Bulletin of the Polytechnic Institute of Iași. Machine constructions Section Feed

Reliability-Screened Human Evaluation of Four Conversational AI Platforms: Chatgpt, Perplexity, Claude and Gemini

Mon, 21 Sep 2026 00:00:00 GMT

Human evaluation of large language models is widely used but its reliability is rarely verified. This study evaluated four conversational platforms, ChatGPT 5, Perplexity, Claude Sonnet 4 and Gemini 2.5 Flash, accessed on 20 August 2025. Twenty prompts covering factual, creative, technical and ethical-critical tasks were submitted identically to all four systems. Ten independent evaluators ranked each set of four responses on four criteria, yielding 800 forced-ranking blocks. Concordance was screened before any comparison. Only the factual set reached agreement above chance (Kendall’s W = 0.0379; mean pairwise agreement 0.561); the creative, technical and ethical-critical sets were statistically indistinguishable from random ranking. On the factual set Perplexity and ChatGPT outranked Claude and Gemini (Friedman chi-square = 22.30, p < 0.001), an effect carried almost entirely by response structure. Three of four dimensions in routine use were therefore unusable as measurements. Reliability screening should precede interpretation in human evaluation of language models.

Hybrid Manufacturing - Compatibility Between CNC Machining and Additive Manufacturing

Mon, 21 Sep 2026 00:00:00 GMT

Hybrid manufacturing can be understood as the integration of additive and subtractive manufacturing processes to obtain parts that benefits from the advantages of both technological approaches. A hybrid process does not mean the simple successive application of two technologies, but the design of the part in relation to the requirements of both processes. A virtual case study made in Autodesk Fusion on a stepped cylindrical shaft is presented and the component is analyzed through CAD modeling, CAM preparation, CNC, Additive Manufacturing (AM) and AM-CNC technological scenarios, and static tensile simulation. This approach is relevant to the current requirements of the industry, where the integration of digital tools in design, simulation, assisted manufacturing and optimization of technological processes is increasingly sought.

Locomotion Mechanisms for Mobile Robots: Classification, Kinematic-Dynamic Modeling, and Motion Control – A Review

Mon, 21 Sep 2026 00:00:00 GMT

The paper presents a comprehensive analysis of mobility and locomotion mechanisms for mobile robots, emphasizing their role in dynamic, human-shared environments. It explores the evolution from traditional Automated Guided Vehicles to autonomous platforms capable of safe interaction with humans, highlighting the importance of mobility as a multidisciplinary concept integrating mechanical design, control, and safety compliance. A detailed classification of locomotion systems is provided, including differential drive, Ackermann steering, omnidirectional wheels, tracked platforms, legged robots, and hybrid configurations, with a focus on their advantages, limitations, and application domains. Furthermore, the study addresses kinematic and dynamic modeling principles essential for accurate motion prediction and control, followed by an overview of advanced motion control strategies such as nonlinear controllers, model predictive control, and learning-based approaches. Performance criteria related to maneuverability, energy efficiency, stability, and compliance with international safety standards are discussed.

Conceptual Digital Twin Architecture for Integrity of HDPE Butt Fusion Welds

Mon, 21 Sep 2026 00:00:00 GMT

High-density polyethylene (HDPE) pipelines are extensively used in water and gas distribution systems due to their excellent corrosion resistance, durability, flexibility, and low maintenance requirements. The long-term reliability of these systems depends primarily on the quality of butt fusion welded joints, which represent the most critical locations in terms of structural integrity. In current practice, the quality assessment of HDPE welded joints is based on monitoring welding process parameters, visual inspection, mechanical testing, and non-destructive testing (NDT) methods. Although each of these approaches provides valuable information regarding weld quality, the resulting data are generated at different stages of the weld life cycle and are typically stored and studied independently. The present paper proposes a conceptual Digital Twin architecture for the integrity assessment of HDPE butt fusion welds, based on the correlation of welding process parameters, characteristic defects, and inspection results within a Knowledge Graph model to support decision-making.

