This work presents a planar curvilinear beam finite element, referred to as the Wire Element, for the geometrically nonlinear analysis of slender one-dimensional structures undergoing large displacements and finite rotations. The proposed formulation differs from conventional displacement-interpolation approaches for it reconstructs the internal kinematics through direct integration of compatibility equations along the beam axis. Within a flexibility framework, the equilibrium relations, cross-sectional constitutive behaviour and kinematic compatibility altogether combine to derive the element response in the current configuration. The element accommodates the evolving geometry through internal section-slice discretization. At every converged configuration of the incremental–iterative procedure, each section-slice updates its orientation forming the one-dimensional curvilinear beam. The flexibility contribution of each slice is evaluated analytically and accumulated to construct the end-to-end element flexibility matrix; this allows to obtain the tangent stiffness matrix directly referred to ending nodes. As a result, the formulation enables the modelling of highly curved beams using a minimal degrees-of-freedom set, whatever curvature complexity is. The proposed element is assessed through comparisons with analytical Elastica solutions and with conventional geometrically nonlinear finite element analysis. The results demonstrate accurate prediction of large displacements and rotations equilibrium paths accompanied with a significant reduction of the required degrees-of-freedom and therefore computational effort. Owing to its compatibility-driven kinematic reconstruction, the formulation is particularly suited for applications where geometric fidelity and computational efficiency must be simultaneously addressed.
This study proposes an enhanced grid-based data-driven transient tire-soil contact model to enable quick and accurate simulation-based assessment of off-road mobility systems. In this model, the compact tire-soil contact kinematic description in the classical terramechanics model is generalized to account for three-dimensional non-uniform transient contact stress distributions on a grid contact patch by leveraging machine learning techniques. Adaptive piecewise functions are introduced to the grid contact model, such that only a small number of control points are learned by neural networks to predict transient contact stress responses within the contact patch, considering the granular soil material flow. By doing so, the learning data can be effectively compressed, and the number of neural network calls in the online mobility simulation can be significantly reduced while maintaining accuracy comparable to the physics-based tire-soil interaction model. The effect of soil eruption, which causes the top surface to rise when a heavy wheel load is applied to cohesive soil, is also incorporated in the proposed model. The accuracy and computational improvement by the proposed approach are examined using transient traction and cornering simulations, involving large soil deformation, as well as bulldozing effects induced by significant sideslips. Furthermore, the proposed model is applied to mobility simulations of an all-terrain vehicle performing obstacle avoidance at high speed on sandy terrain, and the simulation results are validated against field test data in a scenario not considered in the training data. It demonstrates the potential of the proposed approach to support the development of off-road mobility systems through rigorous simulation-based assessment.
This study proposes a physiologically constrained multibody dynamic model based on prismatic-revolute compound joints to characterize the biodynamic responses of the seated human body under whole-body vibration (WBV). The anatomically-inspired model divides the human body into five rigid segments: head-neck complex, upper torso, middle torso, lower torso, and hip-and-thigh. Adjacent segments are interconnected by compound joints that integrate translational and rotational degrees of freedom (DOFs) to replicate the physiological kinematics of the human spine. The governing equations of motion are derived via the virtual power principle; linearization of the nonlinear dynamic equations yields analytical expressions for apparent mass (AM) and seat-to-head transmissibility (STHT). Experimental data from 20 seated subjects exposed to band-limited (0.5–20 Hz) random vertical vibration were adopted for the system model identification. A weighted least-squares objective function incorporating both AM and STHT was constructed, and a two-stage hybrid optimization strategy combining particle swarm optimization global search and gradient descent local refinement was implemented to enable efficient, high-precision identification of joint stiffness and damping parameters. Validation demonstrates strong agreement between model predictions and experimental measurements, with goodness of fit values exceeding 0.968 for seat-to-head transmissibility and 0.926 for apparent mass. Modal analysis identifies five distinct vibration modes in the frequency range from 1.24 to 8.30 Hz, dominated by pitch and vertical motions of the upper-body segments. The identified natural frequencies align well with previously reported experimental modal findings, verifying the model’s capability to capture intrinsic whole-body vibration transmission mechanisms. Compared to existing models, the proposed framework achieves a superior balance among model compactness, computational efficiency, and biomechanical fidelity through physiology-guided DOF reduction. The model provides a reliable numerical tool for seating system design and occupant-seat coupling dynamics analysis, and also supports segment-level vibration absorbed power calculation to facilitate WBV injury risk assessment and segment-specific frequency weighting research.
In recent years, abnormal lateral swinging has frequently occurred in the trailing car of Power Centralized EMUs operating in single-track tunnels. This phenomenon has severely compromised ride comfort and operational safety. The objective of this research is to investigate its underlying mechanism and evaluate effective mitigation strategies. Dynamic tests reveal that the trailing power car exhibits a 1.5 Hz yaw motion, causing the lateral Sperling index to exceed 2.97. Fluid-structure interaction (FSI) dynamics simulations are conducted to reproduce the abnormal vibration and clarify its excitation mechanism. The results indicate that intense unsteady aerodynamic loads, particularly the aerodynamic yaw moment, are the primary excitation source of the low-frequency forced vibration of the car body. Further analysis demonstrates that this forced vibration cannot be effectively mitigated through conventional suspension damper optimization. Three mitigation approaches are evaluated, including bogie suspension damper adjustment, inter-car lateral stiffness optimization, and active lateral actuation. The results show that conventional suspension damper optimization provides only marginal improvement. Therefore, two effective mitigation strategies are identified: optimizing the inter-car lateral stiffness and installing an active lateral Giant Magnetostrictive Actuator (GMA) between the car body and the bogie. Both strategies significantly suppress tunnel-induced yaw oscillations. This research provides a theoretical foundation and practical engineering solutions for addressing aerodynamically induced vehicle oscillations, offering critical guidance for improving train operational performance.
Lower limb exoskeletons are gaining attention in the literature for their ability to enhance end-user muscle torque generation; this is particularly advantageous for assisting in walking actions. However, maintaining upright balance often remains the user’s responsibility. Existing standing balance controllers have primarily focused on regulation of whole-body momentum, instead of using full-state feedback and optimal control. In the current study, a constrained optimal full-state feedback feet-in-place (FIP) balance controller for an active hip-knee-ankle exoskeleton was developed using model predictive control (MPC). An integrated human-exoskeleton multibody dynamic model of upright stance was developed for controller design and evaluation. The model was linearized for controller design. Multibody dynamic simulations assessed balance recovery and controller performance following support-surface translations under two scenarios: 1) without any active torque generation from the simulated human (MPC-only regulates balance); 2) providing a percentage of assistance to a simulated human with variable torque capabilities (human driven via linear quadratic regulator, LQR). In the first set of simulations, exoskeleton performance was assessed comparing MPC against unconstrained LQR. In the second, the exoskeleton was driven by MPC exclusively while human torques were produced via the LQR utilized in the first set. Criteria in the MPC/LQR objectives included those identified from human experiments: regulate center of mass deviations, changes in trunk pitch, and net torque application over the prediction horizon. Multibody dynamic simulation results suggest that applying either of LQR and MPC to exoskeleton control can provide restoring torques post-perturbation during standing. Unlike the LQR however, the MPC consistently maintained the zero moment point within the support polygon and actuator torques within limits, though MPC did in some cases struggle with convergence in the joint kinematics. This was more evident when adding a human-in-the-loop that concurrently tried to regulate balance according the LQR while the exoskeleton used MPC (second set of simulations). We conclude that MPC offers a viable near-optimal control law for constrained standing balance assistance.
