David Ryckelynck

Professor

Keywords

Team

CSM - Mécanique Numériques des Solides

Biography

David Ryckelynck is a researcher specializing in computational mechanics and materials modeling, with significant expertise in the development of advanced model reduction and machine learning methods applied to industrial problems. His work focuses on the intersection of numerical simulation, data science, and structural mechanics, aiming to optimize high-fidelity calculations while preserving their accuracy. Among his major research areas are order-of-size reduction (ROM) for nonlinear problems, multimodal data analysis in materials science, and the integration of artificial intelligence techniques to improve physical simulations, particularly in contexts such as additive manufacturing, material fatigue, or defect characterization via 3D imaging. His research also explores innovative approaches such as variational autoencoders, compressed convolutional neural networks, and mesh morphing methods for applications in fluid dynamics and solid mechanics. His approach often combines rigorous mathematical tools, such as spectral decompositions and clustering methods, with deep learning techniques to process complex, high-dimensional data.

Publication(s)

2026
Functional maps regularization for high quality mesh morphing applied to shape registration in computed tomography DOI : 10.1007/s00366-025-02243-8
2025
Construction of Data Sequence for Model Order Reduction in Thermomechanical Modeling of DED Additive Manufacturing DOI : 10.1002/nme.70005
2025
Multimodal super-resolution for fast image-based simulation of crystal plasticity DOI : 10.1016/j.cma.2025.118210
2025
Quaternion-based vision-transformer for polycrystalline EBSD scans pre-trained on large-scale synthetic data DOI : 10.1016/j.matdes.2025.114599
2024
Coupled Laplacian Eigenmaps for Locally-Aware 3D Rigid Point Cloud Matching DOI : 10.1109/CVPR52733.2024.00331
2024
Manifold Learning: Model Reduction in Engineering DOI : 10.1007/978-3-031-52764-7
2023
A priori compression of convolutional neural networks for wave simulators DOI : 10.1016/j.engappai.2023.106973
2023
A Simple and Robust Framework for Cross-Modality Medical Image Segmentation applied to Vision Transformers DOI : 10.1109/ICCVW60793.2023.00446
2023
Real-time numerical prediction of strain localization using dictionary-based ROM-nets for sitting-acquired deep tissue injury prevention DOI : 10.1016/B978-0-32-389967-3.00027-5
2022
An updated Gappy-POD to capture non-parameterized geometrical variation in fluid dynamics problems DOI : 10.1186/s40323-022-00215-x
2022
Mechanical Fatigue Testing Under Thermal Gradient and Manufacturing Variabilities in Nickel-Based Superalloy Parts with Air-Cooling Holes DOI : 10.1007/s11340-022-00868-0
2022
Deep multimodal autoencoder for crack criticality assessment DOI : 10.1002/nme.6905
2022
Condition Number and Clustering-Based Efficiency Improvement of Reduced-Order Solvers for Contact Problems Using Lagrange Multipliers DOI : 10.3390/math10091495
2022
Uncertainty quantification in a mechanical submodel driven by a Wasserstein-GAN DOI : 10.1016/j.ifacol.2022.09.139
2022
Physics-informed cluster analysis and a priori efficiency criterion for the construction of local reduced-order bases DOI : 10.1016/j.jcp.2022.111120
2022
Multimodal data augmentation for digital twining assisted by artificial intelligence in mechanics of materials DOI : 10.3389/fmats.2022.971816
2022
A Bayesian Nonlinear Reduced Order Modeling Using Variational AutoEncoders DOI : 10.3390/fluids7100334
2022
Deep learning model to assist multiphysics conjugate problems DOI : 10.1063/5.0077723
2022
Uncertainty quantification for industrial numerical simulation using dictionaries of reduced order models DOI : 10.1051/meca/2022001
2021
Data-targeted prior distribution for variational autoencoder DOI : 10.3390/fluids6100343
2021
Mechanical assessment of defects in welded joints: morphological classification and data augmentation DOI : 10.1186/s13362-021-00114-7
2021
Data-driven reduced bond graph for nonlinear multiphysics dynamic systems DOI : 10.1016/j.amc.2021.126359
2021
Real-time data assimilation in welding operations using thermal imaging and accelerated high-fidelity digital twinning DOI : 10.3390/math9182263
2021
A Modular U-Net for Automated Segmentation of X-Ray Tomography Images in Composite Materials DOI : 10.3389/fmats.2021.761229
2021
A pruning algorithm preserving modeling capabilities for polycrystalline data DOI : 10.1007/s00466-021-02075-5
2021
Mechanical dissimilarity of defects in welded joints via Grassmann manifold and machine learning DOI : 10.5802/CRMECA.51
2021
Hyper-reduced arc-length algorithm for stability analysis in elastoplasticity DOI : 10.1016/j.ijsolstr.2020.10.014
2020
Crystal plasticity modeling of the cyclic behavior of polycrystalline aggregates under non-symmetric uniaxial loading: Global and local analyses DOI : 10.1016/j.ijplas.2019.10.007
2020
Hyper-reduced direct numerical simulation of voids in welded joints via image-based modeling DOI : 10.1002/nme.6320
2020
Model order reduction assisted by deep neural networks (ROM-net) DOI : 10.1186/s40323-020-00153-6
2020
Reduced Order Modeling Assisted by Convolutional Neural Network for Thermal Problems with Nonparametrized Geometrical Variability DOI : 10.1007/978-3-030-52246-9_17
2020
A nonintrusive distributed reduced-order modeling framework for nonlinear structural mechanics—Application to elastoviscoplastic computations DOI : 10.1002/nme.6187
2020
Deep Convolutional Generative Adversarial Networks Applied to 2D Incompressible and Unsteady Fluid Flows DOI : 10.1007/978-3-030-52246-9_18
2018
Hybrid hyper-reduced modeling for contact mechanics problems DOI : 10.1002/nme.5798
2018
Computer Vision with Error Estimation for Reduced Order Modeling of Macroscopic Mechanical Tests DOI : 10.1155/2018/3791543
2017
Towards error bounds of the failure probability of elastic structures using reduced basis models DOI : 10.1002/nme.5554
