Why Haven't These Buildings Toppled?

Tinstaafl

What's the pitch & width of that roof? If it's really low-pitched and wide, I could see some over-engineering. Add CA building codes to that, and I wouldn't be too surprised to see them requiring a 2' steel I-beam.

JumboJack

I don't remember it was quite a few years ago.It was a pretty flat roof.Not very wide though....

__________________
in texas with framing and cornish people will do it for 3.00 a foot. What do yall think about that? Just laber

Anti-wingnut

And you would prefer to be in a masonary house built to local standards outside the US, to an American wood framed house in an earthquake?

JumboJack

Tin...Did you thank me in that last post for the avocado paint?Unfortunately I can't take credit for it.....

__________________
in texas with framing and cornish people will do it for 3.00 a foot. What do yall think about that? Just laber

Tinstaafl

Nah, that was for the fine illustration of how to use a stepladder.

S'pose that avocado's ripe yet?

Kanding

Building codes are not a new invention. Some of the oldest known building performance laws are contained in a larger document called the “Code of Hammurabi”, which was written around 1700 B.C. Since then, many other building performance laws were written in various parts of the world for the last several thousand years. I am not a historian, but I would guess that in the US you may find that various local governments have had some type of building code implemented before 1900 as well. Also, the IBC is based not just on the UBC but on a conglomeration of BOCA and SBC.

It is true that prescriptive rules that govern construction not requiring engineering design (such as those found in the IRC) are quite conservative. However, structural standards referenced by the IBC (such as ASCE7 for loads, ACI-318 for reinforced concrete, AISC for steel, NDS for wood, etc.) are certainly not out-of-line in terms of overdesign. These safety factors, or more properly called load and resistance factors in modern LRFD format, were developed based on past experience of acceptable risk, as well as an extensive probabilistic analysis of maximum loads and expected variation of material resistance. For example, structural components designed to modern structural codes generally have a reliability index between 3-4 for a 50 year maximum expected load combination. Practically, this means that there would be a theoretical probability of component failure of roughly 1 in 5,000 over 50 years of service life. Or in other words, roughly 1 out of every 5,000 components built would be expected to fail in this period of time. And these probabilistic assessments do not include any design or construction error.

In fact, most field studies have shown a *higher* rate of structural failure than predicted by these theoretical assessments. Note that failures do not necessarily mean a complete collapse or loss of life.

Codes specify minimum acceptable standards, not best practice. It is always the ethical duty of the engineer to ascertain whether those minimum standards are sufficient or not. I would highly recommend to avoid the notion that building to code results in a structure that is exceedingly safe. Codes are periodically adjusted, among other reasons, to rectify structural failures that occur at rates greater than what was expected or considered acceptable.

Tim Garrison PE

Your points are well taken. It is true that if a big enough earthquake, tsunami, or hurricane strikes even a structure built to code may fail. And structures not built to code may fail as well.

Structural design, like driving a car, is a gamble. You're playing with odds every time you engage in either activity. I do not suggest building to any sub-code standards. My suggestion, stated a couple times in my article, is to build to code but no more.

If anyone wants to be safer than code, they have the freedom to build to a higher standard - codes are minimum standards. Of course, they'll pay more doing so. When I look around at the vast majority of buildings in this country not even close to code and see that they're still standing, I have a hard time suggesting that people spend extra money exceeding code.

I think that code writers and engineers have for far too long been silent on this side of the story. All they ever show are the disaster photos. Photos representing some tiny, tiny fraction of the buildings built.

My clients never tell me, "Tim, please exceed the building code - I'm happy to pay extra to be extra safe." But they do say, "Tim, please do what you can to save me money." Ethically, I'm bound to design to code, which I do. But I don't choose to spend my client's money on things which have an infinitesimal probablility of ever doing them any good.

bwalley

Just because a building is standing, does not mean it is structurally sound.

BuiltByMAC

You keep on designing for the bottom dollar client, I'll keep on building my structures as 'bombproof' as I can so if and when they get overloaded (parties on a deck), the structure doesn't fail because I built to code minimums.
My clients don't tell me they're happy to pay extra to be extra safe - when they hire me to build their project, that's included in my price.
I don't give them another option.

As I read through this thread, I can honestly say that's the worst excuse I've heard of for building a crappy structure -
"Other buildings were built crappy and they haven't fallen down yet."
Well, now Jimmy, if other buildings ran and jumped off a cliff, would yours jump too?

Mac

Tim Garrison PE

Well said, Mac. And actually I agree that no one should be building crappy structures. Your comments are testimonial to the misunderstanding that our industry suffers about building codes.

