ABSTRACT. Effective diagnosis and identification of unbalances in modern, high-speed ro-tating machines still poses a serious challenge for designers and operators of such equipment. For this purpose, in order to evaluate results of on-line moni-toring performed on a real machine, it is extremely important to use the knowledge obtained from analyses of theoretical results performed by means of a reliable digital-twin model of this object. Therefore, this paper proposes an al-ternative approach to identifying rotor shaft static and dynamic unbalances based on using results of on-line measurements of lateral vibration amplitudes and phases and treating them as input data for inverse problem solution by means of a reliable structural theoretical model of the tested device. Such a model of the rotor shaft  bearing supports system is constructed from flexural-ly deformable cylindrical continuous beam finite macro-elements and discrete oscillators mutually connected according to the structure of the object under study. Dynamic responses of this model are sought in the domain of modal functions, the number of which must be appropriately correlated with the num-ber of measurement paths of the rotor shaft monitored on-line. Owing to the ap-propriate processing of measurement signals from the time domain into the fre-quency domain obtained from one rotational speed, the unbalance values and their phase angles are determined once and with relatively high accuracy in the most probable cross-sections of the rotor shaft under testing. The practical ef-fectiveness of the proposed approach depends largely on the number of meas-urement sensors installed on the monitored real object.

ABSTRACT. Monitoring mechanical systems is essential for the safe and efficient operation of rotating machines, mitigating failures that can lead to unplanned downtime, financial losses, and asset damage. In this context, advances in data-driven and predictive maintenance techniques has driven the development of Virtual Sensing (VS), enabling the estimation of variables such as vibration and unbalance forces without exclusive reliance on physical sensors, thereby reducing costs and increasing monitoring flexibility. Rotor unbalance, one of the most critical faults in rotating systems, leads to excessive vibration, reduced dynamic stability, and accelerated wear, making balancing a key maintenance procedure. In this way, the present work proposes an Augmented Kalman Filter (AKF)-based methodology for near-real-time estimation of vibration states and unbalance forces. The approach was experimentally validated on an instrumented rotor test bench integrated with a finite element model developed by using the open-source ROSS library. Unbalance and balancing planes were defined at rigid disk locations, and vibration measurements were acquired at a position close to the bearings. To evaluate resilience under non-ideal instrumentation, measurement planes were intentionally rotated with respect to the model reference axes. After signal preprocessing, the AKF covariance matrices were tuned to ensure observer convergence. The results demonstrated physically consistent unbalance estimates and significant vibration reduction, confirming the applicability of the proposed method for balancing flexible rotors in the industriy.

ABSTRACT. Balancing rotating machinery is essential for stable, reliable operation and extended equipment lifespan. Conventional methods require trial masses and multiple experimental runs, which are time-consuming, costly, and disruptive. This paper proposes a machine learning approach that eliminates the need for trial masses and extensive testing. To overcome the scarcity of real-world data—a common limitation of purely data-driven methods—we developed a data augmentation strategy based on the deterministic relationship between the phase of the vibratory response and the angular position of the unbalance. Synthetic datasets were generated from only two historical balancing records and used to train a feedforward neural network capable of directly predicting the magnitude and angular location of corrective masses from steady-state vibration measurements. Experimental tests on a flexible rotor demonstrated an average unbalance reduction exceeding 70\%, even with minimal data. The approach is efficient, scalable, and practical for industrial applications.

ABSTRACT. Research work presents a high-fidelity, real-time simulation framework for electro – hydraulic mobile machines, featuring a comprehensive case study of PATU655 crane system powered by a Parker servo motor. A detailed dynamic model of the hydraulic subsystem is developed in MATLAB, capturing key aspects such as fluid dynamics, actuator behavior, and mechanical linkages. This model creates a realistic simulation environment suitable for control prototyping and hardware-in-the-loop (HIL) applications. Classical PID-based control architecture is employed to regulate the servo motor and hydraulic actuation system in the heavy mobile machines laboratory, LUT University. The controller tuned to ensure accurate position tracking, stable pressure regulation, and efficient energy use under fixed load conditions. The precision of the Parker servo motor enhances the system’s responsiveness, allowing find-grained control over motion and force transmission. To further enable real-time implementation, the MATLAB-based model is exported as a Functional Mock-up Unit (FMU) and integrated into Ansys Twin Builder, where a dynamic reduce-order model (ROM) is generated for real-time execution. Simulation results validate the accuracy of the model, the effectiveness of PID control strategy, and feasibility of running the FMU-based dynamic ROM in real time. The system demonstrates reliable dynamic performance, robustness, and practical real-time applicability, making it well-suited for integration into modern mobile hydraulic machinery. This work highlights the value of combining high-fidelity with conventional control techniques and real-time capable reduced-order models for developing intelligent, efficient, and adaptive electro-hydraulic systems in industrial applications.

