Aerospace Engineering Commons^™

Multifidelity Dust Erosion Analysis Of Mars Entry Vehicles, Adam Boland, Dominic Zanti, Serhat Hosder, Andrew Hinkle

Mechanical and Aerospace Engineering Faculty Research & Creative Works

The objective of this paper is to present a multifidelity approach for estimating recession rate and kinetic energy impact rate on the surface of planetary entry vehicles operating in dusty atmospheric environments at hypersonic speeds. The multifidelity model used a co-Kriging approach that combined a low-fidelity correlation with a correction factor from high-fidelity CFD solutions. The developed multifidelity model enables efficient and accurate exploration of a design space to determine at what conditions encountering dust is most dangerous to TPS survivability. Two sample problems are used to demonstrate the effectiveness of the approach: the Mars 2020 lander and a HIAD-type …

Improving Neural Network Efficiency With Multi-Fidelity And Dimensionality Reduction Techniques, Vignesh Sella, Thomas O'Leary-Roseberry, Xiaosong Du, Mengwu Guo, Joaquim R.R.A. Martins, Omar Ghattas, Karen Willcox, Anirban Chaudhuri

Mechanical and Aerospace Engineering Faculty Research & Creative Works

Design problems in aerospace engineering often require numerous evaluations of expensive to-evaluate high-fidelity models, resulting in prohibitive computational costs. One way to address the computational cost is through building surrogates, such as deep neural networks (DNNs). However, DNNs may only be an effective surrogate when sufficient evaluations of the high-fidelity model are required such that the up-front training cost is amortized, or in situations that require real-time responses (such as interactive visualizations). Typically, the data requirements for adequately accurate training of DNNs are often impractical for engineering applications. To alleviate this issue, the proposed work utilizes output dimensionality reduction along …

An Adaptive Sampling Strategy On Optimal Takeoff Trajectory Prediction Of Electric Drones, Dheeraj Paramkusham, Samuel Sisk, Jiachen Wang, Shuan Tai Yeh, Xiaosong Du, Nathan Roberts

Mechanical and Aerospace Engineering Faculty Research & Creative Works

Electric vertical takeoff and landing (eVTOL) aircraft transforms future transportation systems by alleviating transportation congestion on the ground. This eVTOL technique possesses unique features, including reduced noise, low pollutant emissions, efficient operating costs, and flexible maneuverability. Meanwhile, battery consumption poses critical challenges to flight task duration. Thus, optimal takeoff trajectory design is essential due to immense power demands during eVTOL takeoffs. Conventional design optimization, however, iteratively evaluates high fidelity simulation models, making the design process computationally intensive. In this work, we implement a machine learning-enabled inverse mapping optimization concept, i .e., directly predicting optimal design based on design requirements (including …

High-Order Dynamic Mode Decomposition For Multidimensional Harmonic Retrieval, Yanming Zhang, Steven Gao, Lijun Jiang

Mechanical and Aerospace Engineering Faculty Research & Creative Works

Estimating channel parameters such as azimuth, elevation, Doppler shift, and delay is a key challenge in wireless communication, often formulated as a multidimensional harmonic retrieval (MHR) problem. To address this, we propose a high-order dynamic mode decomposition (HODMD) framework for robust frequency estimation from high-dimensional signals in noisy environments. The HODMD approach combines high-order singular value decomposition (HOSVD) to decompose tensor data into a core tensor and mode matrices, with dynamic mode decomposition (DMD) to extract frequencies from the imaginary parts of the DMD eigenvalues. Simulation examples validate the effectiveness of the proposed method, demonstrating its efficiency in solving MHR …

Characterizing Intralaminar Distal Crack And Delamination In Multidirectional Laminates Under Low Velocity Impact: Modeling Approach With Ultrasound Testing Validation, Niildiip Chandraa, Rachel Van Lear, Arief Yudhanto, Yuqing Zhao, David A. Jack, Douglas E. Smith

Mechanical and Aerospace Engineering Faculty Research & Creative Works

Multidirectional laminated composites are essential for load-bearing structures, especially under frequent impact loading. While existing modeling techniques simulate various impact damage modes, our understanding of the underlying mechanisms, specifically the initiation and progression of intralaminar distal cracks and interlaminar delamination, needs improvement. Validating these mechanisms is crucial for enhancing modeling accuracy and expediting the design process. In this work, we propose a three-dimensional finite element modeling (3D FEM) approach aimed at elucidating the impact damage mechanisms in multidirectional carbon fiber reinforced polymer (CFRP). Our model incorporates cohesive zone models (CZM) to simulate the behavior of intralaminar cracks at distal points …

