Performance Characterization of Combined Pile–Raft Foundations (CPRF)

A Combined Pile–Raft Foundation (CPRF), also called a piled raft foundation, is a foundation system in which the raft and a selected number of piles work together to support a structure. Piles may be placed strategically to improve ultimate resistance, reduce total settlement, control differential settlement, or achieve a combination of these objectives.

Unlike a conventional piled foundation in which the design may seek to transfer most of the structural load to piles, a CPRF deliberately uses the load-bearing contribution of the raft and the piles together. The optimum solution is therefore not necessarily the one with the maximum number of piles or the highest possible pile load share.

Two important performance indicators are:

1. The pile–raft coefficient, \( \alpha_{pr} \), which expresses the fraction of total vertical load carried by the piles.
2. The settlement reduction ratio, \( R_s \), which compares the settlement of the CPRF with that of an otherwise comparable raft-only foundation under the same specified loading conditions.

These parameters are useful, but they are not sufficient by themselves. A complete assessment must also consider bearing resistance, differential settlement, pile structural capacity, raft bending and shear, soil–structure interaction, groundwater, construction effects and long-term consolidation.

Contents

1. Fundamental behaviour of CPRF
2. Load-transfer mechanism
3. Pile–raft coefficient
4. Settlement reduction ratio
5. Interaction and stiffness theory
6. Ultimate resistance and serviceability
7. Optimum pile location and number
8. Analysis and modelling methods
9. Design workflow
10. Field monitoring and validation
11. Worked illustrative example
12. Published case histories
13. Indian and international references
14. Do's and don'ts
15. Engineering conclusions

1. Fundamental Behaviour of a CPRF

The basic concept is that the raft does not become inactive when piles are introduced. Soil beneath the raft continues to provide contact resistance, while piles transfer load through shaft resistance and base resistance. Compatibility of deformation governs the load distribution between these components.

The total applied load is shared between:

• Raft contribution: contact pressure distributed over the raft–soil interface.
• Pile contribution: axial resistance developed along pile shafts and at pile toes.
• Soil deformation: compression and shear deformation of the ground, including deeper compressible layers.
• Structural deformation: bending and deformation of the raft, piles and supported structure.

The piles are commonly concentrated beneath heavily loaded columns, cores, shear walls or regions where settlement control is particularly important. However, the final arrangement depends on the actual three-dimensional interaction between the soil, piles and raft.

Central engineering principle: The piles and raft must be analysed as a compatible foundation system. It is generally inappropriate to calculate the isolated capacity of a raft and add the sum of isolated pile capacities without considering pile–pile, pile–raft and soil–raft interaction.

2. Load-Transfer Mechanism

2.1 Equilibrium of vertical loads

For a foundation under static vertical loading, neglecting external uplift and other separately modelled vertical actions, equilibrium requires:

\[ Q_{\mathrm{tot}}(s) = Q_{\mathrm{raft}}(s) + \sum_{j=1}^{n}Q_{\mathrm{pile},j}(s) \]

Here:

• \(Q_{\mathrm{tot}}\) = total vertical load transmitted to the foundation.
• \(Q_{\mathrm{raft}}\) = resultant contact force carried by the raft and supporting soil.
• \(Q_{\mathrm{pile},j}\) = axial load transferred through pile \(j\).
• \(n\) = number of piles.
• \(s\) = a specified settlement or deformation state.

The load carried by the raft can be expressed by integrating the contact stress over the effective contact area:

\[ Q_{\mathrm{raft}}(s) = \int_{A_r}q(x,y,s)\,dA \]

For a rectangular coordinate system, \(dA=dx\,dy\), so that:

\[ Q_{\mathrm{raft}}(s) = \int\!\!\int_{A_r}q(x,y,s)\,dx\,dy \]

The quantity \(q(x,y,s)\) represents the contact pressure at the raft–soil interface. It is not necessarily uniform. It may vary considerably below column locations, near raft edges, and around concentrated pile groups.

2.2 Load carried by an individual pile

The axial force in an individual pile is supported by shaft resistance and toe resistance. In simplified form:

\[ Q_{\mathrm{pile},j} = Q_{s,j}+Q_{b,j} \]

where:

\[ Q_{s,j} = \int_0^{L_j}\tau_j(z)\,u_j\,dz \]

\[ Q_{b,j}=q_{b,j}A_{b,j} \]

Here, \(L_j\) is pile length, \(u_j\) is pile perimeter, \(\tau_j\) is mobilised shaft shear stress, \(q_{b,j}\) is mobilised toe pressure and \(A_{b,j}\) is the pile base area.

