Virtual Reality in Chemical Engineering Education

THE APPLICATION OF VIRTUAL REALITY TO CHEMICAL ENGINEERING EDUCATION

John T. BellH. Scott Fogler

Department of Chemical Engineering

University of Michigan

2300 Hayward, 3074 H. H. Dow Bldg.

Ann Arbor, MI 48109-2136JohnBell@umich.edu

http://www.engin.umich.edu/dept/cheme/bell.html

KEYWORDS: Chemical Engineering, Education,Interactive Simulation, Virtual Reality.ABSTRACT

Virtual reality ( VR ) is an emerging

computer interface that has the potential to havetremendous impact on engineering education, byproviding students with new insights into theirstudies and permitting them to explore environmentsthat would be otherwise inaccessible. Howeverbefore that potential can be fully exploited,engineering educators must first learn not only themechanics of VR, but also the intricacies of howbest to apply this new tool to scientific and technicaleducation. In order to develop techniques for theeffective application of VR to engineering education,ongoing research in the department of chemicalengineering at the University of Michigan hasproduced three major and numerous minor VRbased educational modules, designed to aid in theinstruction of chemical engineering topics. Besidesdeveloping effective methodology, another primarygoal of this research is to reach as many studentsas possible on a nationwide basis, which requiresthe use of relatively inexpensive ( studentaffordable ) personal computers as a base platform.Effectively portraying technical information in realtime using minimal computing power requiresspecial simulation techniques that are unique to thisenvironment. This paper provides a brief descriptionof the VR modules developed to date, includingsome of the special simulation techniques that theyincorporate, and discusses steps that are currentlybeing taken to reach a wider audience throughVRML and other world wide web-based techniques.

BACKGROUND:Description of VR

Virtual reality ( VR ) is an emerging

computer interface that strives to increase therealism and impact of simulations by placing theuser in the center of an interactive three dimensionalenvironment, complete with spatialized sound,haptic feedback, and eventually olfactory and tastefeedback as well. Technological devices used todeliver this experience typically include head-mounted displays ( HMDs ) and wired gloves,however the real critical component to deliveringeffective VR is a graphics system capable ofrendering three dimensional graphics at areasonable frame rate. Where computing power islimited, compromises must be made in terms ofmodel details, and special techniques must beincorporated to squeeze the most performance fromthe available equipment.

Benefits of VR to Engineering EducationThere are many educational benefits offered

by VR, as evidenced by hundreds of papers in theliterature (Panteledis) applying VR to K-12education. For engineering educators, one of theappealing features of VR is the ability to takestudents to places that are otherwise inaccessible,such as the surface of Mars, the inside of anoperating reactor, or between the plates of acapacitor. Another strong benefit is to reachstudents who have alternate learning styles (Felder),particularly those who are visual, active, and globallearners. Three dimensional visualization is yetanother strong benefit of VR, as might be useful inCAD classes, mechanical engineering applications,and for viewing atomic crystal structures.

Unique Simulation Problems in EducationalVirtual Reality

Most scientific computer simulations require

first and foremost an accurate result. Whileexecution speed and robustness are also importantcriteria, they are subject to the constraint ofdetermining the correct answer to within a tighttolerance, often six digits of precision or better. VR,on the other hand, requires fast screen updates inorder to be believable. Rather than asking \"Howfast can we get the answer?\must constantly ask \"How much detail and realismcan we add and still complete all calculations( including graphics rendering ) within a tenth of asecond?\"

The classic solution to this problem is to use

a high powered computer, with a graphics systemcapable of delivering whatever performance level isrequired. However an educational simulation willhave little impact if students do not have access tothe computing power necessary to run it. Thereforeeducational VR is faced with the difficulty ofdelivering as much detail and speed as possible oncommonly available inexpensive computerplatforms.

In addition to the cost of the computing

engines, educational VR is also constrained by thelimited performance of student affordable peripheraldevices. For example, high quality HMDs haverecently dropped in price from roughly $70,000 to amere $20,000, which is still far out of the reach ofmost students. There are devices available for lessthan $1000, but their visual resolution is sufficientlypoor to render the user legally blind in many cases.Because of this price-performance tradeoff,

the most successful and well-known VR applicationscurrently fall into one of two categories: high budgetapplications where the benefits justify the high costof quality equipment ( e.g. military flight simulatorsand medical applications ), and the entertainmentindustry. The latter can tolerate poor visual qualitybecause the user is generally too excited and busyshooting things to notice details or lack thereof.Educational VR is stuck in the middle, attempting todeliver high quality technical instruction, with anabsolute minimum of CPU cycles, in an atmospherewhere the user cannot read more than a few words( in huge fonts ) or discern any meaningful details.

