Showing 1–7 of 7 results for author: Tan, G

Numerical investigations on interactions between 2D/3D conical shock wave and axisymmetric boundary layer at Ma=2.2

Authors: Yihui Weng, Qin Li, Guozhuo Tan, Wei Su, Yancheng You

Abstract: Numerical simulation and analysis are carried out on interactions between a 2D/3D conical shock wave and an axisymmetric boundary layer with reference to the experiment by Kussoy et al., in which the shock was generated by a 15-deg half-angle cone in a tube at 15-deg angle of attack (AOA). Based on the RANS equations and Menter's SST turbulence model, the present study uses the newly developed WEN… ▽ More Numerical simulation and analysis are carried out on interactions between a 2D/3D conical shock wave and an axisymmetric boundary layer with reference to the experiment by Kussoy et al., in which the shock was generated by a 15-deg half-angle cone in a tube at 15-deg angle of attack (AOA). Based on the RANS equations and Menter's SST turbulence model, the present study uses the newly developed WENO3-PRM211 scheme and the PHengLEI CFD platform for the computations. First, computations are performed for the 3D interaction corresponding to the conditions of the experiment by Kussoy et al., and these are then extended to cases with AOA = 10-deg and 5-deg. For comparison, 2D axisymmetric counterparts of the 3D interactions are investigated for cones coaxial with the tube and having half-cone angles of 27.35-deg, 24.81-deg, and 20.96-deg. The shock wave structure, vortex structure, variable distributions, and wall separation topology of the interaction are computed. The results show that in 2D/3D interactions, a new Mach reflection-like event occurs and a Mach stem-like structure is generated above the front of the separation bubble, which differs from the model of Babinsky for 2D planar shock wave/boundary layer interaction. A new interaction model is established to describe this behavior. The relationship between the length of the circumferentially unseparated region in the tube and the AOA of the cone indicates the existence of a critical AOA at which the length is zero, and a prediction of this angle is obtained using an empirical fit, which is verified by computation. The occurrence of side overflow in the windward meridional plane is analyzed, and a quantitative knowledge is obtained. To elucidate the characteristics of the 3D interaction, the scale and structure of the vortex and the pressure and friction force distributions are presented and compared with those of the 2D interaction. △ Less

Submitted 24 March, 2023; v1 submitted 22 March, 2023; originally announced March 2023.

RLEKF: An Optimizer for Deep Potential with Ab Initio Accuracy

Authors: Siyu Hu, Wentao Zhang, Qiuchen Sha, Feng Pan, Lin-Wang Wang, Weile Jia, Guangmng Tan, Tong Zhao

Abstract: It is imperative to accelerate the training of neural network force field such as Deep Potential, which usually requires thousands of images based on first-principles calculation and a couple of days to generate an accurate potential energy surface. To this end, we propose a novel optimizer named reorganized layer extended Kalman filtering (RLEKF), an optimized version of global extended Kalman fi… ▽ More It is imperative to accelerate the training of neural network force field such as Deep Potential, which usually requires thousands of images based on first-principles calculation and a couple of days to generate an accurate potential energy surface. To this end, we propose a novel optimizer named reorganized layer extended Kalman filtering (RLEKF), an optimized version of global extended Kalman filtering (GEKF) with a strategy of splitting big and gathering small layers to overcome the $O(N^2)$ computational cost of GEKF. This strategy provides an approximation of the dense weights error covariance matrix with a sparse diagonal block matrix for GEKF. We implement both RLEKF and the baseline Adam in our $α$Dynamics package and numerical experiments are performed on 13 unbiased datasets. Overall, RLEKF converges faster with slightly better accuracy. For example, a test on a typical system, bulk copper, shows that RLEKF converges faster by both the number of training epochs ($\times$11.67) and wall-clock time ($\times$1.19). Besides, we theoretically prove that the updates of weights converge and thus are against the gradient exploding problem. Experimental results verify that RLEKF is not sensitive to the initialization of weights. The RLEKF sheds light on other AI-for-science applications where training a large neural network (with tons of thousands parameters) is a bottleneck. △ Less

Submitted 13 December, 2022; originally announced December 2022.

