Advances in Dynamics of Vehicles on Roads and Tracks II

A multi-degree of freedom dynamic vibration absorber (MDOF DVA) to suppress the vibration of the carbody of high-speed trains is proposed. The MDOF DVA is installed under the carbody, the natural vibration frequency of which are designed as a dynamic vibration absorber for lateral motion, bouncing, rolling, pitching, and yawing of the carbody. A high-speed train dynamics model including an under-carbody MDOF DVA is established. Based on the virtual excitation method, the vibration control effect on each DOF of the MDOF DVA on the carbody is analyzed. Results show that the MDOF DVA can absorb the vibration of the carbody in multiple degrees of freedom.

Within the Shift2Rail projects Pivot2 and NEXTGEAR, an innovative Metro vehicle with single axle running gear and only one suspension step is proposed. A composite material running gear frame is developed to be used both as structural and as suspension element. The design with only one suspension step can significantly degrade the passengers ride comfort. Thus, active modal control is implemented both in lateral and vertical direction to increase the performance of the system. The running gear frame is modelled in Abaqus® as well as the carbody. Structural modes of both elements are implemented in SIMPACK®. A hydraulic actuator model is developed in Simscape®, where two pressure-controlled valves are used to control the pressure inside the chambers of a double acting hydraulic cylinder. A co-simulation environment is then established between SIMPACK® and Simulink®. The vehicle is running with speeds between 10 and 120 km/h. Active modal control makes it possible to maintain ride comfort levels of conventional bogie vehicles with this innovative single axle and single suspension step running gear, promising substantial weight savings of about 400 kg/m. The single axle running gear solution with active comfort control developed here can be an attractive alternative to bogies, providing reduced Life Cycle Costs.

The running gears of DLR’s long-term project Next Generation Train utilize independently rotating wheels with mechatronic track guidance, direct drives close to the wheels and are optimized for low weight. On the basis of encouraging research results so far, DLR decided to design and build a true scale prototype of the NGT running gear and use it as a research facility. It is the intention to improve, validate and demonstrate the mechanical and mechatronic design, sensor and actuator lay-out step by step and finally approach the Technology Readiness Level 6. By the end of 2022, this prototype will be put into operation considering low speed scenarios up to max. 5 m/s at an in-house integration test rig. This is the current task, which is reported on in the paper. However, this work is supposed to prepare advanced performance experiments up to 350 km/h on external roller rigs and at railway test tracks later on.

Utilization of independently rotating wheel attains benefits in low floor transit, while sacrifices the steering performance because of the deficiency of longitudinal creep forces. As a promising structure for improving steering performance of vehicle, independently rotating wheel with negative tread conicity was proposed previously, but the research on such structure still stays at the stage of simulation and scale model, while full scale vehicle experiment has not been implemented so far. To this end, the experimental validation of vehicle assembling independently rotating wheels with negative tread conicity is firstly implemented in present research. The vehicle system compatible with full scale rail is constructed and the stand of better curving performance is supported by experiment results. Moreover, as the traction is necessary for the practical operation of vehicle, the utilization of traction motors for further improving the curving performance of vehicle is naturally proposed in this research. The active steering control is attempted, and the results indicate that the further improvement of curving performance is achieved by means of proposed control scheme.

With the trend of car-body light-weighting, car-body structural vibration is imposing a new challenge to the vertical ride comfort which, however, is difficult to be improved by passive suspension components. Semi-active primary suspension (SAPS) is studied in this work to improve the vertical ride comfort with special attention on the suppression of car-body first bending mode. At first, a coupling effect between car-body bending mode and bogie pitch and longitudinal vibrations is analyzed. Then a simplified vehicle model is established to represent the vehicle dynamic behavior, showing a good agreement with a detailed multibody vehicle model in SIMPACK, integrated with the finite element model of the car-body. Skyhook controller is applied in the SAPS, capable of mitigating vibration from the car-body bending mode. LQG controller is also developed, showing further reduced vibration not only related to the first bending mode but also in a lower frequency range associated with vibrations due to rigid modes.

Railway bogies are generally maintained preventively within certain time periods. Vehicles run for a long time (up to thirty years) so that about one-third of lifecycle costs are caused by maintenance. Condition-based predictive maintenance strategies offer economic improvements of up to 15 percent and an increased availability of up to 100 percent. By implementing a sensor based diagnostic system using artificial intelligence techniques and business analytics, maintenance can be optimized.This paper presents a concept for the detection of faulty mechanical components, such as dampers and springs, of railway bogies. On the one hand, the work deals with the analysis of time series data by evaluating power spectral density and transfer functions. On the other hand, it shows results of the training of machine learning models based on features. Various multi body simulations have been done to develop a physical understanding of the effects of the individual fault modes. Beside this, the available data set for the analysis also includes real measured data from test rides with faulty components.The developed algorithms have been deployed on a Siemens commuter train. In addition to validation results, this work also shows an example of the successful detection of real failures in operational mode.

The on-line condition monitoring system can monitor the condition of the train by measuring the wheel load and lateral force of the train passing through the measurement point over a long period of time. With this system, the derailment coefficient and wheel load balance of the sharp curve section have already been monitored. On sharp curves section, it is very important to evaluate the steering performance of the bogie and the lubrication conditions at the contact point between the wheels and rails. In addition, many types of bogie are running on the same section. It is efficient to evaluate the steering performance of the bogie and the lubrication conditions at the contact point between the wheels and rails from the ground side. In this study, new indicators consisting of wheel load and lateral force were introduced and evaluated to add train steering performance to monitoring system.

The use of Electronic Control Unit (ECU) in vehicle requires them to be widely tested before the first physical prototype is developed. This is even more important when the vehicle is unmanned because most critical tasks are demanded to the control unit without a human supervision. This paper presents a Hardware-in-the-loop (HiL) test bench used to validate the control algorithm of the Vehicle Control Unit (VCU) of a Driverless Railway Vehicle (DLRV). The VCU has the duty to control the traction motor and the pneumatic braking systems but has also to control the hybrid powertrain and its configuration. The test bench is then built in Simulink Real Time environment where the vehicle model is implemented. The model communicates with the VCU through CAN BUS communication as it will operate on the real vehicle.

Light rail vehicles (LRV) are becoming more attractive for urban centres as a sustainable mass transportation solution. The tight curves and short transitions that characterise urban LRV networks lead to high wear and undesirable vehicle dynamics that can be avoided with active suspensions or traction control algorithms. This paper presents a comparison on the dynamic performance and curve negotiation of LRV with solid wheelsets and independently rotating wheels (IRW) with different traction control systems. Two multibody simulations were conducted to compare slip, angle of attack (AoA) and other vehicle dynamics parameters. The traction control was set to operate at the maximum traction conditions with a slip set point. The wheel-rail contact model included the effects of slip-dependent friction variations. It was found that the LRV with IRW reduced the lateral wheel-rail contact forces, traction coefficients and wheel torque when negotiating a curve, while maintaining the AoA performance. The results indicate that with an appropriate traction control algorithm, a LRV with IRW can be more track friendly than a solid wheelsets LRV when negotiating a curve in maximum traction conditions.

This paper presents a data-driven robust controller for the active steering of driven independently rotating wheels (DIRW). Associated with a two-axle DIRW vehicle, a reinforcement learning controller called Deep Deterministic Policy Gradient (DDPG) is applied to improve the guidance and curve-negotiation behaviour of the DIRW system. We implement deep neural networks in DDPG to learn complex vehicle behaviours by training with data generated dynamically from non-linear simulation models. The controller can achieve adaptive optimization through online training episodes. The DDPG controller’s effectiveness is verified by the co-simulation method: the DIRW railway vehicle’s dynamics model is established in SIMPACK, and the data-driven controller is trained and deployed in MATLAB. The simulation results show that for the DIRW system, the proposed control approach can improve the IRW’s running performance and can significantly reduce the wheel-rail wear in both straight and curved tracks.

The wheel transition area in railway crossings is subjected to impact loads that cause an accumulation of structural degradation in crossing panels over time. This degradation leads to high maintenance costs and possibly traffic disturbances. There is therefore a demand from infrastructure managers to monitor the condition and predict maintenance needs for these assets without the need for regular on-site inspections. One solution for operational condition monitoring is to observe the structural response of the crossing under traffic loading via embedded accelerometers. From these measurements, relative changes in track dynamics over time can be observed. To derive a condition or predict maintenance needs, however, these measured accelerations need to be related to the status of the asset. A framework for this where measurement data, simulation models and maintenance history are combined to build an online model that can assess the status and predict future maintenance needs for a material asset is often called a Digital Twin. This paper will present a Digital Twin framework that uses measured accelerations, climate data, scanned running surface geometry and a multi-body simulation (MBS) model to estimate the status and degradation rate of crossing panels. Method developments for this framework are demonstrated for two in situ crossings.

The emerging integration of mechatronic systems in modern railway vehicles enables significant improvements with respect to safety, comfort, and wear reduction. To fully exploit the potential of mechatronic systems, the German Aerospace Center (DLR) complements its validated estimation and control concepts in the field of lateral vehicle dynamics with approaches for longitudinal dynamics. The present work introduces an adhesion-based maximum-seeking brake control that offers a benefit in contrast to slip-based approaches especially in safety critical scenarios.

Coupling vibration between maglev vehicle and track can easily occur when maglev train levitate on a light or low stiffness track beam. To overcome this, track beam nowadays always got big mass and stiffness which directly result in massive increasing of construction cost of maglev line. It would be a much more economical way to solve this vibration problem by improving levitation control strategy. In this paper flexible levitation control strategy and traditional rigid levitation control strategy are tested on a small scale magnet-track beam coupling test rig. For magnetic levitation system, because of the strong nonlinear characteristic, test rig study is more accuracy than computer simulation. The test rig is designed based on similarity theory and have one magnet and one simply supported track beam. Results show that comparing with rigid levitation control strategy, flexible levitation controller can make the system more stable. Coupling vibration can be actively decreased under flexible control strategy.

Wheel polygonalisation, as a common phenomenon in railway vehicles, will worsen the dynamic effect of wheel-rail and affect running safety. Detection of the polygonal wear is essential for railway vehicle maintenance and running safety. Therefore, a novel polygonal wear detection method based on vehicle vibration measurements is proposed in this paper. Firstly, the axle box vertical acceleration signal is decomposed into multiple intrinsic mode functions (IMFs) by the variational mode decomposition (VMD) algorithm. Then, the observed vibration signal composed of multiple IMFs is analyzed by the independent component analysis (ICA) algorithm, and the independent component related to polygonal wear is selected according to their correlation coefficients. Finally, the optimal independent component is used to calculate the order and amplitude of the polygonal wear by the inertia principle. To verify the effectiveness of the proposed method, the simulation signal and axle box acceleration signal of measured data are implemented. Experimental results demonstrate that the proposed method can effectively estimate the order and amplitude of the polygonal wear.

As an important part of the third rail current collection system, current collectors are widely used in urban rail transit. Different from the common urban rail current collector, the APM 300 collector shoe is in contact with the three planes of the unique C-shape power supply rail. Besides, there are numerous rail seams in the power supply rail line, which may force the collector shoe detached from the power supply rail. In this paper, the dynamic characteristics of the current collector running on the continuous power supply rail and rail seams are analyzed, respectively. In the contact module, Point-Plane force element is used to calculate the contact force. The simulation results show that reducing the shoe mass or increasing the stiffness of the restoring tension spring can decrease the shoe-rail contact force and the lateral acceleration of the collector shoe. And the joints’ lateral and vertical deviations need to be controlled within 1.5 mm.

