Thin-Walled Structures 145 (2019) 106393

Contents lists available at ScienceDirect

Thin-Walled Structures
journal homepage: www.elsevier.com/locate/tws

Full length article

Experiments and finite element modelling of screw pattern of self-drilling T

screw connections for high strength cold-formed steel
Krishanu Roya,∗, Hieng Ho Laub, Tina Chui Huon Tingc, Rehan Masooda,d, Ankur Kumara,
James B.P. Lima
a
Department of Civil and Environmental Engineering, The University of Auckland, New Zealand
b
Faculty of Engineering, Computing and Science, Swinburne University of Technology, Sarawak Campus, Kuching, Sarawak, Malaysia
c
Faculty of Engineering and Science, Curtin University Malaysia, Miri, Sarawak, Malaysia
d
Department of Construction Engineering and Management, NIT/SCEE, National University of Science and Technology, Islamabad, Pakistan

ARTICLE INFO ABSTRACT

Keywords: Cold-formed steel (CFS) self-drilling screw connections are popular due to rapid fastening and ease of in-
Cold-formed steel stallation. However, limited research is available in the literature to investigate the effect of screw patterns on
Finite element modelling the connection, where the individual screws can undergo different failure modes including bearing, tilting, pull-
Self-drilling screw connections out and shear failure, particularly for Grade 550 CFS plates. This issue is addressed herein through a combination
Screw pattern
of experimental and numerical tests. In total, 25 new experimental tests are reported that investigate the effect of
Shear failure
screw patterns, number of screws, and screw spacing on the connection strength; in the tests, 12 and 14 gauge
screws are considered. A nonlinear explicit dynamic finite element (FE) model was then developed which in-
cludes a fracture criterion for the steel sheets. The FE model was validated against the experimental test results
and showed good agreement in terms of connection strength and failure modes, including bearing, tilting, pull-
out and shear failure.

1. Introduction failure of the screws (Fig. 3). The experimental results are used to va-
lidate a non-linear explicit dynamic finite element model.
The use of cold-formed steel (CFS) sections as structural members is Research in the literature is available for screw connections but,
increasing steadily and is now commonly used for structures such as when investigating screw patterns, limited to three self-drilling screws.
portal frames and houses, as well as roofing and walling systems [1,2]. Yan and Young [7], Chen et al. [8] and Lu et al. [9] considered the
This is because of its high-strength-to-weight ratio and ease of con- connection strength of such self-drilling screws; it should be noted that
struction. In Australia and New Zealand, CFS is now typically of grade in the case of Yan and Young [7] and Chen et al. [8] the connection
G550, and connections in CFS there are normally formed through self- strength was also determined at elevated temperatures. Moen et al. [10]
drilling screws (Fig. 1), ideally leading to connections that have suffi- also conducted experiments on cold-formed steel-to-steel through-fas-
cient strength, stiffness and ductility [3,4]. Self-drilling screws are es- tened screw connections to correlate tilting, bearing, and shear limit
sentially externally threaded fasteners, as their name implies, are cap- states to single fastener load deformation response. From their study
able of drilling their own hole and form their own mating threads [10], it was concluded that the AISI fastener strength predictions were
(Fig. 2). on average higher than the experimental test values. However, it should
This paper is concerned with the effect of screw patterns, including be mentioned that Moen et al. [10] considered only one screw and it
the number of screws and screw spacing, on the connection strength. In was limited to G330 and G400 steel sheets.
total, 25 experimental test results are presented, all using G550 CFS In terms of the design standards, Rogers and Hancock [11] pre-
steel sheets, using up to five self-drilling screws and for two different sented a summary of self-drilling screws for both the G550 and G300
thicknesses of steel sheets, namely, 1.0 mm and 1.2 mm. All the test sheet steels, but the effect of screw patterns was not considered. They
specimens were designed to meet both the AISI [5] and AS/NZS 4600 found that the AS/NZS 4600 (1996) [12] and AISI (1996) [13] provided
[6] specification for end and edge distance. In the experimental tests, accurate load predictions for when the CFS sheets being connected
failure modes were a combination of bearing, tilting, pull-out and shear through self-drilling screws are of a similar thickness, and that failure is

∗
Corresponding author. University of Auckland, Building 902, Level 4, Newmarket Campus, Auckland, 1010, New Zealand.
E-mail address: [email protected] (K. Roy).

https://doi.org/10.1016/j.tws.2019.106393
Received 15 April 2019; Received in revised form 26 July 2019; Accepted 10 September 2019
Available online 20 September 2019
0263-8231/ © 2019 Elsevier Ltd. All rights reserved.
K. Roy, et al. Thin-Walled Structures 145 (2019) 106393

Notations hole to nearest end

Pu Ultimate test load
CFS Cold-formed steel t1 Thickness of steel sheet member in contact with screw
COV Coefficient of variation head
d Nominal screw diameter t2 Thickness of steel sheet member not in contact with screw
E Young’s modulus of elasticity head
Fy Yield strength 0.2 Static 0.2% proof stress;
FEA Finite element analysis Engineering Strain
e Distance measured in line of force from center of standard

Fig. 1. Different self-drilling screw in CFS framing systems.

