Vessel Design and Analysis 6

[ALK2415]

Hi,
looking for textbook that deals with design of oil storage shell/vessel (similar to below picture)
with special attention to wind and seismic loadings,
Thanks in advance for your guidance and support.
Note, any reference with FEM modeling of such kinds will be also helpful

 

[JStephen]

With smaller pressure vessels, there are numerous manufacturers and information regarding them is fairly widespread and publicly available.
With large spheres, that tends to be more specialized, very few producers, and information tends to be more proprietary.
With that said, check latest ASCE 7. ASCE also has a book available on wind loading on refinery equipment that may be useful.
Some of the vessel handbooks have limited information on leg supports (Moss, Megyesy), although they may not be up-to-date with current building codes.
If you have access to vessel design software that covers the geometry, the documentation and references there may help.
Realistically, it is best left to experienced producers of the product.

https://sp360.asce.org/PersonifyEbusiness/Merchandise/Product-Details/productId/269951644

 

[ALK2415]

thanks JStephen,
its for Teaching-Training young Engineers about these kind of structures
but for large scale facilities few references I found
the Autopipe-Vessel v42 (closed box) for design storage tank with limited options to alter its design
Staad-Advanced foundation almost same
basically doing all FEM modeling step by step for stresses checking (tedious process)

 

[AskTooMuch]

How reliable is Staad FEM for plate/vessel design? I thought it can't do buckling check of plates.

 

[ALK2415]

@ asktoomuch
STAAD,Pro is more than capable ! it versatile
like for example above structure - will check wind drift & seismic loadings and the rest of load types (DL, LL, S, R, even dust !). it also fix the hydrostatic distribution across all spherical shape plate structure.
added many branches like
1- RAM connection, for example most of steel connection (in above structure) can designed easily and according to latest AISC, ASCE, IBC ...etc
2- Advanced foundation, where you could design above structure foundation directly as (Pile-cap, Ring, separate, combined, or Other types). the design includes shop-drawings (high level Dxf and isometric views) and BOQ in details (see attached file)
3- pretensioned cables (bracing) and other stiffening could be added

4- buckling stress check: i think (not sure) max absolute stress (-ve for comp & +ve for tensile stress) will cover the maximum critical stress that will be developed in membrane (plate) vessel structure

5- above example dimensions (25m dia, 16 mm thick) without stiffening (still working on it) ---> max stress = 165 MPa and "270 MPa" in ring area (were thickened plate needs to be adjusted) (see picture)

6- most important thing of integrated design is that (STAAD) will takes all applied load combinations (per CODE) and choose the most critical case of loading per item design (member, connection, or footing )

7- it also has Select command which well optimize your section in (normalized group selection)

- In few words, it need clear thinking and some imagination to solve the "Hard" part of the modeling step

 

[ALK2415]

one more thing,
whole this started because one of the Oil companies, designed oil storage tank (strange) footing of crushed stone on top of very unstable subgrade (sedimentary soil), is that Normal-valid Design ?
While STAAD foundation gives the more reliable design of Ring beam footing ! (last picture)

 

[HTURKAK]

Eng. ALK2415;

Spherical shells is used for the storage of liquified gas products ( LPG, Propan . etc ) and the applicable codes ( ASME VIII, DIN AD2000 -Merkblatt ..) You may analysis the stresses with structural analysis softwares ( FEM model) but design of the PV's totally different story and industry..

I will suggest you to look ;
- GUIDEBOOK FOR THE DESIGN OF ASME SECTION VIII PRESSURE VESSELS (By JAWAD MAAN )
- Pressure Vessel Design Manual ( Dennis MOSSES )

Regarding the pictures at your latest post;

- The first picture is not clear but section shows granular foundation for ground supported tank
- The third picture is Crushed stone ring wall foundation detail for ground supported tank ( See API 650 App. B)
- The last pictures ; 4th and 5th , Concrete Ringwall foundation for ground supported tanks.

You may post your threads to the Storage tank and Pressure vessel forums to get more specific responds.



Use it up, wear it out;
Make it do, or do without.

NEW ENGLAND MAXIM

 

[ALK2415]

Thanks HTURKAK

 

[dig1]

Alk2415,

HTURKAK brings up great points. As a STAAD user I would not design a pressure vessel, with respect to local stress analysis, with STAAD. When looking at local stresses for this type of vessel designed to ASME the rules of Section VII, Division 2, Part 5 Design by Analysis is the proper procedure (EN 13445 has its own criteria and that may be the proper selection for your location). In ASME for a linear elastic analysis you will need to evaluate the linearized stress values against the various failure modes and also include buckling and if necessary fatigue (there is a fatigue screening section). I do not believe that STAAD linearizes the stresses. The alternate is to perform an elastic-plastic analysis and again there are rules for this. For this local analysis you can use Ansys, Abaqus, FE/Pipe, FeMap, or another appropriate program (I'm not endorsing or excluding any other than saying I would not use STAAD) and working through the provisions of the respective pressure vessel or tank Code.

