Denna rapport beskriver vårt projektarbete på ComOpt i Helsingborg. Uppgiften har varit att göra presentationer av företagets produkter. Det centrala i arbetet har varit grafisk design, i det här fallet gränssnittsdesign till de presentationer vi har gjort. I vår rapport reflekterar vi över den designprocess vi har genomgått och ställer frågor som: Vad gör en designer till en bra designer? Och: Var får vi kreativitet och inspiration ifrån? Det vi har kommit fram till är att en bra designer kännetecknas av förmågan att kunna prestera bra trots bristande resurser. En bra designer är kreativ och innovativ. Denna kreativitet åstadkoms genom att designern känner sig själv på ett bra sätt och vet hur han eller hon ska få inspiration. Vi har kommit fram till att mycket handlar om att strunta i det självklara och få inspiration i sin omgivning. En inspirerande omgivning ger ofta upphov till många bra idéer. Under vårt arbete har vi använt oss av metoder såsom intervjuer, feedbackmöten, mock-up och storyboard med flera. Detta har hjälpt oss att sätta oss in i produkterna, identifiera kundgrupp och designa presentationer våra klienter har varit nöjda med. Abstract This report is about our work at the company ComOpt in Helsingborg. Our task there has been to make presentations of the company´s products. The core in our work has been graphic design, in this case on the interfaces of the presentations we have been making. In our report we are reflecting about the designprocess we have been in and asking questions such as: What is it that makes a good designer? And: Were do we get our creativity and innovation from? Our conclusion is that a good designer is able to produce good products even with the lack of important resources. A good designer is creative and innovative. The designer knows him or herself and knows how to get inspired and thereby get great ideas. We have noticed that we get these ideas by ignoring the obvious and receiving inspiration through our surroundings. During our work we have been using methods as interviews, feedbackmeetings, mock-ups and storyboards and etcetera. This methods has helped us getting to know the products, identify the customergroup and designing presentations that has been to the satisfaction of our clients.

Millions of people suffer from heart failure worldwide. The need for heart donations started the development of mechanical circulatory support systems. A suction phenomenon can occur in the artificial heart when not enough blood is available. Due to occlusion, suction in the artificial heart can cause the arteries to collapse and have fatal consequences. This thesis is in collaboration with Scandinavian Real Heart AB and follows the implementation of a preemptive atrial suction detection algorithm for the total artificial heart (TAH) developed by Realheart. The main limitation is the number of sensors available to collect data from, restricted to a pressure and current sensor. The data used in the thesis is collected on a mock loop that simulates the pressures in the human body. The implementation follows an iterative process where different Artificial Intelligence algorithms are tested and evaluated. The final algorithm uses a recurrent neural network (RNN) for classification and is evaluated based on the accuracy and the number of seconds before suction occurs. The results show that the RNN can preemptively classify the data one second before it occurs. The algorithm assumes suction to happen one second before it occurs, preemptively detecting suction. The results from this thesis enable a continuation that can improve the development of TAHs. Future work includes an addition of features for a more accurate and robust algorithm, a more diverse dataset, an improved labelling process and the addition of a time axis to the RNN to improve the time before suction is detected.

