Geotechnical Engineering: Analytical Assessment of Earth Stresses, Foundation Loads, Bearing Capacity, Seepage, Piles and Embankment Widening

A competent geotechnical engineer does not begin with a software model. The engineer first asks a more fundamental question: “What should the soil actually be doing?”

Numerical software is extremely useful for difficult soil–structure interaction problems, but it should not replace engineering mechanics. Before opening a finite-element package, the designer should be able to estimate the magnitude, direction and distribution of the major stresses and load components using hand calculations.

This article develops that approach for shallow foundations, retaining walls, basement slabs, wind-turbine foundations, deep foundations, seepage, batter piles and embankment widening.

Engineering philosophy: A numerical model should confirm or refine a mechanically reasonable conceptual model. It should not be used to discover what the engineer should already understand from equilibrium, effective stress, compatibility and soil strength.

Contents

1. Estimating reasonable earth-stress levels
2. Rigid versus flexible shallow foundations
3. Eccentric loading and effective bearing area
4. Worked eccentricity example
5. Bearing capacity under factored lateral loads in LRFD
6. Deeply embedded basement slabs and embedment effects
7. Granular backfill between wall and self-supporting slope
8. Vertical and lateral effective stress under upward seepage
9. Why settlement must be calculated using refined sublayers
10. Why driven piles may use undrained analysis
11. Why drilled shafts often use drained analysis
12. Why batter piles remain useful despite lower flexural capacity
13. Additional loading from a new wedge fill on an existing embankment
14. Hand-analysis workflow before numerical modelling
15. Common mistakes
16. Expert checklist
17. References and design guidance

1. Estimating Reasonable Earth-Stress Levels Before Detailed Analysis

The first task of a geotechnical engineer is to establish the approximate stress state in the ground.

For a soil element at depth z, the simplest starting point is the total vertical overburden stress:

σv = γ z

where:

• σv = total vertical stress
• γ = unit weight of soil
• z = depth below ground level

If the groundwater table is present, the engineer must distinguish between total stress, pore-water pressure and effective stress.

σ'v = σv - u

u = γw zw

Therefore:

σ'v = γz - γwzw

For saturated soil below the groundwater table, the effective unit weight is approximately:

γ' = γsat - γw

and the effective vertical stress increases approximately as:

σ'v = σ'v(at water table) + γ' Δz

1.1 Estimating horizontal stress

For an approximately normally consolidated soil under at-rest conditions:

σ'h = K0 σ'v

For normally consolidated soil, a commonly used approximation is:

K0 ≈ 1 - sinφ'

For overconsolidated soil, the value of K0 may be substantially higher and should be established using appropriate correlations or laboratory data.

The corresponding total horizontal stress is:

σh = σ'h + u

1.2 Active and passive conditions

If the retaining structure moves sufficiently away from the backfill, the soil may approach active conditions:

Ka = tan²(45° - φ'/2)

For a horizontal backfill and no cohesion:

σ'h = Ka σ'v

For passive conditions:

Kp = tan²(45° + φ'/2)

Important: Ka, K0 and Kp represent different physical states. They are not interchangeable coefficients. A wall that has not moved sufficiently to mobilize active conditions should not automatically be designed using Ka.

1.3 Load decomposition

For a foundation or structure, a useful preliminary exercise is to divide the total load into:

Load component	Typical soil response
Vertical structural load	Vertical bearing stress, compression and settlement
Self-weight of foundation	Additional bearing pressure and stabilizing moment
Earth pressure	Horizontal shear and overturning moment
Wind/seismic load	Horizontal force and overturning moment
Water pressure	Hydrostatic force and uplift
Seepage force	Body force modifying effective stress
Adjacent surcharge	Additional vertical and horizontal soil stress

The resulting foundation actions are then reduced to:

V = vertical resultant

H = horizontal resultant

Mx, My = overturning moments

From these quantities, the engineer can calculate eccentricity, bearing pressure, sliding demand and bearing-capacity demand.

2. Rigid and Flexible Shallow Foundations: Why Is Load Distribution Different?

The assumption that contact pressure below every footing is uniform is one of the most common simplifications in foundation engineering.

The actual pressure distribution depends on:

• relative stiffness of footing and soil,
• footing geometry,
• soil stiffness profile,
• loading pattern,
• eccentricity,
• foundation embedment,
• construction sequence, and
• soil–structure interaction.

