Teaching Computational Fluid Dynamics Using MATLAB: November 2013
Teaching Computational Fluid Dynamics Using MATLAB: November 2013
net/publication/267597038
CITATION READS
1 5,244
2 authors, including:
Amitabha Ghosh
Rochester Institute of Technology
27 PUBLICATIONS 179 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
All content following this page was uploaded by Amitabha Ghosh on 18 August 2015.
IMECE2013-66458
ABSTRACT mechanics but they can still generate a grid on GAMBIT and
A set of MATLAB modules has been developed for an run flow calculations using FLUENT. In this paper these
introductory graduate course on computational fluid dynamics approaches will be called the conventional or layman's
(CFD) at Rochester Institute of Technology (RIT). These approaches to CFD. Without the analytical rigor required to
modules can provide a breakthrough in CFD education because understand the underlying theory of CFD the final user of CFD
they can assist both learning and comprehension, while software does not necessarily understand the mathematical
avoiding analytical mistakes by the CFD learner. With behavior of fluid dynamics, the numerical procedures for
advances in CFD and availability of software, students in upper solving partial differential equations (PDEs), the different
level fluid mechanics classes have less incentive to learn theory discretization schemes behind the codes, or the necessary
and the tools for abstract thinking. This paper proposes for the stability analysis. For this reason, incorrectly computed results –
first time an alternate approach to teaching and learning of CFD which at first sight might seem reasonable– may be accepted by
through the use of symbolic computation in MATLAB, while students as valid. In order to reduce this possibility, students
preserving the accuracy and content of abstract analyses. should be taught to properly select a physical model, understand
applicable governing equations, initial and boundary conditions
INTRODUCTION plus the domain discretization used for the simulation. But with
In the past 30 years, advances in computer technology have the current abundance of software in the marketplace how is
brought significant breakthroughs in engineering and science. this possible? In fact, the engineering advancement has made
Side by side engineering computations were assisted by the students lazy or, disinterested to learn the hard way that our
expansion of both memory and data storage. Most conventional forefathers had to practice. Innovation in educational research
engineering problems in fluid mechanics have been solved and assisted development of software technology to provide clarity
an abundance of flow software is available in the market at in thinking and assist visualization.
nominal costs. Most of the engineering graduate schools have Computer animation is a powerful tool which can enhance
introduced CFD in their curriculum. Student thesis and project student understanding significantly when properly designed.
work are routinely solved using FLUENT or CFX where Traditionally, this is where computers could assist advancing
students need not learn the theory of CFD. As a result CFD has education in a major way. Historically, campus-wide animation
become a buzzword where many people do not even know what driven courseware were developed as early as 1988 under the
it is really about. It is true that the field of CFD is fairly broad project Athena at MIT [1]. All modules developed under that
beginning with formulations and ending with data analysis and project were designed by students with supporting staff experts
display. Most of all, today results can be achieved by pressing in X-windowing software and overseen by faculty mentors. The
one button on the computer keyboard, or a mouse-click. first author was a faculty evaluator, and has since used these
In the past, many educators have used CFD to solve tools (e.g., the TODOR courseware) successfully in teaching of
engineering questions themselves and taught students the fruit aerodynamics and fluid mechanics. Although such tools are
of their learning. Or they have assisted students to find their very user friendly and stand-alone computational marvels, they
own answers, while they have not been trained in the theory of require close monitoring of a seasoned instructor for enhanced
CFD. In the graduate programs this approach is the most student learning in complex topics. Moreover they could not
prevalent. A student may not have a working knowledge of fluid
*
Professor, Member ASME. +Graduate Student, Department of Mechanical Engineering.
PART 1 – THE CONVENTIONAL APPROACH Figure 2. Quiz Question from Convective Phenomena
Connectivity Example A:
Courseware Examples:
Figure 1. Test Question from Convective Phenomena to
Home Work in Introduction to CFD Analysis A. Superposition in Ideal Flows using Panel Method
0.5
-1
-1.5
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5
B. Vortex Panel Method to Calculate Airfoil Results
Table 3: Comparison of Analytical Modified Equation [Ref. 15] with MATLAB generated table through u xxx term