Estimating Displacements

[geotechguy1]

Suppose you have a tank where the base shear applied from fluid moving in the tank during an earthquake is greater than the available resistance from friction / passive resistance of the slab it's sitting on.

Has anyone estimated displacements in such a situation, and if so how? I.e. assuming we just want to let the thing slide around rather than try and anchor it somehow.

 

[HTURKAK]

It is hard to answer this question. My points ,

- The horizontal sliding displacements is not common and should not be big due to rapid reversal of seismic loading,
- I remember a case , 1989 Lorna Prieta earthquake a jet fuel tank slide about 150 mm on its foundation,
- It can be estimated with time history analysis if you can obtain real time -history acceleration reccord,
- API 650 suggests 50 mm displacements for Piping Attachments for Self-anchored tanks,

If this is a real case , i would suggest you the following options;
- provide mechanical anchors to limit the horizontal sliding ( API 650 provision)
- sharpen your pencil and review the friction coeff. ( ASCE 7-16 Suggets V < W tan 30° ),
- incorporate the bottom slope to the calculations ( assumed cone up bottom with max slope 2 percent )

You may prefer to post more info to get better responds..
...


He is like a man building a house, who dug deep and laid the foundation on the rock. And when the flood arose, the stream beat vehemently against that house, and could not shake it, for it was founded on the rock..

Luke 6:48

 

[geotechguy1]

I've pencil sharpened to phi = 40 to 45 degrees and a KP of 6 for a passive wedge in front of a proposed shear key.

For clarity, I'm not the tank designer per se but we're evaluating whether a rigid inclusion or stone column type solution with a gravel load transfer platform is a viable option for the foundation. The structural engineer wants 20m long 1m diameter concrete piles with fixed connections to the tank for the sole purpose of resisting lateral loads. The largest tanks are ~10m diameter so I don't think they are huge by any means. The ground conditions are poor (100+m alluvial deposits) and I'm in NZ so the seismic demands are quite high.

I agree displacements shouldn't be big - everything is flat so unlike a slope there is no static shear bias causing the displacements to ratchet / accumulate with each cycle in the manner they would in a newmark sliding block analysis / pseudostatic analysis of a slope. To me it seems straightforward - it only fails in sliding for the ULS earthquake and the code provision for ULS is life safety, i.e. we should be able to allow connections to break if we have to which is basically the approach being taken for everything else at this site. I'm just mindful I'm going to be asked the question. Was hoping there was a way that didn't involve Plaxis3D and a time history analysis.

 

[HTURKAK]

I disagree with this solution. 1 m dia , 20 m long concrete piles with fixed connections for the subject ground conditions (100+m alluvial deposits) would be useless. You know that the lateral soil displacements due to seismic shaking increases at deeper levels.
I am not sure i got visualize correctly your case. Probably , the use of stone columns ( which will compact the soil with tamping )could be viable option.

Your statement the tanks fails in sliding for the ULS earthquake. I understand this is SSE ( safe shut down EQ with a period of 2500 yrs). In this case , i think it is necessary to use judgement . How important this tanks, what will be consequences if connecting pipes ruptures ..
I would prefer do nothing ( but provide flexiblity to piping ) if this a water tank at a swine farm.
My opinion..
..

He is like a man building a house, who dug deep and laid the foundation on the rock. And when the flood arose, the stream beat vehemently against that house, and could not shake it, for it was founded on the rock..

Luke 6:48