THE EFFECT OF OPENINGS ON OUT-OF-PLANE CAPACITY OF

MASONRY INFILLED REINFORCED CONCRETE FRAMES

Filip ANIĆ1, Davorin PENAVA2, Ivica GULJAŠ3, Vasilis SARHOSIS4, Lars ABRAHAMCZYK5,

Christoph BUTENWEG6

ABSTRACT

Reinforced concrete frames are usually infilled with masonry walls. However, during the seismic design of RC

buildings, the contribution of masonry walls in the overall seismic performance of the frame is mistakenly

neglected. This paper, discusses the effect of openings on out-of-plane behavior of masonry infilled reinforced

concrete frames. A series of numerical analyses were conducted on a 1/2.5 scale model structure of single-bay

single-storey infilled frame containing a door or medium sized window openings, centrically and eccentrically

positioned. Use made of the three-dimensional software ATENA 3D Eng. The numerical model was initially

calibrated against experimental tests. Then, a monotonic out-of-plane uniform load was applied in the structure

and the failure mechanisms and force against displacement relationships obtained. From the results analysis, it

was found that the presence and type of openings significantly influence the out of plane behavior of RC frames

with openings.

Keywords: masonry, infill, RC frame, opening, finite element method, nonlinear analysis, out-of-plane

1. INTRODUCTION

Structural frames, constructed by reinforced concrete (RC) are often infilled with masonry

panels/walls. However, in structural practice, the influence of masonry infill on seismic behavior is

usually ignored, i.e. is considered as a non-bearing element as laid out in Eurocode 8 provisions (CEN

2005). Research undertaken over the last few decades demonstrated that they utterly contribute to the

overall dynamical behavior of the frames (Dowrick 2009). However, most of the researches

investigate the effect of the in-plane (IP) behavior rather than the out-of-plane (OOP) one. With

respect to research in the OOP loading, the main focus was on fully infilled frames (Asteris et al.

2017), with certain exceptions to single openings such as in Akhoundi et al. (2016), i.e. there is no

classification, or analytical models on openings and their effect on OOP behavior.

The first to investigate the effect of the OOP loading on masonry infilled frames was in late 50’s with

the research of (McDowell et al. 1956). There the “arching action” was established. From that point,

nearly the entire field of OOP loading adopted the same loading method of a uniform (or rarely point)

load applied directly onto the infill (Akhoundi et al. 2016; Abrams et al. 1996; Pereira et al. 2011;

1PhD student, Faculty of Civil Engineering Osijek, Josip Juraj Strossmayer University of Osijek, Osijek, Croatia,

fanic@gfos.hr 2Dr., Faculty of Civil Engineering Osijek, Josip Juraj Strossmayer University of Osijek, Osijek, Croatia,

dpenava@gfos.hr 3Professor, Faculty of Civil Engineering Osijek, Josip Juraj Strossmayer University of Osijek, Osijek, Croatia,

iguljas@gfos.hr 4Dr., School of Engineering, School of Engineering, Newcastle University, Newcastle upon Tyne, United

Kingdom, vasilis.sarhosis@newcastle.ac.uk 5Dr., Earthquake Damage Analysis Centre EDAC, Bauhaus University, Weimar, Germany,

lars.abrahamczyk@uni-weimar.de 6Professor, Faculty of Civil Engineering, RWTH Aachen University, Aachen, Germany, butenweg@lbb.rwth-

aachen.de

2

Hak et al. 2014). Those loading methods are suitable for blast modelling as shown in (Parisi et al.

2016; Wu & Hao 2007). For the seismic analysis of infilled frames, this approach may be inaccurate

as the seismic excitement is transferred trough frames and diagraphs and not trough the infill, such as

it is used in IP loading (Nicola et al. 2015).

Consequently, this paper studies OOP response of six infilled RC frames configurations by a 3D

micro-modelling approach using the Atena3D software (Cervenka Consulting 2015). Those six

structural configurations were divided into two groups to account variation in size and location of

openings (Penava 2012) and presented in Table 1. The first group (I) consisted of four masonry

infilled RC frames containing an unconfined opening (e.g. door or window) centrically and

eccentrically positioned. The second group (II) had two reference specimens, i.e., fully infilled framed

and bare frame. Hence, the research includes: a) II/1 RC frame without infill; b) II/2 RC frame with

full infill; c) I/1 RC frame with centric door opening in the infill; d) I/2 RC frame with centric window

opening in the infill; e) I/3 RC frame with eccentric door opening in the infill; and f) I/4 RC frame

with eccentric window opening in the infill. The opening area (𝐴𝑜) was selected to be 2.0 m2 which

falls within the range (i.e., 𝐴𝑜 > 1.5 m2 and 𝐴𝑜 > 2.5 m2) defined by EN 1998-1 (CEN 1998). RC

frames were designed as medium class ductility frames (DCM) according to EN 1992-1-1 provisions

(CEN 2004). Atena 3D (Cervenka Consulting 2015) software used for the analysis. For further details

on the computational models, the reader is kindly requested to refer to (Anić et al. 2017).