Os dispositivos de assistência ventricular (DAVs) podem ser utilizados para a estabilização hemodinâmica de pacientes à espera do transplante cardíaco. Os avanços nas tecnologias e a utilização de materiais biocompatíveis vem contribuindo para o desenvolvimento de dispositivos com dimensões reduzidas e menor trauma ao sangue. A avaliação do desempenho desses dispositivos demanda a utilização de simuladores hidráulicos do sistema circulatório que reproduzam as pressões e fluxos existentes nas condições fisiológicas de interesse. Este trabalho tem como objetivo o desenvolvimento de um simulador da circulação pediátrica com ajustes automatizados de pressões. O simulador é composto por um circuito hidráulico modelando os laços sistêmico e pulmonar e um sistema microcontrolado com uma interface de usuário para medição e visualização dos fluxos e pressões ventriculares e automatização dos ajustes das pressões arteriais aórtica e pulmonar (PAo, PAP) e das pressões atriais esquerda e direita (PAE, PAD). Duas bombas pulsáteis com 15 ml de volume de ejeção são utilizadas para modelar os comportamentos mecânicos dos ventrículos esquerdo e direito. As complacências da aorta e da artéria pulmonar e as pré-cargas dos ventrículos são simuladas por câmaras com volumes ajustáveis de ar e líquido (análogo sanguíneo) utilizando uma bomba de ar. As resistências hidráulicas dos laços são ajustadas por oclusores motorizados. Os sinais instantâneos dos fluxos de entrada e saída dos DAVs e das pressões arteriais, atriais e ventriculares são obtidos por transdutores e digitalizados em um microcontrolador que comanda os oclusores e a bomba de ar. Foram desenvolvidos algoritmos para ajustes das resistências, complacências e pré-cargas. Uma interface gráfica de usuário apresenta os sinais em tempo real (ou gravados) permitindo a escolha dos parâmetros e condições de simulação. O desempenho do sistema de automatização foi avaliado nas simulações de: 1) condições definidas pelo aplicativo da interface e 2) condições fisiológicas (normal e redução na contratilidade do miocárdio). No modelo hidráulico sistêmico as pressões foram ajustadas em ambas as situações com erro máximo de 0,5% para a PAo e 5% para a PAE em aproximadamente 80 segundos. No modelo completo da circulação o erro máximo para as simulações de condições fisiológicas foi de 4% para as pressões arteriais e 5% para as atriais. Os resultados obtidos demonstram que o simulador desenvolvido permite reproduzir adequadamente as características da circulação pediátrica essenciais para a avaliação do desempenho de dispositivos de assistência mecânica. O simulador é portátil, de fácil utilização e pode ser utilizado como ferramenta didática ou para o treinamento de profissionais da saúde envolvidos em assistência a pacientes com suporte circulatório. / Ventricular assist devices (VADs) can be used for the hemodynamic stabilization of patients waiting for heart transplantation. Advances in the technologies and the use of biocompatible materials have contributed to the development of devices with reduced dimensions and blood trauma. Evaluation of the performance of these devices demands the use of hydraulic simulators of the circulatory system that reproduce pressures and flows existing in physiological conditions of interest. This work aims to develop a simulator of the pediatric circulation with automated adjustments of pressures. The simulator consists of a hydraulic circuit modeling the systemic and pulmonary branches and a microcontrolled system with a user interface for monitoring flows and ventricular pressures, and automating adjustments of aortic and pulmonary arterial pressures (AoP, PAP) and left and right atrial pressures (LAP, RAP). Two pulsatile pumps with 15 ml ejection volume are used to model the mechanical behavior of the left and right ventricles. Aortic and pulmonary arterial compliances and preloads of the ventricles are simulated by chambers with adjustable volumes of air and fluid (blood analog) using an air pump. Motorized clamps adjust the hydraulic resistances of the loops. Instantaneous signals of VAD input and output flows and of arterial, atrial and ventricular pressures are obtained by transducers and digitalized in a microcontroller that commands the clamps and the air pump. Algorithms were developed to adjust the resistances, compliances and preloads. A graphical user interface displays signals in real time (or recorded) and allows selection of simulation parameters. The performance of the automation system was tested setting pressures in two situations: 1) simulations of random conditions defined by the interface software and 2) simulations of physiological conditions (normal and low myocardial contractility). In the systemic model, the pressures were adjusted in both cases (maximum error of 0.5% for AoP and 5% for LAP) in approximately 80 seconds. In the complete model, the maximum error for simulations of physiological conditions was 4% for arterial pressures and 5% for atrial pressures. The results obtained demonstrate that the simulator developed allows mimicking the relevant features of the systemic and pulmonary branches of the circulation needed to assess the performance of mechanical circulatory assist devices. The simulator is portable, easy to operate and can be applied in teaching and training of health professionals working with mechanical circulatory support of patients.

Os dispositivos de assistência ventricular (DAVs) podem ser utilizados para a estabilização hemodinâmica de pacientes à espera do transplante cardíaco. Os avanços nas tecnologias e a utilização de materiais biocompatíveis vem contribuindo para o desenvolvimento de dispositivos com dimensões reduzidas e menor trauma ao sangue. A avaliação do desempenho desses dispositivos demanda a utilização de simuladores hidráulicos do sistema circulatório que reproduzam as pressões e fluxos existentes nas condições fisiológicas de interesse. Este trabalho tem como objetivo o desenvolvimento de um simulador da circulação pediátrica com ajustes automatizados de pressões. O simulador é composto por um circuito hidráulico modelando os laços sistêmico e pulmonar e um sistema microcontrolado com uma interface de usuário para medição e visualização dos fluxos e pressões ventriculares e automatização dos ajustes das pressões arteriais aórtica e pulmonar (PAo, PAP) e das pressões atriais esquerda e direita (PAE, PAD). Duas bombas pulsáteis com 15 ml de volume de ejeção são utilizadas para modelar os comportamentos mecânicos dos ventrículos esquerdo e direito. As complacências da aorta e da artéria pulmonar e as pré-cargas dos ventrículos são simuladas por câmaras com volumes ajustáveis de ar e líquido (análogo sanguíneo) utilizando uma bomba de ar. As resistências hidráulicas dos laços são ajustadas por oclusores motorizados. Os sinais instantâneos dos fluxos de entrada e saída dos DAVs e das pressões arteriais, atriais e ventriculares são obtidos por transdutores e digitalizados em um microcontrolador que comanda os oclusores e a bomba de ar. Foram desenvolvidos algoritmos para ajustes das resistências, complacências e pré-cargas. Uma interface gráfica de usuário apresenta os sinais em tempo real (ou gravados) permitindo a escolha dos parâmetros e condições de simulação. O desempenho do sistema de automatização foi avaliado nas simulações de: 1) condições definidas pelo aplicativo da interface e 2) condições fisiológicas (normal e redução na contratilidade do miocárdio). No modelo hidráulico sistêmico as pressões foram ajustadas em ambas as situações com erro máximo de 0,5% para a PAo e 5% para a PAE em aproximadamente 80 segundos. No modelo completo da circulação o erro máximo para as simulações de condições fisiológicas foi de 4% para as pressões arteriais e 5% para as atriais. Os resultados obtidos demonstram que o simulador desenvolvido permite reproduzir adequadamente as características da circulação pediátrica essenciais para a avaliação do desempenho de dispositivos de assistência mecânica. O simulador é portátil, de fácil utilização e pode ser utilizado como ferramenta didática ou para o treinamento de profissionais da saúde envolvidos em assistência a pacientes com suporte circulatório. / Ventricular assist devices (VADs) can be used for the hemodynamic stabilization of patients waiting for heart transplantation. Advances in the technologies and the use of biocompatible materials have contributed to the development of devices with reduced dimensions and blood trauma. Evaluation of the performance of these devices demands the use of hydraulic simulators of the circulatory system that reproduce pressures and flows existing in physiological conditions of interest. This work aims to develop a simulator of the pediatric circulation with automated adjustments of pressures. The simulator consists of a hydraulic circuit modeling the systemic and pulmonary branches and a microcontrolled system with a user interface for monitoring flows and ventricular pressures, and automating adjustments of aortic and pulmonary arterial pressures (AoP, PAP) and left and right atrial pressures (LAP, RAP). Two pulsatile pumps with 15 ml ejection volume are used to model the mechanical behavior of the left and right ventricles. Aortic and pulmonary arterial compliances and preloads of the ventricles are simulated by chambers with adjustable volumes of air and fluid (blood analog) using an air pump. Motorized clamps adjust the hydraulic resistances of the loops. Instantaneous signals of VAD input and output flows and of arterial, atrial and ventricular pressures are obtained by transducers and digitalized in a microcontroller that commands the clamps and the air pump. Algorithms were developed to adjust the resistances, compliances and preloads. A graphical user interface displays signals in real time (or recorded) and allows selection of simulation parameters. The performance of the automation system was tested setting pressures in two situations: 1) simulations of random conditions defined by the interface software and 2) simulations of physiological conditions (normal and low myocardial contractility). In the systemic model, the pressures were adjusted in both cases (maximum error of 0.5% for AoP and 5% for LAP) in approximately 80 seconds. In the complete model, the maximum error for simulations of physiological conditions was 4% for arterial pressures and 5% for atrial pressures. The results obtained demonstrate that the simulator developed allows mimicking the relevant features of the systemic and pulmonary branches of the circulation needed to assess the performance of mechanical circulatory assist devices. The simulator is portable, easy to operate and can be applied in teaching and training of health professionals working with mechanical circulatory support of patients.

Accumulating evidence suggests that the magnetic resonance imaging (MRI) scanner can act as a stressor, eliciting subjective and neuroendocrine stress responses. Approaches to familiarize subjects with the scanner could help minimizing unintended effects on neural activation patterns of interest. Controlled studies on the effects of a scanner training are however missing. Using a comparative design, we analyzed within- and between session changes in subjective and neuroendocrine stress parameters in 63 healthy, scanner-naïve adults who participated in a two-day training protocol in an MRI, mock, or lab environment. A habituation task was used to assess within-session changes in subjective and neuroendocrine (cortisol) stress parameters; between-session changes were indicated by differences between days. MRI and mock, but not lab training were successful in reducing subjective distress towards the scanner. In contrast, cortisol reactivity towards the training environment generally increased during day 2, and the percentage of cortisol responders particularly rose in the mock and MRI groups. Within-session habituation of subjective arousal and anxiety was observed during both days and irrespective of training condition. Present findings demonstrate that training in a scanner environment successfully reduces subjective distress, but may also induce sensitization of endocrine stress levels during repeated scanning. Subjective distress can further be stabilized by acclimating subjects to the environment prior to the MRI assessment, including a short habituation phase into the assessment protocol. If replicated, present findings should be considered by researchers employing repeated measurement designs where subjects are exposed to a scanner more than once.

Cette thèse porte sur le comportement hydromécanique d’un mélange de poudre et pellets de bentonite MX80 avec une proportion 80/20 en masse sèche. Il s’agit d’un matériau étudié par l’Institut de Radioprotection et de Sûreté Nucléaire (IRSN) dans le cadre du projet SEALEX qui a pour objectif principal la vérification de l’efficacité des dispositifs de scellement ou des barrières ouvragées dans le system du stockage géologique des déchets radioactifs. Le comportement hydromécanique du matériau à différentes échelles a été étudié par différents essais en laboratoire. Premièrement, les changements à l’échelle microstructurale d’un seul pellet de bentonite durant l’hydratation a été abordée à l’aide de deux techniques : la porosimétrie au mercure et la tomographie aux rayons-X. Les résultats ont montré que le gonflement d’un pellet peut être expliqué par deux mécanismes : la création des fissures surtout à des succions entre 38 et 9 MPa, et le gonflement des grains de bentonite, correspondant à l’hydratation des smectites à l’échelle nano. A des succions inférieures à 9 MPa, une diminution de l’épaisseur des feuillets d’argile et une augmentation du désordre des ceux-ci sont observées. Des essais de rétention d’eau, de pression de gonflement et de compression à l’odomètre à succion contrôlée ont été effectués sur le mélange de poudre et pellets. Les propriétés de rétention d’eau sous conditions de volume constant et pour un seul pellet sous conditions de gonflement libre apparaissent similaires pour des valeurs de succion supérieures à 4 MPa. Cela implique que la succion physico-chimique est prédominante devant la succion capillaire. Pour des valeurs de succions plus basses, une capacité de rétention plus faible a été observée sous conditions de volume constant, à relier à la disparition des macro-pores par le gonflement des grains de bentonite. Des valeurs de pression de préconsolidation plus petites que celles des mélanges de bentonite pure ont été obtenues pour des succions non-nulles, montrant l’effet granulaire des pellets dans le mélange. Deux colonnes d’infiltration ont été réalisés afin d’étudier deux cas extrêmes avec une densité sèche globale identique (1.49 Mg/m3). Avec la première colonne, un mélange de poudre et pellets relativement homogène, fabriqué en suivant un protocole spécial a été étudié. En revanche, un mélange fortement hétérogène a été fabriqué dans la deuxième colonne d’infiltration. Les résultats montrent que la pression de gonflement radiale dépend fortement de la distribution des pellets et de la poudre ainsi que de l’évolution du front d’hydratation. Une anisotropie de gonflement a été observée dans les deux cas, avec la pression de gonflement axiale inférieure à celle radiale. De plus, la valeur finale de pression de gonflement axiale est différente pour les deux colonnes, bien que les deux échantillons aient fabriquées avec la même densité sèche globale. En parallèle, plusieurs observations à la tomographie aux rayons-X ont été réalisées sur le mélange de pellets et poudre pendant l’hydratation. Un mélange complètement homogène a été observé après 100 jours d’hydratation à l’échelle étudiée (50 μm/voxel). Un nouveau modèle d’endommagement qui prend en compte des fissures observées au sein du pellet pendant hydratation a été développé en adaptant le Barcelona Expansive Model (BExM). L’essai d’infiltration sur l’échantillon relativement homogène a été simulé avec succès en utilisant le modèle développé. L’hétérogénéité initiale de la porosité a été aussi considérée dans la simulation afin de reproduire l’anisotropie de gonflement. Les résultats expérimentaux obtenus dans le cadre de cette étude permettent de mieux comprendre la réponse des ouvrages de scellement avec le mélange de pellets et poudre de bentonite dans le projet SEALEX. De plus, le modèle développé, qui prend en compte des fissures observées au sein du pellet et l’hétérogénéité initiale du matériau, permettra d’améliorer / The present investigation deals with the hydro-mechanical