International Journal of Steel Structures (2018) 18(3):891–909 Online ISSN 2093-6311

https://doi.org/10.1007/s13296-018-0034-x Print ISSN 1598-2351

A Novel Approach for Bolted T‑Stub Connections

Merve Sağıroğlu1 · Mahyar Maali2 · Mahmut Kılıç2 · Abdulkadir Cüneyt Aydın2

Received: 20 March 2017 / Accepted: 28 January 2018 / Published online: 24 April 2018
© Korean Society of Steel Construction 2018

Abstract
This paper reports an investigation into new connection types and their behaviors determined using full-scale experiments.
T-shaped connections were created using the IPE standard profile. The aim of this study was to analyze the influence of T
connections based on the IPE standard profile, height of beam to height of T-stub joint (H) of T-stub joints, and lengths (X)
of T-stub joints on the behavior of steel connections, in order to provide the necessary data for improving Eurocode 3 and
enable efficient use of residue IPE standard profiles and back to the consumption cycle. While the moment resistance values
increased with an increase in H from ­Hmin to ­Hmax in model groups with X of 126 mm, and the energy dissipation increased
with an increase in H from ­Hmin to ­Hmax and also with an increase in the lengths (X) of T-stub joints from 54 to 126 mm.

Keywords Experiment · Moment–rotation curves · IPE standard profile · T-shaped connection · Eurocode 3

List of symbols ӨMj.max Rotation of the connection at maximum load

Fu Ultimate or tensile stress ӨMj.Rd Connection rotation analytical value at which
Fy Yield stress the moment resistance first reaches ­Mj.Rd
Est Strain hardening modulus Өmin K−R Rotation between the lower bound of the knee-
ρy = fy/fu Yield ratio range of the joint moment–rotation curve and
εst Strain at the strain hardening point the rotation capacity
εuni Uniform strain Өsup K−R Rotation between the upper bound of the knee-
εf Strain at rupture load range of the joint moment–rotation curve and
X Cartesian axis; distance the rotation capacity
E Young’s modulus Ψj.maxload Joint ductility index evaluated for rotation at
I Moment of inertia maximum load
Mj.Rd Joint flexural plastic (design) resistance Ψj Joint ductility index
Mj.max Maximum bending moment
Mmin.K−R Lower resistance bound of the knee-range of
the joint moment–rotation curve 1 Introduction
Msup.K−R Upper resistance bound of the knee-range of
the joint moment–rotation curve Traditionally, steel portal frames have been designed assum-
MӨ.Cd Bending moment at fracture of the joint ing that beam-to-column joints are rigid or nominally pinned
Sj.ini Initial rotational stiffness of a joint joints. Rigid joints are capable of transferring axial forces
Sj.p−1 Post-yield rotational stiffness of a joint as well as moments; this means that joints classified as rigid
ӨCd Rotation capacity of a connection may be assumed to have sufficient rotational stiffness to jus-
tify analysis based on full continuity (Fig. 1). In contrast,
nominally pinned joints are capable of transferring internal
* Abdulkadir Cüneyt Aydın axial forces but cannot transfer moments; this means that a
[email protected] nominally pinned joint is capable of accepting the result-
1
Department of Civil Engineering, Faculty of Engineering
ing rotations under the design loads (Fig. 1). Despite these
and Architecture, Erzurum Technical University, Erzurum, facts, it is largely recognized that a great majority of joints
Turkey do not exhibit such idealized behavior. In fact, many joints
2
Department of Civil Engineering, Faculty of Engineering, transfer some bending moments associated with rotations
Ataturk University, 25240 Erzurum, Turkey

13
Vol.:(0123456789)
892 International Journal of Steel Structures (2018) 18(3):891–909

90º
90º

Rigid Pinned Semi-rigid

Fig. 1  Joint types

(De Limaa et al. 2002). These joints are called semi-rigid component that is located in the tension zone of blind-bolted
connections (Fig. 1), and the use of semi-rigid connections connections to concrete-filled tubular steel profiles. The aim
provides a good response under design loads. The behavior was to provide necessary data to improve Eurocode 3 (2005).
of connections should be considered in the design and analy- T-shaped connections created using the IPE standard profile
sis of steel frames because it represents the actual behavior and T-shaped elements are different from those reported in
(Sagiroglu and Aydin 2015). In recent years, a number of the literature, which utilize welded plates. Thus, elimination
studies have been carried out to examine the behavior of of problems occurring at the welds of connections such as
semi-rigid connections. the occurrence of fracture points and the inability to perform
De Limaa et al. (2002) studied an experimental and well in place is expected (Coelho et al. 2004). To recom-
mechanical model for predicting the behavior of minor axis mend the use of weld-less T connections, knowledge of their
beam-to-column semi-rigid joints. This theory implies that behavior is needed. However, unlike in this paper, T-stub
each joint component is represented by a spring possessing connections based on the IPE standard profile are not men-
a nonlinear force versus displacement (F–∆) curve. Coelho tioned or investigated in either Eurocode 3 or the literature.
et al. (2004) assessed the behavior of 32 bolted T-stub con- Thus, the aim of this study was to analyze the influence
nections made of welded plates. Although T-stubs have been of T connections based on the IPE standard profile, height
used for many years to model the tension zone of bolted of beam to height of T-stub joint (H) of T-stub joints, and
joints, the study mainly focused on rolled profiles as T-stub lengths (X) of T-stub joints on the behavior of steel connec-
elements. To extend this model to the case of welded plates tions, in order to provide the necessary data for improving
as T-stub elements, a test program was undertaken. Swanson Eurocode 3 and enable efficient use of residue IPE standard
et al. (2002) conducted several tests on individual T-stubs to profiles and back to the consumption cycle. Moment–rota-
show that the behavior of T-stubs was partly influenced by tion curves were used to evaluate the main parameters char-
the bending of the bolts. Herrera et al. (2013) investigated acterizing the behavior of the tested connections, such as
the behavior of built-up T-stubs subjected to tensile load- the moment resistance, rotational stiffness, rotation capacity,
ing by using numerical and experimental models. Recently, and ductility of a joint.
Aydin et al. (2015a) experimentally assessed the behavior of
semi-rigid connections with angles and stiffeners. This study
was undertaken to analyze the influence of angles with and 2 Experimental Investigation
without stiffeners on the behavior of the beam-to-column
connections. Katula and Dunai (2015) The paper presents 2.1 Test Details
the details and results of an experimental study on bolted
end-plate joints of industrial type steel building frames Liu Eighteen experimental models were developed to predict the
et al. (2014). This paper examines the behaviour of two types behavior of bolted T-stub connections under static loading
of practical open beam-to-tubular column connection details in six groups, labeled T140-2-10, T140-1-10, T160-2-10,
subjected to combined moment, axial and/or shear loads T160-1-10, T200-2-10, and T200-1-10 groups. In the speci-
Drosopoulos et al. (2012). This paper presents a numeri- men name of the form TA-B-C, TA represents the T-stub
cal investigation of the mechanical behaviour of extended connection cutting of IPE standard profile, B is the number
end—plate steel connections including comparison with full of row bolts in T-stub connections, and C is the diagonal
size experiments Pitrakkos and Tizani (2015). This paper of bolts. (Thus, T140 cut from IPE140 standard profile,
presents a component-method model for a structural joint T160 cut from IPE160 standard profile, and T200 cut from

