Journal of Building Engineering 98 (2024) 110993

Contents lists available at ScienceDirect

Journal of Building Engineering

journal homepage: www.elsevier.com/locate/jobe

Compression behavior of square and circular SFRC columns

confined with external steel straps
Julian Carrillo a, * , Matías Hube b , Carlos Blandon c , Ramon Mata d ,
Joaquin Abellan-Garcia e
a
Depart. of Civil Engineering, Universidad Militar Nueva Granada, UMNG, Colombian Earthquake Engineering Research Network, CEER, Colombia
b
Depart. of Structural and Geotechnical Engineering, Concrete Innovation Hub UC, Pontificia Universidad Católica de Chile, Chile
c
Depart. of Civil Engineering, Universidad EIA, UEIA, Colombia
d
Depart. of Civil Engineering, Universidad San Sebastián, Concepción, Chile
e
Depart. of Civil and Environmental Engineering, Universidad Del Norte, Colombia

A R T I C L E I N F O A B S T R A C

Keywords: Steel-Strapping Tensioning-Techniques (SSTT) have been proposed recently as a confinement

Concrete confinement strategy to enhance the strength and deformation capacity of new and existing Reinforced Con­
active confinement crete (RC) columns. Main advantages of SSTT are the low cost, simple installation and modest
Steel-strapping tensioning-technique installation time. Existing methods to quantify the increase of the compressive strength and
SSTT (steel strapping tensioning technique) deformation capacity of actively confined concrete have been obtained considering conventional
Fiber reinforced concrete
deformed steel. Additionally, previous studies have observed that toughness and deformation
Circular hoops
capacity of concrete increase when adding steel fibers to the concrete matrix. However, the
combined effect of SSTT applied to Steel Fiber Reinforced Concrete (SFRC) columns has not been
assessed. Hence, the objective of this study is to evaluate the compressive behavior of SFRC
columns confined with SSTT. A total of 34 short columns with SFRC and plain concrete were
subjected to monotonic compression tests. The studied variables consider the cross-section of the
columns (square and circular), the steel fibers dosage (15, 30 and 60 kg/m3), and the spacing of
the steel straps (37.5, 50 and 75 mm). The obtained results demonstrate that the SSTT increased
the compressive strength between 4 % and 18 %, and the deformation capacity up to 800 %.
Additionally, the use of steel fibers was effective in further increasing the deformation capacity of
columns confined with SSTT. A numerical model based on a database of 150 tests, where 34 are
from this study, is also proposed to predict the compressive behavior of SFRC columns confined
with SSTT.

1. Introduction
One of the motivations of Earthquake Engineering is evaluating new materials and structural systems, aimed to identify novel
solutions that are adequate in terms of structural performance and economy. A novel technique that can be used to build or repair
concrete elements is steel straps installed externally around structural elements, which involves non-destructive works [1]. Steel straps
have been mostly used in the packaging sector to secure and protect goods during transport. However, in 1995 Frangous et al. [2]
proposed the Steel-strapping Tensioning-Technique (SSTT) for construction and rehabilitation of concrete columns. The SSTT consist

* Corresponding author.
E-mail address: [email protected] (J. Carrillo).

https://doi.org/10.1016/j.jobe.2024.110993
Received 1 September 2023; Received in revised form 25 September 2024; Accepted 7 October 2024
Available online 13 October 2024
2352-7102/© 2024 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
J. Carrillo et al. Journal of Building Engineering 98 (2024) 110993

of providing concrete confinement by means of steel straps that are installed around structural elements at discrete spacing. Nazirah
et al. [1], Sarmah et al. [3], Lim and Togay [4], Yooprasertchai et al. [5], among others, have observed that the SSTT increases the
strength and strain of concrete. Frangous et al. [2] emphasized the need to enhance the concrete strength and ductility using a low-cost
and fast-application technique, that allowed the continuous operation of the structure that did not require specialized labor. Nazirah
et al. [1] observed that actively confining straps enhance the brittle nature of concrete since the pre-tensioning stress depends on the
elastic and compressive strength of concrete.
Confinement stresses transverse to the axial loading direction increases the compressive strength and deformation capacity of
concrete [6,7]. Confinement can be applied in the concrete as active or passive mode. Passive confinement is achieved by using
embedded steel hoops or jacketing through fiber reinforced polymers (FRP) or fiber reinforced cementitious matrix (FRCM) that are
activated due to the Poisson effect when the concrete element is subjected to axial loads [8,9]. Active confinement is achieved by
applying an external confinement pressure [10,11,12]. Contrary to the case for passive confinement, the transverse stresses for active
confinement act from the beginning of the application of the axial load. The confining pressure applied throughout the loading phase of
active confinement is larger than that of passive confinement [1,11]. Previous studies have reported about the influence of the SSTT on
the confinement of concrete. Yooprasertchai et al. [5] and Ekkachai et al. [13] proposed a model for estimating the compressive
strength and strain of recycled concrete and brick aggregates confined with SSTT. Yooprasertchai et al. [5] observed that the
compressive strength increased up to 250 % and Ekkachai et al. [13] observed that the compressive strength and strain increased up to
253 % and 414 %, respectively. Nazirah et al. [1] proposed a method to evaluate the influence of actively confining techniques in
degraded concrete. Lim and Togay [4] proposed a unified model to assess the behavior of FRP jacketing and active confinement for
concrete elements. Mander et al. [9] and Saatcioglu and Razvi [14] have proposed models to quantify the increase of strength and
deformation capacity of confined concrete. However, these models cannot be directly applied to concrete confined with SSTT.
Moghaddam et al. [11], Awang et al. [15] and Lee et al. [12] observed that applying existing models for concrete confined with SSTT
generates unreliable results because such models were developed considering that confinement is provided by internal steel hoops that
are embedded in the concrete. When the confinement is provided by SSTT, the confined stresses are applied from the exterior of the
concrete element [2]. The tensioning of the straps in SSTT defines the provided confinement which is activated from the beginning of
the load application. This active confinement increases during the axial load application due to the transverse deformation of the
concrete generated by the Poisson ratio. Hence, the stress-strain characteristics of the straps define the mechanical response of the
confined column.
Steel fibers are currently used to improve the mechanical properties of plain concrete. Steel fibers are uniformly distributed in
concrete and provide strength in multiple directions, whereas reinforcing bars provide strengths only in the direction of the bars [16].
The use of steel fibers in concrete helps to control cracks induced by shrinkage and cracks that appear when the maximum compressive
strength is reached. The use of steel fibers in concrete reduces the fragility and increases the ductility, toughness, and the resistance to
impact, fatigue and abrasion. The use of steel fibers in concrete also enhances the behavior under flexural, tensile and shear loads [17].
Steel fibers slightly influence the compressive strength of concrete, but they increase the deformation capacity by providing concrete
confinement [16,18–21]. Based on the benefits than the use of steel fibers can provide to concrete, ACI 318-19 [22] allows the use of
these fibers to replace minimum shear reinforcement in beams. The use of steel fibers to replace minimum shear reinforcement in
beams is limited to fibers with a length-to-diameter ratio (lf/df) between 50 and 100, and the minimum dosage is 60 kg/m3 [22].
Even though several studies have evaluated RC elements confined with SSTT [2,5,12,15], concrete columns without conventional
reinforcement and SFRC have not been evaluated. This gap highlights the necessity for further research into SSTT systems. The
variability in mechanical characteristics of steel straps and their fastening mechanisms across different locations, as well as the di­
versity in procedures for installing and pre-stressing the steel straps, exhibit the need for exhaustive research in this regard. Addi­
tionally, more research is required to further study the effect of steel fibers when combined with SSTT [18]. Therefore, the objective of
this study is to evaluate experimentally the compressive behavior of SFRC columns confined with SSTT. The studied variables are the
cross-section of the columns (square and circular), the steel fibers dosage (15, 30 and 60 kg/m3), and the spacing of the steel straps
(37.5, 50 and 75 mm). Additionally, a numerical model to predict the compressive behavior of SFRC columns confined with SSTT is
proposed.

2. Experimental program
The experimental program was aimed to evaluate the effect of the cross-section of the columns, the dosage of the steel fibers, and
the spacing of the steel straps on the performance of the columns. Therefore, compression tests were carried out on 34 short columns
having 300 mm height with two different cross-sections. The nomenclature used to identify the column specimens is G-M#-S#-T#,
where G denotes the geometry of the cross-section of the specimens (C for circular and S for square). The parameter M# defines the
concrete material used, namely, PC for plain concrete, and SFRC15, SFRC30 and SFRC60 for SFRC with a dosage of 15, 30 and 60 kg/
m3, respectively. The parameter S# specifies the spacing of the steel straps, where S75, S50 and S37.5 are used for 75, 50 and 37.5 mm
spacing, respectively. For the control specimens that were tested without steel traps, UC (Unconfined Concrete) was used for S#. The
last parameter T# refers to the number of the tested specimen. When this number is excluded from the nomenclature, it characterizes
the average value of the response of the specimens, e.g. C-PC-UC-1 and C-PC-UC are the response of specimen 1 and the average value
of the response of unconfined cylindrical columns made of plain concrete, respectively.

2.1. Geometry
The experimental program included tests of 30 circular columns of 150 mm diameter and 4 square columns of 150 mm × 150 mm

2
J. Carrillo et al. Journal of Building Engineering 98 (2024) 110993

cross-section. The circular columns were constructed with PC and SFRC using the three dosages previously indicated. The square
columns were constructed with PC only. Specimens with circular and square cross-sections were tested and evaluated because previous
studies [7,11,23] have shown that the cross-section affects the stresses within the element and the concrete confinement. According to
Campione et al. [23], the confining stress distribution in circular cross-section elements is continuous and uniform along the perimeter.
On the other hand, the confining stress distribution in square or rectangular cross-sections is not uniform and stress concentrations are
generated at the corners. This stress concentration in non-circular cross sections reduces the effectiveness of confinement up to 33 %
compared to circular cross sections [23]. Therefore, Moghaddam et al. [11], Abbasnia et al. [7] and Campione et al. [23] propose to
round the corners of non-circular cross sections to reduce stress concentrations.
The dimensions of the column specimens with circular and square cross-sections are shown in Fig. 1a and b, respectively. The radius
of the rounded corners (rd ) of the square specimens is shown in Fig. 1b and the resulting length of the straight side is b − 2rd = 90 mm,
where b = 150 mm is the side of the specimens with square cross sections. The rounded corners of the columns with the square cross
sections (Fig. 1b) are also necessary to facilitate the installation and effectiveness of the steel straps. The rounded corners increase the
amount of concrete that is effectively confined by the steel straps. The ratio between the radius of the rounded corners and the size of
the specimen (rd /b) was defined as equal to 0.2, to achieve an effectiveness of the confinement of about 80 % [23]. Using this ratio, the
resulting radius of the rounded corners is rb = 30 mm.