STAAD will give you the local loads from the structural analysis to use in your local analysis.

 

[ALK2415]

Thanks dig1
Iam using this example as teaching/training materials for structural engineers (checking of stress)
if it was for design purposes, I'm sure a large scale oil-company will do that jobe
some time to kill so i added some stiffeners and change the column layout
my new question is the rules of thumb still applied here for columns & beams (initial size of members) given by AISC ? [ATTACHED PDF]

 

[Geoff13]

It's quiet Sunday, so I'm going to pass along a few thoughts.

With large spheres, that tends to be more specialized, very few producers, and information tends to be more proprietary.

As noted by JStephen above, the design of spheres and their supports is largely a proprietary product of a few large companies. There are no textbooks.

However, there are a few obvious issues with your model that I will still comment on.
[ul]
[li]25m diameter is a huge sphere. Perhaps 10 m diameter would be sufficient for a training exercise.[/li]
[li]Based partly on the screen shots in your other thread (thread1452-513586), you are modelling the columns as 2D line elements that are likely attached to a single node of your sphere elements. This will lead to HUGE stress concentrations. The columns must be modelled as a 3D mesh, like the sphere, so the attachment between them can be the actual contact area between the sphere and the column. This will be require very interesting meshing of the sphere and column in the attachment area. However without accurate modelling of the connection you are going to get unrealistically high stresses.[/li]
[li]From a modelling perspective, lots of columns may help reduce the stresses in the attachment area, but the cost to build so many will be too large. Look at real spheres and you will see far fewer support columns than you are modelling.[/li]
[li]As shown in the 3D rendered screen shot in your other thread, the top panel of bracing goes through the sphere. You must get rid of these X-braces since they can't be installed in real life.[/li]
[li]This appears to be a liquid storage sphere, but the shell thicknesses seems to be uniform. Not the way real spheres are built.[/li]
[li]In the 3D rendering in your other thread, the horizontal struts between the columns are shown as multiple segments, and even look radiused. These struts must be straight, or they can't do their job. If I have misinterpreted the rendering I apologize.[/li]
[/ul]

I think that's a good set of basic issues. I expect there are other more detailed issues. I suggest you first get the simple gravity model working. Then you can consider adding wind loads. I'm not sure that seismic design is needed for a training exercise. If a student ends up with one of the large companies building spheres they will provide the needed training.

... designed oil storage tank (strange) footing of crushed stone on top of very unstable subgrade (sedimentary soil), is that Normal-valid Design ?

It is generally a bad idea to ask an unrelated question. Many folks won't scroll through your sphere questions to realize you also have a flat bottom tank question.

The product load is the primary load, and is uniform across the entire tank bottom. If the subgrade is poor then adding a ring wall won't help. My personal experience is that 95% of API tank use this granular berm foundation. While a few tanks require a concrete ring wall, they are too expensive to be used for most tanks.

 

[ALK2415]

Thank you, Geoff13, for your kind answer and for your time spent writing detailed answers.
so trying ABAQUS will be much more accurate and realistic solution with 3D element meshing (thin plate)
tank was for pressurized gas 1.0 MPa (145 Psi) LPG
Note: Those engineers are already working on construction sites (with no experience) and have even taken real college courses due to covid times. So they return to trying to give them a basic analysis of oil facilities, like below:
i did what i could to solve most of the practical design problems (show below) and also working for the rest of structures
If you point me to some textbook that solves high temperatures that affect structural steel sections like flare stacks, power-generating chimneys, etc.
Hopefully, they will understand some basic loading distribution and support reactions as learning materials
Those companies do not encourage or enforce training programs or seminars before going to the field.

 

[Geoff13]

Why would you store LPG as a gas? Bump the pressure up enough to liquify it and your 25 m dia sphere only needs be about 4 m dia!! Much, much, much cheaper. I already said your sphere had too many columns, but if it's for gas only both the number and size of your columns is very, very excessive. This is the only gas sphere I'm aware of (Link). It's not mine, but is visible from the highway. This sphere will not have earthquake loads.

Even a training exercise should be realistic. 99.99% of spheres store liquids, so practising on a gas sphere is a waste of time.

As for the rest of your questions, good luck.

 

[AskTooMuch]

The vertical bracing of the piperack looks crazy. Also that pipe support with Ts every few feet.
Maybe before doing loadings, use structural layout first that is both stable and economical.

 

[rb1957]

is anyone going to comment on the mesh density of the tank ?

"Hoffen wir mal, dass alles gut geht !"
General Paulus, Nov 1942, outside Stalingrad after the launch of Operation Uranus.

 

[lexpatrie]

Those Socrates rules of thumb are intended for building design. It looks like a fair amount of your steel is unbraced, so those rules of thumb would likely underestimate what you actually need, but they are a starting point, and if you get pieces below those estimated sizes it's something to investigate, as I'd be suspicious of those results. This is a wildly complicated model, so try to check portions by hand (with needed simplifications) as best you can. I think the whole process (including the simplified checks) is very teachable. Even for senior engineers.

Regards,
Brian