Composite construction is the dominant form of construction for the multi-storey building sector. This has been the case for over twenty years. Its success is due to the strength and stiffness that can be achieved, with minimum use of materials.
The reason why composite construction is often so good can be expressed in one simple way - concrete is good in compression and steel is good in tension. By joining the two materials together structurally these strengths can be exploited to result in a highly efficient design. The reduced self weight of composite elements has a knock-on effect by reducing the forces in those elements supporting them, including the foundations. Composite systems also offer speed of construction benefits, which were a key reason for the boom in use of steel for commercial buildings in the UK in the 1980s. The floor depth reductions that can be achieved using composite construction can also provide significant benefits in terms of the costs of services and the building envelope.
The scope of this article covers composite beams, composite slabs, composite columns and composite connections. Whilst beams and slabs are very common in UK construction, indeed there exist a number of different basic types of composite beam, composite columns and composite connections are much less so. The reasons for this are considered below.
Design of composite elements and systems
Design of composite beams in the UK has traditionally been according to BS 5950-3-1. Composite slabs have been designed to BS 5950-4 and the profiled decking used for those slabs to BS 5950-6. There was no BS guidance for composite columns. Design (for buildings) is now covered by BS EN 1994-1-1 and BS EN 1994-1-2 (Eurocode 4), noting that the British Standards may still be used.
How and why composite construction works
Whilst steel is a material that works extremely well in tension, when parts (or all) of a cross section are slender (thin, wide plate elements) they fail in local buckling before the yield strength of the material can be reached. This results in inefficiencies in the material used, as part of the cross sectional is ineffective.
The figure below shows the plastic stress distribution in a typical downstand beam acting compositely with a composite slab. The relative proportions of the steel section and slab mean that, as is commonly the case, the plastic neutral axis lies within the concrete. All the steel is therefore in tension.
Concrete is a material that works well in compression but has negligible resistance in tension. Hence for structural purposes it traditionally relies on steel reinforcement to carry any tensile forces (this is the role played by the steel part of a composite cross section, which is effectively external reinforcement), or must be pre-stressed so that even when subject to tension an element is in net compression.
For the concrete part (within the so-called effective width) of a cross section to carry compression, and the steel part to carry tension, the two materials must be structurally tied together. For downstand beams this is achieved using headed shear studs, which are attached to the upper flange of the steel beam. This attachment is normally achieved with so-called through deck welding. The profiled metal decking that forms the basis of the composite slabs is sandwiched between the base of the stud and the top flange, and the welding process joins all three together. Although the upper surface of the top flange must be left unprotected, the presence of galvanizing on the decking does not affect weld quality.
In exceptional circumstances through deck welding can be avoided by using single span lengths of decking (which butt up to rows of studs welded directly to the top flange in the fabrication shop), or cutting holes in the decking so that it can be dropped over the shop welded studs.
Other forms of shear connection are available, including larger diameter studs and shot-fired connectors, but for buildings by far the most common option is 19 mm diameter headed studs. Their resistance according to BS EN 1994, when used with transverse decking, is less than the resistance given in BS 5950-3-1. Also, BS EN 1994 states that not more than two studs can be used per trough when the decking runs transverse to the beam axis.
One of the advantages of welded studs is that they are considered to be ductile, which means that (in the absence of any fatigue considerations) the connection can be designed using plastic principles because it is assumed that force can be transferred between adjacent studs. This greatly simplifies the design process.
When a beam is designed with full shear connection it means that sufficient connectors are present to either fully fail the concrete in compression, or fully fail the steel section in tension (whichever is the smaller force). Fewer connectors may be used, resulting in so called partial shear connection, if the applied loading is at a low enough level, for example the common case where a beam design is governed by construction stage or serviceability considerations. However, codes also specify a certain minimum degree of connection that is needed to prevent excessive slip between the steel and concrete, which would result in failure of the connectors.
