Optical system for real-time measurement of the absolute displacements applied to

flat jack test

M. Cucchi1, C. Tiraboschi2, M. Antico3, L. Binda4

ABSTRACT
During the single and double flat jack tests applied on site to masonry walls, the measurement of
displacements is carried out in order to determine the deformability properties (Young modulus,
Poisson ratio). Traditionally displacement transducers (LVDTs, deformometers, etc.) or a removable
extensometer applied to the masonry surface, are used to measure relative displacements at each stress
increase. Nevertheless some known difficulties are present like: destructive positioning of the devices,
difficulty in the measure of the maximum limit of pressure. Furthermore local anomalies due to the
positioning of the transducers on small stones and bricks (in rubble walls) which can be submitted to
local rotations and displacements during the loading cycles, can influence the test results and
reliability.
The proposed optical system allows the localization in a Cartesian coordinate system of fixed points
within the interested area by using a high resolution video-camera and the analysis of captured frames.
The developed software allows for the continuous acquisition of displacements referred to the fixed
initial image of a high number of points on the chosen area.
The system allows to fix absolute coordinates and to extrapolate information like displacement vectors
of each measured point and also relative displacements between the most representative points of the
interested area.
Keywords: Displacement devices, optical systems, flat-jack

1. INTRODUCTION

The level of damage in a building can be defined only with a deep knowledge of its materials and
structure. Visual inspection, geometrical survey and crack pattern together with mapping of surface
decay are important steps of the necessary experimental investigation when this knowledge is not
available. These operations allow to identify not only the damages but also to formulate preliminary
hypotheses on the causes of damage. At this point other information are needed on the materials and
structure: strength, stress-strain behaviour, constraints and interactions among the structural elements.
Historical masonry is frequently characterised by non homogeneity, a combination of elements
regularly shaped or irregular stones finely cut or roughly cut, but also multiple leaf sections. The
masonry mechanical characteristics important for the structural analysis, cannot be directly deduced
by the ones of their components (brick, mortar, stone) except for rare cases. Therefore tests directly on
the composite have to be carried out. Only two ways can be followed for historical buildings: a)
sampling on site of a masonry prism later on tested in laboratory, b) on site characterisation of the
masonry.
1
Technician DIS- Politecnico di Milano [email protected]
2
Assistant Professor DIS- Politecnico di Milano [email protected]
3
Technician DIS- Politecnico di Milano [email protected]
4
Honorary Professor DIS- Politecnico di Milano [email protected]
Sampling on site is a difficult operation even if the masonry is regularly built and monolithic [1]. Due
to the fact that it is destructive, no more than one or two specimens can be extracted for each building
with a very low statistical reliability given the heterogeneity of the masonry. If sampling has to be
ruled out together with on site destructive testing (on masonry panels), then the choice is to carry out a
low destructive test on site in order to know as much as possible mechanical parameters.
Among the proposed on site tests the single and double flat jack tests set up in 1981 [1] seems to be
the most interesting. In the paper a brief description of the tests will be given. Essentially the test takes
into account the behavior of masonry under compression and the method consists in measuring the
local displacements (hence strains) caused by varying compressive stresses.
Traditionally displacement transducers (LVDTs, deformometers, etc.) or a removable extensometer
applied to the masonry surface, are used to measure relative displacements at each stress increase.
Nevertheless some known difficulties are present like: destructive positioning of the devices, difficulty
in the measure of the maximum limit of pressure, local anomalies due to the positioning of the
transducers on small stones and bricks (in rubble walls) which can be submitted to local rotations and
displacements during the test, can modify the test results and reliability.
The optical system proposed by the authors allows the localization in a Cartesian coordinate system of
fixed points of the interested area by using a high resolution video-camera ( with a matrix of 2448
x2048 pixels) and the analysis of captured frames. The developed software allows for the continuous
acquisition of the absolute displacements of a high number of points on the surface of the chosen area
(between the two jacks in the case of the double flat jack tests) with resolution higher than the
twentieth of a pixel.
The system allows to extrapolate information like displacement vectors of each measured point on the
chosen area and also relative displacements between the most representative points of that area.
The advantage of the system goes from the time spared in comparison with the application of the
traditional LVDTs, to the non invasive approach. Furthermore the principal peculiarities are: the
possibilities of observing movements impossible to be detected by the classical devices, the
knowledge of the maximum limit of the test, the local anomalies, the stress concentration due to local
eccentricity.

