Pyrolysis of thermally-thick biomass particles is a highly complex process involving fluid-particle coupling, anisotropic intra-particle transfer, and detailed pyrolysis reactions in a porous particle. To enable numerical predictions under widely varying operating conditions, a comprehensive biomass pyrolysis model using detailed kinetics coupled with an anisotropic porous media model with resolved fluid-particle coupling is developed. The results show that the detailed model can provide reasonable predictions of biomass pyrolysis in a wide range of conditions, including temperatures ranging from 550 to 1800 K, particle size ranges from 3 mm to 25 mm, moisture contents up to 66.67% (dry basis), various particle shapes, different wood species, as well as radically different flow conditions (natural convection and forced convection flow modes). Intra-particle heat transfer and fluid-particle coupling exert substantial effects on the pyrolysis process, while the influence of anisotropic properties is relatively mild. Sensitivity analysis identifies wood thermal capacity, emissivity, and thermal conductivity as the most critical parameters; three out of 32 reaction rates have significant effects on the modelling results at low-temperature conditions, while all reaction rates show a weak impact on the modelling results at high-temperature conditions, due to a large internal thermal Damko & uml;hler number (Da > 10). These findings provide profound theoretical insights into the pyrolysis mechanisms of thermally thick biomass particles, verifying that the model serves as a powerful tool for predicting the pyrolysis behaviors of thermally thick biomass particles, with direct implications for the design and optimization of biomass pyrolysis reactors.

In response to the critical national demand for upgrading automotive gasoline quality, the concept of dual reaction zones was developed to intensify both olefin generation and conversion. The successful large-scale implementation of this process has yielded substantial economic benefits and spurred the invention and systematic study of the diameter-transformed fluidized bed (DTFB) reactor, leading to a suite of new catalytic processes. This study begins with the conceptual origins of the DTFB reactor. By analyzing unimolecular and bimolecular mechanisms in hydrocarbon catalysis, the key conditions necessary for maximizing target products are identified. Furthermore, it elucidates the scientific and technological challenges in applying diameter variation to partition the reaction section, highlighting that the primary challenge lies in achieving precise coupling between flow and reaction multimodalities, which necessitates a generalized drag model for accurate prediction of flow regime transitions. Since flow structure is influenced by both macroscopic parameters and local dynamics, a two-way coupled energy minimization multi-scale (EMMS) drag model and a corresponding multi-scale computational fluid dynamics (CFD) approach have been proposed, laying a theoretical foundation for quantitative design of diameter-transformed sections. The subsequent development of ancillary technologies has provided the necessary engineering safeguards for flexible control of temperature, density, and gas–solid contact time in each zone, ultimately enabling the industrialization, large-scale operation, and long-term stability of DTFB-based catalytic technology. Finally, the study outlines several typical processes and their application performance, and prospects future work.

The twisted-tape based vortex tube is recognized as a promising reactor for the highly energy-intensive steam cracking process. This study develops a short-length twisted tape, hereinafter termed the diameter-transformed vortex generator (DTVG), to reorganize decaying swirling flow through the coupling of bulk rotating flow and longitudinal vortices. Computational fluid dynamics with the RNG k–ε turbulence model was used to investigate macroscopic transport behavior, vortex structure evolution, sectional thermal–hydraulic performance and entropy generation. Based on vortex evolution, the DTVG vortex tube can be divided into induction, twin vortex and single vortex decay sections. This sectional swirling flow reorganization transforms the induction section into a vortex regulation region rather than a strong mixing region, while shifting the main heat transfer benefit to the twin vortex decay section. There, robust helical longitudinal vortex pairs promote wall-to-core transport, and near-wall shear renews the thermal boundary layer, yielding the highest Nusselt number and performance evaluation criterion values of 306.753 and 1.163, respectively. Entropy generation analysis further provides thermodynamic evidence for this functional transformation: compared with hollow twisted tape and conventional twisted tape, global entropy generation in the DTVG induction section is reduced by 21.7% and 21.3%, and frictional entropy generation by 42.3% and 43.4%, indicating suppressed premature mechanical energy dissipation during vortex initiation.

