Seismic Testing and Modeling of Full-Scale Substandard

RC Columns Retrofitted with Sprayed GFRM with

and without Basalt Mesh under High Axial
Compression and Shear Demand
Downloaded from ascelibrary.org by "Indian Institute of Technology, Hyderabad" on 04/28/25. Copyright ASCE. For personal use only; all rights reserved.

Nima Kian, S.M.ASCE1; Ugur Demir2; Ali Osman Ates3; Oguz C. Celik, A.M.ASCE4;
and Alper Ilki, Aff.M.ASCE5

Abstract: This study presents the experimental and analytical hysteretic behaviors of eight full-scale RC square and rectangular columns.
The columns were designed to have different shear spans that represent: (1) a column that complies with the Turkish Government Ministry of
Reconstruction and Resettlement’s (1975) seismic design code, Turkish Seismic Design Code (TSDC); (2) a substandard column; and (3) two
sprayed glass fiber–reinforced mortar (GFRM)-retrofitted counterparts of the substandard column with and without basalt mesh. The substan-
dard columns were designed to be subjected to relatively high shear ratios (i.e., the ratio of the shear force that corresponds to the moment
capacity to shear strength of the cross section) up to 0.85 and with a high axial load-to-capacity ratio of 0.75. All columns were tested under
constant axial load and reversed cyclic lateral displacement excursions. The results revealed that the columns that complied with the TSDC
showed satisfactory behavior for seismic performance, and the performance of the substandard columns was extremely poor. However, the
hysteretic performance of the substandard columns that were subjected to high axial stress and shear significantly improved after the proposed
retrofitting. Finally, a numerical model was developed in OpenSees to reproduce the hysteresis curves of the specimens. The slip of the
longitudinal bars at the column–foundation interface, strain penetration into the foundation, and buckling of the longitudinal bars in com-
pression were accounted for in the modeling. The results are in good agreement with the experimental hysteresis curves. The performance
levels of the columns are further specified, and the predictions of the current seismic codes were analyzed: (1) the European Committee for
Standardization’s 2005 code, Eurocode 8: Design of structures for earthquake resistance; Parts 1–3: Strengthening and repair of buildings
(EC8-3); and (2) the Turkish Government Ministry of Interior Disaster and Emergency Management Authority’s 2018 code, Turkish Building
Earthquake Code (TBEC). The TBEC provided more accurate estimates of plastic rotation capacities for substandard specimens. In contrast,
EC8-3 overestimated the plastic rotation capacity when shear stresses were relatively high due to lower shear span-to-depth ratios (a/d).
DOI: 10.1061/JCCOF2.CCENG-5227. © 2025 American Society of Civil Engineers.
Author keywords: Basalt mesh; Column; Ductility; High axial load; OpenSees; Reinforced concrete; Retrofitting; Seismic performance;
Sprayed glass fiber–reinforced matrix; Substandard.

Introduction buildings that were constructed before 2000 constitutes approxi-

mately 71.6% of the total in Istanbul (Aydogdu et al. 2023). There-
There are many substandard RC buildings in developing countries fore, many old buildings in Türkiye and other earthquake-prone
that were built long before the modern seismic guidelines were countries do not comply with modern provisions and requirements
published (e.g., 1950–1975). For example, the stock of these and do not satisfy seismic code regulations TSDC (Turkish Gov-
ernment Ministry of Reconstruction and Resettlement 1975) that
1
Ph.D. Candidate, Disaster Management Institute, Istanbul Technical were effective at their year of design and construction. One of
Univ. (ITU), Istanbul 34467, Türkiye (corresponding author). ORCID: the most important deficiencies of these substandard buildings
https://orcid.org/0000-0001-8896-9489. Email: [email protected] is the lower concrete compressive strength than that considered
2
Assistant Professor, Dept. of Architecture, Izmir Institute of Technol-
in the design (Inel et al. 2008) (the minimum allowed concrete
ogy, Izmir 35430, Türkiye. ORCID: https://orcid.org/0000-0001-8319
-2535. Email: [email protected]
characteristic compressive strength mandated by the TSDC
3
Assistant Professor, Dept. of Civil Engineering, Gazi Univ., Ankara (Turkish Government Ministry of Reconstruction and Resettlement
06560, Türkiye. ORCID: https://orcid.org/0000-0003-0808-9495. Email: 1975) was 14 MPa), which caused the columns to be subjected to
[email protected] higher axial stresses, which were sometimes quite close to their
4
Professor, Dept. of Architecture, ITU., Istanbul 34437, Türkiye. axial load capacities [>0.5fcbh where the concrete compressive
ORCID: https://orcid.org/0000-0001-9448-2562. Email: celikoguz@itu strength is fc, and the column dimensions are b (width) and h
.edu.tr (height)]. The second important issue is the light and improperly
5
Professor, Dept. of Civil Engineering, ITU., Istanbul 34469, Türkiye. detailed transverse reinforcement that results in poor ductility due
ORCID: https://orcid.org/0000-0002-4853-7910. Email: [email protected]
to poor confinement of the concrete and a high shear demand-to-
Note. This manuscript was submitted on January 21, 2025; approved on
March 27, 2025; published online on April 24, 2025. Discussion period capacity ratio that leads to premature shear failures in seismic
open until September 24, 2025; separate discussions must be submitted zones. These concurrent deficiencies could worsen the overall
for individual papers. This paper is part of the Journal of Composites structural seismic behavior and result in brittle failures and col-
for Construction, © ASCE, ISSN 1090-0268. lapses under gravity loads or future moderate-to-strong ground

© ASCE 04025022-1 J. Compos. Constr.

J. Compos. Constr., 2025, 29(4): 04025022

 motions (Aydogdu and Ilki 2023). Therefore, these structural mem- reinforcement and FRP). The main differences between this
bers (or structures) are unable to satisfy the specified rotation or study and previous ones are: (1) low concrete compressive strength
strain (displacement) limits that are mandated by the design (approximately 10 MPa) of the substandard columns, and the use of
codes. For instance, the recent earthquakes in Southeastern Türkiye nonribbed or plain bars as reinforcement; (2) high axial stress (0.75
were among the most disastrous, where millions of people suffered of the axial capacity, the range of which is tested for the first time in
and over 50,000 lost their lives (Iş ık et al. 2023; Celik et al. 2024). the literature in this study, to the best knowledge of the authors) on
Considering the numerous substandard structures in Türkiye, par- the substandard columns; and (3) high shear demand at flexural ca-
ticularly in Istanbul, with a very high probability of exposure to se- pacity, which resembles the prevalent coexisting deficiencies in the
vere earthquakes that measure Mw > 7.3 (where the moment substandard building stock. Square and rectangular columns with
magnitude is Mw) with an estimated return period of 475 years shear span-to-depth ratios (a/d ) of 3.80 and 2.21 are tested, respec-
Downloaded from ascelibrary.org by "Indian Institute of Technology, Hyderabad" on 04/28/25. Copyright ASCE. For personal use only; all rights reserved.

(Ansal et al. 2009), an urgent need for retrofitting is inevitable. tively (where the shear span is a, and the effective depth of the cross
Some studies discussed major earthquakes and their potential ef- section is d ). Another originality of this study is the continuation of
fects on substandard buildings and proposed rapid risk assessment the tests until the lateral load resistance or axial load capacity is
techniques (Celep et al. 2011; Tapan et al. 2013; Aydogdu and Ilki lost. In the literature, most tests on RC columns have been termi-
2023; Gurbuz et al. 2023; Aydogdu et al. 2023). Some studies have nated soon after a 15%–20% loss in lateral load capacity. In addi-
enhanced the seismic performance of substandard members using tion, although CNR-DT215 (CNR [Italian National Research
different retrofitting methods in the last few decades (Dai et al. Council] 2018) and ACI 549 (ACI [American Concrete Institute]
2012; Fakharifar et al. 2015; Deng et al. 2018; Lu et al. 2019; 2020) have addressed the design of FRCMs under gravitational
Raza et al. 2019), where different applications that used fiber- loads, there is no information on the seismic retrofitting of existing
reinforced polymer (FRP) composites have been more preferable structural members with a composite material such as FRCM in the
in recent years (Ilki et al. 1998; Mirmiran et al. 1998; Ilki and Kum- regulations and standards, to the best of the authors’ knowledge.
basar 2001; Bousias et al. 2004; Ilki et al. 2004; Farrokh Ghatte Therefore, in this study, the efficiency of an FRCM is investigated
et al. 2019; Kian et al. 2021; Narlitepe et al. 2021; Tore et al. under pure axial stress and bending moments. In addition, in the
2021; Kian et al. 2024). Of note, the FRP method has some draw- section “Analytical Modeling”, to simulate the experimental behav-
backs, such as the emission of harmful greenhouse gasses during ior, a numerical model is developed in the OpenSees framework
the gluing process, the poor behavior of epoxy resins at tempera- (Mazzoni et al. 2006) to predict the hysteretic load–displacement
tures above the glass transition temperature, ineffectiveness on curves of all tested specimens that consider the buckling and slip
wet surfaces or at low temperatures, the difficulty of postearthquake of the nonribbed conventional steel plain longitudinal bars (con-
damage assessment of members that are wrapped with FRP, and the ventional steel bars without deformation), and the effects of strain
high cost of epoxy resin (Triantafillou et al. 2006; Bournas et al. penetration into the foundation. Comparisons between the experi-
2009; Awani et al. 2017). To address the problems that stem mental and numerical results show satisfactory agreement for the
from epoxy resins that are utilized as matrix material, a new branch load–displacement curves, which demonstrates the acceptable per-
in strengthening methods has been burgeoned by the substitution of formance of the proposed modeling approach. Finally, the accuracy
epoxy resins with cement-based mortar. This new composite mate- of seismic assessment documents [e.g., the European Committee
rial consists of a fabric with open mesh geometry and cement-based for Standardization’s 2005 code, Eurocode 8: Design of structures
matrixes. These composite materials are known as fiber-reinforced for earthquake resistance; Parts 1–3: Strengthening and repair of
cementitious matrix (FRCM), which consists of a cementitious buildings (EC8-3) (CEN 2005), the 2018 Turkish Building Earth-
matrix that is reinforced with fibers. This relatively low-cost mate- quake Code (TBEC) (AFAD 2018), and ASCE 41-23 (ASCE
rial is compatible with concrete (Colajanni et al. 2014; Awani 2023)] when estimating the plastic rotation limits for different seis-
et al. 2017; Ortlepp and Ortlepp 2017; Ates et al. 2019; mic performance levels are evaluated.
Gonzalez-Libreros et al. 2019; Faleschini et al. 2020) and masonry
substrates (Mezrea et al. 2017, 2021; Del Zoppo et al. 2019; Napoli
and Realfonzo 2022) and could be implemented at a wider range of Experimental Program
temperatures as well as on wet surfaces. In addition, after the re-
moval of the plaster, the concrete surface does not need to be pre- In the experimental work, eight full-scale RC columns with cross-
pared before strengthening, which eliminates the need for primer sectional dimensions of 300 × 300 mm and 300 × 600 mm were
usage as filler for the surface. Few studies were conducted on the cast. In each group, one of the columns represented the columns
evaluation of the seismic behavior of the FRCM confined columns that were constructed and complied with the TSDC (Turkish Gov-
(Bournas et al. 2009; Luleci et al. 2021; Kian et al. 2022; Ates et al. ernment Ministry of Reconstruction and Resettlement 1975) (i.e.,
2023b). Raoof and Bournas (2017) showed that a cementitious the code-compliant column). The three other columns were de-
composite material could surpass FRP systems at high tempera- signed to resemble the substandard columns that were built in Tür-
tures or in a fire because of the permeability and nonflammability kiye before the 2000s and did not comply with any seismic design
provided by a cement-based matrix. codes. One of the three substandard specimens was maintained
This study aims to investigate the cyclic behavior of the substan- as-is and tested without any intervention. The other two were retro-
dard RC square and rectangular columns that are subjected to high fitted with external jacketing in two different configurations: (1)
axial loads (0.75fcbh) as well as high shear demand (up to 85% of sprayed GFRM with a glass fiber content of 5% of the mixture
the shear capacity of substandard columns at flexural strength) be- weight; and (2) three layers of basalt mesh combined with sprayed
fore and after retrofitting with a sprayed glass fiber–reinforced ma- GFRM with a glass content of 3.5% of the mixture weight. The cy-
trix (GFRM) external jacket with and without basalt mesh. This clic test results of the code-compliant column served as control col-
method saves a great deal of time due to the easy application of umns to give intuition that whether performance improvement of
the matrix material to the existing concrete surface using a spraying retrofitted columns could capture that of the code-compliant one.
gun and requires less labor compared with traditional retrofitting The specimens were designated with two letters, where the first
methods (e.g., enlarging the cross section with concrete jacketing, letters refer to code-compliant (C), substandard (S), substandard
the application of FRCM using a trowel, and adding steel retrofitted with sprayed GFRM only (G), and retrofitted with basalt

© ASCE 04025022-2 J. Compos. Constr.