2.1 Flexible footing

A flexible footing bends relatively easily compared with the supporting soil. The soil reaction therefore tends to follow the applied load distribution and foundation deformation.

For a flexible slab under a concentrated column load, the soil pressure can become highly nonuniform.

The simplified relationship is:

q(x,y) ≈ k · w(x,y)

where k is an idealized subgrade reaction modulus and w is foundation deflection.

Important limitation: The Winkler modulus is not a fundamental soil property. It depends on footing dimensions, soil conditions, loading configuration and the manner in which it is back-calculated.

2.2 Rigid footing

A rigid footing tends to maintain a planar deformation shape. If the footing rotates, the soil reaction becomes approximately linear across the contact area under simplified elastic assumptions.

For a rectangular footing:

q = V/A ± Mx/Zx ± My/Zy

For one-direction eccentricity:

qmax = V/A (1 + 6e/B)

qmin = V/A (1 - 6e/B)

provided the entire base remains in compression.

2.3 The physical difference

Rigid foundation	Flexible foundation
Maintains approximate plane geometry	Can bend significantly
Soil pressure redistributes strongly	Pressure follows local deformation/load
Linear pressure distribution is often a useful approximation	Nonlinear distributions are common
Foundation stiffness dominates local compatibility	Soil stiffness strongly influences deformation

Field insight: If a footing is very stiff relative to the soil, assuming uniform pressure under an eccentric load is usually mechanically inconsistent. The resultant pressure must move toward the compressed side to balance the applied moment.

3. Eccentric Load on a Shallow Foundation: Effective Bearing Area

Wind-turbine foundations are a classic example because overturning moments can be very large relative to the vertical load.

The first calculation is eccentricity:

e = M/V

where:

• M = resultant moment at foundation level
• V = vertical compressive load
• e = eccentricity of resultant

3.1 Effective dimensions

For a rectangular footing of dimensions B × L:

B' = B - 2eB

L' = L - 2eL

where:

eB = My/V

eL = Mx/V

The effective bearing area becomes:

A' = B' L'

and the effective bearing pressure is:

qeff = V/A'

This reduced-dimension approach is consistent with the effective-footing concept used in LRFD foundation analysis. FHWA guidance explicitly describes eccentricity-induced reduced footing dimensions for shallow-foundation bearing and overturning evaluation. 1

3.2 The middle-third concept

For one-direction eccentricity, full contact exists when:

e ≤ B/6

At:

e = B/6

the minimum pressure becomes zero.

When:

e > B/6

the resultant lies outside the middle third and tensile contact would be required to maintain a linear pressure distribution. Soil cannot normally provide tension, so the actual contact becomes partial.

3.3 Partial-contact pressure

For one-dimensional eccentricity with a rectangular footing, the compressed contact width is approximately:

B' = 3(B/2 - e)

The triangular pressure distribution then has:

qmax = 2V/(B'L)

with:

qmin = 0

Do not confuse the two methods: The effective-area method and the partial-contact pressure method are related concepts but should not be mixed indiscriminately. Use the method prescribed by the applicable design standard and load combination.

4. Worked Example: Wind-Turbine Foundation Under Increasing Eccentricity

Consider a circular foundation of diameter:

D = 20 m

and vertical design load:

V = 30,000 kN

Assume the resultant moment is varied to illustrate the effect of eccentricity.

Case	e/D	e (m)	Interpretation
A	0.00	0.00	Concentric
B	0.05	1.00	Moderate eccentricity
C	0.10	2.00	Significant eccentricity
D	0.15	3.00	Severe eccentricity

For a circular foundation, a rectangular effective-area formula should not simply be applied without modification. The effective compressed region is a circular segment whose geometry depends on eccentricity.

For preliminary engineering, however, an equivalent rectangular section can sometimes be used to obtain conservative or screening-level estimates, provided the adopted standard permits it.

4.1 Practical wind-turbine procedure

1. Determine vertical service and factored loads.
2. Determine maximum overturning moment.
3. Calculate e = M/V.
4. Determine whether the foundation remains fully compressed.
5. Calculate the effective bearing area according to the selected code methodology.
6. Calculate maximum contact stress.
7. Check ultimate bearing capacity.
8. Check sliding.
9. Check overturning/uplift.
10. Check settlement and rotation.
11. Check cyclic soil response where relevant.