Although large biomass fuels offer significant potential for deep decarbonization, their use in cement pre-calciners often results in incomplete burnout. Raw meal agglomeration adds further complexity, and together these effects cause reactor instability and inefficiency, as well as downstream operational issues. For the first time, this paper presents an enabling numerical model to address these challenges. For this purpose, a Multi-Fluid Model (MFM) is initially developed to effectively capture the multiphase reactive flows and agglomeration effects in industrial pre-calciners. The model integrates a particle-scale agglomeration mechanism and incorporates the agglomerate sizes into constitutive relations—including granular energy, solid viscosity, drag force, heat and mass transfer, and reactive surface area. Validation against industrial data confirms the model's markedly improved predictive accuracy compared with previous models, and its applicability is further examined under different tertiary air conditions. Based on this, a Sub-Particle-Scale (SPS) model is incorporated to account for intraparticle temperature gradient effects within large biomass particles in industrial systems. After validation using industrial measurements, the integrated model is applied to reveal biomass size effects on incomplete burnout and pre-calciner performance. These new efforts account for fine particle agglomeration and intraparticle temperature gradients in large particles while maintaining suitable computational efficiency, thereby enabling effective industrial-scale simulations. It provides a tool for guiding the operation of biomass-fueled pre-calciners under various conditions, supporting the advancement of low-carbon cement production.

This review systematically presents a combined optimization methodology integrating mesoscale modeling and CFD simulation strategies for gas–solid fluidized catalytic reactors. Taking the diameter‑transformed fluidized bed for clean gasoline production as an illustrative case, the article summarizes recent advances in dynamic structure‑dependent drag, heat transfer, and mass transfer models within the energy-minimization multiscale model framework. Their coupling with reaction kinetics and implementation into the two‑fluid model are discussed, enabling multi‑scale simulations from single units to the full reaction‑regeneration loop. Key findings indicate that mesoscale models accurately predict solid concentration, flow regime transitions, and saturated carrying capacity, allowing quantitative reactor zoning and fast bed regulation. Full‑loop simulations further reveal that non‑uniform mass transfer significantly influences catalyst regeneration behaviors, providing critical heat load data for heat exchanger design and collective regulation. Future integration of high‑fidelity simulation with plant data via machine-learning is highlighted as a promising path toward intelligent industrial reactor operation.

Diesel hydrodesulfurization is a highly exothermic gas-liquid-solid reaction where intraparticle liquid distribution governs overall performance. Conventional methods fail to capture the coupled dynamics of gas-liquid flow, vaporization, and reaction inside a single catalyst particle. We develop a VOF-PMM framework with a capillary-force model to simulate the wetting process under coupled exothermic reaction, vaporization, and mass transfer. Validation against NMR-observed "piston-like" liquid fronts confirms model fidelity under exothermic conditions. Applied to diesel hydrotreating, the results show two regimes. Without vaporization, liquid imbibition is controlled by the balance between capillary force and pore resistance. Spearman analysis identifies the capillary number Ca (viscous-to-capillary ratio) as the dominant inhibitory factor: lower gas velocity or higher surface tension enhances wetting. Under reactive conditions, reaction-induced vaporization significantly reduces internal liquid holdup, producing a "wet-shell/dry-core" structure. Raising the temperature from 593 to 633 K decreases liquid holdup by 15 %, and increasing gas velocity to 3.6 m/s causes a further 16 % reduction. These behaviors arise from the competition between reaction-driven vaporization and capillary-driven imbibition, characterized by the Damko & uml;hler number Davap (the reaction-to-vaporization rate ratio). When Davap >> 1, liquid holdup stabilizes above 0.9; when Davap << 1, sustained "rapid imbibition-slow drainage" cycles occur. This finding provides pore-scale design criteria for catalyst shaping and process optimization.