The research on active control of the pantograph has been intensively studied, however, the control results in practice are not satisfactory as expected, for the existence of time delays coming from the input and output measurement. This paper focuses on the problem of time delays in the designing of the control strategy, including the sensor signal input and the action of the actuator. A multibody dynamic pantograph model is established based on the relative coordinate method, combined with the rigid catenary established by the modal superposition method. Firstly, the performance of the traditional PID controller with signal time delay is studied, including the separation or combination of sensor signal input delay and actuator signal output delay. Then, a novel fuzzy PID controller with an improved rule library is proposed, which further improves the control performance of the PID controller. Finally, to deal with the time delay of sensor and actuator well, a dynamic predictive fuzzy adaptive PID (DMC-Fuzzy-PID) controller is proposed by combining dynamic matrix control (DMC) with the improved fuzzy PID controller, which may support the engineering application of active control of pantograph.

Rail corrugation is a degradation phenomenon that affects almost all the railway lines and for which up to now a definitive solution does not exist. It consists in the appearance of a quasi-periodic irregularity on the running surface of the rail, causing high dynamic loads and, thus, possible failures in vehicle and track components, ground-borne vibrations transmitted to the buildings nearby the infrastructure and acoustic pollution. The only effective treatment to mitigate the rail corrugation problem is the periodic grinding of the rails, performed by dedicated vehicles. A continuous monitoring system for rail corrugation growth has been developed. The roughness profile on the rail running surface is estimated from axle-box acceleration measurements performed by an in-service vehicle. The FRF of the measurement system is calculated from a frequency domain model of the vertical dynamics of the wheel-rail interaction. The estimate of the longitudinal rail profile allows the identification of the track sections where corrugation is present and the monitoring of its evolution. Some differences between estimated and measured peak-to-peak amplitude are present, but the causes have been identified and should be overcome with further experimental tests and refinements of the mathematical model.

While there has been a growing amount of research on modelling and active control of pantograph-catenary systems (PCSs) for stable current collection of electrified rail vehicles, the coupled nature of the pantograph and PCS dynamics has not been explicitly addressed. This work provides a comprehensive illustration of states and their interactions in the vertical degree-of-freedom of coupled PCSs. A displacement-dependent contact wire mass is coupled to its vertical displacement for a more realistic wire behaviour. Conditional masses, one-way damping, and equal and opposite characteristics of the contact force have been modelled for enhanced PCS interaction modelling and contact force control. The geometric contact kinematics has been derived to solve for the vertical wire displacement. The formulation of contact force/position control objectives is presented for the resulting system and is applicable to similar class of systems. Simulation results show that the contact force from typical passive pantographs is lower than desired values and they take longer to stabilize from contact losses. In contrast, a PI controller for an active pantograph setup can better handle initial contact gaps and regulate the resulting contact force to the desired value.

Longitudinal running stability is particularly important for vehicle running safety. The longitudinal running stability of this type of vehicle under braking and traction conditions in straight line is studied in this paper. Based on Eulerian-Lagrange method, the dynamics model of N-module tram was established, and the stability of the system was analyzed by phase-plane method. The results show that under the action of longitudinal force, the tram is in the states of tension and compression: the tension state inhibits the vehicle's gyratory motion, while the compression state promotes the vehicle's gyratory motion. When the tram is in the state of compression, the critical compression force and the critical rotation stiffness have a relationship of 2L times (L is half length of the carbody). When the compression force exceeds the critical value, the vehicle’s rotation angle increases rapidly and finally converges to another stable odd point in the phase-plane space. Finally, SIMPACK software is used to build a dynamic model of a three-module single-carbody tram, and the simulation test of tram emergency braking is carried out. The results show that during the braking process, the tram keeps running in a straight line state when the rotational stiffness of the train is over the critical value, otherwise the yaw angle increases rapidly, and the vehicle shows a zigzag folded state. The simulation results verify the correctness of theoretical analysis and this work provide a theoretical guidance for the design of single-carbody low floor tram.

With the increment of running mileage, polygonal wear appears in the circumferential direction of the wheel, which will aggravate the impact vibration between the wheel and the rail and affect the running safety and stability of the vehicle. Due to the heavy load and the higher running speed, the effect of wheel polygonal wear on the express freight train cannot be ignored. Besides, the running safety and stability may be further deteriorated when the train runs in the complex environment especially on the bridge and in the crosswind. In this paper, the dynamic model of train-track-bridge are established, and the effects of wheel polygonal wear on the vehicle and track-bridge structure when express freight train runs on the bridge in the crosswind are investigated. The results show that the polygonal wear intensifies the dynamic responses of vehicle and bridge. The amplitude and the order have different effects on the safety and stability of the vehicle.

Car-body hunting motion (CHM) has a significant impact on the ride comfort and safety of high-speed trains (HSTs). In this paper, tracking tests of car-body hunting phenomenon occurring on a high-speed train travelling on a Chinese railway line were carried out. In the first stage, the information such as car-body hunting feature and the train running conditions were investigated. The potential causes and key influence factors of this abnormal phenomenon were studied. In the second stage, several countermeasures such as reprofiling wheels (RW), adjusting the height of the car-body supporting point (AHCSP), were put forward. The comparative tests after the adjustment measures were carried out. The test results show that the vehicle sways abnormally during the acceleration stages and high-speed passing station stages. When the CHM appears, the vehicle lateral comfort index is obviously larger than the vertical index, which obviously exceeds the limit value. The main frequency peak of the car-body lateral acceleration appears at 1.22 Hz, which is close to the natural frequency of car-body lower center roll mode. Many frequency doubling components of this frequency appear at the same time. Through the successive implementation of CHM and RW measures, the CHM phenomenon has been greatly improved, the abnormal vibration of the car-body has disappeared, and the train has resumed normal operation.

The carbody shaking phenomenon can greatly reduce the ride comfort of high-speed trains. To investigate the abnormal vibration feature and root cause of the carbody shaking phenomenon occurring on Chinese high-speed trains, extensive field measurements have been conducted on a Chinese high-speed railway line. A three-dimensional rigid-flexible coupling dynamic model of a high-speed vehicle is developed to reproduce the shaking phenomenon and reveal the key influence factors, in which the carbody is modeled by finite element (FE) method. The results show that the main vibration frequency peaks of the car-body lateral and vertical acceleration appear around 9.0 Hz, which is close to the natural frequency of carbody diamond mode. These frequency peaks can also be found in the accelerations of the bogie frames and axle boxes. It is also found that the worn wheel profiles increase the wheel-rail contact equivalent conicity, which can reduce the bogie hunting motion stability and induce the flexible resonance of carbody. This means the carbody shaking phenomenon is the vibration resonance of carbody diamond mode that induced by the hunting motion of bogies with worn wheel profiles.

In the proposed work an evaluation of the dynamic characteristics of an articulated passenger train in comparison with the characteristics of a conventional passenger train is made. Mathematical models of the trains of both types are set. The motion of trains along a path of arbitrary shape in plan and profile is considered both in steady-state and in transient motion regimes. Two combinations of wheel and rail profiles are considered. The results obtained in simulations under the conducted studies of an articulated train dynamics at various motion regimes, as well as the comparison of its dynamic characteristics with the similar by parameters conventional train, have shown the possibility of using articulated passenger trains on the railways of Ukraine and other countries with a 1520 mm track gauge.

The article is devoted to the study of the aerodynamics of the undercar space. The aim of the study is to solve the scientific problem of calculating the movement of air masses in the undercar during the movement of a high-speed train and to determine the mechanism of entrainment of a particle by an air flow from the surface of a railway track. To achieve the aim, the following methods were used: design of solid models, embodiment of cfd analysis by the finite volume method in the Flow Simulation module of the SolidWorks software package and in the Fluent module of the Ansys software package based on the numerical solution of the Navier – Stokes system of equations, also the use of the Motion module of the SolidWorks software package for dynamic analysis. According to the results of the study, a picture of the movement of air flows in the undercar space was established, a stable circulation of air masses in the overhead area was determined. This is dangerous in that the particles caught up, during the lifting process, move to the level bogies and can cause damage to the car body elements. Using the designed dynamic model, there is observed the weakening of the adhesion forces of the particle during the movement of the rolling stock, and a particle of ballast can be carried away by the air flow generated by the high-speed rolling stock.

A recent flange climb derailment suggests that the current lateral/vertical load (L/V) ratio-based criteria may not be adequate to assess the risk of derailment in all circumstances. To better understand and prevent similar derailments, TTCI defined three friction coefficients to analyze the potential issues of the current criteria, developed a new geometric criterion to assist in derailment risk assessment, and modified the instrumented wheelset (IWS) processor to use the wheel profile to determine the wheel/rail (W/R) vertical contact position and contact angle. Analysis shows that the uncertainty of friction coefficient, and the nonmonotonicity of contact angle during the process of flange climb may lead to a misjudgment of the derailment risk based on L/V ratio-based criteria. Meanwhile, the geometric criterion can accurately assess the risk of flange climb derailment by judging whether flange climb is imminent. The W/R contact position is recommended as an index for derailment risk assessment.

A joint investigation between the Swedish Transport Administration (infrastructure manager) and passenger train operator SJ AB has been made to analyse whether there is a difference in ride comfort between reprofiling rails by grinding or by milling. The investigation involves long-time measurements of carbody accelerations of four coaches of SJ’s high-speed tilting train, measurements of track geometry quality, rail profiles and wheel profiles.The carbody accelerations have been analysed by two ride comfort indexes, Wz, sensitive to frequencies 2–9 Hz, and ISO-rms, sensitive to frequencies 0.3–2.5 Hz. The lateral track irregularity quality has been analysed by standard deviation of alignment in domain D1, i.e. with wavelengths 3–25 m. The rail profiles have been analysed by calculating equivalent conicity in combination with measured, worn wheel profiles.The analysis shows that the ride comfort index Wz is higher for track sections with milled rails compared to track sections with ground rails, i.e. having worse ride comfort. There are, however, differences between the four measured coaches, with one coach having overall better (lower) Wz ride indexes, and no difference between milled and ground sections. The ISO-rms index, which is sensitive to lower frequencies, do not show any difference between the milled and ground rails. The milled rails show in general higher equivalent conicities than the ground rails, which is consistent with giving higher bogie vibrations for susceptible vehicles.

Resilient track components are widely used in the modern railway tracks, such as the rail pads, the sleeper pads, the continuous support of the embedded rail system (ERS), etc. Track components made of resilient materials usually possess non-linear dynamic properties. In the present work, the equivalent stiffness and damping of an ERS are characterised in function of preload and frequency through laboratory tests and show an obvious non-linear behaviour. A non-linear rheological model of unit length of the ERS is identified according to the experimental data. The good agreement between the identified and experimental equivalent stiffness and damping proves the rheological model an appropriate way for the modelling of the non-linear properties of the ERS. The rheological model is subsequently integrated into a 2D track model of wheelset-track dynamic interaction simulation. The simulation related to a wheelset passage is performed and the results are compared to the ones obtained with a linear track model. The system responses obtained with the linear and non-linear track models are considerably different from each other. The results imply that the non-linear dynamic properties of resilient track components should be carefully analysed and accounted for in the train-track dynamic interaction when an accurate prediction of the system response is desired.

