The IND subway at 59th Street, constructed circa 1930. The mosaic on the wall isn’t damaged, but rather is supposed to look like that, which might be artistic but can be difficult to read when you’re on a train and looking at it diagonally through a window not at the right height…
The black-painted vaulted concrete is above the track and below the station roof, which supports earth and then the roadbed of Eighth Avenue. The off-white painted concrete is above the platform and below the mezzanine that is the main circulation space in the station, with stairs up the street and down to the three platforms. It’s all steel framed, but the steel for the mezzanine and the roof is encased in concrete. I don’t know about this station specifically, but where I’ve seen the original drawings for station structures, the vaulted concrete is shaped like a true vault but also reinforced. If you were to remove the tile finish on the left, which is the outer wall of the station, you’d see a steel column lined up at each of the beams above, and concrete vaults between them. Also, the columns in the wall and the beams in the roof are probably five feet in center. I didn’t check here, but that’s the standard because it makes figuring out where you are along the line very easy: you count by fives from the nearest marked point.
The main design issue here is earth pressure. I don’t know exactly how much earth is on top of the roof, but it’s probably something like 5 to 10 feet thick, or 500 to 1000 pounds per square foot of loading, plus the theoretical 300 psf of loading from traffic above (if every inch of the avenue were to be covered by heavy trucks, which does not happen even if it sometimes feels like it does). The side walls have to be designed for full earth pressure even though most of the wall area is facing cellars. In short, everything is pushing in on all sides on this buried box.
The reason I took the picture, the thing that caught my eye, is the series of built-up beams over the tracks. They’re in line with the platform columns, at fifteen feet on center, and are carrying nothing. There are two possibilities but one seems to me to be much more likely than the other. The first is that they are compression struts to brace the exterior wall. They would carry load from the wall through the mezzanine floor structure, then through another set of struts on the other side to the opposite wall. That’s a perfectly legitimate form of bracing, but the location of the beams argues against it. Putting such a strut at every third wall column either means only one-third of the wall columns are braced in that manner, which is nonsensical, or it means that there’s a hidden distribution beam running horizontally at the height of the struts, which is possible but simply not the way NYC’s subways structures were designed. They’re caterpillars, with every horizontal segment the same.
The other possibility is that the struts are there to be sure that the mezzanine structure is braced against lateral movement. The platform columns are oriented with their weak axis in that direction, so even if they can take the load (which they probably can) they might be flexible enough to allow a bit of sway. The mezzanine is loaded by people, often moving in one direction (when one or more crowded trains let out) and possible loaded heavily. If you look at an extreme example – the mezzanine loaded to capacity and everyone starts running (a fire panic, for example), it’s pretty easy to see significant lateral load on the mezzanine, and any perceptible sideways movement would be bad.
The struts brace the platform columns against buckling in their weak-axis direction, but the mezzanine floor does that as well, so it’s hard to see the struts being added simply to perform a redundant function. On the other hand, another argument for them as stability braces (the second possibility) rather than carrying significant load (the first possibility) is how slender they are. A built-up I section with a trussed web is not an efficient compression strut, but it’s fine for small loads.