Search

Decay-time-dependent $CP$-violation effects in transitions of neutral $B$ mesons to $CP$-eigenstates can be visualised by oscillations in the asymmetry, as a function of decay time, between decay yields from mesons tagged as initially having $\overline{B}$ or $B$ flavour. Such images, for example for $B^0 \to J/\psi K^0_{\rm S}$ decays where the magnitude of the oscillation is proportional to $\sin(2\beta)$ with $\beta$ being an angle of the Cabibbo--Kobayashi--Maskawa Unitarity Triangle, provide a straightforward illustration of the underlying physics. Until now there has been no comparable method to provide visualisation for the case of decays to multibody final states that are not $CP$-eigenstates, where interference between $CP$-even and -odd amplitudes provides additional physics sensitivity. A method is proposed to weight the data so that the terms of interest can be projected out and used to obtain asymmetries that visualise the relevant effects. Application of the weighting to $B^0_s$ decays, where effects due to non-zero width difference are not negligible, provides a novel method to observe $CP$ violation in interference between mixing and decay without tagging the production flavour., Comment: 17 pages, 5 figures

Counterfactual explanation generation is a powerful method for Explainable Artificial Intelligence. It can help users understand why machine learning models make specific decisions, and how to change those decisions. Evaluating the robustness of counterfactual explanation algorithms is therefore crucial. Previous literature has widely studied the robustness based on the perturbation of input instances. However, the robustness defined from the perspective of perturbed instances is sometimes biased, because this definition ignores the impact of learning algorithms on robustness. In this paper, we propose a more reasonable definition, Weak Robust Compatibility, based on the perspective of explanation strength. In practice, we propose WRC-Test to help us generate more robust counterfactuals. Meanwhile, we designed experiments to verify the effectiveness of WRC-Test. Theoretically, we introduce the concepts of PAC learning theory and define the concept of PAC WRC-Approximability. Based on reasonable assumptions, we establish oracle inequalities about weak robustness, which gives a sufficient condition for PAC WRC-Approximability.

In the realm of Artificial Intelligence (AI), the importance of Explainable Artificial Intelligence (XAI) is increasingly recognized, particularly as AI models become more integral to our lives. One notable single-instance XAI approach is counterfactual explanation, which aids users in comprehending a model's decisions and offers guidance on altering these decisions. Specifically in the context of image classification models, effective image counterfactual explanations can significantly enhance user understanding. This paper introduces a novel method for computing feature importance within the feature space of a black-box model. By employing information fusion techniques, our method maximizes the use of data to address feature counterfactual explanations in the feature space. Subsequently, we utilize an image generation model to transform these feature counterfactual explanations into image counterfactual explanations. Our experiments demonstrate that the counterfactual explanations generated by our method closely resemble the original images in both pixel and feature spaces. Additionally, our method outperforms established baselines, achieving impressive experimental results.

In Lifelong Learning (LL), agents continually learn as they encounter new conditions and tasks. Most current LL is limited to a single agent that learns tasks sequentially. Dedicated LL machinery is then deployed to mitigate the forgetting of old tasks as new tasks are learned. This is inherently slow. We propose a new Shared Knowledge Lifelong Learning (SKILL) challenge, which deploys a decentralized population of LL agents that each sequentially learn different tasks, with all agents operating independently and in parallel. After learning their respective tasks, agents share and consolidate their knowledge over a decentralized communication network, so that, in the end, all agents can master all tasks. We present one solution to SKILL which uses Lightweight Lifelong Learning (LLL) agents, where the goal is to facilitate efficient sharing by minimizing the fraction of the agent that is specialized for any given task. Each LLL agent thus consists of a common task-agnostic immutable part, where most parameters are, and individual task-specific modules that contain fewer parameters but are adapted to each task. Agents share their task-specific modules, plus summary information ("task anchors") representing their tasks in the common task-agnostic latent space of all agents. Receiving agents register each received task-specific module using the corresponding anchor. Thus, every agent improves its ability to solve new tasks each time new task-specific modules and anchors are received. On a new, very challenging SKILL-102 dataset with 102 image classification tasks (5,033 classes in total, 2,041,225 training, 243,464 validation, and 243,464 test images), we achieve much higher (and SOTA) accuracy over 8 LL baselines, while also achieving near perfect parallelization. Code and data can be found at https://github.com/gyhandy/Shared-Knowledge-Lifelong-Learning, Comment: Transactions on Machine Learning Research (TMLR) paper

In this paper, we present our solutions for the 5th Workshop and Competition on Affective Behavior Analysis in-the-wild (ABAW), which includes four sub-challenges of Valence-Arousal (VA) Estimation, Expression (Expr) Classification, Action Unit (AU) Detection and Emotional Reaction Intensity (ERI) Estimation. The 5th ABAW competition focuses on facial affect recognition utilizing different modalities and datasets. In our work, we extract powerful audio and visual features using a large number of sota models. These features are fused by Transformer Encoder and TEMMA. Besides, to avoid the possible impact of large dimensional differences between various features, we design an Affine Module to align different features to the same dimension. Extensive experiments demonstrate that the superiority of the proposed method. For the VA Estimation sub-challenge, our method obtains the mean Concordance Correlation Coefficient (CCC) of 0.6066. For the Expression Classification sub-challenge, the average F1 Score is 0.4055. For the AU Detection sub-challenge, the average F1 Score is 0.5296. For the Emotional Reaction Intensity Estimation sub-challenge, the average pearson's correlations coefficient on the validation set is 0.3968. All of the results of four sub-challenges outperform the baseline with a large margin.

