Abstract
This article reviews research on cold-formed steel structures published in 2013 and 2014 in three leading journals: the Journal of Structural Engineering, ASCE, Thin-Walled Structures and the Journal of Constructional Steel Research. It also reviews papers published in the three main conferences in the area over the same period. These are Eurosteel 2014 (Naples, Italy), the 7th International Conference on Thin-Walled Structures (Busan, Korea) and the 22nd International Specialty Conference on Cold-Formed Steel Structures (St Louis, MO, USA). Three research areas which have recently been incorporated in the North American Specification NAS S100:2012 or are being incorporated in the Australian/New Zealand Standard AS/NZS 4600 have been highlighted. These are the works on the semi-analytical finite strip method for sections in shear by the author and his colleagues at the University of Sydney, net section rupture by Associate Professor Lip Teh at the University of Wollongong and fire design by Professor Mahendran at Queensland University of Technology.
Introduction
Cold-formed steel (CFS) structures are structures composed of structural sections formed by folding at ambient temperature without any heat treatment. They are normally thin-walled but sections up to 25-mm thick are now being cold-formed from plate and strip. The usual manufacturing process is by roll forming where coil steel passes through a series of rollers which progressively form the desired shape. Traditionally, simple channels (Cs), zeds (Zs), hat sections and decking have been used mainly in roof and wall systems, steel storage racks, steel-framed houses (residential) and many other similar secondary applications. However, the sections are now being used more commonly in primary structures such as portal frames and floor systems. Furthermore, as the sections become thinner in higher strength steel, more complex shapes are being created with the inclusion of more complex stiffeners both in the flat elements and at the section lips.
Research into CFS structures has increased considerably in recent years. In 2003, the author of this article published a similar review in the Journal of Constructional Steel Research (Hancock, 2003) which contained 60 publications. This review cites over 200 papers over a similar 2-year period 2013–2014 indicating a more than threefold increase in research in the area. Since the publication of the 2003 paper, there have been significant developments in CFS design specifications and standards. The 2007 and 2012 editions of the North American Specification NAS:S100 and the 2005 edition of the Australian/New Zealand Standard AS/NZS 4600 included the direct strength method (DSM) of design in addition to the effective width method (EWM). In the same period, European Committee for Standardisation published Eurocode 3 Part 1.3 for cold-formed members and sheeting. It is substantially based on the EWM. A recent text explaining this code with examples has been prepared by Dubina et al. (2013b). The Chinese Technical Code of Cold-Formed Thin-Walled Steel Structures GB 50018 was published in 2002 based on the EWM.
Since 2003, there has been a considerable shift in the areas of research. Some areas such as section buckling including generalised beam theory (GBT), the finite strip method (FSM) and the constrained finite strip method (cFSM), the DSM of design, shear walls, fire design, and seismic design have increased considerably. On the other hand, areas such as corrugated and curved panels, torsion and distortion (excluding distortional buckling) and enhanced mechanical properties have seen a reduction in research. A review of cold-formed stainless steel has not been included in this review.
Section buckling and design including GBT, FSM and DSM
With the advent of the DSM of design based on the buckling signature curve concept, there has been a substantial increase in research into the different methods of buckling analysis of thin-walled sections. The two principal methods used are the GBT and the FSM. The finite element method (FEM) can be used, but it is less useful in isolating the separate local, distortional and overall (flexural/torsional) modes at this time and the interaction between them. In preparing this review, it has become difficult to separate buckling analyses from design. However, papers where investigation of the elastic buckling modes are the primary aim of the paper are classified as buckling, whereas papers where the theory or test results are used in design such as the DSM have been classified as design.
Elastic buckling including interaction
The major work in elastic buckling has concentrated on GBT mainly from Professor Camotim and his colleagues at the Universidade de Lisbon (TUL), the cFSM both from Professor Schafer and his colleagues at Johns Hopkins University and Professor Ádány and his colleagues at Budapest University of Technology and Economics and the FSM for shear buckling at the University of Sydney (Pham and Hancock). Other researchers in the area are also cited.