J. Compos. Constr., 2025, 29(4): 04025022

 Table 1. Specimen matrix and characteristics according to ACI 318-19 (ACI 2019) and TBEC (AFAD 2018)
ρsh/ρsh,min Ve/Vr

Specimen Aspect ratio Explanation Axial load ratio ACI 318-19 TBEC a/d ACI 318-19 TBEC
C1 1 TSDC-compliant 0.40 0.68 1.1 3.81 0.38 0.43
S1 1 Substandard 0.75 0.20 0.65 3.81 0.62 0.80
G1 1 GFRM-retrofitted 0.75 0.20 0.65 3.81 0.72 0.75
GB1 1 GFRM + 3layers basalt mesh 0.75 0.20 0.65 3.81 0.71 0.73
C2 2 TSDC-compliant 0.40 0.99 0.85 2.21 0.34 0.33
S2 2 Substandard 0.75 0.63 0.55 2.21 0.80 0.71
G2 2 GFRM-retrofitted 0.75 0.63 0.55 2.21 0.84 0.83
Downloaded from ascelibrary.org by "Indian Institute of Technology, Hyderabad" on 04/28/25. Copyright ASCE. For personal use only; all rights reserved.

GB2 2 GFRM + 3 layers basalt mesh 0.75 0.63 0.55 2.21 0.85 0.84

mesh–reinforced GFRM (GB). The numbers refer to the cross- in the failure. Symmetrically, 10 nonribbed plain bars with a
sectional aspect ratio (e.g., either 1 or 2). Table 1 gives the speci- 16 mm diameter that corresponded to the longitudinal reinforce-
men matrix and details of the tested columns. In addition, ment ratios of 2.2% and 1.1% (calculated according to As/bh
the ratios of the transverse reinforcement of the tested columns to where the total area of longitudinal reinforcements in the cross sec-
the minimum required transverse reinforcement (ρsh/ρsh,min) that tion is As) were used in the square and rectangular specimens, re-
correspond to ACI 318-19 (ACI 2022) and TBEC (AFAD 2018) spectively. Plain bars with 10 mm diameters and spacings of 250
are listed in this table with the ratios of Ve/Vr calculated according and 80 mm were used as the transverse reinforcement that corre-
to these documents, where the shear force that corresponds to the sponded to 0.25% and 0.78% volumetric transverse reinforcement
moment capacity of the column is Ve and the shear capacity of ratios ( ρsh ) for the substandard and code-compliant specimens, re-
the column with the contribution of the concrete and transverse re- spectively. The volumetric transverse reinforcement ratio was cal-
inforcement is Vr. As given in Table 1, the axial load ratio is the culated according to ρsh = Ash/bks where the area of the transverse
ratio of the applied axial load-to-axial capacity of the RC column. reinforcement in the direction of the loading is Ash, the distance be-
All columns were cast together with overdesigned concrete tween the centroids of the extreme stirrups leg perpendicular to the
foundation blocks with plan dimensions of 1,200 × 1,200 mm and loading direction is bk, and the spacing of the stirrups is s. In addi-
1,800 × 14,00 mm for the square and rectangular column cross sec- tion, in the tested substandard columns, the ratio of the stirrup spac-
tions, respectively. The foundations had a height of 700 mm and ing to the maximum allowable stirrup spacing based on ACI
were intentionally over-designed to ensure sufficient strength for 318-19 (ACI 2022) and TBEC (AFAD 2018) were 1.90 and
testing the behavior of the columns. This was done because sub- 2.60, respectively. The geometry, reinforcement details, and
standard columns typically fail during earthquakes before the foun- mounted strain gauge (SG) positions for the square and rectangular
dations, especially when soil properties are not a significant factor columns are shown in Figs. 1 and 2, respectively.

Fig. 1. Reinforcing details and dimensions for the square columns (mm). Note: YFLA5, FLA5 = the commercial names of the strain gauges type that
were used for longitudinal and transverse bars, respectively with a gauge length of 5 mm and water resistance.

© ASCE 04025022-3 J. Compos. Constr.

J. Compos. Constr., 2025, 29(4): 04025022

Downloaded from ascelibrary.org by "Indian Institute of Technology, Hyderabad" on 04/28/25. Copyright ASCE. For personal use only; all rights reserved.

Fig. 2. Reinforcing details and dimensions for the rectangular columns (mm).

Table 2. Concrete mix proportions lower than 14 MPa. The compressive strength of the concrete
that was used for the construction of the footings was 35.6 MPa.
Crushed aggregate Sand
The average compressive strength was calculated from the uniaxial
(5–12 mm) (0–5 mm) Cement Water
Columns (kg/m³) (kg/m³) (kg/m³) (kg/m³)
compressive test results for three standard cylinder specimens (150
× 300 mm) in accordance with ASTM C39/C39M (ASTM 2021a).
Substandard 769 1,153 180 198 The first and last RC column specimens were tested at 62 and
Code-compliant 727 1,090 257 231
106 days after the specimens were cast, which corresponds to an
average concrete age of 84 days. Three tensile tests were carried
out on the longitudinal and transverse nonribbed plain reinforcing
Concrete and Reinforcement Material Properties
bars in accordance with ASTM A615/A615M (ASTM 2022a). The
yielding and ultimate stresses of the longitudinal ( fyl and ful) and
A ready-mixed concrete with average 28 and 90-day compressive
transverse ( fyw and fuw) bars were 324 and 404 MPa, and 361 and
strengths of 9.8 and 10.5 MPa for substandard columns and 19.3
430 MPa, respectively. The yielding and ultimate strains of the lon-
and 19.5 MPa for code-compliant columns were used, respectively.
gitudinal (εyl and εul ) and transverse (εyw and εuw ) bars were
Table 2 gives the concrete mix proportions for the substandard and
0.0015 and 0.28, and 0.0017 and 0.27, respectively. The COV
code-compliant specimens. Table 3 lists the concrete properties.
for the yielding and ultimate stresses and strains of the longitudinal
The concrete elastic modulus is Ec, the strain that corresponds to
and transverse reinforcement was 8%.
the peak stress is εco , the ultimate strain is εc85 (strain that corre-
sponds to 85% of the stress in the descending branch), the average
is μ, and the standard deviation is σ. The modulus of elasticity of
the concrete was calculated according to ASTM C469/C469M Material Properties of Retrofitting System
(ASTM 2022b), which defines the chord modulus as the slope be-
tween the stress levels that correspond to axial strains of 0.00005 The ingredients of the mortar in the GFRM were white standard
and 0.4fc. portland cement (900 kg/m3), fine silica sand (1,000 kg/m3), meta-
The aim of using two types of concrete was to assess the effec- kaolin (100 kg/m3), latex modifier (33 kg/m3), water (320 kg/m3),
tiveness of the retrofitting material when improving the seismic be- and polycarboxylate ether superplasticizer (2.4 kg/m3) with a mix
havior of substandard columns and to evaluate whether retrofitted ratio of 1:1.11:0.11:0.036:0.35:0.0026 by weight. The 28-day aver-
substandard columns could achieve or exceed the performance of age compressive strength, σ, and COV of five standard 50 mm cubic
code-compliant columns. Of note, the TSDC (Turkish Government specimens in accordance with ASTM C109/C109M (ASTM 2021b)
Ministry of Reconstruction and Resettlement 1975) did not allow were 47.5, 1.3 MPa, and 2.8% for the GFRM with a 3.5% (by
for the use of concrete with a characteristic compressive strength weight) short glass fiber content, and 41.4, 0.5 MPa, and 1.4% for

Table 3. Concrete properties

Substandard Code-compliant

Parameter fc (MPa) Ec (MPa) εco εc85 fc (MPa) Ec (MPa) εco εc85

µ 9.8 14,400 0.0023 0.0054 19.3 29,200 0.0021 0.0035
σ 1.2 176.7 0.0004 0.0001 1.1 440.2 0.0001 0.0001
COV (%) 10.9 1.0 18 1.9 5.5 2.0 1.0 1.7

© ASCE 04025022-4 J. Compos. Constr.

J. Compos. Constr., 2025, 29(4): 04025022

 the GFRM with a 5% (by weight) short glass fiber content, respec-
tively. The high compressive strengths of the mortar (41.4 and
47.5 MPa) could be attributed to the low water-to-cement (w/c)
ratio of 0.35 and the inclusion of fine silica, which contributed sig-
nificantly to the mortar’s strength development. Being a retrofitting
system, it requires high strength mortar to adhere firmly to the con-
crete substrate, which creates a cohesive composite material where
the fibers and mortar work effectively together.
The tensile strength, tensile strain, bending strength, and elastic
modulus of the GFRM with a 3.5% glass fiber content and three
Downloaded from ascelibrary.org by "Indian Institute of Technology, Hyderabad" on 04/28/25. Copyright ASCE. For personal use only; all rights reserved.

layers of basalt mesh were 9.7, 0.0091, 19.9 MPa, and 27.4 GPa, (a) (b) (c)
respectively. However, for the GFRM with a 5% glass fiber con-
tent, these were 7.42, 0.0077, 19.4 MPa, and 26.9 GPa (Ates
et al. 2023a). Of note, the cost of the retrofitting material with a
GFRM is approximately 25% less than that of the retrofitting tech-
nique where a combination of GFRM and basalt mesh is utilized.
The Short Nippon electric glass fibers (Shiga, Japan) that were
used in this study contain 17% zirconia, which assures resistance
against alkali in the cement. The elastic modulus, tensile strength,
and failure strain of the glass fibers are 75 GPa, 1,500 MPa, and
0.02, respectively, as provided by the manufacturer. These excep-
tional tensile characteristics contribute to the high tensile strength
of the GFRM, which provided adequate confinement to the col-
umns in this study. In addition, compared with carbon fibers,
glass fibers are cost-effective with alkali resistance. The technical
properties of the basalt mesh reinforcement for the weft and wrap
orientations with the same number of fibers in both directions pro-
vided by the (Spinteks) manufacturer are given in Table 4. The
(d)
height of the column is covered with a total width of basalt mesh
(1,000 mm). Fig. 3. Showing: (a) rounded corners of columns; (b) GFRM spraying;
Using a concrete grinding machine, the four corners of the col- (c) wrapping of first layer of basalt mesh; and (d) retrofitted column
umns were rounded off to a radius of 25 mm before retrofitting to cross section, basalt mesh, and glass fibers (mm). Note: R25 = the ra-
prevent premature stress concentrations on the corners of the retro- dius of rounded off corner.
fitting jacket. Then, using sandpaper, the rounded corners were
smoothed [Fig. 3(a)]. Fig. 3(b) shows the GFRM retrofitting pro-
cess, and Fig. 3(c) shows the basalt mesh–reinforced GFRM appli-
cation. Fig. 3(d) shows the basalt mesh grid and the glass fibers that cross-sectional analyses that were employed in the available mod-
were used in the mortar with cross-sectional representation. Of els in the literature (Triantafillou et al. 2006; Ates et al. 2019) to
note, bidirectional PVC-coated and silane fiber-sized dry basalt achieve the targeted structural seismic performance. To ensure
mesh reinforcement with mesh spacings of 25 mm in two directions the uniform thickness of the GFRM jacket, two wooden sticks,
were used in conjunction with the GFRM with a 3.5% glass fiber each 30 mm thick, were fixed to the top and bottom of the column,
content [Fig. 3(d)]. The mesh spacing was large to permit the which served as guides to maintain a consistent jacket thickness.
proper dispersion of the GFRM in the spaces in the basalt mesh re- The thickness of the GFRM was controlled using a depth gauge.
inforcement. Retrofitting was applied at heights of 1,000 mm from To avoid damage that was induced by cracks above the externally-
the bottom for the columns, 54 days after casting, using the sprayed jacketed plastic hinge zone during the reversed cyclic lateral load-
GFRM with or without the basalt mesh. Glass fibers with average ing, which was reported by other researchers (Tong et al. 2020;
filament diameters of 18 ± 2 μm and 32 mm lengths were chopped
Ates 2022; Ates et al. 2023b), the total height of the column
automatically by the spraying gun and mixed into the mortar that
(1,000 mm) was retrofitted instead of only the potential plastic
came from a tube. Then, the GFRM was sprayed (at a distance of
hinge region. For the specimens that were retrofitted with basalt
approximately 30–50 cm from the columns) onto the surface of
mesh–reinforced GFRM (which is referred to as basalt mesh–rein-
the members with a thickness of 30 mm [Figs. 3(b and c)]. The
forced FRCM), the process began by spraying the first GFRM layer
thickness of the GFRM jacket was determined using numerical
(approximately 7 mm thick) onto the concrete substrate. Then, a
basalt mesh was embedded in the GFRM, which ensured an even
Table 4. Basalt mesh reinforcement technical properties
spacing of 7–8 mm between each layer. This process was repeated
Property Value layer by layer until the final basalt mesh was embedded, and a final
3
Filament density (kg/m ) 2,620–2,650 GFRM layer was sprayed onto the column surface. Finally, an
Specific surface weight (g/m2) 303 overlap of 200 mm was provided in the transverse direction. The
Thickness (mm) 0.8–0.9 GFRM was evenly distributed over the concrete substrate using a
Width (mm) 1,000 roller with smooth, consistent pressure. Wooden sticks were used
Tensile strength (kN/m) 95 to maintain a 30 mm gap between the bottom of the jacket and
Young’s modulus (GPa) >90 the top of the foundation, which prevented early damage from
Note: Technical properties were provided by the manufacturer (Spinteks, the jacket coming into contact with the foundation under high
Denizli, Türkiye). axial loads and lateral displacement.

