LNE - Mise en œuvre et validation d’un jumeau numérique pour une machine à haute exactitude de mesure
Type de poste:
Description
Contexte et objectifs de la thèse :
La cylindricité est une spécification géométrique décrivant la forme tridimensionnelle d'un élément cylindrique. Son contrôle à l'aide de machines commerciales soulève un nombre de problèmes pratiques qui limitent les capacités de mesure actuelles. En effet, il n'existe actuellement aucun instrument de mesure capable de générer une cartographie tridimensionnelle de l’écart de cylindricité avec une incertitude nanométrique. En conséquence, le Laboratoire Commun de Métrologie LNE-CNAM a développé une machine de référence pour la mesure de cylindricité de très haute exactitude « NanoCyl » afin d’améliorer ces capacités actuelles d'étalonnage et de mesure de cylindricité. Les résultats obtenus montrent une incertitude de mesure atteignable de l’ordre de 40 nm.
Pour atteindre des incertitudes de mesure inférieures à 20 nm, le LNE-CNAM conjointement avec le LURPA (ENS Paris-Saclay) proposent un sujet de thèse dont les objectifs scientifiques sont :
(1) implémentation d'un jumeau numérique pour la NanoCyl
(2) développement d’un outil de validation du jumeau numérique
(3) calcul des incertitudes de mesure
(4) réalisation d’un cas d’étude sur un ensemble piston-cylindre industriel.
Les travaux de recherche proposés s’inscrivent dans le cadre du programme européen de recherche 22DIT01-ViDiT.
Cylindricity is a geometric specification that describes the three-dimensional form/shape of a cylindrical element. Its control using commercial machines raises a number of practical issues that limit current measurement capabilities. Indeed, there is currently no measuring instrument capable of generating a three-dimensional representation of the cylindricity deviations with a nanometric uncertainty. The Laboratoire Commun de Métrologie LNE-CNAM has recently developed an ultra-high accuracy cylindricity measuring machine named “NanoCyl”. As a primary standard, the NanoCyl aims to improve the current calibration and cylindricity measurement capabilities. The given results show an achievable measurement uncertainty of the order of 40 nm.
To enhance the measurement uncertainty below 20 nm, the LNE-CNAM in collaboration with the LURPA (ENS Paris-Saclay) propose the following research including:
(1) implementation of one hybrid digital twin for the developed primary NanoCyl
(2) a validation method of the implemented digital twin
(3) a method for measurement uncertainty evaluation
(4) investigation of a case study on one selected industrial piston cylinder assembly
The proposed PhD and scientific objectives are part of the European collaborative project 22DIT01-ViDiT funded by EPM-EURAMET.
The main scientific objectives of the proposed research are:
1. Investigation of the state-of-the-art of DTs applied for measurement instruments and selection and implemention of one hybrid DT (which will be data driven and model driven) for the developed primary NanoCyl. The NanoCyl is based on the comparison of the form of a cylindrical part (measured with at least 2 contactless probes) with the form of a reference cylinder (measured in real time by at least 8 capacitive probes). An additional error model, related to the physical behaviour of the integrated probes, will be investigated and considered in the hybrid DT. Furthermore, the thermal drift of the metrology frame will be measured in situ through PT100 sensors (or other probing systems). An appropriated thermal drift model will be established and considered in the DT.
2. Investigation of the state-of-the-art related to the uncertainty methods that are dedicated to DT evaluation. The target questions to be addressed will be:
• How to define the uncertainty and accuracy of the model?
• How to include the correction and the related uncertainty in the physical measurement’s metrological characteristics and uncertainty?
• How to establish traceability for a DT?
• How to generate synthetic datasets for the validation of the identified uncertainty methods?
3. Selection of an appropriate method, for the validation of the main metrological characteristics (precision, accuracy, sensitivity, etc.) of a DT. The selected method will support and guarantee compliance with the validation approaches. The validation method will also contribute to the evaluation of the accuracy, representativeness and precision estimation of the DT. JCGM:GUM compliant, simulative and numeric, and Bayesian inference-based methods will be considered.
4. Based on the selected fidelity level, and the parameters describing the physical system, influence factors for the NanoCyl that are included in the DT will be identified, characterized and evaluated as to simplify the uncertainty propagation model. The least impacting contributions will be neglected to speed up the evaluation and to reduce the computational effort. One statistical model for the uncertainty influence factors will be proposed in order to study the related statistical distribution and relevant parameters. The modelling will include both empirical observation and the metrological characterization data provided by sensors. The developed method will be applied to evaluate the measurement uncertainty of the implemented DT. The hybrid modelling strategy, as well as JCGM:GUM compliant and Bayesian Monte Carlo methods will be considered.
5. Investigation of one case study such as to demonstrate the applicability of the implemented and validated DT NanoCyl tool to characterize one selected industrial piston cylinder assembly with low uncertainty. This measurement will be complex due to the radial changes, between the piston and the cylinder, in their engagement length as well as the combination of three parameters (radial deviations, median line deviations and cross section deviations) into one 3D measurement.
Profil :
Vous devez faire preuve d'excellentes capacités d'analyse, avoir un intérêt affirmé pour la conception de machines et la métrologie dimensionnelle, la modélisation et la simulation, ainsi que des compétences en matière d’instrumentation.
La maîtrise de l'anglais écrit et oral est exigée pour ce poste.
Une connaissance pratique de l'instrumentation et des outils informatiques pour la modélisation et la simulation est souhaitable.
Vous devez idéalement être titulaire d'un master (ou équivalent) en génie automatique et électrique, génie mécanique, ou génie industriel, avec de solides connaissances en instrumentation, modélisation, simulation et traitement des données.
Candidates must have excellent analytical skills, a strong interest in machine design and dimensional metrology, modeling and simulation methods, expertise in instrumentation, and strong communications skills.
Proficiency in written and oral English is required for this position.
Working knowledge of instrumentation, computational tools for modeling and simulation is desirable.
Applicants should ideally hold a master's degree (or equivalent) in Automation and Control, Mechanical Engineering, Manufacturing Engineering, or Industrial Engineering, with strong background in instrumentation, modeling, simulation and data processing.
Localisations :
- 80 % au LNE : Paris 15ème
- 20 % au LURPA : Gif-sur-Yvette (91190)
Pour candidater, envoyez votre CV et LM à hichem.nouira@lne.fr et recrut@lne.fr en rappelant en objet du mail la référence de l’offre (ML/THJN/DMSI)
URL de l'offre:
https://www.lne.fr/fr/offre-emploi/mise-en-oeuvre-validation-jumeau-numerique-machine-haute-exactitude-mesure
