Using Pre Built Program Templates To Teach Numerical Methods

Abstract

Most engineering programs require students to learn some form of structured programming early in their curriculum, but in many cases, students do not use their newly acquired programming skills again. While outside of computer-related majors there may not be a need to maintain programming skills, programming is still an effective way to make certain that students understand how a numerical method is arriving at a solution. A method called “template-based programming” was developed to enable use of high- level computer languages in courses where programming is not explicitly part of the course objectives. In this method, a student is given a fully functioning program, or template, that only lacks the functional code for a numerical method to solve a particular type of problem. Since the work of developing the interface and other portions of the program has been completed for the student, all they need to concentrate on are the aspects of the programming project that contribute toward the course objectives. Examples of how this approach has been used in numerous chemical engineering courses will be presented, including templates developed in Compaq Visual FORTRAN, Microsoft Visual Studio.NET, and Microsoft Excel.

In the University of Kentucky chemical engineering curriculum, students are required to take a course in computer programming prior to taking their first “core” chemical engineering course. Subsequent to that course, it is possible that a student will never to be required to write a complete program from “scratch” again. This makes some sense in chemical engineering and other disciplines where greater than 90% of practitioners never program in a high-level language.1

Computer programming concepts in some form are still required according to Accreditation Board for Engineering and Technology (ABET) criteria.2 While most curricula still require high-level languages such as C, C++, and Visual Basic,3 an increasing number of programs are choosing to teach scripting languages within mathematics applications such as Maple, MATLAB, and Mathematica.4

Even though most practicing engineers will not program, it is often argued that programming is an effective means of teaching problem formulation and problem