A new moving train–track interaction model is shown in the paper to accurately and efficiently determine long-term high-speed train–track interaction dynamic response. In this model, the rail is represented with a reduced beam model (RBM) and the slabs mainly affected by the train–track interaction are modeled with four-node Kirchhoff-Love plate elements. The fastenings are modeled with spring-damper elements and the Concrete Asphalt (CA) layer is modeled with a Winkler foundation. Since only a small part of the rail and a few slabs from the whole slab track are included, the present model has fewer degrees of freedom than the traditional model using the modal superposition method and the simulation time is significantly decreased. The present model is validated by a long-term train–track interaction case where the results are compared with those from the traditional method. The simulation results show that the present model is accurate and efficient and has great advantages in solving high-speed train–track interaction problems.

An iterative simulation procedure for the prediction of differential settlement of ballast/subgrade in a transition zone between two track forms is presented. The procedure is based on a time-domain model of vertical dynamic vehicle–track interaction to calculate the contact loads between sleepers and ballast in the short-term, which are then used in an empirical model to determine the settlement of ballast/subgrade below each sleeper in the long-term. The procedure is applied for a transition zone between a 3MB slab track and a ballasted track. Each sleeper in the ballasted track section is modelled by a discrete (rigid) mass. Non-linear sleeper support conditions and the possible development of hanging sleepers over time due to settlement are considered. For heavy-haul traffic, the influence of axle load and train speed on the long-term settlement at the transition is studied in a demonstration example.

As large span cable-stayed bridges with ballastless track are pioneerly applied in Chinese high-speed railway, the dynamic performance of the train–track–cable stayed bridge system (TTCSBS) considering influence of environmental factors needs to be paid great attention. This work presents an investigation on the dynamic performance of the TTCSBS subjected to temperature-induced deformation. Firstly, based on multibody dynamics and finite element (FE) theory, an integrated coupled dynamic model of the TTCSBS is established. Then, the temperature-induced deformation of the cable-stayed bridge is obtained by solving the finite element model of the bridge considering the statistical characteristics of monitored temperatures of girder and pylons, and it is applied as the inputs to the TTCSBS for dynamic simulation analysis. On this basis, the coupled vibration features of the TTCSBS considering the track random irregularity and the deformations induced by temperature gradient are studied in detail. The results indicate that the temperature-induced deformation of the cable-stayed bridge would obviously decrease the riding comfort, while it influences the running safety and the dynamic responses of bridge slightly. With increase of the train running speed, the effect of the temperature-induced deformation on the ride comfort of train becomes more remarkable, meanwhile, the running safety indices and the dynamic responses of bridge increase gradually.

The applicable frequency range of an existing code for simulation of vertical dynamic vehicle–track interaction is extended by implementation of a more advanced track model that includes the rail modelled by solid three-dimensional finite elements. In this paper, a previous validation of the software with respect to magnitudes of dynamic wheel–rail contact forces is supplemented by a comparison of measured and simulated rail accelerations at train pass-by. To this end, data collected as part of an extensive field measurement campaign related to the adaptation of the software CNOSSOS-EU to conditions on the Swedish railway infrastructure is used. In particular, the dataset includes measurements of track receptance, rail roughness, wheel out-of-roundness and pass-by rail acceleration. The results calculated when using a simplified two-dof mass-spring-damper wheelset model and a track model accounting for the rail by Rayleigh–Timoshenko beam elements show good agreement with those obtained for more advanced wheelset and track models deploying three-dimensional solid finite elements.

As a major component of the locomotive, wheels have great importance to realize their functions such as load carrying and power transmission through interactions with the track. However, the polygonised wheels will aggravate the dynamic performance of the locomotive and accelerate the component failures. Monitoring of the wheel polygonization is very important for the locomotive maintenance. In this paper, the effects of the wheel polygonization on the dynamic responses of the traction motor electrical system are analyzed based on the established locomotive dynamics system by considering the electromechanical coupling effect between the mechanical transmission subsystem and the electrical subsystem. The results indicate that the impact effect of the wheel polygonization has a great effect on the vibration of the wheelset and has almost no contribution to the vibration acceleration of the bogie frame, and the characteristic frequency of the lower order wheel polygonization can be found in the frequency spectrum of both the vibration responses of the mechanical system and the electrical signal of the traction motor. And this phenomenon can be used for the monitoring of the wheel polygonization.

Full-scale 3D train-track dynamics simulations for long heavy haul trains have not been reported. Successful implementations of such simulations can unlock a series of research topics such as vehicle derailments in train operational environment and track dynamics behaviour under train forces. This paper introduces a method that can be used to develop 3D train-track dynamics models for long heavy haul trains. The method uses parallel computing for 3D train dynamics simulations, in which one computer core is used to compute each vehicle of the train. Rails are modelled using the Finite Element Method and then decomposed into shorter sections by using the Domain Decomposition Method. Parallel computing is then used to simulate individual track sections by using one computer core per track section. Individual computer cores exchange information about coupler forces, vehicle statuses, wheel-rail contact forces and track domain boundary conditions. Hundreds of computer cores are required, therefore, a High Performance Computer or cluster is required to perform such modelling and simulation.

Railway locomotive traction performance is strongly dependent on the friction condition between wheel and rail at the contact interface. One of the major factors affecting the friction condition at that interface is the third body layer. This layer as experienced by each wheelset of a locomotive is partially if not fully eliminated from the contact zone when the subsequent wheelsets run on the rail. Such a self-cleaning phenomenon is generally ignored in classical locomotive dynamics studies, assuming it should not introduce a significant effect. Thus, a constant dynamic friction coefficient is used for all sequential wheels. However, recent investigations show that consideration of the phenomenon is important to accurately reflect the actual physical process at the wheel-rail interface and the exclusion of such phenomenon may result in a significant cumulative error in locomotive dynamics studies.

The modeling of a heavy haul locomotive is a quite challenging task because it requires introducing multidisciplinary knowledge in the final model design. Common methods of modeling heavy haul locomotives are well described in recent publications. However, design and modeling approaches for both bogie traction control and independent wheelset traction control design architectures should not be applied for a situation when the locomotive is not limited by adhesion. In both those approaches the traction power is equally distributed between two traction inverters in the case of bogie traction control design architecture or between six traction inverters in the case of independent wheelset traction control design architecture. This means that an unreasonable limitation of traction power exists in those modeling approaches. To resolve this issue, the right way to model the traction power system is to introduce a supplementary control mechanism that adjusts the torque reference at each inverter of the locomotive based on a physical state estimation of the traction power condition of the locomotive during each time step of a simulation process. This paper discusses the right ways of modeling this scenario and how the proposed improvement affects the traction performance in locomotive adhesion studies.

Within the Shift2Rail project RUN2Rail, an innovative Metro vehicle with single axle running gear is proposed. In NEXTGEAR, also a Shift2Rail project, the work is continued to achieve a higher technical readiness level. For a 2-axle vehicle with a wheelset distance of 8 m the contradiction between stability and curving performance is imminent, and an active wheelset guidance is considered necessary for networks with many small curve radii. For networks with larger curve radii a passive solution might be enough. A nonlinear axle guidance stiffness that has the potential to improve the curving performance is studied here. The running gear frame is modelled in Abaqus® and structural frame modes are implemented in SIMPACK® together with the other suspension elements. To select the properties of the nonlinear spring, simulations are performed to check stability at demanding conditions at 160 km/h on tangent track as well in a curve with 1000 m radius. These two cases give the high and low guidance stiffness of the nonlinear spring. The results show that a nonlinear wheelset guidance can reduce the wear compared to a linear guidance and be an alternative to active wheelset steering for networks with low numbers of narrow curves.

Despite the well-known variability of friction conditions in friction damping systems it is often assumed by necessity that ‘typical’ friction conditions exist in all friction components in a larger system. Following some surprising results in laboratory tests of friction type draft gears modelling of stick slip behaviour has been attempted. The modelling of friction and wedge geometry was carried out as in earlier studies [1–4]. A maximum force criterion was used to model the point at which slip occurred. As it was expected that large dynamical system will usually mask effects from one or two connections exhibiting different friction behavior, the effects were simulated on a large heavy hail train system. It was found that if the stick-slip behaviour was possible in the force range, then the system could be significantly affected, with in-train forces being doubled in the case presented. The case, however, overstates the problem, as it would be very unlikely that such stick-slip behaviour would exist in all connections. The implications are discussed.

Air spring’s modelling is a relevant subject in the railway field, both for bogie and pantograph suspensions. Considering thermodynamic model has a great advantage for its accuracy and reliability, efforts and huge progress have been made before. However, there are still some parts where efforts could be made: although the air spring’s force and displacement’s relationship: F/x is mostly important, the thermodynamic model should theoretically demonstrate the relationship between different parameters, allowing to express the ratio between air spring’s pressure and displacement: Pa/x. By starting from some basic thermodynamic equations, this paper aims at modelling the relation Pa/x, with reference to an air spring of a high-speed train’s pantograph, with a pressure regulator mounted in the pneumatic system. Three different models have been proposed stage by stage. They are all validated by comparing simulation results with experiment results for frequency characteristics, the last one gives the best fit with the most cost. three models’ characteristics have all been explicated and differences have also been summarized.

For trains formed from articulated cars, there are the following basic operational features: an increase in the tonnage per meter while maintaining the standard length of the train, an increase in the train-handling capacity of the railway network and a reduction in the required fleet and in the amount of idle time during the repair. The article describes design features of the articulated cars, running performance numerical simulation models, and provides the results of computer simulation, running dynamic tests and track response tests, which confirm efficiency and safe operation of this type of rolling stock.

Although Independently Rotating Wheels (IRW) meet the low-floor height requirements of the Light Rail Transit systems, its self-centering and self-steering moments resulting from the longitudinal creep forces are small to maneuver sharp curves due to the lack of a rigid rotational speed coupling between its wheels. While the Active Steering control of IRWs can potentially achieve perfect steering with satisfactory running stability, in the present study a simpler passive stabilization control method using a Gyroscopic damper is proposed to realise automatic sharp curve steering and high-speed stability to improve the dynamic performance of a railway vehicle running with IRWs. The present study discusses the stabilization effect of the Gyroscopic damper in three basic configurations of IRWs, namely the conventional IRW, the Negative Tread Conicity Independently Rotating Wheels (NTCIRW) and the EEF (Einzelrad-Einzel Fahrwerk) bogie. In this study, the theoretical investigation and numerical simulations with a full-scale railway vehicle is done which shows the effectiveness of the proposed gyroscopic damper.

The distribution of load placements on wagons, i.e., the Centre of Gravity (CoG) of wagons, and the distribution of loads between wagons in the train configuration, i.e., train make-up, can significantly change the train operational dynamics and, in some circumstances, contribute to derailment of wagons and wheel/rail damage (wear and Rolling Contact Fatigue (RCF)). To reduce derailment risks and wheel/rail damage, the lateral, longitudinal and vertical limits for placing loads on wagons and constraints for the placing of empty or lightly loaded wagons in a consist of loaded wagons should be defined. From the on-going research on this area, a derailment risk and track damage assessment methodology has been developed based on train dynamics and vehicle system dynamics simulations for quick prediction of possible worst operational cases for the investigated train configurations that consider different positions of empty/lightly loaded wagons in a train consist and the CoG for full or partially loaded wagons. Based on all preliminary simulation case studies, the post-processed results show that the effect of vertical and lateral wagon loading CoG shifts to the derailment risk and wheel-rail wear indexes is more significant compared to the longitudinal wagon loading CoG shifts.

