7/25/2019 1-s2.0-S026382311500021X-main
1/12
Experimental investigations of cold-formed steel beams of corrugatedweb and built-up section for anges
Dan Dubina a,b, Viorel Ungureanu a,b,n, Lucian Glia c
a Department of Steel Structures and Structural Mechanics, Politehnica University of Timisoara, Timisoara, Romaniab Laboratory of Steel Structures, Romanian AcademyTimisoara Branch, Timisoara, Romaniac Technical University of Cluj-Napoca, Cluj-Napoca, Romania
a r t i c l e i n f o
Article history:Received 13 January 2015
Received in revised form
18 January 2015
Accepted 19 January 2015Available online 11 February 2015
Keywords:
Corrugated web beam
Cold-formed steel solution
Discrete fasteners
Self-drilling screws
Experimental investigation
a b s t r a c t
The steel beams of corrugated web represent a relatively new structural system which emerged in thepast two decades. The thin corrugated web affords a signicant weight reduction of these beams,
compared with hot-rolled or welded ones. In the solutions existing on the market, the anges are made
ofat plates welded to the sinusoidal web sheet, requiring a specic welding technology. A new solution
is proposed in this paper, in which the beam is composed by a web of trapezoidal cold-formed steel
sheet and anges of built-up cold-formed steel members (e.g. back-to-back lipped channel sections or
angles with turn lips). The connections betweenanges and web can be done by self-drilling screws or
by spot welding. The rst part of the study, summarised in this paper, is devoted to the evaluation and
validation of technical solution, including experimental investigations, carried out at the CEMSIG
Research Centre of the Politehnica University of Timisoara ( http://www.ct.upt.ro/en/centre/cemsig). In
a subsequent paper, numerical investigations aiming to optimise the solution and estimate its technical
limits for applications will be presented.
& 2015 Elsevier Ltd. All rights reserved.
1. Introduction
Corrugated web girders represent a relatively new structural
system emerged in the past two decades especially in Germany
and Austria, used in a large number of applications. Increased
interest of this solution was observed for the main frames of
single-storey steel buildings and in steel bridges. In 1988 the rst
machines for the production of SIN-beams were developed by
Zeman[1]. These semi-automatic machines of the rst generation
were able to produce SIN-beams with parallel anges and web
thicknesses of 2.0 mm, 2.5 mm or 3.0 mm. The machines of latest
generation are able to produce SIN-beams by a fully automated
process. A more variable design of cross-sections, a variety of web
thickness, lower beam heights and smaller ange dimensionsbecame possible. Furthermore tapered beams and machine-made
web openings can be produced.
In Japan has been developed a roll forming process to produce
corrugated web I-beams and partially corrugated webs which
were used in mobile-modular home construction[2]. In the United
States, beams with corrugated webs are more and more widely
used and many bridges and large span buildings are built using
corrugated web I-beam.
The main benets of this type of beams are that the corrugated
webs increase the beams stability against buckling, which may
result in a very economical design via the reduction of web
stiffeners. Due to improvements of the automatic fabrication
process corrugated webs up to 6 mm thickness became possible.
Furthermore, the use of thinner webs results in lower material
cost, with an estimated cost savings of 1030% in comparison with
conventional fabricated sections and more than 30% compared
with standard hot-rolled beams. The buckling resistance of used
sinusoidal corrugated sheet used for webs is comparable with
plane webs of 12 mm thickness or more.
In the existing solutions the anges are at plates, welded tothe sinusoidal web sheet, requiring a specic welding technology
and highly automated manufacturing process. The anges mainly
provide exural strength to the beam with low contribution from
the corrugated web, which provides the shear capacity. Failure of
the web occurs by steel yielding or web buckling. Lateral-torsional
buckling of the girder and local ange buckling, separately or in
interaction, represents other possible failure modes.
The paper presents a new technological solution of such a
system, composed by webs made of trapezoidal cold-formed steel
sheets and anges of built-up cold-formed steel members (e.g.
back-to-back lipped channels, back-to-back angles with turn lips
Contents lists available at ScienceDirect
jo ur nal ho me pa ge: www.elsevier.com/locate/tws
Thin-Walled Structures
http://dx.doi.org/10.1016/j.tws.2015.01.018
0263-8231/&2015 Elsevier Ltd. All rights reserved.
n Corresponding author at: Department of Steel Structures and Structural Mechanics,
Politehnica University of Timisoara, Timisoara, Romania.
E-mail address:[email protected](V. Ungureanu).
Thin-Walled Structures 90 (2015) 159170
http://www.ct.upt.ro/en/centre/cemsighttp://www.sciencedirect.com/science/journal/02638231http://www.elsevier.com/locate/twshttp://dx.doi.org/10.1016/j.tws.2015.01.018mailto:[email protected]://dx.doi.org/10.1016/j.tws.2015.01.018http://dx.doi.org/10.1016/j.tws.2015.01.018http://dx.doi.org/10.1016/j.tws.2015.01.018http://dx.doi.org/10.1016/j.tws.2015.01.018mailto:[email protected]://crossmark.crossref.org/dialog/?doi=10.1016/j.tws.2015.01.018&domain=pdfhttp://crossmark.crossref.org/dialog/?doi=10.1016/j.tws.2015.01.018&domain=pdfhttp://crossmark.crossref.org/dialog/?doi=10.1016/j.tws.2015.01.018&domain=pdfhttp://dx.doi.org/10.1016/j.tws.2015.01.018http://dx.doi.org/10.1016/j.tws.2015.01.018http://dx.doi.org/10.1016/j.tws.2015.01.018http://www.elsevier.com/locate/twshttp://www.sciencedirect.com/science/journal/02638231http://www.ct.upt.ro/en/centre/cemsig7/25/2019 1-s2.0-S026382311500021X-main
2/12
or hat omega). The connections between anges and web are
made with self-drilling screws or by spot welding. It is easy to be
observed that the new solution, as a whole, is 100% composed by
cold-formed steel elements, avoiding the combination of two
types of products, i.e. cold-formed for webs and hot-rolled for
anges. High protection to corrosion due to the fact that all
components are galvanized is a major advantage. On the other
hand, if the fabrication might adopt an automotive fastening
technology such as spot welding, the production of some standar-
dised beam series can be highly automated.
