Resistance of piles to compressive loads 187Method Mean Qc/Qm Standard deviation(s) Coefficient of variation
ICP, all piles 0.99 0.28 0.28
ICP, all open-end piles 1.05 0.30 0.28
API RP2A(1993), all piles 0.87 0.58 0.60Note
Qc/Qmdenotes calculated/measured.Imperial College have assessed the reliability of their method for piles in sands by
comparing the predictions of shaft capacity with those of the American Petroleum Institute
method(3.5)as shown below:
White and Bolton(4.35)re-analysed the IC data base for closed-end piles on the basis that
instead of the criterion of failure being the load causing a settlement of one-tenth of the diam-
eter they assumed that plunging settlement occurred, i.e. beyond the point Din Figure 4.1. They
also made allowance for only partial embedment of some piles into the bearing stratum, and the
presence in some piles of a weaker layer below base level. They found a mean of qb= 0.9qwith
no trend towards a reduction of qbwith increase in pile diameter. They suggested that a reduc-
tion factor to obtain the ultimate bearing capacity of a closed-end pile in sand should be linked
to partial embedment and partial mobilization rather than to absolute diameter. This suggestion
would appear to be part of the methodology of research based on analysis of test pile failures
rather than criteria to be adopted at the design stage of piled foundations. White and Bolton
noted the dearth of high-quality pile load test data in the public domain.
It was generally assumed in past years that no allowance should be made for significant
changes in the bearing capacity of piles driven into coarse soils with time after installation.
Neither increases nor decreases in capacity were considered although the ‘set-up’or tempo-
rary increase in driving resistance about 24 hours after driving was well known. The long-
term effects had not been given serious study. However, the research work at Imperial
College described in the previous section did include some long-term tension tests on piles
at Dunkirk(4.30). Six 465 mm OD 19 m long and one 465 mm OD 10 m long steel tube
piles were tested in tension at ages between 10 days and about 6 months. A progressive
increase in resistance of about 150% was recorded. All the tests were ‘first-time’, that is,
none of the piles were tested a second time.
The 762 mm OD 44 m long test pile at the Jamuna Bridge site was referred to in the
previous section(4.34). There was an increase in tension capacity of about 270% on retest after
the initial test made a few days after driving into medium-dense silty micaceous sand.
Precast concrete piles on the same site showed a progressive increase of about 200% in
compression at various ages up to 80 days after driving. The ultimate resistances were
estimated from dynamic tests and graphical analysis of loading tests not taken to failure.
Jardine et al.(4.30)attributed the increased tension capacity at Dunkirk mainly to relaxation
through creep of circumferential arching around the pile shaft leading to increase in radial
effective stress.
The procedure for determining the resistance of piles driven into sand using CPT values is
wholly empirical and was originally based on uninstrumented loading tests and experience. The
tests were mainly made on piles of small to medium diameter. EC7 rules do not recommend
any particular method of relating qcto base or shaft resistance but state that the method adopted
should have been established from pile loading tests and from comparable experience involving