Discussion of “the effect of heat treatment on Mg2Si coarsening in aluminum 6105 alloy”

Abstract

This interesting article [1] includes measurements of the size distribution, average radius , and volume fraction f of globular Mg 2 Si particles in this Al-Mg-Si alloy vs duration of homogenization treatment for up to 7 days at 590 °C. The authors conclude, on the basis of regression analysis, that the results show a particularly strong correlation with n ϭ 3.8 in [1] where t is time at temperature and is the value of at t ϭ 0, suggesting that grain boundary diffusion (n ϭ 4) dominates volume diffusion (n ϭ 3) in the coarsening of these particles. The present purpose is to offer an alternative interpretation of these results.The experimental data at issue are presented in Table I, together with resulting values of and for each value of t, taking at t ϭ 0. These results give mean K values of 3.6 Ϯ 0.4 ϫ 10 Ϫ20 m 3 /s for volume diffusion and 8.3 Ϯ 0.7 ϫ 10 Ϫ25 m 4 /s for grain boundary diffusion control. Table II presents corresponding values of K predicted from the LSW [2,3] and Kirchner [4] models, respectively:and [3] where C ∞ is solid solubility of the controlling solute; D and DЈ are its diffusivities in the matrix and along grain boundaries; w is grain boundary width; ␥ is dispersoid/matrix interfacial energy; ⍀ is the volume per mol of dispersoid containing one atom of controlling solute; R is the gas constant; T is the temperature (here 863 K); A ϭ (2/3-m ϩ m 3 /3), where m is half the ratio of grain boundary energy to particle/matrix interfacial energy; and B ϭ 0.5 ln (1/f ), where f is the average area fraction occupied by a particle on the grain boundary. Resulting values of K LSW and K GBD are, respectively, factors of 2 to 4 ϫ 10 3 and 3 ϫ 10 5 below their corresponding values obtained experimentally (Table I), indicating that neither solute diffusion through the ␣-Al crystal lattice nor along its grain boundaries can account for the observed rates of coarsening of Mg 2 Si in this alloy at 590 °C. A further possibility arises from the proximity of 590 °C to the solidus temperature (ϳ620 °C according to the phase diagram "typical"Table I. Results from Reference 1 for Mean Radius r vs Time t at 590 °C for Mg 2 Si Particles in 6105 Aluminum Alloy t (days) 0 1 2 3 4 5 6 10.0 16.6 19.8 22.2 23.6 24.1 26.3 K 1 (m 3 /s) -0.041 0.039 0.038 0.035 0.030 0.033 mean 0.036 Ϯ 0.004 m 3 /s K 2 (m 4 /s) -0.76 0.83 0.90 0.87 0.76 0.90 mean 0.83 Ϯ 0.07 m 4 /s Here, and with . r o ϭ 10 mm K 2 ϭ (r 4 Ϫ r 4 )/t K 1 ϭ (r 3 Ϫ r 3 )/t r (mm)

Characterisation of microstructure and thermal stability of rapidly solidified Al–8·5Fe–1·3V–1·7Si alloy during prolonged exposure at 625°C

“…65Fe alloy. The enhanced rates reported by Gilman et al at 425uC and Pontikakos and Jones at 500uC could be a consequence of short circuiting of diffusion paths by matrix grain boundaries or dislocations, while that obtained in the present work at 625uC could be associated with incipient melting close to the liquidus temperature, 22 increasing the effective diffusivity by a factor of 10 3 or 10 4 , similar to the enhancement factor observed. Instances of similarly large enhancements of K EX near to the liquidus have been identified for Al 3 Ti in Al, 23 for Si in Al (Ref.…”

“…65Fe alloy. The enhanced rates reported by Gilman et al at 425uC and Pontikakos and Jones at 500uC could be a consequence of short circuiting of diffusion paths by matrix grain boundaries or dislocations, while that obtained in the present work at 625uC could be associated with incipient melting close to the liquidus temperature, 22 increasing the effective diffusivity by a 12 Cube of mean radius r of equivalent sphere v. time t at 625uC for Al 13 Fe 4 particles in material 22 The evolution and more rapid coarsening of a population of faceted silicide particles from the original population of nanoscale spheroidal silicides could be a consequence of loss of coherency with the matrix when silicides exceed a critical size. Figure 13 shows cube of mean radius v. time at 625uC for the spheroidal silicide particle population, indicating K EX 5(r 3 2r o 3 )/ t,2 nm 3 s 21 compared with predicted values of 370 and 4 .…”

“…24) as well as for Mg 2 Si in 6105 alloy. 22 The evolution and more rapid coarsening of a population of faceted silicide particles from the original population of nanoscale spheroidal silicides could be a consequence of loss of coherency with the matrix when silicides exceed a critical size. Figure 13 shows cube of mean radius v. time at 625uC for the spheroidal silicide particle population, indicating K EX 5(r 3 2r o 3 )/ t,2 nm 3 s 21 compared with predicted values of 370 and 4 .…”