ABSTRACT. This paper investigates elastically mounted rigid air ring bearings, which are designed to improve the stability of high-speed rotor systems. Conventional three-lobe air bearings, where the bearing sleeve is rigidly connected to the housing, typically exhibit a poor stability behavior. One way to overcome this limitation is to introduce external dissipation to the system. Therefore, the bearing sleeve (ring) is elastically mounted in the housing by a beam foil structure, which generates dissipation by dry friction. The dynamics of the system is strongly influenced by the properties of the foil supporting structure and the bore geometry of the air gap between the journal and the inner ring surface. This work presents detailed parametric studies examining the influence of different bearing parameters – especially the effect of a taper-land bore geometry – on the stability and robustness of the rotor/bearing system. Since the nonlinear rotordynamic behavior beyond the critical speed of instability (onset of whirl/whip oscillations) is of particular interest, transient run-up simulations are required. The goal is to operate the system above the critical speed of instability with small whirl/whip amplitudes. To address these challenges, a nonlinear multibody rotor model is used, as well as finite element models to calculate the pressure distribution within the air films of the two radial bearings. The multibody and finite element subsystems are coupled in time domain using a sequential co-simulation approach.

ABSTRACT. The Morton Effect is a synchronous rotor instability caused by uneven heating in a journal bearing. The heating is caused by viscous shearing of the oil film resulting in a “hot spot” developing on the rotor. The “hot spot” causes the rotor to bow, which creates imbalance. Eventually, the rotor bow and resulting imbalance becomes large enough that the clearance increases and the shaft cools and relaxes until the bow is removed. The process then repeats itself.

Elevated rotor vibration due to Morton Effect was found on a centrifugal compressor in natural gas pipeline service. The 2-stage compressor is of “overhung” arrangement with the impellers outboard of the impeller-end bearing. At compressor speeds above 5,500 RPM, the synchronous vibration amplitude would continuously oscillate, and the phase angle would rotate 0 - 360 degrees. The oscillation period was 3 to 5 minutes, depending on operating speed.

Subsequent rotordynamic analysis indicated the Morton Effect was likely to occur with 10 mil bearing clearance. Using the normalized rotordynamic model, potential modifications and their effect on Morton behavior were evaluated. After numerous iterations, an impeller-end bearing with 13 mil clearance and 0.2 preload was predicted to be optimum.

A new impeller-end bearing was machined with the recommended geometry and installed in the compressor. The resulting radial vibration was low and acceptable throughout the entire operating speed range.

ABSTRACT. The dynamic behaviour of rotating machinery, such as centrifugal pumps, compressors, and turbopumps, is strongly influenced by the characteristics of narrow fluid-filled gaps, including journal bearings, annular seals, thrust bearings, and balance pistons. Balance pistons are commonly employed in large centrifugal pumps to compensate for the inherently induced axial forces resulting from pressure increases across the impellers. Within the paper, a revision of the annular gap test rig operated at the Chair of Fluid Systems at Technische Universität Darmstadt is presented. The test rig was revised to experimentally investigate the rotordynamic characteristics of balance pistons and axial thrust-balancing devices. The paper presents the overall setup, as well as the calibration procedure for the active magnetic bearings. Furthermore, preliminary results are presented targeting the determination of the static and dynamic characteristic of double balance pistons in the relevant parameter range for turbulent flows in modern power plant pumps.

ABSTRACT. Air bearings are considered a key technology for high-speed and oil-free turbomachinery due to their low friction losses, high temperature resistance, and contamination-free operation. Bump foil bearings, in particular, are characterized by their elastic structure, which allows among other aspects self-adaptation . Nevertheless, accurate prediction of their dynamic behavior has so far been limited, as the deformation of the foil structure is often only simplified in numerical models. In this work, the deformation of the top foil and the bump foil is simulated using a three-dimensional finite element method with shell elements, which enables a realistic representation of the elastic behavior. This approach allows improved coupling between the pressure field and the structure. Considering foil elasticity indicates a significant influence on load capacity, damping behavior, and stability. By integrating these nonlinear effects and realistic boundary conditions, resonance phenomena and load variations can be analyzed in greater detail. This represents an important step towards reliable and application-oriented simulation of foil-supported rotor systems, opening new possibilities for the design of efficient, oil-free rotating machinery.

ABSTRACT. Constant velocity joint are an important part of a passenger car. The increasing electrification of powertrains is changing the load profiles for drive shafts. In par-ticular, recuperation and the modified speed-torque characteristic curve of an elec-tric motor compared to a combustion engine are causing changes in the load pro-file. This means that previously used designs for joints can cause increased wear or tend to excite vibrations in the axial direction. The latter becomes problematic in electric drives, which already have a low NVH level. To accompany test bench trials for evaluating different designs, a simulation model is to be developed here that maps the individual components of a constant velocity joint. A multibody model with frictional contacts and analytical contact detection between the balls and raceways can be used to investigate the influence of lubricant and raceway geometry and their deviation. Through analytical param-eterization, the model should also serve as a basis for automatic parameter varia-tion or even optimization in order to systematically investigate the influence of geometric deviations.

ABSTRACT. In the modern, highly parallel product development methods it is important to take into consideration the possible dynamic loads on the machine already in the conceptual design phase. Rotors usually are modelled as one-dimensional beams in this phase, which allows easy parametrization and variation of the main design variables based on e.g. finite element (FE) simulation. Including these FE models in multibody simulation (MBS) traditionally has not been easy, since they typically do not handle models from FE software with rotational degree of freedom, therefore one must build the whole model in the MBS environment again. Implementing design changes or multiple design variations and in case there is only limited availability to the model due to intellectual property protection, this work can become quite burdensome and prone to errors. This work aims to propose a method that needs only the consistent mass matrix and the undeformed node positions of the rotor as inputs to first generate a lumped mass matrix via modal lumping, then use that lumped mass matrix to compute the so-called mass invariants via matrix multiplications. In a numerical example, these are used in a floating frame of reference based MBS environment, that can handle rotational degrees of freedom, to simulate the dynamic behavior of a turbine-generator ma-chine in maritime conditions. The results are compared to those calculated by the traditional inertia shape integrals and 3D solid elements.

ABSTRACT. Subsynchronous resonance (SSR) in turbine-generator shaft trains is a serious concern in power plants, potentially resulting in substantial shaft damage or a reduced fatigue life. A critical aspect of SSR analysis is the consideration of system damping. The interaction between an electrical grid system and a mechanical turbine-generator shaft train involves numerous mechanisms and effects that collectively contribute to the coupled system's overall damping. Consequently, accurately determining the contributions of various damping mechanisms to the coupled power systems through numerical simulations remains a complex task. This study introduces an experimental approach for identifying the damping characteristics of turbine-generator shaft trains subject to sub-synchronous torsional vibrations. The damping was estimated using experimental data obtained from the full-scale turbine-generator shaft at the Olkiluoto 3 nuclear power plant in Finland. Band-pass filtering was implemented for a time domain log-decrement damping identification method and the influences of electrical load and vibration amplitude were examined separately. Challenges and limitations associated with traditional viscous damping model were demonstrated and various aspects that must be considered for accurate modelling of the actual damping behavior of the coupled power systems are also highlighted. Additionally, damping was also identified under steady-state operational conditions through the application of Operational Modal Analysis (OMA) for comparison.

ABSTRACT. Periodic stiffness changes can have a positive effect on the vibrations of a mechanical drive train system. To take advantage of that fact, a novel approach is presented which utilises an electric synchronous machine for this purpose because such a machine can easily be applied for implementing this concept.