2. Literature review
There are several types of built-up cold-formed steel beams onthe market, prepared for industrialised fabrication, for which bolts,
screws or spot welds are used for anges-to-web connection.
Zhao [3] at Queensland University of Technology initiated a
research program to investigate the structural behaviour and
design of hollow ange members in compression. The study was
focus on members with rectangular hollow anges, where the
sections are formed from a single steel strip, with various
manufacturing methods such as spot welding, self-pierced riveting
and screw fastening foranges-to-web connections. He found that
the type of fastening and spacing does not affect the member
compression capacity signicantly. Wanniarachchi [4] extended
the work of Zhao[3]and developed a new cold-formed steel beam
with two rectangular hollow anges, rigid in torsion, and a slender
web, cross-section assembled using intermittent screw fastening.He has found that intermittent screw fastening method appears to
be structurally adequate and minimises the fabrication cost.
Landolfo et al. [5] evaluated the applicability of built-up cold-
formed steel beams assembled by laser welding and assessed the load
bearing capacity of the assembled beams. The I-section with hollow
anges is fabricated from two back-to-back special C-proles. The two
proles are joined with connections which are located on the web
and on the anges. Two reinforcing plates are placed inside the top
and bottom hollow anges of the I-section, providing an additional
connection system between the two C-proles.
BEN-VAUTIER S.P.A.[6], patented a modular H-beam comprises
one or more modules, each formed of two half-structural parts of
two pieces of structural steel, forming each a thin sheet, compris-
ing a central part or core, and lateral half-anges. The half-anges
form gaps, inside which plates are introduced in order to
strengthen the anges of the beam, which constitute the regions
more subjected to bending stresses.
Table 1
Types of specimens.
CWB-1 Standard solution:ange-to-web connection in every corrugations and uniformly distributed seam fasteners (see Fig. 1)
CWB-2 Standard solutionsupplementary lipped channel sections under the load application points (see Fig. 2)
CWB-3 Optimized solution by adapting the ange-to-web connections according to the distribution of shear stresses (connections at each second corrugations where
the shear force decreases) (seeFig. 3)
CWB-4 Standard solution by eliminating shear panels and doubling of corrugated webs in the zones with high shear forces (seeFig. 4)
CWB-5 Optimized solution by adapting both the ange-to-web connections and seam fasteners to the distribution of shear stresses (see Fig. 5)
Fig. 2. Conguration of the specimens CWB-2.
Fig. 3. Conguration of the specimens CWB-3.
Fig. 4. Conguration of the specimens CWB-4.
Fig. 1. Conguration of the specimens CWB-1.
D. Dubina et al. / Thin-Walled Structures 90 (2015) 159170160
7/25/2019 1-s2.0-S026382311500021X-main
3/12
Signicant work on girders with corrugated web was devoted to
study the shear capacity. A summary of the research and development
in beams with corrugated webs was reported by Elgaaly and Dagher
[7]. Smith [8] performed four tests on two girders with corrugated
webs, which were welded to the anges using intermittent welding.
He found that the connection between the ange and the web is
critical for the shear strength as the weld used in the test was
subjected to high strength and web was easily ruptured at this point
before it reached its buckling strength. He suggested that intermittent
welding of the corrugated webs to the ange is not advisable.
Hamilton[9]performed 42 tests on 21 beams, which used four
different corrugation congurations and two thicknesses. Unlike
Smiths[8] specimens, the webs were continuously welded to the
anges from one side. It was found that the failure of all specimens
was initiated by local buckling of one of the corrugation folds.
Another conclusion was that dense corrugation proles are more
likely to fail in global shear buckling. Elgaaly et al.[10]veried the
test results done by Smith [8] and Hamilton [9] using nonlinear
FEM and found that the results of the nite element analysis were
very close to the test results.
Luo and Edlund[11]used non-linear nite element analysis to
perform a geometrical parametric study and compared the numer-
ical results with existing empirical and analytical formulae. Within
the parametric range studied, they have found that the ultimate
shear capacity increases proportionally with the girder depth and
does not seem to be dependent on the ratio of girder length over
girder depth, while the post-buckling shear capacity not only
increases with the girder depth, but also appears to be dependent
on the ratio of girder length over girder depth. They have also
found that the corrugation depth did not seem to have much effect
on the ultimate shear capacity but affected the degree of the
localization of the buckling mode.A lot of work has been done on the bending behaviour of steel
girders with corrugated web. It was observed that the contribution
of the web to the ultimate moment capacity of a beam with
corrugated web is negligible, and the ultimate moment capacity
will be based on the ange yield stress.
Elgaaly et al.[12]have performed a series of experimental and
analytical studies. They have experimentally tested six specimens
that had corrugated webs in the centre panel and at panels
adjacent to the support. They cross braced theat panels to ensure
that the failure would occur in the centre panel. All the specimens
failed due to ange yielding followed by vertical buckling of the
compression ange into the web. They found that the web did not
contribute much to the bending capacity of the beam and its
contribution could be neglected.Chan et al. [13] studied the effect of web corrugation on the
bending capacity of the beam using FEM. Beams with plan web,
horizontally corrugated web and vertically corrugated web were
studied. They found that the vertically corrugated web provides a
stronger support against the ange buckling than those with hor-
izontally corrugated and at webs. Also, the corrugation radius was
investigated and found that larger corrugation radius could sustain
higher bending moment. It was also found that, the vertically
corrugated beam had a 10.6% reduction in weight when compared
with the beam with at web.
