Accepted Manuscript

Measurement of Strain on Concrete Using an Ordinary Digital Camera

Reza Aghlara, Mahmood Md Tahir

PII: S0263-2241(18)30456-1
DOI: https://doi.org/10.1016/j.measurement.2018.05.066
Reference: MEASUR 5567

To appear in: Measurement

Received Date: 21 February 2016

Revised Date: 10 April 2018
Accepted Date: 17 May 2018

Please cite this article as: R. Aghlara, M. Md Tahir, Measurement of Strain on Concrete Using an Ordinary Digital
Camera, Measurement (2018), doi: https://doi.org/10.1016/j.measurement.2018.05.066

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
 Measurement of Strain on Concrete

Using an Ordinary Digital Camera

1 2
Reza Aghlara , Mahmood Md Tahir

1 Faculty of Civil Engineering, Universiti Teknologi Malaysia, 81310 Johor Bahru, Johor, Malaysia
2 Institute for Smart Infrastructure & Innovative Construction (ISIIC), Faculty of Civil Engineering, Universiti Teknologi Malaysia,
81310 Johor Bahru, Johor, Malaysia

Abstract

Two-Dimensional Digital Image Correlation (2D-DIC) technique was applied in this experimental
study, to measure the steel-concrete bond strain at concrete surface. The bond strain was caused by
tensile force when it was subjected to embed reinforced bar of concrete block in the pullout test. A
series of digital images from surface of concrete were taken by a semi-professional digital camera
(Nikon D80) during the loading. Acquired images were used as input data for two DIC software to
calculate deformation. After verifying the displacement results by measured LVDT results, full-field
strain components on concrete were calculated at various levels of forces. The results show the
measurement of displacement and strain by this method are acceptable to a great extent.

Keywords: Concrete, Strain, Pullout test, Bond, Full-field measurement, Digital Camera, Digital image
correlation, DIC

1 Introduction

The measurement methods in different fields of science have undergone significant developments
because of recent advances in technology. One of these methods is Digital Image Correlation (DIC). It
has numerous capabilities as compared to the previous methods of measurement. DIC has been
increasingly acclaimed by many researchers recently [1,2,3]. In this technique, deformation of the
specimen subjected to the load, is calculated precisely through comparison of digital images and
mathematical operations.

A large number of investigations in material properties and behaviors are still carried out in the
absence of these novel measurement methods [4,5]. Normally, the conventional methods are
performed by merely measuring certain single point on specimen in order to find related displacement
or strain. Therefore, applying these powerful optical techniques (2D-DIC) with the feature of full-field
measurement not only helps to survey the overall behavior of test object, but may also lead to more
discoveries in the studied field. Besides, it has economic justification when the numbers of tests are
quite considerable.

1/17
Generally, most of the studies on adhesion between concrete and steel have still been conducted by
popular and traditional measuring methods [6,7]. In these studies, measurement was mostly carried
out on some specific points through Strain Gauges or LVDTs, and the obtained results were
generalized to other points of the specimen. By contrast, the 2D-DIC method initially uses full-field
measurement. In other words, the deformation for all points of interested area in the specimen is
directly measured by this method without any generalization and interpolation. That is why this
method of measurement can be considered more accurate. Furthermore, bond strain depends on
some sophisticated factors as the direction of concrete pouring and aggregate size, which cannot be
simulated easily by theoretical and numerical techniques. With regard to such complexity, the 2D-DIC
strain analysis can be also considered as a reliable technique, since its measurement is based on
actual data or digital images gained from experimental works. This study aims to illustrate the
efficiency of 2D-DIC and its capabilities in measuring displacement and strain on concrete by an
ordinary digital camera Nikon D80. Such measurement would either be impossible or very costly by
using the common methods.

