Historical overview

The principle of the network arch finds its origin in the year 1878. In this year the first arch bridge with crossing hangers was built, as shown in Figure 2 on an old postcard.

In 1926 the Danish engineer Octavius F. Nielsen patented the idea of a traditional arch bridge with hangers under an angle, and if necessary crossing each other, to create a net-like hanger arrangement. Around this period some 60 of these Nielsen-bridges have been built, but none of them had crossing hangers.

In the 1950’s Professor Per Tveit (Norway) developed the concept of the network arch when he was investigating the bending moment distribution in Nielsen-bridges. He suggested that the bending moments could be reduced when the hangers cross each other multiple times. He described his idea in an article that was published in the June issue of “The structural engineer”.

The force distribution of a (network) arch bridge can be compared to that of a simply supported beam. The arch and main girder take normal forces thereby acting as the flanges. The shear force is taken by the hangers that act as the web.

The main advantage of the network arrangement becomes clear in the load case “half load”, as is shown in Figure 4. The classical arch bridge shows large horizontal deformations. As a result of that a large number of hangers in the unloaded part will become relaxed. This has disastrous consequences for the moments and buckling lengths of the arch and main girder, also the vertical deflections will be large.

When the hangers are gradually inclined the horizontal deflection will be significantly lower because the hangers in the unloaded part of the span remain tensioned.

The Nielsen bridge (diagonal hanger configuration) also works by the same principle, except that the opposite hangers will become relaxed (see Figure 4, the dotted lines represent the hangers susceptible for compression).

To counteract the problem of relaxation the inclination of the hangers can be increased and more self-weight can be added. With more slanting hangers the distance between the hanger nodes will also increase, resulting in larger bending moments in main girder and arch. For an optimal diagonal hanger arrangement, there are always concessions to be made. Either you accept larger bending moments without relaxing hangers or vice versa.

When an extra set of diagonals is added (see Figure 4) the distance between the nodes becomes smaller thereby reducing the bending moments in arch and main girder. When another set of hangers is added the distance becomes even smaller. This process illustrates the principle behind the network arch bridge: by adding sets of diagonal hangers to the arrangement a network arrangement is obtained, which has overcome the disadvantages of a diagonal arrangement.

The first arch bridge with multiple crossing hangers was designed by Per Tveit and was built in 1963 in Steinkjer in Norway spanning 80m, see Figure 5.

In that same year another two network arches were constructed. The Bolstadstraumen bridge spanning 84m was also designed by Per Tveit and built in Norway (Figure 5). Also the Fehmarnsund bridge in Germany spanning 248m (Figure 6). The Fehmarnsund bridge is clearly a class bigger than the two Norwegian bridges, not only in span, but also in load
carrying capacity. This bridge accommodates two road lanes and a single railway track.

Until the 80’s no network arches are built in Europe. However, thanks to a Japanese professor who was involved in the design of the Fehmarnsund bridge, the idea travels across the globe to become popular in Japan. In 1968 the first Japanese network arch is constructed. In Japan this bridge is called Nielsen-Lohse bridge after the original inventors. The ‘Lohse’ part of the name applies to the principle of the tied arch. This principle was invented by a German railway engineer Hermann Lohse who developed the ‘Lohse-girder’. The Lohse-girder makes use of a tensile element in the deck to counteract the compression forces in the arch.

Since 1968 over 50 Nielsen-Lohse bridges have been built in Japan, see Figure 8 for some spectacular designs.

Worldwide, over a hundred network arch bridges have been built based on an overview of the existing network arches [www.network-arch.com]. In the Netherlands two network arches have recently been constructed. In August 2012, the first Dutch network arch was transported to its final location crossing the Twentekanaal near Zutphen. In April 2013, the Oversteek bridge spanning 285m was moved into position.

When looking at the year of construction in the overview of existing network arches it becomes clear that the network arch is gaining popularity. This shows that 30% of all the network arches are built in the last decade. Also the German Railway authorities (Deutsche Bahn) have recently adopted the network arch bridge in their ‘railway bridge design guide’ as an innovative alternative for classic arch bridges.

The rising popularity could be explained by the development of the engineering software. Because of the absence of computers in the early days of Nielsen-bridge engineering the hangers where placed under constant angle. This significantly simplified the calculation process. For this reason, the first Japanese network arches where built also with hangers under constant angles. 

For an overview of the span range for which the network arch bridges are mostly applied see Figure 9. This graph is based on roughly 70% of the network arches ever built, so the numbers aren’t accurate but show clearly the most popular range, and most popular traffic type.