute to the preventive
maintenance and the improvement of functionality in existing infrastructures.
3.2 Problem in one-dimensional vibration measurement using a LDV
When only one LDV is utilized and the laser beam is perpendicular to the measurement surface,
the velocity component perpendicular to the surface can be measured with very high accuracy.
However, if the laser beam is not perpendicular to the surface, the accuracy of the velocity decreases depending on the laser beam angle.
In order to investigate this problem, a simple experiment was conducted. Figures 12 show the
experimental set up. First, a steel plate with an attached accelerometer was hit by a hammer.
Then, a LDV measured the velocity of the plate from the direction making an angle of 45 degrees with respect to the normal axis.
Figure 13 shows the comparison. The solid blue line is the velocity computed from the acceleration and the thin red dot line is the velocity from the LDV. Comparing these two velocities,
their amplitudes are different, because the laser beam angle is not perpendicular to the surface.
In order to obtain the vibration component perpendicular to the surface, compensation on laser
beam angle is necessary based on Vm = V cos . Where Vm is the velocity perpendicular to the
surface, V is the velocity measured by the LDV and is the laser beam angle of the LDV.
The green thick dot line in Figure 13 is the velocity compensated by the equation. Comparing
the solid and thick dotted lines, their amplitudes agree well. Therefore, when only one LDV is
used, it is necessary to install the LDV perpendicular to the surface to conduct accurate measurements. However, on field measurement using LDV, it is difficult to satisfy such a condition.
3.3 Three-dimensional measurement principle using LDVs
The principle of three-dimensional measurement involves a geometrical transformation of coordinates as shown in Figure 14. Three-dimensional vibration components of the body are determined from measurements of the three LDVs by pre-multiplying by a transformation matrix
consisting of the direction cosines of the laser beam angles.
Vx cos 1

Vy = cos 2
V cos
3
z

cos 1
cos 2
cos 3

cos 1
cos 2
cos 3

V1

V2

V3

(1)

where Vx ,Vy ,Vz are vibration components of the body in each axes, V1 ,V2 ,V3 are vibration
components measured by each LDV and i , i , i ( i = 1, 2,3) are the laser beam angles of each
LDV.
Accelerometer

300 [mm]
Measurement
points

Steel plate

385 [mm]

45

LDV
LDV

(a) Side view

(b) Top view
Figure 12. Configurations of a LDV in one-dimensional vibration measurement

Experimental Vibration Analysis for Civil Engineering Structures

Accelerometer
LDV: 45 degrees
LDV: angle compensation

120
Velocity [mm/sec]

139

60
0
-60
-120
8

8.1

8.2

8.3

Time [s]

8.4

Figure 13. Measurements of one-dimensional vibration measurement using a LDV

Z

Vz
Vy

V1
V2

Vx

V3
LDV1
LDV3
X

LDV2

Figure 14. Three-dimensional vibration measurement using LDVs

3.4 Fundamental investigation of three-dimensional measurement

As fundamental study, three multiple points type LDVs having a scanning system (SLDV) are
set at parallel to a measurement surface. The surface was vertically irradiated with each LDV
when input voltages are 0 V. Therefore, laser beam angles for the measurement points are directly obtained from mirror angles.
The developed system was applied to three-dimensional vibration measurement for a steel
railway bridge. A tri-axial accelerometer was attached on the surface of the main girder for the
purpose of validation. Figure 15 shows the setting of the SLDVs and the measurement axes of
the accelerometer. Each SLDV was set parallel to the main girder and ground. Figures 16 show
the comparison in each axis concerning time histories. These are free vibration components after
a train passes thorough the railway bridge. Since both measurements agree well in each axis, the
system can be applied to on-site measurement.
3.5 Three-dimensional vibration measurement system using SLDVs
The three-dimensional vibration scanning measurement system shown in Figure 17, which is
not necessary to take care of setting of the SLDVs, is investigated for the application to on-site
measurement.
When the SLDVs are arbitrary set, it is impossible to directly obtain laser beam angles from
angles of mirrors necessary. Therefore, it is necessary to identify positions and rotations of the
SLDVs in absolute world coordinates, and then obtain the laser beam angles from the vectors
connecting measurement points with SLDVs.
Laboratory experiment was conducted in order to verify the validation of three-dimensional
vibration measurement employing this system. Three SLDVs were placed arbitrary as shown in
Figure 17 in this experiment. The arbitrary means that it is not necessary for the SLDVs to be
set parallel to the measurement surface and ground. The tri-axial accelerometer was made to be
vibrated by blowing of electric fans as shown in Figure 17. In order to compare the measure-

140

EVACES07

ment by the accelerometer with the measurement by the SLDVs, the acceleration measured by
the accelerometer was integrated numerically, and the comparison was conducted with the velocity.
Figures 18 show the comparison of measurements. Since both results agree well in each
measurement axis, the validation of the measurement was verified. Therefore, we can conclude
that it is necessary to identify the positions and rotations of each SLDV, and obtain the direction
cosines along the direction irradiated with laser in order to conduct accurate three-dimensional
vibration measurement using SLDVs.

