Pub Date : 2025-12-03DOI: 10.1016/j.cherd.2025.12.004
Haifeng Gong , Yaozhong Hu , Ye Peng , Lin Yan , Bao Yu , Hong Yin , Ping Ouyang
In order to further improve the demulsification efficiency of electrostatic demulsification, a chaotic-pulse-width-modulation (CPWM) electric field is proposed for demulsification separation of emulsion oil. Due to the vibration deformation behavior of emulsion oil droplets with different sizes under the action of CPWM field is not clear, this work establishes the deformation dynamics model of oil droplets in CPWM electric field, and analyzes the deformation characteristics of droplets with different sizes. The results show that with the increase of droplet size, the lag degree of droplet deformation is strengthened, which leads to the decrease of droplet vibration times in CPWM electric field. The CPWM electric field can make droplets of different sizes have the opportunity to vibrate and deform at their own resonance frequency. At this time, the droplet achieves the best deformation, and the vibration amplitude is close to the resonance amplitude. However, droplets of different sizes will experience several pulses with similar droplet resonance frequency in CPWM electric field. There will be a slight gap in the amplitude under different pulses. The gap in amplitude is related to the state of the droplet at the end of the previous pulse, that is, the smaller the oil flow rate of the droplet at the end of the previous pulse, the greater the amplitude under this pulse. The degree of chaotic vibration of the droplet in the CPWM electric field decreases with the increase of droplet size, indicating that the disorder of droplet vibration decreases.
{"title":"Analysis of deformation and vibration characteristics of droplets with different sizes in emulsified oil subjected chaotic pulse width modulation electric field","authors":"Haifeng Gong , Yaozhong Hu , Ye Peng , Lin Yan , Bao Yu , Hong Yin , Ping Ouyang","doi":"10.1016/j.cherd.2025.12.004","DOIUrl":"10.1016/j.cherd.2025.12.004","url":null,"abstract":"<div><div>In order to further improve the demulsification efficiency of electrostatic demulsification, a chaotic-pulse-width-modulation (CPWM) electric field is proposed for demulsification separation of emulsion oil. Due to the vibration deformation behavior of emulsion oil droplets with different sizes under the action of CPWM field is not clear, this work establishes the deformation dynamics model of oil droplets in CPWM electric field, and analyzes the deformation characteristics of droplets with different sizes. The results show that with the increase of droplet size, the lag degree of droplet deformation is strengthened, which leads to the decrease of droplet vibration times in CPWM electric field. The CPWM electric field can make droplets of different sizes have the opportunity to vibrate and deform at their own resonance frequency. At this time, the droplet achieves the best deformation, and the vibration amplitude is close to the resonance amplitude. However, droplets of different sizes will experience several pulses with similar droplet resonance frequency in CPWM electric field. There will be a slight gap in the amplitude under different pulses. The gap in amplitude is related to the state of the droplet at the end of the previous pulse, that is, the smaller the oil flow rate of the droplet at the end of the previous pulse, the greater the amplitude under this pulse. The degree of chaotic vibration of the droplet in the CPWM electric field decreases with the increase of droplet size, indicating that the disorder of droplet vibration decreases.</div></div>","PeriodicalId":10019,"journal":{"name":"Chemical Engineering Research & Design","volume":"225 ","pages":"Pages 48-58"},"PeriodicalIF":3.9,"publicationDate":"2025-12-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145693078","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-03DOI: 10.1016/j.cherd.2025.12.002
Qingqing Liu , Georges Chahine , Chao-Tsung Hsiao , Jason Hartwig
The separation of liquid and vapor plays a crucial role in various engineering applications, especially in the use of cryogenic liquids for space exploration. The DynaSwirl® (DS) phase separator is a passive rotation-based phase separator, which provides an effective solution for achieving efficient separation in microgravity environments. The DS phase separator uses multiple elongated tangential injection inlets into the separation chamber, which enable stable operation and allow high tangential velocities and low pressure at the core of the generated vortex leading to effective phase separation. This research delves into the fundamental physics that governs the operation of these systems. A DS phase separator is analyzed through both numerical simulations and experimental methods. The DS phase separator employs centripetal and centrifugal forces generated within a swirl chamber to facilitate separation. The effectiveness of separation improves with increasing swirl strength; however, this enhancement leads to greater pressure losses throughout the system. This paper identifies and analyzes the various flow and liquid characteristics that influence pressure losses. These factors include the physical properties of the working liquid, such as density and viscosity, and the swirl strength, which is influenced by viscosity. Additionally, the study examines the impact of incorporating a vortex blocker/killer (BK) attachment on these losses. To conduct the investigation, we examine a set of six selected liquids of interest to NASA, which includes four cryogenic fluids: LN2, LH2, LOX, and LCH4, along with water and a water/glycerin mixture. Additionally, two sets of six fictitious liquids, whose properties are derived from these, are analyzed to isolate the effects of the liquid properties. The pressure losses primarily occur in three areas of the DS phase separator: 1) the inlet slots, 2) the exit orifice, and 3) the swirling flow within the chamber. For the BK configuration, the pressure loss caused by the swirl accounts for about 70 % of the total pressure loss from the wall to the exit when the Reynolds number exceeds 24,000, whereas without BK, this contribution is approximately 90 %. This highlights the effectiveness of the BK in reducing swirl-induced pressure losses. For a given flow rate, reducing viscosity is shown to decrease friction pressure losses but also to increase swirl pressure losses due to a strengthened vortex. In contrast, decreasing the flow rate and density results in reduced pressure losses across all three components of the system. This paper provides a comprehensive analysis of the flow and liquid characteristics, offering valuable insights for the design and optimization of devices that involve strong swirl flows.
