Allan Cousins writes:
I have recently been thinking about the way in which professionals come to accumulate “knowledge” over their careers and how that process utilizes (read: abuses) the notion of corroboration. I believe this might be of interest to both of you and so I wanted to see if either of you might have any insights or comments.
In particular, I have been thinking about professional endeavours that have dichotomous outcomes where the range of possibilities is restricted to (or perhaps more accurately, viewed as) it either worked or it did not work. For the purposes of this discussion I will look at structural engineering but I believe the phenomenon I am about to describe is just as applicable to other similarly characterized disciplines. In structural engineering: the structure either stood up or it collapsed, the beam either carried the load or it did not, etc. In my experience there are nearly as many theories of how structures work as there are structural engineers. But this wide range of opinions among structural engineers is certainly not because the underlying concepts are not well understood. That may have been true in 1850 but not today. In fact, structural engineering is quite mature as a field and there are very few concepts (except at the edges of the field) where such a diverse range of thought could be justified.
This begs the question of how could this unsatisfactory state of affairs have come to pass? I have often pondered this but only recently have come to what I think to be a reasonable explanation. First, let us rule out the idea that structural engineering professionals are of below average intelligence (or rather below some required intelligence threshold for such endeavors only known to Omniscient Jones). Under such an assumption I believe that the likely answer to our question comes down to an interplay between industry dynamics, an abuse of the concept of corroboration, and the nature of the outcomes inherent to the field.
Even if engineers have never heard of the concept of Philosophy of Science (and most have not) they are apt to act in ways akin to the typical scientist. That is, they go about their enterprise (designing structures) by continuously evaluating their understanding of the underlying structural mechanics by looking at and seeking out corroborating evidence. However, unlike scientists structural engineers don’t usually have the ability to conduct risky tests (in the popperian sense) in their day to day designs. By definition the predicted outcome of a risky test is likely to be wrong in absence of the posited theory and if structural engineers were routinely conducting such field tests newspaper headlines would be replete with structural engineering failures. But today structural engineering failures are quite rare and when they happen they are usually small in magnitude (one of the greatest structural engineering failures in US history was the Hyatt Regency Walkway collapse and it only caused 114 deaths. For comparison that is about the same number of deaths caused by road accidents in a single DAY in the US). Indeed, building codes and governing standards are codified in such a way that the probability of failure of any given element in a system is quite a rare event (global failure even rarer still). What that means is that even if what a structural engineer believes to be true about the structural systems that they design actually has very little verisimilitude (read: is mostly wrong and to a severe degree) their designs will not fail in practice as long as they follow codified guidelines. It is only when structural engineers move away from the typical (where standard details are the norm and codes contain prescribed modes of analysis / design) where gaps in their understanding become apparent due to observed failures. What this means then is that while the successful outcome of each “test” (each new structural design) is likely to be taken by the designer as corroborating their understanding (in the same sense that it does for the scientist), it does not necessarily even provide the most meager of evidence that the designer has a good grasp of their discipline. In fact, it is possible (though admittedly not overly likely) that a designer has everything backwards and yet their designs don’t fail because of the prescribed nature of governing codes.
The above leaves us with an interesting predicament. It seems clear that structural engineers or others in similarly situated disciplines cannot rely on outcomes to substantiate their understanding. Though in practice that is what they largely do; they are human after all.
This lack of ability to conduct risky tests interplays with industry dynamics and in not a particularly promising way. Those who commission structural designs are unlikely to care about the design itself (except to the extent that it doesn’t fail and doesn’t mess with the intended aesthetic), and as a result, structural engineering tends to be treated like a commodity product where the governing force is price. What that means is that there is an overwhelming pressure to get designs out the door as quickly as possible lest a structural engineering firm lose money on its bid. This pressure all but guarantees that even if senior structural engineers have a good understanding of structural principles the demands for their time leave few hours in the day to be spent on mentorship and review of young engineers’ work product. As a result, young engineers are unlikely to be able to rely on senior engineers to correct their misunderstanding of structural principles. That pretty much leaves only one other avenue for the young engineer to gain true understanding and that is via self-teaching of the literature and the like. However, given the lack of ability to construct risky tests (see above) the self-learning route is apt to lead young structural engineers to think that they have a good understanding of certain concepts (because they see corroborating evidence in their “successful” designs) where that is not the case. Though to be fair to my brethren I am assuming that the average young engineer does not have the ability to discern true engineering principles from the literature on their own without aid. However, I believe this assumption to hold, on average.