Table 1 Classification and description of the specimens (Penava et al. 2016)

Specimen Appearance of the

specimen

Opening

Group Mark Type and area Position

I

1

Door Centric

lo / ho = 0.35 / 0.90 m

eo = li / 2 = 0.90 m Ao = 0.32 m2

Ao / Ai = 0.14

2

Window Centric

lo / ho = 50.0 / 60.0 cm eo = li / 2 = 0.90 m

P = 0.40 m Ao = 0.30 m2

Ao / Ai = 0.13

3

Door Eccentric

lo / ho = 0.35 / 0.90 m

eo = hi / 5 + lo / 2 = 0.44 m Ao = 0.32 m2

Ao / Ai = 0.14

4

Window Eccentric

lo / ho = 50.0 / 60.0 cm eo = hi / 5 + lo / 2 = 0.44 m

P = 0.40 m Ao = 0.30 m2

Ao / Ai = 0.13

II

1

Not considered in OOP

2

- -

3

Figure 1 RC frame and masonry unit dimensions

2. DEVELOPMENT OF THE MODEL, MATERIALS AND BOUNDARY CONDITIONS

2.1 Development of the model and materials

For the numerical model, 3D solid elements for concrete, clay blocks and elastic plates were used. In

addition, 2D interface elements to represent block to block contacts and block to frame contact

implemented. Finally, 1D truss rebar elements for concrete were selected as shon in (fig.3).

Material properties used in this paper were adopted from (Penava et al. 2016) (tab.2&3). Bedjoint

interface materials take into consideration the interlocking effect (fig.2) to account for the effect of the

mortar filled between the voids in the masonry blocks. Material model CC Nonlinear cementitious 2

(Cervenka et al. 2012) for clay block in (Penava et al. 2016) incorporates the experimentally obtained

and calculated properties in direction of voids. However, for masonry walls, bending tests conducted

in accordance with EN 1052-2 provisions (CEN 1999), clay block parameters such as tension strength

ft and tension softening function had to be modified in order to simulate the tests. The tests were

carried out as a preliminary study of infills OOP behavior, i.e. to establish the governing OOP bending

parameters. Further on, the tensile strength was changed from that in direction of voids 1.80 MPa to

strength perpendicular to voids 0.38 MPa. The end displacement of tension softening function ranged

from 1∙10-4 m to 1∙10-5 m (fig.1b). Clay blocks fracture energy is depended on the blocks tensile

strength ft (eq.3). Fracture energy calculated with tension strength in direction of voids. Equations 1, 2

and 3 are adopted from (Cervenka et al. 2012). Minimum normal Knn,min and tangential stiffness Ktt,min

are normal and tangential stiffness divided by 1000 as recommended by (Cervenka et al. 2012).

Knn = E / t (1)

Ktt = G / t (2)

Where t is mortar thickness (standard thickness of 10 mm).

Gf = 0.000025 ft (3)

