See discussions, stats, and author profiles for this publication at: https://www.researchgate.
net/publication/315787587
Impact Testing of Concrete - The Measurement Device
Conference Paper · August 2014
CITATIONS READS
0 966
3 authors:
Petr Máca Radoslav Sovják
Technische Universität Dresden Czech Technical University in Prague
29 PUBLICATIONS 172 CITATIONS 41 PUBLICATIONS 168 CITATIONS
SEE PROFILE SEE PROFILE
Petr Konvalinka
Czech Technical University in Prague
83 PUBLICATIONS 224 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Structural behaviour of double shell tunnel linings View project
Experimental development of high-performance concrete with improved resistance against shock loading and high energy absorption and dissipation capacity View project
All content following this page was uploaded by Radoslav Sovják on 05 April 2017.
The user has requested enhancement of the downloaded file.
Proc. of the Intl. Conf. on Advances In Civil, Structural and Mechanical Engineering- CSME 2014.
Copyright © Institute of Research Engineers and Doctors. All rights reserved.
ISBN: 978-1-63248-025-5 doi: 10.15224/ 978-1-63248-025-5-66
Impact Testing of Concrete
The Measurement Device
Petr Máca, Radoslav Sovják, Petr Konvalinka
Abstract—Testing of ultra high performance fiber reinforced Concrete in compression or direct tension subjected to high
concrete under impact loading is described in this paper. In strain rates is usually tested by utilization of split Hopkinson
addition a design process of a novel impact measurement device pressure bar (SHPB). The compression loading can be uniaxial
is presented in this work. There are several ways how to test [5-7] or biaxial. The biaxial method is still in development but
impact resistance of concrete and various groups throughout the promises good results [8, 9]. The basic principle of the method
world developed different testing devices. Most of such devices is that the specimen is placed between two long aluminum
are based on a principle of a falling weight on the concrete bars and an impulse is developed by a pressure gun into one of
specimen. This work however presents an impact machine that is the bars. Impulse transmission is monitored by strain gauges
based on a pendulum principle. Such test configuration has
and the difference between the magnitude of the pressure
several advantages, such as elimination of double hits, easy access
to the sample and high degree of device modulability. In addition,
wave before and after the destruction of the specimen is
the placement of sensors and high speed camera is relatively easy. acquired. The main disadvantage of this method is the fact,
In this experimental work concrete beams were tested in the that the specimens are relatively small, usually cylinders with
testing device, but the test setup can be rearranged to test slabs or a diameter of 50 mm. This paper describes a new machine that
joins. The calibration process of the machine is shown in this has been constructed specifically for dynamic testing of
paper as well as the results from impact resistance measurement concrete. At the moment the machine is in the bending loading
of high performance fiber reinforced concrete. configuration but it is relatively easy to transform it into the
slab testing setup. The instrumentation and construction
Keywords—impact, high performance concrete, horizontal process of the machine is described in the paper.
loading
II. Principle of the impact
I. Introduction
Response of concrete subjected to impact loading is of the
machine
main interest of several military and civil applications. In The required horizontal loading setup and space capacity
recent years there is an increase interest in testing concrete of the laboratories of the Czech Technical University governed
elements by using proper dynamic analysis. When it comes to the design of the impact machine. The target impact kinetic
impact testing of concrete, the main issues encountered by energy was set to 2000 J in the initial design, resulting in
many researchers are insufficiently fast data acquisition machine capable of dropping a 50 kg impactor from a height
systems, vibrations in the loading frame and resonant of 4 m. In general, for impact test machines, there are several
frequency of the sensors. These issues then lead to apparent possible setups that can be divided into two groups, according
energy losses and large variance of the results between the to the direction of impact i.e. horizontal and vertical impact.
laboratories. The large variance of the results is also connected The setup presented in this paper utilizes horizontal impact
with different techniques used for studying mechanical direction with stationary specimen where the impactor is
properties of concrete and concrete structures under dynamic following a circular trajectory using guide elements. The
loading. For instance, several authors used modified Charpy schematic diagram of the pendulum machine is shown in Fig.
impact test [1] to study impact behavior of concrete in 1. The energy of the impact can be varied by placing the hoist
bending. The modified Charpy test is limited by size of the of the impactor in different positions along the left column.