2.2. External SSTT confinement

2.2.1. Spacing of steel straps and volumetric steel ratio
The spacings of the steel straps of the tested specimens were selected in this study to achieve specific ratios of reinforcement volume
to concrete volume. The spacings of the steel straps were selected based on the Colombian seismic design code NSR-10 [24]. In this
code, different reinforcement ratios are specified for three levels of ductility demands (Minimum - DMI, Moderate – DMO, and Special –
DES). The volumetric reinforcing ratios of confining steel for DMI and DES levels of the NSR-10 [24] are equivalent to those of ACI
318-19 [22] for columns of special moment frames. For these two levels, the spacing of the steel straps corresponds to the one specified
by NSR-10 [24] to confine the plastic hinge region of a column. This spacing was selected to evaluate the behavior of columns with
SSTT at the critical regions of columns.
The volumetric steel ratio and the total area of transverse steel straps for the columns with both square and circular cross-sections
(ρs and Ash) were obtained from sections C.7.10.5, C.21.3.5 and C.21.6.4 from NSR-10 [24]. Table 1 summarizes the spacing and the
volumetric steel ratio requirements of NSR-10 for DMI, DMO and DME. Because of the absence of longitudinal reinforcement of the
column specimens, the spacing of the steel straps is controlled by conditions (d) for DMI and (c) for DMO and DES rather than con­
ditions (a) and (b) in Table 1. The resulting spacing of the steel straps for DMI, DMO and DME are 75 mm (S75), 50 mm (S50) and 37.5
mm (S37.5), respectively, since the maximum spacing of transverse reinforcement is the most conservative criteria. To verify the
minimum transverse reinforcement ratio prescribed by NSR-10 [24], Table 1 shows both the minimum values and the values calcu­
lated with the spacings used in this study for DMI, DMO and DES (S75, S50 and S37.5, respectively). Eqs. (1) and (2) show the
transverse reinforcement ratio calculated for square (ρt) and circular (ρs) columns, respectively, expressed in terms of the thickness (az)
and width (bz) of the steel strap, the perimeter (p) and the width (b) of the specimen and, in the case of square columns, the radius of the
rounded corners (rd).
a b p
ρt = [ (z z )] (1)
πr 2
s b2 -4 rd 2 - d4

az bz p
ρs = (2)
π D2
4
s

Table 1 shows that the ratio of transverse reinforcement ρs of circular and squared cross-sections associated with ductility demand

Fig. 1. Specimen dimensions: (a) circular cross-section; (b) square cross-section.

3
J. Carrillo et al. Journal of Building Engineering 98 (2024) 110993

Table 1
Spacing and reinforcing ratio of confining reinforcement specified by NSR-10.

DMI (section C.7.10.5) DMO (section C.21.3.5) DES (section C.21.6.4)

Spacing of transverse reinforcement s (a) 16 db N.A. (a) 8db N.A. (a) 6db N.A.
( )
(b) 48 dbe 188.8 mm (b) 16ϕe 62.9 mm 350 - lo N.A.
(b) so = 100 +
3
(c) x 150.0 mm (c) x/3a 50.0 mma (c) x/4a 37.5 mma
(d) sa-b-c/2 (hinge)a 75.0 mma (d) 150 mm 150.0 mm
( ) ʹ
ρs of circular columns A f fcʹ fcʹ
ρs = 0.45 ch - 1 c ≥ 0.063% ρs ≥ 0.08 = 0.54% ρs ≥ 0.12 = 0.81%
Ag fyt fyt fyt
ρs = 0.43% ρs = 0.65% ρs = 0.86%
⎧ ( ) ⎧ ( )
ρs of square columns ρs ≥ 0.065% ⎪ f ʹ Ach ⎪ f ʹ Ach
⎪ 0.2 c
⎪
⎪ -1 = 0.01% ⎪ 0.3 c
⎪
⎪ -1 = 0.02%
⎨ fyt Ag ⎨ fyt Ag
ρt ≥ ρt ≥
⎪
⎪ fcʹ ⎪
⎪ fcʹ
⎪
⎪
⎩ 0.06 = 0.35% ⎪
⎪
⎩ 0.09 = 0.53%
fyt fyt
ρt = 0.41% ρt = 0.61% ρt = 0.82%
Area of transverse reinforcement Ash = az bz N Ash = az bz N Ash = az bz N
Ash = 48.6 mm2 Ash = 72.9 mm2 Ash = 97.2 mm2
a
Criteria selected to define the external confinement of the specimens; Ach = cross-sectional area of the member measured to the outside edges of transverse
reinforcement; Ag = gross area of concrete section; Ash = total cross-sectional area of transverse reinforcement, including crossties, within spacing s and perpendicular to
dimension bc; bc = cross-sectional dimension of member core measured to the outside edges of the transverse reinforcement composing area Ash; x = shorter overall
dimension of rectangular part of cross section; db = nominal diameter of bar; dbs = nominal diameter of transverse reinforcement; N.A. = the condition is non-applicable;
ρs = ratio of volume of spiral reinforcement to total volume of core confined by the spiral, measured out-to-out of spirals; ρs = ratio of area of distributed transverse
reinforcement to gross concrete area perpendicular to that reinforcement; so = center-to-center spacing of transverse reinforcement within the length ℓo.

DMI are 0.43 % and 0.47 %, respectively, which are significantly higher than the minimum ρs of 0.063 % and 0.065 %, respectively. For
DMO, the table shows that ρs for circular section columns is 0.65 %, which is greater than the minimum ρs of 0.54 %. Similarly, for
square-section columns at DMO, the area Ash is 72.9 mm2, which is larger than the calculated minimum areas (2.2 mm2 and 30.8 mm2).
For DES, the table shows that ρs for circular-section columns is 0.86 %, which is greater than the minimum ρs of 0.81 %. For the square-
section columns at DES, the area Ash is 97.2 mm2, which is larger than the calculated minimum areas Ash1 and Ash2 (2.5 mm2 and 34.6
mm2). Fig. 2a, b and 2c show the distribution of external confining straps for DMI, DMO and DME, respectively.

2.2.2. Lateral confining pressure

The lateral confining pressure that provides conventional transverse reinforcement and SSTT is generated by the interaction be­
tween the transverse deformation of the concrete column and the stiffness of the transverse reinforcement. In the case of conventional
transverse reinforcement, confinement starts to be effective when the compressive axial load is reaching the concrete compressive
strength, because of the significant amount of transverse deformation of the concrete. For these axial loads, the transverse reinforcing
steel elongates and provides restriction to the transverse deformation of the concrete. Thus, lateral confining pressure (fle) is achieved
in the concrete core. In SSTT, the interaction between the transverse deformation of concrete and fle works the same as for conventional
reinforcement but the concrete is already confined before the application of the compressive axial load due to prestress forces applied

Fig. 2. Distribution of steel straps for both circular and squared column specimens. (a) S = 75 mm, (b) S = 50 mm, and (c) S = 37.5 mm.

4
J. Carrillo et al. Journal of Building Engineering 98 (2024) 110993

to the straps during installation. Fig. 3 shows the specimens with circular and square cross-sections and with varying confinement
provided by the steel straps.
In this study, two setups were used to measure the active lateral confining pressure induced by steel straps. The first setup shown in
Fig. 4 consists of a hollow steel cylinder divided into two parts. The two parts of the cylinder are separated by a load cell with a loading
capacity of 50 kN. The load cell registers the force applied to the cylinder when the steel straps are installed and tensioned by the
strapping device. The force measured by the load cell is equal to the sum of the forces applied by the steel straps. The strapping device
used in the first setup (see Fig. 4a) was also used to apply the initial confinement to the specimens. The prestressing process was applied
into two steps. The strap was initially fixed to the specimen and the handle was then turned once to provide a tension force of 1.0 kN
acting on the steel strap. This force is equivalent to a stress of 71.4 MPa and 20 % of the yield stress (fyt) of the steel strap. This first setup
was used to calibrate the strapping device; however, values of pressure were not registered since the uncertainty was limited by the
load cell accuracy and the stress losses during the strapping process.
The second setup used to measure the lateral confining pressure induced by the steel traps consisted of measuring the deformation
of the steel straps during the strapping process of the column specimens. This setup was used to get a more reliable measure of the
initial active confinement pressure by measuring the tensile strains of the straps. Then, the stress of the steel straps was obtained by
multiplying the measured strain by the modulus of elasticity of the steel. The lateral confining stress (fl) is obtained using the volu­
metric ratio of confining steel (ρs) and the tensile stress of the steel straps (fs), as shown in Eq. (3). A target confinement stress of about
20 % of fyt was applied to the column specimens to increase their strength and ductility. The measured values of this initial active
confinement stress varied between 15 % and 25 % of fyt.
ρs ⋅ f s
fl = (3)
2

2.3. Material properties

2.3.1. PC and SFRC
The relevant fibers geometric properties affecting the concrete behavior are the length of the fibers (lf), the diameter of the fibers
(df) and the length-to-diameter ratio (lf/df). The dosage of fibers can be expressed as kilograms per cubic meter of concrete (Df) or as
volume fraction per cubic meter of concrete (Vf). The relationship between these two measures is obtained by using the density of the
steel fibers (7850 kg/m3). The length-to-diameter ratio (lf/df) of steel fibers affects the concrete behavior in both fresh and hardened
state [17]. For example, when the length-to-diameter ratio varies between 50 and 100, the compressive behavior improves after
reaching the compressive strength. When the length-to-diameter ratio is larger than 100, the mixing process of the concrete in the fresh
state is affected, and clusters are generated [19]. The presence of these clusters increases the porosity and decreases the strength and
stiffness of the concrete. This reduction of concrete strength because of the fibers have been reported by Nataraja et al. [20] and
Carrillo et al. [21].
The concrete used to cast the column specimens was mixed in the laboratory using a batch mixer. The specified concrete strength,
the characteristics of the constituent materials, and the proportions for one cubic meter of the used plain concrete are shown in Table 2.
For the SFRC, Dramix steel fibers with end hooks were used with three different fiber dosages. The length of the fibers (lf) was 60 mm,
the diameter (df) 0.90 mm and the length-to-diameter ratio (lf/df) was 67. The SFRC was mixed using the same batch mixer used for the
PC. The dosage of steel fibers used was equivalent to 25 %, 50 % and 100 % of the minimum fiber dosage specified by the NSR-10 and
ACI 318 for replacing transverse reinforcement in beams. Therefore, the used dosages corresponded to 15, 30 and 60 kg/m3. Table 3
summarizes the nominal and the measured dosages of the used SFRC. The actual dosages of fibers (Df in Table 3) were obtained
following the procedure of UNE-EN 14488-7. The table also summarizes the fiber volume fractions (Vf) and the slump of the three
concrete mixtures. The study includes the influence of fibers in terms of reinforcement index (RI), which is the product of Vf and lf/df.