behavior of a mixture composed of pellets and powder of MX80 bentonite with a proportion of 80/20 in dry mass. This is one of the studied materials by the French Institute for Radiation protection and Nuclear Safety (IRSN) within the SEALEX project, which aims at investigating the long-term performance of swelling clay-based sealing systems in the context of geological high-level radioactive waste disposal. This study has been conducted by following an experimental program covering different scales. Firstly, the microstructure changes while wetting of a single pellet was investigated by combining MIP results with μ-CT observations. Results revealed that swelling of a pellet is due to the development of cracks, with significant development between 38 and 9 MPa of suction, combined to swelling of bentonite grains, which is governed by the hydration mechanisms of smectite at nano-scale. The application of suctions below 9 MPa leads to a significant decrease of the platelet thickness and to an increase in the disorder of the platelet assembly. Water retention tests, swelling pressure tests and suction controlled oedometer tests on the pellet/powder mixture were performed. Similar water retention properties were observed for the mixture under constant-volume condition and pellet under free swelling condition under suctions higher than 4 MPa, suggesting that physico-chemical suction prevails on capillary suction. At lower suctions, constant-volume condition defined a lower water retention capacity because of the disappearance of macro-pores. Lower yield stress values than the common pure bentonite mixtures were found for the pellet/powder mixture for non-zero suctions, showing that the volume change behavior is governed by the rearrangement and crushing of pellets, and the loss of the granular structure in the case of zero suction. Two mock-up tests were performed, aiming at studying two extreme cases at a global dry density of 1.49 Mg/m3: a homogeneous pellet/powder mixture fabricated by following a special protocol, and a strong heterogeneous sample. Results revealed that the radial swelling pressure depends strongly on the local pellet/powder distribution combined with the evolution of the hydration front. An anisotropy swelling was found in both cases, being the axial swelling pressure lower than the radial one. Moreover, different values of axial pressure were found between the two tests, even though they have the same global dry density of samples. In parallel, μ-CT observations were carried out on the mixture while wetting, revealing a homogeneous sealed sample after 100 days of hydration. No density gradients were identified at the investigated resolution (50 μm/voxel) after this long time of hydration. A new damage model, which takes into account the development of fissures within a pellet while wetting, was proposed an included to the well-IVknown double porosity Barcelona Expansive Model (BExM) to carry out numerical simulations of one mock-up test. The initial heterogeneous porosity distribution was also considered to reproduce the anisotropy swelling. The experimental results obtained in this study will greatly help well understand the response of seals made up of pellets/powder bentonite mixture in the SEALEX in situ experiment. Moreover, the constitutive model developed taking into account the pellet cracking damage and the initial sample heterogeneity allows significantly improving the prediction of hydomechanical behavior of seals/plugs made up of this mixture, constituting thus an useful tool for the safety assessment of the nuclear waste disposal system

La question de l’incertitude dans la représentation de l’architecture n’est pas nouvelle. Des figurations anciennes montrent des différences de traitement des structures existantes et des éléments disparus. L'émergence des outils numériques et l’utilisation de la 3D en archéologie du bâti depuis 1990 ont offert aux archéologues de nouvelles possibilités de représentation de l’incertitude. Mais les restitutions 3D n’expriment pas toujours l’authenticité historique. Des règles ont été écrites, fondées sur la transparence et la traçabilité du processus d’étude et de construction de la représentation 3D. Les recherches dans ce domaine ont porté sur l'expression et la perception de l’incertitude avec, notamment, le développement de dispositifs visuels qui traduisent l’état des connaissances. Qu’il s’agisse de systèmes d’informations liant représentations 3D et sources documentaires ou de l’application de variables visuelles aux représentations, les propositions apportent des réponses étudiées dans ce travail de thèse. Certaines approches ont été appliquées à la chartreuse de Villeneuve lez Avignon (Gard) et l’abbaye Saint-Michel de Cuxa (Pyrénées-Orientales). A partir des sources documentaires, études archéologiques et relevés numériques, des modèles 3D ont été construits. Nos recherches nous ont poussées à nous interroger sur les incertitudes de ces informations et sur les moyens de les visualiser. Les approches et méthodologies mises en place dans cette thèse soulignent le caractère pluridisciplinaire du travail, mêlant connaissances archéologiques et exploitation des instruments numériques pour l’acquisition, le traitement et la visualisation des données en 3D. / The item of uncertainty into the architecture representation is not new. The ancient figurations showed differences in visual processing between existing structures and missing elements. Since the 1990’s, the emergence of digital tools and the use of 3D in virtual archaeology have offered to the archaeologists new possibilities to represent the uncertainty. But the 3D restitution is not always express the historical authenticity. This is in this context that rules were written, based