more likely to depend on tilting of the screws. They found that when has only been considered by Li et al. [14] and limited to G300 steel of
two different thickness of CFS sheets were connected, bearing failure 1.0 mm thick steel plates. It was found that the screw spacing influ-
was observed, which was not predicted accurately by the design stan- enced the connection strength significantly. They [14] showed that the
dards [12,13]. Rodgers and Hancock [11] proposed a reduced bearing self-drilling screw connections have more strength as screw spacing
coefficient to be used for self-drilling screw connections to limit the increases within a certain range (5 times the diameter of screw spa-
unconservative nature of the design standards [12,13]. It is to be cing), and as the range is exceeded, it has no influence on the strength.
mentioned that this reduced bearing coefficient is now added in the Similar observation was observed by Sapiee et al. [x] However, there is
recent standards [5,6]. work in the literature on the effect of the number of self-drilling screws
With respect to screw patterns, their effect on connection strength on the connection strength. Li et al. [14] found that connections that
have 5 screws parallel to the loading direction can result in steel sheet
fracture. However, this was for G300 steel grade. Li et al. [14] also
concluded that the AISI [5] requires the strength of screws, in addition
to the strength of the connected elements, when determining the con-
nection strength and that the strength limited by shear fracture of the
screws should be obtained by standard testing. This is also important
for different screw patterns, which is considered in this paper.
In terms of finite element modelling, Huynh et al. [3,4] recently
described a nonlinear explicit dynamic finite element (FE) model for
self-drilling screw connections in CFS using the FEA software, ABAQUS
[15] and investigated the effect of up to 3 screws on connection
strength. The FE model used was able to reproduce the experimental
results of Huynh et al. [3] in the form of shear strength, stiffness and
ductility of connections. However, the FEA model did not include any
study on the effects of different screw patterns. Fan et al. [16] had
Fig. 2. Different types of self-drilling screw-connections used in this research.

2
K. Roy, et al. Thin-Walled Structures 145 (2019) 106393

Fig. 3. Failure modes of self-drilling screwed connections.

previously developed a FE model to simulate single lap screw connec- screw connections of G550 sheet steels. In total 25 experiments on
tion in thin steel sheeting of different thicknesses, under static shear. different types of self-drilling screws (12 and 14 gauge) and steel sheet
The model was validated against the experimental test results and can thicknesses (1.0 mm and 1.2 mm). A nonlinear explicit dynamic FE
be used to predict the ultimate resistance, deformation, screw rotation model was then developed which includes a fracture criterion for the
and stress distribution of the connections. However, it is noted that the steel sheets. The FE model was validated against the experimental test
FE model developed by Fan et al. [16] was only for one screw and did results, which gave good agreement against the test results, both in
not consider different screw pattern effects on connection strength. terms of connection strength and failure modes including bearing,
Daudet and Laboube [17] studied the effects of sheet ductility. They tilting, pull-out and shear failure.
[17] concluded that low ductility steel achieved 99% of the expected
capacity. The AISI [5] and AS/NZS [6] could therefore accurately ac- 2. Experimental investigation
count for steel ductility while predicting the connection strength of the
self-drilling screws. 2.1. Test specimens
As mentioned previously, this paper presents an experimental and
numerical investigations on connection strength of CFS self-drilling The test specimens were made from steel sheets of G550 (grade 550)

3
 K. Roy, et al.

Table 1
Specimen dimensions and comparison of experimental and FE strength against the design strengths.
Specimen Plate thickness Plate width Screw diameter Screw spacing End distance Number of specimens tested Experimental results FEA results Failure modes

t w d p e – PEXP P FEA PFEA/PEXP –

(mm) (mm) (mm) (mm) (mm) – (kN) (kN) (kN) –

Type N1: One screw connected

N1-R1C1-B-1 (#12 A2S) 1.2 80.1 5.4 30.1 30.1 1 7.91 8.38 1.06 Tilting + Bearing + Shear
N1-R1C1-B-1 (#14 Tenn JX) 1.2 80.2 6.3 30.2 30.2 1 8.69 9.39 1.08 Tilting + Bearing + Pull out
N1-R1C1-B-1 (#14 ASTeks A3) 1.2 80.2 6.3 30.3 30.3 1 9.20 9.66 1.05 Tilting + Bearing + Pull out
N1-R1C1-B-1 (#14 ASTeks XT) 1.0 80.2 6.3 20.2 20.2 1 5.51 5.90 1.07 Tilting + Bearing + Pull out
N1-R1C1-B-2 (#14 Tenn JX) 1.0 80.1 6.3 25.1 25.1 1 5.67 6.12 1.08 Tilting + Bearing + Pull out
Type N2: Two screws connected
N2-R2C1-B-1 (#12 A2S) 1.2 80.0 5.3 30.1 30.1 1 16.59 18.08 1.09 Tilting + Bearing + Shear

