Trafikverket har ansvaret för ett av Sveriges största infrastrukturprojekt, nämligen att ersätta det nuvarande signalsystemet längs landets järnvägar till ERTMS. Med bytet av signalsystemet kommer behovet av att byta ut 1 500 teknikhus. Syftet med arbetet är att genom en analys av kostnader för den nuvarande grundläggningsstandarden med U-betongplintar för teknikhus, försöka hitta billigare lösningar. Arbetet utförs främst genom en analys av alternativa material- och grundläggningstyper, men hänsyn tas även till tillverkning, installation, logistik, påverkande laster och geologi. Delmålet är att kunna sänka den nuvarande kostnaden för projektet och huvudmålet är att hitta en ny framtida standard för grundläggningen av Trafikverkets teknikhus. De vanligaste grundläggningstyperna för teknikhusen visade sig vara kryp- och plintgrunder, utav plåt respektive betong. I arbetet presenteras även markskruvar, som är en ny idé för teknikhusen. Slutsatsen av rapporten är att markskruvar är det billigaste och smidigaste grundläggningsvalet för teknikhus. Även plåtramar, som idag är ett mer vanligt förekommande alternativ för teknikhus, är billigare än den nuvarande standarden. Arbetets totala besparing för projektet är nära 34 miljoner kronor med markskruvar, och därför har delmålet uppnåtts. Huvudmålet om att hitta en ny standard kräver däremot framtida undersökningar av Trafikverket innan fastställning. En rekommendation för Trafikverket är att vidareundersöka priset för utländska leverantörers markskruvar och avgöra om installationen ska utföras på egen hand. För vidare studier rekommenderas en noggrann analys på transportkostnader för grundläggningar och hur stora kostnader de fribärande teknikhusen tillför. / Trafikverket is responsible for one of Sweden’s largest infrastructure project, namely replacing the current signal system along the countries railway to ERTMS. With the replacement of the signal system comes the need to replace 1 500 equipment shelters. The task is to analyze costs for the equipment shelter’s current foundation standard and try to find cheaper solutions. The work is mainly executed by analyzing alternative materials and foundations, but also manufacturing, installation, logistics, impacting loads and geology. One of the goals is to reduce the current costs of the project, and the main task is to find a new foundation standard that Trafikverket can use for future equipment shelters. The most common foundation types for equipment shelters turned out to be crawling spaces and plinth foundations out of sheet-metal and concrete. The report also presents ground screws, which is a new idea for equipment shelters. The report’s conclusion is that ground screws are the cheapest and most flexible foundation for equipment shelters. Even sheet-metal frames, which are a more common option today, proved to be a more economical choice than the current standard. The total saving turned out to be near 34 million SEK with ground screws and therefore one of the objectives has been achieved. The main goal of possibly finding a new standard, however, requires future investigations by Trafikverket before it can be determined possible. A recommendation to Trafikverket is to further investigate the price of foreign suppliers’ ground screws and determine if the installation is to be carried out on its own. For further studies, a detailed cost analysis of cantilevered houses and transports is recommended.

No description available.

Aluminium alloys have the possibility to be infinitely recycled. By only generating 5% of the emissions compared to primary aluminium, great CO2 savings can be made. One of the issues in manufacturing components entirely from post-consumer scrap is the presence of trace elements and impurities. Such elements can be Fe, Cu, Cr, P and Pb. In sheet metal manufacturing, these elements can also react with process agents such as Ti, B, Na and Sr and affect the recrystallization behavior and in turn mechanical properties.In this work, a derivative of the 3003 alloy made entirely from post-consumer scrap has been analysed. The alloy achieved insufficient formability due to lack of recrystallization and grain growth. With the use of scanning electron microscopy (SEM) equipped with energy dispersive spectroscopy (EDS) and electron backscatter diffraction (EBSD) together with focused ion beam (FIB) lamella preparation, the microstructure was characterized.The characterization shows both larger particles of α-Al15Si2M4 (M=Mn,Fe,Cr) from solidification and dispersoids from heat treatment, pinning the grain boundary movement together with Q-AlCuMgSi. With the use of high throughput computational thermodynamics, Thermo-Calc was used to effectively screen compositions lowering the amount of α-Al15Si2M4 and removing the Q-AlCuMgSi phase. The new alloy was cast using directional solidification at different cooling rates to study the particle morphology, which in turn plays a role in the particle break up and distribution during cold working as the interparticle spacing affects the grain growth.Varying cooling rates was seen to affect morphology and distribution. Hot compression was utilized to examine the particle redistribution before cold work. It was however shown that hot compression was not sufficient in redistributing the particles as would be the case in rolling.