2017
Hyper-reduction of generalized continua DOI : 10.1007/s00466-016-1371-2
2017
Fragmentation modeling of a resin bonded sand DOI : 10.1051/epjconf/201714011004
2016
Hyper-reduction framework for model calibration in plasticity-induced fatigue DOI : 10.1186/s40323-016-0068-6
2016
A posteriori global error estimator based on the error in the constitutive relation for reduced basis approximation of parametrized linear elastic problems DOI : 10.1016/j.apm.2015.11.016
2016
Hyper-reduced predictions for lifetime assessment of elasto-plastic structures DOI : 10.1007/s11012-015-0244-7
2015
POD Preprocessing of IR Thermal Data to Assess Heat Source Distributions DOI : 10.1007/s11340-014-9858-2
2015
Thermo-mechanical numerical simulation of the cooling-down of high zirconia fused-cast refractory blocks
2015
Modelling and prediction of deformation during sintering of a metal foam based SOFC (EVOLVE) DOI : 10.1149/06801.2971ecst
2015
Estimation of the validity domain of hyper-reduction approximations in generalized standard elastoviscoplasticity DOI : 10.1186/s40323-015-0027-7
2014
Micro-mechanical characterization of lead-free solder joints in power electronics DOI : 10.1109/ITHERM.2014.6892271
2014
Architectured materials to improve the reliability of power electronics modules: Substrate and lead-free solder DOI : 10.1007/s11664-013-2662-4
2014
Architectured bimetallic laminates by roll bonding: Bonding mechanisms and applications DOI : 10.1179/1743284713Y.0000000412
2013
A priori hyper-reduction method for coupled viscoelastic-viscoplastic composites DOI : 10.1016/j.compstruc.2012.11.017
2012
Multidimensional hyper-reduction of large mechanical models involving internal variables DOI : 10.1115/ESDA2012-82971
2012
Beremin model: Methodology and application to the prediction of the Euro toughness data set DOI : 10.1016/j.engfracmech.2011.10.019
2012
Numerical simulation of the cooling-down of high-zirconia fused-cast refractories DOI : 10.1016/j.jeurceramsoc.2012.06.004
2012
Bimodal Beremin-type model for brittle fracture of inhomogeneous ferritic steels: Theory and applications DOI : 10.1016/j.engfracmech.2011.10.016
2012
Multidimensional a priori hyper-reduction of mechanical models involving internal variables DOI : 10.1016/j.cma.2012.03.005
2011
A robust adaptive model reduction method for damage simulations DOI : 10.1016/j.commatsci.2010.11.034
2011
Reduced-order modelling for solving linear and non-linear equations DOI : 10.1002/cnm.1286
2011
A priori reduction method for solving the two-dimensional Burgers' equations DOI : 10.1016/j.amc.2011.01.065
2010
Toward "green" mechanical simulations in materials science: Hyper-reduction of a polycrystal plasticity model DOI : 10.3166/ejcm.19.365-388
2010
Multi-level a priori Hyper-Reduction of mechanical models involving internal variables DOI : 10.1016/j.cma.2009.12.003
2010
Truncated Integration for Simultaneous Simulation of Sintering Using a Separated Representation DOI : 10.1007/s11831-010-9055-0
2010
Anisotropic constitutive model and FE simulation of the sintering process of slip cast traditional porcelain DOI : 10.1063/1.3457622
2009
Hyper Reduction of finite strain elasto-plastic models DOI : 10.1007/s12289-009-0424-x
2009
Hyper-reduction of mechanical models involving internal variables DOI : 10.1002/nme.2406
2008
Detection of deviations origins in a heat treatment process using Proper Orthogonal Decomposition (POD) basis DOI : 10.1007/s12289-008-0200-3
2007
Anticipation of gears distortions caused by heat treatments DOI : 10.1063/1.2729681
2007
A new fully coupled two-scales modelling for mechanical problems involving microstructure: The 95/5 technique DOI : 10.1016/j.cma.2006.10.013
2007
Micro-Macro approach for mechanical problems involving microstructure DOI : 10.1063/1.2729701
2006
Adaptive model reduction for sensitivity analysis
2006
On the reduction of kinetic theory models related to finitely extensible dumbbells DOI : 10.1016/j.jnnfm.2006.01.007
2006
A simple error indicator for meshfree methods based on natural neighbors DOI : 10.1016/j.compstruc.2006.04.002
2006
On the a priori model reduction: Overview and recent developments DOI : 10.1007/BF02905932
2006
Deterministic particle approach of Multi Bead-Spring polymer models DOI : 10.3166/remn.15.481-494
2006
An adaptive ROM approach for solving transfer equations DOI : 10.3166/remn.15.589-605
2005
The constrained natural element method (C-NEM) for treating thermal models involving moving interfaces DOI : 10.1016/j.ijthermalsci.2004.12.007
2005
α-NEM and model reduction: Two new and powerful numerical techniques to describe flows involving short fibers suspensions DOI : 10.3166/reef.14.903-923
2005
Treating moving interfaces in thermal models with the C-NEM DOI : 10.1007/3-540-27099-x_14
2005
A priori hyperreduction method: An adaptive approach DOI : 10.1016/j.jcp.2004.07.015
2005
An efficient 'a priori' model reduction for boundary element models DOI : 10.1016/j.enganabound.2005.04.003
2004
A new extension of the natural element method for non-convex and discontinuous problems: The constrained natural element method (C-NEM) DOI : 10.1002/nme.1016
2003
Natural interpolation on non-convex bodies by using the constrained Vorono' diagram: C-natural elements method; [Interpolation naturelle sur les domains non convexes par l'utilisation du diagramme de Voronoï contraint] DOI : 10.3166/reef.12.487-509
2002
Friction modelisation of tool workpiece contact for the finite element simulation of cutting process; [Modeĺlisation du frottement outil-pièce pour la simulation de la coupe par la méthode des éléments finis] DOI : 10.1016/S1296-2139(02)01172-7
2002
An a priori model reduction method for thermomechanical problems; [Réduction a priori de modèles thermomécaniques] DOI : 10.1016/S1631-0721(02)01487-0
2000
Efficient adaptive strategy to master the global quality of viscoplastic analysis DOI : 10.1016/S0045-7949(00)00107-3

