You are on page 1of 17

Article

InPlaneBehaviourofaReinforcementConcrete
FramewithaDryStackMasonryPanel
KunLin1,YuriZarevichTotoev2,HongjunLiu1,*andTianyouGuo1
ShenzhenEngineeringLabforWindEnvironmentandTechnology,ShenzhenKeyLabofUrban&Civil
EngineeringDisasterPrevention&Reduction,ShenzhenGraduateSchool,HarbinInstituteofTechnology,
Shenzhen518055,China;linkun.hit@gmail.com(K.L.);guotianyou1992@gmail.com(T.G.)
2 CentreforInfrastructurePerformanceandReliability,TheUniversityofNewcastle,UniversityDrive,
CallaghanNSW2308,Australia;Yuri.Totoev@newcastle.edu.au
* Correspondence:liuhongjun@hit.edu.cn;Tel.:+8675526033805
1

AcademicEditor:JungHoJe
Received:25December2015;Accepted:3February2016;Published:11February2016

Abstract:Inordertoimprovetheenergydissipationofthemasonryinfilledframestructurewhile
decreasing the stiffening and strengthening effects of the infill panels, a new dry stacked panel
(DSP)semiinterlockingmasonry(SIM)infillpanelhasbeendeveloped.Inthispaper,thematerial
properties of DSP and a traditional unreinforced masonry (URM) panel have been evaluated
experimentally.Aseries of cyclic testswere performed toinvestigate the cyclic behaviour of the
reinforcement concrete (RC) frame with different infill panels. The failure modes, damage
evolution,hystereticbehaviour,stiffnessdegradationandenergydissipationwerecomparedand
analysed. We concluded that DSP is capable of significantly improving the seismic energy
dissipationduetoitshystereticbehaviourwhentheframeisinelasticstagewithoutincreasingthe
stiffnessoftheframe.Therefore,DSPorSIMpanelscanbeconsideredasfrictionaldampers.Based
on the experimental results, the influence of DSP was examined. Using the parallel model, the
hystereticloopsofDSPsubjectedtodifferentloadcaseswereachieved.Thetypicalfullhysteretic
loop for DSP could be divided into three distinct stages of behaviour: packing stage, constant
friction stage and equivalent strutstage. The connection between the paneland theframe hada
great effect on the transferring of different mechanical stages. The constant friction stage was
verified to provide substantial energy dissipation and benefits to the ductility of the structure,
which,therefore,issuggestedtobeprolongedinreality.
Keywords: infilled RC frame; dry stacked panel; semiinterlocking masonry; cyclic test; failure
mode;stiffness;energydissipation;parallelmodel;mechanism

1.Introduction
Althoughmasonryisoneoftheoldestandmostpopularbuildingmaterials,ithaslimiteduse
in seismic areas because of the low tensile strength and brittle behaviour of modern slender
traditionalunreinforcedmasonry(URM).Toachievebetterseismicperformance,masonryhasbeen
combined with other stronger materials. One such dual structural system is the reinforcement
concrete(RC)framestructurewithmasonryinfillpanels.Theseinfilledframesareverycommonall
aroundtheworldbecausetheyarepracticalandeconomical.However,theirseismicperformance
remainsquestionable,especiallyinhighseismicareas.
Masonry infill panels are usually designed as nonstructural elements, whose only role is to
shelter from the elements and to divide space. However, in reality, panels almost unavoidably
interact with the frame during strong earthquakes, which makes the realistic structural response
differentfromthedesignedone,thereforeoftenresultinginaseriousdamage[13].
Materials2016,9,108;doi:10.3390/ma9020108

www.mdpi.com/journal/materials

Materials2016,9,108

2of17

Previous research shows that the confining effect of the frame could improve the seismic
behaviouroftraditionalURMpanels [46].However,seriousdamagewasgenerallyfoundbothin
theRCframeandmasonrypanelsafteranearthquake,whichmadethestructuresunabletocontinueto
beused.Furthermore,infillpanelsaregenerallynotdistributedevenlythroughthestructure,therefore
increasingthecomplexityoftheseismicresponseandsometimesresultinginthesoftstoryeffect.
Addingdamperstotheframecouldimproveitsenergydissipation[79]andseismicbehaviour.
However, the requirement of additional constructional cost made it difficult to popularize. As
masonrypanelsareoftenanintegralpartoftheframestructure,usingthemtoincreasetheenergy
dissipationwithoutsignificantlyaffectingtheoverallstructuralstiffnesscouldbeofgreatbenefitto
theseismicdesign.
According to previous research, the uncertainty in the characteristics of the masonry infills has
greatimpactontheinfilledRCframesperformance,especiallyonitslimitstatesofdamagelimitation
[1012].Furthermore,theinfluenceofthemortarjointontheseismicbehaviourofthemasonryinfilled
frame is significant [13,14].If no mortar joint exist, dry stack masonry (DSM) is expected to exhibit a
morestableperformance.Actually,thedrystackingmethodwasfirstlyusedalmostfivethousandyears
ago;thePyramidsinEgyptaresomeofthemostfamousexamples[15].Inthiskindofmasonry,bricks
areassembledwithoutmortar;therefore,itisalsomortarlessmasonry[16,17].Becauseitiseasytobuild
and is demountable, therefore, reducing the energy used in building, DSM will achieving higher
materialefficiencyandlowercosts.Becauseofitsadvantages,drystackmasonryisstillpopularin
buildings around the world, especially in developing counties [1820]. Previous research of dry
stack masonry focused on the lateral capacity and the seismic behaviour of the unreinforced dry
stackmasonry [21].Resultsshowthatthefailurecriteriaofdrystackstonecanbeconsideredasa
MohrCoulomb failure [2224], and the strength of dry stack units does not make a significant
differenceintheresistancetolateralloads[25,26];theinterlockingandfrictionbetweenunitsgovern
the lateral loadbearing capacity, and the type of wall boundary conditions and the vertical
compressionlevelwereconfirmedastwoimportantfactorsforthefailuremode[25,27].Although
considerablenonlineardeformationshavebeenattained[21,22],unframeddrystackmasonrywalls
exhibitedlittleenergydissipationbecauseoftherockingfailuremechanism.
Outofplane behaviour is one major factor of restricting the application of DSM [28,29].
According to the authors previous research [17], semiinterlockingunitshave been used to build
the dry stack panel. The semiinterlocking masonry panel allows relative sliding inplane of the
panel and maintains outofplane stability, as shown in Figure 1. The dry stack panel (DSP) no
longer remains a nonstructural element; instead, it should be considered as a damper, which
contributesprimarilytotheenergydissipationduringearthquakes.
This paper describes cyclic tests on the RC frame with different infill forms (not infilled,
traditional unreinforced masonry panel (TMP) infilled and DSP infilled) of the panels and
investigates the panel response mechanisms and its contribution to the entire structural energy
dissipation. Series hysteretic loops have been achieved to analyse the frictional behaviour and
mechanismofDSP.

