Tony Mezzacappa
Tony Mezzacappa, a theoretical astrophysicist at ORNL says, "My passion stems from trying to understand how we came to be and understanding our cosmos."

from Knoxville News-Sentinel
October 21, 20002
original URL:,1406,KNS_4257_1484540,00.html

Decoding big bangs

Researchers use massive math capability to attack mystery of how stars die

By Frank Munger, News-Sentinel senior writer
October 21, 2002

Imagine billions and billions and billions of atomic bombs exploding all at once.

The thought is almost incomprehensible. Now imagine charting the physics of that force and trying to reconstruct the catastrophic event and its consequences in great detail, stage by stage.

That's sort of the challenge facing Tony Mezzacappa as he and 17 scientific colleagues try to understand how stars die. They are studying core-collapse supernovae, the death acts of massive stars that are 10 times, even 20 times larger than our sun.

Fortunately, the researchers have some of the top computers in the world to assist them in a task that would have been impossible not so long ago.

"I think we'll be occupied with this problem for at least 5 to 10 years," said Mezzacappa, a theoretical astrophysicist on the staff at Oak Ridge National Laboratory. "We may crack the mechanism (that triggers the collapse of the stars' cores) in a 3- to 5-year time frame, but to pin it all down we'll certainly be active on this problem over the next decade."

The Terascale Supernovae Initiative, which the 44-year-old scientist heads, is the largest project in the world looking at core-collapse supernovae.

The stellar explosions are among the most energetic events in the universe, and that alone makes them worthy of study. But it's their production of elements that is supremely important.

"They are the dominant source of all elements in the periodic table heavier than oxygen," Mezzacappa said. "So, without core-collapse supernovae explosions, you and I literally would not be here."

Mezzacappa has been studying supernovae since 1986, when he was a graduate student at the University of Texas.

ORNL hired him in 1996 to establish a theoretical astrophysics program, and the effort has come a long way in a short time, thanks in part to the concurrent growth of the lab's supercomputing capabilities.

To attack problems of extraordinary complexity, scientists need a lot of number-crunching power.

The dynamics of an exploding star, involving things like fluid flow and radiation transport, are described in equations known as partial differential equations.

"We take those equations in their mathematical form, and we convert them to a form that's programmable," Mezzacappa said. "We build the codes to literally solve those equations by stepping forward in time, one step at a time in small steps. And those codes are then run on these massively parallel (computer) architectures."

Most of the supernovae research currently is done on Eagle, an IBM SP supercomputer capable of a trillion calculations per second, one teraflop. But some activities are being shifted to Cheetah, ORNL's newest and most powerful supercomputer with a capability of 42 teraflops.

As they near extinction, massive stars resemble an onion. There's an iron core with outer shells of silicon, oxygen, carbon and helium wrapped in a blanket of hydrogen.

"Those are just the dominant elements," Mezzacappa said. "It's actually a rather complex mix."

Researchers are focusing on the core, where extremes of nature are reached during the collapse.

Mezzacappa said it is believed that radiating neutrinos - virtually massless particles - power the explosion, with a radiation rate that's the equivalent of a 1 billion billion billion billion billion watts.

The energy is so incredible that if one took a sugar-cube-size chunk of matter from the core and placed it on the surface of Earth it would weigh 100 million tons.

"That's how dense the matter is," the ORNL scientist said.

While much has been learned about these explosions from particle studies and decades of mathematical formulations, scientists still don't understand fully how the stellar events occur.

"We have pieces of the puzzle but not the entire puzzle," Mezzacappa said. "Once we understand the mechanism, how to generate the explosion, we've got to be able to predict all the observables: the element production, gravitational wave production, gamma ray production and on out. ... The problem will not be solved until we can predict all of the phenomena associated with supernovae."

That means a lot of computing lies ahead, and Mezzacappa admits he's more of screen gazer than a stargazer.

"I love the mathematics of it. I love the modeling. But, ultimately, my passion stems from trying to understand how we came to be and understanding our cosmos. The big picture. That's really what drives this for sure."

Copyright 2002, KnoxNews. All Rights Reserved.
Mirrored with permission.

This page is part of the CSMD web archive and is not maintained.
Please visit for the latest CSMD information.