These equations describe the resistance components conceptually. Actual pile resistance must be evaluated using the applicable design method, soil parameters, installation method, loading conditions and relevant pile test data.

2.3 Compatibility of deformation

For a rigid raft under idealised uniform settlement, pile-head and raft displacements may be approximated as equal. In a real flexible raft, settlement varies spatially. Each pile therefore experiences the displacement of its pile-head location, while the surrounding soil and raft redistribute load according to their stiffnesses.

The pile load is not simply a fixed fraction of the applied load. It develops as settlement increases, soil resistance is mobilised and the raft deforms.

3. Pile–Raft Coefficient (\(\alpha_{pr}\))

3.1 Definition

The pile–raft coefficient is defined as the fraction of the total foundation load carried by the piles at a specified loading or settlement state.

\[ \boxed{ \alpha_{pr}(s)= \frac{\displaystyle\sum_{j=1}^{n}Q_{\mathrm{pile},j}(s)} {Q_{\mathrm{tot}}(s)} } \]

Using the equilibrium equation:

\[ Q_{\mathrm{raft}}(s) = \left[1-\alpha_{pr}(s)\right]Q_{\mathrm{tot}}(s) \]

Consequently, the raft load fraction is:

\[ \alpha_r(s)=1-\alpha_{pr}(s) \]

Under the assumptions used in these equations, \(\alpha_{pr}\) is dimensionless and ranges from 0 to 1 for a conventional compressive load-sharing interpretation.

Coefficient	Interpretation	Important qualification
\(\alpha_{pr}=0\)	All the considered load is carried by the raft.	This is the raft-only limiting case.
\(0<\alpha_{pr}<1\)	Both the raft and piles carry load.	This is the usual CPRF condition.
\(\alpha_{pr}\rightarrow1\)	The piles carry nearly all the considered load.	This does not automatically mean the foundation is optimal.

3.2 Interpretation of a high coefficient

A higher \(\alpha_{pr}\) means that a greater proportion of the total load is transferred through the piles. It does not, on its own, demonstrate that the settlement is smaller or that the foundation is safer.

For example, a stiff pile group supporting most of the load may produce a high pile load fraction, but the raft could still experience unacceptable differential settlement due to spatial variations in soil stiffness. Conversely, a well-designed CPRF may allow a substantial portion of the load to be carried by the raft while remaining within settlement and strength limits.

3.3 Dependence on settlement and loading stage

The coefficient should be evaluated at the same defined load or settlement state when comparing alternative designs. Its value may change with:

• Magnitude and distribution of applied load.
• Raft flexural stiffness.
• Pile length, diameter, spacing and axial stiffness.
• Soil stiffness and strength with depth.
• Degree of shaft and toe resistance mobilisation.
• Groundwater variation and consolidation.
• Construction sequence and excavation unloading.

Measurement caution: Published load-sharing ratios may use different denominators, including gross structural load, net foundation load or effective load within a tributary area. Ratios should not be compared directly unless the load definitions and sign conventions are consistent.

4. Settlement Reduction Ratio (\(R_s\))

4.1 Definition

The settlement reduction ratio compares the settlement of the combined pile–raft foundation with the settlement of a comparable raft-only foundation subjected to the same specified loading conditions.

\[ \boxed{ R_s=\frac{s_{\mathrm{CPRF}}}{s_{\mathrm{raft}}} } \]

where:

• \(s_{\mathrm{CPRF}}\) = settlement of the combined pile–raft foundation.
• \(s_{\mathrm{raft}}\) = settlement of the reference raft-only foundation.

Both settlements must be evaluated using a consistent definition of loading, ground profile, foundation geometry, groundwater, drainage conditions, time and settlement measurement location.

4.2 Interpretation

Value of \(R_s\)	Interpretation
\(R_s=1.0\)	No reduction in the defined settlement relative to the reference raft.
\(0<R_s<1.0\)	The CPRF settles less than the reference raft.
\(R_s>1.0\)	The CPRF settles more than the reference raft under the stated comparison conditions.

The percentage reduction in settlement is:

\[ \boxed{ \eta_s=(1-R_s)\times100\% } \]

For example, if the raft-only settlement is 60 mm and the CPRF settlement is 30 mm:

\[ R_s=\frac{30}{60}=0.50 \]

\[ \eta_s=(1-0.50)\times100=50\% \]

The combined foundation therefore reduces the specified settlement by 50% relative to that reference case.

