A critical challenge in mineral processing is the accurate estimation of fundamental sampling error (FSE). Currently this challenge is addressed by two dominant yet disconnected paradigms: Gy’s liberation-based theory and empirical nugget effect analysis. The liberation-corrected model reveals an ore’s intrinsic homogeneity but cannot predict sampling error for intermediate crush sizes. Conversely, the nominal nugget model measures observed heterogeneity but fails to distinguish between intrinsic clustering and liberation state. This paper introduces a novel integrated framework that synergizes these approaches to resolve this disconnect. We re-analyse a comprehensive gold ore dataset (448 samples across 14 size fractions) using both the liberation-corrected FSE model and the Poisson-based nominal nugget model. Our analysis demonstrates that the ore is intrinsically homogeneous (intrinsic homogeneity index = 0.68
Ultrafiltration (UF) membrane technology has emerged as an efficient and energy-saving method for treating algae-laden water, yet membrane fouling remains a significant barrier hindering its widespread adoption. This study proposed a novel pretreatment method combining permanganate, sulfite, and ferrous (Mn(VII)/S(IV)/Fe (II)) to mitigate membrane fouling caused by algae-laden water. The results demonstrated that the Mn(VII)/S (IV)/Fe(II) pretreatment significantly improved membrane flux, reducing reversible and irreversible fouling resistance by 96.95 % and 90 %, respectively, compared to untreated algae-laden water. Blocking model analysis revealed that the Mn(VII)/S(IV)/Fe(II) pretreatment effectively delayed the cake layer formation, with standard blocking emerging as the dominant fouling mechanism. The cake layer formed after Mn(VII)/S(IV)/Fe(II) pretreatment was sparser, rougher, more hydrophilic, and less negatively charged than that formed directly by the untreated algae-laden water. The generated active species from the Mn(VII)/S(IV)/Fe(II) system played a critical role in oxidizing algal organic matter, breaking aromatic structures and functional groups (e.g., C--O, C-N), reducing hydrophobic interactions and electrostatic attraction between foulants and the membrane. Extended Derjaguin-Landau-Verwey-Overbeek analysis confirmed that the pretreatment reduced the total interfacial free energy by 25.3 %, primarily due to weakened van der Waals forces and enhanced electrostatic repulsion. This reduction in adhesion energy effectively inhibited algal contaminant deposition. In summary, the Mn(VII)/S(IV)/ Fe(II) pretreatment approach proposed by us demonstrates significant potential for mitigating UF membrane fouling caused by algae-laden water, offering promising applications for membrane-based drinking water treatment technologies.
This study evaluates the feasibility of a passive phase change material (PCM)-based battery thermal management system intended for use in non-road mobile machines (NRMM). Two 3P4S modules, each with a nominal energy of 200 W h, were assembled from high-power cylindrical NCA Li-ion cells: one module was filled with a commercial paraffin PCM and the other was a reference without PCM. The modules were cycled at 1C, 1.5C, and 2C under identical ambient conditions, and module temperatures were recorded during charge and discharge. The results show that the PCM exhibits distinct thermal behavior at different C-rates. At 1C, the sensible heat of the solid PCM is sufficient to limit the temperature rise of the module. At 1.5C, the module temperature reaches the PCM melting point and remains approximately constant. At 2C, practically all PCM melts, which delays the temperature rise but its effect is constrained by the poor thermal conductivity of the organic PCM. The added PCM mass reduced the module gravimetric energy density from 231 Whkg-1 to 146 W hkg-1. Nevertheless, the results indicate that paraffin-based passive PCM thermal management can effectively limit temperature rise and improve the thermal behavior of high-power NCA battery modules at moderate C-rates, while still maintaining a competitive system-level energy density for small-scale applications.
Cognitive workload is critical to safety and efficiency in industrial human–robot collaboration. Conventional assessments often rely on intrusive sensors, subjective ratings, or task-specific models, which limit generalizability. We propose a large language model (LLM)-driven framework for multimodal workload prediction. It integrates electroencephalography (EEG), remote photoplethysmography (rPPG), and robotic control signals. Data streams are synchronized, tokenized, and fused through the Mixture of Experts (MoE) mechanism. A neural inference module refines the representation and jointly estimates three indices: Fatigue Level (FL), Attention Level (AL), and Control Precision Level (CPL). Predictions are calibrated with NASA Task Load Index (NASA-TLX) ratings for interpretability. Experiments across tasks of increasing complexity show that the framework outperforms unimodal and multimodal baselines, achieving higher accuracy and robustness. The three indices capture both transient fluctuations and sustained workload transitions. These results highlight the value of combining LLM-based representations with multimodal sensing for scalable and interpretable cognitive workload assessment in industrial robotics.
As the preferred material for plasma-facing components in future fusion test reactors, tungsten plays a critical role in ensuring the safe and stable operation of fusion reactors on the first wall of blankets and divertor targets. This paper aims to explore advanced manufacturing methods for pure tungsten and analyze the feasibility of applying additive manufacturing technology in nuclear fusion. Pure tungsten components were fabricated using powder bed fusion electron beam (PBF-EB), followed by annealing heat treatment in this work. The evolution of microstructure and mechanical properties at different annealing temperatures was investigated. Results revealed a distinct polyhedral equiaxed grain structure, with average grain size initially decreasing and then increasing as annealing temperature rose. Optimal performance was achieved at 1100 degrees C, with a density of 99.5%, Vickers hardness of 406 HV0.3, and compressive strength of 1961 MPa. Compared to untreated specimens, these properties showed substantial improvement. The findings provide guidance for developing properties of other refractory materials and improve the application of additive manufacturing in plasma-faced material fabrication.
