COMPUTER INTEGRATED TECHNOLOGIES AND KNOWLEDGE MANAGEMENT:DESIGN TECHNIQUES, TOOLS, AND COMPONENTS

DESIGN TECHNIQUES, TOOLS, AND COMPONENTS

Like traditional design, 2D geometrical processing takes place through the creation of object contours in several 2D views. The views are then further detailed to show sections, dimensions, and other important drawing information. As opposed to 3D CAD systems, 2D CAD design occurs through conventional methods using several views. A complete and discrepancy-free geometrical description, comparable to that provided by 3D CAD systems, is not possible. Therefore, the virtual product can only be displayed to a limited extent. Two-dimensional pictures still have necessary applications within 3D CAD systems for hand-sketched items and the creation of drawing derivations.

Despite their principal limitation compared to 3D CAD systems, 2D systems are widely used. A 3D object representation is not always necessary, for example, when only conventional production drawings are required and no further computer aided processing is necessary. In addition, a decisive factor for their widespread use is ease of handling and availability using relatively inexpensive PCs.

The main areas of 2D design are the single-component design with primitives geometry, sheet metal, and electrical CAD design for printed circuit boards.

Depending on the internal data structure of the computer, classical drawing-oriented and para- metrical design-oriented 2D CAD systems can be differentiated. This distinction has an essential inﬂuence on the fundamental processing principles and methodical procedures followed during the design phase. The drawing-oriented systems are then implemented efﬁciently in the detailing phases while new-generation parametrical systems support the development in early concept phases. With efﬁcient equation algorithms running in the background, extensive calculations and parametrical de- pendencies for simulation purposes can be applied.

The essential goal in using design-oriented, 2D CAD systems is creation of manufacturing plans and documentation. From the exact idea of the topology and dimensions of the technical entity to the input of the geometry and ending with the detailing of the technical drawing and preparation of the parts list, this computer-supported process corresponds to a large degree with conventional strat- egies.

The technical drawings are created with the aid of geometrical elements available within the system. The following geometrical elements can be used: points, straight lines, parallel lines, tangents, circles, arcs, ﬁllets, ellipses, and elliptical arcs such as splines. Furthermore, equidistant lines may be created. Editing commands such as paste, delete, join, extend, trim, and orientation operations

such as translation and rotation, exist for element manipulation purposes. Further design clariﬁcation and simpliﬁcation is provided by the help functions of scaling, duplicating, locating, and mirroring on symmetrical axes. Detail work may be simpliﬁed with the use of zoom functions, and snap points aid in dimensioning and positioning of geometrical elements. The technical drawings are then com- pleted with the use of hatching, dimensioning, and labeling.

The following strategies are recommended for the completion of drawings using 2D CAD systems:

• Do default settings for drawing parameters such as scale ratios, line characteristics and hatching, font, and measurement standards.

• Call up and blend in predeﬁned frames with text and form ﬁelds.

• Use construction lines for added support.

• Create contour and surface in different views.

• Apply eventual cutouts and hatching

• Set measurements, tolerances, etc.

• Transfer generated objects to predeﬁned printing or storage mediums.

An example of a complex assembly drawing is presented in Figure 1.

With the use of layers, drawing information of all types may be distributed over any desired

number of named and editable layers. The layers simplify the drawing process because momentarily

irrelevant information can be easily blended out. Additional text in different languages or measure-

ment systems or norms can also be called up and implemented.

Grouping techniques allow for the summarizing of geometrical elements through various selection

processes and in turn ease further working processes.

The majority of systems provide function libraries for the simpliﬁcation of drawing preparation

and the efﬁcient use of results. The following areas may be simpliﬁed with the use of function

libraries:

• Creation and manipulation of geometrical elements

• Geometrical calculations

• Access to saved elements using search criterion

• Creation of groups and layers

• Utilization of drawing elements such as dimensioning, hatching, and text

• Use of standard dialogs for value, point, or font input

Geometry macros are internal computer representations of dimension and form-variable compo- nent geometries of any complexity. The macros are easily called up from a database and copied by the user. The user may then easily insert the copied macro into the current object representations. The advantages of the geometry macros are seen especially when working with standard and repeat parts.

A spatial representation is desirable for a variety of tasks where primarily 2D representations are utilized. Therefore, in many systems an existing function is available for isometric ﬁgures that provide an impression of the object’s spatial arrangement without the need for a 3D model. This approach is sufﬁcient for the preparation of spare parts catalogs, operating manuals, and other similar tasks. Prismatic bodies and simpliﬁed symmetric bodies may also be processed. For the use of this function it is necessary to present the body in front and side views as well as plan and rear views where required. After the generation of the isometric view, the body must be completed by blending out hidden lines. This happens interactively because the system actually does not have any information available to do this automatically. So-called 21⁄2 systems generate surface-oriented models from 2D planes through translation or rotation of surface copies. The main application for these models lies in NC programming and documentation activities.

To reduce redundant working steps of geometrical modiﬁcations, the dimensioning of the part must be ﬁnished before the computer-supported work phases. This approach, which underlies many CAD systems today, is connected to the software’s technical structure. In geometry-oriented dimen- sioning, the measurement alone is not inferred but rather the totality of geometrical design of the entity. The supporting role played by the system for interactive functioning is limited to automatic or semiautomatic dimension generation. The dimensional values are derived from the geometry spec- iﬁed by the designer. The designer is forced to delete and redo the geometry partially or completely when he or she wishes to make changes. This creates a substantial redundancy in the work cycle for the project.

The principal procedures of dimension-oriented modeling in parametric 2D CAD systems often correspond to conventional 2D systems. For the generation of geometrical contours, the previously mentioned functions, such as lines, circles, tangents, trimming, move, and halving, are used. The essential difference is the possibility for free concentration on the constructive aspects of the design.

In parametrical systems, on the other hand, the measurements are not ﬁxed but rather are displayed as variable parameters. Associative relationships are built between the geometry and measurements, creating a system of equations. Should the geometry require modiﬁcation, the parameters may be varied simply by writing over the old dimensions directly on the drawing. This in turn changes the geometrical form of the object to the newly desired values. The user can directly inﬂuence the object using a mouse, for example, by clicking on a point and dragging it to another location, where the system then calculates the new measurement. The changes cause a veriﬁcation of the geometrical relationships as well as the consistency of the existing dimension in the system.

