(Dr. David Rosen, advisor)
"A Decision Support System for Fabrication Process Planning in Stereolithography"
Stereolithography prototypes are being used for numerous applications in the design process. These rapidly prototyped parts are used more and more to create working prototypes, tooling, and even production grade products. To make full use of stereolithography in the design process, the limitations of stereolithography must be known. To better understand these limitations, relationships between part geometry, tolerances, and process variables must be understood quantitatively. Once understood, process plans may be developed that will result in higher quality prototypes that are able to meet or exceed any expectations of the prototype.
A method of process planning for stereolithography based on empirical models, analytical models, and multi-objective optimization is described in this thesis. Three characteristics of prototypes are investigated: surface finish, accuracy, and time required to build the prototype. These characteristics are quantified through the use of analytical and empirical models that relate parameters of the stereolithography process to the quality of the fabricated prototype. This process planning method allows user requirements in the form of geometric tolerances and surface finish tolerances to be applied to different surfaces and features on a CAD model. Using these requirements, process plans are developed and evaluated in terms of how well the different requirements are met and the amount of time necessary to build the prototype. In the end, several alternative process plans are developed with predictions for both the achievable tolerance values and the required build time.
The process planning method is implemented in C++ with the use of a
solid modeling kernel. Several examples are provided to highlight
the capabilities and limitations of the process planning method.
In addition the models used to quantify the different characteristics of
the prototype are validated. The validation of these models show
that the surface finish models may be used to make fairly good predictions.
The accuracy models while inaccurate appear to provide some benefit in
terms of the trends that are produced. And the build time models
provide a general indication of the actual build time. The results
of these examples also demonstrate that the process planning method does
in fact provide decision support for the development of process plans in
stereolithography. Through the use of the process planning tool a
better understanding of the effect of different process plans on the quality
of the fabricated prototype may be developed.