© ASCE 04025022-5 J. Compos. Constr.

J. Compos. Constr., 2025, 29(4): 04025022

Downloaded from ascelibrary.org by "Indian Institute of Technology, Hyderabad" on 04/28/25. Copyright ASCE. For personal use only; all rights reserved.

(a) (b)

Fig. 4. Test setup: (a) schematic; and (b) typical column.

Test Setup lateral displacement of the column, respectively. Finally, an

LVDT (LVDT-6) with a gauge length of 50 mm was installed on
Square and rectangular columns were subjected to reversed cyclic the side face of the column at the top perpendicular to the load di-
displacement reversals that involved applying a horizontal load rection to measure out-of-plane deformations.
under constant axial loads of 709 and 1,418 kN, which correspond
to 0.75 of their axial load capacity, respectively. The lever arm of
the horizontal load was 1,000 and 1,200 mm for the square and
Experimental Results and Discussion
rectangular columns, respectively. All tests were conducted within
106 days after the columns were cast. The test setup is shown in
Fig. 4. Lateral Load–Displacement Hysteresis Curves and
Each footing had eight holes to fix the specimen to the strong Damage Progression
floor of the outdoor laboratory using 36 mm diameter high strength Fig. 6 shows the lateral load–displacement hysteresis curves that
bolts. The horizontal hydraulic actuator was fixed to a strong RC were obtained at the end of the experiments for tested columns.
reaction wall. The vertical load was applied using a manually con- The critical stages, such as initial flexural cracking, yielding of
trolled hydraulic jack with 2,000 kN capacity in a force-controlled the longitudinal reinforcements, lateral and axial load capacity
mode. The axial load was maintained almost constant during the losses, and buckling of the longitudinal bars, are marked on the
tests with a ± 5% fluctuation in the target axial load. Then, hysteresis curves for a better understanding of the behavior (e.g.,
quasi-static cyclic lateral displacement reversals were applied observed during tests and determined by the analytical modeling).
according to ACI 374 (ACI 2013) with increasing drift levels The near collapse (NC) and collapse prevention (GO) limit states
(±0.125%, ± 0.25%, ± 0.5%, ± 0.75%, ± 1.0%, ± 1.5%, ± 2.0%, according to EC8-3 (CEN 2005) and TBEC (AFAD 2018) are
± 3.0%, and ± 4.0%) using a horizontal actuator with displacement marked in Fig. 6. As seen in Fig. 6, the developed model could suc-
and a force capacities of ± 400 mm and ± 300 kN, respectively. cessfully capture the strength and displacement capacity.
The target lateral displacement cycles were repeated twice, as rec- The damage to the tested columns at distinct drift ratios at the end
ommended by this guideline ACI 374 (ACI 2013). To capture the of the second cycle is shown in Figs. 7 and 8, respectively. During
behavior of the columns within the linear elastic response range, the the tests on the square and rectangular specimens, flexural cracks
initial drift was set to 0.125%, which corresponds to approximately were first observed at the column–footing interfaces at 0.5% and
¼ of the theoretical yield drift. 0.25% drift ratios, respectively. For the C1 and C2 specimens, the
The locations of the linear variable differential transducers cracks were 100–200 mm and 200–400 mm above the bottom of
(LVDTs), which were installed on the columns to measure the dis- the columns, respectively. On the faces of the C1 and C2 columns
placements of their position, are shown in Fig. 5. Six LVDTs perpendicular to the loading direction, slight concrete cover crush-
(LVDT-2, as shown in Fig. 5) were used to measure deformations ing was observed at the 1.5% and 1% drift ratios, respectively.
on both sides of the column in the load direction within the plastic Spalling of the concrete cover was observed on the same faces of
hinge zone, which provided data to calculate the column rotations, C1 and C2 at 2% and 1.5% drift ratios, respectively. Horizontal flex-
moment–curvature relationships, and crack opening or closing at ural cracks were propagated and widened up to 1.1–1.3 mm in the
the column–foundation interface. The deformation within 30 mm plastic hinge zone (which is equal to the depth of the cross section)
of the column base region that was captured by the closest until 4% and 2% drift ratios for the C1 and C2 columns, respectively
LVDT to the foundation was used to calculate the rotation at the (Figs. 7 and 8). Then, the tests on C1 and C2 were terminated by a
base of the column due to bar slip that was induced by strain pen- gradual lateral load bearing capacity loss of approximately 30%, at
etration. Two LVDTs (LVDT-1) were used to measure the rotation 4% and 2% drift ratios, respectively [Figs. 6(a and b)]. The reason
of the foundation, and LVDT-3 was installed to measure the foun- for the higher drift ratio for C1 is that the a/d ratio was 73% higher
dation translation, which were both negligible for all tested col- than C2, which allowed flexurally dominated deformations. How-
umns. LVDT-4 and LVDT-5 measured the midheight and top ever, for C2, the lower a/d ratio resulted in more shear-dominated

© ASCE 04025022-6 J. Compos. Constr.

J. Compos. Constr., 2025, 29(4): 04025022

Downloaded from ascelibrary.org by "Indian Institute of Technology, Hyderabad" on 04/28/25. Copyright ASCE. For personal use only; all rights reserved.

(a) (b)

Fig. 5. Instrumentation of: (a) square; and (b) rectangular columns (mm).

deformation. In addition, the efficiency of the confinement due to transverse reinforcement ratios for the code-compliant and substan-
the smaller cross-sectional aspect ratio of the square column was dard specimens.
considered another reason for the higher drift capacity. For S1, con- By comparing the ultimate drift capacity of the substandard
crete cover crushing and spalling occurred at 1% and 1.5% drift ra- and code-compliant specimens, the necessity of retrofitting to im-
tios, respectively. During the first cycle of the 1.5% drift ratio, the prove the seismic performance of substandard columns was noted.
lateral load capacity loss was 24.2%, and an axial load capacity For the retrofitted Columns G1 and GB1, a vertical crack in the
loss of 47.6% was recorded during the second cycle of the 1.5% external jacket was observed on the edge of the external jacket
drift ratio [Fig. 6(c)]. For S2, concrete cover crushing and spalling for the 2% drift ratio; however, for G2 and GB2, it was observed
occurred at the 0.75% drift ratio. During the first cycle of the on the edge of the jacket for the 0.75% and 1% drift ratios, respec-
0.75% drift ratio, the lateral load capacity loss was 20.2%. Here, tively. The cracks were initiated from the edge of the jacket due to
the axial load decreased by 14%. Therefore, the failure of the sub- stress concentration in these regions. (Ates et al. 2019). Then, the
standard specimens was brittle due to spalling of the concrete crack width was widened, and the crack length was extended in
cover, buckling of the longitudinal reinforcements, and crushing the plastic hinge zone from the column–foundation interface up
of the core concrete sequentially. Of note, core concrete crushing to 300 mm for G1 and GB1 (Fig. 7) and 600 mm for G2 and
was detected after the cover concrete was removed following the GB2 (Fig. 8). This is due to the different plastic hinge zone
completion of testing [Figs. 7 and 8]. In addition, S2 experienced lengths of 300 and 600 mm for square and rectangular columns.
more brittle failure due to more shear-dominated behavior that The failure of G1 and G2 was captured by the increase in crack
was triggered by a lower a/d ratio compared with S1. Comparing width and rupture of the GFRM jacket. The failure modes of
the results between S1 and C1 revealed that cover crushing and GB1 and GB2 were accompanied by basalt mesh rupture at 3%
spalling in S1 occurred three times earlier than C1 in terms of and 1.5% drift ratios after widening the formed crack on the jacket
drift ratio, while failure of S1 occurred 2.66 times earlier than C1 at 2% and 0.75% drift ratios, respectively [Figs. 6(g and h), 7, and
in terms of drift ratio. The same values were obtained for the C2 8]. The basalt mesh rupture could be justified because the confin-
and S2 specimens for crushing and spalling. All these behavioral ing stress exceeded the tensile strength of the jacket. Then, by ex-
differences stem from different applied axial load-to-capacity and tending and widening the crack, the tensile stress transfers

© ASCE 04025022-7 J. Compos. Constr.

J. Compos. Constr., 2025, 29(4): 04025022

Downloaded from ascelibrary.org by "Indian Institute of Technology, Hyderabad" on 04/28/25. Copyright ASCE. For personal use only; all rights reserved.

(a) (b)

(c) (d)

(e) (f )

(g) (h)

Fig. 6. Hysteresis curves of specimens: (a) C1; (b) C2; (c) S1; (d) S2; (e) G1; (f) G2; (g) GB1; and (h) GB2.

through the cracks to the basalt mesh, which results in basalt rup- was promising to overtake and reach the target lateral load capac-
ture. Compared with S1 and S2, a more gradual failure was wit- ity of the code-compliant specimen (C1).
nessed for the retrofitted specimens, similar to C1 and C2. The As shown in Figs. 7 and 8, substandard columns experienced
lateral load capacity of S1 was 26% less than C1. When compared buckling in the longitudinal rebars, which led to early concrete
with S1, the lateral load capacities of G1 and GB1 were increased cover spalling and a brittle failure mode with strength degradation.
by 35% and 32%, respectively. The lateral load capacity of code- This failure occurred due to the high axial load level, which re-
compliant Specimen C1 was successfully achieved by retrofitting sulted in a loss of axial stiffness and led to the loss of the lateral
the square columns. The lateral load capacity of S2 was 27% less stability of the longitudinal reinforcing bar. This instability was ex-
than C2. Compared with S2, the lateral load capacities of the G2 acerbated by the weak lateral confining pressure and the high un-
and GB2 specimens were 15% and 16% higher, respectively. supported length of the longitudinal bar [(s/db) where the
Therefore, in square specimens, the applied retrofitting method diameter of the longitudinal bar is db]. However, in the

© ASCE 04025022-8 J. Compos. Constr.

J. Compos. Constr., 2025, 29(4): 04025022

 in
in
Downloaded from ascelibrary.org by "Indian Institute of Technology, Hyderabad" on 04/28/25. Copyright ASCE. For personal use only; all rights reserved.

Fig. 7. Damage progress in square specimens (DR = drift ratio).

in

in

in

Fig. 8. Damage progress in rectangular specimens.

GFRM-retrofitted columns, although the bars reached their com- reported by Ates et al. (2023b). The results of this study indicate
pressive yield strain and buckled under severe axial load, they that, in addition to the s/db ratio, the onset of bar buckling depends
could sustain axial load at larger drift ratios. This was possible be- on the confinement that is provided by the external jacket (e.g., G1,
cause the external confinement restricted the cover spalling and ex- GB1, G2, and GB2) and internal transverse reinforcement (e.g., C1
erted inward lateral pressure on the concrete, which prevented and C2). A summary of the test results for the square and rectangu-
sudden bar buckling by restricting the lateral deformation, as lar specimens is given in Table 5.

