The picture above is a storefront on the Upper West Side. The building in question was constructed in 1899 and is seven stories tall, which make it quite likely that the structure consists of masonry bearing walls and wood-joist floors. The remarkable thing about the picture is that the big riveted built-up plate girder supporting the wall above the storefront is bare except for paint. It used to be covered, most likely by a sign but maybe by an architectural fascia, as can be seen by strip of common red brick. The facade above and the pier to the right are covered in limestone veneer, so there’s no way that the red brick was supposed to be visible in the original design.
You have to get steel very, very hot before it melts and even hotter before it burns but, relatively speaking, it doesn’t have to be very hot at all to lose strength. Structural steel will start to lose strength pretty rapidly around 1000 degrees Fahrenheit, which sounds hot until you remember that the temperature of a ceiling above a building fire can easily reach 2000F. Steel has to be insulated from the heat or it’s going to fail badly and quickly in a fire.
The plate-girder was never fireproofed. The sign or fascia that used to hide it from view couldn’t possible (based on the geometry of the facade and storefront) have provided proper insulation. So the removal of that sign or fascia didn’t remove protection from the girder, but merely exposed the truth to view: this building was built under an old code that contained provisions, such as a lack fo fireproofing for girders like this in flammable buildings, that we now find unacceptable. Buildings are typically grandfathered when the codes change, so that existing conditions that become non-compliant through changes in the codes are allowed to remain, but it can be shocking when you see it as plainly as this.
I was going to name this post “how engineers learn” but (a) I’ve used a variation on that before and (b) I thought of about a dozen bad jokes in the first minute after that title came to mind. So, no.
The picture above shows the southwest corner of the roof of that building from the 1980s that I’ve been revisiting. I’m standing on the main roof and the two structures in front of me are a stair bulkhead-to-be (rather obviously) and the window washer track (less obviously). I designed the concrete structure for both. The perimeter columns are part of the lateral-load system of the building – it has a “tube frame” where closely spaced perimeter columns work with the spandrel beams to create lateral resistance at the perimeter – and I did not design that.
The need for a window-washer track is a secondary effect of having inoperable windows. How do you clean the outside of a tall building when the windows don’t move? You have window-washer scaffolds mounted at the roof that can be lowered to access the facades. The scaffolds are hung from two wire ropes, which are hung from davits, which are basically very big hooks. One possibility is to have fixed davits at various locations; at this building the davits are mounted on a small flat-bed electric truck, which drives around on a miniature racetrack circling the perimeter above the roof.
In any case, the loads on the window-washer track and on the stair bulkhead roof are small compared to those in the main structure of a 45-story building. And the design of these items is pretty much separate from the design of the main structure. That made them perfect, in 1987, to hand off to a junior designer like me.
The track was pretty boring. It was a span of one-way concrete slab, a short concrete beam, and a longer concrete girder, each repeated a number of times. I had already designed similar beams and slabs in the main structure below, so this was nothing new. It was nice to have it to myself, but the design itself was uninteresting. The bulkhead roof, on the other hand, was fun. It’s designed as a single bent beam, more or less tracking a 180-degree arc up from the roof, across, and then back down again. It had to be designed for (a very small amount of) wind load and even better, had to have the rebar detailed to make one right-angle bend and two obtuse angle bends. I spent a lot of time on the section cut longitudinally through the bulkhead roof that showed all the rebar bends. I feel like the final result, in addition to teaching me something about design, is my (very small) contribution to the art of concrete.
Since we moved to Broad Street – more than two years ago, although it doesn’t seem like it’s been that long – the lot at 45 Broad has been vacant, waiting for a new tower. The foundations for that building are currently being built, but until the superstructure is in place Broad Street pedestrians have the view above, looking at the back of the buildings fronting on William Street. The building I want to talk about is 40 Exchange Place, AKA Lord’s Court, an office building constructed in the 1890s. You can see the westmost wing of its sort-of U-shaped plan in the center of the photo: it’s the tall narrow brick wall with two rows of windows. Here’s a picture of the front, facing the corner of William Street and Exchange Place, shortly after it was built:
Note that new skeleton frame under construction just to the south (left).
William Birkmire, an architect and engineer who is best known today as a chronicler of the technology of early skyscrapers, had this to say about the building: “Lord’s Court Building, situated at the south-west corner of William Street and Exchange Place, [is] on a peculiar-shaped plot of ground containing about 12,000 square feet. The building is 214 feet high from the sidewalk, and contains fifteen stories.” What do I mean by “sort-of U-shaped”? What does Birkmire mean by “peculiar-shaped”? Here’s the plan Birkmire published with his text, showing all the floors but the lowest and highest:
The west facade of Lord’s Court (on the right) of that long wing is the tall narrow wall in my photo. Lower Manhattan used to be full of these strange lots, which were the result of centuries of real-estate speculation playing on the street layout of a Dutch village. Most have disappeared since the 1920s as bigger buildings were built on multiple lots.
