3D printer in a workshop.

Advanced Additive Manufacturing Laboratory

Experimental setup for multimodal monitoring of Laser Directed Energy Deposition (LDED) processes for geometric variation detection.

The Advanced Additive Manufacturing Laboratory's research focuses on advancing smart and intelligent manufacturing technologies by enhancing the quality and efficiency of the fabrication processes through real-time monitoring, offline inspection and AI-driven techniques for quality prediction and evaluation.

While the Advanced Additive Manufacturing Laboratory's primary focus is currently centered on the Directed Energy Deposition (DED) additive manufacturing process, our research scope reaches well beyond DED.

Details of our current projects can be found in the Projects section. 

Capabilities


Our group embraces a wide range of advanced manufacturing technologies as part of our ongoing exploration and innovation efforts:

Advanced Manufacturing 

  • Smart Manufacturing: multi-modal process monitoring of manufacturing processes, such as metal additive manufacturing, polymer based additive manufacturing, integrated with AI based data processing for developing intelligent manufacturing systems. 

  • Quality inspection in manufacturing and assembly lines using advanced sensor systems 

Mechanical Testing and Material Characterization 

  • Mechanical Property Testing: Determining fundamental material properties including tensile strength, compression, and flexural modulus. 

  • Metallographic Sample Preparation: Preparing material samples via precision sectioning, mounting and polishing for microstructural examination.

Microscopy and Microstructural Analysis 

  • Non-destructive, high-resolution 3D imaging of internal structures, porosity, and morphology of materials and components using Micro-CT X-ray Microscopy.  

  • Optical Microscopy and Scanning Electron Microscopy Imaging: High-resolution imaging of polished and etched samples for microstructural analysis, phase identification, and defect detection. Example applications: measuring melt pool depth and width of printed tracks. 

Structural Health Monitoring (SHM) and Non-Destructive Evaluation (NDE) 

  • Real-time Damage Detection: Using Acoustic Emission (AE) to actively monitor and detect the initiation and growth of defects like cracks in real-time. 

  • Defect Identification and Classification: Analyzing AE signals to classify the type and severity of damage occurring within a material or structure. 

  • 3D Scanning and Reverse Engineering: Capturing the precise geometry of physical objects to create digital models for analysis, modification, or replication. 

  • Dimensional Analysis and Quality Control: Performing high-precision geometric measurements and automated inspections to verify part accuracy against CAD models. 

  • Thermal Imaging and Infrared Thermography: Utilizing high-resolution infrared cameras to detect thermal anomalies, monitor heat distribution, and identify subsurface defects, delamination, or areas of stress concentration. It collects melt pool temperatures and thermal gradients in real-time during printing processes.  

Photo of Souran ManoochehriSouran ManoochehriPrincipal Investigator

Souran Manoochehri, Ph.D.
Professor and Chair of the Department of Mechanical Engineering

Souran Manoochehri’s research interests are in areas of design and manufacturing, including additive manufacturing, computer integrated design and manufacturing and intelligent modeling and optimization. His research on additive manufacturing and integrated design integrates mathematical modeling, machine learning methods and experimental studies, providing product and process quality assurance. 

Contact information: [email protected] | 201-216-5562 

About Souran Manoochehri

Co-Principal Investigators

Chaitanya Krishna Vallabh (cvallabh)Chaitanya Krishna Vallabh 
Teaching Assistant Professor, Department of Mechanical Engineering

Chaitanya Vallabh’s research expertise lies in the broad area of system dynamics and vibrations with specialization in microparticle adhesion and manipulation using acoustic techniques coupled with laser interferometry. His research experience and interests also include additive manufacturing (AM), advanced, smart, and scalable manufacturing methods, elastic wave propagation, ultrasonic non-destructive evaluation, structural health monitoring (SHM), systems integration, computer vision and signal processing. 

Current Research Interests: Advanced manufacturing, Additive manufacturing, Engineering Education, Machine Learning Applications in Manufacturing and Automation and Big Data Analysis

Contact information: [email protected] | 201-216-5051

Chan Yu (cyu)Chan Yu
Lecturer, Department of Mechanical Engineering

Chan Yu is extensively involved in teaching courses on manufacturing, which include design for manufacturability, design for additive manufacturing, advanced additive manufacturing and optimization principles in mechanical engineering. His research interests lie in the areas of design and manufacturing, with a focus on design optimization, Design for Manufacturing and Assembly (DFMA), and additive manufacturing. One of his recent projects focused on the development of efficient optimization techniques for defect detection in manufacturing assembly.