on the transparency and the traceability of the process of study and the construction the three-dimensional representation. Researches in the domain of the uncertainty representation have concerned the modalities of expression and perception of the uncertainty with, especially, the development of devices able to translate the state of the knowledge. Whether information systems linking 3D representations and documentary sources or application of visual variables, the different propositions on the definition of uncertainty and its representation modes allowed to experiment some of these approaches to the charterhouse of Villeneuve lez Avignon (Gard, France) and the benedictin abbey of Saint-Michel-de-Cuxa (Pyrénées-Orientales, France). From the documentary sources, archaeological studies and digital data, 3D models have been built. But our researches led us to think about the uncertainties of those information and on ways to visualize them.The approaches and methodologies produced in this thesis underline the multidisciplinary of the work, mixing archaeological knowledge and the exploitation of the digital instruments for the acquisition, the treatment and the visualization of data in 3D.

Il y a peu, j’ai pu lire « qu’on pouvait considérer que les ponts romains de l’Antiquité, pouvaient être considérés comme inefficaces, au regard des standards actuels : ils utilisaient trop de pierre et énormément de travail était nécessaire à leur construction. Au fil des années, pour répondre à une problématique équivalente, nous avons appris à utiliser moins de matériaux et à réduire la charge de travail ». Ces problématiques nous les retrouvons aussi en conception mécanique où l’on essaye en continu de proposer des systèmes de plus en plus performants mais devant être conçus en moins de temps, étant moins cher à produire et fournissant des prestations au moins équivalentes à ce qui a déjà été conçu.Au cours d'un processus de conception classique, les concepteurs définissent une géométrie ne présentant aucun défaut puis, étant donné que les moyens de production ne permettent pas d’obtenir de telles pièces finales, ils spécifient les schémas de cotation définissant les écarts acceptables garantissant le bon fonctionnement du système. Seulement, cela est fait après avoir produit les dessins détaillés, c'est à dire trop tard. Pour répondre à cette problématique, je présenterai l’intégration, très tôt dans le cycle de vie de conception, d’un processus de validation optimisé, basé sur une maquette numérique directement en lien avec sa représentation fonctionnelle (maquette fonctionnelle), et permettant de valider des schémas de cotation 3D standardisés.Je décrirai d'abord ce que l’on entend par « maquette fonctionnelle » et surtout ce que cette nouvelle définition apporte en plus de la définition numérique. Une fois ce point abordé, je détaillerai les liens qui permettent d’avoir une unicité de l’information au sein de l’environnement de travail, tout comme les processus qui permettent de lier les représentations fonctionnelles et numériques.Ensuite, je détaillerai les processus basés sur ces concepts, et qui ont pour but de valider les choix qui sont effectués en avance de phase au niveau des schémas de cotation. Pour ce faire, je commencerai par présenter l’analyse au pire des cas (utilisant les modèles de domaines écarts notamment), permettant de garantir le bon fonctionnement de l’ensemble mécanique, dans le cas ou touts les écarts se retrouvent à l’intérieur des zones respectives (définies par les tolérances).Enfin, je finirai par introduire ce qu’une couche statistique, couplée à l’analyse au pire des cas utilisant les enveloppes convexes, peut amener dans le contexte industriel et notamment sous la contrainte temporelle. / Some time ago, I read "According to our current standards, we could consider Roman bridges of ancient times as ineffective: they used too much stone and hard work during construction. Over the years, in order to respond to similar problems, we learned how to use fewer materials and reduce the workload. These issues can also be found in the mechanical design field, where we continuously try to offer more efficient systems, but which have to be designed in less time, be cheaper to produce and provide benefits at least equivalent to what has already been designed.During a conventional design process, designers define the ideal geometries and - given that the machining tools cannot produce mechanical parts without any geometrical defects - specify the associated tolerancing schemes. These tolerancing schemes define acceptable geometrical deviations, thus providing a well-functioning system. Unfortunately this is done after having designed detailed parts and thus, too late.In order to address this problem, I will begin by introducing the integration, in the first design stages, of a new optimized validation process based on a Digital Mock-Up, directly linked to its functional representation (Functional Mock-Up), in order to validate 3D standardized tolerancing schemes. I'll first describe what is meant by "Functional Mock-Up" (FMU) and specify which information is added to the Digital Mock-Up (DMU). Once that is done, I will detail the relationship that leads to the uniqueness of the information and the processes linking the Functional and Digital representations.Then, I'll detail the processes based on these concepts, which aim to validate the tolerancing schemes, during the early design stages. To do this, I'll begin by introducing the worst case analysis (using the deviation domain model), which ensures the proper functioning of the mechanical system. Finally, I will end this by introducing the benefits that can be brought, by coupling a statistical layer to the worst case analysis (using the convex hull).

