K. KUNC et al.: TENSILE AND COMPRESSIVE TESTS OF TEXTILE 295–299COMPOSITES AND RESULTS ANALYSIS
295–299
TENSILE AND COMPRESSIVE TESTS OF TEXTILE COMPOSITES
AND RESULTS ANALYSIS
NATEZNI IN TLA^NI PREIZKUSI TEKSTILNIH KOMPOZITOV IN
ANALIZA REZULTATOV
Kry{tof Kunc, Tomá{ Kroupa, Robert Zem~ík, Jan Krystek
University of West Bohemia in Plzeò, NTIS, Univerzitní 22, 306 14, Plzeò, Czech Republic
ptaeck@kme.zcu.cz
Prejem rokopisa – received: 2014-08-01; sprejem za objavo – accepted for publication: 2015-05-05
doi:10.17222/mit.2014.167
The presented work is focused on an experimental investigation of the behavior of six types of textile composites subjected to
pure tensile, cyclic tensile and compressive loading according to ASTM standards. Each type was loaded in directions forming
angles between 0° and 90°, with a step of 15° with respect to the reference directions of the weaves. Two types of woven fabrics
were tested (plain and quasi-unidirectional plain-woven fabric). Images of specimens taken during the tests were subsequently
used for the calculation of the so-called locking angle of yarns (bundles) just before failure. Force-displacement dependencies
were recorded during the tensile tests. Ultimate forces were obtained from the compressive tests. The second half of the article
is dedicated to the analysis of the experimental data gathered with nearly 1000 experiments. Special software for automatic
calculation of averaged dependencies, maximum forces and maximum displacements was created. Furthermore, the
methodology for calculating the locking angle was proposed and tested. The obtained results will be used for the identification
of the material parameters of the proposed material model in the following research.
Keywords: textile composites, woven fabric, tensile test, compressive test, result analysis, weave locking
Prispevek je usmerjen v eksperimentalno preiskavo obna{anja {estih vrst tekstilnih kompozitov, obremenjenih z natezno,
cikli~no-natezno in tla~no obremenitvijo skladno z ASTM standardi. Vsaka vrsta je bila obremenjena v smeri, ki je tvorila kot
med 0° in 90°, s koraki po 15°, glede na smer tkanja. Preizku{eni sta bili dve vrsti tkanin (obi~ajna in kvazi enosmerna obi~ajna
tkanina). Posnetki vzorcev med preizkusi so bili uporabljeni za izra~un zapornega kota preje (sve`njev) tik pred poru{itvijo.
Odvisnosti sila-raztezek so bile posnete med nateznimi preizkusi. Kon~ne sile so bile dobljene iz tla~nih preizkusov. Naslednji
del prispevka je bil posve~en analizi eksperimentalnih podatkov iz skoraj 1000 preizkusov. Kreirana je bila posebna programska
oprema za avtomatsko ra~unanje povpre~nih odvisnosti: maksimalnih sil in maksimalnih raztezkov. Poleg tega je bila
predlagana in preizku{ena metodologija za izra~un zapornega kota. Dobljeni rezultati bodo uporabljeni pri nadaljevanju raziskav
za dolo~anje parametrov materiala v predlaganem modelu materiala.
Klju~ne besede: tekstilni kompoziti, tkanina, natezni preizkus, tla~ni preizkus, analiza rezultatov, zaklepanje vezave
1 INTRODUCTION
Textile composites made from carbon, glass and
aramid fibers are nowadays commonly used. However, to
be able to simulate the behavior of these modern mate-
rials as in the case of classical metals, it is appropriate to
use complex mathematical models with many more
material parameters. A significantly non-linear behavior
of composite materials is caused by different properties
of its components and by a complicated manufacturing
process. Sophisticatedly gathered data from many
experimental tests are, therefore, required for identifying
material parameters and designing modern tailored
composite structures.1
2 EXPERIMENTAL PART
Experimental tests were focused on three composite
types – glass, carbon and aramid. Each type was tested in
two woven-fabric versions: a) a plain weave with a 1:1
fiber ratio and b) a quasi-unidirectional plain weave with
a 1:9 fiber ratio.2
The following markings are used to describe all the
tested specimens: GP – glass plain weave, GU – glass
quasi-unidirectional weave, CP – carbon plain weave,
CU – carbon quasi-unidirectional weave, AP – aramid
plain weave and AU – aramid quasi-unidirectional
weave.