4.3 Which settlement should be used?

A meaningful performance comparison must identify whether the reported value is:

• Settlement at the raft centre.
• Maximum settlement across the raft.
• Average settlement.
• Settlement at a particular column or core.
• Long-term settlement after consolidation.
• Settlement at a specified construction stage.

For settlement-control design, maximum settlement and differential settlement are often more informative than a single average value. For a flexible raft, comparing centre settlement in one model with maximum settlement in another can produce a misleading ratio.

4.4 Differential settlement

Let the settlements at two points be \(s_1\) and \(s_2\), separated by a horizontal distance \(L\). The average angular distortion can be represented as:

\[ \beta=\frac{|s_1-s_2|}{L} \]

This is a simplified measure of relative vertical deformation. The acceptable limit depends on the structural system, wall and cladding sensitivity, serviceability requirements, and the applicable project criteria.

5. Interaction Theory and Foundation Stiffness

5.1 Why load sharing cannot be calculated by simple addition

The raft and piles deform the same ground. The settlement caused by one pile influences neighbouring piles and the raft, while raft pressure influences the stresses and displacements around the pile group.

A useful conceptual representation for a simplified linear interaction model is:

\[ \begin{bmatrix} \mathbf{w}_r\\ \mathbf{w}_p \end{bmatrix} = \begin{bmatrix} \mathbf{C}_{rr}&\mathbf{C}_{rp}\\ \mathbf{C}_{pr}&\mathbf{C}_{pp} \end{bmatrix} \begin{bmatrix} \mathbf{q}_r\\ \mathbf{Q}_p \end{bmatrix} \]

Here:

• \(\mathbf{w}_r\) represents raft displacement degrees of freedom.
• \(\mathbf{w}_p\) represents pile-head displacements.
• \(\mathbf{q}_r\) represents raft contact-pressure degrees of freedom.
• \(\mathbf{Q}_p\) represents pile-head loads.
• The \(\mathbf{C}\) matrices represent direct and interaction flexibility relationships.

The compatibility conditions couple raft displacement and pile-head displacement at the pile connections. This schematic equation is a framework for understanding interaction, not a complete constitutive model for nonlinear soil.

5.2 Conceptual stiffness analogy

If the raft and pile system are represented as two parallel linear springs with stiffnesses \(K_r\) and \(K_p\), both undergoing the same vertical displacement \(s\), then:

\[ Q_r=K_rs,\qquad Q_p=K_ps \]

\[ Q_{\mathrm{tot}}=(K_r+K_p)s \]

Therefore, for this idealised model:

\[ \alpha_{pr}=\frac{K_p}{K_r+K_p} \]

and:

\[ s_{\mathrm{CPRF}}=\frac{Q_{\mathrm{tot}}}{K_r+K_p} \]

If the reference raft stiffness remains \(K_r\), the corresponding idealised settlement ratio is:

\[ R_s=\frac{K_r}{K_r+K_p} \]

Thus, in this restricted linear spring model:

\[ \boxed{R_s=1-\alpha_{pr}} \]

Important limitation: \(R_s=1-\alpha_{pr}\) is not a general CPRF design law. Real foundations have nonlinear soil response, nonuniform raft pressure, differential settlement, pile-group interaction, consolidation and load-dependent stiffness. The actual relationship between \(R_s\) and \(\alpha_{pr}\) must be obtained from a suitable analysis or reliable measurements.

5.3 A more realistic conceptual interpretation

In an actual CPRF, increasing the number of piles may reduce settlement while also changing the load share. However, the settlement reduction achieved per additional pile generally depends on the position of that pile, the stiffness of the soil below it, pile spacing and the existing raft–pile interaction.

A pile beneath a heavily loaded central core may have a different effect from a pile near a lightly loaded edge. Piles can also redistribute loads and reduce curvature without producing a proportionate change in average settlement.

6. Ultimate Resistance and Serviceability

6.1 Ultimate resistance

The ultimate resistance of a CPRF is a system-level problem. The foundation must be assessed for relevant failure mechanisms, which may include:

• General or local bearing failure beneath the raft.
• Individual pile geotechnical failure.
• Pile-group failure or block-type failure.
• Combined raft–pile failure involving the surrounding soil.
• Excessive deformation before nominal ultimate resistance is reached.
• Structural failure of the raft, pile or pile–raft connection.

