2nd International RILEM Conference on Strain Hardening Cementitious Composites 12-14 December 2011, Rio de Janeiro, Brazil
399
APPLICABILITY OF COMPOSITE CHARPY IMPACT METHOD FOR STRAIN HARDENING TEXTILE REINFORCED CEMENTITIOUS COMPOSITES
J. Van Ackeren P
(1)P, J. Blom P
(1)P, D. Kakogiannis P
(1)P, J. Wastiels P
(1)P, D. Van Hemelrijck P
(1)P, S.
Palanivelu P
(2)P, W. Van Paepegem P
(2)P, J. Vantomme P
(3)
(1) Vrije Universiteit Brussel, Mechanics of Materials and Constructions, Brussels, Belgium (2) Universiteit Gent, Dept. of Materials Science and Engineering, Gent, Belgium (3) Royal Military Academy, Civil and Materials Engineering Department, Brussels, Belgium Abstract
Since most test methods, used for strain hardening cementitious composites, are adopted
from polymer composites test standards, this paper investigates the applicability of a standard impact test (ISO 179-1:2000), which is used in the polymer composites world, on a specific textile reinforced cementitious composite using an inorganic phosphate cement (IPC) as a matrix material. The aim is to provide a test method to investigate the local impact behaviour of textile reinforced cementitious composites. It was found within this work that this method is applicable to cementitious composites provided some adaptations to the specimens’ dimensions. The Charpy impact strength of an IPC matrix reinforced with different kinds of fibre types is successfully investigated. This test method can be used to qualitatively rank different cementitious composite materials.
1. INTRODUCTION In the last few decades, many new cementitious materials were developed in order to reach
thinner and stronger construction elements. The evolution started with the introduction of loose steel or polymeric fibres into the fresh concrete mixture, leading to fibre reinforced concrete (FRC): the addition of fibres into concrete provides toughness or energy absorption capacity to the inherently brittle concrete. A new generation of FRC’s are the so called textile reinforced concretes (TRC) in which mainly fibre textiles (glass, polymers …) are applied as reinforcement: they can provide also a strain hardening capacity when a sufficiently high quantity of fibres is used. In a recently held conference on these kinds of materials [1], the potential of these new composites was again demonstrated. In an attempt to better define high performance fibre reinforced cementitious composites (HPFRCC), Naaman and Reinhardt [2] proposed a clear classification of these materials according to their possibility to exhibit strain hardening under bending or in tension. Within this classification a distinction is made between the four following categories: (1) crack control, (2) deflection hardening, (3) strain hardening and (4) high energy absorption.
2nd International RILEM Conference on Strain Hardening Cementitious Composites 12-14 December 2011, Rio de Janeiro, Brazil
400
Strain hardening cementitious composites (SHCC) not only show strain hardening under tensile loading, but also can absorb a large amount of energy. The cementitious material studied in this paper can most certainly be classified in the fourth category. It exists of an inorganic phosphate cement (IPC) matrix with different kinds of textile reinforcements. IPC was developed at the Vrije Universiteit Brussel and consists of a liquid component based on phosphoric acid solution containing inorganic metal oxides and a calcium silicate powder component. Next to the advantages of being incombustible (EN13501-1), this material also possesses a neutral pH after hardening, which allows cheap E-glass fibres to be used as reinforcement. It is however also possible to process other types of fibres, such as polymeric fibres or carbon fibres, into this cementitious matrix. Figure 1 shows typical tensile stress-strain behaviour of IPC composites with different kinds of fibre reinforcements. The fibre volume fraction is kept constant for all materials at about 20 %. It is clear from this figure that indeed all tested composite specimens can absorb a significant amount of energy due to their pronounced strain hardening behaviour after multiple matrix cracking. Note that depending on the used fibre reinforcement one can obtain a completely different composite material. For instance the curve of PE-reinforced IPC shows stiff and strong (up to 400 MPa) behaviour, while on the right side of the graph the PVA reinforced IPC shows ductile (maximum strain more than 5 %) but less strong behaviour.