Critical Synthesis of CNC Wood Machining Research in the Industry 5.0 Era

Mon, 21 Sep 2026 00:00:00 GMT

This paper reviews computer numerical control wood machining research published between 2020 and 2025, exploring how to choose cutting parameters so that surface quality, energy use, tool wear, and environmental impact are balanced. Wood’s anisotropy, hygroscopicity, and biological variability make it so different, even for the same species or board. Two clear shifts have emerged. Optimization work is moving from one-factor-at-a-time experiments toward multi-objective methods built on neural networks, genetic algorithms, and hybrid artificial intelligence schemes. In parallel, edge computing and predictive maintenance are bringing real-time decision-making down to the machine itself. Two gaps also stand out: few studies test sensors and artificial intelligence models under dusty, high-vibration conditions, and existing digital twins remain static and ignore the internal stresses, knots, and grain irregularities revealed during cutting. The paper closes by arguing for a total-process view of sustainability, optimizing surface quality, dust emissions, auxiliary energy use, and tool life simultaneously.

Kinematic Modeling of a Small-Tooth-Difference Planetary Involute Reducer

Mon, 21 Sep 2026 00:00:00 GMT

This paper develops a general matrix-based kinematic model of a small-tooth-difference planetary involute reducer operating at the minimum tooth-number difference. The mechanism comprises a fixed internally toothed central gear, two identical externally toothed satellites mounted on oppositely phased eccentric supports of the input shaft, an intermediate ball coupling between the satellites, and a second ball coupling transmitting motion to the coaxial output shaft. The formulation combines generalized coordinates, an angular constraint matrix, a transmission vector, and block-matrix representations of satellite orbital and rigid-body motion. The satellites undergo plane-parallel motion combining circular orbital translation with slow self-rotation, the latter constituting the useful reduced motion transmitted to the output. The ball couplings are treated only through geometric compatibility conditions. A parametric study for several tooth-number differences evaluates transmission-ratio sensitivity, eccentricity variation, reconfiguration using interchangeable eccentric bushings, and numerical consistency of the proposed matrix model over one input revolution under ideal rigid-body assumptions.

Technical Review on Unmanned Aerial Vehicle (UAV) Technologies

Mon, 21 Sep 2026 00:00:00 GMT

Unmanned Aerial Vehicles (UAVs), commonly known as drones, have transitioned from specialized military assets to commonly found tools across agriculture, logistics, surveillance, and disaster management. As UAV technology advances rapidly, establishing a clear framework for understanding their diverse capabilities becomes essential. This paper presents a comprehensive literature review aimed at synthesizing foundational UAV concepts and establishing a standardized classification. Authors examine the core technical characteristics, operational principles, and historical evolution that define modern drone systems. Furthermore, this study categorizes UAVs based on operational parameters, including aerodynamic design (fixed-wing, rotary-wing, and hybrid systems), weight, range, altitude, and autonomy levels. By consolidating fragmented data from contemporary research, this paper provides a streamlined reference model for classification. Ultimately, this review bridges the gap between basic UAV mechanics and complex operational frameworks, offering researchers and industry professionals a foundational guide to navigate the current drone landscape and identify future technological advancements.

Analytical Modeling and Advanced Experimental Analysis of Natural Frequencies in Printed Circuit Boards

Mon, 21 Sep 2026 00:00:00 GMT

This paper presents a combined theoretical, numerical and experimental investigation of the vibrational behaviour of automotive printed circuit boards (PCBs), focusing on natural frequencies, mode shapes and the spatial distribution of the dynamic response. Two classical plate formulations, Kirchhoff–Love and Reissner–Mindlin theory, are implemented in MATLAB to predict modal parameters for two automotive electronic control unit (ECU) case studies (Audi A4 and Volkswagen Polo). Because analytical models cannot fully reproduce the behaviour of populated boards, experimental validation was performed using Laser Doppler Vibrometry, including full-field out-of-plane scanning, three-dimensional (SMART 3D-Scan) vibration analysis and single-point 3D sensing, in accordance with ISO 16750-3. The correlation between numerical and experimental results supports early-stage design assessment, optimized component placement and improved solder-joint reliability, providing a consistent methodology for evaluating PCB dynamic performance under realistic automotive operating conditions.