Many different strategies have been used to study screen time control to deal with the side effects of dark patterns. However, users might be annoyed by those methods that create too strong “friction” and give them up too soon. To explore this design space, I turned to reflective design, which is a systematic approach including participants in the design process for critical reflection and new knowledge. Since TikTok is a representative prevalent social media platform in recent years, in this study, I did a reflective design of 5 interventions of adding “weak friction” to the TikTok platform as a probe and invited 8 participants to test them. The results suggest there is an untapped potential for weak friction to counteract the side effects of dark patterns and support novel self-restriction ways when using social media apps. The probe also resulted in rich material that documented participants’ attitudes, feelings and expectations of friction, in order to raise attention from researchers, UX practitioners and authorities of designing friction in social media apps., Många olika strategier har använts för att studera skärmtidskontroll för att hantera biverkningarna av mörka mönster. Användare kan dock bli irriterade över de metoder som skapar för stark "friktion" och ger upp dem för tidigt. För att utforska detta designutrymme vände jag mig till reflekterande design, vilket är ett systematiskt tillvägagångssätt som inkluderar deltagare i designprocessen för kritisk reflektion och ny kunskap. Eftersom TikTok är en representativ utbredd social medieplattform under de senaste åren, gjorde jag i denna studie en reflekterande design av 5 interventioner för att lägga till "svag friktion" till TikTok-plattformen som en sond och bjöd in 8 deltagare att testa dem. Resultaten tyder på att det finns en outnyttjad potential för svag friktion för att motverka biverkningarna av mörka mönster och stödja nya självbegränsningssätt när man använder appar för sociala medier. Undersökningen resulterade också i ett rikt material som dokumenterade deltagarnas attityder, känslor och förväntningar på friktion, för att väcka uppmärksamhet från forskare, UX-utövare och myndigheter för att designa friktion i appar för sociala medier.

We assess the performance of the hybrid Open Accelerator (OpenACC) and Message Passing Interface (MPI) approach for multi-graphics processing units (GPUs) accelerated thermal lattice Boltzmann (LB) simulation. The OpenACC accelerates computation on a single GPU, and the MPI synchronizes the information between multiple GPUs. With a single GPU, the two-dimension (2D) simulation achieved 1.93 billion lattice updates per second (GLUPS) with a grid number of $8193^{2}$, and the three-dimension (3D) simulation achieved 1.04 GLUPS with a grid number of $385^{3}$, which is more than 76% of the theoretical maximum performance. On multi-GPUs, we adopt block partitioning, overlapping communications with computations, and concurrent computation to optimize parallel efficiency. We show that in the strong scaling test, using 16 GPUs, the 2D simulation achieved 30.42 GLUPS and the 3D simulation achieved 14.52 GLUPS. In the weak scaling test, the parallel efficiency remains above 99% up to 16 GPUs. Our results demonstrated that, with improved data and task management, the hybrid OpenACC and MPI technique is promising for thermal LB simulation on multi-GPUs., Comment: 31 pages, 12 figures

Many different strategies have been used to study screen time control to deal with the side effects of dark patterns. However, users might be annoyed by those methods that create too strong “friction” and give them up too soon. To explore this design space, I turned to reflective design, which is a systematic approach including participants in the design process for critical reflection and new knowledge. Since TikTok is a representative prevalent social media platform in recent years, in this study, I did a reflective design of 5 interventions of adding “weak friction” to the TikTok platform as a probe and invited 8 participants to test them. The results suggest there is an untapped potential for weak friction to counteract the side effects of dark patterns and support novel self-restriction ways when using social media apps. The probe also resulted in rich material that documented participants’ attitudes, feelings and expectations of friction, in order to raise attention from researchers, UX practitioners and authorities of designing friction in social media apps., Många olika strategier har använts för att studera skärmtidskontroll för att hantera biverkningarna av mörka mönster. Användare kan dock bli irriterade över de metoder som skapar för stark "friktion" och ger upp dem för tidigt. För att utforska detta designutrymme vände jag mig till reflekterande design, vilket är ett systematiskt tillvägagångssätt som inkluderar deltagare i designprocessen för kritisk reflektion och ny kunskap. Eftersom TikTok är en representativ utbredd social medieplattform under de senaste åren, gjorde jag i denna studie en reflekterande design av 5 interventioner för att lägga till "svag friktion" till TikTok-plattformen som en sond och bjöd in 8 deltagare att testa dem. Resultaten tyder på att det finns en outnyttjad potential för svag friktion för att motverka biverkningarna av mörka mönster och stödja nya självbegränsningssätt när man använder appar för sociala medier. Undersökningen resulterade också i ett rikt material som dokumenterade deltagarnas attityder, känslor och förväntningar på friktion, för att väcka uppmärksamhet från forskare, UX-utövare och myndigheter för att designa friktion i appar för sociala medier.

Learning the causal structure behind data is invaluable for improving generalization and obtaining high-quality explanations. We propose a novel framework, Invariant Structure Learning (ISL), that is designed to improve causal structure discovery by utilizing generalization as an indication. ISL splits the data into different environments, and learns a structure that is invariant to the target across different environments by imposing a consistency constraint. An aggregation mechanism then selects the optimal classifier based on a graph structure that reflects the causal mechanisms in the data more accurately compared to the structures learnt from individual environments. Furthermore, we extend ISL to a self-supervised learning setting where accurate causal structure discovery does not rely on any labels. This self-supervised ISL utilizes invariant causality proposals by iteratively setting different nodes as targets. On synthetic and real-world datasets, we demonstrate that ISL accurately discovers the causal structure, outperforms alternative methods, and yields superior generalization for datasets with significant distribution shifts., Comment: 16 pages (including Appendix), 4 figures

The development of direct methanol fuel cells is hindered by the issue of methanol crossover across membranes, despite the remarkable features resulting from the use of liquid fuel. Here we investigate the proton-selective conduction behavior across 2D graphyne in an aqueous environment. The aqueous proton conduction mechanism transitions from bare proton penetration to a mixed vehicular and Grotthuss transportation when the side length of triangular graphyne pores increases to 0.95 nm. A further increase in the side length to 1.2 nm results in the formation of a patterned aqueous/vacuum interphase, enabling protons to be conducted through the water wires via Grotthuss mechanism with low energy barriers. More importantly, it is found that 2D graphyne with the side length of less than 1.45 nm can effectively block methanol crossover, suggesting that 2D graphyne with an appropriate pore size is an ideal material to achieve zero-crossover proton-selective membranes.

The development of direct methanol fuel cells is hindered by the issue of methanol crossover across membranes, despite the remarkable features resulting from the use of liquid fuel. Here we investigate the proton-selective conduction behavior across 2D graphyne in an aqueous environment. The aqueous proton conduction mechanism transitions from bare proton penetration to a mixed vehicular and Grotthuss transportation when the side length of triangular graphyne pores increases to 0.95 nm. A further increase in the side length to 1.2 nm results in the formation of a patterned aqueous/vacuum interphase, enabling protons to be conducted through the water wires via Grotthuss mechanism with low energy barriers. More importantly, it is found that 2D graphyne with the side length of less than 1.45 nm can effectively block methanol crossover, suggesting that 2D graphyne with an appropriate pore size is an ideal material to achieve zero-crossover proton-selective membranes.