Another advantageous element of the parametrical system is the ability to deﬁne the above-mentioned constraints, such as parallel, tangential, horizontal, vertical, and symmetrical constraints.

Constraints are geometrical relationships ﬁxed by formula. They specify the geometrical shape of the object and the relevant system of equations necessary for geometry description. Thus, a tangent remains a tangent even when the respective arc’s dimension or position is changed. With the con- straints, further relationships can be established, such as that two views of an object may be linked to one another so that changes in one view occur the related view. This makes the realization of form variations fairly simple.

Modern CAD systems include effective sketching mechanisms for the support of the conceptual phase. For the concept design, a mouse- or pen-controlled sketching mechanism allows the user to portray hand-drawn geometries relatively quickly on the computer. Features aiding orientation, such as snap-point grids, and simplifying performing the alignment and correction of straight, curved, orthogonal, parallel, and tangential elements within predetermined tolerances. This relieves the user of time-consuming and mistake-prone calculations of coordinates, angles, or points of intersection for the input of geometrical elements. When specifying shape, the parametrical system characteristics come into play, serving to keep the design intent. The design intent containing the relative positioning is embodied by constraints that deﬁne position and measurement relationships. Component variations are easily realizable through changes of parameter values and partially with the integration of calculating table programs in which table variations may be implemented.

The existing equation system may be used for the simulation of kinematic relations. Using a predetermined, iterative variation of a particular parameter and the calculation of the dependent parameter, such as height, simple kinematic analysis of a geometrical model is possible. This simpliﬁes packaging analysis, for example.

3D Design

Surface Modeling The geometrical form of a physical object can be internally repre- sented completely and clearly using 3D CAD systems, is contrast to 2D CAD systems. Therefore, the user has the ability to view and present reality-like, complex entities. Thus, 3D CAD systems provide the basis for the representation of virtual products. Through 3D modeling and uniform internal computer representation, a broader application range results so that the complete product creation process can be digitally supported. This starts with the design phase and goes through the detailing, calculation, and drawing preparation phases and on to the production. Because of the complete and discrepancy-free internal computer representation, details such as sectional drawings may be derived automatically. Also, further operations such as ﬁnite element analysis and NC pro- grams are better supported. The creation of physical prototypes can be avoided by using improved simulation characteristics. This in turn reduces the prototype creation phase.

Due to the required complete description of the geometry, a greater design expenditure results. Because of the greater expenditure, special requirements for user-friendliness of component geometry generation of components and their links to assemblies and entire products are necessary.

3D CAD systems are classiﬁed according to their internal computer representations as follows (Gra¨tz 1989):

• Edge-oriented models (wire frames)

• Surface-oriented models

• Volume-oriented models

The functionality and modeling strategy applied depend on the various internal computer represen- tations.

Edge-oriented models are the simplest form of 3D models. Objects are described using end points and the connections made between these points. As in 2D CAD systems, various basic elements, such as points, straight lines, circles and circular elements, ellipses, and free forming, are available to the user and may be deﬁned as desired. Transformation, rotation, mirroring, and scaling possibil- ities resemble those of 2D processing and are related to the 3D space.

Neither surfaces nor volumes are recognizable in wire frame models. Therefore, it is not possible to make a distinction between the inside and outside of the object, and section cuts of the object cannot be made. Furthermore, quantiﬁable links from simple geometrical objects to complex com- ponents are not realizable. This is a substantial disadvantage during the practical design work. An arbitrary perspective is possible, but in general a clear view is not provided. Among other things, the blending out of hidden lines is required for a graphical representation. Because of the lack of infor- mation describing the object’s surfaces, this cannot take place automatically but must be carried out expensively and interactively. Models of medium complexity are already difﬁcult to prepare in this manner. The modeling and drawing preparation are inadequately supported. Another problem is the consistency of the geometry created. The requirements for the application of a 3D model are indeed given, but because of the low information content the model does not guarantee a mistake-free wire frame presentation (Gra¨tz 1989).

Surface-oriented systems are able to generate objects, known as free-form surfaces, whose surfaces are made up of numerous curved and analytically indescribable surfaces. One feature visible in an internal computer representation of free-form surfaces is their interpolated or approximated nature. Therefore, various processes have been developed, such as the Be´zier approximation, Coon’s surfaces, the NURBS representations, and the B-spline interpolation (see Section 4.1).

As with all other models, the computer internal representation of an object implies characteristic methods and techniques used in generating free-form surfaces. In general, basis points and boundary conditions are required for the surface description. Free-form surfaces, such as the one portrayed in Figure 2, can be generated using a variety of techniques. A few examples for the generation of free- form surfaces will be described later.

In addition to the simple color-shaded representations, photorealistic model representations may be used for the review and examination of designs. The models are portrayed and used in a reality- like environment and are supported by key hardware components. This allows for the consideration of various environmental conditions such as type and position of the light source, the observer’s position and perspective, weather conditions, and surface characteristics such as texture and reﬂec- tivity. Also, animation of the object during design is possible for the review of movement character- istics. Through these visualization methods it is possible to assess a product’s design before prototype production begins and aid the preparation of sales brochures and other documents.

A characteristic feature and major disadvantage of surface models is the fact that the surfaces are not automatically correlated to form a body. This has important consequences for the designer. It results in a distinct modeling technique for the user that offers only limited functionality with regard to realization of the design tasks. For example, from the outset it is not deﬁned whether a surface should face the inside or the outside of the object. Therefore, the designer must be careful to ensure a closed and complete model during design phases. Intersections are not recognizable by the system

because of missing information describing the solid during cutting operations. The intersections are represented only by contour lines that are interactively hatched.

Surface models are applied when it is necessary to generate a portrait of free-formed objects. Such models are used in the automobile, shipbuilding, and consumer goods industries, for example. The surface formations of the model, which are distinguished into functional and esthetic surfaces, can be used as ground-level data for various applications employed later on. This includes later applications such as the generation of NC data from multiaxis processing. Another area of application is in FEM analysis, where the surface structure is segmented into a ﬁnite element structure that is subsequent to extensive mechanical or thermal examinations. With surface models, movement sim- ulations and collision analyses can be carried out. For example, mechanical components dependent on each other can be reviewed in extreme kinematic situations for geometrical overlapping and failure detection.

