What are the ASCE 7-10 Seismic Overstrength Load Combinations used for? 2

[jochav5280]

Hello All:

Would someone please enlighten me as to the purpose of ASCE 7-10's Seismic Overstrength Load Combinations? Are they to be applied just to the connections or to both connections and member design. I've not been able to find a good explanation of what these are used for. I've been applying them to my structural models whenever I choose a structural system with "omega" greater than unity, however, I generally also use the "rho" value of 1.3 per Section 12.3.4.2, so unless the Overstrength factor is greater than 1.3, it's load combinations will not control.

I appreciate confirmation that I am understanding these combinations correctly. Many thanks in advance!

Best regards,

jochav5280

 

[KootK]

Some parts of your lateral system are meant to yield and absorb energy and some parts are not. The theory is that the non-yielding parts continue to accrue seismic load until the yielding parts actually do yield. So, logically, a conservative design load for a non-yielding part is the maximum expected load that the yielding parts could reasonably absorb prior to yielding (over strength loads). It generally applies to things like collectors, columns, and connections but varies somewhat from system to system.

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.

 

[JAE]

The Omega (overstrength) factors should only be used where specifically called for in the code. Typically these are applied to connections and other items that would tend to fail in abrupt rupture (i.e. brittle things) vs. elements that can fail via yielding and help the structure absorb the seismic energy.

So look in ASCE 7 where it specifically refers to the load combinations with Omega.

Check out Eng-Tips Forum's Policies here:
faq731-376

 

[rn14]

Seismic design doesn't intend for a structure to remain elastic in an earthquake because doing so would result in grossly over-designed structures. While yielding is allowed certain components are to remain elastic. An easy example here would be to design a connection to the strength of the connecting member. Now the problem is that when we determine the strength of the connecting member we are conservative and under estimate its strength using things like phi and a lower bound material strength. So when your 50ksi steel connecting member is actually 60ksi your connection might be under-designed. Now combine that with the fact the things like strain hardening and strain rate also result in the connecting member having more strength than we accounted for and we end up in a bad spot. The summation of all these effects is "overstrength". Depending on your building system there are certain components that need to be designed to handle this overstrength.

 

[MacGruber22]

jochav5280, pay close attention to the code that governs the material which you are using. Overstrength load combinations in ASCE 7 are only half of the picture.

Example:
[URL unfurl="true"]http://skghoshassociates.com/SKGAblog/viewpost.php?id=49[/url]

 

[jochav5280]

Thank you all for your comments,

So I have some follow-up questions for you.

I understand that seismic designed systems are designed to allow certain components of the system to form fuses in order to dissipate seismic energy, which is how they are able to utilize higher R-values, which leads to lower base shears. What is not clear to me after several AISC seminars is how to actually implement these designs. So here are my questions:

1) Are the Overstrength load combinations meant to size the fuse elements or are they meant to size the elements that the fuse elements connect to in order to ensure that when the fuses fuse, the other elements won't yield? I've not seen anyone tie this all together in a clear fashion to date.

2) What's the workflow for using modern structural modelling software packages such as STAAD.Pro to size the fuse elements and surrounding non-fuse elements? In using the AISC-360 provisions, we typically apply the load combinations to determine the forces and design the members accordingly; with the AISC-341 provisions, are we able to simply run the seismic load combinations associated with the R and Omega values to size the components, or is there some external analysis that needs to take place? If so, what does that analysis look like, and does anyone have a good reference for what the design steps look like, (an example would be appreciated.) I've attended Rafael Sabelli's lectures and he indicates that some external spreadsheets need to be used to simulate the effect of the fuses on the surrounding elements, but the workflow is not clear. This would be a great thing for AISC to cover in an extensive building example.

3) We generally select seismic systems that don't require the AISC-341 Seismic provisions, (i.e. Steel Ordinary Concentrically Braced Frames, Non-building structures similar to buildings, ASCE 7-10, Table 15.4-1.) For those that are designing with fuse elements, how are subsequent replacement of the fuses planned after an event, (what keeps the structure stable when the fuses are removed?) With reinforced concrete structures, how are these structures designed to fuse without making subsequent repair post event possible? It's not clear how you would repair cracked concrete structural supports when they are integral to the structure.

Thank you for your time and help; I'd really like to understand how to design structures per the seismic provisions from the modern commercial software perspective, (it's pretty murky right now.)

jochav5280

 

[BUGGAR]

In short, you only have to use them where some specific (frequently obscure) code section specifically says that you have to use them. But of course they are hidden here and there throughout the code and not neatly collected anywhere. If you miss one, it will cost you.

