GeoTerva Open the module

Formulas and tables · Pile foundations

α method for piles in clay: adhesion factor and skin friction

The unit skin friction of a pile in clay as a fraction of the undrained shear strength: the adhesion factor α for driven and for bored piles at every strength from 10 to 280 kPa, with a worked example.

  • Free to read
  • cu = 10–280 kPa
  • Driven and bored
  • Worked example

1The α methodTotal stress, short term.

A pile in saturated clay is loaded far faster than the clay can drain, so its shaft resistance is usually worked out in total stress, from the undrained shear strength cu. The α method, introduced by Tomlinson (1957) from load tests on piles driven into clay, writes the unit skin friction at any depth as a fraction of that strength:

f=αcu

α is the adhesion factor. It is not a soil property but a number back-figured from load tests, and it stands for everything that makes the clay against the shaft weaker than the clay the sample came from: remoulding during installation, the gap a driven pile leaves near the surface, the softening of the wall of a bored hole. The shaft resistance is summed layer by layer over the perimeter p of the pile:

Qs=∑ipαicu,iΔLi

The ultimate capacity adds the point resistance, Qu = Qp + Qs (see pile end bearing in clay). The α method needs nothing but cu: the unit weights and the water table do not enter it, unlike the λ method and the β method.

2α for a driven pileA curve against cu/pa.

For a driven pile α falls steadily as the clay gets stronger. The values used here are interpolated values after Terzaghi, Peck and Mesri (1996), given against the strength divided by the atmospheric pressure pa ≈ 100 kPa: α = 1.00 up to cu/pa = 0.1, 0.48 at cu/pa = 1 and 0.34 from cu/pa = 2.4 upwards. Between the tabulated points α is read by straight-line interpolation.

α=α(cupa)

The scatter behind any such curve is wide, because α depends on the effective stress as well as on cu. The API recommendations for offshore piles (API RP 2A) therefore write it in terms of the ratio ψ = cu/σ′v:

α=0.5ψ−0.5for ψ≤1;α=0.5ψ−0.25for ψ>1,ψ=cuσv′,α≤1

Neither form is exact, and the two do not give the same α for the same clay. The table below, and the GeoTerva pile modules, use the curve against cu/pa, which needs no effective stress.

3α for a bored pileThe drilled-shaft rule of O’Neill and Reese.

A bored pile displaces nothing: the hole is drilled, its wall softens as it takes up water, and the concrete is cast against it. O’Neill and Reese (1999) give a rule of its own for such a drilled shaft:

α=0.55for cupa≤1.5;α=0.55−0.1(cupa−1.5)for 1.5<cupa≤2.5

Above cu/pa = 2.5 the rule is held at 0.45. The same procedure credits no friction over two lengths of the shaft: the top 1.5 m, which is disturbed by the work and may shrink away from the concrete, and the bottom one diameter, over which the base is already mobilising the soil:

f=0for z<1.5 mand for z>L−D

The two rules cross near cu = 78 kPa: in softer clay the driven-pile curve gives the larger α, in stiffer clay the drilled-shaft rule does. They come from different sets of load tests, and each belongs with its own procedure: the drilled-shaft α with its two excluded lengths, the driven-pile α over the whole shaft.

4Table of α and fcu = 10–280 kPa.

Table 1. Adhesion factor α and unit skin friction f = αcu (pa = 100 kPa)
cu (kPa)cu/paα drivenf driven (kPa)α boredf bored (kPa)
100.101.0010.00.555.5
150.150.9614.40.558.3
200.200.9218.40.5511.0
250.250.8721.80.5513.8
300.300.8224.60.5516.5
400.400.7429.60.5522.0
500.500.6834.00.5527.5
600.600.6237.20.5533.0
700.700.5840.60.5538.5
800.800.5443.20.5544.0
900.900.5145.90.5549.5
1001.000.4848.00.5555.0
1201.200.4250.40.5566.0
1401.400.4056.00.5577.0
1501.500.3958.50.5582.5
1601.600.3860.80.5486.4
1801.800.3664.80.5293.6
2002.000.3570.00.50100.0
2202.200.3475.90.48105.6
2402.400.3481.60.46110.4
2502.500.3485.00.45112.5
2802.800.3495.20.45126.0

Driven: interpolated values after Terzaghi, Peck and Mesri (1996). Bored: O’Neill and Reese (1999); over the top 1.5 m and the bottom diameter of a bored pile f = 0. Values computed by GeoTerva with the functions its pile modules use.

5Worked exampleA driven and a bored pile in the same clay.

A circular pile D = 0.6 m in diameter is embedded L = 18 m in three clay layers: 0–4 m with cu = 30 kPa, 4–12 m with cu = 60 kPa and 12–18 m with cu = 110 kPa. The perimeter is p = πD = 1.885 m. The pile is first driven and then, for comparison, bored.

Table 2. Shaft resistance layer by layer, driven and bored
Layercu (kPa)α drivenf driven (kPa)Qs driven (kN)α boredLength credited, bored (m)Qs bored (kN)
0–4 m300.8224.6185.50.552.577.8
4–12 m600.6237.2561.00.558.0497.6
12–18 m1100.4549.5559.80.555.4615.8
Total130615.91191
  • Driven: α is read from Table 1 for each layer, and f = αcu acts over the full thickness: Qs = 1.885 × (0.82 × 30 × 4 + 0.62 × 60 × 8 + 0.45 × 110 × 6) = 1306 kN.
  • Bored: every layer has cu/pa ≤ 1.5, so α = 0.55 throughout, and friction is credited from 1.5 m to L − D = 17.4 m only: Qs = 1.885 × 0.55 × (30 × 2.5 + 60 × 8.0 + 110 × 5.4) = 1191 kN, 91 % of the driven pile’s.

The point resistance is added to each (see pile end bearing in clay), and the allowable load follows from one factor of safety on the sum. The pile capacity in clay module of GeoTerva does this layer by layer, for a driven or a bored pile, and draws the friction down the shaft.

6QuestionsShort answers.

What is the adhesion factor α?

The ratio of the unit skin friction on a pile shaft to the undrained shear strength of the clay, f/cu, back-figured from pile load tests. For a driven pile it falls from 1.00 in very soft clay to 0.34 in stiff clay.

Is α the same for a driven and a bored pile?

No. A driven pile takes α from a curve against cu/pa; a bored pile (drilled shaft) takes 0.55 up to cu/pa = 1.5, falling to 0.45 at 2.5 (O’Neill and Reese, 1999), and no friction over the top 1.5 m and the bottom diameter.

When is the α method not the right one?

It is a total-stress method for the short-term, undrained condition, which is usually the critical one for a pile in saturated clay. For the long-term, drained shaft resistance the β method works from the effective stress, and for long driven piles the λ method is an alternative fitted to whole piles.

The theory behind it