Due to the important contribution of flexible modal in rigid-flexible coupled multiple body system (MBS) dynamics, model updating of flex-body generation became a significant and necessary procedure. In this paper, a model updating method based on the measured frequency response function (FRF) is proposed to modify the flexible body in finite element analysis (FEA). To overcome the complexity of multiple measured responses and the huge cost of computation, the Kriging surrogate model with multi-objective function which considers the FRF’s magnitude and shape of curves is proposed. Moreover, the FRF measurement of a high-speed railway carbody is conducted and the measured FRF is transferred into proposed model updating procedure. The results show that the proposed method is performed successfully on a high-speed railway vehicle after model updating and the simulated FRFs tend to coincide with the experimental FRFs. Finally, the updated model is transferred to the rigid-flexible MBS model to validate the influence of the model updating with a comparison of the vibration spectrum and ride comfort index. The results show that the updating procedure will make the prediction of vibration and ride comfort index closer to experimental results.

This paper assesses the impact of a number of potential innovations in the running gear of a railway vehicle upon the wheel-rail interface. Innovations including light-weighting of vehicles and better control of the wheel-rail interface as developed in recent research projects are considered. The longer-term implications are explored and the issues of assessment and sharing of the potential benefits and the direction of future rolling stock design are commented upon.

This paper proposes a methodology which generates statistical insights into operating conditions (e.g. coefficient of friction and superstructure conditions) and their influence on the railway vehicle responses (forces and accelerations). This includes an exploratory data analysis to gain knowledge on the general structure of the data and a variable importance analysis to quantify the dependency between each operating condition and vehicle response. The methodology is applicable to measurements and multi body dynamics (MBD) simulations. Thus, it can be utilized to investigate the influence of uncontrolled operating conditions and to gain additional information on the model validation process.The methodology is applied to a data set based on vehicle response measurements from an on-track test and the corresponding track measurements. This results in the determination of the influence of known operating conditions on the vehicle responses. Finally, the methodology is applied to the MBD simulated repetitions and exemplary compared with the result of the measurements.

The problems of registration of dynamic processes of interaction between rolling stock and railway tracks have been reviewed in this work. The ideas of a newly developed technology for continuous registration of forces between wheel and rail by measuring stresses on four concentric circles located on the inner side of the wheel disk have been presented in it. Also, theoretically, using the finite element method, the locations for the installation of strain gauges on the inner side of the wheel disk and the method of processing the received signals were determined, allowing continuous registration of vertical and lateral forces in contact between wheel and rail by measuring the stresses on the wheel disk and determining the values of these forces with sufficient accuracy when conducting tests on the impact on the track, as well as when monitoring the technical condition of the railway track using a strain gauge wheelset and navigation aids.

Vertical track stiffness and its variation as well as the variation of vertical track geometry along the track are one of the main sources for dynamic wheel-rail contact forces. These dynamic forces contribute to the track settlement. If these local settlements vary along the track, geometric irregularities develop further amplifying the dynamic loading of the track caused by the interaction between the vehicle and track. Here, a simple and fast, physical-based dynamic vehicle-track interaction (VTI) modeling approach is presented for predicting the evolution of vertical track geometries. The settlement model implemented into the VTI model is calibrated using the evolution of peak-to-peak values (track quality parameter) measured in the field over 350 days for a track where only concrete sleeper was used. Using this calibrated settlement model, the physical-based VTI model can predict the evolution of an another track quality parameter, standard deviation, for this track section. The model could also describe the different evolution of track geometry quality for another track section where concrete sleeper with Under Sleeper Pads (USP) are used. Finally, the calibrated VTI model is used to assess the track geometry deterioration when the vehicle properties are changed. The efficient, physical-based VTI model can assist in designing and optimizing tracks and in supporting of maintenance activities.

Simulations of vehicle/track interaction (VTI) in switches and crossings (S&C) require taking into account the complexity of their geometry. The VTI can be handled via a co-simulation process between a finite element (FE) model of the track and a multibody system (MBS) software. The objective of this paper is to reduce the computing effort in the co-simulation process. In the proposed approach, the VTI problem is solved inside the MBS software to reduce the computational effort in the track model as well as the flow of input/output between both modules. The FE code is used to supply the matrices of stiffness, damping and mass at the beginning of the simulation. An explicit time scheme is used with mass scaling. A good agreement is found between both approaches with a reduction of the computing time by a factor of 10. This new approach allows the optimisation of the design of S&C in further studies.

Condition monitoring of track geometry irregularities from onboard measurements is a cost-effective method for daily surveillance of track quality. The monitoring of Alignment Level (AL) and Cross Level (CL) track irregularities is challenging due to the nonlinearities of the contact between wheels and rails. Recently, the authors proposed a signal-based method in combination with a machine learning (ML) fault classifier to monitor AL and CL track irregularities based on bogie frame accelerations. The authors concluded that the Support Vector Machine (SVM) fault classifier outperformed other traditional ML classifiers. Thus, an important question arises: Is the previously reported decision boundary an optimal boundary? The objective of this research investigation is to obtain an optimal decision boundary according to theory of probabilistic classification and compare the same against the SVM decision boundary. In this investigation, the classifiers are trained with results of numerical simulations and validated with measurements acquired by a diagnostic vehicle on straight track sections of a high-speed line (300 km/h). A fault classifier based on Maximum A Posterior Naïve Bayes (MAP-NB) classification is developed. It is shown that the MAP-NB classifier generates an optimal decision boundary and outperforms other classifiers in the validation phase with classification accuracy of 95.9 ± 0.2%and kappa value of 80.4 ± 0.6%. Moreover, the Linear SVM (L SVM) and Gaussian-SVM (G SVM) classifiers give similar performance with slightly lower accuracy and kappa value. The decision boundaries of previously reported SVM based fault classifiers are very close to the optimal MAP-NB decision boundary. Thus, this further strengthens the idea of implementing statistical fault classifiers to monitor the track irregularities based on dynamics in the lateral plane via in-service vehicles. The proposed method contributes towards digitalization of rail networks through condition-based and predictive maintenance.

Self excited wheel-set axle vibrations can cause discomfort for the passenger, lead to polygonization or lead to issues of the press-fit between wheel and axle. To investigate them closer, measurement data of more than 5000 vibration events were analyzed. This was then used to develop an energy based methodology to calculate the contact energy for each recorded vibration. The energies were then statistically analysed and different percentile energies were calculated. It was possible to use these energies to parametrize a creep force model, e.g., the Polach model, because the energy of one oscillation cycle is directly related to the shape of the adhesion characteristic. The resulting creep curves were then used in a Multi-Body-Dynamics simulation to validate the results, which resulted in a good agreement. Additionally, a realistic worst-case scenario was defined, which can now be used to investigate the influence of changing vehicle parameters on the maximum vibration amplitude during the design phase.

The FASTSIM algorithm is widely used in multi-body simulation (MBS) software packages for the evaluation of the tangential wheel-rail contact forces in a steady state. As the original algorithm is restricted to Hertzian contact patches, a strip-based local approach is proposed to extend FASTSIM to non-elliptical contact cases. The paper presents this tangential contact approach in detail, which was briefly introduced by Ayasse & Chollet along with the semi-Hertzian method. The contact stresses and their directions are compared with the reference results from the program CONTACT. Different settings for the traction bound are explored to determine their influence on the contact stresses, creep forces, and the limits of the saturation zone in the case of wheel-rail contact. A design of experiments is constructed for a non-Hertzian contact case, with different combinations of the longitudinal, lateral, and spin creepages. The absolute error in the normalised creep forces is used as the quantity of interest and found to be consistent with results in the literature for Hertzian contact cases using FASTSIM.

A novel idea for predicting the equivalent conicity is presented, based on the inclinations of the wheel thread at the running circle and the rail profile at top-of-rail.Both the shape of the wheel profiles and the shape of the rail profiles can be acceptable in today’s standards, but together the wheel/rail profile combination can lead to an unacceptable high value of the equivalent conicity, which can make the vehicle unstable. By introducing two gradient indices, one for the wheel and one for the rail, it is possible to separate the equivalent conicity into two parts, which also make it possible to put limit values on wheel and rail profiles separately. The indices are combined into a joint index, GIP. The new GIP index is compared to the equivalent conicity for a large number of worn rail and wheel profiles, and show promising results.

Most of the existing railway wheel profile optimization methods involve the purpose of reducing wear, where the T-gamma value is often treated as an indicator to evaluate the degree of wear, considering a small T-gamma value representing less material loss. A small T-gamma value, however, implies that the wheel-rail contact points are likely to occur in the vicinity of the wheel’s nominal rolling circle, and the optimized wheel profile obtained with the target of minimizing the T-gamma value may result in a concentrated distribution of wheel-rail contact points around the nominal rolling circle, accelerating the formation of hollow wear. Using the T-gamma value as the optimization target, therefore, cannot guarantee that the wheel profile still has excellent wear performance during long-term service. Aiming at this issue, an indicator, named wear concentration index (WCI), is developed to assess the degree of wheel material loss and the shape stability of wheel profiles. Numerous simulations demonstrate the feasibility and superiority of the index WCI and indicate that this index can be used as an alternative to the target T-gamma in wheel profile optimization.

Polygonal wear is a common type of damage on the railway wheel tread, which could induce wheel-rail impacts and further components failure. This study presents a finite element (FE) thermomechanical model to investigate the causes of wheel polygonal wear. The FE model is able to cope with three possible causes of polygonal wear: thermal effect, initial defects, and structural dynamics. To analyse the influences of the three causes on wheel-rail contact stress and wear depth, different material properties (i.e., elastic, elasto-plastic, thermo-elasto-plastic with thermal softening), and wheel profiles (i.e., round and polygonal) were used in the FE model. The simulation indicates that a high temperature up to 264.20 ℃ could be induced by full-slip wheel-rail rolling contact when the polygonal profile is used. The thermal effect, similar to that induced by tread brake, may then have a significant influence on wheel-rail contact stress and wear depth. In addition, the involvement of initial defects, i.e., polygonal profile, causes wheel-rail impact contact and remarkably increases the contact stress and wear. By reliably considering all the three possible causes, the proposed FE model is believed promising for further explaining the generation mechanisms of wheel polygonal wear.

The railway wheel wear prediction is essential in the optimization of the maintenance strategies of the wheel-rail system which have both economic and safety implications. The computation of the stresses across the contact patch is necessary to determine the wear distribution, which makes the wear simulation computationally expensive and challenging. Therefore, a fast wear computation method is highly desirable for wheel wear estimation in the context of multibody railway vehicle dynamics simulations where many contact problems must be solved at each timestep. This paper proposes a fast computation approach to estimating wear distribution over the wheel-rail contact patch. The effectiveness and efficiency of the proposed method are demonstrated with the selected case studies.

The reliable prediction of wheel wear can help to reduce maintenance costs. With the help of two common approaches (statistical, contact mechanics based), it is possible to predict wheel profile shapes either quickly and precisely, but for a unique operating situation only, or for varying operating scenarios in a more time-consuming, but often less accurate way because so many, sometimes even unknown, input data are needed. There is no method available for predicting worn wheel profile shapes quickly, accurately, and generally. The hybrid approach presented in this work combines the two state of the art approaches mentioned above in order to exploit their advantages and eliminate their disadvantages. The new method was calibrated and validated on wheel measurement data taken from the field. A good agreement between measurements and predictions was observed when using maximum wheel-rail contact shear stresses as the wear measure in the methodology.

In this paper, a modified wear calculation method is developed, which can give less precise but faster results compared to the classic wear calculation method. Besides, a precise contact point detection program is developed to cooperate with this modified method.