This leads to a cycle where young engineers – who have a less than perfect understanding of structural systems that goes unchecked – become senior engineers who in turn are looked up to by a new crop of young engineers. The now senior engineers mentor the young engineers, to the extent time demands allow, and distill their misknowledge to them. Those young engineers eventually become senior. And in the extreme, the cycle repeats progressively until “knowledge” at the most senior levels of the field is almost devoid of any verisimilitude at all. Naturally there will be counterbalancing forces where some verisimilitude is maintained but I do think the cycle, as I have described it, is at least a decent characterture of how things unfold in practice. It’s worth remarking that many on the outside will never see this invisible cycle because it is shielded from them by the fact that structures tend to stand up!
It seems to me that this unfortunate dynamic is likely to play out in any discipline where outcomes are dichotomous in nature and where the unwanted outcome (such as structural failure) is a low probability event by construction (and is unconnected to true understanding of the underlying concepts). It is certainly interesting to think about, and when the above phenomenon is coupled with human tendency to ascribe good outcomes to skill, and poor outcomes to bad luck, the result in terms of knowledge accumulation / dissemination may be quite unsatisfactory.
I think what I have just argued is that professional activities that become commoditized are likely to be degenerative over time. This would certainly accord with my experience in structural engineering and other fields where I have some substantive knowledge. And I wanted to see if you would agree or not. Do you have any stark counter examples from your professional life that you can recall? Do you think I am being unduly pessimistic?
There are two things going on here:
1. Corroboration, and the expectation of corroboration, as a problem. This relates to what I’ve called the confirmationist paradigm of science, where the point of experimentation is to confirm theories. The motivations are then all in the wrong places, just in general. Quantitative analysis under uncertainty (i.e., statistics) adds another twist to the vicious cycle of confirmation, with the statistical significance filter and the 80% power lie, by which effects get overestimated, motivating future studies that overestimate effect sizes, etc., until entire subfields get infested with wild and unrealistic overestimates.
2. The sociological angle, with students following their advisors, advisors promoting former students, etc. I don’t have so much to say about this one, but I guess that it’s part of the story too.
Also relevant to this discussion is the recent book, False Feedback in Economics: The Case for Replication, by Andrin Spescha.
Is there an example of a structure where this happens? Ie, there are many different explanations for why it is stable.
As for the “vicious cycle” part:
Smaldino and McElreath, “The natural selection of bad science”, https://royalsocietypublishing.org/doi/10.1098/rsos.160384
“In structural engineering: the structure either stood up or it collapsed, the beam either carried the load or it did not, etc. ”
I disagree. I live in earthquake country, and there are degrees of failure in structures during earthquakes, ranging from a crack in the sheetrock to complete collapse. Every significant earthquake is an experiment of sorts, with spatially varying degrees of shaking, and every such earthquake, engineers go around inspecting damaged buildings, getting the experimental results.
I agree with you. Further, I found this post extremely difficult and frustrating to read since it seemed full of conjectures/assumptions that just seem wrong to me. I’m prepared to be educated since i don’t really know anything about structural engineers or their training. But what struck me:
What you cite as the false dichotomous outcome – failure or success. Don’t many structures suffer from stress that requires them to be retrofitted, produce leaks, etc. It seems to me to be a continuum.
The statement that most structural engineers have not heard of Philosophy of Science seems wrong to me, and quite condescending. They may not be experts, but I believe most have heard of the term.
The inability to conduct tests seems true but only in the literal sense. Aren’t simulations used extensively? I would think that many designs are quite thoroughly tested, although the physical structure only gets the real test.
If the conjectures are true, then perhaps we should avoid surgeons who are not right out of school, since their skills probably degenerate. I actually try to choose surgeons with considerable experience. Surely there is some kind of countervailing effect here.