Solid Modeling A very important class of 3D modelers is generated by the volume systems. The appropriate internal computer representation is broken down into either boundary rep- resentation models (B-reps) or Constructive Solid Models (CSGs). The representation depends on the basic volume modeling principles and the deciding differences between the 3D CAD system classes.

Volume modelers, also known as solid modelers, can present geometrical objects in a clear, consistent, concise manner. This provides an important advantage over previously presented modeling techniques. Because each valid operation quasi-implicitly leads to a valid model, the creativity of the user is supported and the costly veriﬁcation of geometric consistency is avoided.

The main goal of volume modeling is the provision of fundamental data for virtual products. This data includes not only the geometry but also all information gathered throughout the development process, which is then collected and stored in the form of an integrated product model. The infor- mation allows for a broad application base and is available for later product-related processes. This includes the preparation of drawings, which consumes a large portion of the product development time. The generation of various views takes place directly, using the volume model, whereby the formation of sections and the blending out of hidden lines occurs semiautomatically.

Normally, volume models are created using a combination of techniques. The strengths of the 3D volume modeling systems lie in the fact that they include not only the complete spectrum of well- known modeling techniques for 2D and 3D wire frames and 3D surface modeling but also the new, volume-oriented modeling techniques. These techniques contribute signiﬁcantly to the reduction of costly input efforts for geometrically complex objects.

The volume model can be built up by linking basic solid bodies. Therefore, another way of thinking is required in order to describe a body. The body must be segmented into basic bodies or provided by the system as basic elements (Spur and Krause 1984). Each volume system comes with a number of predeﬁned simple geometrical objects that are automatically generated. The objects are generated by a few descriptive system parameters. Spheres, cubes, cones, truncated cones, cylinders, rings, and tetrahedrons are examples of the basic objects included in the system.

Many complex components can be formed from a combination of positioning and dimensioning of the basic elements.

Complex components are created in a step-by-step manner using the following connecting oper- ations:

• Additive connection (uniﬁcation)

• Substractive connection (differential)

• Section connection (average)

• Complementing

These connections are known as Boolean or set-theoretical operations. For object generation using Boolean operations, the user positions two solids in space that touch or intersect one another. After the appropriate functions are called up (uniﬁcation, average, complementing, or differential), all further steps are carried out by the system automatically and then present the geometrical object.

The sequence of the operations when the set-theoretical operations are applied is of decisive importance. Some CAD systems work with terms used for manufacturing techniques in order to provide the designer with intuitive meanings. This allows for the simple modeling of drilling, milling, pockets, or rounding and chamfering. This results because each process can be represented by a geometrical object and further deﬁned as a tool. The tool is then deﬁned in Boolean form and the operation to perform the removal of material from the workpiece is carried out. To correct the resulting geometry, it is necessary that the system back up the complete product creation history. This is required to make a step by step pattern that can be followed to return the set-theoretical operations to their previous forms.

Set-theoretical operations provide an implementation advantage where complex model deﬁnitions must be created with relatively few input commands. These deﬁnitions are, if even realizable, ex- tremely costly and time consuming by conventional methods. The disadvantage is that a relatively high conceptual capability is required of the user.

Solid models with a number of free-form surfaces are realized using surface-oriented modeling techniques that correspond to the surface model presented. With one of these techniques, sweeping, one attains a 3D body through an intermediate step when creating a 2D contour. The 2D contour is expanded to three dimensions along a set curve. The desired contours are generated with known 2D CAD system functions and are called up as bases for sweep operations.

In rotational sweeping, objects are generated through the rotation of surfaces, as well as closed or open, but restricted, contour lines around a predeﬁned axis. The axis may not cut the surface or contour. A prismatic body originates from the expansion of a closed contour line. Shells or full bodies can be generated with sweep operations. With sweep-generated shells, movement analysis and as- sembly inspection can occur. The analysis takes place by the bodies’ movement along a curve in space.

The modeling strategy implemented depends heavily on the actual task and the basic functionality offered by the CAD system. This means that the strategy implemented cannot be arbitrarily speciﬁed, as most CAD systems have a variety of alternative approaches within the system. For the user, it is important that a CAD system offer as many modeling methods as possible. Modeling, such as surface and solid modeling, is important, and the ability to combine and implement these methods in geo- metrical models is advantageous to the user. Because of the numerous modeling possibilities offered, the CAD system should have as few restrictions as possible in order to allow the user to develop a broad modeling strategy. Figure 3 shows a surface-oriented model of a bumper that was created using a simple sweep operation of a spline along a control curve. Using a variable design history, the air vent may be positioned anywhere on the bumper’s surface without time-consuming design changes during modeling. The cylindrical solid model that cuts and blends into the surface of the bumper demonstrates that no difference exists between a solid and surface model representation. Both models may be hybridized within components. A CAD system supporting hybridized modeling creates a major advantage because the designer realizes a greater freedom of design.

Parametrical Design

Another development in CAD technology is parametrical modeling (Frei 1993). The function of parametrical systems displays basic procedure differences when compared to conventional CAD sys- tems. The principal modeling differences were explained in the section on parametrical 2D CAD systems.

In parametrical model descriptions, the model can be varied in other areas through the selection and characterization of additional constraints.

Boundary conditions can be set at the beginning of the design phase without ﬁxing the ﬁnal form of the component. Therefore, generation of the geometry takes place through exact numerical input and sketching. The designer can enter the initial designs, such as contour, using a 2D or 3D sketcher. In this manner, geometrical characteristics, such as parallel or perpendicular properties, that lie within

set tolerances are automatically and correctly recognized. The characteristics may also be set through predetermined constraints. Subsequently, the user can develop the initial dimensions using operations that essentially resemble those of conventional measurement determination. Through associativity, the adaptation of the model presented takes places automatically.

Beginning with the ﬁrst contour sketches, the designer attains the desired solid object through further modeling using operations such as sweeping. The design history is of great importance at this point of the development process. The history can subsequently be modiﬁed, inﬂuencing the geometry of the model. Because of the associative relationship between the dimensioning and the solid model, the geometry of a component is easily varied in every design phase. Based on the parametrical method implemented, the designer can deﬁne conditions for parallel, tangential, or per- pendicularity characteristics. Also, complex parametrical dependencies from the system can be solved simultaneously using nonlinear equations.

