skip to main content


Search for: All records

Creators/Authors contains: "Cheng, Xinlun"

Note: When clicking on a Digital Object Identifier (DOI) number, you will be taken to an external site maintained by the publisher. Some full text articles may not yet be available without a charge during the embargo (administrative interval).
What is a DOI Number?

Some links on this page may take you to non-federal websites. Their policies may differ from this site.

  1. Modeling unsteady, fast transient, and advection-dominated physics problems is a pressing challenge for physics-aware deep learning (PADL). The physics of complex systems is governed by large systems of partial differential equations (PDEs) and ancillary constitutive models with nonlinear structures, as well as evolving state fields exhibiting sharp gradients and rapidly deforming material interfaces. Here, we investigate an inductive bias approach that is versatile and generalizable to model generic nonlinear field evolution problems. Our study focuses on the recent physics-aware recurrent convolutions (PARC), which incorporates a differentiator-integrator architecture that inductively models the spatiotemporal dynamics of generic physical systems. We extend the capabilities of PARC to simulate unsteady, transient, and advection-dominant systems. The extended model, referred to as PARCv2, is equipped with differential operators to model advection-reaction-diffusion equations, as well as a hybrid integral solver for stable, long-time predictions. PARCv2 is tested on both standard benchmark problems in fluid dynamics, namely Burgers and Navier-Stokes equations, and then applied to more complex shock-induced reaction problems in energetic materials. We evaluate the behavior of PARCv2 in comparison to other physics-informed and learning bias models and demonstrate its potential to model unsteady and advection-dominant dynamics regimes. 
    more » « less
    Free, publicly-accessible full text available July 1, 2025
  2. ABSTRACT

    We revisit the classical KZ problem – determination of the vertical force and implied total mass density distribution of the Milky Way disc – for a wide range of Galactocentric radius and vertical height using chemically selected thin and thick disc samples based on Apache Point Observatory Galactic Evolution Experiment spectroscopy combined with the Gaia astrometry. We derived the velocity dispersion profiles in Galactic cylindrical coordinates, and solved the Jeans equation for the two samples separately. The result is surprising that the total surface mass density as a function of vertical height as derived for these two chemically distinguished populations is different. The discrepancies are larger in the inner compared to the outer Galaxy, with the density calculated from thick disc being larger, independent of the Galactic radius. Furthermore, while there is an overall good agreement between the total mass density derived for the thick disc population and the standard halo model for vertical heights larger than 1 kpc, close to the mid-plane the mass density observed using the thick disc population is larger than that predicted from the standard halo model. We explore various implications of these discrepancies, and speculate their sources, including problems associated with the assumed density laws, velocity dispersion profiles, and the Galactic rotation curve, potential non-equilibrium of the Galactic disc, or a failure of the Navarro-Frenk-White (NFW) dark matter halo profile for the Milky Way. We conclude that the growing detail in hand on the chemodynamical distributions of Milky Way stars challenges traditional analytical treatments of the KZ problem.

     
    more » « less
  3. Complex spin textures in itinerant electron magnets hold promises for next-generation memory and information technology. The long-ranged and often frustrated electron-mediated spin interactions in these materials give rise to intriguing localized spin structures such as skyrmions. Yet, simulations of magnetization dynamics for such itinerant magnets are computationally difficult due to the need for repeated solutions to the electronic structure problems. We present a convolutional neural network (CNN) model to accurately and efficiently predict the electron-induced magnetic torques acting on local spins. Importantly, as the convolutional operations with a fixed kernel (receptive field) size naturally take advantage of the locality principle for many-electron systems, CNNs offer a scalable machine learning approach to spin dynamics. We apply our approach to enable large-scale dynamical simulations of skyrmion phases in itinerant spin systems. By incorporating the CNN model into Landau-Lifshitz-Gilbert dynamics, our simulations successfully reproduce the relaxation process of the skyrmion phase and stabilize a skyrmion lattice in larger systems. The CNN model also allows us to compute the effective receptive fields, thus providing a systematic and unbiased method for determining the locality of the original electron models. 
    more » « less
  4. We present the first detailed chemical analysis from APOGEE-2S observations of stars in six regions of recently discovered substructures in the outskirts of the Magellanic Clouds extending to 20° from the Large Magellanic Cloud (LMC) center. We also present, for the first time, the metallicity andα-abundance radial gradients of the LMC and the Small Magellanic Cloud (SMC) out to 11° and 6°, respectively. Our chemical tagging includes 13 species including light,α-, and Fe-peak elements. We find that the abundances of all of these chemical elements in stars populating two regions in the northern periphery, along the northern “stream-like” feature, show good agreement with the chemical patterns of the LMC, and thus likely have an LMC origin. For substructures located in the southern periphery of the LMC we find more complex chemical and kinematical signatures, indicative of a mix of LMC-like and SMC-like populations. The southern region closest to the LMC shows better agreement with the LMC, whereas that closest to the SMC shows a much better agreement with the SMC chemical pattern. When combining this information with 3D kinematical information for these stars, we conclude that the southern region closest to the LMC likely has an LMC origin, whereas that closest to the SMC has an SMC origin and the other two southern regions have a mix of LMC and SMC origins. Our results add to the evidence that the southern substructures of the LMC periphery are the product of close interactions between the LMC and SMC, and thus likely hold important clues that can constrain models of their detailed dynamical histories.

     
    more » « less
  5. Abstract We report the first 3D kinematical measurements of 88 stars in the direction of several recently discovered substructures in the southern periphery of the Large Magellanic Cloud (LMC) using a combination of Gaia proper motions and radial velocities from the APOGEE-2 survey. More specifically, we explore stars in assorted APOGEE-2 pointings in a region of the LMC periphery where various overdensities of stars have previously been identified in maps of stars from Gaia and DECam. By using a model of the LMC disk rotation, we find that a sizable fraction of the APOGEE-2 stars have extreme space velocities that are distinct from, and not a simple extension of, the LMC disk. Using N -body hydrodynamical simulations of the past dynamical evolution and interaction of the LMC and Small Magellanic Cloud (SMC), we explore whether the extreme-velocity stars may be accounted for as tidal debris created in the course of that interaction. We conclude that the combination of LMC and SMC debris produced from their interaction is a promising explanation, although we cannot rule out other possible origins, and that these new data should be used to constrain future simulations of the LMC–SMC interaction. We also conclude that many of the stars in the southern periphery of the LMC lie outside of the LMC plane by several kiloparsecs. Given that the metallicity of these stars suggests that they are likely of Magellanic origin, our results suggest that a wider exploration of the past interaction history of the Magellanic Clouds is needed. 
    more » « less