This paper presents a basic study on a potential formation mechanism of wheel polygonization at trams. It was found that wheelset vibrations may occur at negotiating curves with small radius of curvature, caused from a self-excitation mechanism. To investigate the influence of system parameters relevant for the appearance of these vibrations, a representative system model is set-up in SIMPACK. Considering the falling friction effect in the creep force–creepage characteristics at high values of lateral creepage, results reveal that the flexibility of both the wheelset and the resilient wheels is essential. Recurring periods of respective vibrations at constant velocity could be able to promote wheel polygonization.

The paper presents a data-driven modeling approach for a variable-geometry suspension (VGS) system. In the optimization process, a learning-based algorithm is used to select the relevant variables from the measured dataset. The dynamics of the VGS as a polytopic Linear Parameter Varying (LPV) system are formed. The optimized model is written into a polytopic form, which is the basis of the control design. Using the resulting model a controller for achieving steering functionality is designed. The effectiveness of the control method on a VGS test-bed using Hardware-in-the-Loop (HiL) simulation is demonstrated.

The Twistbeam axle suspension is a cheap and robust layout for rear axles at front wheel driven midsize cars. Appropriate models have to take the elastic deformation of the torsion beam into account. A Finite Element approach requires detailed informations of the material properties and the shape which are usually only available in the final production stage. This paper presents a lumped mass model which can easily be integrated into a multibody vehicle model and can be used in the early stage of development. An approximation by the design kinematics further reduces the complexity of the model and considers only the kinematic properties of the Twistbeam suspension. Simulations using a nonlinear and three-dimensional vehicle model with different maneuvers, such as steady-state cornering, step steer input, and driving straight ahead on random road, demonstrate the performance and, in particular, the difference of the presented Twistbeam suspension models.

For expanding virtual methods in vehicle dynamics development, precise and real-time capable suspension models are required. Especially the elastokinematic properties have to be represented very thoroughly. This contribution presents a model of an elastokinematic double wishbone front suspension with topology and parameters typical for passenger vehicles. It has a total of 144 degrees of freedom and a high numerical stiffness. To enable computationally efficient integration, a linear-implicit integration method is used. To achieve good scalability of computational effort and accuracy, the determination of the necessary Jacobians is separated from the actual integration, so that both can be performed with different step sizes. Dynamic suspension tests are used to show the influence of the step sizes on the accuracy of the method.

Most simulation tools focus on solving dynamic vehicle models in the time-domain. The time-signals of the solution can be used to analyze steady-state conditions and frequency-response characteristics. However, performing time-domain simulations for these purposes can be computationally expensive and unsuitable for a time-sensitive environment like motor-racing. A vehicle model which can achieve these functionalities efficiently, next to time-domain simulations, is necessary. This paper discusses a framework within which a dynamic vehicle model can efficiently perform steady-state simulations and linearization, next to time-domain simulations. The framework consists of a format to organize the equations of motion and mathematical methods required to achieve the functionalities desired. A four-wheeled vehicle model is created as a test case to apply the proposed approach.

Designing new suspension systems is a complex task which affects directly the tire behavior, impacting the vehicle balance and final performance. As computational resources are getting cheaper and more powerful, simulation tools are being used together with optimization techniques that allow the exploitation of such resources. This work presents a framework for simulation and optimizations of the kinematics of a suspension system. An evolutionary multi-objective optimization algorithm is implemented along with a suspension kinematics simulation software that evaluates the solutions and retrieves the final best compromise between the given objectives inside the design space given to the algorithm. The dependence of the kinematic objectives is modeled through a set of weight functions and scaling factors. Considering the multi-objective nature of this problem and the complex relation between the input and output variables along with the conflicting objective responses, a genetic algorithm was chosen as the optimization approach. The developed framework can potentially reduce the design time by returning a single final sub-optimal solution based on the set of objectives and their respective weights and scaling factors.

This study focuses on an autonomous shuttle (level 4 driving automation) with four-wheel-drive in-wheel motors. We propose an active pitch control, called APC, by a driving/braking force distribution of the front and rear wheels to prevent passengers from falling over inside the shuttle. This study applies a half-car model that considers the driving/braking force and a passenger model using a rigid body. A passenger system's zero-moment point (ZMP) is defined as an index to evaluate the possibility of passengers falling over. The present manuscript applies a 2DOF control system that consists of an optimum regulator as a state feedback controller and a feedforward controller derived from the inverse characteristic of the closed-loop system where the feedback component is substituted into the state equation. Moreover, we designed a motor output limiter and a limiter to avoid tire force saturation. Numerical simulations were performed to verify that the proposed APC can reduce the possibility of the passenger's falling over when the shuttle accelerates/decelerates.

Heat build-up is a significant concern for vehicles with a hydro-pneumatic suspension system. Heat build-up arises due to the dissipation of heat from the damper and results in an increase in gas temperature in the hydro-pneumatic spring. Increased gas temperature results in a change in static ride height that may negatively affect the performance of the vehicle. The force-displacement characteristic of hydro-pneumatic springs is best characterized when a real gas approach is incorporated into the energy equation (EE) of the First Law of Thermodynamics. The thermal time constant (TTC) is included in the EE; it influences the rate of heat transfer between the gas and the surroundings. It is postulated that the TTC can be tuned to mitigate heat build-up in the hydro-pneumatic strut, either by introducing damping to the system in the form of a hysteresis loop or by increasing heat transfer to the surroundings. The effect of different TTCs on the gas temperature and hysteresis loop area was investigated. The effects of excitation frequencies and amplitudes were also investigated. The hysteresis loop can indeed introduce significant additional damping to the system, but this is only realized at large excitation amplitudes at low frequencies. The effect of the TTC is almost negligible at high frequency excitation with small amplitudes. In conclusion, the TTC cannot be tuned to mitigate the heat build-up typically associated with hydro-pneumatic suspensions. However, certain special case scenarios where it may indeed be a useful tool to mitigate heat build-up are discussed.

Despite the ever-increasing studies on active suspensions, most of which illustrate significant improvements over the conventional suspension systems, there have been only a few recent real-life applications in modern production vehicles. The drawbacks of high cost, added weight, added power requirement, and difficulty of maintenance which require specialized tools and technicians, have constrained these applications to the luxury vehicles.This work focuses on comparing the performance of three different active suspension architectures with each other: Linear motors at the front suspensions, linear motors at the rear suspensions, and linear motors at both front and rear suspensions in an attempt to see if a compromise solution can be found. The aim is to investigate if, in view of the benefits of decreased weight, power requirement, and increased cost efficiency, the somewhat reduced but still achievable improvements in the cases of front or rear-only active suspensions can still be a viable solution.In the comparison of the three alternatives, state feedback control considering signal delays and control allocation techniques have been used. The control objectives are set as ride comfort improvement on straight-line driving, braking distance improvement during ABS braking and roll angle mitigation during high-speed steering. Simulation results are assessed and quantified with respect to decisive vehicle dynamics variables, such as sprung mass acceleration, braking distance, reference yaw rate tracking, sideslip angle and roll angle, and conclusions are drawn on benefits of using each three architecture considered. The all active suspension comes out to be the best in ride comfort improvement, braking distance improvement and roll angle mitigation as expected, but the improvements provided by front and rear active suspension architectures in braking distance improvement and roll angle mitigation respectively, are still satisfactory.

The present study investigates the dynamic characteristics of a heavy ground vehicle that is subjected to crosswind on winter roads by two-way coupled simulations of aerodynamics and vehicle dynamics. Three different road friction coefficients, $$\mu $$ μ = 1.0, 0.3 and 0.15, were investigated, representing dry, snowy, and icy road conditions, respectively. Results show that the lateral displacements of the vehicle on the icy road are 13%, 20% and 26% more than the corresponding lateral displacements of the vehicle on the dry road at maximum crosswind velocities of w $$_{cw,max}$$ c w , m a x = 3.5 m/s, 4.5 m/s and 5.5 m/s, respectively. Additionally, at w $$_{cw,max} \ge $$ c w , m a x ≥ 4.5 m/s and a crosswind length of three times the vehicle length, corresponding to the reduced frequency of k = 2.1, the crosswind might cause the driver to lose control of the vehicle, causing it to cross the road’s lateral margins in all road conditions which could result in an accident.

A commonly seen vehicle combination on the European road network is the tractor-semitrailer with a maximum legal length of 16.5 m. In Nordic countries longer tractor-semitrailers are allowed, which makes tractors with larger wheelbase an appealing option. Concerns have been raised that shorter tractors have poorer performance and are involved in more accidents than the Nordic equivalents, especially in winter conditions. To answer the raised questions and concerns, this paper investigates the effects of a tractor wheelbase length on performance of a vehicle combination, as well as other influencing factors such as lubrication of the fifth when, loading condition and road friction level. Six tractor-semitrailer combinations with different wheelbases and axle configurations are modeled and compared. Nordic (truck-dolly-semitrailer) and B-double (tractor-link trailer-semitrailer) combinations, which are common in Nordic countries, are also used as reference vehicles in the comparison. Three maneuvers are simulated to provoke dangerous situations, namely braking with all wheels in a curve, engine braking in a curve and a fast single lane change. The results show that wheelbase alone does not have a major role in the outcome of the simulations, but a combined influence of axle configuration, wheelbase, fifth wheel position and axle loads determines the vehicle performance. Factors such as existence of a working anti-lock braking system or lubrication of the fifth wheel have a large effect on the vehicle performance, under the tested conditions.

An active steering control strategy was formulated to improve the offtracking and high-speed stability of an autonomously driven long combination vehicle, a Nordic combination, which is formed of three units - a truck, a dolly, and a semitrailer. The controller, designed using a new Artificial Flow Guidance (AFG) method, was seen to significantly improve the lateral performance of the actively steered vehicle at a broad range of speeds. In comparison with the conventional vehicle where only the front axle of the truck is steered, and for the scenarios considered, the controller eliminated steady-state offtracking when all axles were allowed to steer. Stability was also improved, with rearward amplification (RWA) of the semitrailer reduced from 1.47 to 1.00. The effect of steering a reduced number of the axles was also considered and it was seen that for low-speed maneuvers, compared to the conventional vehicle, offtracking could be reduced by 72% when the rear axles of the semitrailer are actively steered. At higher speeds, actively steering the dolly axles alone is sufficient to supress the RWA from 1.47 to 1.02.

Current scientific literature classifies the well-known clutch judder phenomenon in friction judder and pressure-induced judder, based on the mechanisms that generate it. Friction judder is associated to a negative gradient of the friction coefficient and to stick-slip phenomena, while pressure-induced judder is mainly caused by geometric disturbances. The peculiarities of the torsional oscillations experimentally observed in dual-clutch transmissions of tractors, incompatible with the explanations found in the current literature, led to consider parametric excitation as a possible cause of instability.Consequently, the clutch was modelled by means of a system of three coupled linear differential equations with time-periodic coefficients. Stability analysis, performed by adopting Floquet theory, allowed to draw stability maps as functions of the two excitation parameters, i.e. the frequency and the amplitude of the dither signal in the clutch actuation pressure. Stability maps proved to be particularly promising to explain the experimentally observed phenomenon. In addition, it was possible to show that clutch judder can also arise for positive gradients of the friction coefficient.

A model is presented for studying the energy used by the braking system in heavy vehicles. The aim is to keep the model as simple as possible, while retaining reasonable fidelity for the problem at hand. All actuators are thus created as ideal, i.e. no losses, then simple dissipative loads are added where applicable in order to capture their effects from a vehicle systems perspective. The vehicle itself is a simple point mass with a replaceable brake system actuator and a proportional speed controller. The model results capture the energy converted by the vehicle actuators with an error of less than 5% compared to the measured data, indicating that the model is capable of estimating the energy used by the two braking system variants, as well as the drivetrain, in other heavy vehicles and driving scenarios.