I find it highly questionable that a structural engineer would know pretty much anything about philosophy of science. Like probably they’ve heard of the term “philosophy of science” but that’s about it. I say this as someone who got a bachelors in math, had most of the courses required for a philosophy minor, then went back to school for a bachelors in Civil Engineering and eventually a PhD in CE. The 20 year old kids I was sitting in class with were smart, no question, but they had a LOT of required coursework and ZERO of it was philosophy. They’d have taken multiple quarters (at UCD we had quarters) of physics coursework so perhaps they’d have heard about it there, but I doubt it would have been meaningful. They also had almost no statistical training. So little that to do lab classes they’d plot stuff by hand on graph paper and fit a ruler through the data by eye.
Things they’d have had a bunch of coursework in would include:
Calculus
Physics, mostly classical mechanics and thermo and waves etc, not much quantum
Mechanics of Materials / solid mechanics
Materials Science (ie. microstructure of materials etc)
Fluid Mechanics
Fluid systems design (pumps, pipes, etc)
Specific building-code based design courses in structures generally, and, wood, concrete, and steel specifically as separate courses.
Engineering / applied Thermodynamics, Electrical Circuits
Some higher level math coursework, probably ODEs and partial differential equations
Possibly chemistry and biochemistry if they were water resources or environmental focused
Design of fluid control structures (weirs, dams, pipelines, channels, etc)
Soil mechanics
Soil structural design
Engineering economic calculations (lifetime costs, present value, investment calculations etc)
Pavement design
Traffic engineering principles, and/or code based design courses
Land Surveying
Economics (micro and macro, most likely to fulfill a social science requirement)
Technical writing (to fulfill a writing requirement)
Engineering students are pretty much the most busy students you’ll find on a campus, they have a LOT of courses required, and very FEW courses that are elective. They’re generally barely hanging on under the course load. they’re rarely going to be doing something like taking a philosophy class.
The “false dichotomy” of “failure vs success” may partially be because of the conservative nature of engineering. If a beam goes into plastic deformation it’s considered “failed” by most of the undergraduate courses. Of course, there may be virtually nothing wrong with the structure at all. Maybe a nail in your beam pulls out 2mm, maybe a bolted metal connection plate bends ever so slightly. That’s a “failed” connection to first approximation. Of course, when a real earthquake comes along like Northridge people go examine the more serious failures. We learned a lot from Northridge because a bunch of steel connections where there was a backing-bar on structural welds failed when they shouldn’t have, because the backing bar itself acted like a “crack” that concentrated stresses. After that they required welders to take extra time and effort to remove the backing bar after welding.
Still, in my lifetime there’s been exactly one Northridge earthquake type situation where major changes came out. So you could go a whole career as an engineer and not personally get much out of the learning that comes from such events. More like the industry as a whole and the educational materials for young engineers change and in 40 years eventually people move to new standards.
Allan’s description sounds more right than wrong to me. The codes protect people from failure by being relatively conservative. Not every building is built like the Disney hall with a 3d model and lots of earthquake simulations for example. Sometimes people just do relatively simple calculations. Sometimes the controlling factor is more like deformation/stiffness than strength. For example they were building a new science building on campus at UCD and had MASSIVE beams for the second floor. The steel design professors assumed they were storing very heavy stuff on the second floor. In fact, they had a light load that included a lot of microscopy. They wanted a floor that wasn’t going to vibrate so the microscope images were more stable. Most likely they could have packed the second floor with floor to ceiling books given the load capacity of those beams. In reality it was probably 1 person per 10ft x 10ft square sitting at a desk with a microscope. Essentially epsilon load compared to capacity.
