Performance-Based Design "Philosophy" 7

[JohnRwals]

Hello!

Performance-base design concept was introduced to architecture study long time ago.
In fact, performance concept is one of hot topics in feminism also.
I wonder anybody can explain the history of concept/terminology 'performance'.

This concept has been impacting structural engineering industry a lot including seismic design.
But, I could find many informative papers released only when this concept was discussed actively in 1990s.
Do all industries including structural engineering accept this concept?
I would like to know how this concept/methodology has influenced building codes and engineering practice.
Is there any paper or book explaining how this performance-based design has been adopted/developed HISTORICALLY
in the seismic design?

Thanks!

_JRW

 

[SE2607]

Pick up a copy of ASCE 41 (-17 is the latest). In there, you will find some history of performance based design including ASCE 31, FEMA 178, NEHRP 1997.

 

[SWComposites]

It’s probably something dreamed up by clueless MBA types with fantasies of saving lots of $ by avoiding annoying codes and regulations. And in the end it costs more $ than its worth.

 

[JoshPlumSE]

It’s probably something dreamed up by clueless MBA types with fantasies of saving lots of $ by avoiding annoying codes and regulations. And in the end it costs more $ than its worth.

Not even close. The introduction of "performance based design" was introduced, I believe, in response to retrofitting of structures due to the Northridge and Loma Prieta earthquakes. When retrofitting these structures, engineers ran into issues associated with the fact that many of these older structures could not be upgraded to current code requirements. The key was to demonstrate that they could be cost effectively improved to an "acceptable" level of performance.

For the most part Performance Based Design was NOT intended to be used for new structures at all. There are a lot of FEMA documents that were part of the process of developing criteria for this. A lot of this was related to the fact that FEMA was giving out A LOT of money as part of disaster relief and no one had good criteria of evaluating what can or should be done for these damaged buildings.

If you're looking for the history of this, I'd start with the following:

FEMA 178 - Handbook for the Seismic Evaluation of Existing Buildings
FEMA 310 - Jan 1998 - Handbook for the Evaluation of Buildings (A Prestandard)
FEMA 356 - Prestandard and Commentary for the Seismic Rehabilitation of Buildings

One of the FEMA documents (not necessarily one of the three above) has an appendix that shows the first "non-linear static push over" analysis of an existing building. It was pretty crude because no software existed at the time to really perform this type of analysis. But, this example spurred the software industry to develop tools more suited for this type of analysis.

 

[bones206]

In fact, performance concept is one of hot topics in feminism also.

 

[WesternJeb]

Thanks for not letting the OP get away with that, lol. I was also bewildered by that statement.

 

[bones206]

Haha I hope it’s a typo

 

[Just Some Nerd]

Not a single thread is safe these days...

----------------------------------------------------------------------

Why yes, I do in fact have no idea what I'm talking about

 

[bones206]

If you squint really hard, feminism = finite elements. So I'm gonna go with that.

 

[JohnRwals]

I guess performance-based method is deeply related with reliability engineering, which is based on statistics/probability.
Do many people agree with performance targets like collapse prevention, immediate occupancy, etc.
Another problem is how we can quantify these goals?
Though people agree with these goals, what about data (sample space) provided for data analysis;
my data (work environment) could be completely different from day to day, sometimes worker to worker.
This means this philosophy is hard to accomplish consensus with confidence
and hardly possible to implement thru codification , though its spirit and theories can be understandable.
Think about the failure history of Metric vs Imperial Unit systems.
Though this conversion makes sense and very convenient, we rejected changing our long tradition!
That's why I am wondering this new experiment may be related with modern philosophy (intellectual landscape)
like feminism or something else.

_JRW

 

[bones206]

Ok, so not a typo.

 

[Luceid]

What do you mean by "do we agree with performance targets?"

Following a performance based design procedure can give you a lot of perspective on what happens post-elastic design, and can often show how code-based seismic designs may not necessarily produce failure patterns that are desirable once things start to hinge. If you design a braced frame building with every brace designed to 90% DCR - if the demand earthquake is bigger - what will happen to the building? What will actually fail first - and what are the consequences?

Agree with Josh's points as well - helpful for retrofit structures that need a boost to be considered "modernized."

 

[JoshPlumSE]

This means this philosophy is hard to accomplish consensus with confidence
and hardly possible to implement thru codification , though its spirit and theories can be understandable.

Yet, we do have codified performance based standards now. See ASCE 31 & 41.

I'm no expert in either of these codes. And, I don't think they're nearly as "cookie cutter" as the IBC is. But, there is a community of structural engineers that do this type of work quite often. I believe most places will require a "peer review" of the assumptions and such.

