Out-of-Plane Behavior of In-Plane Damaged Masonry Infills

Retrofitted with TRM and Thermal Insulation

P. D. Gkournelos, Ph.D.1; and T. C. Triantafillou, M.ASCE2
Downloaded from ascelibrary.org by "Indian Institute of Technology, Hyderabad" on 09/20/23. Copyright ASCE. For personal use only; all rights reserved.

Abstract: The effect of in-plane damage on the out-of-plane response of retrofitted and thermally insulated masonry infills was examined in
this paper through a set of experiments performed on a medium-scale reinforced concrete frame. Structural reinforcement was realized
through the use of textile reinforced mortar (TRM), while expanded polystyrene boards were used for thermal insulation. Various specimen
configurations were tested both in- and then out-of-plane sequentially for each infill specimen. Experimental results have shown that the
TRM-based strengthening scheme can improve the out-of-plane response of masonry infills both in terms of strength and stiffness, especially
in the case of predamaged walls, where strength increases of above 80% were achieved. The addition of insulation arranged in a sandwich
configuration resulted in a slight out-of-plane improvement but was not as effective in the case of predamaged infills due to prior in-plane
loading, which caused partial debonding of the board. An analytical model is also proposed and validated against the experimental data,
which can predict the out-of-plane behavior of a masonry infill while also accounting for the existence of reinforcement and prior damage.
Finally, using the same model in a number of case studies, generalized response diagrams are produced and a set of simplified empirical
equations is suggested. DOI: 10.1061/JCCOF2.CCENG-4324. © 2023 American Society of Civil Engineers.
Author keywords: Arching model; In-plane loading; Masonry infills; Out-of-plane loading; Seismic retrofitting; Textile-reinforced mortar;
Thermal insulation.

Introduction and Background like shear-wall elements, significantly increasing the in-plane
strength and stiffness of RC frames. However, there have been nu-
Masonry infills have been used widely in reinforced concrete (RC) merous cases where ill-positioned or poorly constructed infills have
buildings as a form of isolation from the outside environment or led to partial of even total collapses of RC buildings, due to forma-
just to divide internal spaces, due to their low cost, ease of construc- tion of soft stories, failures of captive columns, torsional side-
tion, and ability to form practically all kinds of shapes. During the effects stemming from irregularities in plan, out-of-plane failures,
design phase, their structural influence is usually ignored, and the and so forth (e.g., Fardis and Panagiotakos 1997; Fardis 2000).
RC elements are detailed to cope with the entirety of the externally Therefore, during the process of structural analysis and retrofitting
applied loads. This simplification is a sensible one and generally of an existing RC building, masonry walls should be considered
leads to a safe-sided, over-design of the RC skeleton. Moreover, it and taken advantage of, while making sure that their existence
does not fall far from the reality in the case of modern buildings, will not cause any adverse effects.
in the sense that their lateral strength and stiffness do not match up The observation of the fact that the masonry infills can have a
to the ones of the primary RC skeleton, the reason being that contem- beneficial effect on the seismic response of RC buildings has led
porary seismic regulations impose strict detailing rules for all the pri- to the development of various retrofitting techniques around
mary structural members. In the case of old RC structures however, them. These methods aim at converting them from secondary ele-
things are reversed: the concrete is weaker, cross sections have ments to primary structural ones, able to successfully resist a signif-
smaller dimensions, and the steel reinforcement density is consider- icant portion of the seismic loads. Some typical examples include
ably less. All the above lead to a lower lateral strength and stiffness, their strengthening using steel-reinforced mortar overlays (e.g.,
thus the existing infills are relatively stronger, stiffer, and their influ- Acun and Sucuoglu 2006), fiber-reinforced cementitious compos-
ence on the global structural behavior is much more significant. ites (e.g., Kyriakides and Billington 2008; Dehghani et al. 2015),
The fact that masonry infills can affect the seismic response of fiber-reinforced polymers (FRPs) (e.g., Triantafillou 1998, 2001;
an existing building is nowadays widely acknowledged and has Ozcebe et al. 2003; Saatcioglu et al. 2005; Yuksel et al. 2006;
Almusallam and Al-Salloum 2007; Altin et al. 2008; Ozden et al.
been documented in the past (e.g., Mehrabi et al. 1996; Fardis
2011), and more recently textile reinforced mortars (TRMs) (e.g.,
and Panagiotakos 1997). When they are evenly distributed in
Koutas et al. 2014, 2015a, b; Akhoundi et al. 2018; Koutas and
plan and height, their effect is mainly favorable because they act
Bournas 2019; Sagar et al. 2019).
Recently, the technique of strengthening of masonry elements
1
Univ. of Patras, Patras 26504, Greece (corresponding author). ORCID: and infills with TRM started to be combined with conventional
https://orcid.org/0000-0002-9054-3900. Email: panagiotisgournelos@ thermal insulation solutions, thus forming an integrated system
gmail.com for the concurrent seismic and energy upgrading of existing struc-
2
Professor, Univ. of Patras, Patras 26504, Greece. ORCID: https://orcid
tures (e.g., Triantafillou et al. 2017, 2018; Pohoryles et al. 2022).
.org/0000-0003-0263-3955. Email: ttriant@upatras.gr
Note. This manuscript was submitted on March 25, 2023; approved on This system has the potential to meet both retrofitting goals at a
July 21, 2023; published online on September 15, 2023. Discussion period cost which is lower than the independent application of the two up-
open until February 15, 2024; separate discussions must be submitted for grading schemes (Bournas 2018; Gkournelos et al. 2019). For the
individual papers. This paper is part of the Journal of Composites for Con- same purpose, other integrated systems involve the use of
struction, © ASCE, ISSN 1090-0268. geopolymer-based inorganic matrix composite systems, in which

© ASCE 04023054-1 J. Compos. Constr.

J. Compos. Constr., 2023, 27(6): 04023054

 the employed mortars themselves serve simultaneously as the tex- different strengthening configurations as well as the respective ex-
tile matrix and the thermal barrier (Longo et al. 2020, 2021). periments that were conducted.
The present paper examines the aforementioned TRM-based in- The RC frame was constructed using small-aggregate C30/37
tegrated retrofitting system and its application on masonry infills concrete, B500C steel for the longitudinal reinforcement, and
through in- and out-of-plane experiments conducted on medium- mild steel (S220 quality) for the transverse one. All the construction
scale, RC, masonry infilled frames. The seismic strengthening works, namely formwork preparation, reinforcement placement,
was realized using a glass fiber textile embedded inside a fiber- and concrete pouring, were performed by experienced personnel.
reinforced cementitious mortar, while for the energy upgrading, As far as the masonry infills are concerned, these were constructed
expanded polystyrene (EPS) plates were employed. Various using six-hole, perforated, clay bricks sourced from a local factory,
strengthening configurations were tested to examine the interaction combined with a regular, in situ mixed building mortar, composed
Downloaded from ascelibrary.org by "Indian Institute of Technology, Hyderabad" on 09/20/23. Copyright ASCE. For personal use only; all rights reserved.