Because of the parametrical relationships created in the background of the process, the user is in a position to modify the dimensions and topology of the geometry later in the project. Costly re- working of the model due to changes of the boundary conditions is avoided.

Complete process chains may be built up associatively. Downstream activities, including drawing creation and detailing, make up a large percentage of the design time. Here the generation of various views directly from the existing solid model may take place because the formation of sections and the blending out of hidden lines occurs automatically. The main advantage of this approach over previous CAD systems is due to the above-mentioned preservation of the bidirectional relationship between the solid model and the drawing derived from the model. This associative connection allows the designer to take geometrical modiﬁcations directly from a drawing, even if the designer has progressed substantiality into the detailing phase. Conversely, changes to the solid model are auto- matically reﬂected in the drawing.

The effect and the functionality of the parametric concepts do not end with the completion of the geometry. Rather, they are encompassed in all areas of product development. Solid modeling, cal- culations, drawing preparation, NC programming, simulation, and documentation are connected to- gether in an associative chain so that modiﬁcations made in one area automatically take place in the other areas.

For the designer, it is important that both parametric and nonparametric geometry can be combined and implemented during the development of a component. Therefore, complex components must not be completely described parametrically. The designer decides in which areas it is advantageous to present the geometry parametrically.

Feature Technology

Most features in commercial CAD systems are built upon a parametrical modeler. The ability to deﬁne one’s own standard design elements also exists. These elements can be tailored to the designer’s way of thinking and style of expression. The elements are then stored in a library and easily selected for implementation in the design environment. Additional attributes such as quantity, size, and geo- metrical status can be predetermined to create logical relationships. These constraints are then con- sidered and maintained with every change made to an element. An example of a design feature is a rounding ﬁllet. The ﬁllet is selected from the features library and its parameters modiﬁed to suit the required design constraints.

The library, containing standard elements, can be tailored to the user’s requirements for production and form elements. This ensures consistency between product development and the ability to man- ufacture the product. Therefore, only standard elements that match those of available production capabilities are made available. Feature deﬁnition, in the sense of concurrent engineering, is an activity in which all aspects of product creation in design, production, and further product life phases are considered.

In most of today’s commercially available CAD systems, the term feature is used mainly in terms of form. Simple parametric and positional geometries are made available to the user. Usually features are used for complicated changes to basic design elements such as holes, recesses, and notches. Thus, features are seen as elements that simplify the modeling process, not as elements that increase in- formation content. With the aid of feature modeling systems, user-deﬁned or company-speciﬁc fea- tures can be stored in their respective libraries. With features, product design capabilities should be expanded from a pure geometric modeling standpoint to a complete product modeling standpoint. The designer can implement fully described product parts in his or her product model. Through the implementation of production features, the ability to manufacture the product is implicitly considered. A shorter development process with higher quality is then achieved.

By providing generated data for subsequent applications, features represent an integration poten- tial.

Assembly Technology

Complex products are made up of numerous components and assemblies. The frequently required design-in-context belongs to the area of assembly modeling. Thereby, individual parts that are relevant to the design of the component are displayed on the screen. For instance, with the support of EDM functionality, the parts can be selected from a product structure tree and subsequently displayed. The part to be designed is then created in the context already deﬁned. If the assembly is then worked on by various coworkers, the methods for shared design are employed. For example, changes to a component can be passed on to coworkers working on related and adjacent pieces through system- supported information exchange.

An example of a complex assembly in which shared design processes are carried out is the engine in Figure 4. The goal is to be able to build and then computer analyze complex products. The assembly modeling will be supported by extensive analysis and representation methods such as interference checks as well as exploded views and section views.

Besides simple geometrical positioning of the component, assembly modeling in the future will require more freedom of modeling independent of the various components. Exploiting the association between components permits changes to be made in an assembly even late in the design phase, where normally, because of cost and time, these modiﬁcations must be left out. Because the amount of data input for assemblies and the processing time required to view an assembly on the screen are often problematic, many systems allow a simpliﬁed representation of components in an assembly, where suppression of displaying details is possible. But often the characteristics of simpliﬁed representation are to be deﬁned by the user. It is also possible for the CAD system to determine the precision with which the component is displayed, depending on the model section considered. Zooming in on the desired location then provides a representation of the details.

A representation of the assembly structure is helpful for the modeling of the assembly. Besides the improved overview, changes to the assembly can be carried out within the structure tree and parts lists are also easily prepared.

The combination of assemblies as a digital mock-up enables control over the entire structure during the design process. Thereby, interference checks and other assembly inspections are carried out using features that make up the so-called packaging analyses of the software. For the check of collision, voxel-based representations are suitable as well. Importing and portraying assemblies from other CAD systems is also advantageous. This is possible with the creation of adequate interfaces.

An important functionality in assembly modeling is the deﬁnition of geometric and functional tolerances that are supported by the respective CAD module (Figure 5). Thereby, the geometrical and functional tolerance deﬁnitions are carried out based on assembly speciﬁcations and international standards. Special analysis functions are utilized for the control of complex assemblies and products.

For example, the ﬂy-through function allows the user to navigate in real time through complex assemblies. This enables visual inspections, which among other things, help in the recognition of component interference.

The design of complex assemblies requires sharing of the design processes through teamwork. This shared group work is supported by software that enables conferences over the Internet / intranet in which text, audio, and video information may be exchanged.

Further Technologies

Further applications in the process chain of product development can be carried out based on the 3D models created in early phases. For example, most commercial CAD systems are composed of mod- ules and are implemented for the realization of a geometric model and other development-oriented solutions. Because the CAD systems are compatible with one another, an unhindered exchange of data is possible right from the beginning of the design through production stages. The systems must meet the requirements for continuity and openness, which are indispensable for the development of complex products. Continuity means that the product data are entered only once. They are saved in a central location and are available for use in carrying out other tasks without extensive conversion being required between the various forms of data. Openness means that, besides the ability to process data on various hardware platforms, it is possible to connect with other program systems within the same company and to external sites such as suppliers and other service providers. Further differen- tiation results from homogeneous and heterogeneous system environments. Using a common database for the application modules, for example in STEP format, means that conversion or interface problems between the individual applications are avoided. External application programs and data exchange between the purchaser and the supplier are simpliﬁed by the use of data exchange formats such as IGES, SET, and VDAFS. A uniform user interface improves the usability and comfort of the user.