2.3.2. Stainless steel straps

The characteristics of the steel straps were determined from four samples of the same batch, tested in tension following ASTM E8
[26] standard. The stress-strain curves of these test on the four steel straps specimens are shown in Fig. 5. The designation of the
stainless steel of the steel straps of this study was UNSS30400. The thickness of the steel straps was 0.77 mm and the width was 15.78

Fig. 3. Column specimens confined with SSTT: (a) circular columns, (b) square columns.

5
J. Carrillo et al. Journal of Building Engineering 98 (2024) 110993

Fig. 4. Strapping device: (a) setup to calibrate the initial confining stress, (b) connection clip.

Table 2
Design parameters of the concrete mixture.

Material Type Quantity

Specified concrete strength, MPa 21

Cement, kg/m3 Gray cement Argos 334.0

Ash, kg/m3 Volcanic ash 25.5

3
Fine aggregates, kg/m Rock sand 190.0
Sand from river 747.5

Coarse aggregates, kg/m3 Gavel from river 920.0

Additives, cm3 Viscoflow 1069

Sikaplast 1190

Water, L ASTM C-1602 [25] 176.5

Water/cement ratio 0.53

Maximum Size Aggregate, MSA, mm 12.7

Slump, mm 200

Table 3
Dosage of steel fibers and slump of SFRC.

Property SFRC15 SFRC30 SFRC60

3
Df-nominal, kg/m 15 30 60
Df, kg/m3 12.7 31.5 65.2
Fiber volume fraction, Vf,% 0.19 0.38 0.76
Reinforcement Index, RI = Vf × lf/df 10.83 26.85 55.58
Slump, mm 185 75 50

Fig. 5. Stress-strain curves of strapping steel.

6
J. Carrillo et al. Journal of Building Engineering 98 (2024) 110993

mm. The measured average yield strength (fyt), tensile strength (fm), and modulus of elasticity (Es) of the steel straps were 357 MPa,
688 MPa and 254 GPa, respectively. The Coefficient of Variation (CV) of the measured values of fyt, fm and Es were 7.3 %, 4.1 % and 3.6
%, respectively, and they were obtained from the tests of four steel straps.

2.4. Test setup of the column specimens

The test setup and the instrumentation of the column specimens with circular and rectangular cross-sections are shown in Fig. 6.
The compression test of the column specimens was conducted following ASTM C39 [27] and ASTM C469 [28] test standards using a
Controls universal testing machine (model MCC8) with a loading capacity of 2000 kN. The horizontal transducer in Fig. 6a was
installed in an opened ring with two pinned supports in opposite sides of the cylinder. The linear displacement of the transducer refers
to the arch length of the ring; therefore, the transverse displacements of the cylinder can be geometrically correlated. The column
specimens were subjected to monotonic compressive axial deformation using displacement control protocol and a loading velocity of
6.5 μm/s.
The deformation of the column specimens with circular cross-sections were obtained using a measuring frame that contained two
longitudinal displacement transducers (50 mm stroke) to measure the axial deformation and one horizontal displacement transducers
to measure the transverse deformation (Fig. 6a). The deformation of the column specimens with square cross-sections was registered
by three displacement transducers with a stroke of 50 mm. The transverse deformation of the concrete was measured with two 30 mm
length strain gauges (model FLA-30-11-2L) that were attached to the adjacent faces of the column. The deformation of the steel straps
in both types of columns was measured with two 5 mm length strain gauges (TML model FLA-5-11-5LJC) that were attached to the
central straps (see Fig. 6b and d). Details of the experimental program are reported by Fierro [29].

3. Test results and discussion

3.1. Failure modes
The characteristic failure modes of unconfined concrete (UC) columns and columns confined with SSTT are shown in Fig. 7.
Similarly to results observed by Ekkachai et al. [13], the failure mode of most UC columns was characterized by a diagonal or
transverse crushing region. The use of steel fibers in these UC columns was effective to improve the integrity of concrete, thus no
significant concrete detachment was observed in these column specimens. The crushing region of the column specimens confined with
SSTT was less localized than that of the UC columns. Ekkachai et al. [13] and Yooprasertchai et al. [5] also observed that specimens of
PC confined with SSTT exhibited crushing, accompanied by the failure of the steel straps in the same region. In addition, concrete
crushing and sliding were observed along the fault planes of the column specimens confined with SSTT.

3.2. Stress-strain relationships

Moghaddam et al. [11] suggest applying an initial confining pressure to columns with SSTT equivalent to 30 % of the yield strength
of the confining steel. This confining pressure is required to achieve yielding of the confining steel while the concrete reaches the peak
compressive stress. However, the target confining stress of 0.3fy of the steel straps may not be accurately achieved in practical ap­
plications. In fact, the confining stresses applied to the SSTT in this study varied between 0.15fy and 0.25fy. Hence, yielding of the steel
straps in the SSTT diverged from the time that peak concrete stress occurred.
Fig. 8a and b shows the compressive behavior of the columns in the longitudinal and transverse directions, respectively. These
figures show the single tests of the square columns and the average behavior of the 30 circular columns, represented as 16 stress-strain
relationships. The strain in the axial direction of the circular columns of Fig. 8a was obtained from the average displacement measured
from the two displacement transducers (Fig. 6a). The strain of the steel straps of the specimens with SSTT, related to the compressive
stress, is shown in Fig. 8c. In these figures, the yield strain measured for the steel straps (εy = 0.0034) is shown with vertical dashed
lines.
The measured mechanical properties of the 34 column specimens are summarized in Table 4. For some cases, one column sample

Fig. 6. Instrumentation of column specimens: (a–b) circular columns, (c–d) square columns.

7
J. Carrillo et al. Journal of Building Engineering 98 (2024) 110993

Fig. 7. Failure modes of selected circular and square column specimens.

was tested and for some others two column samples were tested (refer to the last number of the nomenclature of the specimens).
Table 4 shows the confined concrete compressive strength (f’cc), the strain at peak strength (εcc), the modulus of elasticity (Ec), the
Poisson ratio (υ), the stress and strain at 15 % strength loss (f’85 and ε85), the stress and strain at 50 % strength loss (f’50 and ε50), and
the ultimate strength and strain (f’cu and εcu). The ultimate strength and strain are associated with the concrete failure. Some values of
f’85, f’50, ε85 and ε50 for some specimens are not shown because a stress drop of 85 % and/or 50 % was not measured in these specimens.
Table 4 includes the initial strain of the steel straps (εso), and the strain of the steel straps at the peak compressive stress of the columns
(εsc). The table also shows the stress of the steel straps when the column specimens reach the peak compressive stress (fsc) and the
parameter kep, which represents the ratio fsc/fyt. The measured normalized confined concrete strength (f’cc/f’co), the normalized
confining pressure (fle/f’co) and the normalized strain at peak stress (εcc/εco) are summarized in Table 4. The table also shows the
volumetric steel ratio of confining steel (ρs), the initial confining pressure (fl), the effective confining pressure (fle), and the parameter
RI. The terms kep and keg in Table 4 are the coefficients of effectiveness proposed by Mander et al. [9] for confining pressure and
geometry, respectively. These latter coefficients are explained in section 4.2. The table shows the normalized strain at 85 % of peak
stress (ε85/εcc) and the normalized strain at 50 % of peak stress (ε50/εcc). The ratios ε85/εcc and ε50/εcc are referred to as the plasticity
and the ductility ratios, respectively [30,31].
Five branches can be identified in the stress-strain curves of the column specimens in Fig. 8a. The first three branches (I, II and III) of
the confined specimens (S75, S50 and S37.5) are similar to those observed in non-degraded unconfined concrete tests (UC) in terms of
elastic branch, peak stress, and initial softening. Similarly to results observed by Ekkachai et al. [13], the elastic modulus in the I
branch is enhanced by the active confining pressure, mostly on degraded concrete columns. Narizah et al. [1] observed that high
strength concrete actively confined exhibit strain hardening after the elastic branch. The slops of branches II and III decrease just
before and after the peak compressive strength, respectively. In Fig. 8a, branches IV and V are caused by the confinement with SSTT.
Jianghao et al. [10] argued that SSTT delays the degradation of the concrete because pre-tensioned straps contributes earlier than the

8
J. Carrillo et al. Journal of Building Engineering 98 (2024) 110993

Fig. 8. Stress-strain relationships of selected column specimens: (a) average axial strain, (b) average transverse strain in the concrete, (c) average transverse strain in
the steel straps.

peak strength, improving the material utilization rate (ratio between maximum strength and maximum stress achieved), strengthening
effects, and avoiding stress hysteresis. The first new branch (IV) identified in the column specimens with SSTT (after point f’85) is
characterized by a negative slope due to the strength degradation. The slope of this descending branch depends on the amount of
confinement with SSTT. Finally, branch V is characterized by a less steep descending slope, where strength degradation is less severe.
In some of the specimens (e.g., C-SFRC60-S50 in Fig. 8a5), the slope of this branch is almost horizontal, which has been also observed
by previous studies [1,3,4,8,10,13]. This last branch on the stress-strain curves is attributed to the yielding of the steel traps, where the
concrete deforms under constant confining pressure.