I believe Mandalay Bay (which experienced some severe geotechnical problems) was ultimately deemed acceptable after a Performance Based Evaluation.

We've got the Millennium Tower in San Francisco, which was also deemed acceptable (I believe) after a Performance Based Evaluation.

I believe the retrofit of a number of CA hospitals have used these codes / principles as well.

A couple of other thoughts:
a) Our building codes tend to be a force based design. Meaning you calculate an earthquake based FORCE (based on a lot of assumptions) and then design the structure to resist that force.

b) Transportation industry was one of the first to find the flaw in this. I remember a presentation by Nigel Priestly at one of the SEAoC conventions where he shows a short and very thick column that failed in Northridge. That failure led to the subsequent failure of the entire overpass.

c) The alternative is to use an Energy dissipation type of design or "displacement based" design. Priestly went on to explain how an ENERGY analysis of the system would have resulted in a very different design. The key was to make sure the elements could displace in a way that the earthquake wanted them to displace. And, that they could experience this displacement in a ductile enough manner to absorb the energy that the earthquake would impart.

d) These concepts (force based vs energy based design) are essentially the same as the difference between an Equivalent Lateral Force method of design via ASCE-7 and a Push Over type of analysis using ASCE 41.

 

[JoshPlumSE]

I also wanted to mention that this was sort of a natural transition for the concrete transportation guys. Because they had been retrofitting concrete columns for CALTRANS for years. Wrapping them in steel or composite jackets in order to increase their ductility during a seismic event.

In many ways, the civil / structural industry was slow to adopt these types of methods. If you look at the automobile industry, when did they start building in "crumple zones" in their cars in order to improve their crash protection performance. This is very similar, but easier. You have a known amount of energy for a car with a certain mass that impacts a wall. If you can absorb that energy in a way that doesn't severely injure the occupants then you've solved the problem. It's very similar with earthquakes. It's why "base isolation" was developed for essential facilities.

 

[SE2607]

JoshPlumSE - I believe ASCE 41 addresses displacement based design for ductile elements (displacement controlled components) and uses that force required to achieve that displacement for the design/analysis of the brittle elements (force controlled components). This was an eye opener for me. It made me think NOT to overdesign shear walls because the hold downs would be penalized.

It would be nice if the overstrength factor of ASCE 7 and the factors that determine the load on force controlled components in ASCE 41 would converge, but then I wake up.

 

[ALK2415]

ITS BASED ON ECONOMIC / SAFE DESIGN SO IT WILL BE (KEEP ON) MOST SUITABLE METHOD OF DESIGN (I THINK)
ITS HAD BEEN APPLIED TO ALL KIND OF STRUCTURAL MEMBERS DESIGN
I READ ITS STARTED IN DENMARK SOME WHERE IN 1960's (you should look there)

 

[JohnRwals]

I understand and agree with PBSD's goals.

Another issue is the divergence between ASCE7 and ASCE41.
Do we need to develop separate worlds between new and existing buildings?
How can we reconcile this gap?
My point is that building codes/standards are updated/revised without enough preparation, maturity.
New theories or philosophies are not necessarily better than traditional ones.
In fact, it is hard to catch up with rapidly changing codes/standards.
_JRW

 

[Smoulder]

Haven't read/understood all replies and not American ...

Performance design to me contrasts with prescriptive design. Prescriptive example is to put 0.2% bottom reinforcement is all spanning members. Performance alternative is to avoide brittle collapse which might be done with enough top steel.if top steel is more convenient. Performance opens more doors than the prescriptive tunnel. Is that how it's used in America?

 

[Luceid]

Essentially yes, performance based design typically requires more thinking - less smearing and therefore requires you to know more about your structure's performance.

There is already a gap between new and existing buildings - as it should be. If every building had to be constantly updated with every code cycle - people would stop building buildings and put their money somewhere else. The exception to this rule is the extreme case where the risk is large and widespread enough that it warrants some sort of "rapid" retrofit, such as the soft story upgrades to garages and upgrades to unreinforced masonry structures in California.

 

[IDS]

I don't see "performance based design" as being something totally separate to codified design procedures. Limit state design codes are based on performance based design procedures. In some cases requirements are specifically related to specified probability levels and performance requirements, and in some cases these are hidden in the load factors and materials reduction factors, but in both cases the levels are set with regard to probabilities and consequences of failure.

In my opinion it would be beneficial to make this connection clearer by adopting three limit states, rather than the two currently explicitly stated (with the third currently being implied, but not defined).

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/