A Comparative Study on the Optimization of EMG Systems for Controlling Myoelectric Prosthetics

Mon, 21 Sep 2026 00:00:00 GMT

This paper presents a comparative analysis of two analog modules for acquiring biological electrical signals, the AD8232 and the SEN0240, with a view to their use in an experimental myoelectric control system for an upper limb prosthesis. The study stems from the need to identify a module capable of autonomously distinguishing between muscle rest and voluntary contraction, so that the acquired signal can be used to control a shape-memory (alloy) actuator via a microcontroller and an N-MOSFET transistor. The comparison focuses on criteria such as signal stability, ease of integration with Arduino, sensitivity to artifacts, and practical utility for controlling an actuator. The experimental results indicate that the SEN0240 offers a more direct alignment with the requirements of a myoelectric application, while the AD8232 can be used to detect bioelectric signals but has some drawbacks. Thus, this paper justifies the selection of the SEN0240 module for the continued development of the experimental system for controlling the actuator using myoelectric signals.

Structural Response of a Protective Brake System Shield Under Thermal Load

Mon, 21 Sep 2026 00:00:00 GMT

Protective sheet-metal components located in the engine compartment are exposed to significant temperature variations generated by the powertrain and adjacent heat sources. These thermal loads can affect stress distribution, deformation behavior, and the performance of bolted joints. This study investigates the thermo-mechanical response of a steel protective shield mounted on a valve block support using an M6 class 8.8 bolted connection. A three-dimensional finite element model was developed to evaluate the assembly under an initial bolt preload of 8 kN and three thermal cycles between −40°C and +120°C. The analysis focused on equivalent von Mises stresses, total deformation, contact pressure, and bolt clamp force evolution. The results indicate that thermal cycling causes periodic variations in preload and localized stress concentrations, particularly around the mounting hole and geometric discontinuities. The maximum bolt force remained below the ISO 898-1 proof load limit, confirming elastic behavior of the fastener. Furthermore, no significant preload relaxation was observed, demonstrating that the assembly maintains adequate clamping performance and structural integrity under the investigated thermal conditions.

Current Trends in the Development of Cold Hardening Tools

Tue, 30 Jun 2026 00:00:00 GMT

Hardening by cold plastic deformation represents an essential technology for enhancing the performance of components operating under demanding conditions, ensuring superior resistance to wear, fatigue, and corrosion. Specialized hardening tools, designed for a diverse range of applications—internal, external, flat, or threaded surfaces—play a decisive role in the efficiency of this process. The complexity of these tools stems from the necessity to ensure dimensional precision, the uniformity of the hardened layer, and adaptability to various component geometries. This paper presents an analysis of existing constructive solutions for hardening tools, correlated with modern technological requirements, highlighting trends in arhitecture optimization and their prospective applications in contemporary industry. The study proposes a synthesis of the current state of the art regarding hardening tools, utilizing an idea diagram for a clear and accessible organization of the various possible solutions.

Microwave Heating and Physical Phenomena of Microwave Heating of Materials –A Review

Tue, 30 Jun 2026 00:00:00 GMT

Microwave heating is a major research direction in advanced processing technologies, intensively studied for its superior energy efficiency and improved process performance compared to classical heating methods. The paper contributes to the understanding of the physical mechanisms underlying the interaction between the electromagnetic field and matter and verifies the current state of applications of this technology. The paper deals with the physical phenomena underlying microwave heating, characterizes relevant material properties, such as dielectric permittivity, loss factor and electromagnetic field distribution, and classifies materials according to their response to the microwave field. The research results certify that microwave heating allows for a rapid and volumetric energy transfer, which leads to significant reductions in processing time and measurable improvements in energy efficiency, as well as in the physical and structural properties of materials. From the perspective of strategic forecasting, bibliometric and patent analyses, an accelerated adoption of microwave heating in industrial and research applications is indicated. Microwave heating is a versatile and high-potential technology with diverse application areas.

Development of an Experimental Model for Conventional Electrochemical Machining

Tue, 30 Jun 2026 00:00:00 GMT

For machining surfaces with high hardness, a good method with high processing efficiency is conventional electrochemical machining (C-ECM). In this article, we present the design and testing of an experimental model under laboratory conditions, using equipment and electrolyte solutions in accordance with the process requirements. After the experiments are completed, the next stage involves validating the model by evaluating its performance in relation to various technological parameters, such as the type and concentration of the electrolyte and the applied voltage, as well as by analyzing the quality of the resulting surfaces. The ultimate goal of this work is to verify the proper functioning of the conventional electrochemical machining system and to identify the most efficient ways to use this technology for machining complex materials. In conclusion, the developed experimental model represents an efficient and versatile tool for advanced material machining, supporting both scientific research and future industrial optimization and implementation.

