Mechanics and Advanced Technologies

A Review of Input Strategies and Model Architectures in Machine Learning Approaches for Multiaxial Fatigue Life Prediction

Pavlo Yakovchuk — 2026-06-18

This paper presents a review of machine learning methods for predicting fatigue life under multiaxial loading. The problem addressed in this work is the lack of a unified approach for accurate fatigue life prediction under complex proportional and non-proportional loading paths.
The results of the study consist in a systematic review and classification of modern machine learning approaches based on model architectures and input data representation strategies. In the latter, the analyzed methods are grouped into approaches based on engineered features, time-series data and image-based representations. The literature review covers publications from 2015 to 2025 retrieved from Scopus, Web of Science, and Google Scholar.
It is shown that prediction accuracy depends not only on the choice of machine learning algorithm, but primarily on the way the loading history is represented and integrated into the model. Physically based parameters remain effective in the case of limited experimental data, whereas deep learning methods are more suitable for large datasets and complex loading histories.

The results can be applied in engineering practice for the selection of appropriate machine learning methods for fatigue life prediction, particularly in cases involving complex loading histories, limited experimental data, or the need for interpretable and physically consistent models.

Algorithms for Transforming Three-Dimensional Finite Element Models into Models with Only CHEXA Elements

Konstantin Rudakov — 2026-06-18

There are a large number of developed finite element models of real objects based on finite elements (FE) CTETRA, as well as CHEXA with part of CPENTA. But to obtain better solutions of three-dimensional boundary value problems, it is desirable to have models based on only FE CHEXA. The problem of creating finite element models only with FE CHEXA has not yet been solved at the stage of their initial creation. Therefore, mixed models are usually built: separate parts of the FE CTETRA type and CHEXA type are "glued" into a single model (Connection Property Glued technology in the Connect command group). This can be done "manually". And, for example, the "Hex-Dominant Mesher" algorithm in Simcenter Femap does this automatically. As for existing FE models with the presence of FE CHEXA or CPENTA, they are usually completely deleted and created anew according to the new requirements for them. Such a process is time-consuming in the absence of automation.
Research objective: to create automated effective algorithms and programs for converting three-dimensional finite element models with the presence of FE CTETRA and/or CPENTA into models with only CHEXA elements.
Implementation methodology: idea, programming, testing, numerical experiment.
Research results. A formula was obtained that allows calculating the coordinates of intermediate nodes of the edges of the third direction with second-order accuracy when transforming CPENTA finite elements into CHEXA finite elements. Effective algorithms for transforming three-dimensional finite element models into models with only CHEXA finite elements, of the first or/and second order of approximation, were developed. The proposed algorithms are implemented in the C programming language, which provides maximum execution speed. They were implemented in two versions: built into the author's OKA_3D program, as well as in the form of a separate auxiliary program for servicing Simcenter Femap of all versions, starting from version 4.4. The interface is carried out through files of the .neu format. Testing was carried out.
Conclusions. Effective algorithms for transforming three-dimensional finite element models from FE CTETRA and CPENTA into a model with only CHEXA finite elements have been proposed and implemented.

The Use of Cold Extrusion with Fluid-Assisted Counterpressure for the Production of Blanking Punch Blanks Made of SHX-15 SHD Steel

Volodymyr Kaliuzhnyi — 2026-06-18

In the context of serial and mass production of products using cold deep drawing processes, the primary objective is to improve the durability of the forming tool, which can be achieved by replacing machining with cold plastic forming during the manufacture of such tools. This work employs FEM modeling using an elastic-plastic metal model to determine the parameters of cold hydroextrusion with variable counterpressure for die punch blanks made of SHX-15 SHD steel, and experiments were conducted to produce the punch blanks. For two geometric shapes of punches, the simulation determined maximum fluid back pressures of 200 and 300 MPa, which ensured the hydroextrusion of punch blanks with deformation rates of 40.8% and 50.6%, respectively. The dependencies of the extrusion forces on the displacement of the deforming tool, the forces required to remove the obtained punch blanks from the containers, and the distributions of specific forces on the deforming tool were established. The shape and dimensions of the punch blanks after extrusion, as well as the distributions of temperature and the components of stresses and strains in the deformed metal, were determined. The work hardening of the metal structure due to cold plastic deformation was evaluated based on the distribution of strain intensity. The greatest work is performed at the transition from the cylindrical deforming surface of the punch blanks to the conical support surface, which is necessary to increase the stability of the punches during stamping. Experimental studies on the back-pressure extrusion of punch blanks confirmed the results of the calculations. The results of modeling and experiments on variable-back-pressure extrusion for SHX-15 SHD steels are novel. The proposed counterpressure extrusion can be used in the manufacture of forming tools from low-ductility die and tool steels for metal forming processes.

