Dear sirs,

For sites located outside of USA, if i know the peak ground acceleration (Sp):

a) I must give input to this value in "Peak ground acceleration for Non-ASCE (Sp)" or I must give input to this value in "Mapped max. earthquake short period (Ss)" & "Mapped max. earthquake 1sec period (S1)", according to E.4.3(2)?

b) In both cases: is neccesary input the Q value? Normally Q=1 or Q=0.666 or Q=0?

Thanks

You input Sp, the software willl determine Ss and S1.

You should define Q.

Thanks for your answer,

but if i don't enter any value of Q. TANK take any default?

I'm in the same situation, and not familiar with API 650 details.

1) is Sp expressed as acceleration in m^2/s, or is it a percentage of gravity?

2) should I select the Site class according to Asce (earthwuale type= mapped) even if I'm using a site acceleration? Or should I select "site specific" (but in this case, even if I insert a huge Sp, my analysis is passing anyway..).
I think that "mapped" has to be selected.

Sorry for the "basic level" questions, I'm not really expert on this code and in using Tank too..

My problem is: I'm designing for Siberia (seismic regulation is SNIP II-7-81), so I have to find some way to "force" TANK to analyze with an equivalent seismic acceleration.
I know this acceleration, because I have already worked on the SNIP code: for every vessel I can determine the global shear, I have it's mass, so I find the acceleration.
I think this acceleration should be used as Sp in API650

I've been studying this issue for a while now, and I am in doubt because it seems to me that TANK is not considering the parameter SP.
Infact, in a "mapped" analisys, the site Class will determine the shear forces, regardless Sp (and this is correct, because Sp is used for non-ASCE territories, where no mapping is available).
For "site-specific" analysis, Sa* and Sa0* are used, regardless Sp.

Now, API 650 points to 3 methods in par. E.4
1) for mapped sites
2) site-specific procedures for non-mapped zones, in order to obtain Sa* and Sa0*
3) non-ASCE sites method (Par.E.4.3), in which substitutions are used to apply method 1) starting from some rough non-ASCE parameters like Sp.

It seems to me that TANK right now is not implementing the third option (the one I wanted to use)
Infact as far as I select the Site Class (and I have to select it), this will override Sp input and the consequent Ss and S1 input in accordance with E.4.3

I'm doublechecking my assumptions both with teh code and with a little program that is supposed to work with API650 seismic calculation too.
It's input data (see attachment) clearly point to 4 different methods: mapped, site-specific, Non-ASCE and Non-ASCE again.

I fear that something is missing on Tank or, probably, in the way I'm using it.
Now I'm trying another way to let TANK do what I want (read Sp and use it in the calculation): I just found that, if i select a Mapped analysis, and the I specify an "F" Site Classe, the parameter results will change when I change Sp.
But I wonder: is this the way the program is supposed to work?

I hope this will be considered carefully, because every designer that is working outside ASCE mapped zone will face these difficulties.

I'll look at this and get back to you.

Dear Richard,

I hope you had time to look into the problem.
Meanwhile, I have talked with a couple of engineers that sometimes work with us: they both said that, when dealing with projects located outside ASCE mapped zones, they had to give up with Tank and find some other way to take into account the non-ASCE seismicity.

If you specify Sp, then equations E.4.3-1 and E.4.3-2 are used to determine Ss and S1.

I believe there are a lot of holes in Appendix E. Any requirements, procedures, or conclusions you can provide will help me in further determining what my be missing or not addressed. Hopefully I can make allowances for these in a future version of TANK.

Richard, I know that TANK uses E.4.3-1 and E.4.3-1 to determine Ss and S1.
The point is that:
1) TANK will not use the computed values of Ss and S1 to continue the calculation when specifying a "mapped" earthquake type, with one exception.
2) TANK will not use the computed values of Ss and S1 to continue the calculation when specifying a "site-specific" earthquake type.

So, Sp is not affecting the results and this is why I'm believing that TANK seismic analysis procedures don't cover entirely what API prescribes.

Here is the only procedure I found to obtain any output variation related to changes on the Sp input value:
1) setting a "mapped" earthquake (this means that TANK will not use Sa0* and Sa*, but will try to use Ss and S1).
2) selecting an "F" terrain (TANK will not find any table reporting the values for the above said Ss and S1, so I'm assuming it will "read" Ss and S1 from the E.4.3-1 computation).
3) input the desidered Sp.

With these inputs, a change in Sp will affect the results in SOME way: I'm wondering if this is the way the program is supposed to work.

Please let me know if my explanation is not precise, sometimes I get stuck with English and the result is that what I'm trying to explain sounds confusing to me too

Richard, I'm quite confused right now.
I found that today I can input Mapped earthquakes and Sp acceleration... and it seems that both these parameters are used in the calculation.

I need a holiday....

I'm still working on this.

Here is what TANK is doing:

You select either "Site Specific" or "Mapped".

For "Site Specific", you're also supposed to define Sa* and Sa0*. These values are then used to determine Ai and Ac per E.4.6.2.

