The relation between microstructure, crystallographic texture and hydrogen
diffusion has been studied in an IF-steel. The steel samples were deep drawn
until 10%, 20%, 30% and 40% strain levels and then hydrogen diffusion
coefficients were determined using Helios II system. Light optical
microscope (LOM), scanning electron microscope (SEM) and electron
backscatter diffraction (EBSD) were used for microstructural
characterization and crystallographic texture studies. The dependence of
microstructural parameters was evaluated by Pearson correlation coefficient
(PCC) values. Those evaluations showed that local misorientations,
crystallographic texture and dislocation density depend on each other. The
PCC values show that grain size and dislocation density are the independent
microstructure related parameters and those parameters were used to build a
model to predict hydrogen diffusion coefficient by multiple linear
regression analysis. Using this model sensitivity analysis was also
performed to understand which parameter the hydrogen diffusion is most
sensitive to. The results of this analysis show that hydrogen diffusion is
more sensitive to dislocation density, which indicates that dislocations are
more effective trapping sites for hydrogen atoms. On the other hand, grain
boundaries are less effective trapping sites since they also provide an
additional diffusion mechanism.