Earthquake Resistance question UBC 1

[Enhineyero]

UBC have 3 seismic source types A,B,C. it says that A is for faults that produce Magnitude 7 and above. The forces calculated from UBC is based on a Earthquake with a return period of 50 years. So if I design the structure using a seismic source type A, what is the magnitude of earthquake that the structure can resist? Using static analysis (is there a difference in the resistance of the structure using dynamic analysis?)

In my understanding it can withstand a maximum of Magnitude 7.0 earthquake, anything above that, the safety of the structure is unpredictable? Anyone care to make an input?

 

[ishvaaag]

You may be referring to Table 16-U seismic source types, used to get when applicable the near source factors from tables 16-S and 16-T. As you see, seismic source type A is identified as one able to produce magnitude equal or greater than 7.0; as also is B for some specific cases. Hence, such source types are just code blackboxes where tu throw any seismic source able to produce at least 7.0 magnitude, or more.

So my understanding is not in that buildings designed as per the code can stand only 7.0 magnitude earthquakes. It is pertinent to say that what the code does is to place particular specifications to uniformize and control to some specific levels the effects and damages of the (statitistically determined) expected earthquakes at any given place of the US territory. So all the quoted factors in the code have a say in the seismic solicitation to which structure and other building parts must be analyzed for, and as you see, magnitudes over 7.0 are not excluded, they simply would count as seismic source type A or B.

It also would be rare that the code would do such thing since subduction zones use to show such magnitude earthquakes and the alaskan territory has such kind of seismic source. For subduction type sources it directs to particular determination of the seismic source type, anyway.

Respect the static or dynamic analysis to ensure safety, when so allowed in the codes, you may use both. When the procedures are properly followed for such cases you should get a satisfactory design as per the code by any of the static or dynamic analysis methods.

It is also neccesary to say that no code ensures inmmunity against every earthquake happenstance, just place preventive measures estimated adequate according to the -technically supported- legislative body. This is not a particular feature of the code specifications for earthquakes, for all structural design, and components' specs, are equally established on the basis of statistical prevention of failure, i.e., all along some failures are understood are going to happen even by inception of the code when supposing extremes of rare loadings, undetected low quality materials and inadverted bad or negligent workmanship; it is only the fact of earthquakes being of so massive effects that make the particular limitations on the design and constructions against earthquake more conspicuous than in the other more limited and local structural failures.

 

[MiketheEngineer]

I thought UBC went away with the new IBC??

Just asking??

 

[Enhineyero]

Thanks Ishvaag. The reason i ask this is that due to all the Earthquake happening recently a client of mine asked me what is the magnitude that my structure can resist? I don't know if the guy without engineering knowlegde can understand all our engineering jargon.

So basically how can we determine the threshold of earthquake magnitude that our structures can resist using UBC 97? when using a seismic source A, which produces Magnitude 7 and above, can we say that it can withstand (as per design) say a Magnitude 9 Earthquake? or a Magnitude 7 or 8 for that matter

 

[ishvaaag]

You can say him that as per the code, allowance to resist earthquakes of such magnitude is in the codes, however, this not meaning an entire chance of survival of the building and occupants, and no damage. Then you can proceed to explain that design as per the codes impose reasonable prevention against earthquake damage of any level, but it is obvious that not for any case, such moderate importance buildings placed in the chasms caused by one high magnitude earthquake.

You can also say that in reality, most of what the structural codes prevent from earthquakes is damage caused by the vibration induced by them in the soil, what presupposes some (even if small) distance of the epicenter itself, and even then, it is only statistical prevention, i.e., ensuring that a big percentage of the human life is preserved, and buildings and infrastructure according to their importance, and their attachments (if properly installed as per the code) only suffer proportionate damage (scarce under major eartquakes and very little damage under moderate earthquakes).

So the code has an estimate of what peak ground acceleration has the code-estimated likelihood of being experienced in the design lifetime of the building. You, essentially, design the building to resist some set of ground vibrations thought adequate at the site. These sets of ground vibrations have some correlation with the magnitude of the earthquake and the distance to epicenter; in fact that is what the seismological institutes do backwards, to ascertain the magnitude from the records, a situation akin to the answer your client is asking from you.

