Hydrogen permeation through low alloy steels

A sensitive method of measuring hydrogen permeation is presented, whereby hydrogen is introduced to the entry side of a thin metallic membrane by applying a sinusoidal oscillation about an appropriate value of electrical potential to a cell containing 0.1N NaOH. The corresponding current is taken as the hydrogen incident flux to the surface. The exit side of the membrane is in contact with an ultra-high vacuum chamber where hydrogen partial pressure is measured using a mass spectrometer against a known pump rate in order to obtain a measure of the exit flux. Measurements have been made on a thin nickel membrane as a material with well known and stable hydrogen permeation parameters, and these have been used to assess the sensibility of the technique. A newly developed mathematical model is presented that uses Fourier analysis to determine the phase lag (φ) and amplitude ratio (Λ) between the current or potential and hydrogen partial pressure. Analytical relationships between (φ), (Λ) and the applied frequency (the newly introduced parameter) have been used to determine the relevant thermodynamic and kinetic parameters (such as hydrogen diffusion coefficient and solubility) which enable the effects of both surface and bulk kinetics to be distinguished. Results from the experiments on nickel showed a very good consistency with the literature [diffusion coefficient, D = (3.6±0.5)×10-14 and (1.8±0.29)×10-13 m2/s at 22°C and 60°C, respectively, and solubility of (2.11±0.09)×10-2 and (3.2±0.18)×10-2 mol H2/m3mbar0.5 at 22°C and 60°C, respectively]; this formed a well established calibration for undertaking and interpreting experiments on heat-treatable En24 low alloy steel. Permeation experiments were done for three heat treatment conditions and the results for diffusion coefficient were (6±1)×10-10 m2/s for the annealed, (7.89±1.2)×10-12 m2/s for the quenched and (3.25±0.8)×10-11 m2/s for the quenched and tempered condition. Solubilities were (5.3±1.5)×10-9 for the annealed, (3.28±0.94)×10-5 for the quenched and (1.22±0.6)×10-7 mol H2/m3mbar0.5 for the quenched and tempered condition and trapping parameters measured here were = 7 p kN , = 834 p kN and = 2055 p kN , respectively. Results from these experiments helped to explain the variability reported in the literature for the values of the diffusion coefficient and solubility. The differences in behaviour from classical permeation are more easily attributed to trapping than to surface kinetic effects.EThOS - Electronic Theses Online ServiceGBUnited Kingdo