The application modules can be related to the various tasks within the development process. In the various phases of the virtual product development, different modules come into play depending on the task involved. The following principal tasks are presented:

• Drawing preparation: Drawings can be derived from the solid model and displayed in various views and measurements (Figure 6). This includes the representation of details as well as com- ponent arrangements. For higher work efﬁciency, it should be possible to switch between the 2D drawing mode and the 3D modeling mode at any time without suffering from conversion time and problems. A further simpliﬁcation of the work results if the drawing and measurement routines provide and support the drawings in standardized formats, for example.

• Parts list generator: The preparation of a parts list is important for product development, pro- curement, and manufacturing. This list is automatically generated from the CAD component layout drawing. The parts list can then be exchanged through interfaces with PPS systems.

• Analysis and simulation applications: Analysis and simulation components enable the charac- teristics of a product to be attained and optimized earlier in the development phase. The costly and time-consuming prototype production and product development are thereby reduced to a minimum.

The strength characteristics can be calculated with ﬁnite element methods (FEMs). Besides stress– strain analyses, thermodynamic and ﬂuid dynamic tests can be carried out. Preprocessors enable automated net generation based on the initial geometric model. Most CAD systems have interfaces for common FEM programs (e.g., NASTRAN, ANSYS) or have their own equation solver for FEM tests. Based on the substantial data resulting from an FEM test, it is necessary for postprocessing of results. The results are then portrayed in deformation plots, color-coded presentations of the stress– strain process, or animations of the deformation.

Material constants, such as density, can be deﬁned for the components of the solid model. Then other characteristics, such as volume, mass, coordinates of the center of gravity, and moment of inertia can be determined. These data are essential for a dynamic simulation of the product. For kinematic and dynamic simulations, inelastic solid elements with corresponding dynamic character- istics are derived from the solid model (Figure 7). Various inelastic solid elements can be coupled together from a library. The complete system is built up of other elements such as a spring and damper and then undergoes a loading of external forces. The calculation of the system dynamics is usually carried out using a commercial dynamics package. A postprocessor prepares a presentation of the results for display on a screen. An animation of the system can then be viewed.

• Piping: Piping can be laid within the 3D design using conduit application modules, as shown in Figure 8. Design parameters such as length, angle, and radius can be controlled. Intersections or overlapping of the conduit with adjacent assemblies can also be determined and repaired. The design process is supported through the use of libraries containing standard piping and accessories. The parts lists are automatically derived from the piping design.

• Weld design: Weld design can also be supported by using appropriate application modules (Figure 9). Together, designers and production engineers can determine the required joining techniques and conditions for the assemblies. Thereby, the weld type and other weld parameters such as weld spacing and electrodes are chosen for the material characteristics of the components to be joined. Process information such as weld length and cost and time requirements can be derived from the weld model.

• Integration of ECAD design: The design of printed circuit boards and cable layouts are typical application examples of 2D designs that are related to the surrounding 3D components. With the design of printed circuit boards, layout information, such as board size and usable and unusable area, can be exchanged and efﬁciently stored between and within the MCAD and ECAD systems. For the design of cable layouts (Figure 10), methods similar to those used for conduit design are implemented. This simpliﬁes the placement of electrical conductors within assemblies.

• Integration of process planning: Product data can be taken directly from production planning. The machine tool design and planning can take place parallel to the design of the product. NC programs and control data can be derived from the ﬁnished 3D description of the geometry, although the model requires supplemental information for mountings and other devices. The tool paths are determined under consideration of the additional technical information of the different processes, such as drilling, milling, turning, eroding, and cutting. The generation of NC data in diverse formats, such as example COMPACT II, CLDATA, and APT, is carried out using a postprocessor. The NC programs derived from an optimized design can be checked by simulation.

In addition to the three- and ﬁve-axis milling machines used for conventional production tasks, rapid prototyping processes are applied for quick design veriﬁcation. For example, the process of stereo lithography can be applied. Rapid prototyping reduces the period of devel- opment and allows for the quick generation of physical prototypes for product review and evaluation purposes. A variety of materials, from plastics to metals, are used in rapid prototyp- ing, enabling the prototypes to be presented in the most frequently used materials so that material characteristics can be easily reviewed (Gebhardt and Pﬂug 1995). The control data for the laser stereo lithography are generated from the geometry data.

When molds for injection molding are made, die geometry is taken into consideration be- cause of the possibility of material shrinkage. Rheological calculations are carried out using preprocessors (e.g., Moldﬂow, Cadmould 3D). The set-up of the tools takes place using catalogs provided by the machine manufacturer. Subsequently, NC data are generated for the part shape and machining plates.

With complete product models, simulation and robot programming can take place for the various manufacturing processes. The programming and simulation take place ofﬂine. The pro- grams generated are then sent on to the robots for execution of the process.

Only through comprehensive integration of the complete virtual product development process, with the many different working steps, can the existing potential for development time, costs, and overall productivity be optimized (VDI-EKV 1992, Krause 1992).

CAD Interfaces

General Explanations

Despite the lack of a clear deﬁnition for the term interface, its use has become quite frequent. In general, an interface can be deﬁned as a link forming a common boundary between integrating systems. It is a system of conditions, rules, and agreements that deﬁnes the terms of information exchange between communicating systems or system components (Anderl 1993).

CAD interfaces interconnect internal CAD system components and provide a link to other soft- ware and hardware components. They connect internal CAD system components such as geometrical modeling, object representation, and realization of calculation operations to a CAD database and standardized CAD ﬁles.

CAD interfaces can be classed as device interfaces (hardware interfaces) and software interfaces (Grabowski and Anderl 1990) or, alternatively, as internal and external interfaces. Internal interfaces provide links within the application environment, enabling the exchange of information between other system programs or ensuring undisturbed communication with the user. External interfaces require a common or uniform representation of the information to be exchanged. They transmit product data, by means of graphical peripheries, to the user. The standardization of external interfaces is of special importance for the transmission of product data. Because of the exchange of data between various CAD systems, uniform data formats are required.

In Figure 11, CAD components and interfaces are displayed.

From an application in use, access to the product data and methods of a CAD system is provided through application interfaces. In the course of using a CAD system through a standardized interface, the integration of such an external application interface becomes increasingly important. Standardized application interfaces allow access to the methods employed by the various CAD systems.

