Seismic Requirements 1

[mgg]

In UBC-97, another basic sturtural system is mentioned, cantilevered column building systems. I would like to know if ductility or seismic frame requirements (UBC Section 1921) are applicable for the cantilevered column elements. I am applying this structural system to a precast covered walkway, with hinge-type connections at the top of each column.

 

[mgg]

Does anyone use UBC 1997? Is that why there are no replies for this post?

 

[ishvaaag]

I don't use it, mannyg, nor I have but UBC 94 (I practice in Spain). In any case the structure being in a high risk of strong earthquake zone in my view would mean that any of the structural elements of one structure in such location needs to comply with the requirements for the same stated in the code, that may include specific requirements of ductility for the columns.

 

[JAE]

mannyg:

UBC 1997, Section 1921.2.1 includes wording that tells you when the entire section 1921 is required to apply.

Whether you have a moment frame, shearwalls, or your cantilevered columns, the section (1921) applies if you are in the indicated seismic zones.

In other words, its not the kind of system you're using, its the zone you're in that determines its applicability.

 

[mgg]

You are right, Jae and Ishvaaag. I'm sorry I referred to the whole section 1921. I was really after the detailing requirement for confinement, section 1630.8.2.2.

Just to give background information. I am designing in Guam. We comply with seismic zone 4 and a basic wind speed of 155 mph. And I am evaluating the structural stability of an existing structure.

An Ordinary Moment Resisting Frame is defined (UBC) as a frame not meeting special detailing requirements for ductile behavior as opposed to a Special Moment Resisting Frame. It is possible that an existing building can be in a high seismic area without the proper detailing requirements. To conclude that the structure requires retrofitting may not be warranted. There are a lot of buildings on Guam that do not conform to the aforementioned requirements, but have withstood a few earthquakes (8.2 being the highest richter magnitude) and supertyphoons (sustained wind speeds in excess of 145 mph). The code applies a lower R-factor for a OMRF than a SMRF, essentially increasing the seismic base shear. In other words, penalizing the use of a OMRF. Therefore, applicability is not only based on zonation but also the type of structural system.

I understand when designing a new building you want a MRF to perform in a ductile behavior and take advantage of current technologies and methods. This is why I am at a brick wall when it comes to existing structures. I would appreciate any views on how to approach such evaluations.

And back to my original question. Does cantilevered columns require special seismic confinement similar to elements of a SMRF or not?

 

[Taro]

Good question. I think the reason that you are not receiving a definitive answer is that there isn't one available. The lateral system response factors in Chapter 16 are developed by one organization, while the design and detailing requirements in Chapter 19 are developed by a different organization. Chapter 19 does not specifically address requirements for cantilevered column elements.

For new design, I believe you should comply with the special detailing requirements as a matter of good practice (in the absence of other applicable code requirements). But it's a different issue for existing structures. It is my opinion that you can justify not needing special moment frame detailing because the lateral system response factor, R, is only 2.2 (compared to 8.5 for special moment frames). Therefore, we are already analyzing and designing the system in recognition of limited ductility.

 

[Qshake]

There are many existing facilities in high seismic areas that were never designed to even minimal seismic requirements. I would expect more so in developing countries.

By reducing the "R" factor you're only increasing the elastic force (base shear), which for an existing detail is suspect to begin with as it never had the detailing to achieve the ductility associated with the increase in elastic force. Thus another load path or supplemental load path must be evaluated or the necessary measure taken to ensure that the ductility of the existing element is enhanced.

To fully evaluate the elements look at the plastic capacity of the element and frame, this is, afterall, the maximum load that may be resisted. This may require some assumptions made regarding the connections in order to bracket the response but it should suffice.

The easy thing to do here is to improve the structure as noted above and move forward.

 

[JAE]

mannyg:
First, the correct response here is to point you toward the FEMA documents, specifically FEMA 273 and 274. These are aimed at existing building evaluations regarding seismic viability. The UBC is really intended to deal with new construction although it is used for evaluating existing structures all the time in practice. It gives various levels of performance that the owner can be consulted on as to the building's response to an earthquake. The documents are generally free in the U.S. (in Guam I'm not sure).

Using the UBC for an existing structure will many times put you into a box where you cannot "meet the code" short of a total retro-fit. If you must use only the UBC, I would check your cantilevered columns based on the code, based on the fact that they aren't Special Moment Frames (since you imply they are not), use the appropriate R for the cantilevered columns, and see where that puts you.

 

[ishvaaag]

I have an article in an ACI book stating how chilean buildings relying in much shearwall (more than in equivalent situation in the US) have also performed very well. The mere fact of that the nuclear industry attempts to keep elastic the buildings shows, that giving that enough strength is provided, a stiff building can perform even better than a flexible one, as long of course the earthquake does not overcome its strength.

In any case soem identification of the recommended forces (or how to proceed) by the code should be possible. Then the related analyses would show if what you have is enough or not to meet the intended safety standard.

Put more simply, if the lateral displacement got for your structure under earthquake forces is so small as to be met by the available ductility and strength of your column, you have finished your check from a science of construction viewpoint.

To make it equivalent to code compliance, you have to show that your model and calculations fit with some of the corresponding indications the code makes.

