Analysis of thermal desorption of hydrogen in metallic alloys


Type
Article
Change log
Authors
Abstract

The degree of embrittlement of metallic alloys is sensitive to the concentration of absorbed hydrogen, with hydrogen traps (particularly at grain boundaries) playing an important role. Thermal desorption spectrometry (TDS) is widely used to measure the detrapping and diffusion behaviour of hydrogen in metallic alloys. However, it is problematic to obtain a consistent interpretation of TDS data from the literature, due to the large number of material parameters that influence the measurement, and this results in a wide range of quoted values for trapping parameters such as the number of trap types, trap binding energies and trap densities. In this paper, the governing partial differential equation for hydrogen diffusion with sink and source terms for a single trap is formulated in non-dimensional form, assuming local equilibrium between the hydrogen atoms at the lattice sites and the trap sites. An asymptotic analysis reveals two distinct regimes of diffusion behaviour in TDS tests. Kissinger-type behaviour is expected in a TDS test for low heating rates on an alloy with a low lattice activation energy. Contour maps of maximum hydrogen flux and the corresponding temperature are plotted using axes of trap density and trap binding energy by making use of the full numerical solution (and asymptotic solutions). These maps serve as a useful tool for an accurate and simple determination of the trap binding energy as well as the trap density.

Description
Keywords
Thermal desorption spectrometry, Hydrogen embrittlement, Diffusion, Trap binding energy, Trap density
Journal Title
Acta Materialia
Conference Name
Journal ISSN
1359-6454
1873-2453
Volume Title
144
Publisher
Elsevier BV
Sponsorship
The authors would like to acknowledge the funding and technical support from BP through the BP International Centre for Advanced Materials (BP-ICAM-02ext) which made this research possible. Support from the US Office of Naval Research (grant N62909-14-1-N232) is also gratefully acknowledged.