Consolidation Settlement

Settlement is the clay's voids shrinking under a new effective stress.

A distribution center is planned on a site with 20 ft of soft clay underneath, and the grading plan calls for 3 ft of fill. The settlement estimate comes back at 4.8 in. Then the plan changes to 6 ft of fill. Twice the fill, so 9.6 in?

The new estimate is 8.5 in. The second 3 ft weighs exactly as much as the first, yet it squeezes the clay only about three-quarters as much. By the end of the first section you'll see where both numbers come from, and why the second 3 ft gets the smaller share.

A slope times a count of log cycles

Compressibility and Preconsolidation gave you the two pieces. A clay layer shortens by the void ratio it loses, spread over its height:

ΔH=Δe1+e0 H\Delta H = \frac{\Delta e}{1 + e_0}\,H

And on the virgin line, the void ratio drops by CCC_C for every log cycle of effective stress. Read that definition backward, and Δe\Delta e is the slope times the number of log cycles the stress travels:

Δe=CClog⁡σf′σ0′\Delta e = C_C \log\frac{\sigma'_f}{\sigma'_0}

Here σ0′\sigma'_0 is the effective stress in the clay now, and σf′\sigma'_f is the effective stress after the load, once the excess pore pressure has drained away. That makes this the ultimate settlement, where the clay ends up; how long it takes to get there is the job of Time Rate of Consolidation. Put the two pieces together, and you have the settlement of a normally consolidated clay:

ΔH=CCH1+e0log⁡σf′σ0′\Delta H = \frac{C_C H}{1 + e_0}\log\frac{\sigma'_f}{\sigma'_0}

The log counts how far the stress moves, CCC_C says how much void ratio each log cycle costs, and H/(1+e0)H/(1 + e_0) turns that change in void ratio into a change in height. The Handbook prints this equation just after its Soil Consolidation Curve Over Consolidated Clay figure, writing HoH_o and eoe_o for HH and e0e_0, as the case where σ0′\sigma'_0 and σf′\sigma'_f are both at or above σc′\sigma'_c: a clay already on its virgin line. The logs are base 10.

For a widespread fill, σf′=σ0′+Δσ\sigma'_f = \sigma'_0 + \Delta\sigma, with Δσ=γfillHfill\Delta\sigma = \gamma_{fill} H_{fill} as in Changing Effective Stress. The log wants the ratio of the two whole stresses, final over initial, never Δσ\Delta\sigma by itself.

Now the Hook. Its clay is normally consolidated, with CC=0.30C_C = 0.30, e0=1.00e_0 = 1.00, and σ0′=1,000\sigma'_0 = 1{,}000 psf at its middle, and the fill weighs 120 lb/ft³. Three feet of fill adds 360 psf:

ΔH=0.30(20)2.00log⁡1,3601,000=3.0(0.134)=0.40 ft=4.8 in\Delta H = \frac{0.30(20)}{2.00}\log\frac{1{,}360}{1{,}000} = 3.0(0.134) = 0.40\ \text{ft} = 4.8\ \text{in}

Six feet adds 720 psf, and log⁡(1,720/1,000)=0.236\log(1{,}720/1{,}000) = 0.236, so ΔH=3.0(0.236)=0.71\Delta H = 3.0(0.236) = 0.71 ft, or 8.5 in. With HH in feet, ΔH\Delta H comes out in feet, so multiply by 12 when the question asks for inches.

1,000 1,360 1,720 0.92 0.96 1.00 Effective stress, σ′ (psf, log scale) Void ratio, e First 3 ft +360 psf Next 3 ft +360 psf Δe = 0.040 Δe = 0.031
Two equal helpings of fill on a log axis
Both steps add 360 psf, but the second covers less distance on the log axis, so it squeezes out less void ratio.

Why does the second helping do less? The clay judges each new load against what it already carries. The first 360 psf raises σ′\sigma' by 36%; the second lands on 1,360 psf and raises it by only 26%. On a log axis, equal ratios make equal steps, so the second step is shorter, and so is its drop in void ratio. On a normally consolidated clay, doubling the load always less than doubles the settlement.

Checkpoint

A 5 m layer of normally consolidated clay has e0=1.20e_0 = 1.20 and CC=0.45C_C = 0.45. At its middle, σ0′=65\sigma'_0 = 65 kPa. A widespread fill adds 50 kPa. What is the layer's ultimate settlement?

mm

From the boring log to the settlement

Often a question doesn't hand you σ0′\sigma'_0. It gives a soil profile and a fill instead, so find σ′\sigma' first: the effective stress at the middle of the clay, built just as in Effective Stress.