While I think what Allan writes has some merit at the individual level, if you figure an engineer enters the workforce at say 22 and leaves it at 62, in 40 years they may decline in their ability to understand the fundamentals of WHY things work, that doesn’t mean they don’t know how to *do the job*. You can design a good structure if you know the design code algorithms even if you don’t know why they work. On a day-to-day basis being an Engineer is a lot like being a 1960’s moon-shot “computer” (ie. a human with a hand calculator). A LOT of stuff is done by hand on paper and pencil or using ad-hoc Excel spreadsheets (a scary LOT of these excel spreadsheets, likely with a lot of bugs in them). Carrying out the tasks without remembering why they work the way they do is more or less the job. It’s part of why I’m not an active practicing civil engineer and why the other PhDs I know are split fairly widely outside CE. The “algorithms” for structural design work really fine, the need for PhDs is when there isn’t “codified” methods to do stuff. So active research tends to be on things like optimizing the cost of running a bus network, or monitoring and predicting how sub-subterranean pollution spreads through aquifers, or how to minimize the risk of wildfire in forests or active control of vibrations to reduce the weight required for those microscope rooms… whatever other non-codified areas.
When it comes to picking a surgeon, the more advanced surgeon is someone who has repeatedly done a physical task over and over again, and learned to deal with situations that come up randomly that interfere with that task (like a bleed or a tendon rupture or whatever). You would NEVER EVER IN A MILLION YEARS ask a surgeon to teach people about the biology of how tendons develop or the effect of vitamin C deficiency on blood vessel strength or etc etc. They DO NOT know the fundamentals of the biology the way biologists do, unless they’re one of the rare ones who do some research and regularly interact with biologists who keep them constantly thinking about those topics.
Structural engineers are kind of similar. They can do the design, and figure out how to deal with situations that screw up in the field, but halfway through their career they’re unlikely to remember fundamentals of say microstructural changes after heat treatment of tool steels that they might have learned as an undergrad in a materials science course. For example, when I asked one of my engineering mechanics professors how Lagrangian mechanics worked early on in my studies of engineering they were able to say that they’d done a course in it 10-15 years earlier but they really couldn’t remember any of the content.
“Still, in my lifetime there’s been exactly one Northridge earthquake type situation where major changes came out.”
Daniel, that seems like a Southern California point of view. Up where I live Humboldt County, we had significant damage from an earthquakes as recently as last December, and from another one 30 years earlier. We have had minor damage from a bunch of others (e.g., our water tanks moved in one maybe five years ago, which taught me that the associated water lines need a flexible section.
John, not that there haven’t been other earthquakes, but there haven’t been other earthquakes where suddenly our understanding of a particular mechanical method of failure was changed all at once (that I can think of).
Before Northridge structural engineers thought steel buildings should fail in a very ductile manner due to the ductility of the material. In actuality the sections with the backing bars failed suddenly and brittle-ly due to cracks that only occurred due to the presence of this stress concentrator.
We have to distinguish between “expected failures” and “unexpected failures”. Sometimes a big earthquake comes along, and buildings fail, but in ways that wouldn’t be surprising. You can’t build all buildings to the standards of the Egyptian pyramids, we make trade-offs instead. Sometimes an earthquake is big enough that it exceeds what we expect a building to be able to handle. That’s “expected”. But sometimes it *doesn’t* exceed what we expect a building to handle, and yet the building fails anyway. That’s “unexpected”.
As far as I know there’s only one case like Northridge where all of a sudden a new mechanism of failure was discovered and widely disseminated in the structural engineering field.
In most earthquakes the main thing that might be of interest is that the quake itself could exceed the loads that were considered appropriate to use for design. So then the maps get updated to a higher level of peak acceleration or whatever. There also are situations where people in the past allowed certain types of designs which today we don’t allow. For example garages where there’s no shear reinforcement on the sides of the large door opening, or building a second story over the garage without designing for the new shear capacity. But the mechanics of these cases are all pretty well understood. Structural Engineers aren’t surprised when these older buildings built to laxer standards fall down.
What you’re adding beyond the ultimate failures I specifically mentioned are serviceability level failures. However, even with these types of failures the same trouble results. Naturally in both ultimate and serviceability level failures there are degrees of things going wrong (though all still classified as a failure – see Daniel’s comments below). I never meant to imply otherwise. And in this regard, serviceability level failures can sometimes be even more problematic in that the failure isn’t always so easily tied solely to the original design even if it was deficient. Questions regarding whether the failure was due to workmanship, materials, and the like are more likely to play a significant role. It is easier to blame things for why drywall is cracking other than the structural design than it is for shearing bolts at a brace bay due to unaccounted for thermal expansion.
Moreover, drywall cracking even if solely tied to the original design may actually not be an unexpected level of failure. We don’t design structures so that drywall never cracks – that would be prohibitively uneconomic. We design them knowing that there are some maintenance items that will crop up well before the ultimate life of the structure is reached (serviceability criteria are usually met using 1 in 5 or 1 in 10 year loads rather than 1 in 50 years for ultimate conditions). And sometimes a 1 in 5 year loading happens after or even during original construction (doesn’t mean the design was deficient)! This is similarly true for ultimate loads. If an underground concrete tunnel roof shears off – failing in ultimate limits states – but experienced a surcharge impact type loading that the operator of the property knew they weren’t supposed to apply, then this is a failure but not an unexpected one. In other words, we wouldn’t blame the designer, right? Same is true for natural loads such as hurricanes, earthquakes, etc.
It seems to me, and to most of my colleagues I have discussed failures with, that there are few true failures in the profession, and even fewer that get back to the engineer of record (not all failures are necessarily known and of those that are not all get back to the original designer). In my experience this makes it extremely difficult to appropriately judge one’s own expertise as things given the appearance of working whether your grasp of the underlying mechanics is sound or not. And when they fail, there are many other things to look at with suspicion other than one’s subject matter knowledge. This is generally not true for complete catastrophes, but as Daniel Lakeland points of those are not the norm. Perhaps this is all rather unimportant since by my own admission, things seem to generally work out, and all that this results in is a lower reliability index. Even then, perhaps not, as mistakes may err in the conservative direction on average.
I also admit the negative feedback loop in the OP is not always the case, and I was far too strong in my wording. Dale correctly admonishes my tone and overall style – I was probably having a hard delegated designer day. Learning absolutely does happen through failure and non-failure alike in the profession, and there is knowledge across the ranks. I should also add that in seismic country such as California & British Columbia, the structural engineers I’ve dealt with tend to be much more with it. I would suspect that has to do with the significantly more onerous requirements of seismic design.
After graduating from UC Davis with a bachelor’s in Civil Engineering I worked for a while at a forensic engineering company that did a wide variety of stuff. There was a reasonably smart older engineer near retirement named Wulff (German or Swiss I think) who did calculations and stamped drawings when needed. Mostly when designing repair structures for failing systems we investigated.
One day he came to me and said that a contractor wanted to use two pipes inside each other to hold up a basketball backboard in a school gym and was arguing that it should count as a pipe with walls basically twice as thick. Wulff was a pretty smart guy and intuitively knew that wasn’t right but didn’t know why so he came to me and asked me what I thought. I was quite flattered to be asked by the senior engineer in the firm for an explanation about materials mechanics but I was indeed able to give an answer. In bending the cross section of the material has strong shear forces that hold the material together and resist the deformation of the cross section. No such shear forces could be transmitted across the interface between the inner pipe and the outer pipe. The pipes would resist bending independently so the total force would be closer to the sum or the resistances of the two members not the resistance of a member with the sum of the two cross sections.
I do think university classes explain this stuff pretty well, but after a decade or so of doing civil engineering it would be easy for the practicing engineer to forget these teachings because a lot of what they do shortcuts the basic mechanics. Rather than say doing an integral one looks up in a table the numerical value for a given standardized member etc. Eventually you know the shortcut but don’t remember the long version of the calculation.
Perhaps structural engineering relies on a continuous stream of recently educated engineers to counterbalance the things the experienced guys have unlearned?
Edison did a lot of testing, and that practice was ingrained in the GE engineering departments–until Welch took over circa 1980. We had test data on some turbine materials under load and at temperature for 100,000 hours. When I joined the Large Steam Turbine department there were as many development and test engineers as design engineers (about 1000 of each). Welch told us to get rid of every office that had “Development” in its title (Materials Development, Rotor Development, etc.). We kept one last contingent of about dozen people in a new office we called “Advanced Design”.
Any way, I used to know a lot of engineers who worked with Green’s Functions, Lagrangians, and Laplace Transforms, and could do Fracture Mechanics and Fluid Mechanics calculations. I used Cowper’s Method for shear coefficients to be used for vibration analysis of vane shapes, and so on. If you need to build it, we will come.