Table 5. Summary of test results

Vei μiD EDCi
(%) (%) (%)
Specimen Peak load (Ve) (kN) C1,2/S1,2
Ve δy (mm) δu (mm) μD C1,2/S1,2
μD EDC (kN/m) EDCC1,2/S1,2
C1 130.5 — 6.3 36.6 5.8 — 52.1 —
S1 98.8 −24 4.7 14.1 3.0 −48.3 12.5 −76
G1 133.7 +35 6.0 27.0 4.5 +50 44.4 +255.4
GB1 130.8 +32 5.6 30.0 5.4 +80 40.1 +227.4
C2 305.4 — 3.6 20.9 5.8 — 63.8 —
S2 223.6 −27 1.9 5.9 3.1 −46.6 11.4 −82.1
G2 257.8 +15 4.4 18.4 4.2 +35.5 38.9 +242.5
GB2 259.8 +16 3.5 17.6 5.0 +61.3 42.3 +271.9
Note: Specimen’s decrement values for peak lateral load (Ve), ductility (μD), and EDC for S specimens are written with respect to C specimens, and increments
of the same values for G and GB specimens are written with respect to S specimens for percentage. V ie, EDCi, μiD = shear force, energy dissipation capacity,
and displacement capacity of S1, S2 specimens when divided to C1 or C2 specimens. Similarly they are shear force, energy dissipation capacity, and
displacement capacity of retrofitted specimens when divided to S1 or S2 specimens. i = S1, S2, G1, G2, GB1, GB2 specimens.

© ASCE 04025022-9 J. Compos. Constr.

J. Compos. Constr., 2025, 29(4): 04025022

 confinement pressure. This is because the C1 and C2 specimen
buckling did not occur because of the high transverse reinforcement
ratio and lower axial load-to-capacity ratio (0.40). However, in the
retrofitted specimens, buckling was postponed and largely limited,
and the axial load-to-axial capacity ratio was higher (0.75). Basalt
mesh works as a buried reinforcement inside the GFRM mortar
even after cracking in the mortar. Therefore, although G1 and G2
contain 5% glass content, the GB1 and GB2 specimens with a
lower glass content of 3.5% and three layers of buried textile rein-
forcement had more ductility (20%). This was because, after the ex-
Downloaded from ascelibrary.org by "Indian Institute of Technology, Hyderabad" on 04/28/25. Copyright ASCE. For personal use only; all rights reserved.

ternal jacket fractured in GB1 and GB2, the basalt meshes that bore
the tensile stresses provided confinement to the core concrete and
prevented crack extension and widening in contrast to G1 and
G2. In addition, the ultimate tensile strain of the GFRM with
3.5% glass content and three layers of basalt (GB specimens)
mesh was 18% higher than that of the GFRM with 5% glass content
(G specimens) (Ates et al. 2023a). The ductility increase was 29%
less for the G2 with respect to G1. For GB2, the ductility increase
was 23.4% less compared with the GB1 specimens. This was be-
cause of a lower a/d ratio in the rectangular columns (42%),
Fig. 9. Yield and ultimate displacements, dissipated energy, and secant which decreased the confining efficiency due to predominant
stiffness definition. Note: Pmin = maximum lateral force in pulling di- shear deformations. In addition, the confining efficiency that
rection; Pmax = maximum lateral force in pushing direction; and Pi = stemmed from the retrofitting on the rectangular columns was
the lateral force corresponding to target drift at i-th cycle. less than the square columns, as reported by Demir et al. (2019).

Displacement Ductility Strain Profiles and Strain Penetration

For better comparisons between the column performance when sus- The strain profiles of the longitudinal bars for all columns are
taining their ductility, displacement ductility factors, which were shown in Fig. 10 for different drift ratios. The strain values were
defined as the ratio of the ultimate displacement (δu) to the yield calculated by averaging the captured data in the pushing and pull-
displacement (δy), were calculated (Table 5). To determine the ing directions that corresponded to the final target drifts. Crack
yielding point, the method that was proposed by Priestley (2000) openings in the column–foundation interface were reported in the
was used. The ductility definition is shown in Fig. 9. First, to deter- orders of lower axial load-to-capacity ratios (i.e., <20%) (Verder-
mine the yield displacement, a secant line that passed through the ame et al. 2008; Di Ludovico et al. 2014; Ali and El-Salakawy
point, which corresponded to 0.75 of the maximum load in the as- 2016; Di Ludovico et al. 2014; Seifi et al. 2018). This study re-
cending branch, was drawn until it intercepted the horizontal tan- vealed a similar phenomenon in substandard and retrofitted col-
gent line that was drawn from peak load. From this point, a umns under a high axial load ratio (75% of the capacity) due to
vertical line was drawn toward the displacement axis, where it in- the penetrated tensile strain of the longitudinal bars inside the foun-
tercepts the envelope curve, which is specified as the yield point. dation. As shown in Fig. 10, the yield and postyield strains in the
The ultimate lateral displacement was assumed to be the displace- longitudinal reinforcements spread over a plastic hinge length
ment that corresponds to 80% of the lateral load strength on the de- that was approximately equal to the column depth for all columns,
scending branch of the load–displacement curve. The average except for the S1 and S2 substandard columns. The high axial load
displacement ductility ratios of the columns were calculated and low concrete compressive strength resulted in low tensile stress
using the following equation. The average values of the yield in the longitudinal bars. The weak bond and bar slip led to strain
(δ+y and δ−y ) and ultimate displacements (δ+u and δ−u ) for the push- penetration into the foundation, which caused crack opening at
ing and pulling directions have been considered: the column–foundation interface. This strain penetration contrib-
uted to rigid rotation at the base and influenced top displacement
(|δ+u |+|δ−u |) due to bar slip. This agrees well with the crack opening at the bot-
μD = (1)
(|δ+y |+|δ−y |) tom of the column results (Figs. 7 and 8). Fig. 10 shows that the
longitudinal reinforcing bars of the code-compliant (C1 and C2)
The lower ductility of the substandard specimens was due to the columns yielded in tension; however, the longitudinal bars in the
poor confinement that was provided by the stirrups. The ductilities substandard columns (S1 and S2) did not experience yielding in
of S1 and S2 were 48% and 47% less than C1 and C2, respectively. tension. This was because the high axial load and low transverse
The ductility increments for G1 and GB1 were 50% and 80% with reinforcement in the substandard columns resulted in concrete
respect to S1. For G2 and GB2, the ductility increased by 35% and crushing and poor confinement during the early stages of loading,
62% with respect to S2. Therefore, the retrofitting method compen- which led to a compressive-governed behavior. However, longitu-
sated for the lack of ductility in substandard columns with respect dinal reinforcing bars in the retrofitted specimens (e.g., G1, G2,
to the code-compliant ones by successfully postponing bar buck- GB1 and GB2) yielded in tension as code-compliant specimens
ling, concrete cover spalling, and core concrete crushing. The duc- successfully. This was because the compressive-governed failure
tility of the retrofitted specimens increased by up to 80% compared of substandard columns changed to tension-controlled failure in
with the substandard specimens, which indicates a significant im- retrofitted columns due to the confinement that came from the ex-
provement in the displacement capacity increase. Of note, the ap- ternal GFRM jackets. This can be rooted to the strain profiles as
plication of a thicker cover layer could further enhance the mentioned earlier. For all specimens, the strain values showed a de-
column performance by providing an increase in the lateral creasing trend in the region between h/2 and h. Therefore, above h,

© ASCE 04025022-10 J. Compos. Constr.

J. Compos. Constr., 2025, 29(4): 04025022

Downloaded from ascelibrary.org by "Indian Institute of Technology, Hyderabad" on 04/28/25. Copyright ASCE. For personal use only; all rights reserved.

(a) (b)

(c) (d)

(e) (f )

(g) (h)

Fig. 10. Strain profiles: (a) C1; (b) C2; (c) S1; (d) S2; (e) G1; (f) G2; (g) GB1; and (h) GB2. The horizontal solid line refers to the column–foundation
interface; C = compression; and T = tension.

there is a lower strain distribution along the longitudinal reinforce- Specimens C1 and C2 reached compression yield strain values
ment, indicating that bar slippage inside the column above h was at 3% and 2% drift ratios, respectively [Figs. 10(a and b)]. Yielding
insignificant. This agreed well with the damage pictures (Figs. 7 and postyielding compression strains for S1 and S2 occurred at the
and 8), where above the plastic hinge zone, there was no splitting sections 150 and 300 mm above the footing at 1% and 0.5% drift
or cracking that originated from the bar slippage. ratios, respectively, which was consistent with the observed

© ASCE 04025022-11 J. Compos. Constr.

J. Compos. Constr., 2025, 29(4): 04025022

 compression damage (cover spalling) [Figs. 10(c and d)]. For G1 displacement in the plastic hinge zone. In addition, the total elastic
and GB1, reinforcement did not yield in compression until a displacement above and in the plastic hinge was calculated and
1.5% drift ratio [Figs. 10(e and g)]. At a 3% drift ratio, G1 and added to the flexural deformation that was calculated in the plastic
GB1 experienced postyield compression strains at the footing hinge zone. Similar assumptions were made by Ozcan et al. (2010)
level [Figs. 10(e and g)]. This was consistent with the GFRM jacket and Ates et al. (2023b) for the determination of the displacement
crack propagation that was witnessed at that drift ratio. Specimens components of RC columns with plain bars. Finally, by subtracting
G2 and GB2 experienced yield and postyield compression strains at the sum of the flexural and slip displacements from the total dis-
1.5% drift ratios, which indicated the efficiency of the confinement placement, the shear displacement was determined. Sliding compo-
that was provided by the external jacket that postponed the yield nents were neglected according to the values that were measured by
compression strains three times more than that of S2 [Figs. 10(f the LVDTs. In addition, the elastic shear displacement was calcu-
Downloaded from ascelibrary.org by "Indian Institute of Technology, Hyderabad" on 04/28/25. Copyright ASCE. For personal use only; all rights reserved.

and h)]. These results indicate that a GFRM jacket could drastically lated using 1.2VeLV/AG, where the shear modulus is G and was cal-
restrict the compressive damage to longitudinal bars. culated using Ec/2(1 + νc) (Pujol 2002). Here, the distance from the
It was assumed that the vertical displacement that was measured actuator to the middle of the first gauge length is Lv, the cross-
by the first LVDT from the surface of the foundation with a gauge sectional area is A, and Poisson’s ratio is νc. The elastic shear dis-
length of 30 mm was composed of slip in the longitudinal bars placements for the S1 and S2 specimens were 0.18 and 0.24 mm,
(Rodrigues et al. 2013; Seifi et al. 2018). The vertical displacement respectively.
that was measured by two subsequent LVDTs with gauge lengths The results revealed that a significant proportion of the top dis-
of 120 and 150 mm were assumed to be composed of the flexural placement (24%–51%) originated from the column–foundation

(a) (b)

(c) (d)

(e) (f )

(g) (h)

Fig. 11. Contribution of lateral displacement components on top displacement; (a) C1; (b) C2; (c) S1; (d) S2; (e) G1; (f) G2; (g) GB1; and (h) GB2.

© ASCE 04025022-12 J. Compos. Constr.