(a)

(b)

Figure1.Semiinterlockingmasonryunits:(a)Rectangularinterlocking;(b)Circularinterlocking.

Materials2016,9,108

3of17

2.ExperimentalProgram
In order to investigate the inplane cyclic behaviour of frames infilled with SIM panels, the
testingprogramwascarriedoutonafullscaleRCframewithadrystackedmasonrypanel(DSP).In
thisinitialphaseofresearch,onlytheinplanebehaviourofframeswasconsidered;nooutofplane
loadswereapplied;nointerlockingagainstoutofplanedisplacementsofbrickswasrequired;and
nonewereused.Hence,DSPinthesetestsisconsideredtorepresentprototypeSIMpanel.Testing
wasdoneinthreestages:
(1)ThefirstcyclicdisplacementtestwasperformedonabareRCframe.Displacementswere
limitedtoavoidplasticstrainsintheframe.
(2) DSP was built inside the frame using solid concrete bricks, and the second cyclic
displacement test was performed on the frame infilled with DSP. Displacements were limited to
minimizeplasticstrainsinthereinforcement.Afterthis,theDSPwasdisassembled,andthecyclic
testatthehighestdisplacementlevelachievedinthefirstbareframetestwasrepeatedtoevaluate
thedamagetotheframeduringthesecondtest.
(3)Atraditionalunreinforcedmasonrypanel(TMP)wasbuiltinsidetheframeusingthesame
bricks,andthethirdcyclicdisplacementtestwasperformed.Displacementsinthethirdtestwere
increaseduntildestructionofthespecimen.
2.1.SpecimenDescription
SolidconcretebricksshowninFigure2withdimensionsof227mm113mm80mmwere
usedforbothdrystackpanels(DSP)andtraditionalmasonrypanels(TMP).

Figure2.Thedimensionsofsolidconcretebricks.

Material tests on brick and mortar include standards tests such as the compressive strength
testsonasinglebrick[30],thelateralmodulusofrupturetest[30],thebondwrenchtest[31]theand
compressivestrengthtestonbothdrystackedandtraditionalmasonryprisms[31],aswellasthe
frictiontestondrystackedmasonrytriplets.Themechanicalpropertiesofthebricksandmortarare
summarizedinTable1.ThecyclictriplettestwasavariationoftheEuropeanstandardtripletshear
test[32].ThetestsetupisshowninFigure3a,b.Inthistest,twoindependentactuatorshavebeen
used: one for the vertical cyclic displacement; the other for the horizontal constant pressure. The
realtime vertical shear load was recorded from the embedded pressure sensor in the vertical
actuator; the relative displacement was recorded from the linear variable displacement
transducers(LVDTs).

Materials2016,9,108

4of17

Table1.Mechanicalpropertiesofthespecimens.
Specimen
Bricks
Traditionalmasonry
Drystackmasonry
Basebeam
Columns
Topbeamandslab

Density
(kg/m3)
2250

2250
2370
2280
2350

ElasticModulus
(MPa)
26,365
20,407
7702.3
32,000
25,000
33,000

CompressiveStrength
(MPa)
28.55
21.5
18.3
21.8
23.6
24.1

TensileStrength
(MPa)
3.25
0.17

2.93
2.34
3.30

Three levels of compression were used for the triplet test: 0.1, 0.3 and 0.5 MPa. For each
pressure case, 3 specimens were tested. For each specimen under each pressure level, 4 cyclic
displacementsloadings were applied; these are: 0.8, 1.6, 2.4 and 3.2 mm; the corresponding
velocitiesare:1mm/min,2mm/min,2mm/minand4mm/min.Forthedrystackprism,theshear
and the normal stress comply with the MohrCoulomb criterion; a frictional factor of 0.66 with a
correlation coefficient of 0.99 was determined. The typical hysteretic loop is shown in Figure 3c.
Detailedtestresultscanbefoundintheliterature[33].

(a)

(b)

(c)
Figure3.Compressionsheartestofdrystackedmasonrytriplets:(a)Schematicdiagramofthetest;
(b)Testsetup;(c)Typicaltestresult.LVDT,linearvariabledisplacementtransducer.