{"title":"Effects of liquid properties on pressure loss in a passive rotation-based phase separator","authors":"Qingqing Liu , Georges Chahine , Chao-Tsung Hsiao , Jason Hartwig","doi":"10.1016/j.cherd.2025.12.002","DOIUrl":"10.1016/j.cherd.2025.12.002","url":null,"abstract":"<div><div>The separation of liquid and vapor plays a crucial role in various engineering applications, especially in the use of cryogenic liquids for space exploration. The <span>DynaSwirl</span>® (DS) phase separator is a passive rotation-based phase separator, which provides an effective solution for achieving efficient separation in microgravity environments. The DS phase separator uses multiple elongated tangential injection inlets into the separation chamber, which enable stable operation and allow high tangential velocities and low pressure at the core of the generated vortex leading to effective phase separation. This research delves into the fundamental physics that governs the operation of these systems. A DS phase separator is analyzed through both numerical simulations and experimental methods. The DS phase separator employs centripetal and centrifugal forces generated within a swirl chamber to facilitate separation. The effectiveness of separation improves with increasing swirl strength; however, this enhancement leads to greater pressure losses throughout the system. This paper identifies and analyzes the various flow and liquid characteristics that influence pressure losses. These factors include the physical properties of the working liquid, such as density and viscosity, and the swirl strength, which is influenced by viscosity. Additionally, the study examines the impact of incorporating a vortex blocker/killer (BK) attachment on these losses. To conduct the investigation, we examine a set of six selected liquids of interest to NASA, which includes four cryogenic fluids: LN<sub>2</sub>, LH<sub>2</sub>, LOX, and LCH<sub>4</sub>, along with water and a water/glycerin mixture. Additionally, two sets of six fictitious liquids, whose properties are derived from these, are analyzed to isolate the effects of the liquid properties. The pressure losses primarily occur in three areas of the DS phase separator: 1) the inlet slots, 2) the exit orifice, and 3) the swirling flow within the chamber. For the BK configuration, the pressure loss caused by the swirl accounts for about 70 % of the total pressure loss from the wall to the exit when the Reynolds number exceeds 24,000, whereas without BK, this contribution is approximately 90 %. This highlights the effectiveness of the BK in reducing swirl-induced pressure losses. For a given flow rate, reducing viscosity is shown to decrease friction pressure losses but also to increase swirl pressure losses due to a strengthened vortex. In contrast, decreasing the flow rate and density results in reduced pressure losses across all three components of the system. This paper provides a comprehensive analysis of the flow and liquid characteristics, offering valuable insights for the design and optimization of devices that involve strong swirl flows.</div></div>","PeriodicalId":10019,"journal":{"name":"Chemical Engineering Research & Design","volume":"225 ","pages":"Pages 85-101"},"PeriodicalIF":3.9,"publicationDate":"2025-12-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145693080","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-03DOI: 10.1016/j.cherd.2025.12.001
Zhiqiang Ma , Lu Wang , Pan Zhang , Fei Gao , Guanghui Chen , Jipeng Dong , Jianlong Li
Bubble reactors are widely employed in biochemical and chemical processes due to their simple design, excellent heat- and mass-transfer characteristics, and efficient mixing performance. Bubble size and distribution are critical parameters that directly affect the gas-liquid interfacial area and mixing efficiency. In this study, visualization experiments of gas-liquid two-phase flow were conducted to investigate bubble dynamics and motion behaviors in a bubble reactor. Digital image analysis (DIA) was used to quantify the evolution of bubble size and distribution under varying inlet apparent gas-liquid velocities (UG = 1.89–9.44 cm/s; UL = 0.94–2.83 cm/s). The results indicate that, as the gas-liquid velocity increases, the average bubble diameter decreases by approximately 25–40 %, and the peak bubble-occurrence frequency shifts from 4.0 mm to about 2.0 mm. Increasing liquid velocity enhances turbulent kinetic energy, promoting bubble breakup and inhibiting coalescence, thereby leading to a narrower size distribution and the formation of a characteristic feather-like plume pattern. The apparent gas velocity exhibits a linear correlation with the internal flow velocity, whereas the apparent liquid velocity follows an exponential relationship. The maximum stable bubble diameter observed in this system is approximately 50.0 mm, which agrees with theoretical predictions based on the capillary-length criterion. Furthermore, a new empirical correlation incorporating the gas-liquid velocity ratio (UG/UL) was proposed, achieving a coefficient of determination R² = 0.989. These findings provide both theoretical and practical guidance for the design optimization and operational enhancement of bubble reactors.
