HAL-INSA Toulouse
Not a member yet
34325 research outputs found
Sort by
Simulation of a GaInP 2 -GaAs -GaInAsN -Ge photovoltaic cell for space applications
No full textInternational audienceWe use an open-source simulation software to design III-V multi-junction solar cells containing a dilute nitride subcell. We discuss the different modeling techniques in terms of physical models and numerical methods, and provide an educated guess for a GaInP2 -GaAs -GaInAsN -Ge lattice-matched 4-junction solar cell performance assessment. We suggest a way to ensure an effective computation of the main output values of an opensource Poisson-Drift-Diffusion solver for multi-junction solar cell analysis. An optimization of a multi-junction cell including a n-i-p subcell is performed, starting from basic theoretical considerations, to the use of a double-layer anti-reflective coating to target the most limiting subcell, and then with an optimization of the n-i-p architecture in the 1eV subcell with respect to the GaInAsN layer properties. We show that the detrimental effects of low diffusion lengths in the GaInAsN layer can be alleviated by taking advantage of field-assisted collection
COARSE DISTANCE FROM DYNAMICALLY CONVEX TO CONVEX
No full textChaidez and Edtmair have recently found the first examples of dynamically convex domains in R 4 that are not symplectomorphic to convex domains (called symplectically convex domains), answering a long-standing open question. In this paper, we discover new examples of such domains without referring to Chaidez-Edtmair's criterion in [3]. We also show that these domains are arbitrarily far from the set of symplectically convex domains in R 4 with respect to the coarse symplectic Banach-Mazur distance by using an explicit numerical criterion for symplectic non-convexity.Along with the proof of Theorem 1.1, we discover a family of dynamically convex domains X Ωp (see (3)), parametrized by p ∈ (0, 1], that are not symplectomorphic to convex ones when p is sufficiently small.</div
Uncertainty-aware surrogate modeling for urban air pollutant dispersion prediction
No full textInternational audienceThis study evaluates a surrogate modeling approach that provides rapid ensemble predictions of air pollutant dispersion in urban environments for varying meteorological forcing, while estimating irreducible and modeling uncertainties. The POD–GPR approach combining Proper Orthogonal Decomposition (POD) and Gaussian Process Regression (GPR) is applied to emulate the response surface of a Large-Eddy Simulation (LES) model of the Mock Urban Setting Test (MUST) field-scale experiment. We design and validate new methods for (i) selecting the POD-latent space dimension to avoid overfitting noisy structures due to atmospheric internal variability, and (ii) estimating the uncertainty in POD–GPR predictions. To train and validate the POD–GPR surrogate in an offline phase, we build a large dataset of 200 LES 3-D time-averaged concentration fields, which are subject to substantial spatial variability from near-source to background concentration and have a very large dimension of several million grid cells. The results show that POD–GPR reaches the best achievable accuracy levels, except for the highest concentration near the source, while predicting full fields at a computational cost five orders of magnitude lower than an LES. The results also show that the proposed mode selection criterion avoids perturbing the surrogate response surface, and that the uncertainty estimate explains a large part of the surrogate error and is spatially consistent with the observed internal variability. Finally, POD–GPR can be robustly trained with much smaller datasets, paving the way for application to realistic urban configurations
Time-Reversal Symmetry in RDMFT and pCCD with Complex-Valued Orbitals
No full textInternational audienceReduced density matrix functional theory (RDMFT) and coupled cluster theory restricted to paired double excitations (pCCD) are emerging as efficient methodologies for accounting for the so-called non-dynamic electronic correlation effects. Up to now, molecular calculations have been performed with real-valued orbitals. However, before extending the applicability of these methodologies to extended systems, where Bloch states are employed, the subtleties of working with complex-valued orbitals and the consequences of imposing time-reversal symmetry must be carefully addressed. In this work, we describe the theoretical and practical implications of adopting time-reversal symmetry in RDMFT and pCCD when allowing for complex-valued orbital coefficients. The theoretical considerations primarily affect the optimization algorithms, while the practical implications raise fundamental questions about the stability of solutions. Specifically, we find that complex solutions lower the energy when non-dynamic electronic correlation effects are pronounced. We present numerical examples to illustrate and discuss these instabilities and possible problems introduced by N-representability violations
Towards a machine learning operations (MLOps) soft sensor for real-time predictions in industrial-scale fed-batch fermentation
No full textInternational audienceReal-time predictions in fermentation processes are crucial because they enable continuous monitoring and control of bioprocessing. However, the availability of online measurements is limited by the availability and feasibility of sensing technology. Soft sensors - or software sensors that convert available measurements into measurements of interest (product yield, quality, etc.) - have the potential to improve efficiency and product quality. Machine learning (ML) based soft sensors have gained increased popularity over the years since they can incorporate variables that are measured in real-time, and exploit the intricate patterns embedded in such voluminous datasets. However, ML-based soft sensor requires more than just a classical ML learner with an unseen test set to evaluate the quality prediction of the model. When a ML model is deployed in production, its performance can deteriorate rapidly leading to an unanticipated decline in the quality of the output and predictions. Here a proof concept of Machine Learning Operations (MLOps) to automate the end-to-end soft sensor lifecycle in industrial scale fed-batch fermentation, from development and deployment to maintenance and monitoring is proposed. Using the industrial-scale penicillin fermentation (IndPenSim) dataset that includes 100 fermentation batches, to build a soft sensor based on Long Short Term Memory (LSTM) for penicillin concentration prediction. The batches containing deviations in the processes (91–100) were used to assess concept drift of the LSTM soft sensor. The evaluation of concept drift is evidenced by the soft sensor performance falling below the set threshold based on the Population Stability Index (PSI), which automatically triggers an alert to run the retraining pipeline
Computational protein design
