IACSIT International Journal of Engineering and Technology, Vol. 4, No.

5, October 2012

Displacement Measurement of Bending Tests Using

Digital Image Analysis Method
Ardalan Hosseini, Davood Mostofinejad, and Masoud Hajialilue-Bonab

obtain displacement in the field of structural engineering.

AbstractOver the last few years due to sharp increase in Photogrammetry is the process of obtaining accurate
experimental investigations in the field of structural measurements from physical objects using photography. Few
engineering, the necessity for reliable and precise deformation investigations have been conducted using this technique in
measurement is felt more than ever. The most common
experimental test in the field of structural engineering is
the field of structural engineering [1]. In summary the
bending test in which beams deformation is evaluated in order technique operates by monitoring colored targets which are
to calculate other structural parameters. One of the most usual placed at selected locations on the field surface using
techniques for obtaining beams deformation is utilizing Linear specially designed cameras. The accuracy of the system is
Variable Differential Transformer (LVDT) sensors. However, mainly dependent on the camera hardware, scale of the object
non-mechanical procedures for displacement measurement and geometry of the survey [2].
have shown much interest between researchers.
Photogrammetry, Digital Image Correlation (DIC) method and A. Digital Image Correlation (DIC) method
Particle Image Velocimetry (PIV) are some of the new
Another recent method for obtaining displacement fields is
image-based methods for displacement measurement. Due to
the advantages of PIV method in the field of experimental fluid Digital Image Correlation (DIC) method. In general, DIC is
mechanics and Geotechnics, the current study has been focused an image analysis procedure which uses the mathematical
on the performance of PIV method for displacement correlation functions to analyze digital images of a field
measurement of bending test in structural engineering. To undergoing deformation. This technique offers the advantage
reach the aforementioned goal, one steel I-beam was subjected of obtaining continuous measurements of the whole
to four-point flexural loading and the load-displacement curve
was plotted using PIV method. The experimental results
displacement field. Few studies have been carried out using
obtained from the PIV method showed excellent agreement this technique [3], [4]. In order to obtain the displacement
with LVDTs data. It is concluded that due to the lower cost and field, surface preparation consisting of creating a random
potential of obtaining whole displacement field on common sprayed-on speckle pattern, should be done. After the surface
bending tests, PIV method could be widely used as an preparation, digital images should be recorded during
alternative to conventional measurement techniques in deformation of the object. Finally the correlation between the
structural engineering.
deformed images and the undeformed reference image will
Index Terms Bending test, digital image correlation (DIC),
be used to obtain a two-dimensional displacement field [3].
displacement measurement, linear variable differential The accuracy of this technique is based on the camera
transformer (LVDT), particle image velocimetry (PIV). resolution. Based on experimental investigation, an image
displacement accuracy of approximately 0.01-0.005 pixels
can be expected [3]-[5].
I. INTRODUCTION
B. Particle Image Velocimetry (PIV)
Over the last few years due to extreme increase in
Particle Image Velocimetry (PIV) is originally a
experimental investigations in the field of structural
velocity-measuring technique developed in the field of
engineering, the necessity for accurate deformation
experimental fluid mechanics [6]. The technique was
measurement is felt more than ever. It is obvious that further
originally implemented using double-flash photography of a
calculations could be seriously affected by accuracy of the
seeded flow and the resulting photographs were divided into
evaluated displacement fields. Bending test is the most
a grid of test patches. For PIV analysis, the displacement
common test in the field of structural engineering in which
vector of each patch during the interval between the flashes is
beams deformation should be evaluated in order to calculate
found by locating the peak of the autocorrelation function of
any other structural parameters. Utilizing Linear Variable
each patch. The peak in the autocorrelation function indicates
Differential Transformer (LVDT) is a primitive technique for
that the two images of each seeding particle overlying each
obtaining deformation of bending tests. Although there are
other, so the correlation offset is equal to the displacement
certain benefits of LVDTs such as high accuracy and simple
vector [7]. A modified approach was used to implement PIV
installation, there are also some disadvantages such as high
in geotechnical testing by White et al, 2003. Although fluid
cost, inability to obtain the whole displacement field and
requires seeding with particle to create features upon which
extra charges for data acquisition.
image analysis can be operate, natural sand which is used in
Photogrammetry is another technique which is applied to
geotechnical tests, has its own texture in the form of
different-colored grains and the light and shadow formed
Manuscript received August 3, 2012; revised September 14, 2012. between adjacent grains when illuminated [7]. According to
The authors are with the Civil Engineering Department, Isfahan Whites investigations, the modified PIV technique offers an
University of Technology (IUT), Isfahan, Iran. (e-mail:
[email protected]; [email protected]; [email protected]).
order-of-magnitude increase in accuracy, precision, and

10.7763/IJET.2011.V4.451 642
 IACSIT International Journal of Engineering and Technology, Vol. 4, No. 5, October 2012

measurement array size compared with previous image-based

methods of displacement measurement. An image
displacement resolution of 0.005 pixels can be expected from
the PIV technique [7].
The advantages of PIV method in the field of experimental
fluid mechanics and geotechnical tests and also due to
importance of the bending test in the field of structural
engineering, have led the current study to evaluate the
performance of PIV method for displacement measurement
of bending test by means of experimental investigations.

II. EXPERIMENTAL PROCEDURE

A. Test Setup (a)
In order to evaluate the displacement field using PIV
method, one steel I-beam was subjected to four-point flexural
test. The yielding stress of the steel was equal to 240 MPa
according to the manufacturers catalog. The steel I-beam
dimensions and the test setup are shown in Fig. 1.