{"title":"Influence of inlet apparent gas and liquid velocities on the dynamic behavior of bubbles in a bubble reactor","authors":"Zhiqiang Ma , Lu Wang , Pan Zhang , Fei Gao , Guanghui Chen , Jipeng Dong , Jianlong Li","doi":"10.1016/j.cherd.2025.12.001","DOIUrl":"10.1016/j.cherd.2025.12.001","url":null,"abstract":"<div><div>Bubble reactors are widely employed in biochemical and chemical processes due to their simple design, excellent heat- and mass-transfer characteristics, and efficient mixing performance. Bubble size and distribution are critical parameters that directly affect the gas-liquid interfacial area and mixing efficiency. In this study, visualization experiments of gas-liquid two-phase flow were conducted to investigate bubble dynamics and motion behaviors in a bubble reactor. Digital image analysis (DIA) was used to quantify the evolution of bubble size and distribution under varying inlet apparent gas-liquid velocities (<em>U</em><sub>G</sub> = 1.89–9.44 cm/s; <em>U</em><sub>L</sub> = 0.94–2.83 cm/s). The results indicate that, as the gas-liquid velocity increases, the average bubble diameter decreases by approximately 25–40 %, and the peak bubble-occurrence frequency shifts from 4.0 mm to about 2.0 mm. Increasing liquid velocity enhances turbulent kinetic energy, promoting bubble breakup and inhibiting coalescence, thereby leading to a narrower size distribution and the formation of a characteristic feather-like plume pattern. The apparent gas velocity exhibits a linear correlation with the internal flow velocity, whereas the apparent liquid velocity follows an exponential relationship. The maximum stable bubble diameter observed in this system is approximately 50.0 mm, which agrees with theoretical predictions based on the capillary-length criterion. Furthermore, a new empirical correlation incorporating the gas-liquid velocity ratio (<em>U</em><sub>G</sub>/<em>U</em><sub>L</sub>) was proposed, achieving a coefficient of determination <em>R</em>² = 0.989. These findings provide both theoretical and practical guidance for the design optimization and operational enhancement of bubble reactors.</div></div>","PeriodicalId":10019,"journal":{"name":"Chemical Engineering Research & Design","volume":"225 ","pages":"Pages 36-47"},"PeriodicalIF":3.9,"publicationDate":"2025-12-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145693077","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The phase change heat and mass transfer of gas-liquid two-phase evaporation in cube has always been the research focus of falling film cooling. Especially the gas-liquid heat transfer process under high temperature and high airflow velocity conditions. In order to study the heat and mass transfer process of falling film evaporation in tube under high liquid film Reynolds number conditions, cooling falling film model in the tube was established using fluent software. Study the influence of internal platform height and liquid film Reynolds number of cooling ring on the vertical annular cooling falling film distribution, and through user define function(UDF) to study the gas-liquid two-phase evaporation phase change heat transfer process. The results indicate the higher platform height leads to larger liquid film thickness, the average liquid film velocity is decrease, and turbulence intensity is increase. As the width of the crevice increases, the thickness of the liquid film also increases, but when the width of the gap exceeds 3.5 mm, the thickness of the liquid film decreases. The flow velocity and turbulence intensity of the liquid film decrease with the increase of crevice width. The thickness and flow velocity of the liquid film increase with Reynolds number, but the amplitude of the increase gradually decreases, and the turbulence intensity of liquid film decreases. The higher gas temperature increases the cross-sectional temperature and axis temperature of the gas-liquid two-phase flow, and vapor content increases. The lower inlet water temperature, the lower cross-sectional temperature and axis temperature, and the greater heat transfer coefficient(HTC)between gas-liquid phase. The maximum HTC is about 724 W/(m2·K). The maximum deviation of HTC is about 17.83 %, and the minimum deviation is 12.08 %.The vapor content increases with the decrease of inlet water temperature. The HTC and vapor content between gas and liquid increase with the increase of liquid film Reynolds number. The maximum deviation between the calculated HTC and the reference experimental value is 15.14 %, and the minimum deviation is 12.62 %. The dimensionless HTC is directly proportional to the Reynolds number of the liquid film and the dimensionless temperature, and the fitting curve deviation R2= 0.984.
