Le comportement en fatigue de l’acier à très haute résistance (THR) X23NiCoCrMoAl13-6-3 étudié est caractérisé par une dispersion dans les performances de tenue en fatigue. Deux modes d’amorçages de fissures sont observés sur sa courbe de Wöhler : le premier mode correspondant à des amorçages de fissures à partir d’inclusions non métalliques en surface des éprouvettes menant aux plus faibles durées de vie. Le second mode correspond à des amorçages de fissures en interne sur inclusions menant à de longues durées de vie et à des fissures en forme d’œil de poisson (« fish-eye »). Le lien entre les conditions d’essai (niveau de sollicitation, température et environnement), la nature et la localisation du site d’amorçage et la durée de vie résultante n’est jusqu’alors pas clairement identifié. L’objectif de ce travail de thèse est donc de mieux appréhender les mécanismes physiques à l’origine des différents types d’amorçage et de propagation en fatigue menant à la dispersion en durée de vie.La première partie de ce document présente la caractérisation microstructurale multi-échelle de l’acier THR avec l’identification des entités microstructurales et de leurs dimensions caractéristiques. Ensuite, l’analyse des faciès de rupture a permis de caractériser de manière approfondie les zones associées aux différents stades de propagation de la fissure de fatigue. Pour l’amorçage en surface, on trouve : 1) la particule rompue 2) la propagation sous air 3) la rupture ductile finale. Pour les amorçages en interne deux stades supplémentaires sont observés : 1.a) formation d’une zone à grains fins (« fine granular area, (FGA) ») autour de la particule si sa taille est inférieure à une taille critique et 1.b) propagation sous vide. La mécanique linéaire de la rupture est utilisée pour estimer le facteur d’intensité des contraintes de la fissure, Kmax, pour chacune de ces zones.Le facteur Kmax associé à la transition entre FGA et fish-eye se révèle être constant à une température donnée. Cette transition se caractérise par une taille de zone plastique en pointe de fissure de l’ordre de trois fois la largeur des lattes de martensite. Lorsque la température augmente, le Kmax en fin de FGA diminue mais la taille de la zone plastique reste constante. Le caractère cristallographique, pseudo stade I, de la propagation dans la FGA est mis en évidence. Il a été de plus confirmé qu’une FGA pouvait se former en cours de propagation de fissures à partir d’éprouvettes entaillées (SENT), sous vide sollicitées à Kmax faible. Le caractère cristallographique a été confirmé par analyse EBSD 3D du faciès de rupture. Un calcul de durée de vie en propagation pure sous air et sous vide a montré que les durées de vie peuvent varier d’un ordre de grandeur en fonction de l’environnement. / The fatigue behavior of a very high strength steel X23NiCoCrMoAl13-6-3 shows a large scatter in fatigue life. Two fatigue crack initiation scenarios are found. They are separated by a plateau in the Wöhler curve: the first one corresponds to crack initiation at the specimen surface from non-metallic inclusions leading to short fatigue lives. The second scenario corresponds to internal crack initiation from particles leading to so-called fish-eye cracks. The link between test conditions (load level, temperature and environment), the nature and location of the crack initiation site and the associated fatigue life has yet not been clearly established. The objective of the present PhD project is to better understand the physical mechanisms leading to the different fatigue crack initiation and propagation mechanisms.In the first part of the thesis the microstructure of the very high strength steel was studied at multiple scales and its main constituents and their respective sizes were identified.Further, the analysis of the fatigue tests by fractography has permitted to identify the different characteristic fatigue crack propagation stages. For surface fatigue crack initiation, the following stages were found: 1) a broken particle 2) fatigue crack propagation in air 3) the final ductile fracture. For internal fatigue crack initiation two more fatigue crack propagation stages are observed: 1.a) formation of a fine granular area (FGA) for small particles and 1.b) propagation in vacuum. Linear elastic fracture mechanics are used to evaluate the stress intensity factor, Kmax for each of these zones.Kmax revealed to be constant at the end of the FGA. The size of the plastic zone at the end of the FGA was calculated and corresponded to three times the width of the martensite laths. Tests at different temperatures (20°C, 200°C and 400°C) revealed a decreasing FGA size with increasing temperature at constant applied stress amplitude. As a consequence, the critical stress intensity factor varied as the FGA decreased with temperature. In contrast, the critical plastic zone size remained constant and equal to the sizes of microstructural features. This represents a strong similarity between crystallographic, stage I-like, crack propagation and FGA formation in a vacuum. This was further confirmed as an FGA was formed during crack propagation using SENT samples in vacuum and for low Kmax levels. The crystallographic character was verified by an SEM/FIB analysis of the fracture surface. A fatigue life prediction, assuming propagation only, showed that the fatigue life can vary by an order of magnitude as a function of the test environment corresponding to surface and internal crack initiation.

