Open Access

Article

Article ID: 4103

Thermal confinement effects in laser-polished AM 316L slots: The role of geometry in internal surface response

by Aswin Karkadakattil

Mechanical Engineering Advances, Vol.4, No.1, 2026;

Laser polishing (LP) is widely employed to enhance the surface quality of additively manufactured (AM) metals; however, its behaviour within deep or confined internal geometries remains insufficiently understood. Many high-performance AM components, such as biomedical implants, turbine cooling channels, and metal microfluidic systems, incorporate narrow internal features where heat-transfer conditions differ significantly from open surfaces. In this study, laser powder bed fusion (LPBF)-fabricated 316L stainless steel specimens containing ~10 mm deep slots with widths ranging from 1 to 5 mm were subjected to laser polishing using a continuous-wave fibre laser (power: 80–120 W, scan speed: 450–750 mm/s, spot size: ~80–100 µm, ~60–70% track overlap, single-pass strategy). The influence of internal geometric confinement on microstructural evolution and mechanical response was systematically investigated. A pronounced depth-dependent microhardness gradient was observed along the slot wall, with hardness decreasing from approximately 270 HV in the lower region to ~210 HV near the slot opening, with more significant gradients in narrower geometries. Quantitative grain-size analysis revealed finer grains (~8–12 µm) in the lower region and coarser grains (~18–25 µm) toward the upper region, indicating progressive grain coarsening with increasing height. These variations are attributed to geometry-dependent thermal boundary conditions, where enhanced conductive coupling to the bulk substrate in the lower region promotes higher cooling rates, while reduced thermal extraction near the slot opening results in slower solidification. The results provide clear experimental evidence that internal geometric confinement can significantly influence microstructure–property evolution during laser polishing, even under constant processing parameters. This study highlights the importance of incorporating geometric effects into post-processing strategies for AM components and offers practical insights for achieving more predictable and uniform mechanical performance in confined internal features.

Open Access

Article

Article ID: 4125

PDFP-enhanced physics-informed neural networks for solving higher-order PDEs: Application to engineering beam problems

by Shahbaz Ahmad, Neha Zaman, Muhammad Asif, Muhammad Qasim, Aaqib Hussain Shah, Anam Shahzadi, Faizan Arshid, Muhammad Israr

Mechanical Engineering Advances, Vol.4, No.1, 2026;

This study introduces a modern technique for solving beam equations on elastic foundations, applied to Euler–Bernoulli and Timoshenko beams based on the Winkler foundation. Conventional Physics-Informed Neural Networks (PINNs) have faced major challenges in handling large space-time domains, even when closed-form solutions exist. To address these challenges, we introduced a preconditioned PINNs loss function that incorporates prior knowledge from similar problems, thereby enhancing both productivity and accuracy. This novel method improves the generalization capability of PINNs in structural engineering, specifically for beam dynamics on elastic foundations. The effectiveness of the method is shown through numerical simulations on Euler–Bernoulli beams and extended domains for Timoshenko beams. Comparing with state-of-the-art PINNs shows that our method speeds up the convergence and precisely describes the behavior of the system, surpassing current strategies under the L2-norm metric. Besides, we examine the weight and weight loss plots, and present 3D visualizations of the best weight configurations. Further, we describe the superior performance of the proposed method. The present study explores the application of Preconditioned Devidon–Fletcher–Powell (PDFP)-enhanced Physics-Informed Neural Networks (PINNs) for solving higher-order partial differential equations, with particular emphasis on their implementation in engineering beam problems. It provides an overview of the new approach for addressing complex beam dynamics, including visualizations of the neural network architectures, PINNs convergence behavior, and representative solutions for the beam problems.

Open Access

Article

Article ID: 3942

Numerical investigation of induction hardening of stationary cylindrical steel pins with convective quenching

by Mohammad Yaghoub Abdollahzadeh Jamalabadi

Mechanical Engineering Advances, Vol.4, No.1, 2026;

This paper presents a comprehensive, fully-coupled Multi-physics finite element model for simulating the induction hardening process of stationary cylindrical steel pins, including subsequent convective cooling. The model integrates three interacting physics domains—electromagnetic induction, transient heat transfer, and metallurgical phase transformations—within an efficient two-dimensional axisymmetric formulation. Temperature-dependent material properties for all steel phases (ferrite, pearlite, austenite, martensite) and the surrounding air are implemented, and the formulation accounts for latent heat effects during phase changes. The framework employs a segregated solver approach, ensuring robust convergence between the strongly coupled electromagnetic, thermal, and phase transformation modules. The stationary configuration simplifies the computational approach while retaining high fidelity for industrial applications. The simulation predicts critical process outcomes such as transient temperature distributions, phase evolution, and the resulting spatially-graded hardness profile. It further evaluates the resultant residual stress distribution, providing insight into potential distortion and component performance. Furthermore, it serves as a predictive tool for optimizing key operational parameters, including induction coil current frequency and magnitude, heating time, and forced convective cooling intensity. Model predictions for case depth versus applied power show strong agreement with experimental measurements, validating the framework. The validated model demonstrates its utility as a virtual design platform, reducing the need for costly experimental trials. This integrated model provides a complete and practical computational framework for designing, analyzing, and optimizing stationary induction hardening processes to achieve targeted hardness depths, improve energy efficiency, and ensure consistent product quality in manufacturing.