{"title":"Study on falling film in large-diameter tube and co-current gas-liquid heat transfer under high temperature and velocity of gas flow","authors":"Liang Wang, Xuan-xuan Huang, Yi-fei Wang, Guang-suo Yu","doi":"10.1016/j.cherd.2025.11.034","DOIUrl":"10.1016/j.cherd.2025.11.034","url":null,"abstract":"<div><div>The phase change heat and mass transfer of gas-liquid two-phase evaporation in cube has always been the research focus of falling film cooling. Especially the gas-liquid heat transfer process under high temperature and high airflow velocity conditions. In order to study the heat and mass transfer process of falling film evaporation in tube under high liquid film Reynolds number conditions, cooling falling film model in the tube was established using fluent software. Study the influence of internal platform height and liquid film Reynolds number of cooling ring on the vertical annular cooling falling film distribution, and through user define function(UDF) to study the gas-liquid two-phase evaporation phase change heat transfer process. The results indicate the higher platform height leads to larger liquid film thickness, the average liquid film velocity is decrease, and turbulence intensity is increase. As the width of the crevice increases, the thickness of the liquid film also increases, but when the width of the gap exceeds 3.5 mm, the thickness of the liquid film decreases. The flow velocity and turbulence intensity of the liquid film decrease with the increase of crevice width. The thickness and flow velocity of the liquid film increase with Reynolds number, but the amplitude of the increase gradually decreases, and the turbulence intensity of liquid film decreases. The higher gas temperature increases the cross-sectional temperature and axis temperature of the gas-liquid two-phase flow, and vapor content increases. The lower inlet water temperature, the lower cross-sectional temperature and axis temperature, and the greater heat transfer coefficient(HTC)between gas-liquid phase. The maximum HTC is about 724 W/(m<sup>2</sup>·K). The maximum deviation of HTC is about 17.83 %, and the minimum deviation is 12.08 %.The vapor content increases with the decrease of inlet water temperature. The HTC and vapor content between gas and liquid increase with the increase of liquid film Reynolds number. The maximum deviation between the calculated HTC and the reference experimental value is 15.14 %, and the minimum deviation is 12.62 %. The dimensionless HTC is directly proportional to the Reynolds number of the liquid film and the dimensionless temperature, and the fitting curve deviation <em>R</em><sup>2</sup>= 0.984.</div></div>","PeriodicalId":10019,"journal":{"name":"Chemical Engineering Research & Design","volume":"225 ","pages":"Pages 230-244"},"PeriodicalIF":3.9,"publicationDate":"2025-12-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145734667","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-01DOI: 10.1016/j.cherd.2025.11.025
Diana K. Díaz-Cervantes, Friné López-Medina, Eduardo López-López, Dulce Y. Medina-Velázquez, Fernando Pérez-Villaseñor, Elsa H. Fernández-Martínez, Arturo Elías-Domínguez
This study presents a generalized kinetic model validated through three industrial case studies performed under varying operating conditions for the gas-phase isomerization of light naphtha (C5 and C6 series). The proposed model comprises 29 chemical reactions, fewer than those in the reference models, including reversible isomerization reactions for C4–C6 species, hydrocracking reactions involving C4–C7 hydrocarbons, hydrogenation coupled with ring opening, and benzene saturation reactions. Initially, both kinetic parameters (activation energies and pre-exponential factors) were conventionally adjusted. However, this work introduces a simplification strategy in which the activation energies are fixed, and only the frequency factors are fitted. This approach proved effective in accurately capturing the system's behavior, indicating that such simplification is applicable to gas-phase processes with feed compositions similar to those studied. Across the three industrial scenarios, the proposed model exhibited significantly lower prediction errors compared to existing models. When both kinetic parameters were adjusted, the model achieved an average prediction error of 1.41 %, while the corresponding reference models reported an average error of 8.80 %. Even when only the frequency factors were calibrated (with fixed activation energies), the model maintained a respectable average error of 2.24 %, still outperforming the reference approaches. Additionally, analysis of the final reactor outlet composition for each case revealed that the systems operated under non-equilibrium conditions. The model also accurately reproduced equilibrium conditions, further confirming its robustness. These findings suggest that the proposed model reliably predicts system behavior both far from and at chemical equilibrium.