A preliminary conceptual representation is:

\[ Q_{\mathrm{system}} \neq Q_{\mathrm{raft,ult}} + \sum Q_{\mathrm{pile,ult}} \quad\text{as an automatic design rule} \]

The right-hand sum may overestimate system resistance if it ignores overlapping failure zones, interaction and compatible deformation. A suitable system-level analysis and applicable safety format are required.

6.2 Serviceability limit state

At working loads, the foundation must satisfy project-specific limits for total settlement, differential settlement, tilt, rotation and structural distortion. Long-term consolidation and groundwater-related effects must be considered where relevant.

A useful design checklist is:

Check	Question to answer
Total settlement	Is the predicted maximum or specified reference settlement acceptable?
Differential settlement	Are settlement gradients and angular distortions within allowable limits?
Pile axial resistance	Does each pile satisfy the relevant geotechnical and structural requirements?
Raft structural capacity	Are bending, one-way shear, punching shear and local stresses acceptable?
Load redistribution	Can the system safely accommodate variations in pile and raft load sharing?
Long-term behaviour	Have consolidation, creep where applicable, and groundwater changes been assessed?
Robustness	Is the system sufficiently tolerant of credible variations in soil and construction conditions?

7. Optimum Number and Location of Piles

The objective is not necessarily to maximise \(\alpha_{pr}\). It is to achieve the required performance with an efficient arrangement of piles.

7.1 Strategic pile placement

Piles may be concentrated beneath:

• Heavily loaded columns and shear-wall intersections.
• Lift cores and central structural cores.
• Regions with high calculated raft curvature.
• Locations where differential settlement is critical.
• Regions with weaker soil or a deeper compressible layer.
• Areas requiring additional resistance against relevant lateral or uplift effects.

The final layout must also consider constructability, reinforcement congestion, pile installation tolerances, pile spacing, groundwater and the presence of adjacent structures or utilities.

7.2 Pile spacing

Pile spacing influences group interaction and the load transferred through the raft. Closely spaced piles can interact strongly through overlapping soil deformation zones. Wider spacing may increase the contribution of the raft between piles, but may also increase local raft bending or differential settlement.

No single pile-spacing ratio is universally optimum. Spacing should be investigated against settlement, pile loads, raft moments, soil profile and the governing design requirements.

7.3 Marginal benefit of an additional pile

For preliminary optimisation, the incremental settlement benefit of adding one pile can be defined as:

\[ B_s=\frac{s_n-s_{n+1}}{s_n}\times100\% \]

where \(s_n\) is the selected settlement measure with \(n\) piles and \(s_{n+1}\) is the same measure after adding one pile while retaining consistent comparison assumptions.

An economic comparison may also use the incremental cost per unit settlement reduction:

\[ C_{\mathrm{inc}} = \frac{\text{Incremental installed foundation cost}} {\text{Settlement reduction achieved}} \]

This is a decision-support metric, not a structural design criterion. It should be used alongside all mandatory safety and serviceability checks.

8. Analysis and Modelling Methods

8.1 Preliminary analytical methods

Simplified analytical methods are useful for preliminary sizing and sensitivity studies. They can help estimate raft stiffness, pile stiffness, load sharing and settlement trends.

However, the limitations of assumptions concerning soil homogeneity, linear elasticity, pile interaction and raft rigidity must be documented.

8.2 Equivalent raft or simplified interaction approaches

Equivalent raft and simplified pile–raft interaction methods can provide practical estimates of overall foundation response. Their applicability depends on soil layering, pile geometry, load distribution and the assumptions used in the method.

These approaches are useful for screening alternatives, but highly irregular geometry, strongly layered soil, nonlinear response or complex groundwater conditions may require more detailed analysis.

8.3 Three-dimensional finite-element modelling

A three-dimensional finite-element model can represent the raft, piles and soil as interacting components. Depending on the software and constitutive formulation, the model can evaluate:

• Settlement contours and differential settlement.
• Contact pressure beneath the raft.
• Axial force distribution along individual piles.
• Shaft shear stress and pile toe resistance.
• Raft bending moments, shear forces and deformation.
• Construction-stage effects and groundwater changes.
• Nonlinear soil behaviour and consolidation, where modelled.

The reliability of the output depends on the quality of the soil investigation, constitutive parameters, interface assumptions, mesh, boundary conditions, construction sequence and model calibration.

8.4 Load-transfer methods

Load-transfer approaches use relationships between pile displacement and shaft or toe resistance. These can be combined with raft–soil response and pile-group interaction to estimate load sharing and settlement.

