Approximate Functionally Graded Materials for Multi-Material Additive Manufacturing
Functionally graded materials (FGM) possess superior properties of multiple materials due to the continuous transitions of these materials. Recent progresses in multi-material additive manufacturing (AM) processes enable the creation of arbitrary material composition, which significantly enlarges the manufacturing capability of FGMs. At the same time, the fabrication capability also introduces new challenges for the design of FGMs. A critical issue is to create the continuous material distribution under the fabrication constraints of multi-material AM processes. Using voxels to approximate gradient material distribution could be one plausible way for additive manufacturing. However, current FGM design methods are non-additive-manufacturing-oriented and unpredictable. For instance, some designs require a vast number of materials to achieve continuous transitions; however, the material choices that are available in a multi-material AM machine are rather limited. Other designs control the volume fraction of two materials to achieve gradual transition; however, such transition cannot be functionally guaranteed. To address these issues, we present a design and fabrication framework for FGMs that can efficiently and effectively generate printable and predictable FGM structures. We adopt a data-driven approach to approximate the behavior of FGM using two base materials. A digital material library is constructed with different combinations of the base materials, and their mechanical properties are extracted by Finite Element Analysis (FEA). The mechanical properties are then used for the conversion process between the FGM and the dual material structure such that similar behavior is guaranteed. An error diffusion algorithm is further developed to minimize the approximation error. Simulation results on four test cases show that our approach is robust and accurate, and the framework can successfully design and fabricate such FGM structures.