Date of Award

Summer 2019

Degree Type

Masters Thesis

Degree Name

Master of Science (MS)

School

Mathematics and Natural Sciences

Committee Chair

James Lambers

Committee Chair School

Mathematics and Natural Sciences

Committee Member 2

Haiyan Tian

Committee Member 2 School

Mathematics and Natural Sciences

Committee Member 3

Huiqing Zhu

Committee Member 3 School

Mathematics and Natural Sciences

Abstract

For this thesis, Krylov Subspace Spectral (KSS) methods, developed by Dr. James Lambers, will be used to solve a one-dimensional, heat equation with non-homogenous boundary conditions. While current methods such as Finite Difference are able to carry out these computations efficiently, their accuracy and scalability can be improved. We will solve the heat equation in one-dimension with two cases to observe the behaviors of the errors using KSS methods. The first case will implement KSS methods with trigonometric initial conditions, then another case where the initial conditions are polynomial functions. We will also look at both the time-independent and time-dependent cases for both sets of initial conditions for discrepancies in accuracy and efficiency. Our numerical results will be compared to the results given by Finite Difference methods to show that accuracy can be improved without sacrificing efficiency.

Share

COinS