of the different components and determine their optimal placement. of cement, lime, and sand in a ratio roughly equal to 1:1:5, by vol-
The purpose of the research presented here is to study, for the first ume. The infills were built in a double-layer configuration resulting
time: (a) the effect of in-plane damage on the out-of-plane response in a total wall thickness of 120 mm, which is 20 mm less than the
of the masonry infills with combined structural and thermal retro- width of the frame, and hence they were arranged to coincide with
fitting; (b) the effect of thermal insulation on the structural behav- one of its faces; hereafter referred to as the exterior face. All the in-
ior; (c) the development of a simple analytical model, able to fills were built by an experienced mason and cured for a period of
produce good quality estimates for the out-of-plane response, tak- 28 days.
ing into account the damage due to in-plane loading; and (d) the use For the TRM reinforcement, a commercial glass fiber textile
of the analytical model in parametric studies. combined with a commercial fiber-reinforced cementitious mortar
was employed. The textile was made of polymer-coated, alkali-resistant
glass fiber rovings arranged in two orthogonal directions at a mesh
Experimental Program spacing of 18.1 × 14.2 mm. Its dry weight of 280 g/m2 was almost
equally distributed in the two directions: 145 g/m2 in the warp and
135 g/m2 in the weft direction. That, combined with the fiber density
Test Specimens and Materials of 2,600 kg/m3, results in a nominal thickness of 0.056 mm for the
For the purposes of the experimental campaign, a single 1/3 scale warp and 0.052 mm for the weft. The elastic modulus and the tensile
RC frame was constructed and used during all the subsequent in- strength of the dry glass fibers were equal to 80 GPa and 2,600 MPa
and out-of-plane tests (Fig. 1). For that reason, the frame was de- (characteristic values) according to the manufacturer. The mortar was
signed and detailed to be sufficiently flexible, ductile, and able to mixed in situ during the strengthening phase, at a binder:water ratio
withstand the repeated experiments with minimal damage. Inside of 4.2:1, by weight, resulting in plastic consistency and good workabil-
this frame, the different masonry infills were constructed and re- ity (checked empirically).
ceived seismic and energy retrofitting; in all cases, the retrofitting The TRM reinforcement was applied following a typical wet
was applied a month after the construction and the experiments lay-up procedure. First, any protruding building mortar pieces
were conducted a month after it, to provide enough time for all were broken away from the masonry surface, which was then
the materials to reach their nominal strengths. cleaned from dust and loose particles using high air pressure and
In total, three different configurations were subjected to in- and thoroughly saturated with water. A first mortar layer was applied
out-of-plane experiments. The first included unreinforced masonry over the entire masonry and the surrounding RC frame elements
infills (C, control), the second TRM-reinforced infills (R, rein- and within it, the first textile piece was carefully inserted. Next,
forced), and the third TRM-reinforced and EPS-insulated infills in the case of R specimens, a second mortar layer was added on
(RI, reinforced and insulated). Since the mechanical properties of top along with the second textile piece and finally, a finishing
the thermal insulation are considerably reduced compared to thin layer of mortar was applied. In the case of RI specimens, the
those of the TRM and the masonry, it was decided not to consider 20-mm-thick thermal insulation boards were bonded over the first
any configuration in which the insulation would be placed as the TRM layer using the same mortar, and outside of them, a final tex-
outermost layer. Every configuration included tests on two infills: tile and mortar layer was added, practically bonded on the EPS in-
the first infill was only tested out-of-plane (*_ND, no damage) sulation. No mechanical anchorage of the TRM or the insulation
while the second was first subjected to cyclic in-plane testing and was employed so as not to compromise the simplicity nor the appli-
then tested out-of-plane (*_D, damage). Table 1 lists all the cability of the system for practical cases; however, it is expected
that the employment of such a system would yield even better re-
sults for the TRM-reinforced specimens. In the present study, the
bond of the system relied solely on the mortar, which according
to the manufacturer had a tensile adhesion strength of 1.2 MPa
on concrete, 0.15 MPa on EPS, and 0.20 MPa on extruded

Table 1. Test matrix

Tested Tested
Specimen ID Configuration in-plane out-of-plane
C_ND No Yes
Masonry
C_D Yes Yes
R_ND No Yes
Masonry TRM (×2)
R_D Yes Yes
RI_ND Masonry TRM No Yes
Fig. 1. Dimensions (in mm) and steel detailing of the RC frame. RI_D EPS TRM Yes Yes

© ASCE 04023054-2 J. Compos. Constr.

J. Compos. Constr., 2023, 27(6): 04023054

 Concrete specimens were obtained during the two pouring
phases of the frame; cubic (150 × 150 × 150 mm) during the
pouring of the foundation and cylindrical (150 × 300 mm) ones
during that of the superstructure. The specimens were subjected
to axial compression tests according to EN 12390-3 (CEN
2019b), yielding the mean compressive strength values of
43.8 and 39.2 MPa for the foundation and the superstructure
concrete, respectively, which are in agreement with the nominal
quality of C30/37. Tensile experiments were performed to deter-
mine the steel reinforcement properties according to EN
Downloaded from ascelibrary.org by "Indian Institute of Technology, Hyderabad" on 09/20/23. Copyright ASCE. For personal use only; all rights reserved.

15630-1 (CEN 2019a), and yield strengths of 538 and

259 MPa were obtained for the longitudinal and the transverse
rebars, respectively; these values are in accordance with the
Fig. 2. Stratigraphy of the TRM-strengthening scheme. nominal properties too.
The building mortar as well as the reinforcing mortar were both
subjected to flexural and compression testing to verify their respec-
polystyrene (XPS) (characteristic values). The stratigraphy of the tive properties. Flexural testing was conducted on three 40 × 40 ×
intervention scheme is schematically depicted in Fig. 2. 160-mm prisms, which were collected during the infill construction
The TRM and insulation application was performed by the first or reinforcement phases and prepared within a steel mold with three
author of this paper along with the help of undergraduate and post- identical compartments. They were subjected to three-point bend-
graduate students. This alone highlights the ease of application of ing to determine their flexural strength, whereas their compressive
the proposed retrofitting method, as it does not contain any danger- strength was evaluated by subjecting the two fractured parts of each
ous materials or difficult installations. prism to axial compression applied on a 40 × 40-mm area using
steel plates according to EN 1015-11 (CEN 2006). Through this
Material Characterization procedure, the values of 2.5 and 10.7 MPa were obtained for the
flexural and compressive strength of the building mortar, while
All the materials which were employed during the construction of the respective values for the reinforcing mortar were 5.4 and
the RC frame, the infills, and the retrofitting scheme application 11.2 MPa.
were characterized experimentally. The results are summarized Square masonry assemblages, measuring 450 × 450 × 120 mm,
in Table 2, while more details are provided in the following were subjected to axial compression experiments to evaluate the
paragraphs. masonry axial strength according to EN 1052-1 (CEN 1998).
Table 2. Measured material properties Two specimen types were tested, control and reinforced ones,
with the latter receiving two TRM layers on one side. The axial
Num. of Mean Coefficient of
compression experiments yielded a 6.2- and 7.8-MPa compressive
Material Property specimens value variation
strength for the control and reinforced specimen, respectively.
Foundation fc - MPa 3 43.8 2.2% These simple compression experiments were very important for
concrete the later validation of the analytical predictive arching action
Superstructure fc - MPa 3 39.2 5.7%
model and that is why they were preferred over more elaborate
concrete
Rebar Φ8 fy - MPa 3 538 3.3% ones. Compression experiments were also performed on single
(B500C) fu - MPa 642 0.7% bricks normal to their perforations; a 7.1-MPa compressive strength
ɛu > 10% ∼ was obtained.
Rebar Φ5.5 fy - MPa 3 259 4.8% The tensile properties of TRM were obtained by testing coupons
(S220) fu - MPa 336 0.3% in uniaxial tension according to AC434 (ICC Evaluation Services
ɛu > 15% ∼ 2016). In total, four specimens with one textile layer and five
Building mortar ft - MPa 3 2.5 0.8% with two were tested, with the textile being stretched along its
fc - MPa 6 10.7 4.3%
warp direction. The specimens were constructed by casting large
Reinforcing ft - MPa 3 5.4 3.0%
mortar fc - MPa 6 11.2 16.7% tiles (10 mm thick for single layer and 12 mm for double layer con-
Bricks fc - MPa 3 7.1 5.7% figurations) on molds on a flat surface with the textile(s) preposi-
Wallettes fc - MPa 3 6.2 5.6% tioned in place. These tiles were then cut into rectangular
(control) ɛu 0.18% 0.5% coupons of dimensions 100 × 500 mm using a diamond cutting
Wallettes fc - MPa 3 7.8 6.0% saw, and steel plates with holes were glued on their ends using
(reinforced) ɛu 0.10%a 13.8% epoxy resin, to allow for the application of the tensile load. For
TRM – 1L Ft - kN/m 4 58.2 3.4% the measurements of the longitudinal tensile strains, a video exten-
σt,max - 1,044
someter setup was employed. The full testing configuration is de-
MPa
ɛu 2.05% 2.9% picted in Fig. 3. Experimental results showed that single-layer
TRM – 2L Ft - kN/m 5 70.2 3.2% coupons were able to sustain 58.2 kN/m, while double-layer ones
σt,max - 629 35.1 kN/m per layer (70.2 kN/m in total). In other words, by dou-
MPa bling the number of fibers, only a 21% increase in the ultimate load
ɛu 2.22% 11.2% was achieved. This phenomenon is attributed to the fact that: (a)
Note: fc = compressive strength; ft = tensile strength; fy = yield stress; fu = when two layers coexist, they do not share the same deformations
ultimate tensile strength; ɛu = strain at fu; Ft = maximum tensile force per and therefore cannot reach their capacity at the same time; (b) the
running meter; and σt,max = maximum tensile stress. inorganic matrix is not able to redistribute the stress between the
a
Measurements taken outside the thermal insulation layer are not reliable. two layers at high applied loads.