Traffic disruption modelling with mode shift in multi-modal networks

Authors: Dong Zhao, Adriana-Simona Mihaita, Yuming Ou, Sajjad Shafiei, Hanna Grzybowska, A. K. Qin, Gary Tan, Mo Li, Hussein Dia

Abstract: A multi-modal transport system is acknowledged to have robust failure tolerance and can effectively relieve urban congestion issues. However, estimating the impact of disruptions across multi-transport modes is a challenging problem due to a dis-aggregated modelling approach applied to only individual modes at a time. To fill this gap, this paper proposes a new integrated modelling framework for a… ▽ More A multi-modal transport system is acknowledged to have robust failure tolerance and can effectively relieve urban congestion issues. However, estimating the impact of disruptions across multi-transport modes is a challenging problem due to a dis-aggregated modelling approach applied to only individual modes at a time. To fill this gap, this paper proposes a new integrated modelling framework for a multi-modal traffic state estimation and evaluation of the disruption impact across all modes under various traffic conditions. First, we propose an iterative trip assignment model to elucidate the association between travel demand and travel behaviour, including a multi-modal origin-to-destination estimation for private and public transport. Secondly, we provide a practical multi-modal travel demand re-adjustment that takes the mode shift of the affected travellers into consideration. The pros and cons of the mode shift strategy are showcased via several scenario-based transport simulating experiments. The results show that a well-balanced mode shift with flexible routing and early announcements of detours so that travellers can plan ahead can significantly benefit all travellers by a delay time reduction of 46%, while a stable route assignment maintains a higher average traffic flow and the inactive mode-route choice help relief density under the traffic disruptions. △ Less

Submitted 12 October, 2022; originally announced October 2022.

Comments: 8 pages, 7 figures, 1 table, The 25th IEEE International Conference on Intelligent Transportation Systems (IEEE ITSC 2022)

Improvements to enhance robustness of third-order scale-independent WENO-Z schemes

Authors: Qin Li, Xiao Huang, Pan Yan, Guozhuo Tan, Yi Duan, Yancheng You

Abstract: Although there are many improvements to WENO3-Z that target the achievement of optimal order in the occurrence of the first-order critical point (CP1), they mainly address resolution performance, while the robustness of schemes is of less concern and lacks understanding accordingly. In light of our analysis considering the occurrence of critical points within grid intervals, we theoretically prove… ▽ More Although there are many improvements to WENO3-Z that target the achievement of optimal order in the occurrence of the first-order critical point (CP1), they mainly address resolution performance, while the robustness of schemes is of less concern and lacks understanding accordingly. In light of our analysis considering the occurrence of critical points within grid intervals, we theoretically prove that it is impossible for a scale-independent scheme that has the stencil of WENO3-Z to fulfill the above order achievement, and current scale-dependent improvements barely fulfill the job when CP1 occurs at the middle of the grid cell. In order to achieve scale-independent improvements, we devise new smoothness indicators that increase the error order from 2 to 4 when CP1 occurs and perform more stably. Meanwhile, we construct a new global smoothness indicator that increases the error order from 4 to 5 similarly, through which new nonlinear weights with regard to WENO3-Z are derived and new scale-independents improvements, namely WENO-ZES2 and -ZES3, are acquired. Through 1D scalar and Euler tests, as well as 2D computations, in comparison with typical scale-dependent improvement, the following performances of the proposed schemes are demonstrated: The schemes can achieve third-order accuracy at CP1 no matter its location in the stencil, indicate high resolution in resolving flow subtleties, and manifest strong robustness in hypersonic simulations (e.g., the accomplishment of computations on hypersonic half-cylinder flow with Mach numbers reaching 16 and 19, respectively, as well as essentially non-oscillatory solutions of inviscid sharp double cone flow at M=9.59), which contrasts the comparative WENO3-Z improvement. △ Less

Submitted 1 August, 2022; originally announced August 2022.

Multi-level scalar structure in complex system analyses

Authors: Lipo Wang, Guiwen Tan, Hui Cao

Abstract: The geometrical structure is among the most fundamental ingredients in understanding complex systems. Is there any systematic approach in defining structures quantitatively, rather than illustratively? If yes, what are the basic principles to follow? By introducing the concept of extremal points at different scale levels, a multi-level dissipation element approach has been developed to define stru… ▽ More The geometrical structure is among the most fundamental ingredients in understanding complex systems. Is there any systematic approach in defining structures quantitatively, rather than illustratively? If yes, what are the basic principles to follow? By introducing the concept of extremal points at different scale levels, a multi-level dissipation element approach has been developed to define structures at different scale levels, in accordance with the concept of structure hierarchy. Each dissipation element can be characterized by the length scale and the scalar variance inside. Using the two-dimensional fractal Brownian motion as a benchmark case, the conditional mean of the scalar difference with respect to the length scale shows clearly a power law and the scaling exponent is in agreement with the Hurst number. For the 3D turbulence velocity component, the 1/3 scaling law can be represented. These results indicate the important linkage between the turbulence physics and ow structure, if well posed and defined. In principle, the multi-level dissipation element idea is generally applicable in analyzing other multiscale complex systems as well. △ Less