Contact information: [email protected] | 201-216-5561

Student Researchers

Ke Xu.

Ke Xu

Ph.D. Candidate in Mechanical Engineering

Youmna Mahmoud.

Youmna Mahmoud

Ph.D. Candidate in Mechanical Engineering

Sudharshanan Dhabaseelan Vasugi.

Sudharshanan Dhabaseelan Vasugi

Ph.D. Student in Mechanical Engineering

André Colón.

André Colón

Ph.D. Student in Mechanical Engineering

Alumni

Javid Akhavan, Ph.D.

Javid Akhavan, Ph.D.

AI Engineer at ZS Consultants

Jiaqi Lyu, Ph.D.

Jiaqi Lyu, Ph.D.

Systems Software Engineer at Precision X-Ray, Inc.

Current Projects

Sensor-based Monitoring and Data-driven Quality Prediction

Sensor-based Monitoring and Data-driven Quality Prediction in Additive Manufacturing 

Lead Student Researcher: Ke Xu 

Multimodal Monitoring for Geometric Variation in LDED: This project, developed a novel multimodal monitoring framework that synergistically integrates contact-based acoustic emission (AE) sensing with vision-based coaxial camera monitoring. The primary goal was to enable layer-wise identification and evaluation of geometric variations in Laser Directed Energy Deposition (LDED) fabricated parts, using specimens with and without through-holes as test cases. By applying multiple machine learning algorithms including SVM, neural networks, random forest, gradient boosting, XGBoost, and logistic regression, the integrated multimodal strategy demonstrated superior classification performance of 94.4% accuracy, significantly surpassing individual sensing modalities. This research established a technical foundation for characterizing part variations and manufacturing-induced defects. 

Experimental setup for multimodal monitoring of Laser Directed Energy Deposition (LDED) processes for geometric variation detection.Experimental setup for multimodal monitoring of Laser Directed Energy Deposition (LDED) processes for geometric variation detection.Optomec LDED machine chamber with an integrated AE sensor. Optomec LDED machine chamber with an integrated AE sensor.

Predicting Mechanical Hardness in LDED using AE and Machine Learning: This project aimed to develop a non-destructive method coupled with machine learning (ML) models for predicting the mechanical hardness of samples printed by the Laser Directed Energy Deposition (LDED) process. By leveraging real-time Acoustic Emission (AE) signals captured during the printing process, the study investigated the influence of key printing parameters (overlap ratio, dwell time, and number of layers) on both AE signal characteristics and the resulting mechanical hardness. Three ML regression models—Support Vector Regression (SVR), Gradient Boosting Regression (GBR), and Gaussian Process Regression (GPR) were developed and tested to establish a predictive relationship between AE features and hardness. GBR demonstrated the highest accuracy, achieving a Mean Absolute Error (MAE) of 2.72 and a Mean Absolute Percentage Error (MAPE) of 3.43% on the testing set. This highlights the potential of integrating AE sensors with advanced machine learning algorithms for real-time, non-destructive prediction of mechanical properties in LDED additive manufacturing processes.

Melt Pool Depth Prediction in Directed Energy Deposition (DED)

Melt Pool Depth Prediction in Directed Energy Deposition (DED) Single-Track Prints Using Point Cloud Analysis 

Lead Student Researcher: Youmna Mahmoud

This project introduces a data-driven approach to predicting melt pool depth in Directed Energy Deposition (DED) additive manufacturing using laser-scanned point cloud data. To eliminate the need for destructive and time-consuming inspection methods, we developed a fully automated workflow that extracts key geometric features, such as track width and height, from point cloud scans to estimate melt pool depth. Preprocessing of the raw point cloud data involved advanced filtering and segmentation techniques, including RANSAC (Random Sample Consensus) for plane fitting, DBSCAN (Density-Based Spatial Clustering) for noise and outlier removal, and PCA (Principal Component Analysis) for orientation normalization and feature extraction. 

We trained multiple supervised machine learning regression models, namely Linear Regression, Decision Tree (DT), Support Vector Regression (SVR), Gaussian Process Regression (GPR), and Neural Networks (NN). The dataset was divided into a training set (70%) and a testing set (30%), with 5-fold cross-validation applied during training to ensure robust model evaluation. Among the models, the Gaussian Process Regression model achieved the best performance, with a Mean Absolute Error (MAE) of 18.89 µm and a Root Mean Squared Error (RMSE) of 25.5 µm. This AI-powered pipeline provides a scalable, non-destructive solution for real-time melt pool monitoring and control in metal AM processes. 

