Session: 02-01 Ceramics

Paper Number: 98101

98101 - Triply-Periodic Minimal Surface Infill With Tunable Mechanical Properties 

Additive manufacturing (AM) processes fabricate parts by adding material in a layer-by-layer fashion. This manufacturing method affords novel customisation capabilities from a wide variety of feed material to new types of geometries. In particular, customised lattice structures are of high interest across multiple industries including aerospace, medicine, and electrical components. While lattices are feasible to manufacture with AM, there are current existing challenges within the conventional AM process flow that inhibit design capabilities. Due to the high surface-area-to-volume-ratio (SAVR) and complex topology of lattices, there is a costly trade off of computational memory limits and geometric accuracy.

In the conventional AM process flow there are two methods for designing lattices that occur in either the computer-aided manufacturing (CAM) stage or the computer-aided design (CAD) stage. In the CAM stage, lattices are designed as an ``infill'' option. Infill options, i.e. lattice types and volume fraction, are present in many fused deposition modelling (FDM) CAM software. While this is a computationally efficient means to model lattices the customisation options are generally very limited and dependent on what is provided in the software. The other option is to model lattices in the CAD stage. The CAD stage offers more customisation possibilities for lattices, however, traditional CAD estimates geometries using parametric patches. As the SAVR and topology complexity increases, the number of patches increases, thus increasing the required computational resources. Furthermore, the intermediary file format to transfer the model to the CAM software typically estimates models using a triangular mesh, thus also facing similar challenges to the CAD parametric modelling.

This research harnesses the benefits of both methods in a streamlined alternative AM process flow for designing lattices. The work is specifically focussed on a sub-class of lattices called triply-periodic minimal surfaces (TPMS) that are defined by a single implicit function. These surfaces are infinitely smooth, minimal surfaces, free of self-intersections, and are embedded in 3-dimensional real space. As these lattices are easily defined and manipulated with implicit functions, this work employs meshless implicit-based modelling. Combined with modelling TPMS structures at the CAM stage as customisable infill, the result is a computationally efficient, streamlined alternative process flow.

The benefit of this is this alternative process flow can be used to efficiently create TPMS structures with tunable mechanical properties. Designers would be able to input a desired mechanical property and fabricate a customised TPMS structure. The alternative process flow uses a direct slicing algorithm that calculates toolpaths based on implicit contour lines. The contour lines are calculated based on geometric relationships between the geometric property and implicit function isovalue. The contour lines are then directly translated into GCode for a FDM printer. As mechanical properties and other advance engineering properties, such as compressive mechanical strength, are based on elementary geometric properties, further relationships can be established for specific shape, material, and AM technology applications.

This is demonstrated with compressive mechanical testing of Gyroid sheet structures with varying volume fractions. The the Gibson-Ashby relationship relates the compressive strength to the volume fraction of cellular materials. This was calculated for Young's modulus, thus, designers can use this established relationship to be able to create Gyroid sheet structures to specification. While there are some volume fraction discrepancies between the theoretical calculation and the measured volume fraction, the issues are identified for future developments and expansion of this alternative process flow.

Presenting Author: Soo-Hwa Kim University of Bath, UK

Presenting Author Biography: Soo-Hwa Kim is a PhD student at the University of Bath in the United Kingdom. Her research employs implicit modelling techniques for customising additively manufactured lattice structures, specifically focussed on triply-periodic minimal surfaces. Outside of work, she volunteers at a historic English heritage waterwheel, enjoys running, and loves cats.

Authors: