It’s not too often that I make an engineering commentary on this site, but the characterisation of the NTSB’s conclusion that a design flaw was the "critical factor" in the I-35 bridge failure in Minneapolis, MN, strikes me as a little misleading. (The NTSB’s own announcement is here.)
In its interim report, the FHWA went into great detail about how the gussets that failed were the most "underdesigned," to cut to the chase. From this, the safety recommendation wisely urges the following:
For all non-load-path-redundant steel truss bridges within the National Bridge Inventory, require that bridge owners conduct load capacity calculations to verify that the stress levels in all structural elements, including gusset plates, remain within applicable requirements whenever planned modifications or operational changes may significantly increase stresses.
This is good thinking. The problem is that it’s easy to misinterpret the failure to mean that, if the original designers of this (and any other bridge) had done their job properly, that this wouldn’t have happened. To some extent that’s true, but it needs to be tempered by a few observations.
First, the bridge lasted forty years. Generally speaking, serious design flaws manifest themselves earlier. The most egregious example of this was the Tacoma Narrows Bridge, "Galloping Gertie," which literally tore itself to pieces when the narrow span resonantly interacted with the wind.
Second, the bridge was "beefed up" in the 1990’s. This added weight to the bridge. Any weaknesses in the bridge–design, material quality, maintenance neglect, or otherwise, are magnified by increases in bridge weight.
Third, the entire U.S. Interstate system is operating well over capacity in terms of traffic and weight load, especially truck traffic. That too reduces the factors of safety in bridge components, making them more prone to failure.
Fourth, any design has weaknesses. You can count on a structure, if it fails, to do so at its weakest point.
It’s entirely possible that, had the original conditions the bridge was designed for had continued without the addition of load (live or dead,) this accident would not have happened. That’s the flip side of the NTSB’s recommendation, and that’s why the FHWA and state DOT’s need to take care when adding load to structures.
It’s a tribute to designers when designs persist in conditions beyond what they were intended to. An example of this was the foundations of oil platforms in Katrina, not one of which failed even in that disaster. But it’s not something that those who tend to our infrastructure, working in a society that frequently doesn’t fix what’s broke, can simply afford to take as a given.