Johnson and Cafolla[14]have studied the effect of the vertically
corrugated webs on the local buckling of the compressive ange and
the exural behaviour of beams with corrugated webs via numerically
and in experimental tests. They found that, depending on the shape of
Fig. 5. Conguration of the specimens CWB-5.
Fig. 6. Experimental arrangement.
Fig. 7. Samples cut from at regions and corners.
Table 2
Yield and ultimate strengths.
Type fyM[N/mm2] fuM[N/mm
2] sfy sfu fyk [N/mm2] fuk [N/mm
2]
BM-CF 438.74 517.06 5.69 2.46 425.48 511.33
BM-CW 441.65 521.86 25.53 3.48 382.16 513.05
CM-CF 521.64 585.07 9.89 9.69 498.60 562.49
BM-CW 349.41 394.75 12.67 10.89 319.89 369.37
BM-SP 358.42 419.59 4.09 2.48 348.90 413.81
D. Dubina et al. / Thin-Walled Structures 90 (2015) 159170 161
7/25/2019 1-s2.0-S026382311500021X-main
4/12
the corrugations, the slenderness should be based on the mean or on
the maximum outstand, which means the distances from the hor-
izontal fold to the edges of the ange. Also, they found that the
contribution of the web to exural capacity was small.
Lindner[15]studied by experimental tests the lateral-torsional
behaviour of steel girders with corrugated webs and found thatthe torsional section constant ITfor a beam with corrugated web
doesnt differ from that of a beam with at web, but the warping
section constant Iwis different.
The effect of the corrugation proles of the web on the lateral-
torsional buckling strength of I-girders was also studied [16,17].
Pasternak et al. [18,19]presented a new proposal for Annex D of
EN 1993-1-5:2006[20].
Moon et al. [21] investigated the lateral-torsional buckling
strength of an I-girder with corrugated steel webs under linear
moment gradient by using nite element analysis. It was found
that the buckling behaviour of the I-girder with corrugated steel
webs differed depending on the number of periods of the
corrugation. Simple equation for the moment gradient correction
factor for these types of beams was suggested.
In what concerns girders with trapezoidal corrugated webs
under patch loading, Leiva-Aravena and Edlund[22]performed six
tests, in which three parameters were considered, i.e. the load
patch width, the load path location and the web thickness. From
the comparison between the test and nite element analysis
results, it can be concluded that the FE model is able of depicting
the behaviour of girders with corrugated webs subjected to in
plane compressive patch loading and of calculating the failure load
to a good degree of accuracy.
Elgaaly and Seshadri [23] performed ve tests on four different
corrugation proles. Two distinct modes of failure were observed: web
crippling and web yielding. They also studied, using FEM, the interac-
tion between partial compressive edge loading and bending or shear.
Luo and Edlund [24] performed nonlinearnite element analysis to
study the effect of four factors that inuence the buckling strength of
the beams, i.e.: (1) strain hardening model; (2) corner effect; (3) initial
imperfections; (4) loading position. They used elastic-perfectly plasticand RambergOsgoods models and found that with a Ramberg
Osgood strain-hardening model for webs, the ultimate strength of the
girder is about 812% higher than using an elastic-perfectly plastic
model. Also, the effect of the corners due to cold-forming does not
have any signicant effect on the ultimate strength.
Nguyen et al.[25]investigate the moment modication factors of
I-girder with trapezoidal web corrugations under moment gradient
and various end restraint conditions and proposed closed-form
expressions for the moment modication factors.
Tahir et al. [26]investigated the performance of the strength, the
rotational stiffness, and the ductility of the composite and non-
composite connection using trapezoidal web proled steel sections.
Eight full scales testing of beam-to-column connections comprised of
four specimens for composite and four for non-composite connectionwith different geometrical congurations have been carried out. The
tests results showed good agreement between the experimental and
the predicted values. The test also concluded that composite connec-
tions have higher moment resistance, higher stiffness, and less ductile
compared with the non-composite connections.
Kvesdi et al. [27] investigated of the stress distribution in the
ange of the girders with corrugated webs. During the experimental
tests the stress distributions on different locations (i.e. anges and
web) were measured as a basis for parametric analyses.
Probably, the rst built-up steel girder using corrugated sheets as
web elements and cold-formed sections for anges is the Macomber
Panlweb girder, patented in 1967. The Macomber Panlweb girder
consists of 1.9 mm to 3.8 mm thickness of the corrugated web for
depths of 0.51 to 1.02 m [28,29]. This solution applies shallow
Fig. 8. Failure modes of (a) coupons cut from BM-CW, BM-SP, BM-CW and BM-SP; (b) coupons cut from CM-CF.
Table 3
Types of tested connections.
Name t1[mm] t2 [mm] No. of tests dnom[mm]
T1-1.4 0.7 0.7 6 4.8
T2-1.7 1.0 0.7 5 4.8
T3-3.7 2.01.0 0.7 6 6.3
T4-9.0 1.0 8.0 5 5.5
T5-11.0 2.01.0 8.0 5 M12
T6-2.7 2.0 0.7 10 6.3
T
3
T4-9.0
1.0
T
T2-1.7
3.7
T1-1.4
22.7
Fig. 9. Location of tested connections.
D. Dubina et al. / Thin-Walled Structures 90 (2015) 159170162
7/25/2019 1-s2.0-S026382311500021X-main
5/12
VV-section for anges, which make difcult enough the installing of
ange-to-web fasteners. On these system the rst research activities
were carried out in the 70s by Harrison[30].
The rst attempt of the authors of this paper related to this
type of beams 100% composed by cold-formed steel elements was
a numerical study [31] in order to prove the efciency of such
solution against cold-formed steel trusses.
