Discussion: A practical design approach for piles with negative friction

I would like to thank Professor Poulos for adding to the literature on this subject, and the attempts by himself and Professor Fellenius to get us thinking sensibly about the impact of negative skin friction/downdrag on the performance of bearing piles. The proposed approach puts the problem of dealing with negative skin friction on a rational footing, treating it as a serviceability issue rather than one of ultimate capacity. I would suggest, however, that the ‘alternative design criterion for controlling settlement' proposed in the method would be better conditioned if a margin was provided with respect to the shaft resistance in the stable zone rather than the total (shaft plus base) resistance, R_ug2, in the stable zone.

To clarify this, the following table summarises the calculations presented in the paper, for selected FS₂ values, alongside the equivalent values of factor of safety with respect to shaft resistance, F_s, derived as indicated by the relevant equation.

where UEB is the ultimate end bearing resistance, and other terms are as defined by Poulos.

In the case of the end-bearing pile, when F S₂ is 1·0, the value of F_s is significantly below 1·0, implying that much of the load (P_w + P_Nmax) is carried by the pile base: hence the large settlements that are predicted in Figure 7 of the paper. In all the other cases, the value of F_s is greater than 0·9, in which case it is understandable that the predicted settlements are better controlled; it is well known that much smaller movements are required to mobilise the balancing forces on the shaft (≤ssthan1% of the pile diameter) than on the pile base (>10% of the pile diameter). The values of F_s for the floating pile are larger than those of the end-bearing pile for the same value of F S₂, as a larger proportion of the resistance is provided by the shaft than the base, in the stable zone.

I would suggest that the ‘alternative design criterion for controlling settlement' should therefore be that the factor of safety with respect to shaft resistance in the stable zone, under the combined effect of the pile working load and potential downdrag load, should be greater than, say, 0·9 in order to ensure that settlements under the combined effect of working loads and potential negative skin friction are tolerable (at least for the examples covered in the paper).

Of course if, for example, piles are being driven onto a very dense/strong layer that underlies the potentially settling soils—that is, a truly end-bearing pile—then the proposed criterion is unlikely to be met, and settlement limits would have to be set accordingly, taking into account that the behaviour of the pile will be base dominated, and hence it will settle more—as is the case when there is a significant depth of settling soil, as noted in the paper.

I look forward to hearing Professor Poulos's thoughts in response to this suggestion.

The author is grateful to Dr Bourne-Webb for his thoughtful and constructive discussion. He suggests that an alternative design criterion for control of the settlement of piles subjected to negative friction is to adopt a factor of safety with respect to shaft friction in the stable zone, rather than an overall factor of safety against combined shaft friction and end bearing. A value of 0·9 on the shaft friction within the stable zone is suggested by the discusser. He quite correctly notes that the pile shaft requires relatively small displacements for full mobilisation, thus implying that, once this friction is mobilised, the ‘softer' pile base may experience a rapid rate of increase of settlement with any increase in either applied load or induced downdrag force. The use of the discusser's criterion may lead to slightly longer piles, but this may be prudent if there is concern about the base conditions, for example with bored piles constructed under bentonite.

It is difficult to argue against Dr Bourne-Webb's concept, as it is inherently more conservative than that proposed by the author. However, three points may perhaps be relevant here.

• (a) Suggesting a factor of safety for the shaft alone of less than unity may give rise to concern by checkers or reviewers of the design, especially in some parts of the world where there is a pervasive conservatism in the pile design process.

• (b) As acknowledged by the discusser, the approach may not be practical if the pile is driven to a hard stratum, with little or no penetration of the pile tip into the stratum.

• (c) There is often a tendency for end-bearing resistance to be underestimated, so that the real factor of safety of the pile within the stable zone may well be in excess of that used for design.

A possible compromise approach may be to consider both the proposed criteria—that is, to seek a design that satisfies both of the following requirements.

• (a) A factor of safety against both shaft friction and end bearing in the stable zone of 1·25.

• (b) A factor of safety against shaft friction alone of 0·9 or greater.

If a problem is encountered with respect to satisfying these criteria for truly end-bearing piles, then a settlement analysis can be carried out readily in this case. Finally, it must be borne in mind that the criteria suggested above are still a ‘shortcut' to avoid a more rigorous assessment of the requirements of serviceability.

In the event of any uncertainty regarding the most appropriate shortcut approach, a proper settlement analysis should be undertaken in which both the applied load and the ground settlements are taken into account in an appropriate manner. Several of the references cited in the paper describe various approaches to such an analysis.