For wind turbines, bearing capacity alone is not enough. Excessive rotation can become a serviceability or operational problem even when classical ultimate bearing capacity is apparently adequate.

5. Bearing Capacity of a Shallow Foundation Under Factored Lateral Loads in LRFD

A horizontal load does not simply become another vertical bearing pressure. It influences the foundation through:

• sliding,
• overturning moment,
• eccentricity,
• foundation-soil contact reduction, and
• possible reduction in bearing capacity because of inclined loading.

5.1 Basic load system

V = factored vertical load

H = factored horizontal load

M = factored overturning moment

The resulting eccentricity is:

e = M/V

Then determine effective footing dimensions.

5.2 Bearing-capacity equation

A general classical bearing-capacity form is:

qult = c'Nc sc dc ic + qNq sq dq iq + 0.5γ'BNγ sγ dγ iγ

where:

• c' = effective cohesion
• q = effective surcharge at foundation level
• B = effective foundation width
• Nc, Nq, Nγ = bearing-capacity factors
• s = shape factors
• d = depth factors
• i = load inclination factors

For an LRFD design, the designer does not simply compare an unfactored applied load with an arbitrary allowable bearing pressure. The factored resistance and factored load must be treated consistently with the selected LRFD specification.

Factored resistance ≥ Factored effect

or in generic LRFD notation:

φRn ≥ Σηi γi Qi

The exact resistance factors, load factors and modifiers must come from the governing design code rather than being invented for a project.

5.3 Sliding

A simplified frictional resistance may be estimated as:

Rsliding = V tanδ

with additional components where permitted, such as passive resistance or adhesion.

However, passive resistance should not be credited automatically if excavation, erosion, future utilities, drainage or construction activities can remove the passive soil.

Critical point: Do not use the full passive resistance of soil in front of a foundation unless that soil can reliably remain in place throughout the design life.

6. Can a Deep Basement Slab Benefit from Significant Embedment?

Yes, but the answer depends on what resistance mechanism is actually being mobilized.

A deeply embedded foundation does not automatically obtain an unlimited increase in bearing capacity merely because it is deep.

6.1 Surcharge at foundation level

For a foundation embedded at depth D, the overburden surcharge is approximately:

q = γD

The bearing-capacity term:

qNq

therefore increases with embedment.

6.2 But basement excavation changes the problem

Suppose a basement is excavated to a substantial depth and the slab is constructed at the bottom. The original overburden has been removed.

The engineer must distinguish between:

• gross bearing capacity,
• net bearing capacity,
• stress relief caused by excavation,
• reloading caused by the structure,
• basement wall restraint,
• slab–soil contact, and
• potential uplift from groundwater.

6.3 Basement slab and uplift

For a basement below groundwater level, uplift may become more important than bearing capacity.

U = γw h A

where h is the hydraulic head difference.

The overall stability must therefore consider:

Downward resisting weight + permitted anchorage/friction ≥ Uplift force

A deeply embedded basement can have a substantial geotechnical advantage, but it can also introduce a substantial groundwater problem. Embedment is not synonymous with bearing-capacity reserve.

7. Granular Backfill Narrowly Confined Between a Wall and a Self-Sustaining Vertical Slope

This is a subtle retaining-wall problem.

Suppose a retaining wall is constructed immediately adjacent to a nearly vertical self-supporting cut or rock face, leaving a narrow space that is filled with granular material.

The question is:

Should classical active earth pressure be used?

Not automatically.

7.1 Why classical Rankine pressure may be inappropriate

Rankine active pressure assumes a soil mass capable of developing the required failure mechanism and wall movement.

If the granular material is confined between two relatively stiff boundaries, the soil may develop significant arching.

The stress state can therefore approach an at-rest or confined condition rather than a freely developing active wedge.

7.2 Possible mechanisms

Condition	Potential pressure state
Wall moves freely away from backfill	Active pressure may develop
Wall restrained	At-rest pressure may be more appropriate
Narrow confined gap	Soil arching may significantly modify pressure
Rigid rock boundary	Load transfer may occur to both boundaries
Compacted fill	Construction-induced lateral stress may be significant

USACE guidance emphasizes that horizontal earth pressure in cohesionless backfill depends strongly on wall movement. Active pressure develops only after sufficient wall movement away from the backfill. 2

7.3 Conservative design approach

For a restrained wall, a practical preliminary calculation is often:

σ'h = K0 σ'v

rather than automatically adopting Ka.