© ASCE 04023054-3 J. Compos. Constr.

J. Compos. Constr., 2023, 27(6): 04023054

Downloaded from ascelibrary.org by "Indian Institute of Technology, Hyderabad" on 09/20/23. Copyright ASCE. For personal use only; all rights reserved.

(a)

Fig. 3. Tensile testing of TRM coupons.

(b)

Experimental Procedure
The primary experimental program included in- and out-of-plane
tests on the infilled RC frames, the configurations of which can
be observed in Fig. 4. During the in-plane experiments, a cyclic
time-history protocol was applied to the RC beam using a horizon-
tal actuator, while at the same time, axial load was exerted on the
RC columns using a second vertical actuator and a distributing
steel beam. The total axial force was set to 135 kN (67.5 kN for
each column), corresponding to 10% of the axial compression
strength of the columns. The time history included 20 cycles, start-
ing from a lateral displacement of 1 mm, all the way up to 20.7 mm. (c)
At each displacement level, two cycles were applied, and each next
displacement level was 40% higher than the previous as per the Fig. 4. (a) In-plane experimental configuration; and (b and c)
FEMA recommendations (FEMA 2007). The maximum lateral dis- out-of-plane experimental configuration.
placement of 20.7 mm corresponds to an interstory drift value of
slightly over 2%, which corresponds to a severe damage state
(Hak et al. 2012). That way, the infills were predamaged in prepa- points around the center of the infill, each one on a 100 ×
ration for their subsequent out-of-plane testing. Before conducting 100-mm area, to delay local failure phenomena at the point of
any experiment, the RC frame was subjected to the same in-plane application. At the same time, the beam and footing of the
time history to precrack it, validate its stable behavior, and reassure frame were both supported out-of-plane, so that the boundary
that it can safely withstand the imposed drifts without being dam- conditions would be as realistic as possible (foundation, fixed;
aged. The lateral displacements were measured at the beam height beam, simply supported), but also to make sure that the measured
from both sides using two high resolution cameras connected to a displacements are due to bending of the masonry and not due to
computer running video extensometer software. These cameras the cantilever-like rotation of the whole frame. The out-of-plane
were tracking in real-time the distance between two marks: the displacements were measured using a mesh of three potentiome-
first was placed at the middle of the beam and the second on the ters and three linear variable displacement transducers (LVDTs).
strong frame’s column. The three LVDTs were placed at the middle of the infill along its
Out-of-plane experiments were performed right after the in- height, while the three potentiometers were positioned at the
plane ones for specimens C_D, R_D, RI_D and directly for edge and also along the height. The LVDTs were chosen to mea-
the rest (C_ND, R_ND, RI_ND). The out-of-plane load was ap- sure the middle deformations, which were more crucial, because
plied using the horizontal actuator and by distributing it to four of their much higher accuracy.

© ASCE 04023054-4 J. Compos. Constr.

J. Compos. Constr., 2023, 27(6): 04023054

 Experimental Results as δ+max,i and |δ−max,i |. Moreover, Pj, Pj−1, δj, and δj−1 denote the
loads and displacements for two consecutive points of the response.
In-Plane Experiments + −
Pmax,i + |Pmax,i |
Ki = + − (1)
In total, three masonry infills were subjected to in-plane experi- δmax,i + |δmax,i |
ments, namely specimens C_D, R_D, and RI_D. The results are
given in terms of force–displacement hysteretic loops, lateral stiff-
Ej = E j−1 + 0.5(Pj + P j−1 )(δj − δ j−1 ) (2)
ness, and cumulative hysteretic energy in Fig. 5. Fig. 6 depicts the
envelope curves for all these three specimens. The lateral stiffness A stable hysteretic behavior was recorded for all three speci-
for cycle i is computed according to Eq. (1), while the cumulative mens, without any significant load drop up to the imposed lateral
+
Downloaded from ascelibrary.org by "Indian Institute of Technology, Hyderabad" on 09/20/23. Copyright ASCE. For personal use only; all rights reserved.

dissipated energy according to Eq. (2). In these equations, Pmax,i drift of 2%. TRM strengthening proved to be capable of signifi-
−
and |Pmax,i | are the maximum lateral loads by absolute value in cantly increasing the lateral strength of the RC frame, as specimen
both directions, while the respective displacements are denoted R_D exhibited a 37% higher lateral load, 13% higher initial

(a) (b)

(c) (d)

(e) (f )

Fig. 5. Experimental results from in-plane testing of specimens: (a and b) C_D; (c and d) R_D; and (e and f) RI_D.

© ASCE 04023054-5 J. Compos. Constr.

J. Compos. Constr., 2023, 27(6): 04023054

Downloaded from ascelibrary.org by "Indian Institute of Technology, Hyderabad" on 09/20/23. Copyright ASCE. For personal use only; all rights reserved.

