Robots build your car and computer, and now they’re laying out interior walls. What are the implications for other building components, including curtainwalls? Read on, or be assimilated …
We’ve all seen interior partition subs marking floors with chalk lines for metal stud and gypsum walls. The 4/22/13 issue of ENR had an article about a robot used to lay these out to an accuracy of +/- 1/8”. The CAD or BIM information is entered into the “laybot” robot, which then moves around the floor at 1.8 – 2.2 feet/minute, even around corners, laying out curved or straight walls. I wonder if it can be adapted for layout of curtainwall? Some challenges come to mind.
In any curtainwall job, one of the first tasks the field crews undertake is layout of the wall in relation to the structure. On taller buildings, the forming of the upper floors is proceeding above, while the curtainwall layout, starting on the first floor, has to predict where the building is going to be long before it’s completed.
The biggest challenge is that a continuous, full-height curtainwall is not on the slab, it’s off the slab, hanging out in thin air. So, putting a layout line where the wall will be is not feasible (after all this isn’t drywall we’re talking about) except maybe at the base of the wall. But curbs and overhanging walls can complicate that. Additionally, perimeter columns get in the way of putting the layout lines right at the edge of the floor. That location is further complicated as slab edges often are not straight.
So, layout crews tend to “cheat” the line to miss the columns, moving back in on the floor slab to miss the columns. One of the things the field crews usually discuss with the GC when they first hit the job is: How much room are we going to need for layout and material storage around the perimeter of the floor?
Layout in plan is one facet of this, but the layout crew also has to deal with vertical tolerance. If you’ve ever stood at the base of a tall curtainwall and looked up, it’s surprising to see how straight the wall is (it’s a perspective you can’t get when standing away from the building, taking in the whole of it). The in/out plumb of the wall can only be seen from that vantage point, or possibly from the other end, on the roof looking down. It’s a sign of how good the layout crew was. Specifications often limit plumb deviations to less than 1/8 inch per floor. That’s a lot.
It’s always been a minor miracle to me that more buildings and curtainwalls don’t crash into each other. BIM can help model this prior to start of the work, but when the building actually starts, theory becomes reality, or should, anyway. Granted, tolerances and designed-in allowances accommodate that, but when a steel structure is erected with tolerances of +/-2” or more, and curtainwalls being built to +/-1/32”, why don’t they bump into each other more often?
And concrete’s even more challenging than steel in this regard. At least with steel, you can work from fixed dimensions every time, but concrete’s formwork is repeated over and over again for the height of the building, allowing just a small mistake to have serious repercussions.
And the horror stories about having to chip concrete to allow a mullion to go by are out there. In one project I heard about, crews had to chip the concrete floor beam so far back, they had to cut through the rebar, also. I’ll bet the project structural engineer loved hearing that.
Typically the 1 ½- to 2-inch tolerance between face of slab and back of wall system is cheap insurance. It gives the general contractor (GC) some flexibility with the edge of the slab, and allows the curtainwall to accommodate any variance without having to notch slabs. Anything smaller than 1 ½” ends up with notching requirements later. Architects don’t like it, as they don’t like to think about how to close off the gap. But it’s a necessary “evil” which permits an easier installation of the wall.
Having the discussion with the GC prior to work starting, usually as part of the pre-installation conference, is a good place to get all this figured out. For example, what the structure erection tolerances are, what the curtainwall anchors can reasonably expect to accommodate for the structure tolerance, and what to do when the inevitable clash occurs. It may be beneficial to have this discussion as soon as possible after contract award so that everyone, GC, framing sub and glazing sub all get off on the right foot, knowing where everybody’s supposed to be when the wall starts to go up.
It will be interesting to see who among the curtainwall subs will be the first to set up a robot to do layout work…
Answer to previous trivia: CCNY, 1950. They beat the same school both times, and since I know some Bradley grads, I won’t mention who it was that lost both tourney championship games in the same year.