TheRCframedimensions,reinforcementdetailsandlocationofthestraingagesareshownin
Figure 4. Columns were 120 mm 200 mm; the base beam, 330 mm 600 mm; and a slab,
100 mm 600 mm, was cast as a part of the top beam. The mechsanical properties of the frame
elementsweretestedandaresummarizedinTable1.Theyieldstrengthsofthereinforcement(D12:
fy=400MPa;D10:fy=400MPa;D6:fy=210MPa)werealsodeterminedinapreliminarytest,where
D12, D10, D6 means the reinforcement bar with a diameter of 12 mm, 10 mm and 6 mm
respectively.

Materials2016,9,108

5of17

There were 25 layers and 22 layers in the dry stack panel and traditional masonry panel,
respectively. During construction of the dry stack panel, bricks were cut to achieve a tight fit
between columns; the top layer was cut to fill the frame. Even though, it was difficult to achieve
perfectcontactbetweenthepanelandthetopbeam.Therewasanunevenapproximately1mmgap
leftbetweenthem.Asimilar,butwider(approximately60mm)gapwasformedduringconstruction
ofthetraditionalmasonrypanel,whichwasfilledwithdentalplaster.The1:1:6(cement:lime:sand
byvolume)mortarwasusedforthetraditionalmasonrypanel;thethicknessofmortarjointswas
around8mm.AsmallamountofBycol(acommonAustralianairentrainer,RecochemInc.,Sydney,
Australia)wasaddedtoimprovetheworkabilityofthemortar.

(a)

(b)
Figure4.Framedetails:(a)Instrumentationofframe(unit:mm);(b)Sectiondetails.S,straingage.

2.2.CyclicTests
Based on previous research [34], a series of cyclic tests was set by the authors, shown as
Figure5a.Theframewasboltedtothestrongfloor.Theverticalloadwasappliedbythehydraulic
jackthroughthesteelupperspreaderbeam.Athinneopreneplatewasplacedbetweentheframe
and upper spreader beam to ensure uniform load distribution over the specimen. The horizontal
load was applied by a hydraulic actuator, which was fixed to the reaction wall. The installation
photoisshowninFigure5b.
Two types of instruments were employed during the tests: electrical strain gages and linear
variable displacement transducers (LVDTs). The applied load and displacement of the hydraulic
actuators were also measured. The position of LVDTs (abbreviated as L) and strain gages
(abbreviatedasS)areshowninFigure4aandFigure5a(bracketsmeantheLVDTwaslocatedon
theoppositeside).

Materials2016,9,108

6of17

To simulate the gravity load, a constant compressive stress equal to 0.3 MPa, which is
consideredtypicalfora3storyresidentialbuilding,wasappliedduringthetest.Thespecimenwas
loadedinplaneaccordingtothedisplacementhistoryshowninFigure6.Theeastboundforceand
thedisplacementswereconsideredpositive.ThedisplacementratesarelistedinTable2.

(a)

(b)

Figure5.Experimentalsetup:(a)Testsetup(unit:mm);(b)Photos.L,LVDT.

Figure6.Displacementhistory.
Table2.Displacementrates.
Travel(mm)
0.5 1
Travelspeed(mm/min) 0.5 0.5
Period(min)
4
8

1.5
0.5
12

2
0.5
16

3
1
12

4
1
16

6
2
12

8
2
16

10
2
20

13 16 20
3 4 4
17 20 20

InordertoavoidplasticdamagetotheRCframe,theactuatortravelwaslimitedbothinthe
bareRCframeandthedrystackpaneltest.Themaximumtravelof10mmwasappliedinthebare
frame test, and the maximum travel of 16 mm was applied in the dry stack panel test. The
maximumtravelof20mmwasappliedinthetraditionalmasonrypaneltest.Cracksweremarked
duringcyclingastheyappeared.Aftereachcycle,theloadingwasstopped,andthecrackpattern
wasphotographed.
Theappliedhorizontalloadandthehorizontaldisplacementatthetopoftheframe(measured
byLVDT1)wereusedtoevaluatethecyclicbehaviourandtheenergydissipationinthestructure.
In this way, the contamination of results potentially caused by the slip between the spread beam
andtheframeandbythefreeplayintheexperimentalsetupcanbeeffectivelyavoided.Thestory
driftwascalculatedbydividingthemeasureddisplacementofLVDT1bythepanelheightof2m.
3.ExperimentalResults
3.1.CrackPatterns

Materials2016,9,108

7of17

Astheapplieddisplacementwaslimited,therealmostnovisualdamageoccurredduringthe
first bare frame test. Only when the amplitude of cycles reached a maximum of approximately
9mm,therewasasmallhairlinecoverablecrackobservedatthelefttopcorneroftheframe.
ThereweresomesmallcoverablecracksobservedinthetestwithDSPatbothtopcornersof
theframe.Thefirstcrackwasobservedatthesamelocationandthesameamplitudeasinthebare
frame test. The second crack was observed in the opposite top corner at the next level of
displacementsofapproximately11mm.Fourbrickscrackedduringthetest,asshowninFigure7a.
However, the distribution of cracked bricks appeared to be random. This cracking could be
attributedtothepreexistingdamagetothosebricks.

(a)

(b)

Figure7.Finalcrackpatterns:(a)Reinforcementconcreteframewiththedrystackedpanel(DSP);(b)
RCframewiththetraditionalmasonrypanel(TMP).

