Additive manufacturing (AM) gives designers enormous freedom for creative forms and complex geometries which have been a constraint with conventional manufacturing. Additive manufacturing is an up and coming field but lacks automated and specific design rules for different AM processes. The manual verification of design rules to build accurate parts for AM is both time consuming and cumbersome. Moreover, it needs expert knowledge which is gained over longer periods of time and is difficult to replicate. The objective of this research is to develop a Design for Additive manufacturing (DFAM) framework which is composed of three different complimentary models. These models feed on input part characteristics, knowledge base, and additive manufacturing (AM) machine specifications.

The first model determines the optimal AM technique for each input part. A genetic algorithm using (XTOOLSS) software is implemented to execute the evolutionary computations. The second model determines the compliance of the entry part with AM rules using a decision tree algorithm. The modified J48 classifier in WEKA software is used to increase the accuracy of the data mining procedure. The final model consists of an AM Design Rule Engine to specific recommendations for each part design. A graphical user interface (GUI) is developed using Python programming language to determine deficiencies and recommend design changes to the part for DFAM compatibility. The metal models formulated in this research are extensible to accommodate additional element characteristics, AM machine specifications and updated knowledge base for a high-fidelity DFAM system. This study establishes a foundation for a comprehensive Design for Additive Manufacturing (DFAM) Expert System which automates the complex design and manufacturing processes in the AM field.