Irradiation of astrophysical objects – SED and flux effects on thermally driven winds

2019

ApJ

Self Cite

Multiphase media have very complex structure and evolution. Accurate numerical simulations are necessary to make advances in our understanding of this rich physics. Because simulations can capture both the linear and nonlinear evolution of perturbations with a relatively wide range of sizes, it is important to thoroughly understand the stability of condensation and acoustic modes between the two extreme wavelength limits of isobaric and isochoric instability as identified by Field (1965). Partially motivated by a recent suggestion that large non-isobaric clouds can 'shatter' into tiny cloudlets, we revisit the linear theory to survey all possible regimes of thermal instability. We uncover seven regimes in total, one of which allows three unstable condensation modes. Using the code Athena++, we determine the numerical requirements to properly evolve small amplitude perturbations of the entropy mode into the nonlinear regime. Our 1D numerical simulations demonstrate that for a typical AGN cooling function, the nonlinear evolution of a single eigenmode in an isobarically unstable plasma involves increasingly larger amplitude oscillations in cloud size, temperature and density as the wavelength increases. Such oscillations are the hallmark behavior of non-isobaric multiphase gas dynamics and may be observable as correlations between changes in brightness and the associated periodic redshifts and blueshifts in systems that can be spatially resolved. Intriguingly, we discuss regimes and derive characteristic cloud sizes for which the saturation process giving rise to these oscillations can be so energetic that the cloud may indeed break apart. However, we dub this process 'splattering' instead of 'shattering', as it is a different fragmentation mechanism triggered when the cloud suddenly 'lands' on the stable cold branch of the equilibrium curve.

Section: Introductionmentioning

confidence: 82%

Non-isobaric Thermal Instability

Waters

2019

ApJ

Self Cite

“…Determining the ionization state of the gas is required to self-consistently calculate the force due to line driving, as the force multiplier depends on ξ (Dannen et al 2018 in prep). Further we should go beyond the isothermal approximation and compute heating/cooling from the relevant SED (Dyda et al 2017), because temperature too can affect the number of available lines and thermal driving may alter the global outflow properties.…”

Section: Discussionmentioning

confidence: 99%

Effects of radiation field geometry on line driven disc winds

Dyda

Monthly Notices of the Royal Astronomical Society

2018

Self Cite

We study line driven winds for models with different radial intensity profiles: standard Shakura-Sunyaev radiating thin discs, uniform intensity discs and truncated discs where driving radiation is cutoff at some radius. We find that global outflow properties depend primarily on the total system luminosity but truncated discs can launch outflows with ∼ 2 times higher mass flux and ∼ 50% faster outflow velocity than non-truncated discs with the same total radiation flux. Streamlines interior to the truncation radius are largely unaffected and carry the same momentum flux as non-truncated models whereas those far outside the truncation radius effectively carry no outflow because the local radiation force is too weak to lift matter vertically away from the disc. Near the truncation radius the flow becomes more radial, due to the loss of pressure/radiation support from gas/radiation at larger radii. These models suggest that line driven outflows are sensitive to the geometry of the radiation field driving them, motivating the need for self-consistent disc/wind models.

“…The behavior in the unobservable EUV region, for instance, must be assumed. Additionally, the SED also affects the gas thermal stability (e.g., Kallman & Mc-Cray 1982;Krolik 1999;Mehdipour et al 2015;Dyda et al 2017). In particular, AGN1 and AGN2 both have regions of isobaric thermal instability; we return to this point in §3.…”

Section: Underlying Sedsmentioning

confidence: 96%

“…This paper builds upon several previous studies of AGN gas flows where radiation source terms are included in a progressively more self-consistent manner (e.g., Proga et al 2000;Proga 2007; 2015; Dyda et al 2017). To properly include both radiation heating/cooling and radiation forces, one needs to accurately compute the coupling between radiation and matter, i.e.…”

Section: Underlying Sedsmentioning

confidence: 99%

“…Plasma codes such as XSTAR 1 (Kallman & Bautista 2001) and CLOUDY 2 (Ferland et al 2013) can be used to model these features, as they perform detailed photoionization, radiative transfer, and energy balance calculations. Several groups have begun incorporating these calculations into hydrodynamical codes, and this is currently the most accurate approach for comparing theory with observations (e.g., Salz et al 2015;Ramírez-Velasquez et al 2016;Kinch et al 2016;Dyda et al 2017;Higginbottom et al 2017).…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Photoionization Calculations of the Radiation Force Due To Spectral Lines in AGNs

Dannen

Kallman

et al. 2019

ApJ

Self Cite

One of the main mechanisms that could drive mass outflows in AGNs is radiation pressure due to spectral lines. Although straightforward to understand, the actual magnitude of the radiation force is challenging to compute because the force depends on the physical conditions in the gas, and the strength, spectral energy distribution (SED), and geometry of the radiation field. We present results from our photoionization and radiation transfer calculations of the force multiplier, M (ξ, t), using the same radiation field to compute the gas photoionization and thermal balance. We assume low gas density (n = 10 4 cm −3 ) and column density (N ≤ 10 17 cm −2 ), a Boltzmann distribution for the level populations, and the Sobolev approximation. Here, we describe results for two SEDs corresponding to an unobscured and obscured AGN in NGC 5548. Our main results are the following: 1) although M (ξ, t) starts to decrease with ξ for ξ 1 as shown by others, this decrease in our calculations is relatively gradual and could be non-monotonic as M (ξ, t) can increase by a factor of few for ξ ≈ 10 − 1000; 2) at these same ξ for which the multiplier is higher than in previous calculations, the gas is thermally unstable by the isobaric criterion; 3) non-LTE effects reduce M (t, ξ) by over two orders of magnitude for ξ 100. The dynamical consequence of result (1) is that line driving can be important for ξ as high as 1000 when the LTE approximation holds, while result (2) provides a natural cloud formation mechanism that may account for the existence of narrow line regions. Result (3) suggests that line driving may not be important for ξ 100 in tenuous plasma.