A similar solution has been proposed and analysed in the frame
of PRECASTEEL project [32], but using blind rivets as seam
fasteners for the corrugated web and bolts for web-to-ange
connections. For anges, back-to-back lipped channel or two typesof hat-sections have been used. Deep corrugation web sheeting of
longitudinal intermediate stiffeners have been applied in this
solution. However, looking to the test results, one observes the
sensitivity to distortion of corrugation still remain high.
Another very important aspect related to the cold-formed steel is
the connecting technique. Briskham et al. [33]performed a compara-
tive study on of self-pierce riveting, resistance spot welding and spot
friction joining for aluminium automotive sheet. Quantitative compar-
isons have been made on the basis of tensile strength (shear and peel),
process time, equipment price and running cost. The results identied
resistance spot welding as a more economically favourable option
than self-pierce riveting or spot friction joining for the task of
producing the majority of the joints. The analysis indicates that it is
the ongoing cost of the rivets that makes self-pierce riveting the most
expensive process. For resistance spot welding, the largest cost factors
identied were energy consumption and frequency of electrode
replacement. Even the material is aluminium, similar conclusions
can be drawn for steel too.
Guenfoud et al. [34] tested welded specimens fabricated
through one, two or four layers of steel sheets with thicknesses
ranging from 0.76 mm to 1.52 mm. A total of 72 tension tests and
107 shear tests were completed. The idea was the initiation of a
research program on the shear resistance and tension resistance of
multi-layer arc spot welds. They found that the type of electrode,
high current setting and proper welding technique affect thequality of arc-spot welds in multi-layer connections, and a lower
limit for the net effective weld diameter was proposed.
Snow [35] conducted a similar research in order to establish a
relationship between arc spot weld shear strength and the arc time
used while forming the weld. In this case the arc times were broken
down into three separate categories. The rst category consisted of
full-time welds, the second 2/3-time welds, and the third 1/3-time
welds. Testing was performed on steel gauge sheets of 0.85 mm,
1 mm, 1.3 mm and 1.6 mm. Each gauge material was tested in single-,
double- and four-layer congurations. Two types of diameter arc spot
welds were tested. Comparisons were made between shear strength
and weld geometry, including average diameter, effective diameter
and penetration. The research has proven that arc time has a
tremendous inuence on arc spot weld shear strength.
Fig. 10. Forcedisplacement curves for the tested connections.
D. Dubina et al. / Thin-Walled Structures 90 (2015) 159170 163
7/25/2019 1-s2.0-S026382311500021X-main
6/12
As a nal remark, a beam with corrugated web behaves similarly
to a lattice girder, in which the bending moments and applied forces
are transferred via anges only, while the transverse forces are
transferred through the diagonals and verticals of the lattice girder,
in this case the corrugated web. The dimensioning of corrugated web
beams is ruled by Annex D of the EN 1993-1-5:2006[20], together
with specic aspects of EN 1993-1-1:2006 [36] and EN 1993-1-
3:2006 [37]. At the end, on the purpose of nding an analytical
approach for designing such beams with corrugated web intermit-
tently connected to the anges, the procedure used for calculation
sheathing acting as a diaphragm could be adapted[38]. The experi-
ence, in this case, has shown the most contributing factors, both to
Fig. 11. T3-3.7 connection: (a) at 3 mm displacement corresponding to SLS; (b) at 6 mm; (c) at 12 mm; (d) at failure.
Fig. 12. Deformed shape of the beam end shear panel.
Fig. 13. Deformed shape of CWB-1 beam at failure.
D. Dubina et al. / Thin-Walled Structures 90 (2015) 159170164
7/25/2019 1-s2.0-S026382311500021X-main
7/12
strength and stiffness are the distortion of sheeting corrugations and
seem fasteners[39].
3. Technical solution: Specimens, material and connection
properties
3.1. Description of technical solution
The new technological solution proposed by the authors is
composed by webs made of trapezoidal cold-formed steel sheets
and anges of built-up cold-formed steel members (e.g. back-to-
back lipped channels, back-to-back angles with turn lips or hat
omega). As connecting technique self-drilling screws or spot weld-
ing both for the connections betweenanges and web and as seam
fasteners to ensure the continuity of the web can be used.
Fig. 15. Deformed shape of the beam end shear panel and distortion of the web corrugation.
Fig. 16. Evolution of shear yield lines for beam end shear panel: (a) at 44 mm; (b) at failure.
0
50
100
150
200
250
0 10 20 30 40 50 60
Force
[kN]
Displacement [mm]
CWB - 1
buckling of the shear panel (BSP)
distortions of the corrugated web (DCW)
collapseFig. 12(a)
Fig. 12(b)
Fig. 12(c)
Fig. 14. Loaddisplacement curve for CWB-1 beam.
Fig. 17. Deformed shape of CWB-2 beam at failure.
D. Dubina et al. / Thin-Walled Structures 90 (2015) 159170 165
7/25/2019 1-s2.0-S026382311500021X-main
8/12
In this paper only the solution considering back-to-back lipped
channels for anges and self-drilling screws is investigated. Some
other particularities of this solution compared to the ones pre-
sented in[28]and [32]are:
Small corrugation depths and the thicknesses for corrugated
web in order to reduce the distortion of the corrugation; Reinforcing shear panels where the shear force is maximum;
Trapezoidal or parallel anges sloped beams.
3.2. Specimens and test procedure
The experimental program was carried out at the CEMSIG
Research Centre (http://www.ct.upt.ro/en/centre/cemsig) of the
Politehnica University of Timisoara. Five beams with corrugated
webs with a span of 5157 mm and a height of 600 mm have been
tested, as shown inTable 1, considering different arrangements for
self-drilling screws and shear panels [40,41].