So far, this is just the old NYC story of a large building being awkwardly jammed onto a small lot. But if we continue with Birkmire, it gets more interesting: “The constructive metal-work is of wrought steel through-out, the connections being designed with particular reference to the lateral stresses incidental to such tall structures, especially in the long wing, where heavy beams and knee-braces are used.” In other words, the peculiarities of this building’s lateral-load bracing were worthy of mention in the 1890s.
The long wing is about 30 feet wide at its widest, at the west end. It tapers down to less than 25 feet wide where it joins the main block of the building. At 214 feet tall, that gives this wing, by itself, a slenderness of about seven. To put that in perspective, that’s more slender than the old World Trade center towers, the new One WTC up to the observation deck (excluding the unoccupied spire), or the Empire State Building. If the wing were only connected to the main block at their juncture, it would have to have a state of the art (for its era) lateral-load system. As it turns out, the structural designer was clever in a different way. The photo below is a cropped and slightly enhanced version of the one at the top:
The purple ovals aren’t just doodles. I’ve put them there to highlight a few of the tie beams that connect the wing to the main block. They’re not at every floor and they’re concentrated at the top of the building, but they are definitely part of the lateral-load frame. In short, the designer dealt with the difficulties of designing a slender free-standing wing by making it not free-standing. This type of tie beam can be seen in a lot of old skyscrapers, usually cutting across light-courts, allowing their frames to work as a unit rather than as a bunch of very slender isolated wings.
The moral is something I’ve said before: people in the past were just as clever as us. Given what they accomplished with the primitive tools they had, they may well have been more clever than us.
The phrase “they don’t build them like they used to” is always good for some laughs. There are certainly differences between past and current practice, but there is no consistency in terms of quality. Some things in the past are better than their current-day counterparts (the quality of ornamental brickwork, for example) and some are worse (the solidity of foundations, for example). The cliché assumes that the past was always better, which is simply not true.
The picture above is some twenty years old, and I mention that because the problem it shows was fixed a long time ago. It shows a side yard from a window of the building on the far left. The tall rectangular brick thing in front of us, more or less centered on the photo, is a mid-1800s chimney carrying the boiler flue for a church. You can just make out the sloped roof of the church to the left of chimney, in the narrow gap between it and the building I was standing in. An important fact is that the chimney was built some 20 or 30 years after the church.
If you look at the gap between the chimney and the building, the problem is clear: the chimney is sloped to the left. If you look at the intersection of the sloped roof and the chimney, towards the bottom of the photo, a possible reason is clear: the chimney encloses the edge of the sloped roof. The gutter you can see came much later, but there’s a corralled brick projection at the top of the wall that supports the roof eave, and that brick interrupts the near side of the chimney. Not entirely, but enough so that side (of the four that make up the chimney) is weaker. It’s also the side that gets water, snow, and ice shed onto it from the roof. Here’s a picture looking up that shows the dramatic difference between the chimney below the eave (vertical) and above the eave (not):
How did this happen? There are two obvious possibilities. The first is that people building the chimney constructed the foundation tight to the church’s foundation wall without looking up to see that there was a projecting brick corbel at the eave. The second is that they knowingly built the chimney so that the eave cut into its side. The chimney masonry isn’t tied to the adjacent wall, so the chimney is effectively free-standing, so there’s no reason (other than the fact it would have been funny-looking) that there couldn’t have been a small gap.
I’m trying very hard to not say that those builders some 170 years ago screwed up, but I think I’m failing. They constructed a time bomb that eventually went off.
The subway station at 72nd Street and Broadway is part of the original IRT line of 1904, and has a beautiful miniature head house as its original entrance. That entrance, along with the too-narrow platforms and way-too-narrow stairs meant that this station was crowded for decades. Around 2000, a plan was developed to slightly ease the overcrowding by building a new entrance (which opened in 2002) with new stairs half a block to the north, coupled with closing a couple of traffic lanes to extend Verdi Square. (I played a very small role in that project by designing bracing for the Verdi Monument to prevent damage from vibrations during excavation.)
The picture above is the interior of the new head house, which looks like a modern version of the old one. The outside of it is fine, although not as attractive as the old one. The inside…I have my doubts. You’re looking at three separate planes of glass in the picture: the glass above the fluorescent lights is a skylight, the glass directly below the lights is a clerestory around the base of the skylight, and the glass at the bottom of the picture is the east wall of the structure. All the glass, and the new brick as well, is supported by a steel frame.
The frame needs to be stable when subjected to wind load, but it’s a very small building and the forces aren’t that big. The decisions was made – somewhere, by somebody – to put diagonal-rod bracing in every clerestory panel. The round plate and the clevises were used to avoid the geometric interference where two bracing rods cross. This is all very exposed-and-honest structureish but it’s overkill. There are gusset plates welded at each vertical-to-horizontal connection that required more welding than creating moment connections would have. The amount of connection – gussets, clevises, the round plates – is far out of proportion to what it should be.
Simply put: the structure is unnecessarily complex as structure and not particularly attractive as architectural ornament. The way to have beautiful exposed structure is to have the structure make sense on its own terms.