 

[Taro]

Qshake,
By reducing the "R" factor you are NOT "only increasing the elastic force (base shear), which for an existing detail is suspect to begin with as it never had the detailing to achieve the ductility associated with the increase in elastic force". There are two main strategies for resisting seismic motions: (a) strength and (b) ductility. By using a large R value, you are relying on the structure to have considerable post-yield toughness and energy dissipation capability. By using a low R value, you are taking a strength-based approach and limiting the ductility demand required of the system. As ishvaaag notes, the design approaches are different, but equally valid. The most efficient design results from a compromise between the two extremes. This is the basis behind the current performance-based design philosophy. By choosing how much post-yield behavior is acceptable, you can achieve your desired performance level. Of course, as you say, some structural systems are detailed to be more ductile and you can take advantage of that in your design. This does not invalidate other systems, though.

JAE,
Good point about using the FEMA standards. A couple of points to note, however:
(1)The current guidelines for the seismic rehabilitation of buildings are FEMA 356 and 357.
(2) In some cases "the correct response" might not be to use solely the FEMA standards. Sometimes the goal of an existing structure evaluation is to assess whether it complies with current code requirements.

 

[Qshake]

Taro,

I understand the philosophy of the code very well, but note that with existing structures you can increase the base shear all you want and ultimately you will be designing for the yield case (ductility and significant post yield response) because that is simply all that you have to work with. Because with increased load you will reach a point of yield whereafter you must consider the permanent displacements. And yes, this will dissipate lots of energy. This is, of course, why in areas of high seismic hazard it is more economical/plausible to design/retrofit to the plastic capacity rather than the large elastic forces. Unless, of course, you modify the load path and create new elements for resisting those elastic forces. And that is what I wished to point out.

And I agree also, that the proper direction here should be to aquire FEMA 273 and 274.

 

[JAE]

Taro
We've got 273 etc. ...didn't know about 356 and 357. Are these relatively new? How do they relate to 273?

shows how much we use them

 

[Qshake]

In effect, 356 and 357 will replace 273 and 274.

 

[JAE]

Thanks, Q...we just ordered them.

 

[mgg]

Thank you all for your enlightenment. I actually have the FEMA documents. Although, I have not had the time to read them.

Possibly, the original designer applied safety factors I am not aware of. In this case, my perspective is to verify if the safety factors, if any, will be enough to deem the structure sound. Therefore, evaluating using the UBC can be warranted as long as the strength and toughness of the structure and its components is adequate to handle the elastic base shear forces without attaining permanent displacements. It is rudimetary to determine the strength-based capacity of a structure, however, the problem is to determine the demand or proper base shear forces. This is where the R coefficient for ductility and overstrength comes into play.

The UBC defines a R-value of 2.2 for cantilever column building systems. Does anyone know what assumptions were made in regards to the type of confinement provided (special seismic or not)? Some of you have mentioned that because the value is less than that of an OMRF, it is safe to assume the transverse reinforcement is not specially detailed for a R of 2.2. I don't feel confident with that reasoning because other factors play into the determination of R.

 

[Qshake]

Regarding the confinement issue... If the existing column has transverse reinforcing on 12" centers as was commonly done in the 70s and earlier then you will have a confinement problem. In addition, if there are lap splices in the end region there is the additional possibility for "unzipping" of those splices and undermining the load transfer.

There are many simple ways to mitigate the confinement issue. Prestressed bands, prestressed wires, concrete encasement, steel jackets etc. Probably the most cost effective are the prestressed hoops covered with shotcrete.

 

[ishvaaag]

Albeit it being sure the best is provide confinement reinforcement, and as well being in accord in that other things will be influencing the assignation of a R for a class of structures, I think your guess of the R for "cantilever column buildings" being less than the R of OMRF meaning that just the requirements for OMRF columns might be of application to any columns in your building is not so daring, as long your building correspond to such class.

If the whole building is to be sustained against lateral loads by such main element that it can be classed a column building, most likely important shear or core walls will be part of the structure, and such walls not always have proper confinement but at the boundary elements of some shear walls. Normally, a lesser R response would mean the expected overall lateral stiffness anywhere must be more than in the OMRF structures, so starting from properly determined forces (if the class has been properly identified) it is highly unlikely any columns in the "column building" have to suffer more relative displacement than any OMRF, nor then take more force from lateral distortion under earthquake shakeout.

Except making resource to the R makers (or technical documentation of same, if available) I don't see how I, you, or a reviewing party may find your assumption of not being needed more confinement than for OMRF be unsafe at least for some typifiable cases. In any case, your general line of thought in this regard I see reasonably sustainable in front of the reviewng party if there's no the lesser doubt of your building being a column building.

But this of course won't be as good for the building life safey standard as adding confinement, as Qshake points.

 

[Taro]

Qshake,
I think we're miscommunicating about the R factor and its relation to ductility demand. Let me try again...

Seismic design provisions are based on the concept of an idealized elastic single-degree-of-freedom oscillator responding to random excitation at its base. The assumption is that the displacement of the actual nonlinear oscillator is the same as the idealized counterpart. By increasing the R factor, the nonlinear part of the total displacement is increased, but not the total displacement.

On the other hand, for systems with limited ductility capacity, a lower R factor is required. Yes, you are designing for a higher base shear. But the real effect of this is that the system stays elastic for a greater portion of the imposed displacement and less ductility is required. Therefore, the stringent detailing provisions that are required to develop the ductility required by the use of a high R factor are not applicable when using a lower R factor.

As I stated before, if I was designing a new cantilevered column structure, I would definitely incorporate ductile detailing as a matter of good practice. However, if you're analyzing an existing structure, you sometimes have to think a little more "outside the box".

 

[hakimi]

when working with non ductile structural systems with low range of R factors, you dont need to use special detaillings for ductile structures. So you have to be aware of applicability of detaillings and requirements you work with.