No full textInternational audienceCombining molecular modeling, machine-learned models, and an increasingly detailed understanding of protein chemistry and physics, computational protein design (CPD) and human expertise have been able to produce new protein structures, assemblies, and functions that do not exist in nature. Currently, generative deep-learning based methods, which exploit large databases of protein sequences and structures, are revolutionizing the field, leading to new capabilities, improved reliability and democratized access in protein design. This Primer provides an introduction to the main approaches in CPD, covering both physics-based and machine-learning-based tools. It aims to be accessible to biological, physical, and computer scientists alike. Emphasis is placed on understanding the practical challenges arising from limitations in our fundamental understanding of protein structure and function, and recent developments and new ideas that may help transcend these
A new ISO standard for the experimental characterization of in-plane permeability of fibrous reinforcements
No full textInternational audienceDuring Liquid Composite Molding, a fibrous reinforcement is impregnated with liquid resin. Process design requires knowledge of the reinforcement permeability for fluid flow, but until recently, there has been no standard available for its measurement. In 2023, following decades of benchmarking activities and a standardization project, an ISO standard for the experimental characterization of in-plane permeability of fibrous reinforcements for liquid composite molding was finally published. It focuses on the experimental characterization of unsaturated in-plane permeability and specifies the requirements for test equipment, methods and data analysis. Given the deficiency of standardized procedures within the composites industry, this paper intends to provide an example of the steps towards standardization and summarizes lessons learned. It illustrates the research milestones that led to the establishment of the standard, promotes the standard by detailing its general content and notable features and finally gives explanations and reasoning behind the developed guidelines
Excited-State-Specific Kohn-Sham Formalism for the Asymmetric Hubbard Dimer
No full text14 pages, 10 figuresInternational audienceBuilding on our recent study [https://doi.org/10.1021/acs.jpclett.3c02052, J. Phys. Chem. Lett. 14, 8780 (2023)], we explore the generalization of the ground-state Kohn-Sham (KS) formalism of density-functional theory (DFT) to the (singlet) excited states of the asymmetric Hubbard dimer at half-filling. While we found that the KS-DFT framework can be straightforwardly generalized to the highest-lying doubly-excited state, the treatment of the first excited state presents significant challenges. Specifically, using a density-fixed adiabatic connection, we show that the density of the first excited state lacks non-interacting -representability. However, by employing an analytic continuation of the adiabatic path, we demonstrate that the density of the first excited state can be generated by a complex-valued external potential in the non-interacting case. More practically, by performing state-specific KS calculations with exact and approximate correlation functionals -- each state possessing a distinct correlation functional -- we observe that spurious stationary solutions of the KS equations may arise due to the approximate nature of the functional
Experimental and numerical multiscale testing of CFRP bonded stepped repairs
No full textInternational audienceBonded stepped repairs to aircraft composite structures offer many advantages, such as a smooth aerodynamicsurface, high strength, and low mass addition. However, their design remains challenging due to the varyingstiffness along the bondline in a thin laminate. This study investigates, numerically and experimentally, to whatextent an “equivalent” stepped joint can be used to design a stepped repaired panel. In the proposed case, failureis driven by laminate fracture instead of patch disbonding. Tension tests on stepped repairs at the scale ofcoupons and panels were carried out in 11 different configurations. Specimens were obtained by hot-bonding asit would be done to perform in-situ repairs. Finite element modelling was performed with cohesive zonemodelling to account for disbonding and delamination, and continuum damage mechanics to simulate compositefailure. This multiscale experimental study showed that stepped repaired coupons have a similar behaviour torepaired panels in terms of damage mechanisms, failure onset location, and tensile strength. It supports the ideato use coupons instead of whole panels to carry out experimental testing of stepped repairs. A good agreementwith 2D and 3D numerical simulations was also found. They predicted accurately the strength of the repairs andhighlighted a failure location compatible with the experimental results. As a conclusion, an equivalent steppedjoint can be representative for the strength of a stepped repaired panel, including when failure occurs inside thelaminates.<br /
Easy reversible clustering of gold nanoparticles via pH-Induced assembly of PVP-b-PAA copolymer
No full textInternational audienceThe growing demand of novel hybrid organic/inorganic systems with exciting properties has contributed to an increasing need for simplifying production strategies. Here, we report a simple method to obtain controlled three-dimensional hybrid architectures, in particular hybrid supracolloids (hSC), formed by gold nanoparticles and a double hydrophilic block copolymer, specifically the poly(acrylic acid)-block-poly(N-vinyl-2-pyrrolidone) (PAA-b-PVP), directly in aqueous medium. The ubiquitous pH-sensitive poly(acrylic acid) (PAA) block initiates the assembly through pH changes, while the poly(N-vinyl-2-pyrrolidone) block assures the close affinity with the AuNPs. We demonstrate that the formation of hybrid supracolloids (hSC) is the result of the synergetic behavior of the two specific polymeric blocks. Additionally, the entire process shows spontaneous and fast switchability. The nanostructured copolymer behaves like a highly swollen hydrogel and displays a disordered internal structure. The driving force for the association of the copolymer chains is induced by the synergetic effects of the decrease in solubility of the poly(acrylic acid) block and the formation of inter and intra chains hydrogen bonds. These were demonstrated by using small angle X-ray scattering (SAXS), quartz crystal microbalance with dissipation monitoring (QCM-D) and scanning transmission electron microscopy coupled with energy-dispersive X-ray spectroscopy (STEM-EDX). In turn, the AuNPs are randomly spread all over the polymeric matrix, as demonstrated by field emission gun - scanning electron microscopy (FEG-SEM). A correlation analysis reveals the hSC density depends mostly on the initial concentration of AuNPs. These results can inspire the fabrication of more complex structures with multicomponent composition