The tested specimens (coupons) were cut from six
composite plates, manufactured with the RTM techno-
logy, using a water jet to get seven different groups of
specimens with the principal material orientation  of the
weave (between 0° and 90° with a step of 15°) with
respect to each coupon’s longitudinal axis (and load
direction). Average thicknesses of the coupons are shown
in Table 1.
Table 1: Average thicknesses of composite plates
Tabela 1: Povpre~ne debeline kompozitnih plo{~
Material t (mm) Material t (mm)
GP 1.8 GU 1.8
CP 2.0 CU 1.5
AP 2.2 AU 2.0
Materiali in tehnologije / Materials and technology 50 (2016) 3, 295–299 295
UDK 620.172:621.315.614:67.017 ISSN 1580-2949
Original scientific article/Izvirni znanstveni ~lanek MTAEC9, 50(3)295(2016)
To ensure the objectivity of the measurement, a
minimum of seven specimens for each material, weave
type and weave orientation was prepared.
Three types of tests were performed on a Zwick/
Roell Z050 machine – pure tensile, cyclic tensile and
compressive loading – according to ASTM standards.3
For the following experiment analysis, data from a
minimum of six identical tests were accepted. The total
number of specimens for one experiment (tensile/cyclic
tensile/compressive) is seven coupons for seven material
orientations multiplied by three different materials and
two types of woven fabrics. In total, nearly 1000 speci-
mens were tested, including the preliminary work.
Typical specimen set-ups before the tensile a) and
compressive b) tests and their dimensions are shown in
Figure 1. Cyclic tensile tests were driven by the pre-
scribed displacement with amplitude l = 1 mm. In the
finished cyclic tests, the number of hysteresis loops for
the same specimens was not always the same. This is
evident from the result graphs, where the tangents of a
few last load/unload cycles are missing.
The tensile tests for the plain-woven fabrics were less
problematic than the ones for the quasi-unidirectional
plain-woven fabrics, which had to be equipped with
aluminium pads in numerous cases to avoid a premature
destruction of a coupon by machine grips and to get
acceptable results from the experiment. All the coupons
for the compressive tests had to be prepared with alumi-
nium pads to achieve the proper attachment to the testing
machine.4 The schemes shown in Figure 2 represent
typical experimental set-ups for the tensile (a) and com-
pressive tests (b).
All the specimens subjected to the tensile tests (load-
ing) were photographed to identify the so-called locking
angle.5
The locking angle (Figure 3) is defined by two
angles, 1 and 2, representing the diversions of the
principal material directions, just before the rupture
(axes , , ), from their initial states (axes e1, e2, e3).
This phenomenon was most notable on the specimens
made of the aramid textile and least notable on the glass
textile. Similarly, for the boundary material orientations
296 Materiali in tehnologije / Materials and technology 50 (2016) 3, 295–299
K. KUNC et al.: TENSILE AND COMPRESSIVE TESTS OF TEXTILE COMPOSITES AND RESULTS ANALYSIS
Figure 3: Aramid plain-weave specimen ( = 30°) and locking-angle
interpretation
Slika 3: Vzorec aramida z navadnim valom ( = 30°) in predstavitev
zapornega kota
Figure 2: a) Tensile- and b) compressive-test schemes
Slika 2: Shemi: a) nateznega preizkusa in b) tla~nega preizkusa
Figure 1: Dimensions of specimens for: a) tensile and b) compressive
tests and the principal material orientation
Slika 1: Dimenzije vzorcev za: a) natezne in b) tla~ne preizkuse in
glavna orientacija materiala
Figure 4: Photographs of selected plain-weave specimens after the
tensile test (GP, CP, AP)
Slika 4: Posnetki izbranih vzorcev z navadnim valom po nateznem
preizkusu (GP, CP, AP)
of  = 0° and  = 90° and quasi-unidirectional plain-
weave composites made of carbon and glass, no
significant weave-locking phenomenon was proven. Car-
bon composites with plain weave showed a high strength
during the tests and never ruptured completely (ex-
cluding the boundary orientations). Photographs of
selected specimens taken after the tensile test are shown
on Figure 4 (with plane weave) and Figure 5 (with
quasi-unidirectional plain weave).