13
International Journal of Steel Structures (2018) 18(3):891–909 893

IPE200 standard profile). Figure 2 shows the experimental hand-tightened full-threaded grade 8.8 M10 bolts in 12 mm
program, and Table 1 gives the test details. The behaviors drilled holes were kept constant for all tested specimens.
of the T-stub joints were compared within their groups. Col-
umn stiffeners with a thickness of 10 mm were welded to 2.2 Mechanical Properties
the column by means of a continuous 45° fillet weld. Thus,
the columns were chosen with large cross-section, and the A coupon tension test of the structural steel material of the
use of stiffeners prevented excessive deflection in the flange column stiffener, T-stub profile, and flange and web of the
column. The fillet welding was performed from the upper beam and column were performed in compliance with UNE-
side of the joint in the down-hand position for the workshop. EN 10002-1 (2002). The average characteristic values for
A consumable electrode was used in the manual metal arc structural steels and bolt (8.8) are listed in Table 2. Each
welding process (Maali et al. 2015, 2016). A plate stiffener, bolt (8.8) was tested under tension in order to determine the
T-stub profile, and profile section were S235. HE160B was mechanical properties of the bolt material, in accordance
selected for columns, and IPE160 was selected for beams; with ISO 898-1999(E) (2009).

A
A-A
A A
Beam IPE160
3.5d 2d
2d STR1

STR4 h hy
STR2

x
A
STR3

A STR2
A-A x/2
2d

STR3 hy Beam IPE 160

STR4 h
STR1

x
A

Fig. 2  T-stub geometry, proposed three-dimensional semi-rigid joint, and location of strain gauges (STR1 = parallel to beam on T-stub,
STR2 = vertical to beam on T-stub, STR3 = parallel to column on T-stub, STR4 = vertical to column on T-stub, and d = diagonal of bolts)

13
894 International Journal of Steel Structures (2018) 18(3):891–909

Table 1  Test details Group name Experiment T-stub joint Hmax = hymax/h Hmin = hymin/h Hav = Hyav/h X (mm)

T140-2-10 T140-Hmax-126 IPE 140 1 – – 126

T140-Hmin-126 – 0.63 –
T140-Hav-126 – – 0.82
T140-1-10 T140-Hmax-54 IPE 140 1 – – 54
T140-Hmin-54 – 0.57 –
T140-Hav-54 – 0.79
T160-2-10 T160-Hmax-126 IPE 160 1 – – 126
T160-Hmin-126 – 0.63 –
T160-Hav-126 – 0.82
T160-1-10 T160-Hmax-54 IPE 160 1 – – 54
T160-Hmin-54 – 0.57 –
T160-Hav-54 – – 0.79
T200-2-10 T200-Hmax-126 IPE 200 1 – – 126
T200-Hmin-126 – 0.63 –
T200-Hav-126 – – 0.82
T200-1-10 T200-Hmax-54 IPE 200 1 – – 54
T200-Hmin-54 – 0.57 –
T200-Hav-54 – – 0.79

hy height of beam, h height of T-stub joint, av average, max maximum, min minimum, H height of beam to
height of T-stub joint of T-stub joints, X lengths of T-stub joints

Table 2  Average characteristic values for structural steels and bolt (8.8)
E (MPa) Est (MPa) fy (MPa) fu (MPa) ρy = fy/fu εst εuni εf

10 mm plate 205,352 1798 687 721 0.95 2.71 × 10−2 2.68 × 10−2 1.62 × 10−1
Bolt – – 839.18 971.93 0.86 – – –
Beam web 202,942 1289 371 521 0.71 1.57 × 10−2 1.39 × 10−2 0.87 × 10−1
Beam flange 203,985 1321 528 612 0.86 1.79 × 10−2 1.51 × 10−2 1.00 × 10−1
Column web 204,424 1396 541 637 0.85 1.89 × 10−2 1.63 × 10−2 1.11 × 10−1
Column flange 208,242 1928 831 945 0.88 2.99 × 10−2 2.81 × 10−2 1.78 × 10−1
T140-stub web 201,273 1231 321 452 0.71 1.32 × 10−2 1.11 × 10−2 0.59 × 10−1
T140-stub flange 202,631 1298 478 563 0.85 1.61 × 10−2 1.30 × 10−2 0.91 × 10−1
T200-stub web 203,577 1351 423 611 0.69 1.82 × 10−2 1.49 × 10−2 1.01 × 10−1
T200-stub flange 204,102 1399 543 702 0.77 2.14 × 10−2 1.62 × 10−2 1.19 × 10−1

E Young’s modulus, Est strain hardening modulus, fy static yield, fu tensile stresses, εst strain at the strain hardening point, εuni uniform strain, εf
strain at rupture load

2.3 Test Arrangement and Instrumentation obtain a realistic stress pattern at the connection, and the
fracture of the several specimens, i.e., ultimate load, was
The specimens were subjected to a static force applied by attained with a specific testing machine (Aydin et al. 2015a,
a 900 kN hydraulic jack with a maximum piston stroke of b). Thus, the lengths of the beam and column were chosen
300 mm. Tests were performed under displacement control as 1500 mm.
with a constant speed of 0.01 mm/s up to the failure of the The primary requirements of the instrumentation were the
specimens. In order to prevent the lateral torsional buckling measurement of the applied load (P), which was measured
of the beam while loading, a two-column guidance device by a load cell and hydraulic pump machine; the displace-
was provided near the beam (Aydin et al. 2015a, b), and ments (DT) of the connection, beam, T-stub joint, and flange
lateral torsional buckling of the beam did not, in fact, occur of the column, which were measured by linear variable dis-
during loading in the experiments. Figure 3 illustrates the placement transducers (LVDTs, shown as DT in Fig. 3) with
test arrangements and the mechanism. This study aimed to a maximum displacement of 100 mm); the strains at the

13
International Journal of Steel Structures (2018) 18(3):891–909 895

DT1

Strain gauges

LVDT
Hydraulic Jack
DT5 DT4 DT3 DT2 DT1

DT6-7 DT8
500mm 250mm 250mm 300mm 100mm

DT8
L1

Lload
X
Load cell

DT6-7 DT8

Fig. 3  Locations of the displacement transducers (DT = LVDT)

