Rupture (matériau) | EPFL Graph Search

thumb|Courbe de traction idéale d'un matériau ductile thumb|Courbe de traction typique pour un matériau fragile En science des matériaux, la rupture ou fracture d'un matériau est la séparation, partielle (comme une crique ou une fissure ou une brisure) ou complète, en deux ou plusieurs pièces sous l'action d'une contrainte. Une rupture peut être souhaitée par le concepteur de la pièce comme dans le cas de la conception de dispositifs de sécurité ou au contraire celui-ci cherche à éviter cette rupture en mettant en adéquation la fonction de cette pièce avec les dimensionnements et choix des matériaux utilisés et des procédés de fabrication. La rupture fragile, au contraire de celle ductile, est caractérisée par l'absence de déformation plastique macroscopique, et donc par la propagation très rapide des fissures avec faible consommation d'énergie. Dans le cas d'un cristal, la rupture est bien nette, elle suit des plans cristallographiques, on parle de rupture par clivage. Le matériau

Liquid lead-bismuth embrittlement effects on unirradiated and irradiated ferritic/martensitic steels for nuclear applications

Bin Long

In liquid metal spallation targets such as the MEGAPIE (the megawatt pilot experiment) target and the future ADS (accelerator driven system) spallation targets, which utilize liquid lead-bismuth eutectic (LBE) as the target material, liquid metal embrittlement (LME) effects on the target structural materials are considered as one of the critical issues that determine the lifetime of the targets. However, nowadays, the LBE embrittlement effects are not yet well understood, particularly under irradiation conditions. The aim of this work was: (a) to investigate the effects of LBE embrittlement on the mechanical properties of ferritic-martensitic (FM) steels (candidate materials for applications in spallation targets) in various states, and (b) to study the mechanisms of LBE embrittlement effects on the mechanical properties of FM steels. These two goals have been reached by studying LBE embrittlement effects on the mechanical properties of the T91 steel in various conditions: (i) the standard normalized and tempered condition, (ii) hardened by tempering at lower temperatures, and (iii) after irradiation under high energy proton and spallation neutron mixed spectrum. Mechanical tests such as slow-strain-rate tensile (SSRT) tests and 3-point bending tests have been performed to characterize the mechanical properties of the T91 steel and to determine the effects of LBE embrittlement on these mechanical properties. Microstructural analyses such as transition electron microscopy (TEM) and scanning electron mirocopy (SEM) observations have been conducted to obtain the microstructural information needed for understanding the embrittlement mechanisms. SSRT tests on the T91 steel in the standard metallurgical (SM) state, tempered at 760°C and denoted as HT760, revealed that it may encounter LBE embrittlement in the temperature range of 300 to 425°C. For the first time, a well defined "ductility trough" for a FM steel – LBE system was evidenced from the reduction of the fracture strain (or total elongation) of the T91 FM steel in this temperature range. SEM observations of the fracture surfaces showed that the fracture mode of the specimens suffering embrittlement effects is brittle transgranular cleavage. SSRT tests on the hardened T91 steel, tempered at 500 or 600°C and denoted as HT500 and HT600, respectively, showed more pronounced LBE embrittlement effects. In comparison to that of HT760, the "ductility troughs" of HT600 and HT500 cover a wider temperature range. The LME onset temperature TE is not exactly known but most probably lower than 150°C, which means close to the melting point temperature (125°C) of LBE. The ductility recovery temperature TR of HT500 is higher than that of HT760 and HT600. The results of SSRT tests on the HT600 and HT500 materials demonstrate clearly that the LBE embrittlement effects on the tensile properties of FM steels are strongly enhanced by hardening. Both the influenced temperature range and the degradation level increase with the hardening amount. The fracture mode of HT600 and HT500 materials tested in liquid LBE at T ≤ 450°C showed mainly brittle transgranular cleavage. SSRT tests on the T91 and F82H FM steels irradiated at temperatures in the range of 140-500°C to doses in the range of 0.048 to 20 dpa revealed significant irradiation-induced hardening and embrittlement effects (loss of ductility) as compared to the unirradiated ones. The SSRT tests performed on irradiated specimens in liquid LBE revealed that the specimens undergo a further reduction in ductility. The ductility of some irradiated specimens is reduced to a very low level of about 2-3%. 3-point bending tests on the T91 steel showed that the load-displacement curves of specimens tested in liquid LBE are different from those of specimens tested in Ar. The evaluated J-integral values of specimens tested in liquid LBE are always lower that of the specimens tested in Ar. It means that the crack propagation in a specimen tested in liquid LBE needs less energy as compared to a specimen tested in Ar. These results suggest that the LBE embrittlement is so-called "crack propagation controlled" rather than "crack nucleation controlled". SEM observations revealed that the crack propagation mode undergoes a ductile-to-brittle transition. In liquid LBE the fracture toughness of FM steels can be reduced by about 20% as compared to that in inert Ar environment. 3-point bending tests on the hardened T91 materials HT500 and HT600 demonstrated that the LBE embrittlement effects are much more drastic for hardened materials, similarly to what was observed as a result from SSRT tests. The JQ value is reduced by as much as 95% for both HT600 and HT500 materials tested in liquid LBE. The hardened materials exhibit unstable crack propagation behavior in liquid LBE in a wide temperature range of 150 to 500°C. 3-point bending tests on the irradiated T91 steel revealed not only irradiation-induced embrittlement effects but also LBE embrittlement effects. The fracture toughness of the irradiated specimens was found to decrease further as a result from testing in liquid LBE. SEM observations of the fracture surface showed a mixture of intergranular and transgranular fracture. The LME phenomena evidenced for the T91 steel – LBE couple could be interpreted using the model of adsorption-induced reduction in cohesion of atomic bonds combined with the Kelly-Tyson-Cottrell (KTC) criterion (σ/τ ≥ σmax/τmax). According to the KTC criterion, the tendency to fail by cleavage increases as (σmax/τmax) decreases. The σmax value can be greatly reduced by adsorption of liquid Pb or Bi atoms at crack tips, while the τmax value can be increased by hardening the steel by means of either deformation, or tempering, or irradiation. Practically, the present results indicate that a great attention should be paid to application of FM steels in liquid LBE, since the mechanical properties of FM steels, such as the ductility and the fracture toughness, can be substantially degraded by LBE embrittlement effects, particularly under irradiation conditions as the LBE embrittlement are enhanced by irradiation-induced hardening. Fundamentally, the observations made in this work suggest that the mechanism responsible for liquid LBE-induced embrittlement effects on FM steels can be essentially interpreted by using the adsorption-induced reduction in cohesion of atomic bonds model combined with the KTC criterion for brittle cleavage fracture.