Open Access

Article

Article ID: 3338

A data-driven approach for predicting stress intensity factors of a single-edge cracked plate with a random polygon-shaped void

by Mehrad Zargar Ershadi, Saeid Nickabadi, Majid Askari Sayar, Alireza Alidoust, Reza Ansari

Mechanical Engineering Advances, Vol.3, No.4, 2025;

This study represents a data-driven framework for predicting mode I (K_I) and mode II (K_II) Stress Intensity Factors (SIFs) in single-edge cracked plates with central polygon-shaped voids. Finite element simulations were conducted in Abaqus software to generate a dataset by varying key parameters, including the polygon’s number of vertices, angle, average radius, and crack length. Two machine learning models were employed to analyze the dataset created by the finite element method: Group Method of Data Handling (GMDH) networks and an Artificial Neural Network (ANN). The GMDH networks were optimized using the least squares method and the Root Mean Squared Error (RMSE) criteria, while the ANN, designed as a feedforward fully connected network, was trained with the backpropagation algorithm and the gradient descent optimization technique using TensorFlow and Keras libraries. The ANN demonstrated exceptional accuracy, with a R² value exceeding 0.99 for K_I predictions and 0.98 for K_II, significantly outperforming GMDH models, particularly in capturing the nonlinear behavior of K_II.

Open Access

Article

Article ID: 3350

Enhancing learning experience of manufacturing through metaverse—development and demonstration

by Dhaval Anadkat, Mukhtar Sama, Amit Sata

Mechanical Engineering Advances, Vol.3, No.4, 2025;

Integration of disruptive technologies like Metaverse, AR/VR, Digital Twin, etc., with manufacturing education is revolutionising the learning experience and bridges the gap between traditional and hands-on practice. This paper focuses on the development and demonstration of an immersive interactive environment in manufacturing processes, specifically in the Vertical Centrifugal Casting (VCC) and Tungsten Inert Gas (TIG) welding. Integration of immersive environment with traditional manufacturing processes allows the learners to interact with a real-physical setup, real-world scenarios, and optimize the process in a virtual risk-free environment from a distant place. The proposed system demonstrates the real-time monitoring and controlling of data using IoT, data collection using DAQ, as well as digital twinning of the system for improved learning and operational efficiency. This integration allows users to monitor and operate different process parameters, like tuning of VCC, mold rotation, metal pouring, and practicing metal pouring, as well as adjusting the parameters in real time. This study highlights the significance of imparting manufacturing education through immersive technologies, resulting in improved experiential learning, safer practice opportunities, and enhanced student preparedness for Industry 4.0 environments, as validated through system validation. This study revolutionizes conventional industrial education with Metaverse applications, facilitating a more engaging, accessible, and efficacious training paradigm aligned with the Sustainable Development Goals (SDGs) 4 (Quality Education) and 9 (Industry, Innovation and Infrastructure).

Open Access

Article

Article ID: 3582

A biomimetic design framework for structurally optimized lifting beams using graded geometry and internal ribs

by Jacob Nagler

Mechanical Engineering Advances, Vol.3, No.4, 2025;

This paper presents a comprehensive biomimetic design framework for lifting beams that couples an exponentially graded outer shell with an internal dendritic rib network to maximize stiffness-to-weight performance while meeting serviceability and safety constraints. A multi-fidelity modelling chain is developed: a variable-section Euler–Bernoulli model for rapid sizing, a shear-corrected Timoshenko formulation for regimes in which transverse shear is significant, and a 2D arch-frame finite-element model that resolves local rib–shell interactions and stress concentrations. A density-based SIMP topology-optimization workflow is integrated with parametric regression to extract manufactural rib trajectories, and analytical closed-form expressions are derived for second moment contributions of graded shells and discrete ribs. Morphologically, the final optimized topology converges to a graded cellular arch frame resembling a chiropteran bat-wing, where the internal dendritic lattice functions as a variable-depth Warren truss to effectively decouple shear flow from the bending-dominated outer shell. Extensive analytical investigations: parametric sweeps, one-factor-at-a-time sensitivity, and first-order uncertainty propagation, demonstrate that graded thickness and load-path aligned ribs increase the section modulus and reduce peak bending demands; for representative baseline geometries and materials, the proposed topology yields ~20% reduction in peak bending stress and ~15% reduction in midspan deflection at equal mass compared with conventional solid sections. High-fidelity FEA highlights local saw-tooth stress peaks at rib roots that exceed mean analytical estimates by ≈60%, indicating the necessity of filleting, fatigue-aware detailing, and AM process control. The manuscript concludes with a rigorous experimental validation roadmap (AM prototyping, DIC, static and fatigue testing, CT/NDT) and recommends embedding uncertainty-aware surrogates and single-loop multidisciplinary optimization to ensure robust, certifiable lifting hardware under multi-source variability.