The development of direct methanol fuel cells is hindered by the issue of methanol crossover across membranes, despite the remarkable features resulting from the use of liquid fuel. Here we investigate the proton-selective conduction behavior across 2D graphyne in an aqueous environment. The aqueous proton conduction mechanism transitions from bare proton penetration to a mixed vehicular and Grotthuss transportation when the side length of triangular graphyne pores increases to 0.95 nm. A further increase in the side length to 1.2 nm results in the formation of a patterned aqueous/vacuum interphase, enabling protons to be conducted through the water wires via Grotthuss mechanism with low energy barriers. More importantly, it is found that 2D graphyne with the side length of less than 1.45 nm can effectively block methanol crossover, suggesting that 2D graphyne with an appropriate pore size is an ideal material to achieve zero-crossover proton-selective membranes.

Using the density functional theory method, we demonstrate that the recently synthesized three-dimensional carbon-honeycomb (C-honeycomb) can serve as a promising nanoporous scaffold for both lithium and sodium deposition with good electronic conductivity. Lithium/sodium can insert into the channels of C-honeycomb with low migration energy barriers along the walls of one-dimensional pores (<0.5 eV for lithium and <0.25 eV for sodium) and exist in the form of metal nanorods. A high theoretical capacity (711 mAh/g, almost twice that of graphite) can be achieved when the one-dimensional pores are filled with lithium/sodium atoms. The volume expansion of the C-honeycomb after metal insertion is less than 5 and 15% for lithium and sodium, respectively. Further introduction of defects such as pyridinic-N doping or single vacancy can provide initial nucleation sites for the metal nanorods and increase the open-circuit voltage.

We investigate thermal effects on the sedimentation behavior of elliptical particles via particle-resolved direct numerical simulation. Two scenarios of fluid-particle heat transfer in an infinitely long channel are considered: one is a cold particle settling in a hot fluid, while the other is a hot particle settling in a cold fluid. Results show that when an elliptical particle sediment in a wide channel (i.e., the block ratio is large), in addition to the two sedimentation modes reported in the literature for the particle sediments in isothermal fluids, there exist another three modes arising from thermal effects: the tumbling mode, the anomalous rolling mode and the inclined mode. Specifically, for a cold particle settling in a hot fluid, we found the tumbling mode and inclined mode, while for a hot particle settling in a cold fluid, we found the anomalous rolling mode and inclined mode. The phase diagrams of the sedimentation modes as functions of Archimedes and Grashof numbers are given. Analyses of the relationship between particle Reynolds number and Grashof number indicates that the mode presented depends on the competition between channel wall confinement, combined forced and natural convection. © 2018 Elsevier Ltd

Using the density functional theory method, we demonstrate that the recently synthesized three-dimensional carbon-honeycomb (C-honeycomb) can serve as a promising nanoporous scaffold for both lithium and sodium deposition with good electronic conductivity. Lithium/sodium can insert into the channels of C-honeycomb with low migration energy barriers along the walls of one-dimensional pores (<0.5 eV for lithium and <0.25 eV for sodium) and exist in the form of metal nanorods. A high theoretical capacity (711 mAh/g, almost twice that of graphite) can be achieved when the one-dimensional pores are filled with lithium/sodium atoms. The volume expansion of the C-honeycomb after metal insertion is less than 5 and 15% for lithium and sodium, respectively. Further introduction of defects such as pyridinic-N doping or single vacancy can provide initial nucleation sites for the metal nanorods and increase the open-circuit voltage.

We investigate thermal effects on the sedimentation behavior of elliptical particles via particle-resolved direct numerical simulation. Two scenarios of fluid-particle heat transfer in an infinitely long channel are considered: one is a cold particle settling in a hot fluid, while the other is a hot particle settling in a cold fluid. Results show that when an elliptical particle sediment in a wide channel (i.e., the block ratio is large), in addition to the two sedimentation modes reported in the literature for the particle sediments in isothermal fluids, there exist another three modes arising from thermal effects: the tumbling mode, the anomalous rolling mode and the inclined mode. Specifically, for a cold particle settling in a hot fluid, we found the tumbling mode and inclined mode, while for a hot particle settling in a cold fluid, we found the anomalous rolling mode and inclined mode. The phase diagrams of the sedimentation modes as functions of Archimedes and Grashof numbers are given. Analyses of the relationship between particle Reynolds number and Grashof number indicates that the mode presented depends on the competition between channel wall confinement, combined forced and natural convection. © 2018 Elsevier Ltd

Using the density functional theory method, we demonstrate that the recently synthesized three-dimensional carbon-honeycomb (C-honeycomb) can serve as a promising nanoporous scaffold for both lithium and sodium deposition with good electronic conductivity. Lithium/sodium can insert into the channels of C-honeycomb with low migration energy barriers along the walls of one-dimensional pores (<0.5 eV for lithium and <0.25 eV for sodium) and exist in the form of metal nanorods. A high theoretical capacity (711 mAh/g, almost twice that of graphite) can be achieved when the one-dimensional pores are filled with lithium/sodium atoms. The volume expansion of the C-honeycomb after metal insertion is less than 5 and 15% for lithium and sodium, respectively. Further introduction of defects such as pyridinic-N doping or single vacancy can provide initial nucleation sites for the metal nanorods and increase the open-circuit voltage.