3.3. Effect of confinement on the compressive strength and deformation of the concrete
The square columns of PC confined with SSTT showed a confined concrete compressive strength (f’cc) from 3.5 % to 6.5 % lower
than those of circular columns in Fig. 8a2 (Table 4) because the SSTT is more effective in cylindrical geometries [9]. The results of
Fig. 8a and Table 4 show that the confined concrete compressive strength of the column specimens confined with SSTT (S75, S50 and
S37.5) was larger than that of the unconfined specimens (UC). The confined concrete compressive strength of the circular specimens
with plain concrete and external confinement S75 and S37.5 was 6 % (44.0/41.7) and 16 % (48.5/41.7) larger, respectively, than that
of the unconfined circular specimen (C-PC-UC). For the square columns with plain concrete, the confined concrete compressive
strength of the specimens with external confinement S75 and S37.5 was 4 % (41.9/40.2) and 11 % (44.6/40.2) larger, respectively,
than that of the unconfined square column (S-PC-UC). The same trend was also observed for the circular specimens with SFRC. The
confined concrete compressive strength of the specimens with SFRC15 and external confinement S75 and S37.5 were 10 % (42.5/38.6)
and 18 % (45.5/38.6) larger than that of the specimen C-SFRC15-UC. The confined concrete compressive strength of the specimens
with SFRC30 and external confinement S75 and S37.5 was 9 % (43.0/39.3) and 17 % (45.9/39.3) larger, respectively, than that of the
specimen C-SFRC30-UC. Finally, the confined concrete compressive strength of the specimens with SFRC60 and external confinement
S75 and S37.5 was 7 % (42.9/40.2) and 16 % (46.6/40.2) larger, respectively, than that of the specimen C-SFRC30-UC. It can be
concluded from these findings that the confined concrete compressive strength increases as the confinement also increases for both
circular and square columns.
Regarding the post-peak behavior, the ultimate deformation (εu) of columns confined with SSTT was about eight times larger than
that of the unconfined columns. Additionally, Fig. 8a shows that the strength degradation decreases as the external confinement in­
creases. The average negative slope of the branch III of the stress-strain curve of the circular column specimens with steel strap spacing
of 75 mm, 50 mm, and 37.5 mm were 63 %, 79 % and 82 %, respectively, lesser than that of the control specimen C-PC-UC. The
external confinement and the steel fibers contributed to control the crack generation and propagation. Furthermore, the external
confinement limits the volumetric expansion of the concrete, which is explained by the reduced descending slope in Fig. 8b.
Finally, Fig. 8a y 8b show that use of steel fibers increased the ductility of the columns. This increase can be observed in slopes of the
branches IV and V of the stress-strain curves of SFRC specimens confined with SSTT. The εu of the SFRC column specimens with a fiber
dosage of 60 kg/m3 was approximately 0.023 for columns without SSTT and increased to 0.056 for columns with SSTT. For columns

9
 Table 4

J. Carrillo et al.
Mechanical properties of the 34 column specimens.

Specimen D (L) s RI ρs εo fl εsc fsc kep keg fle f’cc εcc Ec u f’85 ε85 f’50 ε50 f’cu εcu f’cc/ fle/f’c εcc/ ε85/ ε50/
(f’co) (εco) f’c εco εcc εcc
mm mm % mm/ MPa mm/ MPa MPa MPa mm/ MPa MPa mm/ MPa mm/ MPa mm/
mm mm mm mm mm mm

C-PC-UC-1 151.0 0.0 0.00 0.0000 – – 0.0000 0.0 0.00 1.00 – 40.5 0.0032 27848 0.19 34.4 0.0031 – – 38.3 0.0025 1.00 – 1.00 0.97 –
C-PC-UC-2 151.5 0.0 0.00 0.0000 – – 0.0000 0.0 0.00 1.00 – 41.3 0.0028 27736 0.17 36.6 0.0037 – – 32.5 0.0030 1.02 – 0.88 1.32 –
C-PC-S75-1 153.3 75.0 0.00 0.0043 0.00030 0.77 0.00059 148.6 0.42 1.00 0.32 44.9 0.0033 23001 0.22 38.2 0.0041 22.5 0.0244 11.5 0.0435 1.11 0.007 1.03 1.24 7.39
C-PC-S75-2 152.5 75.0 0.00 0.0043 0.00024 0.77 0.00054 137.2 0.38 1.00 0.29 43.0 0.0037 22584 0.19 36.6 0.0054 21.5 0.0150 27.8 0.0256 1.06 0.007 1.16 1.46 4.05
C-PC-S50-1 152.8 50.0 0.00 0.0065 0.00027 1.16 0.00073 185.4 0.52 1.00 0.60 46.0 0.0034 25120 0.16 39.1 0.0053 23.0 0.0278 15.7 0.0301 1.14 0.013 1.06 1.56 8.18
C-PC-S50-2 153.8 50.0 0.00 0.0065 0.00032 1.16 0.00075 190.5 0.53 1.00 0.62 46.8 0.0038 22963 0.20 39.8 0.0051 23.4 0.0108 23.2 0.0108 1.16 0.013 1.19 1.34 2.84
C-PC-S37.5–1 151.5 37.5 0.00 0.0086 0.00021 1.54 0.00106 269.5 0.75 1.00 1.16 47.6 0.0031 25643 0.15 40.5 0.0058 23.8 0.0282 17.6 0.0340 1.18 0.024 0.97 1.87 9.10
C-PC-S37.5–2 152.0 37.5 0.00 0.0086 0.00032 1.54 0.00119 301.6 0.84 1.00 1.30 49.7 0.0031 27836 0.15 42.2 0.0053 24.8 0.0173 28.0 0.0174 1.23 0.026 0.97 1.71 5.58

C-SFRC15-UC-1 152.5 0.0 10.83 0.0000 – – 0.0000 0.0 0.00 1.00 – 35.0 0.0024 25638 0.14 29.7 0.0037 – – 24.2 0.0046 0.86 – 0.75 1.54 –
C-SFRC15-UC-2 152.3 0.0 10.83 0.0000 – – 0.0000 0.0 0.00 1.00 – 42.4 0.0024 24836 0.17 36.0 0.0026 – – 31.1 0.0030 1.05 – 0.75 1.08 –
C-SFRC15-S75-1 152.5 75.0 10.83 0.0043 0.00031 0.77 0.00066 167.4 0.47 1.00 0.36 41.0 0.0029 22336 0.16 34.8 0.0053 20.5 0.0115 5.3 0.0383 1.01 0.009 0.91 1.83 3.97
C-SFRC15-S75-2 152.8 75.0 10.83 0.0043 0.00022 0.77 0.00058 147.6 0.41 1.00 0.32 44.0 0.0038 21279 0.19 37.4 0.0056 22.0 – 35.0 0.0060 1.09 0.007 1.19 1.47 –
C-SFRC15-S50-1 152.8 50.0 10.83 0.0065 0.00028 1.16 0.00078 198.4 0.56 1.00 0.64 42.8 0.0037 22989 0.21 36.4 0.0066 21.4 0.0106 9.7 0.0167 1.06 0.015 1.16 1.78 2.86
C-SFRC15-S50-2 152.8 50.0 10.83 0.0065 0.00021 1.16 0.00076 192.8 0.54 1.00 0.63 44.4 0.004 21494 0.17 37.8 0.0075 22.2 0.0138 22.6 0.0267 1.10 0.014 1.25 1.88 3.45
C-SFRC15-S37.5–1 153.0 37.5 10.83 0.0086 0.00021 1.54 0.00066 167.9 0.47 1.00 0.72 43.9 0.0042 21962 0.13 37.3 0.0079 21.9 0.0182 17.2 0.0373 1.08 0.016 1.31 1.88 4.33
C-SFRC15-S37.5–2 153.0 37.5 10.83 0.0086 0.00024 1.54 0.00078 198.2 0.56 1.00 0.85 47.2 0.0041 22720 0.09 40.1 0.0070 23.6 0.0270 20.5 0.0311 1.17 0.018 1.28 1.71 6.59

C-SFRC30-UC-1 151.8 0.0 26.85 0.0000 – – 0.0000 0.0 0.00 1.00 – 40.2 0.0024 26378 0.16 – – – – 25.3 0.0045 0.99 – 0.75 – –
C-SFRC30-S75-1 153.0 75.0 26.85 0.0043 0.00024 0.77 0.00080 202.9 0.57 1.00 0.44 43.2 0.0032 23280 0.18 36.7 0.0040 21.6 – 32.7 0.0044 1.07 0.010 1.00 1.25 –
C-SFRC30-S75-2 152.5 75.0 26.85 0.0043 0.00030 0.77 0.00086 217.3 0.61 1.00 0.47 42.8 0.0034 22067 0.14 36.4 0.0057 21.4 0.0111 11.8 0.0266 1.06 0.011 1.06 1.68 3.26
10

C-SFRC30-S50-1 152.0 50.0 26.85 0.0065 0.00027 1.16 0.00054 137.7 0.39 1.00 0.45 44.4 0.0037 22657 0.19 – – – – 7.3 0.0803 1.10 0.010 1.16 – –
C-SFRC30-S37.5–1 152.5 37.5 26.85 0.0086 0.00032 1.54 0.00074 188.8 0.53 1.00 0.81 45.4 0.005 21904 0.14 38.6 0.0111 22.7 0.0312 18.9 0.0572 1.12 0.018 1.56 2.22 6.24
C-SFRC30-S37.5–2 152.5 37.5 26.85 0.0086 0.00034 1.54 0.00064 162.0 0.45 1.00 0.70 46.3 0.0041 21606 0.20 39.3 0.0125 23.1 0.0194 13.2 0.0523 1.14 0.015 1.28 3.05 4.73

C-SFRC60-UC-1 151.5 0.0 55.58 0.0000 – – 0.0000 0.0 0.00 1.00 – 36.6 0.0025 25680 0.14 31.1 0.0039 18.3 0.0089 6.9 0.0147 0.90 – 0.78 1.56 3.56
C-SFRC60-UC-1 151.8 0.0 55.58 0.0000 – – 0.0000 0.0 0.00 1.00 – 44.0 0.0028 25835 0.19 37.4 0.0039 22 0.0154 17.7 0.0321 1.09 – 0.88 1.39 5.50
C-SFRC60-S75-1 153.5 75.0 55.58 0.0043 0.00030 0.77 0.00092 232.5 0.65 1.00 0.50 43.2 0.0049 19845 0.15 36.7 0.0101 21.6 0.02 4 0.0552 1.07 0.012 1.53 2.06 4.08
C-SFRC60-S75-2 153.5 75.0 55.58 0.0043 0.00028 0.77 0.00086 218.0 0.61 1.00 0.47 42.5 0.0044 21039 0.17 36.1 0.0084 21.3 0.0221 12 0.0424 1.05 0.011 1.38 1.91 5.02
C-SFRCC60-S50-1 153.0 50.0 55.58 0.0065 0.00028 1.16 0.00100 254.3 0.71 1.00 0.83 45.5 0.0038 23715 0.16 38.7 0.0074 22.8 0.0169 16 0.0838 1.12 0.018 1.19 1.95 4.45
C-SFRCC60-S50-2 152.3 50.0 55.58 0.0065 0.00021 1.16 0.00099 251.3 0.70 1.00 0.82 43.5 0.0041 22421 0.16 37.0 0.0063 21.7 0.012 14.2 0.0212 1.07 0.019 1.28 1.54 2.93
C-SFRC60-S37.5–1 152.8 37.5 55.58 0.0086 0.00034 1.54 0.00099 250.8 0.70 1.00 1.08 47.2 0.0041 23293 0.24 40.2 0.0106 23.6 0.0235 9.8 0.0658 1.17 0.023 1.28 2.59 5.73
C-SFRC60-S37.5–2 152.8 37.5 55.58 0.0086 0.00021 1.54 0.00086 218.2 0.61 1.00 0.94 46.0 0.0037 22805 0.2 39.1 0.0086 23 0.0212 9.9 0.0464 1.14 0.020 1.16 2.32 5.73