Upper-bound estimate of reduced pressure and optimisation of scheme parameters for backward extrusion of parts with a conical cavity

Anton Hranovskyi — 2026-06-18

Backward extrusion of parts with a conical cavity is characterised by complex metal flow, sensitivity of the force regime to tool geometry and friction conditions, and a lack of generalised analytical models suitable for rapid engineering estimation.

The aim of this study is to develop a parametric computational scheme of the deformation zone in the backward extrusion of parts with a conical cavity and, on this basis, to obtain an analytical expression for reduced pressure by the upper-bound method.

The deformation zone was represented by a system of rigid triangular elements in the axial section, with separation of the bottom region and the wall reduction region. The reduced pressure was expressed as the sum of two components. Analytical expressions for the optimal scheme parameters were obtained for the compression and flow-turning zone, while for the wall region relationships were established between the subdivision parameters, the similarity coefficient, wall height, and the number of similar elements. Primary verification was performed using finite element method in QForm software package at an extrusion speed of 1 mm/s.

A generalised analytical expression for reduced pressure in the backward extrusion of parts with a conical cavity was obtained. It was shown that a specific feature of the backward process is a variable normalising quantity, which affects the shape of the plotted dependences and may lead to the appearance of gentle segments or local minima. The greatest influence on reduced pressure is exerted by the relative punch radius, especially in the region of its high values; the friction coefficient also significantly increases the level of reduced pressure. For the cases considered, the mean absolute relative error of the extrusion force was 12.6%.

The developed parametric model based on the upper-bound method is suitable for primary engineering estimation of the force regime in the backward extrusion of parts with a conical cavity within the considered parametric section.

Energy efficiency of pneumatic and electric drive systems and their integration into the Internet of Things

Oleh Levchenko — 2026-06-19

The article addresses the issue of energy efficiency of pneumatic and electric drive systems in modern industry and their integration with Internet of Things (IoT) technologies. The operational features of pneumatic systems are analyzed, along with the main causes of energy losses, including compressed air leaks, pressure drops, and inefficient parameter control. Methods for improving energy efficiency are considered, including optimization of operating modes, the use of modern equipment, and the implementation of monitoring systems. Particular attention is paid to the possibilities of integrating pneumatic networks into IoT environments, which enables continuous monitoring, predictive maintenance, and reduction of operating costs.
The paper also presents an analysis of electric drives, their advantages and disadvantages in comparison with pneumatic drives, particularly in terms of energy efficiency, accuracy, and controllability. The operational characteristics of the main types of electric motors and their areas of application in industry are discussed. Trends in the development of the pneumatic systems and electric drive markets are highlighted, as well as the influence of global factors such as digitalization and decarbonization.
As a result, it is established that the most promising direction is the development of hybrid systems that combine the advantages of pneumatic and electric drives. The integration of IoT technologies significantly improves the efficiency, reliability, and cost-effectiveness of production processes, facilitating the transition to intelligent energy-efficient manufacturing.
It is determined that in modern industry, Internet of Things (IoT) technologies play a key role by transforming approaches to production management and enhancing business efficiency. The advantages and innovations of IoT are considered, demonstrating how their implementation improves production processes and contributes to increasing the energy efficiency of pneumatic and electric drive systems.

Modeling the stress-strain state of metal during stress concentrator formation with a wedge

