Firstly, we give a general introduction to gravitational waves, the instruments used to detect them, potentially interesting sources, and the basics of gravitational wave data analysis with respect to the compact binary search. Following this, we look at a new class of approximants for inspiral waveforms. In these complete approximants, instead of truncating the binding energy and flux functions at the same post-Newtonian order, we instead keep terms such that the approximant corresponds in spirit to the dynamics of the system, with no missing terms in the acceleration. We compare the overlaps with an exact signal (in the adiabatic approximation) for a test mass orbiting a Schwarzchild black hole, for standard and complete approximants in the adiabatic approximation, and beyond the adiabatic approximation using Lagrangian models. A limited extension to the comparable mass case is also given. We then investigate two approaches to performing inspiral searches in a time-critical manner. Both involve splitting the search parameter space across several compute nodes. The first attempts to split the parameter space in an efficient manner by using information from previous runs. The second is balanced dynamically, with slave nodes requesting work off a master node. We then develop a new method for coincidence analysis. In this method, each trigger has associated with it an ellipsoidal region of the parameter space defined by the covariance matrix. Triggers from different detectors are deemed coincident if their ellipsoids are found to overlap. Compared to an algorithm which uses uncorrelated windows separately for each parameter, the method significantly reduces the background rate for comparable detection efficiency. We then give a summary of the current status of the ongoing search for high mass compact binary coalescences in the first calendar year of LIGOs fifth science run.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:584294 |
| Date | January 2008 |
| Creators | Robinson, Craig A. K. |
| Publisher | Cardiff University |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://orca.cf.ac.uk/54665/ |
Page generated in 0.0018 seconds