Sim-to-Real Transfer Learning Approach: Real-Time Melt Pool Analysis in DED

Sim-to-Real Transfer Learning Approach for Real-Time Melt Pool Analysis in DED Process

Lead Student Researcher: Youmna Mahmoud

This project proposes a simulation-to-reality (sim-to-real) transfer learning framework to enable future real-time melt pool analysis in Directed Energy Deposition (DED) additive manufacturing. Real-time estimation of melt pool depth remains a significant challenge due to its subsurface nature and the limitations of in-situ sensing technologies. To overcome this, we develop a hybrid machine learning model that leverages both high-fidelity simulations and limited experimental data. 

The core of the framework is a multi-input hybrid Convolutional Neural Network (CNN), initially trained on a large synthetic dataset generated via thermal finite element simulations in COMSOL. These simulations model surface temperature distributions and corresponding melt pool dimensions (width and depth) across various DED process settings. This forms the pretraining stage, where the CNN learns thermal-process-geometry relationships in a controlled virtual environment. 

To bridge the gap between simulated and real-world conditions, a transfer learning stage follows. The pretrained model is fine-tuned using a small set of real infrared (IR) images captured with the Optris 05M camera during experimental DED runs. This step adapts the model to real-world factors such as emissivity variation, sensor noise, and unmodeled disturbances. 

While real-time deployment is the long-term objective, the current work validates the feasibility of the framework by demonstrating its ability to accurately predict melt pool geometry using experimental data. These results lay the groundwork for integrating the model into closed-loop control systems, where it could enable adaptive process optimization based on live thermal feedback. 

OPTOMEC 3D additive machine tool.OPTOMEC 3D additive machine tool.Infrared camera mounted on the deposition head inside a Directed Energy Deposition machine for real-time process monitoring.Optris 05M IR camera integrated in our Directed Energy Deposition setup.

Multi-Sensor Quality Inspection System

Multi-Sensor Quality Inspection System  

Lead Student Researcher: Sudharshanan Dhabaseelan Vasugi 

This project focuses on the development of a laser scanner-based 3D inspection framework for quality assurance in additive manufacturing. High-resolution point cloud data is captured and processed using machine learning based algorithms to identify and classify surface and geometric defects. The system is also designed to detect assembly misalignments and perform dimensional verification with an accuracy margin under 5%. The objective is to enable fast, automated, and scalable inspection for reliable quality assurance in advanced manufacturing workflows. 

Multi-sensor 3D inspection setup with laser scanners oriented to ensure full surface coverage of the sample, enabling defect detection, dimensional analysis, and assembly verification. Multi-sensor 3D inspection setup with laser scanners oriented to ensure full surface coverage of the sample, enabling defect detection, dimensional analysis, and assembly verification. Multiple views of the multi-sensor 3D inspection system.Another view of the multi-sensor 3D inspection setup.

Equipment

Our research employs covers a host of areas, each of which is furnished with state-of-the-art equipment.

Additive and Advanced Manufacturing

  • OPTOMEC LENS MTS 500 LDED 3D Printer 

  • Bambu Lab X1E 3D Printer 

  • Ultimaker S-Series 3D Printer (S5/S7) 

  • Creality Ender 3 Pro 3D Printer 

  • Formlabs SLA 3D Printing System (4L Printer, Wash, Cure) 

  • Tormach PCNC 440 CNC Mill 

Mechanical Property Testing

  • Instron 5969 Universal Testing System 

  • ADMET eXpert 9300 Rotating Beam Fatigue System 

  • WP 510 Motor-driven Torsion Tester 

Metallographic Sample Preparation

  • TechCut 5x™ Precision High Speed Saw 

  • TechPress 3x™ Mounting Press 

  • MetPrep 3™ Grinder/Polisher 

Microstructure Analysis

  • Optical Microscopy: Olympus BH-2 Microscope  

  • Scanning Electron Microscopy: Zeiss Auriga Small Dual-Beam FIB-SEM 

  • Sky Scan 1272 High resolution 3D Micro CT Xray Microscopy 

Sensors

  • Mistras Micro-SHM System 

  • Mistras Acoustic Emission Sensors (PK6I, PKWDI, PICO) 

  • FLIR Blackfly S Camera 

  • Optris PI 05M Infrared Camera  

  • FLIR A35 Thermal Imaging Camera 

  • 3D scanning system with a prototype setup replicating an assembly line using a conveyor belt (with adjustable speed) 

  • 3D Laser scanners (Revopoint POP series) 

  • Intel RealSense laser scanner 

Publications

Peer-Reviewed Journal Publications

  • Xu, K., Mahmoud, Y., Manoochehri, S. & Vallabh, C. K. P. (2025). Using acoustic emission signal analysis and machine learning algorithms for predicting mechanical hardness in laser directed energy deposition parts. The International Journal of Advanced Manufacturing Technology,138, 4455–4474. 

