The New Era of Stellar Physics

University of Washington Colloquium

Transcript

1. Image: J. Insley (ALCF) and Y. Jiang (KITP) The New Era of Stellar Physics Matteo Cantiello 1,2 1 CCA, Flatiron Institute 2 Princeton University
2. This Talk • How do we model stars • Why is this a “new era” for stellar physics? • A few “frontier” problems: rotation, magnetism, massive stars
3. What is a Star?
4. The timescale on which a star reacts to a perturbation of hydrostatic equilibrium Stellar TimeScales: Dynamical
5. How fast a star reacts to a perturbation of thermal equilibrium. It’s also the timescale of contraction of a star where the nuclear energy generation suddenly disappeared Stellar TimeScales: Thermal Timescale L = ˙ Eint
6. A star can remain in thermal equilibrium for as long as its nuclear fuel supply lasts. The energy source of nuclear fusion is the direct conversion of a small fraction φ of the rest mass of the reacting nuclei into energy Stellar TimeScales: Nuclear Timescale Hierarchy:
7. The Central Role of Stars • Universe’s building blocks • Drive ionization, chemical and dynamical evolution of gas • Responsible for formation of compact remnants (NS and stellar mass BHs) • Stars and their explosions are cosmic yardsticks • They host planets
8. Know thy star, know thy planet
9. Reionization / Stellar Feedback. Chemical Evolution 30 Doradus X-ray: NASA/CXC/PSU/L.Townsley et al.; Optical: NASA/STScI; Infrared: NASA/JPL/PSU/L.Townsley et al.
10. Gravitational Waves Sources
11. The recent past: “Stars” was considered a solved problem in theoretical astrophysics…
12. ! Transient surveys unraveling unpredicted variety of explosive stellar deaths (e.g. PTF/ZTF, ASAS-SN, Pan-STARRS and soon LSST). We do not understand SN progenitors ! We are entering the era of high precision stellar physics (Kepler, BRITE, K2, GAIA, TESS, PLATO). Theory is lagging behind ! Dawn of GW-Astronomy! (LIGO/ VIRGO) ! Probing the epoch of reionization / first stars? (EDGES / JWST) Exciting times for Stellar Physics
13. We still rely on 1D stellar evolution Assumes homogeneous properties across horizontal surfaces. Retains 1D approximation and solve for transport processes using simplified (di usion) models. Latter come from theory and/or multi-D simulations, and are calibrated via observations. Constant, , composition P, ρ, ω + Microphysics (EOS, opacity, nuclear reaction) 1D stellar evolution entered a ‘mature’ phase with community driven, open-source software instruments (e.g. MESA)
14. Multi-D: The Computational Challenge Ratio between largest and smallest scales can be related to the Reynolds number of the flow (assuming Kolmogorov) max min ∼ Re3/4 ∼ 3 × 107
15. Number of cells to model a cubic region from the largest eddies down to the viscous damping scale (direct numerical simulation). See e.g. Meakin 2008 N = max min 3 ∼ 1022 Multi-D: The Computational Challenge
16. Current largest hydrodynamic simulations on petaFLOPS class machines. Still ~11 orders of magnitude away Fugaku (Riken Center) is currently the fastest supercomputer in the world (~7.6M Cores, 442 petaFLOPS) N = (8192)3 ∼ 5 × 1011 1022 Multi-D: The Computational Challenge
17. Moore’s Law Computational power doubles every ~18 months
18. First Direct Numerical Simulation of the Sun? ~55 years from now (Assuming Moore’s law) 1 . 5 log2 1011
19. It’s worse than that… Not only the spatial dynamical range is huge, but the hierarchy of relevant timescale also poses an immense challenge (~1015 time steps to simulate full evolution!) 10
20. It’s worse than that… Not only the spatial dynamical range is huge, but the hierarchy of relevant timescale poses an immense challenge too (~1015 time steps to simulate full evolution!) On ~Dynamical Timescale On ~Thermal Timescale ~ year 2120 Full Evolution ~ year 2145 ~ year 2075
