Keynote Paper

Can be Considered The Thin-Walled Steel Structures Suitable to be Used in

Seismic Regions?

D. Dubina

Department of Steel Structures and Structural Mechanics, Faculty of Civil Engineering and
Architecture, "Politehnica" University of Timisoara, Romania

ABSTRACT

Thin-walled steel structures are usually made by sections of class 4 or 3 and they can be
obtained either by cold-forming or welding of thin plates. Traditionally these structures, being
slender, are considered non-ductile and not allowed for plastic design. As a consequence a
reduction factor q=1 and elastic design can be applied only.
Present paper summarises recent research in the field and demonstrate that thin-walled steel
structures could be considered as “low dissipative” according to the EN 1998-1 classification,
which means a q factor value of 1.5 to 2 may be used for seismic design.

INTRODUCTION

Thin-walled steel structures are usually made by thin-walled sections, of class 4 or, at most,
of class 3. Such types of structures are usually made by cold-formed or thin plate welded
sections. Compared with hot-rolled sections (of class 1 or 2), they are characterised by a reduced
post-elastic strength and, as a consequence, by a reduced ductility (e.g. they do not have
sufficient plastic rotation capacity to form plastic hinges).
The European specific design rules for cold-formed steel design have no recommendations
for seismic design of these structures. In the North American Specification (AISI, 2001), the
provisions in Section G, "Design of cold-formed steel structural members and connections for
cyclic loadings", rather refers to fatigue then to seismic behaviour. The 2003 draft of the
Australian/New Zealand Standard (Revision of AS/NZS 4600:1996) for Cold-formed Steel
Structures in its Section 6 "Fatigue" includes similar provisions as AISI 2001. Therefore, actual
cold-formed steel design codes do not contain specific recommendations for seismic design of
cold-formed steel structures.
EN 1998-1, even does not specifically mention the use of thin-walled steel sections for
seismic resistant structures, provides for low dissipative (e.g. low ductility) structures a
behaviour factor q of values from 1.5 to 2.0 (see Table 1). Assuming that such a type of

83
structures are made by “elastic” sections (e.g. class 3 or class 4, in case) a q factor greater than
1.0 can be justified by overstrength and structural redundancy. The question of which the
present paper attempt to answer is if the thin-walled structures can be or not classified as “low
dissipative”.

Table 1. Behaviour factor “q” (pr EN 1998-1)

Behaviour factor “q”
Design concept Required ductility class
EC8
Highly dissipative
q≥ 4.0 H (high)
structures
Medium dissipative
2.0 ≤ q ≤ 4.0 M (medium)
structures
Structures with limited
q = 1.5 ÷ 2 L (low)
dissipation

ELASTIC-PLASTIC BEHAVIOUR OF THIN-WALLED SECTIONS IN POST-

CRITICAL RANGE

Post-critical strength of thin plates and sections

The behaviour of an ideal and an actual plate is shown in Figure 1. Looking to the path of an
ideal plate stress-deflection curve it can be observed:
1. In the pre-critical range σ < σ cr , the plate has a linear behaviour characterised by a plane
stress state;
2. When the critical stress point is reached, σ = σ cr , the plate suddenly loose its rigidity
(see Figure 1c) and significant increase of deflection occurs;
3. In post-critical range σ cr < σ < f y the behaviour continues to remain elastic and due to
“membrane lag” effect a stabilizing action occurs due to which a post-critical stress
reserve is available. This “membrane lag” is the explanation for the non-linear elastic
behaviour within this range;
4. When the first yield is reached in the point σ = σ pl , the curve change the curvature, and
the plate starts its elastic-plastic behaviour. In the domain σ ≤ σ pl , the unloading path is
fully reversible. For this reason, the point σ = σ pl is also called “reversibility” point;
5. In the range σ > σ pl the plate loses rapidly its stiffness and reaches the ultimate strength,
1u.
When both geometrical and material imperfections are present a continuous deformation
process starts from the beginning (Figure 1); more increase the initial deflection w0, smoother is
the -w curve. It is difficult to capture the points cr and pl, and often during tests u is taken as
cr and vice versa. Therefore for the actual thin plate, the cr-u range is quite reduced.
In case of thin-walled bars, the sectional buckling (e.g. local or distortional buckling) occurs
prior to the initiation of plastification. Sectional buckling is characterised by the stable post-
critical path and bar does not fail when it occurs, but significantly lose from its stiffness. The
yielding starts at the corners of cross-section a few time before the failure of the bar, then