BS 5950-3-1, which was written in the 1980s, took a fairly simplistic approach to the issue of minimum degree of shear connection. BS EN 1994 recognises two additional parameters that influence this minimum degree, namely steel grade and the effect of asymmetry when one of the beam flanges is larger than the other (a smaller top flange is often used as the concrete carries most of the compression, but such asymmetry places higher demands, in terms of slip capacity, on the shear studs). For S275 steel and symmetric sections the limits in BS EN 1994 are considerably less onerous than those found in BS5950. For asymmetric beams they are considerably more onerous. Even BS EN 1994 fails to recognise the considerable benefits when the beam is unpropped during construction, as most are. NCCI produced by SCI (Pn002a ) allows the minimum degree of connection to be relaxed when a beam is unpropped.
The benefit of joining the steel and concrete together structurally is to increase the resistance of the steel beam alone by around a factor of two. The stiffness may increase by up to a factor of three. The relative benefits decrease with span, as the size of the steel beam increases relative to the size of the slab.
The components of a composite beam are as described above, but the same principles apply to composite slabs and composite columns. A slab uses profiled steel decking in place of a steel section, and force is transferred via embossments and certain aspects of the deck geometry (rather than discrete shear studs).
A composite column may be either a hollow section steel tube filled with concrete, or a steel section encased in concrete. Force is transferred between the two materials by friction and, where needed, discrete mechanical connectors, including shear studs that may be attached to an embedded steel section.
With all forms of composite construction it is important for the designer not to forget the construction stage. Assuming that there is no temporary propping, the steel part of a composite cross section must alone resist self weight and other construction loads as the concrete at that stage is ineffective. Not only is the resistance less, but there may be instability phenomena to consider. When acting compositely the top flange of a steel beam is restrained laterally by the slab, but during construction lateral torsional buckling (LTB) may reduce the effective resistance - only when the decking runs transversely and is properly fixed does it prevent LTB - further guidance is available covering both design, in SCI P359 , and detailing, in SCI P300.
Types of composite beam
Three general types of composite beam are considered below. The drivers that are relevant to a particular project will affect which flooring system is the most appropriate.
The most common type of composite beam is one where, as with a traditional non-composite steel framed solution, the concrete slab sits on top of the top flange of the steel beam. This clearly means that the soffit is interrupted at beam locations by a 'downstand'. The effective span range for this type of solution is around 6 to 12 m, which therefore makes it a competitor to a number of concrete flooring options.
The slab itself is typically either composite (in-situ concrete cast on profiled steel decking) or formed from precast units. Particular detailing is required for the shear connection when precast units are used, so that the body of the precast units can be mobilised as part of the concrete compression flange. See SCI P287 for more information.
Composite slabs offer a number of advantages - the decking acts as external reinforcement at the composite stage, and during the construction stage as formwork and a working platform. It may also provide lateral restraint to the beams during construction.
The decking is lifted into place in bundles, which are then distributed across the floor area by hand. This dramatically reduces the crane lifts when compared with a precast based alternative.
A number of variations on the idea of downstand beams are available to meet long-span needs. They provide the opportunity to achieve longer spans (20 m or more) than are possible using a 'standard' solid web, rolled downstand beam.
Shallow floor solutions
Shallow floors offer a range of benefits , which must be considered in the context of a given project to identify when they are most appropriate.
A number of shallow floor solutions are available, including a range of rolled and fabricated options from Tata Steel. The unique ASB - Asymmetric Slimflor Beam - is rolled with a wider bottom flange than that found at the top (see SCI ED003 for detailed design guidance). This geometrical form is common to all shallow floor solutions, as it enables the slab to sit on the upper surface of the bottom flange - rather than the upper surface of the top flange as found with downstand beams. Tata Steel also produces fabricated beams based on open and hollow sections (SFB and RHSFB) with a plate welded to the lower flange to achieve this geometry.
The slab may be either in-situ concrete on steel decking, or precast concrete. Because it sits on the bottom flange a key consideration is torsion applied to the beam. For internal beams this will mainly be a concern during construction, requiring appropriate sequencing to minimise torsion. Edge beams must be designed to carry torsion in the final state, making the RHSFB a good option given the torsional resistance and stiffness of hollow sections.