2 Theoretical Background Of 2D-DIC

2.1 Overview

Measuring deformation on specimen’s surface under load is one of the important aspects of
experimental studies. In addition to using widespread measurement devices such as LVDT, Strain
Gauge and Extensometer, new optical methods have also been introduced in the recent decades.
Generally these techniques can be divided into two major categories, i.e., using interferometer (e.g.
holography interferometry, speckle interferometry and moiré interferometry) and non-interferometer
(e.g. grid method and digital image correlation). The surface deformation in non-interferometer
technique is determined by comparing the grey intensity changes of the object surface before and
after deformation. 2D-DIC method as a representative of non-interferometer optical technique is
widely accepted and commonly used as a powerful tool for the surface measurement. The full-field
displacement and strain are provided directly by comparing digital images of the specimen’s surface in
non-deformed and deformed states respectively. In fact, DIC is an optical metrology created by digital
image processing and numerical computing cooperation. A group of researchers from University of
South Carolina established DIC in the1980s [8].

The experimental setup and specimen preparation are quite simple. For recording the digital images
of specimen before and after deformation, only one fixed Coupled-Charged Device (CCD) camera is
needed. No special preparation of test object is required except for spraying paint onto the surface of
specimen. A white light source or natural light can be used for illumination during loading. It does not
require a laser source. Since this method uses digital images as input data in analysis, every digital
image recorded by different high-spatial-resolution digital image acquisition devices can be processed
in this technique. Nevertheless, the 2D-DIC method also suffers from some weaknesses. Firstly, the

2/17
surface of planar object must have a random grey intensity. Secondly, the quality of imaging system
has a considerable effect on reliability of measurement. It is true that 2D-DIC is limited to in-plane
measurement technique, but with proper design of specimen shape, most of the area can be studied
for measuring the desired factors [9].

2.2 Basics of 2D DIC

Generally, three consecutive steps must be followed in 2D-DIC method: a) specimen and
experimental preparation, b) acquiring digital images of test object surface during loading, c) analysis
of desired displacements and strains, using a computer program to process digital images. In Fig.1, a
classic experimental setup is illustrated schematically using an optical imaging device for 2D-DIC
method. In order to take photos automatically at predetermined intervals, the camera must be
connected to a computer. The surface of the sample, which is illuminated by white light source, must
have an appropriate speckle pattern or it can be artificially prepared by using spray paint to create
white and black spots.

Figure 1. Typical optical image acquisition system for 2D-DIC

The real physical point motion on the specimen surface can be calculated by the estimated motion of
image point and multiplying it by the magnification of the imaging system as long as they fulfil two
necessary requirements for the surface: the surface must be flat and it must be parallel to the CCD
sensor of screen. Any out of plane movement of the object during the experiment leads to changes in
magnification coefficient of the imaging system and errors in displacement calculation of the points. To
decrease the errors, Sutton suggested using Telecentric Imaging System or placing the camera
farther from the specimen [10]. Geometric distortion in image system can also cause errors. These
errors can be avoided using an appropriate camera, or it can be corrected by some techniques [11].

3/17
Images of the specimen are used as input data to 2D-DIC software to determine the displacement of
each point on it. These software such as Rapid Correlator, VIC-2D and Opticist are accessible in the
market. However, some researchers have their own code or use Matlab toolbox. The first step for
calculation in 2D-DIC program is selecting and determining a region of interest (ROI) in the reference
image. This region is divided into equal virtual grids. Calculating the total deformation is conducted
based on displacement of each point on these grids. For this purpose, a small square surface called
"subset" is considered in such a way that the point is located exactly in the centre of the square, (Fig.
2, left) [12]. This subset has a certain number of points and each point has its own relevant grey
value. Therefore, a cross-correlation (CC) or sum-squared difference (SSD) criterion is defined and
calculated for each subset in two consecutive images. After calculating the criteria, their distribution
and selection of the peak, the relevant "subset" can be traced (Fig. 2, right). In this way, the location of
subset in deformed image can be detected for the next operation. With the reference and deformed
subset centres, displacement vector of the desired point is generated on the surface. After this stage,
displacement of the point can easily be calculated numerically through mathematical operations.
Finally, the new location of the other points of subset can be calculated through shape function. This
function is estimated using coordinates of the initial vector of subset centres

Figure 2. Schematic illustrations of a reference square subset before deformation and a deformed
subset after deformation

As stated earlier, a criterion should be defined in order to estimate the degree of similarity between
the reference and deformed subsets, before carrying out the correlation analysis. Numerous
correlation criteria can be found in the literature which is categorized into two groups, namely cross-
correlation (CC) criteria and sum of squared differences (SSD) correlation criteria. For the CC criteria
the commonly used definitions can be named as; normalized cross-correlation (NCC) and zero-
normalized cross-correlation (ZNCC). In SSD, the most applied criterion can be counted as SSD and
normalized sum of squared differences (NSSD), zero-normalized sum of squared differences
(ZNSSD).