4
N2-R2C1-B-1 (#14 Tenn JX) 1.2 80.1 6.3 30.2 30.2 1 16.76 16.57 0.99 Tilting + Bearing + Pull out
N2-R2C1-B-1 (#14 ASTeks A3) 1.2 80.2 6.3 30.0 30.0 1 16.01 17.13 1.07 Tilting + Bearing + Pull out
N2-R2C1-B-1 (#14 ASTeks XT) 1.0 80.2 6.3 20.2 20.2 1 11.83 12.08 1.02 Tilting + Bearing + Pull out
N2-R2C1-B-2 (#14 ASTeks XT) 1.0 80.1 6.3 20.3 20.3 1 11.56 11.82 1.02 Tilting + Bearing + Pull out
N2-R2C1-B-3 (#14 Tenn JX) 1.0 80.3 6.3 25.1 25.1 1 11.91 11.97 1.01 Tilting + Bearing + Pull out
Type N3: Three screws connected
N3-R3C1-B-1 (#12 A2S) 1.2 80 5.35 30 30 1 24.28 25.74 1.06 Tilting + Bearing + Shear + Pull out
N3-R3C1-B-1 (#14 ASTeks A3) 1.2 80 6.30 30 30 2 25.83 27.90 1.08 Tilting + Bearing + Pull out
N3-R3C1-B-2 (#14 ASTeks A3) 1.2 80 6.30 30 30 2 25.63 27.90 1.09 Tilting + Bearing + Pull out
Type N4: Four screws connected
N4-R4C1-B-1 (#12 A2S) 1.2 80 5.35 30 30 1 32.67 35.28 1.08 Tilting + Bearing + Shear + Pull out
N4-R4C1-B-1 (#14 ASTeks A3) 1.2 80 6.30 30 30 2 32.46 35.38 1.09 Tilting + Bearing + Pull out
N4-R4C1-B-2 (#14 ASTeks A3) 1.2 80 6.30 30 30 2 34.33 35.38 1.03 Tilting + Bearing + Pull out
Type N5: Five screws connected
N5-R5C1-B-1 (#14 ASTeks A3) 1.2 80 6.30 30 30 2 42.24 46.46 1.10 Tilting + Bearing + Pull out
N5-R5C3-B-1 (#14 ASTeks A3) 1.2 80 6.30 30 30 2 42.75 46.60 1.09 Tilting + Bearing + Pull out
Mean 1.08
COV 0.02
Thin-Walled Structures 145 (2019) 106393
K. Roy, et al. Thin-Walled Structures 145 (2019) 106393

2.3. Section labelling

The test specimens were labelled such that the label expressed the
number of screws (N), number of screw rows (R), and number of screw
columns (C), arrangement of screw pattern/symmetry (B) and specimen
numbers. For example, the label “N1-R1C1-B-1” as shown in Fig. 5 is
explained as follows:

• “N1” indicates the number of screws in a specimen, which is 1 in this

label.
Fig. 4. Edge distance and pitches.
• “R1” represents the number of rows occupied by the screws, which
is 1 according to this label.
low ductility steels of 1.0 mm and 1.2 mm thick, which were connected • “C1” represents the represents the number of columns occupied by
by ASTeks A2S 12 - 14 × 20 HWF self-drilling screws along with gauge the screws, which is 1 according to this label.
14 screws of Tenn JX, ASTeks A3, and ASTeks XT. The efficiency of self- • “B” represents the arrangement of screw pattern/symmetry.
drilling process is influenced by the CFS sheet thickness and the • “1” represents the number of repetitive tests or specimen numbers,
strength of such screw connections is affected by various parameters which is 1 as per this label.
including the number of screws, spacing between the screws, edge and
end spacing between the screw and steel plates and screw patterns. The 2.4. Material testing
individual CFS sheets consist of a nominal length of 400 mm and
nominal width of 80 mm, whereby both ends were glued with packing Tensile coupon tests were carried out to determine the material
shims of 60 mm long. In total 25 new experimental tests are reported. properties of the test specimens. The tensile coupons were cut from the
All the screw spacings were measured from centre-to-centre distance of centre of the web plate in the longitudinal direction of the untested
the screw diameters. Different types of test specimens and their di- specimens. The tensile coupons were cut in accordance with the AS
mensions are shown in Table 1. 1391 [18]. Five coupons were obtained from both longitudinal and
transverse directions of the steel sheets. The coupons were tested in an
Instron 4469 tensile testing machine, which has a capacity of 50 kN. A
2.2. End and edge distance calibrated extensometer of 50 mm gauge length was used to determine
the tensile strain of the coupons from the results of the tensile coupon
All the test specimens in this research are designed to meet both tests the average values of the Young's modulus and yield strength were
AISI [5] and AS/NZS 4600 [6] specification on end and edge distance. 203 GPa and 586.8 MPa, respectively for the steel sheets. The material
With reference to the AS/NZS 4600 [6], it is stated that the end properties of the CFS sheets are also summarized in Table 2(a).
distance (e1), shown in Fig. 4, is the edge which are parallel to the To determine the tensile material properties of screws, the same
loading. The standard specifies in clause 5.4.2.2 that the centre of a Instron machine was used, but with different grips and textured jaws
screw to the edge of any part shall be not less than 1.5df and the dis- surfaces, which allowed proper griping of the screw threads. Total of
tance between centres of screw should not be less than 3d (e1 ≥ 1.5d, five screws were tested in a tension test. The head of the screws was cut
p1 ≥ 3d). Whereas Table 1 of AISI [5] specifies that the end distance off to enable clamping the shank and threads of the screws directly in
and screw spacing should be 30 mm and 60 mm for fastener nominal the tension testing machine's jaws. Table 2(b) provides the mechanical
diameter of less than 6.4 mm (e1 = 30 mm, p1 = 60 mm). properties of the tested screws.
Edge distance, shown as e2 in Fig. 4, is defined as the free edge
which are placed perpendicular to the direction of the loading (Aus- 2.5. Testing rig and loading procedure
tralian/New Zealand Standard 4600 [6]). The standard specifies in
clause 5.4.2.2 that the screw should be placed at an edge distance at Fig. 6 shows the machine used for the experimental tests and the
1.5d, between the centres of the screw diameter to the free edge and the test-set-up is shown in Fig. 7 for the standard lap-joint test. An Instron
distance between centres of screw should not be less than 3d 5900 machine, which had a capacity of 50 kN, was used to set up the
(e2 ≥ 1.5d, p2 ≥ 3d). However, according to the AISI [5], the edge specimens and apply the loading. A U section of dimensions
distance is stated to require a minimum of 1.5d and less than 2d spacing 9 mm × 9 mm x 9 mm and thickness of 1.0 mm was used to set up the
starting from the centre of the screw to all edge distance, and the screw test specimens in the test rig. A Perspex platform of 30 mm × 30 mm
spacing should be 19 mm (1.5d ≤ e2 < 2d, p2 = 19 mm). was glued to the specimen at both the ends. Four LVDTs were used to