Product data represent all data such as geometry, topology, technological information and organizational data to deﬁne the product. This data is generated during the design process, enhanced by the application and then internally mapped.

Standardization of CAD Interfaces

The primary objectives for the standardization of interfaces are:

• The uniﬁcation of interfaces

• The prevention of dependencies between system producer and user

• The prevention of repeated working of redundant data

The standardization of interfaces leads to the following advantages (Stanek 1989):

• The possibility of an appropriate combination of system components

• Selection between various systems

• Exchange of individual components

• Expandability to include new components

• Coupling of various systems and applications

• Greater freedom in the combining of hard and software products 2.2.2.1. Interfaces for Product Data Exchange The description of product data over the entire product life cycle can be effectively improved through the use of computer-supported systems. For this reason, uniform, system-neutral data models for the exchange and archiving of product data are desired. These enable the exchange of data within CAD systems and other computer-aided applica- tions.

The internal computer models must be converted into one another through so-called pre- and postprocessor conversion programs. To minimize the costs of implementing such processors, a stan- dardized conversion model is applied. The conversion models standardized to date are still of limited use because they can only display extracts of an integrated product model. Interfaces such as

• initial graphics exchange speciﬁcation (IGES)

• standard d’echange et de transfert (SET)

• Verband der Automobilindustrie-Fla¨chenSchnittstelle (VDA-FS)

have been designed and to some extent nationally standardized for geometry data exchange, mainly in the area of mechanical design (Anderl 1989).

IGES is recognized as the ﬁrst standardized format for product-deﬁned data to be applied indus- trially (Anderl 1993; Grabowski and Anderl 1990; Rainer 1992). The main focus is the transfer of design data. IGES is for the mapping of:

• 2D line models

• 3D wire models

• 3D dimensional surface models

• 3D solid models

• Presentation models for technical drawings

Considering the further demands on CAD / CAM systems, the integration of additional data in IGES format has been realized. Examples are the data for FEM, factory planning, and electronic / electrical applications.

The interface standard VDA-FS was developed by the German Automotive Industry Association for the transfer of CAD free-form surface data. VDA-FS is standardized and described in Standard DIN 66301 and has proven efﬁcient for use in many areas, but its efﬁciency has been demonstrated principally in the German automotive industry (Grabowski and Glatz 1986; Nowacki 1987; Scheder 1991).

The STEP product model (Standard for the Exchange of Product Model Data) offers the only standardized possibility for efﬁciently forming the product data exchange and system integration in the CAx world (Wagner and Bahe 1994).

With the development of ISO 10303 (Product Data Representation and Exchange), also called STEP, the objective is to standardize a worldwide accepted reference model for the transfer, storage and archiving, and processing of all data needed for the entire product life cycle (Anderl 1993; Du¨ring and Dupont 1993; Krause et al. 1994; McKay et al. 1994).

STEP can be seen as a toolbox for describing application-oriented product information models using basic elements, so-called integrated resources, while considering previously deﬁned rules and standardized methods (Holland and Machner 1995).

ISO 10303 can be subdivided into:

• Generic and application-oriented information models

• Application protocols

• Methods for speciﬁcation and implementation

• Concepts for examination and testing

Independent application speciﬁcations are referred to as generic models. An overview of the series is as follows (Figure 12):

• Series 0: fundamentals, principles

• Series 10: description or speciﬁcation methods

• Series 20: implementation methods

• Series 30: test methods and criterion

• Series 40: application-independent base models

• Series 100: application-dependent base models

• Series 200: application protocols

• Series 300: abstract test methods

• Series 500: application-speciﬁc, interpreted design

The core of consists of information models in which the representational form of the product data is deﬁned. Information models can be broken down into three categories: generic resources, application-related resources, and application protocols. Similar to a toolbox system, the generic resource models deﬁne application-independent base elements that may be used in application-speciﬁc resources or directly in application protocols. An example is geometrical information, which is re- quired in most application models. Application-related resources are especially tailored to the de- mands of speciﬁc branches but also serve for the deﬁnition of application protocols. Application protocols form implementable parts for STEP. They deﬁne a section of the information structure that is needed to support a speciﬁed application.

The description methods used for deﬁning the information structure, validation, and test methods as well as implementation methods are standardized. The object-oriented modeling language EX- PRESS as well as its graphical form EXPRESS-G are implemented for the presentation of the in- formation structures (Scholz-Reiter 1991). Validation and test methods deﬁne the speciﬁcations with which the STEP processors are examined. The following elements are included in partial models of the generic resources:

• Base model

• Geometric and topological model

• Tolerance model

• Material model

• Product structure model

• Representation model and

• Process model

The signiﬁcance of STEP goes much further than the exchange and archiving of product models. The innovative character of STEP development sets new standards for CAD systems. The functional requirements for new systems stem from partial models such as tolerance, form / feature, or PSCM models. In certain circumstances, STEP may be considered a reference model for the internal com- puter representation of CAD systems.

Engineering Data Management

Engineering data management (EDM, also an abbreviation for electronic document management or enterprise data management) is data modeling of a complete enterprise. In other words, the data modeling goes beyond the intrinsic product data. A common synonym for EDM is ‘‘product data management’’ (PDM), in which emphasis is put on the handling of product-related engineering data. Engineering data management was coined as a broader term. All deﬁnitions combined provide the perception that EDM is the management and support of all information within an enterprise at any point in time (Ploenzke 1997).

For the support of data modeling within a company, software systems, so-called EDM systems, exist. With these systems, various tasks are carried out depending on the product, company, and level of integration, such as the management of (Kiesewetter 1997):

• Drawing data

• CAD model data

• Parts lists

• Standard libraries

• Project data

• NC data

• Software

• Manufacturing plans

• Tools and equipment production facilities

In addition, EDM systems serve for classiﬁcation and object parameter management. Generally, EDM systems provide a set of speciﬁc functions for the modeling of product and process data. Because of their speciﬁc origins, the various EDM systems focus on a variety of main functions and strengths, but all provide a basic functionality for document management as well as functions for the support of change, version, and release management. The complete functionality of an EDM system as a central information system for the product development process is made up of a broad spectrum of functions. These functions can be broken down into application-related and system-overlapping functions (Ploenzke 1997).