  • Mahmoud, Y., Lyu, J., Akhavan, J., Xu, K., & Manoochehri, S. (2023). Thermal history-based prediction of interlayer bond strength in parts manufactured by material extrusion additive manufacturing. The International Journal of Advanced Manufacturing Technology, 126(9), 3813-3829. 

  • Akhavan, J., Lyu, J., Mahmoud, Y., Xu, K., Vallabh, C. K. P., & Manoochehri, S. (2023). Dataset of in-situ coaxial monitoring and print’s cross-section images by Direct Energy Deposition fabrication. Scientific Data, 10(1), 776. 

  • Xu, K., Lyu, J., & Manoochehri, S. (2022). In situ process monitoring using acoustic emission and laser scanning techniques based on machine learning models. Journal of Manufacturing Processes, 84, 357–374. 

Conference Proceedings

  •  Sudharshanan Dhabaseelan Vasugi, Chaitanya Krishna Prasad Vallabh, and Souran Manoochehri, "Multi-Sensor 3D Inspection System for Enhanced Manufacturing Quality," Proceedings of the ASME 2025 International Design Engineering Technical Conferences and Computers and Information in Engineering Conference (IDETC/CIE2025), August 17–20, 2025, Anaheim, CA. 

  • Mahmoud, Y, Lyu, J, Vallabh, CKP, & Manoochehri, S. "Melt Pool Depth Prediction in Directed Energy Deposition Single-Track Prints Using Point Cloud Analysis." Proceedings of the ASME 2024 International Design Engineering Technical Conferences and Computers and Information in Engineering Conference. Volume 2A: 44th Computers and Information in Engineering Conference (CIE). Washington, DC, USA. August 25–28, 2024. V02AT02A024. ASME. https://doi.org/10.1115/DETC2024-141272 

  • Akhavan, J., Mahmoud, Y., Xu, K., Lyu, J., & Manoochehri, S. (2023, July). TDIP: Tunable Deep Image Processing, a Real Time Melt Pool Monitoring Solution. In 2023 Congress in Computer Science, Computer Engineering, & Applied Computing (CSCE) (pp. 1899-1908). IEEE. 

  • Lyu, J, Akhavan, J, Mahmoud, Y, Xu, K, Vallabh, CKP, & Manoochehri, S. "Real-Time Monitoring and Gaussian Process-Based Estimation of the Melt Pool Profile in Direct Energy Deposition." Proceedings of the ASME 2023 18th International Manufacturing Science and Engineering Conference. Volume 1: Additive Manufacturing; Advanced Materials Manufacturing; Biomanufacturing; Life Cycle Engineering. New Brunswick, New Jersey, USA. June 12–16, 2023. V001T01A024. ASME. https://doi.org/10.1115/MSEC2023-105104 

  • Mahmoud, Y, & Manoochehri, S. "In-Situ Temperature Monitoring of ABS During Fused Filament Fabrication (FFF) Process With Varying Process Parameters." Proceedings of the ASME 2021 International Design Engineering Technical Conferences and Computers and Information in Engineering Conference. Volume 3A: 47th Design Automation Conference (DAC). Virtual, Online. August 17–19, 2021. V03AT03A031. ASME. https://doi.org/10.1115/DETC2021-69813 

  • Xu, K., & Manoochehri, S. (2021). Health Monitoring Using Acoustic Emission Technique During Fused Filament Fabrication Printing Process. ASME. International Design Engineering Technical Conferences and Computers and Information in Engineering Conference, V010T10A006–V010T10A006. 

  • Xu, K., & Manoochehri, S. (2019). Job Shop Scheduling Optimization Using Genetic Algorithm With Global Criterion Technique. ASME. International Design Engineering Technical Conferences and Computers and Information in Engineering Conference, V001T02A059–V001T02A059. 

Publications in Preparation

  • Xu, K., Vallabh, C. K. P., & Manoochehri, S. (under review). Integrating Machine Learning with Multimodal Monitoring System Utilizing Acoustic and Vision Sensing to Evaluate Geometric Variations in Laser Directed Energy Deposition.