Numerische Prozessanalysen werden heute standardmäßig zur virtuellen Prozessabsicherung der Herstellung umgeformter Blechformteile eingesetzt. Die dabei notwendige hohe Prognosegüte kann nur mit einer hinreichend präzisen Materialmodellierung realisiert werden. Ein wesentlicher Aspekt ist dabei die Modellierung des Fließverhaltens und der Werkstoffanisotropie. Zur Charakterisierung des Werkstoffverhaltens existieren zahlreiche Versuchs- und Auswertestrategien; diese werden aufgrund des hohen Versuchs- und Rechenaufwands in der industriellen Anwendung jedoch nur vereinzelt eingesetzt. Im Rahmen dieser Arbeit wird eine neuentwickelte halbanalytische Methode zur Charakterisierung von Fließortkurven sowie der Folgefließortkurven vorgestellt. Dazu wurde eine Spannungsanalyse auf Messdaten einer digitalen Bildkorrelation sowie ein Schnittlinienansatz zur Berechnung innerer Kräfte entwickelt. Durch Bilanzierung dieser inneren Schnittkräfte mit gemessenen äußeren Prüfkräften werden Parameter eines gewählten Materialmodells in einer inversen Analyse identifiziert. Die Methode zur Parameteridentifikation wurde zunächst an synthetischen und anschließend an realen Experimenten untersucht. Es wurde eine sequentielle Identifikationsstrategie aufgestellt, die an den Blechwerkstoffen DX54, DP600 sowie AA5182 untersucht wurde. Abschließend wurde die Methode mithilfe von Validierungsversuchen hinsichtlich der identifizierten Materialantworten überprüft.:Inhaltsverzeichnis I Symbolverzeichnis IV Abkürzungen und Begriffe VIII 1 Einleitung 1 2 Stand der Technik 3 2.1 Grundlagen zur Umformung von Blechwerkstoffen 3 2.1.1 Umformprozesse zur Herstellung von Blechformteilen 3 2.1.2 Anforderungen bei der Herstellung von Blechbauteilen 4 2.2 Numerische Prozessauslegung von Blechumformoperationen 5 2.2.1 Anwendung der FEM als Hilfsmittel in der Prozessauslegung 5 2.2.2 Kritische Punkte der numerischen Prozessauslegung 7 2.3 Werkstoffverhalten und Materialmodellierung 8 2.3.1 Kenngrößen zur Beschreibung großer Deformation 8 2.3.2 Grundlagen zur Beschreibung elastoplastischen Materialverhaltens 9 2.3.3 Beschreibung der Verfestigungsregel in Materialmodellen 11 2.3.4 Modellierung anisotroper Fließortkurven 12 2.3.5 Modellierung der Fließortkurvenentwicklung 18 2.3.6 Zusammenfassende Hinweise zu den Fließkriterien 25 2.4 Kennwertermittlung und Werkstoffcharakterisierung 26 2.4.1 Experimentelle Methoden und analytische Verfahren zur Fließortkurvenermittlung 26 2.4.2 Erweiterte Messtechnik zur Identifikation von Fließortkurven 33 2.4.3 Inverse Strategien zur Identifikation von Fließortkurven 36 2.5 Zusammenfassung zum Stand der Technik 43 3 Zielsetzung und Vorgehensweise 45 3.1 Zielsetzung 45 3.2 Vorgehensweise 45 4 Experimentelle Versuchsdurchführung 47 4.1 Versuchsaufbau und Messtechnik 47 4.2 Prüfverfahren zur Werkstoffprüfung 48 4.2.1 Zugversuch 49 4.2.2 Kerbzugversuch 49 4.2.3 Scherzugversuch 49 4.2.4 Biaxialer Zugversuch 50 4.3 Wahl der Versuchswerkstoffe 51 4.3.1 Kaltgewalzter Tiefziehstahl DX54 51 4.3.2 Kaltgewalzter Dualphasenstahl DP600 52 4.3.3 Aluminiumknetlegierung AA5182 52 4.4 Messergebnisse aus den Werkstoffprüfungen 53 4.5 Zusammenfassung zur Versuchsdurchführung 55 5 Entwicklung einer halbanalytischen Methode zur Parameteridentifikation 56 5.1 Spannungsanalyse 56 5.1.1 Spannungsrichtung 57 5.1.2 Spannungszuwachs 58 5.1.3 Elastische Kompensation 59 5.2 Schnittkraftermittlung 60 5.3 Inverse Analyse 62 5.4 Zusammenfassende Darstellung der Entwicklung 64 6 Validierung der Methode an virtuellen Experimenten 67 6.1 Virtuelle Versuche 67 6.2 Validierung der Spannungsanalyse 68 6.2.1 Vorgehensweise zur Prüfung der Spannungsanalyse 69 6.2.2 Ergebnisse der Überprüfung der Spannungsanalyse 70 6.3 Validierung der Schnittkraftermittlung 76 6.4 Sensitivität der Materialantwort gegenüber den Materialparametern 80 6.4.1 Einfluss variierender Fließkurvenapproximationen 80 6.4.2 Einfluss variierender Fließortkurven 82 6.5 Diskussion zur Wahl geeigneter Schnittlinien 88 6.6 Fehlerbetrachtung 89 / Numerical process analysis is widely used today for the virtual process validation of the production of formed sheet metal parts. In this context, sufficiently precise material modeling is essential, especially for the flow behavior and the material anisotropy. Numerous test and evaluation strategies are known for the material characterization. However, these strategies are only used occasionally in industrial applications due to high experimental and computational costs. In the context of this