A typical diagonal cracking failure was achieved in the test with TMP. The crack was found
betweenthetopoftheleftcolumnandthepanelatthestorydriftof0.063%.Thefirstcrackinthe
masonrypanelwasfoundatthebottomrightcorneratthestorydriftof0.099%.Theoriginofthis
crackwassixbricksfromthebottom,anditsteppeddownthroughthemortarjointsallthewayto
the middle of the bottom edge of the panel. No new cracks were observed in the panel until the
storydriftof0.78%.Cracksinthecolumnsbegantodevelopatthestorydriftof0.034%,andtheir
numberandwidthskeptincreasingthroughthetest.
Thespecimenfailedduringthesecondcycleatthestorydriftof0.78%.Thefailurewassudden,
andabouta15mmwidediagonalcrackopenedimmediately.Atthesametime,therightcolumn
developedamajorcrackatthebottom.ThefinalcrackpatternsareshowninFigure7b.
3.2.HystereticBehaviour
Plotsoftheinplanestorydriftversustheinplanelateralforceforallcyclictestsareshownin
Figure8.
In the first bare frame test, the hysteretic loops closely repeated each other for displacement
cyclesofthesameamplitude.Theloopswerealsosymmetricatalllevelsofthedisplacement.Inthe
testwithDSP,thehystereticloopswerestable,butnotsymmetrical.Themaximumforcerecorded
in the negative direction was lower than the positive force at the same displacement. This
asymmetricbehaviourcanbeattributedtotheexistenceoftheunevengapbetweentheframeand
thetopofthemasonrypanel,asmentionedabove.Thepositiveloopshavealargearea,whichis
indicativeofgreaterenergydissipation.Theyareassumedtobetterrepresentthecyclicbehaviour
ofadrystackpanelwithoutthegapatthetop.
InthetestwithTMP,theobservedbehaviourwastypicalforaframedunreinforcedmasonry
panel subjected to cyclic loading. After an initial elastic stage, loops exhibited static hysteresis
attributed to plastic damage in the structure. There was no signal crack in the panel as in some
confinedmasonrycyclictests.Themaximumstrengthwasreachedatdriftlevelscloseto0.5%.At
thispeak,anidealdiagonalcracksuddenlyopenedacrossthemasonrypanel.Afterthepeakload
wasreached,thestorydriftkeptincreasingto0.8%withthestrengthreducedby30%fromthepeak
value(dottedlineinFigure8d).

Materials2016,9,108

8of17

(a)

(b)

(c)

(d)

Figure8.Hystereticloopoftheframewithdifferentinfilledforms:(a)Bareframe;(b)RCframewith
DSP;(c)Doublechecktestonthebareframe;(d)RCframewithTMP.

3.3.LateralStiffness
Stiffnessdegradationisassessedusingsecantstiffnessofthepeakpeakvaluedeterminedfrom
each complete hysteresis loop. In order to analyse the effects of the panel on frame stiffness, the
relativestiffnessKwasused,whichwascalculatedbyEquation(1).

K'

Ki

K0

(1)

whereKiistheaveragesecantstiffnessofeachstorydrift;K0istheinitialsecantstiffnessofbareframe.
The energy dissipation capacity is an important parameter in the evaluation of the potential
seismic resistance of the structure. It was calculated for all three tests as the area inside the
hystereticloops.
Thestorydrift(),relativestiffness(K)andenergydissipationofthreecycles(WD;thiswillbe
discussedinnextsection)arelistedinTable3.ComparingthecontributionofDSPandTMPtothe
overallstructuralresponse,thefollowingobservationscanbemade:
(1)TheDSPincreasedtheinitialstiffnessoftheassemblyslightlytoapproximatelytwotimes
that of the bare frame. The TMP increased the initial stiffness of the assembly massively to
approximately30timesthatofthebareframe;
(2)Intheelasticresponsestageoftheframe(0.18%)[35],thesecantstiffnessoftheframe
withDSPisalmostthesameasbareframes,whilethecorrespondingvalueforthetestwithTMPis
aboutsixtimesthat.

Materials2016,9,108

9of17

Table3.StiffnessandenergydissipationoftheRCframewithdifferentinfilledforms.

(%)
0.02
0.04
0.06
0.09
0.13
0.16
0.25
0.36
0.46

BareFrameTest
WD
F
K
(kN)
(kNmm)
1.90 1.00
0.51
3.80 0.93
1.25
5.50 0.88
2.23
7.16 0.85
3.31
9.40 0.75
6.47
11.08 0.67
15.15
14.03 0.55
38.31
17.13 0.48
64.68
19.19 0.42
91.81

(%)
0.01
0.03
0.06
0.08
0.12
0.17
0.25
0.33
0.44
0.58
0.70

TestwithDSP
WD
F
K
(kN)
(kNmm)
2.19 2.44
0.65
4.39 1.68
5.26
6.5 1.34
12.85
8.21 1.16
22.66
10.54 0.92
51.03
12.65 0.79
75.53
19.38 0.71
147.32
30.28 0.72
255.97
41.46 0.75
441.62
46.10 0.64
555.30
50.50 0.61
758.87

(%)
0.001
0.01
0.02
0.03
0.06
0.1
0.17
0.24
0.32
0.43
0.55
0.78

TestwithTMP
WD
F
K
(kN)
(kNmm)
5.18 30.03
0.47
18.33 27.48
3.07
36.36 21.06
7.32
47
17.6
15.38
61.61 11.25
65.62
70.63 8.62
121.47
87.26 5.94
313.61
99.92 4.75
454.85
113.9 4.09
657.05
129.98 3.40
1381
135.27 2.83
2594
145.99 1.44
6933