Based on the classical modeling of synchronous machines using space vectors with regard to the synchronous rotating reference frame, we discuss the possibilites for achieving periodic stiffness changes in a drive train system. The various control concepts with either an electrical excitation or an excitation with permanent magnets are explained in detail. Advantages and drawbacks of both machine types for such applications are mentioned.

An application example shows how vibration energy can be shifted between modes and how this enables efficient utilization of the existing damping.

ABSTRACT. Turbogenerators consist of very important mechanical and electrical components in power plants. Steam turbines convert thermal energy into mechanical energy causing the rotor to spin, while the electrical generator produces electrical energy by electromechanical interactions. The total energy is transferred via the air gap torque of the generator. For a safe and reliable operation, the mechanical-electrical interactions in the air gap of the generator and the resulting torsional vibrations of the shaft train have to be observed very carefully. In case of disturbances in the electrical system, the excited transient torsional vibrations may become very large due to the low system damping. In such cases it is important to control the torsional vibration behaviour to guarantee acceptable vibration and stress levels. A Digital Twin is a suitable monitoring tool to solve this task. With a Finite Element Model (FEM) as part of the Digital Twin torsional vibrations can be calculated. And the real torsional vibrations are measured at selected locations of the shaft train with sensors. By comparison of measured and calculated torsional vibrations the vibration difference is introduced into a calibration loop of the Digital Twin, in which the system parameters as well as the air gap torque are continuously identified and adjusted. The Digital Twin can be used online, running permanently parallel to the real system as a monitoring system and as a detection and identification system in case of system failures.

ABSTRACT. Additive manufacturing, commonly known as 3D printing, has become a popular technology for producing mechanical parts due to its short cycle times and low costs associated with materials and equipment. This technology is increasingly utilized for creating functional prototypes, conceptual models, pre-surgery models, and customized products. One promising application is the production of low-power or low-loading mechanical components, which can benefit from the quick, low-cost, and on-site manufacturing of spare parts. A key question arises regarding the performance of these 3D-printed spare parts during machine operation. This study investigates the dynamic performance of a rotor manufactured using this technology with ABS material. Experimental tests were conducted under various printing conditions, including different infill densities and raster angles. The lateral vibration of the rotor was measured on a test rig at varying rotational speeds and levels of unbalance. The results indicated that the rotor with 100% infill density exhibited vibration levels comparable to those of a solid machined rotor made from the same material. However, as the infill density decreased, harmonics at 2xΩ and 3xΩ (indicating parametric resonance) appeared in the vibration frequency spectrum. These findings can be attributed to the increasing orthotropy of the rotor’s cross-section due to the 3D printing process.

ABSTRACT. Rotors with longitudinal periodicity present wave band gaps, so that no resonances occur within these wide frequency bands. While previous research has primarily focused on wave propagation in periodic rotors supported by isotropic bearings, which are characterized by equal stiffness in all radial directions, real-world applications frequently involve anisotropic bearings. These bearings have stiffness that varies between radial directions, also including cross-coupling stiffness terms. This introduces additional complexities into the dynamic behavior that must be investigated. In this work, we extend the wave analysis of periodic rotors supported by isotropic bearings to anisotropic bearings, focusing on how different directional stiffness and high cross-coupling stiffness influence elastic wave propagation and band gap formation. High cross-coupling stiffness can drive the rotor system toward instability, a phenomenon that is critical in rotating machinery design. To capture these effects, the methodology combines Bloch wave propagation theory with boundary conditions that incorporate anisotropic bearing properties. The dispersion curves obtained illustrate the impact of anisotropic stiffness and destabilizing effects on the band gap formation and the dynamic behavior of the periodic rotor. Furthermore, the results reveal how changes in bearing characteristics alter the wave propagation. These findings provide valuable insights into the interaction between boundary conditions and periodic structures, paving the way for improved performance and stability analysis of periodic rotating systems.

ABSTRACT. Viscoelastic materials offer promising applications in rotordynamics but pose significant design challenges. This paper demonstrates strategies to address common issues including frequency-dependent stiffness and damping, creep and instability. For this, a test bed with a polyoxymethylene (POM) rotor is investigated. The Numerical Assembly Technique (NAT), an efficient modelling technique based on analytical formulations for each section of the rotor, is used to describe the rotor. The viscoelastic material is assumed to follow a fractional Zener model. To determine the correct parameter for this model, the eigenfrequencies of the system and the bearing supports are assessed using unbalance response and impact-hammer tests. Different sets of parameters of the fractional Zener model for POM from literature are compared with each other and with measurements. The pre-bend of the shaft is compensated with a balancing approach for warped rotors and stability is ensured by the introduction of external damping and anisotropy via 3D-printed bearing shields. These procedures increase the operational range of the system to higher speeds, indicating the effectiveness of the proposed methods for the proper application of viscoelastic rotors.

ABSTRACT. Rotors serve as the core components in critical machinery, however the coupling of nonlinear factors—specifically nonlinear contact stiffness in bolted joints, oil-film forces, and rub-impact forces—often leads to unpredictable instabilities. To address the limitations of traditional vibration analysis in identifying transient bifurcations and complex nonlinear motions, this paper presents a numerical study on the bifurcation forms and stability of a bolted rotor-bearing system from an energy perspective. A dynamic model incorporating rub-impact, contact stiffness and Capone oil-film forces is established. Distinct from conventional Poincaré maps, a energy trajectory method is proposed to analysis the form of the rotor system. The energy trajectory provides a more detailed description of the rotor's motion patterns, capturing energy fluctuations that traditional waveforms miss. The comparison demonstrates that the proposed energy perspective is a sensitive and intuitive indicator for real-time stability monitoring in engineering applications.

ABSTRACT. Annular seals in rotating machines may generate significant dynamic forces, leading to fluid flow-induced instabilities, which detrimentally affect system performance. One way to reduce, eliminate, and postpone such instabilities is the use of tilting-pad journal bearings with active oil radial injection. This work experimentally investigates the feasibility of different control schemes applied to actively-lubricated tilting-pad journal bearings, to counteract destabilising seal-induced forces in a rotor-bearing-foundation system. The theoretical prediction of the global system dynamics is obtained in five steps: i) Obtaining the rotor dynamics and coupling it with the foundation finite element model. ii) Integrating the simulated seal forces as cross-coupling stiffnesses to both the rotor and foundation. iii) To add the fluid film bearing stiffness and damping properties, calculated through the linearization of the thermo-elasto-hydrodynamic bearing model. iv) Inclusion of the servo valve actuator dynamics, which by injecting high-pressure oil, influences both the bearing and rotor behaviour. v) Closed-loop control system design for the oil radial injection using modern control theory and global model state-space representation, to enhance system stability. Instabilities are experimentally generated in a test rig using an active magnetic bearing, with orthogonally placed coils applying controlled destabilising forces based on real-time rotor displacement measurements. Both theoretical and experimental results demonstrate that modal-based controllers effectively reduce, eliminate or de-lay the onset of fluid-induced instabilities, validating the potential of active lubrication in stabilising rotor-bearing-foundation systems.