2.3 Measurement errors

4/17
Absence of any deficiencies in imaging system during test and precision of the correlation algorithm,
are the two factors that play important role in the accuracy of 2D-DIC results. In order to achieve a
high-quality measurement, identifying the source and the estimating rate of errors are very important.
Generally, sources of error in this method are divided into two different groups; testing error and
analysis error [13]. Errors in the first group are mainly due to the specimen loading and imaging.
There are four important factors that fall under this group, namely speckle pattern, test object position,
image distortions and noises. In the second group, the four factors related to correlation algorithm
have significant influence: subset size, correlation criteria, interpolation scheme and shape function. It
is difficult to differentiate one error source from the others, so the effect of these factors is investigated
separately by experts. According to the relevant literature, based on experimental conditions and
algorithm details, the measures given in Table 1 are recommended to achieve high accuracy in
displacement measurement using the 2D-DIC method [12].

Table 1. Useful points to achieve high accuracy in displacement measurement

Experimental conditions
High contrast in speckle pattern
Parallelism of the sensor and surface
Using a tele-centric lens
Using a low noise CCD camera
Stable camera and even illumination
Algorithm details
Using a larger subset
ZNSSD or ZNCC correlation criterion
Optimizing the CC by NR method
Using a high-order shape function
Using a bi-cubic spline interpolation scheme

2.4 Application of 2D-DIC

A large number of applications of 2D-DIC can be found in the literature over the last few years. They
are generally divided into the following four parts: 1) to determine the deformation field of various
materials subjected to loading [14], 2) to determine the different mechanical properties of a material as
Young’s modulus, Poisson’s ratio, etc.[15], 3) to validate the result of FEM (theoretical analysis) to
bridge the gap between experiment simulation and theory [16], 4) to study the strain gradient, strain
localization and strain concentration of various materials [17].

3 Research Methodology

The pull-out test was conducted on a Universal Machine (UM) available at the structure and material
laboratory of Universiti Teknologi Malaysia (UTM). Images of the specimen’s face were captured
during the entire test. In order to have proper intersection between the specimen surface and bond
stress area of reinforcement, 20 mm thickness was considered as a cover on embedded bar in both
sides of specimen. To avoid the bar from yielding and having ultimate bond stress in concrete, the
anchorage length of reinforcements for Ø12 was selected to be 40cm. The dimensions of concrete
specimens were 400×100×55 mm. The design of concrete mix is developed based on the manual of
structural lab, and the quantities calculated per cubic meter for cement, water, fine aggregates

5/17
(particle sizes less than 4 mm) and coarse aggregate (particle sizes of 4 mm to 10 mm) were 410 kg,
250 kg, 997 kg and 693 kg, respectively. The average compressive strength of concrete cylinders at
age of 28 days was 31 MPa. Before mounting the specimens on universal machine for pull-out test,
the specimens’ surfaces were painted white color with black pattern properly to create a random
speckle pattern. Fig. 3 presents the specimen geometry and the experimental setup.

Moving Grip

Steel Frame

LVDT
400 mm
Pull out
Specimen

55 mm
Fixed Grip
100 mm
mmm
m

Figure 3. The pull-out specimen; a) geometry, b) front view of experimental setup, c) side view of
experimental setup

After installing the specimen between the clamps of UM and adjusting the digital camera on level
tripod in front, the tensile load was subjected to the bar at the rate of 0.05 (kN/sec.). To obtain a
dependable criterion for attained results of correlation analysis, one LVDT was installed next to the
specimen to record the displacement of one specific point on concrete (Fig 3-(b)). Digital images were
taken automatically using camera controller software in the computer. The test configuration is shown
in Fig. 4. The tensile load was subjected to the bar until failure of the specimens. LVDT displacements
were recorded at every 10 second intervals.