We investigate thermal effects on the sedimentation behavior of elliptical particles via particle-resolved direct numerical simulation. Two scenarios of fluid-particle heat transfer in an infinitely long channel are considered: one is a cold particle settling in a hot fluid, while the other is a hot particle settling in a cold fluid. Results show that when an elliptical particle sediment in a wide channel (i.e., the block ratio is large), in addition to the two sedimentation modes reported in the literature for the particle sediments in isothermal fluids, there exist another three modes arising from thermal effects: the tumbling mode, the anomalous rolling mode and the inclined mode. Specifically, for a cold particle settling in a hot fluid, we found the tumbling mode and inclined mode, while for a hot particle settling in a cold fluid, we found the anomalous rolling mode and inclined mode. The phase diagrams of the sedimentation modes as functions of Archimedes and Grashof numbers are given. Analyses of the relationship between particle Reynolds number and Grashof number indicates that the mode presented depends on the competition between channel wall confinement, combined forced and natural convection. © 2018 Elsevier Ltd

We assess the performance of the combined Open Accelerator (OpenACC) programming standard and graphics processing unit (GPU) accelerator for lattice Boltzmann (LB) simulations of fluid flow, heat and mass transfer. By optimizing the data layout, minimizing the memory access frequency, and adjusting the number of gangs and vector length, we show that the enhanced parallel computations can result in orders of magnitudes of speedup relative to the serial implementation of the LB algorithm. Based on such implementations, benchmark quality results are obtained with fine grid of 20492 for both two-dimensional lid driven cavity flow with Reynolds number up to 7500, and double diffusive cavity flow with solute Rayleigh number up to 108. © 2017 Elsevier Ltd

Fuel cells and flow batteries are promising technologies to address climate change and air pollution problems. An understanding of the complex multiscale and multiphysics transport phenomena occurring in these electrochemical systems requires powerful numerical tools. Over the past decades, the lattice Boltzmann (LB) method has attracted broad interest in the computational fluid dynamics and the numerical heat transfer communities, primarily due to its kinetic nature making it appropriate for modeling complex multiphase transport phenomena. More importantly, the LB method fits well with parallel computing due to its locality feature, which is required for large-scale engineering applications. In this article, we review the LB method for gas-liquid two-phase flows, coupled fluid flow and mass transport in porous media, and particulate flows. Examples of applications are provided in fuel cells and flow batteries. Further developments of the LB method are also outlined.

An issue with the use Of Metallic lithium as an anode material for lithium-based batteries is,dendrite growth, causing a periodic breaking and repair of the solid electrolyte interphase (SEI) layer. Adding 2D atomic crystals, such as h-BN; as an interfacial layer between the lithium metal anode and liquid electrolyte has been demonstrated to be effective to mitigate dendrite growth, thereby enhancing the Columbic efficiency of lithium metal batteries. But the underlying mechanism leading to the reduced dendrite growth remains unknown. In this work, with the aid of first-principle calculations, we find that the interaction between the h-BN and lithium metal layers is a weak van der Waals force, and two atomic layers of h-BN are thick enough to block the electron tunneling from lithium metal to electrolyte, thus prohibiting the decomposition of electrolyte. The interlayer spacing between the h-BN and lithium metal layers can provide larger adsorption energies toward lithium atoms than that provided by bare lithium or h-BN, making lithium atoms prefer to intercalate under the cover of h-BN during the plating process. The combined high stiffness of h-BN and the low diffusion energy barriers of lithium at the Li/h-BN interfaces induce a uniform distribution of lithium, under h-BN, therefore effectively suppressing dendrite growth. © 2016 American Chemical Society.

Fuel cells and flow batteries are promising technologies to address climate change and air pollution problems. An understanding of the complex multiscale and multiphysics transport phenomena occurring in these electrochemical systems requires powerful numerical tools. Over the past decades, the lattice Boltzmann (LB) method has attracted broad interest in the computational fluid dynamics and the numerical heat transfer communities, primarily due to its kinetic nature making it appropriate for modeling complex multiphase transport phenomena. More importantly, the LB method fits well with parallel computing due to its locality feature, which is required for large-scale engineering applications. In this article, we review the LB method for gas-liquid two-phase flows, coupled fluid flow and mass transport in porous media, and particulate flows. Examples of applications are provided in fuel cells and flow batteries. Further developments of the LB method are also outlined.

Fuel cells and flow batteries are promising renewable energy technologies to address climate change and air pollution problems. Understanding the complex multiscale and multiphysics transport phenomena in these electrochemical systems requires effective numerical approaches. Among various numerical methods, the lattice Boltzmann (LB) method stands out as a powerful tool to simulate fluid flows and associated transport phenomena. This thesis focuses on LB modeling of transport phenomena in fuel cells and flow batteries. It starts with accelerating LB simulation using the combined Open Accelerator programming standard and graphics processing unit (GPU) accelerator. By optimizing the data layout, minimizing the memory access frequency, and adjusting the number of gangs and vector length, a speed up around 50-60 times for multiphysics LB simulation can be achieved comparing with the serial implementations. The enhanced computational performance highlights the potential of LB method to utilize the emerging GPU accelerator for applications in large-scale engineering problems. Subsequently, efforts are devoted to gas-liquid two-phase flows in polymer electrolyte membrane fuel cells, coupled fluid flows and mass transport in aqueous redox flow batteries, and particulate flows in suspension redox flow batteries. In a polymer electrolyte membrane fuel cell system, the transport phenomena involve gas-liquid two-phase flows in the flow channels and in the porous gas diffusion layers on both the anode and cathode. Previous LB models to simulate gas-liquid two-phase flows are applicable to low-density-ratio interfacial problems only, whereas the density ratio is large in practical fuel cell systems. To address this issue, a three-dimensional pseudo-potential-based LB model is developed to simulate gas-liquid two-phase flows with large density ratio. It is demonstrated that this three-dimensional model enables the density ratio to be as large as 700 in static and quasi-static case

An issue with the use Of Metallic lithium as an anode material for lithium-based batteries is,dendrite growth, causing a periodic breaking and repair of the solid electrolyte interphase (SEI) layer. Adding 2D atomic crystals, such as h-BN; as an interfacial layer between the lithium metal anode and liquid electrolyte has been demonstrated to be effective to mitigate dendrite growth, thereby enhancing the Columbic efficiency of lithium metal batteries. But the underlying mechanism leading to the reduced dendrite growth remains unknown. In this work, with the aid of first-principle calculations, we find that the interaction between the h-BN and lithium metal layers is a weak van der Waals force, and two atomic layers of h-BN are thick enough to block the electron tunneling from lithium metal to electrolyte, thus prohibiting the decomposition of electrolyte. The interlayer spacing between the h-BN and lithium metal layers can provide larger adsorption energies toward lithium atoms than that provided by bare lithium or h-BN, making lithium atoms prefer to intercalate under the cover of h-BN during the plating process. The combined high stiffness of h-BN and the low diffusion energy barriers of lithium at the Li/h-BN interfaces induce a uniform distribution of lithium, under h-BN, therefore effectively suppressing dendrite growth. © 2016 American Chemical Society.