work, a newly developed semi-analytical method for the characterization of yield locus curves and subsequent yield locus curves is presented. For this purpose, a stress analysis based on digital image correlation data and a cutting-line approach for internal forces computation was developed. By balancing these internal cutting forces with measured external test forces, parameters of a material model are identified in an inverse analysis. The method for parameter identification was first examined on synthetic and subsequently on real experiments. A sequential identification strategy was set up and examined on the sheet metal materials DX54, DP600 and AA5182. Finally, the method was checked using validation tests with regard to the identified material responses.:Inhaltsverzeichnis I Symbolverzeichnis IV Abkürzungen und Begriffe VIII 1 Einleitung 1 2 Stand der Technik 3 2.1 Grundlagen zur Umformung von Blechwerkstoffen 3 2.1.1 Umformprozesse zur Herstellung von Blechformteilen 3 2.1.2 Anforderungen bei der Herstellung von Blechbauteilen 4 2.2 Numerische Prozessauslegung von Blechumformoperationen 5 2.2.1 Anwendung der FEM als Hilfsmittel in der Prozessauslegung 5 2.2.2 Kritische Punkte der numerischen Prozessauslegung 7 2.3 Werkstoffverhalten und Materialmodellierung 8 2.3.1 Kenngrößen zur Beschreibung großer Deformation 8 2.3.2 Grundlagen zur Beschreibung elastoplastischen Materialverhaltens 9 2.3.3 Beschreibung der Verfestigungsregel in Materialmodellen 11 2.3.4 Modellierung anisotroper Fließortkurven 12 2.3.5 Modellierung der Fließortkurvenentwicklung 18 2.3.6 Zusammenfassende Hinweise zu den Fließkriterien 25 2.4 Kennwertermittlung und Werkstoffcharakterisierung 26 2.4.1 Experimentelle Methoden und analytische Verfahren zur Fließortkurvenermittlung 26 2.4.2 Erweiterte Messtechnik zur Identifikation von Fließortkurven 33 2.4.3 Inverse Strategien zur Identifikation von Fließortkurven 36 2.5 Zusammenfassung zum Stand der Technik 43 3 Zielsetzung und Vorgehensweise 45 3.1 Zielsetzung 45 3.2 Vorgehensweise 45 4 Experimentelle Versuchsdurchführung 47 4.1 Versuchsaufbau und Messtechnik 47 4.2 Prüfverfahren zur Werkstoffprüfung 48 4.2.1 Zugversuch 49 4.2.2 Kerbzugversuch 49 4.2.3 Scherzugversuch 49 4.2.4 Biaxialer Zugversuch 50 4.3 Wahl der Versuchswerkstoffe 51 4.3.1 Kaltgewalzter Tiefziehstahl DX54 51 4.3.2 Kaltgewalzter Dualphasenstahl DP600 52 4.3.3 Aluminiumknetlegierung AA5182 52 4.4 Messergebnisse aus den Werkstoffprüfungen 53 4.5 Zusammenfassung zur Versuchsdurchführung 55 5 Entwicklung einer halbanalytischen Methode zur Parameteridentifikation 56 5.1 Spannungsanalyse 56 5.1.1 Spannungsrichtung 57 5.1.2 Spannungszuwachs 58 5.1.3 Elastische Kompensation 59 5.2 Schnittkraftermittlung 60 5.3 Inverse Analyse 62 5.4 Zusammenfassende Darstellung der Entwicklung 64 6 Validierung der Methode an virtuellen Experimenten 67 6.1 Virtuelle Versuche 67 6.2 Validierung der Spannungsanalyse 68 6.2.1 Vorgehensweise zur Prüfung der Spannungsanalyse 69 6.2.2 Ergebnisse der Überprüfung der Spannungsanalyse 70 6.3 Validierung der Schnittkraftermittlung 76 6.4 Sensitivität der Materialantwort gegenüber den Materialparametern 80 6.4.1 Einfluss variierender Fließkurvenapproximationen 80 6.4.2 Einfluss variierender Fließortkurven 82 6.5 Diskussion zur Wahl geeigneter Schnittlinien 88 6.6 Fehlerbetrachtung 89

L’objectif de ce sujet de thèse est de prédire l’apparition des instabilités plastiques (striction diffuse et striction localisée) dans les matériaux métalliques. Ces matériaux sont décrits par des modèles de comportement élasto-plastique couplés à l’endommagement. L'approche de Lemaitre, reliant l'endommagement à la déformation plastique équivalente et au taux de restitution de la densité d'énergie élastique, est adoptée. Parmi les différents critères et indicateurs qui sont considérés pour la prédiction des instabilités matériau, la théorie de bifurcation et les critères de type force maximum sont tout particulièrement analysés et comparés. Un objectif important de cette étude consiste à déterminer les mécanismes déstabilisants clés associés à cette modélisation du comportement, ainsi que l’impact des différents aspects physiques et des paramètres matériau sur l’apparition de la striction. Les développements résultants sont appliqués à une sélection représentative de matériaux métalliques afin prédire leurs limites de formabilités. Cette approche combinant des lois de comportement et critères de striction peut être utilisée comme outil théorique et numérique d’aide à la