Fig. 1 presents the components of the CWB-1 beam with
corrugated web, the so called standard solution, i.e.:
back-to-back lipped channel sections for anges2C120/2.0
(grade S350GDZ); corrugated web with the corrugation depth of 43 mm and the
thickness of 0.7 mm
A45/0.7 (grade S320GDZ); reinforcing shear panelssupplementary plates of 1 mm thick-
ness and 830 mm length, at the beam ends where the shear
force is maximum (doubling the corrugated web) (grade
S320GDZ); reinforcing U150/2.0 proles used under the load application
points, to avoid excessive local deformations (grade S350GDZ); self-drilling screws for ange-to-web connectionSTP-6.325; self-drilling screws for shear plates to end support with a
nominal diameterSTP-5.525; self-drilling screws as seam fasteners for corrugated webs with
a nominal diameterSTT-4.820; bolts M12 class 8.8 for anges to the end support connection.
Fig. 6 presents the experimental arrangement. Six pointsbending tests, monotonically conducted, were applied for each
specimen with a loading velocity of 2 mm/min.
The full-scale testing program was completed with tensile tests
to determine both the material properties for beam components
and the behaviour of connections.
3.3. Material and connections properties
In order to determine the mechanical properties of the CWB
components, a set of samples were cut out from the lipped
channels, corrugated sheet, both from the at regions and corners
0
50
100
150
200
250
0 10 20 30 40 50 60
Force
[kN]
Displacement [mm]
CWB - 2
BSP + DCW
DCW+ pull out of the screws
increasing of yield lines on SP
collapse
Fig. 15(a+b)
Fig. 15(c)
Fig. 16(a) Fig. 16(b)
Fig. 18. Loaddisplacement curve for CWB-2 beam.
Fig. 19. Distortion of the web corrugation at different levels of the load.
Fig. 20. Deformed shape of CWB-3 beam at failure.
D. Dubina et al. / Thin-Walled Structures 90 (2015) 159170166
7/25/2019 1-s2.0-S026382311500021X-main
9/12
and reinforcing shear panels, according to EN ISO 6892-1:2009
[42] specications, as shown in Fig. 7. A total number of 30
specimens have been tested, 5 for each type of specimen.
Table 2presents the mean values of tensile testes (i.e. yield and
ultimate strengths, fyM, fuM), the corresponding standard devia-
tions (i.e. sfy, sfu) and the characteristic values for yield and
ultimate strengths (i.e. fyk, fuk) for the above samples [40]. The
following abbreviations for coupons have been used: BM-CF
coupon cut from the ange of the lipped channel; BM-CW couponcut from the web of the lipped channel; CM-CF coupon cut from
the ange-web corner of the lipped channel; BM-CW coupon cut
from the at region of the corrugated web; BM-SP coupon cut
from the shear panel. Fig. 8 presents the failure modes for the
tested coupons.
Six types of connections were tested according to ECCS pub-
lication No. 124 [43] in order to determine their behaviour, at a
loading velocity of 1 mm/min, i.e.:
(1) T1-1.4 seam fasteners for corrugated sheets;
(2) T2-1.7, seam fasteners for shear plates and corrugated sheets;
(3) T3-3.7, self-drilling screws for shear plates and anges;
(4) T4-9.0, self-drilling screws for shear plates and end supports;
(5) T5-11.0, bolts for anges to end-supports;(6) T6-2.7, self-drilling screws for anges to corrugated webs at
mid-span,
in order to determine the behaviour of all types of connections
found in the beam[38].Table 3presents the tested specimens and
0
50
100
150
200
250
0 10 20 30 40 50 60 70
Fo
rce
[kN]
Displacement [mm]
CWB - 3
Fig. 19(a)
BSP + DCW
DCW+ tilting of the screws
collapse
Fig. 19(b) Fig. 20
Fig. 21. Loaddisplacement curve for CWB-3 beam.
Fig. 22. Distortion of the web corrugation.
Fig. 23. Deformed shape of CWB-4 beam at failure.
0
20
40
60
80
100
120
140
160
180
200
0 20 40 60 80 100 120 140 160
For
ce
[kN]
Displacement [mm]
CWB - 4
distortions of the corrugated web (DCW)
shear failure of the fastners
collapse
Fig. 22(a)
Fig. 22(b)
Fig. 23
Fig. 24. Loaddisplacement curve for CWB-4 beam.
D. Dubina et al. / Thin-Walled Structures 90 (2015) 159170 167
7/25/2019 1-s2.0-S026382311500021X-main
10/12
the number of tests done for each typology, while Fig. 9presents
the location of these types of connections.
Fig. 10presents the forcedisplacement curves for the six types
of tested connections presented above, with corresponding mean
values, to be used for relevant models in numerical simulations.Very good ductility can be observed in all the cases that being one
of the causes for the signicant redundancy of tested beams.
Fig. 11 presents one of the T3-3.7 tested connections in four
different stages.
4. Testing of specimens. Main results and interpretation
Therst tested specimen was CWB-1 beam and its conguration
has been presented in Fig.1. In this case the rst deformation, which
corresponds to the buckling of shear panel (BSP), appears for a
displacement of 10 mm at 52 kN (see Fig. 12a). At 11 mm small
distortions of the corrugated (DCW) web have been recorded, asshown inFig. 12b.Fig. 12c presents a detail of the shear panel at
failure.
The behaviour was ductile, with an initial stiffness ofK0-Exp
6862.2 N/mm and the maximum load is reached at Fmax
218.9 kN. The collapse appears for a displacement of 58 mm.
Fig. 13presents the deformed shape of the beam at collapse, while
inFig. 14the loaddisplacement curve is drawn.
