Additive Manufacturing
Additive Manufacturing (AM), commonly known as 3D printing, involves fabrication of three dimensional objects in a layer-by-layer and additive fashion. AM has caused a paradigm shift in the traditional design-to-manufacturing loop and has significantly sped-up the prototyping phase. AM provides design freedom to engineers and enables creation of light-weight, high-performing, and bio-inspired objects. Thereby AM is opening-up a whole new design space which was closed to the traditional subtractive manufacturing techniques.
Spreadify: AI-enabled spreading process control in powder-bed AM
Contributors: Akash Mehta, Wentai Zhang
Ph.D. Advisor: Prof. C. Fred Higgs III
Reproduced from: Metals 9(11), (2019)
Powder-bed additive manufacturing (AM) involves two repetitive steps- powder spreading and powder binding/fusion. Final part properties like mechanical strength, surface finish, and density depend proper selection of spreading and binding process parameters. Most commercial 3D printers come equipped to work with a handful number of powders. Owing to the complexity of powder mechanics, an introduction of a new powder system in a commercial printer requires significant trial and error to obtain the correct parameters or a 'recipe' which result in parts with acceptable tolerances. This study focused on the powder spreading step with an intension to provide layer-wise spreading process control. A rheometry calibrated, GPU-parallelized discrete element method (DEM) model was used to study the physics of the spreading process. A surrogate model of the DEM model was developed using artificial intelligence or AI (viz., a fully connected back propagation neural network). This AI-model was used to produce a spreading process map connecting spreader speeds to spread layer properties. This AI-based spreading process control software, named Spreadify, is available for licensing through CMU CTTEC and Rice OTT
PhD defense
Spreadify
WCCM talk
Publications:
- Desai, P., & Higgs III, C. F., Spreading Process Maps for Powder-Bed Additive Manufacturing Derived from Physics Model-Based Machine Learning, Metals 9(11), (2019): 1176.
- Desai, P. S., Mehta, A., Dougherty, P. S. M., & Higgs III, C. F., A rheometry based calibration of a first-order DEM model to generate virtual avatars of metal Additive Manufacturing (AM) powders, Powder Technology 342 (2019): 441-456.
- Higgs III, C. F., & Desai, P. S., Carnegie Mellon University & William Marsh Rice University, Machine learning enabled model for predicting the spreading process in powder-bed three-dimensional printing, Patent pending (2017)
- Zhang, W., Mehta, A., Desai, P. S., & Higgs III, C. F., Machine Learning Enabled Powder Spreading Process Map for Metal Additive Manufacturing(AM), in Proceedings of SFF Symposium, Austin, TX, Aug, 2017, pp. 1235–1249.
Drillogy: AI-guided design of 3D printable drill bits for rock excavation
Contributors: Nicholas Wolf, Joshua Wagner
Co-Inventor: Prof. C. Fred Higgs III
Over the last many years, petroleum companies have been trying to find ways to excavate through the rocks to reach the energy rich geological formations in the least amount of time. The present invention introduces a new process to do exploratory design search, enabled by AI, for lithology-specific drill bits and eventually additively manufacturing or 3D printing the wining drill bit design that minimizes the rock excavation time by taking into account rate of penetration (RoP) and wear of drill bit. This AI-based drill bit design and well planning software, named Drillogy, is available for licensing through CMU CTTEC and Rice OTT
PhD defense
Drillogy
Publication:
- Higgs III, C. F., & Desai, P. S., Carnegie Mellon University & William Marsh Rice University, System and method for design of rapid excavating and wear-resistant drill bits, Patent pending (2019)