Complexity of Engineering 19

[vmirat]

Recently, I received the latest edition of the Steel Design Manual from AISC. While purusing through the pages of this weighty tome, I reflected on how complex the field of engineering, specifically structural engineering, has become over the years. The respective institutes, societies, and such have spent lots of time and money developing more and more exacting methods for determining forces on structures. I'm wondering if we are getting to a point of diminishing returns? The more complex we make the process, the greater the chance for error. We develop computer programs to help us solve these complex problems but at the same time don't trust them. Are we to the point of "measuring with a micrometer, marking with a crayon, and cutting with a chainsaw?" And then I thought about the entire construction process, of which we are only a part. What about the builder? There was recently a very long post concerning additional certification for Structural Engineers (SE) based on the premise that the structural world is so complex that addition qualifications are required to insure quality. But what about the builder? Should there be additional qualifications for those involved with the building of these structures that we spend so much time and mental energy designing?

 

[Bagman2524]

A few years back while working the Turbine Generator side of a power plant outage, one of the Westinghouse Tubine experts/consultants told about a 100 ton turbine moved over a bridge in Spain built by the Romans. I wonder how many of structures built today will be functional in 2000+ years. My guess is not many. Everything today is about being economical which requires more exacting engineering methods, which has its' advantages and disadvantages. Our codes today are good a requiring for various types of loads but don't seem to consider durability over time.

 

[archeng59]

Are we to the point of "measuring with a micrometer, marking with a crayon, and cutting with a chainsaw?"

perhaps. we design building structures using the some of the latest technology available. yet, the contractor employs the same kind of people to build with basically the same technology that was used by our great-grandparents.

 

[lkjh345]

Vmirat:

I don't think I could agree with you more!!

Unfortunately, IMHO, I think the practicing engineers have let the academic folks take over the code writing committees, and they have gone off and made things more and more complicated in some eternal quest for 'economy'.

Designing to the last bit of capacity is fine as long as your not the person who has to stamp and sign the drawings and say that no matter what happens over the next 50 or 75 years, the building isn't going to fall down. I think when you are that person, all of a sudden your not so ready to go right to the edge.

 

[whyun]

Every pages added in newer editions make efforts to justify by rigorous analysis that less material can resist demand. I don't agree with blind conservatism but calculating to anything more than 3 sig-figs is a wasted effort. Surely, optimized member may perform to minimum code standards but one size larger member performs better with less "engineering" effort.

 

[FSS]

I think the "publish or perish" mentality gets mixed into the codes and standards. Unfortunately, academic engineers do not always make for good practicing engineers. Working through the various codes (and constant tinkering) for a living is tough!

 

[JStephen]

It's amusing to me how some of that stuff works. On the seismic, you pull a factor off a map, but eventually have to multiply by 2/3 to get the factor you need. Why didn't they just multiply the map factors by 2/3 in the first place? In certain areas, no seismic design is required. Why don't they just mark that part of the map "no seismic required" instead of making you pull numbers off two different maps and compare to a particular requirement in the text?

When you start reading about where these maps come from, it sure sounds like some voodoo engineering going on. That 2/3 factor, as best I can tell, came from a "That's too high so let's multiply by 2/3" logic. The R and I factors look like they came from a committee process, and could just as easily have been 50% less or more, and no one would know the difference. But hey, we've got maps accurate to two decimal places, so lets calculate all those forces to 8 decimal places and feel good about it.

I find the newer codes read suspiciously like the tax code, and can't help but wonder if there's not some connection. "Design of this widget shall be per some other paragraph as modifed by this paragraph but not less than this other section or greater than this equation", etc.

 

[chichuck]

I give a star to Vmirat,

"measuring with a micrometer, marking with a crayon, and cutting with a chainsaw" is definitely quotable.

I also give a star to JStephen for his insight. I heartily agree with his assessment of the current seismic requirements in ASCE 7-05. And your thoughts on the similarities between engineering codes and the tax code is food for thought.


regards,



chichuck

 

[swearingen]

Although I generally agree with what's been said (compare the thickness of an ASCE 7-88 with a 7-05), I have to say that the code committees have put an honest effort to make some of the codes more readable and useable. Part of the reason the American steel manual has gotten so much thicker is because of the ASD/LRFD debate - they had to include both. Also, they added many tables and User Notes to try and help out the practicing engineer. Comparing the seismic sections of ASCE 7-02 to 7-05, you can see that the effort was to provide separate chapters for separate subjects so that you could just read the one small chapter dealing with your specific area. This replaces reading a large chapter dealing with all seismic issues and trying to pick out what's applicable in your case.

