That’s a building (click on the picture to expand it) I walked by every day I was in Brussels and every day I stopped and stared at it, trying to make sense of one detail.
The main exterior facade (from the second floor up) appears to be structural, made of precast concrete with steel connections. There’s a connection at the midheight of each column, detailed so that it looks like a structural hinge*. That makes perfect sense: if the precast wall is a structural element, it is likely carrying both vertical (gravity) load and lateral (wind) load. The small beams and columns that make up the wall are a moment frame** in which they are all in bending. In that type of frame, there is a theoretical inflection point at*** the midheight of each column where the bending moment is zero. Putting hinges at those inflection points does not change the overall structural action of the frame, but it (a) makes it easier to build the frame and (b) forces the inflection points to be exactly where the hinges are. So far, so good.
It’s the first floor where the design loses me. The architectural designer wanted the first floor to be more open**** so the moment-frame wall gives way to widely-spaced columns and a glass curtain wall. There’s no structural rationale for the columns being set back from the plane of the wall above; rather, structural logic says that the columns should be in that plane. So there’s a structurally-awkward offset within that big beam at the second floor level – the large beam is needed to deal with the eccentricity of the column set-back. If you’re pretending that structural logic has driven the design, why violate it in this manner?
Much, much worse are the hinges at the tops of the first-floor columns. If the moment frame is carrying lateral load, that load has to be transferred to the foundations below the first-floor columns. The most logical form for the columns would be moment-connected at their tops and bottom, the same as the small columns in the frame. Or, the first-floor columns could have hinges at their bottoms, making them part of the moment frame without transmitting moments to the foundations. That’s a pretty common design. Instead, there are hinges where the columns meet the frame. Thus means one of two things is true: (1) the columns are cantilevered up from the foundations or (2) the whole presentation of lateral load support is fake. Option 1 is more likely, and either requires that the foundations be designed for moment***** or that there’s yet another big beam, like the one at the second floor level, to distribute the column cantilever moments before they get to the foundations.
This has been a very long-winded way of saying that if you’re going to play the game of structural expression with your visible architecture, then do it right. Make the whole structure make sense. Or alternately****** accept that no one is really looking for structural expression and avoid the whole idea entirely.
* I.e., capable of transmitting shear and axial force, but not moment, from one side to the other.
** Or, if you prefer, a Vierendeel truss.
*** Or near, if you want to be pedantic. But see the next sentence (above) after the one with this note.
**** Form follows function, or something like that: distinguishing the lobby and public spaces at the first floor from the office spaces above.
***** Which is an enormous pain in the neck.
****** And, in my opinion, better.
You must be logged in to post a comment.