You should come to one of my structural concepts seminars so that you can learn what "Built to Code" means. In short it means that there is a factor of safety of 2.5 in everything built to code. Which means it is 2.5 times stronger than it needs to be to avoid failure. I think 2.5 times stronger is plenty, and that's my only point.

bwalley

The building code gets ammended due to lobbying efforts by large builders, they don't lobby to make the code better, they lobby to change the code in some cases to save money on building costs.

Kanding

No, this is not true. I do not know where you heard that. Safety factors vary substantially depending on the type of component, load, failure mode, and code.

For example, the modern versions of all the major structural codes, which are referenced by IBC (ACI-318 for concrete, AISC for steel, and NDS for wood), are written in LRFD format. A common governing ASCE 7 LRFD load combination is 1.2 DL + 1.6 LL, with resistance factor phi which typically ranges from 0.65 (columns in compression) to 0.9 (bending limit state, tensile-controlled failures). This would give an equivalent ‘safety factor’ anywhere from about 2.46 to 1.33, depending on what DL/LL proportion is applied and the anticipated failure mode.

Moreover, the safety factor itself is not useful for directly assessing practical measures of safety (i.e. chance of a failure event), which requires the introduction of probabilistic methods. No matter how high your safety factor is, there is always some probability of failure--no structure is ever 100% safe. Modern (structural) code design generally provides a theoretical component reliability on the order of 1 in 5000 component failures (i.e. exceedance of a limit state) over 50 years. Given the number of components and structures in existence, this is not a terribly small rate of failure. Again keep in mind that a component ‘failure’ does not usually result in collapse in most redundant structures.

KennMacMoragh

What's the safety factor on this?

Tim Garrison PE

I agree that 2.5 is not a hard, fast number. But in the world of wood, where most viewers of this thread live, it is a reasonable number.

And truth be told, there are a few, select instances where I design in excess of code minimum: floor deflection (bounciness); and roof dead load (I leave allowance for re-roofing over existing). In all other instances I try to not to exceed code.

What about you, Kanding, do you pump up your designs in excess of code? If so, is this something you tell your clients? I'm wondering why you're so resisting my very simple point. Perhaps you think our codes are too lenient - that we need more stringent codes?

KennMacMoragh

I agree, those are two things I think are just under designed in the code. I kind of understand them not upping the standards for floor joists, a bouncy floor isn't a safety issue, just an annoyance issue. But I don't understand the low requirements for roof loads. When we had all the roof collapses in Washington about ten years ago from all the snow, it was investigated and found that the roofs were not overloaded by very much. If that's the case, why not up the standards a little bit? Their excuse was, "Well people didn't maintain their roofs", seems kind of dumb. I'm sure there is a lot of overkill stuff in the code too, but that's why they change it every three years.

JumboJack

See,that's why I just build em like the plans tell me to....

__________________
in texas with framing and cornish people will do it for 3.00 a foot. What do yall think about that? Just laber

Kanding

Tim,

I would agree that SF=2.5 probably falls within the range of some wood components designed according to NDS. To answer your question, I sometimes design beyond code minimums; it depends on the situation.

I would suggest that we are paid to give our analysis and design recommendations, and part of this work involves deciding when to exceed code minimums. I almost always make these decisions without involving the client, and they usually result in rather minimal increases in cost. However, in one instance, I recall designing a slab foundation on what I considered to be questionable soil (high plasticity index). I recommended a slab thicker than usual, a higher strength concrete, and much more steel than usual. In this case, the additional cost to the slab was substantial, and I told the client that my recommendations greatly exceed the typical minimums, but this is what was needed. In that case, he took my recommendations. However, I can tell you that if he did not, I would not have sealed the drawings.

In another instance, I was approached by a developer who wanted to build prefabricated housing units on the Gulf coast after hurricane Katrina. Given the catastrophic failures in the region at the time, I designed the structure conservatively. I gave him a preliminary design, and he was disappointed in the result, as the member sizes were bigger than he expected. I only told him that I could not reduce the member sizes. He paid me at twice the hourly rate that I billed, believe it or not, but I never heard from him again. He may have went to another engineer, or gave up on the project, I don’t know.

It never happened before, but if a client did come to me and insist that I design to code minimums when I though it was inappropriate, I would decline the job and recommend a different engineer.

Of course the adequacy of current standards is a completely subjective question: i.e. what level of risk are you willing to accept? If your opinion is that existing standards are overly-conservative, that is fine, I really have no argument with you on this. Given that code safety standards are rather
arbitrary, I really don’t have a strong opinion on it.

Have a good weekend.

KennMacMoragh

What do you mean? What else do we have to do besides talk about engineering on a Friday night?

hughjazz

who's on first?
.....................................I don't know is on third.

hughjazz

I'd give it a 5.. 5.5 tops