Where field load-test data are available, the adopted load-transfer relationships should be checked against the observed load–displacement response.

8.5 Minimum model validation requirements

• Check total vertical equilibrium.
• Check that integrated raft contact pressure and total pile-head loads reproduce the applied load, subject to explicitly modelled additional forces.
• Check mesh and domain sensitivity.
• Check the response to plausible changes in soil stiffness and strength.
• Compare predicted pile loads and settlements with relevant field measurements, if available.
• Document the difference between model predictions and measured values.

9. Recommended Design and Assessment Workflow

flowchart TD
    A[Define structure loads and performance criteria]
    B[Investigate soil profile and groundwater]
    C[Analyse raft-only foundation]
    D{Does raft satisfy all requirements?}
    E[Adopt raft solution if justified]
    F[Identify settlement or capacity deficiencies]
    G[Select preliminary pile number and locations]
    H[Analyse combined raft-pile-soil interaction]
    I[Calculate pile-raft coefficient]
    J[Calculate total and differential settlement]
    K[Check pile, raft and system resistance]
    L{All safety and serviceability checks satisfied?}
    M[Revise pile layout or foundation geometry]
    N[Optimise cost and constructability]
    O[Prepare construction quality and monitoring plan]
    P[Validate predictions against field performance]

    A --> B --> C --> D
    D -- Yes --> E
    D -- No --> F --> G --> H
    H --> I --> J --> K --> L
    L -- No --> M --> H
    L -- Yes --> N --> O --> P

This workflow is iterative. A satisfactory pile–raft coefficient does not replace the resistance checks, and a satisfactory total settlement does not replace differential-settlement or structural checks.

10. Field Monitoring and Validation

Instrumentation is particularly valuable where settlement reduction is the primary objective or where the foundation is supporting a high-rise, sensitive or heavily loaded structure.

Instrument or observation	Purpose
Settlement markers or precise levelling points	Measure total and differential settlement during and after construction.
Strain gauges or instrumented piles	Estimate axial force distribution and load transfer along piles.
Load cells at pile heads, where appropriate	Measure individual pile-head forces.
Earth-pressure cells beneath the raft	Estimate the distribution of contact pressure.
Piezometers	Monitor groundwater or pore-pressure changes.
Inclinometers, where justified	Monitor relevant lateral ground movement.
Construction records	Relate foundation behaviour to loading, excavation and construction sequence.

10.1 Back-calculating the pile–raft coefficient

Where total applied load, raft contact pressure and pile loads can be estimated consistently:

\[ \alpha_{pr,\mathrm{measured}} = \frac{\sum Q_{\mathrm{pile,measured}}} {Q_{\mathrm{total,measured}}} \]

The raft contribution can independently be estimated from contact-pressure measurements:

\[ Q_{\mathrm{raft,measured}} \approx \int_{A_r}q_{\mathrm{measured}}(x,y)\,dA \]

Instrumentation samples only selected points, so interpolation and integration uncertainties must be acknowledged. Load-cell readings, strain-gauge readings and raft pressure measurements also require appropriate calibration and interpretation.

10.2 Monitoring settlement reduction

A true field settlement reduction ratio requires a credible raft-only reference settlement. That reference is usually obtained from a validated analytical or numerical model, a suitable comparable foundation, or another justified reference assessment.

A raft-only foundation is not normally available for a direct side-by-side field comparison beneath the same completed structure. Therefore, the uncertainty of the reference settlement must be stated.

11. Illustrative Relationship Between \(\alpha_{pr}\) and \(R_s\)

The conceptual graph in the supplied figure shows a possible decrease in settlement ratio as the pile–raft coefficient increases. This is a plausible trend for some foundation systems, but the curve should not be interpreted as a universal empirical relationship.

The following interactive calculator compares two deliberately simplified relationships. The first is a parallel linear-spring idealisation. The second is a user-defined illustrative curve, not a published design correlation.

CPRF performance calculator

Pile–raft coefficient, \(\alpha_{pr}\)

Selected coefficient: 0.60

Reference raft-only settlement (mm) Illustrative model Linear parallel-spring idealisation Illustrative nonlinear relationship

Raft load fraction: 0.40

Illustrative settlement ratio: 0.40

Illustrative CPRF settlement: 24.0 mm

Illustrative settlement reduction: 60.0%

This calculator demonstrates mathematical relationships only. The nonlinear curve is an assumed scenario, not a design prediction or code-approved correlation. Do not use its output for foundation sizing.

