September 1st, 2006

**Categories: ** Applications, Software, Visualization

Argonne National Laboratory is actively pursuing research and design for a Rare Isotope Accelerator (RIA) facility that will aid basic research in nuclear physics by creating beams of unstable isotopes. Such a facility has been labeled as a high priority by the joint Department of Energy and National Science Foundation Nuclear Science Advisory Committee because it will allow more study on the nature of nucleonic matter, the origin of the elements, the Standard Model, and nuclear medicine.

An important part of this research is computer simulations that model the behavior of the particle beam, specifically in the Fragment Separator. The Fragment Separator selects isotopes based on their trajectory in electromagnetic fields and then uses absorbers to separate particles with a certain mass and charge from the rest of the beam.

This project focused on the development of a multivariate, correlated Gaussian distribution to model the distribution of particles in the beam as well as visualizations and analysis to view how this distribution changed when passing through an absorber. The distribution was developed in the COSY INFINITY programming language. The user inputs a covariance matrix and a vector of means for the six phase space variables, and the program outputs a vector of correlated, Gaussian random variables.

A variety of random test cases were conducted in two, three and six variables. In each case, the expectation values, variances and covariances were calculated and they converged to the input values. The output of the absorber code is a large data set that stores all of the variables for each particle in the distribution. It is impossible to analyze such a large data set by hand, so visualizations and summary statistics had to be developed.

The first visualization is a three-dimensional graph that shows the number of each isotope present after each slice of the absorber. A second graph plots any of the six phase space variables against any of the others to see the change in the beam’s distribution. Also, the expectation values, variances and covariances of the phase space variables were calculated after the absorber. The distribution that models the particle beam gives the variability that physicists need to simulate many different situations in the Fragment Separator. The statistics and visualizations will allow quick analysis of the particle beam. Both of these developments will contribute to the overall viability of the RIA proposal.

Rosenthal, C., Erdelyi, B., Modeling and Visualizing the Particle Beam in the Rare Isotope Accelerator, *Journal of Undergraduate Research*, vol 6, Washington, D.C., U.S. Department of Energy, September 1st, 2006. http://www.scied.science.doe.gov/scied/JUR.html