6/17
 Figure 4. The pull-out test configuration

4 Results and Discussions

Universal Machine recorded the amount of tensile load subjected to the specimen in kN and stroke in
mm, which is the rate of entire elongation or the increased distance between clamps of UM during
test. The loading rate is adjustable, and it was tuned at 0.05 kN per second in the test. Fig. 5
illustrates the specimen load-elongation behavior subjected to tensile load. As it is seen, the
specimen’s behavior was linearly under pull-out force with ultimate bond strength of 68 kN at
elongation of 4.40 mm.

Figure 5. Pull-out test results of the specimen with Ribbed Bar (d=12 mm)

LVDT results were registered at every 10 second interval to attain a criterion for acceptability of DIC
analysis. Displacements of a certain point in the right edge of the specimen, recorded by LVDT, are
compared with the displacement results of two DIC software for the same point (Fig. 6) which are
almost the same with a difference of about 0.2 mm. Such variations in results of two different software
can be due to diverse algorithm details of each software as explained in section 2.3.

7/17
 Figure 6. Displacement of one specific point on concrete against force in the specimen

After conducting the analysis, the desired displacements and strains were calculated for region of
interest (ROI) of all obtained images. As an example, Fig. 7 shows the major principal bond strain on
the concrete, when the tensile force is 67 kN.

Figure 7. The major principal strain (e1) contour for the specimen at 67 kN

In order to obtain the detailed assessment of the specimen’s behavior, all the full-field distribution of
three main components of strains are plotted on the surface of the specimen using the numerical
results obtained by the DIC software Vic 2D. Fig. 8 shows such plots when the tensile force was 67
kN.

8/17
 67 kN

(a. Surface of pull-out specimen)

67 kN

(b. Major principle strain, e1)

67 kN

(c. Minor principle strain, e2)

67 kN

(d. Principal strain angle, e12)

Figure 8. Surface of pull-out specimen and three main components of principal strains at 67 kN

9/17
Fig. 8.b shows the full-field major principal strain in which the effect of bond distribution between
concrete and reinforcement can be observed clearly. The distinct growth of strain on the top of the
reinforcement bar in the loading side of specimen is due to existence of the concrete-steel bond. The
minor principal strain (Fig. 8.c) shows that the bond has an influence around the bar, but with less
intensity in comparison with the major strain distribution. The principal strain angle distribution in Fig.
8.d illustrates that this parameter is higher on both the top and the bottom than the middle part of
specimen surface. In other words, these two mentioned areas have opposite rotation angles on the
plane of specimen where they were subjected to tensile force, the fact that seems hard to achieve
through the conventional measurement method.

In order to study the variation of the strain when the loading increases, two perpendicular lines were
selected on critical area of the specimen surface with red lines as shown in the legend of Figs 9-14.
Fig. 9-11 show the strains’ changes against the subjected tensile load at the shown line in the legend
along the width of the specimen located at 100 mm from top.

Fig. 9 shows the evolution of concrete strain on the transversal line of the specimen by increasing
tensile load. The second curve in this figure shows the changes of normal strain ex on the mentioned
line, when the load is 12 kN. The total area of these strains has negative values which show that the
line is under pressure in that direction. It is observed that this component of bond strain on the middle
part of line starts to decrease until the load reaches to 32 kN. But the trend has an abrupt change and
starts to increase until it gets its peak point in the positive area. In other words, this middle part of the
line (about 30 mm length) shrinks at initial loads, but it expands during three last loads in the test. It is
possible that this area shows the crack occurrence area, which is located in the middle of specimen.
Having higher strain more than 0.003 in this area is a strong evidence for this claim. In another study
of steel-concrete bonding in a pullout specimen, a similar shape and trend was presented for variation
of concrete strain, which is measured with seven strain gauges along the width of specimen [18]. The
authors believe that the significant changes in the strain after uniform shapes in early loads are
because of the inception of initial cracks in the concrete at the interface of two materials; the concrete
and bar.

