Water electrolyzers play a crucial role in green hydrogen production.However,their efficiency and scalability are often compromised by bubble dynamics across various scales,from nanoscale to macroscale components.This...Water electrolyzers play a crucial role in green hydrogen production.However,their efficiency and scalability are often compromised by bubble dynamics across various scales,from nanoscale to macroscale components.This review explores multi-scale modeling as a tool to visualize multi-phase flow and improve mass transport in water electrolyzers.At the nanoscale,molecular dynamics(MD)simulations reveal how electrode surface features and wettability influence nanobubble nucleation and stability.Moving to the mesoscale,models such as volume of fluid(VOF)and lattice Boltzmann method(LBM)shed light on bubble transport in porous transport layers(PTLs).These insights inform innovative designs,including gradient porosity and hydrophilic-hydrophobic patterning,aimed at minimizing gas saturation.At the macroscale,VOF simulations elucidate two-phase flow regimes within channels,showing how flow field geometry and wettability affect bubble discharging.Moreover,artificial intelligence(AI)-driven surrogate models expedite the optimization process,allowing for rapid exploration of structural parameters in channel-rib flow fields and porous flow field designs.By integrating these approaches,we can bridge theoretical insights with experimental validation,ultimately enhancing water electrolyzer performance,reducing costs,and advancing affordable,high-efficiency hydrogen production.展开更多
A thermodynamically complete multi-phase equation of state(EOS)applicable to both dense and porous metals at wide ranges of temperature and pressure is constructed.A standard three-term decomposition of the Helmholtz ...A thermodynamically complete multi-phase equation of state(EOS)applicable to both dense and porous metals at wide ranges of temperature and pressure is constructed.A standard three-term decomposition of the Helmholtz free energy as a function of specific volume and temperature is presented,where the cold component models both compression and expansion states,the thermal ion component introduces the Debye approximation and melting entropy,and the thermal electron component employs the Thomas-Fermi-Kirzhnits(TFK)model.The porosity of materials is considered by introducing the dynamic porosity coefficientαand the constitutive P-αrelation,connecting the thermodynamic properties between dense and porous systems,allowing for an accurate description of the volume decrease caused by void collapse while maintaining the quasi-static thermodynamic properties of porous systems identical to the dense ones.These models enable the EOS applicable and robust at wide ranges of temperature,pressure and porosity.A systematic evaluation of the new EOS is conducted with aluminum(Al)as an example.300 K isotherm,shock Hugoniot,as well as melting curves of both dense and porous Al are calculated,which shows great agreements with experimental data and validates the effectiveness of the models and the accuracy of parameterizations.Notably,it is for the first time Hugoniot P-σcurves up to 10~6 GPa and shock melting behaviors of porous Al are derived from analytical EOS models,which predict much lower compression limit and shock melting temperatures than those of dense Al.展开更多
短期预测在智能电网建设中扮演着重要角色,深刻影响电网发输变配用各个环节的智能化改造。短期预测一般基于系统实测数据,而传感器故障,数据传输错误等原因会导致数据质量下降,严重影响短期预测的精确性。为建立数据质量受损情况下的精...短期预测在智能电网建设中扮演着重要角色,深刻影响电网发输变配用各个环节的智能化改造。短期预测一般基于系统实测数据,而传感器故障,数据传输错误等原因会导致数据质量下降,严重影响短期预测的精确性。为建立数据质量受损情况下的精确短期预测模型,提出了结合数据预处理和双向长短期记忆(bi-directional long short-term memory,Bi-LSTM)的短期预测框架Bi-LSTM-DP(bi-directional long short-term memory data preprocessing)。在Bi-LSTM-DP中,采集的数据首先通过均值填补缺失值,进而基于Savitzky-Golay滤波器对数据降噪,最后采用Bi-LSTM提取时间序列的信息,实现短期预测。为了评估所提方法的性能,文中使用实测的公开数据集分别预测风电发电量和负荷需求,与其他参考方法对比表明了所述方法的有效性和鲁棒性。展开更多
Based on the theory of superimposed deformation and the regional tectonic background,the multi-phase non-coaxial superimposed structures in Junggar Basin were systematically analyzed using seismic interpretation,field...Based on the theory of superimposed deformation and the regional tectonic background,the multi-phase non-coaxial superimposed structures in Junggar Basin were systematically analyzed using seismic interpretation,field outcrop observation,and paleo-stress field recovery methods according to the characteristics of the current tectonic framework.Moreover,the tectonic evolution process of the basin was reconstructed using sandbox analogue modelling technology.The results showed that the study area has experienced five phases of non-coaxial deformation with superimposition:The first phase of deformation(D1)is characterized by NNE-SSW extension during late Carboniferous to early Permian,which formed large graben,half graben and other extensional structure style around the basin.The second phase of deformation(D2)is represented by NE-SW compression during the middle to late Permian,and it comprised numerous contraction structures that developed based on D1.The basic form of the entire basin is alternating uplift and depression.The third phase of deformation(D3)is the NW-SE transpressional strike-slip in the Triassic-Jurassic,which produced numerous strike-slip structural styles in the middle part of the basin.The fourth phase of deformation(D4)is the uniform sedimentation during Cretaceous,and the fifth phase(D5)is the compression along NNE-SSW due to the North Tianshan northward thrust,which produced three rows of fold thrust belts and tear faults in the front of the mountain in the southern margin of the basin.The newly established three-dimensional tectonic evolution model shows that,based on the large number of NW-trending grabens and half grabens in the Carboniferous basement of Junggar Basin,multiple level NE trending uplifts have formed with the joint superposition of the late structural inversion and multiple stress fields.This has resulted in the current tectonic units of alternating uplifts and depressions in different directions in the study area.展开更多
基金supported by a grant from the Research Grants Council of the Hong Kong Special Administrative Region,China(Project No.15308024)a grant from Research Centre for Carbon-Strategic Catalysis,The Hong Kong Polytechnic University(CE2X).