 

[jochav5280]

Hello Buggar:

Well put; seems to me the only way one can learn this is to have a mentor who's done it before, but I don't have that, so the more experienced engineers tend to just b.s. a lot. What's worse is that some of them have designed structures with high R-values, yet have not followed the detailing requirements. They've also told me to ignore the Direct Analysis Method and just use normal linear elastic, without applying the amplification factors to account for second-order effects. I've been frustrated that the baton hasn't been passed very well to the younger generation at my office.

Best regards,

jochav5280

 

[KootK]

Regarding your latest questions:

1) The over strength design is intended to protect the non-fuse element.

2) I often design the lateral system elements requiring over strength outside of the big comprehensive software packages with semi-old fashioned spreadsheets. You can also run two versions of your analysis model: one designing to regular load cases and one to over strength cases.

3) Don't sweat the post earthquake fixes. We structural engineers are a creative bunch. We can dream up a way to repair most anything. The probability of seeing a design level wind or seismic event during a repair is very low. We take advantage of that while executing our fixes. Depending on the appetites of the client, we sometimes design our buildings to be easily repairable. Examples include post tensioned, self restoring concrete moment frames and steel coupling beams installed between shear walls thaT can be swapped out easily.

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.

 

[jochav5280]

Thank you KootK:

I've added some follow-up questions to your post below:

1) Thank you for this clarification.

2) So the Overstrength load combinations would be used for designing the non-fuse elements for the expected fuse forces and the regular seismic load combinations would be used to design the fuse elements? When I attended Rafael Sabelli's seminar, he performed an analysis that involved determining the fuse elements capacity by increasing their yield strength to an expected yield strength; was this just for connection design to ensure the connections elements would not yield under the expected fuse force or was this approach used to determine forces that the non-fuse elements need to be designed for? If this analysis was used for determining the forces that the non-fuse elements need to be designed for, then what is the purpose of the Overstrength load combinations?

I believe the analysis he performed was using an external spreadsheet as well; is the structural analysis carried out assuming that all of the fuse elements fuse at the same time for a given seismic load direction in a structures with orthogonal lateral force resisting systems and then take the resulting loads imposed on the surrounding non-fuse elements to determine if the non-fuse elements pass the code check? Do we need to have all of the lateral system fuse at once for both orthogonal directions? If not, how would be combine fuse loads for structure with non-orthogonal lateral force resisting systems that need to be designed for the worst seismic event in any direction?

3) Thank you for your examples; I work in the industrial field where our clients don't have an appetite for being out of production, so designing for a ductile seismic system is not something that we follow; as mentioned previously, we design for systems with lower R-values, which allows us to just utilize the AISC 360 provisions. I've corresponded with AISC's Larry Muir and he indicated that ductile seismic designs could actually be heavier and more expensive to detail/fabricate than non-ductile seismic designs. I'm not really sure how we would know this as the only way to is to design two structures, one with a ductile seismic design and the other with a non-ductile design, (i'm not sure if even universities have done this.)

Thank you again for your valuable input!

Best regards,

jochav5280

 

[Archie264]

MacGruber22 (Structural) 8 Feb 15 14:39
jochav5280, pay close attention to the code that governs the material which you are using. Overstrength load combinations in ASCE 7 are only half of the picture.

Example:

http://skghoshassociates.com/SKGAblog/viewpost.php?id=49

That's a pretty sad commentary on the state of our profession. Here we have conflicting requirements that have to be pointed out by someone who's discovered them because it's unlikely that the average engineer would find them because the codes have become so complex. If the codewriters' goal is to have more Greek letters than the Iliad they're close to succeeding. As our predecessors designed and built stuff without the benefit of such knowledge it's a wonder any building built before, say, 1990 has survived.

 

[JAE]

Archie, I hear what you are saying.
On the other hand, a different perspective might suggest that an "average engineer" who doesn't understand some of the nuances and underlying concepts of seismic design probably shouldn't be attempting to design something over their head.
Our profession should be just that - a higher level of learning developed beyond blinding following a simple code cookbook procedure.
I agree the codes have gotten more complex - and probably stupidly trying to allow design to the knat's ass. But there's something to be said for the knowledge base of true professionals who have taken it upon themselves to educate themselves and know what the methods, philosophies, and procedures are behind the code.

Check out Eng-Tips Forum's Policies here:
faq731-376

 

[BUGGAR]

Regarding mentoring, I used to mentor younger engineers, but with the code changing every three years, they have effectively made us old engineers ignorant every three years and we have to learn that stuff all over again!
Advice? "Love the Rivet" (read Robyn Davidson). Every job I start, I have to Xerox entire code sections of all applicable codes and highlight every little thing that affects my design. When I miss something, and I do, It embarrasses the hell out of me, but I have to respect the young people who catch me.