The HMD also blocks user access to the keyboardand to supporting documents, in exchange fordelivering a high degree of immersion.

DESCRIPTION OF DEVELOPED SIMULATIONSA number of educational applications have

been developed at the department of ChemicalEngineering of the University of Michigan to explorethe use of VR as an effective educational tool (Bell &Fogler, 1995, 1996a-c). The three most significantapplications - Vicher 1, Vicher 2, and Safety - are allsimulations of chemical plants. These applicationsaddress the topics of catalyst decay, non-isothermalreaction conditions, and chemical plant hazardanalysis respectively. Scenes from each of theseapplications are shown below:Vicher1

Vicher 1 consists of three different reactor

rooms, three microscopic areas, and a welcomecenter that serves as a central location. The reactorrooms illustrate slow, medium, and fast mechanismsof catalyst decay, ( time scales of minutes, hours,and days to weeks ) and the industrial methods forhandling each case. The three microscopic areasillustrate the mechanism of solid catalyzed reactionswithin a porous medium, in increasingly strongermagnification. The first microscopic view shows asingle pellet the size of a small planet. Flying insideshows the interior of the porous structure where thereaction mechanism is first illustrated, and zoomingin further shows a close-up of the pore wall with onlya single reacting molecule within view.

Figure 1: The transport reactor room allows students toview the effects of changing reactor conditions. Thecutaway pellet in the corner illustrates the shrinking coremodel of catalyst decay.

Vicher2

Vicher 2 currently consists of three reaction

engineering areas, illustrating different topics in non-isothermal reactor design, and a central welcomecenter. As in Vicher 1, students can operate most ofthe virtual reactor equipment, and can observeinternal conditions either by turning the reactorstransparent, or by stepping inside the equipment.

Figure 2: The non-isothermal packed bed reactor roomuses color-coded 3-D graphs to illustrate the kineticsassociated with a tubular reactor.

Safety

The safety application allows students to

explore a chemical plant in order to analyze thehazards present and the safety systems in place tohandle those hazards. This simulation has muchmore realistic detail than that found in the Vichermodules, because it is based upon photographs andobservations taken at an actual facility. On the otherhand, this is a static world, and the only action thatstudents can take besides visual exploration is tobring up the associated \"help\" documentation,containing a full description of the items present,MSDS information, and equipment photographs.

Figure 3: This safety and hazards analysis moduleincludes a help document ( inset ) containing photographsof the industrial site being modeled.

Exploratory Applications

In addition to the three major applications

described above, a number of smaller applicationshave been developed, in order to explore thecapabilities of VR as an educational tool, and to testand illustrate particular techniques. Theseapplications can be categorized roughly into twoareas: three dimensional spatial relationships, andthe exploration of information space. The formerincludes visualizations of metallic crystal structuresand fluid flows, while the latter includesthermodynamic relationships and azeotropicdistillation diagrams. Some views of these smallerexploratory applications are shown here:

Figure 4: Top row shows body-centered and face-centered crystal structures. Bottom row shows fluid flowand thermodynamic relationships.

Further details regarding all of the

applications described above can be found in thecited references, and on the web sitehttp://www.engin.umich.edu/labs/vrichel.

SPECIAL TECHNIQUES DEVELOPED FORIMPLEMENTING EDUCATIONAL VIRTUALREALITY

In order to overcome the unique problems

associated with delivering educational VR on studentaffordable equipment a number of specialtechniques have been developed. These techniquesinclude tricks for displaying scientific concepts asquickly as possible, as well as methods forovercoming the general inability to read text in astudent-affordable virtual environment. A few ofthese special techniques are described here.

Billboarding Molecules

Figure 5: The interior of a microscopic catalyst pore,illustrating reaction mechanisms in Vicher 1.

