← Latest papers
⚡ engineering

Global sensitivity and identifiability of Structured Cam Clay parameters for cement-treated clays: a variance-based analysis with a staged calibration protocol

This study employs a variance-based global sensitivity analysis and a staged calibration protocol to demonstrate that while key Structured Cam Clay parameters governing cement-treated clays exhibit strong path-dependent influences and identifiability trade-offs, they can be effectively isolated and calibrated using specific laboratory measurements.

Original authors: Alhassan Ahmed Igbeneghu, Ngozi Kayode-Ojo, Samuel Adesina Adegbemileke

Published 2026-11-16✓ Author reviewed ⓘ
📖 5 min read🧠 Deep dive

Original authors: Alhassan Ahmed Igbeneghu, Ngozi Kayode-Ojo, Samuel Adesina Adegbemileke

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Beneath the roads, embankments, and light buildings of many coastal cities lies a layer of soft, squishy clay that cannot support weight on its own. To make this ground safe for construction, engineers often mix in cement. The cement acts like a glue, bonding the soil particles together and giving the earth a rigid internal skeleton it did not have before. This process turns a weak, compressible mud into a stronger, stiffer material capable of holding up a structure. However, once the weight of a building is applied, this new skeleton begins to break down. The strength gained from the cement fades, and the soil starts to settle. Predicting exactly how much the ground will sink or how much load it can hold requires a mathematical model that accounts for both the initial glue and the way that glue eventually fails. For decades, engineers have used a specific set of equations, known as the Structured Cam Clay model, to make these predictions. It is a tool that adds just a few extra numbers to an older, well-known model to describe this cemented behavior. But a lingering question has remained: do we truly understand how each of these numbers controls the soil's behavior, and can we actually find the right numbers from the standard tests engineers perform in the lab?

A team of researchers from Nigeria set out to answer these questions with a rigorous, computer-based investigation. Instead of looking at the model one number at a time, as is common practice, they treated the entire system as a complex machine where all the numbers change simultaneously. They wrote a computer program to simulate thousands of different scenarios, representing a wide range of cement-treated clays found in nature. They ran these simulations under two distinct conditions that mimic real-world construction: one where water is trapped inside the soil and cannot escape, and another where water is allowed to flow out freely. By observing how the soil responded in these thousands of virtual tests, they could measure exactly which numbers mattered most for strength and which mattered most for how much the soil would compress.

The results revealed a surprising truth about how these soils behave. When the soil is tested in a way that traps water, its peak strength is almost entirely controlled by just two factors: the amount of cement glue holding it together and a measure of how the soil particles naturally slide past one another. The other numbers, which describe the soil's internal structure and how that structure breaks down, have almost no effect on the maximum strength the soil can reach in this wet, trapped state. In fact, the researchers found that for soils with a significant internal structure, the maximum strength can be predicted with a simple rule of thumb that is accurate to within five percent for most cases. This rule suggests that the strength is simply a combination of the cement's contribution and the soil's natural friction, divided by two.

However, the story changes completely when the water is allowed to escape. In this drained condition, the numbers that control strength become far less important. Instead, the soil's tendency to shrink or expand is ruled by entirely different factors: the amount of extra space between the particles created by the cement, and a number that describes how quickly the soil's internal structure collapses under pressure. A soil that appears incredibly strong when tested with water trapped inside might turn out to be highly compressible and prone to sinking when the water is allowed to drain away. This distinction is critical because a building might be safe from immediate collapse but still suffer from dangerous, long-term settlement if the wrong numbers are used to predict its behavior.

The study also tackled the problem of finding the right numbers for the model in the first place. The researchers discovered that trying to fit all the numbers at once to a single test curve is a trap. They found that two key numbers—the cement strength and the frictional sliding tendency—can trade off against each other perfectly. This means that a computer could find many different combinations of these numbers that all produce the exact same result on a test graph, even though the physical meaning of those numbers would be completely different. Because of this, the researchers concluded that you cannot simply run a computer program to guess all the values at once. Instead, they proposed a step-by-step protocol. This method suggests measuring specific properties in isolation first, such as testing the soil without cement to find its natural compression rate, and then using separate tests to pin down the cement's specific contribution. Only by fixing these values one by one, using the specific test that isolates them, can engineers avoid the confusion of trade-offs and build a reliable model.

To prove their method works, the team applied this step-by-step approach to five different samples of cement-treated sandy clay from Southwest Nigeria. The results confirmed their theory. One soil sample, which had a very fragile internal structure, showed a modest increase in strength when cement was added. Yet, when the water was allowed to drain, this same soil was the most compressible of all, showing the largest volume change. If an engineer had looked only at the strength in the wet condition, they would have missed the fact that this soil was the most likely to settle significantly over time. The study demonstrates that the influence of any single property depends entirely on the conditions of the test and the specific outcome being measured. By understanding these relationships and using a staged approach to calibration, engineers can better predict how cement-treated ground will behave, ensuring that the structures built upon it remain safe and stable.

Drowning in papers in your field?

Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.

Try Digest →