Fig. 5. Labelling of test specimens.

5
K. Roy, et al. Thin-Walled Structures 145 (2019) 106393

Young's modulus

Young's modulus
GPa

GPa
203

208
E

E
Ultimate stress

(MPa)

697.2
σu

Ultimate stress

(MPa)

963.7
σu
Gauge width

(mm)

12.5
b

Fig. 6. Testing machine.

Yield stress

Yield stress

measure the displacements of the test specimens from both the top and
bottom of the connections. LVDT positions are shown in Fig. 7. Strain
(MPa)

(MPa)
586.8

813.6
σ0.2

gauges were also used to measure the ultimate strain at failure for all
σ0.2

test specimens. The strain gauge locations are shown in Fig. 7. The
readings of the LVDTs and strain gauges were recorded after each in-
crement of loading. The loading rate was kept as 0.5 mm/min for all
test specimens. Two cameras were used to capture the deformed shapes
and ultimate failure modes of the test specimens. One camera was used
in the front side of the test specimens while the other camera was set up
Gauge length

at the back side of the specimens to capture the sequence of failures

during the experimental tests.
(mm)

50
L0

3. Experimental results
Longitudinal (Average results from coupon tests of 5 screws)

The performance of different types of screws connections in the steel

plates was investigated in terms of number of screws, group effect for
more than one screw, and screw pattern. It was found that the beha-
viour of the connection varies with different types of screws. It was also
observed from the experimental tests that the screw connections ex-
Longitudinal (average results of 5 coupons)

perienced multiple screw failures in different sequence. The mixed

failure modes observed in the experimental tests include either single or
Material properties of specimens

combined modes of failures as tilting, bearing, tearing of the steel

plates, pullover and pull-out failure of the screws. Both the screw and
plate failures were observed in the experiments. Further information
regarding the different failure modes are detailed in the following
sections based on the number of screws and screw patterns:
(a) CFS sheets.

3.1. Number of screws

(b) Screws
Direction

Direction
Table 2

3.1.1. Connection strength

The experimental results for test specimens with gauge 12 and 14

6
K. Roy, et al. Thin-Walled Structures 145 (2019) 106393

Fig. 7. Experimental setup and instrumentation.

Fig. 8. Bearing and tilting failure for Gauge 12 screw series specimens.

Fig. 9. Bearing and tilting failure for Gauge 14 screw series specimens.

7
K. Roy, et al. Thin-Walled Structures 145 (2019) 106393

screw connections are shown in Table 1. For some specimens (N1- 3.1.2. Failure modes
R1C1-B-2 (#14 Tenn JX), N2-R2C1-B-2 (#14 ASTeks XT), N2-R2C1-B-3 Most of the test specimens showed tilting of screws when the load
(#14 Tenn JX), N3-R3C1-B-2 (#14 ASTeks A3), N4-R4C1-B-2 (#14 was applied initially (Fig. 8), followed by the bearing failure and slip of
ASTeks A3)), the tests were repeated by 2 or 3 times and the average the screw thread, and curling effect (Figs. 9 and 10). Curling effect
values of test strengths are reported in Table 1. As can be seen from occurred in both the gauge 12 and gauge 14 screw test series, causing
Table 1, the strength of the screw connections increases with the in- the steel sheets to tilt (Figs. 8(b) and 9 and ) and then slip off the screw
creasing number of screws. For gauge 14 screws of ASTeks A3, screw threads (Figs. 8–10(a)) throughout the duration of the loading. Curling
connections yielded at much higher ultimate loads compared to the effect and deformation of the hole caused slipping of the screw thread
Tenn JX screws for both the single or double screw connections. (Figs. 8 and 9). The bearing and tilting failure for Gauge 14 screw series
specimens are shown in Fig. 10. The shear off occurred at the screw
head for gauge 12 screw specimen as shown in Fig. 11. The thread slip
was caused by the bearing failure and curling effect of the specimen
(Fig. 12(a) and (b) and for Tenn JX Screws and ASTeks XT Screws,
respectively)). Sheet piling was also seen in both samples due to tearing
of steel sheets as the screws were pulled out. Generally, the ultimate
failure for the gauge 14 screw was pull-out failure (Fig. 13). Some of the
pull-out failures were observed from the gauge 12 for 3 and 4 screws
only. Hole elongation and sheet piling behaviour was observed for both
the screw gauges of 12 mm and 14 mm series and are shown in Fig. 14.
All the screws which failed under the combination of tilting and bearing
failures, they were not influenced by the number of screws for both the
steel thicknesses of 1.0 mm and 1.2 mm.