Teaching

Elective Course Period (October and January)

Course Director

Finite Elements

Course Director

The course consists of a theoretical component (including lectures and small-group sessions) and a practical component based on a mini-project. The ""lecture"" component includes a presentation of the theoretical framework of the method (10 lecture and tutorial sessions), as well as its implementation (5 lecture and tutorial sessions), and the industrial context (5 sessions). The mini-projects account for one-third of the course. Students will choose their projects from a list of topics focused either on specific applications or on algorithmic or mathematical developments: dynamics, statics, contact, thermal analysis, diffusion, fluid mechanics, mesh adaptation, etc. Content Theoretical formulation and implementation of the method Variational formulation: Sobolev spaces, weak solutions, the Lax–Milgram theorem, connection to virtual work and the calculus of variations. Finite Element Method: Description of the method, examples of finite elements, convergence results. Matrix formulation: Elementary matrices, assembly, boundary conditions, solution, algorithms in mechanics and thermal analysis. Computational environment Connection to CAD: Design workflows. Meshing techniques: Delaunay methods, frontal meshing. Meshing a complex part. Major commercial software packages: Organization of a computational code, some examples. Parallel computing: Machines and associated algorithms, parallel solvers, domain decomposition. Applications Students interested in applications involving real parts will use major commercial software or software from the School’s Research Centers. They will perform a calculation that is realistic from an industrial perspective. Those more interested in applied mathematics can use a development platform, such as FreeFem++.