Within the catalyst pore shown above there

are fifty or more molecules, continuously moving,reacting, and bouncing off the walls. If they were tobe modeled as ball-and-stick figures, they wouldrequire a minimum of 150 polygons each, or 6500moving polygons that must be tested for collisionsand view blockage with each frame update. Instead,each molecule is modeled as a single polygon, onwhich is applied a picture of a molecule. Thistechnique is known as texture mapping, which isalso used for modeling detailed flat objects such asbrick walls and bookshelves. In the case of themolecules, the polygons must also be turned to facethe viewer, which is called billboarding. The rotationstep was optimized by matching the orientation ofthe molecules to the user's view, rather than makingthem truly orthogonal to the vector between the userand the molecules. The result is correct in thecenter of the screen ( focus of interest ), andsufficiently accurate at the edges not to be noticed.Considering that the execution speed varies with thesquare of the number of visible polygons, thereduction from 6500 to 50 is significant.Graph Display

Several different techniques have been

developed to display scientific graphs. The first wasto produce a 3-D graph using externally calculateddata and terrain generation techniques, as shown inFigure 2 above. This surface was then coloredpolygon by polygon to indicate the temperaturecoordinate. ( red=hot; blue=cold. ) By coloring thevirtual reactors with the same color scheme, atangible link was made between the mathematics

and the equipment. \"Lines\" in 3 space weregenerated as long thin cylinders.

Another technique used to portray 2-D

graphs within a virtual world was to first generate thegraphs using a popular spreadsheet program, andthen apply the captured result as a texture map on asingle polygon. When the user changes reactoroperating conditions ( among pre-determinedchoices ), the texture map is replaced with theimage appropriate for the new situation. In order tomake the graph \"grow\" over time in the time-temperature room, the completed graph is displayedpartially blocked by a cover slip that is \"shrunk\" overtime to reveal the graph. While these graphicaltechniques do not provide highly accurate detail oruseful run-time calculations, they do illustrate thequalitative trends to within the visual resolution oflow-cost HMDs.

Figure 6: Graphs in the time-temperature reactor roomshow temperature, catalyst activity, and conversionchanging with time.

Displaying Text ( or Not )

Traditional educational computer modules

present the user with large amounts of text to beread on the screen. ( In extreme cases they canamount to little more than a computerized version ofthe textbook. ) Several methods have beenexplored for displaying text in VR, without yielding acompletely satisfactory solution. Very large fontsare sufficient for one or two word labels, but not forlengthier texts. Auditory narration also has itsbenefits, but is not universally useful. Virtualtelevision sets have proven to be a usefulinformation delivery device, with each \"channel\"having a different still image ( texture map ) andauditory component. Finally the MS Windows \"Help\"

facility has been utilized, allowing users to click on avirtual object to bring up a separate window withtextual and graphical information. By includingphotographs in these help files a level of detail canbe provided that would not be possible otherwise, asshown in Figure 3. However in general theconclusion has been that VR is not an appropriatemedium for delivering textual information, andtherefore educational VR applications should striveto avoid text as much as possible.

FUTURE EFFORTS FOR DELIVERING VIRTUALREALITY TO STUDENTS

Initially it was stated that in order to reach

the widest possible student audience, educationalapplications needed to be designed for commonlyavailable student affordable personal computers.While there is still a strong basis for that argument, anew vehicle has evolved over the past few years forquickly and easily delivering computer-basedinformation to the masses: the world wide web.Now, with the growing popularity of the virtual realitymodeling language ( VRML ), VR can be expectedto ride the wave of the ever expanding Internet. Ourcurrent efforts in exploring this field involveconverting some of our simpler applications toVRML format, and developing new applicationsusing another web-based VR development tool,WorldUp (Sense8). The latter is a GUI basedinterface for the development of virtual worlds, thatis supposed to eventually include a freely availableNetscape plug-in for anyone to view and interactwith the resulting virtual environment. We havestarted to make sample worlds in both VRML andWorldUp format available on our web site:http://www.engin.umich.edu/labs/vrichel. Howeverthe VRML export is still less than satisfactory, andthe WorldUp Netscape plug-in is not yet available.CONCLUSIONS

In order to effectively apply VR as an

educational tool in engineering and other technicalareas, a number of unique simulation difficultiesmust be identified and solved. Key among these isthe shift in emphasis from accuracy to speedimposed by the need to maintain high frame rates onstudent affordable PCs and the low resolution ofinexpensive viewing devices. A number of specialtechniques have been developed in the course of

developing three major and numerous minoreducational VR applications. Some of thetechniques described here include the display ofmoving ( reacting ) molecules, the display of 2D and3D graphs, and a discussion of different methods ofpresenting textual information in a non text-basedenvironment. Future advances in computertechnology will bring ever more powerful graphics tostudents' desktops, which will relieve some of thecurrent simulation restrictions. The ever burgeoningworld wide web will eventually include VR, howeverthe currently implemented solutions are not sufficientfor practical educational use.ACKNOWLEDGMENTS