Tesla had some good engineers. We like to test. Lack of testing comes down from the top, in my experience.
There was a while when testing was the big thing in CE. If I understand correctly in the 1960’s a lot of work went into testing concrete beams and determining a reasonable model for their failure under bending. The “equivalent stress block” method used in concrete design came directly out of fitting the model to data on broken beams if I understand correctly. Properly understood, this was Bayesian statistics being done back in the day.
example of how this model works is here:
https://www.structsource.com/analysis/types/concrete.html
The various parameters in this model such as phi and c and beta1 etc are taken to be point estimates but the choice of those point estimates was likely essentially done by a process of comparing the consequences of different types of errors and choosing values which minimize not the mean squared error, but the tradeoff between errors that cause collapse and errors that cause you to over-engineer the beam. Typically the calculation will estimate failure slightly before it actually occurs, thereby avoiding underestimates of beam strength.
There is a LOT of old engineering test laboratory information and test data available from the 1950’s onward which would be amazing to collect up and shove into the internet archive and/or build some Jupyter notebooks from. It would be a great way to teach Bayesian statistical methods for engineering to actually have massive databases of failure of steel beams, concrete beams, beam-columns, nail connections between plywood sheets under pull-out loads etc etc under real world conditions with measured material properties etc.
Who would fund the creation of such an archive, and who would give up their hoarded data to fill it? Far too many things are only found in scans of 1950’s printouts of hand drawn scatterplots etc.
Allan, if you’re reading this, could you provide more details about the unfortunate state of structural engineering? Specifically, I’m looking for examples of
A few friends of mine are chemical or structural or mechanical engineers with degrees from well-respected, notoriously theoretical programs. They, of course, complained all the way through undergrad about the difficulty and the seemingly pointless theoretical orientation, then arrived in industry and now complain about their bosses and the senior engineers’ lack of understanding. By their accounts, the senior engineers press essentially random buttons in ASPEN or whatever and can’t answer even basic questions. In one troubling case, a process was optimized by scanning through a multiple regression in matlab and picking the variable with the highest R squared value.
I ask for specifics because this has just never seemed possible to me. As you say, buildings still usually stand and products usually get manufactured inexpensively. Furthermore, the building designs seem to be getting more, shall I say, ambitious, and yet more stable over time. By your account, it seems like all this stuff should be getting worse. Are you saying we’re being basically saved by software and regulations designed by people with theoretical understanding 50-80 years ago?
> Are you saying we’re being basically saved by software and regulations designed by people with theoretical understanding 50-80 years ago?
I too would love for Allan to give some examples, he certainly has more real world experience than I do. But in my limited real world experience, yes this is a lot of it.
You could imagine for example that the people who build the software programs have a lot more theoretical understanding than the people who are 8 years out of undergrad school and have spent their entire time in industry pushing the buttons and letting the software tell them what to do.
Also, the building code has a lot of information codified into it, but it’s in the form of “do this not that”. It isn’t in the form of “here is the derivation for why you should do this not that”. So eventually no one knows why you should reduce the effective cross section of the beam by 13% under such and such conditions, but it’s there in the code so everyone does it (or whatever, pick some equivalent issue).
The reason we can do more complicated building designs now is that the software people are using allows them to do a lot more sophisticated analyses than what they would have done by hand 20-40 years ago. And the people who write the software do a very good job of writing it and it is reliable. But there’s a big difference between “SAP2000 tells me I should use these beam sections” and “I know why SAP2000 tells me I should use these beam sections”.
What’s the controlling load on a particular column in a complex building? Does the engineer even know anymore? I’m not sure. SAP2000 told them what to use, they used it and moved on. The pace at which you’re expected to operate is much faster because the automation is there to do it for you.
There is an emphasis in civil engineering on “spot checks”. Basically, pick a few random columns, do a hand calculation, and see if it agrees with what the computer spits out to within a reasonable margin for error. That’s a good and worthwhile thing for people to continue to do.
