skip to main content
US FlagAn official website of the United States government
dot gov icon
Official websites use .gov
A .gov website belongs to an official government organization in the United States.
https lock icon
Secure .gov websites use HTTPS
A lock ( lock ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites.


Title: Magnetohydrodynamics with physics informed neural operators
Abstract The modeling of multi-scale and multi-physics complex systems typically involves the use of scientific software that can optimally leverage extreme scale computing. Despite major developments in recent years, these simulations continue to be computationally intensive and time consuming. Here we explore the use of AI to accelerate the modeling of complex systems at a fraction of the computational cost of classical methods, and present the first application of physics informed neural operators (NOs) (PINOs) to model 2D incompressible magnetohydrodynamics (MHD) simulations. Our AI models incorporate tensor Fourier NOs as their backbone, which we implemented with theTensorLYpackage. Our results indicate that PINOs can accurately capture the physics of MHD simulations that describe laminar flows with Reynolds numbers Re 250 . We also explore the applicability of our AI surrogates for turbulent flows, and discuss a variety of methodologies that may be incorporated in future work to create AI models that provide a computationally efficient and high fidelity description of MHD simulations for a broad range of Reynolds numbers. The scientific software developed in this project is released with this manuscript.  more » « less
Award ID(s):
1931561
PAR ID:
10429345
Author(s) / Creator(s):
;
Publisher / Repository:
IOP Publishing
Date Published:
Journal Name:
Machine Learning: Science and Technology
Volume:
4
Issue:
3
ISSN:
2632-2153
Page Range / eLocation ID:
Article No. 035002
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. AbstractPrevious theoretical and simulation results indicate that anisotropic porous materials have the potential to reduce turbulent skin friction in wall-bounded flows. This study experimentally investigates the influence of anisotropy on the drag response of porous substrates. A family of anisotropic periodic lattices was manufactured using 3D printing. Rod spacing in different directions was varied systematically to achieve different ratios of streamwise, wall-normal, and spanwise bulk permeabilities ($$\kappa _{xx}$$ κ xx ,$$\kappa _{yy}$$ κ yy , and$$\kappa _{zz}$$ κ zz ). The 3D printed materials were flush-mounted in a benchtop water channel. Pressure drop measurements were taken in the fully developed region of the flow to systematically characterize drag for materials with anisotropy ratios$$\frac{\kappa _{xx}}{\kappa _{yy}} \in [0.035,28.6]$$ κ xx κ yy [ 0.035 , 28.6 ] . Results show that all materials lead to an increase in drag compared to the reference smooth wall case over the range of bulk Reynolds numbers tested ($$\hbox {Re}_b \in [500,4000]$$ Re b [ 500 , 4000 ] ). However, the relative increase in drag is lower for streamwise-preferential materials. We estimate that the wall-normal permeability for all tested cases exceeded the threshold identified in previous literature ($$\sqrt{\kappa _{yy}}^+> 0.4$$ κ yy + > 0.4 ) for the emergence of energetic spanwise rollers similar to Kelvin–Helmholtz vortices, which can increase drag. The results also indicate that porous walls exhibit a departure from laminar behavior at different values for bulk Reynolds numbers depending on the geometry. Graphical abstract 
    more » « less
  2. Abstract We report a gravitational-wave parameter estimation algorithm,AMPLFI, based on likelihood-free inference using normalizing flows. The focus ofAMPLFIis to perform real-time parameter estimation for candidates detected by machine-learning based compact binary coalescence search,Aframe. We present details of our algorithm and optimizations done related to data-loading and pre-processing on accelerated hardware. We train our model using binary black-hole (BBH) simulations on real LIGO-Virgo detector noise. Our model has 6 million trainable parameters with training times 24 h. Based on online deployment on a mock data stream of LIGO-Virgo data,Aframe+AMPLFIis able to pick up BBH candidates and infer parameters for real-time alerts from data acquisition with a net latency of 6 s. 
    more » « less
  3. Abstract The production of a pair of τ leptons via photon–photon fusion, γ γ τ τ , is observed for the first time in proton–proton collisions, with a significance of 5.3 standard deviations. This observation is based on a data set recorded with the CMS detector at the LHC at a center-of-mass energy of 13 TeV and corresponding to an integrated luminosity of 138 fb−1. Events with a pair of τ leptons produced via photon–photon fusion are selected by requiring them to be back-to-back in the azimuthal direction and to have a minimum number of charged hadrons associated with their production vertex. The τ leptons are reconstructed in their leptonic and hadronic decay modes. The measured fiducial cross section of γ γ τ τ is σ obs fid = 12.4 3.1 + 3.8 fb . Constraints are set on the contributions to the anomalous magnetic moment ( a τ ) and electric dipole moments ( d τ ) of the τ lepton originating from potential effects of new physics on the γ τ τ vertex: a τ = 0.0009 0.0031 + 0.0032 and | d τ | < 2.9 × 10 17 e cm (95% confidence level), consistent with the standard model. 