3.1.3. Load verses displacement graphs

The load verses displacement graphs for gauge 12 and gauge 14
screw specimens are plotted in Fig. 15(a) and 15(e) and and
Fig. 16(a) and 16(e) and , respectively. All the test specimens experi-
enced elastic deformation in the initial stage of the loadings until it
reached the ultimate strength. After which, non-linear behaviour was
observed. Beyond the ultimate strength, several peaks were observed,
indicating that the curling effect happened, and the screw thread was
consecutively pulled-out, or rupture failure occurred.
For gauge 14 screws, two different types of screw lengths were used
as 20 mm and 25 mm for ASTeks A3 and Tenn JX, respectively. The
longer screw had more threads for the steel sheet to grip on before the
fracture started, thus, yielding more and showing significant extension
compared to the shorter screw. This behaviour is shown in
Fig. 17(a) and (b) and ; whereby, the extension is greater in N1-R1C1-B-
2 compared to N1-R1C1-B-1, and greater in N2-R2C1-B-3 compared to
N2-R2C1-B-1 and N2-R2C1-B-2. For N3 series, only one experiment was
carried out and the ultimate load was on the second peak, due to the
slipping of screw threads prior to the first peak load. The ultimate load
was around 17 kN, which is almost 3 times the ultimate load of the N1
series.

3.1.4. Group effect on strength

The group effect of the screws on connection strength is discussed in
this section. The connection strength increases linearly as the number of
screws increases for both the gauge 12 and gauge 14 screws, until the
first peak was observed. There was no significant group effect for gauge
12 series and only minor effect was observed for gauge 14 series as
shown in Table 3. This is because the low ductility steel has lower Fu/Fy
ratio, and thus contributed to a lower stress distribution, which even-
tually improved the performance of the screws within the steel plates.
Majority of the test results for one to four screws showed slightly lower
connection strength for gauge 12 screws as compared to gauge 14
screws. This is because gauge 12 and 14 screws had a nominal diameter
Fig. 10. Bearing and tilting failure for Gauge 14 screw series specimens for of 5.35 mm and 6.30 mm, respectively.
different types of screws.

8
K. Roy, et al. Thin-Walled Structures 145 (2019) 106393

Fig. 11. Screw shear off for gauge 12 screw series specimens.

3.2. Screw pattern configuration 4. Numerical investigation

Three different screw patterns were designed for the experiments, 4.1. General
which include specimens with three, four and five screws as shown in
Fig. 18. The ultimate loads for all gauge 14 ASTeks A3 screw patterns A nonlinear elasto-plastic finite element model was developed using
are presented in Table 4. As can be seen from Table 4, all test specimens ABAQUS [15]. The finite element model was based on the centre line
failed by tilting and bearing failure. The load vs displacement beha- dimensions of the screws and steel sheets. In the model, the measured
viour of the specimens with 3, 4 and 5 screws are shown in Fig. 19(a) cross-sectional dimensions were used. Specific modelling techniques
and (b) and 19(c), respectively. On the other hand, Table 5 shows the are described in the following subsections.
effects of screw pattern on the connection strength for 3, 4, and 5
screws. The connection strengths of the screw patterns for 3, 4, and 5 4.2. Geometry and material modelling
screws are 5.95%, 1.52%, and 3.91% lesser when compared to the
number of screws. Although the differential percentages are less than The whole geometry of the steel sheets and screws was modelled.
10%, the screw arrangement does contribute to some effects for the However, for some of the FEA models, half the geometry of the screws
screw connection strength. was modelled considering the appropriate symmetry boundary condi-
tions. This is to reduce the computation time to finish the analysis as
compared to the FEA models with full screw models. Coupon tests were
mainly performed to investigate the material properties for both the
screws and steel sheets as described in section 2.3. Yield strength of

9
K. Roy, et al. Thin-Walled Structures 145 (2019) 106393

Fig. 12. Thread slip occurrence on the specimens.

Fig. 13. Pull-out failure for gauge 14 screw series specimens.

10
K. Roy, et al. Thin-Walled Structures 145 (2019) 106393

Fig. 14. Hole elongation and sheet piling.

Fig. 15. Typical load verses displacement curves for 12-gauge screws.

