Tank Roof Rafter Design 4

[massad]

Dear All ,
I am working on a 60m dia tank for fuel oil. Before I have worked on small diameter tanks and hence self supported cone/dome/umbralla roof were satisfactory. But now I need to design a dome roof with rafter on it . Can any one give me an idea about how to proceed for the structural calculations for the roof . API does not give any information this kind of design. If any one has any literature or know any websites or books please let me know .

BR
Assad

 

[DSB123]

massad,
Just a word of caution. If you have not performed calculations before on the structural design of tank roofs then it may be worthwhile searching for a competent design company that has. If you are intending to tackle the job yourself then I would submit your calcs to a competent person, familiar with the process of design, for review.

 

[IFRs]

API 650 covers supported cone roof design in section 5.10 (11th edition) and section 3.10 (10th edition). However, the details are up to you. I agree with DSB123 that your design should at least be reviewed by an expert.

 

[COSTASTSAPROUNIS]

Dear Massad
First of all define the loads an the load combinations.
(snow, live load, wind etc.)
After this distribute the loads to the rafters.
You can examine strength of the rafters - take care for the bracing- and be careful for the buckling effects of the rafters especially the lateral torsional buckling.
The strength of the rafters should be examined under some design code (ASCE, EC3 etc).

After having some basic predesign results use a software to see if you have the same results ( i use SAP2000). But remeber that when you will simulate the material of the roof plates change the young modulus from the value of steel to a value for a very very soft material. Otherwise the roof plates will produce a self support phenomenon with membrane forces.
If you have reasonable reason between your hand made calculations and simulation then you are in a good way.
The above are the first step for the design .
If you find old drawings from other tanks then use them they will have a lot of experience included on them.
Designing such a roof is not something easy but it is not so difficult.
Field engineers can help you also with sketches and drawings.
I ll see if i have AUTOCAD drawings for such roofs to send you.
Work Hard
Costas


Dr. Costas J. Tsaprounis

 

[massad]

Dear Dr. Costas J. Tsaprounis,
Well I have done some basic calculation considering a simply supported straight beam with trapoizoidal loading . Loading increases as we move towards outer compression ring . The loads I have conseded in UDL is as you have suggested live load , snow load etc + the weight of roof plate . I have also considered a lineraly distributed load since the roof plate would be in pie shape having more material , hence weight towards the end . So combining these two loadings I got trapoizoidal loading . Using these loading I calculted the maximum bending moment for the selected section and unsing this moment the maximum bending stresses and compared them with allowable. Is it also necessary to check it for shear force . Also is is ok to consider it as a simply supported beam or should I consided it as fixed beam.

Now the main thing as you have mentioned is lateral torsional buckling . I am still working on that part . If you have any suggestions , let me know .

BR

Assad

 

[COSTASTSAPROUNIS]

A simple supported beam is a good approach. The reality is simply supported at one end and at the other end a torsional spring which represent the torsional stiffness of the ring.
Becarefull on the bracing in order to define exactly the maximum free length.
Costas Tsaprounis

Dr. Costas J. Tsaprounis

 

[carthago]

Let is try to built up a sort of check list ..kind of screening...

Of course feel free to add extra points....


1/-Your tank is 60 m diam.

2/-How is your roof slope ?? (1/6 = 9 degree.. or something similar)
It should appear after calcs that the deflection would not allow to water to accumulate (ponding)

3/How long is your rafter ?
probably longer than 30 meter ...this could be built up out of several standard lengths...

4/-What are your loads ( on the rafter) , if I remember from API 650 it is = 25 psf live load (LD)+ 20 psf dead load (DL) of the roof + own weight of all beams

the 25 and 20 psf will give a trapezoidal distribution on the rafter ( small near centerline..)

5/-What is the distance between two adjacent rafters on the shell side ?
According to API 650 , this shall not exceed (2*3.14 * foot )= 1. 9 meter.

6/-How are the reactions transmitted to the tank walls?
( rafters connected to a brackets , which are welded to the inner wall of the shell or something equivalent)
Do to the roof slope , you will get also radial loads on the walls


7/-Do you have headers ( a ring of beam between centerline and tank edge ( at say OD/4 to support the subrafters?
You might need them , because you can not have all rafters all running from center to tank wall

8/-Do you have subrafters ?
This is because you want to keep those 1.9 m distance...

9/-How are rafters supported near the tank centerline ?? ( on a ring ? on a column ?)
The ring will take the horizontal reactions (again because of the slope) ...


I am trying to do something on roof and trying to exchange info..on another site ( ast-forum) but apprently all the guys are on leave...(lol)
-----
Fathi

 

[massad]

Hello All ,
Well I have almost done with some calculations , just need one suggestion , do I go for dome roof with rafters or dome roof with rafters and internal column. Also after what diamter should internal columns should be used .

Also DSB123 how can I send my calculaitons for review

BR
Assad

 

[IFRs]

The decision to use a column or not should be purely economic - evaluate the materials, fabrication and erection costs for each and the answer should become obvious. If you are considering internal columns, would you consider several columns and a simple cone roof? For a supported dome roof, I'd design the structure independent of the roof plates using FEA. You could use a triangulated structure similar to typical "geodesic" style aluminum dome roofs, or use radial arches with some depth with a compression ring in the middle and a tension ring at the outer edge. This is a large enough roof that you would benefit from engaging a firm or consultant that has done them before in this size. It's not economic for you to reinvent the wheel!