conception de nouveaux matériaux à ductilité améliorée / The aim of the present work is to predict the occurrence of plastic instabilities (diffuse and localized necking) in thin sheet metals. The prediction of these plastic instabilities is undertaken using an elastic–plastic model coupled with ductile damage, which is then combined with various plastic instability criteria theory. The bifurcation-based criteria and the maximum force criterion used in this work are formulated within a general three-dimensional modeling framework, and then applied for the particular case of plane-stress conditions for sheet metals. Some theoretical relationships or links between the different investigated necking criteria are established, which allows a hierarchical classification in terms of their conservative character in predicting critical necking strains. The resulting numerical tool is implemented into the finite element code ABAQUS/Standard to predict forming limit diagrams, in both situations of a fully three-dimensional formulation and a plane-stress framework. This approach, that combines constitutive equations to necking criteria, serves as a useful tool in the design of new materials with improved ductility

Made available in DSpace on 2016-12-08T15:56:19Z (GMT). No. of bitstreams: 1 Ricardo Kirchhof Unfer.pdf: 99775 bytes, checksum: 27709fdea77151b3e21cfdaf92146d6e (MD5) Previous issue date: 2015-08-28 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / The new contributions of present thesis are an alternative experimental waviness method and a new theoretical model for evaluation of limit strains of sheet metal forming in the automotive industry, employing the analysis of roughness and waviness parameters evolution with the equivalent plastic strain and a new ductile rupture criterion as function of stress triaxiality. For validation of proposed experimental waviness method and rupture criterion, specimens of IF steel sheet of thickness 0.72mm, 0.82mm and 0.85mm were tested under tensile test and Nakazima sheet forming test with Marciniak flat punch. The specimens were deformed in incremental steps in different strain path such as: balanced biaxial stretching, plane strain, uniaxial tension and pure shear. Present proposed method consider that roughness and waviness parameters are important surface quality index and useful as indicator of onset of visible local necking and should be monitored in sheet metal forming. Thus, a Waviness Limit Curve, WLC, is proposed to be plotted in Map of Principal Limit Strains (MDLC) of sheet metal as an alternative to the forming limit strain curve, FLC. Various parameters of roughness and waviness such as arithmetic average roughness (Ra), maximum peak-valley roughness (Rt), arithmetic average waviness (Wa), total heigth peak-valley waviness (Wt), maximum peak height waviness (Pp) and maximum valley depth waviness (Pv) were measured during the uniaxial and biaxial testing. Tensile test of IF steel specimens at 0º, 45º and 90º to the rolling direction, RD, and Nakazima type of IF steel specimens at 0º RD only, in which a mesh of 2.5mm circles were printed, were employed. During the uniaxial and biaxial testing, the following plastic deformations were calculated in each straining step from the printed circles in the sheet surface: major true strain (1) and minor true strain (2). Parameters of roughness and waviness versus equivalent plastic strain curves were plotted to obtain a phenomenological equation of 4th and 3rd degree polynomials, using the Hill 1979 yield stress criterion. Furthermore, the growth rate of parameters Wa and Wt with equivalent plastic strain were calculated. From the growth rate, it was possible to conclude that the sheet 12 thickness imperfections, measured by Wa and Wt, evolve during material deformation, being possible to predict with accuracy the influence of equivalent plastic strain in the onset of local necking. From the experimental analysis, it was concluded that the maximum peak-valley waviness parameter (Wt) was the best option to chacaracterized the beginning of local necking in IF steel sheet forming. Thus, from the analysis of waviness growth rate parameters, Wa and Wt, during plastic strains, it was possible to propose a criterion of local necking inception. This allowed to obtain and plot an experimental Waviness Limit Curve (WLC) in the map MDLC: the beginning of local necking was established when the normalized parameter Wt* (=Wt/Wto) attained a critical value Wt*= 2.5, but with distinct values of equivalent strains. A second mathematical model was proposed for local necking criterion to obtain the FLC of IF steel sheets, employing stress triaxiality analysis and the Hill 1979 yield stress criterion. 