Time to revisit some details I omitted in an early blog post. This spring, I talked a bit about the construction sequencing of the Hell Gate Bridge, following up on some pictures I took while crossing it two summers ago. It seems that I haven’t exhausted the topic. The picture above, part of the HAER survey, is a good one: it shows the heavy steel of the bridge, the houses and small apartment buildings of Queens in the foreground, the Queens leg of the Triboro Bridge in the mid-ground, and the midtown-Manhattan skyscraper farm off in the distance.
Hell Gate is a nasty place to build a bridge: the currents in the East River are fast and (since it’s a tidal strait rather than an actual river) reverse direction four times per day; the right-angle bend in the river at this location, combined with the connection to the Harlem River (another tidal strait) and the presence of two islands and a number of smaller rocks mean that there are whirlpools and cross-currents. The river is also deep and, while not as wide as it is further south, not narrow. So using any kind of falsework in the river is not realistic.
The answer for the Hell Gate Bridge was to cantilever the two halves of the arch until they met and became self-supporting. I used the Pennell etching below before, but it’s too good to not use to illustrate the point:
In general, it shouldn’t come as a surprise that the temporary cantilevers were as meticulously planned as the permanent arch – after all, that’s what engineering is about. Given that the engineers were Gustave Lindenthal and Othmar Ammann, both contenders for the title of best American bridge engineer of their respective generations, that meticulous work was well-documented. All of the following illustrations are taken from a paper that Ammann published in the Transactions of the American Society of Civil Engineers in 1918. First, the photographic equivalent of Pennell’s art, showing the arch construction in progress, supported by the temporary cantilever. Note that “Long Island side” means Queens, and Wards Island is one of the two islands in the river at Hell Gate as well as the west end of the arch bridge. An elevated viaduct and two other bridges carry the tracks the rest of the way to the Bronx. Note that the temporary cantilever steel is considerably lighter than the arch: it only had to carry the dead weight of the bridge, the traveller cranes, and wind load, not freight trains.
Here’s the end of the arch construction and the beginning of the deck construction. Note the staged disassembly of the temporary steel.
Ammann used a technique that is fairly common when engineers want to show changes over time: a series of cartoons of his structure at different stages. The two stages at the top right and middle right, for example, coincide with the November and January photos above.
There’s a big issue hidden in the two lower diagrams in the middle column and Ammann spent some time explaining it. The two halves of the arch had to meet exactly to make the geometry of the arched truss work properly, but it was difficult to count on construction over months on opposite sides of a river being that exact. The solution was to construct each half-arch above its final position and lower them to meet once they were both completed. The calculations for this process had to include the deflection of the half-arches and the piggy-backed cantilever members: the elongation of the cantilever back-stays under load, for example. In the diagrams below, we see the nominal position of the steel, the deflected (loaded) positions, and the lowered positions:
Finally, the tension in the long diagonal backstays at the top of the temporary cantilevers caused considerable uplift at their ends. Tension foundations were difficult to build in that era, so the solution was to simply put big counterweights on top of the backstay ends. The top picture below shows the counterweight, the lower picture shows the very beginning of the construction sequence: