DEVICE

Dynamic Environment for Visualization in

Chemical Engineering

Noel Rappin, Mark Guzdial -- College of Computing

William Ernst, Peter Ludovice, Matthew Realff, Dennis Senol -- School of Chemical Engineering

I.The Problem of Connecting Theory to Practice

The goal of education is imparting the ability to apply new knowledge acquiredin the classroom to real-world problems. This requires the student to be ableto make the connections that link the theory to reality. Students mustunderstand what they know and when that knowledge is useful. Hands-onexperience makes the process of creating these connections easier for thestudent. (Kolodner, 1993) Traditional education has attempted to solve thisproblem in a number of different ways, including laboratory classes and co-oplearning. In some fields real experience, even laboratory experience, can beexpensive or dangerous. Furthermore, real world systems often do not encourageobservations that can be directly linked to the classroom theory.

We propose to use computer simulations to offer the students the opportunity todo problem-solving and design in contexts similar to that of the real world.Computer simulation can be an effective way of giving students experience thatcan create these links between theory and practice.

There are two critical problems that need to be addressed in a simulationenvironment designed for learning.

• Context: The simulations must help students connect the simulationactivity to the real world context. If the simulations seem to be no more thanvideo games, they will not help in connecting theory to practice. Our solutionto this is to make the student an active builder of the simulations. Further,we want students to trace through the effects of their design decisions to thefinal results. In real world contexts, measuring the result to determine if theevaluation criteria were is a non-trivial task (real meters are imprecise, forexample). This is part of the real-world context that we want students tolearn. (diSessa 1986)

• Complexity of the simulations: A simulation environment that iscomplex enough to allow creative activity is as complicated as any programming

Chemical engineering is a field where conventional lab techniques are not ableto provide experience with the full range of complexity in a timely fashion.The cost of creating a lab were students can build objects of even mildcomplexity is prohibitive, and the potential danger is high. Because there isa lack of opportunity for hands-on experience, it is not surprising thatchemical engineering students at Georgia Tech graduate proficient in theory,but have ignored certain practical issues that arise in real equipment designand operation. These issues are often associated with the mechanical andeconomic aspects of chemical engineering process units.

• Students design distillation columns which satisfy thermodynamic designcriteria without addressing column hydrodynamics, which would in practice causethe column to be inoperable due to liquid weeping through the trays or floodingdue to high gas velocities.

• In a capstone design course, inattention to the physical setup of the systemcaused students to solve a problem with a pump system that was 1,000 times moreexpensive than necessary. They had failed to connect the theory to the geometryof the problem to determine the height to which the liquid had to be pumped.

In both of these cases, students did solve the problems correctly from thepoint of view of meeting the theoretical requirements of the problem. However,students did not understand real world problem-solving and design whichinvolves meeting all constraints and reducing costs.

Students do learn to deal with real contexts in actual industrial practice, butthat is expensive and only takes place very late in the student's education.If we could help students understand earlier the contexts in which realchemical engineering occurs, their learning might be better made moreauthentic, more integrated, and more usable later. However, it is difficult tohave beginning and intermediate students working with real chemical engineeringplants and simultaneously integrate their theoretical coursework with thispractical experience. A properly constructed computer simulation can fill thisneed, and allow the students to have the contextualized experience they need.

II.Our Solution: Dynamic Environment for Visualization in Chemical Engineering(DEVICE)

We are constructing a simulation environment structure which we call DEVICE(Dynamic Environment for Visualization In Chemical Engineering). The DEVICEstructure provides for the contextualized and scaffolded simulationenvironments needed to meet the needs of chemical engineering students.

• We provide for context through the use of visualizations andmultimedia annotations. Students assemble simulations in DEVICE bymanipulating visualizations of components that chemical engineers use (Figure1). These components are linked to multimedia annotations which providegraphics, videos, and other explanations that serve to strengthen theunderstanding of these visualizations as representative of real-worldsituations.

• We provide for scaffolded, creative activity through a programmablesimulation construction kit for both components and measuring devices.While we are not interested in teaching students software engineering and C++,we are interested in having students program their own simulations to theextent that they (1) define and order pieces of the simulation and (2) debugthe simulation (e.g., get results other than expected, determine the cause,repair, and further debug). This non-traditional programming process allowsthem to creatively design solutions to problems and to trace the effects oftheir design decisions to the final outcomes. We will support programming forsimulation-building through use of software-realized scaffolding techniques tofacilitate the programming process [Guzdial, 1994]. This simulationconstruction kit will be domain-independent.

We are currently developing the first generation DEVICE simulation (Figure 1).This simulation will allow for students to assemble and manipulate parametersof simulation devices (Figure 2), as well as construct and use measuringdevices (Figure 3). We plan to have a functioning prototype of this firstgeneration simulation by February 1995 which we plan to evaluate with chemicalengineering undergraduates during the Winter 1995 quarter. The first pilot testwill involve a small sample of students in laboratory conditions. We intend touse the results of the pilot test in creating a second generation of DEVICE.This version would contain more included components and multimedia annotations,and include a scripting language by which the students could create their owncomponents. We will evaluate the second simulation for a single project usingall students in a chemical engineering class.

Future plans include the development of DEVICE simulations throughout theChemical Engineering curriculum. In addition, we hope to use the core userprogramming interface to create environments for building simulations in otherdomains. The role of DEVICE will be to complement traditional lectures ordemonstrations with an interactive instructional component that stresses thepractical application of the theory and emphasizes student design activity.

WorksCited

Guzdial, M. (1994). Software-Realized Scaffolding to Facilitate Programmingfor Science Learning. Interactive Learning Environments. (When?)

Kolodner, J. (1993). Case Based Reasoning. San Mateo, CA: MorganKaufmann Publishers