This composite material has been investigated in a project that studied the possibility of producing sandwich structures that are able to absorb a significant amount of energy under blast or impact loading. The textile reinforced IPC was meant to protect the energy absorbing core against fire and to distribute the load from the explosion to the sandwich core [3]. Apparently, the material itself is able to absorb energy under blast and impact loading without failing. Under impact loading only local damage can cause failure of the material. Globally, there will be matrix cracking reducing the stiffness, but the material keeps its load bearing capacity. In this paper, the local impact damage and the corresponding energy absorption are investigated by means of impact tests.
Figure 1: tensile stress-strain behaviour of IPC reinforced with different UD-fibres
0
50
100
150
200
250
300
350
400
450
0 1 2 3 4 5 6strain [%]
stre
ss [M
Pa]
Carbon
PE
basalt
E-glass (1)
E-glass (2)
PVA
2
2. CHMany
polymertest: thethat forcorrespomost copendulua quantienergy acapacityother mcomposmaterialdevelopinstrumobtain dthe locain figursingle bdifferenimpact possibleright) annormal interlammechanhoweveflaws thdamage
2nd Internatio
HARPY IMy test methr compositee scale of thr concrete onding enerommon test um tests. In itative wayabsorption cy and the sp
materials. Inites [4]. Thls. For plast
ped (ISO 17ented test) [
design data.al impact enre 2 on the blow of a pnce in potenstrength is e testing dind normal flatwise di
minar shear nics theory er not necesshat will in
ed by the no
Figure 2
onal RILEM12-14
MPACT TEShods for SHes world, rahe specimenor FRC bergy that canmethods codrop weigh
. The Charpcapacity of
pecimen dimn the 1980he specified tics and pol
79-1: Plastic[5]. It is cle The purpo
nergy absorpleft. A spe
pendulum sntial energy defined as t
irections: thor parallel
irection. Fu(ILS) and
the specimsary to notc
nitiate failurtch.
2: left: Char
M Conferenc4 December
ST HCC – like t
ather than tn dimensioneams. In on be absorboncerning iht tests globpy pendulum
f metals undmensions ha’s and 199specimen d
lymer compcs – Determ
early stated ise of the Chption of diffecimen is sistriker. Theof the pend
this energy he specimen
to the lamurthermore d those not mens can bech the specimre. Moreov
rpy impact t
ce on Strain r 2011, Rio d
401
tensile testithe concretens of a ben
order to stubed, a suitabimpact on cbal and locam impact te
der three poiave led to th90’s Charpydimensions posite matermination of in this standharpy test isferent materimply suppoe absorbed dulum befodivided by
n can be temina. All spe
a distinctioexhibiting
e notched. mens since
ver the fibr
test; right: d
Hardening Cde Janeiro, B
ng - are ade world. An
nding test foudy the locble test metcomposites al impact beest was initint bending
he developmy impact teare howev
rials a standCharpy im
dard that thes to providerials. The prorted on boenergy is d
ore and aftery the mid seested flatwiecimens teson is made g this fractuIn case of the materia
res inside t
different tes
edgewise
CementitiouBrazil
dopted fromnother examor SHCC iscal impact thod needs are either d
ehaviour cantially design. Adaptatio
ment of Chaests were der not applidard test me
mpact propere method she a compararinciple of toth sides andetermined r impact. Tection area. ise or edgested in this
between mure mode. cementitiou
al already cothe materia
t configurat
f
us Composite
m, or inspiremple is the much smabehaviour to be selec
drop weightn be investined to test tns of the m
arpy impact developed ficable for thethod has arties – Part hould not beative test to the test is ilnd is impac
by calculaThe reported
There are ewise (see f
work are tmaterials exSimilar to us composiontains mical will be
tions [5].
flatwise
es
ed by the bending
aller than and the
cted. The t tests or igated in the local
machine’s tests for
for FRC hin TRC
also been 1: Non-
e used to evaluate lustrated ted by a
ating the d Charpy different figure 2, tested in xhibiting
fracture ites it is
croscopic severely
2nd International RILEM Conference on Strain Hardening Cementitious Composites 12-14 December 2011, Rio de Janeiro, Brazil
402
The Charpy impact strength can be compared to typical values for toughness that is defined
in fracture mechanics as being the energy that is necessary for the growth of a crack. Toughness (GRcR) is a material parameter, typical values for different materials can be found in Ashby and Jones [6]. For metals, values between 100 and 1000 kJ/m² are given while for most plastics the toughness is less than 10 kJ/m². Typical values for composites start at 5 kJ/m² and can reach more than 100 kJ/m².