ho

lo

180

li = 130

20/20

4

Table 2 CC Nonlinear cementitious 2 material model

Description Symbol Frame concrete Concrete lintel Clay block Unit

Elastic modulus E 4.100 E+04 3.032 E+04 5.650 E+03 MPa

Poisson's ratio μ 0.200 0.200 0.100 /

Tensile strength ft 4.000 2.317 0.380 MPa

Compressive strength fc -5.800 E+01 -2.550 E+01 -1.750 E+01 MPa

Specific fracture energy (eq.3) Gf 1.200 E-04 5.739 E-05 4.500 E-04 MN/m

Crack spacing smax 0.125 0.125 / m

Tensile stiffening cts 0.400 0.400 / /

Critical compressive disp. Wd -5.000 E-04 -5.000 E-04 -5.000 E-04 /

Plastic strain at fc εcp -1.417 E-03 -8.411 E-04 -1.358 E-03 /

Reduction of fc due to cracks rc.lim 0.800 0.800 0.800 /

Crack shear stiffness factor SF 20.000 20.000 20.000 /

Aggregate size 1.600 E-02 2.000 E-02 / m

Fixed crack model coefficient 1.000 1.000 1.000 /

Table 3 CC 3D Interface material model

Description Symbol Mortat bedjoint Mortar headjoint Unit

Normal stiffness (eq.1) Knn 5.65 E+05 8.50 E+04 MN/m2

Tangential stiffness (eq.2) Ktt 2.57 E+05 3.86 E+04 MN/m2

Tensile strength ft 0.20

0.20

MPa

Cohesion c 0.35

0.35

MPa

Friction coefficient

0.24

0.24

/

Interlock function

see fig. 2a (where

applicable) / /

a) Interlocking function b) Tension softening function

Figure 2 Interface function

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

0.0 0.5 1.0 1.5 2.0

c/c 0

d (mm)

0.04, 3

0.0

1.0

0.0E+00 1.0E-05

σ/f

t

d (mm)

5

2.3 Boundary conditions

Infill loaded as shown in (fig.3). First, the 365 kN of vertical force applied to the column tops in five

steps, after which column supports are active in y and z direction. Then, a uniformly distributed load

equal to 0.002 MPa/step applied in y direction and at the entire area of the infill (excluding openings

where applicable); hence the monotonic load. While the OOP load is active, frames beam is supported

on the opposite face in regards to the infills load face (–y direction). Foundation support is active

through the whole calculation. Due to inability to develop interlocking between masonry unit and the

concrete elements, bedjoints on relations infill – frame has interlocking functions (fig.2) turned off.

Also, during the casting of lintel, concrete slips fully into the voids of masonry units, therefore, a

perfect connection on those parts was adopted.

Figure 3 Numerical model setup

Table 4. Crack propagation by stage

Crack

stage Full infill Openings

1st

Front: Diagonal cracks developed under

approx. 45° in the upper area;

Back: Horizontal cracks appear on beam –

infill connection

Front: Crack appeared around lintel; Back:

Horizontal cracks appear on beam – infill

connection

2nd

Front: Lower diagonal cracks start to

develop

Back: Horizontal cracks appear on

foundation – infill connection

Front: Crack around lintel propagate towards the

beam following approx. 45° angle

3rd

Front: as diagonal cracks develop,

horizontal cracks at infills mid height start

to develop. Back: Diagonal cracks on the

upper section start to developed

Back: Horizontal cracks appeared on foundation –

infill connection

4th Back: cracks developed at the infill around

the frame, heavy cracking of the beam

Front: Horizontal crack propagated in about infill

mid height appear. Beam was heavily cracked.

5th Front: fully developed diagonal cracking Front: Cracks started to fill the lower part of the

infill with angle of approx. 45°

y

z

x

73.000

kN/step

73.000

kN/step

0.002

MPa/step

3D solid elements

1D truss elements

support

monitoring point

6

3. RESULTS’ DISCCUSIONS AND CONCLUSIONS

Figure 4 shows the force - displacement relationships. In figure 4, displacements are the maximum

global displacement in the y direction. Figure 4a shows the area load – displacement relations, and

figure 4b shows force – displacement relations as the bearing can be different if the area is taken into

account by formula 1. Horizontal lines represent the value of the best fitting analytical model by

(Klingner et al. 1997), calculated as per eq.2. It is to be noted that analytical models excessively vary

among each other as showed in (Asteris et al. 2013).

W = w (A – Ao) (1)

l

hl

l

x

xMhhlM

lhw

yh

yv

yhyv2/

ln2ln82

(2)

, where: xy is out of plane deflection, My OOP resisting moment, h horizontal, v vertical direction.

Figure 5 shows the minimum principal stresses along the cross-section of the infill. Table 4 shows the

crack propagation at five different stages of applied load. Figure 6 shows the crack development at the

front and back side of the infill wall. Only cracks greater than 0.1mm are shown in the Figure 6 sine

these are visible with naked eye (ATC-43 1998). Load applied at the front side of the masonry wall.

Figure 7 shows the minimum principal stresses at the front and back side of the wall. Figure 5, 6 and 7

show the minimum principal stresses and crack patterns when maximum displacement in the wall is

equal to 6 cm.

From Figures 5, 6, 7 and Table 4 it can be concluded that:

a) Computational models did simulate the expected arching failure, see figure 5. The three points

of clamping that make the compression arch are clearly visible on the full infill model (III/2).