We assess the performance of the combined Open Accelerator (OpenACC) programming standard and graphics processing unit (GPU) accelerator for lattice Boltzmann (LB) simulations of fluid flow, heat and mass transfer. By optimizing the data layout, minimizing the memory access frequency, and adjusting the number of gangs and vector length, we show that the enhanced parallel computations can result in orders of magnitudes of speedup relative to the serial implementation of the LB algorithm. Based on such implementations, benchmark quality results are obtained with fine grid of 20492 for both two-dimensional lid driven cavity flow with Reynolds number up to 7500, and double diffusive cavity flow with solute Rayleigh number up to 108. © 2017 Elsevier Ltd

Fuel cells and flow batteries are promising technologies to address climate change and air pollution problems. An understanding of the complex multiscale and multiphysics transport phenomena occurring in these electrochemical systems requires powerful numerical tools. Over the past decades, the lattice Boltzmann (LB) method has attracted broad interest in the computational fluid dynamics and the numerical heat transfer communities, primarily due to its kinetic nature making it appropriate for modeling complex multiphase transport phenomena. More importantly, the LB method fits well with parallel computing due to its locality feature, which is required for large-scale engineering applications. In this article, we review the LB method for gas-liquid two-phase flows, coupled fluid flow and mass transport in porous media, and particulate flows. Examples of applications are provided in fuel cells and flow batteries. Further developments of the LB method are also outlined.

Fuel cells and flow batteries are promising renewable energy technologies to address climate change and air pollution problems. Understanding the complex multiscale and multiphysics transport phenomena in these electrochemical systems requires effective numerical approaches. Among various numerical methods, the lattice Boltzmann (LB) method stands out as a powerful tool to simulate fluid flows and associated transport phenomena. This thesis focuses on LB modeling of transport phenomena in fuel cells and flow batteries. It starts with accelerating LB simulation using the combined Open Accelerator programming standard and graphics processing unit (GPU) accelerator. By optimizing the data layout, minimizing the memory access frequency, and adjusting the number of gangs and vector length, a speed up around 50-60 times for multiphysics LB simulation can be achieved comparing with the serial implementations. The enhanced computational performance highlights the potential of LB method to utilize the emerging GPU accelerator for applications in large-scale engineering problems. Subsequently, efforts are devoted to gas-liquid two-phase flows in polymer electrolyte membrane fuel cells, coupled fluid flows and mass transport in aqueous redox flow batteries, and particulate flows in suspension redox flow batteries. In a polymer electrolyte membrane fuel cell system, the transport phenomena involve gas-liquid two-phase flows in the flow channels and in the porous gas diffusion layers on both the anode and cathode. Previous LB models to simulate gas-liquid two-phase flows are applicable to low-density-ratio interfacial problems only, whereas the density ratio is large in practical fuel cell systems. To address this issue, a three-dimensional pseudo-potential-based LB model is developed to simulate gas-liquid two-phase flows with large density ratio. It is demonstrated that this three-dimensional model enables the density ratio to be as large as 700 in static and quasi-static case

An issue with the use Of Metallic lithium as an anode material for lithium-based batteries is,dendrite growth, causing a periodic breaking and repair of the solid electrolyte interphase (SEI) layer. Adding 2D atomic crystals, such as h-BN; as an interfacial layer between the lithium metal anode and liquid electrolyte has been demonstrated to be effective to mitigate dendrite growth, thereby enhancing the Columbic efficiency of lithium metal batteries. But the underlying mechanism leading to the reduced dendrite growth remains unknown. In this work, with the aid of first-principle calculations, we find that the interaction between the h-BN and lithium metal layers is a weak van der Waals force, and two atomic layers of h-BN are thick enough to block the electron tunneling from lithium metal to electrolyte, thus prohibiting the decomposition of electrolyte. The interlayer spacing between the h-BN and lithium metal layers can provide larger adsorption energies toward lithium atoms than that provided by bare lithium or h-BN, making lithium atoms prefer to intercalate under the cover of h-BN during the plating process. The combined high stiffness of h-BN and the low diffusion energy barriers of lithium at the Li/h-BN interfaces induce a uniform distribution of lithium, under h-BN, therefore effectively suppressing dendrite growth. © 2016 American Chemical Society.

We assess the performance of the combined Open Accelerator (OpenACC) programming standard and graphics processing unit (GPU) accelerator for lattice Boltzmann (LB) simulations of fluid flow, heat and mass transfer. By optimizing the data layout, minimizing the memory access frequency, and adjusting the number of gangs and vector length, we show that the enhanced parallel computations can result in orders of magnitudes of speedup relative to the serial implementation of the LB algorithm. Based on such implementations, benchmark quality results are obtained with fine grid of 20492 for both two-dimensional lid driven cavity flow with Reynolds number up to 7500, and double diffusive cavity flow with solute Rayleigh number up to 108. © 2017 Elsevier Ltd

In this paper, we present a three-dimensional lattice Boltzmann (LB) model to simulate suspensions that contain both micro- and nanoparticles. The microparticle dynamics is explicitly resolved by tracking individual solid particles, whilst the nanoparticles and base fluid are implicitly described as continua. The application of the LB model to simulate a micro spherical particle sedimentation in a microchannel and migration in a microtube filled with nanofluids show that the model allows the analysis of important parameters, including nanoparticle volume fraction, nanoparticle diameter, microchannel and microtube size. Finally, the model is applied to simulate particles in Couette flow. Through direct calculations of viscous dissipation, a mathematical correlation for viscosity as a function of micro- and nanoparticle volume fraction is proposed for the dilute suspension system. © 2016 Elsevier Inc.

Silicene has been predicted to be an extraordinary anode material for lithium-ion batteries with a large capacity and low lithium migration energy barriers, but the free-standing form of silicene is unstable, virtually requiring a substrate support. In this work, we propose to use graphene as a substrate and a protective layer of silicene, forming a van der Waals heterostructure of silicene and graphene (Si/G) to serve as a prospective anode material for lithium/sodium-ion batteries. Ab initio calculations show that the Si/G heterostructure not only preserves the silicene's large lithium/sodium capacity (487 mA h g-1) and low lithium/sodium migration energy barriers (<0.4 eV for lithium and <0.3 eV for sodium), but also provides much larger lithium/sodium binding energies via a synergistic effect, which can effectively inhibit the formation of dendrites. Density of states results show that the Si/G heterostructure is metallic before and after lithium/sodium intercalation, ensuring a good electronic conductivity. In addition, the mechanical stiffness of the Si/G heterostructure is found to be larger than that of pristine silicene or graphene, which helps preserve the structural integrity and enhance the cycle performance. © 2016 The Royal Society of Chemistry.

