Foundation Engineering / Pile groups, settlement and negative skin friction
Settlement and downdrag
Two problems that are both about movement, not strength.
Lesson review: expert verification remains open
This lesson has received a full-context editorial review and a check of its technical explanations. Selected arithmetic has been checked separately. This is not independent engineering sign-off.
Corrections checked in this pass
- Corrected the pile-plan scale and removed an unsupported total-settlement ratio. The available movement components are labelled separately.
Checks needed for project use
- A geotechnical reviewer must confirm drainage and stress basis, bearing-factor formulation, correlations and settlement influence factors. Partial factors and project-specific ground conditions are outside these teaching checks.
- Equivalent-raft depth/spread are selected empirical assumptions; embedded layered-ground movement and group load sharing need verification. Single-pile shaft-induced soil displacement is omitted, so no complete total or ratio is available.
These are educational models and exercises. Check their assumptions before using the results. The full technical content and every diagram have not yet been independently verified.
Concrete and steel calculations retain their stated first-generation Eurocode teaching basis. Use the standards and amendments specified for your project; do not mix generations without an appropriate design basis.
Read the lesson register and outstanding review notesWhy this matters
Module 12 said most piled foundations exist because of settlement rather than capacity. This is where that is paid for — and where the most-missed load case in pile design lives.
A group settles far more than one pile. A single pile stresses a modest volume of soil around and below it. A group stresses a volume set by the group's plan dimensions, extending well below the pile toes — the same width-and-influence-depth argument as Module 11's raft.
The standard device is the equivalent raft: replace the group with a raft at two-thirds of the embedded length, carrying the whole load over the group's plan area, and compute its settlement by ordinary shallow-foundation methods. Load is then spread below it, conventionally at 4:1.
For a 5.4 m square group of 18 m piles carrying 9000 kN, the equivalent raft sits at 12 m depth under 309 kPa. For an illustrative homogeneous elastic half-space with E = 15 MPa and ν = 0.3, applying the rigid-square factor 0.99 gives 100.1 mm. Applying a surface solution at an equivalent-raft depth is an approximation; embedment and layering are not resolved. The earlier 8.7 mm single-pile figure contains only axial shortening and an illustrative base displacement, omitting soil displacement from shaft-load transfer. It is not a total settlement, so no physical group/single-pile settlement ratio can be inferred.
Negative skin friction is the case where shaft friction changes sign.
If the ground around a pile settles more than the pile does — because fill has been placed over compressible soil, because the water table has been lowered, or because a nearby load is consolidating the ground — then the soil hangs off the shaft instead of supporting it. Over that depth, shaft friction acts downwards as an additional load on the pile.
The depth over which this happens is bounded by the neutral plane: the level at which the pile and the soil settle by the same amount. Above it, downdrag; below it, ordinary positive shaft resistance.
For a 0.6 m pile through 6 m of settling fill with γ′ = 10 kN/m³ and β = 0.3, the downdrag is about 102 kN — which for a pile carrying a few hundred kilonewtons is a substantial and entirely additional load.
Check yourself
A pile passes through 6 m of newly placed fill over soft clay, into gravel below. What does the fill do to the pile?
Practice
For the illustrative rigid-square elastic calculation, take q = 9000/(5.4²) kPa, B = 5.4 m, E = 15,000 kPa, ν = 0.3 and influence factor 0.99. What is qB(1−ν²)I/E in millimetres?
Try it
A pile group, three ways
Nine piles are not nine times one pile. Change the layout and see which mechanism governs.
What is fixed
- 0.6 m piles, 12 m long, α = 1.0 in soft clay.
- Converse–Labarre can never exceed 1 — driven groups in loose sand often do, which the formula cannot express.
- Downdrag uses β = 0.3 and γ′ = 10 kN/m³ over the settling depth.
- Piles
- 9
- Group outer dimension
- 4.20 m
- Single pile capacity
- 1007 kN
- Naive sum
- 9059 kN
- Efficiency η
- 0.727
- By efficiency
- 6585 kN
- As a block
- 14414 kN
- Lower illustrative ultimate resistance
- 6585 kN
- Available73% of the naive sum
- Lost to group effects
Against the naive sum of 9059 kN, the block is NOT critical at 14414 kN — as a group grows, its perimeter rises in proportion to its side while the pile count rises as the square, so the block falls behind. Efficiency is a separate reduction, to 6585 kN. The lower illustrative ultimate resistance is 6585 kN from group efficiency. "Always check the block" is right — "the block always wins" is not.
Summary
- Group efficiency falls with closer spacing and larger groups: 0.73 for 3×3 at 3D, 0.53 for 5×5 at 2D
- Check block and individual-pile routes consistently; the given 5×5 efficiency route is lower than its block resistance
- A group settles far more than one pile because it stresses ground well below the toes
- The equivalent raft at two-thirds of the length is a device for getting that depth roughly right
- Negative skin friction adds load, is bounded by the neutral plane, and is cured by managing movement