Integration of Sustainability in Road Freight Transport Companies in the Republic of Moldova: Between Polycrisis, Compliance and Competitiveness

Tue, 30 Jun 2026 00:00:00 GMT

The last years of sustainable transition have been marked by the imprint of geopolitical, energy, climate and socio-economic crises. Road freight transport remains one of the sectors with a significant impact on the environment on the one hand, on the other hand it is impacted by the global polycrisis. In this context of economic survival, the probability that alignment with climate objectives will become less and less achievable is increasing. A gap appears between the need for sustainable transition and the real capacity of companies.The article starts from the idea that sustainability in road freight transport can no longer be analysed separately from the broader context in which companies operate. Currently, the entrepreneurial environment is marked by an economic, energy, geopolitical and climate polycrisis, which increases uncertainty and makes many companies focus their attention, primarily, on survival, operational continuity and cost control. In these conditions, climate objectives risk being relegated to the background, even if the road freight transport sector remains one of the areas with a significant impact on the environment, through CO₂ emissions, high energy consumption, noise, congestion and infrastructure wear and tear.The purpose of the research is to analyse the capacity of road freight transport companies in the Republic of Moldova to adapt to the requirements of sustainable transition.

Strategic Forecasting Methods for Micro and Nanotechnologies

Tue, 30 Jun 2026 00:00:00 GMT

Strategic forecasting is essential for decision-making in the knowledge-based economy, where rapid technological change demands structured anticipation. This paper critically reviews the main forecasting methods used for micro and nanotechnologies and evaluates their theoretical basis, strengths, and limitations. A systematic literature review covers quantitative approaches, including time-series extrapolation, S-curve diffusion models, bibliometric analysis, and patent-landscape analysis, as well as qualitative methods such as the Delphi technique, scenario planning, strategic roadmapping, and expert judgment. Hybrid and AI-enhanced frameworks are also examined. The review shows that quantitative methods are objective but largely retrospective, whereas qualitative methods are flexible but vulnerable to bias and limited validation. Hybrid approaches appear most promising, yet they remain unevenly implemented. The identified methodological gaps support the development of an integrated, adaptive forecasting framework for emerging technologies.

Perturbation Analysis of Kinematic Trajectories Using Fractal Dynamics for the Detection of Experimental Errors in Patients With Rheumatoid Arthritis

Tue, 30 Jun 2026 00:00:00 GMT

In the context of finger biomechanics and the development of biomechatronic systems, joint motion trajectories exhibit nonlinear variations and local instabilities that cannot always be adequately described by classical deterministic models; therefore, this study proposes a mathematical framework based on the introduction of controlled fractal perturbations generated through Mandelbrot-type iterative dynamics. The time series of MCP, PIP, and DIP joint angles were processed using a graphical user interface developed in MATLAB, designed for the kinematic analysis of the index finger, modeled as a serial kinematic chain with three degrees of freedom, allowing the reconstruction of the distal phalanx trajectory and its comparison with a reference trajectory derived from healthy-subject data. Through the automatic computation of Cartesian errors, normal deviations, and discrepancies in the joint space, the proposed method highlights motion variability and local instabilities, contributing to data validation and to the optimization of objective biomechanical assessment in rheumatoid arthritis.

An Experimental Evaluation of the Operational Capabilities of A Diamond-Tipped Cold Burnishing Tool

Tue, 30 Jun 2026 00:00:00 GMT

Cold burnishing is a finishing process that improves the mechanical properties of metal components without removing material. It works by compressing and smoothing surface irregularities, inducing residual compressive stresses and increasing surface hardness, which enhances durability and surface quality. This paper focuses on slide diamond burnishing (SDB), a method that uses a diamond tool with high hardness, low friction, and good thermal conductivity to achieve superior finishes, even on difficult materials. The study investigates the tool’s performance on external cylindrical surfaces using minimum quantity lubrication (MQL), while varying key parameters such as spindle speed, feed rate, and working depth. A comparison between conventional turning and diamond burnishing is made using roughness parameters Ra and Rz. Results show a significant reduction in surface roughness and the achievement of near mirror-like finishes, demonstrating that diamond burnishing is an effective and sustainable alternative to grinding and polishing.