The separation of clay blocks from each other by bedjoints (developing tension) at maximum

deflection is visible at Figure 7, where gray areas denotes shear sliding and tensile failure at

the joints.

b) Compression arch has also develop in the infill outside the opening in group I (fig.5).

However, there were 4 distinct clamping points: 1. infill – beam, 2. infill – lintel, 3. infills

deflection point, 4. infill – foundation. As the consequence of 2. point of clamping, RC lintel

developed high principal stresses (fig.7).

c) Due to the development of the compression arch together with the loading a single face of the

infill, front and back side of the masonry wall have different stresses and crack patterns.

d) Cracks that developed on full infill model (fig.6a&b) replicate the arching failure when the

infill is connected by all sides to the frame as shown in (Akhoundi et al. 2016; Dawe & Seah

1989; Furtado et al. 2015).

e) The location of openings in the masonry infill wall significantly influence the bearing capacity

(fig.4), crack (fig.6) and stress (fig. 7) development. Masonry infilled wall panels containing

centric openings had higher bearing capacity than those with eccentric openings.

f) Masonry infill wall panels with centric window opening and full infill behaved in a similar

manner (fig.4b). Similar results also obtained from (Akhoundi et al. 2016). It should be

highlighted that for the model developed by (Akhoundi et al. 2016), a slightly higher capacity

observed for the infill wall with centric window than the one with the full infill (2%

difference).

g) Comparing the analytical results by (Klingner et al. 1997) to the numerical results obtained in

this study, it can be concluded that approx. 15% difference obtained for the full infill model,

3% for centric window model and 17% for centric door model.

h) As noted and calculated by (Moghaddam & Goudarzi 2010), there are two possible failure

modes: a) as a results of bending and excessive shear (slender infills)– shear failure wcr =

0.098 MPa; and b) due to crushing on clamping points wmax = 0.378 MPa. Seemingly, all

frames would fall under the category of crushing failure mode no matter the openings.

Likewise, infills cannot be considered slender.

7

i) In accordance with points b), f) & h) it can be concluded that in comparison with the full infill

model, openings did not affect the failure mode.

a) Area load versus displacement b) Force versus displacement

Figure 4. Load – displacement diagram

a) II/2 I/1 I/2 I/3 I/4

Absolute deformation ×1

Figure 5. Minimum principal stress at max. displacement cross section

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.00 2.00 4.00 6.00

w (

MP

a)

d (cm)

I/1 I/2I/3 I/4III/2 Klingner

0

100

200

300

400

500

600

0 2 4 6

W (

kN

)

d (cm)

I/1 I/2

I/3 I/4

III/2 Klingner full infill

Klingner door Klingner window

0

-1

-3

-4

-5

-6

-8

-9

-10

σmin (MPa) < −10 >0

8

a) III/2 front b) III/1 back

c) I/1 front d) I/1 back

e) I/2 front f) I/2 back

g) I/3 front h) I/3 back

i) I/4 front j) I/4 back

Min. crack width = 0.1 mm ; deformation × 1 ; crack width multiplier ×1

Figure 6 Crack patterns

9

a) III/2 front b) III/1 back

c) I/1 front d) I/1 back

e) I/2 front f) I/2 back

g) I/3 front h) I/3 back

i) I/4 front j) I/4 back

Figure 7. Minimum principal stress

0

-3

-5

-8

-10

-13

-15

-18

-20

σmin (MPa) < −20 >0

10

4. ACKNOWLEDGMENTS

The research presented in this article is a part of the research project FRAmed-MAsonry composites

for modelling and standardization (HRZZ-IP-2013-11-3013) supported by Croatian Science

Foundation and its support is gratefully acknowledged. The authors are grateful to Professor Vladimir

Sigmund (1956 – 2016) for his significant contribution and direction towards the research described in

this paper.

5. REFERENCES

Abrams, D.P., Angel, R. & Uzarski, J., 1996. Out-of-Plane Strength of Unreinforced Masonry Infill

Panels. Earthquake Spectra, 12(4), pp.825–844.

Akhoundi, F. et al., 2016. Out-of-plane response of masonry infilled RC frames : Effect of

workmanship and opening. , pp.1147–1154.

Anić, F., Penava, D. & Sarhosis, V., 2017. Development of a three-dimensional computational model

for the in-plane and out-of-plane analysis of masonry-infilled reinforced concrete frames. In 6th

International Conference on Computational Methods in Structural Dynamics and Earthquake

Engineering. Rhodes Island, Greece.