(a) (b)

Fig. 6. (a) Envelope curves; and (b) stiffness and energy dissipation characteristics.

stiffness, and 21% more total dissipated energy, compared to the application of the strengthening scheme reduces the effect of in-
control one. However, specimen RI_D had a considerably different plane damage on the out-of-plane capacity of an infill.
response. In this case, only a 10% increase was recorded for the lat- As far as the absolute response values are concerned, the un-
eral strength, the initial stiffness was 12% lower, and the total dis- damaged reinforced specimens (R_ND) reached practically the
sipated energy was practically the same. These results suggest that same strength as the respective control specimen (C_ND), while
the outermost TRM layer was not adequately activated, something in the case of damaged ones (R_D versus C_D), there was a 65%
that was evident during the experiment, as debonding was observed strength increase. Specimens with both TRM strengthening and
at the insulation–masonry interface. thermal insulation outperformed the control ones by 29% and
81% in the undamaged and damaged case, respectively. The
slightly better performance of the RI specimens is attributed to
Out-of-Plane Experiments the increased lever arm of the textile reinforcement, which results
All six infills, with and without damage, were subjected to mono- in its activation at lower displacement levels. The initial stiffness
tonic out-of-plane testing up to their failure. The results of these ex- of the infills dropped substantially, by a factor of 8–12, due to
periments are outlined in Fig. 7 in terms of out-of-plane load versus the existence of in-plane damage. This was expected though, as
mid-displacement curves and maximum load. by the end of the in-plane tests, the masonry-frame interface had
As it can be observed in Fig. 7, all specimens resisted quite high sustained a significant amount of damage, thus leading to a reduced
out-of-plane loads, thanks to the formation of a horizontal arching activation of the arching mechanism. It is also important to note
mechanism. This mechanism was evidenced by the crack distribu- that for both reinforced specimen types (R and RI), localized de-
tion as it was observed at the end of the experiments and is depicted bonding was observed at the frame–reinforcement interface by
in Fig. 8 for the case of the C_ND specimen; similar cracking pat- the end of the in-plane experiments. Because of this phenomenon,
terns were observed for all the specimens which were tested the efficiency of the reinforcement layer was reduced during the
out-of-plane. Moreover, the existence of in-plane damage can sig- subsequent out-of-plane experiments.
nificantly reduce the out-of-plane capacity of a masonry infill. This The failure mode of all specimens was local crushing of bricks
reduction is 53% for the control specimens (C_ND and C_D), 23% at the four points of the load application, something undesirable but
for the reinforced (R_ND and R_D), and 34% for the reinforced also unavoidable with the employed experimental configuration.
and insulated ones (RI_ND and RI_D). In other words, the This means that the recorded maximum loads do not represent

(a) (b)

Fig. 7. Comparative results of out-of-plane experiments: (a) force–displacement curves; and (b) out-of-plane strength results.

© ASCE 04023054-6 J. Compos. Constr.

J. Compos. Constr., 2023, 27(6): 04023054

 Analytical Modeling

Following the completion of the experimental campaign, an analyt-

ical modeling approach was developed for the prediction of the
out-of-plane arching response of masonry infills, accounting for
the possible existence of prior in-plane damage and/or tensile rein-
forcement. The procedure starts by assuming a predefined response
mechanism, compatible with the geometry and the boundary con-
ditions, and the corresponding set of fracture lines on the infill sur-
face. Then, arching is assumed to take place via the rigid-body
Downloaded from ascelibrary.org by "Indian Institute of Technology, Hyderabad" on 09/20/23. Copyright ASCE. For personal use only; all rights reserved.

rotation of the formed panel segments around their boundaries

and up to the adopted fracture lines, thus fully defining the displace-
ment field. The next step consists of dividing the panel into hori-
zontal and vertical arching strips, similarly to the procedure
followed by Dawe and Seah (1989), which are assumed to be
(a) able to transfer only axial loads, also known as thrusts. These
loads travel along the body of the strip segment at an oblique
angle and therefore can be transformed into a triplet of axial,
shear, and moment acting on its centroid, as depicted in Fig. 9. If
this internal state can be estimated, then the internal work can be
computed for every strip and consequently for the whole infill.
Then, the external virtual work may also be estimated easily, for
the given external loads and displacement field, and finally, by
equating the internal and external virtual works, the response of
the infill can be determined.

Formulation
For the derivation of the internal state of every infill strip, a proce-
dure similar to that described by Angel (1994) is followed. It is as-
sumed that the boundaries are immovable, the compression zones
(b) have the same depth both at the supports and at the intersections
with the fracture lines, and that the axial strains vary linearly within
Fig. 8. (a) Crack distribution; and (b) approximate horizontal arching these compression zones as well as along the outer fibers of the in-
mechanism (C_ND specimen). fill strip (Fig. 10). Then, the horizontal force that acts on it per unit
width can easily be evaluated by integrating the normal stresses
which act on the contact area:
the actual out-of-plane strengths of the masonry infills, which
would be achieved if this failure mode could be avoided (e.g., by dis-
tributing the load uniformly over the entire surface). This justifies the c θ0
fact that specimen R_ND reached the same maximum load as speci- N = σ avg ⇒ N = σ avg c (3)
cos θ
men C_ND, as the local failure occurred before the full activation of
the arching mechanism and the reinforcing textile. Despite this
limitation of the experimental configuration, it was still possible The axial shortening of the extreme fibers can be estimated
to extract numerous valuable conclusions, as outlined previously. based on the assumption of the linear distribution of strains. If

Fig. 9. Division of infill panel in strips and internal forces in a strip segment.

© ASCE 04023054-7 J. Compos. Constr.

J. Compos. Constr., 2023, 27(6): 04023054

 Δc Δc
tan θ = ⇒ θ = tan−1 (9)
c c
Therefore, based on the above procedure and set of assump-
tions, given the maximum strain at its outermost fibers, the internal
state of a rotating infill segment can be fully tracked. Next, for the
calculation of the internal work, assuming that only the axial loads
are contributing, it can be stated that for every rotating segment of
width Δw, the internal work ΔWint is
⎫
Mtot = M + M ′ − Nd ⎬ t − 2δG c
− δ ⇒ Mtot = N −d
Downloaded from ascelibrary.org by "Indian Institute of Technology, Hyderabad" on 09/20/23. Copyright ASCE. For personal use only; all rights reserved.

t/2 c
M = M′ = N
G
⎭ cos θ
cos θ
≈ N (t − 2δG c − d) (10)

ΔWint = Mtot θΔw = [N (t − 2δG c − d)]θΔw (11)

In Eqs. (10) and (11), Mtot is the total equivalent moment acting
on the infill segment and δGc is the distance of the application point
of the thrust force from the most stressed fiber. For the total internal
work Wint, the individual works are summed for all the rotating in-
fill segments (two for each strip), in both directions (for two-way
arching). As a result, it can be stated that
nx ny
Wint = ΔWint,ix + ΔWint,iy (12)
ix =1 iy =1