The authors gratefully acknowledge the

undergraduate student programmers that haveassisted in the development of the Vicher modules,specifically Christian Davis, Darren Obrigkeit, ShawnWay, Jeroen Spitael, Paul Sonda, Anita Sujarit,Scott Whitney, Adam Deedler, and Pieter Spitael,( in chronological order. ) Thanks are also given toDr. Joseph Louvar and Lawrence James of BASFChemical Corporation and to Tom Pakula ofMarathon Oil Company for the valuable resourcesthat they have provided, and to the University ofMichigan Department of Chemical Engineering for

initial funding.

This project was supported, in part

by the

National Science Foundation

Opinions expressed are those of the authors

and not necessarily those of the Foundation

REFERENCES

Bell, John T., and H. Scott Fogler, 1995, \"TheInvestigation and Application of Virtual Reality as anEducational Tool\Proceedings of the AmericanSociety for Engineering Education AnnualConference, Anaheim, CA, June 1995.

Bell, John T., and H. Scott Fogler, 1996a,\"Preliminary Testing of a Virtual Reality BasedModule for Safety and Hazard Evaluation\Proceedings of the 1996 Illinois / Indiana ASEESectional Conference, March, 1996, BradleyUniversity, Peoria, IL.

Bell, John T., and H. Scott Fogler, 1996b, \"RecentDevelopments in Virtual Reality Based Education\Proceedings of the American Society forEngineering Education Annual Conference,Washington, DC, June 1996.

Bell, John T., and H. Scott Fogler, 1996c, \"Vicher:A Prototype Virtual Reality Based EducationalModule for Chemical Reaction Engineering\Computer Applications in Engineering Education,4(4), October, 1996.

Felder , R. M. and L. K. Silverman, 1988, \"Learningand Teaching Styles in Engineering Education\Journal of Engineering Education, 78(7), 674-681,April, 1988.

Pantelidis, Veronica S., \"Virtual Reality andEducation: Information Sources\.washington.edu/pub/scivw/citations/VR-ED.html.Sense8 Corporation, 100 Shoreline Highway Suite282, Mill Valley, CA 94941, (415) 331-6318,http://www.sense8.com.

AUTHOR BIOGRAPHIESJohn T. Bell

( Lecturer, Department of Chemical

Engineering, University of Michigan, 3074 H.H. DowBldg, Ann Arbor, MI 48109-2136, ( 313 ) 763-4814,JohnBell@umich.edu, http://www.engin .umich.edu/dept/cheme/bell.html ) John received his BS inChemical Engineering in 1984 from Georgia Instituteof Technology, followed by a year of graduate studyat l'Institut du Génie Chimique, in Toulouse France,where he received a Diplôme des ÉtudesApprofondes. At the University of Wisconsin,Madison, he received an MS in ChemicalEngineering in 1987, an MS in Computer Science in1988, and a Ph.D. in Chemical Engineering in 1990.John taught advanced computer courses at a privatecomputer training firm for four years.

His current research work combines his chemicalengineering and computer science skills to study theapplicability of virtual reality to chemical engineeringand education. His other research interests alsoinvolve the application of emerging computertechnologies to chemical engineering and education.H. Scott Fogler

( Vennema Professor of Chemical Engineering,Department of Chemical Engineering, University ofMichigan, 3074 H.H. Dow Bldg, Ann Arbor, MI48109-2136, ( 313 ) 763-1361, H.Scott.Fogler@umich.edu, http://www.engin.umich.edu/dept/cheme/ fogler.html ) Scott has over 130 researchpublications, including \"The Elements of ChemicalReaction Engineering\" ( the most used book on thissubject in the world) and “Strategies for CreativeProblem Solving.” Scott was the 1995 Warren K.Lewis award recipient of the AIChE for contributionsto chemical engineering education.

In 1980, Professor Fogler was a first

recipient of the newly instituted award forOutstanding Research from the University ofMichigan College of Engineering, and also in 1980,received the Chemical Engineer of the Year Awardfrom the Detroit section of the American Institute ofChemical Engineers. In 1987, he received theUniversity of Colorado Distinguished AlumnusAward, and in 1988, he was elected President of theComputer Aids for Chemical Engineering (CACHE)Corporation.