This dissertation presents a rapid manufacturing process for sand casting patterns using a hybrid additive/subtractive approach. This includes three major areas of research that will enable highly automated process planning; a critical need for a rapid methodology.

The first research area yields a model for automatically determining the locations of layers, given the slab height, material types and part geometry. Layers are chosen such that it will avoid catastrophic failures and poor machining conditions in general. First, features that are possible thin material machining positions are defined, and methods for detecting these feature positions from an STL model are studied. Next, a layer thickness calculation model is presented according to positions of these features.

The second area focuses on tools and parameters for the subtractive side of processing each layer. A tool size and machining parameter selection model is presented that can automatically select tool sizes and machining parameters, given layer thickness, part geometry, and material types. Machining strategies and related machining parameters are studied first. Then the method for Stepdown parameter calculation is presented. Finally, an algorithm based on both accessibility and machining efficiency is proposed for the selection of tool sizes for the rough cutting operation, finish cutting operation and optional semi-rough cutting operation.

The final research area focuses on a cutting force analysis for thin material machining with additional layer thickness & tool size interaction. Popular cutting force models are reviewed, and a suitable model for cutting force calculation in this process is evaluated. Then, a cantilever beam model is used to analyze the thin material machining failure problem, and a minimum layer thickness model is presented. Third, a combined layer thickness & tool size model is constructed based on the machining tool deflection under cutting forces.

This rapid pattern manufacturing process and related software has been implemented, and experimental data is presented to illustrate the efficacy of this system and its process planning methods.