Skip links

Wrong In Several Ways

I’ve written before about wood bowstring trusses (good, bad, and weird) but I did not have this patent in front of me when I did, and it raises a few more questions. It plays a bit with the timeline of the US origin bowstring trusses and it brings up, again, earlier work in Ireland.

Since it’s well established that bowstring trusses were used in Ireland as early as the mid-1800s, should this patent have been issued at all? I don’t know if there were built examples of the Irish-style bowstring – called Belfast trusses – in the US before 1907, but claiming something meets the criteria for patenting when it’s well known in another country is, at best, weak. The big difference between Belfast trusses and American bowstring trusses is the geometry of the lattice webs: Belfast trusses maintain a constant angle between the diagonals in the lattice and the curved upper chords, while the US examples typically use the simpler geometry of having the diagonals maintain a constant angle with the straight lower chord. If you look at the picture above you can see that the geometry follows the Belfast pattern. The text of the patent talks a great deal about the geometry as if it were new, but given that John McKeown was one of the owners of a reasonably prominent Chicago construction company, McKeown Brothers, its hard to believe he did not know of the Irish examples.

There is an innovation in the drawing, but it was abandoned by McKeown for something better. Figure 2 is a section through the upper chord, showing the web diagonals E extending past the multiple pieces of wood (the 1s and 2s) making up the chord. The purlins of the roof (B) are supported on the top chord but also wedged between the ends of the diagonals, which is a bit of cleverness that is difficult to build because it requires the geometry of the lattice to be fairly precise. The innovation was having the chord members interlock by means of square corrugations (the 3s and 4s) cut into their sides. That works great in theory but would be an enormous pain to build, especially for the curved top chord. The McKeown company is known today for being among the early adopters of glue laminating, specifically for this purpose, which is a much better idea. The bowstrings I’ve seen in New York, not built by McKeown, have the chords laminated with nails or bolts.

I’ve never seen a documented first use of bowstrings in the US, probably because no one knows. Most sources talk vaguely about “shortly after 1900” which is a reasonably good description for 1907. Maybe this patent marks the beginning of the type here. Given the generally bad history of this truss type in this country, it’s unfortunate that the design aspects weren’t taken more seriously.