• Application-related functions: The main priority for an EDM system is the management of all product data and documentation. Application-related functions provide task-speciﬁc support of the data management. Classical data management functions such as additions, modiﬁcations, and deletions are extended to integrate additional capabilities. For example, in the case of drawing management, automatic extraction of metadata from the CAD drawing header is im- plemented in the EDM database or used for classiﬁcation of the components.

• System-overlapping functions: Data and document management requires an application-neutral infrastructure that provides functions for an organized handling of the management processes. These overlapping system and application functions support the data and document management through functions created for variant and version management and ensuring the editing status of the document. Also, these functions support the provision of central services such as user management, privacy, and data protection.

Support provided by the use of an EDM system is aimed at the integration of the information ﬂow and processes into the business processes. Integration and transparency are the essential aspects

for the EDM system to process every demand and provide the right information at the right time. The storage, management, and provision of all product-related data create the basis for:

• The integration of application systems for technical and commercial processes as well as for ofﬁce systems in a common database

• The task-oriented supply of all operations with actual and consistent data and documentation

• The control and optimization of business processes

Metadata are required for the administration of product data and documentation, enabling the identiﬁcation and localization of product data. Metadata represent the information about the creator, the data of generation, the release status, and the repository. The link between all relevant data is made by the workﬂow object, which is a document structure in which documents of various formats are ﬁlled out along a process chain. This structure is equivalent to part of the product structure and, after release, is integrated into the product structure. In general, this is the last step of a workﬂow.

The modeling and control of a process chain are the task of the workﬂow management, whereas access control is the task of document management. The users concerned in the workﬂow are man- aged by groups, roles and rights. The systems applied are integrated in the EDM system. Linking a special document with the editing system is the responsibility of the document management program.

Architecture and Components

The reference architecture, from Ploenzke (1997) (Figure 13) describes the individual components of an EDM system from a functional point of view and illustrates the connections between the respective system components.

The core of the reference architecture is the engineering data. The application-oriented and system- overlapping functions are applied to the core data. The engineering data are found within the model data and broken down into product-deﬁned data and metadata. Model data, such as drawing data, parts lists, text ﬁles, raster data from document, and data in other formats, are not interpreted by the EDM system but are still managed by the EDM System. Metadata is interpreted by the EDM system and contains information regarding the management and organization of the product data and doc- uments.

The application-oriented functions support the management of the model data. Although the model data may be made up of various kinds of data, the management of these data, with the help of application oriented functions, must be able to summarize the data logically in maps and folders. Also, relationships between the documents and product data can then be determined. Important application-oriented functions are (Ploenzke 1997):

• General data management

• Drawing data management

• Classiﬁcation

• Parts list management

• Standard parts management

• Project planning and management

• Production plan management

• NC data management

• Machine tool and equipment management

• Method management

System-overlapping functions support the business processes over numerous steps, such as change and workﬂow management. Furthermore, central system services such as user management and data security are provided. Important system-overlapping functions are (Ploenzke 1997):

• Change management

• Workﬂow management

• User administration

• Data security

• Data protection

• Communication

• Archiving

Besides the core components of the reference architecture, the following system components belong to the system environment (Ploenzke 1997):

• User interface

• Machine environment

• Interfaces

• Basic software and hardware environment

The user interface forms the direct interface to the user and thus must provide the look and functions desired by the user. The application environment provides methods for customizing the interface. This allows the adaptation of the EDM system to company-speciﬁc conditions and for maintenance of the system. Programming interfaces known as application procedural interfaces (APIs) enable the launching and integration of application modules for the functional extension of an EDM system. The basic software and hardware environment forms the respective operation platform of an EDM system. With the status of the technology today, client–server-based EDM systems come into operation with connecting databases. These provide the various users with the necessary client pro- grams, increasing the economical utilization of today’s workstations (Krause et al. 1996).

EDM can also be termed an enabling technology. The reference architecture is an integration platform for systems, functions, and data. Company-speciﬁc conditions and a dynamic control pro- cess, however, cause the transformation of each EDM system installation into an almost nontrans- ferable case. Nevertheless, the EDM reference architecture forms the basis of system integration for CAD integration, PPS coupling, or archiving, regardless of the company-speciﬁc focus.

The necessity for integration and transparency of the data leads to the broad scope of functionality enabled by an EDM system. The primary functions are combined in numerous modules. The level of performance of the individual modules varies from system to system and is determined by the company philosophy and strengths of the supplier (Figure 14).

Many functions that serve for the basic control of data in the EDM already stem from other systems such as CAD or PPS systems. These functions are combined in the EDM system and make feasible the creation and management of an information pool. This supports company-wide availa- bility and enables faster information searches (Figure 15). The information pool, known as the vault, is a protected storage area that enables the availability of all documents and ensures that no unau- thorized access occurs. Access to documents held in other systems is possible through logins to the various systems. After logging off, the document is available only in a read-only form. Therefore, logging off copies the document to local memory. Logging in and opening a changed document is associated with the saving of a new version of the document. All customary EDM systems on the market are able to manage a centralized or decentralized vault.

Calculation Methods

General Explanation

Increasing demands for product performance necessitates a secure component conﬁguration. The factor of safety is determined by calculations performed on the system. The following goals should be achieved:

• Assurance against failure

• Testing of product functionality

• Assessment of external effects

• High strength-to-weight ratio

• Optimal material utilization

• Achievement of an economical production process

The most important computer-supported calculation processes are:

• Finite element methods (FEM)

• Boundary element methods (BEM)

• Finite different methods (FDM)

These methods can be applied for problems in which differential equations describe any continua. The methods have different approaches for performing calculations. FEM assumes a variational for- mulation, BEM a formulation by means of integrals, and FDM uses differential equations. Due to the mathematical formulation, it is irrelevant whether the computational problem comes from me- chanics, acoustics, or ﬂuid mechanics. In all three methods, the discretization of the structure is common and is required in order to derive the mathematical formulations for the desired tasks.

The processes commercially available have integrated interfaces that enable them to work with geometries already developed in CAD systems. To prepare the geometry model, various support mechanisms for the particular calculation systems are offered.

Beyond FEM, BEM, and FDM, there are other calculation systems that are based on problem- speciﬁc model creation. These systems, however, are usually applied only in conjunction with an associated model. Two examples are the calculation of suspensions and the determination and layout of weld contacts.

