PSI - Issue 13

Zahreddine Hafsi et al. / Procedia Structural Integrity 13 (2018) 210–217 Hafsi et al. / Structural Integrity Procedia 00 (2018) 000–000

214

5

Fig. 4. Geometry of the model and boundary conditions (a) 3D geometry; (b) simplified 2D geometry

The evolution of the inner surface concentration Ci is obtained through the outlet pressure evolution (Fig.3.b) using equation (7). Obtained solubility as well as concentration as functions of time are plotted in Fig.5.

Fig. 5. (a) solubility; (b) concentration

Table1 summarizes modelling data used to solve hydrogen gas diffusion through the pipeline wall via COMSOL Multiphysics. Constant diffusion coefficients are assumed for the three tested pipe materials (Cineros et al., 2003; Peng, X. Y., Cheng Y. F., 2014)

Table 1. Parameters of the diffusion model Parameter

Value

Unit

Internal diameter D Pipe wall thickness e

0.6

m m

0.01

m 2 s -1 m 2 s -1 m 2 s -1

Diffusion coefficient Dc : unnitrided API X52 Diffusion coefficient Dc : nitrided API X52

4.32x10 -11 9.41x10 -12 1.7x10 -11

Diffusion coefficient Dc : API X80

As shown in Table 1, the diffusion coefficient is lower for X80 than for X52 steel. It can be predicted that API X80 steel is more likely to resist to hydrogen embrittlement than API X52 steel under same conditions of exposure to pressurized hydrogen gas. Furthermore, a validation of this observation is still required by solving numerically the diffusion equation. Also, numerically solve diffusion equation allows quantifying to what extent hydrogen gas can permeate trough the material lattice under the assumed working conditions. The simulation is performed on the half of the pipe section as illustrated in Fig 4.b with a symmetry boundary condition in the borders. An extremely fine mapped mesh is parameterized and applied to the geometry. The simulation is conducted using a time dependent solver with a time range of 24 hours.