Somebody:
Building codes are made by a small number of smart people with deep knowledge to ensure that no unexpected structural failures occur. But most of the workaday engineering community is, shall we say, down the curve. Because the smart people make the codes and continue to revise them as new knowledge and innovations emerge, buildings are robust, even though the work of the people down the curve is poor when not constrained by codes. IOW, the quality of the codes hides the chump work.
I’d think a real project would have a lead engineer who knows his stuff, a bunch of mid-level guys who get it that the lead knows his stuff (some of these would be specialists of some sort), and a bevy of fresh out of an undergrad program blokes who are still figuring out what’s going on but do what they’re told*. In real life, this isn’t a problem. (The other day, I asked a dumb question about a building that was going up. The bloke I asked didn’t know, but he called over the guy in charge, and I got a really great answer. (About piles: they dig holes until they find a layer that’s strong enough (using a soil-equivalent of a Rockwell hardness tester), and that’s how deep the piles are. (17 meters in my part of Tokyo.))
The idea that there’s no one around who knows what’s going on is really really silly. IMHO, of course.
If you went in and did a random sample of the bodies on the project, you’d get what the original quote thought he saw. And you’d be wrong to think that there’s anything wrong with that. As I think the original quote is.
*: I have a nephew who’s one of these. He’s doing air conditioning systems for server farms. Something that, presumably didn’t come up in his undergrad mechanical engineering course. So he’s quite bemused by the whole game, and does what he’s told, and pulls down a really fat salary. If you asked him to build you such a system, you’d be unhappy. But I doubt his employer’s customers are unhappy in the slightest.
As a point of clarification, the building codes themselves are not usually written by the wise people you’re thinking about. The detailed and highly sophisticated work generally comes in the form of standards (e.g. CSA S16, CSA A23.3 or for my American brethren ACI 316, etc.) produced by professional bodies. Sometimes these are specifically referenced by the state or provincial building codes, and sometimes not. But the profession has a standard of care and if most engineers are designing to S16, even if not specifically referenced in your governing building code, then you better be doing that (or have good reason not to).
When I suggested the other day that we ditch in person teaching and replace it with on-line education, I suggest that because I’m banking on the same phenomenon that comes from building codes: the top people would design the software used in education, just as the top people design the codes used in building, so education can continue to improve just like buildings continue to improve, even though the average educator is slipping down the curve.
I’m hesitant to provide specific examples in this venue as they would be too identifying of other involved parties (among a few other things), which would violate my licensure board’s code of ethics. Though if you’re ever in Southern Ontario I’d be happy to grab a pint and regale you with some interesting stories. I understand that this means you have to take me at my word, and it’s understandable that you’d put a low prior on the accuracy of some stranger’s comments about their profession.
As far as being protected by standards, codes and software is concerned I would agree with Daniel in saying that yes, they hide a lot of ignorance (no doubt my own included).
The thing with FEM software is that if you know to wield it properly, you can design almost anything accurately enough. The firms doing the kind of ambitious projects you’re referring to are likely to have true modeling specialists, which are different from structural engineers who do modelling as part of their practice. There is a division of labour between engineering and modelling, and it allows them to confidently do these kinds of projects. This is different from the typical industrial project which has the structural engineer of record as the modeler, with details being delegated to other designers retained by the contractor (term of art is delegated design – basically EOR doesn’t get paid enough or wants to reduce their fee to obtain work so they download crucial details to the contractor).
The formula of use software + spot check that Daniel mentions is what I am accustomed to. But of course, the checks are usually for A) individual members and B) things that are easy to do by hand. That leaves a lot of room for there more complicated behavior of overall systems (e.g. lateral resisting system) to be wildly out of whack. But the thing is, even if they are deficient there is a lot of redundancy baked into our general building practices, and the use standard detailing helps save the day. Just because we don’t design for certain load paths doesn’t mean they don’t exist! And the structure usually has plenty that we don’t consider that can save us. Think of old 100-year-old barns that haven’t been maintained. Lots still standing but if you asked a structural engineer to quantify the lateral load path they would be pressed (and almost certainly not put a seal on it even if they came up with something). I’ve also seen load tests performed on concrete slabs with no reinforcing steel that somehow lasted for decades without an ultimate limit states failure.
