Fluidized bed coal gasification is a key technology for clean coal utilization, yet its inherent nonlinearity and multi-physics coupling challenge traditional modelling approaches. Artificial intelligence (AI) offers a promising pathway to capture these complex relationships from industrial data, enabling accurate prediction and process optimization. Based on 2018 operational samples from a 15MWth KEDA (R) circulating fluidized bed coal gasifier, this study developed and compared four models for gas flow rate prediction: polynomial regression, random forest, DNN, and CNN-DNN hybrid. The CNN-DNN hybrid model achieved the best performance (R 2 = 0.8004, MAPE = 0.5044%) with the most consistent residual variance, and while its accuracy gain over DNN (R 2 = 0.7872) was modest, it reduced trainable parameters by 62.7% and learned relationships consistent with gasification principles. SHAP analysis identified coal feed rate (26.5%), top pressure (22.7%), and air flow rate (15.4%) as the three key parameters, and partial dependence plots informed an optimization strategy of 'stabilizing coal feed, increasing top pressure, and raising air flow appropriately'. This work offers an interpretable data-driven approach for soft sensing and operational optimization of coal gasification processes.
During hydraulic fracturing of unconventional reservoirs, slickwater entering narrow hydraulic fractures undergoes a complex flow evolution from the near-wellbore jet region to a fully developed state in the deep fracture; however, its single-phase flow behaviour within fractures remains insufficiently understood. In this study, a visualized experimental system for flat-plate fractures was established and combined with particle image velocimetry (PIV) to measure the flow field from the near-wellbore region to the distal fracture region. By quantitatively characterizing the velocity and vorticity fields, the effects of injection rate, slickwater viscosity, fracture width, and perforation configuration on in-fracture flow-field development were systematically investigated. The results show that the flow field can be divided into a high-speed jet disturbance zone, a turbulent transition zone, and a fully developed flow zone. As the fluid moves away from the perforations, velocity non-uniformity and vorticity fluctuations continuously decrease, perforation-induced vortices gradually dissipate, and the flow field ultimately stabilizes. Slickwater suppresses spanwise development and mixing and weakens large-scale vortex structures, thereby promoting downstream flow stabilization, with higher-viscosity slickwater exhibiting a more pronounced drag-reduction effect. Higher injection rates, lower viscosity, smaller fracture widths, and single-perforation injection all increase the flow-field stabilization distance. Meanwhile, perforation configuration controls flow development mainly by altering local jet velocity, jet-jet interactions, and jet-wall interactions. These results reveal the spatial evolution of single-phase slickwater flow in narrow hydraulic fractures and provide an experimental basis for investigating proppant-slickwater two-phase transport in complex fractures and optimizing hydraulic-fracturing parameters.
To address the unclear heat transfer mechanisms and lack of systematic parametric studies for supercritical CH4-N2 mixtures in hydrogen liquefaction precooling systems, a numerical investigation using the RNG k-epsilon turbulence model is conducted. The effects of N2 ratio (0%-40%), heat flux (30-60 kW/m2), and mass flux (120-210 kg/(m2 & centerdot; s)) on heat transfer performance are analyzed. Results show that increasing N2 ratio weakens overall heat transfer, with top wall deterioration changing significantly, while bottom wall heat transfer always exceeds that on the top wall. Higher heat flux reduces heat transfer coefficients on both walls and worsens deterioration. Top wall deterioration is dominated by thickening of the thermal resistance layer; bottom wall behaviour reflects competition between buoyancy driven local enhancement and overall attenuation. Increasing mass flux improves coefficients on both walls and suppresses deterioration. Top wall heat transfer is jointly influenced by a low-density fluid and pseudo-critical effects, whereas the bottom wall maintains good performance. Based on simulation data, a concentration parameterized heat transfer correlation is established with prediction errors within +/- 15%. Compared with an existing concentration weighting correlation, the new correlation reduces the mean relative deviation by 9.253%, enabling more accurate prediction. This study provides theoretical basis and data support for optimizing hydrogen liquefaction processes and designing precooling heat exchangers.
Abstract This study examines the combined effects of foaming temperature and carbon fillers on the morphology, porosity, electrical resistivity, and compressive properties of polyvinylidene fluoride (PVDF)‐based composite foams. PVDF/graphite (GR) composites containing 60 wt.% carbon fillers were modified by partially replacing GR with carbon fibres (CF), carbon black (CB), or a CF/CB system. The reference PVDF/GR (40/60 wt.%) composite was foamed using supercritical CO 2 at 210, 220, and 230°C to identify the best foaming condition. Foaming at 210°C yielded limited cell development, attributed to restricted PVDF chain mobility, whereas 230°C promoted pore coalescence and partial structural collapse. The best cellular structure was obtained at 220°C, reaching 20% porosity and in‐plane resistivity of 0.14 Ω · cm. At this temperature, secondary fillers further improved porosity and electrical performance. CF increased porosity to 24% and promoted conductive bridging, while CB increased porosity to 30% and reduced interparticle resistance. The CF/CB system exhibited the strongest synergistic effect, achieving 32% porosity and the lowest in‐plane resistivity of 0.03 Ω · cm. Improved electrical performance was accompanied by reduced compressive properties. Supercritical CO 2 foaming combined with hybrid filler design therefore produces highly conductive PVDF‐based foams, although optimization remains necessary to balance electrical conductivity and mechanical integrity for proton exchange membrane fuel cell (PEMFC) bipolar plates. The novelty lies in elucidating the foaming‐driven redistribution of a graphite‐rich, multiscale CF/CB/graphite network within the PVDF matrix and its coupled effects on porosity, through‐plane and in‐plane resistivities, and compressive behaviour.
Abstract The solid–liquid fluidization system has garnered significant scientific and industrial attention in recent years due to its broad applicability in process engineering. Contemporary global initiatives for the treatment of industrial effluents and wastewater have motivated researchers to explore inverse fluidization as a promising alternative to conventional fluidization techniques. This preference arises from its inherent advantages, including operational simplicity, ease of handling, lower process cost, reduced energy consumption, and enhanced overall efficiency. The present study focuses on the determination of the minimum inverse fluidization velocity using four different polymeric particles, such as low density polyethylene grade 1 (LDPE1), low density polyethylene grade 2 (LDPE2), high density polyethylene (HDPE), and polypropylene (PP), respectively, which are dispersed in a non‐Newtonian dilute solution of sodium salt of carboxymethyl cellulose at four varying concentrations. Empirical correlations have been developed to predict the minimum inverse fluidization velocity as a function of relevant physical and dynamic parameters of the system. A Euler–Euler multiphase model has been employed to discretize the system, solving two sets of Navier–Stokes equations, corresponding continuity equations, and a transport equation. The effects of variations in the concentration of the continuous phase, the initial mass of the dispersed phase, and the particle type on the volume fraction and velocity distribution have been systematically examined. The simulation results provide comprehensive insights into the hydrodynamic behaviour of the inverse fluidization process, offering a deeper understanding of system dynamics and performance characteristics.