In case of CWB-2 beam, detailed in Fig. 2, i.e. standard solution
(CWB-1) and supplementary lipped channel sections under the load
application points, the rst deformations correspond to the buckling
of shear panel combined with the distortion of the corrugated web,
and appear for a displacement of 14 mm (see Fig. 15a and b). At
29 mm displacement the distortion of the corrugated web increase
simultaneously with the pull out of the screws (see Fig. 15c).
Fig. 16a presents a detail of the shear panel for a displacementof 44 mm and at failure (seeFig. 16b).
The behaviour was ductile, with an initial stiffness ofK0-Exp
7831.5 N/mm and the maximum load is reached atFmax231.3 kN.
The collapse appears for a displacement of 54 mm. Fig. 17 shows
the deformed shape of the beam at collapse, while in Fig. 18 the
loaddisplacement curve is plotted.
Beam CWB-3 beam is the optimized solution by adapting the
ange-to-web connections according to the distribution of shear
stresses (connections at each second corrugations where the shear
force decreases), as shown in Fig. 3. For this beam the rst
deformation appears for a displacement of 15 mm, which corre-
sponds to the buckling of shear panel and distortion of the
corrugated web (seeFig. 19a). At 44 mm displacement the distor-
tion of the corrugated web is accompanied by the tilting of the
Fig. 25. (a) Distortion of the web corrugation; (b) buckling of shear panel.
Fig. 26. Deformed shape of CWB-5 beam at failure.
0
50
100
150
200
250
0 10 20 30 40 50 60 70 80 90 100
Force
[kN]
Displacement [mm]
CWB - 5
distortions of the corrugated web (DCW)
buckling of the shear panel (BSP)
collapse
Fig. 25(a)
Fig. 25(b)
Fig. 26
Fig. 27. Loaddisplacement curve for CWB-5 beam.
D. Dubina et al. / Thin-Walled Structures 90 (2015) 159170168
7/25/2019 1-s2.0-S026382311500021X-main
11/12
screws (see Fig. 19b), in the regions where the shear force is
signicant, but the number of screws is optimised.
The behaviour is ductile, with an initial stiffness of K0-Exp
7184.9 N/mm and the maximum load is reached at Fmax
209.5 kN. The collapse appears for a displacement of 62 mm.
Fig. 20 presents the deformed shape of the beam at collapse,
whileFig. 21shows the recorded loaddisplacement curve.
Beam CWB-4 is the standard solution, i.e. CWB-1, but eliminat-
ing shear panels and doubling of corrugated webs in the zones
with high shear forces, i.e. ends of the beam (see Fig. 4). In case of
this beam, the rst deformation, which corresponds to the distor-
tion of the corrugated web near supports, appears for a displace-
ment of 21 mm, as shown in Fig. 22a. At 74 mm displacement
shear failure of the fasteners was recorded followed by 10%
reduction of the beam capacity (seeFig. 22b).
The behaviour is ductile, with an initial stiffness of K0-Exp
3985 N/mm and the maximum load is reached at Fmax181.9 kN.
The collapse appears for a displacement of 164 mm. Fig. 23
presents the deformed shape of the beam at collapse, while
Fig. 24shows the recorded loaddisplacement curve.The last beam is CWB-5, and represents the optimized solution
by adapting both the ange-to-web connections (i.e. CWB-3) and,
supplementary, seam fasteners to ensure the continuity of corru-
gated web, according to the distribution of shear stresses (see
Fig. 5). The rst deformation corresponds to the distortion of the
web corrugation in the region with the reduced number of screws
for a displacement of 21 mm (see Fig. 25a), while at 35 mm
buckling of shear panels appears (Fig. 25b).
The behaviour is ductile, with an initial stiffness of K0-Exp
5516.2 N/mm and the maximum load is reached atFmax214.6 kN.
The collapse appears for a displacement of 88 mm. Fig. 26presents
the deformed shape of the beam at collapse, while Fig. 27shows
the recorded loaddisplacement curve.
Finally, Fig. 28 shows comparatively, for all the ve testedspecimens, the loaddisplacement curves and the ultimate (ULS)
and serviceability limit state (SLS) levels.
5. Conclusions
A large experimental program carried out at the CEMSIG
Research Centre (http://cemsig.ct.upt.ro) of the Politehnica Uni-
versity of Timisoara on ve beams with corrugated webs with
different arrangements for self-drilling screws and shear panels
was presented.
Very good agreement can be found between beams CWB-1 and
CWB-3, both in terms of initial stiffness and ultimate force. A
slightly increase in stiffness can be observed in case of CWB-2
beam, due to the supplementary lipped channel sections added
locally at the load application points. CWB-4 beam is the most
exible solution compared to the other four and has half initial
stiffness compared to CWB-2 beam. At this level, CWB-5 beam
represents the best solution in terms of optimisation.
Finally, even the results looks promising, signicant work has
to be done in order to investigate, validate and optimise such a
solution for mass production, i.e.:
numerical models for calibration and validation of experimen-
tal models[44];
to optimise the number of self-drilling screws used for connections;
tests using spot welding are the next step of research; there are
no differences expected at the level of global behaviour, but
some are estimated in terms of strength and stiffness;
numerical simulation of large span beams in order to study the
sensitivity to lateral-torsional buckling of such elements[44].
Based on that, standardized beams can be designed, calibrated
for series of vertical loading intensities, accounting or not for
lateral-torsional effects and the potential of this solution for
industrialized fabrication has to be, once more, emphasized.
It has to be notice that, by extending the application of thetechnical solution described within present paper for parallelanges girders, promising experimental results have been very
recently obtained in a PhD study on trapezoidal beams made of
cold-formed steel proles and corrugated web[45].
References
[1] Zeman & Co GmbH: SINBEAMCorrugated web beam, technical document;2012. www.zeman-stahl.com/.
[2] Hamada M., Nakayama K., Kakihara M., Saloh K., Ohtake F. Development ofwelded I-beam with corrugated web. The Sumitomo search, No. 29; 1984,p. 75-90.
