"Our work is about comparing the data we collect in the STAR detector with modern calculations, so that we can write down equations on paper that exactly describe how the quark-gluon plasma behaves," says Jerome Lauret from Brookhaven National Laboratory. "One of the most important assumptions we've made is that, for very intense collisions, the quark-gluon plasma behaves according to hydrodynamic calculations in which the matter is like a liquid that flows with no viscosity whatsoever."
Elliptic flow as a function of transverse momentum for identified particles as measured at RHIC. At low transverse momentum the pattern of elliptic flow is well reproduced by hydrodynamic calculations (lines).
Image Courtesy Art Poskanzer and Paul Sorensen.
Proving that under certain conditions the quark-gluon plasma behaves according to such calculations is an exciting discovery for physicists, as it brings them a little closer to understanding how matter behaves at very small scales. But the challenge remains to determine the properties of the plasma under other conditions.
"We want to measure when the quark-gluon plasma behaves like a perfect fluid with zero viscosity, and when it doesn't," says Lauret. "When it doesn't match our calculations, what parameters do we have to change? If we can put everything together, we might have a model that reproduces everything we see in our detector."
Measuring the properties of the quark-gluon plasma, including the elliptic flow and viscosity, requires the analysis of very large data samples. STAR uses the OSG's software stack to federate its computing facilities into a uniform computational pool. STAR researchers have also adapted the Storage Resource Manager/DataMover tool, developed in collaboration with the Lawrence Berkeley National Laboratory, to run on the OSG infrastructure. This tool allowed STAR to transfer a very large dataset—20 terabytes of data comprising 110 million collisions—in near-real time from Brookhaven National Laboratory on Long Island to the National Energy Research Scientific Computing Center in California.
"This work is going on all over the world," adds Lauret. "We have sites as close as Massachusetts and as far away as Sao Paulo and China. The OSG and SRM/DataMover have been a terrific enabler of STAR data analysis."
Efficient data movement has allowed STAR to spread its data analysis over a computing pool that includes resources from many of these sites. This has allowed researchers to quickly analyze properties such as the patterns of elliptic flow in the particles that reach the STAR detector. Results on elliptic flow and other phenemona, aided by the use of the OSG software stack and SRM/DataMover, will be presented at the Quark Matter 2006 conference in Shanghai this November.
This article, an Open Science Grid Research Highlight, was reprinted courtesy of OSG.
e-mail this article