J. Compos. Constr., 2025, 29(4): 04025022

 interface rotations. Particularly for S1 and S2, a rocking behavior
that was governed by the opened crack in the column–foundation
joint was observed. This could be one of the drawbacks of using
smooth or plain rebars in older RC buildings that constitute a sig-
nificant portion of the existing building inventory. Fig. 11 shows
the contribution proportion of the lateral displacement components
on the top displacement. For the retrofitted and code-compliant
specimens, the contribution of the flexure and slip deformations (a)
with an increase in the drift ratio increased and decreased, respec-
tively [Figs. 11(a, b, and e–h)]. By assuming that the rotational slip
Downloaded from ascelibrary.org by "Indian Institute of Technology, Hyderabad" on 04/28/25. Copyright ASCE. For personal use only; all rights reserved.

values were prevalent at the column–footing interface, the average

curvature values at 0–30 mm were used to calculate the share of
slip on the top displacement. In all specimens, the share of top
displacement from the shear deformations was approximately
10%–18%. The slip component of the top displacement (Δslip) is
obtained according to the following equation:
(b)
Δslip = φaverage · LG · LV = θslip · LV (2)
Fig. 12. Stiffness versus drift ratio of: (a) square; and (b) rectangular
where θslip = measured rotation due to slip at the bottom of the col- specimens per drift ratio.
umn; and φaverage and LG = average curvature of first LVDTs and
the height of the measurement segment, respectively.
Average curvatures are obtained using the following equation:
that an axial load ratio from zero up to 12% did not have a consid-


|dl | + |dr | 1 erable effect on stiffness degradation. In this study, a prominent dis-
φaverage = × (3) crepancy in the stiffness degradation rate was observed, comparing
Lh LG
substandard (axial load ratio of 0.75) and code-compliant specimens
where dl and dr = measured left and right displacements using (axial load ratio of 0.4). All square and rectangular columns up to
LVDTs at the same level; and Lh = horizontal distance between the 1.5% and 0.5% drift ratios, respectively, showed approximately
two LVDTs. The base moment was calculated by the production the same stiffness degradation trend. At a constant drift ratio, S1 and
of the measured lateral load and the distance between the middle S2 had lower stiffness compared with the other specimens, which
of the gauge length to the applied lateral load. indicated rapid stiffness degradation. This sudden stiffness degrada-
The plastic hinge length was considered to be identical to the tion in the substandard column was attributed to the high axial load
depth of the cross section based on: (1) crack propagation (Figs. stresses under which the reinforcing bars buckled. In addition, due
7 and 8) until 300 and 600 mm from the column–foundation inter- to their outward bending, the concrete cover spalled completely, and
face for square and rectangular columns, respectively; and (2) SG the core concrete was crushed. Therefore, the concrete cross-
data in which compression and tensile yielding of the reinforcing sectional area was significantly reduced in the plastic hinge zone.
bars are witnessed at distances equal to the section dimensions. Continuous and smooth stiffness degradation was observed in the
However, the assumed plastic hinge length according to the retrofitted and code-compliant specimens due to their ductile behav-
TBEC (AFAD 2018) is half the section dimension (h/2) regardless
ior. This was due to enhanced confinement, which delayed the con-
of the axial load level. In addition, Paulay and Priestley (1992) pro-
crete spalling and bar buckling and allowed the column to sustain
posed an equation for the plastic hinge length (lp = 0.08l + 0.022db-
higher drift ratios with maintained structural integrity, as shown
fyl), which results in lp ≈ 0.5 − 1.0 h in many typical cases (240 mm
in Fig. 12. The retrofitted and code-compliant square columns
for the tested columns in this study). Where the plastic hinge length
is lp and the clear height to the horizontal load application (shear
span) is l. In addition, the calculated plastic hinge length, which
considers the effect of axial load according to Bae and Bayrak
(2008), was 150 mm. Therefore, it could be modified for sub-
standard columns with high axial stress and poor seismic detailing.
According to EC8-3 (CEN 2005), the plastic hinge length formula-
tions are only applicable for seismically designed columns with
deformed bars. Of note, for all specimens, the highest contribution
of the slip component could be seen in the 0–30 mm measure- (a)
ment zone due to plain bar slip, which was consistent with the
strain profiles (Fig. 10). The high level of slip deformation contri-
bution on the top displacement gave extra displacement capacity to
columns.

Stiffness Degradation
The average peak-to-peak secant stiffness (Ki) versus drift ratio of
the second cycle in the push and pull directions for the square and (b)
rectangular specimens are shown in Fig. 12. The secant stiffness
Fig. 13. Residual displacement of (a) square; and (b) rectangular spec-
is defined as the ratio of the lateral load to the corresponding dis-
imens per drift ratio.
placement at each cycle (Fig. 9). Rodrigues et al. (2013) showed

© ASCE 04025022-13 J. Compos. Constr.

J. Compos. Constr., 2025, 29(4): 04025022

 with a/d of 3.81 have higher stiffness with respect to the substandard Lobatto integration points along the column height were modeled,
ones. However, this was not same for the rectangular columns with a which accounted for the distributed plasticity along with the
lower a/d of 2.21, which could be attributed to the lower efficiency element.
of the external jacketing confinement.
Constitutive Models
Residual Displacement Ratio
Core and cover concrete, as well as the steel reinforcement behav-
The average residual displacement ratios at zero lateral force after un- ior, were incorporated into the analysis by defining the stress–strain
loading at each cycle (δres/δi) versus the drift ratio in the push and pull relationships for the relevant fibers. The parameters for the confined
directions of the second cycle for all specimens are shown in Fig. 13. core and unconfined cover concrete of S1, C1, S2, and C2 were cal-
In this figure, the residual displacement after unloading at zero lateral
Downloaded from ascelibrary.org by "Indian Institute of Technology, Hyderabad" on 04/28/25. Copyright ASCE. For personal use only; all rights reserved.

culated based on the constitutive models that were proposed by

force is δresand the target displacement of the ith cycle is δi. The per- Saatcioglu and Razvi (1992) for low compressive strength concrete
manent plastic deformation was much more pronounced for the sub- and employed using the Concrete02 command. When the retrofit-
standard columns and increased quickly. Because the S1 and S2 ted Specimens G1, GB1, G2, and GB2 were modeled, it was as-
columns could survive until 1.5% and 0.5% drift ratios, the residual sumed that the cover concrete was confined by the external
displacement at these specific drifts could be compared. For instance, jacket. The behavior was modeled by considering the stress–strain
at a 1.5% drift ratio, the residual displacement ratios for G1, GB1, and model that was proposed by Triantafillou et al. (2006). The dual
C1 were 66%, 50%, and 55% less than S1, respectively. Similarly, at confinement effect for the core concrete that was provided by the
a 0.5% drift ratio, these values for G2, GB2, and C2 were 68%, 63%, external jacket and stirrups was considered based on the model
and 55% less than S2, respectively. This was due to the efficient con- that was proposed by Ilki et al. (2008). For modeling the steel
finement that stemmed from either a sufficient transverse reinforce- bars, the reinforcement tensile test data were used to implement
ment ratio (C1 and C2) or sprayed GFRM jackets that restricted the uniaxialMaterial Hysteretic model in OpenSees, which consid-
buckling of the longitudinal reinforcing bars and spalling of the con- ers strain hardening. Because the substandard specimen bars suf-
crete cover. This decreased the damage intensity, as shown in Figs. 7 fered from buckling due to the widely spaced stirrups, the
and 8 (G1, G2, GB1, and GB2). As shown in this figure, the retrofit- buckling model that was proposed by Dhakal and Maekawa
ting method significantly reduced the residual displacement com- (2002) was implemented. In this model, the tension part of the en-
pared with substandard columns. Low residual displacement in velope curve is divided into: (1) elastic range; (2) yield plateau; and
retrofitted columns enhances structural integrity and safety by mini- (3) hardening zone. The compression envelope is composed of a
mizing permanent deformations after seismic events, which reduces linear elastic range followed by a nonlinear buckling model that
the need for extensive repairs. This ensures that buildings remain op- was formulated by the unsupported length of the longitudinal bar
erational postearthquake, which lowers maintenance costs and im- in height [stirrup spacing (Lu)], db, and fyl as variables to calculate
proves overall resilience. the compressive stress. The slenderness ratio could be calculated by
dividing the unsupported length (Lu) by the diameter of the longi-
tudinal bar (db), and this value in the tested substandard columns
Energy Dissipation Capacity
was 15.6.
The energy dissipation capacity (EDC) is calculated based on the Fig. 14 shows the model that was developed in OpenSees with
sum of the enclosed area (shaded area) of the hysteresis curves the strain penetration into the footing. This figure shows the uniax-
up to failure (Fig. 9). The results revealed that C1, GB1, and G1 ialMaterial Pinching4 model that is associated with the zero-length
dissipated 3.17, 2.27, and 2.55 times more energy compared with element as well as the employed reversed cyclic stress–strain mod-
S1, respectively. These results for C2, GB2, and G2 were 4.62, els for the concrete and longitudinal reinforcing.
2.72, and 2.43, respectively, compared with S2. The achievement Of note, the ratios (Ve/Vr ) for the code-compliant, substandard,
of a higher EDC by the retrofitted columns demonstrates their ca- and retrofitted specimens were between 0.34 and 0.38, 0.62 and
pability to retain lateral load capacity at higher drift ratios com- 0.80, and 0.71 and 0.85, respectively (Table 1). Based on the
pared with substandard columns. shear strength model of Sezen et al. (2004) and the column failure
The experimental results show that although a high axial load mode classification of Zhu et al. (2007), columns with Ve/Vr > 1.0
and shear demand caused undesired brittle failure, a weakness of are expected to experience significant shear deformations that
substandard RC structures, there was a promising enhancement cause shear failure with diagonal shear cracks, and columns with
in the performance of GFRM-retrofitted RC columns under a con- Ve/Vr < 0.7 and 0.7 < Ve/Vr < 1.0 experience flexure and shear–flex-
siderably high axial load, which was due to the confinement that ure failure modes, respectively. In addition, the observed crack
was provided by the jackets. However, this improvement is less propagation in the tested columns, along with the insignificant
for the rectangular columns compared with the square columns proportion of shear displacement (Fig. 11), indicates that the
due to a decrement in the confinement efficiency at higher cross- failure was not governed by shear distress. Therefore, in the pro-
sectional aspect ratios. In addition, the smaller a/d in rectangular posed analytical model, the shear deformation contribution was
specimens resulted in more shear demand and, therefore, less effi- neglected.
ciency due to the external jacket confinement.
Modeling of Rotations due to Strain Penetration into the
Foundation
Analytical Modeling
In the proposed model, the slip displacement component that was
Numerical models of the columns were created using OpenSees induced by the longitudinal reinforcements’ slip in the column–
(Mazzoni et al. 2006). Columns were modeled by employing a foundation interface of the tested columns was considered, because
force-based nonlinear beam–column element that accounted for they were not negligible, as shown by the column tests (Fig. 11). A
geometrical nonlinearity with the converging of the force equilib- zero-length rotational element was connected in a parallel manner
rium in every step. Fiber-discretized sections on seven Gauss– to the nonlinear force-based beam–column element to include

© ASCE 04025022-14 J. Compos. Constr.

J. Compos. Constr., 2025, 29(4): 04025022

Downloaded from ascelibrary.org by "Indian Institute of Technology, Hyderabad" on 04/28/25. Copyright ASCE. For personal use only; all rights reserved.

(a)

(b) (c)

Fig. 14. (a) Modeling details that consider the slip proportion in the top displacement; (b) uniaxial Material Pinching 4 model; and (c) defined stress–
strain curve for concrete and reinforcement. Note: Sec. 7 = section 7 at which 7th integration point was assigned; and N, M = normal applied axial
load and bending moment.

rotations in the column–foundation interface that were induced by stress–strain curve of concrete until the stress level of 0.2fc
strain penetration. Then, using the uniaxialMaterial Pinching4
model, a moment–rotation spring was defined and associated Gfc 0.8fc
ε20 = − + εco (4)
with the zero-length element. The parameters for the moment–rota- 0.6fc LIP Ec
tion spring model of the specimens were determined using several
steps. First, theoretical tensile force values for the longitudinal bars where LIP = length of the softening integration point. Of note, the
were derived from the moment–curvature analysis for various mo- fracture energy in tension is defined as the required energy to be ab-
ment values. Next, to determine the longitudinal rebar slip that cor- sorbed to create a unit area of crack (Oshtolagh et al. 2023). How-
responded to each tensile force, the so-called good bond stress (τb)– ever, it is accepted for localization in compression (Coleman and
slip curve given in the Fib Model Code (fib 2010) was selected Spacone 2001). In Eq. (4), the compression fracture energy (Gfc )
using a trial-and-error process. Then, this curve was applied to is defined as the area under the postpeak stress–strain curve of con-
19 joint link elements that are associated with a single bar in crete compressive strength, which is from the peak stress (fc ) until
SAP 2000 (CSI 2010). A nonlinear static displacement-controlled the end of the softening branch (0.2fc). Therefore, by calibrating the
load case was defined to the bar’s end node, which was in the col- ε20 (strain that corresponds to 0.2fc in the descending branch), a
umn–foundation interface to implement the tensile analysis by pull- constant energy release was ensured. Therefore, the numerical
ing it out to obtain the appropriate force–slip curve. The other end model’s postpeak behavior improved, as shown in Fig. 15, and a
of the bar was fixed using constraints in all degrees of freedom due good agreement was achieved with the experimental result.
to sufficient anchorage that was provided by enough development As shown in Fig. 16(a), the inclusion of buckling in bars refined
length. Then, by dividing the obtained bar slip by the neutral axis the postpeak response of the substandard columns. In addition,
depth, the rotation of the column due to the longitudinal bar slip Figs. 16(b and c) show the example recorded stress–strain curve
was calculated. Finally, with the rotation and corresponding mo- of longitudinal reinforcement for the S1 and G1 specimens. Ac-
ment values at each step, the moment–rotation curve for the rota- cording to the experiments, the reinforcing bar could reach a
tional spring was obtained. yield plateau in the retrofitted and code-compliant columns; how-
ever, they buckled under severe axial stresses in the substandard
columns. Figs. 16(b and c) show that they agreed well with the ob-
Employed Regularization Method and Numerical served behavior in the tests. As shown in Fig. 16(b), the bar buck-
Refinements led severely after reaching its peak stress. However, as shown in
Fig. 16(c), the bar did not buckle until larger drift ratios and strains
Strain softening due to localization in force-based elements during were experienced.
the postpeak response occurred in S1. Therefore, the postpeak Of note, the modeling of substandard RC columns that have ma-
load–displacement of the numerical model had a steeper descend- terial properties out of the engineering range (e.g., a compressive
ing branch than the experimental result. Coleman and Spacone concrete strength of less than 12 MPa) is extraordinarily cumber-
(2001) proposed two methods of regularization to obtain the objec- some. Therefore, capturing all occurrences that were witnessed dur-
tive response in the postpeak zone: (1) constant fracture energy cri- ing the test was challenging. Considering buckling of the
terion on the element force–displacement level; and (2) geometric longitudinal reinforcements, solving the strain and curvature local-
scaling on the local moment–curvature level. Of note, the number ization at the first integration point as well as accounting for the
and location of the integration points affect the overall section cur- longitudinal rebar slip contribution to the top displacement are
vature demand and element response. In addition, they provided an some superiorities (e.g., ability to capture key damage mechanisms
equation to control the linear postpeak softening branch of the or failure modes, strain penetration, and reinforcement buckling in

© ASCE 04025022-15 J. Compos. Constr.