For "Mapped", the software checks the input values of Ss, S0, and S1. If any of these are zero (i.e. not defined), then Section E.4.3 is used to determine Ss and S1. E.4.6.1 states that Sp can be used directly for S0. Once S1 and Ss are known, E.4.4 is used to obtain Fa and Fv. Now everything necessary to compute Ai and Ac per E.4.6.1 is available.

Richard, thanks for your explanation.
Yesterday I performed some further tests starting from scratch on this issue, and I came to the same conclusion.

The problem I have previously faced were caused by my wrong assumption that, even when specifying Sp, I also had to manually compile the relative Ss, S0 and S1.
So I inserted these values once and, with those input already set, a change in Sp did not have any influence.

One last doubt: Sp is in %g, right? This is not a "true" acceleration, I guess it is:
ground peak acceleration /(9.81*m/s^2)

Yes, that is my understanding.

I try to give some remarks in the problem of seismic parameters and which Q must be considered. Anyway this is what I was able to understand from the cryptic API 650 and what I’m writing is my opinion.

There are two methods in API 650.

One is related to Non_ASCE methods. This method is addressed to the places of the world not covered by ASCE maps. So it must start with a value which is not given by maps.
In explanatory appendix EC API says "Assuming the only parameter given is the 475 year peak ground acceleration (damping = 5%). This is comparable to the ‘Z’ used in the earlier editions of the UBC."
In fact, under this method Sp=Z unless more accurate value is available.

To continue the calculation under non- ASCE method, we may consider API 650. E.4.3 SITES NOT DEFINED BY ASCE 7 METHODS paragraph 2: “ If no response spectra shape is prescribed and only the peak ground acceleration, SP, is defined, then the following substitutions shall apply:
Ss = 2.5 SP (E.4.3-1)
S1 = 1.25 SP (E.4.3-2)"

To conclude, non ASCE method is based on an Sp comparable to or identical with Z of UBC. Many seismic codes following the "old” UBC have a “Z” acceleration. UBC intended that structures be designed for “life-safety” in the event of an earthquake with a 10-percent probability of being exceeded in 50 years (commonly referred to as the 475-year earthquake). The problem with this design earthquake, as it was recognized during the development of the "new" ASCE spectral response acceleration maps, was that it did not provide adequate protection for the infrequent but very large seismic events.

That is why another method (more accurate) is related to ASCE code where Ss is the mapped, maximum considered earthquake, 5% damped, spectral response acceleration parameter at short periods (0.2 seconds). ASCE intends design for “collapse prevention” in a much larger earthquake, with a 2-percent probability of being exceeded in 50 years (commonly referred to as the 2,500-year earthquake).

That explain why Ss considered by ASCE method is not identical with Ss considered by Non- ASCE methods!
As a detail, a comparison between Ss of ASCE and Ss of UBC (in fact a non-necessary one, since Ss is mapped under ASCE!) would be to consider for ASCE a Ss as given by
Ss = (1.5) * Z * Na * (2.5) = 3.75 * Z * Na.

Na = acceleration-dependent near-source factor of the 1997 UBC and is applied only for Zone 4- high seismic areas, so for zones less than 4, Ss = 3.75 * Z .
However, this formula is just a detail in discussion, highlighting that Ss under ASCE is- for moderate seismic zones, 50% larger than Ss in Non-ASCE method. Or, as a conclusion, we may say there is a factor of 3/2 for ASCE method versus a factor of 1 for non-ASCE method!

Again, the methods difference is due to the different reference considered: it is clear that Non- ASCE and UBC refers to a “475-year earthquake” whereas ASCE refers to a “2,500-year earthquake”. However, it MUST be a final agreement between the FINAL results for moderate earthquakes. The intention was to change something for large earthquakes, not to have different results between methods working with moderate earthquakes.

The correction has been solved by API 650 applying a Q factor in "design parameters".
Q was set by API 650 to 1 for Non- ASCE method and to 2/3 for ASCE method.

See also API 650 "For sites where only the peak ground acceleration is defined, substitute SP for S0 in Equations E.4.6.1-1 through E.4.6.2-1. The scaling factor, Q, is defined as 2/3 for the ASCE 7 methods. Q may be taken equal to 1.0 unless otherwise defined in the regulatory requirements where ASCE 7 does not apply".

This Q factor applies to Ss when calculate Sds- so in the end of calculation Sds has identical or comparable values for both methods.

Hello Dear Richard, I´m from Venezuela and I have worked a lot in designing tanks.
The most difficult part when I´m dealing with clients is that sometimes they want to use the ASCE code and the API-650 appex E to do the seismic analysis.
In Venezuela we have our standard to evaluate the seismic load in the tanks, and from my point of view ,that standard, is one of the most clear and complete standard about seismic.
I have read several seismics standard but I always end up in the same point, the Famous table E-1 and E-2.
As the API and ASCE standard define for the non ASCE zones, engineers can use the Sp value, and then, making some sustitution the will have the So and S1. BUT when you use both table you have to "match" those values (s0 and S1) with some soil condition to get Fa and Fv.
HERE is the big problem , It is very difficult to make that match and the worst part , FROM MY POINT OF VIEW, is that Engineers Try to "model" the soil response or the Spectral response just with the valu of Sp.
I think that using this method or this approach would raise some mistakes or will give a non appropiate model of response...
I would like know you opinion about it.
BEst Regards.
JM