viewing point through the model, creation of animated motion sequences to illustrate maintenance or service operations, and even the creation of movie sequences for pop popula ularr con consum sumpti ption on or for tra traini ining ng pu purpo rposes ses.. Continuing improvements in both hardware and soft ware will make these more realistic and less expensive. Fig. 9 (see Color Page CP-2) illustrates the dashboard of the bus shown in Fig. 3. This dashboard is seen as the viewing viewing poi point nt for the the vis visual ualiza izatio tion n that that does does a
design, engineering analysis, detailed design, drafting and documentation, and manufacturing design. This article describes the application of computers to engineering analysis.
‘‘fly through’’ of the interior of the bus—all aspects of the bus, including the dashboard, exist solely in the digital model.
ical ical mulated fun functi ctions ons, , approx appprevious roxima imatio n method met hods, dat data a accu accumula ted from prev ioustion experien expe rience ce s, andand phys physical ical testing to simplify their analyses. Some analyses were so time-consuming that, when done at all, they could be completed only for one simplified example. This frequentl frequ ently y led to underunder- and over-desi over-designed gned systems systems.. The first case resulted in systems that did not work proper pro perly ly or fai failed led outrig outright ht.. In the the second second cas case, e, the systems were more expensive than necessary or too heavy heav y to meet their goals. Physical Physical prototyp prototypes es were (and remain) very costly and time-consuming to build and and te test st—an —and d they they ofte often n ha have ve to be recr recrea eate ted d as designs are changed.
Bibliography 1988. Mantyla, M. Introduction M. Introduction to Solid Modeling . Potomac, MD: Computer Science Press. 1990. Rogers, D. F., and Adams, J. A. Mathematical Elements for Computer Graphics , 2nd Ed. New York: McGraw-Hill. 1996. Foley, J. D., Van Dam, A., Feiner, S. K., and Hughes, J. F. Computer Graphics: Principles and Practice , 2nd Ed. Reading, MA: Addison-Wesley. 1996. Senerson, J., and Curran, K. Computer K. Computer Numerical Control: Operation and Programming . Upper Saddle River, NJ: Prentice Hall. 1997. Hearn, D., and Baker, M. P. Computer Graphics , 2nd Ed. Upper Saddle River, NJ: Prenti Prentice ce Hall. 1997. Rehg, J. A. A. Introduction to Robotics in CIM Systems , 3rd Ed. Upper Saddle River NJ: Prentice Hall. 1997. Rogers, D. F. Procedural Elements for Computer Graphics , 2nd Ed. New York: McGraw-Hill. Barry Flachsbart, David Shuey, and George Peters
COMPUTER-AIDED COMPUTER-AID ED ENGIN ENGINEERING EERING (CAE) For articles on related subjects see C OMPUTER-AIDED DESIGN/COMPUTER-AIDED M ANUFACTURING ; CONTROL APPLICATIONS; DIGITAL D ESIGN A UTOMATION ; FINITE ELEMENT M M ETHOD; ROBOTICS; and SIMULATION.