sample and weight of the impact pendulum. Another way how The impactor is shown in three possible positions-heights in
to assess impact-bending properties of concrete are methods Fig. 1. The specimen is placed near the bottom return point of
based on a principle of weight falling down on the sample the impactor trajectory. This setup has several advantages and
from specified height, so called drop tower tests [2-4]. The disadvantages, compared to the more conventional vertical
disadvantage of these methods is the fact that a so called impact. The advantage is that it is easier to avoid double
double hit caused by rebound of the impactor can occur in impact just by geometric arrangement. The specimen is placed
case when the specimen is not fully broken by the first blow of little further on the impactor trajectory, and due to the circular
the impactor. shape, little above, i.e. the point of impact is not the least
potential point on the trajectory. Once the impactor hits the
specimen, it rebounds and tends to balance in the bottom-
return point of trajectory, not hitting the specimen again.
Petr Máca, Radoslav Sovják, Petr Konvalinka
Czech Technical University in Prague, Faculty of Civil Engineering
Czech Republic
63
Proc. of the Intl. Conf. on Advances In Civil, Structural and Mechanical Engineering- CSME 2014.
Copyright © Institute of Research Engineers and Doctors. All rights reserved.
ISBN: 978-1-63248-025-5 doi: 10.15224/ 978-1-63248-025-5-66
Figure 3. Specimen support group, top view.
To eliminate the resonant frequencies, the design change lead
to replacing the rigid guide with flexible cable guides. Such
solution smoothed out the acceleration readings, but resulted
in the impossibility of dropping the impactor from heights
above the circular trajectory axis.
The specimen itself is connected with steel yokes to the
Figure 1. Schematic design of an impact pendulum machine. The impactor is supports that are instrumented with load cells. The placement
shown in several possible launching positions.
of the specimen is shown in detail in Fig. 3, the view is from
The machine itself consists of several structural groups as above. The steel yokes allow the specimen to rotate freely
shown in Fig. 1. These groups are the impactor support group, along the support as is shown in Fig. 4. This setup has an
impactor guide group (Fig. 2) and specimen support group that additional advantage as it restrains the uplift of the specimen.
is shown in detail in Fig. 3. These groups are highly modular At the moment the machine is prepared for measuring beams
and can be easily replaced. The impactor itself can be replaced with a cross-section of 100×100 mm. The span of the beam
as well. There were already five designs tested so far with can be from 300 mm to 1000 mm.
different weights and cross-sections from welded lightweight
I-beams to solid rectangular steel bars. The impactor itself is
A. Measurement
instrumented with a set of accelerometers as shown in Fig. 2. Proper dynamic analysis that takes into account inertial
It is possible to place the accelerometers to various positions forces needs to be performed to adequately measure impact
along the impactor. The most interesting placement is near the behaviour of larger concrete elements where high
tup of the impactor where four positions are available – top, accelerations exist. There are several methods how to diminish
bottom, right and left from the tup. It is also possible to place the influence of the inertial forces. Sometimes an attenuator,
an accelerometer to the rear face of the impactor to measure usually made of plywood, is introduced in the contact zone
the vibration of the impactor itself. between the loading impactor and the specimen [10, 11]. To
overcome the inertia, it is also possible to carry out a
As long as the machine was intended as a prototype from numerical analysis but this is fairly complex task. An
the beginning, all parts are designed completely experimental way how to deal with inertia was suggested in
interchangeable and modular. It is possible to adjust the their work by Bentur et al. [12]. They proposed direct
position of the axis of the impactor circular trajectory in measurement of acceleration along the concrete specimen. In
500 mm increments, from 1.5 to 3.0 m above the specimen case a linear distribution of acceleration along the specimen is
position. Initially, the design was supposed to utilize drop assumed, it is possible to calculate the acceleration at the mid-
height of double the axis height (with top point of trajectory span of the specimen that is denoted ̈ . If the mass density
above the axis level), using rigid guides. The rigid guides were and cross-sectional area remain constant throughout the
designed as lightweight as possible, not to interfere with the length of the beam, it is possible to calculate the inertial load
impact mass, using duralumin profiles. During the testing of as
the machine, it was found out that the resonant frequency of
the rigid guides was very near to the measured values during
impact on concrete. For this reason it was impossible to filter ̈ [ ]
this resonant frequency from the accelerations measured at the
impactor tup and the design had to be changed. Where is the inertial load at time , is the mass
density, is cross-sectional area, ̈ is acceleration at the
mid-span, is the length of the mid-span and is overhanging
part of the beam. The actual bending load can be then
obtained by subtracting the generalized inertial load
from the observed tup load under the assumption of
single degree of freedom system.