Design And Testing Of An Intrusive Arm For Supersonic Flow Sampling, Caleb Roberts, Davide Viganò

Mechanical and Aerospace Engineering Faculty Research & Creative Works

Intrusive probes are critical tools in supersonic wind tunnel testing, enabling direct measurement of flow properties such as pressure and heat flux. To support these measurements, a single-axis intrusive arm was designed and implemented in the Missouri S&T Supersonic Wind Tunnel. The system is designed to maintains precise probe positioning, while withstanding the aerodynamic loads, and preserving a dynamic seal between the test section and ambient environment. Its modular design accommodates a range of probe types for various research applications. This paper presents the design methodology, including structural, mechanical, and aerodynamic considerations, and details the integration process. System performance was …

Physics-Constrained Generative Artificial Intelligence For Rapid Takeoff Trajectory Design, Samuel Sisk, Xiaosong Du

Mechanical and Aerospace Engineering Faculty Research & Creative Works

The urban air mobility (UAM) industry is rapidly growing to alleviate regional transportation congestion. Electric vertical takeoff and landing (eVTOL) aircraft plays a critical role in this growth due to their efficiency and reduced operating cost. However, excessive energy demands by the takeoff phase impact the practicality of the aircraft. Conventional multidisciplinary analysis and optimization (MDAO) identifies a minimum energy trajectory but can be computationally intensive due to iteratively evaluating high-fidelity simulation models. In addition, complex constraints in practical MDAO pose crucial challenges to the already demanding process. Surrogate models enable efficient design optimization, but complex constraints could prohibit surrogate-based …

Versatile Injector Platform: A Modular Design For Supersonic Flow Research, Connor Bell, Davide Viganò

Mechanical and Aerospace Engineering Faculty Research & Creative Works

Fuel–air mixing remains a critical challenge in the development of high-speed air-breathing propulsion due to the extremely short residence times in supersonic combustors. Experimental studies are essential to investigate the underlying mixing mechanisms, but they require injection systems that can deliver repeatable and controlled flow conditions. In this work, we present the design of a modular strut-type injection platform developed for the Missouri S&T Supersonic Wind Tunnel. In its initial configuration, the platform features a planar trailing-edge slit to generate a two-dimensional jet and includes interchangeable trailing-edge modules to support a wide range of geometries and flow-control strategies. Designed with …

Deep Reinforcement Learning-Based Optimal Takeoff Trajectory Design Of An Evtol Drone, Nathan M. Roberts, Bingling Huang, Xiaosong Du

Mechanical and Aerospace Engineering Faculty Research & Creative Works

The continuing development of electric vertical take-off and landing (eVTOL) aircraft presents promising opportunities to alleviate transportation congestion. However, designing and executing takeoff trajectories that optimally balance energy usage, passenger comfort, and aircraft constraints remains challenging. Conventional design optimization provides a solution for pre-defined flight conditions and environments but is not ideal in real-world applications. In contrast, deep reinforcement learning (DRL) implements optimal policy making real-time decisions with no assumption of mathematical models. Seeing the lack of literature on DRL-based takeoff trajectory design of eVTOL aircraft, we implement and conduct DRL-based optimal takeoff trajectory designs for the Airbus³ Vahana …

Optimizing Uav Swarm Deployment For Efficient Communication Signal Strength Alignment In Disaster Scenarios, Mina Khalilzadeh Fathi, Chaoying Pei

Mechanical and Aerospace Engineering Faculty Research & Creative Works

In disaster scenarios, establishing reliable communication infrastructure is critical, and unmanned aerial vehicle (UAV) swarms offer a promising solution as temporary base stations. This study models communication demand in disaster-affected areas by applying Gaussian kernels to building data, forming a spatial demand distribution. Signal strength is estimated using the normalized inverse Free Space Path Loss (FSPL) to account for realistic attenuation. To guide UAV placement, we extract high-demand regions from the demand distribution using a gradient-based thresholding method. Based on this information, we develop a greedy algorithm to iteratively position UAVs for optimal coverage in areas with the greatest communication …