12. Worked Illustrative Example

Consider a hypothetical foundation with the following assumed service-load results. These numbers are chosen to demonstrate the calculations; they do not represent an actual project or a calibrated soil model.

Parameter	Assumed value
Total applied vertical load	100 MN
Raft contribution at the considered stage	40 MN
Combined load carried by all piles	60 MN
Raft-only reference settlement	60 mm
Maximum CPRF settlement	30 mm

12.1 Calculate the pile–raft coefficient

\[ \alpha_{pr}=\frac{60}{100}=0.60 \]

Therefore, the piles carry 60% of the total load at the assumed stage.

12.2 Verify the raft contribution

\[ \alpha_r=\frac{40}{100}=0.40 \]

\[ \alpha_{pr}+\alpha_r=0.60+0.40=1.00 \]

The load fractions satisfy equilibrium under the stated assumptions.

12.3 Calculate settlement reduction

\[ R_s=\frac{30}{60}=0.50 \]

\[ \eta_s=(1-0.50)\times100=50\% \]

The assumed CPRF reduces the specified maximum settlement by 50% compared with the raft-only reference settlement.

12.4 Engineering interpretation

The coefficient \(\alpha_{pr}=0.60\) and ratio \(R_s=0.50\) describe two different aspects of performance. The coefficient describes load distribution; the ratio describes relative settlement.

These values do not establish that the foundation is safe. Further verification is necessary for pile resistance, raft structural design, differential settlement, group behaviour, long-term consolidation, groundwater and all relevant ultimate and serviceability limit states.

13. Published Case Histories and Real-World Evidence

13.1 Japanese buildings monitored during and after construction

Yamashita, Yamada and Hamada reported five piled-raft case histories in Japan in a 2011 paper published in Soils and Foundations. The buildings ranged from 19 m to 162 m in height. Field monitoring covered construction and approximately 17 to 60 months after completion.

The reported settlements were approximately 19–24 mm, while the pile load-sharing ratios were estimated at approximately 0.61–0.93 under the authors' adopted effective-load definition. The authors reported generally good agreement between measured and predicted settlement and load sharing.

The same study examined ten case histories and discussed the relationship between pile spacing and load sharing. The findings illustrate the importance of interaction and spacing, but they do not establish a universal optimum coefficient or spacing ratio.

Source: Yamashita, Yamada and Hamada (2011), Investigation of Settlement and Load Sharing on Piled Rafts by Monitoring Full-Scale Structures.

13.2 A 162 m residential tower in Japan

A published case history of a 47-storey, 162 m high residential tower describes a piled raft supported by 36 cast-in-place piles, each 50 m long. The raft was founded on dense diluvial sand and gravel, and the average raft contact pressure was reported as approximately 600 kPa.

Instrumentation included settlement observations, pile axial-load measurements, raft contact pressures and pore-water pressures. The investigation was undertaken to verify that the selected piled-raft solution met the foundation's stringent settlement requirements.

Source: Yamashita, Hamada and Soga, Settlement and Load Sharing of Piled Raft of a 162 m High Residential Tower.

13.3 Messeturm, Frankfurt, Germany

The Messeturm case history is an important example of piled-raft design in Frankfurt's overconsolidated clay. Published descriptions discuss a large raft supported by strategically positioned bored piles, with the pile arrangement influenced by the heavily loaded central core and the need to control settlement and raft deformation.

A later numerical study describes a 256 m building with a raft approximately 58.8 m by 58.8 m and 64 large-diameter bored piles arranged in three rings. The cited study examines settlement, pile shaft friction and the effects of groundwater conditions.

The broader lesson is that load sharing is not constant under every groundwater and loading condition. Groundwater drawdown can change effective stresses, uplift and pile resistance, making hydrogeological assumptions an important part of foundation assessment.

Source: Numerical Investigation of Optimized Piled Raft Foundation for a High-Rise Building in Germany.

13.4 Piled raft combined with ground improvement in Japan

Published research on a large base-isolated building in Japan describes a piled raft combined with grid-form deep mixing walls. The ground improvement was intended to address liquefiable sand and reduce settlement in underlying soft cohesive soil.

The reported monitoring included settlement and load sharing over several years, as well as seismic observations. The case illustrates that a piled raft may need to be evaluated together with ground improvement where liquefaction or deep soft layers govern performance.