Serhii Karnaukh — 2026-06-18

The scientific article addresses the problem of improving the efficiency of sectional rolled product separation processes by enhancing the methods of stress concentrator formation during cold bending fracture. The relevance of the study is driven by the need to reduce energy-force expenditures, improve workpiece quality, and implement waste-free technologies in metallurgical production. The aim of the work is to optimize the parameters of the wedge tool indentation process based on an analysis of the stress‑strain state of the metal. The methodological framework of the study consists of analytical modeling using the slip line field method (Hill’s solution) for a rigid‑plastic medium. An improved algorithm for determining the contact pressure and the technological force, accounting for the geometry of the plastic zone of the displaced metal, is proposed. Special attention is paid to a comparative analysis of two approaches for approximating the free boundary of the plastic region-straight and concave. To verify the adequacy of the analytical relationships, numerical modeling was carried out in the Deform 3D environment, which enabled the investigation of the influence of the main technological parameters (wedge angle, indentation depth, loading rate) on the force and energy characteristics of the process. Additionally, experimental studies were performed using a specially designed setup, providing verification of the obtained results. It was established that the use of the model with a concave free boundary of the plastic zone yields a more accurate description of the stress‑strain state of the metal and reduces the calculated wedge indentation force compared to the traditional model. The regularities of the influence of the process parameters were identified: as the wedge angle, indentation depth, and loading rate increase, the force and deformation work rise, and there exists a rational range of their values beyond which a sharp increase in energy consumption occurs. The discrepancies between the analytical, numerical, and experimental results do not exceed acceptable engineering limits. The scientific novelty lies in the development of approaches for describing localized plastic deformation during the formation of stress concentrators and in refining the geometry of the plastic zone. The practical significance of the results lies in their applicability for optimizing wedge tool parameters, improving workpiece quality, and enhancing the efficiency of waste-free rolled product separation technologies.

Accounting for the peculiarities of abrasive machining of carbon-carbon materials and tools for its implementation

Mykhailo Molnar — 2026-06-18

Problem: Processing carbon-carbon (C/C) composites is accompanied by significant difficulties due to their quasi-brittle fracture mechanics and the tendency of fine carbon dust to clog the cutting zone when using diamond cutting tools. Existing continuous models treat the elastic cutting tool as a vibrating string, which does not allow for the microdynamic interactions between grains and bonds that are critical for understanding the mechanisms of sludge removal and process stability.

Objective: The main objective is to develop a comprehensive 8-degree-of-freedom coupled model with concentrated parameters for modeling the stochastic dynamics of abrasive processing and investigating the hypothesis of grain rotational motion as a mechanism for improved slurry removal.

Methodology: The work models a localized segment of a rope with four active abrasive grains as a discrete system with 8 degrees of freedom. The grains are considered as rigid bodies with translational and rotational compliance, connected by a steel core and fixed in a viscoelastic nickel bond. The dynamics of the system are described by coupled differential equations using the Reissner-Sagotsi rotational stiffness and Rayleigh damping formulas. The cutting process is simulated as a stochastic sequence of pulses reflecting the heterogeneity of the material.

Results: A mathematical modeling method was proposed that can investigate the rotational and translational movements of grain during cutting, which will allow finding the parameters of the cutting tool that will facilitate the removal of chips from the cutting zone.

Conclusions: The proposed model is a theoretical tool for finding the parameters of diamond cutting tools, in particular the viscoelastic properties of the matrix, to achieve a balance between material removal rate and dynamic stability.

Criteria for Evaluating the Quality of DPM Marking on Heterogeneous Symbolic Images Formed by Electrochemical Etching of Metal Surfaces

Oleksandr Vasylkivskyi — 2026-06-18

This analytical review is devoted to the development of a system of criteria for evaluating the quality of direct marking of stainless-steel products using the electrochemical etching method. It is shown that the current requirements of ISO/IEC 29158:2020, which are oriented toward two-dimensional alphanumeric barcodes, do not fully account for the peculiarities of electrochemical marking of arbitrary images (symbols, logos, pictograms).

The paper systematizes the optical criteria according to the current standard (contrast parameters) and proposes assumptions regarding their dependence on the state of the near-surface layer of stainless steels formed during electrochemical etching. Based on this, an expanded multi-level system of criteria is proposed, which includes geometric, optical, physical-mechanical, and operational parameters.

An analysis of the applicability of the criteria to the direct marking of arbitrary images was performed, and the unsuitability of parameters requiring fixed structural templates for assessing the quality of electrochemically etched symbols and logos was substantiated.

The proposed system of criteria is considered as a basis for combining standard methods for classifying direct marking symbols according to ISO/IEC 29158:2020 with phenomenological models of the electrochemical etching process and for the further development of methods for experimental validation of marking quality under industrial production conditions.

Structural and Technological Principles of Modular Design of Unmanned Aerial Vehicles

Oleksii Babich — 2026-06-18

The object of the study is the structure and architecture of quadcopter unmanned aerial vehicles (UAVs) as complex engineering systems operating under variable conditions and requiring a high level of adaptability and maintainability.

