1A correlation from load testsTwenty-four driven piles in sand.
Coyle and Castello (1981) analysed 24 large-scale field load tests on piles driven in sand and fitted two charts to them: one for the bearing capacity factor of the point, one for the earth-pressure coefficient of the shaft. Both vary with the embedment ratio L/D as well as with the friction angle φ′.
Unlike Meyerhof’s and Vesic’s point factors, these are not derived from a failure mechanism but fitted to measured capacities, so they already carry whatever limit the field imposed: there is no separate limiting value and no critical depth. Because the tests were on driven piles, the method belongs to driven piles. The GeoTerva sand module does not offer it for a bored pile, and cautions when the pile is jacked.
2The point resistanceNq* against L/D and φ′.
The point resistance takes the usual form, with the effective vertical stress q′ at the tip and the area Ap of the section:
The factor is much smaller than Meyerhof’s, and it does not rise steadily with the embedment: at φ′ = 34° it peaks at 44.9 near L/D = 20 and falls to 20.1 at L/D = 70. A longer pile of the same diameter therefore earns a smaller factor on a larger q′, which is how the measured point resistance levels off with depth.
| L/D \ φ′ | 30° | 32° | 34° | 36° | 38° | 40° |
|---|---|---|---|---|---|---|
| 0 | 23.6 | 28.8 | 34.5 | 45.5 | 60.1 | 79.5 |
| 5 | 26.6 | 32.3 | 38.7 | 51.1 | 67.5 | 89.2 |
| 10 | 28.7 | 34.9 | 41.8 | 55.2 | 73.0 | 96.4 |
| 15 | 30.1 | 36.6 | 43.9 | 58.0 | 76.3 | 100.8 |
| 20 | 30.8 | 37.5 | 44.9 | 59.5 | 78.2 | 103.3 |
| 25 | 30.5 | 37.1 | 44.5 | 59.1 | 78.0 | 102.7 |
| 30 | 28.9 | 35.2 | 42.2 | 56.4 | 75.1 | 98.5 |
| 35 | 26.7 | 32.6 | 39.0 | 52.4 | 70.2 | 91.8 |
| 40 | 24.5 | 29.9 | 35.8 | 48.1 | 64.4 | 84.5 |
| 45 | 22.2 | 27.1 | 32.4 | 43.6 | 58.4 | 77.0 |
| 50 | 20.0 | 24.4 | 29.2 | 39.4 | 52.6 | 69.9 |
| 55 | 18.0 | 21.9 | 26.2 | 35.4 | 47.2 | 63.3 |
| 60 | 16.1 | 19.6 | 23.5 | 31.8 | 42.3 | 57.2 |
| 65 | 14.4 | 17.5 | 21.0 | 28.4 | 37.7 | 51.5 |
| 70 | 13.8 | 16.8 | 20.1 | 27.2 | 36.1 | 49.4 |
Read on the curves of Coyle and Castello (1981). Along a curve and between the curves the GeoTerva modules interpolate linearly, hold the value below φ′ = 30° and above 40°, and continue the last segment of a curve beyond L/D = 70. Values computed by GeoTerva with the function its modules use.
3The shaft resistanceOne average stress over the whole length.
For the shaft, Coyle and Castello take the average effective vertical stress over the embedded length, a pile–soil friction angle tied to φ′ and a coefficient K from their second chart:
p is the perimeter of the section and σ̄′o the mean of the effective vertical stress from the head to the tip. δ′ = 0.8φ′ is part of the correlation, not a free choice. On the chart each curve of K is a straight line on a logarithmic K axis; the chart is drawn to L/D = 36.1. There is no critical depth in this method: the averaging over the whole length, and a K that falls with L/D, play that part (compare the K–δ method with L′ = 15D).
| L/D \ φ′ | 30° | 31° | 32° | 33° | 34° | 35° | 36° |
|---|---|---|---|---|---|---|---|
| 0 | 1.364 | 1.574 | 1.849 | 2.220 | 2.773 | 3.563 | 4.542 |
| 5 | 1.112 | 1.283 | 1.507 | 1.810 | 2.261 | 2.905 | 3.703 |
| 10 | 0.906 | 1.046 | 1.229 | 1.475 | 1.843 | 2.368 | 3.019 |
| 15 | 0.739 | 0.853 | 1.002 | 1.203 | 1.503 | 1.930 | 2.462 |
| 20 | 0.602 | 0.695 | 0.817 | 0.981 | 1.225 | 1.574 | 2.007 |
| 25 | 0.491 | 0.567 | 0.666 | 0.799 | 0.999 | 1.283 | 1.636 |
| 30 | 0.400 | 0.462 | 0.543 | 0.652 | 0.815 | 1.046 | 1.334 |
| 35 | 0.326 | 0.377 | 0.442 | 0.531 | 0.664 | 0.852 | 1.088 |
| 36.1 | 0.312 | 0.360 | 0.423 | 0.508 | 0.635 | 0.815 | 1.040 |
Read on the curves of Coyle and Castello (1981), straight lines in log K against L/D. Between whole degrees the GeoTerva modules interpolate linearly; beyond L/D = 36.1 they continue the straight line and report that the value is a projection. Values computed by GeoTerva with the function its modules use.
4Worked exampleA driven pile, 0.45 m in diameter, 13.5 m long.
A driven concrete pile of circular section, D = 0.45 m, is embedded L = 13.5 m in a sand with φ′ = 34°, γ = 18 kN/m³ above the water table at 3.0 m depth and γsat = 20 kN/m³ below it. L/D = 30.0, Ap = 0.1590 m², p = πD = 1.414 m. The effective vertical stress is 54.0 kPa at the water table and 161.0 kPa at the tip.
- Point: Table 1 at L/D = 30.0 and φ′ = 34° gives Nq* = 42.2, entered as 42 (a chart reading to whole units, as the module fills it in): qp = 161.0 × 42 = 6762 kPa, Qp = 0.1590 × 6762 = 1075 kN.
- Mean stress on the shaft: σ̄′o = [½ × 54.0 × 3.0 + ½ × (54.0 + 161.0) × 10.5] / 13.5 = 89.6 kPa.
- Shaft: Table 2 gives K = 0.815, entered as 0.81; δ′ = 0.8 × 34° = 27.2°; fav = 0.81 × 89.6 × tan 27.2° = 37.3 kPa; Qs = 37.3 × 1.414 × 13.5 = 712 kN.
- Capacity: Qu = 1075 + 712 = 1787 kN; with a factor of safety of 3, Qall = 596 kN.
Meyerhof’s point for the same pile is 617 kN, and the K–δ shaft is worked out on pile skin friction in sand. The spread between the methods is the honest measure of how well the capacity of a pile in sand can be known before a load test.
5QuestionsShort answers.
Can the Coyle and Castello method be used for bored piles?
No. Their correlation was fitted to load tests on driven piles, and boring loosens the sand instead of densifying it. The GeoTerva sand module refuses the method for a bored pile; Meyerhof or Vesic for the point and the K–δ method for the shaft are used instead.
Why does Nq* fall for long piles?
Because it is fitted to measured capacities, and measured point resistance stops growing past a certain depth while q′ keeps growing. At φ′ = 34° the chart peaks at Nq* = 44.9 near L/D = 20 and falls to 20.1 at L/D = 70.
Is there a critical depth in the Coyle and Castello shaft method?
Not explicitly. The shaft resistance uses the average effective stress over the whole length, and K falls with L/D (at φ′ = 34° from 2.77 at L/D = 0 to 0.64 at L/D = 36.1), which plays the part the critical depth L′ = 15D plays in the K–δ method.