Why the middle? σ0′\sigma'_0 grows with depth while the fill adds the same Δσ\Delta\sigma everywhere, so every slice of the clay has its own ratio σf′/σ0′\sigma'_f/\sigma'_0. Near the top, where σ0′\sigma'_0 is small, the ratio is large and the slices squeeze a lot; near the bottom they squeeze less. The equation uses one ratio for the whole layer, and the slice at mid-layer stands in for all of them. The Handbook doesn't say which depth to use, but one mid-layer calculation is the standard practice and what exam questions expect.

The fill raises σ\sigma, and in the long run the water table is where it was, so uu at mid-layer is unchanged and σ′\sigma' gains the full Δσ\Delta\sigma. Keep both stresses effective. CCC_C was measured against effective stress, and a total stress in the log shrinks the ratio and the settlement with it.

Worked Example From the profile to the settlement
1 of 2

A site has 8 ft of sand over a 16 ft layer of normally consolidated clay, which rests on rock. The water table is 3 ft below the ground surface. The sand weighs 110 lb/ft³ above the water table and 122 lb/ft³ below it. The clay has γsat=115\gamma_{sat} = 115 lb/ft³, e0=1.10e_0 = 1.10, and CC=0.38C_C = 0.38. A 5 ft fill with γ=120\gamma = 120 lb/ft³ is spread over the whole site. Find the clay's ultimate settlement in inches.

A Fill, γ = 120 lb/ft³ Original ground Sand, γ = 110 lb/ft³ γsat = 122 lb/ft³ Clay, γsat = 115 lb/ft³ e0 = 1.10, CC = 0.38 Rock 5 ft 3 ft 5 ft 8 ft 8 ft
The site, with the new fill on top

When the clay remembers more

Everything so far assumed a normally consolidated clay, already on its virgin line. An overconsolidated clay starts on the flat recompression branch, and the whole calculation turns on whether the load carries it off that branch. So before you write any equation, compare σf′\sigma'_f with σc′\sigma'_c.

Normally consolidated σ′0 = σ′c Void ratio, e σ′ (log scale) σ′0 = σ′c σ′f Δe CC Overconsolidated, stays below σ′c σ′f ≤ σ′c Void ratio, e σ′ (log scale) σ′0 σ′f σ′c Δe CR Overconsolidated, loaded past σ′c σ′0 < σ′c < σ′f Void ratio, e σ′ (log scale) σ′0 σ′c σ′f Δe CR CC
One load, three clays
All three clays start at the same σ0′\sigma'_0 and end at the same σf′\sigma'_f; only σc′\sigma'_c differs. The highlighted path is the part of each curve the load travels.

On the left, a normally consolidated clay travels the steep virgin line the whole way: the equation above. In the middle, the clay's σc′\sigma'_c lies beyond σf′\sigma'_f, so the load never leaves the flat branch. Its Δe\Delta e is the flat slope times the same log cycles, and the equation simply trades CCC_C for CRC_R:

ΔH=CRH1+e0log⁡σf′σ0′(σf′≤σc′)\Delta H = \frac{C_R H}{1 + e_0}\log\frac{\sigma'_f}{\sigma'_0} \qquad (\sigma'_f \le \sigma'_c)

With CRC_R about a sixth of CCC_C, that's about a sixth of the settlement, which is how the second warehouse in Compressibility and Preconsolidation got away with less than 2 in.

Checkpoint

A 6 m clay layer has e0=0.85e_0 = 0.85, CC=0.30C_C = 0.30, and CR=0.05C_R = 0.05. At its middle, σ0′=70\sigma'_0 = 70 kPa and σc′=160\sigma'_c = 160 kPa. A widespread fill adds 60 kPa. What is the layer's ultimate settlement?

mm

On the right, the load pushes past σc′\sigma'_c, and the path turns a corner. It runs along the flat branch from σ0′\sigma'_0 to σc′\sigma'_c, then down the virgin line from σc′\sigma'_c to σf′\sigma'_f. Each leg is its own slope times its own log cycles:

ΔH=H1+e0[CRlog⁡σc′σ0′+CClog⁡σf′σc′](σ0′<σc′<σf′)\Delta H = \frac{H}{1 + e_0}\left[C_R\log\frac{\sigma'_c}{\sigma'_0} + C_C\log\frac{\sigma'_f}{\sigma'_c}\right] \qquad (\sigma'_0 < \sigma'_c < \sigma'_f)

The virgin leg starts at σc′\sigma'_c, so its ratio is σf′/σc′\sigma'_f/\sigma'_c. Take care when you open the Handbook here: it prints this case with CClog⁡(σf′/σ0′)C_C\log(\sigma'_f/\sigma'_0) as the second term. Read literally, that charges the stretch from σ0′\sigma'_0 to σc′\sigma'_c twice, once at each slope, and overstates the settlement. Expect a question on this case to state the form it wants; the one that matches the curve is σf′/σc′\sigma'_f/\sigma'_c.