EPFL2009

Tragverhalten und Bemessung von Holz-Glas-Verbundträgern unter Berücksichtigung der Eigenspannungen im Glas

Klaus Kreher

The construction principles of "Timber-Glass-Composite-Girders" demand practical and theoretical research to ensure reliable designs. Toward this objective the mathematical description of load-bearing and the non-rigid bond is the subject of this work, i.e. to define the contributing parameters for the design of these girders. The non-rigid-bond of the adhesive joint is of great importance because of its influence on the load-bearing of the system as a whole. The loading of girders is dominated by permanent forces thus leading to the obligation of knowing the long-term behaviour of the system and its components. The definition of the characteristic values, taking into account the long-term load-bearing, is an intrinsic part of this investigation. It can be shown that the load-bearing of the girders depends on the fracture of the glass panes. The properties of glass-fractures are a function of the residual internal prestresses due to the heat-strengthening of the panes. The load-bearing, particularly the post-cracked behaviour, changes with respect to the intensity of these internal prestresses. The timber is able to reinforce the cracked glass, leading to a ductile load-bearing behaviour as in the girders, with a dependency upon the size of the remaining fractures. A model based on the differential equations of the non-rigid bond is defined in order to calculate the loading-peaks of the adhesive joint and the material loading itself. The characteristic values of the non-rigid bond, such as the number and distances between cracks and the load introducing length, were evaluated in tests on composite slabs and small scale girders. The existing calculation models which take into account non-rigid bonding were modified to adapt the influence of the width and thickness of the adhesive joint. In order to calculate the stressing of the material the existing calculation models were adapted to calculate the post-cracked situation. The load-bearing behaviour of glass-panes bent with respect to its strong neutral axis needs other safety-considerations than panes, bent off their plane. It is shown that the existing safety considerations subject to the use of glass can not easily be adapted onto the composite girders. The bending with respect to the strong neutral axis and the reinforcement of the glass of the timber demands a different hypothesis to adapt the fractural mechanical analysis and to establish a safe limit conception. This is part of the performed research of this document. The following descriptions present briefly the obtained results: Depending on the quality of the glass (residual stresses due to heat-strengthening, e.g. annealed glass, heat-strengthened glass, fully tempered glass), the ductility and the mode of failure of the girders does change. As a result of its failure mode, annealed glass without internal prestresses offers the highest remaining load-carrying potential after the first crack has appeared. This ductility and thus the structural safety, diminishes with an increasing degree of internal prestressing due to thermal treatment. Heat-strengthened glass (with various degrees of prestressing, various residual stresses) shows a decrease in remaining load-carrying capacity with an increasing degree of prestressing until it fails in a brittle mode as fully toughened glass does. This has to be well considered in respect to safety considerations. Girders with panes made of glass having internal residual stresses due to heat strengthening below 50N/mm2 are considered to collapse ductily, residual stresses bigger than 50N/mm2 cause brittle failure. The effective stiffness of the girders decreases under permanent loads in function of the bondage. Both, timber and adhesive take part in this decrease; the influence of both of the involved materials has been defined. Since the design of conventional glass constructions which uses the concept of principle stresses cannot be adapted, unidirectional shear stresses had to be determined for the shear strength of glass, which is defined as 25N/mm2 for annealed glass. The characteristic values of the adhesive to define the non-rigid bonds have been determined, likewise the influence of the width and thickness of the adhesive joint. The load introducing length, the distances between cracks and the number of cracks are defined with tests on composite slabs. A model to describe the theoretical bond with the defined parameters based on the differential equations was developed. This model allows the calculation of the material stressing at the load introduction and on both sides of a crack. To calculate the stressing of glass and timber existing calculation models which take into account non-rigid bonding were modified to adapt the influence of the width and thickness of the adhesive joint. In order to calculate the stressing of the material the existing calculation models were adapted to calculate the post-cracked situation. The safety factors for the influences of the glass surface, environmental conditions, load duration and the parameters of fracture mechanical analysis as developed as they are known for semi-probabilistic safety concepts for glass constructions cannot easily be adopted. The bending with respect to the strong neutral axis demands a different hypothesis in order to adapt the fractural mechanical analysis and to develop the safety parameters. For the reliable design of timber-glass-composite-girders a contribution to the appropriate use of probabilistic, semi-probabilistic and deterministic safety concepts are given. A realised construction (Hotel "Palafitte" in Monruz (NE)) shows an example and the experience made in designing girders.