Orientador: Anselmo Eduardo Diniz / Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Engenharia Mecanica / Made available in DSpace on 2018-08-05T19:57:39Z (GMT). No. of bitstreams: 1 Pivetta_CarlosSergio_M.pdf: 2572259 bytes, checksum: d75f09a259545ac56a46ecc1d5870ba0 (MD5) Previous issue date: 2005 / Resumo: O fresamento de cavidades de moldes e matrizes em material endurecido usando ferramenta de topo esférico tem limitações em termos de profundidades das cavidades, da taxa de remoção de cavaco e da vida da ferramenta. Algumas operações de fresamento em cavidades profundas não podem ser realizadas, devido à necessidade de se ter ferramenta com alto balanço, o que gera vibração e, em conseqüência, danos à rugosidade da peça e à vida da ferramenta. Outro ponto ainda não totalmente compreendido é a influência da rugosidade deixada na peça pela operação anterior, na rugosidade da operação de acabamento, já que, devido às baixas profundidades utilizadas nesta operação, a rugosidade da operação anterior pode influenciar a profundidade real removida. O principal objetivo deste trabalho é verificar a influência da relação comprimento/diâmetro da ferramenta (balanço), da rugosidade da operação anterior e da velocidade de corte na rugosidade da peça e na vida da ferramenta de fresamento em acabamento. Para isto foram realizados ensaios de fresamento em aço endurecido AISI D2 (58 HRC) com ferramenta de ponta esférica de metal duro recoberto, variando-se os citados parâmetros de influência. Dentre as conclusões deste trabalho, pode-se destacar: nem a rugosidade da operação anterior, nem o balanço da ferramenta influenciaram a rugosidade da peça, pelo menos nos níveis utilizados neste trabalho; com balanço da ferramenta pequeno, consegue-se baixos valores de desgaste e, com isso, longa vida da ferramenta, mas quando se utiliza conjuntamente balanço da ferramenta e velocidades de corte altas, a vida da ferramenta é bastante reduzida / Abstract: The milling of molds and dies cavities carried out on hardened steel using ball end mill presents limitations in terms of cavity depth, metal removal rates and tool life. Some milling operations on deep cavities are not possible due to the require of long tool overhang, what generates vibrations and, consequently, resulted in poor surface roughness and short tool life. Other point which is not totally understood up to the present is the influence of the surface roughness obtained in the previous operation on the results of finishing operation because on finishing operations, the actual depth of cut is not much larger than the height of roughness left on the surface by the previous operation. The main objective of this work is to verify the influence of the ratio tool overhang/ tool diameter, the roughness of the previous operation and cutting speed on the tool wear in finishing operations. Besides, it also intends to verify the influence of tool wear on the tool vibration, measured through the real time tool deflection. Milling experiments on AISI D2 hardened steel (58 HRC) were carried out using coated (TiNAl) carbide ball end mills with the input variables already mentioned. The main conclusions of this experiments, for the used conditions, were: the previous surface roughness neither the tool overhang have influence on the resulting surface roughness; when the tool overhang was the lowest one, cutting speed did not influence so much tool wear, but when the highest tool overhang was used, cutting speed presented a strong influence on tool wear / Mestrado / Materiais e Processos de Fabricação / Mestre em Engenharia Mecânica