Few-Shot Learning-Enhanced Tiered Path Planning For Mars Rover Navigation, Ziyi Wang, Di Yu, Mina Khalilzadeh Fathi, Chaoying Pei

Mechanical and Aerospace Engineering Faculty Research & Creative Works

Path planning for Mars rovers presents significant challenges due to the diverse terrain, ranging from easily navigable areas to hazardous zones. Traditional methods typically classify terrain simply as passable or impassable, failing to account for the nuances of more moderately challenging areas. In this paper, we introduce a tiered terrain-aware path planning strategy, employing few-shot learning to classify and segment Martian terrain into levels of difficulty. The few-shot learning model, trained on Earth, is sent to the rover, enabling real-time processing of images from satellites or helicopters. The flexibility of few-shot learning, which requires minimal data and training time, enables …

Vibration Modeling Of Rectangular Plates And Application To Solar Cell Dust Mitigation, Jeremiah John Rittenhouse

Doctoral Dissertations

"Many engineering structures can be modeled as rectangular plates, including solar cells. Carefully controlled plate vibration can be useful to solve an engineering problem, such as lunar dust accumulation on solar cells, which blocks light and reduces power generation. For this application, plate vibration at resonance is studied and used experimentally to eject lunar dust from a solar cell.

Piezoelectric actuator placement on the inactive side of a solar cell to induce vibration dust mitigation from the active side of the cell is presented in this work. Three solar cell prototypes were created and tested for efficacy, resulting in an …

Additive Manufacturing Of Multifunctional Materials And Composites, John Michael Pappas

Doctoral Dissertations

"This research focused on developing fundamental knowledge of process-material interactions for additive manufacturing of multifunctional materials that are highly desirable in high-tech industries for potential weight and space savings. Multifunctional materials are very sensitive to defects, which severely hinder performance. Thus, defect formation was systematically studied to develop strategies and methodologies to efficiently fabricate high-performance materials and composites. Blown-powder-based laser additive manufacturing of transparent spinel ceramics was investigated to reduce porosity and cracking, defects that hindered transparency and mechanical properties. A systematic study of processing parameters revealed that laser power and powder flow rate had the largest effects on defect …

Harnessing Extrinsic Dissipation To Enhance The Toughness Of Composites And Composite Joints: A State-Of-The-Art Review Of Recent Advances, Gilles Lubineau, Marco Alfano, Ran Tao, Ahmed Wagih, Arief Yudhanto, Xiaole Li, Khaled Almuhammadi, Mjed Hashem, Ping Hu, Hassan A. Mahmoud, Fatih Oz

Mechanical and Aerospace Engineering Faculty Research & Creative Works

Interfaces play a critical role in modern structures, where integrating multiple materials and components is essential to achieve specific functions. Enhancing the mechanical performance of these interfaces, particularly their resistance to delamination, is essential to enable extremely lightweight designs and improve energy efficiency. Improving toughness (or increasing energy dissipation during delamination) has traditionally involved modifying materials to navigate the well-known strength-toughness trade-off. However, a more effective strategy involves promoting non-local or extrinsic energy dissipation. This approach encompasses complex degradation phenomena that extend beyond the crack tip, such as long-range bridging, crack fragmentation, and ligament formation. This work explores this innovative …

Microstructured Microstructure: Effects Of Concentration Variations On Shock Initiation Of Hmx-Based Plastic Explosives, Dana D. Dlott, Lawrence Salvati, Siva Kumar Valluri

Mechanical and Aerospace Engineering Faculty Research & Creative Works

We studied shock-compressed plastic-bonded explosives (PBX) consisting of HMX (cyclotetramethylene-tetranitramine) with PDMS (poly-dimethylsiloxane) binder, where we varied the HMX weight percent (wt%) up to a maximum of 60 wt%. In this lower-concentration regime, HMX particle clustering causes the PBX structure to consist of HMX clusters and PDMS islands. Structural analysis by optical microscopy allowed us to compute a radial correlation function that gives the mean distance from an average HMX particle to the nearest polymer island. Short-duration 20 GPa shocks (4 ns) created hot spots, and time-resolved thermal emission was used to obtain the growth rate of the subsequent deflagration. …

Hyperspectral Imaging For Temperature Measurements Of Hot Spots In Shocked Plastic-Bonded Explosives, Dhanalakshmi Sellan, Siva Kumar Valluri, Dana D. Dlott

