FlowVisual: A Visualization App for Teaching and Understanding 3D Flow Field Concepts

The study of fluid behaviors has been a challenging topic. Flow visualization enables us to visually acquire qualitative and quantitative flow information. There exist various software tools performing different flow visualization tasks. However, we lack tools that help students learn important flow field concepts. In this paper, we present a visualization app, named FlowVisual which runs on iOS devices, to illustrate basic flow field concepts in 3D. In order to meet a comprehensive learning goal for students, we integrate a number of techniques into FlowVisual design, including field-line tracing, field-line comparison, critical point detection and classification, template-based seeding, and surface visualization. We evaluate and demonstrate the effectiveness of FlowVisual by conducting a formal user study including an introduction and training session, an auto-grading test, and a post-questionnaire survey. Introduction Fluid mechanics and computational fluid dynamics (CFD) are among the core courses in many engineering majors such as mechanical engineering, aerospace engineering, biomedical engineering, chemical engineering, and civil engineering. In these courses, it is important for students to acquire the knowledge of fundamental flow field concepts. Many of those concepts are not straightforward to learn. For instance, it is not easy for beginninglevel students to fully understand the differences between various kinds of field-lines and critical points. Commonly, these materials are taught by instructors through explaining concepts and definitions, drawing diagrams and illustrations, and occasionally, playing custom-made animations or video clips. Using intuitive and real flow examples proves to be an excellent way of learning. However, most examples available today are only designed for lecture or demonstration but not for student interaction or selflearning. Developing a pedagogical visualization tool holds the potential to help students better learn these essential flow field concepts through interactive exploration. In this paper, we present FlowVisual, an educational app running on iOS devices, to illustrate basic flow field concepts in 3D. This app is an extension of the desktop version of FlowVisual for 2D flow fields [14]. The desktop version has been used in classroom teaching of CFD course for multiple times and has received positive feedback from students. From our user study, we found that the app helped students with no previous 2D flow field knowledge understand concepts to the similar degree of students who had studied those concepts before. This new mobile FlowVisual is developed to illustrate the concepts in 3D space as cases in 3D are more common yet more challenging to understand in practice. Besides different kinds of field-lines, we also implemented stream surfaces in this app to enrich the perception of the flow field characteristics in a more continuous fashion. Our key deliverable is an app for classroom demonstration and for self-study by students and professionals. Its implementation on iPad makes it highly portable and accessible by anyone who is interested in learning and exploring key flow field concepts. The app has been used in a classroom environment and its effectiveness was evaluated through a formal user study involving students from mechanical engineering, electrical engineering, and computer science. The mobile FlowVisual is freely accessible in the App Store. The tutorial and evaluation materials are also available online so that instructors and students who are interested in our work can make use of them. Related Work Flow visualization plays a vital role in many scientific, engineering, and medical disciplines, offering users a graphical representation of their vector data for visual understanding, interpretation, and decision-making. For over two decades, flow visualization has been a central topic in scientific visualization, and a variety of techniques including glyph-based [10], texture-based [6], integration-based [8], topology-based [7], partition-based [12], and illustration-based [1] visualizations have been presented. Our tool focuses on integration-based flow visualization as it is most widely used in practice. For integration-based flow visualization, particles or seeds are placed in a vector field and advected over time. The traces or field-lines that the particles follow, e.g., streamlines for steady flow and pathlines for unsteady flow, depict the underlying vector data. Teaching the core concepts of fluid dynamics has not experienced significant changes over the years. A few published works discussed some recent advances. Hertzberg and Sweetman [5] designed a flow visualization course focusing on studio/laboratory experiences for mixed teams of students. The course content includes fluid flow physics, history of photography with respect to the relationship with science and art, as well as flow visualization and photography techniques. Their course proved to be very successful in attracting both graduate and undergraduate students, engineering women in particular. Settles et al. [13] argued that fluid mechanics is fundamentally visual, and visual topics can be taught by modern multimedia methods. They described a new series of 10-15 minutes narrated videos that use flow visualization to illustrate basic fluid mechanics concepts. Rossmann and Skvirsky [11] developed a sophomore-level seminar that exposes students to flow visualization techniques and the science of fluid mechanics, and to the photographic methods needed to create effective images. The fundamentals of fluid flow and photography were taught and practiced in a studio setting. As an interactive visualization app for learning flow field concepts, our FlowVisual builds on the solid education tool of iPad and provides an alternative to the above methods. Terms We give a brief introduction to some important concepts of flow fields. These concepts are incorporated into our FlowVisual design. Flow Field A flow field (or vector field) is an assignment of a velocity vector to each point in the domain to represent the movement of the flow. Essentially, it is a mapping