Insights and Challenges in Stellar Evolution Lars Bildsten

Insights and Challenges in Stellar Evolution Lars Bildsten Kavli Institute for Theoretical Physics University of California Santa Barbara

Though much is known about how stars evolve, we are only just beginning to probe rotation, interior states and hydrodynamics in a meaningful way. Most of the observational progress is from the asteroseismic data from the Kepler and Co. Ro. T satellites, while theoretical progress is driven by people and new computational tools. Matteo Cantiello (KITP), Joergen Christensen. Dalsgaard (Aarhus Univ. ), Jim Fuller (Caltech/KITP), Phil Macias (UCSB=>UCSC), Chris Mankovich (UCSB=>UCSC), Kevin Moore (UCSB=>UCSC), Bill Paxton (KITP), Dennis Stello (U. Sydney) & Rich Townsend (U. Wisconsin)

After the Main Sequence: Red Giant Branch and Clump Stars Clump stars Paxton et al. ‘ 11 • M< 2 M develop degenerate Helium cores that increase in mass with time until ignition in a flash => lifting degeneracy => stable He burning in core (Thomas 1967). Is this true? ?

Space-Based Photometry Co. Ro. T 27 cm diameter Launched December 2006 Kepler 95 cm diameter Launched March 2009

Non-Radial Stellar Oscillations • P-modes (acoustic waves): In the high wavenumber limit=> Evenly spaced in Frequency (in Envelope!) • G-modes (gravity waves): In the high wavenumber limit=> Evenly spaced in Period (in Core!)

Only Acoustic Waves seen in the Sun Christensen-Dalsgaard

Acoustic Waves (p-modes) in Giants • Persistent convection in the outer parts of the giant excites standing acoustic waves (i. e. modes with n radial nodes for each l) =>Measures mean density • The pulsation amplitudes were estimated (e. g. Christensen-Dalsgaard; Kjeldsen & Bedding ‘ 05) based on earlier solar work (c. f. Goldreich & Keeley ’ 77), but ground based tests were a challenge. . . as amplitudes were low. . .

Co. Ro. T finds p-modes !

Kepler Observations Bedding et al. ‘ 10 (Kepler) Kepler

Bedding et al. ‘ 10 (Kepler) • Large frequency spacing is well measured and collapsing these allows for identification of l=0, 1, 2 and often l=3 acoustic modes • These give mean density measurements straight away! • n~10 -15. . . WKB nearly valid

Highest Observed Frequency Huber et al. 2011 • Highest observed frequency is at the acoustic cutoff of the photosphere. • Higher frequency waves have large damping due to wave escape • Combined with frequency spacing, M and R inferred! • However, internal state not probed. . .

Degenerate Core => Burning Core Bildsten et al. ‘ 12 • Time spent on the Red Giant Branch (RGB) at L>30 L is comparable to that spent on the Red Clump. • Hard to distinguish a clump star from an RGB star in the field, but let’s see what seismology can do. .

Propagation Diagrams and Mixed Modes • Scuflaire ’ 74; Osaki ’ 75 and Aizenman et al. ’ 77 noted that the acoustic waves couple to the non-radial g-modes, which are uniformly spaced in period at: • Coupling is strongest for l=1, and many g-modes exist between each successive acoustic mode

Burning vs. Degenerate Cores RGB Clump

Internal Gravity Waves in the Stellar Core then Detected The g-mode spectrum is very dense in the core, but the modes couple to the envelope well enough to emerge and be detected as oscillations evenly spaced in period. Very stable and long-lived CLOCK! A=p-dominated mode (np, ng=8, 476) B=g-dominated mode (np, ng=7, 505) Star near the RGB bump.

Bedding et al. 2011 Luminosity Mosser et al. 2011 • Distinction of stars on the Red Giant branch from those doing He burning in the core (clump stars)!! • Discovered now in nearly all RGB stars (Stello et al. 2013)

Q: How does the Electron Degenerate Helium Core Transition to the Burning State in a M<2 M Star? A: Through a series of thermonuclear instabilities over a 2 million year period (Thomas 1967) that is often overlooked in the literature and never yet tested. . .

MESA is open source: anyone (over 700 users!) can download the source code, compile it, and run it for their own research or education purposes. Bill Paxton, Father of MESA

Temperature Evolution of First Flash Macias et al. 2013 H Burning Layer Macias et al. 2014

Outer Envelope Undergoes Kelvin Helmholtz Contraction Macias et al. 2014 • The radial expansion leads to adiabatic T decline in the H burning layer, shutting it off. • Leads to rapid KH contraction of the giant to the clump.

Core Flash Sequence from MESA Bildsten et al. ‘ 12

Propagation Diagrams Relic Layer Bildsten et al. ‘ 12 • Contraction leads to outer envelope profiles during the flash nearly identical to the red clump • Coupling to core modes during the flash will be as strong as on clump. • Core is in an intermediate state => g-modes dif’t Bildsten et al. ‘ 12

Repercussions Bildsten et al. ‘ 12 • The period of successive flashes is 2 million years, thermal time through the core. • Implies that one in ~50 stars near the clump should be in an unusual state with a core structure intermediate to the two states. . • ~5000 giants studied by Kepler, so many examples expected. . . Clear IDs starting to emerge. .

Mosser et al. 2014 f=possible flashers

Solar Model “Trivially” Evolved Courtesy C. Mankovich See also Tayar & Pinsonneault ‘ 13 • Rotation rate in the core can impact the later evolution of the star, especially for massive stars • Certainly sets rotation for the He burning core and eventually the white dwarf

RGB Power Spectrum: Rotation! G G Beck et al. ‘ 12 Stello et al. 2013 P G

Inferred Core Rotation Mosser et al. 2012

Core loses 95% of its Angular Momentum after Leaving MS

Calculations with Magnetic Dynamos Cantiello et al. ‘ 14, Still not enough angular momentum loss from the shrinking core!

Conclusions • New astronomical tools are revealing the interiors of stars in ways previously impossible. • Rotation can now be explored, an important player in how stars finally collapse. • Theory remains key in at least three ways: – Reliable physical modeling (MESA+GYRE) needed to fully interpret the data – Novel analytics and mechanisms (Cantiello et al. 2014; Fuller et al. 2014) to transport J – Suggestions of new ways to look at the data to reveal rare physical states (e. g. core flash)