Mechanical and Aerospace Engineering Faculty Research & Creative Works

Hot spots are formed when energetic microstructures are shocked, and they play a critical role in shock sensitivity. It is important to know both the time-dependent size and temperature of the hot spots, in order to generate a kinetic model to describe reaction growth in plastic-bonded explosives (PBX). Here we use a recently developed technique where PBX is fabricated in the form of a thin wafer, embedded within a transparent polymer binder, and shocked with a laser-launched flyer plate that produces pressures of about 30 GPa range. In this method, every crystal of the cyclotetramethylene-tetranitramine (HMX)-based PBX can be observed …

Accelerating Shock-Driven Reactions In Metal Nanocomposites, Siva Kumar Valluri, Edward L. Dreizin, Dana D. Dlott

Mechanical and Aerospace Engineering Faculty Research & Creative Works

Metal powders are sought as energetic additives to conventional explosives. However, due to their sluggish reaction kinetics with external gaseous oxidizers, pure metals are replaced with metallic composites with intimately mixed condensed phase oxidizers prepared by arrested reactive milling (ARM). Such composites can be initiated by non-thermal, mechanical means; through shear mixing of fuel and oxidizer under a shock compressive load. Since the ARM preparatory technique allows for tuning multiple powder attributes such as fuel/oxidizer of interest, their degree of mixing, amount of oxidizer, particle porosity, among others, a parametric study is crucial in identifying suitable traits for fast reactions …

Multiphysics Modeling And Experimental Validation Of High-Strength Steel In Laser Powder Bed Fusion Process, M. Rangapuram, S. Babalola, J. W. Newkirk, L. N. Bartlett, F. W. Liou, K. Chandrashekhara, Stephen R. Cluff

Materials Science and Engineering Faculty Research & Creative Works

Laser powder bed fusion (LPBF) is a subset of the additive manufacturing process in which a laser beam selectively joins the metal powder into a desired part in a sequential layer process. Owing to its complex nature of rapid heating and cooling of the melt pool, there is a need to understand the melt pool behavior and its effects on the final manufactured part. a densely packed powder bed is highly desirable for fabricating a superior part using the LPBF process. in this work, discrete element model was used to generate powder beds with realistic powder properties and various factors …

Biaxial Strain Tuned Upconversion Photoluminescence Of Monolayer Ws2, Shrawan Roy, Xiaodong Yang, Jie Gao

Mechanical and Aerospace Engineering Faculty Research & Creative Works

Monolayer tungsten disulfide (1L-WS₂) is a direct bandgap atomic-layered semiconductor material with strain tunable optical and optoelectronic properties among the monolayer transition metal dichalcogenides (1L-TMDs). Here, we demonstrate biaxial strain tuned up conversion photoluminescence (UPL) from exfoliated 1L-WS₂ flakes transferred on a flexible polycarbonate cruciform substrate. When the biaxial strain applied to 1L-WS₂ increases from 0 to 0.51%, it is observed that the UPL peak position is redshifted by up to 60 nm% strain, while the UPL intensity exhibits exponential growth with the up-conversion energy difference varying from − 303 to − 120 meV. The measured …

Uncovering Upconversion Photoluminescence In Layered Pbi2 Above Room Temperature, Sharad Ambardar, Xiaodong Yang, Jie Gao

Mechanical and Aerospace Engineering Faculty Research & Creative Works

As a van der Waals (vdW) layered semiconductor material, lead iodide (PbI2) possessing a direct bandgap with strong photoluminescence emission in visible range has gained wide attention in applications of photonic and optoelectronic devices. Here, up conversion photoluminescence (UPL) in exfoliated PbI2 flakes is demonstrated at room temperature and elevated temperatures. The linear power dependence of UPL emission with 532 nm excitation suggests the one-photon involved multiphonon-assisted UPL emission process, which is revealed by the temperature-dependent UPL emission measurement. Meanwhile, the nonlinear power dependence of UPL emission with 561 nm excitation indicates the transition of UPL emission mechanism from linear …

Iterative Correction Of Robotic Grinding Using Spatial Feedback For Precision Applications, Philip A. Olubodun, Joseph D. Fischer, Douglas A. Bristow

