How do I determine the spring constant for a pile?

[seenote2]

My pipe rack in STAAD is generating too much loads on the piles. So instead of doing a fix support I want to do a FIX BUT support with spring constants. How do I determine the spring constant for my support? I know the pile tolerances and some soils information.

 

[EireChch]

settlement of pile head/load on pile

 

[ukengineer58]

take the load (say in newtons if thats the programs units). Divide by settlement and elastic shortening (which wont be much). Normally use about 10mm in total. So for 600 kN pile 600000/10. assuming the spring unit is N/mm

 

[EireChch]

my mistake, I was wrong way around! - UKeng is right.

 

[Mccoy]

It also depends on the software. I've been working mainly with a FEM-based software. The procedure is to discretize piles in finite-length cylinders. The reaction of each cylinder is modeled by 3 Winkler springs, horizontal, torsional and vertical (all applied to the lateral surface). The Base cylinder has its own vertical spring applied to the base circular surface.

Other softwares model the pile equating it to a single vertical spring, plus a horizontal one. Other ones employ a value called the 'Distributed rigidity' (Fleming, 1992, Salgado, 2008). There may be more which I don't know.

The OP doesn't specify which is the case.

A simplified formula for the basal spring, from Castelli & Motta,2010, is the following:


Kbi about= G0/R0(1.25), where G0 = shear stress at small strain, R0 = pile diameter/2, all in compatible units. G0 should be reduced to G if the strain is not negligible. Lateral reaction may govern though, depending on soil.

Units are particularly critical when dealing with spring reactions, once I mistook the final value by two orders of magnitude, due to a missing conversion in my spreadsheets.

On the other side, the shear values are not very sensitive to the spring constant so mistakes or uncertainties do not appear to influence significantly the results, usually.