ABSTRACT. The study of rotating machinery has long relied on the finite element method for its design and analysis under operational conditions. However, when addressing the interactions between rotating shafts and their highly nonlinear supports, this approach reaches its limitations. While detailed finite element simulation of a bearing is feasible and allows accurate estimation of load distribution within the system, this is not possible at the scale of an assembly incorporating bearings, gears and shafts. Besides, the method makes assumptions about the amplitude of displacements, rotational speeds and often demands intensive computational resources. Finally, joint elements such as gears and bearings generate internal, angularly periodic excitations that can modify the rotor behavior. They exhibit complex physics and require dedicated models. An alternative tool to understand these systems is flexible multibody dynamics, which expands the modeling capabilities for rotating machinery due to its versatility in describing mechanical systems at different scales. This work leverages flexible multibody dynamics for the simulation of bearing components inducing cyclic perturbations in shafts. Semi-analytical, modular models of ball and roller bearings from literature are implemented into a multibody solver. The models allow the representation of the bearing compliance variation, high speed phenomena and internal faults with realistic geometry. This study focuses on the critical steps for accurately representing the system dynamics at constant and variable speeds, to identify the opportunities and limitations of this approach in comparison with more traditional methods.

ABSTRACT. Simulations are used in mechanical system design to understand how a system will behave to ensure intended behaviour and durability. Performing simulations is more cost effective than physical testing and prototyping. In vibration problems complex systems consisting of components from multiple manufacturers need to be modelled together due to the coupled nature of the problem. However detailed simulations require detailed models that may risk the intellectual property of the component manufacturer. This paper presents an approach to model the interfaces of a model adhering to the Functional Mock-up Interface (FMI) standard. With this interface a complete driveline model can be built from Functional Mock-up Units (FMUs) and torsional vibration analysis can be performed for the system. The main results of the paper are highlighted in torsional vibration analysis, where a traditional lumped element model is built from FMU submodels and compared against more traditional method of constructing the model. This paper presents a method for modelling a lumped element model as an FMU and a method for defining the interfaces of these FMUs, with domain overlapping approach, allowing for fast simulation times.

ABSTRACT. To address the weak strength and stiffness of acoustic black holes (ABHs) in vi-bration reduction applications, this paper proposes a rotating shell with an addi-tive ABH, combined with a damping layer to achieve vibration attenuation under rotating conditions. Initially, considering the rotating effect, the Rayleigh-Ritz method is used to establish the equations of motion, combining with Sanders shell theory. Different artificial spring groups are used to simulate the coupling and boundary conditions. Then, the proposed method is verified by comparing the numerical results with simulation analysis and experimental results. Finally, par-ametric studies are carried out to investigate the influences of ABH dimensions on the vibration reduction effect. The results show that the ABH effectively re-duces vibration in the high-frequency region of the rotating disk-shell structure, but its ultimate vibration reduction performance decreases with increasing rota-tional speed. At low speeds, a larger length and smaller uniform cross-sectional thickness are more conducive to vibration reduction, while at high speeds, a smaller length and smaller uniform cross-sectional thickness achieve excellent vi-bration reduction.

ABSTRACT. This study investigates the rotordynamic performance of composite shafts as lightweight alternatives to conventional steel shafts in rotating machinery applications. A comprehensive comparative analysis is conducted between a reference steel shaft and carbon fiber reinforced polymer (CFRP) composite shafts utilizing four different fiber types (T-300, HMU, F-600, and E-130) at varying volume fractions (30%, 40%, 50%, and 60%). The internal damping characteristics of both materials are modeled using viscoelastic models namely, a three-element model for steel and a four-element model for composite materials. The rotordynamic equivalence is evaluated through three key performance metrics: stability limit of spin speed (SLS), unbalance response (UBR) amplitude at rated speed, and static deflection. Results demonstrate that composite shafts achieve substantial mass reductions ranging from 76% to 82% compared to steel, while maintaining comparable or superior rotordynamic characteristics. Higher modulus carbon fibers (F-600 and E-130) exhibit enhanced stability limits and reduced vibration amplitudes, with E-130 at 60% volume fraction showing an SLS of 6236 RPM compared to 4405 RPM for steel. The findings establish that appropriate selection of fiber type and volume fraction enables designers to achieve significant weight savings without compromising rotordynamic performance, offering valuable insights for energy-efficient rotating machinery design.

ABSTRACT. The structure of honeycomb seal (HCS) has a significant impact on its seal-ing performance. Our previous research proposed a novel honeycomb seal structure called the wall-hole-honeycomb seal (WHHCS). The WHHCS can further improve the sealing performance of HCS, including double-wall-hole-honeycomb seal (D-WHHCS) and single-wall-hole-honeycomb seal (S-WHHCS). To delve into the influence of wall hole quantity on the WHHCS, the leakage characteristics of the two configurations are investigated using the computational fluid dynamics (CFD) method. The vortex structures and dis-tributions in both seals were examined based on the Q-criterion. The results show that increasing the number of wall holes improves the leakage suppres-sion capability of WHHCS, with a 36.5% enhancement observed when changing from a single to a double wall hole configuration. The number of wall holes also affects the vortex and streamline distributions in cavities, lead-ing to a U-shaped vortex structure in S-WHHCS at lower Q values. At high-er Q values, different vortex morphologies are observed in the left and right wall hole regions of D-WHHCS, with the left wall hole exhibiting a more pronounced effect on leakage reduction. Furthermore, D-WHHCS exhibits a 10.3% larger vortex volume in the sealing gaps and a 16.7% larger vortex volume in the cavities than S-WHHCS. The larger volume of high-Q vortex regions gives the D-WHHCS a clear advantage in leakage reduction over the S-WHHCS. This study provides guidance for further structural optimization of WHHCS.