Module 2 · 13 min read · Free chapter
Stress below foundations and site investigation
Footing load spreads and fades with depth, so soil is explored down to where the added stress is still about 10% of the footing pressure. Boussinesq gives stress in an elastic half-space; the 2:1 method is a quick estimate. SPT N, CPT qc and fs, and samples with low area ratio describe the ground.
Start from zero: what lies under a foundation
Stand on soft sand in bare feet and you sink. Put a wide plank under your feet and you stay on top. The plank does not make you lighter; it spreads your weight over more sand. Every foundation works on the same idea.
A foundation is the part of a structure that passes its load safely into the ground. To design one, an engineer must answer two questions:
- how does the load spread through the soil below, so that we know which layers feel it, and
- what soil lies there, and how strong and compressible it is.
The first question is answered by stress distribution theory. The second is answered by site investigation: boring holes, taking samples, and testing the soil in the field and the laboratory. This chapter covers both, because each one tells the other how deep to look.
Why it matters. Most foundation failures do not come from wrong formulas. They come from soil that nobody looked at, such as a soft clay layer 6 m down that no borehole reached.
Stress below a point load: Boussinesq
In 1885, Boussinesq solved a clean mathematical problem. He took a single point load Q on the surface of a material that is elastic, homogeneous, isotropic and extends without limit downwards and sideways, called a half-space. He found the extra vertical stress at any point at depth z and horizontal distance r from the load:
σz = (3Q / 2πz²) × [1 / (1 + (r/z)²)]^(5/2)
Directly below the load, r = 0, and the bracket becomes 1. The formula then reduces to σz = 0.4775 Q/z². So the stress under a point load falls with the square of the depth: at twice the depth, it is one quarter.
Real soil is neither perfectly elastic nor uniform, so why do engineers still use this? Because the extra stress depends mostly on geometry, not on the soil's stiffness. Boussinesq's answer does not even contain the elastic modulus. It is a good, slightly safe estimate for most soils, and it can be added up over many small loads to give the stress under a whole footing.
Westergaard later solved the case of a soil reinforced by many thin, stiff horizontal layers that stop sideways strain, as in some varved clays. His solution gives smaller stresses than Boussinesq's under the load.
Why it matters. Settlement of a clay layer depends on how much extra stress reaches it. Boussinesq tells you that a layer 10 m down feels far less than one 3 m down, which decides whether a footing is enough or piles are needed.
Quick methods: the 2:1 rule and pressure bulbs
For hand calculations, engineers often use a simpler picture. They imagine the load spreading out at a slope of 2 vertical to 1 horizontal on every side. At depth z, a footing of size B × L has spread its load over (B + z) × (L + z), so:
σz = Q / [(B + z)(L + z)]
For a strip footing of width B carrying q per metre, this becomes σz = qB / (B + z). The method assumes the stress is uniform across the spread area, which it is not, but it gives a quick average that is good enough for a first check.
A pressure bulb, or isobar, is a line joining all points below a footing that carry the same extra stress, such as 0.5q or 0.1q. The bulbs look like onions hanging from the footing. Engineers treat the zone inside the 0.1q bulb as the soil that matters; for a square footing it reaches roughly 2B deep, and for a strip footing deeper still.
This gives a simple rule for exploring a site. The borehole should reach at least the depth where the added stress drops to about 10% of the footing pressure, which is usually 2B to 3B below the footing. If a soft, compressible layer is still there at that depth, the borehole must go on until it ends, because that layer can still settle.
Why it matters. Two footings with the same pressure but different widths behave differently. A large raft on a site with a deep soft layer can settle a lot even when small test footings did not, because its bulb reaches the soft layer.
Planning a site investigation
A site investigation sets out to find three things: the layers of soil and rock and their thickness, the ground water level, and the engineering properties of each layer. It moves from cheap and broad to costly and precise.
It begins with a reconnaissance, a visit by an experienced engineer without instruments. The engineer looks at the terrain and access for drilling rigs, exposed soil in road cuts, stream banks and pits, seepage from cut faces, signs of slope movement, and flood marks on banks. Cracks in nearby buildings and pavements are a strong clue; small houses that crack every year often sit on expansive clay that swells when wet and shrinks when dry.
A desk study runs alongside: geological and topographic maps, air photos, earlier reports and published case histories, and local knowledge. Local well drillers know at what depth water and hard strata are found. Municipal offices know which foundation types have worked in the area.
Only then comes sub-surface exploration, which has three steps: boring a hole, sampling the soil as the hole advances, and testing the samples and the ground. The specification for the drilling contractor must fix the number, location, spacing and depth of boreholes, the samplers to use and the sampling interval.
Typical spacing used in practice:
| Project | Borehole spacing |
|---|---|
| Normal building on level ground | about 15 m both ways, plus corners and heavily loaded points |
| Preliminary study for site development | 50 to 150 m |
| Cuts along a canal or road | about 30 m |
| Earth dam | about 60 m |
| Road | about 100 m |
Boring methods and samples
Different ground needs different tools to advance a hole:
- Hand auger: a cutting blade on a handle, turned by hand. It suits small jobs to about 5 or 6 m.
- Truck-mounted continuous-flight auger: a powered helical auger that brings cuttings up as it turns. It works well above the water table in sand, silt and clay, but the hole may collapse in soft soil or below water.
- Wash boring: a chopping bit and a jet of water break the soil, and the mud rises to a settling tub. It is quick in soft soil, and the driller watches the colour of the returning mud to see where the layers change. Casing is driven first to stop the sides falling in.
- Rotary drilling: a powered cutting bit or core barrel with circulating fluid, used for rock and hard strata.
- Percussion drilling: repeated blows, used in gravel and soils with boulders, with casing to keep the hole open.