ABSTRACT. In certain design configurations, rotating machines can have speed dependent resonances which are difficult to accelerate through. In other designs, rotors may have resonances which occur in close proximity. Under certain forcing conditions, when resonances occur closely, each resonance may have very different optimum levels of damping, and in some cases, the optimum damping coefficient for one can be several orders of magnitude different from the other. Under these conditions, it can be highly favourable to have a damper which can vary its damping coefficient between resonances. This paper simulates such a system and investigates attenuating both resonances using various types of variable damper at one particular DoF and forcing condition, and compares this with the ’Pareto-Optimum’ constant damping coefficient. The first type is a variable rate damper whose damping coefficient is a function of speed; it has damping that varies linearly with frequency between the optimum levels for each resonance (VLD). A second proposed variable damper type can simply switch damping coefficients at specific frequencies (VSD), in this case between mode 1 and mode 2 optimums. It is well known that Invariant Points can be used to find or tune optimum parameters at resonance, but some types of IPs are also possible ’switching locations’ to change the damping coefficient between resonances. The VSD investigates such switching. Employing variable dampers have been found to have strong benefits in such systems, with some small considerations.

ABSTRACT. Ball bearings (BBs) exist in a large variety of types and configurations and are widely applied in power transmission layouts. Ball bearings support a rotating shaft applying impedance forces which are approximated by linear or nonlinear functions of inner and outer ring kinematics, depending on the operating conditions. Ball bearings may develop defects or higher frictional loss when load or angular/lateral misalignment is not in the predicted range. This paper introduces an Actively Controlled Ball Bearing (ACBB) with the feature to displace the outer ring of the ball bearing with respect to the bearing housing. Three piezoelectric actuators are embedded at the respective radial direction and at three locations along the circumference of the outer ring; the outer ring executes planar displacements at two directions. In this way, the ACBB can accommodate lateral and/or angular misalignment of the shaft with respect to a reference position, and as a consequence, the radial impedance force of the ACBB to the rotor is adjusted in optimal values in order to achieve lower vibration energy, lower frictional power loss in the power train, and optimize further operating parameters and dynamic characteristics. A two-rotor co-axial transmission including a u-joint coupling is the simulated system to check the effectiveness of the ACBBs on reducing vibration amplitude applying Extremum Seeking Control (ESC). The results demonstrate the ability of the ACBB to adjust the relative position of the shafts at an optimal placement, and to reduce the vibration energy of the system.

ABSTRACT. The paper presents results of research on predictive control algorithms applied in the magnetic bearing support system of a high-speed rotor machine. The theoretical characteristics of the control system, utilising both parametric and non-parametric algorithms, are presented. The Extended Horizon Adaptive Control, the Extended Prediction Self-Adaptive Control, and the Generalised Predictive Control belong to the first group. Moreover, the Model Algorithmic Control, the Model Predictive Control, and the Dynamic Matrix Control use a non-parametric object model and are included in the second group. A theoretical model of the magnetic suspension system consists of a massive, single-disk rigid rotor-shaft supported on two active radial magnetic bearings and one active axial (thrust) magnetic bearing. The model adopted in this way took into account the limitations of the real test rig, including the dead zone of the sensors, their bandwidth, and the current limitations of the power amplifiers. The analysed characteristics included the system's responses to kinematic excitations, both at constant rotor speed and during startups. The theoretical studies were carried out by means of Matlab and Simulink software. Additionally, some of the computational results were confirmed by experimental measurements carried out on the dedicated test rig. Finally, there was performed an analysis on the power consumption of the magnetic bearing system using predictive control algorithms, with particular focus on how the tuning factor and control horizon affect the dynamic performance of the rotor.

ABSTRACT. This study presents a performance analysis of a hybrid tilting pad journal bearing (TPJB) designed for reusable liquid methane turbopumps. As the demand for reusability and reliability in modern rocket engines grows, rolling-element bearings face limitations due to wear, lubrication challenges, and short service life, motivating the adoption of fluid-film bearing technologies. The hybrid TPJB, which combines hydrostatic and hydrodynamic effects, offers superior load capacity, low friction, and enhanced dynamic stability—key advantages for high-speed cryogenic turbomachinery. In this work, a comprehensive analytical model incorporating temperature and pressure dependent fluid properties of liquid methane based on NIST data was developed and verified through polynomial curve fitting with a coefficient of determination of 0.998. Parametric studies conducted to examine the effects of static load, length-to-diameter, preload, radial bearing clearance ratio, supply pressure, and ratio of recess area to total bearing area on bearing performance. The results showed that increasing preload and supply pressure improved load carrying capacity and stiffness but reduced damping, while larger clearances and recess area weakened hydrostatic effects and system stability. The derived rotordynamic coefficients provide a technical foundation for future rotor-bearing system analyses, supporting quantitative evaluations of critical speed, vibration response, and overall stability in reusable liquid me-thane rocket engine turbopumps.

ABSTRACT. This paper experimentally investigates the influences of journal misalignment on the static and dynamic performances of a water-lubricated journal bearing (WJB). A test rig is developed to impose controlled misalignment angles on the test bearing under various applied load and rotational speed conditions. Measurements are focused on the journal eccentricity ratio, journal attitude angle, minimum film thickness, bearing stiffness and damping coefficients. The static test results show that increasing misalignment consistently reduces the minimum film thickness. The effect is more pronounced at low bearing loads where hydrodynamic pres-sure generation is weaker. At high loads, the influence of misalignment diminish-es but remains non-negligible. The journal attitude angle decreases with increasing misalignment ratio, and the effect is more significant at low rotor speeds, whereas the stronger hydrodynamic pressure lessens this effect at high rotor speeds. Impact tests demonstrate that misalignment reduces the magnitudes of both the direct and cross-coupled, stiffness and damping coefficients. The direct mass coefficients also decrease under misaligned conditions, implying a reduction of the fluid inertia contribution. The cross-coupled mass coefficients are small compared to the direct ones. The current results provide comprehensive experimental evidence that the bearing misalignment not only decreases the load capacity but also degrades the dynamic performance.

ABSTRACT. This study focuses on the modeling of a tilting pad journal bearing (TPJB) of a vertical hydropower unit and investigates its impact on the shaft response to unbalance forces. Traditionally, guide bearings are modeled with a quasi-static approach, considering instantaneous equilibrium of the pads’ tilt. This approach, which neglects pad inertia, is commonly extended to dynamic loading conditions assuming the quasi-static approximation holds. The validity of this assumption is investigated by developing a model that includes pad inertia and applying it to a full-scale hydropower unit. Scale effects are assessed against a small size TPJB. Results reveal that neglecting even low pad inertia underestimates forces and overestimates displacements at the bearings. For a large-scale TPJB, these discrepancies amplify for all tested conditions. Also, pad angular motion is reduced and occurs with a greater phase lag.

ABSTRACT. The determination of the static equilibrium position is a fundamental and computationally expensive step in the performance and dynamic stability analysis of hydrodynamic journal bearings. This work proposes a new strategy to modify traditional Newton iterations, aiming to minimize processing time while maintaining physical precision. The developed methodology integrates four synergistic approaches: (i) pressure distribution reuse (warm start); (ii) Jacobian stabilization via a Two-Stage solver; (iii) adaptive convergence tolerance; and (iv) dynamic mesh refinement (grid sequencing). Results for a representative test case demonstrate a Time Speedup Factor of 8.9 and a Computational Cost Speedup Factor of 9.1 compared to the standard algorithm. Furthermore, global performance mapping reveals that the proposed algorithm is particularly effective in numerically stiff regions (high eccentricity/low Sommerfeld numbers), achieving speedups of up to 30x. The method robustness is validated throughout the useful operating domain, limited only in regimes of extreme eccentricity, where the physical hypotheses of the continuous hydrodynamic model are no longer valid.

