In this study, the compatibility of polycarboxylate ether-based water-reducing admixtures (PCE) with cements produced with different types and dosages of grinding aids (GA) was experimentally and statistically investigated. A total of 203 paste mixtures were prepared using seven different types of GA and one type of PCE at different dosages. The Marsh funnel flow time and mini-slump values of the mixtures were compared. A modeling study was performed using the experimental data. In this direction, Classical Regression Analysis (CRA), Multivariate Adaptive Regression Splines (MARS), and Artificial Neural Networks (AOMA-ANN) were applied. Innovative approaches, AOMA-ANN (AIP) and AOMA-ANN (ONIP), were introduced. The results showed adverse effects on flow performance with increased GA utilization, except for TEA-based GA. TEA-type GA had the lowest flow performance. AOMA-ANN (ONIP) exhibited the best performance in modeling. Marsh funnel flow-time modeling with AOMA-ANN (ONIP) considered parameters such as sieve residue at 60 microns, the number of molecules per fineness, the density of GA, the pH value of GA, and the PCE dosage. Mini-slump modeling with AOMA-ANN (ONIP) considered parameters such as sieve residue at 60 microns, the density of GA, the pH value of GA, and the PCE dosage.

To develop more environmentally friendly and sustainable cementitious systems, the use of grinding aids (GAs) during the clinker grinding process has increasingly gained attention. Although the mechanisms of the action of grinding aids (GAs) are known, the selection of an effective grinding aid (GA) can be difficult due to the complexity of appropriate selection criteria. For this reason, it is important to model the effect of GA properties on grinding performance. In this study, seven different types of GAs were used in four different dosages, and time-dependent grinding was performed. The Blaine fineness values of cements were compared after each grinding process. In addition, the modeling of these parameters using machine learning and ensemble learning methods was discussed. The Synthetic Minority Over-sampling Technique (Smote) was used to generate artificial data and increase the number of data for the grinding efficiency experiment. The data were modeled using methods such as Artificial Neural Networks (ANNs), Attentive Interpretable Tabular Learning (TabNet), Random Forests (RFs), and the XGBoost Regressor (XGBoost), and the ranking of the parameters affecting the Blaine properties was determined using the XGBoost method. The XGBoost method achieved the best results in the MAE, RMSE, and LogCosh metrics with values of 21.0384, 33.7379, and 15.4846, respectively, in the experimental modeling studies with augmented data. This study contributes to a better understanding of the relationship between GA selection and milling process performance.

This study explores the application of machine learning (ML) techniques—gradient boosting (GB), ridge regression (RR), and support vector regression (SVR)—for estimating the consumption of energy (CE) and Blaine fineness (BF) in cement clinker grinding. This study utilizes key clinker grinding parameters, such as maximum ball size, ball filling ratio, clinker mass, rotation speed, and number of revolutions, as input features. Through comprehensive preprocessing, feature selection methods (mutual info regression (MIR), lasso regression (LR), and sequential backward selection (SBS)) were employed to identify the most significant variables for predicting CE and BF. The performance of the models was optimized using a grid search for hyperparameter tuning and validated using k-fold cross-validation (k = 10). The results show that all ML methods effectively estimated the target parameters, with SVR demonstrating superior accuracy in both CE and BF predictions, as evidenced by its higher R² and lower error metrics (MAE, MAPE, and RMSE). This research highlights the potential of ML models in optimizing cement grinding processes, offering a novel approach to parameter estimation that can reduce experimental effort and enhance production efficiency. The findings underscore the advantages of SVR, making it the most reliable method for predicting energy consumption and Blaine fineness in clinker grinding.

The clinker grinding stage in Portland cement production is highly energy-intensive, primarily due to particle agglomeration and intensified interparticle attractive forces that hinder efficient comminution. Grinding aids (GAs) are routinely employed to mitigate these issues, enhancing grinding efficiency and improving cement performance. However, the undesirable side effects associated with conventional GAs on cementitious systems have spurred interest in modification strategies that can concurrently optimize grinding efficiency and final product quality. In this study, widely used commercial GAs, triisopropanolamine (TIPA), diisopropanolamine (DEIPA), and diethylene glycol (DEG), were chemically modified via esterification with organic acids of different carbon chain lengths. Cement specimens incorporating these modified GAs were produced at two dosages (0.05% and 0.1% by mass of clinker + gypsum), resulting in 24 distinct Portland cement formulations alongside a control mix. The influence of modification on grinding efficiency, particle size distribution (PSD), and powder flowability was investigated. Furthermore, scanning electron microscopy (SEM) was utilized to analyze particle morphology and concrete microstructural characteristics with powder flow behavior. The results indicate that organic acid modification not only facilitates achieving target fineness with lower energy consumption but also markedly improves both the PSD profile and the powder’s flow properties. Specifically, hexanoic acid-modified TIPA and DEIPA, along with propanoic acid-modified DEG, delivered the most favorable outcomes across the evaluated parameters. These findings underscore the potential of developing next-generation, modified GAs that simultaneously enhance energy efficiency and powder handling in cement grinding operations.