Introduction computer-aided engineering (CAE) are: The goals of computer-aided •
improved produ product ct quality
•
improved safety
•
reduced enginee engineering ring time, achieved through fewer design iterations
•
improved produc productt functionality and usability
•
reduced number of prototypes prototypes,, ultimately leading to their elimination in many cases
•
reduced produ product ct cost
Historically, Historica lly, engi engineer neerss analy analyzed zed desig designs ns by building building and test testing ing phys physical ical prot prototype otypes, s, performin performing g calcu calculalations by hand or with some computing aid such as a slide rule. They frequently used tabulated mathemat-
History Th The e adve advent nt of anal analog og (q.v.) and dig digita itall compu computer terss provided engineers with systems capable of analyzing desi design gnss mu much ch more more qu quic ickl kly y and and allo allowe wed d them them to undertake unde rtake analyses analyses that were previousl previously y impractica impracticall to atte attempt. mpt. However, However, early comput computer er systems systems were too slow and limited in capac capacity ity (memory, storage, storage, I/O spe speed) ed) to handle handle extre extreme mely ly lar large ge or comple complex x mechanical systems. While they provided a base for new, more extensive design evaluations, many of the historical problems remained and new problems arose. These The se inc includ luded ed lim limite ited d access access to expen expensiv sive, e, hig highhpowered computing systems and difficulties describing the physical form of designs in a way that computers could work with them efficiently efficiently.. Therefor Therefore, e, many early analysis programs used unrealistically simplified, schemati sche matic-lik c-like e descripti descriptions ons of the phys physical ical syste system. m. It was impossible to describe any but the simplest system’s geometry within the computing environment. Wi With th the the adve advent nt of co comp mput uter er-a -aid ided ed de desi sign gn and and computer comp uter-aide -aided d manufact manufacturin uring g (CAD (CAD/CAM /CAM)) in the earl early y 1970 1970ss and and th the e rapi rapid d adva advanc ncem emen ents ts in co commputer system performance from 1960 to the present, most mo st te tech chno nolo logi gica call ba barr rrie iers rs to CAE CAE have have fall fallen en.. Engineers Engin eers can now have enough comput computing ing powe powerr on their desks to solve any but the largest of problems. For ext extreme remely ly comp complex lex probl problems, ems, supe supercom rcom-puters (q.v.) may be employed.
Engineering analyses can be used to evaluate and pre-
Engineering workstations (q.v.) pro provid vide e extens extensive ive
dict the behavior of new designs, as well as to evaluate the perf performa ormance nce of exis existing ting designs. Engineer Engineerss use computers for a number of tasks, including conceptual
computin comp uting g powe powerr with high high-reso -resolutio lution, n, high-spe high-speed ed graphics systems graphics systems at very reasonable reasonable and continua continually lly decreasing cost, below $5,000 in 1999.
COMPUTER-AIDED ENGINEERING
275
Three broad areas of the enginee Three engineering ring disciplin discipline e are supported by CAE: mechanical, civil, and electrical. In a typical situation, an engineer will use a CAD/CAM syst system em to deve develo lop p a mo mode dell of a syst system em (b (be e it me me-chanical, electrical, electromechanical, or otherwise) that is to be analyzed. Other required data, such as a finite finite ele eleme ment nt mesh, mesh, mechan mechanica icall pro proper pertie ties, s, con con-straints, and loading, are then developed on or linked to this this geom geomet etri ric c desc descri ript ptio ion n of th the e syst system em.. Th The e
In FE FEA, A, a mode modell of a co comp mpon onen entt or asse assemb mbly ly (see Fig. Fig. 1, Co Color lor Pa Page ge CP-3 CP-3)) is de deco comp mpos osed ed into into di disscrete cret e pieces pieces called called the finite finite eleme element nt mesh (FEM). Loads and supporting constraints are applied at mesh loca locati tion ons. s. Th The e finit finite e elem elemen entt mo mode deli ling ng soft softwa ware re creates a series of simultaneous linear equations that rela relate te ea each ch mesh mesh elem elemen entt to it itss neig neighb hbor ors. s. Very Very complex problems can be analyzed by solving these simu simult ltan aneo eous us eq equa uati tion onss it iter erat ativ ivel ely y (see MATRIX
analysis software is used to analyze this combination of model and related data, with its results presented to the engineer in various forms: tabular, graphical, animation, changes to the geometric model, etc.
COMPUTATIONS).
Mechanical Engineering Applications V OLUME OLUME PROPERTIES Vari Variou ouss volu volume metr tric ic prop proper erti ties es ca can n be comp comput uted ed direct directly ly for sol solid id and surface surface mod models els in most most CAD/ CAD/ CAM systems. These properties include lengths, areas, and volumes as well as mass, centroid, first and second moments of inertia, and products of inertia. In many CAD systems the results of this analysis can be transferred directly into structural and mechanism analysis applications where they are required as data. A few systems can compute volumetric properties for components consisting of composite materials.