T-stub connections, which were measured by strain gauges

(TML YEFLA-5, maximum strain of 15–20%). The results
M = PLload (1)
were collected using a data logging device that recorded The rotational deformation of the joint (Ɵ) is equal to
all measurements and the load cells at 1-s intervals. All the connection rotation. The beam rotation is approximately
of the data were recorded for the duration of the test. Four given by (Fig. 3):
strain gauges (STR) were attached to the T-stub connection ( ) (( P )( XDT1 3 2
Lload XDT1
))
(horizontal and vertical) as shown in Fig. 3 to observe the arctan 𝛿DT1 − 𝛿DT5 − − EI 6
− 2
strain distribution. For a good comparison of the results, all Θ=
L1
specimens used the same arrangement for the location of (2)
the strain gauges and the same strain gauges and measuring where I is the moment of inertia and E is the Young’s mod-
devices. ulus of beam. Some differences among the results from
The designed connections form various moment–rotation LVDTs DT1–DT2 are identical, as expected. Therefore, all
curves that describe the relationship between the applied of the deformation values presented in the remainder of the
moment (M) and the corresponding rotation (Ɵ) between paper refer to the readings from DT1 (Aydin et al. 2015b).
the members according to the elements of connection and
the shape of placement (Sagiroglu and Aydin 2015). The
rotation and the bending moment are predicted using the
3 Test Results
displacements of the beam or top-and-seat angle connection
and multiplication of the distance between the load applica-
The M–Ɵ curve of a connection may be characterized using
tion point and beam end, bolted to column ­(Lload), and the
the aforementioned relationships. In particular, the following
applied load, respectively:
characteristics were assessed for the different experimental

13
896 International Journal of Steel Structures (2018) 18(3):891–909

tests (Aydin et al. 2015a, b), as shown in Fig. 4. Eurocode 1 to 2. Thus, the plastic flexural resistance values were
3 (2005) gives quantitative rules for predicting the joint obtained with an increase in the lengths (X) of T-stub
flexural plastic resistance and initial rotational stiffness joints and an increase in the number and row bolts of 1–2
for major beam-to-column joints of end-plate connections. rows.
These structural properties are evaluated below using the
geometric and mechanical nominal properties in Eurocode 3. A comparison of the T140-2-10, T160-2-10, and T200-
Figure 5 and Table 3 present the moment–rotation 2-10 groups showed that the plastic flexural resistance val-
responses for the eighteen full-scale specimens of steel- ues increased in the H ­ max and H­ av model groups with an
bolted, beam-to-column connections with T-stub joints. increase in the thickness of the web and flange in the T140,
T160 and T200 T-stub joints. The plastic flexural resistance
3.1 Moment Resistance values decreased in the H ­ min model groups with an increase
in the thickness of the web and flange in the T140 to T200
3.1.1 Plastic Flexural Resistance ­Mj.Rd T-stub joints. A comparison of the T140-1-10, T160-1-10,
and T200-1-10 groups showed that these groups had the
The plastic flexural resistance M ­ j.Rd corresponds to the inter- same behavior as that of the T140-2-10, T160-2-10, and
section point of the previous two regression lines obtained T200-2-10 groups. Therefore, ­Hmin models are not suitable
for the initial stiffness ­(Sj.ini) and for the post-limit stiffness for use in the industry.
­(Sj.p−l). Figure 5 and Table 3 show the following: Generally, the plastic flexural resistance values increased
with an increase in H from ­Hmin to ­Hmax in model groups
• The plastic flexural resistance values for the T140-2- with X of 126 mm. In addition, the plastic flexural resist-
10, T160-2-10, and T200-2-10 groups with X (lengths ance values increased with an increase in the lengths (X) of
of T-stub joints) of 126 mm increased by about 11.15– T-stub joints from 54 to 126 mm and an increase in the num-
18.47, 5.04, and 32.84–60.04%, respectively, with ber and row bolts of 1–2 rows. Thus, the T140-2-10 model
an increase in H from H ­ min to ­Hmax. The plastic flex- with the lengths (X) of T-stub joints of 126 mm and height
ural resistance values for the T140-1-10 model with X of beam to height of T-stub joint (H) of T-stub joints equal
(lengths of T-stub joints) of 54 mm decreased by about to ­Hmax is an ideal model for use in the industry.
11.23–36.22% with an increase in H from ­Hmin to ­Hmax.
However, the plastic flexural resistance values for the 3.1.2 Maximum Bending Moment ­Mj.max
T160-1-10 and T200-1-10 models with X (lengths of
T-stub joints) of 54 mm increased with an increase in H The maximum bending moment ­Mj.max is equal to the maxi-
from ­Hmin to ­Hmax. Therefore, the T140-1-10 models are mum moment of the design moment–rotation characteris-
not suitable for joints. tic (Fig. 4). Figure 5 and Table 3 show that the maximum
• The plastic flexural resistance values for the T140, bending moment values for the T140-2-10, T160-2-10,
T160, and T200 groups with ­Hmax, ­Hmin, and ­Hav mod- and T200-2-10 groups with X (lengths of T-stub joints) of
els increased by about 44.58, 23.44, 55.20, 21.00, 82.52, 126 mm increased by about 4.4–18.71, 27.16–71.26, and
78.30, 44.56, 57.59, and 52.33%, respectively, with an 36.39–64.78%, respectively, with an increase in H from
increase in the lengths (X) of T-stub joints from 54 to ­Hmin to ­Hmax. The maximum bending moment values for the
126 mm and an increase in the number of row bolts from T140-1-10 and T200-1-10 models with X (lengths of T-stub

Fig. 4  Moment–rotation curve

characteristics

13
International Journal of Steel Structures (2018) 18(3):891–909 897

Fig. 5  Moment–rotation curve for six groups

joints) of 54 mm decreased with an increase in H from ­Hmin in the number and row bolts of 1–2 rows. In other words,
to ­Hmax. The maximum bending moment values for the the maximum bending moment for ­Hmin model groups was
T140, T160, and T200 groups with H ­ max, ­Hmin, and H
­ av mod- lower than for ­Hmax and ­Hav model groups.
els increased by about 56, 38.26, 67.72, 57.35, 7.4, 59.01, A comparison of the T140-2-10, T160-2-10, and T200-
65.75, 35.18, and 68.56%, respectively, with an increase in 2-10 groups showed that the maximum bending moment
the lengths (X) of T-stub joints from 54 to 126 mm and an values for the ­Hmin and ­Hav model groups decreased by
increase in the number and row bolts of 1–2 rows. Thus, the about 43.86–60.18 and 13.75–14.15%, respectively, with an
maximum bending moment values were obtained with an increase in the thickness of the web and flange in the T140 to
increase in the lengths (X) of T-stub joints and an increase T200 T-stub joints. The maximum bending moment values

13
 898

13
Table 3  Main characteristics of moment–rotation curves
Group name Experiment Moment resistance (KN. m) Stiffness (KN m/rad) Rotation (rad) ӨCd
KR (knee range) Mj.Rd Mj. max MӨCd Sj.ini Sj.p − l Sj.ini/Sj.p − l ӨMj.Rd ӨMin.K.R ӨMsup.k.R ӨMj. max