Three specimen types are defined in ISO 179-1 [5] as illustrated in Table 1. Note that the given tolerances are rather small for cementitious materials. In the experimental section the average values of the dimensions together with their standard deviation will be given. Specimen type 1 is used when the material does not show interlaminar shear. Specimen type 2 is designed to induce typical bending failure modes for materials that undergo ILS. These specimens fail in the middle section of the span either by tensile failure on the back of the specimen or compressive failure or buckling of fibres at the impacted side. The span to thickness ratio is fixed to 20 for this specimen type. To test the energy absorption capacity of interlaminar shear phenomena, specimen type 3 is designed with much smaller span to thickness ratio of 6 or 8. The value of 8 can be chosen if it is not possible for practical reasons to test at such small spans.
Table 1: Specimen types, dimensions and spans proposed by ISO-179 [5]
Specimen Length l Width b Thickness h Span L 1 80 ± 2 10.0 ± 0.2 4.0 ± 0.2
2
3 25h
11h or 13h 10 or 15 3 20h 6h or 8h
3. EXPERIMENTAL STUDY All tests in this work were performed on a Ceast Impactor II 7611 pendulum testing
machine. The pendulum has a capacity of 15 J and is equipped with a processing unit that immediately calculates the Charpy impact strength. The speed at impact is 3.8 m/s. Before every series of measurements the machine is calibrated by determination of the friction in the pendulum.
3.1 2D RANDOM GLASS FIBRE MAT REINFORCEMENT
The first series of specimens, which contains 2D random chopped strand glass fibre mats as reinforcement, were manufactured by hand lay-up technique [7]. The volume fraction is about 20 %. Note that only part of this fraction is contributing to the load carrying capacity of the specimen. Five different series of 20 specimens were manufactured (see table 2): the three different types of specimens according to the ISO 179-1 standard. For type 2 and 3 the influence of the width was tested (10 or 15 mm) by adding two extra series.
The results of these tests are made visible in figure 3 in form of boxplots. The the bold line in the middle of the grey box represents the median. Next to the boxplots, a figure shows the tested specimens and their failure modes.
5.00.062+
−
2
Table 2
In fig
types amainly bendingspecimespecimefailure pdue to tor multi50 kJ/mapplicabexplanafibre buFinally width onot sign
For random
Fig
2nd Internatio
: Specimen
Spet
tytypetypetypetype
gure 3 (righnd consequfail throug
g failure moen of type ens (43.5 kphenomenothe larger friple shear, tm² up to 1ble for thiation for thisundles. Thisthe effect o
of 15 mm senificant (figucomparison
m fibre reinfo
gure 3: left:
onal RILEM12-14
dimensions
ecimen type
ll
ype 1 e 2 – 10 e 2 – 15 e 3 – 10 e 3 – 15
ht) it is clearuently variah compressode is how1 (see figu
kJ/m²) compon which abree surface athere is a lo120 kJ/m² is kind of s is that thes means thaof the specieem to be sure 3, left). n with thesorcements. T
boxplots o
M Conferenc4 December
s of IPC rei
ength (mm)
spa(m
80 675 675 633 133 1
r that indeeant Charpy sion and ten
wever combure 3), whpared to typbsorbs morearea that is
ot of scatter were obserrandom fi