Mechanical and Aerospace Engineering Faculty Research & Creative Works

Since the advent of robots, many tasks that were originally performed by humans have now been tasked to industrial robots. From a manufacturing standpoint, robots have primarily been used in pick-and-place or other non-machining operations that require high repeatability. However, with the increasing availability of CAD/CAM software and the development of high-precision metrology, comes the opportunity to integrate robots into a wider variety of manufacturing processes through the use of feedback control. One such machining operation that is being explored is precision grinding of metal parts. Most other work in this area has focused on force regulation to improve grind …

A Feedforward Kinematic Error Controller With An Angular Positioning Deviations Model For Backlash Compensation Of Industrial Robots, Mitchell R. Woodside, Tian Hao Cui, John Emelko, Soichi Ibaraki, Robert G. Landers, Douglas A. Bristow

Mechanical and Aerospace Engineering Faculty Research & Creative Works

The accuracy of industrial robots, typically on the order of several millimeters, has inhibited their adoption in advanced aerospace manufacturing applications, such as robotic machining. For this reason, extensive investigations have been conducted to develop solutions to improve their accuracy. These solutions can be categorized into offline and online compensation strategies, both of which have advantages and limitations. Offline compensation has been shown to improve the accuracy of industrial robots by two to three orders of magnitude; however, the sensitivity of these solutions to environmental changes and external disturbances can degrade their performance. In contrast, online compensation, which utilizes real-time …

Wind Shear Induced Tornados With Full Thermodynamic Effects, Kakkattukuzhy M. Isaac

Mechanical and Aerospace Engineering Faculty Research & Creative Works

Wind shear-induced Tornado behavior under full thermodynamic formulation using the ideal gas equation is investigated. The computational fluid dynamics model that includes the energy equation is used to obtain solutions of a natural-scale tornado under pressure and temperature boundary conditions that account for their variations vs. altitude in the atmosphere. The present results show significant differences from previous isothermal simulations and those using the Boussinesq model for buoyancy under small temperature differences. Results show that the downdraft observed in tornados originating from supercells, often accompanied by rain and hail can be captured using the present model that simulates the genesis, …

Impact-Free Gaits For Planar Bipeds: Changing Walking Speed And Gait, Aakash Khandelwal, Nilay Kant, Ranjan Mukherjee

Mechanical and Aerospace Engineering Faculty Research & Creative Works

The problems of changing the walking speed and stride length of impact-free gaits for point-foot planar bipeds are addressed. The impact-free gaits are designed using an approach developed in prior work. It is shown that the impulse controlled Poincaré map (ICPM) approach can be modified to transition between orbits defining gaits with different walking speeds, and the continuous controller can be changed during the swing phase to transition between gaits that have distinct stride lengths. The effectiveness of the approaches is demonstrated using simulations carried out on a five-link biped.

Interlaboratory Study Of The Operational Stability Of Automated Sorption Balances, Samuel L. Zelinka, Samuel V. Glass, Eleanor Q. D. Lazarcik, Emil E. Thybring, Michael Altgen, Lauri Rautkari, Simon Curling, Jinzhen Cao, Yujiao Wang, Tina Kunniger, Gustav Nystrom, Christopher Hubert Dreimol, Ingo Burgert, Mohd Khairun Anwar Uyup, Tumirah Khadiran, Mark G. Roper, Darren P. Broom, Matthew Schwarzkopf, Arief Yudhanto, Mohammad Subah, Gilles Lubineau, Maria Fredriksson, Marcin Strojecki, Wieslaw Olek, Jerzy Majka, Nanna Bjerregaard Pedersen, Daniel J. Burnett, Armando R. Garcia, Els Verdonck, Frieder Dreisbach, Louis Waguespack, Jennifer Schott, Luis G. Esteban, Alberto Garcia-Iruela, Thibaut Colinart, Romain Remond, Brahim Mazian, Patrick Perre, Lukas Emmerich, Ling Li

Mechanical and Aerospace Engineering Faculty Research & Creative Works

Automated sorption balances are widely used for characterizing the interaction of water vapor with hygroscopic materials. These instruments provide an efficient way to collect sorption isotherm data and kinetic data. A typical method for defining equilibrium after a step change in relative humidity (RH) is using a particular threshold value for the rate of change in mass with time. Recent studies indicate that commonly used threshold values yield substantial errors and that further measurements are needed at extended hold times as a basis to assess the accuracy of abbreviated equilibration criteria. However, the mass measurement accuracy at extended times depends …