From what I've heard and read, the code committees have listened to these arguments and have put much effort into making the codes as practical as possible while still being state-of-the-art.

 

[DaveVikingPE]

Clearly the worst thing to happen to structural engineering was ultimate strength design...

 

[RARWOOD]

Thread reminds me of an engineer at United Steel Products over ten years ago. At that time the engineer had submitted a report to ICBO to obtain a NER on a beam hanger. The report contained both testing data and design calculations for the hanger.

The evaluation person wanted USP to run a finte element analysis on the hanger. The USP engineer response was that finite element was just another approximation of an approximation.

I am in agreement with the comments above, and feel that the codes have gotten so complex to make them practically useless for the design of the vast majority of structures that most engineers deal with.

 

[JAE]

DaveVikingPE - don't be so quick to condemn ultimate strength design. I seem to have read somewhere that "ultimate strength" was actually the "first" method used back in the 1700's or 1800's or some such time. It was only after further research helped engineers understand the concept of stress that the use of an "allowable" stress was developed and the ultimate concept was discarded. After statisticians got involved with research, the idea of a uniform probability of failure then reared its ugly head and ultimate strength design came back to the front and center.

 

[UcfSE]

If society was willing to pay $2 for something they could get built for $1, then we wouldn't need engineers and their complexity. Going a little further, why continue to build for $1 when we could build for 90¢? Why continue to use stone-age technology if we have the knowledge and tools to do a better job? We have the processing power to actually do things right, to actually have a grasp of what is really going on with material behavior without spendign weeks working out matrices by hand. I see no reason to rely only on equations that can be worked out in your head or on your fingers and toes. Yes, we should be aware, blah blah, I know all that stuff about rough checking, error checking, hand checking, garbage in = garbage out ad nauseam. The key word being check, not use. My point is only if we can do a better job or a more true-to-reality job, I say we should do so. When we stop advancing, we start stagnating.

 

[L775]

Codes are only minimum requirements. The EOR is still responsible for his design and plans, regardless whether the code was applied accurately or not, and whether economic objectives were achieved or not.

 

[rb1957]

related to this thread, airplanes have long since lost the capability of lifting their own paperwork.

in a place where i worked for many years, there was a microfiche filing cabinet ... a model airplane from the late 40s to up part of one drawer; in the 50s a design required 1 whole drawer, in the 60s several drawers, 70s a whole side of the cabinet (1/2 the cabinet), in the 80s they had to but a neew (much bigger cabinet (and this didn't include the computer models which would have been on a bushel of tapes), and so it goes ... today (well some time ago) they gave up on microfiche and now everything is on-line, thankfully disc memory is very cheap.

 

[cvg]

and I hear "they don't build em like they used to"...

 

[DaveVikingPE]

Also, we'd be better off without calculus.

 

[henri2]

When I was introduced to limit state structural design in college, one thing that was hammered home repeatedly, was that there had to be a comensurate improvement in QC/QA in the construction phase. This begs the question: are construction QA/QC methodlogies being enhanced to match the increasing level of sophistication in structural analysis/design?

 

[JAE]

henri2 - absolutely not.

 

[YoungTurk]

"airplanes have long since lost the capability of lifting their own paperwork"

I like that one rb1957! Funny part is in the end it really comes down to those folks signing/stamping the paperwork and how conservative they really want to be, like lkjh345 says.

Are the codes too complex? Try reading a FAR or similar regulation and tell me how many times through it takes to get the gist of it. Then see if how you interpreted it matches the next guy's interpretation. Seems even the FAA doesn't have much luck even getting internal agreement.

I've always watched in amazement as legislators continue to pass law after law, lengthening the legal codes ad infinitum - and to what end? I guess its just dawning on me that the same goes for professional codes and standards. At some point, you've just got to start over; pick through the rubble for the real substance and make new. Perhaps all this bureaucracy has something to do with the life span of the "great" (read large/dominant) organizations and societies? Stagnation indeed.