Skin friction of drilled piers (Piles) 1

[JaleelurRehman]

Why there is difference in upward and downward skin friction of drilled piers (piles)? If someone can explain theoretically.

Jaleel Ur Rehman
Geotechnical Engineer
Osaimi Engineering Consulting Office, KSA

 

[SlideRuleEra]

I can't offer a theoretical explanation, but at one time having been a Contractor who both drove and extracted piling, I will make an observational educated guess.

Assumption: A single friction pile's uplift capacity is lower than its compression capacity.

For compression loading, all of the soil surrounding the pile is more or less supported by underlying soil. Therefore, the total surface area of the pile is mobilized to resist movement by friction.

For uplift, only a portion of the pile's surface area is available to resist movement. The soil closest to the surface has nothing above it to directly resist upward movement. The soil in contact with the pile tends to move upward with the pile.

In summary, my guess is that for compression loading there is both more useful surface area and a higher coefficient of friction than there is for uplift loading.

I'm sure this is too simplistic an explanation; I would not be surprised if these observations point in the right direction.

http://www.SlideRuleEra.net

http://www.VacuumTubeEra.net

 

[BigH]

M.J. Tomlinson discusses this in his book Foundation Design and Construction - section 7.16. He says that published data suggests a reduction of 50% for granular soils and same for long term loading in clay - be careful of short piles subject to vibratory horizontal loading. He goes much more into detail in his Pile Design and Construction book - 5th edition, Chapter 6.

Poulos and Davis' Pile Foundation Analysis and Design suggests for granular soils a reduction of 1/3 can be assumed - but best to do a pile load test. They also say that the published reports have quite a bit of scatter and in some cases in conflicting.

 

[Ron]

I work in coastal plains soils, mostly sands, clayey sands, silts and silty sands.... and the general rule is that tension capacity is about half the compression capacity.

 

[Mccoy]

The theoretical causes may not be clearly known apparently.
This article by England provides interesting hypotheses based on bidirectional load tests.

Basically, he concludes that there is no real significant difference from tensile and compressive capacity and observed differences may be due to locked stresses and interpretation of load tests in relation to definition of ultimate failure. Interesting and heretical views, summarized just before chapter 4, page 13 of the linked PDF.

http://www.loadtest.com/INT_media/680-693 England - On the subject of Piles in Tension.pdf

http://www.mccoy.it

 

[Ron]

McCoy....our premise is based on load tests as noted. The results of the load tests seem to correlate with SRE's explanation of lower resistance to uplift in the upper part of the pile...also fits with basic soil mechanics....less skin friction with shallower depth.

 

[Mccoy]

Ron, agreed, that's why I wrote 'heretical views'. The heretic author's conclusions though are also based on load tests, many of'em and pretty sophisticated, 'bidirectional', as to avoid inherent bias.
If 'the proof is in the pudding' then we should eat that pudding and judge by ourselves. Of course, no way I wouldn't follow what the relevant codes suggest. the Italian building code, which follows the European code, has a safety factor for shaft lateral friction in tension which is 9% larger than lateral friction in compression.

2 quick non-heretical references, the sources I use most:

For side resistance, there has been speculation in teh literature that the side resistance in uplift would be less than that in the compression. However, examination of possible Poisson effects for shafts in soils has shown that these effects are negligible. Similarly, examination of available load test data has not shown any discernible difference. Some of these data are given...which show that the uplift and compression data are uniformly distributed and that the same conclusions would be reached for either data set...the author goes on explaining that occasionally a cone of soil may develop in uplift that effectively acts as a part of the foundation; the author suggests a Qsu reduction when D/B<6 and either beta or Alpha su/gammaD >1. A correction factor is given

In this excellent textbook, I cannot yet find a description of this problem. May be it has escaped me, maybe it's not cited because the author believes the two cases should not be treated differently. It's a very good theoretical and practical book

http://www.mccoy.it