e specimen at fibre bunimen width slightly high
se results tThe applied
of test result
ce on Strain r 2011, Rio d
403
inforced wit
an L mm)
wb
62 9.960 9.860 14.318 9.818 14.2
ed the differimpact str
nsion (in abined with hich explainpe 2 specime energy cocreated. Alon the resurved. This
fibre reinfodimensions
ndles were cwas invest
her, but due
two other md matrices a
ts, right: test
Hardening Cde Janeiro, B
th random c
width b stdev
99 0.0184 0.0833 0.09
83 0.0829 0.12
rent specimerength resu minor waya small are
ns the highmens (27 k
ompared to lthough thesults of type
specimenrced cemen
s exceed thecut and thutigated: meae to the larg
matrix mateare a very fin
ted specime
CementitiouBrazil
chopped stra
thicknesh
4.26 3.283.483.14 3.20
en types cauults. Type 1y) in the mea of interlher absorbekJ/m²). Intera pure bendse specimen3 specimentype migh
ntitious mae longitudinus damaged an values oge scatter th
erials are ine grained c
ens, differen
us Composite
and glass fib
ss h stdev 0.20 0.14 0.09 0.19 0.20
use differen1 and 2 sp
middle sectiolaminar shed energy frlaminar shding type ons all fail unns. Values oht thereforeaterial. A
nal dimensioduring pro
of specimenhese differe
impregnatedcement mor
nt failure typ
ILS
es
bre mats
nt failure pecimens on. This
ear for a for these hear is a of failure nder ILS of almost e not be
possible on of the oduction. ns with a ences are
d in the rtar with
pes
S
2
F
a maximenergy Hexion)(490.0 g
Onlyshow incompletten impThe epoaverageseem tothis mufibres a0.02 % at the bcause fothe morthat of specimebe deriv
3.2 UThe
differenshould could bbundlestogether24h at 6rovings TOHO,
Sinceboth spenominalhoweve
2nd Internatio
Figure 4: im
mum grain sand boat bu). The cemg), water (24y results of snterlaminar te break of
pacted specioxy specimee of ten testo fail on the uch higher eand shows a
for mortar back of the or higher enrtar or the IP
the epoxy en is smalleved from the
Unidirectioaim of this
nt kinds of contain the
be well cons. Once imr. Specimen60°C. The d
are manufPVA from
e specimensecimen typl dimension
er, due to ma
onal RILEM12-14
mpacted spe
size of 0.6 muilding (EPent mortar 45.0 g), fly specimens oshear failu
their middlimens was fens on the ots is 93.54 tensile side
energy absoa large straiand IPC whspecimens
nergy absorpPC. The fleresin is 1
er than that oe properties
onal reinfors section is
fibre typese same amontrolled by
mpregnated, ns are then different fibrfactured by
Nordifa ans made withes 2 and 3 ns of 10mmanufacturin
M Conferenc4 December
cimens with
mm and an PIKOTE Re
consists ofash (175.0
of type 1 arure. The spee section (ffound to beother hand kJ/m² with
e of the specorption capin to failurehile it is 7 tos and thus lption is the exural streng10 to 130 of polymer s of the matr
rcement to validate
s as reinforount of fibre
using a sethe bundlecured for 2
re types witOwens Co
nd PE from Dh IPC and Ufor all fibrem x 3mm g, averagely
ce on Strain r 2011, Rio d
404
h mortar ma
epoxy resinsin L 135 a
f quartz (71g), silica-fu
re given, sinecimens co
figure 4, lefte 9.67 kJ/m²show a muc
h a standardcimens (see acity: epox
e. For compo 10 % for larger strainhigh flexurgth of the mMPa. Althcomposite mrix material
e this standrcement. Ines. The fibrelf-made pues were pu24h at roomth their prop
orning, basaDSM.