2nd degree polynomials were obtained from plotted and adjusted phenomenological curves in graphs of equivalent strain at rupture versus stress triaxiality, considering the different plastic strain path in the formability testing. The IF steel specimens showed ductile rupture, according to scanning electronic microscope examinations, which showed that rupture occurred due to formation, growth and coalescence of large quantity of pores in many specimens of balanced biaxial (BB) and plane strain (DP). Also, mixed rupture of ductile type due to shear mechanisms in certain regions of tensile test (TS) and Nakazima uniaxial tension (UTN) specimens were shown. / As propostas de ineditismo da presente tese foram mostrar um método experimental alternativo de rugosidades e um modelo teórico de ruptura novo para se avaliar as deformações limites de conformação de chapas metálicas da indústria automobilística, empregando-se uma análise da evolução dos parâmetros de ondulação da rugosidade com a deformação plástica equivalente e um critério de ruptura dúctil em função da triaxialidade das tensões. Para validação do método experimental de rugosidades e do critério de ruptura propostos, utilizou-se corpos de prova de chapas de aço IF de 0,72mm, 0,82mm e 0,85mm de espessura, que foram submetidas aos ensaios de tração simples e ensaio de estampagem tipo Nakazima com punção plano de Marciniak. As amostras foram sujeitas a deformação incremental em diferentes trajetórias de deformação, tais como: estiramento biaxial balanceado, deformação plana, estiramento uniaxial e cisalhamento puro. O presente método proposto considera que parâmetros de rugosidade e ondulação são índices importantes de qualidade da superfície e servem como indicador preciso do início da estricção local visível e devem ser monitorados na conformação de chapas metálicas. Portanto, propõe-se traçar uma Curva Limite de Ondulação da Rugosidade, CLOR, no Mapa das Deformações Principais Limite de Conformação (MDLC) das chapas metálicas como alternativa da curva limite de conformação, CLC. Vários parâmetros do perfil de rugosidade e da ondulação da rugosidade tais como a rugosidade média aritmética (Ra), altura máxima pico-vale (Rt), ondulação média aritmética (Wa), a altura máxima da ondulação pico-vale (Wt), altura máxima de pico da ondulação (Pp) e profundidade máxima de vale (Pv) foram medidos durante os ensaios uniaxiais e biaxiais. Foram utilizados corpos de prova de aço IF em tração simples a 0º, 45º e 90º da direção de laminação, DL, e corpos de prova tipo ensaio Nakazima de aço IF somente a 0º DL nos quais foram impressos uma malha de círculos de 2,5 mm de diâmetro. Durante os ensaios uniaxiais e 10 biaxiais, as seguintes deformações plásticas foram calculadas em cada etapa a partir da malha de circulos impressos na superficie da chapa: a deformação verdadeira longitudinal maior (1) e a deformação verdadeira transversal menor (2). Curvas dos parâmetros de rugosidade e ondulações versus deformação plástica equivalente, empregando-se o critério de escoamento plástico de Hill (1979), foram traçados para se obter equações fenomenológicas do tipo polinomial de 4° e 3° graus. Além disso, foram calculadas as taxas de crescimento dos parâmetros Wa e Wt com a deformação plástica equivalente. A partir das curvas de taxas crescimento, foi possível concluir que as imperfeições na espessura da chapa, medidas por meio de Wa e Wt, evoluem durante a deformação do material, sendo possível prever com precisão a influência da deformação plástica equivalente no início do surgimento da estricção local. Concluiu-se então que o parâmetro da altura máxima total pico-vale das ondulações (Wt) é a melhor opção para caracterizar o início da estricção local em chapas de aço IF. Portanto, a partir da análise das taxas de crescimento dos parâmetros de ondulação, Wa e Wt, durante as deformações plásticas, foi possível propor um critério para o surgimento da estricção local. Isto permitiu obter e traçar a curva experimental de Limite de Ondulação da Rugosidade, (CLOR), no Mapa MDLC: o início da estricção local foi quando o parâmetro normalizado Wt* (=Wt/Wto) atingiu o valor crítico Wt*=2,5, mas com valores distintos da deformação equivalente. Um segundo modelo matemático foi proposto como critério de estricção