CAM: From Fragmented Disciplines to Systems Architecture
A system architecture for integrated manufacturing engineering is emerging from the point solutions now in use after three decades since the inception of APT. Today's conventional process planning, NC programming, documentation, offline and CMM programming, and other applications offer powerful functionality but encourage compartmentalization that results in errors, rework and duplication of effort. There are normally two parallel and concurrent process threads, one for "parts" and one for "assemblies," that simultaneously move from the virtual to the real spheres. There is little or no integration of the manufacturing engineering functions in either of the two process threads.
Fragmentation of manufacturing engineering systems in the part and assembly process threads.
Today's fragmentation of manufacturing software into dozens of different heterogeneous applications is an inevitable result of niche-based strategy of most software CAM system developers. While all three major automobile manufacturers have designated a single CAD/CAM system, the reality today within each company is the existence of a wide range of CAM products for core and niche applications. As a result, new copies of the original design model must be continually generated for each of the multitude of downstream operations required to define the manufacturing process. A body panel, for example, typically involves three to five forming operations, each of which requires die structures, die face, machining of casting and the associated patterns, checking fixtures, welding fixtures, process documentation, and so on. The irony is that manufacturing engineering has only 20% of the CAD/CAM seats yet they generate 80% of the data.
The result is that the design model is duplicated for various manufacturing and documentation functions, usually with no updateable links in place between the various disciplines. This explosion of data generated by various manufacturing engineering applications has overwhelmed the rudimentary data management infrastructure currently in place. More and more energy has to be devoted to communicating engineering revisions to downstream processes, yet the number of errors continues to grow. While the niche applications do a good job at capturing the geometry of parts, assemblies and tooling, little has been achieved to date in standardizing the concurrent engineering process. The result is a lack of repeatability in concurrent engineering and the persistence of substandard practices in many areas of the organization. The emergence of new niche applications and the isolation of engineering from the shop floor has further encouraged the splintering of manufacturing engineering into specialties that are more and more divorced from the shop floor.
The continual duplication of the design model for downstream processes means that manufacturing engineering typically generates four times as much data as the design.
These problems are on their way to being solved by an emerging architecture that incorporates: (1) a manufacturing model consisting of an associated design model, raw material and the basic process outline; (2) detailed set-up instructions to define the machine tool, tooling and operations that produce the part; (3) libraries of raw materials, operations templates, tools, etc. that serve as reference for the process definition; (4) application software for verification, simulation and optimization purposes; and (5) processors to generate the actual NC code and shop documentation.
Enablers Emerge.
The last several years have witnessed the emergence of key enabling technologies for the integration of manufacturing engineering under a single unified systems architecture. The crystallization of the STEP standard and its incorporation into leading CAD/CAM systems means that it is now possible to transmit solid models between different systems. The ability to convert process instructions, also part of the STEP standard, will become a reality in the future.
New tolerant modeling functionality allows manufacturing applications to work with existing solid and surface models without the need for modifications that were previously required to clean up gaps, overlaps, holes, etc. This innovation has the potential to allow manufacturing engineers to attach toolpaths to and perform other operations on the original solid or surface model generated during the design process, avoiding the need for duplicating the model. Another important new development is manufacturing templates that incorporate an outline of various standardized machining processes and can be attached to the part geometry to streamline and standardize the CNC programming process.
While software used to generate CNC code was the first and is still probably the most mature CAM tool, a significant new development has recently occurred. That is the availability of complex general-purpose processors that can be easily simplified so that a machine operator can use them with minimal training. The fact that they operate within the context of the core CAD/CAM system eliminates the geometry duplication problem. Another recent improvement in processors is the ability to keep track of material left uncut due to tool clearance issues so that it can easily be cleaned up on a subsequent pass.
The PDM Connection.
Data management capabilities have also been substantially improved. New library structures have been developed that can be linked to the shop floor for direct distribution of CNC programs and manufacturing documentation. Data management systems originally developed to manage subassemblies and assemblies now have the intelligence to update all affected subassemblies and assemblies whenever a component changes. Product data management (PDM) systems, just now beginning to see widespread implementation in the automotive industry, provide a means to access design and manufacturing models throughout the organization. These advancements have provided the foundation for a new integrated manufacturing engineering systems architecture that has the potential to pull together all of the fragmented systems of the past. A key requirement of this architecture is that it be able to handle the "process thread" from beginning to end with each step of the process leveraging the information contained in the digital "master model."