Source: Field Measurements on Piled Rafts with Grid-Form Deep Mixing Walls on Soft Ground.

Evidence qualification: The case histories above demonstrate practical piled-raft applications. They should not be treated as direct design precedents unless the geology, structural loading, pile geometry, groundwater, construction method and performance criteria are sufficiently comparable.

14. Indian and International Codal References

For an Indian project, the CPRF-specific standard should be checked first, followed by the relevant foundation, pile, settlement, investigation and structural standards. Always confirm the current applicable edition, amendments and project specifications from the official issuing authority.

Standard or publication	Relevance to CPRF
IS 19117:2025 Design and Construction of Combined Piled Raft Foundations — Code of Practice	The directly relevant Indian Standard for combined piled-raft foundations. Consult the complete official text for its provisions and applicability.
IS 1892:2021 Subsurface Investigation for Foundations — Code of Practice	Ground investigation and characterisation of the soil profile.
IS 1904:2021 General Requirements for Design and Construction of Foundations in Soils — Code of Practice	General foundation design and construction considerations.
IS 2950 (Part 1):1981, including applicable amendments Code of Practice for Design and Construction of Raft Foundations: Part 1 — Design	Raft design considerations; apply alongside the CPRF-specific standard.
IS 2911 (Part 1), relevant sections	Design and construction requirements for the applicable type of concrete pile.
IS 2911 (Part 4):2013 Load Test on Piles	Relevant pile load-testing procedures and evaluation, subject to the current applicable edition.
IS 8009 (Part 1):1976	Settlement calculation for shallow foundations under symmetrical static vertical loading.
IS 8009 (Part 2):1980	Settlement calculation for deep foundations under symmetrical static vertical loading.
IS 6403:1981	Determination of bearing capacity of shallow foundations.
IS 456:2000, including applicable amendments	Reinforced concrete design and detailing of the raft and piles, as applicable.

Official reference portals:

• BIS preview: IS 19117:2025
• Bureau of Indian Standards

14.1 International technical references

• Poulos, H. G. (2001), Piled raft foundations: design and applications, Géotechnique, 51(2), 95–113. https://doi.org/10.1680/geot.2001.51.2.95
• Poulos, H. G. (1994), An approximate numerical analysis of pile–raft interaction, International Journal for Numerical and Analytical Methods in Geomechanics, 18(2), 73–92. https://doi.org/10.1002/nag.1610180202
• Yamashita, K., Yamada, T. and Hamada, J. (2011), Investigation of Settlement and Load Sharing on Piled Rafts by Monitoring Full-Scale Structures, Soils and Foundations, 51(3), 513–532. https://doi.org/10.3208/sandf.51.513

These publications provide theoretical background and field evidence. They do not replace the mandatory requirements of the applicable Indian Standard, project design basis or statutory approvals.

15. Do's and Don'ts in CPRF Performance Assessment

Do's

• Use a realistic geotechnical model based on adequate investigation, groundwater observations and appropriate soil parameters.
• Analyse the raft-only alternative before selecting a CPRF, so the benefit of adding piles can be quantified.
• Calculate \(\alpha_{pr}\) at a clearly defined load or settlement stage.
• Use the same loading basis, geometry and settlement definition when calculating \(R_s\).
• Evaluate maximum settlement, differential settlement and angular distortion, not just average settlement.
• Check the structural capacity of the raft, piles and their connections.
• Consider pile-group interaction and the effect of pile position on raft bending and soil response.
• Assess long-term consolidation and credible groundwater changes where relevant.
• Perform sensitivity analyses on soil stiffness, strength and other uncertain parameters.
• Plan instrumentation and compare measured performance with predictions where the project warrants it.
• Document assumptions, model limitations, design load combinations and acceptance criteria.
• Use the current applicable editions of the relevant Indian Standards and the project's approved design criteria.

Don'ts

• Do not assume that the highest possible pile load fraction represents the optimum CPRF.
• Do not assume that \(R_s=1-\alpha_{pr}\) applies to real foundations without demonstrating that its restrictive assumptions are justified.
• Do not simply add isolated raft and pile ultimate capacities without assessing interaction and system failure mechanisms.
• Do not compare pile load shares from different studies without checking how the denominator and load state were defined.
• Do not compare maximum settlement in one analysis with average settlement in another.
• Do not ignore the bending, punching shear and one-way shear demands in the raft.
• Do not assume that every pile carries the same axial load.
• Do not ignore construction staging, excavation unloading, negative skin friction where applicable, or groundwater effects.
• Do not calibrate a numerical model solely to match one settlement value while ignoring pile loads and raft contact pressures.
• Do not adopt illustrative curves, empirical trends or case-history values as universal code-prescribed design limits.