FINITE ELEMENT ANALYSIS (FEA) Finite element analysis methods are used to perform several types of engineering analyses. These include: •
Struct Structura urall ana analys lysis is of a com compon ponent ent’s ’s beh behavi avior or under various kinds of applied loads and supports. Linear Line ar static, static, modal, modal, dynam dynamic, ic, forc forced ed resp response onse,, and buckling conditions can be analyzed. Special programs progr ams also exis existt for analyzing beam and grid structur stru ctures, es, such as thos those e used in ship ships, s, bridg bridges, es, frames of buildings, and other similar systems.
Triangular mesh elements are the easiest to create and analyze and are the most often used, but engineering workstations can process much more complex mesh types, typ es, lea leadin ding g to hig highe herr acc accur uracy acy and, and, in som some e cases, cases, to more easily defined models. Parts modeled as plates, surface shells, or solid models (see Fig. Fig. 2, Color Page CP-3) can be meshed with elements that include twoand thre three-dim e-dimensi ensional onal tria triangula ngularr and rect rectangu angular, lar, parabolic, tetrahedral, quadrilateral, shell, solid, and beam elements. The mesh can be created manually (usually in an interactive mode with the mesh building program) or automatically by the program with manual refinements. Loading cond Loading condition itionss may inclu include de point point,, distribu distributed, ted, to torq rque ue,, hy hydr drau auli lic, c, and and ot othe hers rs.. Su Supp ppor orts ts ca can n be anchor anc hored, ed, free, free, pinned pinned,, hinged hinged,, slidin sliding g as well well as most other kinds of mechanical connection (see Fig. Fig. 3, Color Page CP-3). Several schemes are used to increase Several increase finite elem element ent analysis accuracy and reduce computational requirements. These include feature suppression and adaptive meshing . In feature suppression, the geometry of a part model that is to be analyzed is simplified by temporarily removing features (fillets, small holes, bosses, flanges flan ges,, et etc.) c.) that that repres represent ent det detail ailss that that the the analys analystt feels feels are not going to have an appreciable impact on the
•
Thermal analysis of a structure structure’s ’s behavior when it is subjected to heating and cooling.
•
Combined structu structural ral and thermal analysis.
•
Pl Plas asti tic c mo mold ld and and part part anal analys ysis is th that at ex exam amin ines es various factors having to do with mold filling and the the shap shape e of the the mold molded ed part part.. Mo Mold ld anal analys yses es includ inc lude e pla plasti stic c flow into into th the e mol mold, d, materi material al and mold temp temperatu erature re gradi gradients ents,, and mold cooli cooling. ng. Sh Shri rink nkag age e an and d warp warpag age e ca can n be pred predic icte ted d fo forr finished parts.
•
Fluid mechanics analysis of th the e flow of fluids, such
overall validity of the FEA results. Since these types of feature feat ure usually usually prod produce uce high highly ly comp complex lex mesh structures that unnecessarily increase the number of simultaneous equations to be solved, their removal shortens the solution process. Adaptive FEM systems automatically refine the mesh definition in areas of detail that contribute to high stress or other qualities or coarsen the mesh in areas of lesser stress. These refinements provide prov ide increase increased d accuracy accuracy and shor shorter ter processin processing g time,, respecti time respectively. vely. This process is done as a closedclosedloop—analyze loop—analyz e the object, refine the mesh, analyze, refine, refi ne, etc.—u etc.—unti ntill the the res result ultss con conve verge rge to a user-d user-defi efined ned tolerance. A few systems offer design optimization, in which the results of the analysis are used to modify the geometri geom etric c model model automati automatically cally in orde orderr to match a
as air, water, and lubricants around the surfaces of an object. Pressure, fluid velocity, and other factors can be determined.
design goal such as, for example, obtaining the lowest weight that will withstand the specified loading conditions ditio ns at a particula particularr safety safety factor. factor.