L’objectif de cette thèse est de contribuer au développement d’approches méthodologiques et d’outils informatiques pour développer les chaînes d’intégration numériques en entreprise étendue. Il s’agit notamment de mieux intégrer et d’optimiser les activités de conception, d’intégration et de simulation dans le contexte du développement collaboratif des produits/systèmes complexes.La maquette numérique (DMU) – supportée par un système de gestion de données techniques (SGDT ou PDM) – est devenue ces dernières années un environnement fédérateur clé pour échanger et partager une définition technique et une représentation 3D commune du produit entre concepteurs et partenaires. Cela permet aux concepteurs ainsi qu’aux utilisateurs en aval (ceux qui sont en charge des simulations numériques notamment) d’avoir un accès à la géométrie du produit virtuel assemblé. Alors que les simulations numériques 3D et 2D prennent une place de plus en plus importante dans le cycle de développement du produit, la DMU offre de nouvelles perspectives à ces utilisateurs pour récupérer et exploiter les données CAO appropriées et adaptées pour les analyses par éléments finis. Cela peut ainsi permettre d’accélérer le processus de préparation du modèle de simulation. Cependant, les environnements industriels de maquettes numériques sont actuellement limités dans leur exploitation par : - un manque de flexibilité en termes de contenu et de structure, - l’absence d’artefact numérique 3D permettant de décrire les interfaces des composants de l’assemblage, - un manque d’intégration avec les données et activités de simulation.Cette thèse met notamment l’accent sur les transformations à apporter aux DMU afin qu’elles puissent être utilisées comme données d’entrée directes pour les analyses par éléments finis d’assemblages volumineux (plusieurs milliers de pièces). Ces transformations doivent être en cohérence avec le contexte et les objectifs de simulation et cela nous a amené au concept de « vue produit » appliquée aux DMUs, ainsi qu’au concept de « maquette comportementale » (BMU). Une « vue produit » définit le lien entre une représentation du produit et l’activité ou le processus utilisant ou générant cette représentation. La BMU est l’équivalent de la DMU pour les données et les processus de simulation. Au delà des géométries discrétisées, la dénommée BMU devrait, en principe, lier toutes les données et les modèles qui seront nécessaires pour simuler le comportement d’un ou plusieurs composants. L’élément clé pour atteindre l’objectif d’élargir le concept établi de la DMU (basée sur des modèles CAO) à celui de la BMU (basée sur des modèles CAE), est de trouver un concept d’interface bidirectionnel entre la BMU et sa DMU associée. C’est l’objectif du « Design-Analysis System Integration Framework » (DASIF) proposé dans cette thèse de doctorat. Ce cadre a vise à être implémenté au sein d’environnements PLM/SLM et doit pouvoir inter-opérer à la fois avec les environnements CAD-DMU et CAE-BMU. DASIF allie les fonctionnalités de gestion de données et de configuration des systèmes PDM avec les concepts et formalismes d’ingénierie système basée sur les modèles (MBSE) et des fonctionnalités de gestion des données de simulation (SDM). Cette thèse a été menée dans le cadre d’un projet de recherche européen : le projet CRESCENDO qui vise à développer le « Behavioural Digital Aircraft » (BDA) qui a pour vocation d’être la« colonne vertébrale » des activités de conception et simulation avancées en entreprise étendue. Le concept du BDA doit s’articuler autour d’une plateforme collaborative d’échange et de partage des données de conception et de simulation tout au long du cycle de développement et de vie des produits aéronautiques. [...] / The aim of this doctoral thesis is to contribute to the facilitation of design, integration and simulation activities in the aeronautics industry, but more generally in the context of collaborative complex product development. This objective is expected to be achieved through the use and improvement of digital engineering capabilities. During the last decade, the Digital Mock-Up (DMU) – supported by Product Data Management (PDM) systems – became a key federating environment to exchange/share a common 3D CAD model-based product definition between co-designers. It enables designers and downstream users(analysts) to access the geometry of the product assembly. While