local para obtenção da CLC de chapas de aço IF, empregando-se análise da triaxialidade de tensões e o critério de escoamento plástico de Hill (1979). Foram obtidos polinômios de 2º grau por meio de curvas fenomenológicas ajustadas e traçadas nos gráficos de triaxialidade de tensões versus deformação equivalente de ruptura, considerando-se as diferentes trajetórias da deformação plástica dos ensaios de conformabilidade. As amostras de chapa de aço IF mostraram ruptura dúctil conforme exame no microscópio eletrônico de varredura, o qual mostra que a fratura ocorreu devido a formação, crescimento e coalescência de uma grande quantidade de espaços vazios ou poros na maioria dos corpos de prova tipo Biaxial Balanceado (BB) e Deformação Plana (DP). Também, as amostras apresentaram fraturas mistas do tipo dúcteis e com mecanismos de cisalhamento em determinadas regiões das amostras fraturadas em Tração Simples (TS) e Tração Uniaxial de Nakazima (UTN).

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / In order to get a better part quality in sheet metal forming, the rate of the material flow into the die cavity must be efficiently controlled. This control is made by a restraining force supplied either by the blankholder, the drawbeads or both. When the restraining force required is too high, the use of drawbeads is necessary, although excessive deformations may be produced. Some others disadvantages, such as difficulties of adjustment during die try-outs in order to determine the actual Drawbead Restraining Force (DBRF), may also be emphasized. To solve these problems and to reduce the number of die try-outs, which are very time consuming, accurate enough drawbeads concepts are necessary. Aiming to understand the influence of the most important parameters on the DBRF and to establish a pre-estimate DBRF theory, in this study a methodology has been developed using similitude. The data bases were achieved by Finite Element (FE) simulations done with an explicit code. Two different materials were used: A-K Steel and 2036-T4 Aluminum. The results have been compared with experimental databases of Nine(1978, 1982) and with the analytical model of Stoughton(1988). The average of absolute error with respect to experimental data bases was about 6 % and, for those cases studied, the maximum discrepancy was found to be less than 11%. For analytical ones, the average of absolute error was about 5 % and, for the cases studied, the maximum error was about 7%. Predictions made with this approach have a very good precision when compared with analytical and experimental results. For this reason, it was used as a contribution for STAMPACK®, an explicit finit element code used to simulate forming process. / Para se obter peças com a melhor qualidade possível em um processo de estampagem de chapas metálicas, a taxa de fluxo de material para dentro da matriz deve ser eficientemente controlada. Este controle é feito por uma força de retenção (FR) originada no prensachapas, nos freios de estampagem ou em ambos. Quando a FR requerida é muito alta, o uso dos freios se torna ainda mais necessário, embora excessivas deformações possam ocorrer na peça estampada por causa do contato com os freios. Outros tipos de efeitos indesejáveis decorrentes do uso deste tipo de dispositivos ainda podem ocorrer, tais como dificuldades para se determinar o valor adequado da FR, o que pode consumir muito tempo. Para se resolver estes problemas e reduzir o número das tentativas de ajustes, são necessários conceitos mais precisos sobre os freios de estampagem. Com a finalidade de se avaliar a influência dos parâmetros mais importantes na FR e de se estabelecer uma teoria para se fazer a predição da FR, desenvolveu-se neste estudo uma metodologia híbrida, empregando-se a teoria da similitude com bases de dados gerados através de simulações numéricas pelo Método dos Elementos Finitos (MEF). Os resultados foram comparados com os experimentos de Nine (1978, 1982) e com o modelo analítico de Stoughton (1988). A média dos desvios absolutos com respeito aos dados experimentais foi de 6% e, para os casos estudados, a discrepância máxima foi sempre menor ou igual a 11%. Em relação ao modelo analítico, a média dos desvios absolutos foi de 5% e, para os casos estudados, o desvio máximo nunca foi superior a 7%. Predições feitas com esta abordagem tiveram, portanto, uma boa precisão, quando comparadas com o modelo analítico e com os dados experimentais. Por este motivo, esta teoria foi aceita como contribuição para o programa STAMPACK®, um código de solução explícita utilizado na simulação de processos de estampagem de chapas metálicas. / Doutor em Engenharia Mecânica