Where the geometry is narrow and arching is important, more sophisticated approaches such as silo/Janssen-type stress transfer or numerical soil–structure interaction analysis may be warranted.

Expert judgment: A narrow soil gap between two rigid surfaces is not simply a miniature retaining-wall backfill problem. It is a confined soil problem.

8. Vertical and Lateral Effective Stress Under Upward Seepage

This is one of the most important effective-stress problems in geotechnical engineering.

The fundamental relationship remains:

σ' = σ - u

But under seepage, pore-water pressure varies differently from the hydrostatic condition.

8.1 Seepage gradient

i = Δh/L

where:

• Δh = hydraulic head difference
• L = seepage path length

The seepage body force per unit volume is approximately:

fs = iγw

8.2 Upward seepage

Under upward seepage, the seepage force acts upward and therefore reduces the effective stress produced by the soil skeleton.

For one-dimensional vertical flow, a simplified effective unit-weight representation gives:

γ' effective = γsat - γw(1 + i)

or:

γ'effective = γ' - iγw

Thus, as i increases, effective stress decreases.

USACE guidance similarly represents upward seepage through an increased effective water body-force term, with the effective soil unit weight reduced accordingly. 3

8.3 Critical hydraulic gradient

The classical critical gradient is approximately:

ic = γ'/γw

For a saturated cohesionless soil:

ic ≈ (Gs - 1)/(1 + e)

At approximately:

i = ic

the effective vertical stress can approach zero.

This is the mechanism associated with quick condition or boiling in susceptible granular soils.

8.4 Lateral effective stress

If the soil remains in an at-rest condition:

σ'h = K0 σ'v

Therefore, a reduction in effective vertical stress caused by upward seepage also affects the effective horizontal stress.

However, one must not mechanically calculate total lateral pressure using only the reduced effective stress. Total stress is:

σh = σ'h + u

The pore-water pressure can be increasing while effective stress is decreasing.

This distinction is crucial: Upward seepage can simultaneously increase pore-water pressure and reduce soil effective stress. Saying “water reduces earth pressure” without specifying total or effective stress is incomplete.

9. Why Settlement Should Be Calculated Using Refined Sublayers

A common shortcut is to calculate the stress increase at the middle of an entire compressible layer and multiply it by the layer thickness.

This can be seriously inaccurate when:

• the compressible layer is thick,
• the foundation is relatively small,
• stress decreases rapidly with depth,
• soil properties vary with depth, or
• groundwater conditions change within the layer.

9.1 Stress influence varies with depth

For a foundation, the vertical stress increment is greatest close to the foundation and decreases with depth.

Δσz = q I(z)

where I(z) is an influence factor.

Therefore, the assumption:

Δσz = constant

through an entire 10 m compressible layer is usually poor engineering.

9.2 Sublayer method

Divide the compressible soil into smaller layers:

S = Σ [mv,i × Δσ'i × Hi]

or, using constrained modulus:

S = Σ [(Δσ'i / M'i) Hi]

where:

• Hi = thickness of sublayer i
• Δσ'i = effective stress increase
• mv,i = coefficient of volume compressibility
• M'i = constrained modulus

9.3 Why the middle-point method can fail

Consider a 12 m clay layer beneath a 2 m wide footing.

If the stress increase at the top is 100 kPa, it may reduce substantially by the bottom of the layer.

Using the stress at 6 m depth for all 12 m assumes:

Δσ(0) = Δσ(6m) = Δσ(12m)

which is physically incorrect.

The refined sublayer approach integrates the stress distribution:

S ≈ ∫ [Δσ'(z)/M'(z)] dz

The numerical summation is simply a practical approximation of this integral.

Expert insight: Settlement calculation is fundamentally an integration problem. The more rapidly stress and soil stiffness vary with depth, the more important sublayer refinement becomes.

10. Why Undrained Analysis Often Governs Driven Piles

The phrase “driven piles are undrained” should not be interpreted as an absolute rule.

It is a statement about the loading/construction timescale and soil response.

