Misleading Patterns

If we interpret the visible pattern literally, it looks like the leak that led to the rusting is worst at the bottom and least bad at the top. Thinking about it for a second leads to the almost-obvious conclusion that we can’t really tell about the top, because water will run or drip off the underside of the beam top flange. But then what?

Secondary influences can greatly change patterns of damage. The classic example is looking at burn patterns after a fire. It might look like the fire started and burned hottest at the back of a house when in reality the origin was at the front but so was the easiest access for fire-fighters. The secondary influence of geometry of access blurs the primary question of where the fire was actually located.

With secondary influences in mind, what do we see in the photo above? We can’t tell where the water was coming from because it was trapped by the thick ceiling (visible on the left) and therefore the most heavily rusted areas are there. Since water can move sideways in the space between the ceiling and the beams, we can’t even tell if the water was coming from the right, where the rust is worst.

There’s no real correlation between damage and the source of the damage. There is a correlation between the type of damage and the type of source: water rusts steel, differential settlement causes diagonal cracking, and rusting frames cause vertical and horizontal cracks.