10/17
 Figure 9. Normal Strain ex on the certain surface line in various tensile forces

As shown in Fig. 10, the normal strain ey was decreased uniformly in the line, approximately one tenth
of the strain in perpendicular direction. Three useful conclusions can be drawn from this graph. Firstly,
the effect of the bond cannot be distinguished in this component as clearly as the previous strain, ex.
Secondly, looking in the negative area of all curves shows that, as expected, the selected line is in
compression. Thirdly, the trend of this strain decreases as the load increases before the last load, in
which the trend is reversed. This behavior also could be due to the formation of some initial cracks on
bond surface, when the tensile load reaches over 62 kN.

Figure 10. Normal Strain ey on the certain surface line in various tensile forces

Fig. 11 shows the variation of the shear strain of the same line by increasing the load. It can be clearly
seen that the amount of normal shear strain on top of the bar has minimum value compared to other
areas of specimen.

11/17
 Figure 11. Shear Strain exy on the certain surface line in various tensile forces

Figs 12-14 illustrate the strain component changes in a certain line (x=50 mm) on the length of the
specimen by growing tensile load until failure. In Fig. 12, during two initial loads, there is a constant
small decrease in ex at whole line except in the top of specimen that is due to reaction of the support.
The ex of the bottom part has a big drop in strain as the load increases from 22 to 32 kN. It is due to
the undesirable out-of-plane motion at the free part of the specimen. The positive sign of the strain on
this line shows that this part is on tension and is located on top of the bar. The shape and trend of
calculated curves in this figure are comparable to some of the results presented in other investigation,
which are carried out in the bonding field with applied strain gauges as measurement devices [19].

Figure 12. Normal Strain ex on the certain surface line in various tensile forces

Fig. 13 displays the evolution of longitudinal strain (ey) on the same mentioned line along the length of
the specimen. As expected, most parts of the specimen are in compression except a short length on
top near to the support. It appears that the strain increase linearly in the line as the tensile load
increases. Hence, it could be said that the bond doesn’t have significant effects on ey strains along the

12/17
specimen when compared with considerable variation of ex. Out-of-plane motion at the end of
specimen is clearly noticeable just after 22 kN.

Figure 13. Normal Strain ey on the certain surface line in various tensile forces

Fig. 14 illustrates the shear strain on the same line on the specimen. The amount of shear strain is the
highest at the maximum load in most parts of the length. The dispersion of the shear strain on the
right part of the graph is more than the other parts. It can be seen that the bond action can have
considerable changes on the variation of the shear strains. Interestingly enough, considering the Figs
12-14, the most variations on the strains (due to the bond) belong to the transversal strain on the
specimen length with increasing nonlinearly and at the peak point of the loading end. Regarding the
last three figures (Fig. 12-14), it can be deduced that the bond strength has the greatest effect in
transversal strain on concrete, while the longitudinal strains have not been influenced much due to the
bond.

Figure 14. Shear Strain exy on the certain surface line in various tensile forces

13/17
 In some investigations, the study of stress is the main objective and more desirable, which can be
calculated from the obtained results of 2D-DIC measurement. For this experimental work, the stress
distribution on the surface of concrete has been calculated by considering the specimen as plain
stress [20, 21].

5 Accuracy of measurements

In addition of the specimen presented in this study, a total of ten other specimens were also tested in
the same procedure that explained in this paper [21]. The size of all specimen were same but different
diameter of the bars were used. Instead of the normal concrete (NC), Steel Fibber Concrete (SFC)
were utilized in three specimens. The objective of the study was to measure deformation on concrete,
i.e. displacement and strain, through DIC technique by two introduced software: Rapid Correlator and
Vic 2D. Table 2 highlights some of the best obtained results for the accuracy of performed
measurements. These accuracies were determined based on the relevant LVDT results. As it is seen
from the table, the results by one of the software have a lower accuracy relative to the other software.
Since the same digital images were used as input data of the two software, these differences in the
accuracy can be attributed to the type of the correlation algorithm applied by these software.
According to the obtained measurements accuracy by one of the software, it can be claimed that the
proposed method is capable to measure the deformation on the concrete to a great extent.
Additionally, based on the acquired full-field measurement contours, the location of the crack in
specimens can be predicted by this method, as shown Fig.7.