A further description of implementing the measured
data into the identification process and the details of the
material model can be found in another paper of the
above co-authors.6
3 ANALYSIS OF THE EXPERIMENTAL DATA
Standalone application ploTra was written in the
Python programming language for the processing of a
large amount of experimental data. The application is
designed to read experimental data from the Zwick/Roell
software (in the TRA format) and execute multiple ope-
rations resulting in the following outputs:
Averaged force-displacement dependencies (dark
curves in the presented graphs). The application accepts
data from one sorted set of measurements and calculates
the average using one of the various available methods,
e.g., 2D averaging, averaging in a given interval, or the
arc length. The results are saved as graphs (PNG/PDF)
and binary files for future usage.
Averaged tangents (slopes) of unload/load cycles
(straight lines in the presented graphs). All the hysteresis
loops from the cyclic tensile tests are identified; their
lowest and highest points are connected to form lines,
the tangents of which are averaged for each experiment
group including the specimens with the same orientation.
It was observed that these tangents were not constant
during the tests.
Materiali in tehnologije / Materials and technology 50 (2016) 3, 295–299 297
K. KUNC et al.: TENSILE AND COMPRESSIVE TESTS OF TEXTILE COMPOSITES AND RESULTS ANALYSIS
Figure 7: Averaged force-displacement dependencies for GU compo-
site
Slika 7: Povpre~na odvisnost sila-raztezek pri GU kompozitu
Figure 5: Photographs of selected quasi-unidirectional plain-weave
specimens after the tensile test (GU, CU, AU)
Slika 5: Posnetki izbranih vzorcev s kvazi-enosmerno, obi~ajno
vezavo po nateznem preizkusu
Figure 6: Averaged force-displacement dependencies and tangent of
unload/load cycle for GP composite
Slika 6: Povpre~na odvisnost sila-raztezek in tangenta na neobreme-
njen/obremenjen cikel pri GP kompozitu
Figure 8: Averaged force-displacement dependencies and tangent of
unload/load cycle for CP composite
Slika 8: Povpre~je odvisnosti sila-raztezek in tangenta cikla neobre-
menjeno/obremenjeno pri CP kompozitu
Tables with averaged statistical data – the maximum
forces and displacements including standard deviations.
4 RESULTS
The tables and graphs below represent the outputs
from the ploTra application. The maximum tensile and
ultimate forces of the compressive tests are shown in
Tables 2 and 3. Averaged force-displacement dependen-
cies (Figures 6 to 11) are suitable for the use of the com-
plex material model including damage. The weave-
locking angles shown in Tables 4 and 5 were collected
using the common tools available in graphics editors.
Table 2: Tensile tests – average maximum forces in (kN)
Tabela 2: Natezni preizkusi – povpre~je maksimalnih sil (kN)
 GP GU CP CU AP AU
0° 9.50 18.45 20.20 39.71 16.01 29.50
15° 5.68 5.63 9.33 3.66 8.25 7.03
30° 4.33 2.78 5.96 1.80 7.33 3.66
45° 3.63 1.91 5.28 0.99 6.63 2.73
60° 3.66 1.67 5.77 0.82 7.01 2.30
75° 4.23 1.39 9.18 0.60 8.37 2.67
90° 6.40 1.36 19.09 0.57 17.37 2.87
Table 3: Compressive tests – average ultimate forces (kN)
Tabela 3: Tla~ni preizkusi – povpre~je kon~nih sil (kN)
 GP GU CP CU AP AU
0° 8.32 13.21 12.01 11.70 4.89 5.36
15° 6.71 4.71 7.32 5.74 4.69 5.37
30° 4.03 4.25 4.91 3.75 3.77 3.99
45° 3.67 4.01 4.51 3.16 3.70 3.46
60° 3.96 3.96 4.75 2.81 4.01 3.37
75° 4.85 4.04 7.13 2.55 4.68 3.45
90° 4.20 4.05 11.6 2.54 4.70 3.52
Table 4: Averaged weave-locking angles 1 and 2 for plain-weave
composites in (°)
Tabela 4: Povpre~je kota tkanja 1 in 2 za obi~ajno tkane kompozite
v (°)
 GP CP AP
0° 0.00 0.00 0.00 0.00 0.00 0.00
15° 2.30 2.00 8.60 1.30 17.80 2.60
30° 5.00 3.50 12.80 5.80 26.70 12.30
45° 7.50 7.10 13.30 12.50 17.30 18.00
60° 3.10 3.30 8.50 20.60 9.80 22.50
75° 1.10 3.20 1.50 11.70 1.30 17.00
90° 0.00 0.00 0.00 0.00 0.00 0.00
5 CONCLUSION
All the materials showed a complex non-linear
behavior.