276 COMPUTER-AIDED ENGINEERING ENGINEERING
Both standalone FEA systems and CAD/CAM systems support supp ort simulati simulation on and displ display ay of resu results. lts. A typic typical al display for a structural analysis might show the geometric model overlaid with stress contours, with color indicating the magnitude of the stress ( see Fig. Fig. 4, Color Page CP-3). Another type of display frequently used is an animated view of the model as it deforms under cyclical loading. These types of display can be com bined to produce simulations that are relatively easy for the engineer to understand, making the reading of large tables of stress values unnecessary.
MECHANISM ANALYSIS Mechanism analysis studies the behavior of mechanical systems undergoing motion. These systems may be comprised of rigid and flexible parts and their interconnecti conn ections. ons. Typic Typically, ally, the geom geometry etry,, mass mass,, iner inertia, tia, complianc comp liance, e, stiffness stiffness,, and damp damping ing of the syste system’s m’s compon com ponent entss as wel welll as the the for forces ces and loa loads ds app applie lied d from outside the system must be defined. Equations of motion are developed and solved. The results of the analysis analy sis may incl include ude posit positions, ions, velo velocitie cities, s, acce acceleralerations, forces (applied, reactive, and inertial), determination of equilibrium positions, and other computed parameters. These results can be displayed as tables, charts char ts and graphs, overlaid drawing drawingss of posit position ion vs. time,, and animatio time animations ns (see Figs. Figs. 5 and 6). Historically, analyzing analy zing syst systems ems of rigid bodies und undergoi ergoing ng large large-amplitude motions was impractical. Early computerized systems could handle rigid body motions involving large-amplitude rotations and translations in two dimensions. Current systems are capable of kinematic, static, and dynamic analysis of three-dimensional rigid and flexi flexible ble bodie bodiess undergoin undergoing g large displ displacem acements ents and coupled rotations and translations. Mechanism analysis prog Mechanism programs rams can acce accept pt geom geometri etric c data dat a from from FEA progr programs ams,, analyz analyze e the moti motion on of a system, and return appropriate loading and force data
Figure 6. A time series of the positions of a truck truck,, its suspension, and wheels, as it passes over a bump (courtesy of Mechanical Dynamics Inc.)
to the the FE FEA A prog progra ram m for for use use in de dete term rmin inin ing g co commponent pone nt deflection deflections. s. Mechanism Mechanism analysis can also be lin linked ked to contro controll syste system m design design and ana analys lysis. is. The control system’s responses to mechanism behavior can be programmed and fed back to control the mechanism’s reaction. Control system modeling coupled with the ability of engineering workstations to compute and display real-time animation make it possible to simulate mech mechanism anismss such as robot robotss and manufactu manufacturing ring workcells ( see R R OBOTICS OBOTICS). Human–m Huma n–machi achine ne interact interactions ions can be analyzed analyzed using using mechanistic models of the human body. The physical characteristics of the model can be varied according to popu populatio lation n stat statistic isticss for height, weight, age, and gender. Android models are now being used to analyze vehicles, machine tools, and other systems in which a human is an integral part of normal operation.
RAPID PROTOTYPING Rapid prototy prototyping ping allows allows a spe specia ciall machin machine e tool tool to
produce physical prototypes of very complex objects. Although several technologies are now used for this process, stereo lithography was the first and remains popular. In stereo lithography, the prototype is built up in thin layers by slicing the geometric model into cross-sec cross -sections tions and usin using g a compute computer-dri r-driven ven laser to harden layer upon layer of a polymer solution, each in the shape of a particular cross-section of the solid model. These machines can create operational mechanis anisms ms of mo mova vabl ble e pa part rts. s. Th The e mode models ls are are us used ed to evaluate the appearance of the designed part and to verify its fit with other parts and its manufactu manufacturability. rability. New materials such as powdered metals and special plasti pla stics cs contin continue ue to be develo developed ped and all allow ow rap rapid id protot pro totypi yping ng mac machin hines es to produ produce ce more more comple complex, x, Figure 5. This overl overlay ay drawing shows shows the positions of the pilot, seat, canopy, and aircraft during an ejection sequence (courtesy of Mechanical Dynamics Inc.).
accurate accu rate,, and usable prot prototype otypes. s. Some rapid protoprototyping systems are being used as low-quantity manufacturing systems.
