Caroline Smith's Research Interests

Jaguar Engine Reliability
Jaguar is known throughout the world as a manufacturer of quality 'classic' cars. They recently developed a new engine - the AJ26. As part of the development of this new engine, they applied statistical reliability modeling ideas to test data collected from the new engine as it goes through various phases of development. The aims are to identify key features affecting engine reliability and to make predictions of future field reliability. Such information, for example, will allow manufacturers to predict how many repairs they should expect to make for each car during the warranty period - an invaluable aid to financial planning. The statistical techniques developed during this project are applicable in a wide variety of situations.

Aeroacoustics of External-Coanda Waste Gas Flares
The title of my doctoral thesis was An Investigation into the Aeroacoustics of External-Coanda Waste-Gas Flares. In this work, a number of new methods were proposed for modeling the noise emitted by plane and curved wall jets, and these techniques applied to external-coanda waste-gas flares of the type commonly used in the petrochemical industry. The research was sponsored by British Petroleum, plc. . In addition, a program of practical work was carried out, leading to the development of novel new flare nozzles which considerably reduced the turbulent mixing noise associated with such flares. Some of this research has already been reported by the American Institute of Astronautics and Aeronautics, and the Journal of Sound and Vibration.

Crossflow Microfiltration
During my postdoctoral work, I developed mathematical models of various crossflow microfiltration processes. Unfortunately, due to the commercially-sensitive nature of this work, the results are not able to be published in publicly available literature, although related work of a more general nature was presented at a number of international conferences and meetings.

Logistics
Many of the standard mathematical and statistical techniques associated with reliability, maintainability and supportability prediction are unable to accurately integrate common practical considerations such as life-cycle costs and time into the models. As a consequence, when attempting to model logistics characteristics which are subject to a multitude of conflicting constraints, engineering practitioners have generally used rather ad hoc techniques. I am currently investigating techniques for integrating these important practical considerations within established statistical models. This work is being carried out in collaboration with a multidisciplinary team of researchers from other academic institutions and industry. In particular, I am working closely with colleagues from the MIRCE Akademy, UK. A number of papers have already been published in this area. In addition, the work has been presented at a number of international conferences.

Conceptual Design Discrimination
Another area of investigation is the application of fuzzy logic to the design process. The aim is to quantify qualitative characteristics, in order to provide an objective means of evaluating and comparing alternative conceptual designs. This work is being conducted jointly with a colleague from Lockheed Martin, Manassas, USA., as an extension of his Ph.D. work.

A more detailed CV exists.