Application of Seismic Loads to Saddle 2

[jproj]

I am trying to apply a horizontal seismic / wind load to my vessel supports (vertical reaction load). Unfortunately, the only reference I have available is the Pressure Vessel Design Manual by Dennis Moss. Aside from the errors and non-existent derivation of equations, the outlined equation doesn't apply to my vessel.

Basically, I have a horizontal vessel with five saddles. The center saddle is fixed while the outer four saddles are sliding. I am trying to calculate the vertical reaction force at each saddle baseplate due to a seismic event.

In the book (two saddles), the transverse vertical load (Qt) equation is given as Qt = (3*Ft*B)/A where...
Ft = transverse seismic force
B = distance from saddle base to vessel center of gravity
A = baseplate length (spanning vessel diameter)

Can anyone confirm (or dispute) that..

1) Qt in this equation is the "per saddle" vertical reaction load (acting up on one half of each saddle, down on the other half of each saddle)?

2) for my vessel (five saddles), Qt = (3*Ft*B)/(5*A)?

I'm also confused about the vertical reaction load for a longitudinal force, but this post is long enough already...

Any help is greatly appreciated!

jproj

 

[roca]

Must be a big vessel to be sitting on 5 saddles.
I've never seen that before

 

[JoeTank]

The vertical load distribute for five equally spaced saddles will not be uniform; therefore, I doubt the solution to your problem will easy. Do not try to adapt the Moss equations. They have no applicability to a five saddle support system. Ditto what roca said. That must be one whopper of a vessel!!!

Joe Tank

 

[LSThill]

Mr. jproj (Chemical)

Do you have Pressure Vessel Design Manual by Dennis Moss 3rd edition.

If you need the Technical Paper:
"A Simple Solution to Cylindrical Pressure Vessels on Multiple Support Saddle".

Contrct me

Stephen
Leonard Stephen Thill

 

[jproj]

The vessel is certainly large (13' OD x 141' T-T), but IMO 5 saddles are unnecessary. It was a customer requirement and likely specified in order to reduce the structural steal per beam load.

As far as the dead weight, I have already estimated the distribution (based on assuming the vessel acts as a continuous beam with 5 support points).

As far as adapting the equations for the transverse seismic force, I figure it should just be a matter of applying the horizontal force at the center of gravity and determining the resulting reaction force at the base plate (up on one side, down on the other).

For two saddles, it is my understanding that the Qt = (3*Ft*B)/A equation is a condensed version of....

Qt = (Ft*B) / [(2/3) * (A/2)] where Ft*B is the (per saddle) moment created between the horizontal force acting at the vessel center of gravity. A/2 is the length of the base plate resisting the moment force (triangular distributed load). 2/3 is the centroid of the triangle, or the reaction point force.

So for five saddles, I would think that the same equation could be applied with at least ballpark accuracy (Ft equally distributed between the 5 saddles). This load would then be added to the dead weights I previously calculated. Agree / disagree?

The longitudinal load, however I agree would be much more difficult to apply.

Stephen, to answer your question, no I don't have the third edition. After the minimal error resolution / poor editing between the first and second edition, I saw no reason to waste more money on the third edition.

jproj

 

[Tomtation]

Two saddles is recommended and not more because due to settling over time the loads at each saddle will not be the same as those calculated. Since the vessel as a beam is very strong two saddles is should be enough. You can adjust their location to reduce the mid-span moment and use stiffening rings to reduce the saddle-horn stress.

Treat the unit as a beam to determine the moments at the supports and the maximum mid-span moment and the saddle reactions. The original paper for calculating horizontal vessels on saddles is by Zick. Contact me if you need a copy. A good description for combining the loads on a horizontal vessel can be found in the back of TEMA and a very good pressure vessel book is "Pressure Vessel Design Handbook" by H. Bednar.

Regarding the equations found in all of the books, they have been simplified and are based on the moments for a simply-supported beam with two equidistant supports with overhangs which represent saddles at equal distance from each head of a simple horizontal vessel. For this situation the reactions at each saddle are the same and in the case of a vessel with more weight on one saddle, you must calculate the actual moments and reactions and use these results in the approriate formulae.

 

[MJCronin]

jproj..

The advice given above is very good abd the Bedar text is excellent.

Is the vessel made of metal or is it FRP or something ?

Your vessel is, of course, extremely large, although many horizontal vessels in the range of 100 ft are commonly specified and used for propane storage.

At 141 feet long, you are paying extra for field allignment inspection, pressure testing and assembly.

I submit, would two "shop assembled" and tested pressure vessels be cheaper ?

My opinion only...

-MJC

 

[BarryEng]

The USBR (United States Bureau of Reclamation) has a book called 'Power Penstocks' (big pipes for the power industry). Also published as ASCE 79 (Manual).

There is a section in the text for analysing earthquake loads on an above ground steel penstock (essentially a pressure vessel on many supports). There is also a section at the back, on a design example, with detailed calculations (about 20 pages long).

I obtained an original copy of the Zick paper (about 1951, I think) from this (Eng-tips) web site. I also use Bednar for info on original derivations of formulas.

I'm not sure if Troitsky - Tubular Steel Structures (the third volume of the excellent Lincoln welding series, the first two by Blogett) has any info.

I just looked it up - it does have some earthquake info but related to a guyed stack, with no info on a pressure vessel. I also could not find any info in the section on above ground pipelines.

The best reference I have, is the USBR (ASCE 79) for some details of earthquake design of an above ground pipeline (pressure vessel).

 

[HowJoh]

Stephen (lsthill)

If you could post a link to, or a citation for, the paper "A Simple Solution to Cylindrical Pressure Vessels on Multiple Support Saddle" I would greatly appreciate it.

Regards,
Howie

 

[MJCronin]

jproj,

With the considerable expense of this huge horizontal vessel and the real financial risk taken with it's unique "five saddle" configuation, I have to ask you this:

Wouldn't it be reasonable to engage a competent consulting firm to develop a FEA model of the vessel, supporting structure etc ?

My feeling about this particular case is that it is so large and unique that the cost of a more intensive anaysis is justified...

Many firms exist all over the USA that can do this work.

What is the shell material, operating pressure, liquid contained etc...?

-MJC

 

[LSThill]

MJCronin (Mechanical)

This is a typical design and was review Engineering Surveillance, Sung Jin, Ulsan Korea 8July06

FEA: Saddle Pro by Paulin Research Group Tapered Saddles and Pipe Shoes Integral & Non-Integral Wear Plates Included for Either Saddles or Pipe Shoes Include 1-10 Web Plates on Saddles

http://www.paulin.com

Leonard Stephen Thill

 

[jehan17718]

Jproj,

Have a look at this thread too....

http://www.eng-tips.com/viewthread.cfm?qid=194550&page=10

-jehan

 

[rhg]

Forget this design

Use Spherical Vessel ( zero problems )

regards

rhg