T140-2-10 T140-Hmax-126 1.61–7.51 3.14 7.91 7.54 14.38 0.57 25.27 0.006 0.001 0.14 0.152 0.153
T140-Hmin-126 1.01–6.11 2.56 6.43 6.15 2.97 0.48 6.17 0.014 0.006 0.16 0.169 0.17
T140-Hav-126 1.21–7.10 2.79 7.56 7.22 3.82 0.68 5.63 0.012 0.007 0.18 0.194 0.1942
T140-1-10 T140-Hmax-54 0.55–3.37 1.74 3.48 3.33 5.63 0.46 12.26 0.0124 0.002 0.16 0.171 0.1714
T140-Hmin-54 0.29–3.71 1.96 3.97 3.52 3.45 0.48 7.15 0.018 0.001 0.15 0.166 0.173
T140-Hav-54 0.57–2.12 1.25 2.44 2.22 8.67 0.33 26.41 0.005 0.0018 0.072 0.098 0.11
T160-2-10 T160-Hmax-126 1.82–8.35 2.38 8.91 7.64 3.05 0.55 5.68 0.013 0.0094 0.206 0.219 0.219
T160-Hmin-126 1.56–2.45 2.26 2.56 1.97 3.62 0.096 37.78 0.029 0.0129 0.074 0.1105 0.125
T160-Hav-126 0.49–6.27 3.65 6.49 6.38 5.29 0.51 10.39 0.013 0.003 0.189 0.186 0.186
T160-1-10 T160-Hmax-54 0.51–3.61 1.88 3.80 3.43 0.66 0.39 1.69 0.11 0.019 0.137 0.15 0.155
T160-Hmin-54 0.25–2.23 0.395 2.37 2.265 5.91 0.68 8.69 0.007 0.003 0.197 0.202 0.204
T160-Hav-54 0.35–2.16 0.792 2.66 2.55 2.97 0.58 4.26 0.014 0.006 0.127 0.164 0.168
T200-2-10 T200-Hmax-126 2.07–8.55 4.78 10.25 9.23 3.29 0.46 7.2 0.022 0.0108 0.147 0.195 0.195
T200-Hmin-126 1.44–2.99 1.91 3.61 2.78 6.52 0.43 15.12 0.0189 0.0109 0.122 0.168 0.17
T200-Hav-126 2.29–5.78 3.21 6.52 6.27 6.84 0.37 18.64 0.011 0.0078 0.144 0.176 0.176
T200-1-10 T200-Hmax-54 1.15–3.12 2.65 3.51 3.26 2.3 0.29 7.93 0.018 0.0188 0.048 0.069 0.077
T200-Hmin-54 0.40–1.98 0.81 2.34 2.2 3.93 0.5 7.86 0.01 0.0059 0.1407 0.18 0.18
T200-Hav-54 0.35–1.87 1.53 2.05 2.05 3.07 0.45 6.77 0.01 0.01 0.17 0.173 0.173
International Journal of Steel Structures (2018) 18(3):891–909
International Journal of Steel Structures (2018) 18(3):891–909 899

for the H
­ max model groups increased by about 13.07–22.83% ues decreased for the H ­ min and ­Hav model groups with
with an increase in the thickness of the web and flange in the an increase in the thickness of the web and flange in the
T140 to T200 T-stub joints. A comparison of the T140-1-10, T140 to T200 T-stub joints.
T160-1-10, and T200-1-10 groups showed that the maxi-
mum bending moment values increased for the H ­ max and H
­ av Generally, the bending moment capacity values increased
model groups with an increase in the thickness of the web with an increase in H from H ­ min to H
­ max for model groups
and flange in the T140 to T200 T-stub joints. The maximum with X of 126 mm. In addition, the bending moment capac-
bending moment values decreased for the ­Hmin model groups ity values increased with an increase in the lengths (X) of
with an increase in the thickness of the web and flange in the T-stub joints from 54 to 126 mm and an increase in the num-
T140 to T200 T-stub joints. Therefore, H ­ min models are not ber and row bolts of 1–2 rows.
suitable for use in the industry.
Generally, the maximum bending moment values 3.1.4 Knee Range
increased with an increase in H from ­Hmin to ­Hmax for model
groups with X of 126 mm. In addition, the maximum bend- The knee range of the M–Ɵ curve is defined as the transi-
ing moment values increased with an increase in the lengths tion zone between the initial and post-limit stiffness, with
(X) of T-stub joints from 54 to 126 mm and an increase in its lower boundary at ­Mmink−R and rotation Ɵmink−R and
the number and row bolts of 1–2 rows. Thus, the T140-2-10 upper limit at ­Msupk−R and rotation Ɵsupk−R (Fig. 4). Fig-
model with lengths (X) of T-stub joints equal to 126 mm ure 5 and Table 3 show that the knee range values for
and height of beam to height of T-stub joint (H) of T-stub the T140-2-10, T160-2-10, and T200-2-10 groups with X
joints equal to ­Hmax is an ideal model for use in the industry. (lengths of T-stub joints) of 126 mm increased by about
13.56, 75–86.37, and 46.14–76.08%, respectively, with an
increase in H from H­ min to H
­ max. The knee range values for
3.1.3 Bending Moment Capacity ­MƟ.Cd the T140-1-10 and T160-1-10 models with X (lengths of
T-stub joints) of 54 mm decreased by about 17.54–54.67
The bending moment capacity M ­ Ɵ.Cd is equal to the maxi- and 8.5%, respectively, with an increase in H from ­Hmin to
mum moment capacity of the design moment–rotation char- ­Hmax. The knee range values for the T140 and T200 groups
acteristic (Fig. 4). Figure 5 and Table 3 show the following: with the ­Hmax, ­Hmin, and ­Hav models increased by about
52.20, 32.94, 73.68, 69.90, and 56.44%, respectively, with
• The ­MƟ.Cd values for the T140-2-10, T160-2-10, and an increase in the lengths (X) of T-stub joints from 54
T200-2-10 groups with X (lengths of T-stub joints) of to 126 mm and an increase in the number and row bolts
126 mm increased by about 18.43, 16.49–74.21, and of 1–2 rows. Thus, the knee range values were obtained
32.07–69.88%, respectively, with an increase in H from with an increase in the lengths (X) of T-stub joints and
­Hmin to H
­ max. The M­ Ɵ.Cd values for the T140-1-10 model an increase in the number and row bolts of 1–2 rows. A
with X (lengths of T-stub joints) of 54 mm decreased comparison of the T140-2-10, T160-2-10, and T200-2-10
by about 5.39–36.93% with an increase in H from ­Hmin groups showed that the knee range values for the ­Hmin and
to ­Hmax. However, the M ­ Ɵ.Cd values with X (lengths of ­H av model groups decreased by about 69.90–82.54 and
T-stub joints) of 54 mm increased with an increase in H 2.03–40.8%, respectively, with an increase in the thick-
from ­Hmin to ­Hmax for the T160 and T200 models. There- ness of the web and flange in the T140 to T200 T-stub
fore, models with X (lengths of T-stub joints) of 54 mm joints. The knee range values for the H ­ max model groups
are not suitable for joints. increased with an increase in the thickness of the web and
• The ­MƟ.Cd values for the T140, T160, and T200 groups flange in the T140 to T200 T-stub joints. In addition, a
with ­Hmax, ­Hmin, and H ­ av models increased with an comparison of the T140-1-10, T160-1-10, and T200-1-10
increase in the lengths (X) of T-stub joints from 54 to groups showed that the knee range values decreased for the
126 mm and an increase in the number and row bolts ­Hmax, ­Hmin, and H
­ av model groups with an increase in the
of 1–2 rows. Thus, the M ­ Ɵ.Cd values were obtained with thickness of the web and flange in the T140 to T200 T-stub
an increase in the lengths (X) of T-stub joints and an joints. Generally, the knee range values increased with an
increase in the number and row bolts of 1–2 rows. increase in H from H ­ min to H
­ max for model groups with X
• A comparison of the T140-2-10, T160-2-10, and T200-2- of 126 mm. Thus, the T140-2-10 model with lengths (X)
10 groups showed that the bending moment capacity val- of T-stub joints of 126 mm and height of beam to height of
ues increased for the ­Hmax model groups with an increase T-stub joint (H) of T-stub joints equal to H ­ max is an ideal
in the thickness of the web and flange in the T140 to model for use in the industry.
T200 T-stub joints. The bending moment capacity val-