ABSTRACT. Rotor unbalance is a critical issue in electrical machines used for high-precision applications. It generated the transverse vibrations at the synchronous rotating speeds, and if left undetected, it would result in serious mechanical failures and unexpected system downtime. To ensure the system's reliability, it is essential to identify such faults at an early stage. To enable early detection of rotor unbalance faults, this paper introduces a Condition Health Monitoring (CHM) framework. It provides predictive diagnosis of rotor unbalance in Induction Motors (IMs). This study employs Motor Current Signature Analysis (MCSA) of circulating currents between the parallel branches as an effective diagnostic tool for detecting rotor unbalance. To detect the inherent mass unbalance in the system, the unbal-ance is modelled as a dynamic eccentricity (DE) in numerical simulations. The FEM model is validated through experimental investigations, and the relationship between DE predicted by FEM simulations and the current is examined to assess fault severity using a linear regression model. The proposed methodology is first validated under known mass unbalance conditions and subsequently extended to previously unseen cases. The inherent mass imbalance of the system at its most concentric conditions is evaluated using this method, which demonstrates its pre-dictive diagnostic capabilities. The results indicate that the CHM framework pro-vides a non-intrusive, effective means of early fault detection and severity as-sessment, thereby supporting condition-based maintenance strategies for IMs.

ABSTRACT. Electrical machines play a critical role in diverse applications, often operating under variable and demanding conditions. Even slight disturbances can compromise mechanical integrity, necessitating a thorough understanding of dynamic behavior for optimized design and reliable performance. While in-duction motors are prized for their robustness, they remain vulnerable to faults, particularly rotor-stator eccentricity, which diminishes natural frequency and overall efficiency. To counteract vibration-induced instability, this study ex-plores the use of a bridge-configured winding (BCW) as an integrated force ac-tuator. A finite element method (FEM)-based analysis evaluates the induction motor's response to varying operational parameters, including variable slip at standard operating speeds, and variable operating speeds, with both static and dynamic eccentricity. The effectiveness of the BCW in mitigating vibration is systematically assessed for these operating conditions. Experimental validation, conducted on a dedicated test rig under no-load conditions, compares perfor-mance with and without the BCW actuator. Results demonstrate that the BCW configuration markedly improves system stability by attenuating rotor vibra-tions, offering a promising solution for fault resilience in induction motors.

ABSTRACT. Improving energy efficiency in industrial systems increasingly involves replacing direct-on-line electric motors with variable-speed electric drives, composed of a frequency converter and motor. To minimize retrofit costs, existing driven machine, drivetrain components and motor foundations are often retained. However, this introduces challenges, as vibration designs originally optimized for a fixed speed must now accommodate a wide range of operating speeds.

This paper presents a novel approach for active control of both lateral and torsional vibrations in variable-speed electric motors. The proposed method utilizes one or more vibration sensors, the rotor core as an actuator mass, and precise control of electromagnetic airgap torque and force via the frequency converter. A key innovation is the introduction of an open software platform that enables external partners to create, validate, and realize advanced active vibration control applications.

The paper outlines the conceptual framework and shares preliminary results demonstrating the potential of this new approach to enhance vibration performance across variable operating conditions.

ABSTRACT. Unbalanced magnetic pull (UMP) and rotor–stator eccentricity are critical factors influencing the operational reliability and vibration behavior of induction machines. This paper presents a method for quantifying UMP forces by analyzing bearing‑housing vibration data collected during both ramp‑up and coast‑down speed transitions. The approach compares amplitude ratios of resonance peaks and selected harmonic features in the vibration signals to isolate the influence of UMP from the mechanical baseline. Finite element method (FEM) simulations of a rotordynamic model, parameterized by static and dynamic eccentricity, are used to generate response maps that relate the measured amplitude ratios to equivalent UMP and eccentricity levels, enabling direct comparison with experimental measurements. Results indicate that amplitude‑ratio analysis provides a reliable means of identifying UMP and assessing its impact on machine dynamics. The results are intended to enhance operational safety, support quality control and acceptance testing, and improve parameterization of predictive analytics and digital‑twin models for induction machines.

ABSTRACT. Situations arise in which it is appropriate to use a journal bearing to support a rotor spinning around an axis which is itself orbiting around a stationary centreline. If it is assumed that the orbiting speed of the journal bearing axis around the stationary centreline remains absolutely constant, a stability analysis may indicate that the system is stable. However, if it is recognised that the overall angular momentum of the system about the stationary centreline is what actually remains constant (or nearly constant) and that the orbiting speed may vary, then a stability analysis may return a very different result – indicating instability. This paper has been inspired by testing on an orbiting journal bearing. During this testing, a low frequency instability was observed in certain running conditions and the rumble remained unexplained for some time. A first analysis treated the journal bearing using only two degrees of freedom and this analysis suggested strong stability. However, when a third degree of freedom was added to the system (accounting for the ability of the orbit of the journal bearing axis to have non-constant velocity), a very different conclusion was found.

ABSTRACT. Compared with conventional flux-switching permanent-magnet (FSPM) motors, partitioned-stator (double-stator) FSPM motors offer higher torque density and reduced core saturation. By splitting the stator into inner and outer sections, a larger cross-sectional area is available because windings and permanent magnets are distributed across two stators rather than concentrated in one. This lowers peak flux density in teeth and yokes, mitigates saturation, and enables approximately twofold torque density relative to a conventional FSPM of equal size. However, placing the rotor between the inner and outer stators introduces structural challenges related to rotor dynamics, vibration, and deformation. The need to improve rotor dynamics and to ensure high reliability and continuous torque under fault conditions affecting the motor phases and magnets motivates this work. This paper introduces a novel multi-sector double-stator FSPM motor featuring a unique three-set, two-phase winding scheme (six-phase winding) that enables radial force generation and fault tolerance, thereby maintaining safe torque production. Coordinated current sharing synthesizes controllable radial-force vectors for active vibration damping in mechanically bearing-supported implementations, and for magnetic suspension in bearingless implementations. Finite element analysis of the motor validates the proposed concept.

ABSTRACT. In a rotating shaft supported by sliding bearings, self-excited vibrations known as oil whip and oil whirl occur at high rotational speeds. To prevent these vibrations, many high-speed rotating machines employ tilting pad journal bearings (TPJBs). TPJBs effectively suppress unstable vibrations by allowing the pads to tilt appro-priately, thereby eliminating the cross-coupled stiffness component of the oil film, which is the primary cause of oil whip and oil whirl. However, if the pads fail to tilt properly due to factors such as wear on the pad backing or the ingress of for-eign particles, the system’s stability is expected to decrease. Such malfunctions are likely to occur in specific pads rather than all pads simultaneously. Neverthe-less, the impact of a single malfunctioning pad on the stability of the shaft-bearing system has not been sufficiently investigated. This study numerically examines how the stability of the shaft-bearing system changes when a malfunction pre-vents a pad in a TPJB from tilting. Specifically, the stiffness and damping coeffi-cients of the TPJB are theoretically determined when a specific pad is fixed at its tilt angle in the static equilibrium state. Based on these coefficients, the linear sta-bility of the shaft-bearing system is analyzed. The results indicate that even when a single pad is constrained, the coupling terms of the stiffness coefficient do not become zero. Consequently, at high rotational speeds, the rotating shaft may be-come unstable, potentially leading to overall system instability.