84
sectional buckling changes into local plastic mechanism quasi-simultaneously with global
buckling occurrence.
Localization of buckling patterns and the local buckling mechanism
For stub columns, multiple local buckling modes may occur simultaneously under the same
critical load. For a slender member, multiple local buckling modes, e.g. m-1, m, m+1,
characterised by corresponding Lm-1, Lm and Lm+1 half wave length, may interact each other and
give rise to an unstable post-critical behaviour called “localisation of the buckling pattern”.
The localised buckling mode is in fact an interactive or coupled mode. This “first” interaction
may occur prior the overall buckling of the member. The “second” interaction, between the
localised mode and the overall one is really dangerous because it is accompanied by a very
strong erosion of critical bifurcation load. When localisation of buckling patterns occurs, then
member post-buckling behaviour is characterised by large local displacements, in the inelastic
range, which produce the plastic folding of walls and the member falls into a plastic mechanism
(Figure 2). This kind of behaviour is specific for cold-formed steel sections and is confirmed by
both tests and numerical simulations (Figure 3), and could be the source for some ductility.

σ
fy perfect elastic plate
perfect elastic-plastic plate
P
σu σ=
b ⋅t

σpl y
P
Actual elastic-plastic plate
σcr u
w0
First yield
w
t
a

x
b
w
w0,1 w0: initial imperfection
w0,2
w0,3
a) Stress vs. deflection
σ kt/k

ideal elastic plate Ideal elastic plate (w0/t=0)

w0,1/t
1.0
σcr
Actual elastic plate drop of stiffness in
σu w0,1/t postcritical range
w0,1 < w0,2 < w0,3
w0,1/t
0.5
actual elastic-plastic
plate

0.0 P/Pcr

ε=u/a 0.0 0.5 1.0 1.5 2.0

b) Stress vs. axial strain c) Axial stress vs. stiffness

Fig. 1. Behaviour of ideal and actual simply supported plate in uniaxial stress

85
 In case of slender bars, when local buckling firstly appears, it always changes into local
plastic mechanism when the member fails (see Figure 4).

P=P/A⋅fy

Flange buckling
Pcr,2
Web buckling

Pu P

Pcr,1

Fig. 2. Local plastic mechanism failure of a plain channel stub column

(a) members in compression

(a) members in bending

Fig. 3. Numerical and experimental evidences of plastic mechanism failure

Plastic rotation capacity of thin-walled sections

Ungureanu and Dubina (2004) used the local plastic mechanism theory developed by Murray
and Khoo (1981) to characterise the local and interactive buckling of cold-formed steel sections.
Moldovan et al. (1999) used the same theory to evaluate the plastic rotation capacity (e.g.
ductility) of built-up U and C cold-formed steel beams. They developed the DUCTROT – TWM

86
computer code, calibrated via test results obtained at the University of Naples (1992) and found
an available cross-sectional ductility, µ, equal to 1.7:
µ = χu / χ y (1)
where χy is the beam curvature corresponding to the initiation of plastic deformations, while χu
is the ultimate curvature.

Fig. 4. FEM simulation of plastic-elastic interaction between the local plastic mechanism
(roof type) and flexural buckling for a lipped channel section

Figure 5 shows the moment-rotation curves for two sections studied by Moldovan et al.
(1999). One can see for these particular sections the “flip” type local mechanism approximates
better the ultimate bending moment. Based on these results, the authors suggested to use a
behaviour factor q=1.7 when design cold-formed steel structures.

70 35 roof mechanism
roof mechanism
60 30
flip mechanism
flip mechanism
50 25
M[kNm]

experimental curve
M[kNm]

40 20 200x50x5

30
200x100x5 15
experimental curve
10
20
5
10
0 -3
0 0 θel 20 40 θpl 60 80 100 x 10
θ [rad]
-3
0 θel 20 θpl 40 60 80 100 120 x 10

θ [rad]
Fig. 5. M-θ curves for built-up sections

Post-elastic behaviour and fatigue

North American Specification (AISI, 2001) in the Section G “Design of cold-formed
structural members and connections for cycling loadings”, and also the actual revised version of
Australian and New Zealand Standard (Revision of AS/NZS 4600:1996) in Section 6 “Fatigue”,
as already mentioned in INTRODUCTION, are the only modern cold-formed steel design codes
to address the problem of fatigue behaviour (e.g. cycling loading effect). However, the relevant