547 MPa and ultimate stress of 659 MPa, along with Young's modulus of true = (1 + ) (1)
208 GPa, was used in the numerical modelling of the steel plates. For
the screws, the material properties were also determined in the ex- true
perimental tests and used in the FEA models. The mechanical properties = ln(1 + )
true (pl)
E (2)
conducted from the tensile tests for both the steel sheets and screws,
were used for the engineering stress-strain curves. As per the ABAQUS where E is the Young's modulus, σ true is the true stress, σ u is the tensile
manual [15], the engineering material curves were converted into true ultimate strength, and are the engineering stress and strain, re-
material curves by following the equations below: spectively in ABAQUS [15].

11
K. Roy, et al. Thin-Walled Structures 145 (2019) 106393

Fig. 16. Typical load verses displacement curves for 14-gauge screws.

4.3. Type of element and finite element mesh screw model and half width model of the screws.

Three dimensional eight noded solid elements with reduced in- 4.4. Boundary conditions and load application
tegration point (C3D8R) were used to model the screws and steel plates.
An element size of 1.0 mm by 1.0 mm was found to be appropriate for In explicit analysis the surfaces between two sheets having surface
screws, based on the results of a convergence study. Mesh refinement to surface interaction was considered along with screw heads and upper
was made around the screws and near the holes in the steel plates to steel plate, and screw threads and drilled hole. Furthermore, inter-
incorporate the fracture behaviour accurately. On the other hand, a locking between two steel plates and the screw thread was modelled by
mesh size of 2.0 mm by 2.0 mm was used for the steel plates. A mesh using a friction coefficient. It was ensured through calibration that the
sensitivity analysis was performed to confirm the number of elements friction should be enough to hold the screws having less effect to the
for both the screws and steel plates. Typical finite element meshes are load displacement curves.
shown in Fig. 20 for specimens with 12-gauge screws for both the full For all the FEA models, full geometries of the steel sheets were

12
K. Roy, et al. Thin-Walled Structures 145 (2019) 106393

Fig. 17. Typical load verses displacement curves for AsTeks XT and Tenn JX screw series.

modelled. On the other hand, for some FEA models, full screw geometry 4.5. Modelling of fracture
was models with screw head and threads. However, for some FEA
models only half the screws were modelled using the symmetry A fracture criterion was used to model the fracture behaviour of the
boundary conditions. The boundary conditions used in both the full and G550 steel sheets in ABAQUS [15] based on the previous work on
symmetry models are shown in Fig. 21. fracture [19]. The ductility of G550 steel is less due to cold-reduction
It has been observed that for some models, material piling happened process, which leads to lower values of displacements and fracture
in front of screws while keeping in contact, which is mainly due to strains, which was the reason behind choosing the specific fracture
bearing failure. However, with the initiation of fracture in the specimen criteria as developed by Choung et al. [19]. “Ductile damage” was used
on screws, different types of contact occur with the exposure of inner as the fracture model in ABAQUS [15]. The fracture strain and stress
material. The load was applied at the ends of the sheet sheets to si- triaxiality data was considered from the Choung et al. [19] and shown
mulate the experimental failure loads. in Fig. 22. Limited deformation was observed in thick steel sheets;

13
K. Roy, et al. Thin-Walled Structures 145 (2019) 106393

Table 3
Group effect of the screw arrangements on strength of screw connections.
Numbers of screws, n Specimen Experimental results PEXP/n Group effect ratio
P EXP

(kN) (kN) –
Gauge 12
1 N1-R1C1-B-1 (#12 A2S) 7.91 7.91 1.00
2 N2-R2C1-B-1 (#12 A2S) 16.59 8.30 1.04
3 N3-R3C1-B-1 (#12 A2S) 24.28 8.09 1.02
4 N4-R4C1-B-1 (#12 A2S) 32.67 8.17 1.03
Gauge 14
1 N1-R1C1-B-1 (#14 Tenn JX) 8.69 8.69 1.00
N1-R1C1-B-1 (#14 ASTeks A3) 9.20 9.20 1.00
N1-R1C1-B-1 (#14 ASTeks XT) 5.51 5.51 1.00
N1-R1C1-B-2 (#14 Tenn JX) 5.67 5.67 1.00
2 N2-R2C1-B-1 (#14 Tenn JX) 16.76 8.38 0.86
N2-R2C1-B-1 (#14 ASTeks A3) 16.01 8.00 0.90
N2-R2C1-B-1 (#14 ASTeks XT) 11.83 5.92 106
N2-R2C1-B-2 (#14 ASTeks XT) 11.56 5.78 1.05
N2-R2C1-B-3 (#14 Tenn JX) 11.911 5.96 1.04
3 N3-R3C1-B-1 (#14 ASTeks A3) 25.83 8.61 1.01
N3-R3C1-B-2 (#14 ASTeks A3) 25.63 8.54 0.95
4 N4-R4C1-B-1 (#14 ASTeks A3) 32.46 8.12 0.91
N4-R4C1-B-2 (#14 ASTeks A3) 34.33 8.58 0.96
5 N5-R5C1-B-1 (#14 ASTeks A3) 42.24 8.45 0.94
N5-R5C3-B-1 (#14 ASTeks A3) 42.75 8.55 0.96

Fig. 18. Screw patterns for specimens with three, four and five screw.

