Redundancy of Structural Systems in the Context of Structural Safety 6

[rlflower]

In light of various discussions on this forum regarding recent failures of decks, balconies, etc., I would like to open a discussion on the topic of redundancy in structural systems in the context of structural safety. My impression is that this design consideration is not well defined and is too often overlooked in critical situations.

A paper has been included in The Proceedings of the Twelfth East Asia-Pacific Conference on Structural Engineering and Construction — EASEC12, entitled "Redundancy of Structural Systems in the Context of Structural Safety". Please review the attachment. I would like to hear your comments on this paper in particular and on the subject in general. Thank you.

Link

Richard L. Flower, P. E., LEED Green Associate
Senior Structural Engineer
Complere Engineering Group, Inc.

 

[rb1957]

consider the approach in my industry ... damage tolerance (multi-loadpath structures, the emphasise on finding damage before it becomes critical)

another day in paradise, or is paradise one day closer ?

 

[IDS]

The link is well worth reading (also click on the Open Access link to the right of the title for links to other papers from this conference, at least some of which look like they should be worth a look).

This paper, and many of the responses to the recent threads on collapses in buildings, reinforce in my mind some changes that I think need to be made in the way we approach structural design:
[ul]
[li]Structural design codes state that prevention of collapse is a key objective of design, but almost all code provisions relate to the strength of individual sections or members in as-constructed condition, assuming good construction practice, with a nominal capacity reduction factor.[/li]
[li]There are limited provisions for consideration of load redistribution and collapse mechanisms, but these are either stated in very general terms, or are limited to certain specific extreme load conditions.[/li]
[/ul]

I believe that the Limit State design approach needs to be made more rational and consistent, and those codes that do not specifically adopt the Limit State approach should do so:
[ul]
[li]Design for strength and design for prevention of collapse should be recognised as two separate limit states, requiring completely different design approaches.[/li]
[li]Design for the Collapse Limit State should consider all potential causes of collapse, including material degradation, poor construction, loss of support, and extreme overloading, as well as earthquake loading.[/li]
[li]The level of reliability required for these events would be related to the potential consequences of failure, just as is already done for earthquake design.[/li]
[/ul]

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[IRstuff]

"My impression is that this design consideration is not well defined and is too often overlooked in critical situations."

This is not true at all in the context of any seismic design. Or, for that matter, in most high rises; the World Trade Center buildings were designed for catastrophic failures, up to and including airplane collisions. And even though the buildings were only designed with the existing aircraft of the day in mind, they survived the initial impacts with large planes long enough to keep fatalities to less than 4000. What the designers did not anticipate was a deliberate attack with planes with full fuel loads.

I won't, however, claim that a balcony is a "structural" system, so to that degree, balconies are not necessarily designed to survive something like an earthquake. I would, however, expect that the apartment building itself, is designed to handle something like a magnitude 7 or 8 earthquake, and should survive sufficiently that the people inside could effect a safe evacuation after a large earthquake.

That said, it should be pointed out that building design has a continual process improvement in the sense that previous systems deemed adequate may be revealed as inadequate, as demonstrated during the Loma Prieta quake, where it was revealed that a new, heretofore, unexpected shaking mode made certain design standards pathetically inadequate and fatal. Those structures have been torn down and replaced with better ones.

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529


Of course I can. I can do anything. I can do absolutely anything. I'm an expert!
There is a homework forum hosted by engineering.com:

http://www.engineering.com/AskForum/aff/32.aspx

 

[IDS]

"My impression is that this design consideration is not well defined and is too often overlooked in critical situations."

This is not true at all in the context of any seismic design.

I agree that procedures for prevention of collapse due to earthquake loads are well defined, at least in high seismic zones (much less so in areas where large earthquakes are very infrequent, but do happen).

My point is that the same approach should be applied to all the other events that can cause collapse.

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[thaidavid40]

"...the same approach should be applied to all the other events that can cause collapse."