Figure 2. Location of accelerometers along the impactor.
64
Proc. of the Intl. Conf. on Advances In Civil, Structural and Mechanical Engineering- CSME 2014.
Copyright © Institute of Research Engineers and Doctors. All rights reserved.
ISBN: 978-1-63248-025-5 doi: 10.15224/ 978-1-63248-025-5-66
with 8 channels that can be sampled by rates up to 2 MHz at
16bit resolution.
III. Calibration of the machine
During the calibration process of the machine a steel
specimen was used. This allowed to measure the response of
the beam without the influence of fracture. Steel with yield
strength of 235 MPa and modulus of elasticity of 210 GPa was
used during this process. Impactor was dropped from the
height of 0.3 m resulting in impact velocity equal to 2.42 m/s.
The impact energy was 120 J. Both reactions exhibited almost
identical signal which indicated that the sample was hit in the
mid-span. During the experimental work it was found that
Figure 4. Instrumented supports that allow rotation of the sample and
prevent an uplift of the specimen during an impact. sampling rates of 500 kHz are adequate enough.
The force measured from the load cells and the total
In some cases such as highly fibre reinforced concrete as well reaction over time is shown in Fig. 5. Four accelerometers
as concrete with continuous steel reinforcement, the data were mounted on the impactor’s tup to measure the
obtained from the accelerometers mounted on the specimen acceleration of the impactor during the impact process. During
are heavily influenced by reflected pulses and vibrations of the the calibration process, it was found that it is very complicated
whole impact machine. For this reason an impact pendulum to read the acceleration on the impactor due to the high
machine presented in this paper was instrumented with load stiffness of the sample and the fact that high resonance
cells under supports (Fig. 4) to measure the reaction forces. occurred. For this reason an impact attenuator was used. This
These reaction forces are not influenced by the inertia of the approach is also adopted by other researchers [10, 11]. The
sample, while the load measured at the tup of the impactor attenuator was made of 4 mm thick plywood and was placed
contains the inertial forces of the specimen. in the contact area between the sample and the impactor. The
measured impact is easily readable as is shown in Fig. 6.
B. Instrumentation used Maximal acceleration was determined to be 3,000 m/s2 and the
Measurement of dynamic properties of materials is quite duration of the force impulse was 2 ms.
complex problems and requires proper instrumentation.
Inadequate data acquisition systems and low sampling rates Using Newton´s second law of motion the force response
can significantly influence the results. Two piezoelectric force gained from the impactor could be compared to the force
sensors were used to measure the reaction forces between the signal gained from the load cells attached to the support
support and the specimen as shown in Fig. 4. The load cells reactions as is shown in Fig. 7. No signal filtering was used on
200C50 were produced by PCB piezotronics and the maximal the signal from accelerometers and therefore a slight deviation
capacity of each load cell is 222 kN. Because these load cells between the signals can be seen as shown in Fig. 7. The time
are not capable of measuring tensile forces, the whole setup difference between the maximal peek measured by the
was prestressed to acquire these tensile forces. The accelerometers on the impactor and the maximal force at the
acceleration of the specimen as well as accelerations of the supports is caused by the length of the trajectory. The distance
impactor can be measured by up to five accelerometers that the impulse wave needs to travel from the impact incident
350B04 produced by PCB with a maximal capacity of to the reactions is approximately 0.3 m.
50,000 m/s2. These accelerometers can be placed into different
locations as described above depending on the experimental
setup. The high resonant frequency of the instrumentation is
particularly important because it can significantly influence
the results. The resonant frequency of the accelerometers is
higher than 100 kHz.
The data acquisition system is triggered by two diffuse
electric sensors OGT 500 that also allow measuring the
accurate speed of the impactor and these are located at the
lowest point of the impactor trajectory. The distance between
the light gates is 200 mm and their switching frequency is
2 kHz. The gates were produced by IFM Electronic. The data
acquisition system consists of a signal conditioner (483C05)
produced by PCB with a frequency response higher than
1000 kHz. The signal conditioner decouples the AC signal
from the DC bias voltage and provides constant current
excitation to the PCB sensors. The decoupled signal is then
acquired by a high precision analogue DAQ board ME-5265
65
Proc. of the Intl. Conf. on Advances In Civil, Structural and Mechanical Engineering- CSME 2014.