10.1 Installation creates rapid disturbance

When a displacement pile is driven into saturated fine-grained soil:

1. the pile displaces soil laterally;
2. the soil experiences rapid shear deformation;
3. excess pore-water pressure is generated;
4. there is insufficient time for drainage during the instantaneous installation process.

Therefore, short-term behavior of cohesive soil is often represented using undrained strength:

τf = su

The pile shaft resistance may be represented approximately using:

Qs = α su As

where α represents an adhesion factor or equivalent empirical parameter.

10.2 But long-term pile behavior may be drained

After installation, excess pore pressures can dissipate.

Consequently, pile capacity can change with time, particularly in cohesive soil.

Therefore:

Stage	Typical condition
Driving	Rapid, essentially undrained response in saturated fine soil
Short-term static loading	May be undrained in cohesive soil
Long-term loading	Effective-stress/drained considerations may become important

FHWA's driven-pile guidance treats pile design as a combination of geotechnical and structural limit states and emphasizes construction effects, static and dynamic testing and installation response. 4

11. Why Drilled Shafts Often Use Drained Analysis

Again, this requires nuance.

FHWA guidance explicitly notes that shafts in cohesive soils can be designed using either total-stress or effective-stress approaches for undrained and drained conditions, respectively, while shafts in cohesionless soils are designed using effective-stress methods for drained loading. 5

11.1 Main distinction from driven piles

A drilled shaft does not displace the surrounding soil in the same manner as a driven displacement pile.

The construction process involves:

1. excavation;
2. removal of soil;
3. possible use of casing or drilling fluid;
4. placement of reinforcement;
5. concreting.

The stress path around the shaft is therefore fundamentally different.

11.2 Cohesionless soils

Granular soils normally have negligible long-term excess pore-pressure storage compared with cohesive soils, so effective-stress analysis is generally appropriate.

τ = σ'n tanδ

or using a β-type approach:

τ = βσ'v

11.3 Cohesive soils

For cohesive soil, both approaches can be relevant:

Undrained: τ ≈ αsu

or:

Drained: τ = σ' tanδ

The appropriate method depends on soil type, construction, loading duration, drainage conditions and the governing design standard.

The correct question is not “Are drilled shafts drained?” The correct question is “What drainage condition corresponds to the governing load stage and soil response?”

12. How Can a Batter Pile Be Used if Its Flexural Lateral Capacity Is Lower?

This is a classic example of confusing structural capacity with foundation-system behavior.

A batter pile is inclined intentionally.

If a horizontal load acts on the pile group, part of the horizontal load can be resisted through axial compression/tension in the inclined pile.

12.1 Simple force decomposition

For a pile inclined at angle θ from vertical:

Horizontal component = P sinθ

Vertical component = P cosθ

Therefore, to resist a horizontal force:

P = H/sinθ

and the vertical component becomes:

V = H cotθ

12.2 Why this can be advantageous

A vertical pile must resist horizontal load largely through:

• soil reaction;
• pile bending;
• pile shear;
• group interaction.

A batter pile converts a portion of horizontal load into axial force.

Since piles are often much more efficient in axial compression/tension than in bending, the overall foundation system can become more efficient.

Key concept: A batter pile may have lower individual lateral-flexural capacity than a vertical pile, yet the foundation system may have greater horizontal resistance because load is redirected into axial pile resistance.

12.3 Example

Suppose:

H = 1000 kN

and the batter angle is:

θ = 15°

The axial pile force required for pure geometric resolution is:

P = 1000/sin15° ≈ 3864 kN

The associated vertical component is:

V = 3864 cos15° ≈ 3732 kN

Thus, the batter pile introduces significant axial demand.

The designer must therefore check:

• axial compression capacity;
• uplift capacity;
• structural axial capacity;
• pile bending;
• pile-head connection;
• group effects;
• settlement;
• horizontal displacement;
• global stability.

13. Additional Load From a New Wedge Fill on a Pre-existing Embankment Slope

This is an important highway widening problem.

Imagine an existing embankment with a sloping side.

A new shoulder is proposed by placing additional fill outside the existing embankment.

The question is:

How can the additional loading be estimated without finite-element software?

13.1 First approximation: added fill weight

If the new fill has cross-sectional area Afill per metre length:

Wfill = γfill Afill

This gives the additional vertical load per metre length of road.

