Supernova Science Center (SNSC) Challenges and Collaborations

Stan Woosley, UCSC
Adam Burrows, University of Arizona
Chris Fryer, LANL
Rob Hoffman, LLNL (plus 21 others)

Despite six decades of study, we still do not understand how supernovae explode. Further advances will require calculations on the largest available computers coupled to developments in multi-dimensional radiation hydrodynamics, turbulent combustion, nuclear physics, and computer science. This will require an unprecedented multi-disciplinary collaboration.

Few events in nature match the grandeur of supernovae and none surpass their raw power. Viewed on a cosmic scale, supernovae light up galaxies in spectacular fireworks that stir the interstellar and intergalactic media. They make most of the elements that form our own planet and bodies, and they give birth to the most exotic states of matter known – neutron stars and black holes. Yet, despite their importance, no one understands, in detail, how they work.

There are two kinds of supernovae, thermonuclear (or Type Ia) and gravity-powered (Type II and Ib), each with its own special challenge. The challenge posed by thermonuclear supernovae is the realistic simulation of turbulent (nuclear) combustion for low Prandtl number and extremely high Rayleigh number. A fusion flame is born somewhere near the center of a white dwarf. Just how, where, and how often it ignites is an important issue (consider an automobile engine with two or more spark plugs in random locations and try to calculate the fuel efficiency!). Once the flame is born, its hot ashes lie beneath cool, denser fuel and are Rayleigh-Taylor unstable. Non-linear growth of that instability leads to shear and increased turbulence. We need to learn just how the flame forms and how fast it moves. This determines the energy of the supernova and how bright it is. These are the same supernovae used by cosmologists to show evidence for the accelerated expansion of the universe, so causes for their diversity have cosmic significance.

In a gravitational supernova, the iron core of a massive star collapses to a neutron star radiating a flood of neutrinos (roughly 20% of the rest mass of the neutron star converted to pure energy by E = mc2). The inefficient coupling of these neutrinos to the overlying stellar material launches a shock wave that explodes the star. Convective motions, powered by neutrino energy deposition, are thought to be central to the success of the model. The computational challenge here is the multi-dimensional simulation of fluid flow coupled to the transport of radiation (neutrinos) that have a non-thermal spectrum and are making a transition from optically thick to thin.

Each of these studies taxes currently existing computers. Were it not for the computer access provided by SciDAC, our group could make no progress. Even given state-of-the-art computers, these problems require resources and manpower beyond what a small group can muster. Some of our most important activities during the first year have thus involved the forging of alliances. These collaborations have been greatly facilitated by our involvement with SciDAC.

1. The Algorithmic Framework ISIC at LBNL – this has been our principal interaction with another SciDAC Center. Using a low Mach number code developed at LBNL to study chemical combustion, the SNSC is studying the propagation of fusion flames in supernovae.
2. The Terascale Supernova Initiative (TSI) – is the other SciDAC Center for the study of (core-collapse) supernovae. We are working together on the high-density equation of state and on organizing workshops.
3. SNSC team members in the U. Arizona in the ACIMS are creating a Parallel Programming and Debugging Facility (ADViCE, PAMS) and a run-time and execution infrastructure (CORBA, DEVS-DOC) that will enable us to better understand and debug our complicated codes. A collaborative proposal involving several of the software ISICs has been submitted to the NSF-ITR Program and has passed the pre-proposal stage.
4. LANL – a supernova research group has been set up at LANL with SNSC co-I Chris Fryer as its leader. The SNSC has initiated a collaboration with the LANL RAGE code group. LANL is also providing computer hardware and access.
5. LLNL – SNSC team members are working within N-Division to improve the nuclear database used for nucleosynthesis studies.
6. The Chicago FLASH Center – SNSC/UCSC postdoc Zingale is a member of the FLASH team and is taking the lead in adapting the FLASH code for supernova studies (Fig. 2).
7. The Joint Institute for Nuclear Astrophysics (JINA) – this NSF Frontier's Center has now been funded and is collaborating with us.
8. The Max Planck Institut fuer Astrophysik, Germany – this is the major European center for the study of supernovae. Two members have joined our Advisory Committee and we are spending time at each others institutions.
9. Sandia, Livermore – SNSC scientists are working with members of the Combustion Facility on modeling turbulent nuclear flames in novae and supernovae.

Figure 1. While modest in appearance, this is a calculation in 2D of the Landau-Darrieus instability in a white dwarf star using the LBNL low Mach number code. Complexity, e.g., turbulence can be easily added.

Figure 2. Mixing behind the shock wave in a Type II supernova due to the Rayleigh-Taylor instability. Calculated at UCSC using the Chicago FLASH code in two-dimensions with adaptive mesh.

Supernova research might have gone on at some level at each of these places anyway, but SciDAC funding and computer time have energized a focused collaborative effort that would otherwise have been lacking. It has also attracted funding at the collaborating institutions that has greatly leveraged DOE’s own modest investment.

For further information on this subject contact:
Stan Woosley – woosley@ucolick.org
Department of Astrophysics, UCSC
Phone: 831-459-2976
http://www.supersci.org/

back to project page