Measurement accuracy (%)

Specimens Description
RapidCo. Vic 2D
1 Overlap 10 97.2 97.3 two 10 mm bars overlapped 10 cm
2 NC T10 65.4 87.4 normal concrete with ribbed bar 10 mm
3 NC T12 86.1 93.2 normal concrete with ribbed bar 12 mm
4 SFC T16 63.3 95.7 steel fibber concrete with bar 16 mm

6 Conclusions

Three components of steel-concrete bond strain are measured and studied on surface of pull-out
specimen using an ordinary digital camera and 2D-DIC method. According to the results obtained in
this study, it could be concluded that this method has a high capability in measuring and investigating
the strain at concrete surface and it has its preferences over to the conventional methods. Some of
these key features of this technique were observed as follows:

1. All components of the bond strain can be measured together at concrete surface by using 2D-DIC
method at each point during the experiment in the same time.

14/17
2. The exact behaviour of bond strain at concrete surface and its growth on specimen can be monitored
by 2D-DIC during the loading.
3. The proposed non-contact DIC method is capable to measure all the components of the strains on all
points of the specimen in form of full-filled distribution at any step of loading, contrary to the
conventional methods that can be performed just for one single point.
4. On average, the obtained accuracy for the proposed measurement method is more than 93% in
determining the displacements for the specimens.
5. Mapping full-field distribution of strains using 2D-DIC method can help one to predict cracking zone on
concrete before occurrence of the failure.

It can be also concluded that by taking some measurements into consideration (such as precise
camera setting, proper specimens shape and experimental setup, high contrast speckle pattern,
having enough control instruments and efficient software), the measurement of the strain at concrete
surface by an ordinary digital camera and 2D-DIC method give appropriate results to a great extent.

7 Acknowledgment

The work reported in this study was graciously supported by Translational Research Grant Scheme
(TRGS) under Ministry of Education awarded to Universiti Teknologi Malaysia with grant number
RJ130000.7809.4L858. The authors remain indebted for the support given by them. Dr Suhaimi Abu
Bakar were deeply appreciated for insightful comments about this study.

7 References

[1] S. Biswal, A. Ramaswamy, Measurement of existing prestressing force in concrete structures through an
embedded vibrating beam strain gauge, Measurement, Volume 83, April 2016, Pages 10-19, ISSN 0263-2241,
DOI:10.1016/j.measurement.2016.01.031

[2] Zhenning Liang, Bo Yin, Hongtao Liu, Jinqiu Mo, Shigang Wang, Displacement measurement of specimen
surfaces with damaged areas by digital image correlation, Measurement, Volume 76, December 2015, Pages
183-188, ISSN 0263-2241, DOI:10.1016/j.measurement.2015.08.026

[3] Bora Gencturk, Kazi Hossain, Aadit Kapadia, Emad Labib, Yi-Lung Mo, Use of digital image correlation
technique in full-scale testing of prestressed concrete structures, Measurement, Volume 47, January 2014,
Pages 505-515, ISSN 0263-2241, DOI:10.1016/j.measurement.2013.09.018

[4] Lee B., Mulheron M.: Fluctuation of bond stress-slip behaviour of deformed bar under displacement control,
Magazine of Concrete Research, 64(10), pp. 863-875, 2012.
DOI: 10.1680/macr.9.00209

[5] Marčiukaitis G., Šalna R., Jonaitis B.,Valivonis J.: A model for strength and strain analysis of steel fiber
reinforced concrete, Journal of Civil Engineering and Management,17(1), pp. 137-145, 2011.