Table 4
Comparsion of experimental and finite element results against the design strengths for different types of screw patterns.
Specimen Experimental results FEA results

PEXP PFEA PFEA/PEXP Failure modes

(kN) (kN) (kN)

Type N3: Three screws connected

N3-R3C1-B-1 (#14 ASTeks A3) 23.51 25.39 1.08 Tilting + Bearing + Pull out + Crack on screw cap
N3-R3C1-B-2 (#14 ASTeks A3) 24.89 25.39 1.02 Tilting + Bearing + Pull out
Mean 1.05
COV 0.04
Type N4: Four screws connected
N4-R4C1-B-1 (#14 ASTeks A3) 34.23 33.37 0.97 Tilting + Bearing + Pull out + Crack on two screw caps
N4-R4C1-B-1 (#14 ASTeks A3) 31.55 33.37 1.06 Tilting + Bearing + Pull out + Crack on one screw cap
Mean 1.02
COV 0.06
Type N5: Five screws connected
N5-R5C1-B -1(#14 ASTeks A3) 40.96 42.56 1.04 Tilting + Bearing + Pull out
N5-R5C3-B -1(#14 ASTeks A3) 40.72 42.56 1.05 Tilting + Bearing + Pull out
Mean 1.05
COV 0.01

14
K. Roy, et al. Thin-Walled Structures 145 (2019) 106393

Fig. 19. Load against displacement graph #14-gauge AsTek A3 screw pattern configuration.

Table 5
Screw connection strength percentage difference for pattern of screws.
Number of screws Mean experimental results for number of Mean experimental results for screw Screw connection strength percentage difference for
connected screws series (kN) pattern series (kN) pattern screw connection (%)

3 25.73 24.20 −5.95

4 33.40 32.89 −1.52
5 42.50 40.84 −3.91

Fig. 20. Typical finite element meshing (N1-R1C1-B-1 (#12 A2S)).

15
K. Roy, et al. Thin-Walled Structures 145 (2019) 106393

Fig. 21. Boundary conditions applied in the FE model (N1-R1C1-B-1 (#12 A2S)).

however, significant deformation of steel sheets was observed near the 4.7. Finite element model validation
vicinity of the screw holes for thinner sheets which led to more elon-
gation and finely bearing failure of the steel sheets. The comparison of failure loads from the laboratory and the finite
element analysis for all specimens are shown in Tables 1 and 4 for
experiments of number of screws and screw patterns, respectively.
4.6. Analysis procedure Close comparison is obtained between the tests and finite element re-
sults in terms of connection strength and failure modes. Fig. 23 shows
In ABAQUS [15], both implicit and explicit solver is available to the failure modes of the specimens including the screw connections
analyse any non-linear models. The implicit solver uses current time to obtained from the finite element analysis, which showed good agree-
solve each step of the problem using the method of iterations. On the ment with experimental failure modes (Table 1).
other hand, explicit solver uses the parameters of current time to ana- Fig. 24(a) and 24(c) compare the load-displacement curves from
lyse the parameters of next time interval, while analysing the model experimental tests and FEA for N2 series of 14-gauge screws. In some of
and therefore no iterations are required in the explicit solver. In this the experiments, some initial slip of screws was observed (Fig. 24a and
study, dynamic explicit was used as the analysis method. The main c), due to which the initial stiffness obtained from the FEA was higher
reason for choosing the explicit analysis was to avoid the convergence than that of the test results. Overall, good agreement is achieved be-
problem due to the complicated interactions between the threads of the tween the experimental and numerical values of shear, tilting and
screws and the holes in the steel plates. The explicit solver can solve bearing strength and their failure modes.
highly discontinuous problems with higher degree of efficiency [15].
6. Conclusions

A detailed experimental investigation on the behaviour of screw

connections of gauge 12 and 14 is presented in this paper to understand
their behaviour and failure modes under shear for G550 steel sheets. A
total of 25 experimental tests are reported on different specimen con-
figurations. Two different types of steel sheet thicknesses (1.0 mm and
1.2 mm) were used to connect each other with the help of self-drilling
screws. The number of screws and screw patterns were the main vari-
ables in the experimental program. The following conclusions can be
drawn from the work presented in this paper:

1. The connection strengths of the screw patterns for 3, 4, and 5 screws

are 5.95%, 1.52%, and 3.91% lesser when compared to the number
of screws. Although the differential percentages are less than 10%,
the screw arrangement does contribute to some effects for the con-
nection strength.
2. All the screws which failed under the combination of tilting and
bearing failures, they were not influenced by the number of screws
for both the steel thicknesses of 1 mm and 1.2 mm.
Fig. 22. Fracture criteria used in the FE model (Choung et al. [19]).
3. Majority of the test results for one to four screws showed slightly

16
K. Roy, et al. Thin-Walled Structures 145 (2019) 106393

Fig. 23. Failure modes of the specimens from the experiments and FEA.