Papers published using GBT (Basaglia et al., 2013, 2014b; Bebiano et al., 2014; De Miranda et al., 2013; Taig et al., 2014) cover the areas of continuous purlins, distortional postbuckling, inclusion of extension and shear modes and the new GBTUL-2.0 software. Papers published using the cFSM (Ádány, 2013, 2014; Ádány and Schafer, 2014a, 2014b, 2014c; Li et al., 2014e) include generalisation to arbitrary cross-sections and extension to the FEM. Alternative methods (Becque and Li, 2014; Karakonstantis and Becque, 2014) for modal decomposition have recently been developed based on polarisation. Further work on including perforations in the FSM has been performed by Smith and Moen (2014).
Recent developments in the FSM include shear and localised loading (Hancock and Pham, 2013, 2014a, 2014b; Pham and Hancock, 2013b). A new development in these latter papers is the ‘shear signature curve’ as shown in Figure 1(b) for sections in pure shear as shown in Figure 1(a).

Lipped channel section in shear: (a) shear flow distribution and (b) shear buckling curves (Hancock and Pham, 2013).
The curve semi-analytical finite strip method (SAFSM; program bfinst7.cpp) is the buckling stress versus half-wavelength ‘signature curve’ of a lipped channel in pure shear with unrestrained end sections, and the curve labelled reSAFSM (program bfinst8.cpp) is the elastic buckling curve versus length for a section restrained with simply supported ends so that it is prevented from cross-sectional deformation at its ends. A shear buckling mode at a half-wavelength of 200 mm from the SAFSM is shown in Figure 2(a) and at with simply supported ends at a fixed length of 200 mm from reSAFSM is shown in Figure 2(b). When the reSAFSM model length is increased to 1000 mm, then the buckling mode shown in Figure 3 with multiple half-wavelengths results. In this case, the end conditions become less important so that the buckling stress is close to the value for the unrestrained single half-wavelength in Figure 2(a) as can be observed from the stresses in Figure 1(b) by comparing the reSAFSM curve at 1000 mm with the SAFSM at 200 mm.

Lipped channel buckling modes in pure shear (200 mm) (Hancock and Pham, 2013): (a) unrestrained buckling mode (SAFSM) and (b) restrained buckling mode (reSAFSM).

Simply supported lipped channel shear buckling mode at L = 1000 mm (reSAFSM) (Hancock and Pham, 2013).
Considerable research has been undertaken in the area of mode interaction between local, distortional and/or flexural/flexural–torsional or combinations of these (Dinis et al., 2014a; Dubina et al., 2013a; Loughlan and Yidris, 2014; Martins et al., 2014a, 2014b; Rizzi et al., 2013; Santos et al., 2014; Ungermann et al., 2014). Based on this research, there is a need to review some of the design rules in existing standards and specifications to further account for the interactions.
DSM and other design methods
The DSM has been further investigated in detail for compression members, flexural members, combined bending and compression, members in shear and the interaction between the modes. The references are for column design (De Miranda et al., 2014; He et al., 2014; He and Zhou, 2014; Landesmann and Camotim, 2013; Young et al., 2013), for beam/purlin design (Basaglia and Camotim, 2013; Basaglia et al., 2014a; Gao and Moen, 2014; Keerthan and Mahendran, 2014a; Kumar and Kalyanaraman, 2014; Pham and Hancock, 2013a), for shear design (Pham et al., 2014a, 2014b, 2014d; Pham and Hancock, 2014) and for structural systems (Camotim and Basaglia, 2014). As for buckling, there is a need to review some of the design rules in existing standards and specifications to further account for interaction. A rig for measuring and classifying global, distortional and local imperfections in cold-formed members is described by Zhao and Schafer (2014).