ABSTRACT. Shaker-based vibration testing of structures is carried out in different engineering disciplines to characterise their dynamic behaviour. The purpose of this study is to enhance the force range of electrodynamic shakers by means of vibration exciters. Vibration exciters are investigated at the interface between the shaker and a test structure, aiming to achieve force amplification at tuned frequencies. Chains of mass-spring systems are tuned at particular frequencies. While keeping their spectral properties the same, their eigenvectors are tuned to satisfy equal modal magnitude criteria based on isospectrality. It is demonstrated that for the diagonal terms of the force transmissibility matrix, the modal magnitudes can be tuned to desired values. A vibration exciter for force amplification is optimised with both frequency and amplitude targets satisfied. A symmetric mechanical vibration exciter is designed to fit a particular test rig. The experimental test on the symmetric mechanical systems indicates amplification is achieved as desired.

ABSTRACT. This contribution presents an atypical approach to coupling a rotor system with a hybrid tilting pad and fixed pad journal bearing, aiming to suppress common oil-induced instabilities such as oil whirl/whip associated with fixed bearings and pad fluttering in tilting pad designs. The proposed configuration strategically combines the advantages of both designs. Specifically, the bottom part of the bearing is designed as tilting pads to ensure dynamic stability under varying load conditions, while the upper pads remain fixed to prevent pad fluttering and reduce structural damage caused by elastic contact between the rotor and movable upper pads. The computational model of the proposed hybrid bearing design is presented in this paper. It is based on an in-house solver of the Reynolds equations applied to each fluid film.

ABSTRACT. This study presents the experimental validation process of rolling and ball bearings designed to enhance the rotordynamic performance of the LM9000 Free Power Turbine. The LM9000, a high efficiency aeroderivative gas turbine, demands exceptional mechanical reliability and dynamic stability under varying operational conditions. To meet these requirements, a series of comprehensive test campaigns were conducted to evaluate bearings behavior under representa-tive loads, speeds, and thermal environments. The validation process included both component-level testing and full-system integration trials, focusing on crit-ical parameters such as vibration response, stiffness and damping characteristics, and thermal performance. Advanced instrumentation and data acquisition sys-tems enabled high-resolution monitoring of bearing dynamics, facilitating corre-lation with analytical models and finite element simulations. The findings confirm that optimized bearing selection and validation are key en-ablers for achieving robust turbine performance and reliability. This work con-tributes to the broader understanding of bearing dynamics in high-performance turbomachinery and supports future design iterations aimed at further enhancing operational efficiency and durability.

ABSTRACT. This work aims to provide further insight into the dynamic behaviour of electric drivetrain ball bearings in heavy machinery, with particular focus on slip occurrence and its impact on localised heating. Bearings are critical components within drivetrains, yet they operate under high loads and speeds that create harsh dynamic conditions, potentially leading to slip and accelerated wear. To investigate this phenomenon, a transient flexible cage bearing model is developed. Furthermore, a detailed local heating model based on the discrete convolution and FFT (DC-FFT) methodology is introduced to evaluate the severity of slip events. Real-world duty cycle data are employed to determine when and how slip occurs during operation. Findings indicate that slip can occur frequently under real operating conditions; however, its severity in terms of localised heating is strongly influenced by local speeds and loads at the slipping contact. Qualitative analysis reveals that, during nominal operation, localised heating remains relatively mild; in contrast, under more severe conditions, peak temperatures can rise to dangerously high levels, posing a risk of damage or even immediate failure.

ABSTRACT. Self-acting gas bearings are a popular solution for small scale turbomachinery. Many designs have been studied, and among those, three are very diffused and promising: foil bearings (FBs), herringbone- grooved bearings (HGBs), and tilting pad bearings (TPBs). Despite the many studies investigating these architectures individually, a direct benchmark between them is still absent. This paper proposes a side-by- side comparison of the three concepts specialized to one common applica- tion: a 14 mm rotor with a nominal speed of 150 kRPM. To achieve this, a numerical model was implemented, using a 2-D point mass rotor. Per- formance metrics were then calculated: static equilibrium positions and fluid film height, Campbell diagrams, logarithmic decrements, friction losses, and the full orbit response under a selected unbalance. The FB exhibits the largest fluid film heights and lowest losses, but higher rotor eccentricities; the HGB shows minimal eccentricities with higher losses; and the TPB provides superior dynamic stability, also at the expense of higher losses.

ABSTRACT. Modeling transmission systems equipped with Variable Geometry Torque Converters (VGTC) demands a trade-off between accuracy and computational efficiency. While detailed Computational Fluid Dynamics (CFD) simulations offer high fidelity, they are computationally intensive. By contrast, steady-state analytical models suitable for rigid body dynamics require geometric parameters that are difficult to obtain, often resulting in inaccurate torque predictions. This work presents an inverse design methodology that uses a multi-start optimization algorithm to estimate the fixed and variable geometric parameters of a VGTC from its reference torque map. The resulting parameters are refined to minimize error across the operating range for implementation in a rigid body model of a transmission system composed of the VGTC and two planetary gear sets, where the VGTC is used to control the system’s output. The method proved highly effective, achieving a mean squared error of only 1.5% compared to the reference data. This approach is a practical alternative in applications where CFD simulations are computationally prohibitive, transient dynamics are negligible, or direct parameter measurement is impractical.

ABSTRACT. This paper presents a systematic, stepwise methodology for control design, commissioning, model updating, optimization, and experimental validation of active magnetic bearing (AMB)–suspended rotor systems. Two critical challenges in AMB deployment are addressed by assessing the feasibility of preliminary designs, and mitigating inaccuracies and uncertainties during machine commissioning. The proposed approach begins with existing rotor-AMB system de- sign and modeling, followed by an analysis routine to validate feasibility, including observability, controllability, and the required control structure’s capability to provide adequate damping. The obtained base-line controller is initially implemented for commissioning purposes and subsequently updated based on an identified grey-box rotor–MIMO model. A multi-objective genetic algorithm (MOGA) is then employed to optimize the final controller, accounting for uncertainties and ensuring robust stability and performance. The paper generalizes the design steps from initial modeling to final solution implementation and demonstrates the routines for rotor–AMB model updating, including one bending mode, as well as controller optimization within PID-control framework. The methodology is validated on a 10 kW induction motor through both simulations and experimental testing. The results highlight a practical pathway from early design to robust implementation, improving the technical and economic viability of rotor-AMB system.

ABSTRACT. The torsional vibration behavior of rotating machinery depends heavily on the specific choice of coupling. However, the large number of discrete catalog variants and potential system combinations makes it impractical to manually navigate the design space. This complexity limits the ability to use automated search methods since the system and coupling options are not integrated. Furthermore, this disconnect prevents the effective integration of coupling selection into a holistic Model-Based Systems Engineering (MBSE) workflow. This paper presents a methodology that unifies these tasks using SysML v2, where both continuous parameters of the system such as the diameter of shafts and discrete elements like the couplings can be optimized. We developed a parametric library that formalizes vendor catalogs as computable model elements and a rule-based transformation that converts these system definitions into shaft-line finite-element models. This integration enables a multi-objective genetic algorithm to simultaneously optimize continuous system parameters and discrete coupling variants. A marine powertrain case study demonstrates that this approach systematically identifies catalog-realizable solutions while significantly reducing design cycle time.

