Abstract
A finite element analysis is proposed to study the effect of specimen dimensions on lateral diffusion of hydrogen during hydrogen permeation flux measurements. The error of measurement on thick specimens because of 1D diffusion approximation may be as much as 70%. A critical condition for accurate measurements is to designate the area of hydrogen monitoring/exit surface smaller than the area of hydrogen charging/entry surface. For thin to medium thickness specimens (ratio of thickness to specimen radius of 5:10 and below), the charging surface should be maximized and the monitoring surface should be minimized. In case of relatively thick specimens (ratio of thickness to specimen radius above of 5:10), use of a hydrogen-diffusion barrier on the specimen boundaries is recommended. It would completely eliminate lateral losses of hydrogen, but cannot eliminate the deviation towards 2D diffusion near the side edges. In such a case, the charging surface should be maximized and the monitoring surface should be as closer in dimension as the charging surface. A regression analysis was carried out and an analytical relationship between the maximum measurement error and the specimen dimensions is proposed.