I spent a week walking the construction edges of the I-4 corridor in Orlando with a notebook, and I came away with a much lower opinion of my own ability to reason about a system under change.
The project added lanes through a corridor that already carried well over two hundred thousand vehicles a day, and it did it without closing the road. Everything I want to describe here is a consequence of that one constraint.
The visible unit is the pier, not the bridge
Standing along the corridor, what repeats is the pier.
• Cylindrical or twin-column shafts, cast in place, rising from a footing buried under the work area
• A pier cap on top, wider than the columns, which is what the deck actually sits on
• bearings between the cap and the girder, so the deck can expand and contract without pushing the column
• precast box girders laid side by side across the cap, then stitched together
The girder is the interesting one. A box section is stiff for its weight, it can be shipped to site as a unit, and it can be set without a falsework tower in the lanes below. That last part matters more than any structural elegance, because falsework in the road means closing the road.
The temporary network is the real project
What I underestimated completely was the number of things that had to exist only once.
Detour roadways paved and then torn up. Steel plates over utility cuts, welded at the edges so they do not walk. Temporary retaining behind the ramp abutments. Silt fence along every drainage path, because the site sits over the aquifer and runoff rules are not a suggestion. Haul routes that avoid the school at the hour the buses move. Light towers angled so the glare does not hit the driver who is trying to merge.
Each of those has a design life of a few months and a maintenance schedule of a few days. Nothing about them shows up in the finished structure, and all of them were on the critical path.
Working at night because the road is a resource
Several of the lifts happened at two in the morning, when the traffic count is a fraction of the daytime number. That is the whole logic of the shift pattern: the resource being protected is not the equipment or the crew. It is the lane. A closure that costs four hours of peak traffic would be measured in economic terms somewhere in a planning document, and the schedule gets written backwards from it.
Getting a coherent view of a corridor you cannot see all at once
The problem I kept hitting is that the corridor is kilometres long and the work faces change weekly. My notes were a list of disconnected places.
I stopped taking notes and boxed the whole corridor into Shapezo instead. You drag a rectangle on a map and it generates a 3D model of that region, so the terrain, water, roads, and blocks are there, and you can orbit it. What that gave me was alignment. I could place each work face I had visited on a real surface, see which ones shared a haul route, and understand why one ramp was being built from three ends at once.
It is not a substitute for drawings. It will not tell you a girder weight or a bearing type. It fixed the thing I was actually bad at, which was holding a long linear site in my head.
What I would tell someone covering a project like this
Ask about the temporary works first. Ask what the site looked like six months ago and what it will look like when the plates come out. The permanent structure is a small fraction of the total decisions, and the interesting ones are the ones designed to be destroyed.
Top comments (0)