文摘Water electrolyzers play a crucial role in green hydrogen production.However,their efficiency and scalability are often compromised by bubble dynamics across various scales,from nanoscale to macroscale components.This review explores multi-scale modeling as a tool to visualize multi-phase flow and improve mass transport in water electrolyzers.At the nanoscale,molecular dynamics(MD)simulations reveal how electrode surface features and wettability influence nanobubble nucleation and stability.Moving to the mesoscale,models such as volume of fluid(VOF)and lattice Boltzmann method(LBM)shed light on bubble transport in porous transport layers(PTLs).These insights inform innovative designs,including gradient porosity and hydrophilic-hydrophobic patterning,aimed at minimizing gas saturation.At the macroscale,VOF simulations elucidate two-phase flow regimes within channels,showing how flow field geometry and wettability affect bubble discharging.Moreover,artificial intelligence(AI)-driven surrogate models expedite the optimization process,allowing for rapid exploration of structural parameters in channel-rib flow fields and porous flow field designs.By integrating these approaches,we can bridge theoretical insights with experimental validation,ultimately enhancing water electrolyzer performance,reducing costs,and advancing affordable,high-efficiency hydrogen production.
基金Project supported by the National Natural Science Foundation of China(Grant Nos.12205023,U2230401,12374056,U23A20537,11904027)。
文摘A thermodynamically complete multi-phase equation of state(EOS)applicable to both dense and porous metals at wide ranges of temperature and pressure is constructed.A standard three-term decomposition of the Helmholtz free energy as a function of specific volume and temperature is presented,where the cold component models both compression and expansion states,the thermal ion component introduces the Debye approximation and melting entropy,and the thermal electron component employs the Thomas-Fermi-Kirzhnits(TFK)model.The porosity of materials is considered by introducing the dynamic porosity coefficientαand the constitutive P-αrelation,connecting the thermodynamic properties between dense and porous systems,allowing for an accurate description of the volume decrease caused by void collapse while maintaining the quasi-static thermodynamic properties of porous systems identical to the dense ones.These models enable the EOS applicable and robust at wide ranges of temperature,pressure and porosity.A systematic evaluation of the new EOS is conducted with aluminum(Al)as an example.300 K isotherm,shock Hugoniot,as well as melting curves of both dense and porous Al are calculated,which shows great agreements with experimental data and validates the effectiveness of the models and the accuracy of parameterizations.Notably,it is for the first time Hugoniot P-σcurves up to 10~6 GPa and shock melting behaviors of porous Al are derived from analytical EOS models,which predict much lower compression limit and shock melting temperatures than those of dense Al.
文摘短期预测在智能电网建设中扮演着重要角色,深刻影响电网发输变配用各个环节的智能化改造。短期预测一般基于系统实测数据,而传感器故障,数据传输错误等原因会导致数据质量下降,严重影响短期预测的精确性。为建立数据质量受损情况下的精确短期预测模型,提出了结合数据预处理和双向长短期记忆(bi-directional long short-term memory,Bi-LSTM)的短期预测框架Bi-LSTM-DP(bi-directional long short-term memory data preprocessing)。在Bi-LSTM-DP中,采集的数据首先通过均值填补缺失值,进而基于Savitzky-Golay滤波器对数据降噪,最后采用Bi-LSTM提取时间序列的信息,实现短期预测。为了评估所提方法的性能,文中使用实测的公开数据集分别预测风电发电量和负荷需求,与其他参考方法对比表明了所述方法的有效性和鲁棒性。
基金supported by the National Natural Science Foundation of China,(Grant No.42072144)Shengli Oilfield,SINOPEC,China(Nos.30200018-21-ZC0613-0030 and 30200018-20-ZC0613-0116)。
文摘Based on the theory of superimposed deformation and the regional tectonic background,the multi-phase non-coaxial superimposed structures in Junggar Basin were systematically analyzed using seismic interpretation,field outcrop observation,and paleo-stress field recovery methods according to the characteristics of the current tectonic framework.Moreover,the tectonic evolution process of the basin was reconstructed using sandbox analogue modelling technology.The results showed that the study area has experienced five phases of non-coaxial deformation with superimposition:The first phase of deformation(D1)is characterized by NNE-SSW extension during late Carboniferous to early Permian,which formed large graben,half graben and other extensional structure style around the basin.The second phase of deformation(D2)is represented by NE-SW compression during the middle to late Permian,and it comprised numerous contraction structures that developed based on D1.The basic form of the entire basin is alternating uplift and depression.The third phase of deformation(D3)is the NW-SE transpressional strike-slip in the Triassic-Jurassic,which produced numerous strike-slip structural styles in the middle part of the basin.The fourth phase of deformation(D4)is the uniform sedimentation during Cretaceous,and the fifth phase(D5)is the compression along NNE-SSW due to the North Tianshan northward thrust,which produced three rows of fold thrust belts and tear faults in the front of the mountain in the southern margin of the basin.The newly established three-dimensional tectonic evolution model shows that,based on the large number of NW-trending grabens and half grabens in the Carboniferous basement of Junggar Basin,multiple level NE trending uplifts have formed with the joint superposition of the late structural inversion and multiple stress fields.This has resulted in the current tectonic units of alternating uplifts and depressions in different directions in the study area.