J. Compos. Constr., 2025, 29(4): 04025022

 four different parameters that include columns controlled by: (1)
flexure; (2) shear; (3) insufficient splice or development length;
and (4) axial force that exceeds 0.7fcbh. The tested substandard col-
umns in this study met the last criterion (4), in which there are two
subdivisions that include: (1) conforming hoops over the entire
length; and (2) all other cases. Substandard columns are classified
by the second subdivision (2). According to this subdivision, all
given limit states are zero, which indicates that these columns
(a) with a low transverse reinforcement ratio and high axial load
ratio (S1 and S2) have no plastic rotation capacity. However, the
Downloaded from ascelibrary.org by "Indian Institute of Technology, Hyderabad" on 04/28/25. Copyright ASCE. For personal use only; all rights reserved.

results showed that S1 and S2 have a limited plastic rotation capac-

ity of 0.012 and 0.004.
Collapse prevention [GO in TBEC (AFAD 2018) and NC in
EC8-3, CEN 2005] corresponds to a level at which the load bearing
elements of the building experience severe damage just before the
partial or total collapse. Controlled damaged [KH in TBEC (AFAD
2018) and significant damage (SD) in EC8-3, CEN 2005] corre-
sponds to a limit state in which the load bearing elements of the
building have no severe damage and could mainly be retrofitted.
Limited damage [SH in TBEC (AFAD 2018) and damage limita-
(b) tion (DL) in EC8-3, CEN 2005] corresponds to a limit state in
which the load bearing elements of the building have very limited
Fig. 15. Utilization of regularization method in S1: (a) calibrating the damage. The GO and NC levels, according to the TBEC (AFAD
concrete stress–strain by compression fracture energy; and (b) revised 2018) and EC8-3 (CEN 2005), are shown in Fig. 6, respectively.
postpeak load–displacement curve. Hereafter, all these limits will be stated as GO for brevity. The re-
sults showed that EC8-3 (CEN 2005) remained unconservative for
the substandard columns, as shown in Figs. 6(c and d). For in-
substandard columns) of the developed model. Therefore, this type stance, Column S2 fails at a 0.5% drift ratio, although EC8-3
of model is required to evaluate and assess the seismic performance (CEN 2005) predicts the GO level as 0.8%, which is 60% higher
of substandard existing columns. than the test result. The TBEC (AFAD 2018) showed more conser-
vative predictions for the substandard columns GO level earlier
than failure. For the retrofitted columns, TBEC (AFAD 2018)
Performance-Based Assessment of Code-Compliant
and EC8-3 (CEN 2005) remained conservative when the GO per-
(Complying with the TSDC), Substandard, and
formance limit was predicted.
GFRM-Retrofitted Columns According to TBEC and EC8-3 The rotational performance limits of all columns based on
The aim of this section is to evaluate the success of the codes during EC8-3 (CEN 2005) and the TBEC (AFAD 2018) are presented
the performance assessment of the tested columns according to the in Table 6. The rotational limit states given as superscripts in
given limit criteria. In ASCE 41-23 (ASCE 2023) and ACI 369.1 Table 6 correspond to DL, SD, and NC based on EC8-3 (CEN
(ACI 2022), the column limit states categorization are based on 2005). Here, θd represents the demand rotation of the specimen
that corresponds to the 0.8 of maximum bending moment in the de-
scending branch of the moment–rotation curve. Of note, the perfor-
mance limit (PL) is defined by EC8-3 (CEN 2005) and TBEC
(AFAD 2018).

Preliminary Design Recommendations

To enhance the structural performance of retrofitted substandard
RC columns with sprayed GFRMs, several key recommendations
are proposed. The GFRM layer should have a minimum thickness
(a) of 30 mm to ensure proper confinement and structural integrity. Be-
fore application, the edges of the columns must be rounded off with
a radius of approximately 25 mm to mitigate stress concentrations
and avoid premature cracking. A glass fiber content of at least 5%
(by weight) in the GFRM mix is necessary to achieve the desired
tensile strength, particularly when other reinforcement materials,
such as basalt mesh, are not integrated. Although the inclusion of
mesh reinforcement, such as basalt mesh, does not significantly en-
hance the strength, it effectively increases ductility and improves
the EDC, which makes it suitable for applications that require
(b) (c) higher ductility. It is recommended that the height of the column
be retrofitted rather than just the plastic hinge zone to prevent dam-
Fig. 16. Showing: (a) comparison between buckling inclusion in S1;
age propagation and to ensure uniform confinement under seismic
(b) reinforcing bar stress–strain curves of S1; and (c) G1.
loading. The corners of the GFRM layer should be rounded with an

© ASCE 04025022-16 J. Compos. Constr.

J. Compos. Constr., 2025, 29(4): 04025022

 Table 6. Experimental chord rotations that correspond to limit states based on EC8-3 (CEN 2005) and TBEC (AFAD 2018)
EC8-3 TBEC

Specimen θd θ DL θ SD θ NC PL θ SH θ KH θ GO PL
S1 0.012 0.0038 0.008 0.011 NC 0 0.003 0.004 GO
C1 0.033 0.0033 0.017 0.022 NC 0 0.009 0.012 GO
G1 0.024 0.0039 0.008 0.011 NC 0 0.004 0.006 GO
GB1 0.028 0.0039 0.008 0.011 NC 0 0.004 0.006 GO
S2 0.004 0.0032 0.007 0.009 DL 0 0.002 0.003 GO
C2 0.021 0.0032 0.011 0.015 NC 0 0.008 0.010 GO
G2 0.016 0.0033 0.007 0.009 NC 0 0.003 0.005 GO
Downloaded from ascelibrary.org by "Indian Institute of Technology, Hyderabad" on 04/28/25. Copyright ASCE. For personal use only; all rights reserved.

GB2 0.016 0.0033 0.007 0.009 NC 0 0.003 0.005 GO

Note: θ DL, θSH = rotational limit state corresponding to damage limitation and limited damage; θSD, θKH = rotational limit state corresponding to significant
damage and controlled damage; θNC, θGO = rotational limit state corresponding to near collapse and collapse prevention; DL, SH = identical limit states with
different naming style in two mentioned codes; and SD, KH, NC, GO = equivalent limit states in two different codes.

oval profile to promote uniform stress distribution and prevent lo- displacement that stemmed from rotational slip (bond degrada-
calized failures. tion) at the column–foundation interface due to strain penetra-
tion (24%–51%).
3. The achieved displacement ductilities for the retrofitted speci-
mens were in the range of μD = 4.2–5.4 for the G and GB
Conclusions specimens. The ductility of G1 and GB1 increased up to 80%
compared with the substandard Specimen S1, which ap-
Eight full-scale RC columns with varying cross-sectional aspect
proached the 94% increase in ductility that was exhibited by
and a/d ratios were tested under a constant axial load and reversed
C1 with respect to S1. Similarly, the ductility of the
cyclic lateral displacements. A novel retrofitting method was im- GFRM-retrofitted rectangular columns increased up to 62%
plemented using a GFRM with a 5% glass content as well as a compared with their substandard counterpart S2, with a lower
GFRM with a 3.5% glass content that was reinforced with three a/d and higher shear demand. Of note, GB1 and GB2 showed
layers of basalt mesh. This retrofitting significantly enhanced the slightly higher ductility increments of 30% and 27% compared
performance of substandard columns under high axial load ratios with G1 and G2, respectively.
(0.75) and shear demand (Ve/Vr of 0.71 and 0.80), which brought 4. The numerical model was validated accurately against experi-
their performance closer to that of code-compliant columns. The mental results that reproduced the hysteretic lateral load–
experimental results were discussed comparatively, which focused displacement curves and incorporated the main failure mecha-
on key behavioral metrics, such as strength, stiffness, energy dissi- nisms that were observed in the tests, such as longitudinal bar
pation, and observed failure mechanisms. An accurate numerical buckling and slip due to strain penetration at the column–foot-
model that incorporates the nonlinear effects, which includes strain ing interface.
penetration and buckling of longitudinal rebars in OpenSees, was 5 The rotational plastic demand of the retrofitted specimens in-
developed, and global hysteretic responses (e.g., ultimate displace- creased compared with the substandard ones. In addition, the
ment and lateral load capacity) were validated against the experi- success of the codes when evaluating the rotational plastic de-
mental data. The success of EC8 (CEN 2005) and TBEC (AFAD mand of the specimens was scrutinized across the three col-
2018) in estimating the performance limits of columns was as- umns. The results indicated that the plastic rotation demand of
sessed. Based on the experimental and modeling results, the fol- substandard columns could be properly predicted by TBEC
lowing major conclusions are drawn. (AFAD 2018), although EC8-3 (CEN 2005) overestimated the
1. The S1 and S2 columns showed a brittle failure mode that was plastic rotation demand by 60% compared with the test results
characterized by cover spalling, concrete crushing, and longitu- of S2. This discrepancy was attributed to the high axial
dinal bar buckling, which highlights the necessity of retrofitting load-to-capacity ratio of 0.75 in the substandard columns,
to mitigate the lack of confinement. Retrofitted square columns which requires more careful consideration in EC8-3 (CEN
(G1 and GB1) with a higher a/d ratio of 3.81 and lower shear 2005). In addition, ASCE 41-23 considers no plastic rotation ca-
demand (Ve/Vr of 0.72) exhibited a 35% increase in lateral pacity for S1 and S2, despite the fact that these columns demon-
load bearing capacity compared with S1 (Ve/Vr of 0.62). The lat- strated a limited but observable plastic rotation capacity.
eral load bearing capacity of C1 (Ve/Vr of 0.38) was 32% higher
than that of S1. For rectangular retrofitted columns (G2 and
GB2) with a lower a/d of 2.21 and higher shear demand (Ve/
Vr of 0.85), the lateral load capacity increased by 16% compared Data Availability Statement
with S2 (Ve/Vr of 0.80); however, C2 (Ve/Vr of 0.34) had a lat-
eral load capacity 36% higher than that of S2. The greater effi- All data that support the findings of this study are available from the
ciency of retrofitting in square columns was attributed to the corresponding author upon reasonable request.
20% lower shear demand due to the higher a/d ratio and the ad-
verse relationship between the confinement efficiency and
cross-sectional aspect ratio. Acknowledgments
2. Despite the relatively high shear demand in substandard col-
umns, the top displacement was primarily governed by flexural The authors are most thankful to the Fibrobeton company and its
displacement, along with significant contributions from board member, Muhammed Marasli, for their continuous support

© ASCE 04025022-17 J. Compos. Constr.