ABSTRACT. Journal bearings play a critical role in providing damping and stability in rotor systems, yet conventional designs often exhibit instability at low eccentricity ratios. Helical groove journal bearings (HGBs) offer a promising alternative, though their dynamic characteristics remain insufficiently explored. In this study, a finite element formulation based on the Reynolds equation and the smoothed-pressure assumption is developed to investigate the static and dynamic performance of HGBs. Stiffness and damping coefficients are evaluated using small perturbations in displacement and velocity, and the influence of the eccentricity ratio and groove angle is examined. The results show that the HGB generates significant direct and cross-coupled dynamic coefficients even at low eccentricity ratios. Direct stiffness and damping generally increase with eccentricity due to enhanced hydrodynamic pressure at reduced film thickness, while cross-coupled terms exhibit trends governed by the groove-modulated pressure distribution. Variation of the groove angle within the investigated range of 25$^\circ$ to 45$^\circ$ demonstrates that smaller angles produce larger overall dynamic coefficients, indicating a strong sensitivity of force generation to groove geometry. Rotordynamic analysis using a simplified rigid-rotor model illustrates how eccentricity influences orbit evolution within the helical configuration. The findings highlight the importance of groove-geometry optimisation in tailoring dynamic coefficients and provide quantitative insight into the behaviour of oil-lubricated HGBs.

ABSTRACT. Torsional vibration is not typically measured or monitored in industrial machine trains, where lateral (radial) vibration is the primary focus. However, in systems with gearboxes, torsional vibration can manifest as lateral vibration due to torsional-lateral coupling. Fluctuating torque in the drivetrain induces lateral vibration through gear mesh interactions, resulting from gear rotor off-sets and transmitted torque, which generate radial (separation) and tangential forces. This paper investigates torsional-lateral coupled vibration in an integrally geared compressor within an air separation plant. The machine train consists of an induction motor, a gearbox, and a multi-stage centrifugal com-pressor. The compressor exhibited significant vibration amplitude fluctuations, with a predominant subsynchronous frequency, particularly when a nearby steel processing plant was operational. To analyze this phenomenon, oscillating torque on the motor shaft was measured using a strain gauge with a telemetry system, while power supply quality was assessed using a current transformer. The findings reveal a strong correlation between fluctuating lateral vibration, oscillating torque, electrical noise in the motor power supply, and the operational periods of an electric arc furnace. Both the compressor and the electric arc furnace are connected to the same power network. Additionally, the oscillating torque frequency coincides with the drivetrain’s first torsional natural frequency. A case study is presented to further illustrate these observations.

ABSTRACT. Electric machines, among other systems of high industrial relevance, are prone to dynamic instabilities (bifurcations) induced by the coupling differ-ent effects, e.g., mechanical vibrations and electromagnetic fields. Safe opera-tion and design rely, therefore, on being able to anticipate and mitigate such events over the full range of relevant system parameters and operating condi-tions. This task is usually formidable for rotor systems due to the size, complexi-ty and nonlinear nature of the corresponding mathematical models, especially when strong coupling with other physics is involved, but can be accomplished efficiently with the right numerical methods. To this end, the authors have re-cently developed the Angular Harmonic Balance Method (AHBM), which ex-tends the traditional HBM in such a way as to provide the basis for a full stabil-ity and bifurcation analysis of rotating machines wherein instantaneous angular speed is not assumed to remain constant. In this contribution, this approach is showcased through the numerical study of a flexible shaft coupled to a perma-nent magnet synchronous motor, leading to a detailed cartography of all possible dynamic regimes attainable under the variation of crucial parameters.

ABSTRACT. Unbalanced magnetic pull (UMP) can arise in electric motors due to asymmetry in magnetic flux. The use of parallel windings to reduce UMP has been studied for both induction and permanent magnet (PMSM) machines and it has been shown that with such windings, oscillatory forces occur in the lateral direction. It was also shown that the circulatory forces occur under some conditions which could have a destabilising effect. Bridge configured windings (as introduced by Khoo et al [1]) present a different option for connecting windings within a phase. It can be shown that when the bridge is shorted, bridge configured windings become equivalent to parallel windings, and when the bridge is open, they become equivalent to series windings. In this work, the effect of bridge windings on UMP is investigated on an example of a permanent magnet motor, with different values of impedance across the bridge considered. The ability to vary resistance across the bridge is considered as an option to reduce the destabilising effect of parallel-type connection in the operating regions where circulatory forces occur.

[1] Khoo, W.K.S., Kalita, K. and Garvey, S.D., 2011. Practical implementation of the bridge configured winding for producing controllable transverse forces in electrical machines. IEEE Transactions on Magnetics, 47(6), pp.1712-1718.

ABSTRACT. To understand the behavior of a gear system, it is important to study the dynamic response, which adds a requirement that must be fulfilled in gear design. Moreover, the load requirements must be met to ensure the system’s integrity under the applied load spectrum, and thus, material selection is decided accordingly. However, the use of light materials, if possible, is in high demand for weight reduction and energy-saving purposes. Gear dynamic behavior depends on the variation of tooth mesh stiffness. Then, material stiffness and geometry have a big role in the generated vibration response. The current paper studies the dynamic behavior of combined material gear systems. That means, for lighter gear systems, plastic (polyamide) material is used for one or both of the engaged gears. Gear mesh stiffness is calculated analytically for four different material cases of the pinion-gear system: namely, steel-steel, polyamide-polyamide, steel-polyamide and polyamide-steel cases. The calculated stiffness is introduced to a gear model for generating the dynamic response. The system dynamic response is generated using a 6 DOF gear dynamic model. Introducing a combination of materials has the advantage of reducing weight if the load requirement is met. However, the time-varying gear mesh stiffness and the dynamic behavior are changed accordingly.

ABSTRACT. An input-driven inverted slider–crank mechanism with an internal morphing slider is investigated for tilting the spin axis of a rotor–propeller assembly mounted at the distal end of the coupler link. The mechanism kinematics im-pose a time-varying propeller axis direction, such that axis reorientation in-troduces gyroscopic coupling through the rotor angular momentum. A cou-pled model is developed by combining the closed-loop kinematics of the in-verted slider–crank linkage with a minimal rotordynamic formulation based on rotor angular momentum balance, enabling prediction of gyroscopic mo-ments and transmitted support loads as functions of rotor speed and axis-tilt (morphing) rate. The governing equations are evaluated in MATLAB to compute the tilting-axis response, gyroscopic moments, and equivalent sup-port loads transmitted through the mechanism during morphing, using thrust and reaction torque models derived from published motor–propeller perfor-mance data. The results indicate that gyroscopic effects and reaction torque can dominate transmitted loads during rapid axis tilt, highlighting the need for rotordynamics-aware sizing of joints, bearings, and actuators in propul-sion-integrated morphing mechanisms.

ABSTRACT. This paper presents a design methodology for solving a specific problem that arose in the context of a research programme exploring the dynamic behaviour of different forms of bearing support dampers over a range of temperatures. The necessity to explore the dynamic behaviour requires a stiff connection between the external surface of the test article and a set of force measuring devices mounted on a large surrounding seismic mass. However, to allow testing over a range of temperatures, we require to allow the testpiece to expand and contract without transmitting strong forces across the force sensors. Taking these two requirements together, we see that it is necessary to produce a “connection” between the seismic mass and the testpiece that is stiff for motions characterised by 1 nodal-diameters (N=1) but flexible for motions characterised by 0 nodal-diameters (N=0). The proposed solution employs flexure features in the form of ligaments and slots integrated into a thermally loaded, prestressed component. The ligaments function as springs, providing flexibility for thermal expansion absorption (0D-flexible) while maintaining high stiffness for conveying lateral/transverse forces (1D-stiff). The testing requires that alternating forces can be applied to the inner portion of each testpiece while the outer portion of that testpiece is held fixed. The frequencies of the alternating loads are expected to reach up to 500 Hz. The design of the flexure features has to be optimized for stiffness, stresses, modal frequencies, and life. Finite Element Analyses were performed using ANSYS to evaluate the component’s strength and stiffness across all loading conditions, including the interference-fit prestress. Due to limitations of life analysis within ANSYS, stress results were exported and analysed using a bespoke Python script. The Manson-McKnight method was employed to calculate mean and alternating stresses, followed by the Goodman criterion to assess the infinite life safety coefficient. This paper details the design process, optimization, and analysis procedures, that was used in a solution that effectively allows thermal flexibility with mechanical stiffness and meets durability requirements.