13
900 International Journal of Steel Structures (2018) 18(3):891–909

3.1.5 Summary of Moment Resistance Results increase in the thickness of web and flange joints, increase in
the lengths (X) of T-stub joints, and increase in H from H­ min
Figure 6 shows the moment–rotation characteristic curves to ­Hmax in T connections. Figure 6 also shows that the ­Mj.Rd
for all groups. These curves show that the T200-Hmax model value of the T140-Hmax-126 model is equal to that of the
with lengths (X) of T-stub joints equal to 126 mm had the T200-Hav-126 model. Therefore, a model with low thickness
greatest ­Mj.Rd among the T140-2-10, T160-2-10, and T200- and maximum height of beam to height of T-stub joint (H)
2-10 model groups. In other words, ­Mj.Rd increased with an could be used instead of a model with large thickness and

Fig. 6  Comparison of moment–rotation curves for six groups

13
International Journal of Steel Structures (2018) 18(3):891–909 901

average height of beam to height of T-stub joint (H). Thus, (X) of T-stub joints from 54 to 126 mm and an increase
the T140-Hmax model with lengths (X) of T-stub joints of in the number and row bolts of 1–2 rows. The rates of
126 mm is the ideal model in all groups. A comparison of increase in the initial stiffness to the post-limit stiffness
the groups showed that ­Mj.Rd increased with an increase in for the T160 and T200 groups with H ­ min and H
­ av mod-
the lengths (X) of T-stub joints and an increase in the num- els decreased by about 76.99, 58.99, 48.02, and 63.68%
ber of bolt row of 1–2 number rows. The ­Mj.Rd values for respectively, with an increase in the lengths (X) of T-stub
the TA-2-10 model groups were twice those of the TA-1-10 joints from 54 to 126 mm and an increase in the number
model groups. and row bolts of 1–2 rows.
Figure 6 shows that ­Mj.max increased with an increase in
the height of beam to height of T-stub joint and an increase Generally, the rates of increase in the initial stiffness to
in the thickness of web and flange joints. For example, the the post-limit stiffness for the T140-2-10, T160-2-10, and
T200-2-10 model groups with a maximum height of beam T200-2-10 model groups increased with an increase in the
to height of T-stub joint (­ Hmax) had the highest values of number and row bolts of 1–2 rows and an increase in the
­Mj.max; the T160-2-10 and T140-2-10 models with H ­ max had thickness of the web and flange in the T140 to T200 T-stub
the next highest ­Mj.max values. The models with ­Hmin had the joints. However, the rates of increase in the initial stiffness
lowest values of ­Mj.max. Therefore, an ideal model could be to the post-limit stiffness for the T140-1-10, T160-1-10,
selected from the lower beam section with ­Hmax and ­Xmax. and T200-1-10 model groups were opposite to those for the
Figure 6 shows that ­Mϴ.Cd is the same as ­Mj.max and ­Mj.Rd. T140-2-10, T160-2-10, and T200-2-10 model groups.

3.2 Stiffness 3.3 Rotation

The initial rotational stiffness ­Sj.ini is given by the slope of 3.3.1 Rotation Corresponding to Plastic Flexural Resistance
the elastic range of the design moment–rotation character- Ɵj.Rd
istic, and the post-limit stiffness ­Sj.p−l is given by the slope
of the plastic range of the design moment–rotation charac- The plastic flexural resistance moment rotation ƟMj.Rd is
teristic (Fig. 4). Table 3 shows the following: equal to the rotation of the plastic flexural resistance moment
(Fig. 4). Table 3 shows that the ƟMj.Rd values for the T140-
• The rate of increase in the initial stiffness to the post-limit 2-10, T140-1-10, and T160-2-10 groups decreased by about
stiffness for the T140-2-groups with X (lengths of T-stub 133.3, 45.16, and 123.07%, respectively, with an increase
joints) of 126 mm increased by about 75.58–77.21% with in H from ­Hmin to ­Hmax. The ƟMj.Rd values for the T200-
an increase in H from H ­ min to H
­ max. The rate of increase 2-10 and T200-1-10 models increased by about 14.89 and
in the initial stiffness to the post-limit stiffness for the 54.54%, respectively, with an increase in H from H ­ min to
T160-2-10 model with X (lengths of T-stub joints) of ­Hmax. The ƟMj.Rd values for the T140 groups with ­Hmax and
126 mm decreased by about 565.14% with an increase ­Hmin models decreased by about 106.66 and 28.57%, respec-
in H from ­Hmin to H ­ max. However, the rate of increase tively, with an increase in the lengths (X) of T-stub joints
in the initial stiffness to the post-limit stiffness for the from 54 to 126 mm and an increase in the number and row
T160-2 groups with X (lengths of T-stub joints) of bolts of 1–2 rows. The ƟMj.Rd values for the T140 groups
54 mm increased by about 41.68% with an increase in H with ­Hav models increased by about 58.33% with an increase
from ­Hmin to ­Hmax. in the lengths (X) of T-stub joints from 54 to 126 mm and
• The rates of increase in the initial stiffness to the post- an increase in the number and row bolts of 1–2 rows. The
limit stiffness for the T140 groups with H ­ max models ƟMj.Rd values for the T160 and T200 groups with H ­ max and
increased by about 51.48% with an increase in the lengths ­Hmin models increased by about 46.15, 75.86, 18.18, and
(X) of T-stub joints from 54 to 126 mm and an increase 47.08%, respectively, with an increase in the lengths (X)
in the number and row bolts of 1–2 rows. However, the of T-stub joints from 54 to 126 mm and an increase in the
rates of increase in the initial stiffness to the post-limit number and row bolts of 1–2 rows. A comparison of the
stiffness for the T140 groups with ­Hmin and ­Hav models T140-2-10, T160-2-10, and T200-2-10 groups showed
decreased by about 15.88 and 369.1%, respectively, with that the ƟMj.Rd values for the ­Hmax and ­Hmin model groups
an increase in the lengths (X) of T-stub joints from 54 increased by about 266.66 and 35%, respectively, with an
to 126 mm and the increase in the number and row bolts increase in the thickness of the web and flange in the T140
of 1–2 rows. In addition, the rates of increase in the ini- to T200 T-stub joints. In addition, a comparison of the T140-
tial stiffness to the post-limit stiffness for the T160 and 1-10, T160-1-10, and T200-1-10 groups showed that the
T200 groups with ­Hmax models decreased by about 70.25 ƟMj.Rd values for the ­Hmax and ­Hav model groups increased
and 10.14%, respectively, with an increase in the lengths by about 74.42 and 100%, respectively, with an increase in