J. Compos. Constr., 2025, 29(4): 04025022

 from the construction phase of the specimens to the testing stage. In εco = concrete strain that corresponds to peak stress;
addition, the support of ITU is appreciated. The contributions of εc85 = concrete strain that corresponds to 85% stress in
Furkan Narlitepe, Merve Nur Demir, Alihan Baltaci, Bilal Sari, descending branch;
Muhsin Can Luleci, Safiye Gundogan, Dr. Erkan Tore, and the εul = ultimate strain of longitudinal rebar;
summer trainees are gratefully acknowledged. This research was εuw = ultimate strain of transverse rebar;
granted in the scope of the bilateral collaboration between the re- εyl = yield strain of longitudinal rebar;
search councils of Türkiye (TUBITAK) and the UK (RCUK) εyw = yield strain of transverse rebar;
under Grant No. 216M379. ε20 = concrete strain that corresponds to 20% of peak stress
in descending branch;
θd = demand rotation of column;
θslip = rotation due to slip at bottom of the column;
Downloaded from ascelibrary.org by "Indian Institute of Technology, Hyderabad" on 04/28/25. Copyright ASCE. For personal use only; all rights reserved.

Notation
μ = average;
The following symbols are used in this paper: μD = displacement ductility;
A = cross-sectional area; νc = concrete Poisson’s ratio;
As = total area of longitudinal rebars; ρsh = volumetric transverse reinforcement ratio;
Ash = area of the transverse reinforcement alongside the ρsh,min = minimum required volumetric transverse
lateral load; reinforcement ratio;
a = shear span; σ = standard deviation;
b = column dimension (width); τb = bond stress; and
bk = distance between the centroid of the extreme stirrups φaverage = average curvature.
leg;
d = effective depth of cross section;
d0 = distance from the first LVDT to bottom of the column;
db = diameter of longitudinal bar; References
dl = measured left displacement;
dr = measured right displacement; ACI (American Concrete Institute). 2013. Guide for testing reinforced con-
Ec = elastic modulus of concrete; crete structural elements under slowly applied simulated seismic loads.
Es = the slope of compression stress-strain curve of ACI 374.1. Farmington Hills, MI: ACI.
reinforcement in descending branch after buckling; ACI (American Concrete Institute). 2019. Building code requirements for
Ets = the slope of tensile stress-strain curve of concrete in structural concrete. ACI 318. Farmington Hills, MI: ACI.
ACI (American Concrete Institute). 2020. Guide to design and construction
descending branch;
of externally bonded fabric-reinforced cementitious matrix and steel-
fc = concrete compressive strength;
reinforced grout systems for repair and strengthening concrete struc-
ful = ultimate stress of longitudinal rebar; tures. ACI 549. Farmington Hills, MI: ACI.
fuw = ultimate stress of transverse rebar; ACI (American Concrete Institute). 2022. Seismic evaluation and retrofit of
fyl = yield stress of longitudinal rebar; existing concrete buildings—Code and commentary. ACI 369.
fyw = yield stress of transverse rebar; Farmington Hills, MI: ACI.
G = shear modulus; AFAD (Ministry of Interior Disaster and Emergency Management
Gfc = fracture energy in compression; Authority). 2018. Turkish building earthquake code. Ankara,
h = column dimension (height); Türkiye: AFAD.
Ki = peak-to-peak ith cycle secant stiffness; Ali, M. A., and E. El-Salakawy. 2016. “Seismic performance of
LG = height of the measurement segment; GFRP-reinforced concrete rectangular columns.” J. Compos. Constr.
Lh = horizontal distance between two LVDTs; 20 (3): 04015074. https://doi.org/10.1061/(ASCE)CC.1943-5614
LIP = length of softening integration point; .0000637.
Ansal, A., A. Akinci, G. Cultrera, M. Erdik, V. Pessina, G. Tönük, and G.
Lu = unsupported length of longitudinal bar in height
Ameri. 2009. “Loss estimation in Istanbul based on deterministic earth-
(stirrup spacing);
quake scenarios of the Marmara Sea region (Turkey).” Soil Dyn.
Lv = distance from actuator to middle of the first gauge Earthquake Eng. 29 (4): 699–709. https://doi.org/10.1016/j.soildyn
length; .2008.07.006.
l = clear height of the column to horizontal load (shear ASCE. 2023. Seismic evaluation and retrofit of existing buildings.
span); ASCE41-23. Reston, VA: ASCE.
lp = plastic hinge length; ASTM. 2021a. Standard test method for compressive strength of cylindri-
Mw = moment magnitude; cal concrete specimens. ASTM C39/C39M-21. West Conshohocken,
s = stirrup spacing; PA: ASTM.
Ve = shear force that corresponds to the bending moment; ASTM. 2021b. Standard test method for compressive strength of hydraulic
Vr = shear capacity of cross section; cement mortars (using 2-in. or [50-mm] cube specimens). ASTM
Δslip = contribution of slip displacement component; C109-21. West Conshohocken, PA: ASTM.
Δt = total displacement; ASTM. 2022a. Standard specification for deformed and plain carbon-steel
bars for concrete reinforcement. ASTM A615/A615M-22. West
Δflexure = The contribution of flexural displacement;
Conshohocken, PA: ASTM.
δi = target displacement of ith cycle; ASTM. 2022b. Standard test method for static modulus of elasticity and
δres = residual lateral displacement at zero lateral force after Poisson’s ratio of concrete in compression. ASTM C469/C469M.
unloading; West Conshohocken, PA: ASTM.
δ+u = ultimate displacement in pushing; Ates, A. O. 2022. “Improvement of seismic performance of reinforced con-
δ−u = ultimate displacement in pulling; crete columns using glass fiber reinforced sprayed mortar with/without
δ+y = yield displacement in pushing; textile reinforcement.” Ph.D. thesis, Dept. of Civil Engineering,
δ−y = yield displacement in pulling; Istanbul Technical Univ.

© ASCE 04025022-18 J. Compos. Constr.

J. Compos. Constr., 2025, 29(4): 04025022

 Ates, A. O., G. Durmuş , and A. Ilki. 2023a. “Tensile and flexural behaviors composites.” Constr. Build. Mater. 27 (1): 206–217. https://doi.org/10
of textile reinforced sprayed glass fiber mortar composites.” Materials .1016/j.conbuildmat.2011.07.058.
16 (12): 4251. https://doi.org/10.3390/ma16124251. Del Zoppo, M., M. Di Ludovico, A. Balsamo, and A. Prota. 2019.
Ates, A. O., S. Hajihosseinlou, A. Nasrinpour, C. Demir, M. Comert, M. “Experimental in-plane shear capacity of clay brick masonry panels
Marasli, and A. Ilki. 2023b. “Seismic retrofit of substandard RC col- strengthened with FRCM and FRM composites.” J. Compos. Constr.
umns using sprayed glass fiber-reinforced mortar and basalt textile rein- 23 (5): 04019038. https://doi.org/10.1061/(ASCE)CC.1943-5614.0000965.
forcement.” J. Earthquake Eng. 28 (6): 1615–1642. https://doi.org/10 Demir, U., M. Ispir, Y. Sahinkaya, G. Arslan, and A. Ilki. 2019. “"Axial
.1080/13632469.2023.2247491. behavior of noncircular high-performance fiber-reinforced cementitious
Ates, A. O., S. Khoshkholghi, E. Tore, M. Marasli, and A. Ilki. 2019. composite members externally jacketed by CFRP sheets.”.” J. Compos.
“Sprayed glass fiber–reinforced mortar with or without basalt textile re- Constr. 23 (4): 04019022. https://doi.org/10.1061/(ASCE)CC.1943
inforcement for jacketing of low-strength concrete prisms.” J. Compos. -5614.0000940.
Deng, M., Y. Zhang, and Q. Li. 2018. “Shear strengthening of RC short
Downloaded from ascelibrary.org by "Indian Institute of Technology, Hyderabad" on 04/28/25. Copyright ASCE. For personal use only; all rights reserved.

Constr. 23 (2): 04019003. https://doi.org/10.1061/(ASCE)CC.1943

-5614.0000922. columns with ECC jacket: Cyclic behavior tests.” Eng. Struct. 160:
Awani, O., T. El-Maaddawy, and N. Ismail. 2017. “Fabric-reinforced ce- 535–545. https://doi.org/10.1016/j.engstruct.2018.01.061.
mentitious matrix: A promising strengthening technique for concrete Dhakal, R. P., and K. Maekawa. 2002. “Path-dependent cyclic stress–strain
structures.” Constr. Build. Mater. 132: 94–111. https://doi.org/10 relationship of reinforcing bar including buckling.” Eng. Struct. 24 (11):
.1016/J.CONBUILDMAT.2016.11.125. 1383–1396. https://doi.org/10.1016/S0141-0296(02)00080-9.
Aydogdu, H. H., C. Demir, T. Kahraman, and A. Ilki. 2023. “Evaluation of Di Ludovico, M., G. M. Verderame, A. Prota, G. Manfredi, and E.
rapid seismic safety assessment methods on a substandard reinforced Cosenza. 2014. “Cyclic behavior of nonconforming full-scale RC col-
concrete building stock in Istanbul.” Structures 56: 104962. https:// umns.” J. Struct. Eng. 140 (5): 04013107. https://doi.org/10.1061/
doi.org/10.1016/J.ISTRUC.2023.104962. (ASCE)ST.1943-541X.0000891.
Aydogdu, H. H., and A. Ilki. 2023. “Developments in seismic safety of RC Fakharifar, M., G. Chen, M. Arezoumandi, and M. ElGawady. 2015.
building stock of Istanbul according to the years and seismic design “Hybrid jacketing for rapid repair of seismically damaged reinforced
codes.” In Vol. 350 of Proc., Building for the Future: Durable, concrete columns.” Transp. Res. Rec. 2522 (1): 70–78. https://doi.org/
Sustainable, Resilient. fib Symp. 2023. Lecture Notes in Civil 10.3141/2522-07.
Engineering, edited by A. Ilki, D. Çavunt, and Y. S. Çavunt, 879– Faleschini, F., M. A. Zanini, L. Hofer, K. Toska, D. De Domenico, and C.
888. Cham, Switzerland: Springer. Pellegrino. 2020. “Confinement of reinforced concrete columns with
Bae, S., and O. Bayrak. 2008. “Plastic hinge length of reinforced concrete glass fiber reinforced cementitious matrix jackets.” Eng. Struct. 218:
columns.” ACI Struct. J. 105 (3): 290. 110847. https://doi.org/10.1016/J.ENGSTRUCT.2020.110847.
Bournas, D. A., T. C. Triantafillou, K. Zygouris, and F. Stavropoulos. Farrokh Ghatte, H., M. Comert, C. Demir, M. Akbaba, and A. Ilki. 2019.
2009. “Textile-reinforced mortar (TRM) versus FRP jacketing in seis- “Seismic retrofit of full-scale substandard extended rectangular RC col-
mic retrofitting of RC columns with continuous or lap-spliced deformed umns through CFRP jacketing: Test results and design recommenda-
bars.” J. Compos. Constr. 13 (5): 360–371. https://doi.org/10.1061/ tions.” J. Compos. Constr. 23 (1): 04018071. https://doi.org/10.1061/
(ASCE)CC.1943-5614.0000028. (ASCE)CC.1943-5614.0000907.
Bousias, S. N., T. C. Triantafillou, M. N. Fardis, L. Spathis, and B. A. fib (Fédération Internationale du Béton). 2010. Fib model code 2010. fib
O’Regan. 2004. “Fiber-Reinforced polymer retrofitting of rectangular MC. Lausanne, Switzerland: fib.
Gonzalez-Libreros, J., M. A. Zanini, F. Faleschini, and C. Pellegrino. 2019.
reinforced concrete columns with or without corrosion.” ACI Struct.
“Confinement of low-strength concrete with fiber reinforced cementi-
J. 101 (4): 512–520. https://doi.org/10.14359/13337.
tious matrix (FRCM) composites.” Composites, Part B 177: 107407.
Celep, Z., A. Erken, B. Taskin, and A. Ilki. 2011. “Failures of masonry and
https://doi.org/10.1016/J.COMPOSITESB.2019.107407.
concrete buildings during the March 8, 2010 Kovancılar and Palu
Gurbuz, T., A. Cengiz, S. Kolemenoglu, C. Demir, and A. Ilki. 2023.
(Elazığ ) Earthquakes in Turkey.” Eng. Fail. Anal. 18 (3): 868–889.
“Damages and failures of structures in İ zmir (Turkey) during the
https://doi.org/10.1016/J.ENGFAILANAL.2010.11.001.
October 30, 2020 Aegean Sea Earthquake.” J. Earthquake Eng.
Celik, O. C., M. B. C. Ulker, C. Gocer, S. Guntepe, O. Koz, M. M.
27 (6): 1565–1606. https://doi.org/10.1080/13632469.2022.2086186.
Eyupgiller, M. Aksu, and A. Yaz. 2024. “Multidisciplinary reconnais-
Ilki, A., K. Darilmaz, I. Bakan, M. Zorbozan, E. Yuksel, H. Saruhan, and F.
sance investigation covering structural, geotechnical, and architectural
Karadogan. 1998. “Jacketing of prefabricated columns.” In Proc., 2nd
based damage to mid-rise residential buildings following the February
Japan-Turkey Workshop Earthquake Engineering, 329–336, Istanbul,
6th, 2023 Kahramanmaraş , Türkiye Earthquake Doublets (Mw 7.8,
Türkiye: Istanbul Technical University and Japan Society of Civil
Mw 7.6).” Soil Dyn. Earthquake Eng. J. 182: 108738. https://doi.org/
Engineers.
10.1016/j.soildyn.2024.108738. Ilki, A., and N. Kumbasar. 2001. “Cyclic behavior of concrete members re-
CEN (European Committee for Standardization). 2005. Design of struc- habilitated and strengthened by using carbon fiber sheets.” In Vol. 2 of
tures for earthquake resistance—Part 1–3: Strengthening and repair Proc., 2nd Int. Conf. Materials Science, 483–492, Montreal/Tokyo:
of buildings. European prEN 1998-1-3. Eurocode 8. Brussels, CSCE/JSCEE.
Belgium: CEN. Ilki, A., N. Kumbasar, and V. Koc. 2004. “Low strength concrete members
CNR (Italian National Research Council). 2018. Guide for the design and externally confined with FRP sheets.” Struct. Eng. Mech. 18 (2): 167–
construction of externally bonded fibre reinforced inorganic matrix 194. https://doi.org/10.12989/SEM.2004.18.2.167.
systems for strengthening existing structures. CNR-DT 215. Rome: Ilki, A., O. Peker, E. Karamuk, C. Demir, and N. Kumbasar. 2008. “FRP
CNR. retrofit of low and medium strength circular and rectangular reinforced
Colajanni, P., F. De Domenico, A. Recupero, and N. Spinella. 2014. concrete columns.” J. Mater. Civ. Eng. 20 (2): 169–188. https://doi.org/
“Concrete columns confined with fibre reinforced cementitious mortars: 10.1061/(ASCE)0899-1561(2008)20:2(169).
Experimentation and modelling.” Constr. Build. Mater. 52: 375–384. Inel, M., S. M. Senel, and H. Un. 2008. “Experimental evaluation of con-
https://doi.org/10.1016/J.CONBUILDMAT.2013.11.048. crete strength in existing buildings.” Mag. Concr. Res. 60 (4): 279–89.
Coleman, J., and E. Spacone. 2001. “Localization issues in force-based https://doi.org/10.1680/macr.2007.00091.
frame elements.” J. Struct. Eng. 127 (11): 1257–1265. https://doi.org/ Iş ık, E., et al. 2023. “Structural damages in masonry buildings in Adıyaman
10.1061/(ASCE)0733-9445(2001)127:11(1257). during the Kahramanmaraş (Turkiye) earthquakes (Mw 7.7 and Mw
CSI (Computers and Structures). 2010. Integrated software for structural 7.6) on 06 February 2023.” Eng. Fail. Anal. 151: 107405. https://doi
analysis and design-application programming interface. .org/10.1016/J.ENGFAILANAL.2023.107405.
SAP2000-OAPI. Berkeley, CA: CSI. Kian, N., U. Demir, C. Demir, M. Marasli, and A. Ilki. 2022. “Seismic per-
Dai, J. G., L. Lam, and T. Ueda. 2012. “Seismic retrofit of square RC col- formance of substandard RC columns retrofitted with sprayed GFRM.”
umns with polyethylene terephthalate (PET) fibre reinforced polymer In Proc., 10th Int. Conf. on FRP Composites in Civil Engineering.