The general sequence of a calculation is equivalent to that of an information handling process. Input in the form of the geometry, material, and forces is transformed using mathematical and physical rules to calculated results. In the majority of cases, the object geometry and material properties are simpliﬁed. Also, the stress–strain or loading characteristics are often idealized, which eases the cal- culation task and reduces time. In many cases assumptions and simpliﬁcations are made because otherwise the calculation of the problem might be too extensive or impossible.

Finite Element Methods

FEM is the most commonly used calculation process today. Their implementation spectrum covers many different areas. For example, they are applied in:

• Statics in civil engineering

• Crash research in the automotive industry

• Electromagnetic ﬁeld research for generator design

• Material strength and life cycle determination in mechanical engineering

• Bone deformation in biomechanics

FEM has found its place in the ﬁeld of structural mechanics. It is used for calculations in:

• Stress and deformation

• Natural shape and eigenfrequency

• Stability problems

Because complex structures, in respect to their mechanical or thermal behavior, are no longer solvable analytically, the structure must be broken down into smaller elements. The FEM process enables the breakdown of a larger structure into elements, thus enabling the description of component behavior. Therefore, very complex entities are solvable. Calculation problems from real-world applications are usually quite complex. For example, in crash testing it is necessary to reproduce or simulate the complete vehicle structure even when it consists of many different components and materials (Figure 16).

The ﬁnite elements are described geometrically using a series of nodes and edges, which in turn form a mesh. The formation of the mesh is calculable. The complete behavior of a structure can then be described through the composition of the ﬁnite elements.

An FEM computation can be divided into the following steps:

• Breakdown of the structure into ﬁnite elements

• Formulation of the physical and mathematical description of the elements

• Composition of a physical and mathematical description for the entire system

• Computation of the description according to requirements

• Interpretation of the computational results

The FEM programs commercially available today process these steps in essentially three program phases:

1. Preprocessing: Preparation of the operations, mesh generation

2. Solving: Actual ﬁnite element computation

3. Postprocessing: Interpretation and presentation of the results

Preprocessing entails mainly geometric and physical description operations. The determination of the differential equations is implemented in the FEM system, whereby the task of the user is limited to selecting a suitable element type.

In the following sections, the three phases of an FEM computation will be presented from a user’s point of view.

FEM Preprocessing The objective of the preprocessing is the automation of the mesh- ing operation. The following data are generated in this process:

• Nodes

• Element types

• Material properties

• Boundary conditions

• Loads

The data sets may be generated in various ways. For the generation of the mesh, there are basically two possibilities:

• Manual, interactive modeling of the FEM mesh with a preprocessor

• Modeling of the structure in a CAD system with a subsequent automated or semiautomated mesh generation

The generation of the mesh should be oriented towards the expected or desired calculation results. This is meaningful in order to simplify the geometry and consider a ﬁner meshed area sooner in the FEM process. In return, the user must have some experience with FEM systems in order to work efﬁciently with the technology available today. Therefore, completely automatic meshing for any complex structure is not yet possible (Weck and Heckmann 1993).

Besides interactive meshing, commercially available mesh generators exist. The requirements of an FE mesh generator also depend on the application environment. The following requirements for automatic mesh generators, for both 2D and 3D models, should be met (Boender 1992):

• The user of an FE mesh generator should have adequate control over the mesh density for the various parts of the component. This control is necessary because the user, from experience, should know which areas of the part require a higher mesh density.

• The user must specify the boundary conditions. For example, it must be possible to determine the location of forces and ﬁxed points on a model.

• The mesh generator should require a minimum of user input.

• The mesh generator must be able to process objects that are made up of various materials.

• The generation of the mesh must occur in a minimal amount of time.

The mesh created from the FE mesh generator must meet the following requirements:

• The mesh must be topologically and geometrically correct. The elements may not overlap one another.

• The quality of the mesh should be as high as possible. The mesh can be compared to analytical or experimental examinations.

• Corners and outside edges of the model should be mapped exactly using suitable node posi- tioning.

• The elements should not cut any surfaces or edges. Further, no unmeshed areas should exist. At the end of the mesh reﬁnement, the mesh should match the geometrical model as closely as possible. A slight simpliﬁcation of the geometry can lead to large errors (Szabo 1994).

• The level of reﬁnement should be greater in the areas where the gradient of the function to be calculated is high. This is determined with an automatic error estimation during the analysis and provides a new point for further reﬁnement.

Various processes have been developed for automatic mesh generation. The motivation to automate the mesh generation process results from the fact that manual generation is very time consuming and quite prone to mistakes. Many processes based on 2D mesh generation, however, are increasingly being suggested for 3D processes.

The most frequently used ﬁnite element forms for 2D are three- and four-sided elements and for 3D are tetrahedrons and hexahedrons. For automatic mesh generation, triangles and tetrahedrons are suitable element shapes, whereas hexahedrons provide better results in the analysis phase (Knothe and Wessels 1992).

FEM Solution Process In the solution process, equation systems are solved that are dependent on the type of examination being carried out. As a result, various algorithms must be provided that allow for efﬁcient solution of the problem. The main requirements for such algorithms are high speed and high accuracy.

Linear statistical examinations require only the solution of a system of linear equations. Dynam- ically nonlinear problems, on the other hand, require the application of highly developed integration methods, most of which are based on further developments of the Runge–Kutta method.

To reduce the computing time, matrices are often converted. This permits a more effective han- dling of the calculations. The objective is to arrange the coefﬁcients with a resulting diagonal matrix. An example of an FEM calculation sequence is shown in Figure 17.

FEM Postprocessing Because the calculation results of an FEM computation only de- liver nodes and their displacement and elements with the stresses or eigenforms in numerical rep-

resentation, postprocessors are required in order to present the results in a graphical form (Figure 18).

The postprocessing manages the following tasks:

• Visualization of the calculated results

• Plausibility control of the calculation

The performance and capability of postprocessors are constantly increasing and offer substantial presentation possibilities for:

• Stress areas and main stresses

• Vector ﬁelds for forces, deformation and stress characteristics

• Presentation of deformations

• Natural forms

• Temporary deformation analysis

• Temperature ﬁelds and temperature differences

• Velocities

It is possible to generate representations along any curve of the FEM model or, for example, to present all forces on a node. Stress ﬁelds of a component can be shown on the surface or within the component. The component can thus be reviewed in various views.

Search This Blog

Operation Research course