Compression members including wall studs
Considerable work has taken place in the area of compression members including angles, sections with perforations and holes, sections built-up from single members, wall stud member with sheathing attached and eccentric loading (beam–columns). The references are for angles (Shifferaw and Schafer, 2014; Silvestre et al., 2013), members with dimples and other indentations, perforations and holes (Ekmekyapar et al., 2014; Kubde and Sangle, 2014; Kulatunga et al., 2014; Kulatunga and Macdonald, 2013, 2014; Macdonald and Kulatunga, 2014; Nguyen et al., 2014; Xu et al., 2014), Carbon Fiber Reinforced Plastic (CFRP) strengthening (Kalavagunta et al., 2013), for members built-up from channel sections (Crisan et al., 2014a; Dabaon et al., 2014; Fratamico and Schafer, 2014; Li et al., 2014b; Piyawat et al., 2013; Selvaraj and Madhavan, 2014; Ting and Lau, 2014a, 2014b), for fixed-ended columns (Dinis et al., 2014c; Gunalan and Mahendran, 2013b), for sheathed members (Peterman and Schafer, 2014) and for members in combined bending and compression (Li et al., 2014a; Torabian et al., 2014). Clearly, the most recent research is directed away from unperforated and single members to more complex arrangements including perforated members.
Flexural members including purlins, sheeting and decking
Research into flexural members is wide ranging from purlins with sleeves and different types of sheeting restraint, novel shapes and sections including corrugated and stiffened webs, shear including shear with holes, flexure with holes, web crippling, composite floors including Oriented Strand Board (OSB), and sheeting and decking. The references are for purlins and beams (Gelji et al., 2014; Georgescu and Ungureanu, 2014; Gutierrez et al., 2013; Kujawa and Szymczak, 2014; Loureiro and Calvo, 2014; Moen et al., 2013; Pham et al., 2014c; Seek, 2014; Uzzaman et al., 2014; Ye et al., 2013), novel shapes including stiffened and corrugated webs (Dubina et al., 2014; Dubina and Ungureanu, 2014; Laím et al., 2013, 2014a; Łukowicz and Urbanska-Galewska, 2014; Paczos, 2014; Siahaan et al., 2014a, 2014b; Tondini and Morbioli, 2014; Wang et al., 2014c; Wang and Young, 2014a, 2014b, 2014c), combined bending and compression (Cheng et al., 2013), shear including combined bending and shear (Acharya et al., 2013; Bruneau et al., 2014; Keerthan et al., 2014a; Keerthan and Mahendran, 2013a, 2013b, 2013c, 2014b, 2014c), web crippling (Gunalan and Mahendran, 2014c; Keerthan et al., 2014b; Keerthan and Mahendran, 2014a, 2014d; Natário et al., 2014a, 2014b; Uzzaman et al., 2013), composite floors including OSB board (Chatterjee et al., 2014b; Zhou et al., 2014) and sheeting and decking (Ádány et al., 2013; Casariego et al., 2014; Danilov and Tusnina, 2014; Guo et al., 2014; Lawson and Popo-Ola, 2013). The influence of sheeting, decking, sandwich panels and OSB board on flexural member behaviour including both purlins and floors is being further investigated in detail to improve reliability and performance.
Connections and fasteners
Research into connections has focussed on five areas. These are bolted connections which include the majority of the papers, clinching, screwed connections, fastener reliability and moment connections. The references are for bolted connections including section rupture and bearing failure (Bolandim et al., 2013; Clements and Teh, 2013; Liu et al., 2014; Teh and Gilbert, 2013a, 2013b, 2014a, 2014b; Yu and Panyanouvong, 2013; Yu and Xu, 2013), clinching and pinned connections (Di Ilio, 2014; Lambiase and Di Ilio, 2014; Mathieson et al., 2014; Mucha and Witkowski, 2014), screwed connections (Moen et al., 2014; Sivapathasundaram and Mahendran, 2014), fastener reliability (Chatterjee et al., 2014a) and moment connections (Bučmys et al., 2014; Lim et al., 2014; Sabbagh et al., 2013). The design of bolted connections in net section rupture has had the most investigation and is being incorporated in the NAS S100 and AS/NZS 4600 standards. However, less research on connections seems to be being undertaken than in the past.
Net section tension rupture is probably the most familiar failure mode of bolted connections to both engineers and non-engineers. Examples of this failure mode in a flat sheet and in a channel section, both of which are composed of cold-reduced sheet steel, are shown in Figure 4. The net section tension capacity of such a bolted connection may be reduced by the existence of shear lag, relative to bolted connections in more ductile steel members.

Net section tension rupture in a flat sheet and a channel section.