© ASCE 04025022-19 J. Compos. Constr.

J. Compos. Constr., 2025, 29(4): 04025022

 Lecture Notes in Civil Engineering, edited by A. Ilki, M. Ispir, and P. Priestley, M. J. N. 2000. “Performance based seismic design.” In Proc.,
Inci, 1317–1328. Cham, Switzerland: Springer. 12th World Conf. on Earthquake Engineering. Wellington, New
Kian, N., M. Farzam, and M. Rezaie Oshtolagh. 2021. “Carbon fiber rein- Zealand: New Zealand Society for Earthquake Engineering.
forced polymer strengthened reinforced concrete square columns under Pujol, S. 2002. “Drift capacity of reinforced concrete columns subjected to
pre-existing eccentric loads.” Adv. Struct. Eng. 24 (15): 3420–3432. displacement reversals.” Ph.D. thesis, Dept. of Civil Engineering,
https://doi.org/10.1177/13694332211029730. Purdue Univ.
Kian, N., A. Ilki, and O. C. Celik. 2024. “Numerical study on inelastic Raoof, S. M., and D. A. Bournas. 2017. “Bond between TRM versus FRP
cyclic behavior of CFRP wrapped substandard RC columns.” composites and concrete at high temperatures.” Composites Part B 127:
Procedia Struct. Integrity 64: 1049–1056. https://doi.org/10.1016/j 150–165. https://doi.org/10.1016/J.COMPOSITESB.2017.05.064.
.prostr.2024.09.398. Raza, S., M. K. Khan, S. J. Menegon, H. H. Tsang, and J. L. Wilson. 2019.
Lu, Y., S. Yi, H. Liang, T. Gong, and N. Li. 2019. “Seismic behavior of RC “Strengthening and repair of reinforced concrete columns by jacketing:
Downloaded from ascelibrary.org by "Indian Institute of Technology, Hyderabad" on 04/28/25. Copyright ASCE. For personal use only; all rights reserved.

square columns strengthened with self-compacting concrete-filled State-of-the-art review.” Sustainability 11 (11): 3208. https://doi.org/10
CFRP-steel tubes.” J. Bridge Eng. 24 (2): 04018119. https://doi.org/ .3390/su11113208.
10.1061/(ASCE)BE.1943-5592.0001345. Rodrigues, H., A. Arêde, H. Varum, and A. G. Costa. 2013. “Experimental
Luleci, M. C., B. Sari, U. Demir, M. Marasli, and A. Ilki. 2021. “Seismic be- evaluation of rectangular reinforced concrete column behaviour under
havior of substandard extended rectangular RC columns jacketed with biaxial cyclic loading.” Earthquake Eng. Struct. Dyn. 42 (2): 239–
sprayed GFRM.” In COMPDYN Proc., 8th Int. Conf. on Computational 259. https://doi.org/10.1002/EQE.2205.
Methods in Structural Dynamics and Earthquake Engineering, 111– Saatcioglu, M., and S. R. Razvi. 1992. “Strength and ductility of confined
119, Athens, Greece: ECCOMAS Proceedia. concrete.” J. Struct. Eng. 118 (6): 1590–1607. https://doi.org/10.1061/
Mazzoni, S., F. McKenna, M. H. Scott, and G. L. Fenves. 2006. “Opensees (ASCE)0733-9445(1992)118:6(1590).
command language manual.” Pac. Earthquake Eng. Res. Center Seifi, A., A. Hosseini, M. S. Marefat, and M. Khanmohammadi. 2018.
264 (1): 137–158. “Seismic retrofitting of old-type RC columns with different lap splices
Mezrea, P. E., M. Ispir, I. A. Balci, I. E. Bal, and A. Ilki. 2021. “Diagonal by NSM GFRP and steel bars.” Struct. Des. Tall Spec. Build. 27 (2):
tensile tests on historical brick masonry wallets strengthened with fabric e1413. https://doi.org/10.1002/TAL.1413.
Sezen, H., M. Asce, and J. P. Moehle. 2004. “Shear strength model
reinforced cementitious mortar.” Structures 33: 935–946. https://doi
for lightly reinforced concrete columns.” J. Struct. Eng. 130 (11):
.org/10.1016/J.ISTRUC.2021.04.076.
1692–1703. https://doi.org/10.1061/(ASCE)0733-9445(2004)130:11
Mezrea, P. E., I. A. Yilmaz, M. Ispir, E. Binbir, I. E. Bal, and A. Ilki. 2017.
(1692).
“External jacketing of unreinforced historical masonry piers with open-
Tapan, M., M. Comert, C. Demir, Y. Sayan, K. Orakcal, and A. Ilki. 2013.
grid basalt-reinforced mortar.” J. Compos. Constr. 21 (3): 04016110.
“Failures of structures during the October 23, 2011 Tabanlı (Van) and No-
https://doi.org/10.1061/(ASCE)CC.1943-5614.0000770.
vember 9, 2011 Edremit (Van) earthquakes in Turkey.” Eng. Fail. Anal.
Mirmiran, A., M. Shahawy, M. Samaan, H. E. Echary, J. C. Mastrapa, and
34: 606–628. https://doi.org/10.1016/J.ENGFAILANAL.2013.02.013.
O. Pico. 1998. “Effect of column parameters on FRP-confined con-
Tong, T., H. Lei, S. Yuan, and Z. Liu. 2020. “Experimental investigation
crete.” J. Compos. Constr. 2 (4): 175–185. https://doi.org/10.1061/
and seismic vulnerability assessment of low flexural strength rectan-
(ASCE)1090-0268(1998)2:4(175). gular bridge piers retrofitted with ultrahigh-performance concrete
Napoli, A., and R. Realfonzo. 2022. “Compressive behavior of masonry jackets.” Eng. Struct. 206: 110132. https://doi.org/10.1016/J
columns confined with FRCM systems: Research overview and analyt- .ENGSTRUCT.2019.110132.
ical proposals.” J. Compos. Constr. 26 (3): 04022019. https://doi.org/10 Tore, E., C. Demir, M. Comert, and A. Ilki. 2021. “Seismic collapse perfor-
.1061/(ASCE)CC.1943-5614.0001200. mance of a full-scale concrete building with lightly reinforced col-
Narlitepe, F., N. Kian, U. Demir, C. Demir, and A. Ilki. 2021. “Seismic umns.” J. Struct. Eng. 147 (12): 04021207. https://doi.org/10.1061/
strengthening of sub-standard rc columns through a novel hybrid thin (ASCE)ST.1943-541X.0003178.
jacketing method.” In Proc., 18th fib Symp. Concrete Structures: New Triantafillou, T. C., C. G. Papanicolaou, P. Zissimopoulos, and T.
Trends for Eco-Efficiency and Performance, 1223–1232, Lausanne, Laourdekis. 2006. “Concrete confinement with textile-reinforced mor-
Switzerland: International Federation for Structural Concrete (fib). tar jackets.” ACI Struct. J. 103 (1): 28–37. https://doi.org/10.14359/
Ortlepp, R., and S. Ortlepp. 2017. “Textile reinforced concrete for strength- 15083.
ening of RC columns: A contribution to resource conservation through TSDC (Turkish Government Ministry of Reconstruction and
the preservation of structures.” Constr. Build. Mater. 132: 150–160. Resettlement). 1975. Specifications for structures to be built in disaster
https://doi.org/10.1016/J.CONBUILDMAT.2016.11.133. areas. Turkish Seismic Design Code. Ankara, Turkey: Turkish
Oshtolagh, M. R., M. Farzam, N. Kian, and H. Sadaghian. 2023. “Behavior Government Ministry of Reconstruction and Resettlement,
of recycled steel fiber-reinforced concrete beams in torsion- experimental Earthquake Research Institute.
and numerical approaches.” Comput. Concr. 32 (2): 173–184. https://doi Verderame, G. M., G. Fabbrocino, and G. Manfredi. 2008. “Seismic re-
.org/10.12989/cac.2023.32.2.173. sponse of r.c. columns with smooth reinforcement. part II: Cyclic
Ozcan, O., B. Binici, and G. Ozcebe. 2010. “Seismic strengthening of rect- tests.” Eng. Struct. 30 (9): 2289–2300. https://doi.org/10.1016/J
angular reinforced concrete columns using fiber reinforced polymers.” .ENGSTRUCT.2008.01.024.
Eng. Struct. 32 (4): 964–973. https://doi.org/10.1016/j.engstruct.2009 Zhu, L., K. J. Elwood, and T. Haukaas. 2007. “Classification and seismic
.12.021. safety evaluation of existing reinforced concrete columns.” J. Struct.
Paulay, T., and M. J. N. Priestley. 1992. Seismic design of reinforced con- Eng. 133 (9): 1316–1330. https://doi.org/10.1061/(ASCE)0733
crete and masonry buildings, 768. New York: Wiley. -9445(2007)133:9(1316).

© ASCE 04025022-20 J. Compos. Constr.

J. Compos. Constr., 2025, 29(4): 04025022