Journal of Building Engineering 98 (2024) 110993

S-PC-UC-1 149.2 0.0 0.00 0.0000 – – 0.0000 0.0 0.00 0.75 – 40.2 0.0019 31774 0.26 34.2 0.0019 – – 24.7 0.0019 0.99 – 0.59 1.00 –
S-PC-S75-1 149.3 75.0 0.00 0.0041 0.00024 0.73 0.00066 168.3 0.47 0.75 0.26 41.9 0.0027 26517 0.2 35.6 0.0042 20.9 0.006 15.1 0.0128 1.03 0.006 0.84 1.56 2.22
S-PC-S50-1 148.5 50.0 0.00 0.0061 0.00022 1.09 0.00120 303.9 0.85 0.75 0.70 42.6 0.0031 25357 0.17 36.2 0.0050 21.3 0.0161 20 0.0190 1.05 0.016 0.97 1.61 5.19
S-PC-S37.5–1 147.5 37.5 0.00 0.0082 0.00024 1.46 0.00157 398.2 1.12 0.75 1.22 44.6 0.0039 20931 0.23 37.9 0.0081 22.3 0.0252 23 0.0252 1.10 0.027 1.22 2.08 6.46

- Data for this specimen is not available; D = diameter of the circular column; L = width of the square column; f’co = unconfined concrete compressive strength; f’cc = confined concrete compressive strength; εco = deformation at
peak stress of unconfined concrete; εcc = deformation at peak stress of confined concrete; Ec = modulus of elasticity of the concrete; υ = Poisson ratio; f85 = post-peak stress corresponding to 85 % of the maximum stress; ε85 = strain
at f85; f50 = post-peak stress corresponding to 50 % of the maximum stress; ε50 = strain at f50; f’cu = ultimate compressive stress; εcu = strain at fcu.
J. Carrillo et al. Journal of Building Engineering 98 (2024) 110993

with Plain Concrete (PC) without SSTT, εu was approximately 0.003 and rose to 0.034 for PC columns with SSTT. In addition, values of
εu of SFRC column specimens with a fiber dosage of 60 kg/m3 were, on average, nine times larger than the ultimate deformation of PC
column specimens. The results of ε85 and ε50 in Table 4 also show that the ductility increases as the dosage of the fibers increases
because larger strain values were measured for larger dosage of fibers.

3.4. Behavior of steel straps

The stress-strain curves in Fig. 8c show that the steel straps started to work effectively before the peak confined concrete
compressive strength was achieved. Beyond the peak strength the strain of the steel straps continued to increase until yielding.
Therefore, the confinement of the columns was improved beyond the point of peak compressive strength. This fact allows that the
yielding of the steel straps occurred later than the point of compressive peak strength. The later yielding of steel confinement has also
been observed when using conventional reinforcing steel [9,11,14,32]. Fig. 9 shows the compressive stress of the confined square and
circular columns shown in Fig. 8c1 to 8c4, as a function of the strains measured at the top and bottom steel straps, up to a strain limit of
0.01 mm/mm. In Fig. 9, the final letter refers to the location of the strap in the column samples, where B is used for a bottom strap and T
for a top strap. The strains in the straps S-PC-S50-1B, C-PC-S37.5–1T, C-PC-S37.5–2T, C-SFRC15-S75-1B, C-SFRC15-S37.5–2B,
C-SFRC60-S37.5–1T and C-SFRC60-S37.5–1B are not shown in Fig. 9 because the strain gages were damaged during the tests.
As shown in Fig. 9, relative initial positive strains up to 0.00005 were measured in the steel straps of the columns. These initial
strains, which cannot be observed in Fig. 9, are explained by micro-cracks that are generated in the concrete before reaching the peak
stress [12]. The micro-cracks create discontinuities in the concrete, causing to shortening of the steel straps towards their initial
configuration. This shortening of the steel straps leads to a decrease in the confining pressure exerted on the columns; however, the
initial confining pressure prevents the distension of the system. The positive relative initial strains of steel straps were not measured for
the column specimens with a fiber dosage of 60 kg/m3 (Fig. 9d). The presence of steel fibers delays crack formation in the concrete and
prevents the loss of the confining pressure that is provided by the steel straps. Therefore, the confining pressure provided by the steel
strap is more efficient in columns with SFRC when compared with PC columns which, as seen in Fig. 9b. Fig. 9 also shows that the
post-peak slopes are noticeably influenced as either the fiber dosage or the confining pressure increases. Most of the straps exceeded
the yielding strain (εy) either in PC or SFRC columns; however, the softening slope gets more uniform when the fiber content of the
columns increases, which represent the increasing efficiency of the SSTT in SFRC. Fig. 9a shows that, at the same strain in the straps,
the compressive stress of square columns is higher than that of circular columns. This difference in the curves of straps in circular and
square columns is mainly related to higher strains in the straps of square columns. The straps in the square columns showed higher
strains due to the higher concentration of confinement stresses at the corners, e.g. the confinement stress at the cross-section of the
circular columns due to the straps is distributed uniformly but this confinement stress tend to be concentrated at the corners of square
columns.

4. Analytical model
The model proposed in this study for estimating the confined concrete compressive strength and strain is developed in this section.
To describe the origin of the model, the results of the experimental program are evaluated in terms of the estimated results from
previous models. Finally, box and whisker plots and overpredictions are used to compare statistically the results of the proposed model
proposed with those available in the literature. The proposed model was calibrated using a database of 150 tests, where 34 are from
this study and the remaining 116 from Sarmah et al. [3], Ekkachai et al. [13], Moghaddam et al. [11], Chin et al. [33,34] and Neupane
et al. [35]. Table 5 summarizes the main characteristics of the latter 116 tests.

4.1. Comparison between measured results and existing analytical models

The confined concrete compressive strength (fcc’) measured from column specimens with SSTT, versus the normalized confining
pressure (fle/f’co) is shown in Fig. 10a. The confined concrete compressive strength (f’cc) is normalized using the unconfined concrete
strength (f’co), as shown previously in Table 4. The figure also shows the confined concrete compressive strength predicted by four
existing analytical models [9,11,14,32], and the average trend observed in this study. For PC, Fig. 10a shows that the compressive
strength predicted by the numerical models is, in general, lower than the measured compressive strength. The analytical model by
Saatcioglu et al. [14] is the only one that accurately predicted the measured compressive strength of the column specimens with PC.
Equivalent to previous results [11,14,36], the confined concrete strength measured in this study increases as the confining pressure
increases. A single linear relationship for the column specimens with PC and SFRC was proposed because, as shown in Fig. 10b, the
confined concrete compressive strength is independent of the RI parameter of steel fibers. Although Carrillo et al. [18] observed that
fibers may increase the confined concrete strength (f’cc), Fig. 10b shows that f’cc was independent on RI. Finally, Fig. 10c shows the
relationship between the strain at peak compressive strength (εcc), normalized with respect to the peak compressive strain of un­
confined concrete (εco), and the term RI. The figure shows that εcc increases as the term RI increases, and that larger values are obtained
for the columns confined with SSTT than for columns with UC.

4.2. Proposed analytical model

The analytical equation that is proposed in this study to estimate the peak strength of SFRC columns confined with SSTT is shown in
Eq. (4). The equation with the first two terms corresponds to the one proposed by Richard et al. [36] to estimate the confined concrete
compressive strength. This equation considers a linear relationship between fcc’ and fle. Because of the non-correlation observed be­
tween fcc’ and RI in Fig. 10b, the proposed numerical model does not consider the fiber dosage to estimate fcc’ of SFRC columns

11
J. Carrillo et al. Journal of Building Engineering 98 (2024) 110993

confined with SSTT. The last term in Eq. (4) is based on Moghaddam et al. [11]. These authors proposed that the peak stress of confined
concrete with SSTT decreases as the parameter (fle/f’co)w where the parameter w can be obtained from statistical regression) and D is a
constant coefficient. Based on the measured data from this study, the exponent ω in Eq. (4) increases the correlation coefficient from
0.90 to 0.93 and reduces the dispersion of samples around the linear trend.
( ) ( )ω
fccʹ fle fle
ʹ=A+C ʹ − D (4)
fco fco fcoʹ
The confining effective pressure fle for Eq. (4) is obtained following Mander et al. [9] recommendation for circular columns using
Eq. (5), where As is the area of the transverse reinforcement and bc is the dimension of the concrete core, measured from
center-to-center of the transverse reinforcement. The confining effective pressure is affected by the coefficient of effectiveness (ke),
which depends on the passive confinement provided by conventional stirrups [9]. Based on Eq. (5), Eq. (6) is proposed in this study to
estimate the effective confining pressure of columns confined with SSTT. In this equation, the term ke from Mander et al. [9] was
replaced by kep and keg , where kep depends on the active lateral confining pressure provided by the steel straps and keg is a geometrical
parameter. In Eq. (6), ρs is the transverse reinforcement ratio calculated using Eqs. (1) and (2) for square (ρt) and cylindrical columns
(ρs). Finally, previous analytical models [11,14,36,32] consider that the confining pressure fle can be estimated using the yield stress fyt
of the transverse reinforcement. Hence, fs is considered as the yield stress of the steel straps. Eqs. (5) and (6) include fyt instead of fs
because fs is already considered in the coefficient kep.
fle = ke fl (5)

ρs fyt
fle = kep keg fl = kep keg (6)
2
The parameter kep in Eq. (6) is defined as the ratio of the stress of the steel straps when the column specimens reach the peak
compressive stress (fsc) and the yield stress of the confining steel (fyt), as shown in Eq. (7). In this study, the parameter kep was measured
in each specimen to observe its dependence with the volumetric ratio of confining steel (ρs). Based on the values of fsc and kep reported
previously in Table 4, Fig. 11 shows the measured values of kep as a function of ρs, and the resulting linear regression (Eq. (7)). The
Person correlation coefficient (r) of the linear regression is 0.85, which is classified as a high correlation [37].
fsc
kep = = 94 ρs (7)
fyt

Fig. 9. Strain of the steel straps of circular columns (a) PC, (b) SFRC15, (c) SFRC30, (d) SFRC60.