{"title":"Generalized kinetic model of the C6 series isomerization process","authors":"Diana K. Díaz-Cervantes, Friné López-Medina, Eduardo López-López, Dulce Y. Medina-Velázquez, Fernando Pérez-Villaseñor, Elsa H. Fernández-Martínez, Arturo Elías-Domínguez","doi":"10.1016/j.cherd.2025.11.025","DOIUrl":"10.1016/j.cherd.2025.11.025","url":null,"abstract":"<div><div>This study presents a generalized kinetic model validated through three industrial case studies performed under varying operating conditions for the gas-phase isomerization of light naphtha (C<sub>5</sub> and C<sub>6</sub> series). The proposed model comprises 29 chemical reactions, fewer than those in the reference models, including reversible isomerization reactions for C<sub>4</sub>–C<sub>6</sub> species, hydrocracking reactions involving C<sub>4</sub>–C<sub>7</sub> hydrocarbons, hydrogenation coupled with ring opening, and benzene saturation reactions. Initially, both kinetic parameters (activation energies and pre-exponential factors) were conventionally adjusted. However, this work introduces a simplification strategy in which the activation energies are fixed, and only the frequency factors are fitted. This approach proved effective in accurately capturing the system's behavior, indicating that such simplification is applicable to gas-phase processes with feed compositions similar to those studied. Across the three industrial scenarios, the proposed model exhibited significantly lower prediction errors compared to existing models. When both kinetic parameters were adjusted, the model achieved an average prediction error of 1.41 %, while the corresponding reference models reported an average error of 8.80 %. Even when only the frequency factors were calibrated (with fixed activation energies), the model maintained a respectable average error of 2.24 %, still outperforming the reference approaches. Additionally, analysis of the final reactor outlet composition for each case revealed that the systems operated under non-equilibrium conditions. The model also accurately reproduced equilibrium conditions, further confirming its robustness. These findings suggest that the proposed model reliably predicts system behavior both far from and at chemical equilibrium.</div></div>","PeriodicalId":10019,"journal":{"name":"Chemical Engineering Research & Design","volume":"224 ","pages":"Pages 386-395"},"PeriodicalIF":3.9,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145614965","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-01DOI: 10.1016/j.cherd.2025.11.020
Kimiya Ramezani, Cláudio P. Fonte, Ashwin Kumar Rajagopalan
Accurate prediction of particle size distribution (PSD) is critical for optimal design of crystallization processes. The interplay between crystallization kinetics and local hydrodynamics complicates predictive process modeling. Computational fluid dynamics (CFD) provides hydrodynamic characterization of the system, which when integrated with population balance equation (PBE) models, leads to a comprehensive representation of the system. Here, we develop a CFD–PBE framework using a compartmental modeling approach, and conduct a systematic evaluation of the impact of hydrodynamics on crystallization processes. CFD simulations of a batch crystallizer provide mean flow field and its turbulence properties, which are mapped onto a coarse-grained structure of well-mixed compartments. PBEs are solved over the compartments, capturing mixing effects on the evolution of the PSD, obtained within few minutes (compared to a few seconds with a perfectly-mixed assumption). Crystallization case studies are conducted to examine mixing effects under varying kinetics, operating conditions, and impeller types. Results show that mixing-driven phenomena are system-specific, and hydrodynamic effects must be considered when transferring insights between systems and scales. The framework is system and geometry-agnostic, and computationally efficient, providing a generalizable approach for scale-up, offering a practical alternative to models that assume perfect mixing.
{"title":"Toward predictive modeling of industrial crystallizers: A compartmental framework for coupling hydrodynamics and population dynamics","authors":"Kimiya Ramezani, Cláudio P. Fonte, Ashwin Kumar Rajagopalan","doi":"10.1016/j.cherd.2025.11.020","DOIUrl":"10.1016/j.cherd.2025.11.020","url":null,"abstract":"<div><div>Accurate prediction of particle size distribution (PSD) is critical for optimal design of crystallization processes. The interplay between crystallization kinetics and local hydrodynamics complicates predictive process modeling. Computational fluid dynamics (CFD) provides hydrodynamic characterization of the system, which when integrated with population balance equation (PBE) models, leads to a comprehensive representation of the system. Here, we develop a CFD–PBE framework using a compartmental modeling approach, and conduct a systematic evaluation of the impact of hydrodynamics on crystallization processes. CFD simulations of a batch crystallizer provide mean flow field and its turbulence properties, which are mapped onto a coarse-grained structure of well-mixed compartments. PBEs are solved over the compartments, capturing mixing effects on the evolution of the PSD, obtained within few minutes (compared to a few seconds with a perfectly-mixed assumption). Crystallization case studies are conducted to examine mixing effects under varying kinetics, operating conditions, and impeller types. Results show that mixing-driven phenomena are system-specific, and hydrodynamic effects must be considered when transferring insights between systems and scales. The framework is system and geometry-agnostic, and computationally efficient, providing a generalizable approach for scale-up, offering a practical alternative to models that assume perfect mixing.</div></div>","PeriodicalId":10019,"journal":{"name":"Chemical Engineering Research & Design","volume":"224 ","pages":"Pages 452-466"},"PeriodicalIF":3.9,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145614959","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-01DOI: 10.1016/j.cherd.2025.11.047