87
provisions in these codes refer mainly to connecting details and not to the member strength
under cycling loading, either in case of high or low fatigue circumstances.
Both codes consider:
• the occurrence of full design wind or earthquake loads is to infrequent to warrant
consideration of fatigue design building;
• evaluation of fatigue resistance is not required if the member of cycles of application
of live load is less than 20000;
• calculated stresses shall be based upon elastic analysis.
For the purpose of this paper it is useful to emphasise that, in recent years, the low fatigue
approach, accounting for cumulative plastic deformations of steel members or connections
under repeated and reversal loadings, is considered as a better way to evaluate the seismic
response of steel structures.
Related to the particular problem of fatigue behaviour of thin-walled cold-formed steel
sections, Lindner and Gläβer (2004) shown that, in such a case, the fatigue damage is rather a
plastic problem than an elastic one. In fact, the tests carried out by the authors clearly
demonstrated the failure under cycling loads occurred only if plastic strains are present. When
the ultimate load in the test was reached, cracks occur along the plastic hinges of a local plastic
mechanism, as shown in Figure 6.

Fig. 6. a) flip disc in web; b) flip disc in test specimen X; c) Crack start in web hinges (Lindner
and Gläβer, 2004)

Practically, even after 2 millions of elastic cycles no damage occurred in the tested specimens.
This means thin-walled cold-formed steel members are no sensitive at all to the effect of cyclic
loading providing it remain in elastic range only.

BEHAVIOUR FACTORS. DUCTILITY, OVERSTRENGTH AND REDUNDANCY

Earthquake force reduction factors are widely used in design codes to reduce elastic spectral
demand to design ones. Structural design to earthquake forces lower than those necessary for an
elastic response are derived from the observation that most structures are able to survive a major
earthquake due to dissipation of energy by plastic excursions and overstrength.
In traditional “capacity design” procedure, a single reduction factor is generally used.
However, distinction and quantification of different components of the force reduction factors
are useful for a better understanding of the seismic response of structures. Figure 7 presents a

88
typical relationship between base shear and top displacement of a structure. Based on the
bilinear idealisation of the real response, the ductility may be defined as:
µ = Du Dy (2)
where Du is the ultimate top displacement, Dy is the top displacement at global yield. Other
terms used in the figure are: Fe – elastic base shear; Fy – yield base shear; F1 – base shear at the
first plastic hinge; Fd – design base shear.

Fe - ductility factor
elastic response qµ = Fe Fy
qµ - overstrength factor
q
real response qd qS = Fy Fd
Fy - redundancy factor
qR
F1
idealized response
qS qR = Vy V1
q Sd
Fd - design overstrength factor
qSd = F1 Fd
Dy De Du D
Fig. 7. Definition of force reduction factors (Fischinger and Fajfar, 1994)

The total reduction factor, used in design is:

qd = qµ ⋅ qS = qµ ⋅ qSd ⋅ qR (3)
The redundancy factor qR used herein represents the plastic redistribution capacity of the
structure (the αu/α1 ratio of Eurocode 8). It is well known that thin-walled steel sections do not
possess a significant post-elastic strength. Therefore, assuming local plastic mechanisms instead
of plastic hinges, the available redundancy of structures made by such a type of sections is based
on their hyperstatic character only.
For slender structures, using an equivalent static elastic-plastic analysis, q factor can be
evaluated with the following formula (Mazzolani and Piluso, 1996):
α
q = u (1 − β ' )α cr + β '  (4)
α1
where β ' = 1 − T ; β ' ≥ 0.5 ; T - is the fundamental period of structure; αcr - is the critical load
multiplier of gravitational loads, V (e.g. αcr = Vcr / V).

COLD-FORMED STEEL WALL STUD SHEAR PANELS

Light gauge steel buildings, both for residential and non-residential purposes, are usually
made by cold-formed sections framing, designed to carry gravitational loading and shear walls,
to resist horizontal forces from wind and earthquake actions. In what concerns the seismic
response, the performance of shear walls is crucial. Significant researches in the field and
important findings have been previously obtained in US (Serette and Ogunfunmi, 1996; Serette,
1998, Salenicovich, 2000), in Japan (Kawai et al., 1999), in Australia (Gad et al. 1999, 2000)
and Europe (de Matteis, 1998). They have been reviewed by Fulop and Dubina (2004a), too.

89
 On the following, the research programme carried out at P.U. Timisoara on the purpose to
characterise the seismic performance of wall-stud shear panels (Fulop and Dubina, 2004a,b) will
be summarised.

Experimental program at P.U. Timisoara

Six series of full-scale wall panels (3600mm x 2440mm), made by cold-formed wall-stud
skeleton and different cladding arrangements, commonly used for both residential buildings,
have been tested in the Laboratory of Steel Structures of the Politehnica University of Timisoara,
Romania.