The calculation of the external work Wext is a far simpler pro-

cess. If di is the out-of-plane displacement at the application
point of the force Fi, L and H are the infill dimensions (assumed
rectangular), q is the uniform area load, and davg is the average
Fig. 10. Assumptions for the internal state of an infill strip segment. out-of-plane displacement over the whole infill area, then for a
set of concentrated forces or a uniform area load respectively, the
external work is
⎧ n
the maximum strain is denoted as εmax , then this shortening will be ⎨F d for n concentrated loads
i i

l
l  Wext = i=1 (13)
εmax  εmax l ⎩
qLHdavg for uniform load
Δc = ε(x)dx = x dx = (4)
0 0 l 2
Since the total internal and external works must be equal, it fol-
Then, for the calculation of the compressive zone width, we can lows that the following relation must hold:
use the following geometric relations based on Fig. 10: Wint = Wext (14)
l l 1 − cos θ The above analysis can be employed using the following algo-
l + 2δ = ⇒δ= (5)
cos θ 2 cos θ rithm to predict the out-of-plane response of a given masonry infill
in the form of a load–displacement curve.
1. Assume a reasonable arching mechanism, compatible with the
δ cos θ Eq. (5) l 1 − cos θ
a= =δ ⇒ a= (6) boundary conditions, and divide the infill into horizontal and/
tan θ sin θ 2 sin θ or vertical strips. The selection can be made on the following
grounds:
t
⎫
t l 1 − cos θ ⎪ • for long infills with aspect ratio (length/height) above 2, ver-
c= −a= − ⎪
2 2 sin θ ⎪ ⎪ tical (one-way) arching action can safely be assumed, pro-
2 ⎪
⎪ 
Δc ⎪
⎪ vided there are no gaps along the top edge;
sin θ =  ⎬ t l Δc + c − c
2 2
• for infills constructed in contact with the surrounding frame,
Δc + c
2 2 ⇒ c = − (7)
⎪
⎪ 2 2 Δc the more common horizontal-arching action can generally be
⎪
⎪
c ⎪
⎪
adopted. If the effect of shrinkage is negligible, then the full,
cos θ =   ⎪
⎪
⎭ two-way arching action may be assumed;
Δ2c + c2 • if there are any large, visible gaps above 10 mm along an
edge which can compromise the activation of axial thrusts,
Eq. (7) can be solved for the compressive zone using any itera- then no arching action may be assumed perpendicular to
tive numerical procedure. Then, the vertical displacement d sus- that edge.
tained by the rotating infill strip and the rotation angle θ can be 2. Assume a material model for the masonry.
computed from the following equations: 3. Assume a specific level of the lateral out-of-plane displacement
at the position where it is maximum (usually at the center);
d = l sin θ (8) this displacement is denoted as dmax. With this assumption,

© ASCE 04023054-8 J. Compos. Constr.

J. Compos. Constr., 2023, 27(6): 04023054

 the displacements along all the fracture lines are fully defined. masonry are appropriately configured. Τhe addition of TRM in-
For the preset maximum displacement level and for every infill creases the axial strength of masonry, therefore the out-of-plane re-
strip rotating segment: sponse of a retrofitted infill would also be indirectly affected by the
• assume a maximum axial strain εmax for the outermost com- adoption of an improved material model. That means that the TRM
pressed fiber; is practically not utilized as tensile reinforcement, and this assump-
• for that specific εmax , calculate the quantities Δc, c, θ, and fi- tion is reasonable because:
nally the respective displacement d. If d is not equal to the • the TRM reinforcement in most practical applications will be
displacement that corresponds to the assumption made in single sided (as in the performed experiments). Therefore, in
Step 2, repeat this step for a different value of εmax until con- the case of cyclic loads (e.g., earthquakes) or loads that may re-
vergence is reached. sult in tension to the unretrofitted side (e.g., wind loads), only its
Downloaded from ascelibrary.org by "Indian Institute of Technology, Hyderabad" on 09/20/23. Copyright ASCE. For personal use only; all rights reserved.

4. Calculate the internal work for every rotating strip. Sum for all indirect influence can be considered;
the strips to calculate the total internal work. • full anchoring of the reinforcement is almost impossible to
5. Calculate the intensity of the external forces that is needed to achieve and debonding is very likely to occur, especially around
achieve a total external work equal to the internal one. the infill perimeter, thus reducing the efficiency of the TRM as
At the end of the last step, both the total external load Fext and tensile reinforcement.
the maximum out-of-plane displacement dmax will be known. The In the lack of experimental evidence, a simple approach would
process can then return back to Step 2 and be repeated for a higher be to calculate the compressive strength of the retrofitted masonry
level of maximum lateral displacement, until the whole response is by accounting for the contribution of the strengthening mortar layer
plotted. This procedure is also outlined in detail in Fig. 11. on the compressive characteristics of the section. This makes the
safe-sided assumption that no compressive strength increase is
caused by the existence of the textile, something that is reasonable
Effect of TRM Reinforcement given the fact that the textile does not fully wrap the loaded ele-
ment, thus resulting in a reduced confinement effect. Considering
If the analyzed masonry infill is reinforced with TRM, then the pro- that both the masonry and the strengthening mortar have similar ul-
cedure is still valid, as long as the material properties of the timate strains, the following relation for the compressive strength of
the retrofitted masonry walls can be extracted:
tTRM fc,TRM + tm fm
fm,r ≈ (15)
tm
where tm and tTRM are the thicknesses of the masonry and the TRM
layer, respectively, and fm and fc,TRM are the respective compressive
strengths. The predicted strength of the retrofitted masonry fm,r is
calculated based on its initial thickness.

Effect of Damage Due to In-Plane Loading

A simple yet effective way to consider the effect of in-plane dam-
age would again be through an appropriate modification of the
stress–strain material model for the masonry. The modification pro-
posed herein is based on the introduction of a damage parameter D,
which ranges between 0 and 1 (1 represents the fully damaged
state) and softens the response through the following formula:
σ D (ε) = (1 − D)E0 ε ≤ σ(ε) (16)
where E0 is the initial, undamaged stiffness of the material, and σ(ε)
represents its complete undamaged response. The modification
does not need to be applied for the entirety of the infill, but for
those only portions of it which have actually suffered damage.
Therefore, after discretizing the infill into horizontal and vertical
strips, we could assign a different material model to each one of
them, depending on their damage state. Undamaged strips should
be evaluated normally, while a reduced material model could be
adopted for the damaged ones. Moreover, since each strip is han-
dled independently, the transition from the undamaged to the dam-
aged strips can even be gradual, as depicted for example in Fig. 12.
In the context of this work, a smooth damage distribution is
adopted, which is maximum at the four corners and minimum
around the infill center using Eq. (17). This was chosen based on
the observation of in-plane damage from the respective experi-
ments, during which a corner crushing mechanism was activated;
for other response mechanisms (El-Dakhakhni et al. 2003), differ-
ent damage distributions should be adopted. The Dmin and Dmax
quantities should be correlated to some engineering property,
Fig. 11. Flow chart of the analytical process.
which is known to be directly related to the damage, like for

© ASCE 04023054-9 J. Compos. Constr.