The figure right shows an ASB with composite slab. The shallowness of the floors is achieved by placing the slabs and beams within the same zone. An added benefit is that a flat soffit is achieved - there are none of the interruptions found with downstand beams. Encasing the steel sections within the slab also has benefits in terms of fire performance, with (often) no need to use added fire protection.
Shallow floor solutions are valid for a span range of around 4 to 9 m. Some solutions achieve composite action between the steel beams and concrete by using discrete mechanical shear connectors - much like a downstand beam - whilst others achieve this primarily by wedging the concrete between the steel flanges. For ASBs the composite action is enhanced by the presence of a profiled pattern rolled into the upper surface of the top flange. Friction then prevents interface slip as the beam is loaded and deflects, providing a reliable form of shear connection to assure the strength and stiffness of the composite element.
A generic design method for shallow floor solutions is not currently given in BS EN 1994-1-1 However, extensive guidance and software (ED003, SIDS ) is available for all of Tata Steel's shallow floor solutions, based on extensive testing (to satisfy both British Standard and Eurocode requirements) of their specific products.
Composite slabs comprise reinforced concrete cast on top of profiled steel decking, which acts as formwork during construction and external reinforcement at the final stage. The decking may be either re-entrant or trapezoidal, as shown in the figures below. Trapezoidal decking may be over 200 mm deep, in which case it is known as deep decking. Additional reinforcing bars may be placed in the decking troughs, particularly for deep decking, or to meet fire design requirements (such bars are more effective than the decking in the fire condition because they are insulated within the concrete).
The figure below shows the geometry of a typical 80 mm trapezoidal deck. The steel is galvanized and around 1 mm thick - hence the need for stiffeners to avoid local buckling when it is acting as a bare steel section to support the wet weight of concrete and other construction loads. The re-entrant stiffener shown at the top of the decking not only stiffens the upper flange but can also be used to support hangers for relatively lightweight items suspended from the soffit. Interlock is achieved through embossments (dimples) that are rolled into the decking profile, and by trapping the concrete around the re-entrant parts of the profile. There are no standard decking profiles, so the interaction achieved by the embossments, etc of each propriety deck is different. It is determined by tests undertaken by the deck manufacturer.
The results of such tests have traditionally been translated into so-called m and k empirical constants that define the performance of a particular deck. BS EN 1994 also includes an option to determine the shear bond per unit area of slab (τ), which can then be used as part of a more sophisticated approach (the τ value is analogous to the resistance of a shear stud). Designers obtain the relevant information (implicitly) from software or brochures provided by the decking manufacturers.
The profiled decking is often designed to be continuous over two spans when acting as formwork. Composite slabs are normally designed to be simple spanning at room temperature, but continuous under fire conditions. This continuity is achieved thanks to nominal reinforcement, which also fulfils other roles such as crack control, that continues over intermediate supports (its influence - assumed to be beneficial - is ignored for room temperature design).
Re-entrant or trapezoidal decking of 50 to 60 mm depth can span around 3 m unpropped, 80 mm deep trapezoidal profiles can span up to around 4.5 m unpropped, and deep decking can achieve around 6 m. Overall slab depths range from 130 mm upwards. Two hours fire resistance can be achieved without the need to fire protect the steel decking.
It is possible to form openings in composite slabs, although this should be planned and the openings formed at the construction stage rather than having to cut out concrete. Openings up to 300 mm square require no additional provisions, those up to 700 mm require additional reinforcement locally around the opening, and those in excess of 700 mm require the use of trimming steel to support the opening.
Composite columns may take a range of forms, as shown in the figure below. As with all composite elements they are attractive because they play to the relative strengths of both steel and concrete. This can result in a high resistance for a relatively small cross sectional area, thereby maximising useable floor space. They also exhibit particularly good performance in fire conditions.