This paper investigates the hierarchical vehicle dynamics control performance for path-following of a longer combination vehicle. The vehicle, an A-double, has four units – tractor, two semi-trailers and a dolly between the trailers. In the upper-level, LQR control is used to command the total forces and moments (virtual control inputs - V) for path-following. The lower level uses a weighted pseudo-inverse control allocation method to distribute the desired virtual controls to the multiple steering axles. It is found that there is a degree of redundancy in the definition of the V, and the paper focuses on the effects of choosing different sets of V on vehicle responses. It is found that the magnitude of constraint forces at the articulation joints and the robustness of the response are both heavily influenced by the choice of V. The evaluation is carried out by simulation using both linear and high-fidelity truck models.

This article presents an autonomous driving control strategy for multi-trailer articulated heavy vehicles (MTAHVs). To design this strategy, a 5 degrees of freedom (DOF) yaw-plane model is generated to represent a MTAHV with the configuration of A-train double, and a nonlinear model predictive control (NLMPC) based controller is designed for direction and speed control of the MTAHV. Due to multi-unit structures, large sizes, and high centers of gravity (CGs), MTAHVs exhibit poor low-speed maneuverability and low high-speed lateral stability. Especially, MTAHVs often experience amplified lateral motions of trailing vehicle units in transient curved path negotiations. To enhance the safety performance of the autonomous driving control strategy, the self-driving A-train double design is featured with active trailer and dolly steering. To evaluate the effectiveness of the innovative autonomous driving control strategy, co-simulations are conducted by combining the integrated controller designed in MATLAB/ SIMULINK and a nonlinear A-train double model generated in TruckSim.

Autonomous rail Rapid Transit (ART) is a novel trackless-tram urban transport system with high flexibility, capacity, and low implementation cost. All-axle steering capability brings additional control freedom to tackle the manoeuvrability, lateral stability, and off-tracking issues for such a long combination vehicle. This paper presents a lateral control approach with three levels: 1) Ackermann based feedforward; 2) articulation feedback; and 3) yaw rate feedback control. The relative contributions of these controllers are analysed with simulation, and the overall tracking and stability performance of the proposed controller is found to be satisfactory, even without parameters re-tuning for different operating conditions.

A newly designed gantry virtual track train (G-VTT) with all-wheel steering control was proposed, which can realize the trajectory of each axle following the front axle of the head car. The analysis method of the matching between vehicle architecture and path-tracking control algorithm and the concept of path-tracking independence were proposed based on decoupling control theory to verify the rationality of G-VTT architecture. The advantage of path-tracking independence is that there is no motion or force interference between each axle, which means the control of each axle can be performed independently and the existing single degree of freedom driver model can be utilized. A path-tracking control strategy based on angular transmission ratio (ATR) was proposed and compared with another two methods–the pure pursuit and Stanley algorithm. The result shows that: the G-VTT architecture has the characteristics of path-tracking independence and the decoupling control of lateral motion of each axle can be realized; compared with the pure pursuit and Stanley algorithm, the ATR algorithm with steering delay (SDATR) and the optimized SDATR algorithm can achieve considerable accuracy with less parameter requirements.

In this paper, a new architecture vehicle is introduced, which is called active-steering articulated virtual rail train (ASVRT). To realize the running control of ASVRTs under virtual rail constraints, a new active-steering system including a target trajectory generator and two controllers is designed. Firstly, the ASVRT kinematic model with n-units is established. Then a target trajectory generator is addressed based on coordinate calculation, transformation, and curve fitting. Furthermore, a rear-axle steering controller is designed to realize all hinge points track on the target trajectory, and then a front-axle coordinated steering controller is designed for units 2-N to achieve coordinated movement of the train and avoid large hinge stress. Finally, a simulation model with five units is built in software ADAMS, and then the simulations are carried out under the typical rails with different speeds. The simulation results show the proposed active-steering system has good tracking control performance.

The possibility of designing the vehicle cornering response has gained plenty of interest lately, due to its importance in terms of vehicle safety and driveability, and to the progressive vehicle electrification. A number of methods has been so far proposed to define the desired cornering response, normally based on a reference yaw rate that is function of vehicle speed and steering input. Torque vectoring techniques (based e.g. on brakes or electric motors) are then used to enforce the desired vehicle behaviour. This paper proposes a methodology to define the vehicle cornering behaviour by means of the Maps of Achievable Performance (MAPs). In particular a curvature-steering wheel angle MAP is selected, allowing the designer to directly intervene on certain aspects of vehicle driveability which are only indirectly affected by existing methods. A formal design procedure is provided, along with a numerical example. Simulations allow to compare the proposed approach with existing ones, also in terms of practical feasibility (related to the amount of required yaw moment). Although this is a preliminary study, results are encouraging, thus endorsing further investigation.

This methods paper is an attempt to place a few of the classical vehicle dynamics analysis methods in a common framework and how the methods can be used in combination to analyze different aspects of the planar motion of road vehicles. The methods described are the handling diagram for analysis of the non-linear steady-state characteristics, the phase plane analysis method to analyse the dynamic characteristics of a particular fixed point (or several) and the Milliken moment method diagram. In summary these methods are used to analyse the planar motion of the vehicle. The purpose of this methods paper is to present each method in detail and how they can be computed using a common vehicle dynamics model. Furthermore, the relationship between different points in the different representations are presented.

Steering feedback is one essential aspect to provide real world information, and can influence driving performance during remote driving. In this work, the classical feedback models based on physical characteristics (Physical Model) and modular characteristics (Modular Model) of the steering system are constructed separately, and the influences of it on the remote drivers are studied. Objective and subjective measurement methods are separately used for evaluating the performance of the feedback models. In the subjective assessment, a multi-level assessment method is used for studying the influence of steering models on driver’s intuitive feeling. In the objective assessment, lane following accuracy, steering reversal rates, vehicle speed, time consumption, and throttle engagement are studied for different feedback models and scenarios. Moreover, the human biological information of electroencephalogram and heart rate variability are measured for studying the workload differences. The results showed that the physical model gave drivers a better steering characteristic feel and confidence in remote driving while the modular model could provide better real world feel. Returnability was an important parameter in remote driving, and the level of feedback force and returnability speed could be lower in remote driving compared to real car driving. It was also found that drivers had a higher workload in remote driving compared to real car driving.

Steering feel, or steering torque feedback, plays an important role in the steering control task. This paper is concerned with developing a mathematical model of the driver-steering-vehicle system with steering feel and identifying the unknown parameters of the model using a fixed-base driving simulator, aiming to replicate real-world steering scenarios. The hypothesis is that a human driver obtains an internal mental model of the steering and vehicle dynamics, which is central to the functions of perception, cognition and action. The model complements earlier work that focused on modelling the role of vestibular (motion) feedback. It is shown that the model is capable of reproducing human drivers’ on-centre steering task of following a target path while compensating for disturbances on the vehicle. The identified driver model parameters are within reasonable ranges. Further work is proposed.

An occupant body motion capture system is developed for the purpose of investigating discomfort during transient vehicle motion. The system is based on probabilistic sensor fusion of visual and inertial measurements using an extended Kalman state estimator and is capable of estimating the translational and rotational states of the occupant body position, velocity, and acceleration. Experimental results validate the feasibility of estimating the occupant head and torso motion states in an accelerating vehicle. The proposed system enables further research into the objective and subjective response of vehicle occupants.

The introduction of autonomous vehicles is expected to change the transportation system radically. One of the essential factors that affect the acceptance and choice of autonomous driving is passenger comfort. All people in the autonomous vehicle will be passengers and be able to perform non-driving tasks like reading etc. which increases the likelihood of motion sickness. This makes accurate estimation of motion sickness a necessity in the design stages of autonomous vehicles. The aim of this work is to review and apply two motion sickness prediction models (ISO-2631 and the 6D-SVC model) and evaluate their ability to capture individual motion sickness feelings using measured data and subjective assessment ratings from field tests. The comparison with the experimental results shows that the applied estimation models can be tuned to capture the individual motion sickness feelings. The results also show that habituation of motion sickness is an important property that needs to be taken into consideration and modelled.

This paper investigates the tuning of motion cues in dynamic driving simulators for EPAS (Electronic power assisted steering) tuning in winter conditions. The study investigates the differences in frequency content of the vehicle states yaw rate and lateral acceleration between dry and winter EPAS tuning. Based on the results from this investigation, which shows an increased spectral density of low frequency content in both yaw rate and lateral acceleration, coordinated tilt is added to the motion cueing, to give the driver low frequency lateral acceleration feedback. The tilt coordination filter is tuned using offline optimisation based on logged data. The resulting MCA is evaluated objectively using a linear model of a driving simulator and subjectively through driving around a winter test track with six test drivers. The test is conducted using a pairwise comparison of two different settings, one setting without and another with added tilt coordination. Objective metrics shows reduction in lateral false cues, increased correlation between actual vehicle acceleration and simulator acceleration and an increased spectral density below 0.30 Hz. The pairwise comparison and the commentary feedback shows potential in adding tilt coordination with more drivers favouring tilt coordination, however statistical significance cannot be reached due to the low number of drivers.

The evolution of mobility is led by automated vehicles (AVs), as they are expected to decrease commute time and vehicle fuel consumption as well as significantly increase safety. One of the main limitations they face is motion sickness (MS), which could jeopardise AVs acceptance by the society. On one hand, AVs driving style is expected to be perceived more aggressive by AV users, which will cause more head and body motion. Hence, the control of the velocity and its minimisation are an efficient countermeasure of motion sickness mitigation in AVs. On the other hand, the excessive reduction of the velocity can significantly affect user’s dissatisfaction due to longer journey time. Therefore, additional approaches of mitigating MS have to be considered without affecting journey time. In this direction, this paper proposes an active integrated seat suspension for both longitudinal and vertical isolation to minimise MS. The model is compared with a conventional passive seat design for vertical isolation only, and a passive integrated seat design. All the seat models are excited by vehicle responses obtained from IPG/CarMaker when a vehicle is driven over a real countryside road with Class B road roughness. The results illustrate more than 50% improvement in comfort and 20% more MS mitigation compared to the conventional passive seat.

The paper presents performance analysis of passive, active and semi-active seat and chassis suspension controls based on covariance, frequency- and time-domain approaches and a half-car model. In the case of active suspensions, a linear quadratic regulator (LQR) is considered with an option of road preview control, while the semi-active suspension controller design is based on a clipped-optimal LQR approach. A practical, decoupled LQR structure is considered, which contains separate chassis and seat suspension controllers and individual cost functions accounting for conflicting criteria related to ride comfort, vehicle handling and suspension stroke limits. The decoupled LQR design is compared with an ultimate, coupled design, which simultaneously accounts for all criteria and relies on full state feedback. The analysis shows that the decoupled controller performance approaches that of the coupled controller, which is of particular interest when considering passive chassis suspension where no chassis state measurement is available.

In addition to the heaving acceleration the vehicle body jerk is also considered as passenger discomfort parameter. In this work a more realistic half-car model equipped with aerodynamic surfaces is investigated for the applications of anti-jerk optimal control strategy. Our purpose is to eliminate the body jerk, therefore the derivative of control input is included in the performance index. To evaluate the effectiveness of the proposed anti-jerk control law (AJCL) with respect to the moment of aerodynamic surfaces (AS), a bump velocity input subjected to the right side of the vehicle is considered as road disturbance. The optimal preview control strategy is proposed which has the ability to generate anticipative actions against the future road disturbances. The simulation results are carried out for two different weighting sets, which show that a maximum jerk reduction can be obtained with a minor decrease in rms tyre and suspension deflection, thus improve the ride comfort as well as vehicle handling capability.

