Search

International audience

International audience; According to researchers and practitioners, modular products are considered as key enablers for Mass Customization (MC). Modular design is a strategy for designing products composed by many subparts (modules), which are developed separately and assembled later in different ways, enabling the configuration of several product variants. Customers do not always know how to translate their requirements into product configurations, thus an alternative could be 'co-configuring' the product with them. This is done by an automatic modules selection based on customer requirements to get a first product configuration, which can be subsequently adjusted by the customer, if necessary. Currently, the optimal module selection is mainly conducted without considering the process planning optimization. Nevertheless, integrating both optimizations would increase product/process flexibility and agility, reducing the overall manufacturing costs. Thereby, it is important to integrate the product configuration, driven by individual customer requirements, with the process planning for MC. This paper proposes a 0-1 integer linear programming model to optimize, in terms of cost-minimization, the integration of product configuration and process planning, for MC. Results obtained with a 2-steps approach composed by an exhaustive search algorithm (ESA) and CPLEX-solver show that total cost can be considerably minimized if both optimizations are considered together.

International audience

International audience

International audience

International audience

International audience; Short abstract: The implementation of Additive Manufacturing (AM) digital flow and product data management system rely more on data formats and standards that nowadays convey Geometric Dimensioning and Tolerancing (GD&T) information directly in the 3D model. The consideration of geometric deviations and variations is an important issue for tolerancing and geometric quality assessment in AM through product and manufacturing information (PMI). This paper reviews current challenges of geometric models and standard formats for AM technologies, which largely impede its advancement. Two criteria to enhance AM geometric models and standards are presented in this paper: tolerancing and compatibility of PMI. The analysis presented in this paper highlights that STEP standard can be a good basis for future research to integrate geometric models and standards for AM. Additive manufacturing (AM) technologies have been applied to the automotive, aerospace and medical industries for many years [1-2]. However, it is a difficult technology for AM that geometric models describe precision and accuracy of product shape, dimensioning and tolerancing. Geometric models have become ubiquitous and offered essential and different tools for the digital product development [3]. In the 1960s Herbert Voelker considered the possibilities of using computer-aided machine to operate machine through CAD geometry. In the 1970s mathematical tools for describing 3D solid modelling had been invented in the early algorithms [4]. At present, geometric deviation has become a major issue on the tolerancing research [5]. Skin model concept has been applied to describe tolerancing representation and analysis [6]. Standards play an important role in the adoption of many technologies. Some of standardized formats include Stereo Lithography (STL), Additive Manufacturing File (AFM) and Standard for Exchange Product (STEP) [7-8]. The theoretical geometrical model has been developed to consider geometric deviations [9]. And the developing and using standards (i.e. ASTM AMF, 3MF, ISO 10303 STEP) and tolerances (ASME Y14) [10-11] have integrated information related to AM process [12]. In order to realise this AM process: design, simulation, build plan, monitoring and control and verification, standard organizations have begun to establish AM file format for data exchange[4]. The design and manufacturing of high performance product may lead to an inevitable need for geometric deviations and tolerancing management [13-14]. The geometric tolerancing models have been proposed [15-18]. Tolerancing analysis has been used in predicting inevitable geometric deviations on the function and quality of product [19]. Skin model shapes can extract valuable information for geometric deviation [20]. Product manufacturing information (PMI) includes the geometric dimensioning and tolerancing (GD&T) [21]. These PMI standards, such as ISO TC213, ASME Y14.5, can define tolerancing [22]. Next sections will highlight challenges of geometric models and standards for AM technologies. These challenges mainly focus on two issues: geometric deviations and tolerancing management, and refer to two criteria: tolerancing and compatibility with PMI for AM formats. In section 3, skin model is introduced, and PMI standard can define tolerancing in AM systems. In section 4, different AM standards (i.e. STL, AMF, STEP) are compared and we emphasize on STEP in future works. In section 5, a conclusion and outlook are given.

Colloque avec actes et comité de lecture. Internationale.; International audience; L'article propose un nouveau concept pour assurer le retour d'information depuis les Machines-Outils à Commandes Numériques vers les systèmes FAO et la simulation d'usinage. Le but principal de la proposition est de gérer les connaissances d'usinage extraites de la machine pour les réutiliser au niveau de la simulation d'usinage et ainsi fournir une aide à la décision pour la réalisation de futur programme d'usinage. Pour assurer la capitalisation des connaissances d'usinage, la proposition s'appuie sur une structure basée sur les entités d'usinage. Bien que la structure du standard STEP-NC permette l'exploitation d'une telle structure, son utilisation est impossible du fait de sa faible implémentation. C'est pourquoi la proposition se base sur la chaîne numérique actuelle et propose une reconnaissance des entités d'usinage directement depuis les fichiers STL extrait de la simulation d'usinage et propose une conversion des programmes de Code-G en STEP-NC. Cela permet alors d'interagir avec l'implémentation d'OntoSTEP-NC, une ontologie basée sur STEP-NC, comme base de connaissance.

International audience

International audience; Recent evolutions of computer-aided product development and massive integration of numerical simulations to the design process require new methodologies to decrease the computational costs. Numerical design of experiments is used to increase quality of products by taking into account uncertainties in product development. But, this method can be time-consuming and involves a high computational cost. This paper presents a literature review of design of experiments methodology to identify potential improvements for simulation process shortening. By means of metamodelling and adaptive design of experiments, the computational cost of each simulation and the number of required experiment can be reduced. However, methods involved are still time-consuming to be set up.