Table 2. Description of wall specimens

Exterior Interior Testing Load Vel. No.
Ser. Open. Brac.
Cladding Cladding Method (cm/min) Test
O - - - - Monotonic 1 1
Corr. Sheet Monotonic 1 1
I - - -
LTP20/0.5 Cyclic 6;3 2
Corr. Sheet Gypsum Monotonic 1 1
II - -
LTP20/0.5 Board Cyclic 6;3 2
Monotonic 1 1
III - Yes - -
Cyclic 3 1
Corr. Sheet Monotonic 1 1
IV Door - -
LTP20/0.5 Cyclic 6 ;3 2
OSB Monotonic 1 1
- - 10 mm OSB -
I Cyclic 3 1
OSB Monotonic 1 1
Door - 10 mm OSB -
II Cyclic 3 1
Total Number of Specimens 15

Table 2 describe the specimens. Details about materials and fabrication technology of
specimens are presented in Fulop and Dubina (2004a). Cyclic loading was introduced according
to ECCS Recommendations (1985).
Figure 8 summarises the main results of cyclic tests and comparison with monotonic curves
of tested specimens of the six series.

Seismic performance
Test results have been used to calibrate a FEM model based of which an Incremental
Dynamic Analysis was applied in order to obtain the values of reduction factors qµ, qSd, and q,
respectively [19]. The main outputs of this research are summarised below.
Shear-resistance of wall panels is significant in terms of both rigidity and load bearing
capacity, and can be effective against lateral load. The hysteretic behaviour is characterized by
very significant pinching, and therefore reduced energy dissipation.
The seam fasteners represent the most sensitive part of the corrugated sheeting specimens;
damage is gradually increased in seam fasteners, until their failure causes the overall failure of
the panel. Much of the post-elastic deformation of the panel is in the region of seam fasteners,

90
therefore increasing the load capacity and ductility of the seams will improve the behaviour of
the panels.
Characteristic Curves - Series I Characteristic Curves - Series II
17000 17000

Force (lbf)
Force (lbf)

-6 6 -6 II-1 6
I-1
II-2
I-2
II-3
I-3
-17000 -17000
Displacement (in) Displacement (in)
Characteristic Curves - Series III Characteristic Curve - Series IV
17000 17000
Force (lbf)

Force (lbf)

-6 6 -6 IV-1 6
III-1
IV-2
III-2 IV-3
-17000 -17000
Displacement (in) Displacement (in)

Characteristic Curves - Series OSB I Characteristic Curves - Series OSB II

17000 17000
Force (lbf)

Force (lbf)

-6 6 -6 6
O. I-1 O. II-1
O. I-2 O. II-2
-17000 -17000
Displacement (in) Displacement (in)

Fig. 8. Experimental curves for all specimens

Figure 9 (a and b) show the failure modes for panels with corrugated sheeting and OSB,
respectively.

Fig. 9. Typical deformation pattern of corrugated sheet (Serie I, II) and OSB sheeted specimens
(Series OSB I)

91
 An additional testing program, on connection specimens has been carried out in order to find
design criteria for fasteners. Tests on sheeting-to-frame fasteners, sheeting-to-sheeting fasteners
(seam) and OSB-to-frame fasteners have been performed using two different loading velocity,
i.e.
v1 = 1mm/min (quasi-static)
v2 = 420mm/min (high seismic strain rate)

Figures 10 and 11 summarise these tests.

Influence of Loading Velocity Influence of Loading Velocity
- Sheeting to Frame Fasteners - Seam Fasteners
Force (N)

Force (N)
4000

2000 1000
De

Slip (mm)
Dr

Slip (mm)
De

Dr
0
0 5 10 15 20 25 0
0 10 20
Experimental (v1=1mm/min)
Experimental (v2=420mm/min) Experimental (v1=1mm/min)
Experimental (v2=420mm/min)
Characteristic curve (v1) Characteristic curve (v2)
Characteristic curve (v2) Characteristic curve (v1)

Fig. 10. Performance criteria at steel-to-steel connection level

Characteristic curve of OSB-to-Frame

Connections

2000
Force (N)

1000
De

Slip (mm)
0
0 5 10
Experimental (v1=1mm/min)
Characteristic curve (v1)

Fig. 11. Performance criteria for OSB-to-steel connections

In case of the wall panels with corrugated sheeting the main damage was concentrated in the
seam fasteners. It is important to establish an acceptable level of deformation at connection level
and, for different wall typologies, related this to the overall deformation of the wall panel. To

92
establish global performance criteria the following acceptable deformations in the seam
fasteners are suggested:
- If slip of the seams does not exceed the elastic limit (De, Figure 11), corresponding to
0.6Fmax of the seam connection, damage is limited and can be considered negligible. In
this case the integrity of the cladding is fully preserved, no repairs are required; it
corresponds to serviceability conditions.
- If slip is limited to the diameter of the screw (Dr=4.8mm, Figure 11) the cladding
requires some repair. There is damage, but not excessive and by minor interventions, like
replacing screws with larger diameter ones, the structure can be repaired. This could
correspond to immediate occupancy.
- In case of life safety criteria any kind of damage is acceptable, without endangering the
safety of occupants. This criterion, Du, corresponds to the attainment of the ultimate
force (Fult) and the starting of the downwards slope.