To forfeit reverse engineering entirely the code, you can say your client, according to the code and in its terms, if respected, your building that is at x distance from some known fault able to produce earthquakes of magnitude such is able to resist them. To say that, either the seismic source is some fault of known position, or if subduction plate origin you exhibit some info showing (how typically) they can generate magnitude 7 and more magnitude earthquakes,

Even then, you are mostly only dealing with the effects of vibration,

Earthquake effect Strategy
Fault rupture Avoid
Tsunami/seiche Avoid
Landslide Avoid
Liquefaction Avoid/resist
Ground shaking Resist

As you see the only way for most other issues related with earthquake the recommendation is to keep away from the seismic problem. If you can't avoid that you stay subject to the vagaries of nature. Think, for example, that it is known tsunamis to have encroached to over 400 m over sea level in narrowing valleys, and blankets of water also hundreds of meters are thought to have occurred in the past in continental wide cataclysmic events; you don't make building codes for cataclysmic events, but for what recent experience has shown be recurring in modern times. Whole prevention would require building space Noah's arks able to leave the problem behind at automatic short notice.

 

[delagina]

A Structural Engineering Firm claimed they designed their Structure to resist magnitude 8.0 earthquake. I heard it on the news. This was right after the Haiti Earthquake.

How they come up with that "claim", I am also interested to know.

 

[Enhineyero]

Delagina, that is a question i was asking. How can we answer the question of a lay person on how safe is our structures in an event of an earthquake of a certain magnitude.

ishavaag's answer is the idea i have in mind, because we can say that, "statistically" the structure can resist the seismic activity of the nearest fault and or strongest/active fault. within a return period of 50 yrs. And based on data and relevant experiments we "assume" a "design" seismic force to be resisted by the structure.

But in the end it is all an "assumption", and it is based on "Statistical Data". So there is a probablity that the current data may be exceeded and/or the Data in hand might be not enough to determine the "statistical force" a certain fault can produce.

Therefore no Structure is safe in an event of an earthquake. although we create an illusion that such structure can resist a certain earthquake force, all these are based on "theoretical assumption". What we have made is a code compliant structure, a code which is patterned to make a "statistically safe" design based on a certain set of data and many assumptions.

 

[ishvaaag]

It is worth to mention that there are many studies and programs trying to emulate the vibratory effects of earthquakes. EQPACK is one DOS program used to such purpose.

In one of such programs you can make simulations of what would occur (in vibratory terms) if one earthquake of x magnitude would happen at y distance from your site. So if you are ready to accept that the seismic effects are properly emulated by the program, and use the output of such programs for your seismic design, you can instil in your client the belief on your structure having been particularly targeted to sustain an earhquake of x magnitude somewhere. All the comments about the code above would stand, only that the input data now would be derived from these intelligent programs used to emulate earthquake ground motion at some point.

However modern codes insist in using better natural records (or spectra of the same) rather than proceed from those produced in artificial ways, as by such programs.

 

[hawkaz]

Ishvaaag,
That program sounds interesting from our perspective, but the information sounds dangerous to release to clients. It may give some un-realistic expectations.

As you know, there are too many variables to account for (different types of earthquakes, different soil conditions, depth of bedrock, depth of quake, distance from epicenter.....)

Unless this program is very sophisticated, I doubt it could have predicted the damage that Loma Prieta caused to San Francisco

 

[ishvaaag]

Well, the programs, their way (coming from earthquake research) are sophisticated even if of the eighties (there are other more modern in the same intent), however you see the codemakers, on the same viewpoint than you, see dangerous allowing them be used for general purpose, then so relegating them at the time to research purpose.

This must mean that as an average the predictions when checked against actual records differ so much that they can't yet be seen as a substitute of the actual records for design purposes, or, even if such is not the case, the codemakers feel not assured of gaining the required statistical reliability when using the predictions to generate the seismic input for the structural design.

This seems reasonable in that we would be using two layers of uncertainty, once when inferring the (prediction) motion records or spectra and then again with the statistical uncertainties that mess our design procedures. The codemakers prefer to get rid of the first one by just using actual records, of which there are plenty (if not always adequate).