Due to increased environmental issues and for economic reasons, the automotive industry intensifies the research towards energy efficient driving, by also taking into consideration the comfort and road holding aspects. In this direction, efforts of modeling and minimising tyre wear, one of the main non-exhaust traffic related sources, are investigated in the literature. This work investigates the suspension and tyre optimisation for tyre wear minimisation while the vehicle is driving on different road surfaces. Nevertheless, given that the road holding and ride comfort are important design criteria in the design process, they are incorporated in the optimisation as objectives by configuring a multi-objective problem with all of the above objectives. Conclusions are extracted about the behavior of tyre wear regarding ride comfort and road holding, and also about where the optimum design variables have converged trying to compromise the above performance aspects under different road roughness profiles.

This paper describes a method that enables comparative rolling resistance measurements of bicycle tyres on different road surfaces. The test equipment is a one degree of freedom two-wheeled pendulum with a rigidly coupled eccentric weight. Derived from the conversion of potential energy into dissipated heat by means of rolling resistance losses the amount of energy can be measured and correlated to the distance travelled. The method enables to distinguish small differences in rolling resistance. As an example the difference between two different sets of asphalt is shown.

Performance, development and validation of vehicle dynamics applications such as anti-lock braking systems (ABS) or indirect tyre pressure monitoring systems (iTPMS) rely on a sufficiently accurate consideration of tyre properties such as transient dynamics and tyre kinematics. Previous investigations showed that the tyre rotation has a distinct effect on the vertical tyre stiffness as well as on the three characteristic tyre radii, the unloaded, static and effective (dynamic) tyre radius. Based on the fundamentals of the TMeasy 5 handling tyre model, an enhanced semi-physical modelling approach was developed to consider rotational speed dependent tyre stiffness and tyre radii in an effective and numerically efficient manner. In the present study, a detailed experimental validation is conducted to verify, evaluate and validate the previously identified rotational speed induced effects and the developed enhanced modelling approach. Based on the results of an extensive tyre testing series, it is shown that the rotational speed induced tyre behaviour can be taken into account effectively by considering a linear dependence of the vertical tyre stiffness and a non-linear progressive one of the unloaded radius on the rotational speed. The resulting rotational speed induced behaviour of the static and effective tyre radius is approximated with sufficient accuracy by the enhanced model. With the presented semi-physical modelling approach and identified rotational speed induced tyre behaviour, applications such as ABS, iTPMS or driving simulators can be enhanced.

When steering at very low forward velocities or standstill the lateral force and self-aligning moment developed by the tire are influenced by turn slip. Steering angle sweep measurements executed on a drum show the impact of forward velocity on these tire characteristics. Although extensions to the base Magic Formula exist to model these effects, it requires additional tests and increases the complexity of the model. An alternative formulation is explored whereby multiple Magic Formula models are used in parallel across the width of the tire. Using this approach the low speed characteristics can be modeled with fair accuracy, in particular the self-aligning moment is improved, whereas at high speeds the results are identical to the base Magic Formula. A disadvantage is the increase in computing time. The parallel formulation may also be suitable to simplify the modeling of the self-aligning moment resulting from camber.

The road surface influences the friction potential. Thus, including road surface information in a high definition map can improve a friction potential estimation system. On highways, road maintenance measures the wet friction potential regularly using special trucks. The measurements of the German system SKM and the Austrian system RoadSTAR are analysed focusing on the questions, if the results relate to the deceleration capabilities of a test vehicle and if the measurement method can be applied to urban streets.

Determining the surface on which a vehicle is moving is vital information for improving active safety systems. Performing the surface classification or estimating adherence through tire slippage can lead to late action in possible risk situations. Currently, approaches based on image, sound, and vibration analysis are emerging as a viable alternatives, though sometimes complex. This work proposes a methodology based on the use of low-cost accelerometers combined with Deep Learning techniques to achieve continuous surface estimation. The performance of the proposed system is evaluated with real tests, where high percentages of accuracy are obtained in the classification task.

Tire and road wear are a major source of emissions of non-exhaust particulate matter (PM) and make up the largest share of microplastics in the environment. To reduce tire wear through numerical optimization of a vehicle’s suspension system, fast simulations of the representative usage of a vehicle are needed. Therefore, this contribution evaluates if instead of a full simulation of a representative test drive, only specific driving maneuvers resulting from a clustering of the driving data can be used to predict tire wear. As a measure for tire wear, the frictional work between tire and road is calculated. It is shown that enough clusters result in negligible deviations between the total frictional work of the full simulation and the cluster simulations as well as between the distributions of the frictional work over the tire width. The calculation time can be reduced to about 1% of the full simulation.

Tyres are a vital component for handling and load carrying while also contributing to the operating cost and environmental impact. The innovations in tyre design are driven by the need to reduce greenhouse gases and to make a better compromise between conflicting tyre properties. To accurately simulate tyres and to make these compromises a representative rubber model needs to be incorporated with strain amplitude dependency for the storage and loss modulus (the Fletcher-Gent effect). Prony series is a commonly used viscoelastic model in tyre simulations but it does not take into account the Fletcher-Gent effect and e.g. possible nonlinearities due to axle load variations are not feasible to simulate. The Fletcher-Gent effect can be modelled using parallel rheological framework (PRF), which can consist of any combination of parallel material models. Nonlinear viscoelastic models have strain amplitude dependency for the storage modulus but single nonlinear parameters lose their clarity in a PRF. Another approach is to combine a linear viscoelastic model with plasticity as is done in this article. Here, an FE truck tyre is developed and used with a viscoplastic PRF model that utilises Prony series with Mooney-Rivlin hyperelasticity and multiple plastic networks. The benefit of this combination is that the strain amplitude and frequency dependency of the storage and loss modulus are separated, which makes parameter studies simpler. The article shows that an FE truck tyre with a viscoplastic PRF model can be used in different simulations to study e.g. steady-state rolling, footprint, vertical stiffness and longitudinal tyre forces.

The TMeasy is a tyre model suitable for vehicle handling analyses and enables easy parametrisation. Recently, a convenient interface to Modelica was implemented by DLR to support the TMeasy also for vehicle modelling in multi-physical domains. This paper focuses especially on the particular problem of reliable reproduction of the tyre’s bore torque which occurs during parking manoeuvres. It outlines the theory behind it, discusses the Modelica interface implementation, and presents the results of parameter identification which were achieved based on real experiments with DLR’s research platform ROboMObil.

Measurement methods to determine the rolling resistance of tyres during different operation conditions are essential in the work towards more energy efficient vehicles. One of the influential parameters is the tyre temperature distribution, which has a large impact on the rolling resistance. Today, the standardised test procedure to measure rolling resistance is steady-state measurement on drums. However, the steady-state temperature on a drum is not the same as the temperature during ordinary driving conditions. The aim of this work is to develop a measuring method that enables to set a desired measurement temperature, which would create the possibility to study the relationship between tyre temperature and rolling resistance in more detail. The measurement method was developed by the use of a flat track equipment but should be applicable to other rolling resistance measurement equipment such as drums. The resulting method gives a repeatable tyre temperature and rolling resistance and can be used for measurements on tyres heated to a chosen measurement temperature.

This paper presents a comparison of continuous longitudinal wheel slip control formulations based on classical control. The study investigates the influence of relevant parameters such as longitudinal vehicle velocity and vertical tire load on the wheel dynamics. A proportional-integral wheel slip controller is designed and three tuning methods are applied: i) a model-free tuning technique, i.e., frequency-based loop shaping of the open-loop transfer function; ii) a model-based tuning technique, i.e., tuning rules based on closed-loop performance criteria like damping and bandwidth incorporating the change of wheel dynamics with the relevant parameters; and iii) an optimization-based tuning technique, i.e., tuning based on the minimization of a cost function defined by objective performance indicators. The controllers are numerically verified in multiple anti-lock braking system scenarios with varying tire-road friction conditions using an experimentally validated high-fidelity simulation model.

The automatic activation of emergency calls (eCalls) in the case of serious accidents is mandatory for all new cars sold in EU since 2018. It is desirable to investigate similar devices also for motorcycles. A challenge specific to single-track vehicles is the fall/crash detection. Indeed, non-negligible roll angles are involved in normal riding and the vehicle dynamics may include peculiar phenomena such as wheelie and stoppie, that should not trigger an eCall. The aim of this work is to devise and employ a multibody model to simulate the main fall and crash patterns of motorcycles. The model can be used to test the robustness of fall/crash detection algorithms as well as to support the development of the algorithms themselves.

A minimal three degrees of freedom model is developed and used to investigate unstable modes during high roll angle and acceleration maneuvers. During high roll angle maneuvers, it is found that there are unstable modes in the 6 Hz to 9 Hz range. The modes share similar characteristics to that of the unstable driveline or “chatter” mode found in braking maneuvers, except at a lower frequency caused by the higher inertia about the roll axis, which is the facilitating factor to the tire vertical oscillations. Power analysis shows that a combination of tire frictional characteristics and governing relationships can lead to a phase lag between tire forces and tire slip velocities that causes to instability, and eigenvalue sensitivity analysis shows which modeling parameters have the most effect on the stability boundaries.

Vehicle dynamics modelling is a widely used tool to run the design phase without the need of expensive prototypes. Models allow to understand the dynamic properties of the system and run sensitivity analysis faster, cheaper, and easier way with respect to physical testing. Being kick e-scooters a breakthrough introduction in the vehicular field, widely accepted mathematical models are still not available. In this paper a lumped parameters model for the vertical dynamics simulation of an electric kick scooter is presented. The mechanical impedance of the driver is accounted in the model since it deeply influences the dynamical properties of the whole system. For model validation purposes, acceleration results from computer simulations are compared with those acquired in an experimental campaign on distributed road irregularity. Dealing with a random input, power spectral densities of vehicle acceleration at most interesting points are used to perform the comparison, thus validating the model. The model proposed in this paper can thus be an important tool for the vehicle design stages, allowing to estimate vehicle’s road holding properties and driver’s comfort while riding kick e-scooter.

A minimal, three degrees of freedom model for the rear of a racing motorcycle is developed to study the effects of roll angle on the stability of the driveline mode, also known as “chatter”, under a heavy braking maneuver. It is found that the added roll angle degree of freedom does not play a large role in the unstable mode, but its addition to the linearized equations of motion does add to the instability of the system. This is attributed to the working point of the tire tangential forces changing as more lateral force is introduced. The mode becomes unstable during the given maneuver with a frequency of around 19 Hz. Power analysis reveals the source of instability, and that the longitudinal tire force gradients can change the stability by changing the phase relationship of the longitudinal force and slip velocity. Eigenvalue sensitivity analysis shows that splitting the rear-hop and driveline mode frequencies also leads to stability, and so does reducing the chain-to-swingarm angle.

