steel Silo compression ring

[hawkaz]

I am trying to determine the appropriate allowable compressive stress in a steel silo compression ring.

Gaylord and Gaylord indicate 10 ksi, but note that this could increase to 18 ksi for low frequency usage silos.

Troitsky indicates 10 ksi.

Both of those just used blanket statements with no backup.



I found another source that indicate 12.8 ksi and a fourth that used .5Fy.

The fourth source appears to be the only one that indicated that they were taking Fy into account.

Does anyone have any insight that would be helpful?

 

[dhengr]

Really read both Gaylord and Gaylord & Troitsky a little deeper, rather than just a compressive stress allowable, like .5Fy, it is a curved plate buckling problem.

 

[ToadJones]

Is this "compression ring" actually in tension, but meant to induce a compressive load on the silo?

 

[ToadJones]

Yea, I'd agree that just saying "10 ksi" is a bit asinine. So, does that mean you can use a bamboo ring and 10 ksi is still cool?

 

[BAretired]

I don't have any experience in silo design, but it is far from simple. Symmetrical loads such as you get when filling result in membrane stresses which are easy to comprehend, but unsymmetrical loads which can occur as a result of the flowing of the stored materials are very difficult to analyze by hand methods.

The compression ring gets its compression from the tension in the conical hopper, but it also gets bending, torsion and "rolling" due to eccentricity of the hopper shell from the centroid of the ring. A compression ring can buckle, so the buckling length enters into the calculation as well.

There have been many failures of silos over the years. The following is a very interesting link to an article written by a firm which has investigated numerous silo failures.

http://www.jenike.com/TechPapers/load-dev-silo-design.pdf

BA

 

[hokie66]

My version of Gaylord, admittedly the 1968 one, said the recommendation of 10 ksi was "used to minimize deflection, and hence the secondary bending stress". Fy wouldn't have anything to do with deflection. This is in the "Steel Tanks" section of the Handbook, so bamboo is excluded.

This section was written by two engineers, including the Chief Structural Engineer, of Chicago Bridge and Iron Company. I think their advice would have been based on a lot of experience.

 

[hokie66]

I think the article BA linked amply demonstrates the need to be very conservative in bulk material storage bins, particularly if you have limited experience in this field, as do I.

 

[JStephen]

For compression rings in tanks, API-620 allows 15,000 psi, and also specifies limitations on thickness and width intended to preclude buckling, and on location to limit rolling. The low stresses are mainly an attempt to be conservative due to the approximate nature of the analysis.

If a compression ring is actually in tension, then it's a tension ring, not a compression ring. It is usually a flat plate or an assembly of elements, not a curved plate.

Gaylord and Gaylord wrote both the Structural Engineering Handbook and also a bin and silo book, so "G&G" can mean two different books.

 

[hokie66]

The API standard is not for bulk solids storage. Solids create much more variable conditions than liquids and gases.

 

[hawkaz]

Thanks for all the input.

To answer a couple of the questions from above:
1. This is the compression ring where the hopper meets the shell.
2. This silo is containing solids
3. I have Gaylord and Gaylord's Structural Engineering Handbook. I haven't been able to get my hands on their bin and silo book yet.


As Hokie indicated, these books recommend limiting the compressive stress to 10 ksi to minimize deflection and buckling- I just hate magic numbers and was hoping to find out how they got to that number- and just how conservative they are being.

Thanks

 

[SAIL3]

I believe hokkie66 is correct. I have used 15ksi as an allowable stress for A36 when I have a good handle on the loading ie: uniform press, dead, live, wind,seis etc.
However, when it comes to loading from the flow of solids, I am not as confident that I have addressed all the possible load variations so I would tend to use 10ksi as a extra factor of safety to reflect this. The ref. given by BA indicates that this is a very complicated loading condition and I would certainly be very careful in addressing it. The 10ksi allowable is really a shot in the dark and , in hindsight, may not be a good substitute for a more rigorous analysis.