13
902 International Journal of Steel Structures (2018) 18(3):891–909

the thickness of the web and flange in the T140 to T200 characteristic (Fig. 4). Figure 5 shows that the Ɵcd values
T-stub joints. Generally, with an increase in H from ­Hmin to for all groups with ­Hmin models decreased with an increase
­Hmax, the ƟMj.Rd values decreased for T140 and T160 model in the lengths (X) of T-stub joints from 54 to 126 mm and
groups and increased for T200 model groups. In addition, an increase in the number and row bolts of 1–2 rows. The
the ƟMj.Rd values increased with an increase in the thickness Ɵcd values for all groups with ­Hav models increased with
of the web and flange in the T140 to T200 T-stub joints. an increase in the lengths (X) of T-stub joints from 54 to
126 mm and an increase in the number and row bolts of 1–2
3.3.2 Rotation Corresponding to the Maximum Bending rows. Therefore, the ideal model for the rotation capacity is
Moment ƟM.j.max ­Hav model groups.

The maximum bending moment rotation ƟM.j.max of a joint is 3.4 Ductility

equal to the maximum rotation of the design moment–rota-
tion characteristic (Fig. 4). Figure 5 shows the following: The ductility of a joint (Ψj) is a property that reflects the
length of the yield plateau of the moment–rotation response.
• The ƟM.j.max values for the T140-2-10, T140-1-10, T160- The proposed definition of the ductility of a joint is the dif-
2-10, and T200-2-10 groups increased by about 12.89, ference between the rotation value corresponding to the joint
2.92, 49.54, and 13.85%, respectively, with an increase in plastic resistance, ƟMRd, and the total rotation capacity, ƟCd
H from ­Hmin to ­Hmax. The ƟM.j.max values for the T200-1- (Schleich et al. 1998; Gil et al. 2003). Thus, the ductility of
10 model decreased by about 160.86% with an increase a joint relates the maximum rotation of the joint, ƟCd, to the
in H from ­Hmin to ­Hmax. rotation value corresponding to the joint’s plastic flexural
• The ƟM.j.max values for the T140 groups with ­Hmin and resistance, ƟMRd (Aydin et al. 2015a, b):
­Hav models increased by about 1.77 and 49.48%, respec-
ΘCd
tively, with an increase in the lengths (X) of T-stub joints Ψj = (3)
ΘMRd
from 54 to 126 mm and an increase in the number and
row bolts of 1–2 rows. In addition, the ƟM.j.max values In addition, the rotation values at the maximum load and
for the T160 and T200 groups with H ­ max and H
­ av models corresponding ductility levels, Ψj.maxload, can be derived
increased by about 31.56, 11.83, 64.62, and 1.7%, respec- from
tively, with an increase in the lengths (X) of T-stub joints
ΘMj ⋅ max
from 54 to 126 mm and an increase in the number and Ψj⋅maxload = (4)
row bolts of 1–2 rows. The ƟM.j.max values for the T160 ΘMRd
and T200 groups with ­Hmin models decreased with an Eurocode 3 (2005) gives quantitative rules for predict-
increase in the lengths (X) of T-stub joints from 54 to ing the joint flexural plastic resistance and initial rotational
126 mm. Therefore, use of ­Hmin models does not improve stiffness for major beam-to-column joints of end-plate con-
the ƟM.j.max values in T-stub connections. nections. These structural properties are evaluated below
• A comparison of the T140-2-10, T160-2-10, and T200- using the geometric and mechanical nominal properties in
2-10 groups showed that the ƟM.j.max values increased for Eurocode 3. Table 4 shows the following:
the ­Hmax model groups with an increase in the thickness
of the web and flange in the T140 to T200 T-stub joints. • The Ψj and Ψj.maxload values for the T140-2-10, T160-
The ƟM.j.max values decreased for the H ­ min model groups 2-10, T200-2-10, T140-1-10, and T160-1-10 groups
with an increase in the thickness of the web and flange in increased with an increase in H from ­Hmin to ­Hmax. The
the T140 to T200 T-stub joints. In addition, a compari- Ψj and Ψj.maxload values for the T200-1-10 model with
son of the T140-1-10, T160-1-10, and T200-1-10 groups X (lengths of T-stub joints) of 54 mm decreased with an
showed that these groups had the opposite behavior to the increase in H from ­Hmin to ­Hmax.
T140-2-10, T160-2-10 and T200-2-10 groups. • The Ψj and Ψj.maxload values for the T140 groups with
­Hmax and ­Hmin models increased with an increase in the
Generally, the ƟM.j.max values increased with an increase lengths (X) of T-stub joints from 54 to 126 mm and an
in H from ­Hmin to ­Hmax in model groups with X of 126 mm. increase in the number and row bolts of 1–2 rows. The Ψj
In addition, the use of ­Hav increased the ƟM.j.max values. and Ψj.maxload values for the T160 and T200 groups with
­Hmin models decreased with an increase in the lengths
3.3.3 Rotation Capacity Ɵcd (X) of T-stub joints from 54 to 126 mm and an increase
in the number and row bolts of 1–2 rows.
The rotation capacity Ɵcd of a joint is equal to the bend- • A comparison of the T140-2-10, T160-2-10, and T200-
ing moment capacity in the design moment–rotation 2-10 groups showed that the Ψj and Ψj.maxload values

