Difficult to rely on full lateral brace from poor soil, especially near the surface where you need all of it you can get, and the buckling length is 2x L.

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

It has nothing to do with lateral support of pile in poor soils. 0.25Fy accounts for damage to the pile tip during driving. Per AASHTO 15th Edition of Standard Specifications for Highway Bridges, Section 4.5.7.3: The maximum allowable stress on a pile shall not exceed the following limits in severe subsurface conditions. Where pile damage or deterioration is possible, it may be prudent to use a lower stress level than the maximum allowable stress. For steel H-piles, the maximum allowable stress shall not exceed 0.25Fy over the cross-sectional area of the pile, not including the area of any tip reinforcement. The maximum allowable stress may be increased to 0.33Fy in conditions where pile damage is unlikely.
This is taken from a very old, 1992 AASHTO manual. For new work, you probably need to use newer or current AASHTO manual specifications.

www.PeirceEngineering.com

Thanks. Thought it did.

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

1503-44 -- I think that's a fair concern and it was only recently that I saw it to be a normal (or at least not unusual) assumption, for example in FHWA's example here (Step P.6), "Ultimate axial compressive resistance is determined in accordance with either equation 6.9.4.1-1 or 6.9.4.1-2. The selection of equation is based on the computation of l in equation 6.9.4.1-3 which accounts for buckling of unbraced sections. Since the pile will be fully embedded in soil, the unbraced length is zero and therefore l is zero. Based on this this, use equation 6.9.4.1-1 to calculate the nominal compressive resistance.":

https://www.fhwa.dot.gov/bridge/lrfd/us_dsp.cfm#de...

Or in Bethlehem Steel's old H-Pile Manual where for a pile with "entire length embedded in any soil or combination of soils, other than virtually fluid material...the pile is supported throughout its length, and no reduction in load is required because of slenderness ratio." And so they end up a with a capacity based on a fully supported column. I think normally in the past when I've looked at stuff like this I've assumed an unbraced length based on soil-structure interaction results from something like LPILE or FB-Multipier using one of the many methods to determine a theoretical point of fixity, or assuming some depth to a point of fixity and finding the effective unbraced length using an assumed K between 1.2 and 2.1 depending on the pile head fixity.


PEinc -- right, if this were for new design definitely I would use the newest edition of the LRFD specifications. We're just being asked to evaluate these old existing H-piles using a service load/allowable stress methodology. I guess I was wondering if anyone had any insight as to the "why" of 0.25Fy (or 0.33Fy) for piles specifically.

For example, was it the case that engineers used to just find the pile reaction from simplified analysis (Pile Reaction = P/# piles + M*x/Iy_pilegroup + M*y/Ix_pilegroup) and they used a lower allowable stress because they weren't accounting for any soil-structure interaction or moment in the piles? In the "pile damage is unlikely" case where you can use 0.33Fy, it seems inconsistent that you'd allow such a low percentage of yield, where if you considered these piles as a typical compression element, the allowable stress would be a function of 0.55Fy and the slenderness ratio, which will probably result in an allowable closer to 0.55Fy than 0.33Fy. So I was thinking the 0.25 (or 0.33) allowable may've been some way to conveniently account for those kind of effects, or to sort of guide the engineer toward choosing a pile that wouldn't have issues during driving.

I've always taken some unbraced length towards the surface above the point of inflexion, Here I thought it was that same issue, Not knowing AASHTO, I should have just passed on this question. I also don't ever recall seeing a pile tip damage factor. Guess we assumed the driving caps would not be damaged.

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

The stresses in the pile can be twice as high (or more) during driving as the design load, due to the impact.

The resistance factors for LRFD are similarly low. We use 0.5 Fy, based on the driving measurement techniques we specify.

Is it the case then that the 0.25Fy (or 0.33Fy) is primarily to make sure the piles can be driven safely? For evaluation of an existing pile for design loads, wouldn't using 0.25Fy be pretty conservative in that case?

The original design assumption would have been that driving it with a force approximately equal to .5*A*Fy would result in a (nominal) bearing capacity of .25*A*Fy. I wouldn't be comfortable making a different assumption after the fact.

Gotcha, that makes sense. I think I was having what might amount to some magical or wishful thinking that the capacity might be limited to 0.25Fy so that during driving they could get these things into the ground without damaging the pile, but after they were driven to refusal and survived the driving process, they would be limited to the smaller of the structural or geotechnical capacities -- and that the actual nominal structural capacity in bearing (after we're no longer concerned about driving) could be a good deal higher, since it might not be materially different from a laterally supported column at that point.

Quote:

...after they were driven to refusal and survived the driving process, they would be limited to the smaller of the structural or geotechnical capacities -- and that the actual nominal structural capacity in bearing (after we're no longer concerned about driving) could be a good deal higher,

If that could be assumed, there would be no reason to limit the design capacity to the.25A*Fy in the first place. I'm fairly sure the design assumption is that the .25A*Fy is the geotechnical resistance limit after driving.