Kinetic Modeling Of Dust Grain Dynamics In Electrostatic Sieving, Aaron Berkhoff, Easton Ingram, Fateme Rezaei, Jeffrey Smith, David Bayless, William Schonberg, Daoru Han

Chemical and Biochemical Engineering Faculty Research & Creative Works

A new kinetic particle modeling framework was developed to investigate electrostatic transport of lunar regolith dust particles with applications to the concept of electrostatic sieving. the new approach is based on kinetic particle dynamics and includes major modules of sampling the particle size distribution, solving electric fields, and tracking motion of charged dust grains. a case study for a concept of electrostatic sieving was chosen to validate the new model. the simulation achieved similar performance of particle size classification as reported in the literature. the new model is computationally efficient (takes a few minutes on a PC-type laptop computer) so …

Boosting Predictive Accuracy Of Single Particle Models For Lithium-Ion Batteries Using Machine Learning, Emmanuel Olugbade, Jonghyun Park

Mechanical and Aerospace Engineering Faculty Research & Creative Works

The accuracy of single particle (SP) models for lithium-ion batteries at high C-rates is constrained by lithium concentration gradients in the electrolyte, which affect ionic conductivity, overpotential, and reaction rates. This study addresses these limitations using extreme gradient boosting machine learning (ML). By training our ML model with data from a comprehensive electrochemical (P2D) model and performing sensitivity analysis on key battery parameters, we enhance predictive accuracy. Compared to conventional SP and P2D models under constant current loading, our ML-based SP model achieves similar predictive accuracy to P2D, with significant improvements in computational efficiency. Additionally, the ML-based SP model demonstrates …

Ground Vacuum Facility To Simulate Low Earth Orbit Plasma Environment, Emmanuel Kofi Asuako Wie-Addo, Jacob Ortega, Daoru Han

Mechanical and Aerospace Engineering Faculty Research & Creative Works

This paper presents a large vacuum facility (6-ft-diameter, 10-ft-long chamber) equipped with a magnetic filter-type low Earth orbit plasma source. A recommended operating envelope for the plasma source was established through single-point measurements by varying the discharge currents of the plasma source and the gas flow rates. A three-axis translation stage was fabricated and tested with 2D and 3D scans of the simulated plasma environment. The measured plasma density during this study was between 1:63 x 1012 − 3:1 x 1012 m−3 for the electrons and 7:54 x 1012 − 1:82 x 1013 m−3 for the ions, whereas the electron …

Surrogate-Based Multidisciplinary Optimization For The Takeoff Trajectory Design Of Electric Drones, Samuel Sisk, Xiaosong Du

Mechanical and Aerospace Engineering Faculty Research & Creative Works

Electric vertical takeoff and landing (eVTOL) aircraft attract attention due to their unique characteristics of reduced noise, moderate pollutant emission, and lowered operating cost. However, the benefits of electric vehicles, including eVTOL aircraft, are critically challenged by the energy density of batteries, which prohibit long-distance tasks and broader applications. Since the takeoff process of eVTOL aircraft demands excessive energy and couples multiple subsystems (such as aerodynamics and propulsion), multidisciplinary analysis and optimization (MDAO) become essential. Conventional MDAO, however, iteratively evaluates high-fidelity simulation models, making the whole process computationally intensive. Surrogates, in lieu of simulation models, empower efficient MDAO with the …

A Robust Recurrent Neural Networks-Based Surrogate Model For Thermal History And Melt Pool Characteristics In Directed Energy Deposition, Sung Heng Wu, Usman Tariq, Ranjit Joy, Muhammad Arif Mahmood, Asad Waqar Malik, Frank Liou

Mechanical and Aerospace Engineering Faculty Research & Creative Works

In directed energy deposition (DED), accurately controlling and predicting melt pool characteristics is essential for ensuring desired material qualities and geometric accuracies. This paper introduces a robust surrogate model based on recurrent neural network (RNN) architectures—Long Short-Term Memory (LSTM), Bidirectional LSTM (Bi-LSTM), and Gated Recurrent Unit (GRU). Leveraging a time series dataset from multi-physics simulations and a three-factor, three-level experimental design, the model accurately predicts melt pool peak temperatures, lengths, widths, and depths under varying conditions. RNN algorithms, particularly Bi-LSTM, demonstrate high predictive accuracy, with an R-square of 0.983 for melt pool peak temperatures. For melt pool geometry, the GRU-based …