Traditional swirl vane demisters are widely used in industrial wet flue gas desulfurization (WFGD) systems. However, they suffer from high operational pressure drop, severe wall erosion, and a sharp decline in demisting efficiency at high gas velocities due to the breakup and re-entrainment of wall liquid film under strong gas shear, making it difficult to meet increasingly stringent ultra-low emission requirements. To address these problems, this paper designs a novel guide vane demister with the comprehensive optimization objectives of high efficiency, low resistance, and erosion resistance. Based on the Eulerian-Lagrangian multiphase flow framework, a gas-liquid two-phase flow numerical model is established. The SST k-ω turbulence model is adopted to describe the turbulent gas flow, the Discrete Phase Model (DPM) is used to track droplet trajectories, and the Discrete Random Walk Model (DRWM) is coupled to account for the effect of turbulent fluctuations on droplet dispersion behavior. The droplet motion equation incorporates the coupled effects of drag force, centrifugal force, and gravity. Using the experimental data of a swirl vane demister from the literature as a benchmark, the pressure drop and demisting efficiency under different inlet gas velocities (1.76–3.49 m/s) are compared to validate the accuracy of the numerical model. On this basis, the structural design of the arc-shaped guide vane demister is completed, with key geometric parameters including a cylinder diameter D0=284 mm, a central column diameter Di=142 mm, a vane outlet angle α=45°, and five guide vanes. Three-dimensional numerical simulations are then performed, and a systematic performance comparison with the traditional swirl vane demister is conducted from multiple dimensions, including velocity field, pressure field, turbulent kinetic energy, and droplet concentration distribution.The model validation results show good agreement between the simulated pressure drop, demisting efficiency and the experimental values, with relative errors within 5%, confirming the reliability of the numerical model. Flow field analysis reveals that a stable "Rankine vortex" structure is formed inside the guide vane demister, with the tangential velocity exhibiting a typical "low at the center, high near the wall" distribution. Compared with the traditional swirl vane demister: (1) the peak tangential velocity decreases from approximately 7.5 m/s to approximately 4 m/s, a reduction of about 40%, and the velocity distribution is more uniform with more gradual axial decay, significantly reducing the impact energy of droplets on the wall and the risk of erosion; (2) the operational pressure drop is reduced by 30%–40% across the entire simulated gas velocity range, with substantially decreased flow resistance; (3) the demisting efficiency is consistently higher than that of the traditional swirl vane demister, reaching 100% when the inlet gas velocity exceeds 2.78 m/s, achieving complete droplet removal.Owing to its streamlined geometry, the guide vane demister effectively eliminates the large vortex dead zone on the leeward side of the vanes and reduces local flow resistance, providing a sustained and stable centrifugal force field. While maintaining high-efficiency gas-liquid separation, it significantly reduces operational energy consumption and wall erosion risk, successfully achieving the comprehensive performance optimization characterized by "high demisting efficiency, low operational energy consumption, and low wall erosion risk." The findings of this study reveal the flow field evolution and droplet separation mechanisms of the guide vane demister, providing a reliable theoretical basis for the innovative design and engineering application of demisters for industrial flue gas purification.

Mesoscale structure in the form of clusters in fluidized beds strongly affects the interphase drag, heat/mass transfer, and reaction processes. To address the sub-grid, mesoscale effects in coarse-grid simulations, this study presents a multiscale mass transfer model based on the steady-state Energy-Minimization Multi-Scale (EMMS) approach. The model incorporates four structure-specific, mass balance equations, including mass source and sink terms, to maintain a steady-state mass transfer. By further assuming a steady equilibrium between the mass transfer and reactions, a reaction model is also developed. The resulting mesoscale drag, mass transfer, and reaction models are incorporated into coarse-grid simulations within the two-fluid model (TFM) framework. The simulation cases cover naphthalene sublimation (purely mass transfer), ozone decomposition (reaction-controlled) and methane oxidation (mass-transfer-constrained reaction) processes over a wide range of flow regimes from the bubbling to fast fluidization beds. The results demonstrate a significant improvement in agreement with experimental data compared to conventional homogeneous approaches. These findings highlight the critical importance of mesoscale modeling of mass transfer and reaction, even when the mesoscale effects on flow are adequately accounted for.

Amidst the global shift from petroleum refining to petrochemical production, novel olefin cracking reactors require elevated-temperature, high-density catalysts capable of selective C-C scission while suppressing radical side-reactions at 600-700 degrees C-a critical gap between conventional catalytic and thermal cracking regimes. This study systematically investigates 1-pentene cracking over tailored commercial ZSM-5 catalysts (fresh F-TCC-2; steamed-aged A-TCC-1/A-TCC-2), elucidating cooperative regulation by temperature, acid strength, and confinement effects. Through Delplot analysis, we propose a synergistic mechanism integrating confined catalytic radical pathways with classical carbocations, establishing a particle-scale kinetic model for high-temperature confined thermal catalysis. Nonlinear least-squares regression determined eighteen kinetic parameters. The more acidic A-TCC-2 catalyst exhibited 30-45 % lower apparent activation energies and enhanced adsorption enthalpies versus A-TCC-1, achieving > 85 % combined light olefins (ethylene, propylene, butylene) selectivity at 650 degrees C. These results provide fundamental kinetics for industrial reactor optimization and demonstrate confinement's pivotal role in governing cracking pathway selectivity.