16. Advanced Interpretation: What Should Be Optimised?

The most useful design question is not, “How can the pile–raft coefficient be maximised?” It is, “What pile arrangement provides acceptable system resistance and deformation at an appropriate whole-life cost?”

A rational optimisation study can compare several layouts against the following indicators:

Indicator	Purpose
\(\alpha_{pr}\)	Quantifies the proportion of load transferred through the piles.
\(R_s\)	Quantifies the settlement reduction relative to the defined raft-only reference.
\(s_{\max}\)	Measures maximum settlement.
\(\Delta s\) and \(\beta\)	Assess differential settlement and angular distortion.
Maximum pile load	Identifies piles that may govern resistance or structural design.
Raft bending and shear	Identifies structural demands and local concentration effects.
System resistance	Checks relevant ultimate limit states.
Installed cost and construction risk	Supports selection of an economical, constructible solution.

An effective design may deliberately allow the raft to carry a substantial share of the load. The piles can then be used to reduce deformation, improve load redistribution and provide additional resistance where required.

The optimum solution depends on the relative stiffness of the raft and piles, the soil profile, the structural loading pattern, construction constraints and the governing acceptance criteria. No single value of \(\alpha_{pr}\) or \(R_s\) is suitable for every project.

17. Conclusions

1. The pile–raft coefficient and settlement reduction ratio describe different aspects of performance. The former quantifies load sharing; the latter quantifies settlement relative to a defined raft-only reference.
2. Both parameters are state-dependent. They depend on load level, deformation, soil properties, pile arrangement, groundwater and construction history.
3. Settlement reduction must be evaluated using consistent reference conditions. Total settlement, maximum settlement and differential settlement should not be interchanged without explanation.
4. Interaction is fundamental. The raft, piles and soil form a coupled system, so independent component capacities cannot automatically be added to obtain the system resistance.
5. The optimum CPRF is not necessarily the system with the highest pile load share. The objective is to satisfy all applicable strength and serviceability requirements efficiently.
6. Field evidence is valuable. Settlement measurements, pile loads, contact pressures and groundwater observations help validate the design model and identify departures from predicted behaviour.
7. For Indian projects, consult IS 19117:2025 directly. Apply it together with the relevant current foundation, pile, settlement, soil investigation and structural standards.

Final engineering principle: A CPRF is successful when the raft and piles act as an intentionally designed, compatible system that satisfies ultimate resistance, total settlement, differential settlement, structural integrity and durability requirements—not merely when a target pile–raft coefficient is achieved.

References and Further Reading

1. Bureau of Indian Standards. IS 19117:2025, Design and Construction of Combined Piled Raft Foundations — Code of Practice. Consult the official standard for detailed provisions.
2. Bureau of Indian Standards. IS 1892:2021, Subsurface Investigation for Foundations — Code of Practice.
3. Bureau of Indian Standards. IS 2950 (Part 1):1981, Code of Practice for Design and Construction of Raft Foundations: Part 1 — Design, with applicable amendments.
4. Poulos, H. G. (2001). Piled raft foundations: design and applications. Géotechnique, 51(2), 95–113.
5. Poulos, H. G. (1994). An approximate numerical analysis of pile–raft interaction. International Journal for Numerical and Analytical Methods in Geomechanics, 18(2), 73–92.
6. Yamashita, K., Yamada, T. and Hamada, J. (2011). Investigation of Settlement and Load Sharing on Piled Rafts by Monitoring Full-Scale Structures. Soils and Foundations, 51(3), 513–532.
7. Yamashita, K., Hamada, J. and Soga, Y. Settlement and Load Sharing of Piled Raft of a 162 m High Residential Tower. ASCE proceedings publication.

Disclaimer: This educational article is intended to explain CPRF performance assessment. It is not a substitute for a project-specific geotechnical investigation, detailed foundation analysis, independent design review or compliance with the complete current standards and statutory requirements.

Yogendra Gopal Borse

Civil Engineer | Assistant Engineer Grade-I, Maharashtra PWD

B.Tech (Civil) from VJTI Mumbai. Experienced in bridge design, road works, estimation, project monitoring and digital engineering tools. Creator of YogiPWD – practical technical resources for civil engineers.

Read full About →