12
J. Carrillo et al. Journal of Building Engineering 98 (2024) 110993

The parameter keg in Eq. (6) is the confinement effectiveness coefficient and is defined as the ratio of the area of the effective
confined core and the gross section area of the columns (Ae/Acc) [9]. Previous studies have proposed that the confinement effectiveness
coefficient depends on the cross section of the column and the detailing of the confining reinforcement [9,14,23]. In this study, the
coefficient proposed by Campione et al. [23] is used. Campione et al. [23] proposed a value of keg equal to 0.75 and 1.0 for columns
with square and circular cross-sections, respectively. The value of keg = 0.75 for square columns is based on columns with rounded
corners with rd/b = 0.2.
The coefficients A, C, D and ω of Eq. (4) were obtained using linear regressions. To estimate of coefficients A, C, D and ω, Eq. (4) was
considered as a linear functional form using assumed values of D and ω, as shown in Eq. (8). The coefficients D and ω were estimated
using iterative linear regressions such that the largest Person correlation coefficient was obtained. Therefore, the coefficients D = 10
and ω = 1.0 produced the higher Person correlation coefficient (r = 0.97) of Eq. (8). Then, the coefficients A and C were estimated
using the linear regression that is shown in Fig. 12a. After estimating the coefficients A, C, D and ω, Eq. (4) can be rewritten as Eq. (9).
This latter equation can be used to estimate the peak compressive stress of SFRC columns confined with SSTT.
( )ω ( )
fccʹ fle fle
ʹ+D =A + C (8)
fco fco ʹ
fcoʹ
( )
fccʹ fle
ʹ = 1 + 5.0 (9)
fco fcoʹ
Models to estimate the strain at confined concrete compressive strength (εcc), the strain at 85 % of the peak stress (ε85), and the
strain at 50 % of the peak stress (ε50) for SFRC columns confined with SSTT were not identified in the literature. According to previous
studies with conventional columns [14,32] the strain at peak stress of confined concrete εcc depends on the ratio fle/f’co. Since in this
study it was identified that εcc depends on fle, fcc and RI, Eq. (10) is proposed to estimate εcc for SFRC columns confined with SSTT. This
equation considers the coefficient G equal to 1, relating the original strain of unconfined PC when fcc’/fco’ = 1, the ratio fle/f’co
generated by the SSTT raised to a power φ, and a final term related to RI. Regarding the strains ε85 and ε50, it was shown that these
strains increased as the confined concrete compressive strength of the columns increased (Fig. 8a and Table 4). Since these strains
depend on the confined concrete compressive strength, the functionals of Eqs. (11) and (12) are proposed to estimate ε85 and ε50,
respectively. Eqs. (11) and (12) include the contribution of the effective confinement pressure, the dosage of the steel fibers, and the
ratio f’cc/f’co. In Eq. (11), the H coefficients affect the strains ε85 and ε50, and the coefficients I, J and K considers the contribution of the
SSTT, the steel fibers, and the ratio (f’cc/f’co).
( )φ ( ʹ)ηcc
εcc fle fcc
=E + F × RI + G (10)
εco fco ʹ
fcoʹ
( ) ( )η85
ε85 fle fccʹ
= H85 + I85 + J85 RI + K85 (11)
εco fco ʹ
fcoʹ
( ) ( )η50
ε50 fle fccʹ
= H50 + I50 + J50 RI + K50 (12)
εco fco ʹ
fcoʹ
Similar to Eq. (8), Eqs. (10)–(12) were adapted to a linear functional form, as shown in Eqs. (13)–(15), respectively. In Eqs. (13) and
(14), the independent variable is fle/fco’ while, in Eq. (15), the independent variable is fcc’/fco’ to obtain a higher correlation coefficient
(r). The coefficients E, F, G, φ and ηcc of Eq. (10), as well as the coefficients H, I, J, K and η of Eqs. (11) and (12) were determined from an
iterative linear regression analysis, such that the highest possible Pearson coefficients (r) were obtained. Values of E, F, G, φ and ηcc of
2.8, 1/240, 1, 0.5 and 0.5, respectively, generated an r of 0.94 in Eq. (13), Fig. 12b. Moreover, values of I85, J85, k85 and η85 of 25, 1/
500, 1/10 and 0.5, respectively, generated an r of 0.64 in Eq. (14), Fig. 12c. Finally, values of I50, J50, k50 and η50 of 1, 1/500, 1 and 2,
respectively, generated an r of 0.69 in Eq. (15), Fig. 12d. The low value of r in the regressions of Fig. 12c and d is due to the fact that the
deformations measured after cracking presented large variations compared to the parameters measured at the point of maximum
compressive strength. The H50, H85, I50 and I85 coefficients were determined from the equations obtained from the iterative linear
regression analyses in Fig. 12c and d. Replacing in Eqs. (10)–(12) the estimated coefficients with the linear regressions of Fig. 12b, c
and 12d, Eqs. (16)–(18) are obtained to estimate εcc, ε85 and ε50.
( ʹ) ( )φ
εcc fcc fle
− G RI − =E +F (13)
εco fco ʹ
fcoʹ
( )η85 ( )
ε85 fccʹ fle
− J85 RI − K85 = H85 + I85 (14)
εco fcoʹ fco ʹ

( ) ( )η50
ε50 fle fccʹ
− J50 RI − I50 = H50 + K50 (15)
εco fcoʹ fcoʹ

13
J. Carrillo et al. Journal of Building Engineering 98 (2024) 110993

Table 5
Database of compressive tests in actively confined columns.

Study Geometry Concrete properties Peak conditions

Shape D or L, mm H, mm fco’, MPa εco’, mm/mm fl’, MPa fcc’, MPa εcc’, mm/mm
Sarmah et al. [3] Cylindrical 150 300 21.7 0.0016 0.00 21.7 0.0016
21.7 0.0016 1.09 27.1 0.0032
21.7 0.0016 1.76 30.3 0.0040
21.7 0.0016 2.44 36.6 0.0048
26.3 0.0017 0.00 26.3 0.0017
26.3 0.0017 1.09 33.2 0.0031
26.3 0.0017 1.76 36.6 0.0037
26.3 0.0017 2.44 40.4 0.0045
30.3 0.0018 0.00 30.3 0.0018
30.3 0.0018 1.09 39.5 0.0029
30.3 0.0018 1.76 43.6 0.0035
30.3 0.0018 2.44 48.3 0.0046

Ekkachai et al. [13] Cylindrical 150 300 11.4 0.0067 0.00 11.4 0.0067
11.4 0.0067 1.37 16.3 0.0078
11.4 0.0067 2.46 19.0 0.0115
11.4 0.0067 5.25 27.0 0.0183
7.4 0.0042 0.00 7.4 0.0042
7.4 0.0042 1.37 13.0 0.0133
7.4 0.0042 2.46 16.2 0.0165
7.4 0.0042 5.25 23.1 0.0184
6.6 0.0045 0.00 6.6 0.0045
6.6 0.0045 1.37 13.4 0.0122
6.6 0.0045 2.46 15.4 0.017
6.6 0.0045 5.25 22.5 0.0247
9.0 0.0037 0.00 9.0 0.0037
9.0 0.0037 1.37 13.8 0.0065
9.0 0.0037 2.46 19.7 0.0117
9.0 0.0037 5.25 24.5 0.0187
5.6 0.004 0.00 5.6 0.004
5.6 0.004 1.37 11.9 0.008
5.6 0.004 2.46 13.3 0.0148
5.6 0.004 5.25 19.7 0.0180

Moghaddam et al. [11] Cylindrical 100 200 50.0 – 0.00 50.0 1.00 εo
50.0 – 2.44 74.0 1.54 εo
50.0 – 1.22 60.0 1.18 εo
50.0 – 0.81 64.0 0.78 εo
50.0 – 0.61 54.0 1.01 εo
50.0 – 2.44 103.0 3.89 εo
50.0 – 0.81 60.0 1.23 εo
50.0 – 2.44 80.5 1.90 εo
50.0 – 1.22 77.0 1.06 εo
50.0 – 0.81 64.5 0.95 εo
50.0 – 0.61 50.0 0.73 εo
50.0 – 2.44 116.0 1.68 εo
50.0 – 0.81 77.0 1.12 εo
50.0 – 0.81 65.5 1.57 εo
50.0 – 0.81 76.5 1.43 εo
50.0 – 1.22 68.5 1.43 εo
50.0 – 0.81 65.5 1.37 εo
50.0 – 6.50 86.0 13.57 εoa
50.0 – 2.17 69.0 1.68 εo

Prismatic 100 200 50.0 – 0.00 50.0 1.00 εo

50.0 – 1.76 92.5 1.07 εo
50.0 – 0.88 84.5 2.03 εo
50.0 – 0.59 72.5 0.79 εo
50.0 – 0.44 56.5 1.14 εo
50.0 – 1.76 97.5 1.66 εo
50.0 – 0.88 73.0 1.10 εo
50.0 – 0.59 70.0 1.17 εo
50.0 – 0.44 63.5 1.00 εo
50.0 – 0.88 79.0 0.93 εo
50.0 – 0.88 71.0 1.17 εo
50.0 – 2.35 90.0 1.14 εo
50.0 – 1.18 75.5 0.93 εo

(continued on next page)

14
J. Carrillo et al. Journal of Building Engineering 98 (2024) 110993

Table 5 (continued )

Study Geometry Concrete properties Peak conditions

Shape D or L, mm H, mm fco’, MPa εco’, mm/mm fl’, MPa fcc’, MPa εcc’, mm/mm
Cylindrical 152 305 10.0 – 0.00 10.0 1.00 εo
10.0 – 1.63 33.3 7.63 εo
10.0 – 1.63 59.3 12.54 εoa
10.0 – 0.40 21.1 3.28 εo

Prismatic 150 300 10.0 – 0.00 10.0 1.00 εo

10.0 – 1.18 23.7 44.17 εoa
10.0 – 1.18 20.1 25.83 εoa
10.0 – 2.35 30.0 50.83 εoa
10.0 – 1.18 17.6 68.33 εoa
10.0 – 0.78 14.5 3.33 εo
10.0 – 1.18 16.7 48.33 εoa
10.0 – 1.76 14.2 5.00 εo
10.0 – 0.88 13.7 3.33 εo
10.0 – 1.76 26.0 74.17 εoa
10.0 – 0.88 14.1 40.83 εoa
10.0 – 0.59 13.7 2.50 εo
25.0 – 0.00 25.0 1.00 εo
25.0 – 0.88 32.8 1.28 εo
25.0 – 1.76 57.5 2.50 εo
25.0 – 2.35 54.8 10.94 εo
25.0 – 1.18 41.8 6.28 εo
35.0 – 0.00 35.0 1.00 εo
35.0 – 1.76 64.8 1.68 εo
35.0 – 0.88 63.4 1.73 εo
35.0 – 1.18 58.1 1.36 εo
35.0 – 2.35 78.4 3.95 εo

Chin et al. [33] Cylindrical 100 200 55.0 – 1.29 105.2 –

55.0 – 1.29 102.3 –
55.0 – 8.12 114.5 –
55.0 – 8.12 108.8 –

Chin et al. [34] Prismatic 100 200 69.9 0.257 0.00 68.6 0.249
69.9 0.257 0.00 71.1 0.265
69.9 0.257 1.23 73.0 0.258
69.9 0.257 1.23 72.4 0.262
69.9 0.257 1.76 80.9 0.333
69.9 0.257 1.76 82.5 0.318
69.9 0.257 2.71 77.4 0.535
69.9 0.257 2.71 79.8 0.552
69.9 0.257 2.71 85.9 0.485
69.9 0.257 2.71 84.5 0.608
69.9 0.257 2.71 91.2 0.453
69.9 0.257 2.71 91.6 0.456

Neupane et al. [35] Cylindrical 150 300 17.8 0.005 0.00 17.8 0.005
17.8 0.005 0.73 23.0 0.009
17.8 0.005 1.64 26.7 0.009
17.8 0.005 3.84 36.5 0.009
17.8 0.005 15.36 52.8 0.012
22.2 0.004 0.00 22.2 0.004
22.2 0.004 0.73 25.3 0.008
22.2 0.004 1.64 29.2 0.007
22.2 0.004 3.84 39.4 0.009
22.2 0.004 15.36 57.7 0.008

- Not reported in the study.

a
Values excluded from the calibration of the proposed model.

