How the steel beams cracked at the new Transbay Terminal | Hacker News Reader

How the steel beams cracked at the new Transbay Terminal

mercurynews.com

11 pointstlb8 comments

This sounds like the kind of failure that happens when designers design structures to the hairy edge and then small mistakes in fabrication. assembly, or later design changes push the design into failure.

That is what it sounds like to me. They were shaving down costs down to the pennies and didn't account for any additional wear and tear during actual construction so now it will cost them a lot more than they saved.

Quite the opposite of cost saving, the Transit Center was a test of cutting-edge, unproven structural engineering. If they wanted it cheap, they could have built a much simpler building.

It’s kind of fascinating and depressing how culture around this has changed. The Brooklyn bridge was so overbuilt due to safety concerns that despite being constructed with bolts that turned out to be far weaker than the ones specced (due to an unethical supplier), it’s still much stronger than it needs to be, and capable of surviving modern traffic (and at one time, subway trains) that’s much more demanding than the horse and foot traffic it was built for despite occasional decades of neglected maintenance.

If modern structures are designed to within an inch of their tolerances, how are they going to handle the way we totally fail to maintain our infrastructure? How are they even going to handle normal wear and tear or changes in traffic patterns (like the way the past few decades have seen a shift in dominant personal vehicle types from sedans to far heavier SUVs)?

If we want our infrastructure to last a couple hundred years without being replaced - instead of a decade or two - shouldn’t we be overbuilding instead of underbuilding? Isn’t putting ourselves in a position to have to replace rather than maintain large expensive structures penny-wise/pound-foolish?

Then people complain about how expensive and over-engineered infrastructure projects are. Also how are you supposed to know what infra will be relevant in the future and is even worth overbuilding in the first place?

Regionally, I can't think of any infrastructure that's under-utilized rather than over-utilized and under-maintained. There are a couple proposed projects that would be stupid to build in the first place, let alone over-build; but I can't think of anything existing that isn't already at or over capacity and very obviously going to be heavily used for the next 50 years and probably more.

In fact, our entire transportation network is badly screwed and about to become far worse because major elements of infrastructure we've been using for between 70 and 120 years years and have taken for granted has seen increasing use but has not been sufficiently maintained.

If it had been built with cost-saving instead of robustness in mind, we'd have had to spend 2/3 as much three times over to rebuild it... or more likely, not have been able to muster the political will to rebuild it at all and instead seen our economy fail due to the lack of infrastructure necessary to support economic activity.

The same goes for our water delivery infrastructure (it's taking 50 years to build a new tunnel so we can even take the existing ones offline for their first maintenance in a century), our schools and post offices, and probably a bunch more things that aren't coming to mind just now.

I actually suspect these attempts at material savings are one of the things leading to ballooning costs. Spending 200 engineering hours to cut $5k worth of steel is still a net loss.

Holes being cut into structural beams, that aren't in the schematic at all reminds me a bit of the famous Hyatt walkway collapse:

https://en.wikipedia.org/wiki/Hyatt_Regency_walkway_collapse

Basically instead of using one continuous rod, they changed the design on-the-fly to use multiple shorter rods, connected at the hanging bridge.

https://www.enr.com/articles/38400-why-engineers-must-rememb...

Reply on news.ycombinator.com