UD glass fibe types. Thex 75mm wy 3.3 mm th
Hardening Cde Janeiro, B
atrix (left) a
n used in lamand EPIKU13.6 g), quaume (70.0 gnce it was foontaining mft). The aver² with a stach larger end deviation figure 4, ri
xy resin hasparison: thethe epoxy rn energy leal strength o
mortar and though the ematerials, itl itself.
dard test men order to bre volume fultrusion se
ut straight im temperatuperties can balt fibres ar
bres, all shoe first serie
with a spanhick, which
CementitiouBrazil
and epoxy m
minating apURE Curing artz powderg) and superound that bo
mortar are torage Charpyandard devianergy absorof 13.8 kJ
ght). There s a much bee elongationresin. Tensilevels are reof the epoxythe IPC is aenergy absot absorbs m
ethod for IPbe comparafraction (apet-up to imin a mouldure and posbe found inre from Bas
wed ILS, ites of tests an of 60 mm
lowered the
us Composite
matrix (right
pplications fAgent H 1
r (499.5 g),r plastifier (oth materialoo brittle cay impact stration of 1.8rption capacJ/m². All sp
are two reaetter bond
n at break isle failure w
eached. They resin com
about 10 MPorption of
more energy
PC specimeable the sp
pproximatelmpregnate td which is st-cured forn table 3. Glsaltex, carb
t was decideare type 2 tem. Specimee span to th
es
t)
for wind 34 from , cement 10.5 g). ls do not ausing a rength of 84 kJ/m². city. The pecimens asons for with the s around
will occur e second
mpared to Pa while the IPC than can
ens with pecimens y 20 %) the fibre
pressed r another ass fibre
bon from
ed to test ests with ens were hickness
2
ratio to observe
To tafailure, nominaltheoretithis serispecimeimpact median from vicompresused fibbond becan also
The specimelength oright ha
2nd Internatio
Fibre t
glass S1200
glass S1500basacarboPVAPE
P
1P straight fibr
18 in steaded (figure 5)ackle this pthe ratio spl thickness cal span to ies had an aens all failestrength arvalues are
isual inspecssive side obre reinforcetween fibreo improve thlast series o
ens is arounof 43 mm aand graph of
onal RILEM12-14
Table 3
type De(k
SE-0 P
1 2
SE-0 P
2 2
alt 2on 1A 1
E 9re bundles, P
2P twis
Figure 5
d of 20. In m). This is proproblem thepan to thickof the specthickness r
average thicd in the midre shown innot signific
ction of theof the specicement. In es and matrhe performaof tests are
nd 3.3 mm wand a span f figure 6. T
M Conferenc4 December
: different f
ensity kg/m³)
2600
2600
2700 1800 1300 975 sted fibre bundles
5: observed
more than 6obably due
e following kness shouldcimens to 2 ratio now bckness of 2.2ddle sectionn Figure 6 cant except e tested speimen, whicorder to in
rix should bance of thes performedwith a standof 26.4 mm
The specime
ce on Strain r 2011, Rio d
405
fibre types w
Tex (g/km)
600
600
1600 250 250 176
s
ILS failure
65 % of theto the relatisolution wad be increamm, while
becomes 60/2 mm, whicn under ben
(left). Duefor the glas
ecimens. Alh is more r
ncrease the be increasedse materialsd on type 3 dard deviatim accordingens all failed
Hardening Cde Janeiro, B
with their pr
Strength (MPa)
950
950
1100 1275 1000 1870
in type 2 sp
tested specive weak boas found: tosed. This w
e the span is/2 = 30. Thch lowers thnding. The oe to scatterss fibres SEll specimenrelated to thenergy abs
d. Local rein. specimens.
ion of 0.1 mg to table 1d under ILS
CementitiouBrazil
roperties
Stiff(G
7
7
7173
10
pecimens
cimens, inteond betweeno encourage
was obtaineds kept cons
he manufacthis ratio to 2obtained var, the differE-1200. Thisns failed unhe matrix m
sorption of nforcement
. The averamm. This re. The result
S as can be s
us Composite
fness GPa)
75
75
78 70
33 05
erlaminar shn fibres ande a bendingd by decreatant at 60 mtured specim27.3. Thesealues for therences betws could be ender bendinmaterial thathese mateat the impa
age thicknesesults in a spts are showseen in the e
es
hear was d matrix. g type of asing the mm. The mens for adapted
e Charpy ween the expected ng at the an to the rials the act point
ss of the pecimen
wn in the example
2
Fig
in figureunder ILthe fibrfibre typ
4. COSince
from poapplicabThe aimcompos
-
-
-
In or
their locfuture wcompare
ACKNORese
Innovat
2nd Internatio
gure 6: resul
e 6 (right). LS than thees and the pes.