Asteris, P.G. et al., 2013. Mathematical micromodeling of infilled frames: state of the art. Engineering

Structures, 56, pp.1905–1921.

Asteris, P.G. et al., 2017. Numerical modelling of out-of-plane response of infilled frames: State of the

art and future challenges for the equivalent strut macromodels. Engineering Structures, 132,

pp.110–122. Available at: http://linkinghub.elsevier.com/retrieve/pii/S0141029616308896.

ATC-43, 1998. FEMA 306. EVALUATION OF EARTHQUAKE DAMAGED CONCRETE AND

MASONRY WALL BUILDINGS. Basic Procedures Manual,

CEN, 2004. Eurocode 2: Design of concrete structures - Part 1-1: General rules and rules for

buildings (EN 1992-1-1:2004), Brussels: European Committee for Standardization.

CEN, 2005. Eurocode 6: Design of masonry structures - Part 1-1: General rules for reinforced and

unreinforced masonry structures (EN 1996-1-1:2005), Brussels: European Committee for

Standardization.

CEN, 1998. Methods of Test for Masonry - Part 1: Determination of Compressive Strength (EN 1052-

1:2004), Brussels: European Committee for Standardization.

CEN, 1999. Methods of Test for Masonry - Part 2: Determination of Flexural Strength (EN 1052-

2:1999), Brussels: European Committee for Standardization.

Cervenka, V., Jendele, L. & Cervenka, J., 2012. ATENA Program Documentation Part 1 Theory,

Prague: Cervenka Consulting Ltd.

Cervenka Consulting, 2015. ATENA for Non-Linear Finite Element Analysis of Reinforced Concrete

Structures.

Dawe, J.L. & Seah, C.K., 1989. Out-of-plane resistance of concrete masonry infilled panels. Canadian

Journal of Civil Engineering, 16(6), pp.854–864. Available at:

http://www.nrcresearchpress.com/doi/abs/10.1139/l89-128.

Dowrick, D.J., 2009. Earthquake Resistant Design and Risk Reduction 2nd ed., John Wiley & Sons

Ltd.

Furtado, A. et al., 2015. Experimental Characterization of the In-plane and Out-of-Plane Behaviour of

Infill Masonry Walls. Procedia Engineering, 114, pp.862–869. Available at:

http://linkinghub.elsevier.com/retrieve/pii/S187770581501680X.

Hak, S., Morandi, P. & Magenes, G., 2014. Out-of-plane Experimental Response of Strong Masonry

Infills. In 2nd European Conference on Earthquake Engineering and Seismology. Istanbul,

11

Turska.

Klingner, R.E. et al., 1997. Evaluation of analytical verification of shaking table data from infilled

frames. The Masonry Society Journal (TMS), 15(2), pp.33–41.

McDowell, E.L., McKee, K.E. & Sevin, E., 1956. Arching Action Theory of Masonry Walls. Journal

of the Structural Division, 82(2), pp.1–8.

Moghaddam, H. & Goudarzi, N., 2010. Transverse resistance of masonry infills. ACI Structural

Journal, 107(4), pp.461–467.

Nicola, T. et al., 2015. Masonry infilled frame structures: State-of-the-art review of numerical

modelling. Earthquake and Structures.

Parisi, F., Balestrieri, C. & Asprone, D., 2016. Out-of-plane blast capacity of load-bearing masonry

walls. , (2010), pp.991–998.

Penava, D., 2012. Influence of openings on seismic response of masonry infilled reinforced concrete

frames.

Penava, D., Sigmund, V. & Kožar, I., 2016. Validation of a simplified micromodel for analysis of

infilled RC frames exposed to cyclic lateral loads. Bulletin of Earthquake Engineering, 14(10),

pp.2779–2804. Available at: http://link.springer.com/10.1007/s10518-016-9929-0.

Pereira, P.M.F. et al., 2011. Behavior of masonry infill panels in rc frames subjected to in plane and

out of plane loads. In 7th international conference amcm. Kraków, Poland.

Wu, C. & Hao, H., 2007. Safe Scaled Distance for Masonry Infilled RC Frame Structures Subjected to

Airblast Loads. Journal of Performance of Constructed Facilities, 21(6), pp.422–431. Available

at: http://ascelibrary.org/doi/10.1061/%28ASCE%290887-

3828%282007%2921%3A6%28422%29.