2. TRADITIONAL DISPLACEMENT MEASUREMENTS IN FLAT-JACK TESTS

2.1. Single flat jack

The single flat jack test is based on the relaxation of stresses caused by a horizontal cut in a masonry
subjected to a uniform distribution of compressive stresses (Fig.1a). Due to this relaxation the cut
tends to close up. Measurement of the distance between four points across the cut before and after
cutting gives a geometry variation as consequence of cut (Fig.1c) [3],[4].

a) b) c)
Fig. 1 a) horizontal cut by semicircular saw, b) flat jack, c) effect of relaxation after cutting
If a flat jack is inserted into the cut and continuously inflated pumping oil, the cut will slowly reopen
until the primitive geometry has been recovered. The pressure given to the flat jack can be measured at
each step so as the variation of the distance between the four reference points, until this distance
reaches zero; this value multiplied by two flat jack constants obtained by calibration allows to
calculate the stress value.

2.1.1 Displacement measurement

The displacements of the masonry are collected by a mechanical gage extensometer which measures
the distance between fixed gage points on the masonry as shown in Fig. 2. The device used shall be
capable of deformation measurements up to 5 mm. Deformation measurements shall have an accuracy
of at least 0.005 % of gage length [2].
Stress [N/mm2]
250
1 2 3 4 0 (after cutting)
200 0 (with flat jack)
0.08
0.16
150
0.23
0.31
DL [mm]

100 0.39
0.47
50 0.55
0.62
0 0.70
0.78
0.86
-50
0.94
1.01
a) b) -100 1.09 c)
Fig. 2 Flat jack Test Setup for In Situ Stress Measurement

At least four pairs of equally spaced embedded plugs vertically aligned above and below the slot are
fixed as shown in Fig. 2. Alternative instrumentation configurations are acceptable if controlled
laboratory tests are conducted to verify the validity of the alternate instrumentation approach. [2]

2.2. Double flat jack

Flat jacks can be used also to study the stress-strain behavior of masonry under compression. After a
first horizontal cut has been done, a second parallel cut is carried out in order to isolate a sort of
masonry specimen of 350x500mm dimensions. The masonry block is subjected to mono-axial
compression test by inserting two jacks into the two slots, connected to the same hydraulic pressure
system (Figs.3) [3], [4],[5],[6].
Loading-unloading cycles are carried out by sending oil in pressure into the two flat-jacks and
increasing the pressure step by step. It is advisable to reach (in order to calculate the elastic
parameters) and exceed the elastic limit of the material and possibly the onset of cracking. In this way
the stress-strain diagram of the material can be traced (Fig.4b).

2.2.1 Displacement Measurement

Displacement devices are mounted after the points are fixed. Measurement brackets for mounting
electrical displacement measuring devices are attached securely to the surface of the masonry to
prevent movement and ensure the required measurement accuracy (Fig. 3a, b). Also horizontal
displacements can be measured (fig. a, b, c).
Pressure is slowly increased and displacement measurements at small increments of pressure are
taken (Fig.3c).
 2.23 m 3.0

1 2 3 4 2.5

Stress [N/mm2]
2.0

1.5
1,02 N/mm2
5 1.0
Local stress

0.5

1.70 m eh ev
0.0
-0.5 -0.25 0 0.25 0.5 0.75 1 1.25 1.5
a) b) Strain [mm/mm] c)
Fig. 3 a), b) Slots, flat jacks and measurement system, c) Pressure system in the jacks

