Silo design using FEA - applying loads to steel silo

[MikeG7]

I am designing a silo for limestone and am doing an FEA using the variable pressure in the barrel and the conical hopper along with the values of wall friction angle, internal angle, bulk density, hopper angle etc. according to the pressure distribution methods described in relevant Codes(e.g. Jannsen's equation , "Walker" theory for the hopper and so on...)

In the FEA, I have applied the circumferential pressure distribution that will give the circumferential stresses. I am at the point where I am going to add the wall traction friction force in the barrel using a surface load in FEA to check the buckling due to compression.

Any idea how I would add the tensile load of the weight of contents and the wall normal friction in the hopper? The weight of the contents is obviously a fixed value but where to apply / how to apply to FEA model.? I need obviously this to check the tension force in the junction of the cylinder and cone hopper.

Much appreciation for any comments.

Using RFEM software (trial).

Regards
MikeG7

 

[JStephen]

I don't know that it matters a whole lot. All of the weight that is not supported by skin friction in the barrel is supported by the cone, and primarily by the cone acting in membrane tension. So you should get very similar membrane tension at the top of the cone regardless of what load distribution you assume in the cone- provided that overall load on the cone is the same.

Some general observations, for what they're worth:
In Gaylord & Gaylord's silo design book, they use Janssen's equation, and say "So-and-so recommends using these coefficients" and give one set of numbers/equations. The next page over "But ol' Whatshisname recommends THESE coefficients" and give another set of numbers/equations. And if you plug them into Jannsen's equation, you get completely different load distribution. The lesson learned, there is apparently a lot of variation in actual and calculated loads in these silos, which means there's not a lot of sense in getting too carried away with the precision of the analysis of the silo itself.

A few years ago, I took the ASME class on "flow of solids in bins", etc. One thing pointed out there is if you want accurate numbers, you have to do actual testing on the product, not just using the value for "limestone" from somebody's table.

If the silo is "small", specifically small enough that minimum plate thickness governs the design, you can save a lot of time by checking for hoop tension using hydrostatic assumption, check for compression assuming entire weight of contents is supported by shell, and check compression ring assuming entire contents is supported by the cone. If the silo is not "small", this will be overly conservative.

 

[MikeG7]

thanks for your comments. Let's say I take the "simplified" approach as described and wish to apply the "entire contents supported by the cone". If it was a flat bottom silo (elevated on leg for example) that would be easy - by applying the pressure over the bottom. But in a cone bottom there is sloping sides and an opening. The normal wall pressures are applied normal to the hopper walls. That part I understand. But if the hopper has an opening at the bottom then there is no surface on which to apply the pressure. I'm thinking that this "missing force" could represent a large portion of the contents weight if the opening is significantly large. In turn the reaction forces on the supports are going to be too small. Any ideas? Maybe I'm overly complicating things which is a bad habit of mine sometimes.

 

[structSU10]

you can apply the 'missing force' as a load around the perimeter of the opening as a line load if you can.

 

[MikeG7]

Thanks fellas I got it now...I think.