Existing pile ok on rigid pile cap analysis. Not ok using FEA RisaFdn 1

[AskTooMuch]

I have an existing pile cap. When I checked the pile load, I'm getting 10kips difference between rigid and FEA analysis using RisaFdn.

Is it ok if only one method pass? Let's say it was reversed and pile allowable fails in rigid pile cap analysis and pass using FEA, is this acceptable too?

Pile is skin friction. I would think the pile in question will just settle a little and load gets redistributed among all piles.

 

[JoshPlumSE]

Could you have modeled the piles as being 100% rigid in RISAFoundation? I've seen a lot of people do this and come up with some odd results.

Ultimately it's an engineering judgment call. Different types of analyses will give different forces. So, it's up to you to decide which is the most appropriate method for each situation.

 

[AskTooMuch]

Yes, it was by default rigid. I assumed 1000kips/in pile stiffness and now have lower than allowable, even lower than the rigid pile cap analysis.

I think the "rigid" pile is messing this calculation up because 1000kips/in is close to rigid yet I get significant difference in pile reaction.

 

[r13]

I would call RISA to discuss the discrepancy, as it must have done program validation against the time tested, well established method.

 

[AskTooMuch]

The discrepancy is expected because one is rigid cap analysis, the other is semi-rigid analysis.
The question is, what if one method results in pile failing while the other method all piles passed.
Can I just switch analysis to whichever pass?

 

[KootK]

By what percentage do the two methods differ? Dirt stuff always lacks precision and I like your settlement / redistribution argument. As long as the difference isn't excessive, I'd role with the method that works

 

[Bowsers]

In electrical / substation design there is generally a hierarchy of analysis depending on the amount of power relative to the piece of equipment. A rigid analysis is considered the least sophisticated method, where an FEA is the next level, and from there different types of loading on the FEA is the most rigorous analysis.

Is there standard documents or methodology within this design type that outlines a requirement? If this is an offshore oil application I would hope this is at least a class III structure or similar.

 

[TLHS]

The rigid pile cap assumption isn't messing up the calculation, it's just possibly not valid. This all sounds like expected behaviour to me?

If you do an analysis of the cap where you assume the cap is even slightly flexible but the piles are perfectly rigid you'll get distribution based on a flexible cap assumption (flexible cap, rigid piles). That may be reasonable if you're on something pretty rigid like bedrock. If your piles are founded on something that moves even a fraction of an inch, that pile movement might be an order of magnitude larger than the bending and shear deflection in the cap. In that case, loads will be distributed reasonably evenly across the the piles (rigid cap, flexible piles).

Pile axial stiffness under load, including the geotechnical movement, is a hugely critical thing to model. You likely don't have to be all that accurate, but you need to at least play with the sensitivity.