COMPUTER-AIDED ENGINEERING
Civil Engineering Applications The analytic tools mentioned previously are also used in the field of civil engineering. However, a few special areas are as exist exist tha thatt do not have have direct direct count counterp erpart artss in mechanic mech anical al engineer engineering. ing. These These inclu include de surv surveying eying,, earthworks, piping, and mapping. In the area of piping plant design, structural engineers use systems similar to those used by mechanical engineers. In roadway design, cut-and-fill and other earth moving movi ng computat computations ions are done in comb combinat ination ion with digital digit al terrain terrain mapping. mapping. Most surv surveyin eying g func functions tions,, such as triangulation and elevation computations, are now computerized, with data being collected in computer form in the field via electronic instruments. Map making and analysis are also largely computerized (see GEOGRAPHICAL INFORMATION S YSTEM).
Electrical/Electronic Engineering Applications Electrical and electronic engineering applications use some structural structural and thermal thermal analy analyses, ses, as desc described ribed above; however, this discipline uses many specialized analyses analy ses for circuit circuit design, VLSI device design (see DIGITAL DESIGN AUTOMATION), and and sim simula ulatio tion. n. A few o of f the major mechanical CAD/CAM systems have some electrica elec tricall CAD capab capabiliti ilities, es, but electron electronic ic CAD (E-CAD) (E-CAD) products are the best choice for electrical/electronic analysis. Many types of electrical analog, digital, and mixed mixe d devices devices can be simu simulated lated.. Simu Simulation lationss allow design engineers to test a design for a circuit board, VLSI chip, or other electrical device before it is committed to manufacturing. In some cases the computer simulation will supplant altogether the building of a prototype prot otype.. Available Available analyses analyses incl include ude gate leve level, l, swit switchching, electrical level, analog-to-digital conversion (q.v.), statistic stat istical al (worst-ca (worst-case), se), and sens sensitiv itivity-b ity-based ased simu simulalations. tions. Con Contro troll sys system temss that that com combin bine e electr electrica icall and mecha me chanic nical al or hydrau hydraulic lic con contro trols ls can also also be sim simula ulated ted.. Other electrical analyses examine transient waveforms and signal signal freque frequency ncy res respon ponse se to deter determin mine e signal signal characte char acterist ristics, ics, such as band bandwidth width and rise and fall times; analyze interference between parallel traces on printed circuit boards; and determine cooling and flow requirements in forced convection systems. E-CAD systems provide libraries of standard electrical components,, including their physical characteristics as components well as performance criteria and specifications. The use of these standard-parts libraries greatly simplifies the process of setting up a simulation.
