Moment Based RT Methods Nathan Butcher with Dusan Keres and Phil - - PowerPoint PPT Presentation

▶

Nov 19, 2022 243 likes •427 views

Moment Based RT Methods Nathan Butcher with Dusan Keres and Phil Hopkins 8/24/2017 Radiative Transfer Two closure methods implemented, OTVET and M1. Chemistry models that assume an equilibrium chemical state or allow for nonequilibrium

SLIDE 1

Moment Based RT Methods

Nathan Butcher with Dusan Keres and Phil Hopkins 8/24/2017

SLIDE 2

Radiative Transfer

◮ Two closure methods implemented, OTVET and M1. ◮ Chemistry models that assume an equilibrium chemical state

r allow for nonequilibrium evolution.

◮ Can track multiple energies of ionizing radiation. ◮ Check correctness on suite of tests outlined in Iliev, et al.

2006 (static) and Iliev, et al. 2009 (with hydro).

SLIDE 3

Test One

◮ Monochromatic 13.6 eV radiation, 1048 photons per second. ◮ Gas is fixed at 104 K, no photoheating or any other cooling

physics.

◮ Final radius of ionization, called the Str¨

mgren radius, has an

analytic calculation.

◮ Both methods underpredict final Str¨

mgren radius but results

are consistent with other codes (Iliev, et al. 2006).

SLIDE 4

Test One

SLIDE 5

Test Two

◮ Source spectrum is a 105 K blackbody emitting 1048 photons

per second.

◮ Effective single bin energy of 29.65 eV. ◮ Photoheating and cooling implemented, no hydro. ◮ Final radius is larger than Str¨

mgren sphere due to higher

temperature (around 3 × 104 K compared to 1 × 104 K).

◮ RT in GIZMO is consistent with other codes in ionization

front and temperature profiles.

SLIDE 6

Test Two

SLIDE 7

Test Two

Figure: Temperature. Top: M1 Multifrequency. Bottom: Comparison in Iliev, et al. 2006

SLIDE 8

Test Two

(a) One Bin (b) Four Bins Figure: Multiple energy bins captures preheating by high energy photons.

SLIDE 9

Test Three

◮ Shadowing by a dense clump, no hydro. ◮ Clump: n = 4 × 10−2 cm−3, T = 40 K ◮ Environment: n = 2 × 10−4 cm−3, T = 8000 K ◮ M1 produces shadow that lasts the entire test runtime.

OTVET fails to produce a shadow, as expected.

SLIDE 10

Test Three

(a) M1 (b) OTVET Figure: Projection of Neutral Hydrogen Fraction at 2 Myr

SLIDE 11

Test Three

(a) M1 (b) OTVET Figure: Projection of Neutral Hydrogen Fraction at 15 Myr

SLIDE 12

Test Four

◮ Static cosmological density field at z = 9. ◮ 16 sources in the 16 highest density halos, with source

luminosity proportional to the density.

◮ Recreate provided grid ICs in GIZMO. Only grid-based codes

ran this test in the comparison paper.

◮ This test completely failed until June due to the ionization

front stalling at high densities.

◮ Consistent disagreement between M1 and OTVET in this test.

SLIDE 13

Test Four

Figure: In comparison paper, neutral fraction of the four codes tested ranged between 10% and 30%

SLIDE 14

Test Five

◮ Same as Test Two, but with live hydro. ◮ Reproduce ionization front position and gas outflow well when

compared to comparison paper.

◮ Only tested with M1 so far.

SLIDE 15

Test Five

SLIDE 16

Test Five

Figure: Number Density. Top: M1 Multifrequency. Bottom: Comparison in Iliev, et al. 2009

SLIDE 17

Test Five

Figure: Mach number. Top: M1 Multifrequency. Bottom: Comparison in Iliev, et al. 2009

SLIDE 18

Future Work

◮ Finish the last two tests in the Iliev suite. Ionization front

expansion from a dense core into an isothermal sphere and the photoevaporation of a dense clump.

◮ Reproduce tests on radiation pressure vs. photoionization

pressure around a stellar source. (Sales, et al. 2014)

◮ Simulate an interesting HII region, such as 30 Doradus in the

LMC (Lopez, et al. 2011)

◮ Largest HII region in local group, with 2400 OB stars ◮ Complicated shape with substructure