Submitted 26 May, 2020; v1 submitted 21 November, 2019; originally announced November 2019.

Comments: This paper will be revised and new results will be added

Effect of cap** layer on spin-orbit torques

Authors: Chi Sun, Zhuo Bin Siu, Seng Ghee Tan, Hyunsoo Yang, Mansoor B. A. Jalil

Abstract: In order to enhance the magnitude of spin-orbit torque (SOT), considerable experimental works have been devoted to studying the thickness dependence of the different layers in multilayers consisting of heavy metal (HM), ferromagnet (FM) and cap** layers. Here we present a theoretical model based on the spin-drift-diffusion (SDD) formalism to investigate the effect of the cap** layer properties… ▽ More In order to enhance the magnitude of spin-orbit torque (SOT), considerable experimental works have been devoted to studying the thickness dependence of the different layers in multilayers consisting of heavy metal (HM), ferromagnet (FM) and cap** layers. Here we present a theoretical model based on the spin-drift-diffusion (SDD) formalism to investigate the effect of the cap** layer properties such as its thickness on the SOT observed in experiments. It is found that the spin Hall-induced SOT can be significantly enhanced by incorporating a cap** layer with opposite spin Hall angle to that of the HM layer. The spin Hall torque can be maximized by tuning the cap** layer thickness. However, in the absence of the spin Hall effect (SHE) in the cap** layer, the torque decreases monotonically with cap** layer thickness. Conversely, the spin Hall torque is found to decrease monotonically with the FM layer thickness, irrespective of the presence or absence of SHE in the cap** layer. All these trends are in correspondence with experimental observations. Finally, our model suggests that cap** layers with long spin diffusion length and high resistivity would also enhance the spin Hall torque. △ Less

Submitted 19 April, 2018; originally announced April 2018.

Visible Photocatalytic Degradation of Methylene Blue on Magnetic Semiconducting La modified M-type Strontium Hexaferrite

Authors: Debesh Devadutta Mishra, Yao Huang, Guolong Tan

Abstract: Methylene blue (MB) is a representative of a class of dyestuffs resistant to biodegradation. This paper presents a novel photocatalytic degradation of methylene blue (MB) by La modified M-type strontium hexaferrite (La0.2Sr0.7Fe12O19) compound, which is a traditional permanent magnet and displays a large magnetic hysteresis (M-H) loop. The remnant magnetic moment and coercive field are determined… ▽ More Methylene blue (MB) is a representative of a class of dyestuffs resistant to biodegradation. This paper presents a novel photocatalytic degradation of methylene blue (MB) by La modified M-type strontium hexaferrite (La0.2Sr0.7Fe12O19) compound, which is a traditional permanent magnet and displays a large magnetic hysteresis (M-H) loop. The remnant magnetic moment and coercive field are determined to be 52emu/g and 5876Oe, respectively. UV-Visible optical spectroscopy reveals that La0.2Sr0.7Fe12O19 is simultaneously a semiconductor, whose direct and indirect band gap energies are determined to be 1.47 eV and 0.88 eV, respectively. The near infrared band gap makes it a good candidate to harvest sunlight for photocatalytic reaction or solar cell devices. This magnetic compound demonstrates excellent photocatalytic activity on degradation of MB under visible illumination. The colour of MB dispersion solution changes from deep blue to pale white and the absorbance decreases rapidly from 1.8 down to zero when the illumination duration extends to 6 hours. Five absorption bands did not make any blue shifts along with the reaction time, suggesting a one-stepwise degradation process of MB, which makes La0.2Sr0.7Fe12O19 a unique magnetic catalyst and differs from TiO2 and other conventional catalysts. △ Less

Submitted 12 December, 2017; originally announced December 2017.

Comments: 5 Figures