J. Compos. Constr., 2023, 27(6): 04023054

 recorded loads are listed in Table 3. In this table, one can see that
the analytical estimates have an error that ranges from −11% to
+15% for the undamaged specimens and from −18% to +18%
for the damaged ones.
The prediction of the response of the undamaged specimens was
reasonably good for all the three specimen configurations, both in
terms of initial stiffness and maximum strength. For the initial part
of the response, the analytical predictions were almost coincident
with the experimental curves, but they started to drift apart slightly
at the points where the experimental ones manifest a sudden stiff-
Downloaded from ascelibrary.org by "Indian Institute of Technology, Hyderabad" on 09/20/23. Copyright ASCE. For personal use only; all rights reserved.

ness drop. This change is attributed to the initiation of sliding that

takes place at the masonry joints, both the inner and the perimetric
Fig. 12. Damage distribution over horizontal stripes due to in-plane ones, with respect to the RC frame. Since the analytical model does
loading. not account for this phenomenon, it is reasonable that this differ-
ence appears. Nevertheless, this shortcoming did not seem to com-
promise the maximum load predictions (error from −11% to
example the maximum in-plane drift. The determination of this re- +15%), which were very close to the experimentally obtained
lation was out of the scope of the present study, and for simplicity, values.
the limiting values Dmin = 0 and Dmax = 1 were chosen instead, as However, the prediction of the response of the damaged speci-
very high drifts were imposed during the in-plane experiments. mens was not as successful in terms of their initial stiffness. The
analytical model systematically overestimated the out-of-plane
D(dn ) = Dmin + (Dmax − Dmin ) cos2 (πdn ) stiffness of the infill configurations by a factor of more than
0 ≤ Dmin ≤ Dmax ≤ 1 three. This increased stiffness inaccuracy is once again attributed
⎧ to sliding phenomena, which were now greatly magnified due to
⎪ d (17)
⎨ for horizontal strips the accumulation of damage after imposing high in-plane drifts
dn = H of above 2%. At the beginning of the out-of-plane experiments,
⎪
⎩ d for vertical strips the masonry joints were already cracked, therefore sliding started
L taking place right from the start and at a considerably higher
grade. Despite all these observations, maximum strength predic-
tions were only slightly worse than those of the undamaged ones
Application on Experimental Data (±18% error). Therefore, for strength predictions, the proposed
Based on experimental observations, the primarily horizontal arch- method is considered adequate also for the case of damaged infills.
ing mechanism depicted in Fig. 8(b) was adopted, according to It is important to note here once again that during the
which there are three rotating segments contributing to the total re- out-of-plane testing of infills, failure was always due to local crush-
sponse. The infill was divided into 50 horizontal and 62 vertical ing, which would occur at the load application points. Therefore,
stripes, and a simple, bilinear material model was assumed for the actual out-of-plane strengths were not reached in any case. It
the masonry, comprising an initial elastic branch up to the maxi- is argued though that they would probably not be considerably
mum strength followed by a subsequent linear softening one. The higher than the reported values, since at the point when the speci-
maximum strength was obtained from the material characterization mens failed, their response was close to reaching a plateau. In any
experiments and it was set to occur at a compressive strain of 0.2%, case, since the above comparisons were made based on deflated ex-
while the softening branch was assumed to reach zero stress at a perimental strengths, any further deviation of the proposed model
strain of 0.25%. The 0.2% value was adopted according to the rec- would be safe sided.
ommendations of Eurocode 6 (EN 1996; CEN 2005) and is in
agreement with the material characterization results for the unrein-
Parametric Studies
forced masonry (0.18% strain @ maximum stress). However, the
0.25% limiting value which governs the slope of the softening After the encouraging results from the application of the proposed
branch was selected after performing an initial fitting of the analytical model on the experimentally obtained data, the model
model using the available experimental data. The goal was to was used to run a set of parametric studies. For these case studies,
have a steep softening branch, thus simulating the brittle nature the same bilinear material model was adopted, and various values
of the material, without compromising the numerical stability of of the infill slenderness (H/t) and aspect ratio (L/H ) were assumed.
the iterative procedure. All the results presented hereafter are given normalized; the
For the reinforced, and the reinforced and insulated specimens, out-of-plane pressure is given as a fraction of the masonry strength
the same stress–strain relation was adopted. The only difference be- (q/fm), while the respective displacements are normalized by the in-
tween the control (C) and the reinforced (R, RI) infills was the value fill height (d/H ).
of the compressive strength (again obtained from the material char- In total, more than 100 simulations were run covering a plethora
acterization tests), simulating that way the existence of the TRM re- of values of infill aspect and slenderness ratios for the cases of ver-
inforcement in a simplistic way. The entire process was automated tical, two-way, and horizontal arching. These revealed that the most
to be able to perform a large number of simulations in a fast and important geometric property of an arching masonry infill is by far
reliable manner. its slenderness, as revealed in Fig. 14(a) for the case of vertical
The experimentally obtained and analytically calculated arching; decreasing the slenderness ratio from 20 to 15 (33%
out-of-plane response for all the tested specimens in terms of thicker masonry) results in more than twofold increase of the
total force versus midpoint displacement are provided in Fig. 13. out-of-plane capacity and initial stiffness. A comparatively less im-
On the left side, the diagrams of the undamaged specimens are pre- portant but still significant geometric factor seems to be the aspect
sented and on the right, those of the damaged ones. The maximum ratio, which is plotted in Fig. 14(b) for the case of two-way arching.

© ASCE 04023054-10 J. Compos. Constr.

J. Compos. Constr., 2023, 27(6): 04023054

 Table 3. Experimentally and analytically determined out-of-plane As depicted in the figure, its variation between the two limiting val-
strengths ues of 1 and ∞ results in a 50% lower expected out-of-plane
Out-of-plane strength (kN) strength. Moreover, it is interesting to observe that the predicted
vertical arching response of an infill is identical to that predicted
Specimen ID Experiment Analysis for a two-way arching infill with the same slenderness and infinite
C_ND 64.1 58.6 (−9%) aspect ratio. As far as the effect of damage is concerned, the ex-
C_D 30.1 35.5 (+18%) pected drop of out-of-plane capacity for damaged infills was
R_ND 64.3 73.7a (+15%) found to range from 30% to 50% for two-way and vertical arching,
R_D 49.7 44.7a (−10%) and from 37% to 43% for horizontal arching. Therefore, if the pre-
RI_ND 83.0 73.7a (−11%) sent model is adopted, a simple and safe way to quickly estimate
RI_D 54.5 44.7a (−18%)
Downloaded from ascelibrary.org by "Indian Institute of Technology, Hyderabad" on 09/20/23. Copyright ASCE. For personal use only; all rights reserved.

the effect of in-plane damage on the arching response is to simply

a
The analytical predictions for RI and R specimens are the same. reduce the predicted capacity by 50%.

(a) (b)

(c) (d)

(e) (f )

Fig. 13. Comparison between the experimental and analytical response of specimens: (a) C_ND; (b) C_D; (c) R_ND; (d) R_D; (e) RI_ND; and
(f) RI_D.

© ASCE 04023054-11 J. Compos. Constr.

J. Compos. Constr., 2023, 27(6): 04023054

 either continuous infilled frames or sufficiently stiff frame
elements;
• a reasonable arching mechanism must be assumed, respecting
the boundary conditions;
• the slenderness ratio H/t must not be too high (values up to 25
are permitted);
• the aspect ratio L/H must be at least 1 for two-way arching and
below 2 for horizontal arching.
Downloaded from ascelibrary.org by "Indian Institute of Technology, Hyderabad" on 09/20/23. Copyright ASCE. For personal use only; all rights reserved.