17
K. Roy, et al. Thin-Walled Structures 145 (2019) 106393

Fig. 24. Comparison of experimental and FEA strengths for different self-drilling screw connections.

lower connection strength for gauge 12 screws as compared to [4] M.T. Huynh, C.H. Pham, G.J. Hancock, On the finite element modeling of the
gauge 14 screws. This is because gauge 12 and 14 screws had a screwed connections of cold-formed steel, Proceedings of the 4th Congress
International de Geotechnique-Ouvrages-Structures, 26-27 October 2017. Ho Chi
nominal diameter of 5.35 mm and 6.30 mm, respectively. Minh City, Vietnam, CIGOS, 2017.
4. A non-linear explicit dynamic finite element model was then de- [5] AISI, North American Cold-Formed Steel Specification for the Design of Cold-
veloped which includes a fracture criterion for the steel sheets. The Formed Steel Structural Members, AISI S100, American Iron and Steel Institute,
Washington DC, U.S.A, 2016.
finite element model includes 3D modelling of the screws with both [6] Australian/New Zealand Standard, Cold-Formed Steel Structures, AS/NZS
actual dimensions and half width screws using symmetry boundary 4600:2018, Standards Australia/Standards New Zealand, Sydney, Australia, 2018.
conditions were modelled which also included a ductile metal da- [7] S. Yan, B. Young, Screwed connections of thin sheet steels at elevated temperatures
– Part I: steady state tests, Eng. Struct. 35 (2012) 234–243.
mage criterion based on fracture strain formulas. In the ductile da- [8] W. Chen, J. Ye, M. Zhao, Steady-and transient-state response of cold-formed steel-
mage model, the stress triaxiality and strain rate were included. The to-steel screwed connections at elevated temperatures, J. Const. Steel Research. 144
finite element model could effectively simulate the different failure (2018) 13–20.
[9] W. Lu, Z. Ma, P. Makelainen, J. Outinen, Behaviour of shear connectors in cold-
modes including bearing, tilting, pull-out and shear failure of the
formed steel sheeting at ambient and elevated temperatures, Thin-Walled Struct. 61
screw fasteners, which were similar to the experimental failure (2012) 229–238.
modes. [10] C.D. Moen, D.A. Padilla-Llano, S. Corner, C. Ding, Towards load-deformation
5. The experimental test results and the finite element model presented models for screw-fastened cold-formed steel-to-steel shear connections, Proceedings
of 22nd International Specialty Conference on Cold-Formed Steel Structures, 5th -
in this paper can be used by the future researchers and practicing 6th November, 2014 St. Louis, Missouri.
engineering to predict the failure modes and connection strengths of [11] C.A. Rogers, G.J. Hancock, Screwed connection test of thin G550 and G300 sheet
such self-drilling screw connections in shear. steels, J. Struct. Eng. ASCE. 125 (1999) 128–136.
[12] Australian/New Zealand Standard, Cold-Formed Steel Structures, Standards
Australia/Standards New Zealand, Sydney, Australia, 1996 AS/NZS 4600:1996.
Acknowledgement [13] AISI, North American Cold-Formed Steel Specification for the Design of Cold-
Formed Steel Structural Members, AISI S100, American Iron and Steel Institute,
Washington DC, U.S.A, 1996.
This research was supported by the Malaysia Ministry of Education [14] Y. Li, R. Ma, X. Yao, Shear behavior of screw connections for cold-formed, thin
(MoE) under the Fundamental Research Grant Scheme (FRGS), FRGS/ walled teel structure, Proceedings of 20th International Specialty Conference on
1/2018/TK01/CURTIN/01/1. The authors would like to thank for their Cold-Formed Steel Structures, 3rd -, 4th November, 2010 St. Louis, Missouri.
[15] ABAQUS Analysis User's Manual-Version 6.14-2, ABAQUS Inc., USA, 2018.
support. [16] L. Fan, J. Rondal, S. Cescotto, Finite element modelling of single lap screw con-
nections in steel sheeting under static shear, Thin-Walled Struct. 27 (2) (1997)
References 165–185.
[17] R. Daudet, R.A. LaBoube, Shear behaviour of self-drilling screws used in low duc-
tility steel, Proceedings of the 13th International Conference on Cold-Formed Steel
[1] P. Kyvelou, L. Gardner, D.A. Nethercot, Design of composite cold-formed steel Structures, University of Missouri-Rolla, USA, 17-18th Oct 1996St. Louis, Missouri.
flooring systems, Structure 12 (2017) 242–252. [18] AS 1391, Metallic Material's - Tensile Testing at Ambient Temperature. Standard
[2] R.A. LaBoube, M.A. Sokol, Behaviour of screw connections in residential con- Australia, (2007).
struction, J. Struct. Eng. ASCE. 38 (2002) 115–118. [19] J. Choung, C.-S. Shim, H.-C. Song, Estimation of failure strain of EH36 high strength
[3] M.T. Huynh, C.H. Pham, G.J. Hancock, Experiments on screwed connections in marine structural steel using average stress triaxiality, Mar. Struct. 29 (1) (2012)
shear using high strength cold-reduced sheet steel, Proceedings of the 18th 1–21.
International Conference on Thin Walled Structures, vol. 24-27 July, 2018. Lisbon,
Portugal, ICTWS, (2018).

18