( )0.5 ( ʹ)0.5
εcc fle 1 fcc
= 2.8 + RI + (16)
εco fcoʹ 240 fcoʹ
( ) ( )0.5
ε85 fle 1 1 fccʹ
= 1 + 25 + RI + (17)
εco fco ʹ 500 10 fcoʹ
( ) ( )
ε50 fle 1 fccʹ
=3+ + RI + (18)
εco fco ʹ 500 fcoʹ

15
J. Carrillo et al. Journal of Building Engineering 98 (2024) 110993

Fig. 10. Peak stress and strain at peak stress: (a) fcc’ versus normalized confining pressure and comparison with existing analytical models, (b) fcc’ versus RI, (c) εcc
versus RI.

Fig. 11. Relationship between kep and volumetric ratio of confining steel.

4.3. Comparison between proposed and existing models

The ratios between the predicted values and the measured values for f’cc and εcc using the proposed equations are compared with
the ratios estimated using ten existing models [1,3,4,9,11,13,14,32,33,35]. It is important to note that the models of Mander et al. [9],
Saatcioglu et al. [14], Hoshikuma et al. [32] and Moghaddam et al. [11] are based on passive confinement. Although some models are
currently proposed to estimate the compressive behavior of SFRC [6,20,21], these models exclude the interaction between steel fibers
and concrete confined using SSTT. The ratio between predicted versus observed (P/O) values is evaluated using boxplots; these
boxplots allow comparing the mean (X) of the P/O ratios, the standard deviation (S), the maximum and minimum values, and the
overprediction (Op). The overprediction is equal to the quotient between the number of points with P/O larger than 1.05 and the total
number of points and is expressed as a percentage.
The boxplots for the P/O ratios for f’cc are shown in Fig. 13a. As expected because of the regression model, Fig. 13a shows that the
average P/O ratio for f’cc calculated with the proposed model is equal to 1.0. The figure also shows that the average P/O ratio for f’cc
using the models of Mander et al. [9], Saatcioglu et al. [14], Hoshikuma et al. [32], Moghaddam et al. [11], Sarmah et al. [3], Nazirah
et al. [1], Ekkachai et al. [13], Lim and Togay [4], Chin et al. [33] and Neupane et al. [35] are equal to 1.02, 0.92, 0.76, 1.04, 1.05,
0.49, 0.78, 1.03, 0.87 and 0.87, respectively. Additionally, the standard deviation obtained with these ten models are 53.2 %, 19.7 %,
19.8 %, 19.6 %, 27.2 %, 109.3 %, 21.3 %, 19.6 %, 20.5 % and 26.2 %, respectively. These standard variations are higher than the
standard deviation of 19.3 % obtained with the proposed model. A drawback identified for the Mander et al. [9], Moghaddam et al.,
Sarmah et al., and Lim and Togay models is their large overprediction (Op) of 27.9 %, 27.2 %, 27.9 % and 33.1 %, respectively.
The overprediction of the models of Saatcioglu et al. [14], Hoshikuma et al. [32], Nazirah et al. [1] and Ekkachai et al. [13] (11.8 %,
0.7 %, 14.0 % and 3.7 %, respectively), as well as their average P/O ratios (0.92, 0.76, 0.49 and 0.78, respectively) show an over­
conservative estimate of the compressive strength. This overconservative estimate is related to the nature of these models which are
based on either passive confinement or active confinement in degraded or recycled concrete.
Fig. 13b shows the comparison of the P/O boxplot for εcc obtained with the proposed model in this study and those obtained with
the eight considered models. The figure shows that the average P/O ratio obtained for the proposed model is 1.0 and the average ratio
obtained with the model by Mander et al. [9] and Saatcioglu et al. [14] are 1.28 and 1.49, respectively. Additionally, the over­
prediction for these two models is large and equivalent to 62.9 % and 68.9 %, respectively. Therefore, the models proposed by Mander
et al. and Saatcioglu et al. are not appropriate to estimate the strain at peak strength of SFRC columns with SSTT. On the other hand, the
average P/O obtained with the models by Hoshikuma et al. [32], Moghaddam et al. [11], Nazirah et al. [1] and Ekkachai et al. [13] are
0.95, 0.85, 0.88 and 0.60, respectively. These values smaller than 1.0 indicate that these models provide conservative estimations of

16
J. Carrillo et al. Journal of Building Engineering 98 (2024) 110993

Fig. 12. Estimation of the regression coefficient: (a) f’cc/f’co, (b)εcc/εco, (c)ε85/εco, (d)ε50/εco.

εcc. This underestimation is explained because these models do not consider the SSTT and the SFRC. It is concluded that the proposed
model is adequate to estimate the strain at peak strength because of the accurate P/O, the lowest standard deviation of the ratios (40.4
%) and overprediction (38.6 %).
Fig. 13c shows the boxplot obtained with the models proposed to estimate the strains ε85 and ε50. These results are not compared
with other models because models to estimate the strains ε85 and ε50 of confined concrete were not identified in the literature. The
average P/O ratio obtained by the proposed model for ε85 is 0.90, which indicates that the model is conservative to predict the strain at
85 % of the peak strength. Additionally, the standard deviation of the P/O ratio for ε85 (16.1 %) and the overprediction (10 %) are
relatively small. The average P/O ratio obtained by the proposed model for ε50 is close to one (0.97), which indicates that the proposed
model is slightly conservative to estimate the strain at 50 % of the peak strength. Moreover, the standard deviation of the P/O ratio and
the Op for ε50 are larger than for ε85 by 33.8 % and 35 %, respectively.
For code-based design, this study also proposes Eqs. (19)–(22) to estimate fcc’, εcc, ε85 and ε50, respectively. Eqs. (19)–(22) are
similar to Eqs. (9) and (16)–(18) but they represent a characteristic response for which the over-prediction is lower than 5 %.
( ) ( )1.5
fccʹ fle fle
ʹ = 1 + 2.6 ʹ - 0.5 (19)
fco fco fcoʹ
( )0.5 ( ʹ)0.5
εcc fle 1 fcc
=2 + RI + 0.6 (20)
εco fcoʹ 240 fcoʹ
( ) ( )0.5
ε85 fle 1 1 fccʹ
= 1 + 10 + RI + (21)
εco fcoʹ 500 20 fcoʹ
( ) ( )
ε50 fle 1 fccʹ
=2+ + RI + 0.5 (22)
εco fcoʹ 500 fcoʹ

17
J. Carrillo et al. Journal of Building Engineering 98 (2024) 110993

Fig. 13. Evaluations of the predictions of: (a) f’cc, (b) εcc, (c) ε85 and ε50.

5. Conclusions
This study reports the confined concrete compressive strength and strain (f’cc and εcc), the stress and strain at 15 % strength loss (f’85
and ε85), the stress and strain at 50 % strength loss (f’50 and ε50), and the ultimate strength and strain (f’cu and εcu) of square and
circular steel fiber reinforced concrete (SFRC) columns confined with steel strapping tensioning technique (SSTT). Test results showed
that the f’cc of the SFRC column specimens confined with SSTT increased between 4 % and 18 % when compared to the unconfined
specimens with plain concrete (PC). Additionally, the ductility of the columns confined with SSTT was about eight times larger than
that of the unconfined columns. The column specimens with unconfined PC showed a brittle compressive failure because εcc was only
0.0027 mm/mm and the columns were unable to reach ε85 and ε50.
Experimental results also provide evidence of contribution of steel fibers to reducing the strength degradation of the column
specimens. Additionally, the use of fibers aided at increasing the ductility of the column specimens. When adding steel fibers, the ε50 of
the unconfined SFRC columns increased to 0.0242 mm/m. In addition, the average ε50 of the SFRC columns confined with SFRC further
increased to 0.0447 mm/mm. The ultimate deformation of the SFRC column specimens with a fiber dosage of 60 kg/m3 was nine times
larger than the ultimate deformation of the column specimens of PC. The use of steel fibers also contributed to the effectiveness of the
confinement provided by the SSTT by controlling the microcracks of the concrete and avoiding the loss of stress of the steel straps.
Furthermore, the experimental results showed that the contribution of the steel fibers to the increase of the confined concrete strength
(fcc’) was less effective in the column specimens with larger spacing of the steel straps.
A database of 116 tests from previous studies and 34 test from experimental program reported herein, resulting in a total of 150
tests was used to propose models for estimating fcc’ and strains related to the maximum strength (εcc) and the post-peak strains at 85 %
and 50 % degradation (ε85 and ε50, respectively) for SFRC columns confined with SSTT. The functional form of the fcc’ estimation
model demonstrated the noticeable relationship between fcc’, fco’ and fle as suggested by the models of Richard et al. [36], Moghaddam
et al. [11] and Saatcioglu et al. [14]. The equation to estimate fle include coefficients of effectiveness by geometry (keg) and by lateral
confining pressure (kep). To consider the large dispersion of the results, a set of code-based equations calibrated to estimate the
response of the specimens with maximum over prediction of 5 % was also proposed. The equations proposed in this study can be used
for assessing the response of normal-weight concrete with fc’ values between 38.6 and 46.6 MPa, and SFRC with fibers with a hook at
the ends and dosages lower than 60 kg/m3. The equations proposed in this study are also limited to the use of SSTT system with steel
straps with characteristics similar to those used in this study.