Based on these assumptions at connection level the following performance criteria, in terms
of panel drift, δ, are suggested for wall panels clad with corrugated sheet: (1) fully operational
(/<0.003); (2) partially operational (/<0.015); (3) safe but extensive repairs required (/<0.025).
The first performance level does not provide ductility, because shear panel work is elastic.
This could be the design criteria for frequent, but low intensity earthquakes. In case of rare but
severe earthquakes, the last two design criteria can be used and some ductility will be available.
In case of OSB-to-steel connections, which are characterised by a fragile behaviour, the
design has to be controlled by the elastic limit only (De – Figure 10). In such a case multiple
performance levels can not be applied.
It is very important to underline that in case of all tested specimens the wall stud system
proved a very good redundancy. Even when damages were very important no collapse occurred.
This is of real importance for buildings located in seismic areas. For corrugated sheet specimens,
and similarly for others, relevant performance design criteria can be suggested.
The effect of overstrength (design overstrength of connections and redundancy of skeleton)
was identified to be important in the post-elastic behaviour of panels and the main source of a
possible earthquake design force reduction. According to IDA procedure spectral accelerations
and displacements, instead of forces and displacements, have been used to calculate qµ and qS
factors (Figure 12):
q S = S yield / S el (5)
qµ = S ult / S yield (6)
where Sel, Syield and Sult are the spectral acceleration corresponding to the equivalent elastic
displacement (initiation of pseudo-inelastic behaviour), to the attainment of equivalent plastic
capacity and to the ultimate capacity, respectively.
The average resulting factor qS of 2.2-2.6 is harmonizing reasonably with the values 1.5-5
suggested by Gad et al. (1999). The possibility of design force reduction due to ductility and
energy dissipation qµ, seems to be more limited (e.g. qµ=1.4-1.6) probably due to low energy
dissipation capacity of the hysteretic loops. This value is also in agreement with the findings of
Gad et al. (1999). One can observe that resulting q-factor value is, at least, 3.0!

93
 D el D yield D ult

S ult
S yield

Sa
S el
0
Displacement

Fig. 12. IDA behaviour curve for shear wall panels

PITCHED ROOF PORTAL FRAMES OF CLASS 3 SECTIONS

Pitched roof portal frames, largely used for light industrial steel buildings, are usually made
by slender welded sections of class 3, even of class 4. Practical seismic design considers these
structures as non-dissipative and assumes q=1 to evaluate the seismic force. However, if all
three components of the force reduction factors are considered, i.e. ductility, overstrength, and
redundancy, one can demonstrate availability of q factors of values between 1.5 and 3, even
non-ductile sections are used in these structures.

Analysed frames
On this purpose, four frames were analysed (Dubina et al., 2002). They have the same span
and height, two different roof slopes, haunched rafters, and tapered or constant section columns.
Frame properties are shown in Table 3. In case 1C tapered column frames, cross-sections are,
along the members, of Class 1 to 3. Columns of 3C frames are of Class 1 cross-sections. Rafter
cross-sections are of Class 1 to 3 for all frames. Steel S235 was used.
Both 2D and 3D elastic-plastic analyses have been performed. For 2D analysis, PEP-Micro
computer program was used, while 3D analysis was performed with ANSYS. PEP-Micro
operates with linear plastic bar elements, while in ANSYS, plastic large strain SHELL43
elements were used. In both analyses was introduced the material behaviour by a bilinear
elastic-perfectly plastic model. In case of 3D analysis, lateral restrains by purlins and side rails
were considered. The lateral restraints are of 3 types, as shown in Figure 13, i.e. type 1 – no
lateral restraints, type 2 – restraining of lateral displacement only, and type 3 – restraining of
both lateral displacement and rotation.
Rafter-to-column connections are nominally rigid. Vertical loads from permanent and snow
actions were introduced at the purlin location. Horizontal load acting at the top of the left side
column was taken 12% from vertical load. To evaluate the dissipative capacity of the structure a
non-linear pushover analysis was performed. During pushover analysis, the horizontal load only,
was incremented. Initial sway and flexural imperfections were considered in both models.