So, can you define ALL the types of events which could possibly cause collapse - that is, beyond the types of loads, and load cases already required? This is beginning to sound a lot like making all structures "foolproof". I don't think such an approach is feasible, for as we used to say in the mechanical design of operable equipment: "You can't make a thing foolproof, because fools are so damn ingenious."

Thaidavid

 

[SlideRuleEra]

Redundancy is good, I've incorporated the general concept into a number of structures. However... suppose you have to address a number of unrelated risks on specific projects. Say:

1. Subsidence at any time due to poor soils.
2. Category 5 hurricane winds (157 MPH +)
3. Magnitude 7.0 earthquake

How to blend technical compromises and economic considerations for optimum results?

http://www.SlideRuleEra.net

http://www.VacuumTubeEra.net

 

[IDS]

So, can you define ALL the types of events which could possibly cause collapse - that is, beyond the types of loads, and load cases already required? This is beginning to sound a lot like making all structures "foolproof". I don't think such an approach is feasible, for as we used to say in the mechanical design of operable equipment: "You can't make a thing foolproof, because fools are so damn ingenious." wink

No, that is not what I am suggesting (in fact it is pretty close to the opposite).

I am saying that when the unexpected happens, structures should be designed to limit the consequences.

The extent to which you do that depends on the extent of the consequences.

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[IDS]

How to blend technical compromises and economic considerations for optimum results?

That's our job, isn't it?

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[SlideRuleEra]

Sometimes, until what should be done (for redundancy) exceeds the budget.

http://www.SlideRuleEra.net

http://www.VacuumTubeEra.net

 

[ChiefInspectorJeff]

I think if most things were built to codes and engineering standards and were maintained the problems would go away.
Florida had proposed that Architects inspect their own jobs. Now this would pose a problem in the real world, but would make a big difference in safety.
Government inspectors have no vested interest in babysitting incompetent contractors and being attacked for doing their jobs.
Contractors can not possibly know how most things must go together, only how they can.There will be errors on the order of 15-80%.
A contractor should have some knowledge of engineering and architecture and codes, but from my experience 15% do.
So structural redundancy might be an extra 100% (actually 300-400%)factor of safety, but may not matter when water attacks the members.
If the joist in Berkeley were doubled how much time would that have bought? Water it seems equalizes factors of safety.
So weep holes and wicks that can be seen would be a better option. Hot tar between membranes and up the wall would work. A balcony sees maximum loads 99% more that a standard roof or floor assembly.
So all aspects of the construction should be redundant.

Inspector Jeff

 

[rlflower]

ChiefInspectorJeff: Suggesting "all aspects of the construction should be redundant" may be overkill. I am suggesting focus upon specific situations where it may well matter. For example, in the specific case of the recent deck collapse in North Carolina, the more conventional method of floor framing - as described in the building code - would be to lap-splice the joists over the beam, which I would describe as a more redundant approach than the one employed.

Richard L. Flower, P. E., LEED Green Associate
Senior Structural Engineer
Complere Engineering Group, Inc.

 

[cvg]

Many larger and arguably more risky designs (dams, power plants, sky scrapers, large bridges, aircraft and automobiles) are evaluated using an Failure mode and effects analysis (FMEA) approach. Sometimes this evaluation will recommend redundant systems for risk reduction. Sometimes it will recommend a different type of design. It might recommend higher design loads or demand ductile failure. Regardless, for any type of project, if there is a failure mode that can be anticipated, it is up to the designer to evaluate the risk and if that failure mode is potentially life threatening, it should be eliminated (regardless of budget).

It is clear that from review of the recent deck and balcony failures that potential failure modes for both structures could have been very easily identified and the failure probability for both structures could have easily been reduced by the designer. In either case, the cost to the project would have been negligible.

 

[rlflower]

"It is clear that from review of the recent deck and balcony failures that potential failure modes for both structures could have been very easily identified and the failure probability for both structures could have easily been reduced by the designer. In either case, the cost to the project would have been negligible."

Agreed.

Richard L. Flower, P. E., LEED Green Associate
Senior Structural Engineer
Complere Engineering Group, Inc.