21. It’s worse than that… … but unexpected advancements in hardware and algorithms could substantially reduce the timeline. Think out of the box! On ~Dynamical Timescale On ~Thermal Timescale ~ year 2120 Full Evolution ~ year 2145 ~ year 2075
22. Models are still useful! Image: J. Insley (ALCF) and Y. Jiang It is likely that many of the resulting flow features captured by incompletely resolved numerical hydro calculations are still robust/useful to understand real astrophysical situations. Particular attention to MHD calculations! Jiang, MC et al. 2015, 2017, 2018 3D Radiation Hydro Simulation of an LBV star
23. Unprecedented synergy between high precision, big data (in particular photometry and astrometry) and theory (in particular asteroseismology and 3D numerical simulations). Together with the advent of open source, community-driven stellar evolution codes like MESA (developed at KITP/UCSB), this is empowering a better scrutiny of established 1D results. With many exciting surprises… A new era of stellar physics 1D Calculations Multi-D Models High-precision, big data
24. The road towards 4D Stars
25. Open Questions: Internal Rotation & Magnetism
26. Probing Stellar Interiors “It would seem that the deep interior of the Sun and stars is less accessible to scientific investigation than any other region of the universe” Sir Arthur Eddington, 1926 Seems to prevent the possibility of measuring important internal properties of stars, like rotation and magnetism (essential to e.g. understand some endpoint of stellar evolution, SLSNe, GRBs etc)
27. Enter Asteroseismology
28. Solar-Like Oscillations Huber et al. 2016 Pre CoRot/Kepler Dupret et al. 2009 Credit: BBSO/NJIT
29. Solar-Like Oscillations Huber et al. 2016 Pre CoRot/Kepler Kepler Telescope Dupret et al. 2009
30. g-mode cavity p-mode cavity νmax N2 evanescent zone Mixed Modes p-mode cavity (envelope) g-mode cavity (core) evanescent zone
31. p-mode cavity (envelope) g-mode cavity (core) Since mixed modes live both as p-mode (in the envelope) and as g-mode (in the core), if observed at the surface their properties (e.g. rotational splitting) can give information about e.g. rotation rate in di erent regions of the star! (Beck et al. 2012, Mosser et al. 2012) Asteroseismology: Mixed Modes Kepler
32. Evolution of Core Rotation Mosser et al. 2012 Early RGB H-burning Clump Stars He-burning P~R2 P~R0.7
33. ! Hydrodynamics instabilities ! Rotationally induced circulations ! Magnetic torques ! Internal gravity waves Angular Momentum Transport Di erent classes of mechanisms have been proposed: e.g. Rogers et al. 2013 e.g Maeder & Meynet 2002 e.g. Spruit 2002 e.g. Heger et al. 2000 Possible to treat the problem in 1D under the assumption of shellular rotation law Fuller et al. 2014, 2015
34. SLOW FAST 1.5MSun Vini =50 km/s Cantiello et al. (2014)
35. SLOW FAST 1.5MSun Vini =50 km/s Cantiello et al. (2014)
36. Asteroseismology now allows to probe the deep interiors of stars Important result: ✴ Internal J-transport not fully understood Cantiello et al. (2014) Large coupling core-envelope seems required. Most compact objects should be slowly-rotating Maeder & Meynet Back to the blackboard
37. Fuller, Piro & Jermyn (2019) Revised “Tayler” Magnetic Torques
38. Fuller & Ma (2019) Revised “Tayler” Magnetic Torques
39. Observed E ective Spins from LIGO/Virgo Stellar evolution in AGNs: Cantiello et al 2020 Stellar Mergers. Renzo, Cantiello et al. 2020 Abbott+ 2020 (O3a)
40. A mystery: Depressed Dipole Modes
41. Depressed Dipole Modes Power Density Frequency ( ) μHz Stello, MC, JF et al. (Nature 2016) Normal Depressed
42. p-mode cavity (envelope) g-mode cavity (core) Mixed Modes interacting with B-Fields?
