HAL Arts et Métiers
Not a member yet
14127 research outputs found
Sort by
As-scanned point cloud generation using structured-light simulation and machine learning-based coverage prediction
No full textInternational audienceAlthough several methods have been proposed for generating as-scanned point clouds, i.e. point clouds incorporating various realistic artefacts that would appear if the corresponding real objects were digitized for real, most of them still fail to take into account the complex phenomena that occur in a real acquisition devices. This paper presents a new way of artificially generating point clouds by combining simulation and machine learning. Starting from the CAD model of the object to be virtually scanned and from a scan configuration, structured light simulation first allows reconstructing a preliminary 3D point cloud. Then, a coverage prediction network is used to predict the regions that would be acquired if a real acquisition was to be done. The prediction model has been trained from a large database of scan configurations and point clouds scanned for real. Finally, filtering and cropping are performed to fine-tune the generated point cloud. Experiments confirm that this method can generate point clouds very close to those that a real scanner would acquire, as shown by several metrics characterizing both local and global similarity. Such a virtual scanning technique enables the rapid generation of large quantities of realistic point clouds, especially when compared to the time-consuming and costly processes involved in using physical acquisition systems. This opens up new perspectives in terms of access to realistic point cloud databases, in particular for the development of various AI-based approaches
Squeal occurrence classification using a harmonic balance vector signal model.
No full textInternational audienceBrake squeal is an instability that generates self-excited limit cycles which vary with time and operating conditions in real experiments. To analyze test results, it is proposed to use a Harmonic Balance Vector (HBV) signal model. It combines Harmonic Balance Method and analytic signal methodologies. From the Harmonic Balance Method, one uses the space-time decomposition where spatial distribution of each harmonic is described by a complex vector and frequency is common to all sensors. From analytic signal, one keeps the assumption that quantities are slowly varying in time. Synchronous demodulation and principal coordinate definitions are combined in a multistep algorithm that provides an HBV estimation. On an industrial brake test matrix, HBV estimation is shown to be robustly applicable. The HBV signal being slowly varying, time sub-sampling reduces the volume of test data by two orders of magnitude. Limit cycle frequency, amplitude and shapes can thus be added to the parallel coordinates that associate to each time sample the operating parameters: pressure, velocity, temperature, torque, disk position, disk/bracket distance, ... This opens a path to a range of analyzes otherwise difficult to perform. Classification of squeal occurrences is first discussed showing pressure and amplitude dependence. The effect of amplitude on both frequency and shape is next demonstrated. The entry and exit of instability when parameters change are then analyzed by proposing a transient root locus built from test. Thus squeal test results are related to the classical complex eigenvalue analysis. Intermittent growth/decay events are shown to be correlated with wheel position. Furthermore, distance measurements indicate that disk shape variations of a few microns play a clear parametric role. Parametric testing and clustering are then used to map the instability region and its edges. Pressure is shown to have an effect dominating other variations. Prospective uses of these results to combine test results and finite element models are discussed last
Wave propagation in laminated structure through wave finite element method
No full textInternational audienceIn this paper, the wave finite element(WFE) method is briefly presented and applied in order to extract the dispersion curves. The formulation of the laminated structure is detailed through the Timoshenko theory. The finite element technique is used to model the laminated beam and extract the mass and stiffness matrices for the bending vibration. The bending vibration of the laminated beam is simulated and discussed. The travelling and evanescent modes are illustrated to characterize the flexural wave propagation in laminated structure. The resolution of the equilibrium equation leads to the extraction of the analytical wave number as a function of the frequency in order to validate the dispersion curves simulated through the WFE method. The question of the influence of the layers thickness on the wave propagation is detailed. An uncertainty is introduced in the thickness as a Gaussian variable and the mean and the standard deviation of the dispersion curves are extracted through the Monte Carlo simulation. Among the contributions of this article, the laminated structures are modeled through the Abaqus software and the mass and stiffness matrices are extracted for the multimodal propagation. The multimodal wave number is presented and discussed for the travelling and evanescent modes
Thermal field estimation in CFRTP composites using an attention-enhanced U-Net
No full textInternational audienceThis study presents a surrogate model based on the convolutional U-Net architecture to predict the thermal field in a carbon fibre-reinforced thermoplastic tape at the microscale during brief and localized heating. Leveraging microstructure data within a machine learning framework, the proposed model aims to enhance the accuracy of temperature field predictions at a low computational cost. The incorporation of a co-attention mechanism to handle image channels of different nature significantly improves precision, resulting in a strong correlation between the model’s predictions and the ground truth obtained from the numerical solution of the heat equation. This capability enables rapid assessment of diverse microstructures, facilitating optimization and real-time applications in manufacturing settings