277
those who do not have access to CAD and CAE systems, to see product designs and simulations of how the products products operate. operate. The increase increased d communic communication ation through thro ughout out product product developme development nt organ organizatio izations ns pro vided by this technology enhances the design process and encourages innovation. As in other areas of computer-aided product development, the World Wide Web (q.v.) continues to play an incre inc reasi asingl ngly y import important ant role role in ana analys lysis. is. Man Many y CAE system systemss can post analys analysis is re resul sults ts as We Web b pages pages in HTML, HTM L, VR VRML ML (see MARKUP L ANGUAGES ANGUAGES), an and d ot othe herr forformats. People without access to the CAE system can view and evaluate these on any computer and get more involved in the product development process. The conceptual design loop is still not completely computerized. At this time, very few computer products are able to couple couple both both the CAD geo geomet metric ric modelmodeling and eng engine ineeri ering ng ana analys lysis is funct function ionss req requir uired ed to develo dev elop p engine engineeri ering ng conce concepts pts rapidl rapidly y and easily. easily. Detailers and drafters remain in the process because it can be too time-consuming for engineers to develop sufficient suffic iently ly detailed detailed mode models ls in curr current ent CAD syst systems. ems. This situation is beginning to change with the introduction duct ion of engi engineer neering ing workb workbench ench systems that are specifically tailored to create solid models easily and feed those models directly into structural and mechanismss analy anism analyses, ses, with the results results imme immediate diately ly availavailable to the designer. This allows engineers to create better developed designs without the costs incurred in the past. Manufacturing Manufactu ring engin engineers eers remain remain outs outside ide the early design des ign loo loop. p. To app apprec reciat iate e the the advant advantage agess of CAE most fully, manufacturing processes and their effects on the product (such as warping caused by machining and stru structur ctural al inte integrit grity y for clamp clamping ing and hand handling) ling) must mu st be ana analyz lyzed ed bef before ore the the design design re reach aches es the the det detail ailed ed layout stage. As mentioned, a few systems are beginning to provide conceptual design tools coupled with sophisticated but simplified analysis tools that can be used by desi design gn engi engineer neerss and manufactu manufacturing ring engin engineers eers,, without the need for the design to be fully detailed. Developments Developm ents in para parallel-p llel-proce rocessin ssing g (q.v.) comp computer uter architectures bring additional power to the engineer’s desktop. Many of the analysis techniques used today (most notably FEA) are good candidates for parallel processing. Faster processing will continue to reduce the design cycle and/or allow additional design iterations, resulting in improved products. More integration between disciplines is to be expected.
Technological Trends
Simulation Simulatio n and visualizat visualization ion prod products ucts are beco becoming ming commonplace. These allow all types of people, notably
In partic particula ular, r, sever several al pro produc ducts ts now com combin bine e th the e electrica elec trical/ele l/electro ctronic nic and the mechanica mechanicall aspects aspects of design and analysis.
278 COMPUTER-AIDED SOFTWARE SOFTWARE ENGINEERING
Bibliography 1980. Timmer, H. G., and Stern, Stern, J. M. ‘‘Com ‘Computati putation on of Global Geometric Properties of Solid Objects,’’ Computer-Aided Objects,’’ Computer-Aided 301–304. Design , 12, 12 , 6 , 301–304. 1982. Huebner, K. H., and Thornton, E. A. The Finite Element Method for Engineers , 2nd Ed. New York: John Wiley. 1984. 198 4. Chace,M. Chace,M. A. ‘‘Metho ‘Methods ds and Exp Experi erienc ence e in Comput Computer er Aid Aided ed Design of Large-displacement Mechanical Systems,’ Systems,’’’ ComputerAided Aide d Analysis Analysis and Optimizat Optimization ion of Mecha Mechanic nical al Syste System m Dynamics , NATO NATO ASI Ser Series ies,, F9. F9. Berlin Berlin:: Spring Springer-Ver er-Verlag. lag. 1984. Erdman, A. G., and, Sandor, G. N. Saddle Mechanism Analysis and Synthesis Vol. 1. Upper River,Design: NJ: Prentice Hall. 1988. Turner, P. R., and Bodner, M. E. ‘‘Optimization and Synthesis for Mechanism Design,’’ Paper MS88-711, Proceedings of the Society of Manufacturing Engineers (October), AUTOFACT 88 Conference and Exposition (October), Chicago, IL. 1989. Sapidis, N., and Perucchio, R. ‘‘Advanced Techniques for Automatic Finite Element Meshing from Solid Models,’’ Computer-Aided Design , 21, 21 , 4, 4 , 248–253. 1991. Zeid, I. I. CAD/CAM Theory and Practice . New York: McGraw-Hill. 1998. Adeli, H., and Kumar, S. Distributed S. Distributed Computer-Aided Engineering . Boca Raton, FL: CRC Press. 1999. Garrett, J. H. Jr., and Rehak, D. R. (eds.) Bridging the Generations: The Future of Computer-Aided Engineering . Pittsburgh, PA: Carnegie Mellon University Press. John MacKrell and Bertram Herzog
COMPUTER-AIDED MANUFACTURING See A A UTOMATION; and COMPUTER -A -AIDED D ESIGN / COMPUTER AIDED M ANUFACTURING.