ABSTRACT. Circular journal bearings are widely used in industry. Although their simple geometry offers advantages such as low cost and ease of manufacture, these bearings can be prone to oil whirl and oil whip phenomena at high speeds and low loads. The search for alternatives to prevent the appearance of previously mentioned instabilities has encouraged the development of non-circular bearings that operate with more than one active oil film zone. Besides improving the shaft stability, these bearings also reduce the power loss and increase the oil flow, resulting in a lower rise of lubricant film temperature. Therefore, the present study aims to investigate the thermohydrodynamic behavior of a twin-grooved two-lobe journal bearing based on a CFD model developed in OpenFOAM, which considers the cavitation, turbulence, and conjugate heat transfer effects. Given the load applied to the bearing, the equilibrium position of the shaft was determined by using the Newton-Raphson iterative method. In general, the comparison between the CFD results and experimental data reported in the literature showed good agreement, with most relative differences remaining below 7%.

ABSTRACT. Supercritical CO2 (SCO2) power cycles offer higher potential cycle efficiency compared to steam Rankine and air breathing Brayton cycles, especially at elevated temperatures. CO2 is inert, inexpensive, and relatively safe compared to other working fluids. The high mo-lecular weight results in relatively small equipment with impressive power densities re-ducing the overall size of the power plant. In fact, SCO2 power turbines represent the highest power density (power-to-weight ratio) of any terrestrial turbines, second only to liquid rocket engine turbopumps. The high pressure, speed, and density of the operating fluid creates many rotordynamic challenges for the compressors and turbines used in the cycle. This paper outlines many of these challenges and describes the designs to mitigate these concerns including rotordynamic stability, synchronous response traversing bending critical speeds, and non-synchronous excitation from the flow field. The equipment de-scribed were part of the U.S. Department of Energy programs name Apollo (centrifugal compressor develop) and STEP (10 MWe power plant demonstration). Advanced technol-ogies developed for the oil and gas industry including squeeze-film damper bearings, laby-rinth seal swirl brakes, and hole pattern damper seals were employed in the design. Vibra-tion data from full power tests will be presented including upsets like emergency shut-downs (ESDs).

ABSTRACT. Flywheels as an energy storage system hold kinetic energy in a fast-rotating mass. They have several advantages like low maintenance, high efficiency, and long lifetime. Modern systems magnetically levitate the rotor in a vacuum under normal operation. Should the magnetic bearings malfunction, mechanical fallback bearings – the so-called touch-down bearings – prevent destruction of the system. The out-er rotor design often applied in modern flywheels requires special planetary touch-down bearings. Instead of one rather large bearing, several smaller bearings are distributed equally around the circumference of the stator. In this design, the bear-ing forces are often in the range of their static load rating. Hence, it is of high rele-vance to be able to gather data on the forces occurring during a dropdown event to assess its severity and decide whether the touch-down bearings must be replaced. Force sensors are generally not integrated into the bearings due to their cost. Thus, this work aims at estimating the contact forces based on the rotor position data that must be measured in any case to control the magnetic bearings. It establishes that with some simplifications the maximum bearing contact forces are proportional to the maximum translational velocity of the rotor and validates this assumption with experimental data from nearly 200 drop-down experiments. Additionally, this work shows that the relation between maximum force and velocity can be calculat-ed based on the mass of the rotor and the contact stiffness.

ABSTRACT. The growing demand for high-efficiency power conversion systems in the megawatt range raises the question of which machine topology is most suitable. While sub-megawatt systems are typically built as integrated turbogenerators, multi-megawatt systems are often realized as separated turbine–generator units. A recently introduced hybrid concept integrates the turbine shaft into the generator structure using a third radial bearing and flexible coupling. This configuration combines the rotor-dynamic benefits of separated units with the compactness of an integrated structure, while eliminating the need for shaft sealing. This study investigates the dynamic behavior of the hybrid configuration with three radial active magnetic bearings, compared to a conventional two-bearing setup. Four shaft designs are analyzed, featuring different placements of the turbine segment and two-stage turbo-compressor on generator and extension shafts. The analyses focus on critical speeds, load capacity, and vibration response under unbalance excitation. Among the examined designs, a layout with favorable rotor-dynamic characteristics is identified. The results provide insight into the trade-offs and advantages of alternative configurations, supporting future design decisions for few-megawatt-scale turbomachinery.

ABSTRACT. This paper presents the rotordynamic modelling and analysis of a vertical rotor designed for a flywheel energy storage system (FESS) demonstrator. The study focuses on developing an analytical and semi-empirical model to predict the dynamic behaviour of the high-speed rotor. The modelling framework incorporates gyroscopic effects and bearing stiffness to evaluate critical speeds under varying operational conditions. Attention is given to the influence of aerodynamic drag and heating within a vacuum chamber, which remain significant at high rotational speeds. These aerodynamic losses and associated heating phenomena are embedded in the model, as they directly affect the system performance, material thermal constraints, and rotor stability. The study uses these effects to improve the overall energy retention capability of the system. In addition, a law governing the turbopump duty cycle is formulated to reduce parasitic power consumption during vacuum generation and thermal control. This duty cycle optimization is expressed as a function of the chamber pressure, heat generation rate, and rotor speed. Also, the combined rotordynamic and thermofluidic modelling provides an integrated understanding of the interactions within a family of rotating machines, offering a valuable guideline for the design of high-efficiency systems for energy-related applications.

ABSTRACT. Dynamic load profile emulation is essential for characterizing and validating electric drive systems under realistic operating conditions. Conventional test benches typically rely on a single load motor with torque control to repro-duce the time-varying torque demands experienced in application-specific duty cycles. However, this approach can be limited when the required torque exceeds the capacity of the load motor, or when the load dynamics are too fast relative to the achievable bandwidth of the torque controller. In this study, we investigate alternative strategies for accurate load emulation by combining physical inertia and torque control, using sinusoidal speed and torque profiles as a representative test case. Three configurations are identi-fied: (i) pure inertial load using a flywheel, which is effective for reproduc-ing fast dynamic profiles where motor-based control is inadequate, (ii) pure motor-based load, suitable when the load motor can deliver the required peak torque and bandwidth, and (iii) hybrid configurations combining a fly-wheel with a smaller load motor, where the inertia passively supplies part of the torque demand while the motor compensates the residual torque to track the profile. The hybrid approach has not been widely considered in literature but offers significant potential, since it enables the use of a load motor with substantially lower rated power than the device under test, reducing test bench cost and complexity. The main outcome of this work is a methodology for selecting between these three configurations, depending on the character-istics of the load profile and the limitations of the test bench.

