Date of Award

8-2018

Degree Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Electrical and Computer Engineering

Committee Chair

Gerhard Klimeck

Committee Co-Chair

Tillmann Kubis

Committee Member 1

David Ebert

Committee Member 2

Krishna Madhavan

Abstract

Nowadays, complex simulations of a variety of processes are extensively used in academia and industry. Particularly in academia, powerful scientific software tools are constantly developed to simulate complex systems; for instance, simulations of quantum transport using the non-equilibrium greens Function formalism. The potential impact of these scientific tools in industry is huge, but it is hindered by the lack of usability of the software by those who are not deeply familiar with it. Visual analytics is a new field that has shown the positive impact of interactive visualizations in software usability and the cognitive process of the user. This research investigates whether the implementation of interactive visual aids also improves the usability and the cognitive processes of research codes users, particularly those used for simulation design. To accomplish this goal, this study defines a framework for simulation design in scientific research, identifies the stages in which visual aids can be implemented to increase usability, and implements an interactive visualization system (NemoViz). NEMO5, a tool for designing atomistic simulation, is used as a case study to measure the effectiveness, efficiency, and user satisfaction of the use of visual aids in scientific simulation design. The results from this research provide a framework of reference for development of user-friendly simulation design tools, and will shed light on strategies that scientific developers might implement to broaden the impact of their simulation codes.

Share

COinS