In this paper, the obstacle avoidance performance of PMV is studied from the viewpoint of active safety. Because of the PMV tilts inward when turning, the inward tilt and the height of obstacles, those are not required in general automobiles, are considered, in addition to comparing FWS and RWS. In open loop driving, the lateral displacement is delayed in RWS compared to FWS, therefore FWS is superior in the obstacle avoidance capability. In closed loop driving, the driver complements the driving trajectories, therefore obstacles can be avoided by both of FWS and RWS. However, the load for the driver to avoid obstacles is heavier and the vehicle behavior is less stable on RWS than FWS. Therefore, it can be said that FWS is also superior than RWS in the obstacle avoidance capability in the conditions close to the real world. In the case of road obstacles such as holes on the road surface, PMVs, both of FWS and RWS, can avoid them easier at the tire contact points due to the inward tilts. In case of high obstacles such as vehicles, the inward tilt is generally disadvantageous due for the roof height to approaching the obstacle. However, FWS with active tilting mechanism does not lean inward at closest point to the obstacle, and therefore, it does not become disadvantageous.

Preliminary results of a comprehensive study on the dynamic behaviour of e-scooters are presented, referring to straight-line stability and emergency braking. Both scenarios are of particular interest from a vehicle safety perspective. Simulation studies and on-road tests are performed. Model parameters are obtained from laboratory measurements as well as from on-road measurement data. Stability analyses based on the benchmark bicycle model reveal a self-stable behaviour for a range of usual forward speeds. Extensions of the model w.r.t. rider body lean and a linear tyre model are discussed. Braking strategies for effective emergency braking manoeuvres are shown to essentially involve the body position of the rider. Due to limited braking forces generated by the front regenerative braking system, riders should be encouraged to shift their body position further to the rear and rely on rear brake as well, which is shown to be an effective strategy to reduce stopping distances.

In the past few decades, braking systems with carbon discs have become the dominant technology for many racing applications, such as in the MotoGP class. Indeed, they provide higher friction coefficients and, thanks to their lightweight materials (with respect to conventional steel brakes), the unsprung masses and the gyroscopic effects can be reduced, thus improving the motorcycle dynamic performance. However, carbon brakes can work properly only within a relatively narrow operating temperature range. Hence, an accurate assessment of their actual thermal behavior is mandatory. After a brief introduction to Bayesian inference theory, this paper focuses on the development of the Unscented Kalman Filter (UKF) algorithm as a suitable mathematical tool for assessing the temperature gradient of the carbon disc. The UKF belongs to a category of optimal estimation algorithms (called Bayesian recursive filters) that use both the sensor measurement and a physical model to calculate the optimal posterior estimate for the state of the system. This paper will go through the simplified one-dimensional finite element model representative of the thermodynamics of the brake that is employed by the UKF for the posterior optimal temperature estimation. Finally, an unconventional usage of the UKF is proposed which turned out to be particularly effective for the identification of the parameters in the thermal model that cannot be directly measured.

In this paper, a neural network-based feedback linearization method is presented for automated vehicle application purposes. The main idea behind this algorithm is to match the dynamics of the nonlinear system to a linear model. The linearization is achieved by the data-driven training process of a neural network in the control structure. Through the proposed feedback linearization the nonlinear dynamical properties of the vehicle can be effectively handled in the control design process. It can provide advanced functionalities for the automated vehicle, e.g. maneuvering in critical situations. Moreover, the benefit of the presented method is that the design of the control system without the exact formulation of the nonlinear vehicle dynamics can be performed. The effectiveness of the method through the control design for achieving path following functionality is presented.

A path planning and tracking framework based on model predictive control to avoid obstacles is proposed for autonomous vehicles. Firstly, a vehicle in road coordinate system is established to describe the relationship between the vehicle and the reference path. Secondly, to deal with multi obstacles, an efficient search-based method along the reference path is used to build collision-free driving corridors as state constraints. Then, a multi-constrained model predictive controller based on vehicle kinematic and dynamic model is employed to compute the optimal steering angle. Finally, the simulation results show that the proposed path planning and control framework approach are effective for various driving scenarios.

The objective of this paper is to explore the potential of combining hardware and software design parameters in the optimization for reduced energy consumption in an electric vehicle. The vehicle used has four electric motors connected to one wheel each through a single speed transmission and a controllable coupling. The software design parameter consists of a control allocator based on a quadratic optimization program that distributes torque and wheel angle to minimize power losses in the electric motors, inverters, transmission and tires. The control allocation algorithm is then used in combination with the hardware design parameters, consisting of a controllable coupling and different transmission ratio setups between front and rear motors, in order to evaluate the joint effect on energy consumption. It was found that the energy consumption can be reduced by 8.4 $$\%$$ % compared to equal torque distribution on all wheels.

Trajectory planning is a crucial component of autonomous driving systems. However, using simple vehicle models for trajectory planning may result in unrealistic reference trajectories, especially in critical driving conditions, endangering the safe driving of autonomous vehicles. This study explores the effect of model complexity on the trajectory planning performance of autonomous vehicles in critical driving scenarios. Five trajectory planners of various levels of model complexity, including Planner $${\text {STK}}$$ STK (single-track kinematic model), Planner $${\text {STDL}}$$ STDL (single-track dynamic vehicle model with a linear tyre model), Planner $${\text {STD}}$$ STD (single-track dynamic vehicle model with a simplified Pacejka tyre model), Planner $${\text {DTB}}$$ DTB (double-track vehicle model with the brush tyre model), and Planner $${\text {DTMlt}}$$ DTMlt (double-track vehicle model with load transfer consideration and the Pacejka tyre model), are designed. The trajectory planners are formulated as optimal control problems, where constraints for obstacle avoidance, yaw stability and the physical limits on vehicle actuators are explicitly considered. These planners are assessed in two severe driving manoeuvres, i.e. the double-lane change and single-lane change manoeuvres. Results indicate that Planner DTMlt outperforms DTB with higher passing velocity as well as smaller peak yaw rate and sideslip angle, and that Planners STD, STDL and STK are not suitable for use in these critical driving scenarios.

This paper investigates and applies a Nonlinear Model Predictive Control (NMPC) method for an Evasive Manoeuvre Assist (EMA) function. Evasive manoeuvres typically involve limit handling situations, in which tyres operate within the nonlinear region. To take into account the nonlinearities, this paper focuses on solving the evasive manoeuvre problem from a nonlinear optimisation perspective. To solve the nonlinear optimisation problems that arise with an NMPC approach, two state-of-the-art nonlinear solving methods are compared in simulation tests: the Interior Point Method (IPM) and Sequential Quadratic Programming (SQP). Moreover, an SQP approach is implemented and tested in a rapid-prototyping test vehicle.

The paper introduces an adaptive Torque Vectoring (TV) controller for all-wheel-drive electric vehicles. The main focus of this study lies in tuning procedures of controller gains in accordance with the manoeuvre conditions. For this purpose, a pre-trained neural network predicts the vehicle behaviour and adjusts the PID gains of the TV controller. The proposed method extends the applicability of the TV system and increases its efficiency as compared to the non-adaptive baseline control methods.

Ensuring safety of use is one of the challenges for autonomous driving functions and means that the system should drive at least as safely as a real driver. Actual development tools are time-consuming and expensive (on-road test) or the real car with all distributed systems is not well represented (simulation). The study investigates the quality of a new approach, the Vehicle-in-the-Loop (ViL) test bench, which strives for combining the advantages of real test drives and numerical simulation. A vehicle equipped with a commercial camera and a Lane Keeping (LK) function is tested on a test track and on the ViL test bench to compare the behavior and proving the usage as a new development tool.

The implementation of vehicle safety control systems in modern cars requires accurate vehicle dynamic state estimation. The estimation of the vehicle sideslip angle is essential to improve the performance of current and future Electronic Stability Control (ESC) systems. An innovative approach to estimate the sideslip angle directly from GPS measurements is developed. Given an existent kinematic sideslip estimator, with solved numerical integration issues during straight line driving, a GPS drift-free sideslip estimate improves such kinematic sideslip angle estimate during steady-state cornering conditions. As both methods do not require any tire-road friction nor other vehicle parameters, the two estimators are fused together using a Fuzzy Logic approach. Both require only basic sensors: wheel speeds, Inertial Measurement Unit (IMU) measurements, steering angle, and GPS module. Extensive experimental validation is used to assess the quality of the developed estimators and their robustness to a wide range of driving scenarios.

Accurate and cost-effective state estimation is needed to reach self-driving. The well-known GNSS and IMU fusion can be improved by the integration of wheel odometry. The robustness of this type of odometry is increased if both the rear and front wheels are utilized. Furthermore, the method is cost-effective, but for accurate motion estimation, the vehicle model parameters have to be calibrated. This paper presents the whole calibration task, such as input estimation, filtering of reference outputs, and parameter identification. The proposed estimation method is a unique version of the Gauss-Newton method, to mitigate the distortion effect of pose initialization. The effectiveness of the proposed calibration process is illustrated through vehicle test experiments. The validation demonstrates that the calibration results in below than 1% relative estimation error, thus the front-odometry can be integrated into the state estimation layer of a self-driving vehicle.

When discussing semi- or fully autonomous driving systems, safety issues must be considered seriously. The focus of this work-in-progress paper is on the design and implementation of a safety and an enhanced E-Gas monitoring concept in a dual-microprocessor system, which is utilized for a fail-operational Steer-By-Wire system in autonomous vehicles. Furthermore, based on ISO 26262, safety requirements for a Steer-By-Wire system and its decomposition are figured out. Meanwhile, an example of the limitation of ISO 26262 towards autonomous driving is discussed and its corresponding solution is proposed. In addition, several test results of the above-mentioned dual-microprocessor system are demonstrated.

Regenerative braking reduces the use of friction brakes for electric vehicles. This can offer a potential for downsizing or, in extreme cases, even elimination of the friction brakes particularly on the rear axle of vehicles with reduced dynamic requirements. However, the elimination of wheel individual rear axle brakes makes wheel individual brake interventions of electronic stability control impossible jeopardising state-of-art vehicle safety. This paper investigates how driving safety can be enhanced for such vehicles by specifically designed torque distribution devices that allow for uneven distribution of centrally generated brake torque to the individual wheels of an axle.

Electric vehicles with the wheels individually driven by e-motors have promising potential for improving performance through finer control over the power distribution among the wheels. Due to the absence of a mechanical driveline to connect the wheels to the transmission and engine, the virtual driveline system (VDS) is proposed as a conceptual framework to connect virtually the individual electric motors and, thus, to optimize and analyze the dynamics and performance of vehicles. Conceptually, the VDS is based on vehicle-generalized parameters (VGP), which are used in the VDS principle to establish relationships between VGPs and, thus, to manage the wheel power split and set up interactive/coordinated controls of the e-motors to optimize and improve energy efficiency, terrain mobility performance, maneuver, etc.

The growth of electric mobility showed the possibility of reinventing the vehicle powertrain layout. The adoption of one motor per wheel allows to precisely control the driving and braking torque on each wheel. However, conventional friction brakes are necessary to guarantee top braking performance of the car since, in general, the braking toque by electric motor is not enough to perform maximum deceleration manoeuvres. A suitable blended braking distribution strategy must be designed to distribute the torques on the wheel by accounting for the vehicle state, the driver request, and the torque vectoring request by stability control algorithm. This paper presents and optimal control strategy that distributed the braking torques on the wheel accounting for electric and hydraulic brakes. The control algorithm considers than the regulations requirements, the wheel vertical load condition both in longitudinal and lateral dynamics conditions, and the request by driver and stability control.

The purpose of this paper is to build a transmission system model with Electric Disconnect Differential (EDD) and give the corresponding control strategy. A physical model of the dog clutch is established to discuss the force analysis in the coupling process. The speed controller of the motor is designed by the pole assignment method, and the speed difference is determined by a random model. Finally, the simulation is verified using MATLAB/Simulink. The results prove that the control strategy can quickly and stably control the coupling and disconnection of the entire transmission system.