Copyright © Institute of Research Engineers and Doctors. All rights reserved.
ISBN: 978-1-63248-025-5 doi: 10.15224/ 978-1-63248-025-5-66
Figure 5. Response of the support reaction to impact loading of the concrete
specimen.
The beams were tested in three point bending, where the
impactor struck the beam in the middle of the deflection. The
reaction forces were measured by the piezoelectric load cells.
In addition, acceleration of the beam was measured in the
middle of the span. Afterwards the deflection of the beam was
calculated by double integration of the measured acceleration.
This was also compared to the data gathered form a high-
speed camera that was installed above the specimen. The
sampling frequency for the camera was 8 kHz and for the
piezoelectric sensors 500 kHz. No filtration was used in this
particular case. The difference between the deflection obtained
from the camera image analysis and from the accelerometers
is shown in Fig. 8. It shows a very good correlation between
those methods of measurement.
Fig. 9. shows a comparison between there point bending
behavior for UHPFRC beams loaded by different loading
speeds. The quasi-static loading was performed by standard
Figure 6. Response of the impactor to the impact usage of an attenuator.
hydraulic testing system. A closed loop deformation control
system was used during the experimental work with a constant
loading rate of 0.2 mm/min. Under the assumption of linear
elastic behavior of the material, this corresponds to the strain
rate of d/dt = 2.2×10-5 s-1 at mid-span on the bottom of the
specimen. The deflection was measured by two linear variable
differential transformers (LVDT) positioned in the middle of
the span at the sides of the specimen.
The dynamic loading was produced by the horizontal
impact machine described in detail in the previous paragraphs.
The impact height in this case was 70 cm and the impactor
weight was 37 kg. Neglecting friction and other losses, this
corresponds to 254 J of kinetic energy at the loading point.
The average quasi-static strain rate was observed to be in the
range of d/dt = 3.2 s-1 and according to [15] is comparable
with low velocity impacts such as vehicle crash or falling
Figure 7. Response of the impactor to impact loading of the specimen rocks. The specimen didn’t completely fail under the impact
compared to the support reactions with using attenuator. loading. After the initial impact, it started to vibrate and
impactor continued to increase its deflection. This can be seen
as a vibration response in the Fig. 9 for the dynamic loading
IV. Impact Resistance of Concrete curve.
After the successful calibration of the machine, the impact
resistance of ultra high performance fiber reinforced concrete
(UHPFRC) was measured. For the purpose of this
experimental work, beams with the dimensions 100×100×400
mm were casted using the UHPFRC mixture with 2% of steel
fiber reinforcement. The mixture components as well as the
mixing procedure is described elsewhere [13, 14]. The basic
UHPFRC mechanical properties, are summarized in Tab. 1.
TABLE I. BASIC MECHANICAL PROPERTIES OF UHPFRC
Property UHPFRC
Fiber type 0.15×13 mm
Workability – spread 225 mm
Compressive strength 150 MPa
Flexural strength 40 MPa
Direct tensile strength 10 MPa
Modulus of elasticity 56 GPa
Figure 8. Beam defelction in time as obtained from acceleration and image
analyis
66
Proc. of the Intl. Conf. on Advances In Civil, Structural and Mechanical Engineering- CSME 2014.
Copyright © Institute of Research Engineers and Doctors. All rights reserved.
ISBN: 978-1-63248-025-5 doi: 10.15224/ 978-1-63248-025-5-66
The authors would also like to acknowledge the assistance
given by the technical staff and participating students.
References
[1] S.K. Al-Oraimi and A.C. Seibi, "Mechanical characterisation and impact
behaviour of concrete reinforced with natural fibres," Composite
Structures, vol. 32, pp. 165-171, 1995.
[2] P. Máca, P. Konvalinka and M. Curbach, "Behaviour of Different Types
of Concrete under Impact and Quasi-Static Loading," Applied
Mechanics and Materials, vol. 486, pp. 295-300, 2014.
[3] X.X. Zhang, G. Ruiz and R.C. Yu, "A New Drop-Weight Impact
Machine for Studying Fracture Processes in Structural
Concrete," Strain, vol. 46, pp. 252-257, JUN. 2010.