A Comprehensive Review of Bearing Current Phenomena in Electric Vehicle Drive Systems

Tue, 30 Jun 2026 00:00:00 GMT

The increasing adoption of electric vehicles (EVs) has brought renewed attention to the reliability and durability of drivetrain components, particularly electric motor bearings. Bearing currents induced by high-frequency voltage potentials from inverter-fed motors have become a significant source of premature bearing failure in electric powertrains. Rolling bearings represent essential elements in electric machines, having a direct influence on their efficiency, operational stability, and lifespan. The rapid expansion of pulse width modulation (PWM) inverter-based drive systems has introduced new electrical stress factors within motors, among which bearing currents have become a critical concern. These parasitic currents can pass through the bearing contacts, causing surface deterioration such as pitting and fluting. Over time, this degradation accelerates wear, diminishes the operational lifetime of the motor, and may ultimately lead to unexpected system failure Current research on bearing current-related degradation primarily focuses on three main aspects: the physical mechanisms responsible for current generation, the key influencing parameters governing these phenomena, and the development of equivalent circuit models to describe current paths through the motor system. This paper presents a comprehensive analysis of the electrical currents passing through bearings used in electric vehicles. The mechanisms of current generation, including common-mode voltage, capacitive coupling, and electromagnetic induction, are examined in depth. Furthermore, the effects of various operating conditions such as motor speed, load, inverter switching frequency, and temperature on the amplitude and frequency of bearing currents are evaluated. The resulting electrical and thermal stresses that accelerate bearing degradation are discussed based on both experimental results and simulation models. Current mitigation strategies, including the use of insulated bearings, grounding techniques, and advanced inverter control methods, are assessed in terms of their effectiveness and limitations. Finally, the paper outlines future research directions aimed at developing integrated design and diagnostic approaches to predict, monitor, and minimize bearing current effects, thereby enhancing the reliability and lifespan of electric vehicle drivetrains.

Finite Element Analysis of Cylindrical Specimens in Tensile Stress Under the Influence of Stress Concentrators

Tue, 30 Jun 2026 00:00:00 GMT

This paper presents a finite element analysis (FEA) of cylindrical specimens subjected to a tensile load, focusing on the influence of stress concentrators and geometric variations. Three types of specimens were studied: a reference specimen with a diameter of 8 mm, a specimen with a stress concentrator located at the midpoint of the gauge length, and a specimen with a diameter increased to 10 mm. The simulations were performed in ANSYS under linear static conditions. The numerical results indicate that the maximum equivalent elastic strain reached values of approximately 0.0031 for the reference specimen and 0.0039 for the specimen with a stress concentrator, highlighting a significant increase in local strain. The maximum values of the equivalent stress obtained were 659.92 MPa for the 8 mm specimen, 782.27 MPa for the specimen with a stress concentrator, and 627.36 MPa for the 10 mm specimen. These results clearly demonstrate stress amplification in the presence of geometric variation of the sample’s cross section. The study confirms that stress concentrators significantly influence stress distribution and represent critical points for fracture initiation in tensile specimens.

CAD/CAM Approaches for Machining A Mechanical Part on a Vertical CNC Machining Center

Tue, 30 Jun 2026 00:00:00 GMT

The study presents the analysis and optimization of the CAM technology applied to the machining of a flange-type part on a vertical CNC machining center. The study aims to reduce machining time while maintaining the final geometry and the required machining accuracy of the part. The analyzed component has a medium level of geometric complexity and requires machining in two setups. The methodology included part modelling, definition of the machining operation sequence, cutting parameter setting, toolpath simulation, and machining time determination. Based on the assessment of the initial CAM variant, two machining operations with a dominant influence on the overall process duration were identified, for which alternative technological solutions and toolpath generation strategies were comparatively analyzed. The results showed a reduction in total machining time of approximately 9.37%, confirming the effectiveness of optimizing critical operations in improving CNC machining productivity.