• Force-displacement dependencies are non-linear even
for the plain-weave material orientations of  = 0°
and  = 90°. Hardening was noticed in the case of
carbon and aramid textiles (exhibiting a convex
298 Materiali in tehnologije / Materials and technology 50 (2016) 3, 295–299
K. KUNC et al.: TENSILE AND COMPRESSIVE TESTS OF TEXTILE COMPOSITES AND RESULTS ANALYSIS
Figure 11: Averaged force-displacement dependencies and tangent of
unload/load cycle for AU composite
Slika 11: Povpre~je odvisnosti sila-raztezek in tangenta cikla neobre-
menjeno/obremenjeno pri AU kompozitu
Figure 10: Averaged force-displacement dependencies and tangent of
unload/load cycle for AP composite
Slika 10: Povpre~je odvisnosti sila-raztezek in tangenta cikla razbre-
menjeno/obremenjeno pri AP kompozitu
Figure 9: Averaged force-displacement dependencies for CU compo-
site
Slika 9: Povpre~je odvisnosti sila-raztezek pri CU kompozitu
load-displacement curve when zoomed) and soften-
ing (a concave curve) was noticed in the case of glass
textile.
• Specimens made of aramid fibers reached the highest
strength during the tests; on the other hand, glass-
fiber specimens reached the lowest strength.
• Weave-locking phenomenon has significant impacts
on the plain-weave orientations of  = 45° – plastic
behavior of the tested materials was observed.
• Unsymmetrical results for the plain-weave compo-
sites are probably caused by an imperfect technology
of manufacturing the textiles or by the preparation of
the specimens.
Acknowledgements
This publication was supported by the project
L01506 of the Czech Ministry of Education, Youth and
Sports and by the project of University of West Bohemia
SGS-2016-038.
6 REFERENCES
1 R. Böhm, W. Hufenbach, Experimentally based strategy for damage
analysis of textile-reinforced composites under static loading,
Composites Science and Technology, 70 (2010) 9, 1330–1337,
doi:10.1016/j.compscitech.2010.04.008
2 T. Kroupa, P. Janda, R. Zem~ík, Linear two scale model for
determination of mechanical properties of textile composite material,
Mater. Tehnol., 46 (2012) 2, 97–101
3 ASTM Standard D 3039 / D 3039M – 08, 2008, Standard Test Me-
thod for Tensile Properties of Polymer Matrix Composite Materials,
ASTM International, West Conshohocken, PA 19428–2959, 2003,
www.astm.org
4 ASTM Standard D 3410 / D 3410M – 03, 2008, Standard Test
Method for Compressive Properties of Polymer Matrix Composite
Materials, ASTM International, West Conshohocken, PA
19428–2959, 2003, www.astm.org
5 N. Hamila, P. Boisse, Locking in simulation of composite rein-
forcement deformations, Analysis and treatment, Composites Part A,
Applied Science and Manufacturing, 53 (2013), 109–117
doi:10.1016/j.compositesa.2013.06.001
6 T. Kroupa, K. Kunc, R. Zem~ik, T. Mandys, Non-linear finite
elements simulations of the tensile tests of textile composites, Mater.
tehnol., 49 (2015) 4, 509–515, doi:10.17222/mit.2014.117
Materiali in tehnologije / Materials and technology 50 (2016) 3, 295–299 299
K. KUNC et al.: TENSILE AND COMPRESSIVE TESTS OF TEXTILE COMPOSITES AND RESULTS ANALYSIS