Li, Yijia2007-08-032007-08-032007-08-032007http://hdl.handle.net/10012/3150This thesis provides an introduction to and analysis of the problem of determining nuclear magnetic resonance (NMR) relaxation times of porous media by using the so-called Carr-Purcell-Meiboom-Gill (CPMG) technique. We introduce the principles of NMR, the CPMG technique and the signals produced, porous effects on the NMR relaxation times and discuss various numerical methods for the inverse problem of extracting the relaxation times from CPMG signals. The numerical methods for solving Fredholm integral equations of the first kind are sketched from a series expansion perspective. A method of using arbitrary constituent functions for improving the performance of non-negative least squares (NNLS) is developed and applied to several synthesized data sets and real experimental data sets of saturated porous glass gels. The data sets were obtained by the author of this thesis and the experimental procedure will be presented. We discuss the imperfections in the assumptions on the physical and numerical models, the numerical schemes, and the experimental results, which may lead to new research possibilities.4491412 bytesapplication/pdfenNMRporous mediaseparation of exponentialsinverse problemMaster ThesisApplied Mathematics