Overstressed using temp. saddles during hydrotest? 1

[Ebardt]

I have a large thin-walled vertical Offgas Drum. The shop just called to say they have the vessel laid horizontally on temp. saddles but they are afraid to start filling with water because they think the vessel will be overstressed.

The vessel is 92'-0" Sm/Sm, 3/8" thick CS, 144" ID. They have it sitting on six (6) temporary steel saddles spaced 180" apart. The saddles are 12" wide and are shimmed with plywood to get a full 120degree contact area. The vessel will weigh about 750,000 lbs full of water, about 125,000lbs per saddle.

Does anyone have a quick method for checking the stability of this vessel prior to hydro or a method for determining the minimum number and spacing of temporary saddles required for hydro?

Any suggestions are greatly appriciated.

 

[Moersleutel]

Ebart

For anything more than 3 saddles we use the three moment equation to determine the loads at the saddles and to evaluate the stresses at the saddles we use PD5500 Annex G, as it allows for saddles not welded to the vessel. It is not a quick procedure if you are starting from scratch.

I have seen FEA been done to determine the stresses, but I believe it to be an overkill, since nobody can guarantee the saddles will be perfectly level and that they will carry the loads predicted by your calcs.

Regards

 

[JoeTank]

The Zick method of analysis is typically used for the design of saddles for horizontal vessels. It is well known and is available in just about every pressure vessel handbook I have seen. As noted above, the load on each saddle much be determined. Even after determining load, it is advisable to consider an additional margin for non-uniform support/contact between the saddle and the shell.

Steve Braune
Tank Industry Consultants

http://www.tankindustry.com

 

[SnTMan]

I find Zick to be troublesome for anything but for what it is intended for, that is:

A symmetrical vessel supported on two saddles.

The loads, reactions, stresses, etc are all bound up in the geometry of a symmetric vessel on two saddles. I don't recall for sure, but I don't believe the vessel weight is even considered, it is assumed negligable compared to the contents. I am referring to "classic" Zick. I am aware that Codecalc (for example) has extended tha analysis somewhat, but I find it useless for heat exchangers. Over the years I have come to have less and less faith in it for the type of work I do.

Doesn't mean I never run the calculatons though.

 

[codeeng]

Generally the biggest problem with shop hydro's is the structure of the floor rather than the saddle/shell stresses. You can check your shell stresses using Zick as mentioned above but more importantly check the bearing stress at grade.

 

[Geof]

It's been twelve years since I did an analysis of this sort, but I used this book for a reference:

Tubular Steel Structures, Theory & Design - Troitsky

It covers pipeline supports, which is similar to this application. I just checked & it is for sale ( at $24.00 ) & apparently recently updated at:

https://ssl.lincolnelectric.com/foundation/

This foundation sells a number of other reference books at very low cost in order to promote welding. They also sponsor(ed) training in welding and structural design that may be of interest to you.

Please post how things go...