Diffusion was measured in the low-alloy (3335 series) ferritic steel (3Ni-0.78Cr-0.65wt%Mn) at 673 to 873K; where a temper embrittlement effect took place. The HIRBS technique was used. The results were compared to previous ones obtained in pure α-Fe. The As, Sb and Sn diffusivities were higher than that of Fe self-diffusion. The results (table 3) could be described by:
As: D (m2/s) = 5 x 10-6 exp[-202(kJ/mol)/RT]
Sb: D (m2/s) = 5 x 10-5 exp[-233(kJ/mol)/RT]
Sn: D (m2/s) = 3 x 10-6 exp[-207(kJ/mol)/RT]
It was found that As exhibited higher diffusivity values in the steel, while Sn was fastest in pure α-Fe. All of the impurity diffusion coefficients lay in a narrow band, along the measured temperatures, both in the pure α-Fe structure (large grain size) and in the martensitic-like morphology of the steel. Only As in steel deviated notably. The HIRBS technique revealed the true bulk diffusion coefficients in these matrices and the influence upon the diffusivity of the various elements. At longer annealing times, the influence of the microstructure was very important; showing the need to use effective bulk diffusion coefficients. The danger of extrapolating data from a pure matrix, instead of measuring diffusion coefficients in real structures, was emphasized.
HIRBS Measurements of As, Sb and Sn Diffusion in a Low-Alloy Ferritic Steel. D.N.Torres, M.J.Iribarren, R.A.Pérez, F.Dyment: Materials Letters, 2006, 60[3], 301-4
Table 3
Bulk Diffusivities of As, Sb and Sn in 3335 Steel
Diffusant | Temperature(K) | D(m2/s) |
As | 873 | 3.2 x 10-18 |
As | 796 | 4.3 x 10-19 |
As | 729 | 1.1 x 10-20 |
As | 673 | 1.1 x 10-21 |
Sb | 873 | 5.5 x 10-19 |
Sb | 796 | 2.7 x 10-20 |
Sb | 729 | 9.8 x 10-22 |
Sn | 823 | 1.3 x 10-18 |
Sn | 793 | 8.2 x 10-20 |