 

[MacGruber22]

Buggar: I agree. We have no choice but to take the time to track-down every possible related code provision whenever we start on a design. Now, I may not be able to bill that time to the project, but for me that task falls within my own continuing education, regardless of how annoying it can be. I would rather spend my entire weekend to bone-up on code provisions so that I walk in Monday morning confident about what I am doing.

Also, for me, being acutely aware of upcoming code versions well in advance eases that process. I find that I feel better about code changes when I become aware of them before the code has become public. Subscribing to all the code email newsletters, social media, etc. keeps me in the loop before I have a specific design task. A lot of engineers I know spend so much time huffing, puffing, whining, complaining, and resisting all the new nuances of the code changes. I think that general attitude sucks, regardless how annoying it can be. Replacing annoyance with curiosity will serve us, those we mentor, and our profession much better.

To go back to the OP's original topic; for those very practiced engineers who are feeling weak in the theory of overstrength design and how it relates to attempting to "force" a building's components to react in a certain manner during a seismic event, it gives an opportunity for both people to learn something. I remember asking a similar question to an "old" engineer when I got out of school, and he scoffed at the topic and dismissed it as useless. Well, that is the crappy attitude that does nothing for young engineers. Sure there are plenty of times when engineering judgement can rationally justify not following the overstrength code provision (or others like direct analysis) to the "T" - when base shear is very low, when you recognize a lot of system redundancy, inherent conservatism, etc. etc. BUT!! A "green" engineer doesn't have that engineering judgement yet, so they need to treat every little part of code as unyielding LAW, and THEN slowly build-up the engineering judgement that is bestowed upon them by a good mentor, who reviews calculations, thought process, etc.

Sorry to digress.

 

[KootK]

So the Overstrength load combinations would be used for designing the non-fuse elements for the expected fuse forces and the regular seismic load combinations would be used to design the fuse elements?

Yes with one caveat. Only the non fuse elements that are part of the lateral load resisting system get over strength design. Your typical infill beam etc does not.

When I attended Rafael Sabelli's seminar, he performed an analysis that involved determining the fuse elements capacity by increasing their yield strength to an expected yield strength; was this just for connection design to ensure the connections elements would not yield under the expected fuse force or was this approach used to determine forces that the non-fuse elements need to be designed for? If this analysis was used for determining the forces that the non-fuse elements need to be designed for, then what is the purpose of the Overstrength load combinations?

I can't really say, for certain, what Sabelli was up to without having been there. From what you've described, it sounds as though he may have been using his special analysis to demonstrate how over strength forces arise in a structure with "blown" fuses. I wouldn't expect over strength factors to be applied over and above the forces from Sabelli's special analysis.

is the structural analysis carried out assuming that all of the fuse elements fuse at the same time for a given seismic load direction in a structures with orthogonal lateral force resisting systems and then take the resulting loads imposed on the surrounding non-fuse elements to determine if the non-fuse elements pass the code check? Do we need to have all of the lateral system fuse at once for both orthogonal directions? If not, how would be combine fuse loads for structure with non-orthogonal lateral force resisting systems that need to be designed for the worst seismic event in any direction?

This is a whole lot of question. ASCE7-10 provides guidance on directionality and orthogonal combinations. In general, the concept is to consider the formation of a complete mechanism in your LFRS and base your over strength considerations on that. It's tough to get more specific without considering a specific structure. And there are exceptions an nuances. For example, in a 30 story chevron braced frame, you don't have to design the first floor columns for a monster axial load resulting from over strength yielding of the braces up all 30 stories. That is considered improbable. I highly recommend the following references if you crave more detail:

1) AISC Seismic Design Manual
2) 2012 IBC Seismic Design Manuals

The first is ridiculously expensive but worth it. The second is very reasonably priced.

'm not really sure how we would know this as the only way to is to design two structures, one with a ductile seismic design and the other with a non-ductile design, (i'm not sure if even universities have done this.

Yeah, it can be a trade off that's difficult to evaluate. Here're my personal guidelines to which there are many, many exceptions:

1) If the project is in a low seismic area, it rarely pays to design for high ductility.
2) If the project is in a high seismic area but low rise, it rarely pays to design for high ductility.
3) If the project is in a high seismic area and is mid-rise, high ductility design usually does pay dividends.
4) If the project is very tall, wind governs in most any seismic regime so it's a moot point.

If your clients are very concerned about downtime, you might look into performance based design. At the 2500 year return period event, you may still be designing to high ductility demands. At a 50 or 100 year event, however, you may choose an essentially elastic design.

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.