43. In the presence of strong B- fields, magnetic tension forces can become comparable to buoyancy Critical Field Strength Lorentz Force ~ Buoyancy Force Fuller + Cantiello et al. (Science 2015) Lecoanet, Fuller, MC et al. (2016) See also Loi & Papaloizou (2017,2018)
44. Numerical Simulations ⇢0@tu + rp0 tot = g⇢0 + 1 4⇡ B0 · rB0 + B0 · rB0 r · u = 0 @t⇢0 = ⇢0N2 0 g ez · u @tB0 = B0 · ru u · rB0 damp damp driving x z Solving the linearized magneto-Boussinesq equations using DEDALUS Lecoanet, Fuller, MC et al. 2017
45. damp damp driving Constant B, variable N Lecoanet, Fuller, MC et al. 2017
46. Simulation vs Analytical Theory Next: 3D and more realistic B-field configurations IGW perfectly transmitted into Alfvénic Waves. High wavenumber: They dissipate Vertical Velocity Lecoanet, Fuller, MC et al. 2017
47. Strong Magnetic Fields A ect IGWs Fuller + Cantiello et al. (Science 2015) Lecoanet et al. 2017, Cantiello + Fuller et al. 2016 Dipolar waves scattered to higher l or “transmitted” into Alfvén waves Magnetic fields break spherical symmetry in the core Waves trapped or dissipate quickly Reese et al. 2004, Rincon & Rieutord 2003, Lee 2007,2010, Mathis & De Brye 2010,2012 Typical Critical B- field ~ 105 G See also Loi & Papaloizou 2017
48. Stars evolve this direction Stello, MC, JF et al. (Nature 2016) A mystery: Depressed Dipole Modes
49. Stello, Cantiello, Fuller et al. (Nature 2016) Fraction of stars with strong internal B-fields From a sample of 3000+ stars But See also Mosser et al. 2016 At least 50-60% have strong internal B-fields!
50. Augustson et al. 2016 ! Convective core dynamos on the MS: Beq ~105 G ! Magnetic field topology is complex ! Flux conservation can easily lead to B~106-107 G on the RG ! Stable magnetic configurations of interlocked poloidal+toroidal fields exist in radiative regions Brun et al. 2005 2Msun Prendergast 1956, Mestel 1984, Braithwaite & Nordlund 2006, Duez et al. 2010, Bonanno & Urpin 2008,2013
51. B-Fields 101 Energy Equipartition = MHD Sims: Courtesy of K.Augustson See Cantiello et al. 2016 for more… Magnetic Flux Freezing & Conservation Magnetar-level fields possible/common!
52. Asteroseismology now allows to probe the deep interiors of stars Important results: ✴ Internal J-transport not fully understood Cantiello et al. (2014) Large coupling core-envelope seems required. Most compact objects should be slowly-rotating ✴ Strong core B-fields potentially ubiquitous in stars above ~1.5MSun Fuller, MC et al. (2015), Stello, MC et al. (2016) *Not* included in stellar evolution Conclusions (I) Maeder & Meynet Augustson
53. Open Questions: Massive Stars Evolution
54. Choi et al. 2016 To Understand CCSNe, GRBs, First Stars and LIGO/VIRGO GWs sources, we need to understand structure, mass loss and binary interactions in massive stars
55. !Stability and energy transport !Mass loss !Rotation !Magnetic Fields !Binary interactions Massive Stars: The most uncertain physics
56. Massive Stars: Rotation & Magnetic Fields • The final rotation rate and magnetization of stellar cores are important for the physics of central engines (SLSNe, LGRBs…) • Current models for angular momentum transport relies on 1D di usion approximation of some (local) physical mechanisms. • Large scale magnetic fields are usually not included Millisecond Magnetar Usov 1992 Collapsar Model Woosley 1993 See e.g. Paxton+ 2013
57. !Stability and energy transport !Mass loss !Rotation !Magnetic Fields !Binary interactions Massive Stars: The most uncertain physics
58. Understanding Massive Stars Most massive stars are in multiple systems, and it is clear that binary interactions are extremely important. However, these stars evolve for a large fraction of their lives as ~single stars. The problem is that we do not understand many basic physical processes in single, massive stars. Need to focus on understanding the basic physics first.