13
International Journal of Steel Structures (2018) 18(3):891–909 903

Table 4  Experimental evaluation of joint ductility indexes Ψj and in H from ­Hmin to ­Hmax. The energy dissipation values for
Ψj.maxload and energy dissipation the T140-1-10, T160-1-10 and T200-1-10 models with X
Group name Experiment Ψj Ψj.maxload Energy dissipa- (lengths of T-stub joints) of 54 mm decreased by about
tion (KNm rad) 17.24, and 55.55%, respectively, with an increase in H from
­Hmin to ­Hmax. Therefore, T-stub connections with X (lengths
T140-2-10 T140-Hmax-126 25.50 25.33 0.61
of T-stub joints) of 54 mm are not suitable for joints. The
T140-Hmin-126 12.14 12.07 0.55
energy dissipation values for the T140 and T200 groups with
T140-Hav-126 16.18 16.17 0.73
­Hmax, ­Hmin, and H
­ av models increased by about 51.66, 37.07,
T140-1-10 T140-Hmax-54 13.82 13.79 0.30
82.19, 86.66, 30, and 70.17%, respectively, with an increase
T140-Hmin-54 9.61 9.22 0.34
in the lengths (X) of T-stub joints from 54 to 126 mm and
T140-Hav-54 22.00 19.60 0.13
an increase in the number and row bolts of 1–2 rows. In
T160-2-10 T160-Hmax-126 16.85 16.85 0.98
addition, the energy dissipation values for the T160 groups
T160-Hmin-126 4.31 3.81 0.16
with ­Hmax and ­Hav models increased with an increase in
T160-Hav-126 14.31 14.31 0.60
the lengths (X) of T-stub joints from 54 to 126 mm and an
T160-1-10 T160-Hmax-54 1.40 1.36 0.29
increase in the number and row bolts of 1–2 rows. How-
T160-Hmin-54 29.14 28.86 0.24
ever, the energy dissipation values for the T160 groups with
T160-Hav-54 12.00 11.71 0.22
­Hmin models decreased with an increase in the lengths (X),
T200-2-10 T200-Hmax-126 8.86 8.86 1.00
because of bolt failure and less rotation. A comparison of
T200-Hmin-126 8.99 8.89 0.31
the T140-2-10, T160-2-10, and T200-2-10 groups showed
T200-Hav-126 16.00 16.00 0.57
that the energy dissipation values decreased for the H ­ min
T200-1-10 T200-Hmax-54 3.50 3.14 0.14
model groups with an increase in the thickness of the web
T200-Hmin-54 18.00 18.00 0.21
and flange in the T140 to T200 T-stub joints. Moreover,
T140-Hmax-54 17.30 17.30 0.18
the energy dissipation values decreased for the H ­ min model
groups with an increase in the thickness of the web and
flange in the T140 to T200 T-stub joints for the T140-1-
decreased for the H ­ max, ­Hmin, and H­ av model groups 10, T160-1-10, and T200-1-10 model groups. Therefore, a
with an increase in the thickness of the web and flange model with ­Hmin is not suitable for use in joints. Generally,
in the T140 to T200 T-stub joints. A comparison of the energy dissipation increased with an increase in H from ­Hmin
T140-1-10 and T200-1-10 groups showed that the Ψj and to ­Hmax, an increase in the lengths (X) of T-stub joints from
Ψj.maxload values decreased for the H­ max and H
­ av model 54 to 126 mm, and an increase in the number and row bolts
groups with an increase in the thickness of the web and of 1–2 rows.
flange in the T140 to T200 T-stub joints. However, the
T160-1-10 groups showed that the Ψj and Ψj.maxload val- 3.6 Failure Modes of Test Specimens and Strain
ues increased for the ­Hmin model groups with an increase
in the thickness of the web and flange in the T140 to Figures 7 and 8 show the moment–strain responses for the
T200 T-stub joints. eighteen experimental models for an investigation of bolted
T-stub connections. Figure 7 shows that all strains for mod-
Generally, the Ψj and Ψj.maxload values increased with an els with X of 126 mm were in the plastic regions and the
increase in H from H ­ min to ­Hmax. In addition, the Ψj and strain at STR1 was greater than other strains. The reason the
Ψj.maxload values decreased with an increase in the thickness strain at STR1 was greater could be that STR1 was parallel
of the web and flange in the T140 to T200 T-stub joints. to the beam. Therefore, a large change was obtained in this
region, and this strain value was produced. Figure 8 shows
3.5 Energy Dissipation that all strains had the same behavior, but the strains at STR1
and STR2 were greater than other strains in these models.
Semi-rigid connections such as bolted joints allow better The reason the strains at STR1and STR2 were greater could
energy dissipation and provide good response under static be that STR1 was parallel and vertical to the beam. There-
loading. In a structural framework with semi-rigid joints, fore, a large strain occurred at this location. A comparison
the characteristics of connections play a significant role in of Figs. 7 and 8 showed that the strain values for models
the energy dissipation mechanisms (Babu and Sreekumar with X of 126 mm were greater than those for models with
2012) Table 4 shows that the energy dissipation values for X of 54 mm, likely because the failure of models with X of
the T140-2-10, T160-2-10, and T200-2-10 groups with X 126 mm did not occur until the rupture of T-stub connection
(lengths of T-stub joints) of 126 mm increased by about on the beam where the STR1 and STR2 strain gauges were
11.48, 83.51, and 69.69%, respectively, with an increase attached. However, the reason the strain values for models

13
904 International Journal of Steel Structures (2018) 18(3):891–909

Fig. 7  Moment–strain curves

for T140-2-10, T160-2-10, and
T200-2-10 groups

13
International Journal of Steel Structures (2018) 18(3):891–909 905

Fig. 8  Moment–strain curves

for T140-1-10, T160-1-10, and
T200-1-10 groups

13
906 International Journal of Steel Structures (2018) 18(3):891–909

Fig. 9  Failure modes for all groups

13
International Journal of Steel Structures (2018) 18(3):891–909 907

Fig. 9  (continued)

with X of 54 mm were the lowest could be that failure of was the rupture of the T-stub connection on the beam. There-
these models did not occur until the shear of bolts on the fore, the thickness of joints was important in determining
column of T-stub connection where no strain gauges were the failure modes. The failure mode of T140, T160, and
attached. T200 groups with X of 54 mm was the bolt being directly
Therefore, the failure mode of T160 and T200 groups overloaded by the applied forces on the beam of the T-stub
with X of 126 mm was the shear of top bolts on the col- connection (Fig. 9). Therefore, the thickness of joints and
umn of T-stub connection and a V-shape appeared between lengths (X) of T-stub joints were important in determining
the T-stub and column at the beam-to-column connection the failure modes.
(Fig. 9). The failure mode of T140 groups with X of 126 mm

