Selected Publications: Accelerator Lattice and Model Services, P. Chu et al., 2013 International Conference on Accelerator and Large Experimental Control Systems (2013)
Database for Accelerator Modeling, P. Chu et al., 2013 International Particle Accelerator Conference (2013)
Online Physics Model Platform, P. Chu et al., 2012 International Particle Accelerator Conference (2012).
Generic Model Host System Design, P. Chu et al., 2010 International Particle Accelerator Conference (2010).
Service Oriented Architecture for High Level Applications, P. Chu et al., 2010 International Particle Accelerator Conference (2010).
XAL-based Applications and Online Model for LCLS, P. Chu et al., 2009 Particle Accelerator Conference.
High-level Application Framework for LCLS, P. Chu et al., 2007 International Conference on Accelerator and Large Experimental Control Systems (2007).
Drifting Beam Application for SNS Superconducting Cavity Setting, P. Chu et al., 2006 Linear Accelerator Conference (2006).
Effects of overlapping parametric resonances on the particle diffusion process. C. M. Chu et al., Phy. Rev. E 60, 1 (1999).
Unexpectedly Wide rf-Induced-Synchrotron Sideband Depolarizing Resonances. C. M. Chu et al., Phys. Rev. E 58, 4973 (1998).
My recent research has been covering accelerator physics software. Before joining NSCL, I was involved in the commissioning of two recent large accelerator projects in the U.S. - the Linac Coherent Light Source (LCLS) at SLAC National Accelerator Laboratory and the Spallation Neutron Source at Oak Ridge National Laboratory. In order to deliver the right beam for cutting-edge experiments, one of the many technical challenges is to control the particle beams at all times in order to satisfy many separate physics parameters along the beam lines. The complexity of any modern accelerator makes manual beam tuning virtually impossible. Computer software, then, is heavily used for beam tuning and operation automation. This software effort includes data handling, interaction with the control systems and physics modeling for the beam. Quick online physics modeling calculation provides a powerful tool for such beam tuning. Additionally, in order to achieve high availability for an accelerator as a user facility, high performance and reliability for software is crucial.
For FRIB, the beam tuning is even more difficult than any of the existing similar accelerators, because multi-charge-state beams will travel concurrently in narrow beam pipes to be delivered to a tiny spot on a target. An online model is a quick but not detailed method to simulate a particle beam. A modified online model is under development to take into account FRIB-specific hardware. The figure shown here is an example of online model calculation for beta function along the LCLS beam line.
If a simple online model is not sufficient for the sophisticated FRIB beam online control, a more detailed beam particle tracking can be run on High Performance Computing (HPC) clusters. A recent study showed that low-cost Graphical Processing Unit (GPU) based parallel computing is feasible to push large amounts of particle tracking computation for practical online purposes.
For beam tuning, online model serves as a computer library. There are many applications based on the online model, such as beam trajectory correction, emittance matching and fast feedback. In addition to these physics-related applications, there are many other general purpose tools ,such as data archiving and electronic logbook, to facilitate the operation.
In order to operate the FRIB at its design goal, a complete software solution is essential. There will be numerous applications developed. A well-designed architecture to host these applications is a must. An integrated software infrastructure with solid physics support is our research interest. To improve software performance and reliability, an industrial standard approach would be the most effective way. Service-oriented architecture (SOA) is introduced. It is to separate heavy computation and data processing from the rest of the software so the computation part can take advantage of fast computing provided on servers. The SOA approach will be the foundation for future accelerator control with cloud computing. Overall the research is a combination of beam physics and computer engineering.