Peak Ground Acceleration 2

[alphaxy]

Fellows,

I am having a confusion with the PGA issue.

Our Geotechnical contractor submitted a report to us for a particular project where the engineer in charge stated a value of 0.52g

The peer reviewer keeps on asking in our meeting about the value we have used for Ca = 0.44Na = 0.44(1.2) = 0.528 where seismic source type A was considered and soil profile type D is selected as per UBC 1997. The peer reviewer of the project is asking me if the 0.52g given to us by the geotechnical contractor is in allowable(service) value or in strength(ultimate) value already (in which, for me it is an allowable value) but the geotechnical engineer in charge cannot determine if it is allowable or ultimate (strength) already but he did just say that it is the final value (as is)

With this, if the allowable value is 0.52g, and what we have used is strength value of 0.528, if we divide this by 1.4 to make it an allowable value, this means that our design is deficient by 4% and that is questionable (as per reviewer's point of view).

Do you have any idea where this "1.4" factor from UBC 1997 comes from? How did it arrived to 1.4? what is the basis of that?

I keep on thinking if the coefficients has something to do with the ultimate or allowable..The code only stated that "seismic force levels and R-factors are at strength levels", considering UBC 1997 but in UBC 1994 "seismic force levels and R-factors are at allowable stress levels".

UBC 1994 R-value 12, UBC 1997 R-value 8.5, divide 12/8.5 = 1.41 close to 1.4...I still cannot understand the relationship..

I hope someone could give me a clear solution in this. Many experienced engineers I asked didn't know about this.

Ideas are deeply appreciated.

 

[ishvaaag]

If your code was as in Spain is, it would be ultimate. However, the coefficient of E for combinations where earthquake forces are present, is 1. Furthermore, all the other components of the hypothesis at hand where an earthquake component enters are NEITHER factored; that is you check for 1 times gravity, 1 times a part of the standing live load and snow, 1 times the part of wind assumed standing (usually 0, so 0) and 1 times eartquake at the expected effects of the given peak ground acceleration... not a single safety factor when you are checking for a hypothesis with earthquake. This is what is called a "probabilistic" hypothesis, where the loads are ALL taken at the values that, following the code, can likely exist when earthquake strikes. Then these forces derived from loadings with safety factor = 1 are checked in strength against LIMIT or ULTIMATE values (and apart, for the compatibility displacement issues as mandated by the code).

And this is the part where if you would be using checks in allowable or ultimate enters; if you were checking with ultimate, you would be comparing some actual force against some limit strength; everything would be right. But if you are checking say steel in an ASD context, using the allowable limits for comparison would be unjust comparatively, because you would be checking against a value lesser in the overall safety factor than in the prior case, that is the case for which the concept is developed: a case where the forces and strength are expected what they are expected to be, the expected force and the expected strength.

So a way to bring back the things to order is when working in an ASD scheme to use for EVERYTHING lesser solicitations in the safety factor value. You can do this in different ways but the thing is to state the solicitations that compare justly to the ultimate "probabilistic" strength checks, that may be being done by division of the forces (earthquake AND all others) by the safety factors, something it seems is being done in your case by division of 1.4 of the earthquake forces (and the others?). Hence...

1. It is likely the peak ground acceleration is given as to be taken in an ultimate "probabilistic" strength viewpoint. Your geotech is right in the point that it is a single value to be given ... he simply states at much here the peak ground acceleration needs to be considered at THIS value.

2. If something, it is likely that if you are to compare to ASD allowables, you need to divide the peak ground acceleration as received by some safety factor, and, as practiced here, all other loadings (dead, live, snow) AS WELL, for otherwise you would be keeping an unwarranted safety factor on them that is not warranted by this criteria, that is to meet with actual strength actual probabilistic solicitations.

 

[haynewp]

I would expect values derived from PGA's and other data from the USGS to be at what you are considering as an "ultimate" value. A 1.4 divider normally takes the earthquake value from the USGS and brings it to the code expected "allowable" level for someone choosing to use allowable stress design.

Even if it were a site specific value determined by the geotech by shear wave testing, I would take it as a strength requirement (ultimate) value unless the geotech says otherwise. The ground response to an earthquake is what it is.

 

[msucog]

question: the stated PGA=0.52g? if you don't mind, where is the general location and do you know to what return period the value is referenced? (these questions are more for my own curiosity)

i would consider pga as "ultimate" even though i'm not so sure that is the most appropriate way of looking at it. also bear in mind that pga is not necessarily the highest/peak of anything and pga can differ in the spectral period depending on the location.

i'm more experienced with ibc but i interpret the 1.4 is similar to the 2/3 factor (or divided by 1.5) applied to Sms (ibc). the ibc factor is applied to avoid having a doubly factor of safety applied since the structural system already has a factor of safety inherently built in to its design.

maybe this will help since i did a quick search to find more on how ubc considers the topic. (i searched google for "ubc factor 1.4")-see top left of 2nd page

http://www.structuremag.org/OldArchives/2004/september/Codes and Standards.pdf

i hope i haven't added another layer of confusion to your question. good luck.

 

[shin25]

The PGA is the maximum ground acceleration which is expected during a design maximum earthquake event. Depending on the design methodology used, PGA can provide service level or strength level force values. Bear in mind that if the design structure is expected to remain elastic (which is ASD) during the design earth quake then the force for which the structure will be designed for will be really high and therefore this will be a costly structure. The current codes however require that a properly designed structure does not have to be elastic during a design earthquake, rather the structure can yield and sustain some damage to dissipate seismic forces. This method actually reduces design forces by the use of Response Modification Factor or R. R value is dependent on the type of structure used. A very large R value means the structure has a very large deformation capability or high energy dissipation capability. This is why the design seismic force is divided by R for strength level design. Therefore, using an R value of 1 you see will yield a service level design force which is not the intent. Therefore PGA has very little to do with ASD or strength level design per current codes.

 

[mmillerpe]

The 1.4 factor comes from 1 divided by 0.7. In Chapter 2 of ASCE 7 the factor applied to the earthquake load (E) in strength design load combinations is 1.0E whereas in Allowable Stress Design the factor on the earthquake load is 0.7E. When running some quick numbers you will often see people take the strength design, and divide by the 1.4 factor to get the ASD value when trying to get a feel for the structure.

I am not familiar with the UBC load combinations or how they relate to the ASCE, nor am I familiar with how the UBC derives the PGA, however if the IBC method is used, generally the geotech would provide the value for use in strength design.