Lateral
Axis1 Axis2

Longitudinal

SLDV2

Axis3

SLDV1

RLDV

Vertical

SLDV3

ACC1
LDVy

1
0
-1

-2
10 10.1 10.2 10.3 10.4 10.5
Time [s]

ACC2
LDVx

2
0
-2

-4
10 10.1 10.2 10.3 10.4 10.5
Time [s]

(a) Horizontal In-Plane

(b) Horizontal Out of-Plane
Figure 16. Verification of three-dimensional vibration measurement
Calibration Board

Acceleration [m/s2]

Acceleration [m/s2]

Acceleration [m/s2]

Figure 15. Configuration of field measurement

4

ACC3
LDVz

2
0
-2

-4
10 10.1 10.2 10.3 10.4 10.5
Time [s]

(c) Vertical In-plane

Tri-axial accelerometer

SLDV1
SLDV2
SLDV3

0.044

ACC1
LDVy

0.02
0
-0.02
-0.04
Time [s]

Velocity [m/s]

Velocity [m/s]

0.04

ACC2
LDVx

0.022
00
-0.022
-0.044
0

10
15
Time [s]

20

Velocity [m/s]

Figure 17. Laboratory experiment

0.03
0.02
0.01
0
-0.01
-0.02
-0.03

ACC3
LDVz

2 3
4
Time [s]

(a) Horizontal In-Plane

(b) Horizontal Out of-Plane
(c) Vertical In-plane
Figure 18. Verification of three-dimensional vibration measurement

141

Experimental Vibration Analysis for Civil Engineering Structures

4 DEVELOPMENT OF REMOTE NON-CONTACT MEASUREMENT SYSTEM BY

COMBING LASER DOPPLER VIBROMETER AND TOTAL STATION (Kubota et al. 2007)
4.1 Introduction
In this Research, a non-contact measurement system which combines a LDV and a total station
(TS) was developed. This system is capable of measuring a lot of cable members for long-span
cable supported bridge applications, since it is especially suited for remote non-contact vibration
measurements.
4.2 Measurement at Tatara Bridge
The measurement was conducted by measuring the ambient vibration of the cables using conventional accelerometers and LDV simultaneously. Typical cable vibration frequency plots obtained from the LDV and the accelerometer data are shown in Figure 19.
At first, there was some concern that it will be difficult to determine the natural frequency of
the cables due to the influence of sag and the contamination of the recorded data with many
low-frequency vibrations. However, we were able to get good results, verifying the effectiveness of the LDV in identifying the natural frequency of cable elements.
4.3 Combined LDV with TS system

1500
1000
500
0
0

5
frequency [Hz]

10

Fourier Amplitude Spectral[m/s]

Fourier Amplitude Spectral[m/s ]

When the measurement points are located far from the LDV, it is difficult to confirm the irradiation of the LDVs laser point during on-site measurements. To improve this situation, the measurement system with the LDV attached to a TS, having a highly accurate, remote measurement
position identification ability, was developed. The TS posses 1 mm accuracy for 100m distance,
making it possible to determine the exact position of any point (Fig. 20).
The measurement consists of the following steps: 1) The reflection tapes are attached at the
measurement points, 2) The orientation of the system is adjusted so that the reflection level of
the LDV reaches maximum level, 3) The measurement time is set, 4) Each measurement point is
identified with the TS and positional information is stored, 5) When the TS automatically aims
at a measurement point using the stored position information, the data acquisition program detects a signal and automatically records vibration from the LDV, 6) Once the vibration measurement using LDV finishes, the TS automatically moves to the next measurement point. Once
the coordinates of the points are identified, this system can repeat the measurements automatically.
At Kohei bridge in Hiroshima city, automatically repeated vibration measurement of hanger
ropes was conducted. The measurement of a hanger rope from a distance of about 80m using
LDV is shown in Figure 21(a). Moreover, the result of this hanger rope using an accelerometer
is shown in Figure 21(b). It is can be seen that the frequency peaks obtained from both sensors
are in good agreement. Also, we are confirming that there is no change in the first and second
natural frequencies for repeated measurements.

40
30
20
10
0
0

(a) Configuration
(b) Accelerometer
(c) LDV
Figure 19. Comparison of frequency plots from a Tatara Bridge Cable

5
frequency [Hz]

10

142

EVACES07

Measurement object Hanger rope

LDV Sampling rate 1000Hz

TS Measurement time 0.8 seconds

LDV Measurement time 30 seconds

Longest
Distance
About 80m

TS laser
LDV laser

Fourier Amplitude Spectral[m/s ]

Fourier Amplitude Spectral[m/s]

Figure 20. Combined LDV and TS system

100

LDV

50

0
0

10
20
Frequency [Hz]

30

10000

Accelerometer

5000

0
0

10
20
Frequency [Hz]

30

(a) LDV from the distance of about 80m (b) Accelerometer

Figure 21. Measurement using LDV and TS system

5 CONCLUSIONS
This paper introduces three studies of advanced vibration measurement using LDVs for the purpose of structural monitoring: 1) Understanding of high-speed train induced local vibration of a
railway steel bridge. 2) Development of three-dimensional vibration measurement system using
LDVs. 3) Development of remote non-contact measurement system by combing LDV with total
station.
REFERENCES
Miyashita et al. 2007a Understanding of high-speed-train-induced local vibration of a railway steel bridge
using laser measurement and its effect by train speed, Journal of structural mechanics and earthquake
engineering JSCE, Vol.63, No.2, pp.277-296. (in Japanese)
Matsuura 1976: A study of dynamic behavior of bridge girder for high speed railway, Proceedings of the
Japan Society of Civil Engineers, No.256, pp.35-47. (in Japanese)
Miyashita 2005: Advanced Measurement System using Laser Doppler Vibrometers for Monitoring Civil
Infrastructures, PhD dissertation, University of Tokyo. (in Japanese)
Kubota et al. 2007: Development of a remote non-contact measurement system combining laser Doppler
vibrometer and total station for monitoring of structures, The 3rd international conference on structural
health monitoring of intelligent infrastructure. (in press)