The Hauptbahnhof (Main Train Station) of Leipzig is not only interesting due to its monumental dimensions, but also due to a number of smaller technological advances that were made during its construction. These advances were crucial in setting the stage for the later development of reinforced concrete structural design. This structure was one proving ground for the design of the Centennial Hall in Breslau.
The City of Leipzig and German National Railway officials had decided to concentrate the six existing Leipzig train stations and their yards, some of which had been in existence since the first railway lines in Germany were lain, into one large station. They chose the site of the existing Dresden, Magdeburg and Thüringen Stations in the center of the city. The razing of part of the existing buildings and earthwork was begun in 1897(1). 96 entrees were submitted to a very rigid competition that was held in 1906 for the design of the new train station. The winner was the team of Architects Professor William Lossow & Max Hans Kühne of Dresden(2). Their design foresaw an Entrance Building which would house the administration, restaurants, bars, hotels, etc. with a monumental entrance vault nestled between the building and the train platforms.
The vault of the entrance hall spanned 34 m, had a projected total length of 273 m and height of 29 m. Its axis was perpendicular to the 26 railroad tracks and 203 m long platforms. The competition jury was especially intrigued by this entrance hall and noted that "it decisively promised a powerful impression and will most likely become the symbol of the new train station." The jury is also noted to have stated implicitly that the first prize would be given to the best project proposed in reinforced concrete(3). There is no mention of the reason why.
The commission was given to the architects and the project forged ahead. All further decisions were made by a committee composed of the architects, Lossow & KŸhne (for their "artistic consultation") Dyckerhoff & Widmann A.G. of Dresden (structural designers and contractors), and representatives of the two State Railway Commissions (Preussische and SŠchsische). The final decision to build the Entrance hall of concrete had not yet been made, but with Dyckerhoff & Widmann on the above mentioned committee, it was obvious which material would be chosen. According to the director of Dyckerhoff & Widmann, Regierungs-Baumeister Gehler(4), the reason that concrete was chosen for the entrance hall was not that he was on the committee, rather that,
"...there had not yet been a successful design which joined the massive stone buildings to the front and the light steel-glass construction of the platform roofs towards the tracks. Here we see the first attempt to design this transition with a material that is composed of the two: stone(concrete) and steel."According to F. Kögler, the choice of reinforced concrete was made based on three issues: first, "...(reinforced concrete) gives the impression of mass, but is able to be articulated in an airy and logical manner..."; second, the choice of steel would not have coincided with the design concept of the architect; and third, the decision had already been made to build the platform roofs of steel, so that equity called for the second major structure to be of another material. It is very interesting to note that the choice of structure for one of the largest halls in the world at that time could have been so politically motivated. The construction of the Preussische part of the station was completed by Dyckerhoff & Widmann (with the design engineer being Gehler, the same engineer of the Jahrhunderthalle in Breslau) on the 1st of December, 1911 and opened to service on the 1st of May, 1912. The Sächsische part of the station was constructed by Max Pommer & Rudolphe Wolle, both of Leipzig, and dedicated on the 1st of October, 1915.
The monumentally vaulted space of the entrance hall was articulated into six segments which reflected the septil division of the train platform arched roof structure beyond it. Each of these segments was illuminated by glazed skylights which covered one third of the ceiling surface. The vaults spanned between the last arch of the 203 m long platform roof and the entrance building. Due to the absurdity of resting a massive concrete vault on a steel arch, the first, or last, arch of the platform roof was of reinforced concrete. This unique row of arches will be called the End Arches.
The structural system of the hall is not simple; it is between a simple linear frame and a T-Beam. The obvious advantages of the implied structural system were negated by the spatial requirements. The most obvious structural system would have been a stiff frame in which its vertical members would have continued to grade. This system could not be applied due to the differential heights of the base hinges and the resulting difficulty in the transfer of the horizontal shear forces from the frames loading the End Arches perpendicular to their axis. Additionally, it was noted(5) that the dictums of the architectural requirements on the dimensions of the vertical members, which intruded into the entrance building, would have restricted size of the openings in the hall sides. The next logical choice of structural system would have been simple arches. These arches would have had to span from the roof of the entrance building to the end arches. Unfortunately, this was impossible because the walls of the entrance buildings were not designed to resist the magnitude of horizontal loading that the end reaction of an arch would create. Also, the properties of the soil, partially infill, were such that one could not eliminate the possibility of differential settlement. Thus, "left with no other alternative" and with a very heavy heart(6) the idea of a frame construction was left behind in favor of a simple beam system with a very strong curve to them. The fixed support is on the end arches and the free support on the buildings. The free-roller hinges were a type known from bridge construction. However, the enormous load that the main beams were to resist dictated that their hinges be replaced with concrete "spindles."
The 34.675 m vault was spanned with two types of simple beams: the main beams, of which there were seven, and the intermediate beams, the four between the main beams. The only difference between the two was their height and the elevation of their supports. The main beams had a total height of 5.55 m and extended from the exterior skin to project into the space below, creating the segments of the hall. The intermediate beams had a total height of 3.83 m and filled the In-Between. They were given the name of Plattenbalken (this would be "plate-beams" in a direct translation from German) but were really no more than I-beams with an enlarged compression zone. (Later descriptions of the building no longer used this term in describing these structural elements.) A true Plattenbalken would not have the compression flange separated from the slab spanning between the beams, but would have been an integral part of that slab. The compression zone is the part of the slab which is arbitrarily defined by the structural designer in order to calculate the load bearing capacity of the element. Due to the fact that the beams had a large spacing (9.00 m), and wide span (34.675 m), they were very heavily loaded. The depth to which either beam type could penetrate the space of the hall was dictated by the architects. The height to which the beams could extend above this limit was dictated by the city building codes. That is, the structural designers were totally limited in their options as to the size of the beams.
