More Like this

1. Physics-Informed Machine Learning of Argon Gas-Driven Melt Pool Dynamics

   https://doi.org/10.1115/1.4065457  

   Sharma, R; Guo, Y B; Raissi, M; Guo, W Grace (August 2024, Journal of Manufacturing Science and Engineering)

   Abstract Melt pool dynamics in metal additive manufacturing (AM) is critical to process stability, microstructure formation, and final properties of the printed materials. Physics-based simulation, including computational fluid dynamics (CFD), is the dominant approach to predict melt pool dynamics. However, the physics-based simulation approaches suffer from the inherent issue of very high computational cost. This paper provides a physics-informed machine learning method by integrating the conventional neural networks with the governing physical laws to predict the melt pool dynamics, such as temperature, velocity, and pressure, without using any training data on velocity and pressure. This approach avoids solving the nonlinear Navier–Stokes equation numerically, which significantly reduces the computational cost (if including the cost of velocity data generation). The difficult-to-determine parameters' values of the governing equations can also be inferred through data-driven discovery. In addition, the physics-informed neural network (PINN) architecture has been optimized for efficient model training. The data-efficient PINN model is attributed to the extra penalty by incorporating governing PDEs, initial conditions, and boundary conditions in the PINN model. 

   more » « less
2. Physics-Based Regression vs. CFD for Hagen-Poiseuille and Womersley Flows and Uncertainty Quantification

   Li, H.; Islam, M.; Yu, H.; Du, X. (July 2022, Eleventh International Conference on Computational Fluid Dynamics (ICCFD11))

   Brehm, Christoph; Pandya, Shishir (Ed.)

   Computational fluid dynamics (CFD) and its uncertainty quantification are computationally expensive. We use Gaussian Process (GP) methods to demonstrate that machine learning can build efficient and accurate surrogate models to replace CFD simulations with significantly reduced computational cost without compromising the physical accuracy. We also demonstrate that both epistemic uncertainty (machine learning model uncertainty) and aleatory uncertainty (randomness in the inputs of CFD) can be accommodated when the machine learning model is used to reveal fluid dynamics. The demonstration is performed by applying simulation of Hagen-Poiseuille and Womersley flows that involve spatial and spatial-tempo responses, respectively. Training points are generated by using the analytical solutions with evenly discretized spatial or spatial-temporal variables. Then GP surrogate models are built using supervised machine learning regression. The error of the GP model is quantified by the estimated epistemic uncertainty. The results are compared with those from GPU-accelerated volumetric lattice Boltzmann simulations. The results indicate that surrogate models can produce accurate fluid dynamics (without CFD simulations) with quantified uncertainty when both epistemic and aleatory uncertainties exist. 

   more » « less
3. Computational Fluid Dynamics Modeling of Material Deposition in Pneumatic Micro-Extrusion Additive Manufacturing

   O’Malley, Ethan; Salary, Roozbeh “Ross” (July 2025, American Society of Mechanical Engineers (ASME) - International Mechanical Engineering Congress and Exposition (IMECE 2025))

