Silo Walls-Seismic Forces

[IJR]

How do we determine wall forces due to seismic loading for a silo?.

I think if the silo were resting on columns, column forces can easily be determined. But my case is a silo on ground.

Thanks in advance.

 

[Riz]

IJR,
Provided seismic loading is onerous and the dynamic effect of wind loading has been overshadowed then u should consider base shear and lateral force distribution over the silo. Since your silo is continuous u may elect to create pseudo joints at defined intervals. This will assist in the evaluation of cumulative moments the structure is to resist. With this consideration extreme fibre stresses can be calculated for the case of in plane moments and axial loads at every level. The silo should be considered as one unit as it will resist forces from any direction as a lumped mass. L is the circumference and h is wall thickness. Furthermore, a torsional response to seismic forces needs to be evaluated.

Foundation design has to include the resulting base moment.

I have assumed that silo walls have been designed to withstand internal forces and u have pertinent knowledge in design of eq resisting structures.
Riz

 

[IJR]

Thanks Riz,

I can proceed the way you suggested, but what kept me from doing that is as follows.
1) Design seismic shear force is usually assumed to concentrate at the centre of mass. With this assumption base shear and overturning moments are basically known.

2) But the distribution of this shear force and moments along the height of the wall is not anywhere clearly defined. For tanks of normal height this wont probably matter, but for taller tanks, should I not make any seismic considerations for levels above centre of mass?

3) API 650 which covers similar tanks containing fluids has relatively more rigorous analysis for wall forces due to seismic forces and I assumed there could be similar considerations for tanks storing grain.

If you have any comment about those points, please do comment. Currently I am doing exactly what you recommended.

 

[Riz]

IJR,

The distribution of lateral force is as mentioned earlier. Say if u have a 30m high silo, then the pseudo joints could be placed at 3m intervals. Basically u have a situation of fictitious diaphragms at 3m intervals vertically. The idea is to design walls with an associative moment so that the section does not get over-reinforced.
Note that the location of seismic force is not at the centre of mass vertically but acting at its centrodal axis in plan. The above covers your items 1 and 2.

The torsion effects should be in the order of 2.5% width x lateral force at every joint created.

Overturning moments need seperate consideration. This should be applicable at base and roof levels only.

U also need to make a stick model of the silo and load the joints with lateral force for obtaining deflections. A pdelta analysis can be carried out using staad for instance.

Being not familiar to API 650, I am in the opinion that the requirements for rigorous analysis is absolutely prudent and its a concept I am trying to build 4 u.
Riz

 

[IJR]

Thanks once more Riz. I am right now implementing your ideas.