A disturbed sample keeps the soil's grains but not its structure. Auger cuttings and split-spoon samples are of this kind, and they serve to identify the soil and run index tests such as grain size and Atterberg limits. An undisturbed sample keeps the structure and water content close to their state in the ground. It is needed for strength and consolidation tests, especially in clay.
How much a sampler disturbs the soil depends on how thick its wall is compared to the hole it cuts. This is measured by the area ratio:
Ar = (Do² − Di²) / Di² × 100
Here Do is the outside and Di the inside diameter of the tube. For an undisturbed sampler, Ar should be 10% or less; a thick split spoon has a much larger ratio. The cutting edge is also made slightly smaller than the tube, giving a small inside clearance so the soil core does not rub against the wall as it enters. Thin-walled tubes, piston samplers and hand-cut blocks from a pit give the best undisturbed samples, sealed with wax so they keep their water.
Why it matters. Consolidation and strength tests on a badly disturbed clay sample understate its strength and overstate its compressibility. The design then becomes either unsafe or needlessly costly.
The standard penetration test
The standard penetration test (SPT) is the most widely used field test in India and much of the world. It is simple, needs little skill, works in most soils, and gives a sample along with the result.
At the chosen depth, a split-spoon sampler, about 600 mm long with 50 mm outside and 35 mm inside diameter, is lowered on drill rods. A 65 kg hammer falls freely through 750 mm onto an anvil at the top. The sampler is driven 450 mm in three steps of 150 mm. The first 150 mm is a seating drive through soil disturbed by boring, so it is not counted; the number of blows for the next 300 mm is the N value.
The field N is then corrected. The most important is the correction for overburden: the same sand gives more blows at depth simply because it is more confined. A widely used factor is:
CN = √(100 / σ′v), with σ′v in kPa
So at shallow depth, where σ′v is small, CN is more than 1, and deep down it is less than 1. Other corrections allow for hammer energy, borehole diameter, sampler type and rod length. For very fine or silty sand below the water table, a dilatancy correction is often applied when N exceeds 15: N = 15 + (N − 15)/2, because such sand resists fast driving more than its true density deserves.
The value of SPT lies in its correlations, built over decades by comparing N with lab results:
| Soil | N | What it suggests |
|---|---|---|
| Sand | 0 to 5 | very loose, Dr below about 5%, φ about 26° to 30° |
| Sand | 5 to 10 | loose, Dr about 5% to 30%, φ about 28° to 35° |
| Sand | 10 to 30 | medium, Dr about 30% to 60%, φ about 35° to 42° |
| Sand | 30 to 50 | dense, Dr about 60% to 90% |
| Clay | 2 to 5 | soft, qu about 25 to 50 kPa |
| Clay | 10 to 20 | stiff, qu about 100 to 200 kPa |
| Clay | above 30 | hard, qu above 400 kPa |
N above 50 is usually taken as refusal, often a sign of rock or boulders. In sand, N is used for bearing capacity and settlement; in clay, it is only a rough guide, and lab or vane tests are preferred.
Descriptive lens: how this topic is asked in written papers
Written papers test this chapter in four main ways:
- State and use Boussinesq's equation or the 2:1 method for a footing or a point load.
- Explain a pressure bulb and how it fixes the depth of exploration.
- Plan a site investigation for a given structure: number, spacing and depth of boreholes, samples and tests.
- Describe the SPT with its corrections, or compare SPT, CPT and geophysical methods.
Practice questions (our own, not previous-year questions).
- A point load of 500 kN acts on the ground surface. Find the vertical stress at 4 m depth directly below it and at a radial distance of 2 m. State the assumptions behind the formula used. (10 marks)
- What is a pressure bulb? Explain why a wide raft and a small footing at the same bearing pressure need boreholes to different depths. (10 marks)
- Plan the site investigation for a 10-storey building on a 60 m × 40 m plot in an alluvial area with a high water table. Give the number, layout and depth of boreholes, the samples to be taken and the field and laboratory tests. (15 marks)
Answer plan for question 3.
- Begin with reconnaissance and desk study: old boring logs in the area, nearby building performance, flood and water-table data.
- Lay out boreholes at about 15 m centres, which gives a grid of about 5 × 4, plus extra holes at lift cores and heavily loaded columns.
- Fix the depth from the pressure bulb of a likely raft or pile group, about 1.5 to 2 times the building width or 3 to 5 m below the expected pile tips, and deeper if soft clay continues.
- Take SPT every 1.5 m and at each change of layer; thin-walled tube samples in clay with area ratio within 10%; record the water table.
- List lab tests (index, triaxial, consolidation, chemical tests of water and soil) and add CPT between boreholes to catch thin soft layers.
- Conclude with the deliverable: a soil profile, design parameters and a recommendation on foundation type.
Quick revision
- Footing load spreads with depth; explore the soil it still stresses.
- Boussinesq: σz = (3Q / 2πz²)[1 / (1 + (r/z)²)]^(5/2); below the load, σz = 0.4775 Q/z².
- 2:1 method: σz = Q / [(B + z)(L + z)].
- The 0.1q pressure bulb sets the depth of significant stress, usually 2B to 3B.
- Investigation: reconnaissance and desk study, then boring, sampling and testing.
- Spacing about 15 m for buildings; depth to the 10% stress level, 3 m into rock, 3 to 5 m below pile tips.
- Area ratio within 10% for undisturbed samples, which are needed for strength and consolidation tests.
- SPT: 65 kg hammer, 750 mm fall, blows for the last 300 mm of 450 mm; above 50 is refusal.
- Overburden correction CN = √(100/σ′v); dilatancy correction N = 15 + (N − 15)/2 for fine sand below water.
- CPT gives a continuous qc and fs record; seismic refraction velocity = 1/slope.