Conclusions

The current paper presented the results of an experimental program,

(a) which included in- and out-of-plane experiments on medium scale,
masonry infills built within RC frames, and retrofitted with TRM
and thermal insulation. The main goal of the research was to inves-
tigate the effect of in-plane damage on the out-of-plane response of
such elements and to test the proposed retrofitting scheme under
real-life conditions. Given the reduced scale of the experiments,
the presented results are indicative of the behavior from a qualita-
tive point of view, and not necessarily representative in terms of ab-
solute magnitudes. For that matter, a further validation is needed
against a more extended experimental campaign.
The TRM strengthening method can significantly increase the
in-plane capacity of RC frames and transform them from secondary
elements to reliable load resisting members. The addition of ther-
mal insulation between successive TRM layers seems to render
the outer ones inactive, as observed during the experimental testing.
Concerning the out-of-plane behavior, it was observed that when
(b) TRM layers are applied with higher lever arm (e.g., above the ther-
Fig. 14. Effect of infill: (a) slenderness ratio (vertical arching); and mal insulation), they are activated sooner, thus leading to slightly
(b) aspect ratio (two-way arching with H/t = 15). increased capacities. However, since this benefit is completely
lost in the case of damaged infills and the out-of-plane infill capac-
ities are already high enough, there is no apparent reason to sacri-
The analytically computed predictions of the above case studies fice their in-plane behavior.
were statistically analyzed and led to the development of the two Following the presentation of the experimental results, an ana-
simple empirical equations suggested in the following. These equa- lytical procedure was developed for the prediction of the
tions are composed of two terms: the first one within the brackets out-of-plane response of masonry infills. This method can produce
controls the effect of the aspect ratio, while the second one controls good response predictions with minimal input parameters, while
the more prominent effect of the infill slenderness. They both di- also accounting for the existence of TRM reinforcement and
rectly predict the out-of-plane strength of a TRM-retrofitted ma- prior in-plane damage in a simple manner. The proposed model
sonry infill for the cases of two-way and horizontal arching. For was applied on the experimental data and yielded reasonably
the case of vertical arching, the first equation may also be used good strength estimates for both the undamaged and the damaged
after setting the aspect ratio value as infinite, thus transforming specimens, with an error ranging from −11% to +15 for the undam-
the first term of the equation to the plain numeric value of 1.94. aged and from −18% to 18% for the damaged specimens. After its
validation, several parametric studies were conducted to investigate
⎧ −4

−2.2 how various properties can affect the out-of-plane behavior of in-
⎪
⎪ L H
⎪ 1.6
⎪ + 1.94 two − way arching fills, including the response mechanism, slenderness, aspect ratio,
q ⎨ H t
=   and the existence of prior in-plane damage. Using the results of
fm ⎪
⎪ L −2.5
H −2.2 the method, generic response diagrams were produced and simple
⎪
⎪ + 0.5
⎩ 1.5 horizontal arching equations were proposed for the rapid evaluation of the
H t
out-of-plane capacity of an infill for the most typical configura-
(18) tions. As a final remark, it has to be stressed though that further val-
idation is needed against more experimental data and particularly
Eq. (18) applies to undamaged masonry infills only. If damage for the case of slenderer infills than those presented here.
needs to be accounted for, then, as explained in the previous para- The proposed retrofitting scheme can successfully be imple-
graphs, a simple solution would be to reduce the predicted capacity mented on existing RC buildings with masonry infills and that
by 50%. Apart from this, the use of the above equations is subject to way improve their global response as well as eliminate the possibil-
the following limitations due to the model assumptions as well as ity of out-of-plane failures. Insulation layers are suggested to be
the domain of the data used to perform the statistical regressions: placed externally, above any strengthening ones, as any benefits
• the infill boundaries must be sufficiently fixed, so that the full that may arise from the increased lever arm are counteracted by
arching mechanism assumed can actually form. This premises the reduced in-plane efficiency of the reinforcement.

© ASCE 04023054-12 J. Compos. Constr.

J. Compos. Constr., 2023, 27(6): 04023054

 Data Availability Statement 129 (2): 177–185. https://doi.org/ 10.1061/(ASCE)0733-9445(2003)
129:2(177).
The data from the in- and out-of-plane experiments of this study are Fardis, M. N. 2000. “Design provisions for masonry-infilled RC frames.”
available from the corresponding author upon reasonable request. In Proc., 12th World Conf. on Earthquake Engineering. Upper Hutt,
New Zealand: New Zealand Society for Earthquake Engineering.
Fardis, M. N., and T. B. Panagiotakos. 1997. “Seismic design and response
of bare and masonry infilled reinforced concrete buildings, Part II:
Acknowledgments Infilled structures.” J. Earthquake Eng. 1 (3): 475–503.
FEMA (Federal Emergency Management Agency). 2007. Interim testing
protocols for determining the seismic performance characteristics of
The research work was partially supported by the Hellenic Founda-
structural and nonstructural components. FEMA 461. Washington,
tion for Research and Innovation (HFRI) under the HFRI PhD Fel-
Downloaded from ascelibrary.org by "Indian Institute of Technology, Hyderabad" on 09/20/23. Copyright ASCE. For personal use only; all rights reserved.