3.4.EnergyDissipation
TheenergydissipationinthetestwithDSPwascausedmainlybytherelativeslipandfriction
between bricks inside the panel. There are two distinct stages of the response with two different
mechanisms of energy dissipation. The story drift of approximately 0.2% was a threshold. Before
that,theenergydissipationintheframewasduetotheinternalthermalfrictionintheelasticstage
of response and some microstructural damage in the plastic stage. The energy dissipation in the
panel was due to the constant friction between bricks caused bypanel selfweight only. After the
gap between the frame and panel was closed, the energy dissipation was increasing partially
because more microstructural cracks were developing in the frame, but mainly because
compressionand,asaresult,frictionforceswereincreasingbetweenbricks.
The energy dissipation in the traditional masonry test was caused mainly by the
macrostructural damage to mortar joints, concrete frame and bricks. The first visible cracking of
the frame was observed at the story drift of 0.06%. Before that, energy dissipation was mainly
elasticorcausedbymicrostructuralcracking.
ThetotalenergydissipationinthetestwithTMPbeforetheonsetofcrackingintheframeat
thestorydriftof0.06%was91.87kNmm.Inthetestwiththedrystackpanel,thenewcrackingin
the frame was observed at the story drift of 0.7%. The total energy dissipation before that was
1568.2kNmm,whichisapproximately17timeshigherthaninthetraditionalmasonrytest.
The maximum horizontal load for thebareframe was18.05 kN.The total energy dissipation
beforereachingthisloadwas91.812kNmm.Atthesameloadlevel,thetotalenergydissipationin
the traditional masonry test was 3.07 kNmm. In the dry stack panel test, the total energy
dissipationbeforereachingthisloadlevelwas147.32kNmm,whichisalmost50timeshigherthan
inthetraditionalmasonrytest.
3.5.ResponseMechanisms
Theresponseofthistypeofdualstructuralsystemdependsonthecombinedactionofboththe
RCframeanddrystackmasonrypanel.Asaclassicaltypeofstructure,thelateralcapacityofthe
RC frame has been comprehensively researched. The dry stack masonry behaviour attracted less
researchattention.Theinteractionbetweenthetwoisnotyetfullyunderstood.Itappearsthatthe
DSPexhibitsMohrCoulombbehaviour.However,thecombinedactionoftheframeandthepanel
is not a simple superposition of two mechanisms; the gap between the top of the panel and the
frameplaysakeyroleintheframepanelinteraction.
The envelope response curve for the structure was attained from Figure 8 and shown as
Figure9.Comparedtotheenvelopecurveforthebareframe,thestructuralresponsecanbedivided

Materials2016,9,108

10of17

into three stages: OA (0% to ~0.2% storey drift); AB (~0.2% to ~0.4%); BC (beyond ~0.4%). Three
correspondingresponsemechanismsareshowninFigure10.
Firststage(OA),constantfrictionstage:TheRCframeisinteractingwiththedrystackpanel
compressedbyitsownweightonly.Theframeisnotincontactwiththetopofthepanel,because
the gap between them is open. Frictional forces between bricks are therefore relatively small and
constant.
Second stage (AB), equivalent frictional strut stage:The RC frame is now in contact with the
topofthepanel.Thishastwosignificanteffects:(i)thefrictionbetweenbricksisincreasingdueto
increasing compression of the panel by the frame; and (ii) a type of compressive strut is formed
withinthepanelsimilartothetraditionalinfilledmasonrypanel.Comparedtothefirststage(OA),
thestiffnessofthestructurehasincreased.

Figure9.Envelopecurves.

(a)

(b)

(c)
Figure 10. Response mechanisms: (a) Constant friction stage; (b) Equivalent frictional strut stage;
(c)Plasticstage.

Materials2016,9,108

11of17

Thirdstage(BC),plasticstage:Oneofthecomponentsofthedualstructuralsystemexceedsits
yield limit and is cracking. In this case, the RC frame is cracking and exhibits plastic behaviour.
Comparedtothesecondstage(AB),thestiffnessoftheassemblyhasdecreasedtoaboutthesameas
inthefirststage(OA).TheobserveddamageoftheRCframeisassumedtobethereason.Atthis
stage,thecracksdevelopedatcolumn/beamconnections,asshowninFigure7a.Thisstagecannot
becomparedtobareframeexperimentalresultsonlybecausethebareframetesthadbeenstopped
atthe10mmjacktraveltoavoiddamagingtheframe.
4.EvaluationofDryStackedPanel(DSP)Behaviour
4.1.EvaluationMethod
ComparingthedifferencebetweenthebareframeandtheframewithDSP,wecanseethatthe
contributionofDSPtothecombinedstructuralresponseissignificant.
According to the series of tests and theoretical analysis, five different failure modes of RC
frameswithtraditionalinfillpanelshavebeenidentified[36,37].DSPinfillsbehavequitedifferently
compared to TMP. However, traditional infill panels with weak mortar framed by a strong RC
frameareinsomerespectssimilartoDSP.ThoseTMPsarepronetodiagonal/slidingshearfailure
oftheinfillandtheshearfailureofthewindwardcolumn.Inthiskindofdualstructure,theframe
andtheinfillpanelareoftenconsideredastwoparallelsystemswithdisplacementcompatibilityat
thecompressioncorners.Thisapproachiscalledtheparallelmodel.
IntheDSPstructure,thereisnobondingbetweenbricks,andtheconnectionisweak;therefore,
theparallelmodelwasadoptedfortheanalysisoflateralbearingcapacity.Thebearingloadon
thestructureisresistedjointlybytheframeandthepanel,asexpressedinEquation(2).