Perhaps the most basic capability offered by the new systems architecture is a tight link between the design and the manufacturing models. By the very definition of concurrent engineering, the design model continues to evolve during manufacturing process definition, and the new architecture maintains control of and updates the manufacturing model to reflect these changes without the need for manual intervention or book-keeping. The "master model" concept, an extension of the associative approach, enables an automotive manufacturer and its suppliers to work concurrently and collaboratively by giving the data management system the responsibility for insuring that each group is using the latest version of the part and assembly geometry.
With the power of the new architecture inevitably comes complexity, which is the reason why it is important that the system be scalable and configurable to offer a simple, icon-driven interface tailored to specific applications. An example is a die machining system that allows users to navigate step by step through the machining process. A system of this type is installed at an automotive component manufacturing plant in Detroit; it is being used by more than 100 machinists, die and tool makers.
More generally, constructing scalable products from individual architectural elements provides the ability for automotive manufacturers to extend the general system to cover specialized tasks. The result is applications that are fully integrated with the core architecture at a minimum expense.
What's Required.
The new architecture must be flexible enough to adapt to new manufacturing processing techniques. For example, high-speed machining technology requires, in addition to new cut patterns, a new gentle part entry method known as "helical engages" to avoid the potential for the damage to the rapidly advancing cutting tool. Another recent innovation is the development of NURBS-based toolpaths that take advantage of the ability of a new generation of controllers to take over responsibility for converting curved geometries into thousands of line segments needed to provide instructions for the machine's servo motors. Reductions in machining time of up to 30% have been achieved by eliminating controller wait time and unnecessary cutter slowdown at path discontinuity. A sixteen-fold improvement in accuracy has also been accomplished.
Key requirements of a systems approach include the ability to handle the complete "process thread" from concept through to production; full associativity between design, tooling and manufacturing models; and an open architecture that can be extended to handle new manufacturing technologies. It's important to note that while this discussion has focused on the part and tooling examples, the same architectural elements are applicable to the assembly process engineering. Implementation of this architecture does not require throwing away the existing systems; it can proceed in steps and is easily scalable from a single department to the enterprise level. These features make the "systems" view for integrated manufacturing engineering the most promising concept now emerging in the CAD/CAM systems used in the automotive industry.
Fragmentation of manufacturing engineering systems in the part and assembly process threads.
Today's fragmentation of manufacturing software into dozens of different heterogeneous applications is an inevitable result of niche-based strategy of most software CAM system developers. While all three major automobile manufacturers have designated a single CAD/CAM system, the reality today within each company is the existence of a wide range of CAM products for core and niche applications. As a result, new copies of the original design model must be continually generated for each of the multitude of downstream operations required to define the manufacturing process. A body panel, for example, typically involves three to five forming operations, each of which requires die structures, die face, machining of casting and the associated patterns, checking fixtures, welding fixtures, process documentation, and so on. The irony is that manufacturing engineering has only 20% of the CAD/CAM seats yet they generate 80% of the data.
The result is that the design model is duplicated for various manufacturing and documentation functions, usually with no updateable links in place between the various disciplines. This explosion of data generated by various manufacturing engineering applications has overwhelmed the rudimentary data management infrastructure currently in place. More and more energy has to be devoted to communicating engineering revisions to downstream processes, yet the number of errors continues to grow. While the niche applications do a good job at capturing the geometry of parts, assemblies and tooling, little has been achieved to date in standardizing the concurrent engineering process. The result is a lack of repeatability in concurrent engineering and the persistence of substandard practices in many areas of the organization. The emergence of new niche applications and the isolation of engineering from the shop floor has further encouraged the splintering of manufacturing engineering into specialties that are more and more divorced from the shop floor.
The continual duplication of the design model for downstream processes means that manufacturing engineering typically generates four times as much data as the design.
These problems are on their way to being solved by an emerging architecture that incorporates: (1) a manufacturing model consisting of an associated design model, raw material and the basic process outline; (2) detailed set-up instructions to define the machine tool, tooling and operations that produce the part; (3) libraries of raw materials, operations templates, tools, etc. that serve as reference for the process definition; (4) application software for verification, simulation and optimization purposes; and (5) processors to generate the actual NC code and shop documentation.
