A physics-informed machine learning framework is developed to predict temporal force–displacement responses and three-dimensional deformation histories of hexagonal crash boxes under multi-angle loading. To reduce the computational cost of nonlinear explicit finite element simulations, a three-dimensional convolutional autoencoder (3D-ConvAE) encodes crash box geometries into a low-dimensional latent representation, while a long short-term memory (LSTM) network predicts their temporal deformation evolution. A physics-informed loss function incorporating kinematic admissibility and energy conservation is introduced to suppress non-physical responses and improve mechanical consistency. Validation against benchmark numerical data demonstrates high predictive accuracy, with mean errors of 7.24% for peak force and 6.31% for energy absorption. The framework also maintains geometric fidelity across complex failure transitions, including the shift from progressive folding to global bending, achieving a mean Chamfer distance of 4 mm. The proposed approach provides a robust and computationally efficient tool for predicting nonlinear crash responses and supporting crashworthiness optimization.
Modular cross-laminated timber (CLT) construction offers rapid, low-carbon housing, yet its seismic viability is governed by connection performance. This PRISMA-based systematic review of 50 publications compares the mechanical characterization of twenty connection systems for volumetric CLT across the complete load path, intra-module (InMC), inter-module (IMC) and module-to-foundation (MFC), and sets it against reusability. The evidence separates two families. Adhesive-free, timber-only interlocking joints support circularity but dissipate energy through irreversible timber crushing, so their cyclic ductility is limited and non-repeatable, whereas hybrid steel-timber systems and friction dampers concentrate damage in replaceable components and give stable, repeatable hysteresis; together they define a damage-avoidance hierarchy. Full-scale validation remains scarce, durability unquantified, and the available data biased toward moderate-seismicity codes. The way forward is strong fastener-weak connector design, with stiffness-compatibility guidelines that treat InMC, IMC and MFC as one system.
Prestress force and bending rigidity are coupled in prestressed concrete beams, while prestress-force information is weak in deflection responses. This study proposes an incremental second-order-effect-guided physics-informed neural network (iSOE-PINN) for sequential estimation. A first-order model identifies an equivalent bending rigidity, which is transferred to a second-order PINN for prestress-force estimation. Increasing load scales displacement but does not change the nondimensional axial-force parameter or normalized second-order contribution; any benefit therefore arises from measurement SNR rather than physical decoupling. Independent tests give maximum errors of 0.6% for rigidity and 6.3% for prestress force under 1% Gaussian noise. A 0.26% rigidity-transfer bias predicts an 11.3% prestress-force error, consistent with the sequential result. Across 10 initializations, the mean prestress-force error is 10.89% (SD 0.81 percentage points; 95% CI 10.31%-11.47%). Nine experimental cases yield errors of -3.3% to 3.3% for rigidity and -39.2% to 29.5% for prestress force. The method is therefore a preliminary baseline rather than a field-ready diagnostic tool.
The purpose of this study is to propose a new understanding of the influence of plastic softening of strength parameters on the mechanical properties of the coal surrounding borehole. Firstly, a thorough analysis of experimental data from previous studies reveals that the plastic softening behavior of strength parameters essentially manifests as a reduction in cohesion and a variation in internal friction angle. Subsequently, a theoretical and numerical model for deformation and stress distribution of coal surrounding borehole is established, which considers the softening effect of strength parameters and covers different regions. The maximum error between the numerical and theoretical results is found to be within 12%, which validates the reliability of the proposed analytical model. Finally, field verification results confirm that the mechanical response results derived from the model considering only the plastic softening effect of cohesion are smaller than the field-measured data, indicating that the influence of the internal friction angle must also be adequately accounted for in related analyses.
The design of occupant restraint system (ORS) parameters is directly related to the crash safety performance of vehicles. However, traditional iterative simulation and manual parameter tuning methods suffer from long design cycles and high computational costs, while existing deep learning methods mostly focus on indirect modeling or injury prediction, lacking direct parameter design capability for multimodal data. To address this issue, this paper proposes a CNN-based ORS parameter inverse design model (ORS-PDM) that achieves unified input and joint representation of three types of multimodal parameters: time-series, image, and numerical data. Specifically, the crash pulse (time-series modality), the grayscale image of the human-machine layout (image modality), and the HIC metric (numerical modality) are taken as heterogeneous inputs. After unified preprocessing, these data are encoded into tensor channels and concatenated at the channel level, and then fed into the network to directly predict the optimal ORS parameter combination, thereby realizing inverse parameter design.
Crashworthiness surrogates are usually trained on successful finite element (FE) simulations, although automated full-vehicle workflows may include mixed-variable designs that fail during reconstruction, meshing, solution, or response extraction. This study proposes a simulation-admissibility-aware surrogate-assisted framework for electric-vehicle side-pole crashworthiness design. A battery-electric SUV was parameterized with 32 continuous geometry and 36 discrete material/thickness variables. Of 600 SFE–LS-DYNA designs, 528 were admissible. A workflow-admissibility classifier was evaluated by repeated stratified nested cross-validation. In the admissible domain, the structure-informed MS-GIN surrogate achieved a mean R2 of 0.91 and NRMSE of 6.5%. Bootstrap SHAP identified influential variables in the B-pillar and lower body. NSGA-II optimization with FE verification yielded a compromise design reducing body mass, B-pillar chest intrusion, and battery-neighborhood intrusion by 3.6%, 9.5%, and 14.6%, while satisfying the other constraints. The framework captures admissibility risk, structural sensitivity, and lightweighting–crashworthiness trade-offs in mixed-variable crash design.
Residual vertical velocity during late-stage airbag cushioning can cause localized hard contact between an airdrop-equipment subframe and the ground, which whole-vehicle overload measurements cannot resolve. This study characterizes region-specific hard-impact responses and identifies vulnerable subframe regions. An equivalent contact framework based on residual-velocity equivalence considered contact reaction force, load-transfer paths, and regional dynamic responses. Free-drop tests (100, 150, 200 mm) used a four-region subframe and six uniaxial accelerometers at 50 kHz. Records (0.4 s) were wavelet-detrended and 10 kHz low-pass filtered. Peak acceleration, RMS acceleration, and pseudo-velocity shock response spectrum (PV-SRS) were evaluated. Responses concentrated in the wheel-arch beam and central circular load-bearing regions. At 200 mm, C6 had peak acceleration 707 g, RMS 86 g, and PV-SRS peak 6.9 m/s, with peak frequency 322–335 Hz; C5's dominant peak shifted from 1 kHz to 276 Hz. The framework reveals hard-impact differences obscured by whole-vehicle measurements and supports landing-impact safety and structural reliability assessment of airdrop equipment.
Circular holes influence stress fields and raise local stresses in finite width components. This work is devoted to the analytical modeling and the finite element analysis of a rectangular plate with a central circular hole under uniaxial tension. Stress-concentration data from a classical design curve were fitted to a third-order polynomial by ordinary least squares and combined with the net-section nominal stress to produce a dimensionless maximum-stress model that accounts for local amplification and ligament reduction. The formulation recovers the Kirsch limit for a plate of infinite width and predicts an asymptotic rise in stress as the remaining ligament approaches zero. The stationary points were classified according to their mathematical and physical admissibility. The physically admissible stationary points were x=0.0195 (normalized-stress-only case) and x=0.2158 (equal-weight case) which were both treated as objective-function-dependent geometric indicators rather than universal optima. Analytical–numerical differences were in the range of 1.40–3.65% with a mean of 2.69% from Abaqus/CAE simulations using locally refined CPS8R plane-stress elements.