Jianing Liang , Rulong Ma , Xu Yan , Zhenhua Hao , Pei Wang , Yongchun Shu , Jilin He
The production of low-oxygen fine titanium powder is crucial for powder metallurgy but remains challenging due to severe sintering during high-temperature deoxidation. In this study, we address this issue by introducing MgO as a novel dispersant in the hydrogen-assisted magnesium reduction (HAMR) process for deoxidizing fine TiH2 powder. The effects of MgO addition on the oxygen content, powder morphology, and chemical state were systematically investigated. The results demonstrate that MgO effectively inhibits particle agglomeration and sintering by forming nanoscale inert barriers between TiH2 particles. Consequently, well-dispersed deoxidized powder was obtained. The oxygen content of the deoxidized powder exhibits a non-monotonic dependence on MgO addition, initially decreasing and then increasing. Surface chemistry analysis reveals the partial reduction of the native oxide film, with a shift of Ti species from higher (Ti4+, Ti3+) to lower valence states (Ti2+, Ti). Under optimized conditions (750°C, 9 h, mass ratios of TiH2:Mg:MgCl2:MgO = 1:0.18:0.3:0.2), fine TiH2 powder with good dispersibility and a low oxygen content of 0.0831 wt% was successfully obtained. This work provides a practical strategy for producing low-oxygen fine titanium powder from cost-effective precursors.
{"title":"MgO-assisted deoxygenation enabling the production of low-oxygen TiH2 powder","authors":"Jianing Liang , Rulong Ma , Xu Yan , Zhenhua Hao , Pei Wang , Yongchun Shu , Jilin He","doi":"10.1016/j.cherd.2025.11.047","DOIUrl":"10.1016/j.cherd.2025.11.047","url":null,"abstract":"<div><div>The production of low-oxygen fine titanium powder is crucial for powder metallurgy but remains challenging due to severe sintering during high-temperature deoxidation. In this study, we address this issue by introducing MgO as a novel dispersant in the hydrogen-assisted magnesium reduction (HAMR) process for deoxidizing fine TiH<sub>2</sub> powder. The effects of MgO addition on the oxygen content, powder morphology, and chemical state were systematically investigated. The results demonstrate that MgO effectively inhibits particle agglomeration and sintering by forming nanoscale inert barriers between TiH<sub>2</sub> particles. Consequently, well-dispersed deoxidized powder was obtained. The oxygen content of the deoxidized powder exhibits a non-monotonic dependence on MgO addition, initially decreasing and then increasing. Surface chemistry analysis reveals the partial reduction of the native oxide film, with a shift of Ti species from higher (Ti<sup>4+</sup>, Ti<sup>3+</sup>) to lower valence states (Ti<sup>2+</sup>, Ti). Under optimized conditions (750°C, 9 h, mass ratios of TiH<sub>2</sub>:Mg:MgCl<sub>2</sub>:MgO = 1:0.18:0.3:0.2), fine TiH<sub>2</sub> powder with good dispersibility and a low oxygen content of 0.0831 wt% was successfully obtained. This work provides a practical strategy for producing low-oxygen fine titanium powder from cost-effective precursors.</div></div>","PeriodicalId":10019,"journal":{"name":"Chemical Engineering Research & Design","volume":"225 ","pages":"Pages 12-19"},"PeriodicalIF":3.9,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145651717","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-01DOI: 10.1016/j.cherd.2025.11.030
Syahira Ibrahim , Norhaliza Abdul Wahab , Zakariah Yusuf
Accurate predictive modeling is vital for optimizing membrane filtration processes in wastewater treatment. This paper proposes a hybrid model that integrates an artificial neural network with response surface methodology to predict and simulate permeate flux in a submerged membrane bioreactor treating palm oil mill effluent. The framework integrates response surface methodology in both the experimental design and hyperparameter optimization stages, enabling systematic and efficient development of two neural architectures: feed-forward and radial basis function neural networks. Model performance was benchmarked against the conventional one-variable-at-a-time method and evaluated using independent training and testing datasets. The hybrid approach achieved high predictive accuracy, with coefficients of determination ranging from 0.9142 to 0.9981, while reducing computational time by nearly 49 % and the number of optimization iterations by approximately 50–64 %. The Levenberg–Marquardt training algorithm combined with the rectified linear unit activation function produced the most accurate and efficient configuration. This integrated modeling strategy provides a structured, reproducible, and data-driven approach for simulating complex nonlinear membrane behavior. The findings demonstrate the method’s potential for scalable and intelligent process optimization in wastewater treatment, supporting sustainable water management and aligning with the objectives of sustainable development goal 6 on clean water and sanitation.