ABSTRACT. Spur gear-rotor systems play a crucial role in power transmission in automotive, aerospace, and industrial machinery. Real-world operation often introduces im-perfections, such as the angular misalignment and pinion-gear profile errors, re-sulting from manufacturing tolerances, assembly inaccuracies, or operational loading. These imperfections lead to transmission error (TE) which degrade effi-ciency and excite vibrations in the system. In this work, profile errors and eccen-tricity are considered on the pinion and gear, and are analyzed using a 14-degree-of-freedom (DOFs) mathematical model that represents six translational and six rotational DOFs of the pinion-gear assembly; and one each torsional DOF for the motor and the load, respectively. An analytical model is used to model the profile error and eccentricity, which are then added to the 14 DOFs model. The parame-ters, such as pinion accelerations and TE are obtained, which clearly show effect of errors on the dynamics of the system.

ABSTRACT. This paper presents the kinematic and dynamic analysis of a novel mechanism for piston engines based on a rigid-body formulation; tribological and thermal insights are included. The proposed concept aims to replace the conventional crank-rod-piston mechanism with a simplified layout for converting reciprocating piston motion into rotational output. In the new mechanism, neither a crank nor a connecting rod is required. Instead, the piston undergoes a combined motion of axial translation and continuous rotation about its own axis, resulting in a system composed of only two rigid links: the piston(s) assembly and the fixed frame. The inherent simplicity of the mechanism, together with the rotating-piston principle, leads to several potential advantages. Continuous rotation of the piston promotes a persistent hydrodynamic lubrication regime between the piston rings and the cylinder liner, significantly reducing friction losses by avoiding zero-velocity conditions, boundary lubrication, and squeeze-film effects. Moreover, piston rotation alters the thermal efficiency within the cylinder, potentially increasing mixture enthalpy. The analysis presented focuses on the kinematic relationships and dynamic behaviour of the mechanism under rigid-body assumptions, and on approximate analytical formulations for the associated tribological and thermal phenomena. This work establishes the theoretical foundation for subsequent multi-body, elastohydrodynamic, and experimental investigations.

ABSTRACT. Permanent Magnet Synchronous Machines (PMSMs) are increasingly deployed in traction and aerospace applications due to their high efficiency and power density. The reliability of these machines is a primary concern, where mechanical integrity under electrical faults is critical. A significant risk in PMSMs is the occurrence of a short circuit between the phases of the stator winding. The resulting short circuit torque can be severe, with the initial peak often exceeding 200\% of the rated torque. To further understand the effect of the short circuit torque on the torsional response of the rotor, this study investigates the electromechanical interactions occurring during a short circuit event, using two different approaches and comparing them. First, the short circuit torque is analyzed for constant speed, and it is provided as an input to a transient torsional rotordynamic model. In second method, a feedback loop approach is implemented by transferring the torque data from electromagnetic analysis to the rotor dynamic model, and sending rotational velocity data back to the electromagnetic analysis. This paper provides a critical bridge for enhancing the mechanical robustness and reliability of PMSM rotors under short circuit fault. The contribution is the demonstration of two methodologies to analyze the mechanical consequences of electrical short-circuit faults; first as a direct input only, and second, as a feedback loop. By analyzing the fault condition and its effect through these two approaches, the most suitable method can be recommended to ensure long-term reliability of PMSM machines and prevent catastrophic mechanical failures.

ABSTRACT. Electromechanical systems increasingly need efficient surrogates to replace time-consuming traditional models while offering real-time and accurate solutions. Conventional methods mainly depend on finite element method (FEM) and computationally very expensive, making it difficult to fulfill real-time demands of simulation and dynamic control. Machine learning–based surrogates provide a promising alternative by learning complex nonlinear relationships directly from data, while incorporating physical priors and preserving the accuracy necessary for reliable system modeling. This paper focuses on neural network (NN)-based surrogates for active magnetic bearing (AMB) modeling by developing two NN-based surrogate models to replace the FEM-based lookup table: a data-driven NN and a physics-informed neural network (PINN). The developed models use FEM datasets for training, with coil current and rotor position as inputs for force estimation. These surrogate models are first validated against FEM lookup tables referenced as benchmarks through closed-loop testing, integrating them into a position-controlled AMB system with inner current control loops. Offline validation with experimental data from an axial AMB machine shows that both NN surrogates accurately replicate FEM lookup table behavior while enabling real-time evaluation for rotordynamic control. The results demonstrate that both data-driven and PINN have comparable prediction accuracy; however, PINN, due to physics-based constraints in its loss functions, offers more stable and physically consistent behavior.

ABSTRACT. The conventional design and optimization of tilting pad journal bear-ings(TPJBs) highly rely on expert experience and repetitive, high-fidelity numerical computations. This heuristic-driven iterative process is both time-consuming and inefficient. Although machine learning-based approaches have been applied to this domain, their dependence on high-quality datasets and the black-box nature of their decision-making processes limit their devel-opment in terms of physical interpretability and generalization capabilities. To address these challenges, this paper proposes a collaborative TPJB optimiza-tion framework that positions a Large Language Model (LLM) as the intelli-gent core and utilizes a high-fidelity thermo-elasto-hydrodynamic (TEHD) coupled solver of the TPJB as the evaluation module. Within this framework, the LLM, requiring no domain-specific fine-tuning, performs deep reasoning on the performance feedback from the FEM analysis based on fundamental theories of fluid dynamics and oil-film lubrication to drive the evolution of bearing structural designs. This study demonstrates that, compared to tradi-tional heuristic optimization methods, the collaborative optimization frame-work centered on the LLM enables a rapid search for optimal TPJB struc-tures and facilitates effective performance trade-offs. It effectively integrates multi-modal information streams, opening up a new pathway for the rapid design and optimization of complex TPJBs.

ABSTRACT. The lubrication condition of journal bearings plays a key role in the performance of rotating machines. A high oil supply to the bearings reduces shaft eccentricity and vibration, while a controlled degree of oil starvation can decrease power losses and bearing temperature. Therefore, precise control of the oil supply is essential for optimal performance. However, oil inlets are not always monitored individually, making it difficult to identify and locate leakages or clogging in the oil lines. This paper investigates the identification of oil supply flow rates in journal bearings based on rotor displacement measurements. Experiments conducted on a test rig with a rotor supported by two journal bearings provided vibration data under varying oil inlet flow rates. A detailed analysis of the vibration signals was performed in both the time and frequency domains, highlighting the main changes in the rotor’s orbit and directional spectrum. Subsequently, representative features were extracted from the vibration responses and used as inputs to a neural network, which was trained to predict the delivered flow rates based on the measured vibration. This work enhance the understanding of the dynamic effects of oil starvation in hydrodynamic journal bearings and contributes to the development of an intelligent diagnostic approach for identifying lubrication-related faults.

ABSTRACT. Shrink fits are commonly used in many rotor systems for mounting discs/machine elements. The shrink fitting mounting on the rotor alleviates the stress raisers such as a keyway but needs to have a tighter interference fit for a positive drive. The contact pressure from the interference fit generates localized stress and deformation in the shaft and the mounted element. The contact stress in the axial direction leads to local stiffening conditions and influences the eigenvalues of the rotor. This talk will highlight a new approach to account for the contact pressure and associated stress conditions at the shaft-disc interface. The influence of different interference fit levels and shaft-disc configurations (disc diameter and thickness) on the natural frequencies will be detailed. The proposed approach is validated using an experimental modal analysis and the results confirm the accuracy of the new approach to model the shrink-fit assembly. In contrast to the empirical approach of modeling the shrink fit from the past literature, the present approach offers a more systematic and better way of modeling and of accurately predicting the associated bending natural frequencies of the shaft-disc assembly.