The most expensive number in engineering

For more everyday civil engineering, I think the safety factor "covers up" a lot of systemic inaccuracies everywhere in the system, from modeling to design to manufacturing to unintended uses. Some of those you might account for in a probabilistic model. But it's very difficult to probabilistically model errors in the model itself, as the financial industry found out the hard way.

When driving over a bridge built the way suggested here, how comfortable can we be that certain stresses aren't correlated in ways that the engineers didn't anticipate? Or that a certain distribution isn't actually as well approximated by a Gaussian as it was assumed to be? Intuitively, it's a lot harder to be wildly wrong with the factor of safety approach.

To put this another way, a more complex way to reason about safety necessarily has more moving parts, and is thus more likely to be wrong. So in effect, adopting more complicated safety models introduces a safety risk all on its own. I think that needs to be considered as well.

For example in the JJ Abrams movie Star Trek Into Darkness we see the Enterprise operating at depth in an ocean, then dramatically zooming away into space. Then later another ship falls from orbital height and plows through San Francisco without losing its hull shape.

In Star Wars the Millenium Falcon is constantly doing things that would seem to be outside a normal design for a spacecraft, and it survives (aside from the radar dish).

Even as far back as the movie 2001, the monolith is made out of a material that humans can’t dent or cut. Why so strong? It’s basically just an automated radio.

The idea is: this advanced civilization has such command over physical technology, that they can effortlessly engineer unnecessary strength without losing any of their designed performance.

"""

This is _exactly_ LRFD (Load Factor Resistance Design) which has been in the Civil Engineering building codes since the mid 80's, and became common in use the 90's when I was an Engineer (in training).

(It's the difference between the older green book and the newer (at the time) silver steel design handbook)

It was absolutely drilled into us in school though, that Safety Factors and LRFD factors covered material and other uncertainty, they did not cover blunders.

It makes sure we calculate the loads correctly and use appropriate materials. You can't fix a bad design.

The Arkansas bridge that has been in the news lately probably would have collapsed if it wasn't for the F.S. https://www.ardot.gov/divisions/public-information/40-ms-riv...

Particular kudos to thinking harder about whether we've truly tested the actual environment where our product might be used.

I wish social scientists would do the same in controlled studies of human behavior -- which are then extrapolated to the ways that people make real-world decisions. A particularly vexing examples involves the way that psychology students make decisions in short experiments involving small amounts of money or other rewards. (Endless variations on the "marshmallow test," etc.)

Knowing what a college student will/won't do for a whimsical $5 reward says almost nothing about how an adult on the brink of poverty will balance bigger, more difficult decisions. Yet we apply a 95% confidence level to the college-student experiment and think we've learned something about the power of all financial incentives

I'm almost as sure that everyone keeps using the number people used before they joined the profession. Because if you decide to lower it, and a disaster happen, you are in very deep shit.

So once set, the number will tend to stick until forced to change by something extraordinary.

Which makes me very curious about how the number varies between independent domains. Do Japanese, Norwegian and US bridge builders all use the same number? Do builders of bridges, skyscrapers, and dams use similar numbers?

The answer would tell us something about how arbitrary these numbers are.

I asked whether it gave different answers than the standard 1.5 safety factor. As I expected, the answer was yes. It turns out that in a conventional skyscraper, there is a tiny proportion of the structure that needs to be done right. This is good news as you can x-ray those beams, and check and double check that all the bolts are installed correctly. The cost to do this is tiny. The rest of the building can be built with an effectively smaller safety factor, and it will be fine. This leads to overall cost reductions.

I'm not a real Civil Engineer but I was a graduate one from 1991 - I'm now a IT bod. Anyway, Civ Eng uses established factors of safety or safety multipliers or safety factors or whatever. Structural steel uses 1.2 I recall - so you work out your worst case (in 100 years - look up tables) bending moment and mult by 1.2. Civ Eng is one thing and despite our bridges still failing after 2000 years of really solid knowledge. Tacoma Narrows (who knew the bloody things fly and shake) or London Millenium bridge - lol - shake, shake, shake the room - BOOM.

The thing about safety factors is that they need to be derived conclusively. In Civ Eng - wood is a bit wayward so the safety factor for it is quite large compared to steel.

I have no idea what you do for space thingies (yes I do) but I would expect my first 50 experiments to blow up - I need to explore the extremities.

If I ran a Space Agency I would say something like: "Soz, we are going to make some cracking firework displays first and then we will know what to avoid."

We called this the “Factor of Lloyd” and we had a few sayings about it.

If safety factor adds mass and additional mass requires additional force to accelerate is a lower safety factor safer since you'll lower the amount of force required thus increasing the structural safety?

Calculating safety factor for a given scenario feels like a complex multivariable equation. Is that the case?

Many comments here relate more to one-off design. There is also the medium-high-volume manufacturing end. There, a prototype run might be in the dozens or hundreds of units, more than the entire manufacturing run in other heavy industries.