{"title":"Hybrid artificial neural network and response surface methodology modeling and hyperparameter optimization for submerged membrane bioreactor filtration","authors":"Syahira Ibrahim , Norhaliza Abdul Wahab , Zakariah Yusuf","doi":"10.1016/j.cherd.2025.11.030","DOIUrl":"10.1016/j.cherd.2025.11.030","url":null,"abstract":"<div><div>Accurate predictive modeling is vital for optimizing membrane filtration processes in wastewater treatment. This paper proposes a hybrid model that integrates an artificial neural network with response surface methodology to predict and simulate permeate flux in a submerged membrane bioreactor treating palm oil mill effluent. The framework integrates response surface methodology in both the experimental design and hyperparameter optimization stages, enabling systematic and efficient development of two neural architectures: feed-forward and radial basis function neural networks. Model performance was benchmarked against the conventional one-variable-at-a-time method and evaluated using independent training and testing datasets. The hybrid approach achieved high predictive accuracy, with coefficients of determination ranging from 0.9142 to 0.9981, while reducing computational time by nearly 49 % and the number of optimization iterations by approximately 50–64 %. The Levenberg–Marquardt training algorithm combined with the rectified linear unit activation function produced the most accurate and efficient configuration. This integrated modeling strategy provides a structured, reproducible, and data-driven approach for simulating complex nonlinear membrane behavior. The findings demonstrate the method’s potential for scalable and intelligent process optimization in wastewater treatment, supporting sustainable water management and aligning with the objectives of sustainable development goal 6 on clean water and sanitation.</div></div>","PeriodicalId":10019,"journal":{"name":"Chemical Engineering Research & Design","volume":"224 ","pages":"Pages 433-451"},"PeriodicalIF":3.9,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145614960","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-01DOI: 10.1016/j.cherd.2025.11.039
Mohammad Shakil Ahmmed , Christian K. Otten , Michael V.W. Kofoed , Damien J. Batstone
Mixing and mass transfer are critical to achieve the desired throughput in gas–liquid interactions and thus biological conversion efficiencies. Both mixing and mass transfer are significantly dependent on the size of the gas bubble. The non-Newtonian characteristics of wastewater result in a highly dynamic behaviour of these gas bubbles throughout the reactor volume. While studies so far have been conducted on gas–liquid hydrodynamic behaviour, the impact of wastewater rheology on gas bubble diameter and their population has been overlooked, leading to uncertainty in mass transfer rate prediction. To address this gap, we developed both computational, using coupled computational fluid dynamics (CFD) and population balance model (PBM), and experimental approaches, using high-speed imaging techniques. Unlike existing studies, these approaches were deployed on an industry representative system- a pilot-scale venturi mixed reactor- commonly applied for mixing and mass transfer in both low-viscosity Newtonian and high (and shear-dependent) viscosity non-Newtonian liquids. An excellent agreement was observed between the numerical and experimental results of bubble diameter and population, demonstrating a diverse range of bubbles resulting from bubble coalescence and breakage within the venturi and reactor in Newtonian and non-Newtonian liquids. The results identified that volumetric mass transfer rate (ka) is highly variable throughout the reactor due to variations in bubble size. A larger average bubble size with higher viscosity results in a substantial decrease in ka from 21 day (for water viscosity) to 3 day at higher viscosity, despite an increased gas holdup.
