High-Pressure Grinding Process Modeling

High-pressure grinding rolls (HPGR) have emerged as a cornerstone of energy-efficient comminution in mining and cement industries. Modelling of the HPGR process seeks to predict key performance metrics – particle breakage, product size distribution, energy consumption and equipment wear – as functions of operating parameters and material properties. Mechanistic approaches couple the discrete element method (DEM) with multi-body dynamics (MBD) and particle fracture models to resolve inter-particle and particle–roll forces, characterise breakage events and quantify the influence of roll geometry and wear. Population balance models (PBM) integrate breakage and selection functions calibrated by experiments or simulations to predict downstream product size distributions under varying specific forces, roll gaps and feed sizes. In parallel, data-driven frameworks, often based on explainable artificial intelligence (AI), have been developed to translate large operational datasets into predictive models of power draw and size reduction. Such “conscious lab” systems employ advanced regression algorithms – for example eXtreme Gradient Boosting (XGBoost) enhanced by SHapley Additive exPlanations (SHAP) – to elucidate variable importance and to furnish real-time decision-support for process optimisation. Together, these modelling paradigms enable a deeper understanding of HPGR dynamics, inform roll maintenance strategies, and underpin the design of control systems to reduce energy consumption and enhance throughput consistency on an industrial scale.

Research from Nature Portfolio

A comprehensive discrete element study has been conducted to probe the mechanical characteristics of ore during roll crushing. By calibrating particle breakage models against compression and impact tests, researchers simulated an industrial-scale HPGR configuration, examining the roles of roll structure, operational pressure and feed size on stress distributions and material flow. Novel analysis of particle velocity fields within the compression zone revealed that normal forces dominate breakage, while regions adjacent to cheek plates exhibit heightened wear susceptibility. The findings identify a vulnerable zone where roll and cheek-plate interactions must be managed to optimise crushing efficiency and limit equipment degradation.

An innovative “conscious lab” approach has been applied to model energy consumption in a vertical roller mill (VRM), illustrating the potential of explainable AI for comminution circuits. Operational variables from a cement plant were fed into an XGBoost framework, with SHAP analysis isolating working pressure and gas flow as principal drivers of power draw. The model achieved high predictive accuracy (R² > 0.80) and demonstrated superior performance to conventional statistical and machine-learning techniques. This work underscores the value of transparent AI in demystifying process-energy relationships and guiding operational adjustments to curtail electrical demand.

High-Pressure Grinding Process Modeling publication trend

The graph below shows the total number of articles in high-pressure grinding process modeling across all publications each year (not limited to Nature Index journals).

Technical terms

Discrete Element Method (DEM): A numerical technique that simulates the motion and interaction of individual particles under contact and force laws.

Population Balance Model (PBM): A mass-balance framework that represents the generation and selection of particle size classes via breakage and classification functions.

Edge Effect: The tendency for roll ends in HPGR mills to exhibit differing wear and breakage characteristics compared to the central region, affecting product uniformity.

Explainable AI (EAI): Machine-learning models enhanced with interpretability tools (such as SHAP) to clarify the influence of input variables on predictions.

SHapley Additive exPlanations (SHAP): A method to attribute the contribution of each input variable to the output of a predictive model, based on cooperative game theory.

References

1. Quantify the edge effect of HPGR mills with DEM modelling. Particuology (2025).
2. DEM analysis of wear evolution and its effect on the operation of a lab-scale HPGR mill. Minerals Engineering (2023).
3. Mechanical characteristics of roll crushing of ore materials based on discrete element method. Scientific Reports (2025).
4. Investigation of Lateral Confinement, Roller Aspect Ratio and Wear Condition on HPGR Performance Using DEM-MBD-PRM Simulations. Minerals (2021).
5. Modeling of energy consumption factors for an industrial cement vertical roller mill by SHAP-XGBoost: a "conscious lab" approach. Scientific Reports (2022).
6. Unifying high-pressure grinding rolls models. Minerals Engineering (2022).