The study addresses the problem of limited flexibility and low maintenance efficiency of traditional integrated UAV designs, which complicates their modernization and increases life-cycle costs.

As a result of the research, the composition and structure of the main functional modules of a quadcopter UAV (at least seven core modules) were determined, the role of the integration (load-bearing) module was substantiated, and the types of inter-module interfaces were systematized. A comparative analysis of modular and non-modular designs based on key operational criteria demonstrated a reduction in maintenance time by 30–50% and an increase in recovery efficiency due to fault localization at the module level.

The obtained results are explained by the application of principles of functional decomposition, unification, and interface standardization, which ensure subsystem independence and interchangeability, as well as minimize inter-module dependencies.

A distinctive feature of the proposed approach is the integrated combination of structural and technological principles of modular design, focused not only on the development stage but also on the entire UAV life cycle, enabling simultaneous improvement of maintainability, adaptability, and operational efficiency of the platform.

The practical application of the results is relevant for the design of multipurpose, research-oriented, and operation-intensive quadcopter UAVs, particularly under conditions of frequent reconfiguration, field maintenance, and modernization, provided that unified modules and standardized interfaces are used.

Toward AI-Enabled Production Management: A Decision-Layer Framework with Simulation-Based Structural Validation

Aswin Karkadakattil — 2026-06-18

The increasing adoption of artificial intelligence and simulation in production management has largely concentrated on improving algorithmic performance at isolated decision points. However, relatively limited attention has been given to how decisions interact across strategic, tactical, and operational levels of an organization. This study adopts a decision-centric perspective to examine the structural alignment of production decisions across these interconnected layers. A decision-layer decomposition framework is proposed to explicitly distinguish strategic, tactical, and operational decision domains and to clarify how decisions at each level collectively shape overall system performance. Rather than pursuing algorithmic optimization or empirical benchmarking, discrete-event simulation is employed as a structural validation tool to examine decision coherence across layers. Using a simplified production system, the simulation experiments investigate how strategic capacity decisions propagate through tactical planning and operational execution, influencing key performance indicators such as throughput, lead time, and work-in-process. The results reveal a non-linear performance response in which throughput reaches saturation before lead time and work-in-process are fully minimized. This divergence highlights the limitations of throughput-driven capacity planning when evaluated in isolation and underscores the importance of multi-KPI assessment. By framing AI and simulation as decision-support infrastructures rather than purely optimization mechanisms, the proposed framework provides a transferable approach for evaluating decision coherence and aligning managerial objectives with operational outcomes. The study contributes to production management research by formalizing decision

Effective Damping of Resonant Liquid Fuel Splashes in Aircraft Tanks

Vasyl Kovalev — 2026-06-18

The article is devoted to the study of methods for effective damping of resonant splashes of liquid fuel in aircraft tanks. Modern designs of dampers, their advantages and disadvantages are considered. An important role is played by the mass and geometry of the baffles, which can cover up to 73 % of the volume of the entire tank, while significantly increasing its mass, which has a very negative effect on the hydrodynamic mechanism of the development of inertial flows. The influence of fuel splashes on the flight dynamics and stability of the aircraft is analyzed. Optimal design solutions are proposed to reduce the amplitude of liquid oscillations in tanks of various configurations and force effects on the structures of internal damping devices, which can occupy almost 45 % of the volume of the entire tank.
The results of numerical modeling and experimental studies are presented, confirming the effectiveness of the developed methods. The accuracy of numerical modeling is also determined by the size of the cells in the most dynamic flow regions, where, for example, there are large differences in parameters, in particular velocities. It is shown that the use of optimized damping devices allows you to reduce the amplitude of resonant oscillations by 75–80 % with an increase in the mass of the structure by no more than 3 %.
Analysis of the hydrodynamic mechanism of the development of inertial flows in the tank allows you to predict the distribution of such force parameters as hydrostatic pressures on the tank walls, the distribution of average velocity in characteristic zones, in particular over time. Such flow parameters allow you to establish quite complex nonlinear features of flows and predict their development and force effects on the design of a tank with baffles. Resonant fluctuations of the liquid in tanks allow you to fairly accurately determine the force effects of the flow on the internal structures of tanks and establish the magnitude of shock pressures on the internal structures.