As the OP hints, "safety factor" is not the only term to use. A design margin (including reduction in margin) can be planned with one or more motivations: safety, reliability, weight, volume, reduced BOM costs, unit costs of repair, fleet costs of repair, logistic and warehouse costs of parts for repair, planned obsolescence, and so on.

Probabilistic design, also realized through "Monte Carlo" analysis, can take into account multiple simultaneous non-linearity in various models, where symbolic or formula-based analysis might fail.

For example (and roughly speaking), if one has millions of miles of over-the-road data, say of wheel-to-road forces or geometric road or track profiles, then one might manage to calibrate the following together: 1) a specific vehicle physical model, including parts tolerances and probabilistic discrete flaws; 2) material cycle-fatigue damage properties; and 3) some set of Weibul-distribution-like parameters as an intermediate in predicting failure rates and "lifetimes." .. AFAIK the kind of business analysis one might do could include predicting how many parts one should overproduce and warehouse (in a one-time batch) to service in-warranty and post-warranty repairs out to N years.

At that scale it can also become sociological. "Safety margin" is a loaded term when it comes to liability and imprecision in intent. You reduced the safety margin, as it says right here?!!

Not a bad article, but there could be a whole article on ramifications of different margin-related wording, high-N statistics, and explicit accounts of simultaneous goals.

Why do different fields have different definitions? Because they're different fields! Aerospace doesn't really worry about material imperfection because they do very intensive quality inspections that aren't feasible for bridges (e.g. X-rays that can't effectively be done outside). And early nuclear design use focused on "imperfect theory" because (of course) the theory at the time was somewhat uncertain.

I think this article is overly dismissive of a proven way for a whole industry to learn over time about risk management. e.g. Boeing doesn't want to share the distribution of their material strengths. But they're happy to share some safety factors that don't reveal a lot about their business but help out the whole industry.

So, reducing the padding from 5x to 1.5x was already most of the benefit. If you were at 1.2, there are probably better ways to shave costs than reducing your unexpected force multipliers. It’s definitely attractive to lower cost / increase speed / whatever if you truly think it’s “free”, but the benefits are diminishing.

A NASA project I was related to we logged every single bolt that went on the device, where it came from, the batch number, and had to keep all the old software around in the event we had to reconstruct something.

The amount of overhead was pretty amazing.

Most of that is for safety.

And yeah, the mentioned finance "risk factors" also generally keep going down over the decades/centuries. In a similar fashion: the markets dare to use a slightly lower number, over time nothing too bad happens, more people jump the bandwagon.

The "cost" of using higher than needed "financial risk factor" is easily in the trillions per year.

EDIT: What's usually limiting in elevators (and that's why they say "max 4 people") are the breaks.

Because in many applications, the full cost of failure can be unimaginable.

Also fudge factors have a tangible benefit: time. It permits declaring a design "good enough" sooner.

This could include changing requirements, poor communication, illness, being blocked by other changes, etc.

I learned that you can get away with giving people extended deadlines as long as you hit them.

(My point is that the scaling of the importance of quantities is arbitrary so a single safety factor doesn't make sense to be applied to every quantity.)

The probably of failure should be calculated considering material defects, forces larger than predicted, simulation errors, and all other causes the factor of safety is designed to protect against.

Then the engineering process can allocate those probabilities in the most efficient way.

For example, in a rocket it might make sense to make the engine bells stronger (decreasing probably of failure) while making the fuel tanks weaker (increasing probability of failure). The overall probability of failure remains the same, but perhaps the craft ends up lighter/cheaper/better than it would be if all components just built in a fixed factor of safety.

Because design constraints are not universal across all projects. Human error for something like the space shuttle is fairly minimal considering the mind boggling amount of training that's done. If you have an unreliable or unproven materials provider then that present a design constraint that must be accounted for. Etc.

For example, I want to build a small building. I want to be able to calculate things like the snow load of the roof, make a suitable truss, take into consideration things like the safety factor, and build with the materials to meet those requirements.

I can buy plans for similar constructions, but i don't see how they decide what size wood to use where, etc - and I want to gain an understanding of that.

It might be a good idea to lie to the contractors for a project about the margin of safety so they take better care to make it right. But that may not be possible.

However, I’d hazard the most unnecessarily costly variable in engineering (over time, in aggregate; as well as on most any given substantial project) is the number of days later a project starts than it could have if it had just gone ahead and started.

The public sees destructive flight testing as failures so if your program relies on public money, that's a problem.

Bake in a factor of safety into your estimates depending on the type of work, the track record of the team that’s doing the work etc

So, if the unit of your design is linear but the unknowns operate on the square or the cube of that then the effective safety factor is much smaller.