Balancing Confidence and Caution: Artificial Intelligence's Integration in Lean Profession
No full textInternational audienceWith artificial intelligence (AI) transforming operations management, lean professionals—ranging from in-house practitioners to external consultants—face both opportunities andtensions. This study explores how AI influences their practices and roles by applying a Delphi-Régnier method with experts from industry, academia, and consulting. It examinesorganizational, technological, informational, and people challenges. We aim to offerpreliminary insights into the challenges faced by lean professionals, including how theynavigate between executional tasks and strategic advisory roles. It also investigates how AIcomplements human expertise within hybrid decision-making systems. The study will proposepractical guidelines for aligning AI integration with lean principles
A new micro HATI second order preserving for transient dynamics
No full textThe authors would like to thank Framatome and ANRT for funding this research (CIFRE grant # 2019/1124)International audienceSubdomain decomposition along with Heterogeneous Asynchronous Time Integrator, such as the GC (Gravouil and Combescure) method, allows us to adopt the most appropriate time integrator and time step size depending on subdomains. The paper presents an improved version of the GC method employed for various applications ranging from impact dynamics to fluid-structure interaction problems. The proposed method, called BLG method, is based on the kinematic continuity at the fine time scale, as inherited from the progenitor GC method, guaranteeing an easy application to problems whose nonlinearities are taken into account at the fine time scale. Properties of the new BLG method in terms of stability, local truncation error, numerical damping and period elongation are thoroughly assessed thanks to a detailed spectral study of the amplification matrix related to the split oscillator. The BLG method turns out to be second-order accurate when coupling second-order accurate time integrators from the Newmark family, gaining one order with respect to the GC method. Multi-time step explicit/implicit and explicit/explicit coupling with the BLG method are considered as the most interesting cases for promising applications in nonlinear structural dynamics with impacts and complex fluid-structure interaction problems
Pyromechanics: A solid mechanics approach to deformation during pyrolysis
No full textInternational audienceDuring pyrolysis, organic materials undergo morphological changes that are important to predict, particularly in the field of thermochemical conversion. This work proposes an approach to model deformations during pyrolysis combining elastic, thermal expansion, and pyrolysis contributions. A three-dimensional anisotropic pyromechanics model is derived for porous media by volume averaging. It includes pyrolysis kinetics, mass, momentum and energy conservation for both solid and gas phases. A key advantage of the pyromechanics model lies in its physical framework, which effectively captures the effect of internal stresses on the overall deformation. Implemented as open-source within the Porous material Analysis Toolbox based on the OpenFOAM framework (PATO) using an incremental approach, the model is specifically applied to wood in this study. Two experimental validations are conducted using pyrolyzing cylindrical wood particles to verify the model’s predictions concerning temperature profile evolution and shrinkage effects. Importantly, the model’s capability to estimate stress distribution holds promise for further investigations into crack distribution and propagation
Virtual Reality for Surface Topography Analysis
No full textInternational audienceImmersive technologies, such as virtual reality (VR), have seen widespread adoption across multiple sectors, including product design and usage simulation, owing to their interactive and intuitive capabilities that have transformed visualization and user engagement in these fields. Nevertheless, the application of VR in surface topography analysis remains notably limited. Although VR has significant potential to facilitate detailed and immersive examination of surface characteristics, its integration within this domain has not yet matched its use in other industries. This paper presents a VR-based framework for surface analysis, providing an interactive platform for in-depth exploration of surface features. The proposed tool is intended to advance both the understanding and practical application of surface topography.</div
Three-dimensional dynamic criterion to quantify gait instability during slip-like perturbations
No full textInternational audienceDespite numerous attempts at finding the best parameter to quantify gait instability, there is still no consensus. This study introduces a dynamic criterion by measuring the vector between the body center of mass and the minimal moment axis of the external mechanical action (dBCoM-MMA), which is linked to the whole-body angular momentum (WBAM). This study aimed at determining its three-dimensional orthogonal projection onto the MMA during treadmill walking under external perturbations. So far, eight asymptomatic participants walked on a dual-belt instrumented treadmill under controlled slip and trip perturbations. 3D dBCoM-MMA was increased along the mediolateral axis on the perturbed cycle. Further research is needed to establish thresholds for fall prediction
A novel 3D printed structure with encapsulated air gaps for power electronic component
No full textInternational audienceIn high frequency inductors, fringing effect around air gaps could lead to a serious increase of AC copper loss. The addition of bridges allows to encapsulate the air gaps in the magnetic circuit. Representative cores were printed using Fused Deposition Modeling (FDM) process. A significant reduction of the fringing field (-30% near the edges of winding window) has been observed with finite element analysis