Attempts to account for the shear lag effect in flat members by the major CFS design specifications worldwide have led to anomalies (Teh and Gilbert, 2014a), which persisted despite various amendments over the past several decades. For a bolted connection where the bolt spacing is at least twice the bolt diameter, all the prevailing code equations predict the net section tension capacity to increase with decreasing net section area. This anomaly was recently resolved via the use of calculus, and a new equation (Teh and Gilbert, 2014a) that avoids the known anomalies has been adopted into the 2016 edition of the North American Specification for the Design of Cold-Formed Steel Structural Members.
Comparisons between the in-plane shear lag factors given by the prevailing design standards and the new equation for a bolted connection in a flat brace can be made in Figure 5. The variable d denotes the bolt diameter, and W is the bolt spacing in the direction perpendicular to loading, or, for a single bolted connection, the connected member width.

Shear lag factors for bolted connections in flat sheets.
Bolandim et al. (2013) found that the net section tension rupture provisions in NAS S100:2012 for bolted connections in angle and channel braces led to reliability indices much lower than the target index of 3.5. Based on reliability analyses, they calculated the required resistance factors to be as low as 0.30 for the specification’s design equations.
Teh and Gilbert (2013b) considered the three factors affecting the net section efficiency of a channel brace bolted at the web, namely the in-plane shear lag, the out-of-plane shear lag and the interaction between the detrimental moment due to connection eccentricity and the counteracting moment provided by the bolted connection. They proposed a design equation that has since been shown to provide reasonable estimates for channel braces composed of G450 and SSC400 sheet steels having various aspect ratios and bolting configurations (Teh and Gilbert, 2014a). This equation will be adopted into the 2016 edition of the North American Specification for the Design of Cold-Formed Steel Structural Members NAS S100.
Comparisons between the net section efficiencies of typical channel braces bolted at the web given by the underlying equation in NAS S100:2012 and the new equation to be used in NAS S100:2016 can be seen in Table 1. The results of a regression analysis equation derived by the Center for Cold-Formed Steel Structures (CCFSS) at the University of Missouri are also shown in the table. The variable Ww
is the overall web depth, Wf
is the clear flange width, t is the wall sheet thickness,
Net section efficiencies of channel braces bolted at the web.
For angle braces bolted at one leg, Teh and Gilbert (2014b) proposed a design equation which will be adopted into the 2016 edition of the North American Specification for the Design of Cold-Formed Steel Structural Members NAS S100. It has a similar form to that proposed by Teh and Gilbert (2013b) for channel braces bolted at the web.
Shear walls
There has been a significant increase in research into steel-framed clad walls, both in shear and compression. Investigations of both seismic and dynamic behaviour have become prominent. The references are for clad shear walls in compression (Vieira and Schafer, 2013), blast load (Bondok et al., 2013), static shear (Baldassino et al., 2014; Hernandez-Castillo et al., 2014; Shakibanasab et al., 2014; Tian et al., 2013; Yanagi and Yu, 2014) and seismic/dynamic shear (Baldassino et al., 2014; Balh et al., 2014; Bian et al., 2014; Buonopane et al., 2014; Iuorio et al., 2014a, 2014b; Javaheri-Tafti et al., 2014; Lin et al., 2014; Macillo et al., 2014; Shahi et al., 2014; Shamim and Rogers, 2013; Shimizu et al., 2013; Vigh et al., 2013; Yu et al., 2014). It is clear that clad and braced shear walls are useful for resisting seismic shear loads on CFS frames.
Storage racks
The vast majority of storage racks are constructed from CFS so that much of the research in the area of the stability of steel framing is covered by storage racks. The references are for rack uprights including buckling (Bernuzzi and Maxenti, 2014; Casafont et al., 2014; Crisan et al., 2014b; Dinis et al., 2014b; Nedelcu et al., 2014; Ren and Zhao, 2014; Trouncer and Rasmussen, 2014), analysis of frames (Gilbert et al., 2014; Rasmussen and Gilbert, 2013) and connections (Wang et al., 2014d; Zhao et al., 2014). Interaction buckling in the uprights of storage rack frames is clearly an active area of investigation.