13.2 Approximate surcharge below the existing embankment

The simplest conservative approach is to treat the new fill as an equivalent surcharge over the affected zone.

For a uniformly loaded infinite area:

Δσz ≈ q

near the loaded region.

But the wedge is finite and irregular, so stress decreases with depth and horizontal distance.

13.3 2:1 stress distribution approximation

A practical hand method is the approximate 2:1 distribution:

Δσz = Q / [(B+z)(L+z)]

for an equivalent rectangular loaded area.

For a long highway embankment, the problem may be approximated per unit length:

Δσz = qB/(B+z)

where:

• q = equivalent surface pressure
• B = loaded width
• z = depth below the loaded zone

13.4 More realistic wedge approach

For a triangular or trapezoidal wedge, divide the fill into vertical strips.

Strip	Width	Height	Area	Weight
1	b1	h1	b1h1	γb1h1
2	b2	h2	b2h2	γb2h2
3	b3	h3	b3h3	γb3h3

Then calculate the additional vertical stress at the foundation or soil layer of interest by superposition.

13.5 Why slope stability must also be checked

The additional wedge does not merely increase vertical stress.

It can:

• increase driving moments;
• increase pore pressure;
• reduce factor of safety against sliding;
• change the failure surface geometry;
• increase settlement;
• activate weak interfaces;
• increase lateral spreading.

Therefore: The hand calculation of additional stress is useful for preliminary assessment, but it is not a substitute for slope-stability analysis where the widened embankment has a credible global failure mechanism.

14. A Unified Hand-Calculation Workflow

The following workflow is useful before any numerical modelling.

Step 1 — Establish geometry

• foundation dimensions;
• embedment;
• soil-layer thickness;
• groundwater level;
• slopes;
• retaining-wall geometry;
• adjacent loads.

Step 2 — Establish soil parameters

• γ, γsat, γdry;
• c', φ';
• su;
• K0;
• permeability;
• compressibility;
• OCR;
• modulus.

Step 3 — Establish groundwater and seepage

u = γw h

i = Δh/L

Step 4 — Resolve structural actions

V, H, Mx, My

Step 5 — Calculate eccentricity

ex = My/V

ey = Mx/V

Step 6 — Calculate effective dimensions

B' = B - 2ex

L' = L - 2ey

Step 7 — Check contact pressure

qavg = V/(B'L')

Step 8 — Check bearing capacity

Use the applicable bearing-capacity formulation with appropriate:

• shape factors;
• depth factors;
• inclination factors;
• groundwater corrections;
• eccentricity treatment.

Step 9 — Check sliding

Demand ≤ Available resistance

Step 10 — Check settlement

Divide compressible soil into sufficiently refined sublayers.

Step 11 — Check overall stability

For slopes, retaining systems, basements and large foundations, examine global failure mechanisms separately.

15. Why Software Should Come After the Hand Calculation

Finite-element software can calculate complicated stress fields, but it can also produce impressive-looking results from an inappropriate model.

A geotechnical engineer should first estimate:

Parameter	Hand estimate	Numerical model
Vertical stress	γz	Calculated stress field
Pore pressure	Hydrostatic/seepage estimate	Coupled flow analysis
Earth pressure	Ka/K0/Kp	Soil–structure interaction
Bearing pressure	V/A and eccentricity	Contact stress distribution
Settlement	Sublayer summation	Constitutive-model deformation
Pile response	Axial/lateral hand calculations	p-y / t-z / finite-element modelling

If a numerical model predicts a vertical stress of 900 kPa where a simple overburden calculation suggests 150 kPa, the correct first response is not “the software is sophisticated.” The correct response is “What mechanism in the model produces the additional 750 kPa?”

16. Common Geotechnical Mistakes

16.1 Treating total and effective stress as the same quantity

Always separate:

σ = σ' + u

16.2 Using Ka without checking wall movement

A restrained wall may require K0-type assessment.

16.3 Ignoring foundation eccentricity

Large moments can make the effective bearing area much smaller than the geometric foundation area.

16.4 Checking only bearing capacity

Settlement, sliding, rotation, uplift and global stability can govern.

16.5 Using one average soil parameter for a thick deposit

Soil properties can vary substantially with depth.

16.6 Using one stress increment for an entire compressible layer

Stress influence normally decreases with depth.