3. DIFFICULTIES IN MEASUREMENT AND THEIR CAUSES

Elaboration of data and its interpretation represent a serious difficulty for the comprehension of in situ
test in general. Data collected during the in-situ test are influenced by the masonry typology, the test
position, the presence of consistent discontinuity inside the masonry.
In the case of historic masonry buildings, especially irregular stone walls it is frequently difficult to
position the devices as suggested by the standards due to: a) the presence of courses of irregular
thickness, b) irregular vertical profiles due to irregularly shaped stones which do not allow to read the
displacements unless smoothing of the surface is not carried out, c) variable length of the distance
between the gage points. This last point causes frequently large differences in the stress-strain
diagrams and hence in the calculation of the elastic modulus. It should be advisable to set the gage
points as near as possible to the jacks; nevertheless in case of irregular stone masonry it is possible that
the stone elements position is disturbed by the cuts and that stones are submitted to rotation and
displacements amplified. To increase the number of gage points does not help. Only the possibility of
referring the measured displacements to absolute movements of the gage points can individuate
independent movements of the single stone.
Since with this test it is impossible to reach failure without being destructive, the double flat jack test
is useful at least to reach a state of stress beyond the elastic range and eventually the onset of cracking.
The value of the stress induced by compression can be greater than the stress measured by the single
flat jack thanks to the stress distribution to masonry portions larger than the flat jack area.

4. DESCRIPTION OF THE OPTICAL SYSTEM

In order to create an alternative method able to follow all the movements of the specimen points by
referring them to an initial X,Y system, an optical method is proposed by the authors and described
below. Similar methods were proposed in [7] and [8]
A high resolution monochrome video camera (Mod: uEye UI-2280SE-M) is used with sensor type:
CCD 2448 x 2048 Pixel (5.01 Megapixel), electronic global shutter, maximum frame rate: 6,5/sec.
The camera is directly connected to a personal computer via USB interface. The video signal is
converted into 8 Bit digital format and analysed in real time. Image processing algorithms are used,
real time XY calculations from camera images for continuous displacement measurement are made.
The load history is synchronized with displacements tracks; the pressure gauge is connected to a PC.
The video optical extensometer principle is based on tracking markers fixed to the specimen surface.
The video camera is mounted on a tripod and focused on reference points fixed on the wall. Each
frame analysed gives information about the position X, Y of the markers respect to an origin fixed on
the first frame. The markers are black dots printed on white paper (to obtain the best contrast), their
size has to be proportional to the FOV (Field of View). To measure the absolute displacement it is
imperative that the distance between the video camera and the wall remains constant during the test.
The interested area should be illuminated at a constant level during the test. The aberration of the
lens and perspective errors are corrected using a calibration grid.
The software system proposed (load acquisition and image processing) is based on a virtual instrument
developed with LabView Vision platform (Fig.4).

a) b) c)

a) Marker and area of research

b) Found markers
c) Horizontal displacements
d) Istantaneous vectors
e) Relative displacements between
selected points
f) Absolute vertical displacements
g) Displacement and force values

g)
d) e) f)
Fig. 4 Program screenshot

The developed software allows controlling and optimizing the camera settings such as pixel clock,
exposure time, contrast, gain. For tracking markers, pattern matching techniques are used. These
techniques include normalized cross-correlation, one of the common methods for finding a template in
an image. Due to the large amount of data obtained from each picture acquired, it is essential that these
are efficiently processed. The process of locating and tracking markers was optimized defining small
areas of interest around each reference marker (fig 4b).
The program allows viewing: captured images and graphs, absolute displacements X and Y, relative
displacements between points selected and vectors displacements for each reference markers (Fig. 4 c,
d, e, f). To follow the test also pressure gauge readings are stored and displayed at the same time.
System calibration. The video optical system was calibrated in laboratory by means of a dedicated
calibration device (Fig.5). Nine marker displacements were compared with a displacement transducer.
To move the markers a system driven by a smart motor with a step resolution larger than 1 mm was
used.

Fig. 5 Calibrating device

A controlled history of displacements was imposed up to 1000 mm with steps equal to 10mm from 0 to
200 mm and with steps equal to 100 mm from 200 to 1000 mm .
During the calibration the system tracks 1 marker as reference and 9 markers to measure relative
displacements.
1000

0.10
Measured Displacement (mm)

10
800 0.08
8

Absolute Error (Pixel)

0.06

Absolute Error (mm)

6
4 0.04
600
Applied 2 0.02
Marker 1
0 0.00
Marker 2
-2 -0.02
400 Marker 3
Marker 4
-4 -0.04
Marker 5 -6 -0.06
Marker 6 -8
200 -0.08
Marker 7 -10
Marker 8 -0.10
Marker 9 0 200 400 600 800 1000
0 Applied Displacement (mm)
0 200 400 600 800 1000
Applied Displacement (mm)
a) b)