International audience

Colloque avec actes et comité de lecture. Internationale.; International audience; Dans l'industrie, le temps alloué au processus de simulation est souvent concentré sur la phase de prétraitement. Afin de diminuer le temps consacré à cette phase, il est important de réutiliser autant que possible les modèles numériques. Dans un contexte d'entreprise étendue et de gestion du cycle de vie des produits (PLM), ces modèles proviennent de différents partenaires et sont créés à partir d'outils généralement hétérogènes. Dans un but de réutilisation, il est essentiel de capitaliser les données et informations de simulation dans un format neutre, structuré, commun et standardisé. Basé sur une combinaison de l'approche de la gestion du cycle de vie des simulations (SLM) et de la méthodologie Vérification & Validation (V&V), ce travail de recherche propose une démarche favorable à la réutilisation des modèles de simulation numérique.

International audience; Product Lifecycle Management (PLM) aims at an integrated management of all product-related information and processes through the entire lifecycle. Within an extended enterprise, systems interact in order to share data and information throughout a product lifecycle. Beyond this computing point of view, individuals may be spread throughout the world and that is the reason why the same information on the same product shared through the same PLM system within the same extended enterprise may lead to different interpretations. Individual cognition, knowledge and particularly tacit knowledge have to be considered during the design of such computing systems. The aim of this work is to outline a semantic interoperability between a collaborative platform and a PLM system. Such interoperability allows individuals to construct a shared understanding, supporting tacit knowledge sharing, i.e. knowledge that cannot be made explicit. Our approach leads then to focus on the impact of individual cognition on PLM systems.

6 Pages; National audience; De nos jours, l'interopérabilité des systèmes est primordiale pour assurer une stratégie de gestion du cycle de vie du produit (PLM). Cependant il existe encore certaines barrières à l'interopérabilité à la fois techniques et scientifiques qui inhibent l'échange entre les différents systèmes d'informations. Dans un premier temps cet article défini l'interopérabilité de manière générique et identifie les verrous à lever. Il présente ensuite un état des lieux des solutions développés pour réaliser l'interopérabilité entre les phases de conception, de fabrication et d'assemblage. Enfin des perspectives quant à l'interopérabilité de la chaîne numérique conception/industrialisation sont proposées.

International audience; Product Lifecycle Management (PLM) is one way to improve productivity in all manufacturing companies and this is truer in mechanical Small and Medium Enterprises (SMEs). Nevertheless the implementation of a PLM system is not an easy thing for such companies. SMEs have lots of difficulties to create the right information model in order to improve design efficiency. In this paper we propose a PLM model that could help SMEs of mechanical engineering field. Based on a literature review, on the analysis of the mechanical SMEs needs that would favour efficiency during design process, a new PLM model for mechanical SMEs is proposed.

International audience; PLM is today a reality for mechanical SMEs. Some companies implement PLM systems very well but others have more difficulties. This paper aims to explain why some SMEs do not success to integrated PLM systems analyzing the needs of mechanical SMEs, the processes to implement to respond to those needs and the actual PLM software functionalities. The proposition of a typology of those companies and the responses of those needs by PLM processes will be explain through the applications of a demonstrator applying appropriate generic data model and modelling framework.

In this thesis, we are interested in engineering information management on the whole product lifecycle and through the extended enterprise. The application area is the implementation of information system support to PLM in mechanical SME. The goal is the improvement the PLM implementation by two contributions, on one hand a modeling method and on the other hand a meta-model. With this intention we propose an inductive enterprise modeling method based on the needs. This method takes into account the enterprise needs, the process to realize those needs and the data model to be implemented. It was used into tree different early adopters SME from three distinct domains of mechanical industry. A framework has been designed to make easier the implementation of this method. It includes a map of needs, a product development process model and a data generic model. Each of these deliverables get three levels of genericity: a generic level, encompassing all the enterprise objects, a partial level, containing the objects of one activity domain, and the particular level, containing the objects of one specific enterprise. The main innovation of this study is the inductive aspect of the modeling method that allows to explicit implicit needs, and the specialization of the generic model that proposes an enterprise specific model with the guarantee of interoperability. The application of this methodology through an information system that supports our approach has been tested on different use cases.; Ce travail s‟intéresse à la gestion d‟informations techniques sur l‟ensemble du cycle de vie du produit et à travers l‟ensemble de l‟entreprise étendue. Il a pour champ d‟application le domaine des systèmes d‟informations supports au PLM dans les PME mécaniciennes. L‟objectif est de faciliter la mise en œuvre de système PLM par l‟apport d‟une part d‟une méthode de modélisation et d‟autre part d‟un cadre de modélisation. Pour ce faire, nous avons mis en place une méthode inductive de modélisation d‟entreprise basée sur les besoins. Cette méthode prend en compte les besoins de l‟entreprise, les processus à mettre en œuvre pour réaliser ces besoins et le modèle de données à implémenter. Elle a été utilisée dans trois entreprises pilotes correspondant à trois domaines distincts de l‟industrie mécanique chez les PME. Un cadre de modélisation a été conçu pour faciliter la mise en œuvre de cette méthode. Il inclut une carte des besoins PLM, un modèle de processus de développement produit et un modèle de données générique. Chacun de ces livrables a trois niveaux de généricité distincts : un niveau générique, englobant tous les objets d‟entreprise, un niveau partiel, regroupant les objets d‟un domaine d‟activité donné, et le niveau particulier, contenant les objets spécifiques à une entreprise particulière. L‟apport principal de cette étude réside dans l‟aspect inductif de la méthode de modélisation qui permet d‟expliciter les besoins implicites, et le cadre de modélisation proposant une spécialisation poussée du modèle de données, tout en en garantissant l‟interopérabilité. L‟application de cette méthodologie au travers d‟un système d‟informations supportant notre approche a été testée dans différents cas d‟études.