13
908 International Journal of Steel Structures (2018) 18(3):891–909

4 Conclusion 8. The energy dissipation increased with an increase in

H from ­Hmin to ­Hmax, an increase in the lengths (X) of
The aim of this study was to analyze the influence of T con- T-stub joints from 54 to 126 mm, and an increase in the
nections based on the IPE standard profile, height of beam number and row bolts of 1–2 rows.
to height of T-stub joint (H) of T-stub joints, and lengths
(X) of T-stub joints on the behavior of steel connections, in
order to provide the necessary data for improving Eurocode Acknowledgements The authors wish to thank the BAP Project
(2015/125) of Ataturk University, Turkey, for Project funding, the
3 and enable efficient use of residue IPE standard profiles financial support given by the Tübitak (2215) for Mahyar Maali,
and back to the consumption cycle. and the Gençler Metal steel company for making the test specimens
available.
1. The moment resistance values increased with an increase
in H from H ­ min to ­Hmax in models groups with X of
126 mm. In addition, the moment resistance values
increased with an increase in the lengths (X) of T-stub References
joints from 54 to 126 mm and an increase in the number
Aydın, A. C., Kılıç, M., Maali, M., & Sağıroğlu, M. (2015a). Experi-
and row bolts of 1–2 rows. Thus, the T140-2-10 model mental assessment of the semi-rigid connections behavior with
with lengths (X) of T-stub joints of 126 mm and height angles and stiffeners. Journal of Constructional Steel Research,
of beam to height of T-stub joint (H) of T-stub joints 114, 338–348.
equal to ­Hmax is an ideal model for use in the industry. Aydın, A. C., Maali, M., Kılıç, M., & Sağıroğlu, M. (2015b). Experi-
mental investigation of sinus beams with end-plate connection.
2. Mj.Rd increased with an increase in the thickness of web Journal of Thin Walled structures, 97, 35–43.
and flange joints, an increase the lengths (X) of T-stub Babu, S. S., & Sreekumar, S. (2012). A study on the ductility of bolted
joints, and an increase in H from H ­ min to H
­ max in T con- beam-column connections. International Journal of Modern Engi-
nection. Therefore, a model with small thickness and neering Research (IJMER), 2(5), 3517–3521.
Coelho, A. M. G., Bijlaard, F. S. K., Gresnigt, N., & da Silva, L. S.
maximum height of beam to height of T-stub joint (H) (2004). Experimental assessment of the behaviour of bolted
could be used instead of a model with large thickness T-stub connections made up of welded Plates. Journal of Con-
and average height of beam to height of T-stub joint (H). structional Steel Research, 60, 269–311.
3. Mj.max increased with an increase in the height of beam De Limaa, L. R. O., de Andradea, S. A. L., da Vellascob, P. C. G. S.,
& da Silvac, L. S. (2002). Experimental and mechanical model
to height of T-stub joint and an increase in the thick- for predicting the behaviour of minor axis beam-to-column semi-
ness of web and flange joints. Therefore, the ideal model rigidjoints. International Journal of Mechanical Sciences, 44,
could be selected from the lower beam section with ­Hmax 1047–1065.
and ­Xmax. Drosopoulos, G. A., Stavroulakis, G. E., & Abdalla, K. M. (2012). 3D
finite element analysis of end-plate joints. Steel and Composite
4. With an increase in H from H ­ min to H­ max, the ƟMj.Rd Structures, 12, 93–115.
values decreased for T140 and T160 model groups and European Committee for Standardization (CEN). (2005). Design of
increased for T200 model groups. In addition, the ƟMj.Rd steel structures part 1.8: Design of joints. Stage 49 draft, Brussels.
values increased with an increase in the thickness of the Gil, B., Cabrero, J. M., Goñi, R., & Bayo, E. (2003). An assessment
of the rotation capacity required by structural hollow sections for
web and flange in the T140 to T200 T-stub joints. plastic analysis. In M. A. Jaurrieta, A. Alonso, & J. A. Chica
5. The ƟM.j.max values increased with an increase in H from (Eds.), Tubular structures X (pp. 277–292). Lisse: A.A. Balkema
­Hmin to H
­ max for model groups with X of 126 mm. In Publishers.
addition, the use of ­Hav increased the ƟM.j.max values. Herrera, R., Bravo, M., Gómez, G., & Aedo, G. (2013). Performance
of built-up T-stubs for double T moment connections. Journal of
6. The Ɵ cd values for all groups with ­H min models Constrructional Steel Ressearch, 88, 289–295.
decreased with an increase in the lengths (X) of T-stub Katula, L., & Dunai, L. (2015). Experimental study on standard and
joints from 54 to 126 mm and an increase in the num- innovative bolted end-plate beam-to-beam joints under bending.
ber and row bolts of 1–2 rows. The Ɵcd values for all Steel and Composite Structures, 18(6), 1423–1450.
Liu, Y., Málaga-Chuquitaype, C., & Elghazouli, Y. (2014). Behaviour
groups with Hav models increased with an increase in of open beam-to-tubular column angle connections under com-
the lengths (X) of T-stub joints from 54 to 126 mm and bined loading conditions. Steel and Composite Structures, 16(2),
an increase in the number and row bolts of 1–2 rows. 157–185.
Therefore, the ideal model for the rotation capacity was Maali, M., Aydın, A. C., & Sağıroğlu, M. (2015). Investigation of inno-
vative steel runway beam in industrial building. Sadhana, 40(7),
­Hav model groups. 2239–2251.
7. The Ψj and Ψj.maxload values increased with an increase Maali, M., Kılıc, M., Sagiroglu, M., & Aydin, A. C. (2016). Experi-
in H from H
­ min to H
­ max. In addition, the Ψj and Ψj.maxload mental model for predicting the semi-rigid connections behavior
values decreased with an increase in the thickness of the with angles and stiffeners. Advances in Structural Engineering,
20(6), 884–895.
web and flange in the T140 to T200 T-stub joints.

13
International Journal of Steel Structures (2018) 18(3):891–909 909

Pitrakkos, T., & Tizani, W. (2015). A component method model for Schleich, J. B., Chantrain, P., Chabrolin, B., Galéa, Y., Bureau, A., &
blind-bolts with headed anchors in tension. Steel and Composite Anza, J. (1998). Promotion of plastic design for steel and com-
Structures, 18(5), 1305–1330. posite cross sections: New required conditions in Eurocodes 3
Sagiroglu, M., & Aydin, A. C. (2015). Design and analysis of non- and 4, practical tools for designers. European commission; 1998.
linear space frames with semi-rigid connections. Steel and Com- Swanson, J. A., Leon, R. T., & Kokan, D. S. (2002). Advanced finite
posite Structures, 18(6), 1405–1421. element modeling of bolted T-stub connection components. Jour-
SB EN ISO 898-1:1999. (2009). Mechanical properties of fasteners nal of Constructional Steel Research, 58(5), 1015–1031.
made of carbon steel and alloy steel. Bolts, screws and stud. UNE-EN 10002-1. (2002). Materiales met´alicos. Ensayos de
Published: October 1999 Replaced By: SB EN ISO 898-1:2009, tracci´on. Parte 1: Método de ensayo a temperatura ambiente.
Revised, Withdrawn. AENOR, 2002.

13