CRediT authorship contribution statement

Julian Carrillo: Writing – original draft, Formal analysis, Conceptualization. Matías Hube: Writing – review & editing, Data
curation. Carlos Blandon: Writing – review & editing, Formal analysis. Ramon Mata: Writing – review & editing, Validation. Joa­
quin Abellan-Garcia: Writing – review & editing, Validation.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to
influence the work reported in this paper.

Data availability

Data will be made available on request.

Acknowledgements
The authors thank to Vicerrectoría de Investigaciones of UMNG for the financial support of the IMP-ING-3743 project. The authors

18
J. Carrillo et al. Journal of Building Engineering 98 (2024) 110993

also would like to thank Research Assistants Juan D. Vargas and Alejandro Fierro at Universidad Militar Nueva Granada (UMNG) for
data and figure processing. The authors also thank the companies Argos-Colombia and Proalco-Bekaert-Colombia for supplying the
concrete and steel fibers, respectively. The authors acknowledge USI Library of Institute of Engineering at UNAM for providing some
references. All data that support the findings of this study are available from the corresponding author upon reasonable request. The
views expressed in this paper are solely those of the authors and do not necessarily reflect the views of the sponsors.

References
[1] M. Nazirah, M. Chau-Khun, C. Chee-Loong, Z. Abdullah, W. Warid, Stress strain response of pre-damaged concrete confined with recycled steel straps - a green
confining material, Building Engineering 75 (2023) 1–22, https://doi.org/10.1016/j.jobe.2023.106760.
[2] M. Frangous, K. Pilakoutas, S. Dritsos, Structural repair/strengthening of RC columns, Construct. Build. Mater. 9 (5) (1995) 259–266, https://doi.org/10.1016/
0950-0618(95)00013-6.
[3] M. Sarmah, A. Dutta, S. Deb, Axial stress–strain model for concrete actively confined with Fe-SMA strips, Materials in Civil Engineering 35 (12) (2023) 1–15,
https://doi.org/10.1061/JMCEE7.MTENG-1608.
[4] J. Lim, O. Togay, Unified stress-strain model for FRP and actively confined normal-strength and high-strength concrete, Composites for Construction 19 (4)
(2015) 1–14, https://doi.org/10.1061/(ASCE)CC.1943-5614.0000536.
[5] E. Yooprasertchai, A. Ejaz, P. Saingam, A. Wai, P. Joyklad, Development of stress-strain models for concrete columns externally strengthened with steel clamps,
Construct. Build. Mater. 377 (2023) 131155 259–266, https://doi.org/10.1016/j.conbuildmat.2023.131155.
[6] A. Ezeldin, P. Balaguru, Normal and high-strength fiber-reinforced concrete under compression, Materials in Civil Engineering 11 (1) (1992) 21–29, https://doi.
org/10.1061/(ASCE)0899-1561(1992)4:4(415).
[7] R. Abbasnia, H. Ziaadiny, Experimental investigation and strength modeling of CFRP confined concrete rectangular prisms under axial monotonic compression,
Mater. Struct. 48 (2015) 485–500, https://doi.org/10.1617/s11527-013-0198-y.
[8] K. Toska, F. Faleschini, A new confinement model for FRCM confined concrete, Mater. Struct. 56 (98) (2023) 1–22, https://doi.org/10.1617/s11527-023-
02186-w.
[9] J.B. Mander, J.N. Priestley, R. Park, Observed stress-strain behavior of confined concrete, Struct. Eng. Earthq. Eng. 114 (1988) 1827–1849, https://doi.org/
10.1061/(ASCE)0733-9445(1988)114:8(1827).
[10] J. Jianghao, H. Tianhao, D. Zhiqiang, Z. Hong, W. Gang, W. Yang, S. Chee-Kiong, Performance of concrete columns actively strengthened with hoop
confinement: a state-of-the-art review, Structures 54 (2023) 461–477, https://doi.org/10.1016/j.istruc.2023.05.038.
[11] H. Moghaddam, M. Samadi, K. Pilakoutas, Compressive behavior of concrete actively confined by metal strips; Part B: analysis, Mater. Struct. 43 (2010)
1383–1396, https://doi.org/10.1617/s11527-010-9589-5.
[12] H.-P. Lee, A. Awang, W. Omar, Experimental investigation on SSTT confined concrete with low lateral pre-tensioning stresses, Teknologi (Sciences and
Engineering) 69 (3) (2014) 43–50, https://doi.org/10.11113/jt.v69.3142x.
[13] Y. Ekkachai, E. Ali, S. Panumas, W. Anne, J. Panuwat, Development of stress-strain models for concrete columns externally strengthened with steel clamps,
Construct. Build. Mater. 377 (2023) 1–14, https://doi.org/10.1016/j.conbuildmat.2023.131155.
[14] M. Saatcioglu, S. Razvi, Strength and ductility of confined concrete, Struct. Eng. Earthq. Eng. 118 (6) (1992) 1590–1607, https://doi.org/10.1061/(ASCE)0733-
9445(1992)118:6(1590).
[15] A. Awang, W. Omar, H.-P. Lee, C. Ma, Behavior of externally-confined high strength concrete column under uniaxial compression load. The 8th Asia Pacific
Structural Engineering and Construction Conference (APSEC 2012) and the 1st International Conference of Civil Engineering Research (ICCER 2012), Surabaya,
Indonesia, 2012.
[16] P. Foraboschi, Optimal design of seismic resistant RC columns, Concrete and Construction Materials 13 (8) (2020) 1919, https://doi.org/10.3390/ma13081919.
[17] J. Carrillo, J. y Alcocer S. Vargas, Model for estimating the flexural performance of concrete reinforced with hooked end steel fibers using three-point bending
tests, Struct. Concr. 22 (2021) 1760–1783, https://doi.org/10.1002/suco.202000432.
[18] J. Carrillo, E. Osorio, C. Graciano, Behavior of square and low-strength concrete columns reinforced with hybrid steel bars and micro-fibers, Arabian J. Sci. Eng.
45 (2020) 8443–8456, https://doi.org/10.1007/s13369-020-04748-4.
[19] ACI-544, State-of-the-art Report on Fiber Reinforced Concrete, American Concrete Institute, ACI, USA, 2018, pp. 1–44. ISBN 9781641950190.
[20] M. Nataraja, N. Dhang, A. Gupta, Stress-strain curve for fiber reinforce concrete in compression, Cement and Concrete Composite 21 (1999) 383–390, https://
doi.org/10.1016/S0958-9465(99)00021-9.
[21] J. Carrillo, G. Gonzáles, L. Gallo-Arciniegas, Behavior of ZP-306 steel fiber reinforced concrete subjected to compressive stresses, Ciencia e Ingeniería
Neogranadina 23 (2013) 117–133. ISSN 0124-8170.
[22] ACI 318, Building Code Requirements for Structural Concrete, American Concrete Institute, ACI, 2019, pp. 1–684. ISSBN 9781641950602.
[23] G. Campione, N. Miragli, Strength and strain capacities of concrete compression members reinforced with FRP, Cement Concr. Compos. 25 (2003) 31–41,
https://doi.org/10.1016/S0958-9465(01)00048-8.
[24] NSR-10, Colombian Code for Earthquake-Resistant Construction, AIS, Colombian Association of Earthquake Engineering, Colombia, 2010, pp. 1–2130. ISBN:
978-958-97609-3-2 (in Spanish).
[25] ASTM C-1602, Standard Specification for Mixing Water Used in the Production of Hydraulic Cement Concrete, American Society of Testing Materials, ASTM,
2022, pp. 1–5, https://doi.org/10.1520/C1602_C1602M-22.
[26] ASTM E8-22, Standard Test Methods for Tension Testing of Metallic Materials, American Society of Testing Materials, ASTM, 2022, pp. 1–31, https://doi.org/
10.1520/E0008_E0008M-22.
[27] ASTM C39, Standard Test Method for Compressive Strength of Cylindrical Concrete Specimens, American Society of Testing Materials, ASTM, 2021, pp. 1–8,
https://doi.org/10.1520/C0039_C0039M-21.
[28] ASTM C469, Standard Test Method for Static Modulus of Elasticity and Poisson’s Ratio of Concrete in Compression, American Society of Testing Materials,
ASTM, 2022, pp. 1–6, https://doi.org/10.1520/C0469_C0469M-22.
[29] A. Fierro, Confinement of Plain and Fiber Reinforced Concrete Using Steel Straps, Bch Thesis, Universidad Militar Nueva Granada, Colombia, 2020 (in Spanish).
[30] V. Jensen, The plasticity ratio of concrete and it’s effect on the ultimate strength of beams, Journal Proceedings 39 (1943) 565–584, https://doi.org/10.14359/
8644.
[31] F. Raoul, L. Stéphane, D. Fabrice, 5 - effects of reinforcement corrosion on the mechanical behavior of reinforced concrete, Corrosion and its Consequences for
Reinforced Concrete Structures (2018) 105–133, https://doi.org/10.1016/B978-1-78548-234-2.50005-6.
[32] J. Hoshikuma, K. Kawashima, K. Nagaya, A.W. Taylor, Stress-strain model of confined reinforced concrete in bridge piers, Struct. Eng. Earthq. Eng. 123 (5)
(1997) 624–633.
[33] C. Chin, C. Ong, J. Tan, C. Ma, A. Awang, W. Omar, Confinement-concrete interaction in pre-tensioned partial steel-confined concrete, Structures 23 (2020)
751–765, https://doi.org/10.1016/j.istruc.2019.12.006.
[34] C. Chin, C. Ma, A. Awang, J. Tan, C. Ong, W. Omar, Confining stress path dependent stress-strain model for pre-tensioned steel-confined concrete, Eng. Struct.
201 (2019) 109769, https://doi.org/10.1016/j.engstruct.2019.109769.

19
J. Carrillo et al. Journal of Building Engineering 98 (2024) 110993

[35] R. Neupane, T. Imjai, R. Garcia, J. Alengaram, A new constitutive model for recycled aggregate concrete cylinders actively confined with post-tensioned metal
straps, Sustainable Structures 4 (2) (2024) 000049, https://doi.org/10.54113/j.sust.2024.000049.
[36] F.E. Richard, A. Brandtzaeg, R.L. Brown, A study of the failure of concrete under combined compressive stresses, Bulletin No. 185, Engineering Experimental
Station 26 (12) (1928) 1–116.
[37] D. Rowntree, Introduction to Statistics: a Non-mathematical Approach, first ed., Norma Ed., 1984. Bogotá (in Spanish). ISBN 9780141987491.

20