94
 Table 3. Properties of analysed frames

Main tw

h
Frame dimensions
Geometry [mm] tf Column base
type
b

Rafter Column

1C-1 pin constant

L=12.0 m
h=270 pinned
H H=4.8 m
b=135
=5.7°
variable
1C-1 sem tf=10
h=240...600 semi-rigid
L tw=5
b=180 Sj=2785 kNm/rad
tf=12
1C-2 pin variable
tw=8
L=12.0 m
h=270...600 pinned
H
H=4.8 m
=11.3°
b=135
1C-2 sem tf=10
tw=6 semi-rigid
L
Sj=2785 kNm/rad

3C-1 sem semi-rigid

constant
L=12.0 m
H=4.8 m h=270 Sj=7710 kNm/rad
H
=5.7°
b=135
constant
3C-1 rig L tf=10
h=400 rigid
tw=5
b=180 Sj=30715 kNm/rad
tf=12
variable
tw=8
3C-2 sem h=270...600 semi-rigid
L=12.0 m
b=135 Sj=7710 kNm/rad
H
H=4.8 m
=11.3°
tf=10
tw=6
3C-2 rig L
rigid
Sj=30715 kNm/rad

(a) type 1 (b) type 2 (c) type 3

Fig. 13. Types of lateral restraints

95
 Seismic performance and reduction factor values
The seismic global performance of the frame can be evaluated by using an equivalent
nonlinear inelastic static analysis (pushover analysis). The seismic lateral force was evaluated
following Eurocode 8 provisions. In order to evaluate q-factor values eqn. (4) was used. The
results are summarized in Table 4.

Table 4. Seismic performance, q-factor.

Frame Lateral Pcr Py Pu T q
Failure mechanism
type restraints [kN] [kN] [kN] [sec] factor
type 1 25.5 *
1C-1 pin type 2 299 77 85 0.60 1.39
type 3 693 2.48
type 1 16.8 *
1C-2 pin type 2 346 77 85 0.62 1.52
type 3 695 2.49
type 1 26.1 *
1C-1 sem type 2 308 87 100 0.51 1.46
type 3 728 2.67
type 1 17.9 *
1C-2 sem type 2 361 91 105 0.52 1.62
type 3 731 2.68
type 1 21.8 *
3C-1 sem type 2 321 127 148 0.48 1.50
type 3 774 2.76
type 1 14.4 *
3C-2 sem type 2 395 127 148 0.46 1.68
type 3 872 2.96
type 1 21.5 *
3C-1 rig type 2 326 154 189 0.40 1.53
type 3 799 2.69
type 1 14.8 *
3C-2 rig type 2 407 154 189 0.39 1.72
type 3 906 2.91

-pinned column base -rigid column base

-semi rigid column base -plastic zones

Capacity spectrum method has been used to evaluate the drift demand of the frames (Dubina
et al., 2002). It is very important to underline that drift demand doesn’t overcome the plastic
rotation capacity in relevant sections. The ultimate rotation capacity of both columns and rafters
of the analyzed frames, in the sections where plastic zones occurred, was evaluated with
DUCTROT-M program, on the base of local plastic mechanism theory (Gioncu and Mazzolani,
2002), and the corresponding values are 0.042rad and 0.053rad respectively.

96
 SEISMIC PERFORMANCE OF COLD-FORMED PORTAL FRAMES

Extended research on this subject was performed by Calderoni and co-authors (1994). The
authors started their research from the experimental findings of Ono and Suzuki (1986) who
proved by tests significant post-elastic strength and ductility of some cold-formed steel frames.
Figure 12 shows the frame tested by the Japanese researchers, and the corresponding numerical
model proposed and calibrated by Calderoni et al. (1994) in order to study the behaviour of such
a type of structures. By using this kind of cyclic load-displacement law, a lot of numerical step-
by-step dynamic analyses were performed with reference to some built-up channel section portal
frames. Geometrical and mechanical properties of frames were selected to provide monotonic F-
D curves characterised by elastic stiffness, slope of the softening branch, and residual strength.
The dynamic response of the analysed frames (performed by taking into account also the
geometrical degradation due to second order effects) was obtained by using, as load conditions,
thirty different real accelerograms, recorded during some Italian earthquakes. They were
selected in such way that the corresponding average elastic response spectrum (50% of
probability to be exceeded) to be similar to that provided by EN 1998-1 for soil type A and PGA
equal to 0.15g.
The results of this wide numerical investigation (about 1000 analyses) showed that the
seismic behaviour of thin-walled portal frames was not so different on respect of the
corresponding ideal elasto-plastic structure, provided that the slope of softening branch of the
monotonic F-D curve is reasonable (e.g. 30o around). In these cases, it seemed that a q factor
greater than 1, varying in the range 2 to 3, could be used in low-seismicity zones, if the available
ductility exhibited by the frame is about equal to 3 (Calderoni et al., 1994). Nevertheless it
clearly appeared that the shape of the F-D curve, e.g. the lateral elastic stiffness, affects the
structural response and consequently the judgement on the possibility of using the light gauge
structures in seismic zones. If the traditional capacity method is applied to design these
structures, eqn. (4) is recommended to calculate the q factor. However, the lowest limit of q
factor, suggested by the authors of this study is 1.8 (the q value equal to 3, for the frame of
Figure 14, is really to big to be used in practice, since the sections are unusually thick). This
value approaches to the previous one proposed by Moldovan et al. (1999), and both are
practically of the same order of magnitude with the value given in EN 1998-1 for non-
dissipative structures, i.e. q=1.5-2.0.