DC: FEMA.
lowship grant (Fellowship Number: 62) and under the 1st Call for Gkournelos, P. D., D. A. Bournas, and T. C. Triantafillou. 2019.
Research Projects to support Faculty members and Researchers and “Combined seismic and energy upgrading of existing reinforced con-
the procurement of high-cost research equipment (Grant Number: crete buildings using TRM jacketing and thermal insulation.”
1962). Earthquakes Struct. 16 (5): 625–639.
Hak, S., P. Morandi, G. Magenes, and T. J. Sullivan. 2012. “Damage con-
trol for clay masonry infills in the design of RC frame structures.”
J. Earthquake Eng. 16: 1–35. https://doi.org/10.1080/13632469.2012
References .670575.
ICC Evaluation Services. 2016. Acceptance criteria for masonry and con-
Acun, B., and H. Sucuoglu. 2006. “Strengthening of masonry infill walls in crete strengthening using fabric-reinforced cementitious matrix
reinforced concrete frames with wire mesh reinforcement.” In Proc., 8th (FRCM) and Steel Reinforced Grout (SRG) composite systems.
U.S. National Conf. on Earthquake Engineering. San Francisco, CA: AC434. Los Angeles: ICC Evaluation Services.
Earthquake Engineering Research Institute (EERI). Koutas, L., S. N. Bousias, and T. C. Triantafillou. 2015a. “Seismic
Akhoundi, F., G. Vasconcelos, P. Lourenco, L. M. Silva, F. Cunha, and R. strengthening of masonry-infilled RC frames with TRM:
Fangueiro. 2018. “In-plane behavior of cavity masonry infills and Experimental study.” J. Compos. Constr. 19 (2): 04014048. https://
strengthening with textile reinforced mortar.” Eng. Struct. 156: 145– doi.org/10.1061/(ASCE)CC.1943-5614.0000507.
160. https://doi.org/10.1016/j.engstruct.2017.11.002. Koutas, L., A. Pitytzogia, T. C. Triantafillou, and S. N. Bousias. 2014.
Almusallam, T. H., and Y. A. Al-Salloum. 2007. “Behavior of FRP “Strengthening of infilled reinforced concrete frames with TRM:
strengthened infill walls under in-plane seismic loading.” J. Compos. Study on the development and testing of textile-based anchors.”
Constr. 11 (3): 308–318. https://doi.org/10.1061/(ASCE)1090 J. Compos. Constr. 18 (3): A4013015. https://doi.org/10.1061
-0268(2007)11:3(308). /(ASCE)CC.1943-5614.0000390.
Altin, S., Ö. Anil, E. M. Kara, and M. Kaya. 2008. “An experimental study Koutas, L., T. C. Triantafillou, and S. N. Bousias. 2015b. “Analytical mod-
on strengthening of masonry infilled RC frames using diagonal CFRP eling of masonry-infilled RC frames retrofitted with textile-reinforced
strips.” Composites, Part B 39 (4): 680–693. https://doi.org/10.1016/j mortar.” J. Compos. Constr. 19 (5): 04014082. https://doi.org/10
.compositesb.2007.06.001. .1061/(ASCE)CC.1943-5614.0000553.
Angel, R. 1994. “Behavior of reinforced concrete frames with masonry in- Koutas, L. N., and D. A. Bournas. 2019. “Out-of-plane strengthening of
fills.” Ph.D. thesis, Dept. of Civil Engineering, Univ. of Illinois. masonry-infilled RC frames with textile-reinforced mortar jackets.”
Bournas, D. A. 2018. “Concurrent seismic and energy retrofitting of RC J. Compos. Constr. 23 (1): 04018079. https://doi.org/10.1061/(ASCE)
and masonry building envelopes using inorganic textile-based compos- CC.1943-5614.0000911.
ites combined with insulation materials: A new concept.” Composites, Kyriakides, M. A., and S. L. Billington. 2008. “Seismic retrofit of
Part B 148: 166–179. https://doi.org/10.1016/j.compositesb.2018.04 masonry-infilled non-ductile reinforced concrete frames using spray-
.002. able ductile fiber-reinforced cementitious composites.” In Proc., 14th
CEN (European Committee for Standardization). 1998. Methods of test for World Conf. on Earthquake Engineering, 1–7. Beijing, China: WCEE.
masonry—Part 1: Determination of compressive strength. EN 1052-1. Longo, F., A. Cascardi, P. Lassandro, and M. A. Aiello. 2020. “A new fab-
Brussels, Belgium: CEN. ric reinforced geopolymer mortar (FRGM) with mechanical and energy
CEN (European Committee for Standardization). 2005. Design of masonry benefits.” Fibers 8 (8): 49. https://doi.org/ 10.3390/fib8080049.
structures—Part 1-1: General rules for reinforced and unreinforced Longo, F., A. Cascardi, P. Lassandro, and M. A. Aiello. 2021. “Energy
masonry structures. EN 1996. Eurocode 6. Brussels, Belgium: CEN. and seismic drawbacks of masonry: A unified retrofitting solution.”
CEN (European Committee for Standardization). 2006. Methods of test for J. Build. Pathol. Rehabil. 6: 31. https://doi.org/ 10.1007/s41024-021
mortar for masonry—Part 11: Determination of flexural and compres- -00121-6.
sive strength of hardened mortar. EN 1015-11. Brussels, Belgium: Mehrabi, A. B., P. B. Shing, M. P. Schuller, and J. L. Noland. 1996.
CEN. “Experimental evaluation of masonry-infilled RC frames.” J. Struct.
CEN (European Committee for Standardization). 2019a. Steel for the rein- Eng. 122 (3): 228–237. https://doi.org/10.1061/(ASCE)0733
forcement and prestressing of concrete—Test methods—Part 1: -9445(1996)122:3(228).
Reinforcing bars, rods and wire. EN ISO 15630-1. Brussels, Ozcebe, G., U. Ersoy, T. Tankut, E. Erduran, O. Keskin, and C. Mertol.
Belgium: CEN. 2003. Strengthening of brick-infilled RC frames with CFRP.
CEN (European Committee for Standardization). 2019b. Testing hardened TUBITAK Structural Engineering Research Unit Rep. No. 2003-1.
concrete—Part 3: Compressive strength of test specimens. EN 12390-3. Ankara, Turkey: Middle East Technical University.
Brussels, Belgium: CEN. Ozden, S., U. Akguzel, and T. Ozturan. 2011. “Seismic strengthening of
Dawe, J. L., and C. K. Seah. 1989. “Out-of-plane resistance of concrete ma- infilled reinforced concrete frames with composite materials.” ACI
sonry infilled panels.” Can. J. Civ. Eng. 16: 854–864. https://doi.org/10 Struct. J. 108 (4): 414–422.
.1139/l89-128. Pohoryles, D. A., D. A. Bournas, F. Da Porto, A. Caprino, G. Santarsiero,
Dehghani, A., F. Nateghi-Alahi, and G. Fischer. 2015. “Engineered cemen- and T. Triantafillou. 2022. “Integrated seismic and energy retrofitting of
titious composites for strengthening masonry infilled reinforced con- existing buildings: A state-of-the-art review.” J. Build. Eng. 61:
crete frames.” Eng. Struct. 105: 197–208. https://doi.org/10.1016/j 105274. https://doi.org/10.1016/j.jobe.2022.105274.
.engstruct.2015.10.013. Saatcioglu, M., F. Serrato, and S. Foo. 2005. “Seismic performance of ma-
El-Dakhakhni, W. W., M. Elgaaly, and A. A. Hamid. 2003. “Three-strut sonry infill walls retrofitted with CFRP sheets.” In Proc., 7th Int. Symp.
model for concrete masonry-infilled steel frames.” J. Struct. Eng. on Fiber-Reinforced Polymer (FRP) Reinforcement for Concrete

© ASCE 04023054-13 J. Compos. Constr.

J. Compos. Constr., 2023, 27(6): 04023054

 Structures, Paper 20, edited by C. Shield, J. Busel, S. Walkup, and D. walls using textile reinforced mortars combined with thermal in-
Gremel, 341–354. Farmington Hills, MI: American Concrete Institute. sulation: In-plane mechanical behavior.” J. Compos. Constr.
Sagar, L. S., V. Singhal, and D. C. Rai. 2019. “In-Plane and out-of-plane 22 (5): 04018029. https://doi.org/10.1061/(ASCE)CC.1943-5614
behavior of masonry-infilled RC frames strengthened with .0000869.
fabric-reinforced cementitious matrix.” J. Compos. Constr. 23 (1): Triantafillou, T. C., K. Karlos, K. Kefalou, and E. Argyropoulou. 2017.
04018073. https://doi.org/10.1061/(ASCE)CC.1943-5614.0000905. “An innovative structural and energy retrofitting system for URM
Triantafillou, T. C. 1998. “Strengthening of masonry structures using walls using textile reinforced mortars combined with thermal insulation:
epoxy-bonded FRP laminates.” J. Compos. Constr. 2 (2): 96–104. Mechanical and fire behavior.” Constr. Build. Mater. 133: 1–13. https://
https://doi.org/10.1061/(ASCE)1090-0268(1998)2:2(96). doi.org/10.1016/j.conbuildmat.2016.12.032.
Triantafillou, T. C. 2001. “Seismic retrofitting of structures with Yuksel, E., A. Ilki, G. Erol, C. Demir, and H. F. Karadogan. 2006. “Seismic
fibre-reinforced polymers.” Prog. Struct. Mater. Eng. 3 (1): 57–65. retrofitting of infilled reinforced concrete frames with CFRP compos-
Downloaded from ascelibrary.org by "Indian Institute of Technology, Hyderabad" on 09/20/23. Copyright ASCE. For personal use only; all rights reserved.

https://doi.org/10.1002/pse.61. ites.” In Advances in earthquake engineering for urban risk reduction,

Triantafillou, T. C., K. Karlos, P. Kapsalis, and L. Georgiou. 2018. edited by T. Wasti, and G. Ozcebe, 285–300. Dordrecht, Netherlands:
“Innovative structural and energy retrofitting system for masonry Springer.

© ASCE 04023054-14 J. Compos. Constr.

J. Compos. Constr., 2023, 27(6): 04023054