16.7 Assuming all pile problems are drained or all are undrained

Drainage condition depends on soil type, loading rate and construction/loading history.

16.8 Giving credit for passive resistance that can disappear

Future excavation, erosion or utility installation may remove passive soil.

16.9 Ignoring construction sequence

Earthworks and foundation construction can change the stress path substantially.

17. Expert-Level Interpretation of the Eleven Questions

Question	Core engineering principle
Earth-stress estimation	Start with equilibrium and effective stress.
Rigid vs flexible footing	Relative stiffness controls pressure redistribution.
Eccentric wind-turbine foundation	Moment creates eccentricity and reduces effective contact area.
LRFD bearing with lateral load	Lateral load creates inclination, eccentricity, sliding and overturning effects.
Deep basement slab	Embedment may improve confinement but excavation and groundwater alter the stress state.
Narrow granular backfill	Wall movement and soil arching control pressure; Ka is not automatic.
Upward seepage	Seepage force reduces effective stress while pore pressure increases.
Settlement sublayers	Stress increment and soil stiffness vary with depth.
Driven piles	Rapid displacement can produce excess pore pressure and short-term undrained response.
Drilled shafts	Drainage condition depends on soil, loading and construction; effective-stress methods are common for granular soils.
Batter piles	Inclination converts part of lateral demand into axial pile force.
Wedge fill widening	Added fill produces additional stress and may change global slope stability.

18. Practical Design Checklist

Before approving a geotechnical design, ask:

• Have total and effective stresses been separated?
• Is the groundwater level realistic?
• Is seepage present?
• What is the actual wall movement condition?
• Is Ka, K0 or Kp physically justified?
• Has eccentricity been calculated?
• Does the footing remain fully in compression?
• Has the effective bearing area been considered?
• Are factored loads and resistance factors consistent?
• Has settlement been integrated over realistic sublayers?
• Are soil parameters representative at the relevant strain level?
• Is the pile analysis compatible with drainage conditions?
• Has pile installation altered the surrounding soil?
• Has construction sequence been considered?
• Has uplift been checked?
• Has global stability been checked?
• Has the hand calculation been compared with the numerical model?

19. Final Engineering Perspective

Advanced geotechnical engineering is not primarily about memorizing equations. It is about understanding the physical mechanism represented by the equation.

The same soil can exhibit dramatically different behavior depending on:

• drainage condition;
• stress history;
• loading rate;
• wall movement;
• foundation stiffness;
• groundwater;
• seepage;
• construction sequence;
• soil–structure interaction;
• strain level.

A strong geotechnical engineer therefore develops a preliminary “mental model” before performing the detailed calculation.

For a foundation, think:

Load → eccentricity → contact area → stress → strength → deformation

For a retaining wall:

Soil movement → stress state → earth pressure → wall response

For seepage:

Head difference → hydraulic gradient → seepage force → pore pressure → effective stress

For piles:

Installation → stress path → drainage → shaft/tip resistance → load transfer

For embankment widening:

Added geometry → added weight → stress increase → deformation → stability

These simple chains are the foundation of reliable geotechnical judgment.

Final rule: Never allow a sophisticated numerical model to replace a simple equilibrium calculation. The best practice is to use analytical calculations to establish the expected order of magnitude and mechanism, and then use numerical analysis to investigate effects that cannot reasonably be represented by hand methods.

20. Selected Technical References

• FHWA Geotechnical Engineering Circular No. 6 – Shallow Foundations.
• FHWA Geotechnical Engineering Circular No. 9 – Design and Analysis of Laterally Loaded Deep Foundations.
• FHWA Geotechnical Engineering Circular No. 10 – Drilled Shafts: Construction Procedures and LRFD Design Methods.
• FHWA Geotechnical Engineering Circular No. 12 – Design and Construction of Driven Pile Foundations.
• AASHTO LRFD Bridge Design Specifications – Foundation provisions.
• USACE EM 1110-2-2502 – Flood Walls and Other Hydraulic Retaining Walls.
• Terzaghi, Peck & Mesri – Soil Mechanics in Engineering Practice.
• Das & Sivakugan – Principles of Foundation Engineering.

FHWA's current geotechnical foundation resources list dedicated guidance for shallow foundations, driven piles, drilled shafts and laterally loaded deep foundations. 6