F Fb Fp

(2)

where F, Fb and Fp are the lateral capacity of the whole structure, the bare frame and the DSP,
respectively.Thebearingloadonthepanelcouldbeattainedbysubtractingthebearingloadonthe
frameonlyfromthebearingloadontheframe/panelassemblyatthesamehorizontaldisplacement,
asshowninFigure11.

Figure11.Parallelmodelapproach.

Strictlyspeaking,theprincipleofEquation(2)isapplicableonlyaslongasthebehaviourofthe
frame remains linear. This evaluation based on the parallel model is extended here beyond the
linearrangeoftheframeresistanceonlytogaininsightintoenergydissipationintheDSP.Forthis
analysis,thehystereticloopshavebeenprocessedinthefollowingthreesteps:
(1) Discretization step: The two groups of curves (bare frame and frame with DSP; see
Figure8a,b)werediscretisedusingdisplacementincrementsof0.001mmwithlinearinterpolation
betweenthesepointstoensurethattheshapeofthecurveisnotdistorted.
(2) Regularization step: The bare frame curve was scaled up or down to match its peak
displacementvaluetothatoftheframewiththeDSPcurve.Theoriginalandregularizeddataare
listed in Table 4, where D, F is the horizontal displacement and corresponding lateral force

Materials2016,9,108

12of17

achieved from DSP infilled frame experiment; dB and FB is the horizontal displacement and
correspondinglateralforceachievedfrombareframeexperiment.
(3)Decouplingstep:SubtractingthebareframedatafromtheframewithDSPdataattainsthe
hystereticloopsofDSP.
Table4.CalculationoftheforcecontributionofDSP.
LoadCases
1
2
3
4
5
6
7
8
9

FramewithDSP
D(mm)
0.21
0.15
0.58
0.51
1.11
0.84
1.55
1.24
2.34
2.23
3.43
3.00
4.97
5.10
6.50
6.92
8.56
8.49

F (kN)
2.19
2.08
4.39
4.86
6.50
6.64
8.21
8.24
10.54
10.85
12.65
13.38
19.38
17.03
30.28
22.45
41.46
27.27

BareFrame
OriginalData
RegularizedData
dB(mm) FB (kN) dB (mm) FB (kN)
0.39
1.90
0.21
1.02
0.17
1.10
0.15
0.96
0.84
3.80
0.58
2.61
0.47
2.50
0.51
2.69
1.27
5.50
1.11
4.84
0.87
4.10
0.84
3.96
1.70
7.16
1.55
6.53
1.31
5.75
1.24
5.41
2.63
9.40
2.34
8.38
2.19
8.81
2.23
9.00
3.17
11.08
3.43
11.99
3.42
11.36
3.00
9.96
5.07
14.03
4.97
13.75
5.19
15.13
5.10
14.86
7.09
17.13
6.50
15.70
6.90
17.04
6.92
17.10
9.13
19.19
8.56
17.99
8.65
19.38
8.49
19.02

ForceContributionofDSP
AbsoluteValue (kN)
1.17
1.12
1.78
2.17
1.66
2.68
1.68
2.83
2.16
1.85
0.66
3.42
5.63
2.17
14.58
5.35
23.47
8.25

AverageValue (kN)
1.15
1.97
2.17
2.26
2.01
2.04
3.90
9.97
15.86

The frame behaviour during the constant friction and the equivalent friction strut stages of
response described above is almost linear elastic. However, some degradation of stiffness of the
structurewasobservedbetweencyclesatthetwohighestlevelsofdisplacement.Thissignificantly
nonlinear behaviour of the frame is considered unsuitable for analysis based on the accepted
parallel model. Hence, the data for those two cycles was not used in the analysis. The achieved
hystereticloopsforDSPareshowninFigure12.

(a)

(b)
Figure12.Cont.

Materials2016,9,108

13of17

(c)

(d)

(e)

(f)

(g)

(h)

(i)
Figure12.Hystereticloopsfordrystackedpanelatdifferentlevelsofcyclingdisplacement:(a)Load
Case1;(b)LoadCase2;(c)LoadCase3;(d)LoadCase4;(e)LoadCase5;(f)LoadCase6;(g)Load
Case7;(h)LoadCase8;(i)LoadCase9.

Materials2016,9,108

14of17

4.2.SimplifiedMechanicalModel
It is apparent that the hysteretic behaviour of DSP changes as the displacement increase. The
typicalfullhystereticloopforDSPcanbeassumedasinFigure13a.Therearethreedistinctstagesof
behaviour:
(1)Packingstage:Atthelowlevelofthestoreydrift(<0.03%),theinducedhorizontalshearforces
in the panel are generally lower than the static friction forces between bricks. Hence, there is little
relativeslidingbetweenbricks;justtopackbrickstightlyandreachthefullstaticfrictionpotential.The
DSPslightlyincreasesthestiffnessoftheassembly,ascanbeseeninLoadCases1and2.
(2)Constantfrictionstage(CFstage):Withtheincreaseofstorydrift(0.03%<<0.2%),bricks
start sliding, and the relative displacement gradually increases. DSP shows typical frictional
damper behaviour, which is speed independent and has the peakpeak value, which depends on
theselfweightofthepanelandisalmostconstantatabout4kN.Duringthisstage,theDSPactsas
aneffectivefrictionalenergydissipationdevice(seeLoadCases3to6);thebehaviourofDSPinthe
CF stage was decided by the constantfrictionforceFp0, which depends on the selfweightof the
panel.
(3)Equivalentstrutstage(ESstage):Asthelateraldisplacementincreases,thegapbetweenthe
topofthepanelandtheframestarttoclose(>0.2%),thecompressionintheDSPincreaseswitha
consequent increase in the friction between bricks. This point signifies the formation of the
equivalentcompressive/frictionalstrutintheDSP.Inadditiontotheincreasedenergydissipationin
thepanelatthisstage,theequivalentstrutsignificantlyincreasesitsstiffnesscontribution,ascanbe
seeninLoadCases7to9.TheasymmetryofthehystereticcurvesintheESstageofourtestwas
causedbytheunevengapbetweenthetopofthepanelandtheRCframe.Asthegapisuneven,the
hystereticloopsshowasymmetricbehaviourabouttheverticalaxis.