COMPUTER-AIDED COMPUTER-AID ED SOFTW SOFTWARE ARE ENGINEERING ENGINE ERING (CASE) For articles on related subjects see O BJECT -O -ORIENTED ANALYSIS AND D ESIGN; PROGRAMMING S UPPORT
•
meets business and syste system m requirements
•
is completed within a predictable sche schedule dule
•
is available within budge budgett guidelines
•
allows for easy maintenance and enhanc enhancement ement
The Evolution of CASE The term CASE was was coined in the early 1980s with the automation (q.v.) of manual manual stru structur ctured ed analysis and design desig n tech techniqu niques es and methods. methods. By the mid-1980 mid-1980s, s, code generation linked to analysis and design tools was put under the CASE umbrella. The early CASE tool vendors worked hard to differentiate differentiate their tools from tradition tradi tional al software software developme development nt tools (i.e. editors, editors, compilers, debuggers). As a result, even today many people think that CASE refers to a separate class of tools. This narrow view of CASE resulted in expectations that that were not met by earl early y uses of the tec technolo hnology. gy. For CAS CASE E to be app applie lied d succes successfu sfully lly,, a mor more e glo global bal view view is ne need eded ed.. CASE CASE shou should ld refe referr to any any to tool ol used used to develop, change, and maintain software, and manage software projects. In the late 1980s, this broader view of CASE began to be recognized with the advent of integrated project support environments, integration frameworks, the convergence of existing system maintenance and new development, object-oriented design approaches, and the need to control the software engineering neer ing process. process. In the 199 1990s, 0s, soft software ware engi engineer neering ing became a formal engineering discipline, like mechanical and electrical engineering. Thus all software engineering computer-based tools should be called CASE tools. Starting with the I-CASE software tool conferencess in 199 ence 1994, 4, objec object-or t-orient iented ed developm development, ent, clien client– t– server (q.v.) development, configuration management, framework, repository, and testing tool vendors have
ENVIRONMENTS; SOFTWARE C ONFIGURATION M ANAGEMENT ; SOFTWARE E NGINEERING ; SOFTWARE M AINTENANCE ; SOFTWARE M ETRICS; SOFTWARE P ROJECT M M ANAGEMENT ; OFTWARE EUSABILITY OFTWARE ESTING . S R ; and S T ESTING
been exhibiting together. These vendors are concerned with integration of their tools with the tools of other vendors (interoperability) and the portability (q.v.) of their tools across multiple platforms.
Computer-aided software engineering (CASE) (CASE) encompassess comp passe computer uter-base -based d proc procedur edures, es, tech techniqu niques, es, and tools which can be used to develop, maintain, and reengineer software. CASE is to the software engineer as computer-aided computer-aide d design/computer-aid design/computer-aided ed manufactu manufacturring (CAD/CAM) (q.v.) is to the mechanical engineer and computer-aided computer-aided electrical engineering (CAEE) is to the electrical engineer. Although the variety of technological alternatives can be bewildering, the concepts of CASE provide a common-sense approach to engineering quality software more productively.
Thus all the dimensi dimensions ons of soft software ware enginee engineering ring are coming together to form integrated environments. The four key components comprising these environments analysis is and desig design n dimension, are are show shown n inFig.1. Th The e analys sometime some timess referred referred to as front-end or upper CASE, includes inclu des tools and tech techniqu niques es for plan planning ning systems, defining requirements, and designing systems. Starting with system functions, workflow and business process analysis tools are used to understand and define requirements, rules, and processes. Structured or object-
The application of CASE is intended to allow teams of software engineers to produce software that:
ori orien ented ted analys analysis is and desig design n too tools ls are then then used used to define and design the software systems which support those system functions.