Offshore Integrated Energy Hubs (OIEHs) are emerging as an important infrastructure for offshore renewable energy integration, hydrogen production, and multi-energy coordination, but their development is accompanied by significant technical, economic, and operational risks. To support systematic risk evaluation, this study proposes a risk assessment framework that combines Triangular Intuitionistic Fuzzy Numbers (TIFNs), the Entropy Weight Method (EWM), Grey Relational Analysis (GRA), and the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS). Unlike existing studies that mainly focus on system optimization or project feasibility, this study explicitly incorporates expert hesitation through TIFNs and integrates EWM with GRATOPSIS to improve the robustness and discriminatory power of OIEH risk assessment under data-scarce conditions. The results show that OIEHs are particularly sensitive to risks associated with system integration complexity, cable damage, and cost overruns. The calculated overall risk value is 0.63, indicating a relatively high-risk level. These findings provide useful support for risk-informed planning, implementation, and management of next-generation offshore integrated energy systems.

Pyrolysis of thermally thick biomass particles holds great potential for reducing carbon footprint in various industries. However, the intraparticle temperature gradient greatly influences the heating and conversion, and ultimately, reactor performance. This study presents a novel sub-particle scale (SPS) model that comprehensively incorporates the intraparticle temperature gradient effects on reaction kinetics, heat transfer, and material properties. The SPS model is first validated against experimental data and a spatially resolved model for singleparticle pyrolysis. Further analysis demonstrates that fully incorporating all correction factors is essential, and higher Biot number amplifies the variation of these factors. The SPS model integrated into a Eulerian multifluid model (SPS-MFM) is used for simulating a fluidized bed pyrolyzer. While it does not explicitly resolve the subparticle details, SPS-MFM effectively incorporates the impacts of intraparticle temperature gradient in pyrolyzer simulations. The results accurately capture the conversion behavior and show higher predictability than conventional multifluid models. The intraparticle temperature gradient causes slower temperature increases and delayed conversion, whereas increasing fluidizing gas velocity enhances interphase heat transfer, thereby reducing these effects. This work provides a reliable and efficient approach for understanding industrial-scale biomass pyrolyzers.

Regulatory authorities face challenges in efficiently extracting essential information from the extensive corporate business data distributed across various locations. This study proposes an enterprise hierarchical portrait scheme utilizing a deep learning model combining BERT and a bi-directional long and short-term memory network (BiLSTM) to address this issue. The process involves data preprocessing, utilizing BERT and BiLSTM for feature extraction to model text sequences and extract relevant semantic information, and generating hierarchical labels through a full connectivity layer to create a comprehensive company portrait. This approach aims to provide authorities with a framework for categorizing businesses effectively.

Mass transfer in gas-solid fluidized bed reactors, which links momentum transfer and reactions, plays a critical role in reaction regulation. Dynamic structure-based mass transfer models provide a more accurate depiction of the reaction process than steady-state models by considering transient variables, despite the increased computational challenges. This study develops a solution method for the EMMS-based dynamic mass transfer model and validates it through simulating ozone decomposition process in both turbulent and fast fluidized beds. Further sensitivity analysis of the model parameters indicates that mass transfer between the dilute and dense phases is a key factor affecting model accuracy. Then, the Random Forest algorithm is used to explore the relationship between various variables and their effects on reactions, finding that the solid concentration is the most influential factor in both reactors, followed by pressure for the turbulent fluidized and ozone concentration for the fast fluidized bed.

To investigate the mesoscale structure characteristics of the middle-Stokes-number particle-laden jet (MSPJ) and the evolution law of scale-independent particle volume fraction, as well as to assess the applicability of a particle volume fraction model based on the self-similarity theory to MSPJ, we have established a specialized laser-camera measurement setup and conducted six groups of experiments. The characteristic scale and spatial scale-independent particle concentration are analyzed using the Voronoï method and the scale-independent particle cluster characterization method. The results indicate that the particle clusters in MSPJ exhibit a dynamic stability. As the Stokes number decreases from a large value to 1, the preferential concentration of particles becomes more pronounced. The long axis of the particle cluster serves as the primary characteristic size. Larger clusters display more irregular shapes, while smaller clusters appear to be circular and adhere to the power-law distribution described by the osmosis theory. The particle volume fraction of the jet reveals a distribution pattern characterized by a higher concentration at the center and lower concentrations at the edges. Along the jet centerline, the particle concentration initially decreases and then increases as one moves away from the nozzle. This phenomenon is attributed to the decay of particle velocity at the jet’s far end, where high-velocity particles from upstream catch up with lower-velocity particles downstream, leading to gradual accumulation. Validation of the particle volume fraction model demonstrates that the value of ηCθ3 is influenced by turbulence. The particle distributions and clusters of different jets are self-similar. These findings provide a valuable foundation for enhancing the performance and efficiency of jet combustion systems.