(a)

(b)

(c)
Figure13.Modellingofatypicalhystereticloopfordrystackinfillpanel:(a)Fullhystereticloopfor
DSP;(b)Constantfrictionstage;(c)Equivalentstrutstage

Materials2016,9,108

15of17

5.Conclusions
AnewmasonrysystemdesignedtoimprovetheseismicbehaviourofRCframewithmasonry
panelshasbeendeveloped.Inthissystem,adrystackpanelisbuiltwithmasonryunitscapableof
slidinginplaneofapanel.Aseriesofmaterialandcyclictestswascarried,andthecontributionof
DSPwasinvestigated.
The cracking patterns and hysteretic behaviours of the frame with different infilled panels
have been researched. Typical diagonal cracking failure was achieved in the test with TMP.
DifferentfromtheframewithTMP,therearealmostnocracksoccurringontheframeduringthe
elasticstage;aseriesoffullhysteresiscurveswasachievedfortheframewithDSP,whichindicates
asignificantenergydissipationoftheDSP.
Compared to the traditional unreinforced masonry panel (which increase the stiffness of the
RC frame 30 times), the DSP increases the initial stiffness of frame about two times. During the
elasticresponsestage(0.18%),thesecantstiffnessoftheDSPframeisalmostthesameasthatof
thebareframe,whilethisvalueisaroundsixtimeshigherfortraditionalmasonrypanels.
The addition of DSP to a light RC frame could increase the structural energy dissipation
accordingtotheexperimentalresults.Thesemiinterlockingmasonrypanelscouldbegoodenergy
dissipation devices for frame structures in seismic regions. Their influence requires further
experimentalresearch.
ThetypicalhystereticloopsofDSPwereachievedaccordingtotheparallelmodelbasedonthe
experimental results. The hysteretic loop of DSP exhibits typical frictional behaviour under low
horizontaldrift,whichindicatesthattheDSPhasthepotentialtobedesignedasafrictionaldamper
in a reality program. The mechanism of the DSP frame has been investigated, and two distinct
stages of the response have been identified: they are the constant friction response and the
equivalent frictional strutresponse. The gap between the frameand the panelwasfound to have
significantinfluenceonthecompositeresponseofthestructure.
Author Contributions: Kun Lin, Yuri Zarevich Totoev and Hongjun Liu conceived of and designed the
experiments. Kun Lin performed the experiments. Yuri Zarevich Totoev contributed materials and
experimentaltools.KunLinandHongjunLiuanalysedthedata.HongjunLiuwrotethepaper.Tianyou Guo
assistedintheanalysisoftheexperimentalresults.
ConflictsofInterest:Theauthorsdeclarenoconflictofinterest.

References
1.
2.

3.
4.
5.
6.
7.
8.
9.

Zhao,B.;Taucer,F.;RossettoT.Fieldinvestigationontheperformanceofbuildingstructuresduringthe
12May2008WenchuanearthquakeinChina.Eng.Struct.2009,31,17071723.
Romo,X.;Costa,A.A.;Pauprio,E.;Rodrigues,H.;Vicente,R.;Varum,H.;Costa,A.Fieldobservations
andinterpretationofthestructuralperformanceofconstructionsafterthe11May2011Lorcaearthquake.
Eng.Fail.Anal.2013,34,670692.
Muvafik, M. Field investigation and seismic analysis of a historical brick masonry minaret damaged
duringtheVanEarthquakesin2011.Earthq.Struct.2014,6,457472.
Akpnar, U.; Binici, B. The effect of infill wall collapse on the deformation estimations of reinforced
concreteframes.Int.J.Civ.Eng.Sci.2013,2,171177.
AlChaar,G.;Issa,M.;Sweeney,S.Behaviorofmasonryinfillednonductilereinforcedconcreteframes.J.
Struct.Eng.2002,128,10551063.
TenaColunga, A.; Jurezngeles, A.; SalinasVallejo, V.H. Cyclic behavior of combined and confined
masonrywalls.Eng.Struct.2009,31,240259.
Seong,J.Y.;Min,K.W.;Kim,J.C.AnalyticalinvestigationofanSDOFbuildingstructureequippedwitha
frictiondamper.Nonlinear.Dynam.2012,70,217229.
Occhiuzzi, A. Additional viscous dampers for civil structures: Analysis of design methods based on
effectiveevaluationofmodaldampingratios.Eng.Struct.2009,31,10931101.
Lopez,I.;Nijmeijer,H.Predictionandvalidationoftheenergydissipationofafrictiondamper.J.Sound
Vib.2009,328,396410.

Materials2016,9,108

10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.

25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.