Fire design
Fire design research has become more prominent in recent years especially as more CFS is used in residential construction. Fire design rules in EC3 Part 1.2 (EN 1993-1-2, 2005) are considered to be applicable to CFS members within the scope of EC3 Part 1.3 (EN 1993-1-3, 2006). However, they were specifically developed for hot-rolled steel structures, and despite some special provisions given in Annex E of EC3 Part 1.2 for thin-walled sections, past studies have demonstrated the need for specific research on CFS members exposed to fire conditions in order to develop suitable fire design rules. CFS members are not always subjected to a uniform temperature exposure. For example, CFS wall and floor systems are often protected by fire-resistant gypsum plasterboards (Figure 6), and hence, their lipped channel members (studs and joists) will be subjected to a non-uniform temperature distribution when exposed to fire on one side (Figure 7).

CFS wall system.

Non-uniform temperature distribution in wall studs.
The simplified method in EC3 Part 1.2 (EN 1993-1-2, 2005) requires that the design action effects in a design fire is less than or equal to the corresponding design capacity of the steel members at any given time during the design fire, that is, subject to a particular non-uniform or uniform elevated temperature exposure. This design approach requires thermal performance evaluation of members to determine their temperature at any given time during the design fire (Figure 7), a good understanding of the behaviour and reduced capacities of CFS steel members (columns and beams) subject to various buckling modes such as local, distortional and flexural and flexural–torsional at elevated temperatures (Figure 8) as well as the reduced mechanical properties of CFSs at elevated temperatures (Figure 9). Fire research on CFS members has been addressing the above issues so that improved fire design rules can be developed. Most research was aimed at using the ambient temperature design rules with appropriately modified elevated temperature mechanical properties of CFSs.

CFS member and wall failures in fire.

Elevated temperature mechanical properties of CFS (Dolamune Kankanamge and Mahendran, 2011).
During 2013–2014, Gunalan et al. (2013) and Gunalan and Mahendran (2013a) extended the fire research on CFS members at the Queensland University of Technology to CFS wall systems using both full-scale fire tests and finite element analyses to predict their structural and thermal performances in standard fire conditions and developed suitable fire design rules within the Australian, North American and European CFS design provisions. These design rules can be used to predict the capacity of wall studs in fire and the associated fire resistance ratings of wall systems. Ariyanayagam and Mahendran (2014) expanded this research to include CFS wall systems exposed to more realistic fire time–temperature curves including those based on parametric fires given in EC3 Part 1.2. Professor Y.C. Wang from the University of Manchester continued his work on CFS wall systems to develop a simple method to determine the non-uniform temperature distributions in the wall stud sections exposed to fire on one side without the need to use finite element simulations (Shahbazian and Wang, 2013). Shahbazian and Wang (2014) then proposed a fire design method for CFS walls exposed to parametric fires. In this design method, their simple temperature distribution prediction method is first used to determine the wall stud temperatures in a given parametric fire, and their DSM-based design rules are then used to calculate the capacity of wall studs subject to local, distortional or global buckling effects, which provide the required ultimate load of wall studs versus time in such parametric fires.
Chen et al. (2013) and Chen and Ye (2014) conducted full-scale fire tests of CFS walls made of different configurations and confirmed Gunalan et al.’s (2013) findings that the use of external insulation instead of cavity insulation improved the fire performance of walls. Chen et al. (2013) investigated the effects of using different fire protection boards, based on which suitable recommendations were made to improve the fire performance.
Research on individual CFS columns and beams was also continued during 2013–2014. Craveiro et al. (2014) conducted fire tests of lipped channel and built-up channel columns with restrained thermal elongation to investigate the effects of cross-section, end support conditions, surrounding structure stiffness and applied load level. Their results identified the critical parameters that reduced the critical temperature significantly. Laím et al. (2014b) conducted a similar study for CFS lipped channel and built-up channel beams to investigate the effects of four different profiles, axial restraint to thermal elongation and rotational stiffness of beam supports on the failure modes, temperatures and times. Their tests showed that any axial restraint to thermal elongation was detrimental to fire performance while the use of closed built-up profiles improved the fire performance. Cheng et al. (2014) investigated the fire performance of CFS members under axial and transverse loading while Gunalan et al. (2014) investigated the flexural–torsional buckling behaviour of CFS columns using uniform elevated temperature tests and finite element analyses to develop suitable fire design rules, which showed their adequacy when appropriately reduced mechanical properties were used.