enhancing 3D and 2D simulations in a collaborative and distributed design process, the DMU offers new perspectives for analysts to retrieve the appropriate CAD data inputs used for Finite Element Analysis (FEA), permitting hence to speed-up the simulation preparation process. However, current industrial DMUs suffer from several limitations, such as the lack of flexibility in terms of content and structure, the lack of digital interface objects describing the relationships between its components and a lack of integration with simulation activities and data.This PhD underlines the DMU transformations required to provide adapted DMUs that can be used as direct input for large assembly FEA. These transformations must be consistent with the simulation context and objectives and lead to the concept of “Product View” applied to DMUs andto the concept of “Behavioural Mock-Up” (BMU). A product view defines the link between a product representation and the activity or process (performed by at least one stakeholder) that use or generate this representation as input or output respectively. The BMU is the equivalent of the DMU for simulation data and processes. Beyond the geometry, which is represented in the DMU,the so-called BMU should logically link all data and models that are required to simulate the physical behaviour and properties of a single component or an assembly of components. The key enabler for achieving the target of extending the concept of the established CAD-based DMU to the behavioural CAE-based BMU is to find a bi-directional interfacing concept between the BMU and its associated DMU. This the aim of the Design-Analysis System Integration Framework (DASIF) proposed in this PhD. This framework might be implemented within PLM/SLM environments and interoperate with both CAD-DMU and CAE-BMU environments. DASIF combines configuration data management capabilities of PDM systems with MBSE system modelling concepts and Simulation Data Management capabilities.This PhD has been carried out within a European research project: the CRESCENDO project, which aims at delivering the Behavioural Digital Aircraft (BDA). The BDA concept might consist in a collaborative data exchange/sharing platform for design-simulation processes and models throughout the development life cycle of aeronautics products. Within this project, the Product Integration Scenario and related methodology have been defined to handle digital integration chains and to provide a test case scenario for testing DASIF concepts. These latter have been used to specify and develop a prototype of an “Integrator Dedicated Environment” implemented in commercial PLM/SLM applications. Finally the DASIF conceptual data model has also served as input for contributing to the definition of the Behavioural Digital Aircraft Business Object Model: the standardized data model of the BDA platform enabling interoperability between heterogeneous PLM/SLM applications and to which existing local design environments and new services to be developed could plug.

Inom en stor del av all systemutveckling sker testerna av systemet som sista punkt innan systemet sjösätts. Testdriven utveckling är en systemutvecklingsmetod där testerna istället skrivs först och också är det som driver utvecklingen framåt. Metoden höjs till skyarna av vissa och avfärdas omedelbart som onödigt omständig av andra. Vi vill med denna uppsats undersöka hur det ser ut i verkligheten och vilka faktorer som påverkar användandet, inlärningen och inställningen till testdriven utveckling. Vi genomförde intervjuer på tre stycken Örebrobaserade organisationer och tittade utifrån ramverket method-in-action på vilka faktorer som påverkade användningen och varför. Vi fann att utvecklarna närmade sig testdriven utveckling på väldigt olika sätt och grundade sin inställning mycket beroende på tidigare erfarenhet och inlärning – oavsett hur lång eller kort den varit. Utvecklarna förväntas ofta bedriva självstudier utanför arbetstid – något som inte alltid funkar som kunskaputvecklingsform då tiden utanför jobbet ser olika ut beroende på var i livet man är. Det finns inte heller något klart program eller best-practices att följa för att lära sig metoden i någon av organisationerna. Vi såg också att det finns tekniker utanför själva metoden som utvecklare ganska omgående behöver bli bekanta med för att kunna utveckla testdrivet: dependency injection och mock.