In a dilute particle-laden jet, the drag force is the most important factor determining the momentum exchange between the gas and particles. In this work, different drag correlations are used to predict the jet velocity and compared with experimental data. In addition to our previously reported large-particle, high-Stokes-number jet case, the experimental data of small-particle, intermediate-Stokes-number jet are obtained with particle image velocimetry. The comparison shows that the prediction is sensitive to the choice of drag correlation. The LWL-E (Li & Wang & Li extrapolation) drag model derived from large-particle experiments is more suitable for predicting the velocity of large particles with high Stokes number, while the Rudinger-E model derived from small particle experiments is more suitable for the simulation of particle velocity of small particles with intermediate Stokes number. The standard drag model and the Gidaspow model overpredict slip velocity and underpredict particle velocity, and are not suitable for the particle velocity prediction in a dilute particle-laden jet.

Scale-up has always been the bottleneck to the development of new industrial processes. This study aims to assess the scale-up effects of a diameter-transformed fluidized bed (DTFB) reactor through three-dimensional, multi-phase particle-in-cell (MP-PIC) simulation with the Energy-Minimization Multi-Scale (EMMS) drag and solid stress model. Four 3.5 Mt/a DTFB reactors are designed by scaling up a 1.2 Mt/a one with different scale-up schemes and simulated after validation. It is found the Glicksman’s rule shows the most similarity in solid concentration distribution to the benchmark case while the FixedOperation rule under-predicts the solid concentration, meaning that only keeping constant Ug and Gs cannot guarantee the same distribution of solid concentration when scaling up the fast fluidized bed. In addition, all four scale-up designs ensure the same gas velocity, yet they exhibit varying solid velocities throughout the scale-up process. A more rational scale-up rule is required for the elaborate reactor scale-up

This study presents a comprehensive, transient, 3-D Computational Fluid Dynamics (CFD) model of a multi-cylinder flat engine's lubrication system, simulated using Simerics-MP+. Engine lubrication system is crucial for reducing friction, cooling, and cleaning engine components. Understanding its performance is essential for optimal engine operation. The model was applied to the lubrication system of a 4-cylinder, reciprocating internal combustion engine. The computational domain includes the positive displacement gerotor pump, pressure regulation valve, bearings, piston cooling jets, oil cooler, oil filter, and other relevant components. The gerotor pump and gallery bearings were modeled using real 3-D geometries without any assumptions. Bearing deformation and orbiting due to force imbalance were prescribed in the simulations. The simulation was conducted at 5000 rpm of engine speed. The simulated flow pressure distributions closely matched experimental data.

Dynamic vehicle operation, such as acceleration, deceleration, and tilting, can cause severe oil sloshing in the engine oil pan. This can lead to oil starvation at the pickup tube, compromising lubrication pump performance, and potentially damaging engine components. This study presents a Computational Fluid Dynamics (CFD) multiphase model of an engine oil pan and a system of lubrication pumps, simulated using Simerics-MP+®. A series of numerical simulations are conducted at a given pump speed and extreme oil pan tilt angles or accelerations relevant to a high performance vehicle. Time-dependent oil distributions are visualized, and real-time oil flow rates are monitored at the pickup tubes to assess the impact of oil dynamics and pan position on pick-up tube starvation. This CFD model provides valuable insights into oil pan and pump behavior under extreme vehicle operation conditions, aiding in the design and optimization of lubrication systems to mitigate the risk of oil starvation and improve overall engine safety and performance.

This study pioneers a three-dimensional, transient reactive simulation of an industrial fluid catalytic cracking full-loop system. Within a two-fluid model framework, the simulation incorporates the Energy Minimization Multiscale (EMMS)-based models to account for the effects of mesoscale flow structures on drag and heat transfer, and integrates a 12-lumped kinetics model and a coke combustion model to describe catalytic cracking reactions and catalyst regeneration, respectively. It finds the significant impact of reactions on solid concentration and gas velocity distributions throughout the system, particularly in the first reaction zone. The first reaction zone achieves 80% conversion of feedstock oil, with the second reaction zone contributing an additional 19% conversion. These variations in product concentration along the bed height reflect substantial differences in reaction types under varying environments. Furthermore, the simulation captures temperature changes along the solid circulation path, facilitating the determination of the heat exchanger power required to control the reaction temperature.