Local buckling effects of CFS members at elevated temperatures are currently accounted for using the conventional EWM with elevated temperature yield strength based on 0.2% proof strength. Couto et al. (2014) used a numerical study to investigate the accuracy of this approach and proposed modified effective width expressions for internal and outstand elements of CFS profiles exposed to uniform elevated temperatures. Many research studies mentioned above used nonlinear 3D finite element analyses to study the fire performance of CFS members and to investigate the accuracy of elevated temperature design rules. Ellobody (2013) presents more details of this approach and the main parameters to be considered for the heat transfer and structural analyses of CFS columns in fire conditions.
Limited work has been undertaken on CFS connections in fire. Yan and Young (2013) address this issue through a detailed experimental study of double shear bolted connections of thin steels exposed to uniform elevated temperatures. They showed that the use of ambient temperature design equations for connection strengths provided conservative predictions when elevated temperature mechanical properties were used.
Assessing the residual strength of CFS structures following a fire event is important and thus Gunalan and Mahendran (2014a, 2014b) investigated the post-fire mechanical properties of CFSs and proposed suitable predictive equations for this purpose.
Fire research on CFS members, connections and structural systems is continuing, which will lead to accurate design methods to predict their structural fire performance for inclusion in future design standards.
Seismic design
As CFS structures are used in more active seismic areas, there is an increasing need to carry out research on cold-formed members and structural systems subject to cyclic loading. Shear walls under cyclic load are already covered in section ‘Shear walls’. The references are for moment frames (Bai and Lin, 2013; Li et al., 2014d), strap-braced frames (Dao and Van de Lindt, 2013; Pali et al., 2014; Terracciano et al., 2014), mid-rise construction (Ozaki et al., 2013; Yuan and Xu, 2014) and framing members (Padilla-Llano et al., 2014a, 2014b, 2014c). It is interesting to see increasing research on mid-rise and moment-resisting frames.
Frames
Portal frames composed entirely of cold-formed members are being used frequently and so more research on their structural behaviour is being undertaken. The references are for residential and framed buildings (Li et al., 2013, 2014c; Peterman et al., 2014), moment and portal frames (Hanna, 2014; Zhang and Rasmussen, 2014) and stressed skin action (Phan et al., 2014a, 2014b; Wrzesien et al., 2014). A significant new area of research is clad framed and residential buildings and the effect of the cladding on the frame behaviour especially during seismic action.
Optimisation
Research into section shape optimisation has been carried out for a long time and continues as new algorithms are developed. The references are Ostwald and Rodak (2013), Moharrami et al. (2014), Franco et al. (2014), Leng et al. (2014) and Wang et al. (2014a, 2014b).
Conclusion
This article provides a bibliographical review of papers published in the area of CFS structures in the Journal of Structural Engineering, ASCE; Thin-Walled Structures; Journal of Constructional Steel Research; Eurosteel 2014 Conference, Naples; 7th International Conference in Thin-Walled Structures, Busan, Korea, 2014; and the 22nd International Specialty Conference on Cold-Formed Steel Design and Construction, St Louis, MO, USA, 2014. The notable feature of this article is the threefold increase in the number of papers between a similar review in 2003 and this review from 60 to over 200. The highlight in the period has been the inclusion of the DSM of design in the North American Specification NAS:S100 and the Australian/New Zealand Standard AS/NZS 4600. This has led to a significant increase in stability research using the GBT and the FSM including mode interaction. Two other areas with a major increase in research are fire design and shear walls particularly under seismic load. There is an increased need to incorporate much of this research into new design specifications and standards. At the time of writing, new editions of NAS:S100 and AS/NZS 4600 are under preparation using much of this research.
The author has noted several areas that need further research during this review. These are the inclusion of localised loading/web crippling in the DSM of design so that all modes of failure are covered, more studies on structural systems and system effects including the newly developing method of modular construction, more research on the use of advanced analysis methods for frames and further investigation of seismic design of systems as opposed to simply shear walls.
Footnotes
Acknowledgements
The author is grateful for the material provided by Professor Mahendran (QUT) and Associate Professor Lip Teh (UOW).
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship and/or publication of this article.