For example I could design a 12 inch bucket to hold water and then say I'd better make the handle twice as strong in case somebody fills it past the line indicating it is full but of course that will be super-linear in its effect and my safety margin is way less than a factor of 2.

https://www.youtube.com/watch?v=7PphbSFZWuU

http://holyjoe.org/poetry/holmes1.htm

The Deacon’s Masterpiece

or, the Wonderful "One-hoss Shay":

A Logical Story

by Oliver Wendell Holmes (1809-1894)

Have you heard of the wonderful one-hoss shay, That was built in such a logical way It ran a hundred years to a day, And then, of a sudden, it — ah, but stay, I’ll tell you what happened without delay, Scaring the parson into fits, Frightening people out of their wits, — Have you ever heard of that, I say?

Seventeen hundred and fifty-five. Georgius Secundus was then alive, — Snuffy old drone from the German hive. That was the year when Lisbon-town Saw the earth open and gulp her down, And Braddock’s army was done so brown, Left without a scalp to its crown. It was on the terrible Earthquake-day That the Deacon finished the one-hoss shay.

Now in building of chaises, I tell you what, There is always somewhere a weakest spot, — In hub, tire, felloe, in spring or thill, In panel, or crossbar, or floor, or sill, In screw, bolt, thoroughbrace, — lurking still, Find it somewhere you must and will, — Above or below, or within or without, — And that’s the reason, beyond a doubt, A chaise breaks down, but doesn’t wear out.

But the Deacon swore (as Deacons do, With an “I dew vum,” or an “I tell yeou”) He would build one shay to beat the taown ’N’ the keounty ’n’ all the kentry raoun’; It should be so built that it couldn’ break daown: “Fur,” said the Deacon, “’tis mighty plain Thut the weakes’ place mus’ stan’ the strain; ’N’ the way t’ fix it, uz I maintain, Is only jest T’ make that place uz strong uz the rest.”

So the Deacon inquired of the village folk Where he could find the strongest oak, That couldn’t be split nor bent nor broke, — That was for spokes and floor and sills; He sent for lancewood to make the thills; The crossbars were ash, from the straightest trees, The panels of white-wood, that cuts like cheese, But lasts like iron for things like these; The hubs of logs from the “Settler’s ellum,” — Last of its timber, — they couldn’t sell ’em, Never an axe had seen their chips, And the wedges flew from between their lips, Their blunt ends frizzled like celery-tips; Step and prop-iron, bolt and screw, Spring, tire, axle, and linchpin too, Steel of the finest, bright and blue; Thoroughbrace bison-skin, thick and wide; Boot, top, dasher, from tough old hide Found in the pit when the tanner died. That was the way he “put her through.” “There!” said the Deacon, “naow she’ll dew!”

Do! I tell you, I rather guess She was a wonder, and nothing less! Colts grew horses, beards turned gray, Deacon and deaconess dropped away, Children and grandchildren — where were they? But there stood the stout old one-hoss shay As fresh as on Lisbon-earthquake-day!

EIGHTEEN HUNDRED; — it came and found The Deacon’s masterpiece strong and sound. Eighteen hundred increased by ten; — “Hahnsum kerridge” they called it then. Eighteen hundred and twenty came; — Running as usual; much the same. Thirty and forty at last arrive, And then come fifty, and FIFTY-FIVE.

Little of all we value here Wakes on the morn of its hundreth year Without both feeling and looking queer. In fact, there’s nothing that keeps its youth, So far as I know, but a tree and truth. (This is a moral that runs at large; Take it. — You’re welcome. — No extra charge.)

FIRST OF NOVEMBER, — the Earthquake-day, — There are traces of age in the one-hoss shay, A general flavor of mild decay, But nothing local, as one may say. There couldn’t be, — for the Deacon’s art Had made it so like in every part That there wasn’t a chance for one to start. For the wheels were just as strong as the thills, And the floor was just as strong as the sills, And the panels just as strong as the floor, And the whipple-tree neither less nor more, And the back crossbar as strong as the fore, And spring and axle and hub encore. And yet, as a whole, it is past a doubt In another hour it will be worn out!

First of November, ’Fifty-five! This morning the parson takes a drive. Now, small boys, get out of the way! Here comes the wonderful one-hoss shay, Drawn by a rat-tailed, ewe-necked bay. “Huddup!” said the parson. — Off went they. The parson was working his Sunday’s text, — Had got to fifthly, and stopped perplexed At what the — Moses — was coming next. All at once the horse stood still, Close by the meet’n’-house on the hill. First a shiver, and then a thrill, Then something decidedly like a spill, — And the parson was sitting upon a rock, At half past nine by the meet’n-house clock, — Just the hour of the Earthquake shock! What do you think the parson found, When he got up and stared around? The poor old chaise in a heap or mound, As if it had been to the mill and ground! You see, of course, if you’re not a dunce, How it went to pieces all at once, — All at once, and nothing first, — Just as bubbles do when they burst.

End of the wonderful one-hoss shay. Logic is logic. That’s all I say.

* "poorly representative material test data available" that's 5+

* "extremely challenging environment" 5+ again

* "models are crude aproximation" is another 5+

So we should be able to get a cardboard spaceshuttle, if we only use a safety factor of 125+! Moar cardboard! Great job team!