For a Vessel supported on support lug PV Elite gives below warning-

This is a vertical vessel on support lugs and the earthquake type selected is a building code. Building codes assume all of the mass is located above the point of support. This makes their use on lug supported vessels questionable. It is strongly suggested that based on the seismic zone and applicable parameters, an appropriate lateral (Gx) and (+-) vertical(Gy) acceleration be applied using the G loading option.

My question is, how to compute Gx & Gy accelaration based on seismic zone and applicable parameter.

Hello Rajendra,

Gx, Gy and Gz are acceleration in the directions x, y and z. You don't actually compute them. Instead, you give values to PV Elite. Now, g is the acceleration due to gravity being 9.81 m/s^2 or 32.18 ft/s^2. Let use say you set Gx to 0.4, that would mean the vessel is subjected to an acceleration of 0.4 x 9.81 = 3.924 m/s^2 in the x direction. You as the user must tell PV Elite what value to use. Typically a value of 0.4 is quite conservative for Gx and Gz.

rajendra_g,

It simply says that the responsibility of establishing G factors is yours and following blind the building Code would be not a valid argument.

In my opinion you can follow the building Code but you must be prudent with R value (or any other form as a reduction factor). Some Clients would ask you to consider R=1 as the calculation remains in the elastic domain, others would accept a value of R valid for inverted pendulum.

Thank you Mariog. You mean to say that, the client should provide value for Gx, Gy and Gz.

This is not I'm saying. You are responsible to calculate G's.

What that warning says is that the building code may be not proper fitting with the vessel behavior.
However, you can evaluate the fundamental period of your vessel and apply a judgement as building code make sense for your case.

The main question would be- IMO- the R value. Adopting R=1 is over-conservative but a Client would ask for it. You can adopt other value, but here you have no support from the building code as it is not referring to your case. Your Client would ask you (or not) to document your choice.

R value is the ductility factor and a building code would associate this factor to the structure's capacity to develop some non-collapsing plastic hinges, which is not your case. In the same time, the foundation-soil interaction during seismic event is an argument for R>1.
I mean R value is an sensitive subject and depends on many factors, case by case. It may be a subject to deal with Client or you can take the responsibility to adopt a value.

Hello Rajendra,

You need not worry about the R value or any other values that come from a building code. Just enter reasonable values for Gx and Gz and you will get a simple result.

Dear Mr. Delaforce,

To comment the original question, it was one of Open-Ended category.
"My question is how to compute Gx & Gy acceleration based on seismic zone and applicable parameter."

Even it wasn't my question, I do not understand the answer so I'd like to comment it.

First about the sentence "Just enter reasonable values for Gx and Gz and you will get a simple result." I can say that acting as a Client representative I would reject a calculation based on such approach. Using software and "just" feeding software with numbers as input is the modern "plague" in engineering. It is the Software user responsibility for the input values. A part of the responsibility is proved by the user explaining the input values.

Second, what Rajendra said is the fact Software warned that applying a building code [which assumes that all of the mass is located above the point of support] makes its use on lug supported vessels questionable. I think the warning is correct in intentions, however I would be not so convinced that the root of the problem is the mass distribution vs. supporting point. May be the rigidity of the legs vs. vessel, as well. In my opinion what I need to know better is the first mode of vibration and the relationship with the seismic spectrum- mass distribution, rigidity of components, etc- all have consequences here.

Third, you mentioned 0.4G as quite conservative and probably thought as reasonable conservative. It is reasonable as ground peak acceleration and the vessel will "see" such acceleration in case has a quite rigid behavior. It is not reasonable in case has enough flexible behavior and the vessel would be subject to the maximum value of the design spectrum acceleration, which can be 2.5 to 3 times the peak ground acceleration.

And last- your advice was to not worry about the R value when calculation a vessel. An R=1 value means the calculation is developed in the elastic domain. A value of R=2 for example is currently associated with inelasticity based on structure ductility.
Here I have an open question: what exactly we accept as inelastic behavior when adopting such value for a vessel calculation?

Thank you and best regards.

Hello Mariog,

Respectfully, as a point of interest, how would you assign values for Gx and Gz? Where would we find a reliable source. In other words, what values would you advise Rajendra to use.

