Grain boundaries effects on hydrogen diffusion and trapping in fcc materials – CRISTALHYD

The methodology consists of 3 distinct steps to get experimental and numerical data at different scales.

Task 1: Measures at local scales (experimental approach)
It consists in studying the structural monocrystals of different orientation and bicrystals to obtain the experimental values ??of the diffusion coefficients.
Tools: TEM, SEM-EBSD, electrochemical permeation
Purpose: Construction of a diagram (statistical and tensor) computable in an FEM code

Task 2: Mechanisms at local scales
- Development of joint representative grain
- Evaluation of energy diffusion and trapping
Tools: DM and DFT
Purpose: To get for every grain boundary
- the diffusion coefficient tensor
- the Enthalpy of activation of the diffusion and trapping

Task 3: Scale of polycrystalline aggregate
- Construction of structures polycrystalline 2D and 3D
- Statistical distribution of properties on the grains, grain boundaries and triple junctions
- Development of statistical analyzes
- Building homogenization schemes
Goal: Obtaining
- a homogenizing model integrating diffusion and Trapping phenomena
- Statistical data on the distribution of different hydrogen states

We obtained several significant results:
The results do show a dependence of the hydrogen diffusion coefficient with the crystallographic orientation. We developed a tensorial diffusion formalism to obtain the diffusion coefficient for any crystallographic direction. This allowed us to calculate the iso-values of the diffusion coefficient and represent them on a reverse pole figure

Moreover, in the objective to study the relationship between the microstructure of the grain boundary and the creation of point defects (vacancies, interstitial solutes), we studied the stability of the gaps in the vicinity of grain boundaries. Our calculations show that the atomic level interaction between the vacancies and the plane of the grain boundary is more pronounced for high free volume grain boundaries. Indeed, the grain boundaries with a high expansion volume, and consequently, a large interfacial energy source are considered sites with high potential energy interaction. Beyond this analysis on the capacity of these structures to absorb point defects such as vacancies, we also showed a difference in behavior between the energy distribution of these defects and the strain field induced by their creation around the plane of the boundary.

On the other hand, in our finite element calculations, the spatial distribution of high diffusivity boundaries was characterized statistically with connectivity parameters. By correlating these parameters with the diffusion coefficient, we have shown that there are three diffusion regimes depending of the network structure of high diffusivity boundaries. In parallel, we have developed a homogenization approach that can take into account the connectivity of grain boundaries. This study was extended in 3 dimensions, to validate the results.

The study of single crystals is being finalized. that of bicrystals nickel S3, S5 and S11 (special Boundaries) were characterized by EBSD to accurately know their characteristics (boundary plane, misorientation angle, axis of rotation, ...) and the misorientation gradient in the vicinity of each of them. The study of the segregation of the hydrogen on the latter is in progress. The diffusion part will only start after the single crystal will be fully characterized and the apparent solubility for the different boundaries will be determined.
Regarding our calculation on the atomic scale, we focused on the energy of formation of vacancies in the vicinity of grain boundaries and the study of the deformation field caused by the creation of vacancies in these grain boundaries. All these results must be reproduced in the presence of hydrogen so as to separate the impact of crystal defects in the vicinity of the joints on the segregation of the hydrogen from the impact of hydrogen on the formation of the same shortcomings.
To make a more realistic analysis, development of a digital microstructure reconstructed from Nickel samples was started through a combination of EBSD and a code of Finite Element. This approach consists in producing, by scanning electron microscopy, several successive maps of the same sample surface, removing the material between each step. These maps are then used for the reconstruction of a digital microstructure in 3 dimensions.
A finite element model reconstructed from these microstructures would provide a richer model in terms of information on the nature of grain boundaries. This study is coupled with a local study on the distribution of hydrogen, produced by SIMS analysis.

Scientific journals
1. B. Osman Hoch, A. Metsue, J. Bouhattate and X. Feaugas, Effects of grain-boundary networks on the macroscopic diffusivity of hydrogen in polycrystalline materials. Computational Materials Science, 2015. 97(0): p. 276-284.
2. E. Legrand, J. Bouhattate, X. Feaugas, «Generalized model of desorption kinetics: Characterization of hydrogen trapping in a homogeneous membrane«, International Journal of Hydrogen Energy 39 (2014), 8374-8384.
3. E. Legrand, A. Oudriss, C. Savall, J. Bouhattate, X. Feaugas, «Towards a better understanding of hydrogen measurements obtained by thermal desorption spectroscopy using FEM modeling«, International Journal of Hydrogen Energy 40 (2015), 2871-2881.
4. A. Oudriss, J. Bouhattate, C. Savall, J. Creus, X. Feaugas, F.A. Martin, P. Laghoutari, J. Chêne, On the Implication of Hydrogen on Inter-granular Fracture, Procedia Materials Science, 2014. 3 :p. 2030-2034.

6 international conferences where 4 as a guest speaker
5 national conferences where one was a key note

Hydrogen embrittlement (FPH) is a major cause of failures in the industry structures. This multi-physical phenomenon often leads to cracks initiation and propagation highly dependent on the environment. The severe failure but also the human and economic consequences that may result have been the source of many studies and publications on this topic since the first evidence of the phenomenon, a century ago. Despite many national and international projects on the HE, the mechanisms of hydrogen diffusion and trapping linked to the materials microstructure are relatively unexplored. In other words, if the mechanisms of embrittlement are now well identified, the kinetics associated with transport and segregation of the species (hydrogen) at the origin of the damage remain poorly apprehended. Unlike other defects, such
as dislocations or vacancies, there is a clear misunderstanding of the grain boundaries effect on hydrogen diffusion and trapping. In the most general case of polycrystalline structures, the average behavior of the aggregate (grain and grain boundaries) and the deviation from it are key elements concerning the diffusion and trapping processes which affect partly the material durability.This scale transition stage is often neglected however it can have important consequences on the measurement of the diffusion coefficient and/or the trapping energy.
We propose to develop electrochemical permeation tests (EP) and thermal desorption spectroscopy (TDS) of nickel alloys (single crystal and bi-crystrals and finally polycrystals). In parallel, we will proceed to the development of "robust" finite element codes and more "efficient" calculation means for the calculation of diffusion on 3D aggregates. For this work will serve as a guide to the development of scale transition model of non-stationary phenomena (diffusion process) coupled with the trapping process associated to the heterogeneities of time and space. This project is a "coupled" way of thinking through continuous exchange between the experiment and model development efforts with simple structures to disentangle the various phenomena associated with leading transport and trapping of hydrogen. Treatment results according to the different variabilities will allow us to build tools for a reflection support in the development of microstructures less inclined to hydrogen embrittlement depending on whether it is a function of diffusible or trapped hydrogen (grain boundaries, development of dipole dislocation structures…).