Enablers Emerge.
The last several years have witnessed the emergence of key enabling technologies for the integration of manufacturing engineering under a single unified systems architecture. The crystallization of the STEP standard and its incorporation into leading CAD/CAM systems means that it is now possible to transmit solid models between different systems. The ability to convert process instructions, also part of the STEP standard, will become a reality in the future.
 |
| The new integrated systems architecture for manufacturing engineering developed by EDS Unigraphics can integrate the process threads, such as those illustrated for an automotive inner door panel. |
New tolerant modeling functionality allows manufacturing applications to work with existing solid and surface models without the need for modifications that were previously required to clean up gaps, overlaps, holes, etc. This innovation has the potential to allow manufacturing engineers to attach toolpaths to and perform other operations on the original solid or surface model generated during the design process, avoiding the need for duplicating the model. Another important new development is manufacturing templates that incorporate an outline of various standardized machining processes and can be attached to the part geometry to streamline and standardize the CNC programming process.
While software used to generate CNC code was the first and is still probably the most mature CAM tool, a significant new development has recently occurred. That is the availability of complex general-purpose processors that can be easily simplified so that a machine operator can use them with minimal training. The fact that they operate within the context of the core CAD/CAM system eliminates the geometry duplication problem. Another recent improvement in processors is the ability to keep track of material left uncut due to tool clearance issues so that it can easily be cleaned up on a subsequent pass.
 |
 |
 |
 |
| Machining time can be reduced and accuracy increased by exporting NURBS surfaces to controllers that are equipped to take over interpolation responsibility. |
The PDM Connection.
Data management capabilities have also been substantially improved. New library structures have been developed that can be linked to the shop floor for direct distribution of CNC programs and manufacturing documentation. Data management systems originally developed to manage subassemblies and assemblies now have the intelligence to update all affected subassemblies and assemblies whenever a component changes. Product data management (PDM) systems, just now beginning to see widespread implementation in the automotive industry, provide a means to access design and manufacturing models throughout the organization. These advancements have provided the foundation for a new integrated manufacturing engineering systems architecture that has the potential to pull together all of the fragmented systems of the past. A key requirement of this architecture is that it be able to handle the "process thread" from beginning to end with each step of the process leveraging the information contained in the digital "master model."
Perhaps the most basic capability offered by the new systems architecture is a tight link between the design and the manufacturing models. By the very definition of concurrent engineering, the design model continues to evolve during manufacturing process definition, and the new architecture maintains control of and updates the manufacturing model to reflect these changes without the need for manual intervention or book-keeping. The "master model" concept, an extension of the associative approach, enables an automotive manufacturer and its suppliers to work concurrently and collaboratively by giving the data management system the responsibility for insuring that each group is using the latest version of the part and assembly geometry.
With the power of the new architecture inevitably comes complexity, which is the reason why it is important that the system be scalable and configurable to offer a simple, icon-driven interface tailored to specific applications. An example is a die machining system that allows users to navigate step by step through the machining process. A system of this type is installed at an automotive component manufacturing plant in Detroit; it is being used by more than 100 machinists, die and tool makers.
More generally, constructing scalable products from individual architectural elements provides the ability for automotive manufacturers to extend the general system to cover specialized tasks. The result is applications that are fully integrated with the core architecture at a minimum expense.
What's Required.
The new architecture must be flexible enough to adapt to new manufacturing processing techniques. For example, high-speed machining technology requires, in addition to new cut patterns, a new gentle part entry method known as "helical engages" to avoid the potential for the damage to the rapidly advancing cutting tool. Another recent innovation is the development of NURBS-based toolpaths that take advantage of the ability of a new generation of controllers to take over responsibility for converting curved geometries into thousands of line segments needed to provide instructions for the machine's servo motors. Reductions in machining time of up to 30% have been achieved by eliminating controller wait time and unnecessary cutter slowdown at path discontinuity. A sixteen-fold improvement in accuracy has also been accomplished.
 |
| A key part of the new systems architecture are focused products designed to provide simplified user interfaces for specialized applications. |