[3] Zhao W. Behaviour and design of cold-formed steel hollow ange sectionsunder axial compression. (PhD thesis). Brisbane, Australia: School of Civ. Eng.,Queensland Univ.; 2005.
[4] Wanniarachchi S. Flexural behaviour of cold-formed steel beams with rectan-gular hollow anges. (PhD thesis). Brisbane, Australia: School of Civ. Eng.,Queensland University; 2005.
[5] Landolfo R, Mammana O, Portioli F, Di Lorenzo G, Guerrieri MR. Laser weldedbuilt-up cold-formed steel beams: experimental investigations. Thin-WalledStruct 2008;46(7-9):78191.
[6] WO 00/17463, International Patent Classication E04C 3/07, BEN-VAUTIER S.P.A.,Italy, Modular H-beam; 2000.
[7] Elgaaly M., Dagher H.: Beams and girders with corrugated webs. In: Proceed-ings of the SSRC annual technical session. St. Louis, MO; 1990.
[8] Smith D. Behavior of corrugated plates subjected to shear. (MSc thesis). Orono,ME: Dept. of Civil Engineering, University of Maine; 1992.
[9] Hamilton R. Behavior of welded girders with corrugated web. (Ph.D. thesis).Orono, ME: Dept. of Civ. Engr., Univ. of Maine; 1993 .
[10] Elgaaly M, Hamilton R, Seshadri A. Shear Strength of Beams with CorrugatedWebs. J Struct Eng ASCE, 122; 1996.
[11] Luo R, Edlund B. Numerical simulation of shear tests on plate girders withtrapezoidally corrugated webs. Sweden: Div. Steel and Timber Structures,
Chalmers Univ. of Tech.; 1995.[12] Elgaaly M, Seshadri A, Hamilton RW. Bending strength of steel beams withcorrugated webs. J Struct Eng ASCE 1997;123(6):77282.
[13] Chan CL, Khalid YA, Sahari BB, Hamouda AMS. Finite element analysis ofcorrugated web beams under bending. J Constr Steel Res 2002;58:1391406.
[14] Johnson R, Cafolla J. Local ange buckling in plate girders with corrugatedwebs. Proc Inst Civil Eng Struct Build 1997:14856.
[15] Lindner J. Lateral-torsional buckling of beams with trapezoidally corrugatedwebs. In: Proc of the fourth international colloquium on stability of steelstructures. Budapest, Hungary; 1990.
[16] Jiho M, Jong W, Byung HC, Hak-Eun L. Lateral-torsional buckling of I-girder withcorrugated webs under uniform bending. Thin Walled Struct. 2009. p. 2130.
[17] Moon J, Yi JW, Choi BH, Lee HE. Lateral-torsional buckling of I-girder withcorrugated webs under uniform bending. Thin-Walled Struct 2009;47:2130.
[18] Pasternak H., Robra J., Kubieniec G. New proposals for EN 1993-1-5, Annex D:plate girders with corrugated webs. In: Codes in structural engineering, jointIABSE-b conference, 2, 1365-1372, Dubrovnik, Croatia; 2010.
[19] Pasternak H., Kubieniec G. Flange buckling of sinusoidal corrugated girders. In: Intconf in thin walled structuresICTWS2011. Timisoara, Romania; 5-7.09.2011. 625-
633; 2011.
0
50
100
150
200
250
0 20 40 60 80 100 120 140 160 180
Force
[kN]
Displacement [mm]
CWB - 1
CWB - 2
CWB - 3
CWB - 4
CWB - 5
Fig. 28. Loaddisplacement curves for the tested specimens.
D. Dubina et al. / Thin-Walled Structures 90 (2015) 159170 169
http://www.zeman-stahl.com/http://www.zeman-stahl.com/http://www.zeman-stahl.com/http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref1http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref1http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref1http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref1http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref1http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref1http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref2http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref2http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref2http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref2http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref2http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref2http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref3http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref3http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref3http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref3http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref3http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref3http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref4http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref4http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref4http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref5http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref5http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref5http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref6http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref6http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref6http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref7http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref7http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref7http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref7http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref8http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref8http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref8http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref8http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref8http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref9http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref9http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref9http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref9http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref9http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref10http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref10http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref10http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref10http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref10http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref10http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref10http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref11http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref11http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref11http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref11http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref11http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref12http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref12http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref12http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref12http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref12http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref12http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref12http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref11http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref11http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref10http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref10http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref9http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref9http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref8http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref8http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref7http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref7http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref7http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref6http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref6http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref5http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref5http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref4http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref4http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref3http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref3http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref3http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref2http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref2http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref2http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref1http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref1http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref1http://www.zeman-stahl.com/7/25/2019 1-s2.0-S026382311500021X-main
12/12
[20] EN1993-1-5:2006. Eurocode 3: design of steel structuresPart 1-5: Plated
structural elements. CEN, Brussels.[21] Moon J, Lim NH, Lee HE. Moment gradient correction factor and inelastic
exuraltorsional buckling of I-girder with corrugated steel webs. Thin-
Walled Struct 2013;62:1827.[22] Leiva-Aravena L., Edlund B. Buckling of trapezoidally corrugated webs.
In: Proceedings of the ECCS colloquium on stability of plates and shells.