ONCLUSIOe most test olymer combility of a s
m is to provsites. The fo
The proplocal imhardeninDue to damage Specimeinterlam
rder to be acal impact bwork this poe these resu
OWLEDGearch fundedtion through
onal RILEM12-14
lts of adapte
In this stude carbon anmatrix. The
ONS methods, u
mposites testtandard impide a test m
ollowing conposed test m
mpact resistang TRC mat
the poor fmechanism
ens of type 2inar shear e
able not onlbehaviour inossibility wiults with mo
GEMENTS d by a PhD
h Science an
M Conferenc4 December
ed specimen
dy the specifnd basalt fibe basalt fib
used for strt standards, pact test fro
method to innclusions camethod canance and eterials. fibre-matrix
m. 2 should be
effects
ly to rank dn detail, insill be explor
ore global im
D grant for nd Technolo
ce on Strain r 2011, Rio d
406
ns of type 2
fic used polbres. This cbres seem to
rain hardeniit was sugg
om the polynvestigate than be drawn
n indeed be energy abso
x bond, in
e thin enoug
different mastrumented red in order
mpact tests
the first auogy in Fland
Hardening Cde Janeiro, B
(left) and s
lymer and gan be linkeo be less w
ing cementigested with
ymer compohe local impn: applied for
orption cap
terlaminar
gh (span to
aterials, butCharpy impr to be able like drop w
uthor of the ders (IWT-V
CementitiouBrazil
specimens o
glass fibres aed to the bowell impregn
itious comphin this papeosites worldpact behavi
r qualitativeacity betwe
shear beco
depth ratio
t also to bepact tests cato quantify
weight tests.
Institute foVlaanderen)
us Composite
of type 3 (rig
absorb morond quality nated than
posites, are er to investi
d on these mour of ceme
e comparisoeen differen
omes an im
above 30)
e able to invan be perfory the damag
or the Prom).
es
ght)
e energy between all other
adopted igate the
materials. entitious
on of the nt strain
mportant
to avoid
vestigate rmed. In
ge and to
motion of
2nd International RILEM Conference on Strain Hardening Cementitious Composites 12-14 December 2011, Rio de Janeiro, Brazil
407
REFERENCES [1] W. Brameshuber, Proceedings of the International RILEM Conference on Materials Science
(MatSci), Volume 1: 2P
ndP ICTRC, 2010, Aachen, Germany ISBN: 978-2-35158-106-3
[2] A.E. Naaman, H.W. Reinhardt, High performance fiber reinforced cement composites HPFRCC-4: International workshop Ann Arbor, Michigan, June 16-18, 2003. Cement and Concrete Composites, Volume 26, Issue 6, 757-759 (2004)
[3] S. Palanivelu, W. Van Paepegem, J. Degrieck, B. Reymen, J. Nadambi, J. Vantomme, D. Kakogiannis, J. Van Ackeren, D. Van Hemelrijck, J. Wastiels, Experimental and numerical blast energy absorption study of composite tubes for sacrificial cladding structures, ECCM-14, 7-10 June, 2010, Budapest, Hungary
[4] V. Gopalaratnam, S. Shah, R. John, A modified instrumented charpy test for cement-based composites, Experimental Mechanics, volume 24, issue 2, p. 102-111 (1984)
[5] ISO 179-1: 2000, Plastics – Determination of Charpy impact properties – Part 1: Non-instrumented test
[6] M.F. Ashby, D.R.H. Jones, “Engineering materials 1: An introduction to properties, applications and design”, Elsevier Butterworth-Heinemann, Burlington, 2005, ISBN 0-7506-63804
[7] J. Blom, P. Van Itterbeeck, J. Van Ackeren, J. Wastiels, Textile reinforced inorganic phosphate cement composite moulds, NCTAM-2009, 2009, Brussels, Belgium