Density functional theory calculations and ab initio molecular dynamics simulations are performed to study the feasibility of using borophene, a newly synthesized two-dimensional sheet of boron, as an anode material for sodium-ion and sodium-oxygen batteries. The theoretical capacity of borophene is found to be as high as 1,218 mAh g(-1) (Na0.5B). More importantly, it is demonstrated that the sodium diffusion energy barrier along the valley direction is as low as 0.0019 eV, which corresponds to a diffusivity of more than a thousand times higher than that of conventional anode materials such as Na2Ti3O7 and Na3Sb. Hence, the use of borophene will revolutionize the rate capability of sodium-based batteries. Moreover, it is predicted that, during the sodiation process, the average open-circuit voltage is 0.53 V, which can effectively suppress the formation of dendrites while maximizing the energy density. The metallic feature and structural integrity of borophene can be well preserved at different sodium concentrations, demonstrating good electronic conductivity and stable cyclability.

Rutile RuO2 has been widely regarded as an excellent catalyst for the oxygen reduction reaction (ORR) and oxygen evolution reaction (OER) in nonaqueous lithium-oxygen batteries and achieved superior performance, but the catalytic activity of RuO2's polymorph, RuO2 monolayer, has been less studied. In this work, we study the catalytic activities of both rutile RuO2 and RuO2 monolayer for ORR and OER in the battery using density functional theory method. Computational results show that the RuO2 monolayer exhibits a higher catalytic activity than the rutile RuO2 does. More interestingly, it is found that during discharge a similar lattice structure between RuO2 monolayer and Li2O2 {0001} surface can induce the formation of crystallized Li2O2 with the conductive {0001} surface exposed, whereas during charge the RuO2 monolayer can attract the remaining Li2O2 to its surface spontaneously, thus maintaining the solid-solid reaction interface. Our results suggest that the RuO2 monolayer is a promising catalytic material for nonaqueous lithium-oxygen batteries.

Because of the involvement of solid-state discharge product Li2O2, how a catalyst works in nonaqueous lithium-oxygen batteries is yet to be determined, although the question has undergone fierce debate. In this work, we take an effective and widely used catalyst, rutile RuO2, as a representative and studied its catalytic mechanism in lithium-oxygen batteries via ab initio calculations. For the oxygen reduction reaction (ORR), it is found that rutile RuO2 can provide large adsorption energies toward LiO2 and Li2O2, thus resulting in high initial discharge voltages. Moreover, the normalized degree of unsaturation of surface oxygen is identified as a descriptor for the ORR catalytic activity. For the oxygen evolution reaction (OER), we propose that, in addition to the three-phase interface, the OER may also occur at the two-phase interface of Li2O2/RuO2, where rutile RuO2 provides pathways for the lithium ions while oxygen evolves from the exposed surfaces of Li2O2. Calculation results show that our proposed catalytic scenario is both thermodynamically and kinetically viable. Along with the charge process, the remaining Li2O2 can be attracted to the catalytic surfaces spontaneously, which can effectively preserve the reaction interface. © 2016 American Chemical Society.

Because of the involvement of solid-state discharge product Li2O2, how a catalyst works in nonaqueous lithium-oxygen batteries is yet to be determined, although the question has undergone fierce debate. In this work, we take an effective and widely used catalyst, rutile RuO2, as a representative and studied its catalytic mechanism in lithium-oxygen batteries via ab initio calculations. For the oxygen reduction reaction (ORR), it is found that rutile RuO2 can provide large adsorption energies toward LiO2 and Li2O2, thus resulting in high initial discharge voltages. Moreover, the normalized degree of unsaturation of surface oxygen is identified as a descriptor for the ORR catalytic activity. For the oxygen evolution reaction (OER), we propose that, in addition to the three-phase interface, the OER may also occur at the two-phase interface of Li2O2/RuO2, where rutile RuO2 provides pathways for the lithium ions while oxygen evolves from the exposed surfaces of Li2O2. Calculation results show that our proposed catalytic scenario is both thermodynamically and kinetically viable. Along with the charge process, the remaining Li2O2 can be attracted to the catalytic surfaces spontaneously, which can effectively preserve the reaction interface. © 2016 American Chemical Society.

Silicene has been predicted to be an extraordinary anode material for lithium-ion batteries with a large capacity and low lithium migration energy barriers, but the free-standing form of silicene is unstable, virtually requiring a substrate support. In this work, we propose to use graphene as a substrate and a protective layer of silicene, forming a van der Waals heterostructure of silicene and graphene (Si/G) to serve as a prospective anode material for lithium/sodium-ion batteries. Ab initio calculations show that the Si/G heterostructure not only preserves the silicene's large lithium/sodium capacity (487 mA h g-1) and low lithium/sodium migration energy barriers (<0.4 eV for lithium and <0.3 eV for sodium), but also provides much larger lithium/sodium binding energies via a synergistic effect, which can effectively inhibit the formation of dendrites. Density of states results show that the Si/G heterostructure is metallic before and after lithium/sodium intercalation, ensuring a good electronic conductivity. In addition, the mechanical stiffness of the Si/G heterostructure is found to be larger than that of pristine silicene or graphene, which helps preserve the structural integrity and enhance the cycle performance. © 2016 The Royal Society of Chemistry.

Density functional theory calculations and ab initio molecular dynamics simulations are performed to study the feasibility of using borophene, a newly synthesized two-dimensional sheet of boron, as an anode material for sodium-ion and sodium-oxygen batteries. The theoretical capacity of borophene is found to be as high as 1,218 mAh g(-1) (Na0.5B). More importantly, it is demonstrated that the sodium diffusion energy barrier along the valley direction is as low as 0.0019 eV, which corresponds to a diffusivity of more than a thousand times higher than that of conventional anode materials such as Na2Ti3O7 and Na3Sb. Hence, the use of borophene will revolutionize the rate capability of sodium-based batteries. Moreover, it is predicted that, during the sodiation process, the average open-circuit voltage is 0.53 V, which can effectively suppress the formation of dendrites while maximizing the energy density. The metallic feature and structural integrity of borophene can be well preserved at different sodium concentrations, demonstrating good electronic conductivity and stable cyclability.