Design rules for composite columns in structural frames are given in BS EN 1994-1-1. This is the first time that guidance has been given in a code for use in the UK, which may explain why composite columns have been rarely used to date. Rules are provided for composite H sections, either fully or 'partially encased' (web infill only), and for concrete filled hollow sections. Typical cross sections are shown in the Figure left. Composite columns requiring formwork during execution tend not to be viewed as cost-effective in the UK.
Concrete filled hollow section compression members need no formwork and they use material more efficiently than an equivalent H section. Concrete infill adds significantly to the compression resistance of the bare steel section by sharing the load and preventing the steel from buckling locally. The gain in fire resistance may be at least as valuable, especially if it permits the column to be left unprotected or only lightly protected. Infill concrete retains free water which in other situations would be lost; its latent heat of evaporation significantly delays temperature rise.
Rectangular and circular hollow sections can be used. Rectangular sections have the advantage of flat faces for end plate beam-to-column connections (using Flowdrill or Hollo-bolt connections). Ordinary fin plates can be employed with either shape.
A programme, FireSoft, for the design of concrete filled hollow sections in ambient and fire conditions has been developed.
Although design guidance exists for composite connections (SCI P213 ), they have been very little used in the UK (or indeed elsewhere in Europe). In theory they appear to be rather attractive, as slab reinforcement can be used to avoid the need to add to the steelwork connection, for example with extra rows of bolts in an extended end plate. However, it is difficult to achieve the correct detailing for composite connections, because the needs for strength, stiffness and ductility can border on the mutually exclusive - too little reinforcement will reduce connection ductility (rotation capacity) because of potential rebar failure, too much will reduce ductility because of concrete crushing failure.
In an effort to overcome some of the practical issues, so that the inherently attractive features of composite connections can be more widely exploited, research work is on-going in Europe and may result in the inclusion of specific guidance in a revised version of BS EN 1994-1-1 planned for around 2018.
- ^ 1.0 1.1 1.2 1.3 BS 5950-3-1: 1990+A1:2010 Structural use of steelwork in building . Design in composite construction. Code of practice for design of simple and continuous composite beams. BSI
- ^ BS 5950-4: 1994 Structural use of steelwork in building. Code of practice for design of composite slabs with profiled steel sheeting. BSI
- ^ BS 5950-6: 1995 Structural use of steelwork in building Part 6. Code of practice for design of light gauge profiled steel sheeting. BSI
- ^ 4.0 4.1 4.2 4.3 4.4 4.5 4.6 4.7 4.8 4.9 BS EN 1994-1-1: 2004 Eurocode 4. Design of composite steel and concrete structures. General rules and rules for buildings. BSI
- ^ BS EN 1994-1-2: 2005: Eurocode 4. Design of composite steel and concrete structures. General rules, Structural fire design. BSI
- Steel Designers' Manual 7th Edition. Editors B Davison & G W Owens. The Steel Construction Institute 2012, Chapters 21, 22 and 23
- Johnson R.P, Composite structures of steel and concrete, volume 1 2004 Blackwell Scientific Press.
- Johnson R.P, Designers' guide to Eurocode 4 Design of Composite Buildings, 2nd edition. ICE.
- Nethercot, D. Composite Construction. Spon Press.
- SCI P300, Composite Slabs and Beams using Steel Decking: Best Practice for Design and Construction, (Revised Edition), 2009
- SCI P359, Composite Design of Steel Framed Buildings, 2011
- SCI P287, Design of Composite Beams using Precast Concrete, 2003
- PN002a, NCCI: Modified limitation on partial shear connection in beams for buildings SCI
- SCI ED003, Design of Asymmetric Slimflor Beams to Eurocodes, anticipated 2014
- SIDS2 - Slim Flor Integrated Design Software
- Slimdek® residential pattern book - For multi-storey residential buildings, 2013, Tata Steel
- SCI P365, Steel Building Design: Medium Rise Braced Frames, 2009
- SCI P375, Fire Resistance Design of Steel Framed Buildings, 2012
Member design tools:
- Steel construction products
- Steel material properties
- Floor systems
- Long-span beams
- Design codes and standards
- Simple connections
- Moment resisting connections
- Design of composite steel deck floors for fire