Digital Engineering of Complex Systems (IDSC) track

Course Director

Over the past decade, artificial intelligence (AI) has profoundly transformed the industrial world by revolutionizing the ways in which systems are designed, optimized, and operated. One of the most striking aspects of this evolution is the digital modeling of complex systems. Traditionally based on methods with a strong physical or multiphysical focus, this field now benefits from advances in artificial intelligence, particularly through machine learning and deep learning algorithms. The Digital Engineering of Complex Systems track at Mines Paris aims to provide students with a solid, multidisciplinary foundation in AI, enabling them to gain a deep understanding of its applications and apply them to real-world industrial challenges. The disciplines studied in the specialization are: machine learning and its applications in engineering, computer vision, and high-performance computing. The specialization’s curriculum is structured around three weeks of coursework, supplemented by two weeks of a Data Challenge and a one-week field trip. More information is available on the track’s webpage: https://bruno-figliuzzi.github.io/Option/

Data, Images, Physical Models, and Machine Learning (Research Quarter)

Course Director

Data analysis is playing an increasingly important role in our society, both professionally and personally. This data is often complex—including text, images, videos, genomes, point clouds, and graphs, for example. It serves as the raw material for the digital industries. Automatically extracting useful information from these massive datasets is a major challenge. The goal of this research term is to provide engineering students with their first experience in research on the automatic analysis of complex data. Images and other structured data (graphs, trees, sequences, etc.) will be the focus of this work. The scientific disciplines involved will include, in particular, machine learning, image analysis, robotics, physics, and bioinformatics. Depending on the projects, students will have the opportunity to use kernel methods, deep learning, or mathematical morphology, among other techniques. The fields of application will be very diverse, ranging from autonomous driving to healthcare, including non-destructive testing and materials characterization. Some research projects will take place in collaboration with industry partners.

PhD supervision

2025 Development of deep learning-based digital twins for hot forming processes of Zr alloys NGUYEN Thanh Chung
2025 Data-driven learning of an electric vehicle battery aging model HOLLER Colin
2024 Machine learning for anomaly detection via nanoindentation to identify elastoplastic properties of subgrain-level crystals BELAMRI-REGENPIED Pierre
2023 Calculation on data from automatic tomography for large geometric variations: applications to process simulation and verification of mechanical functions FERHAT AMELIA
2023 Modeling and AI methods to estimate and predict the health states of electric vehicle batteries. VU Germain
2022 Advances in 3D vision and deep learning for shape analysis and synthesis: application to interspecies biomechanical modeling of the knee joint BASTICO Matteo
2021 Deep Learning Application to Prediction and Modeling of Multiphysics Fluid Flows EL HABER George
2021 Identification of crystal plasticity laws using digital twin and statistical learning MESBAH Daria
2020 Hyper-reduction of model order in contact mechanics under potentially non-symmetric mixed formulation: theoretical and numerical analysis. LE BERRE Simon
2019 Real-time prediction of mechanical stresses in welding using a metamodel library PEREIRA ALVAREZ Pablo
2019 Machine learning-driven mechanical submodels: application to structural dynamics BOUKRAICHI Hamza
2019 Functional metrology by 3D imaging and digital twin learning – application to thermo-mechanical fatigue in single-crystal superalloys. AUBLET Axel
2018 Machine learning reduced-order models for studying defect harmfulness LAUNAY Hugo
2018 Statistical learning for nonlinear model reduction DANIEL Thomas
2017 Model Reduction applied to interfacial flows TEYOU TCHUIENKAM Patrick Lionnel
2016 Numerical study of the harmfulness of defects in welds LACOURT Laurent
2016 A reduced 0D unsteady and non-linear thermal model of vehicle cabin for automotive energy optimization HAMMADI Youssef
2016 Crystalline plasticity applied to polycrystals under asymmetric cyclic loading: mechanical analysis and model order reduction FAROOQ Harris
2015 Reduced-order models for simplified welding simulation as a substitute for out-of-reach calculations. DINH TRONG Tuan
2015 Modeling and simulation of polyurethane-bonded sand fragmentation HILTH William
2015 Reduced-order model in contact mechanics. Application to the simulation of nuclear fuel behavior. FAUQUE DE MAISTRE Jules