   Pneumatic micro-extrusion (PME) presents inherent challenges in achieving precise control over material deposition due to the small scale and high viscosity of the material flow. Traditional experimental methods often fall short in capturing the complex mechanisms, interactions, and dynamics governing material transport and deposition in PME, highlighting the necessity for advanced computational approaches to unveil these intricate physical phenomena. The primary objective of this study is to develop a robust computational framework for simulating material transport and deposition in PME, offering detailed insights into the fluid dynamics, flow complexities, and deposition characteristics intrinsic to the PME process. This involves systematically investigating the influence of key process parameters, such as material properties, print speed, and flow pressure, on deposition dynamics. Three-dimensional (3D) computational fluid dynamics (CFD) modeling was employed using ANSYS Fluent, with boundary conditions defined to replicate pneumatic extrusion on a moving substrate. The CFD simulations captured three distinct deposition regimes: (i) under-extrusion, (ii) normal extrusion, and (iii) over-extrusion. Results show that decreasing print speed at constant inlet velocity produced over-extrusion, with increased bead height and width due to material overflow. At normal extrusion, bead geometry remained stable, while under-extrusion reduced bead width and ultimately led to discontinuities when flow stresses exceeded cohesive strength. Notably, bead width decreased significantly between 10 mm/s and 15 mm/s, but showed little difference between 15 mm/s and 20 mm/s. Instead, the higher speed produced discontinuities consistent with experimental observations. Centerline velocity profiles revealed that the flow inside the cartridge was very slow at approximately 0.058 mm/s, whereas the velocity increased sharply within the nozzle throat, reaching nearly 12.88 mm/s. These predictions were further supported by validation experiments, where the measured nozzle velocity of 12.95 mm/s closely matched the CFD-simulated value of 12.88 mm/s, demonstrating strong agreement between simulation and experiment. Additionally, pressure decreased slightly with increasing print speed due to reduced backflow, while nozzle velocity and wall shear stress remained unchanged under fixed inlet velocity conditions. Overall, the outcomes of this study are expected to inform parameter optimization strategies and enhance PME process efficiency, providing critical insights into how variations in PME process dynamics influence print morphology and quality, advancing the path toward optimized fabrication of scaffolds for tissue engineering applications. 

   more » « less
4. Drag coefficient of bent-awn plumegrass ( Saccharum contortum ) seeds in wind

   https://doi.org/10.1063/5.0231717  

   Faraji_Dizaji, Farzad; Wang, Binbin; Sullivan, Lauren L; Kellogg, Elizabeth A (October 2024, Physics of Fluids)

   We present a combination of laboratory experiments and computational fluid dynamics (CFD) simulations to understand the wind-induced drag force and drag coefficient for Saccharum contortum seeds. Seed drop experiments indicate that the settling fall velocities of hair-equipped seeds are within 1–2 m/s, compared to 2.34 times higher settling fall velocity of the seed without hairs. The experimental data illustrate a power-law relationship between drag coefficient (Cd) and Reynolds number (Re) under the free fall condition: Cd∼Re−1.1. CFD simulations show that both viscous and pressure drag force components are important in contributing to wind drag. The presence of hairs substantially increases pressure drag, and its relative importance depends on hair number and orientation. Seed morphology including hair number and orientation influences the drag coefficient under different flow directions relatively to the seed body. The lower drag coefficient observed with crossflow wind compared to free fall suggests that seeds encounter less air resistance while drifting horizontally in the wind, favoring extended flying time and distance. Based on the varying drag coefficients under different conditions, we propose the incorporation of varying drag coefficients in future wind-driven seed dispersal models. 

   more » « less
5. Predicting Milk Flow Behavior in Human Lactating Breast: An Integrated Machine Learning and Computational Fluid Dynamics Approach

   https://doi.org/10.1115/1.4068077  

   Olapojoye, Abdullahi O; Zaheri, Shadi; Nostratinia, Aria; Hassanipour, Fatemeh (May 2025, Journal of Biomechanical Engineering)

   This study develops a comprehensive framework that integrates computational fluid dynamics (CFD) and machine learning (ML) to predict milk flow behavior in lactating breasts. Utilizing CFD and other high-fidelity simulation techniques to tackle fluid flow challenges often entails significant computational resources and time investment. Artificial neural networks (ANNs) offer a promising avenue for grasping complex relationships among high-dimensional variables. This study leverages this potential to introduce an innovative data-driven approach to CFD. The initial step involved using CFD simulations to generate the necessary training and validation datasets. A machine learning pipeline was then crafted to train the ANN. Furthermore, various ANN architectures were explored, and their predictive performance was compared. The design of experiments method was also harnessed to identify the minimum number of simulations needed for precise predictions. This study underscores the synergy between CFD and ML methodologies, designated as ML-CFD. This novel integration enables a neural network to generate CFD-like results, resulting in significant savings in time and computational resources typically required for traditional CFD simulations. The models developed through this ML-CFD approach demonstrate remarkable efficiency and robustness, enabling faster exploration of milk flow behavior in individual lactating breasts compared to conventional CFD solvers. 

   more » « less