Someone asked me to list what I thought were some of the most interesting curtain wall projects in the world. That’s a pretty long list, which gets shorter if you look at a wall and wonder “how’d they do that?” So, here goes with one armchair architecture critic’s take on intriguing curtain walls …
Let me say that none of these buildings have flat curtain walls. We used to say if it were easy, we’d have kids doing it. There’s no challenge in simple flat walls. I used to think a building with corners and step-backs was hard. Been there, done that. Do the first couple of floors with odd modules, get to the typical floors where everything repeats for X number of floors, repeat the bottom at the top with some oddball modules and then wrap it up. Been there, done that, too.
The projects that give me pause offer a bit more in the way of a challenge from a system design point, stretch the field guys to handle material they’re not using every day, and do something that’s not been done before. In the mid-80s, putting a hole in the middle of 2200 Ross in downtown Dallas was such a challenge.
But that pales in comparison to what some of the architects and leading curtain wall subcontractors are doing now. For example, projects where no piece of glass is the same size as the one above or below it and where there is NOT a lot of repetition.
The first one like that I was fortunate enough to work on is the Getty Museum in Los Angeles. The estimated contract value was two to three times what the original contract was for, due to the complexity and design intent that came to light as the project progressed. There was nothing typical about that job. There were sunshades, curved metal panels, curved glass and curved stone. The largest extrusion I’ve ever seen was a sunshade outrigger that had to be all of 5” wide x 20” deep, and was a 5-hollow die. There were more than 7,100 different unitized wall panels on it. If I recall correctly, the duplication dropped to one panel thus at about unit number 50, if not sooner. At its peak, there were about 12 different part drawing teams producing fabrication tickets. On a high-rise unitized curtain wall job there might be 600, maybe 800 max different types of units, and the quantities would have dropped to one thus or two thus at about unit number 400 or 500. So the degree of complexity was seriously over the top from anything previously experienced. Never to be repeated, hopefully, but the job paid the salaries of a lot of people for the better part of two years or more.
Anybody really want to go 3D?
The egg-shaped building in London, Gherkin Tower, has triangular glass lites. (As an aside, I wonder if they got environmental credits for the drop-offs, both glass and metal, generating post-manufacturing recycled content, cutting triangular lites from standard rectangular sheets of glass?) And the surveying that had to occur onsite so the field crews could set the next units, where every line represented angularity changes in the plane of glass, every piece of glass is slightly out of plane with the one next to it.
There’s an hourglass-shaped building in Doha, Qatar, called “Tornado Tower.” Its floor plans are circular. The floors start out wide at the base, get smaller in plan in the middle stories, then wider again at the top.
But that’s nothing compared to some of the derivatives of that design. The one I’m thinking about is basically triangular in plan, but each of the triangle’s edges is curved, and saw-toothed three times per edge. In addition to the hourglass profile, the floors are slightly twisted from each other like 0.50 from the one above/below. Now THAT would be a fabrication challenge. There’s no way you could ever do such buildings without modeling them in 3D CAD or BIM.
I was involved in a project one time that had a sloped roof on a curved building – the Sama Dubai Tower. You’d think sloped roofs would be easy. But when they’re on the roof of a building that in plan is an arc, they’re not. Thanks to Joe Callahan for finally making me see this. Think of a circular stairway: the inside handrail, due to the smaller circle diameter, falls at a much steeper rate than the outside rail. It has the same rise, but the length of the two handrails (in stair lingo, the run) is radically different between the two. When doing this over the 80’ width of the building, the surface needed to be warped 1 ½” for every mullion placed roughly 4’ on-center across the width.
Think of picking up a sheet of plywood ever so slightly at one corner while the other three corners are still on the floor. That’s the 1 ½” warp this geometry wanted to force us into. But to do that in flat glass is nearly impossible. When we called a glass manufacturer, they would only allow their IGUs to be warped ¼” over any dimension. Eventually the glass had to be laid out in a triangular grid in order to make the plane changes required by the building’s overall shape.
A last project to note was one I saw modeled in the architect’s office before it was built, Capital Gate in Abu Dhabi. This one reminded me of the “Slinky” I had as a kid. The building leans 18o – that’s 14o more than the Tower of Pisa. Now THAT would have been a challenge – little or no repetition in that one.
These projects really define what “work” means while in progress, but what a trophy to hang on the wall in the office. My hat is off to the talented folks on such projects – those doing the system design, the shop and fabrication drawings, and the fabrication and installation crews. It really does take a collaborative organization to pull these off. I wonder how many people woke up at 2 a.m. during the project thinking of something they missed, or had an epiphany that clarified some minutia that made the rest of the project go so much better.
Any other intriguing curtain wall examples you’d like to throw into the mix?