EPFL2004

Creep Properties and Related Microstructural Changes of a Lamellar Titanium Aluminide Alloy before and after Helium-Implantation

Per Magnusson

The creep properties of a cast gamma-alpha-2 lamellar titanium aluminide of type Ti-46Al-2W-0.5Si (at. %) containing beta-particles was investigated using both helium-implanted and non-implanted material. Creep tests were performed in vacuum under constant stress using small miniaturized dog-bone shaped samples with gauge dimensions of 10 × 2 × 0.2 mm3. Creep temperatures ranged from 700 °C to 1000 °C and applied stresses ranged from 75 MPa to 400 MPa. Helium implantation was carried out with 0-24 MeV helium ions homogeneously implanting the miniaturized samples from 6.3 to 1333 appm He, at temperatures of 150 °C and 630 to 1000 °C. Non-implanted samples showed very good creep resistance up to temperatures of roughly 850 °C; at temperatures above 900 °C a decrease in creep resistance was observed. For temperatures up to 900 °C the activation energy of creep was determined to 405 kJ/mol and the stress exponent of 7.2. Size and geometry effects of miniaturized samples were observed in form of large scatter in creep strain rates and lower strains at fracture. Electron microscopy showed that the initial microstructure was unstable and that the alloy had a tendency to decompose the alpha-2 phase and form beta-gamma structures, loss of initial microstructure was accelerated with increasing temperature. The creep deformation mechanism was determined to be a combination of stress-induced phase change and dislocation creep. A gradual change in fracture behavior was observed, at lower temperatures fracture was brittle with little necking, at the highest temperature the fracture was completely ductile. Material crept and implanted at 700 and 800 °C exhibited embrittlement at helium contents above 10 appm, characterized by a strong loss of ductility and creep lifetime. Samples crept and implanted at 900 °C experienced embrittlement for all tested helium contents. Investigations with transmission electron microscopy revealed a substantial clustering and growth of helium bubbles at interfaces in the alpha-2 phase. An average bubble size for embrittlement between 5.3 and 6.7 nm was determined. The behavior and evolution of helium bubbles were studied with post implantation annealing experiments. Samples were implanted at 150, 630, 800 and 1000 °C and subsequently annealed at temperatures from 600 to 900 °C. Material implanted at 150 °C showed no bubbles before annealing, only after post-implantation annealing was widespread clustering observed. Material implanted at temperatures of 630 °C or higher, exhibited bubbles before post-implantation annealing. Depending on the implantation temperature, two different branches were observed in the size and distribution of helium bubbles even after post-implantation annealing. Samples implanted at 150 or 630 °C and then annealed at higher temperatures had high bubble density with small bubbles found in the entire microstructure. Samples implanted at 800 or 1000 °C had low bubble densities and the bubbles were only observed at alpha-2/beta and alpha-2/gamma interfaces, with sizes larger than the average bubble radius for embrittlement. Based on the post-implantation annealing experiments, a temperature limit for the onset of helium embrittlement between 630 and 700 °C is predicted. The helium embrittlement, the loss of initial microstructure and decreased creep resistance at temperatures above 900 °C suggests a limited applicability of lamellar TiAl in advanced nuclear reactor environments.

EPFL2012