Dans cette thèse, deux méthodes de couplage sont proposées pour la simulation multi-échelle des procédés de mise en forme. Dans la première partie, une procédure simplifiée (couplage indirect) est adoptée pour coupler les codes éléments finis (Abaqus et LAM3) au modèle polycristallin avec un schéma de transition autocohérente basée sur le comportement élastoplastique du monocristal écrit dans le formalisme des grandes déformations. Cette procédure simplifiée consiste à lier le modèle polycristallin avec l'analyse EF par l'extraction de l'histoire de l'incrément de déformation et de contrainte macroscopique, obtenue à partir d'une simulation EF préliminaire avec une loi phénoménologique, et à l'utiliser comme trajet de chargement dans le modèle polycristallin. Cette méthode est appliquée pour la simulation multi-échelle du procédé de skin-pass. Le suivi du trajet de chargement extrait dans la demi-épaisseur de la tôle a permis de prédire l'évolution des grandeurs physiques associées au modèle de plasticité en particulier la texture cristallographique, la texture morphologique et l'écrouissage. Dans la seconde partie de cette thèse, un modèle polycristallin élastoplastique du type autocohérent en petites déformations est couplé au code EF Abaqus via la routine utilisateur UMAT. Ce couplage (dit couplage direct) consiste à utiliser la théorie de la plasticité cristalline comme loi de comportement à chaque point d'intégration du maillage EF. Le polycristal est représenté par un ensemble de N monocristaux. Chaque fois que le code EF a besoin d'information sur le comportement mécanique aux points d'intégration de chaque EF, le modèle polycristallin est appelé. Pour valider ce couplage développé, nous avons effectué des cas tests de simulation de trajets rhéologiques. Les résultats issus de ce couplage ont été validés avec des modèles de référence. À la différence des modèles phénoménologiques, ce couplage permet non seulement d'avoir des informations sur le comportement macroscopique de la structure mais aussi d'obtenir des informations sur l'état de la microstructure du matériau. / In this thesis, two coupling methods are proposed for the multiscale simulation of forming processes. In the first part, a simplified procedure (indirect coupling) is adopted to couple the finite element codes (Abaqus and LAM3) with a polycrystalline selfconsistent model based on the large strain elastoplastic behavior of single crystals. This simplified procedure consists in linking the polycrystalline model with the FE analysis by extracting the history of the increment of macroscopic strain and stress, obtained from a preliminary FE simulation with a phenomenological law, and then using it as loading path prescribed to the polycrystalline model. This method is applied to multiscale simulation of skin-pass processes. By following on the loading path extracted at the halfthickness of the sheet, we can predict the evolution of some physical parameters associated with the plasticity model, in particular the crystallographic texture, the morphological texture and hardening. In the second part on this thesis, a small strain version of the elastoplastic polycristalline self-consistent model is coupled to the Abaqus FE code via the user material subroutine UMAT. This coupling (called direct coupling) consists in using crystal plasticity theory as constitutive law at each integration point of the FE mesh. The polycristal is represented by a set of N single crystals. Each time the FE code needs information on the mechanical behavior at the integration points considered, the full polycrystalline constitutive model is called. In order to validate this coupling, simulations of simple mechanical tests have been conducted. The results of this coupling have been validated through comparison with reference models. Unlike phenomenological models, this coupling provides not only information on the overall macroscopic response of the structure, but also important information related to its microstructure