Rutile RuO2 has been widely regarded as an excellent catalyst for the oxygen reduction reaction (ORR) and oxygen evolution reaction (OER) in nonaqueous lithium-oxygen batteries and achieved superior performance, but the catalytic activity of RuO2's polymorph, RuO2 monolayer, has been less studied. In this work, we study the catalytic activities of both rutile RuO2 and RuO2 monolayer for ORR and OER in the battery using density functional theory method. Computational results show that the RuO2 monolayer exhibits a higher catalytic activity than the rutile RuO2 does. More interestingly, it is found that during discharge a similar lattice structure between RuO2 monolayer and Li2O2 {0001} surface can induce the formation of crystallized Li2O2 with the conductive {0001} surface exposed, whereas during charge the RuO2 monolayer can attract the remaining Li2O2 to its surface spontaneously, thus maintaining the solid-solid reaction interface. Our results suggest that the RuO2 monolayer is a promising catalytic material for nonaqueous lithium-oxygen batteries.

Because of the involvement of solid-state discharge product Li2O2, how a catalyst works in nonaqueous lithium-oxygen batteries is yet to be determined, although the question has undergone fierce debate. In this work, we take an effective and widely used catalyst, rutile RuO2, as a representative and studied its catalytic mechanism in lithium-oxygen batteries via ab initio calculations. For the oxygen reduction reaction (ORR), it is found that rutile RuO2 can provide large adsorption energies toward LiO2 and Li2O2, thus resulting in high initial discharge voltages. Moreover, the normalized degree of unsaturation of surface oxygen is identified as a descriptor for the ORR catalytic activity. For the oxygen evolution reaction (OER), we propose that, in addition to the three-phase interface, the OER may also occur at the two-phase interface of Li2O2/RuO2, where rutile RuO2 provides pathways for the lithium ions while oxygen evolves from the exposed surfaces of Li2O2. Calculation results show that our proposed catalytic scenario is both thermodynamically and kinetically viable. Along with the charge process, the remaining Li2O2 can be attracted to the catalytic surfaces spontaneously, which can effectively preserve the reaction interface.

In this paper, we present a three-dimensional lattice Boltzmann (LB) model to simulate suspensions that contain both micro- and nanoparticles. The microparticle dynamics is explicitly resolved by tracking individual solid particles, whilst the nanoparticles and base fluid are implicitly described as continua. The application of the LB model to simulate a micro spherical particle sedimentation in a microchannel and migration in a microtube filled with nanofluids show that the model allows the analysis of important parameters, including nanoparticle volume fraction, nanoparticle diameter, microchannel and microtube size. Finally, the model is applied to simulate particles in Couette flow. Through direct calculations of viscous dissipation, a mathematical correlation for viscosity as a function of micro- and nanoparticle volume fraction is proposed for the dilute suspension system.

Rutile RuO2 has been widely regarded as an excellent catalyst for the oxygen reduction reaction (ORR) and oxygen evolution reaction (OER) in nonaqueous lithium-oxygen batteries and achieved superior performance, but the catalytic activity of RuO2's polymorph, RuO2 monolayer, has been less studied. In this work, we study the catalytic activities of both rutile RuO2 and RuO2 monolayer for ORR and OER in the battery using density functional theory method. Computational results show that the RuO2 monolayer exhibits a higher catalytic activity than the rutile RuO2 does. More interestingly, it is found that during discharge a similar lattice structure between RuO2 monolayer and Li2O2 {0001} surface can induce the formation of crystallized Li2O2 with the conductive {0001} surface exposed, whereas during charge the RuO2 monolayer can attract the remaining Li2O2 to its surface spontaneously, thus maintaining the solid-solid reaction interface. Our results suggest that the RuO2 monolayer is a promising catalytic material for nonaqueous lithium-oxygen batteries.

Silicene has been predicted to be an extraordinary anode material for lithium-ion batteries with a large capacity and low lithium migration energy barriers, but the free-standing form of silicene is unstable, virtually requiring a substrate support. In this work, we propose to use graphene as a substrate and a protective layer of silicene, forming a van der Waals heterostructure of silicene and graphene (Si/G) to serve as a prospective anode material for lithium/sodium-ion batteries. Ab initio calculations show that the Si/G heterostructure not only preserves the silicene's large lithium/sodium capacity (487 mA h g-1) and low lithium/sodium migration energy barriers (<0.4 eV for lithium and <0.3 eV for sodium), but also provides much larger lithium/sodium binding energies via a synergistic effect, which can effectively inhibit the formation of dendrites. Density of states results show that the Si/G heterostructure is metallic before and after lithium/sodium intercalation, ensuring a good electronic conductivity. In addition, the mechanical stiffness of the Si/G heterostructure is found to be larger than that of pristine silicene or graphene, which helps preserve the structural integrity and enhance the cycle performance.

Density functional theory calculations and ab initio molecular dynamics simulations are performed to study the feasibility of using borophene, a newly synthesized two-dimensional sheet of boron, as an anode material for sodium-ion and sodium-oxygen batteries. The theoretical capacity of borophene is found to be as high as 1,218 mAh g-1 (Na0.5B). More importantly, it is demonstrated that the sodium diffusion energy barrier along the valley direction is as low as 0.0019 eV, which corresponds to a diffusivity of more than a thousand times higher than that of conventional anode materials such as Na2Ti3O7 and Na3Sb. Hence, the use of borophene will revolutionize the rate capability of sodium-based batteries. Moreover, it is predicted that, during the sodiation process, the average open-circuit voltage is 0.53 V, which can effectively suppress the formation of dendrites while maximizing the energy density. The metallic feature and structural integrity of borophene can be well preserved at different sodium concentrations, demonstrating good electronic conductivity and stable cyclability.

In this paper, we present a three-dimensional pseudo-potential-based lattice Boltzmann (LB) model with an improved forcing scheme for multiphase flows. The Chapman-Enskog multiscale analysis shows that the proposed forcing scheme allows the lattice Boltzmann equation to recover the three-dimensional hydrodynamical equations with additional terms that correspond to the mechanical stability condition and surface tension. Validations of the present LB model with Maxwell construction, Laplace's law and oscillation dynamics demonstrate that the model enables the density ratio to be as large as 700 in static and quasi-static cases while maintaining variable surface tension. Finally, the application of the model to simulation of the droplet motion in a microchannel shows that the model allows the analysis of important effects, including droplet surface tension, channel surface wettability, and channel surface roughness. © 2015 Elsevier Inc.

