Canada Spends Logo

Explore our Budget 2026 analysis

Grant: Discovery Grants Program - Individual

The University of British Columbia

$31,000.00

Three problems in computational science

FY 2024-25

Summary
Much of modern design and optimization in engineering and finance is done with computer simulations. Mathematical descriptions (models) are made of the physical, chemical, or financial phenomena. The mathematical models involve equations that are too complex to be solved with paper and pencil, and so approximate solutions are sought using computer algorithms. My expertise is in mathematical modelling and the development of efficient computational approximation tools. My past research has led to improvements in design in the fuel cell industry and scientific insight into the behaviour of activated polymer materials. Three new projects will be pursued in the proposed work. Phase change is of interest in many industrial and scientific processes like metal casting and the formation of sea ice. Mathematical models of phase change involve conservation of heat and mass that can be expressed in equations and approximated numerically on a grid of points. The main numerical difficulty is in keeping track of the phase at each point. In my recent work on models of this type for sea ice formation, I have discovered a computationally efficient iteration strategy to find the interfaces between the phase regions. The proposed work is to understand theoretically this strategy which will have impact on many industrially relevant computational models. I will also continue my development of a model of phase change in sea ice to be of comprehensive use to climate change and remote sensing applications. The second project is the development of a computational method for the movement of curves relaxing to configurations that minimize energy. The energy that drives the motion of the curves varies between applications, and a generic framework is proposed that encompasses many of the phenomena in biological applications and interfaces in material microstructure. The resulting simulation tool will be made available as an open license computational package to be of use to others. This tool will allow scientists to explore the behaviour of membranes under different physical models to allow them to match theory to experimental observations.   The third project is an investigation into the approximation properties and efficient training of deep neural networks of Neural Ordinary Differential Equation (NODE) type. Deep neural networks underlie many of the recent advances in artificial intelligence. These networks are trained on large amounts of data and the training, done with large computer systems, can be very expensive. There is a way to formulate a NODE so that its training process resembles well understood computational methods for physical problems, which is my area of expertise. The approximation and training properties of a particular NODE architecture will be studied. Insight into efficient, adaptive training for commercial sized networks will be gained.
Program
Discovery Grants Program - Individual
Location
British Columbia, Canada
Recipient Type
Academia
Type
Grant
Source
NSERC's Awards Data

As published

Application id
RGPIN-2024-06021
Name nom
Wetton, Brian
Department département
Mathematics
Organization id
2
Institution établissement
The University of British Columbia
Province en
British Columbia
Province fr
Colombie-Britannique
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2024
Competition year année de concours
2024
Award amount
31000
Program id
RGPIN
Program name en
Discovery Grants Program - Individual
Program name fr
Programme de subventions à la découverte - individuelles
Group en
DISCOVERY RESEARCH
Group fr
RECHERCHE AXÉE SUR LA DÉCOUVERTE
Committee code
1508
Committee name en
Mathematics and Statistics
Committee name fr
Mathématiques et statistique
Area of application code
401
Area of application group en
Environment
Area of application group fr
Environnement
Area of application en
Climate and atmosphere
Area of application fr
Climat et atmosphère
Research subject code
2950
Research subject group en
Applied mathematics
Research subject group fr
Mathématiques appliquées
Research subject en
Applied mathematics
Research subject fr
Mathématiques appliquées
Field of research list names en
Not available
Field of research list names fr
Non disponible
Socioeconomic objective list names en
Not available
Socioeconomic objective list names fr
Non disponible
Application title
Three problems in computational science
Keywords
finite difference methods; imex methods; enthalpy methods; machine learning; materials science; sea ice; asymptotic error analysis
Application summary
Much of modern design and optimization in engineering and finance is done with computer simulations. Mathematical descriptions (models) are made of the physical, chemical, or financial phenomena. The mathematical models involve equations that are too complex to be solved with paper and pencil, and so approximate solutions are sought using computer algorithms. My expertise is in mathematical modelling and the development of efficient computational approximation tools. My past research has led to improvements in design in the fuel cell industry and scientific insight into the behaviour of activated polymer materials. Three new projects will be pursued in the proposed work. Phase change is of interest in many industrial and scientific processes like metal casting and the formation of sea ice. Mathematical models of phase change involve conservation of heat and mass that can be expressed in equations and approximated numerically on a grid of points. The main numerical difficulty is in keeping track of the phase at each point. In my recent work on models of this type for sea ice formation, I have discovered a computationally efficient iteration strategy to find the interfaces between the phase regions. The proposed work is to understand theoretically this strategy which will have impact on many industrially relevant computational models. I will also continue my development of a model of phase change in sea ice to be of comprehensive use to climate change and remote sensing applications. The second project is the development of a computational method for the movement of curves relaxing to configurations that minimize energy. The energy that drives the motion of the curves varies between applications, and a generic framework is proposed that encompasses many of the phenomena in biological applications and interfaces in material microstructure. The resulting simulation tool will be made available as an open license computational package to be of use to others. This tool will allow scientists to explore the behaviour of membranes under different physical models to allow them to match theory to experimental observations.   The third project is an investigation into the approximation properties and efficient training of deep neural networks of Neural Ordinary Differential Equation (NODE) type. Deep neural networks underlie many of the recent advances in artificial intelligence. These networks are trained on large amounts of data and the training, done with large computer systems, can be very expensive. There is a way to formulate a NODE so that its training process resembles well understood computational methods for physical problems, which is my area of expertise. The approximation and training properties of a particular NODE architecture will be studied. Insight into efficient, adaptive training for commercial sized networks will be gained.