As I acknowledge in a response up thread, my original comments were far too grandiose and all encompassing. The entire profession is not devoid of knowledge. There are people that know what they’re doing, and I’ve had the privilege to learn from some great engineers. But it is my impression that on average, things continue to get worse at the top and I think it has a lot to do with our inability to perform risky tests. Those with grey hairs have left or are retiring, and their replacements don’t know nearly as much and are without great ways to calibrate themselves. Perhaps this is due in some small way to the nature of delegated design as well. The devil is in the details but engineers of record don’t do much of those anymore – in certain North American jurisdictions at least.
Also, to touch on a point Daniel has brought up a few times. Sometimes being a productive engineer in practice doesn’t require one to have significant or detailed knowledge as long as the tools employed are routine, thoroughly verified, and generally the projects are of standard fair. Most industrial projects are like this! In my line of work, I usually deal with one-off, highly complicated restoration projects involving repairs to existing high-rise buildings where occupants remain during the restoration effort. Such as removing full building columns under the main footprint of a 20 story building (got 2 of these I’m doing right now). And I suppose it’s the poor engineering judgment displayed by my colleagues on these types of projects that fuels my rant in the OP, and general sadness for the profession; in my eyes there should be a greater standard of care for these projects given the very real risk of loss of life.
In undergrad soil mechanics classes we saw an entire video on the Kansai airport. This was built on landfill in the middle of Osaka bay in Japan. It was hailed as a great civil engineering project where a lot of heroic engineering efforts came together to make it happen.
Of course, looked at in any reasonable non-propagandistic light, Kansai airport is one of the great disasters of civil engineering. The ultimate maximum amount of soil settlement at infinite time calculated prior to the construction was something like 4m. Actual settlement of 4m had occurred before they even finished land-filling if I remember correctly. https://ascelibrary.org/doi/10.1061/%28ASCE%29GT.1943-5606.0001224 gives some details. Ultimately in that 2014 paper they predict something like 17-24 meters of total settlement by the end of the 21st century, basically 4-6 times as much as the maximum predicted at infinity. For the americans on the blog that’s about 80ft of settlement, or the height of a 8 story building.
In order to accommodate this settlement, the building was redesigned to have a bunch of jacking points, every few months they jack up the columns in the building and insert extra steel plates everywhere in the airport. If they don’t do this differential settlement of different columns puts tremendous strain on the structure and its connections and it would fail and collapse. So the whole building stays up by people literally running around the building with surveying equipment to figure out how much settlement there is, and then hydraulic jacking up columns essentially continuously. The overall cost of construction went through the roof compared to initial cost estimates, and ultimately for a long time there was almost no traffic because the cost of landing a plane there was so high. That might be different now.
So, yeah, there’s some feedback sometimes. :-) But the hilarious part is where they take credit for this being an engineering marvel because of all the amazing engineering they did like jacking up the building every few days.
You could look up the numbers: (tl;dr: it’s been 3 inches per year since 2007, it’s the 3rd busiest airport in Japan, and doing just fine, thank you.)
Sure, the problems were nasty but:
“Much of what was learned went into the successful artificial islands in silt deposits for New Kitakyushu Airport, Kobe Airport, and Chūbu Centrair International Airport. The lessons of Kansai Airport were also applied in the construction of Hong Kong International Airport.[22]”
And:
“On 17 January 1995, Japan was struck by the Great Hanshin earthquake, the epicenter of which was about 20 km (12 mi) away from KIX and killed 6,434 people on Japan’s main island of Honshū. Due to its earthquake engineering, the airport emerged unscathed, mostly due to the use of sliding joints. Even the glass in the windows remained intact. On 22 September 1998, the airport survived a typhoon with wind speeds over 60 m/s (130 mph).[25]
On 19 April 2001, the airport was one of ten structures given the “Civil Engineering Monument of the Millennium” award by the American Society of Civil Engineers.[26]”
Of course, it’s not all good news: global warming caused sea level rise will be worse than the sinking problem over the next 50 years (“officially”: my bet is that it’ll take less than 20 years for it and much of eastern Tokyo to be under water.)