Density functional theory calculations are made for bulk thermodynamic properties and surface energies of Li2O2, a primary discharge product, and Li3O4, a possible byproduct in the discharge products, of the non-aqueous lithium-oxygen batteries. Results show that the standard formation Gibbs free energy of bulk Li3O4 is marginally higher than that of Li2O2, but the surface energy of Li3O4 is much lower. Low surface energy results in both lowered nucleation energy and formation Gibbs free energy in the nanometer regime, allowing the Li3O4 nano particles to nucleate ahead of Li2O2 during the discharge process and to exist stably when particle sizes are smaller than about 40 nm. The scanning transmission electron microscopy (STEM) image of Li3O4 crystals is simulated and compared with the measured STEM image of the discharge product particles. The consistency between the simulated and measured STEM images suggests that the Li3O4 phase can exist stably as a discharge product. The energy profile of the oxygen evolution reaction (OER) occurring on the most abundant surfaces of Li3O4 is also calculated. The predicted overpotential for the OER on the {0001} surface (0.30 V) shows a good agreement with experimental data. The presence of more electronically conductive Li3O4 nano particles in the primary discharge product Li2O2 tends to decrease the charge overvoltage of the batteries, explaining why the lower voltage area (<3.5 V) was widely observed during the charging of the batteries. An increase in the oxygen pressure or a decrease in temperature enhances the stability of the Li3O4 phase and increase the proportion of the Li3O4 phase in the discharge products, consequently leading to a lower overall charge overvoltage. © the Owner Societies 2015.

Density functional theory calculations are made for bulk thermodynamic properties and surface energies of Li2O2, a primary discharge product, and Li3O4, a possible byproduct in the discharge products, of the non-aqueous lithium-oxygen batteries. Results show that the standard formation Gibbs free energy of bulk Li3O4 is marginally higher than that of Li2O2, but the surface energy of Li3O4 is much lower. Low surface energy results in both lowered nucleation energy and formation Gibbs free energy in the nanometer regime, allowing the Li3O4 nano particles to nucleate ahead of Li2O2 during the discharge process and to exist stably when particle sizes are smaller than about 40 nm. The scanning transmission electron microscopy (STEM) image of Li3O4 crystals is simulated and compared with the measured STEM image of the discharge product particles. The consistency between the simulated and measured STEM images suggests that the Li3O4 phase can exist stably as a discharge product. The energy profile of the oxygen evolution reaction (OER) occurring on the most abundant surfaces of Li3O4 is also calculated. The predicted overpotential for the OER on the {0001} surface (0.30 V) shows a good agreement with experimental data. The presence of more electronically conductive Li3O4 nano particles in the primary discharge product Li2O2 tends to decrease the charge overvoltage of the batteries, explaining why the lower voltage area (<3.5 V) was widely observed during the charging of the batteries. An increase in the oxygen pressure or a decrease in temperature enhances the stability of the Li3O4 phase and increase the proportion of the Li3O4 phase in the discharge products, consequently leading to a lower overall charge overvoltage. © the Owner Societies 2015.

In this paper, we present a three-dimensional pseudo-potential-based lattice Boltzmann (LB) model with an improved forcing scheme for multiphase flows. The Chapman-Enskog multiscale analysis shows that the proposed forcing scheme allows the lattice Boltzmann equation to recover the three-dimensional hydrodynamical equations with additional terms that correspond to the mechanical stability condition and surface tension. Validations of the present LB model with Maxwell construction, Laplace's law and oscillation dynamics demonstrate that the model enables the density ratio to be as large as 700 in static and quasi-static cases while maintaining variable surface tension. Finally, the application of the model to simulation of the droplet motion in a microchannel shows that the model allows the analysis of important effects, including droplet surface tension, channel surface wettability, and channel surface roughness. © 2015 Elsevier Inc.

In this paper, we present a three-dimensional pseudo-potential-based lattice Boltzmann (LB) model with an improved forcing scheme for multiphase flows. The Chapman-Enskog multiscale analysis shows that the proposed forcing scheme allows the lattice Boltzmann equation to recover the three-dimensional hydrodynamical equations with additional terms that correspond to the mechanical stability condition and surface tension. Validations of the present LB model with Maxwell construction, Laplace's law and oscillation dynamics demonstrate that the model enables the density ratio to be as large as 700 in static and quasi-static cases while maintaining variable surface tension. Finally, the application of the model to simulation of the droplet motion in a microchannel shows that the model allows the analysis of important effects, including droplet surface tension, channel surface wettability, and channel surface roughness. © 2015 Elsevier Inc.

Density functional theory calculations are made for bulk thermodynamic properties and surface energies of Li2O2, a primary discharge product, and Li3O4, a possible byproduct in the discharge products, of the non-aqueous lithium-oxygen batteries. Results show that the standard formation Gibbs free energy of bulk Li3O4 is marginally higher than that of Li2O2, but the surface energy of Li3O4 is much lower. Low surface energy results in both lowered nucleation energy and formation Gibbs free energy in the nanometer regime, allowing the Li3O4 nano particles to nucleate ahead of Li2O2 during the discharge process and to exist stably when particle sizes are smaller than about 40 nm. The scanning transmission electron microscopy (STEM) image of Li3O4 crystals is simulated and compared with the measured STEM image of the discharge product particles. The consistency between the simulated and measured STEM images suggests that the Li3O4 phase can exist stably as a discharge product. The energy profile of the oxygen evolution reaction (OER) occurring on the most abundant surfaces of Li3O4 is also calculated. The predicted overpotential for the OER on the {0001} surface (0.30 V) shows a good agreement with experimental data. The presence of more electronically conductive Li3O4 nano particles in the primary discharge product Li2O2 tends to decrease the charge overvoltage of the batteries, explaining why the lower voltage area (<3.5 V) was widely observed during the charging of the batteries. An increase in the oxygen pressure or a decrease in temperature enhances the stability of the Li3O4 phase and increase the proportion of the Li3O4 phase in the discharge products, consequently leading to a lower overall charge overvoltage. © the Owner Societies 2015.

[徐璈撰 ; 龔橙校], 獨山莫氏藏書., Duo shan Mo shi cang shu., Block print., In oriental style., [Xu Ao zhuan ; Gong Cheng jiao]