Research
The Laboratory for Advanced Manufacturing Engineering & Systems (LAMES) is a research lab in the CECS at Marshall University, established with the goal to develop fundamental research as well as disseminate knowledge in the area of Advanced/Additive Manufacturing. The LAMES research thrusts include: (i) Biomedical Fabrication; (ii) Polymer-based Additive Manufacturing (based on fused filament fabrication and polyjet deposition); (iii) Printed Electronics (in collaboration with the Center for Advanced Microelectronics Manufacturing (CAMM) at State University of New York at Binghamton. The current focus of the LAMES is to forward research platforms for modeling, experimental characterization, optimization, and ultimately control of direct-write additive manufacturing processes. Currently, the following projects are being pursued.
1. Biomedical Manufacturing
The goal of this research work is to fabricate defect-free anatomical models as well as surgical patient-matched implants. The figures illustrate porous bone scaffolds, fabricated at the Lab for Advanced Manufacturing Engineering & Systems (LAMES). This research relates to the area of advanced bio-fabrication, linking high-fidelity physics-based modeling and simulation with experimental process characterization. Please contact Dr. Salary directly if you are interested in collaboration.
2. Advanced Manufacturing Process Modeling
The aim is to understand the underlying, causal phenomena behind material transport, deposition, and free-form formation/fabrication, using advanced modeling methods, such as Computational Fluid Dynamics (CFD) as well as Finite Element Analysis (FEA). The figures demonstrate CFD simulations of materiel transport and deposition in Aerosol Jet Printing (an additive manufacturing technique, used for the high-resolution fabrication of electronics).
3. In-Situ Process Monitoring and Control
Additive manufacturing processes are complex in nature, prone to gradual drifts in machine behavior and deposited material. The goal of this project is to forward avenues for real-time process monitoring and closed-loop control. In pursuit of this goal, various control techniques, e.g., Model Predictive Control (MPC), Fuzzy Control, and Robust Control, are implemented to control the dynamics of additive manufacturing processes. The outcomes of this project pave the way for fabrication of parts with improved uniformity in both size and functional performance. This is unavoidable in critical applications such as defense, aerospace, printed electronics, and biomedical fabrication.
4. Digital Image Processing and Computer Vision
The aim is to forward digital image processing algorithms (developed in MATLAB, Python, and C++ environments) to extract certain morphology features toward in situ process monitoring and control. The figure exemplifies the use of digital image processing for quantification of overspray as well as the width of a printed line.
5. Digital Image Correlation (Experimental Mechanics)
Digital Image Correlation (DIC) is a scale-agnostic, non-contact technique, utilized for measuring material deformation. The goal of this project is to monitor the reliability of flexible electronics during tensile cycling. The objective is to establish a platform to map displacement and strain fields, which not only pinpoint the onset of crack formation, but also allow for modeling of crack propagation. This work is in collaboration with State University of New York at Binghamton - Center for Advanced Microelectronics Manufacturing. The video demonstrates heterogeneous deformation of a flexible electronic structure (composed of a polymer substrate and silver nanoparticle tracks) during tensile cycling, simulated using FEA.
6- Machine Learning (ML) and Artificial Intelligence (AI)
In this project, the aim is to develop machine learning methods - such as Artificial Neural Network (ANN), Naïve Bayes (NB), k-Nearest Neighbors (kNN), Support Vector Machine (SVM), and Ensemble of Learners - to estimate the functional properties of additively manufactured parts as a function of their critical traits as well as process parameters (with an accuracy of ≥ 90%).