Radiative Mixing Layers: Insights from Turbulent Combustion. (arXiv:2008.12302v1 [astro-ph.GA])
<a href="http://arxiv.org/find/astro-ph/1/au:+Tan_B/0/1/0/all/0/1">Brent Tan</a>, <a href="http://arxiv.org/find/astro-ph/1/au:+Oh_S/0/1/0/all/0/1">S. Peng Oh</a>, <a href="http://arxiv.org/find/astro-ph/1/au:+Gronke_M/0/1/0/all/0/1">Max Gronke</a>
Radiative mixing layers arise wherever multiphase gas, shear, and radiative
cooling are present. Simulations show that in steady state, thermal advection
from the hot phase balances radiative cooling in the front. However, many
features are puzzling. For instance, hot gas entrainment appears to be
numerically converged despite the scale-free, fractal structure of such fronts
implying that they should be unresolved. Additionally, the hot gas heat flux
has a characteristic velocity $v_{rm in} approx c_{rm s,c} (t_{rm
cool}/t_{rm sc,c})^{-1/4}$ whose strength and scaling are not intuitive. We
revisit these issues in 1D and 3D hydrodynamic simulations. Low resolution
leads to numerical diffusion (from truncation error) and dispersion (from stiff
source terms). We find that over-cooling only happens if numerical diffusion
dominates thermal transport; convergence is still possible even when the Field
length is unresolved. A deeper physical understanding of radiative fronts can
be obtained by exploiting parallels between mixing layers and turbulent
combustion, which has well-developed theory and abundant experimental data. A
key parameter is the Damkohler number ${rm Da} = tau_{rm turb}/t_{rm
cool}$, the ratio of the outer eddy turnover time to the cooling time. Once
${rm Da} > 1$, the front fragments into a multiphase medium. Just as for
scalar mixing, the eddy turnover time sets the mixing rate, independent of
small scale diffusion. For this reason, thermal conduction often has limited
impact. We show that $v_{rm in}$ and the effective emissivity can be
understood in detail by adapting combustion theory scalings. Mean density and
temperature profiles can also be reproduced well by mixing length theory. These
results have implications for the structure and survival of cold gas in many
settings, and resolution requirements for large scale galaxy simulations.
Radiative mixing layers arise wherever multiphase gas, shear, and radiative
cooling are present. Simulations show that in steady state, thermal advection
from the hot phase balances radiative cooling in the front. However, many
features are puzzling. For instance, hot gas entrainment appears to be
numerically converged despite the scale-free, fractal structure of such fronts
implying that they should be unresolved. Additionally, the hot gas heat flux
has a characteristic velocity $v_{rm in} approx c_{rm s,c} (t_{rm
cool}/t_{rm sc,c})^{-1/4}$ whose strength and scaling are not intuitive. We
revisit these issues in 1D and 3D hydrodynamic simulations. Low resolution
leads to numerical diffusion (from truncation error) and dispersion (from stiff
source terms). We find that over-cooling only happens if numerical diffusion
dominates thermal transport; convergence is still possible even when the Field
length is unresolved. A deeper physical understanding of radiative fronts can
be obtained by exploiting parallels between mixing layers and turbulent
combustion, which has well-developed theory and abundant experimental data. A
key parameter is the Damkohler number ${rm Da} = tau_{rm turb}/t_{rm
cool}$, the ratio of the outer eddy turnover time to the cooling time. Once
${rm Da} > 1$, the front fragments into a multiphase medium. Just as for
scalar mixing, the eddy turnover time sets the mixing rate, independent of
small scale diffusion. For this reason, thermal conduction often has limited
impact. We show that $v_{rm in}$ and the effective emissivity can be
understood in detail by adapting combustion theory scalings. Mean density and
temperature profiles can also be reproduced well by mixing length theory. These
results have implications for the structure and survival of cold gas in many
settings, and resolution requirements for large scale galaxy simulations.
http://arxiv.org/icons/sfx.gif