In order to increase section efficiency, the compression and tension zones of a beam should be separated as far as possible from each other. (The form of the beams would suggest a direct relationship to the well known, and very efficient, steel I-beam.) In this way, one reduces the self-weight of the element as much as possible. It is interesting to note the way in which the section thickness is increased towards the supports. It is only 20 cm thick at the pinnacle, and increases in thickness to 25 cm and then to 30 cm in the field directly next to the support. This is a clear response to the increase in shear, but also reminds one of the built up plate-steel sections of the last century. In order to prevent the web from buckling, stiffening ribs with a thickness of 55 cm were placed every 2.4 m along the length of the beam(7). The structural design of the beams was based on known concepts transferred from well known steel bridge constructions for this span. In fact, the on-site control of the construction was given to the Bridge Division of the Office of the National Railroads.
Due to the fact that this was indeed a new dimension for reinforced concrete, the three contractors (Dyckerhoff & Widmann, Max Pommer, and Rudolphe Wolle) built a 1:1 model of the main beam and tested it with loadings that were 1.5 times those required by the building code. These tests commenced in May of 1910 and were completed in July of the same year. The beam was rammed continuously for two days and taken out of its formwork only after 45 days. The beam was loaded to failure and the results compared to the projected strengths. The first conclusion from the experiment was that the calculated and observed stresses in the concrete and steel did not coincide at all. This did not seem to bother the commission due to the fact that the tests proved that with very little deformation the beams could carry a significantly greater load than required. The second conclusion was the result of observations of the movement of the supports. The movement of the rolling support was observed and noted so that when the plate of that support was to be laid in the concrete, it could be adjusted so that the walls of the entrance building would have to carry a minimal horizontal load. Thirdly, more care had to be taken not only in the construction of the formwork, but also during construction. It was recommended that the ends of the beams should be cabled together until the formwork was removed in order to reduce formwork deformation, and that Australian hardwoods be used in the supporting structure instead of the local pine(8). Lastly, the most important result of the tests was the observation of the crack pattern on the beam. Due to the distribution of the cracks, it was observed that the shear in the beam was much higher than was calculated. Due to this, the amount of shear reinforcement was increased, and it was decided to use a greater number of smaller diameter bars of a better quality of steel.
Indeed, a considerable amount of thought was dedicated to the design and placement of the reinforcement. Great care was taken at the site to bend the steel to a maximum radius of 17d, (d being the diameter of the bar) and bar connections were made with screw-on mufflers instead of overlapping. In addition to this, Dyckerhoff & Widmann used a machine to bend the steel reinforcing bars for the first time.(9)
The infill of the vaults which was carried by the beams also deserve some attention. There were two different types, the weather shell (defining the outside) and the coffered interior shell (defining the inside). Two layers were chosen for two reasons: first, so that there would be an insulating volume of air between the interior and exterior; and second, so that the full depth of the main beams would not be seen. If the full 5.55 m had been exposed, the space would have been distinctly cut into six separate vaulted rooms instead of the slightly articulated continuous space that it is.
The exterior skin was planned to be of a light-weight concrete (Bimsbeton), but due to the weight that was saved by the I-shape of the beams, it was fabricated of normal concrete. The secondary structure between the beams was steel. These were encased in concrete and a 4.5 cm thick plate was rammed in place on top of them. The interior skin was also a mixture of systems. A secondary structure of steel I-Beams was laid between the beams. The lower flanges of these were encased in concrete to the point that they created a support rib for prefabricated 1.75 m square rammed concrete panels. The panels were fabricated at the factory of Dyckerhoff & Widmann in Cossebaude to a thickness varying between 2.5 and 8.5 cm.
The End-Arches provided the transition between the steel and glass construction of the platform roofs and the monumental vaults of the main hall. Due to the fact that the soil was unstable, the arches were designed to be statically determinate. The form of the arches was determined to be exactly the catenary of its greatest loading condition. These arches span 45 m with their thickness varying from 1.25 to 0.75 m, and having a continuous width of 4.00 m. The hinges were 5 mm thick, soft lead squares which were already set in place as the arch was rammed. Each arch took five days to complete. These arches provided the opportunity to test a structural system that was again to be seen in the beautiful hall to be built by Dyckerhoff & Widmann, in Breslau, a few years later.
The train station at Leipzig was built so that expensive land in the center of the city could be released to be used for other purposes. The city could realize an economic gain in consolidating the various stations into one. The hall is another example of the form being dictated by the chosen architectural expression. The monumentality was "guaranteed" with the coffered ceilings. The structural design was inhibited not only by the less than ideal functional prerequisites of the various parts of the station, but also by the choice of site. The structural solution was on one hand simple, but on the other clever and unique. The choice of the simple beams to give the visual impression of act "vaults" is a total contradiction to the way the elements carry their loads. The structural design entered a realm of dimensions not yet tested by contemporary reinforced concrete technology. This monumental construction was a proving ground for not only structural design, but also for construction methodology.