 

[KootK]

In either case, the cost to the project would have been negligible.

This is not quite so cut and dry in my opinion. I believe that the answer to the cantilevered wood balcony problem is that there simply should not be cantilevered wood balconies. Building envelope concerns and structural concerns for cantilevered balconies are always at odds and the results have been catastrophic. Of course, the "cost" of not having cantilevered wood balconies is a major sacrifice in aesthetic.

It is important to recognize an important truth about our profession: it is, and always has been, reactive. We don't check and rule out every conceivable mode of failure. We're not even smart enough to know what all the possible failure modes are. Rather, we check the failure modes that have given us grief in the past and for which we have devised methods of assessment.

To fly headlong into trying to preclude all modes of failure, including those that haven't yet given us grief, is to incur unnecessary costs and fall prey to hubris. I truly believe that there will come a point where the code bodies decide to preclude the use of cantilevered wood balconies. On the flip side, I would like it very much if the same code bodies would stop requiring me to calc out diaphragm capacities for wood structures as nothing ever goes wrong in that realm. I feel a bit ridiculous knocking my self out trying to determine diaphragm chord stresses while people seem to be plumetting to their death from cantilevered wood balconies every other week.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[cvg]

nobody said to "preclude all modes of failure" - only to eliminate those that are potential life safety issues or reduce the probability of failure to nil. Of course not all failure modes could be predicted. but the recent failures could have easily been. I will take hubris over a wrongful death lawsuit any day of the week.

 

[KootK]

nobody said to "preclude all modes of failure"

So what if nobody said this? I was expressing my own opinion, not railing against anyone else's.

I will take hubris over a wrongful death lawsuit any day of the week.

That's the trouble with hubris, it rarely makes anybody any safer. I'll take humility and faith in the collective wisdom of those that came before me.

I also believe in not attempting to fix things that aren't broken. 99% of the building columns out there are utterly non-redundant. Yet column failures are extremely rare. Nothing there to fix in my opinion.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[IDS]

I have given cvg a star, however:

if there is a failure mode that can be anticipated, it is up to the designer to evaluate the risk and if that failure mode is potentially life threatening, it should be eliminated (regardless of budget).

I think this overstates the case. Literally any structure has a non-zero probability of failure resulting in loss of life under some circumstances. There are three key points that I think tend to get overlooked in these discussions:
[ol 1]
[li]All structures should be designed to eliminate or reduce risk to life "So Far As Is Reasonably Practicable". Indeed, this is now a legislated requirement in Australia.[/li]
[li]Following existing codes of practice to the letter does not come close to satisfying this requirement.[/li]
[li]Determination of what is "reasonably practicable" is the hard part, but that does not mean we can ignore the requirement. In my opinion codes of practice need a radical change of approach so that engineers who have followed the requirements of the code ("So Far As Is Reasonably Practicable") will not be found to be negligent, even if the structure does fail with a resulting loss of life.[/li]
[/ol]

Earthquake design provisions (in high seismic regions) show that codes can be written to satisfy this requirement. There is no reason why codes should not adopt a similar approach to other unpredictable risks.

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[AnimusVox]

I'm a bit uneasy with all this talk of lack of redundancy from the structural side when it seems like the culprit is either negligence (I don't believe the wood was pressure-treated, at the very least maintained) or is an outlier such as the deck (residential, 25 years of salt-spray corrosion). I feel like there's already redundancy in most versions of the code in the form of safety factors, minimum strengths, and so-on. A different importance category is assigned to buildings deemed as critical, and some bodies such as the government have superseded the code and have their own, more conservative, requirements. We are also humble enough to revise the code if it proves inadequate in light of a large seismic/wind event. All this said, I believe that these recent problems lie outside of the structural domain.

 

[canwesteng]

I still don't follow how a simply supported deck would be much better, as if one support fails the whole thing comes down anyway. Or should wood balconies be continuous over the exterior support? I still see the same type of issue occurring if the building envelope doesn't properly protect the balcony.