15/17
DOI: 10.3846/13923730.2011.561521

[6] Torre-Casanova A., Jason L., Davenne L., Pinelli X.: Confinement effects on the steel–concrete bond strength
and pull-out failure, Engineering Fracture Mechanics, 97, pp. 92-104, 2013.
DOI: 10.1016/j.engfracmech.2012.10.013

[7] Sharaky I.A., Torres L., Baena M., Miàs C.: An experimental study of different factors affecting the bond of
NSM FRP bars in concrete, Composite Structures, 99, pp. 350-365, 2013.
DOI: 10.1016/j.compstruct.2012.12.014

[8] Peters W. H., Ranson W. F.: Digital imaging techniques in experimental stress analysis, Optical Engineering,
21, pp. 427-431, 1981. DOI:10.1117/12.7972925

[9] Ouglova A., Berthaud Y., Foct F., François M., Ragueneau F., Petre-Lazar I.: The influence of corrosion on
bond properties between concrete and reinforcement in concrete structures, Materials and Structures, 41, pp.
969-980, 2008. DOI: 10.1617/s11527-007-9298-x

[10] Sutton M. A., Yan J. H., Tiwari V., Schreier W. H., Orteu J. J.: The effect of out-of-plane motion on 2D and 3D
digital image correlation measurements, Optics and Lasers in Engineering, 46, pp. 746-757, 2008. DOI:
10.1016/j.optlaseng.2008.05.005

[11] Pan B., Xie H. M.: Assessment and correction of lens distortion for digital image correlation, Acta
Meteorological Sinica, 30, pp. 62-67, 2009.

[12] Pan B., Qian K., Xie H., Asundi A.: Two-dimensional image correlation for in-plane displacement and strain
measurement: a review, Measurement Science and Technology, 20, 062001, 2009. DOI:10.1088/0957-
0233/20/6/062001

[13] Haddadi H., Belhabib S.: Use of rigid-body motion for the investigation and estimation of the measurement
errors related to digital image correlation technique, Optics and Lasers in Engineering, 46, pp.185-196, 2008.

[14] Shibahara Masakazu, Ikushima Kazuki, Itoh Shinsuke, Fukasawa Toichi.: Time-History Measurement of
Welding Deformation Using Digital Image Correlation Technique, International Journal of Offshore and Polar
Engineering, 23(2), pp.152-159, 2013.

[15] Keunecke D., Niemz P., Tollert M., Haensel A.: Mechanical Properties and Failure Behaviour of Yew and
Spruce Determited with a Compact Tension Test and Digital Image Correlation, Wood Research, 57(4), pp. 669-
675, 2012.

[16] Malesa M., Szczepanek D., Kujawinska M., Swiercz A., Kolakowski P.: Monitoring of civil engineering
structures using Digital Image Correlation technique, EPJ Web of Conferences, 6, 31014, 2010. DOI:
10.1051/epjconf/20100631014

[17] Godara A., Raabe D.: Influence of fiber orientation on global mechanical behavior and mesoscale strain
localization in a short glass-fiber-reinforced epoxy polymer composite during tensile deformation investigated
using digital image correlation, Composites Science and Technology, 67,pp. 2417-2427, 2007. DOI:
10.1016/j.compscitech.2007.01.005

16/17
[18] Pedziwiatr J.: Influence of internal cracks on bond in cracked concrete structures, Archives of Civil and
Mechanical Engineering, 8(3), pp. 91-105, 2208.

[19] Yi Na H., Nam Jin W., Kim Sung B., Kim In S., Kim Jang-Ho J.: Evaluation of material and structural
performances of develop Aqua-Advanced-FRP for retrofitting of underwater concrete structural members,
Construction and Building Materials, 24, pp. 566-576, 2010.
DOI: 10.1016/j.conbuildmat.2012.07.046

[20] Aghlara R., Abdullah R., Tahir M. Md.: Mapping Full-field Bond Stress Distribution on concrete using Digital
Image Correlation Results, Journal of Applied Sciences, 12(21), pp. 2235-2243, 2012.
DOI: 10.3923/jas.2012.2235.2243

[21] Aghlara R, Application of 2-dimensional digital image correlation for mapping bond strain and stress
distribution in concrete, Master Dissertation, Universiti Teknologi Malaysia, 2010

17/17