Fig. 6 Calibration of the system: a) calibration curve, b) estimated absolute errors

The obtained results show a resolution higher then 1/20 pixel and a noise less than 1/50 pixel; with a
CCD matrix of 2448x2048 pixel the system resolution is large than FOV-length/48960. In this
configuration the size of the Field of View is equal to 300x250 mm that correspond to a resolution
higher than 6mm. The maximum absolute error found is equal to 0.036 Pixel which correspond to 4mm
(Fig. 6a, b).
The values found are compatible with the specifications of ASTM and therefore it was decided to
apply the Video Optical Extensometer system in the case of single and double flat jack tests.

5. RESULTS FROM APPLICATIONS AND COMPARISON WITH THE

TRADITIONAL SYSTEM

5.1. Single flat jack

The system is not yet utilized for the single flat jack test, due to the difficulties in either protecting the
video camera or repositioning the markers; in fact during the cutting the initial reference can be lost.
Nevertheless some markers were positioned symmetrically across the cut at progressively increasing
distances in order to control the symmetry of the displacements during the re-loading phase to reach
the original stresses.
The results of a test carried out with the traditional system (4 gauge points across the cut and use of
removable extensometer) are reported in Fig7a. Fig.7b shows a screen shot made during a loading
step. In detail it is possible to see: a) the reference marker, b) the grid of 20 markers (90° rotated), c)
the horizontal trajectories of the 20 markers along the time, d) the markers displacement vectors, e) a
stress-relative displacement between two markers on two adjacent stones, f) the opposite vertical
trajectories of the markers across the cut. The optical system confirms the symmetry of the
displacements measured by the extensometer larger at the central part of the cut and decreasing with
increasing distances from the cut, as expected.
 150
1 2 3 4
Displacement [ mm]

100

50

-50
0.0 0.5 1.0 1.5 2.0
Stress [N/mm2]
b)
a)

Fig 7 Single flat jack: acquisition and elaboration: a) traditional system by extensometer, b) optical system

5.2. Double flat jack test on site

The test was carried out at L’Aquila in the Spanish Fortress. Figs 8a and b show the stress-strain
diagram obtained with both systems for basis 1 and 2 on which also the markers were fixed. These
figures show a nearly perfect coincidence of the measured displacements.

2.5 2.5
P2 P3 P4

2.0 2.0
Stress [MPa]

Stress [MPa]

1.5 1.5

1.0 1.0

0.5 0.5
P3
5-6
0.0 0.0
0.0 0.5 1.0 1.5 2.0 0 2 4 6 8 10
Vertical Strain [mm/mm] Vertical Strain [mm/mm]
c)
a) b)
Fig 8 a) e b) Comparison between stress-strain curve obtained with Gefran rectilinear displacement transducer
(potentiometers P) and optical system; c) transducers and markers on the wall A4

Fig.9 b reports the results obtained with the optical system for two couple of points shown in fig.9a.
The two stress strain curves obtained by the relative displacements between the point 14 and 10 and 7-
13 (diagram in fig.9b) are very different; the reason is explained by the trajectory of the single
points, shown in fig. 9c. In fact the stone with the markers 14 and 13 rotates during the test and falses
the results.
 2.5 -1E+005

Vertical Displacement (micron) x 200

Stress [M Pa]
7-13 0
0.24
2.0 10-14 0E+000 0.48
10 7 0.73
1.01
Stress [MPa]
4 1 1.21
1.5 1E+005 1.57
1.73
2.05
2.17
1.0 2E+005

14
0.5 3E+005
13 12 11

0.0 4E+005
0 1 2 3 4 5
Vertical Strain [mm/mm]
a) b) c)

Fig. 9 Optical system: (a) observed masonry; (b) vertical deformations; (c) amplified vertical displacements.