F
15 kN 15 kN
2C 200x75x20x3.2 Flim
Fel (Flim-F)
F

Del IxDel
Del Fres
Dr Dlim
170

D
2C 200x50x3.2 RxFel

(1-R)xFel
R=1-(Flim-F)/Flim
Flim
100 100 100 I=1+(1-R)

Fig. 14. Analysed frame and analytical cyclic behaviour

97
 Test results obtained at P.U. Timisoara (Dubina et al., 2004) provided evidence for the very
low ductility of typical joints used in pitched roof portal frames. Figures 15 and 16 show the
most ductile and strongest joint typology from those which have been tested and the
corresponding monotonic and cyclic curves. Looking to curves, clearly results that in order to
take benefit from some reduction q-factor, overstrength of joints is absolutely necessary because
their plastic rotation capacity is quite poor (θpl = 0.007÷0.013 rad).

Fig. 15. Joint specimens

200 200
150 150
100 100
50 50
M, kNm

M, kNm

0 0
-0.100 -0.050 -500.000 0.050 0.100 -0.100 -0.050 -500.000 0.050 0.100

-100 RIFB-C1-env+ -100 KIP-FB-C1-env+

RIFB-C1-env- KIP-FB-C1-env-
-150 -150
RIFB-C1 KIP-FB-C
-200 -200
φ TOT, rad φ TOT, rad

Fig. 16. Joint moment-rotation curves

Also it is important to emphasise that a significant improvement of seismic performances of

cold-formed steel framing can be obtained if the co-operation with structural envelopes is
considered (e.g. interaction with shear panels, both for roof and cladding).

98
 CONCLUDING REMARKS

Light gauge steel structures, made by class 3 or class 4 sections, fabricated by cold-forming
or thin plate welding can be effectively used in seismic resistant structures mainly due to their
reduced weight/strength ratios.
Traditional capacity design based on equivalent elastic static analysis with reduction factors q
of values 1<q≤2 can be used provided the overstrength of joints and structural redundancy are
available.
Seismic response of light-gauge steel framing can be significantly improved if shear walls are
used to resist horizontal forces.
Both experimental and numerical results sustain to classify light-gauge steel structures as
low-dissipative. Practically an elastic design has to be conducted, but the seismic force can be
evaluated applying a reduction factor q of 1.5-2, corresponding to “L” ductility class, as
specified in EN 1998-1. This is in fact a “pseudo ductility” because it is mostly based on the
overstrength and structural redundancy than on the post-elastic strength reserve of members an
connections.

REFERENCES

AISI–American Iron and Steel Institute. North American Specification for the Design of Cold-
Formed Steel Structural Members with Commentary, Washington, D.C., 2001.
AS/NZS 4600–Australian Standards/New Zealand Standards. Cold-formed Steel Structures,
Sydney, Australia, 1996.
Calderoni, B., De Martino, A., Landolfo, R., Ghersi, A. On the seismic resistance of light gauge
steel frames. Behaviour of Steel Structures in Seismic Areas - STESSA'94. Ed. by F.M.
Mazzolani and V. Gioncu, E&FN Spon, London, 1994.
De Martino, A., Ghersi, A., Landolfo, R., Mazzolani, F.M. Calibration of a bending model for
cold-formed section. XIth International Specialty Conference on Cold-Formed Steel
Structures, St. Louis, 1992, 503-511.
Dubina, D., Cristutiu, M., Ungureanu, V., Nagy, Zs. Stability and ductility performances
of light steel industrial building portal frames. Proceedings of the 3rd European
Conference on Steel Structures - Eurosteel 2002, 19-20 September 2002, Coimbra,
Portugal, Vol. I, 635-644.
Dubina, D., Stratan, A., Ciutina, A., Fulop, L., Nagy, Zs. Monotonic and cyclic performance of
joints of cold formed steel portal frames. Fourth International Conference on Thin-Walled
Structures - ICTWS 2004, 22-24 June 2004, Loughborough Leicestershire, UK.
ECCS - European Convention for Constructional Steelwork. Recommended Testing Procedures
for Assessing the Behaviour of Structural Elements under Cyclic Loads. Technical
Committee 1, TWG 1.3 – Seismic Design, No.45, 1985.
EN 1993-1-3: Eurocode 3. Design of Steel Structures, Part 1.3: General Rules, Supplementary
Rules for Cold-Formed Thin-Gauge Members and Sheeting, CEN/TC 250/SC3 – European
Committee for Standardisation, Brussels, 2001.
EN 1998-1: Eurocode 8. Design of structures for earthquake resistance. Part 1: General rules,
seismic actions and rules for buildings. Draft no. 6, January. CEN - European Committee for
Standardization, 2003.