[4] S.M. Soleimani and N. Banthia, "A Novel Drop Weight Impact Setup
for Testing Reinforced Concrete Beams," Experimental Techniques,
vol. 37, 02-14. 2012.
[5] V. Mechtcherine, O. Millon, M. Butler and K. Thoma, "Mechanical
behaviour of strain hardening cement-based composites under impact
Figure 9. Force deflection diagram for UHPFRC beams loaded by different loading," Cement and Concrete Composites, vol. 33, pp. 1-11, 1. 2011.
speeds. [6] Z.L. Wang, H.H. Zhu and J.G. Wang, "Repeated-Impact Response of
Ultrashort Steel Fiber Reinforced Concrete," Experimental Techniques,
vol. 36, pp. no-no, 04-01. 2011.
V. Conclusions [7] M. Quast, A. Hummeltenberg and M. Curbach, "High Performance
Lightweight Aggregate Concrete under Impact Loading," Beton-
Stahlbetonbau, vol. 107, pp. 15-22, JAN. 2012.
A novel design for impact testing machine of concrete
[8] A. Hummeltenberg and M. Curbach, "Design and construction of a
specimens was presented in this study. It was demonstrated biaxial Split-Hopkinson-Bar," Beton- Stahlbetonbau, vol. 107, pp. 394-
that it is more convenient to have horizontal impact in order to 400, JUN. 2012.
avoid so called double hit and also to provide more [9] M. Just, M. Quast and M. Curbach, "Concrete at high dynamic biaxial
comfortable space for instrumentation of the sample. Design loads," in Performance, protection and strengthening of structures under
of the impact machine is highly modular and it is very simple extreme loading, pp. -, 2013.
to change its configuration for testing the slabs. Steel yokes [10] S.G. Millard, T.C.K. Molyneaux, S.J. Barnett and X. Gao, "Dynamic
were successfully applied in order to avoid uplift of the enhancement of blast-resistant ultra high performance fibre-reinforced
concrete under flexural and shear loading," Int.J.Impact Eng., vol. 37,
specimen. After comparing several impact tests sampled by pp. 405-413, 4. 2010.
different rates, it was stated that 500 kHz sampling frequency
[11] K. Habel and P. Gauvreau, "Response of ultra-high performance fiber
is sufficiently enough for measuring response of materials reinforced concrete (UHPFRC) to impact and static loading," Cement
such steel or concrete even in case without any attenuator. In and Concrete Composites, vol. 30, pp. 938-946, 11. 2008.
the case of attenuator usage, the sampling frequency can be as [12] A. Bentur, S. Mindess and N. Banthia, "The behaviour of concrete under
low as 10 kHz. It was also established that flexible cable impact loading: Experimental procedures and method of
guides should be provided for connection of the impactor to analysis," Materials and Structures, vol. 19, pp. 371-378, 1986.
the rest of the machine in order to avoid eigen frequencies of [13] P. Máca, R. Sovják and P. Konvalinka, "Mix Design of UHPFRC and its
the guide system similar to the eigen frequency of the sample. Response to Projectile Impact," Int.J.Impact Eng., vol. 63, pp. 158-163,
January. 2014.
Samples made of UHPFRC were loaded by impact and its
[14] P. Maca, J. Zatloukal and P. Konvalinka, "Development of Ultra High
three point bending response was compared to beams loaded Performance Fiber Reinforced Concrete mixture," in IEEE Symposium
by standard loading machine. It was found that the maximal on Business, Engineering and Industrial Applications (ISBEIA), pp.
force during impact loading is more than double compared to 861-866, 2012.
the maximal force measured at quasi-static conditions. [15] B. Beckmann, A. Hummeltenberg, T. Weber and M. Curbach, "Concrete
However, this conclusion needs to be confirmed during further Under High Strain Rates: Local Material and Global Structure Response
research on a statically more significant number of samples. In to Impact Loading," International Journal of Protective Structures, vol.
2, pp. 283-294, 09/01. 2011.
the future research the authors plan to test different types of
concrete under impact using the described machine. The
loading strain rate will be varied by using an attenuator and
different impact heights. This should lead to the deeper
understanding of concrete behavior under impact loading at
intermediate strain rates.
Acknowledgment
The authors gratefully acknowledge the support provided
by the National Science Agency of the Czech Republic
(GAČR) under the project No.: P105/12/G059 Cumulative
time dependent processes in building materials and structures.
67
View publication stats