{"title":"Impact of rheology on mass transfer and bubble diameter in wastewater treatment systems","authors":"Mohammad Shakil Ahmmed , Christian K. Otten , Michael V.W. Kofoed , Damien J. Batstone","doi":"10.1016/j.cherd.2025.11.039","DOIUrl":"10.1016/j.cherd.2025.11.039","url":null,"abstract":"<div><div>Mixing and mass transfer are critical to achieve the desired throughput in gas–liquid interactions and thus biological conversion efficiencies. Both mixing and mass transfer are significantly dependent on the size of the gas bubble. The non-Newtonian characteristics of wastewater result in a highly dynamic behaviour of these gas bubbles throughout the reactor volume. While studies so far have been conducted on gas–liquid hydrodynamic behaviour, the impact of wastewater rheology on gas bubble diameter and their population has been overlooked, leading to uncertainty in mass transfer rate prediction. To address this gap, we developed both computational, using coupled computational fluid dynamics (CFD) and population balance model (PBM), and experimental approaches, using high-speed imaging techniques. Unlike existing studies, these approaches were deployed on an industry representative system- a pilot-scale venturi mixed reactor- commonly applied for mixing and mass transfer in both low-viscosity Newtonian and high (and shear-dependent) viscosity non-Newtonian liquids. An excellent agreement was observed between the numerical and experimental results of bubble diameter and population, demonstrating a diverse range of bubbles resulting from bubble coalescence and breakage within the venturi and reactor in Newtonian and non-Newtonian liquids. The results identified that volumetric mass transfer rate (k<span><math><msub><mrow></mrow><mrow><mi>L</mi></mrow></msub></math></span>a) is highly variable throughout the reactor due to variations in bubble size. A larger average bubble size with higher viscosity results in a substantial decrease in k<span><math><msub><mrow></mrow><mrow><mi>L</mi></mrow></msub></math></span>a from 21 day<span><math><msup><mrow></mrow><mrow><mo>−</mo><mn>1</mn></mrow></msup></math></span> (for water viscosity) to 3 day<span><math><msup><mrow></mrow><mrow><mo>−</mo><mn>1</mn></mrow></msup></math></span> at higher viscosity, despite an increased gas holdup.</div></div>","PeriodicalId":10019,"journal":{"name":"Chemical Engineering Research & Design","volume":"224 ","pages":"Pages 523-532"},"PeriodicalIF":3.9,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145614961","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-01DOI: 10.1016/j.cherd.2025.11.038
Donghua Ji , Ming Zhao , Hongyan Li
In this study, thermogravimetric analysis (TGA) was utilized to investigate the combustion behavior and kinetic characteristics of individual samples of soybean straw (SS), anthracite (AN), and oil sludge, as well as a mixed sample of soybean straw and anthracite (SS/AN) at varying heating rates of 10, 20, and 30 °C/min. The results indicate significant differences in the combustion stages of the individual samples. Specifically, SS and oil sludge, characterized by their high volatile matter content, exhibit lower ignition temperatures (Ti) and burnout temperatures (Tb), with their combustion processes divided into three distinct stages. Conversely, AN, with its high fixed carbon content, demonstrates relatively higher Ti and Tb, and its combustion process is categorized into two stages. An increase in the heating rate enhances the combustion intensity of the samples, significantly elevating the maximum weight loss rate and combustion indices (C, S). In the mixed samples, as the proportion of SS increases, the Ti of SS/AN decreases while the combustion indices increase. Kinetic analysis reveals that the order of apparent activation energy is SS>oil sludge> 30 %SS/AN>AN. The apparent activation energy of the 30 %SS/AN is reduced due to synergistic effects, including volatile matter promotion and alkali metal catalysis. These findings confirm that co-combusting SS with AN can enhance combustion performance, providing a valuable reference for bio-coal mixed combustion technology.
{"title":"Investigation on the co-combustion characteristics and kinetics of soybean straw and anthracite","authors":"Donghua Ji , Ming Zhao , Hongyan Li","doi":"10.1016/j.cherd.2025.11.038","DOIUrl":"10.1016/j.cherd.2025.11.038","url":null,"abstract":"<div><div>In this study, thermogravimetric analysis (TGA) was utilized to investigate the combustion behavior and kinetic characteristics of individual samples of soybean straw (SS), anthracite (AN), and oil sludge, as well as a mixed sample of soybean straw and anthracite (SS/AN) at varying heating rates of 10, 20, and 30 °C/min. The results indicate significant differences in the combustion stages of the individual samples. Specifically, SS and oil sludge, characterized by their high volatile matter content, exhibit lower ignition temperatures (<em>T</em><sub><em>i</em></sub>) and burnout temperatures (<em>T</em><sub><em>b</em></sub>), with their combustion processes divided into three distinct stages. Conversely, AN, with its high fixed carbon content, demonstrates relatively higher <em>T</em><sub><em>i</em></sub> and <em>T</em><sub><em>b</em></sub>, and its combustion process is categorized into two stages. An increase in the heating rate enhances the combustion intensity of the samples, significantly elevating the maximum weight loss rate and combustion indices (<em>C</em>, <em>S</em>). In the mixed samples, as the proportion of SS increases, the <em>T</em><sub><em>i</em></sub> of SS/AN decreases while the combustion indices increase. Kinetic analysis reveals that the order of apparent activation energy is SS>oil sludge> 30 %SS/AN>AN. The apparent activation energy of the 30 %SS/AN is reduced due to synergistic effects, including volatile matter promotion and alkali metal catalysis. These findings confirm that co-combusting SS with AN can enhance combustion performance, providing a valuable reference for bio-coal mixed combustion technology.</div></div>","PeriodicalId":10019,"journal":{"name":"Chemical Engineering Research & Design","volume":"224 ","pages":"Pages 533-543"},"PeriodicalIF":3.9,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145614957","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号