Prediction of Mechanical Properties of Rock Materials

The mechanical behaviour of rock materials underpins the design of tunnels, foundations, slopes, dams and hydrocarbon wells. Key properties such as uniaxial compressive strength (UCS), Young’s modulus (E), shear modulus (G), cohesion (c) and friction angle (ϕ) determine rock-mass stability and deformation. Conventional determination relies on laboratory compression and tensile tests, which are costly, time-consuming and sensitive to sampling and scale effects. Over the past decade, indirect estimation methods have gained prominence, employing empirical correlations with index tests (e.g. P-wave velocity, Schmidt hammer rebound), machine learning and soft computing to capture non-linear interrelations between physical parameters and strength indices. Recent advances in deep learning architectures and explainable algorithms have further improved predictive accuracy and transparency, enabling more reliable site-specific models. The global push towards sustainable resource extraction and infrastructure resilience has driven the development of efficient, non-destructive methodologies that can be applied in remote or hazardous environments. Interdisciplinary efforts now integrate geomechanics, computational intelligence and big data analytics to support real-time decision-making and risk assessment in geotechnical engineering.

Research from Nature Portfolio

No recent Nature Portfolio content available.

Prediction of Mechanical Properties of Rock Materials publication trend

The graph below shows the total number of articles in prediction of mechanical properties of rock materials across all publications each year (not limited to Nature Index journals).

Technical terms

Uniaxial compressive strength (UCS): Maximum axial stress that a rock sample can withstand under uniaxial loading.

Young’s modulus (E): Stiffness of a rock expressed as the ratio of axial stress to axial strain in the elastic regime.

P-wave velocity (Vp): Speed at which primary compressional waves travel through a rock, indicative of elastic properties and density.

Brazilian tensile strength (BTS): Indirect measure of tensile strength obtained by diametral compression of a rock disc.

Porosity: Fraction of a rock’s volume occupied by void spaces, affecting strength and permeability.

Artificial neural network (ANN): Computational model inspired by biological neural networks, used to capture non-linear relationships in data.

Deep neural network (DNN): A class of ANN with multiple hidden layers enabling the extraction of complex features from large datasets.

Extreme gradient boosting (XGBoost): Ensemble tree-based learning algorithm that sequentially builds models to minimise prediction error.

Explainable AI (XAI): Set of techniques such as SHAP to interpret and visualise the decision-making process of complex models.

References

1. Closed-Form Equation for Estimating Unconfined Compressive Strength of Granite from Three Non-destructive Tests Using Soft Computing Models. Rock Mechanics and Rock Engineering (2022).
2. A Deep Learning Method for the Prediction of the Index Mechanical Properties and Strength Parameters of Marlstone. Materials (2022).
3. Prediction of uniaxial compressive strength and modulus of elasticity for Travertine samples using an explainable artificial intelligence. Results in Geophysical Sciences (2021).