    more » « less
  4. Abstract A measurement is presented of a ratio observable that provides a measure of the azimuthal correlations among jets with large transverse momentum$$p_{\textrm{T}}$$ p T . This observable is measured in multijet events over the range of$$p_{\textrm{T}} = 360$$ p T = 360 –$$3170\,\text {Ge}\hspace{-.08em}\text {V} $$ 3170 Ge V based on data collected by the CMS experiment in proton-proton collisions at a centre-of-mass energy of 13$$\,\text {Te}\hspace{-.08em}\text {V}$$ Te V , corresponding to an integrated luminosity of 134$$\,\text {fb}^{-1}$$ fb - 1 . The results are compared with predictions from Monte Carlo parton-shower event generator simulations, as well as with fixed-order perturbative quantum chromodynamics (pQCD) predictions at next-to-leading-order (NLO) accuracy obtained with different parton distribution functions (PDFs) and corrected for nonperturbative and electroweak effects. Data and theory agree within uncertainties. From the comparison of the measured observable with the pQCD prediction obtained with the NNPDF3.1 NLO PDFs, the strong coupling at the Z boson mass scale is$$\alpha _\textrm{S} (m_{{\textrm{Z}}}) =0.1177 \pm 0.0013\, \text {(exp)} _{-0.0073}^{+0.0116} \,\text {(theo)} = 0.1177_{-0.0074}^{+0.0117}$$ α S ( m Z ) = 0.1177 ± 0.0013 (exp) - 0.0073 + 0.0116 (theo) = 0 . 1177 - 0.0074 + 0.0117 , where the total uncertainty is dominated by the scale dependence of the fixed-order predictions. A test of the running of$$\alpha _\textrm{S}$$ α S in the$$\,\text {Te}\hspace{-.08em}\text {V}$$ Te V region shows no deviation from the expected NLO pQCD behaviour. 
    more » « less
  5. Abstract A search is reported for charge-parity$$CP$$ CP violation in$${{{\textrm{D}}}^{{0}}} \rightarrow {{\textrm{K}} _{\text {S}}^{{0}}} {{\textrm{K}} _{\text {S}}^{{0}}} $$ D 0 K S 0 K S 0 decays, using data collected in proton–proton collisions at$$\sqrt{s} = 13\,\text {Te}\hspace{-.08em}\text {V} $$ s = 13 Te V recorded by the CMS experiment in 2018. The analysis uses a dedicated data set that corresponds to an integrated luminosity of 41.6$$\,\text {fb}^{-1}$$ fb - 1 , which consists of about 10 billion events containing a pair of b hadrons, nearly all of which decay to charm hadrons. The flavor of the neutral D meson is determined by the pion charge in the reconstructed decays$${{{\textrm{D}}}^{{*+}}} \rightarrow {{{\textrm{D}}}^{{0}}} {{{\mathrm{\uppi }}}^{{+}}} $$ D + D 0 π + and$${{{\textrm{D}}}^{{*-}}} \rightarrow {\overline{{\textrm{D}}}^{{0}}} {{{\mathrm{\uppi }}}^{{-}}} $$ D - D ¯ 0 π - . The$$CP$$ CP asymmetry in$${{{\textrm{D}}}^{{0}}} \rightarrow {{\textrm{K}} _{\text {S}}^{{0}}} {{\textrm{K}} _{\text {S}}^{{0}}} $$ D 0 K S 0 K S 0 is measured to be$$A_{CP} ({{\textrm{K}} _{\text {S}}^{{0}}} {{\textrm{K}} _{\text {S}}^{{0}}} ) = (6.2 \pm 3.0 \pm 0.2 \pm 0.8)\%$$ A CP ( K S 0 K S 0 ) = ( 6.2 ± 3.0 ± 0.2 ± 0.8 ) % , where the three uncertainties represent the statistical uncertainty, the systematic uncertainty, and the uncertainty in the measurement of the$$CP$$ CP asymmetry in the$${{{\textrm{D}}}^{{0}}} \rightarrow {{\textrm{K}} _{\text {S}}^{{0}}} {{{\mathrm{\uppi }}}^{{+}}} {{{\mathrm{\uppi }}}^{{-}}} $$ D 0 K S 0 π + π - decay. This is the first$$CP$$ CP asymmetry measurement by CMS in the charm sector as well as the first to utilize a fully hadronic final state. 
    more » « less