Cutkosky and Tenenbaum [63] developed NEXT-Cut: a system for the design and manufacture of machined parts. Using NEXT-Cut, the designer can create a design by subtracting volumetric machining features corresponding to machining operations from a piece of stock material. As features are subtracted from the workpiece, the system uses its knowledge base to analyze the design's manufacturability. If any of a variety of manufacturability considerations are violated, the designer is warned of the violating features. This system works directly with features defined by the designer and so it is incumbent upon the designer to describe the design in terms of the most appropriate set of features. This requires the designer have good knowledge about machining processes in order to select the most appropriate set for machining, as failure to do so may produce incorrect analysis.
Gupta et al. [25] describe a methodology for early evaluation of manufacturability for prismatic machining components. Their methodology identifies all machining operations which can be used to create a given design. Using those operations, different operation plans for machining the parts are generated. For each new operation plan generated, it is examined whether the plan can produce desired shape and tolerances. If the plan is capable of doing so, the manufacturability rating for the plan is calculated. If no operation plan can be found that is capable of producing the design, then the given design is considered unmachinable; otherwise, the manufacturability rating for the design is the rating of the best operation plan. The rating is based on estimated machining time for the part. Based on this approach, Das et al. [64] reported a methodology of suggesting improvements to a given design to reduce the number of setups to machine a part. Their approach involved using different machining operations to satisfy the geometric constraints put on the part by the designer. These constraints are based on the functionality of the part. Later different modifications are combined to arrive at redesign suggestions.
Yannoulakis et al. [65,66] developed a manufacturability evaluation system for axis-symmetric parts to be machined on turning centers. The part types under consideration did not include axis-symmetric features such as threads and splines. They created a feature-based description of the part and evaluated the manufacturability index of each feature. This manufacturability index was based on the estimated machining time of the feature. They employed a number of empirical techniques to estimate cutting parameters and machining time but did not consider geometric tolerances or the possibility of alternative features. The final result from the manufacturability evaluation procedures employed by them is a set of different indices. These indices gives different types of indications for manufacturability of individual features and complete parts. Some of the indicators deal with the comparative time spent in loading-unloading, fixturing and tool change. One of the feature of their system is that based on the result of the analysis it ranks the features as candidates for redesign. A number of research issues such as feature accessibility, precedence constraints, setups, etc., need to be addressed in order to scale up their approach to prismatic parts.
Lu and Subramanyan [67] developed a manufacturability evaluation system for bearing cages. They addressed several aspects of manufacturability that included fixturing, tooling, gaging, and material handling. They used a multiple cooperative knowledge sources paradigm that separated domain knowledge from the control procedure. Their domain was restricted to parts with axis-symmetric features which can be manufactured on a lathe.
Priest and Sanchez [68,69] developed an empirical method for measuring the manufacturability of machined parts. Their approach involves rating a design based on producibility rating factors. The producibility rating factor is calculated from considerations that influence producibility and observed production difficulties. They defined producibility rating factors for a variety of manufacturing considerations such as material availability, machinability tooling, material/process risk compatibility etc.
Hsiao et al. [70] developed a knowledge-base for performing manufacturability analysis of machined parts. Their approach is capable of incorporating user-defined features and represents machining processes by their elementary machining volumes and limitations on the tool motion. For each design feature, they defined constraint-face sets for representing various machining faces and any neighboring faces that restrict the accessibility of the feature. These constraint-face sets are evaluated to determine if the feature can satisfy the conditions imposed by the elementary machinable volume and tool motion for the machining process. While their approach is capable of handling a limited number of accessibility constraints and tolerances, it does not consider the possibility of alternative features and does not provide any scheme for computing a manufacturability rating.
Anjanappa et al. [71,72] developed a rapid prototyping system for machined parts. Their work emphasized using existing standards and available databases. For example, the design is stored as an IGES file and a rule-based feature extractor is used to find machining features. The set of features is limited and no intersections among features are allowed. The manufacturability analyzer performs analysis based on the specific machining cell configuration for which the system was designed. The manufacturability rating does not calculate machining cost and time but matches the features with tools, machines and fixtures. In addition, it lists those features that are non-manufacturable and those that are potentially difficult to manufacture. From these features, it also creates the NC machining code for machining of the component. This system does not investigate the possibility of alternative ways of machining the same part.
Hitachi corporation [73] extended their design for assembly methodology to also take into account machining processes. Together with their AEM method, this results in an overall producibility evaluation system. Boothroyd et al. [74] published a report on the evaluation of machining component during early design stage. They described two methodologies for arriving at cost estimates. The first methodology takes into account only part and stock geometry, batch size, material and component type. The second methodology uses more shop floor information. The feedback is in terms of manufacturing cost.
There are many other research works reported for manufacturability analysis for machining. We briefly mention some of those here. Chen et al. [75] has developed a system for setup generation and feature sequencing. They use multiple objective functions for setup and tool sequence generation. Mill et al. [76] devised a simultaneous engineering workstations.