Le procédé de mise en forme des tôles métalliques, généralement oxydées, par emboutissage est en pleine évolutionpour répondre à de nouveaux besoins : intervalles de tolérances dimensionnelles de plus en plus étroits, productionen grande série à faible coût, apparition de nouvelles nuances de matériaux, souci écologique. La réussite d’une telleopération est prédominée par le contrôle du frottement au niveau du contact entre tôle à former et outil enmouvement relatif. Pour ce faire, nous proposons dans cette thèse de fonctionnaliser la surface de la tôle en ygreffant des molécules organiques auto-assemblées, par simple immersion dans une solution contenant les moléculesréactives. Le couple métal/molécule retenu pour cette étude est un substrat massif de cuivre pur avec des moléculesd’acides alkylphosphoniques (CnH2n+1PO(OH)2) dans le but d’établir une corrélation entre les propriétésphysicochimiques des films ainsi formés et leur comportement tribologique. La réponse en frottement a été étudiéepar deux moyens de simulation expérimentale des conditions de contact (tribomètres bille/plan et étirage-plan). Nousavons considéré l’influence des trois facteurs de préparation de substrat suivant μ la nature de surface juste avant safonctionnalisation (présence d’une couche d’oxyde ou d’hydroxyde), le nettoyage de surface aux ultrasons après safonctionnalisation (élimination des espèces physisorbées) et la longueur de chaîne alkyle de la moléculeorganophosphonique (n=4, 12 et 16). Bien que la formulation de cette fonctionnalisation soit simple (élaborée àpartir d’une seule molécule), les techniques d’investigation de surface (PM-IRRAS, XPS, électrochimie, MEB,…)montrent que les films auto-assemblés sont constitués d’une bicouche de bonne qualité de greffage et de propriétésanticorrosion. Cette bicouche est composée d’une première monocouche de molécules chimisorbées à la surface (ouSAM) sur laquelle est déposée une deuxième couche d’amas physisorbés. D’autre part, notre étude a montréégalement que les amas physisorbés ont un rôle primordial dans le comportement tribologique du film auto-assemblé(abaissement et stabilité du coefficient de frottement et prévention contre le grippage). / Sheet metal forming processes are quickly evolving to meet new requirements of deep drawing such as: continuousminimization of dimensional tolerances intervals, cost efficiency at mass production, appearance of new materialstypes, environmental concern. The success of such operation is predominated by controlling friction at the contactbetween the metal sheet and forming tool in relative motion. In the present work, we propose to functionalize themetal sheet surface using grafting of self-assembled organic molecules by simple immersion in a solution containingthe reactive molecules. Consequently, thin film of alkylphosphonic acid (CnH2n+1PO(OH)2) was grafted on copperoxide substrate in order to establish a correlation between the physicochemical properties of the films andtribological behavior. Frictional response was studied by two experimental means of contact conditions (stripdrawing testing and ball-on-plane tribometer). The influence of three factors for the preparation of substrate wasconsidered. Firstly, the nature of its surface just before functionalization including the presence of an oxide orhydroxide layer was studied. Then, the ultrasonic surface cleaning after functionalization (removal physisorbedspecies) and finally, the alkyl chain length of the organophosphonic molecule (n = 4, 12 and 16). Although theformation of this functionalization is simple (made from a single molecule), the surface investigation techniques(PM- IRRAS , XPS, electrochemistry , SEM, .., etc.) show that self-assembled films consist of a bilayer with a goodgrafting quality and anticorrosion properties. The bilayer is composed of a first monolayer of molecules chemisorbedon the surface (or SAM) on which a second layer of physisorbed species is deposited. Furthermore, this study alsoshowed that the physisorbed species have a key role in the tribological behavior of self-assembled film (lower andstable friction coefficient and high resistance to galling).

This diploma thesis deals with surface forming methods, especially accuracy and measurement of sheet metal bending on Van Mark brakes. The aim of the work was to make bends of samples in the range of 30 °, 60 °, 90 ° and 120 °. The size of the bend of the examined samples is influenced by the suspension of the sheet, which is not taken into account in the currently used type of protractor. The samples were then analyzed on two types of measuring devices - tactile and optical. After measuring, results were evaluated and plotted in the appropriate graphs. Based on the measurement results, a new type of protractor was designed, which will make the user's work easier and more efficient.