4 ms·
"But what I found out at that meeting were that all factors of safety were gone." Many catastrophic "accidents", and I use quotes because they could have been
by sitkack 13y ago
"But what I found out at that meeting were that all factors of safety were gone."
Many catastrophic "accidents", and I use quotes because they could have been averted had people not cut corners w/o knowing the full context.
* Chernobyl (after hours test by an untrained crew with an inverted fail safe design)
* http://en.wikipedia.org/wiki/Hyatt_Regency_walkway_collapse (design change in the field, very similar to the citicorp flaw)
* 3 Mile Island (indicator that triggered on switch rather than valve)
* Fukashima (cost cutting on seawall and generator snorkels)
* Ariane-5 (code reuse, dead code)
If you want to look at good engineering, look at the Brooklyn Bridge[5] and the DC-3[6].
Too many people don't design with proper safety factors. You build it, you test it, you test it till it fails and you understand those failures. I would trust another citicorp wouldn't happen because we can do realistic wind model, we can do an earthquake model, an anything model. Maybe we can get to a safety factor of 1 when everything is automated, when everyone has an off-site backup of their own brain but until then. Safety factor 6.
[5] http://www.asce.org/People-and-Projects/Projects/Landmarks/Brooklyn-Bridge/ http://www.asce.org/People-and-Projects/Projects/Landmarks/B...
[6] http://en.wikipedia.org/wiki/Douglas_DC-3 http://en.wikipedia.org/wiki/Douglas_DC-3
- sitkack 13y agoThis is fascinating, http://www.onlineethics.org/cms/24528.aspx http://www.onlineethics.org/cms/24528.aspx It also reminds me of the Mars Climate Orbiter. http://en.wikipedia.org/wiki/Mars_Climate_Orbiter http://en.wikipedia.org/wiki/Mars_Climate_Orbiter
- darkmighty 13y agoSafety factors also exist because we don't know what we don't know. Models sometimes break down unexpectedly, so a factor >1 is sort of a "meta safety factor 1", taking into account that our models/people somewhat predictably fail.
- hrktb 13y agoAnother direction would be to build by default with as much margins as realistic. The original design with the welded joints would have allowed the building to withstand the corner winds, but it was replaced by bolts because builders thought it was overkill. Perhaps the consensus (or law?) should be to build with at least 2-3 times the calculated resistance, with only some justified exceptions. Buildings would cost way more, but Intuitively it wouldn't be a bad thing (they would last longer and they would a bit less of a "mine's bigger" attitude)
- Someone 13y agoThe Brooklyn bridge is over-engineered. Proof of that is that it handle today's loads. Of course, the advantage of that is that it hasn't collapsed yet, like a few under-engineered bridges built way later (http://www.washingtonpost.com/wp-dyn/content/article/2007/08/04/AR2007080401439.html http://www.washingtonpost.com/wp-dyn/content/article/2007/08...) But I doubt the ones paying for the Brooklyn Bridge would have chosen this design if they could have had something for half the price that would last for 75 years.
- m_mueller 13y agoAs a European this argument is a bit strange to me (since we have so many structures that are hundreds of years old and are still in use today). Don't people sometimes just want something that works and keeps on working? A monument of engineering to be proud of? Something to show to visitors? And on the economic side of things I'm not at all convinced that rebuilding every century would be cheaper for something like a bridge, considering the cost of closing it down.
- Someone 13y agoWant? Yes, but want to pay for? I like it that I can see a 600+ year old church from my living room, too, and that affected the price of my apartment, but I don't think I would be willing to pay much to have a new building that will be considered iconic in 500 years time in my view. Most modern bridges (and buildings, for that matter) get designed for 75, maybe 100 years of life. I don't think that is different in Europe. Older ones that still stand typically are sturdier, partly due to the use of larger safety margins by engineers who (according to today's knowledge) didn't know much about materials science, partly due to natural selection (bad designs collapsed or were taken down decades ago) For every centuries old bridge, scores have been demolished because they couldn't handle increased traffic or just became too expensive to operate. And that happens in Europe, too, even in old-stuff loving Great Britain (http://en.wikipedia.org/wiki/London_Bridge http://en.wikipedia.org/wiki/London_Bridge) Looking at bridges, I think the only ones that will stand for centuries are stone and masonry ones, and those rarely are built anymore. It is way easier to get large spans using reinforced concrete or steel. Both contain metal that rusts. Preventing that is expensive; fully preventing it probably prohibitively so. Even for landmarks bridges such as the Firth of Forth the Brits do not aim for eternal life (http://en.wikipedia.org/wiki/Forth_Bridge#Maintenance http://en.wikipedia.org/wiki/Forth_Bridge#Maintenance: "Network Rail has estimated the life of the bridge to be in excess of 100 years. However, this is dependant [sic] upon NR’s inspection and refurbishment works programme for the bridge being carried out year on year")
- abduhl 13y agoThe idea of a "factor of safety" has gone away in structural engineering to be replaced by a similar concept of "factoring" (look up allowable stress design, ASD, and load resistance factor design, LRFD, for more information). Factors are based on statistical analysis and criticality of failure. No combination of factors approaches anything near what you would call a safety factor of 6. Typically things are designed with an analogous safety factor of 2.5 or less. This isn't to say that higher factors of safety aren't possible, it's just that they're not worthwhile. We can design for anything. We just can't pay for it most of the time.