5.3 Double flat jack in laboratory

Fig.10 shows the stress-strain diagram obtained from different couples of markers chosen on purpose
for a double flat jack test on a three leaf stone masonry wall tested at the laboratory of Department of
Construction and Transportation Engineering of the University of Padova in a common research. This
method allows for controlling the displacements in many different positions and of obtaining reliable
diagrams.
5
Y10-Y11
Y6-Y0
4 Y4-Y14
Stress (N/mm2)

Y9-Y1
3 Y5-Y7
X15-X12
X2-X11
2 X5-X10

eh ev
0
-8 -6 -4 -2 0 2 4 6 8
Strain (mm/mm) a) b)
Fig. 10 a) Stress-strain diagram obtained from different couples of markers on the specimen A5; b) Markers
fixed on the wall and evidence of a vertical crack

Figures 11a (specimen A4) and b (specimen A5) show the movements of some chosen points under
the loading-unloading cycles. The diagrams representing the trajectories of the points allow an
intuitive interpretation of the behaviour of masonry under compression and show that together with
vertical displacements also large horizontal displacements are present locally. In Fig.11b large
horizontal displacements are present in the right part of the specimen. In fact a vertical crack
developed in that area (Fig.11c).
 Vertical
Displacement
Y Y [mm]

X X

a) b)
Fig. 11 Amplified trajectories of markers and map of final displacements in A4 (a) and A5 (b)

6, CONCLUSIONS
The traditional measurement methodology used to follow the displacements in chosen points of the
specimen under stress variation is not exhaustive for the movement description in the case of irregular
stone walls, because it is detecting the relative displacements between the gage points without taking
into account the displacements referred to an initial configuration.
Therefore an optical system is proposed based on the use of a video camera as a video optical
extensometer. The related software allows for the continuous acquisition of the displacements referred
to the initial system on the first image in a high number of points, with a resolution higher than the
twentieth of a pixel.
Furthermore the advantage of the proposed system goes from the time spared in comparison with the
traditional system to the non-invasive approach.
Examples of application of the optical system on site and in laboratory are reported from which it is
possible to compare the results with LVDT measurements. The comparison shows the advantages of
the proposed method which allows getting more useful information on the point movements, detecting
vertical and horizontal displacements. Also limits of the method have emerged, which require to
continue the research.

ACKNOWLEDGEMENTS
Authors wish to thank dr. G. Cardani for the assistance during the tests, Dr. Lorenzo Cantini for the
help during the on-site tests and Dr. Mariarosa Valluzzi for the collaboration in the laboratory tests
which were carried out at the DCT, University of Padua. The work was supported by the funds of the
DPC - ReLUIS 2010-2013 Contract (Rete dei Laboratori Universitari di Ingegneria Sismica), linea di
ricerca 1, area tematica 1, task 1 and NIKER (New Integrated Knowledge based approaches to the
protection of cultural heritage from Earthquaque-induced Risk) contracts.

REFERENCES
[1] Rossi, P.P. (1982) Analysis of mechanical characteristics ofbrick masonry tested by means of
non-destructive insitu tests. In: 6th International BrickMasonry Conference, Rome
[2] ASTM C 1196 (2009) In situ compressive stress within solid unit masonry estimated using flat
jack measurements
[3] ASTM C 1197 (2009) Standard test method for in-situ measurement of masonry deformability
properties using the flatjack method
[4] Schuller, M.(2001) Flatjack Methods for Diagnosis of Modern Masonry. In: Proceedings, On-Site
Control and Evaluation of Masonry Structures, Binda, L., deVekey, R., editors, RILEM
[5] Coombs J., Tanner J.E. (2008) Development of Laboratories for Masonry Testing and Non-
Destructive Evaluation The Masonry Society Journal, Vol 26, No. 2, pp. 9-20.
[6] Noland, J.L., Atkinson, R.H. and Schuller, M.P.(1990) A review of the flat-jack method for
nondestructive evaluation. In: Proceedings of Non-destructive Evaluation of Civil Structures and
Material, Colorado
[7] Francesco Bianconi (2009) Sul controllo della fessurazione nelle strutture civili Tesi di Laurea
specialistica in Architettura, Politecnico di Milano, Facoltà di Architettura e Società (advisors:
R.Felicetti, G. Cardani; M. Lamperti), Milano
[8] V. Quaglini , C. Corazza, R. Contro, C. Poggi (2003) An Integrated Optical – Analytical
Approach For High Deformable Textiles And Tissues. In: 16th AIMETA Congress of
Theoretical and Applied Mechanics, Ferrara