59. !Stability and energy transport !Mass loss !Rotation !Magnetic Fields !Binarity Massive Stars: The most uncertain physics (Strong internal B-fields ubiquitous?) (Strong internal coupling not fully understood) (Most massive stars are in binary systems!)
60. ! Line Driven Winds: a ected by wind clumping. Likely overestimated by a factor ~3. Z-dependent ! Cool (RSG) winds: Poorly understood. Poorly constrained observationally. Z-dependent? ! Continuum-driven winds, eruptions... Not understood LDW Eruptions / LBVs RSG Mass Loss
61. See e.g. Smith et al. 2004 Unstable Massive Stars: Luminous Blue Variables (LBVs) LBV in quiescence LBV in outburst
62. Unstable Massive Stars: Luminous Blue Variables (LBVs) 1D Stellar Evolution Tracks Polygons: Location of 3D models Jiang, MC et al. (2018 Nature)
63. Massive Star Envelopes ! Massive stars can develop radiation dominated, loosely bound envelopes e.g Joss et al. 1973, Paxton et al. 2013 ! In 1D models such super Eddington envelopes are characterized by: ! Superadiabatic Convection ! Density Inversions (e.g. Grafener et al. 2012) ! Gas Pressure Inversions ! Envelope Inflation (e.g. Sanyal et al. 2015) ! What about 3D?
64. Di erent regimes in Radiation Dominated Convection Di Rad Flux Advection Flux (“convection”…) Critical optical depth Optical depth where radiation di usion timescale = dynamical timescale Mixing Length Theory not supposed to work!
65. The Opacity At fixed density around Iron Opacity peak. Neighboring lines: x10 in rho Fe Jiang, MC et al. 2015 H He Paxton, MC et al. 2015 Cantiello et al. 2009 Iglesias & Rogers 1996
66. The Opacity: Iron Peak 7.0 5.0 5.5 6.0 6.5 7.0 log T 0.0 0.5 1.0 1.5 2.0 k (cm2 g°1) 60 MØ ZAMS profiles Z=0.02 Z=0.01 Z=0.004 Z=0.001 Z=0.0001 Fe Paxton, MC et al. 2015 Cantiello et al. 2009 Iglesias & Rogers 1996 Strong Metallicity Dependence (Pop III)
67. 3D Radiation Hydro Calculations of LBVs Global Calculations and Mass Loss
68. 3D Athena++, Radiation HD (VET) Jiang, MC et al. (2018 Nature) Our simulations can naturally reproduce the HRD location and mass loss properties of (some) LBVs during outburst (not including line driving)
69. The role of He opacity peak He Fe High density clumps rise, expand and cool, reaching low temperatures and relatively high densities. High He opacity
70. Time Evolution Steady State Initial phase
71. LBV in outburst x 1.5 - 2 Jiang, MC et al. (2018 Nature) Initial phase Steady State (not including line driving)
72. LBV in quiescence ~20% Jiang, MC et al. (2018 Nature) Initial phase Steady State (not including line driving) In agreement with observations, see e.g. Elliott et al. 2022
73. Conclusions (II) 1. Massive stars evolution still very uncertain. Even for single stars 2. Large source of uncertainties comes from our lack of understanding of envelope energy transport and mass loss 3. First 3D global radiation hydro calculations used to study the stability and mass loss of very luminous stars. One step closer to understanding mysterious LBVs
74. Thanks! Image: J. Insley (ALCF) and Y. Jiang (KITP)