As we are always interested in the opinions of the forum members, who in many cases are far more knowledgeable than I, how would you suggest the masses of the components be distributed. Could we for example be more precise in the manner that PV Elite employs to distribute the masses?

What I wrote is not a criticism on how PV Elite distributes masses. To be fair, I do not understand why PV Elite/you make the central point of discussion the mass distribution.

Where the mass distribution interferes with a building code philosophy, to be more precise?

I think the mass distribution affects directly the modal parameters and indeed the fundamental period of vibration is a central figure in building code.
Do you think the fundamental period is not calculated properly because vessel with legs mass distribution is not as for a building? This is the sense of the PV Elite warning?

Hello Mariog,

Normally in the case of a building code, the mass is distributed with a weighted average. When 'G' loading is the chosen method of analysis, the tower is considered a rigid structure, thus, the weighted average no longer applies. The purpose of my original response to this question was to suggest a simple solution without any extraneous information.

I hope this is not true. Applying uniform G's over a tower (considering the tower as a rigid structure) underestimates the seismic overturning moment and definitely is not what a building Code is asking for.
A building code asks to consider first an equivalent lateral force and it is true that Cs of ASCE 7 can be considered as an equivalent acceleration applied to building. However, after that, the base shear force is distributed across the height of the structure considering a vibration shape that is representative of the first modal response of the structure- usually the distribution shape is an inverted triangle.

Dear Mr. Delaforce,

I will try to explain what is the story behind the R coefficient.

A building is not a rigid under a seismic event because- first- it is uneconomical to build it up as a rigid. Unfortunately a building is flexible enough to experiment G's accelerations much bigger than the maximum recorded with an instrument on ground i.e. the peak ground acceleration. For a building code, the building must be modeled as a cantilever with mass concentrated at stories. A vessel may be modeled similarly, with mass concentrated in GC of elements. A fine "mesh" would help.
And no doubts, a cantilever deflects under the seismic event.

In case the first period of vibration is "short" enough (and a first modal frequency of a couple of Hertz means it is), ASCE 7 asks to calculate a seismic coefficient as CS=SDs/(R/Ie). CS is applied to W- seismic weight, so CS can be seen as an equivalent lateral acceleration (G's fraction) for a global calculation of (Base) Seismic force. However CS is not seen as an uniform acceleration over the building height.
SDS is roughly 2.5 times the peak ground acceleration specific to a site, including specific soil condition influence and corresponding to a probability to have that seismic event.

For Ie=1 as importance and R=1- as elastic response, CS=SDs. For example if PGA=0.4G, it follows that CS=1. In other words, this means the lateral acceleration is the gravity acceleration. It is dramatic in design and complete unrealistic to be considered and for a tall building never considered.

That's why we accept some supplementary help by other practical aspects.
For a building structure, R>1 is an instrument for Cs reduction, possible for a number of reasons. As the structure begins to yield and deform in-elastically, the effective period of response of the structure lengthens which, for most structures, results in a reduction in strength demand.
Furthermore, the inelastic action results in a significant amount of energy dissipation in addition to other sources of damping. This is a combined effect we call ductility reduction.

The only remark I could have is that R=2.5 as ductility factor will bring back Cs to the level of the peak ground acceleration but we need to be sure R=2.5 can be obtained. The R factors for buildings are obtained by observations and experiments and published for buildings, there is no formula to calculate them.

For our case, I think the above written explanations would rise the question of the degree of plasticity we accept when calculating the seismic effect of a vertical vessel and what exactly R=2.5 means for such case.

Hello Mariog,

I cannot disagree with you. However, I addressed the question posed, and covered the salient points that would help the enquirer solve the problem quickly and efficiently. That really is the purpose of the forum. We are not analysing a building, so the issues raised are, though of interest, do not give a succinct answer.

There is no succinct answer, that's the only issue in this "debate".

We can choose to consider the peak ground acceleration as G's applied uniformly over the height vertical vessel, however this does not correspond with vessel behavior during a seismic event.

We can enter deeply in details and conclude that a building code offers us more information about the actual behavior of vessel (and consequently give us realistic estimation about the horizontal seismic force and overturning seismic moment) but we need to face the issue of an appropriate R which is undocumented yet. That was my original warning in the thread.

Hello mariog,

You are getting into issues that are beyond the scope of the original question.

OK, one may ignore totally what I said in case is not useful. Having no option to delete my posts it's all I can say.