Sometimes I go looking for a specific building or building type in research and serendipitously come across a structure that I had not heard of and wish I’d seen sooner. An example is the small truss bridge above, which spans a railroad line and carries a small local road in rural Pennsylvania. It was moved to its current location in the late 1800s by the Reading Railroad, which had built it in 1846. It is among the earliest bridges in the United States with all-metal construction, consisting of cast and wrought iron in roughly equal quantities.
How early is 1846 with regard to structural engineering in the US? Squire Whipple’s book on bridges, which is the first home-grown text to have a correct and useable method of analyzing trusses, was a year away from publication. Herman Haupt‘s book, which is the first home-grown text to have an organized, readily-comprehensible, correct, and useable method of analyzing trusses, was five years away from publication. There were, of course, texts available in the US that were imported from Europe, many of which were correct and usable, but the lack of American engineering literature gives the correct impression about the primitive state of the profession at the time.
The form of the bridge is a pair of through pony trusses: the deck is approximately at the level of the truss bottom chords and the truss tops are free-standing rather than being connected by horizontal-plane bracing. The description calls these Howe trusses, which is to say trusses with compression diagonals, but they have double diagonals in each panel, so it seems like double-Warren is just as accurate a description. But enough of categorization…let’s look at some beauty. First, the cast-iron compression members have “Egyptian revival” ornament:
The HAER measured drawings give a sense of how well designed this bridge really is:
1846 was also before wrought iron I-beams were being commercially produced by mills, which is why rail sections were used. The note on the left elevation about the centerlines not meeting at a point is shown to be true in the drawing, but the secondary bending stresses created by that mistake are in the wrought-iron top and bottom chords rather than in the cast-iron, so the effect of the mistake is limited. I have to wonder if that was intentional.
Finally, the elegance of the panel-point connections, showing the kind of connections that sand-casting allowed:
The full HAER survey, which is worth a read, is here.