On The Evolution Of Steel-Concrete Composite Construction

Little has been written so far about the historical development of the joining of rigid elements made from steel and concrete – steel-concrete composite construction. This article therefore describes the individual phases in such a way that the reader gains an overview of the evolution of composite construction with steel and concrete. The composite beam in the sense of Emperger’s ideal of the friction/adhesion bond between the materials marked the initial phase (1850–1900). This was followed by the constitution phase (1900–1925) with its constructional separation of the elements of the cross-section. During the establishment phase (1925–1950) it was gradually realized that the elements of the cross-section had to be connected structurally, initially as positional restraint, later as mechanical shear connector. The quantified connection of the elements of the cross-section through standardized testing and the formation of theories in the classical phase (1950–1975) enabled the realization of multiple forms of steel-concrete composite construction for industrial buildings and bridges. Figure 1: Steel beams with welded bar connectors for the new Herdecke Bridge (1951) over the River Ruhr (source: German Federal Ministry of Transport & Digital Infrastructure) 1 Hessen Mobil, Roads & Traffic Management, Wiesbaden, Germany, post@mobil.hessen.de 2 Ernst & Sohn, Berlin, Germany, karl-eugen.kurrer@wiley.com On The Evolution Of Steel-Concrete Composite Construction 5 International Congress on Construction History INITIAL PHASE (1850–1900): THE COMPOSITE BEAM The first structures with iron sections and concrete preceded the emergence of reinforced concrete with non-rigid round bars as propagated by Monier (Bracher 1949). As early as 1808, Ralph Dodd (1756–1822) was granted a patent for suspended floors: “malleable iron” tubes with “ears or flanges” were filled with “artificial stone” to form a composite beam. The floor patent of James Frost already includes “iron ribs” with the intermediate “compartments” filled with “cement”. The suspended floor of Nathaniel Beardmore (1816–1872), patented in 1848, uses riveted I-beams with concrete in the intermediate bays and permanent iron formwork. The “fireproof” floors of Henry Hawes Fox were successful; they were patented in 1844 and marketed by James Barrett, and employed cast iron, from about 1851 wrought iron, upturned Tbeams or I-beams (Fig. 2). Fox shifted the iron beams to the tension zone of the floor (Hurst 2001), and sometimes placed them outside the concrete of the floor itself: “The force of compression acts upon the joists, only through the medium of the concrete.” Fox and Barrett were aware of the underlying stress flows in their floors. This aspect was perfected in 1873 by William E. Ward in the suspended floors of his “Ward Castle”, which are in the form of a Tbeam with rigid tension flange, albeit still encased in concrete. Figure 2: Suspended floor with cast iron joists by Fox & Barrett, pre-1851 (Hurst 2001) A contemporary state of the art report by Paul Christophe (1902) shows a number of slab and beam-and-slab systems with steel sections available on the European market at that time. For example, the Steinbalkenkonstruktion of Fritz Pohlmann, formed by a T-beam over a perforated, asymmetric rolled section, the iron bulb beam, with flat iron hoops forming a shear-resistant connection to the concrete compression zone (Emperger 1904) (Fig. 3). Another example is Mathias Koenen’s (1849–1924) flat-soffit, later (from about 1892) ribbed, floor in which the underlying steel sections carry the tension and the arching concrete infill sections are solely responsible for carrying the compression. Despite the lack of shear connectors, tests confirmed Koenen’s assumptions regarding the structural behaviour (Christophe 1902). E. Pelke, K.-E. Kurrer 5 International Congress on Construction History Figure 3: The Pohlmann floor, 1901 (Emperger 1904) CONSTITUTION PHASE (1900–1925): CONSTRUCTIONAL SEPARATION OF THE ELEMENTS OF THE CROSS-SECTION As soon as reinforced concrete started to be used as an engineering material in structures (a date we shall take as 1886, when Mathias Koenen’s design equation for reinforced concrete slabs was published), so it was regarded as a composite material. No distinction was made between non-rigid and rigid reinforcement; instead, the magnitude of the composite action was the important issue. Consequently, both types of reinforcement had to be fully encased in concrete. Extensive tests between 1907 and 1909 by Carl von Bach (1846–1931) at the MaterialsTesting Institute in Stuttgart (Emperger 1912) resulted in a much smaller “resistance to slip” for steel sections compared with non-rigid round bars. In addition, at the onset of movement, the steel sections burst open the concrete. In a brief submission, Koenen, too, drew attention to “the dangerous shearing-off behaviour” of steel sections encased in concrete. These concerns of von Bach and Koenen may suffice as the first indications of the mutual displacement of the different rigid elements. Both recognized that bond is a poor way of transferring load between rigid concrete and steel sections. Notwithstanding, Emperger (1912) specified the same design rules for both types of reinforcement. However, he did mention that rigid reinforcement is less consistent with the nature of reinforced concrete and the strains in both materials influence each other – a fact that unfortunately had to be left out of the design proposal. Hager was more specific (1916); he explained that the interaction of large steel beam sections in slabs, encased in concrete, could only be clarified by carrying out tests. So the ultimate load model remained within the confines of classic steel construction, assigning the longitudinal load-carrying capacity solely to the steel sections, the transverse load-carrying capacity to the unreinforced concrete. The higher neutral axis of the cast-in steel beam section permitted the permissible stresses in the steel to be increased by 10%. Around 1920 Rudolf Saliger (1873–1958) remarked that “particular measures presume the consistent action of concrete and rigid reinforcement”. (Saliger illustrated this with a flat steel bar riveted in place and bent up at 45°.) However, although this remark pointed the way forward, its significance was initially neglected (Saliger 1920). Immediately after reinforced concrete started to be used as an engineering material, Joseph Melan (1853–1941) was the first to realize that concrete and encased steel trusses contributed On The Evolution Of Steel-Concrete Composite Construction 5 International Congress on Construction History jointly to carrying the loads (Melan 1911) – initially for vaulted floors (1891), later in arch bridges in the mid-1890s in the USA. In his Bogenbrücken in Eisenbeton, Melan relieved the load on the concrete arch by using suspended centering such that between a third and half of the concreting load was carried permanently by rigid reinforcement. At about the same time, cast-in rolled beams started to replace pure steel bridges for beam bridges with short spans because this form of construction was more economic (Wolff 1907). Whereas Wolff (1907) wanted to see this form of construction optimised by using suspended centering, Kommerell (1911) no longer considered the composite action for his standard bridges introduced by ministerial decree of the Royal Prussian Railways and assigned the reinforced concrete solely a transverse distribution action. The load-carrying reserves were taken into account through a moderate increase in the permissible steel stresses. Kommerell does not explain the function of the transverse connections (round steel bars) shown in Fig. 4. Following tests some 25 years later, Combournac (1932) assigned the transverse bars “passing through the webs of the steel sections” the function of a shear connector. Figure 4: Typical section through railway bridge with rolled beams encased in concrete (Kommerell 1911) The Aqueduc de l’Eau Froide near Villneuve, Switzerland, built in 1901, still has the classic, concrete-encased main beam. Short steel angles – more positional restraints than shear connectors – are riveted to the steel beams. Following the traditions of Grubenmann timber bridges, they reveal that the Swiss engineer realized the importance of using separate connectors to join rigid components instead of relying on friction. Acheregg Bridge (1914) on Lake Lucerne in Switzerland was one of the first beam bridges in Europe to have a reinforced concrete deck that helps to carried the load through a friction bond (Rohn 1915). Its modern-looking beam-and-slab cross-section is made up of an approx. 23 cm deep reinforced concrete deck on top of two heavy, 800 mm deep rolled steel beams at a spacing of 3250 mm (Fig. 5). Lightweight steel cross-trusses link the rolled Differdinger “Greyträger 100 B” sections every 1.10 m. Both top flanges and about 20 cm of each web are cast into the underside of the reinforced concrete deck. 2 transverse links (∅20) reqd. for beams ≥ 40 cm deep min. height for timber sleepers min. height for steel sleepers Top of rail Top of rail Tr ac k ce nt re -li ne Tr ac k ce nt re -li ne Inner beam Outer beam Is ol at io n jo in t 30 m m m as tic a sp ha lt Is ol at io n jo in t Clay bricks Asphalt sacking E. Pelke, K.-E. Kurrer 5 International Congress on Construction History ESTABLISHMENT PHASE (1925–1950): STRUCTURAL CONNECTION BETWEEN THE ELEMENTS OF THE CROSS-SECTION – FROM POSITIONAL RESTRAINT TO SHEAR CONNECTOR Reinforced concrete slabs not cast monolithically with their supports require positional restraint; emergence of first constructional shear connectors. Developments in Europe In Europe the interaction of the two materials of the composite initially continued to focus on the friction/adhesion bond between them. The first significant measurements of this were carried out between 19