Ghent, Belgium; 1987.[23] Elgaaly M, Seshadri A. Girders with corrugated webs under partial compres-
sive edge loading. J Struct Eng ASCE 1997;123(6):78391.[24] Luo R, Edlund B. Ultimate strength of girders with trapezoidally corrugated
webs under patch loading. Thin-Walled Struct 2006;24:13556.[25] Nguyen ND, Han SR, Kim JH, Kim SN, Kang YJ. Moment modication factors of
I-girder with trapezoidal web corrugations under moment gradient. Thin-
Walled Struct 2012;57:112.[26] Tahir M, Sulaiman A, Anis S. Experimental tests on composite and non-
composite connections using trapezoid web proled steel sections. Steel
Struct 2008;8:4358.[27] Kvesdi B, Jger B, Dunai L. Stress distribution in the anges of girders with
corrugated webs. J Constr Steel Res 2012;79:20415.[28] US3444664A Patent, Ribbed web girder, Macomber Inc., Fink Jr. HR, Scott VP
(inventors),http://www.google.com/patents/US3444664; 1967.[29] Yu W-W. Cold-formed steel design. 3rd ed.. New York, NY: John Willey&Sons;
2000.[30] Harrison JD. Exploratory fatigue tests of two girders with corrugated webs.
Br Weld J 1965(Mar):.[31] Neagoie B., Ungureanu V., Dubina D. Beams with cold-formed steel sections
for anges and corrugated web. In: Stability and ductility of steel structures
recent research. Academic days; 27 May 2005, Timisoara, Romania, 978-973-
661-977-9, 57-70 (in Romanian).[32] Prefabricated steel structures for low-rise buildings in seismic areas (Pre-
casteel), nal report. isbn:978-92-79-29011-4, Research Fund for Coal and
Steel, RFSR-CT-2007-00038; 2013.
[33] Briskham P., Blundell N., Han L., Hewitt R., Young K., Boomer D. Comparison of self-pierce riveting, resistance spot welding and spot friction joining for aluminiumautomotive sheet. SAE 2006 congress. Technical paper 2006-01-0774.
[34] Guenfoud N., Tremblay R., Rogers C.A. Arc-spot welds for multi-overlap roofdeck panels. In: Twentieth international specialty conference on cold-formedsteel structures. St. Louis, MO, USA; November 3 & 4, 2010. p. 535-549.
[35] Snow G. Strength of arc spot welds made in single and multiple steel sheets.(MsC thesis). 2008.
[36] EN1993-1-1:2005. Eurocode 3: design of steel structuresPart 1-1: Generalrules and rules for buildings. CEN; Brussels.
[37] EN1993-1-3:2006. Eurocode 3: design of steel structures. Part 1-3: General
rules. Supplementary rules for cold-formed thin gauge members and sheeting.CEN, Brussels.
[38] ECCS 1995. European recommendations for the application of metal sheetingacting as a diaphragm. Publication P088, ECCS, Brussels, Belgium.
[39] Flp LA, Dubina D. Design criteria for seam and sheeting-to-framing connectionsof cold-formed steel shear panels. J Struct Eng ASCE 2006;132(4):58290.
[40] Glia L. Structural performances of steel beams with back-to-back cold-formedlipped channel sections for anges and corrugated web. (PhD thesis (Promo-ters: Kiss Zs, Dubina D). Romania: Technical University of Cluj-Napoca; 2012(in Romanian).
[41] Dubina D, Ungureanu V, Glia L. Cold-formed steel beams with corrugated weband discrete web-to-ange fasteners. (May). Steel Constr 2013;6:7481.
[42] EN ISO 6892-1:2009. Metallic materialstensile testingPart 1: Method oftest at room temperature. CEN, Brussels.
[43] ECCS 2008. The Testing of connections with mechanical fasteners in steelsheeting and sections. Publication 124, ECCS, Brussels, Belgium.
[44] Dubina D., Ungureanu V., Glia L. Cold-formed steel beams of corrugated weband built-up section chords. In: Proceedings of the seventh European
conference on steel and composite structures
EUROSTEEL 2014; September1012, 2014, Naples, Italy, 978-92-9147-121-8, p. 429-430.[45] Ballok RI. Structural systems with trapezoidal beams made of cold-formed
steel proles and corrugated web. (PhD thesis (Promoters: Kiss Zs, Dubina D).Romania: Technical University of Cluj-Napoca; 2014 (December).
D. Dubina et al. / Thin-Walled Structures 90 (2015) 159170170
http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref13http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref13http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref13http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref13http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref13http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref13http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref13http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref13http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref13http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref14http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref14http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref14http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref14http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref14http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref15http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref15http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref15http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref15http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref15http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref16http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref16http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref16http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref16http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref16http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref16http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref16http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref16http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref17http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref17http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref17http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref17http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref17http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref17http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref17http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref17http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref18http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref18http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref18http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref18http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref18http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref18http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref18http://www.google.com/patents/US3444664http://www.google.com/patents/US3444664http://www.google.com/patents/US3444664http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref19http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref19http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref19http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref19http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref20http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref20http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref20http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref21http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref21http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref21http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref22http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref22http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref22http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref22http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref22http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref23http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref23http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref23http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref23http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref23http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref23http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref23http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref24http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref24http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref24http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref24http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref24http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref24http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref24http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref25http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref25http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref25http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref25http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref25http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref25http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref25http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref25http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref25http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref24http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref24http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref23http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref23http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref23http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref23http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref22http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref22http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref21http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref21http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref20http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref20http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref19http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref19http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref19http://www.google.com/patents/US3444664http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref18http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref18http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref17http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref17http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref17http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref16http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref16http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref16http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref15http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref15http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref14http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref14http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref13http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref13http://refhub.elsevier.com/S0263-8231(15)00021-X/sbref13