99
Fischinger, M., Fajfar, P., Seismic force reduction factors. Earthquake Engineering. A.
Rutenberg (editor), 1994, Balkema, 279-296.
Fülöp, L.A., Dubina, D. Performance of Wall-Stud Cold-Formed Shear Panels under
Monotonic and Cyclic Loading, Part I: Experimental Research, Thin Walled Structures, Vol.
42, No 2, 2004a, 321-338.
Fülöp, L.A., Dubina, D. Performance of Wall-Stud Cold-Formed Shear Panels under
Monotonic and Cyclic Loading, Part II: Numerical Modelling and Performance Analysis,
Thin Walled Structures, Vol. 42, No 2, 2004b, 339-343.
Gad, E.F., Chandler, A.M., Duffield, C.F., Hutchinson, G. Earthquake Ductility and
Overstrength in Residential Structures. Structural Engineering and Mechanics, 1999, 8:4,
361-382.
Gad, E.F., Duffield, C.F. Lateral Behaviour of Light Framed Walls in Residential Structures,
12th World Conference on Earthquake Engineering, 2000.
Gianfranco de Matteis. The Effect of Cladding in Steel Buildings under Seismic Actions, PhD
Thesis, Universita degli Studi di Napoli Federico II, 1998.
Gioncu, V., Mazzolani, F.M. Ductility of Seismic Resistant Steel Structures. Spon Press, London
and New York, 2002.
Kawai, Y., Kanno, R., Uno, N., Sakumoto. Y. Seismic resistance and design of steel framed
houses. Nippon Steel Technical Report, No. 79, 1999.
Lindner, J., Gläβer, K. Fatigue behaviour of thin-walled cold-formed steel profiles. Proceedings
of International Colloquium Recent Advances and New Trends in structural Design, 7-8 May
2004, Timisoara, Romania, 71-82.
Mazzolani, F.M., Piluso, V. Theory and design of Seismic Resistant Steel Frames. Chapmen &
Hall, 1996.
Moldovan, A., Petcu, D., Gioncu, V. Ductility of thin-walled members. Stability and Ductility of
Steel Structures SDSS'99. Ed. by D. Dubina and M. Ivanyi, Elsevier Science, 1999, 299-307.
Murray, N.W., Khoo, P.S. Some basic mechanism in the local buckling of thin-walled steel
structures. International Journal of Mechanical Sciences, Vol. 23, No. 12, 1981, 703-713.
Ono, T., Suzuki, T. Inelastic behaviour and earthquake - resistance design method for thin-
walled metal structures. Proceedings on IABSE Coll. on Thin-Walled Metal Structures in
Building, Stockholm, 1986, 115-122.
Salenicovich, A.J. et al. Racking Performance of Long Steel-Frame Shear Walls. Fifteenth Int.
Speciality Conference on Cold-Formed Steel Structures, St. Louis, Missouri, October 19-20,
2000, 471-480.
Serette, R.L., Ogunfunmi, K. Shear Resistance of Gypsum-Sheeted Light-gauge Steel Stud Walls.
Journal of Structural Engineering, ASCE, 122:4, 1996.
Serrette, R.L. Seismic Design of Light Gauge Steel Structures: A discussion. Fourteenth Int.
Speciality Conference on Cold-Formed Steel Structures, St. Louis, Missouri, October 15-16,
1998.
Ungureanu, V., Dubina, D. Post-elastic strength and ductility of cold-formed steel sections.
Fourth International Conference on Thin-Walled Structures - ICTWS 2004, 22-24 June 2004,
Loughborough Leicestershire, UK.

100