Modelling supernova-driven turbulence

Research output: Contribution to journalArticleScientificpeer-review

Standard

Modelling supernova-driven turbulence. / Gent, F. A.; Mac Low, M. M.; Käpylä, M. J.; Sarson, G. R.; Hollins, J. F.

In: GEOPHYSICAL AND ASTROPHYSICAL FLUID DYNAMICS, 01.01.2019.

Research output: Contribution to journalArticleScientificpeer-review

Harvard

APA

Vancouver

Author

Bibtex - Download

@article{845ede27378a47ba8996596e5aa88b2c,
title = "Modelling supernova-driven turbulence",
abstract = "High Mach number shocks are ubiquitous in interstellar turbulence. The Pencil Code is particularly well suited to the study of magnetohydrodynamics in weakly compressible turbulence and the numerical investigation of dynamos because of its high-order advection and time evolution algorithms. However, the high-order algorithms and lack of Riemann solver to follow shocks make it less well suited to handling high Mach number shocks, such as those produced by supernovae (SNe). Here, we outline methods required to enable the code to efficiently and accurately model SNe, using parameters that allow stable simulation of SN-driven turbulence, in order to construct a physically realistic galactic dynamo model. These include the resolution of shocks with artificial viscosity, thermal conductivity and mass diffusion; the correction of the mass diffusion terms and a novel generalisation of the Courant condition to include all source terms in the momentum and energy equations. We test our methods with the numerical solution of the one-dimensional (1D) Riemann shock tube, also extended to a 1D adiabatic shock with parameters and Mach number relevant to SN shock evolution, including shocks with radiative losses. We extend our test with the three-dimensional (3D) numerical simulation of individual SN remnant evolution for a range of ambient gas densities typical of the interstellar medium and compare these to the analytical solutions of Sedov–Taylor (adiabatic) and the snowplough and Cioffi et al. results incorporating cooling and heating processes. We show that our new timestep algorithm leads to linear rather than quadratic resolution dependence as the strength of the artificial viscosity varies, because of the corresponding change in the strength of interzone gradients.",
keywords = "artificial diffusivity, high Mach number shocks, instabilities, Numerical methods, supernova-driven turbulence",
author = "Gent, {F. A.} and {Mac Low}, {M. M.} and K{\"a}pyl{\"a}, {M. J.} and Sarson, {G. R.} and Hollins, {J. F.}",
year = "2019",
month = "1",
day = "1",
doi = "10.1080/03091929.2019.1634705",
language = "English",
journal = "GEOPHYSICAL AND ASTROPHYSICAL FLUID DYNAMICS",
issn = "0309-1929",
publisher = "Taylor and Francis Ltd.",

}

RIS - Download

TY - JOUR

T1 - Modelling supernova-driven turbulence

AU - Gent, F. A.

AU - Mac Low, M. M.

AU - Käpylä, M. J.

AU - Sarson, G. R.

AU - Hollins, J. F.

PY - 2019/1/1

Y1 - 2019/1/1

N2 - High Mach number shocks are ubiquitous in interstellar turbulence. The Pencil Code is particularly well suited to the study of magnetohydrodynamics in weakly compressible turbulence and the numerical investigation of dynamos because of its high-order advection and time evolution algorithms. However, the high-order algorithms and lack of Riemann solver to follow shocks make it less well suited to handling high Mach number shocks, such as those produced by supernovae (SNe). Here, we outline methods required to enable the code to efficiently and accurately model SNe, using parameters that allow stable simulation of SN-driven turbulence, in order to construct a physically realistic galactic dynamo model. These include the resolution of shocks with artificial viscosity, thermal conductivity and mass diffusion; the correction of the mass diffusion terms and a novel generalisation of the Courant condition to include all source terms in the momentum and energy equations. We test our methods with the numerical solution of the one-dimensional (1D) Riemann shock tube, also extended to a 1D adiabatic shock with parameters and Mach number relevant to SN shock evolution, including shocks with radiative losses. We extend our test with the three-dimensional (3D) numerical simulation of individual SN remnant evolution for a range of ambient gas densities typical of the interstellar medium and compare these to the analytical solutions of Sedov–Taylor (adiabatic) and the snowplough and Cioffi et al. results incorporating cooling and heating processes. We show that our new timestep algorithm leads to linear rather than quadratic resolution dependence as the strength of the artificial viscosity varies, because of the corresponding change in the strength of interzone gradients.

AB - High Mach number shocks are ubiquitous in interstellar turbulence. The Pencil Code is particularly well suited to the study of magnetohydrodynamics in weakly compressible turbulence and the numerical investigation of dynamos because of its high-order advection and time evolution algorithms. However, the high-order algorithms and lack of Riemann solver to follow shocks make it less well suited to handling high Mach number shocks, such as those produced by supernovae (SNe). Here, we outline methods required to enable the code to efficiently and accurately model SNe, using parameters that allow stable simulation of SN-driven turbulence, in order to construct a physically realistic galactic dynamo model. These include the resolution of shocks with artificial viscosity, thermal conductivity and mass diffusion; the correction of the mass diffusion terms and a novel generalisation of the Courant condition to include all source terms in the momentum and energy equations. We test our methods with the numerical solution of the one-dimensional (1D) Riemann shock tube, also extended to a 1D adiabatic shock with parameters and Mach number relevant to SN shock evolution, including shocks with radiative losses. We extend our test with the three-dimensional (3D) numerical simulation of individual SN remnant evolution for a range of ambient gas densities typical of the interstellar medium and compare these to the analytical solutions of Sedov–Taylor (adiabatic) and the snowplough and Cioffi et al. results incorporating cooling and heating processes. We show that our new timestep algorithm leads to linear rather than quadratic resolution dependence as the strength of the artificial viscosity varies, because of the corresponding change in the strength of interzone gradients.

KW - artificial diffusivity

KW - high Mach number shocks

KW - instabilities

KW - Numerical methods

KW - supernova-driven turbulence

UR - http://www.scopus.com/inward/record.url?scp=85068577165&partnerID=8YFLogxK

UR - http://adsabs.harvard.edu/abs/2018arXiv180601570G

U2 - 10.1080/03091929.2019.1634705

DO - 10.1080/03091929.2019.1634705

M3 - Article

JO - GEOPHYSICAL AND ASTROPHYSICAL FLUID DYNAMICS

JF - GEOPHYSICAL AND ASTROPHYSICAL FLUID DYNAMICS

SN - 0309-1929

ER -

ID: 35442078