Fig. 1. Dimensions of steel I-beam specimen and loading arrangement

(b)
As it was mentioned earlier, it is necessary for the images Fig. 2. (a) Test setup; (b) camera view
to have a texture to create features upon which image
processing can operate. Thus, since the steel I-beam
originally does not show a suitable texture, natural colored III. RESULTS AND DISCUSSION
sand between sieve no. 50 and 100 was stuck to the beams The load-displacement curves for the steel I-beam which
face using suitable two-component epoxy adhesive. were obtained from the PIV method using 10.0 megapixel
The beam was subjected to four-point flexural loading in and 21.0 megapixel cameras are shown in Figs 3 and 4,
elastic zone, using a 2000 kN displacement control hydraulic respectively. In order to evaluate the PIV results,
jack. A Linear Variable Differential Transformer (LVDT) load-displacement curves obtained from the LVDT have also
having a resolution of 0.1 mm was mounted at the mid-span
been plotted in Figs 3 and 4.
of the specimen for plotting the load-displacement curves.
The camera was placed perpendicular to the beam face at a
distance equal to 1.0 m. In order to eliminate any probable LVDT PIV analysis
10.0
parasitic lights, the specimen was illuminated using two
white light projectors. Digital images were taken using a
remote control at regular intervals and a digital data logger 8.0
was used to monitor the load cell, LVDT and image numbers.
Load (kN)

Test setup and the camera view are shown in Fig. 2. 6.0

B. Image Analysis Using Particle Image Velocimetry (PIV)

4.0
Digital images were taken of the deformed steel I-beam
using Nikon D80 digital camera with resolution 10.0
megapixel (3872 2592 pixels) during the loading procedure. 2.0
The images then were processed using GeoPIV8 software,
developed at Cambridge University [8], and 0.0
load-displacement curves were plotted. In order to 0.0 0.5 1.0 1.5 2.0 2.5 3.0
investigate the impact of camera resolution on displacement Displacement (mm)
accuracy, the test was repeated using Cannon 5D Mark II Fig. 3. Load-displacement curves for tested steel I-beam using LVDT and
digital camera with resolution 21.0 megapixel (5616 3744 10.0 megapixel camera.
pixels) and the results were compared.

643
 IACSIT International Journal of Engineering and Technology, Vol. 4, No. 5, October 2012

LVDT PIV analysis IV. CONCLUSIONS

10.0
In this paper, an experimental study was conducted on the
8.0
performance of Particle Image Velocimetry (PIV) method for
displacement measurement of bending tests. Based on the
Load (kN)

6.0 experimental results of the current study, the following

conclusions are drawn:
4.0 1) Displacement obtained from the image analysis method
extremely matches LVDTs data; hence,
2.0 load-displacement curves can be easily drawn for
bending tests using PIV method.
0.0
2) Camera resolution has no significant effect on
0.0 0.5 1.0 1.5 2.0 2.5 3.0 displacement data obtained from the image analysis
Displacement (mm)
method, although high camera resolution increases the
Fig. 4. Load-displacement curves for tested steel-I beam using LVDT and
21.0 megapixel camera. resolution and accuracy of the displacement field.
3) Due to the lower cost and potential of obtaining the
As it is shown in Figs 3 and 4, the load-displacement whole displacement field on common structural tests,
curves for the test beam obtained from the image analysis PIV method can be widely used as an alternative to
method, very well match the load-displacement curves conventional measurement techniques.
obtained from LVDTs data. In addition, as it is obvious in
Figs. 3 and 4, using 10.0 megapixel or 21.0 megapixel REFERENCES
camera does not affect the image analysis results. Moreover, [1] T. K. Lee and R. Al-Mahaidi, An experimental investigation on shear
displacement field can be also obtained from PIV method. behavior of RCT-beams strengthened with CFRP using
photogrammetry,Composite Structures, vol. 82, pp. 185193, 2008.
For this purpose, the PIV analysis was undertaken using [2] E. M. Mikhail, J. C. McGlone and J. S. Bethel, Introduction to modern
patches of 128 128 pixels, spaced at 64 pixels centers. photogrammetry, Wiley, New York, 2001.
Mesh generation and displacement vectors for the central [3] M. Ali-Ahmad, K. Subramaniam and M. Ghosn, Experimental
investigation and fracture analysis of debonding between concrete and
zone of the tested steel-I beam are shown in Figs. 5 and 6, FRP sheets,Journal of Eng. Mech. ASCE, vol. 132, no. 9, pp. 914-923,
respectively. 2006.
[4] C. Czaderki, K. Soudki and M. Motavalli, Front and side view image
correlation measurements on FRP to concrete pull-off bond
tests,Journal of Composites for Construction. ASCE, vol. 14, no. 4, pp.
451-463, 2010.
[5] M. A. Sutton and H. A. Bruck, Digital image correlation using the
Newton-Raphson method of partial differential
correlation,Experimental Mechanics, vol. 29, no. 3, pp. 261-267,
1989.
[6] R. J. Adrian, Particle imaging techniques for experimental fluid
mechanics,Ann. Rev. Fluid Mechanics, vol. 23, pp. 261-304, 1991.
[7] D. J. White, W. A. Take and M. D. Bolton, Soil deformation
measurement using particle image velocimetry (PIV) and
photogrammetry,Geotechnique, vol. 53, no. 7, pp. 619-631, 2003.
[8] D. J. White and W. A. Take,GeoPIV: Particle image velocimetry (PIV)
software for use in Getechnical testing,Cambridge University
Engineering Department Technical Report, D-SOILS-TR322, 2002.
Fig. 5. Mesh generation for the test beam

Fig. 6. Displacement vectors obtained from PIV analysis (magnification

factor = 5)

644