A micromechanical interpretation of the temperature dependence of Beremin model parameters for french RPV steel

Abstract

International audienceLocal approach to brittle fracture for low-alloyed steels is discussed in this paper. A bibliographical introduction intends to highlight general trends and consensual points of the topic and evokes debatable aspects. French RPV steel 16MND5 (equ. ASTM A508 Cl.3), is then used as a model material to study the influence of temperature on brittle fracture. A micromechanical modelling of brittle fracture at the elementary volume scale already used in previous work is then recalled. It involves a multiscale modelling of microstructural plasticity which has been tuned on experimental inter-phase and inter-granular stresses heterogeneities measurements. Fracture probability of the elementary volume can then be computed using a randomly attributed defect size distribution based on realistic carbides repartition. This defect distribution is then deterministically correlated to stress heterogeneities simulated within the microstructure using a weakest-link hypothesis on the elementary volume, which results in a deterministic stress to fracture. Repeating the process allows to compute Weibull parameters on the elementary volume. This tool is then used to investigate the physical mechanisms that could explain the already experimentally observed temperature dependence of Beremin's parameter for 16MND5 steel. It is showed that, assuming that the hypothesis made in this work about cleavage micro-mechanisms are correct, effective equivalent surface energy (i.e. surface energy plus plastically dissipated energy when blunting the crack tip) for propagating a crack has to be temperature dependent to explain Beremin's parameters temperature evolution

Introducing stress random fields of polycrystalline aggregates into the local approach to fracture

“…The empirical distribution function of this failure load level set shows the failure probability of the specimen in increasing load level. The detail of this approach is showed in [7,3]. Figure 2 shows the 20 failure probability curves estimated from 20 bi-dimensional specimen simulations with 20 different microstructures using this approach.…”

“…The fitting of the failure probability curve by the Beremin model gave us the brittle properties of the material. This approach has been recently coupled with polycrystalline aggregates simulations at the microscopic scale [3]. The main idea of such an approach is to model a single material representative volume element (RVE) as a polycrystalline aggregate and compute the principal stress field under given load conditions.…”

“…Within the weakest link theory the failure of a single critical carbide induces the failure of the RVE. Then the failure probability curve of the RVE is fitted using the Beremin model [3,4] to explain the dependence on temperature of material properties. However it is believed that numerous parameters such as grain geometry and orientation may influence the stress field and thus the final result.…”

Quantification and localization of the liquid zone of partially remelted M2 tool steel using X-ray microtomography and scanning electron microscopy

“…A remarkable Bauschinger effect was observed and reflects the presence of significant internal stresses. A measure of the Bauschinger effect can be provided by the so-called ''Bauschinger stress'' denoted X and defined as X ¼ ðS ð1Þ þ S ð2Þ Þ=2 where S (1) and S (2) are respectively the forward macroscopic flow stress at the end of the tension stage and the reverse macroscopic yield stress in the compression stage. Let us note that this definition of X is half the quantity defined in Ref.…”

“…Nowadays, a particular attention is paid in the nuclear industry to the influence of the carbide particle size on material's behavior and local strains and stresses with the aim of better understanding and deriving fracture toughness in the 16MND5 steels used in nuclear reactor vessels [1]. Mean field approaches are often used in designing heterogeneous materials, as they can lead to reasonable estimates of the stress-strain levels in the matrix and in the inclusions, while being very effective in terms of computation costs.…”