When a question gives only the liquid limit, the indexes come from the Handbook's correlations, CC=0.009(LL−10)C_C = 0.009(LL - 10) and CR=CC/6C_R = C_C/6, and drop straight into these equations.

Worked Example A clay loaded past its memory
1 of 2

A 12 ft clay layer has e0=0.80e_0 = 0.80 and a liquid limit of 45%. At its middle, σ0′=1,100\sigma'_0 = 1{,}100 psf and σc′=1,500\sigma'_c = 1{,}500 psf. A widespread fill adds 900 psf. Estimate CCC_C from the liquid limit, take CR=CC/6C_R = C_C/6, and find the layer's ultimate settlement in inches.

Read the equation before you compute

Choosing the expression comes down to where σc′\sigma'_c sits against the path. All three cases share ΔH=Δe H/(1+e0)\Delta H = \Delta e\,H/(1 + e_0) and differ only in Δe\Delta e:

Δe\Delta e for each case
ClayConditionΔe\Delta e
Normally consolidatedσ0′=σc′\sigma'_0 = \sigma'_cCClog⁡(σf′/σ0′)C_C\log(\sigma'_f/\sigma'_0)
Stays below σc′\sigma'_cσf′≤σc′\sigma'_f \le \sigma'_cCRlog⁡(σf′/σ0′)C_R\log(\sigma'_f/\sigma'_0)
Loaded past σc′\sigma'_cσ0′<σc′<σf′\sigma'_0 < \sigma'_c < \sigma'_fCRlog⁡(σc′/σ0′)+CClog⁡(σf′/σc′)C_R\log(\sigma'_c/\sigma'_0) + C_C\log(\sigma'_f/\sigma'_c)

Adapted from the FE Reference Handbook, Soil Consolidation Curve Over Consolidated Clay, with the third case's virgin term written as σ′f/σ′c

An overconsolidated clay can respond to extra load in a way that surprises you. Double the fill on the 6 m layer from the earlier Checkpoint, so it adds 120 kPa, and σf′=190\sigma'_f = 190 kPa passes σc′=160\sigma'_c = 160 kPa. The settlement triples, from 43.6 mm to about 131 mm, because the extra load reached the steep line.

Checkpoint
1 of 3

At the middle of a clay layer, σ0′=1,400\sigma'_0 = 1{,}400 psf, and a consolidation test gives σc′=2,100\sigma'_c = 2{,}100 psf. A widespread fill will raise σ′\sigma' there to 2,600 psf. Which expression gives the clay's Δe\Delta e?

A.
CClog⁡(2,600/1,400)C_C\log(2{,}600/1{,}400)
B.
CRlog⁡(2,600/1,400)C_R\log(2{,}600/1{,}400)
C.
CRlog⁡(2,100/1,400)+CClog⁡(2,600/1,400)C_R\log(2{,}100/1{,}400) + C_C\log(2{,}600/1{,}400)
D.
CRlog⁡(2,100/1,400)+CClog⁡(2,600/2,100)C_R\log(2{,}100/1{,}400) + C_C\log(2{,}600/2{,}100)

Recap

  • Settlement is Δe\Delta e spread over the layer, ΔH=Δe H/(1+e0)\Delta H = \Delta e\,H/(1 + e_0), and Δe\Delta e is a slope times the log cycles the effective stress travels.
  • Normally consolidated: ΔH=CCH1+e0log⁡(σf′/σ0′)\Delta H = \frac{C_C H}{1 + e_0}\log(\sigma'_f/\sigma'_0), with σ0′\sigma'_0 at mid-layer from the profile and σf′=σ0′+Δσ\sigma'_f = \sigma'_0 + \Delta\sigma, both effective.
  • Compare σf′\sigma'_f with σc′\sigma'_c first. Below it, use CRC_R the whole way; past it, CRC_R up to σc′\sigma'_c and CCC_C from σc′\sigma'_c to σf′\sigma'_f.
  • Settlement scales with the index and HH, but the load sits in a log, so doubling a fill on normally consolidated clay less than doubles the settlement.
  • Taking σ0′\sigma'_0 at the top of the layer overpredicts; the Handbook's σf′/σ0′\sigma'_f/\sigma'_0 in the two-leg case's virgin term should be σf′/σc′\sigma'_f/\sigma'_c.