Field-Calibrated Design Equations Promise Safer, Cheaper Bored Pile Foundations

Beneath many of the world’s largest bridges, high-rise towers, and port facilities lies an unglamorous but critical piece of engineering: the large-diameter bored pile. These deep foundations, drilled and cast in place rather than driven, carry enormous structural loads down through thick deposits of soil into the weak or weathered rock formations that underlie them. In such layered ground, a substantial share of the pile’s load-bearing capacity comes not from its tip but from the friction and adhesion mobilized along its shaft, where concrete grips the surrounding clay and sand. Estimating that shaft resistance accurately has long been one of the most stubborn uncertainties in geotechnical design, and a new study from Japan now offers engineers a field-verified way to shrink that uncertainty.

The research, led by Professor Shinya Inazumi of the College of Engineering at Shibaura Institute of Technology, was made available online on July 25, 2026 and published in Volume 32 of the journal Results in Engineering on December 1, 2026. Its central contribution is a set of empirical design correlations for shaft resistance, back-calculated from twenty full-scale instrumented static load tests conducted along a 180-kilometer infrastructure corridor in northeastern Thailand. Rather than relying on scattered values from the literature, the team built its design equations directly from measurements of how load actually transfers from pile to soil, layer by layer, in real foundations subjected to very high test loads.

To understand why this matters, it helps to look at how shaft resistance is conventionally estimated. For cohesive soils such as clay, engineers commonly use the adhesion factor method, in which the unit shaft resistance is taken as a fraction, denoted α, of the soil’s undrained shear strength. For cohesionless soils such as sand, the friction factor method applies, expressing unit shaft resistance as a factor β multiplied by the effective vertical stress at the depth of interest. Both approaches are elegantly simple, but their empirical factors are notoriously variable. Published α and β values differ widely because they are influenced by soil type, geological history, construction technique, and in-situ stress conditions, and most published correlations are derived from ordinary soils alone.

That last limitation is where the new study makes its most distinctive contribution. Between ordinary clay or sand and genuine weak rock lies a class of materials that geotechnical engineers call intermediate geomaterials, or IGMs. These stiff transitional soils are neither fully soil nor fully rock, yet they are extremely common in regions where thick sedimentary deposits grade into weathered siltstone or sandstone. Because most design guidance was developed for conventional soils, engineers have had little reliable basis for assigning α or β values to IGMs, forcing conservative assumptions that can inflate foundation costs or, worse, mischaracterize how a pile will actually behave.

The field campaign behind the new correlations was unusually extensive. The team’s site investigation program involved 158 boreholes drilled at an average spacing of roughly 1.1 to 1.2 kilometers along the 180-kilometer corridor, with soil profiling carried out using the Standard Penetration Test. The bored piles themselves were constructed using a wet process with bentonite slurry supporting the borehole during excavation, a common technique for large-diameter piles in difficult ground. Twenty instrumented static load tests were then performed, with maximum applied loads ranging from 16.25 to 25 meganewtons, corresponding to 2.5 times the allowable working load. The piles were embedded in layered ground consisting of either clay or sand overlying weak siltstone or sandstone.

What transformed these tests into a research dataset was the instrumentation. Each test pile was fitted with vibrating-wire strain gauges and extensometer rods, allowing the researchers to monitor how load transferred along the shaft as the test progressed. By analyzing strain measurements at different depths, they could isolate the unit shaft resistance mobilized within individual soil layers. This layer-by-layer approach produced a dataset covering 32 clay and IGM clay layers and 10 sand and IGM sand layers. From each layer’s stress-displacement curve, the team applied a mobilization criterion to select representative maximum shaft resistance values, ensuring that the back-calculated factors reflected fully mobilized resistance rather than partial, load-level-dependent readings.

The resulting adhesion factors for clay and IGM clay ranged from 0.11 to 1.04 and showed a clear, systematic trend: α decreases as the normalized undrained shear strength increases. The empirical equation derived from this trend offers what the researchers describe as a practical mid-range estimate for clays overlying weak rock. Particularly revealing was the comparison across material classes. When α values for ordinary clay, IGM clay, and weak rock were plotted together, they revealed a gradual transition, with IGM clay forming the upper envelope of the clay response. In other words, IGM clay genuinely behaves as an intermediate material, sitting between ordinary clay and weak rock, and treating it as either end member would misrepresent its capacity.

For the cohesionless side, the calculated friction factors ranged from 0.45 to 1.06 across the investigated sand layers, and the proposed empirical equation for β showed good agreement with correlations already reported in the literature. Encouragingly, the β values obtained for IGM sand followed the same overall trend as ordinary sand, suggesting that conventional β-method formulations can be extended to IGM sand provided that stress history and site-specific ground conditions are appropriately considered. This finding is significant because it means engineers may not need entirely new frameworks for sandy intermediate geomaterials, only careful calibration of the parameters they already use.

The practical implications reach well beyond academic curve-fitting. The study demonstrates that intermediate geomaterials should not be treated as conventional soils without accounting for their transitional behavior, and the researchers propose that a normalized undrained strength of approximately 10 can serve as a practical reference point for identifying where IGM clay gives way to weak rock. Such a threshold gives designers a simple, measurable criterion for classifying ground along soil-rock transitions, which are widespread across Southeast Asia and many other regions where heavy infrastructure is built on deep sedimentary basins. Applying these correlations during preliminary design, the team notes, can produce more reliable capacity estimates without costly additional testing, potentially allowing shorter piles and reduced concrete volumes, which lowers construction costs while supporting safer structures.

For a field that has long balanced safety against economy using factors inherited from decades-old case histories, the value of twenty fully instrumented, full-scale load tests along a single well-characterized corridor is hard to overstate. By replacing guesswork with field-verified evidence, the study gives foundation engineers a defensible basis for predicting how much load the shaft of a large bored pile can carry in layered ground, and it clarifies, with data rather than assumption, where ordinary soil ends and weak rock begins.

Subject of Research: Field-calibrated prediction of shaft resistance for large-diameter bored piles in clay, sand, and intermediate geomaterials

Article Title: Towards more reliable shaft resistance prediction for bored piles

Article References: Towards more reliable shaft resistance prediction for bored piles. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: bored piles, shaft resistance, adhesion factor, friction factor, intermediate geomaterials, static load tests, geotechnical engineering, foundation design, soil-rock transition, instrumented piles, clay, sand