16of17

Celarec,D.;Ricci,P.;Dolek,M.Thesensitivityofseismicresponseparameterstotheuncertainmodelling
variablesofmasonryinfilledreinforcedconcreteframes.Eng.Struct.2012,35,165177.
Dolek, M.; Fajfar, P. The effect of masonry infills on the seismic response of a fourstorey reinforced
concreteframeAdeterministicassessment.Eng.Struct.2008,30,19912001.
Dolek,M.;Fajfar,P.Effectsofmasonryinfillsontheseismicresponseofafourstoreyreinforcedconcrete
frameAprobabilisticassessment.Eng.Struct.2008,30,31863192.
Porco,F.;Porco,G.;Uva,G.;Sangirardi, M.Experimentalcharacterizationofnonengineeredmasonry
systemsinahighlyseismicpronearea.Constr.Build.Mater.2013,48,406416.
Mojsilovi,N.;Stewart,M.G.Probabilityandstructuralreliabilityassessmentofmortarjointthicknessin
loadbearingmasonrywalls.Struct.Saf.2014,52,209218
Ngowi,J.V.StabilityofDryStackMasonry.Ph.D.Thesis,UniversityoftheWitwatersrand:Johannesburg,
SouthAfrica,2005.
Lin,K.;Totoev,Y.Z.;Liu,H.J.;Page,A.W.Modelingofdrystackedmasonrypanelconfinedbyreinforced
concreteframe.Arch.Civ.Mech.Eng.2014,14,497509.
Totoev,Y.Z.;Wang,Z.Y.InPlaneandoutofplanetestsonsteelframewithSIMinfill.InProceedingsof
theTwelfthCanadianmasonrysymposium,Vancouver,BC,Canada,25June2013.
Bansal,D.Interlockingdrystackedmasonry.InProceedingsofthe8thInternationalMasonryConference,
Dresden,German,47July2010.
Adedeji,Y.M.D.Sustainablehousingprovision:Preferencefortheuseofinterlockingmasonryinhousing
deliveryinNigeria.Arch.Res.2012,2,8186.
Ramamurthy,K.;Nambiar,E.K.K.Acceleratedmasonryconstructionreviewandfutureprospects.Prog.
Struct.Eng.Mat.2004,6,19.
Uzoegbo, H.C.; Senthivel, R. An effective retrofitting system for drystack masonry subject to seismic
loading.J.Int.Mason.Soc.2009,22,7184.
Lourenco,P.B.;Ramos,L.F.Characterizationofcyclicbehaviorofdrymasonryjoints.J.Struct.Eng.2004,
130,779786.
Murray,E.B.DryStackedSurfaceBondedMasonry:StructuralTestingandEvaluation.MastersThesis,
BrighamYoungUniversity,Provo,Utah,2007.
Dhanasekar,R.;Ferozkhan,M.;Dhanasekar,M.;Holt,W.Behaviourofdrystackconcretemasonryblocks
under eccentric compression. In Proceedings of the Fourteenth International Brick & Block Masonry
Conference,Sydney,Australia,1720February2008.
Vasconcelos, G.; Lourenco, P.B. Inplane experimental behavior of stone masonry walls under cyclic
loading.J.Struct.Eng.2009,135,12691277.
Uzoegbo,H.C.;Senthivel,R.;Ngowi,J.V.Loadingcapacityofdrystackmasonrywalls.Int.Masonry.Soc.J.
2007,25,4152.
Dyskin,A.V.;Pasternak,E.;Estrin,Y.Mortarlessstructuresbasedontopologicalinterlocking.Front.Struct.
Civ.Eng.2012,6,188197.
RestrepoVlez,L.F.;Magenes,G.;Griffith,M.C.DrystonemasonrywallsinbendingPartI:Statictests.
Int.J.Archit.Herit.2014,8,128.
Vaculik,J.;Griffith,M.C.;Magenes,G.DrystonemasonrywallsinbendingPartII:Analysis.Int.J.Archit.
Herit.2014,8,2948.
Standards Australia. Masonry Units and Segmental Pavers and FlagsMethods of TestDetermining Lateral
ModulusofRupture,AS/NZS4456.15:2003;StandardsAustralia:Sydney,Australia,2003.
StandardsAustralia.MasonryStructures,AS37002001;StandardsAustralia:Sydney,Australia,2001.
European Commission for Standardization. Methods of Test for Masonry, EN 10523:2002; European
CommitteeforStandardization:Brussels,Belgium,2002.
Lin, K.; Totoev, Y.Z.; Liu, H.J.; Wei, C.L. Experimental characteristics of dry stack masonry under
compressionandshearloading.Materials2015,8,87318744.
Mojsilovic,N.;Simundic,G.;Page,A.Masonrywalletteswithdampproofcoursemembranesubjectedto
cyclicshear:Anexperimentalstudy.Constr.Build.Mater.2010,24,21352144.
Ministry of Housing and UrbanRural Development. Code for Seismic Design of Buildings, GB 500112010;
ChinaArchitecture&BuildingPress:Beijing,China,2010.
Shing,P.B.;Mehrabi,A.B.Behaviourandanalysisofmasonryinfilledframes.Prog.Struct.Eng.Mat.2002,
4,320331.

Materials2016,9,108

37.

17of17

Fiore, A.; Netti, A.; Monaco, P. The influence of masonry infill on the seismic behaviour of RC frame
buildings.Eng.Struct.2012,44,133145.
2016 by theauthors; licensee MDPI, Basel,Switzerland. This articleisan open access
article distributed under the terms and conditions of the Creative Commons by
Attribution(CCBY)license(http://creativecommons.org/licenses/by/4.0/).

You might also like