Multiscale Modeling of Theta ' Precipitation in Al-Cu Binary Alloys
Multiscale Modeling of Theta ' Precipitation in Al-Cu Binary Alloys
Multiscale Modeling of Theta ' Precipitation in Al-Cu Binary Alloys
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
V. Vaithyanathan et al. / Acta Materialia 52 (2004) 2973–2987 2985<br />
average aspect ratio<br />
50<br />
40<br />
30<br />
20<br />
10<br />
0<br />
MF 225C(R)<br />
MF 250C(R)<br />
FP/PFM<br />
FP/PFM(avg)<br />
0 10 20 30 40 50 60 70<br />
time (hrs)<br />
Fig. 13. Average aspect ratio <strong>of</strong> h 0 precipitates from the <strong>Al</strong>–2%<strong>Cu</strong> alloy<br />
is compared with the experimental results <strong>of</strong> Merle and Fouquet [38].<br />
The <strong>in</strong>dividual aspect ratios from the simulation (FP/PFM) are displayed<br />
along with the average aspect ratio <strong>in</strong> solid l<strong>in</strong>e. The equilibrium<br />
aspect ratio from the simulation results is <strong>in</strong> good agreement with<br />
the late stage equilibrium value from the experiments.<br />
cence. <strong>Al</strong>so, the same <strong>in</strong>teraction resists coalescence<br />
between unlike variants (which would result <strong>in</strong> an antiphase<br />
doma<strong>in</strong> boundary or tw<strong>in</strong> boundary formation)<br />
and <strong>in</strong>hibits their growth, similar to the experimental<br />
observations [6,30].<br />
From our simulations <strong>of</strong> microstructural evolution<br />
(Fig. 10), we have extracted quantitative precipitate features<br />
like the thickness, length and aspect ratio as a<br />
function <strong>of</strong> time. The thicken<strong>in</strong>g and lengthen<strong>in</strong>g k<strong>in</strong>etics<br />
obta<strong>in</strong>ed from an <strong>Al</strong>–2%<strong>Cu</strong> alloy are shown <strong>in</strong> Figs. 11<br />
and 12. These multiscale model<strong>in</strong>g (FP/PFM – firstpr<strong>in</strong>ciples<br />
phase-field model) k<strong>in</strong>etics are compared with<br />
experimental results available from the literature for<br />
precipitation temperatures between 200 and 250 °C. The<br />
average (solid l<strong>in</strong>e) and maximum (dotted l<strong>in</strong>e) values <strong>of</strong><br />
length and thickness from the simulation are compared<br />
with average values from experiments. It should be<br />
po<strong>in</strong>ted out that the experimental measurements selectively<br />
measure a few hundred precipitates from the large<br />
number <strong>of</strong> available precipitates. <strong>Al</strong>so, experimental<br />
images (<strong>in</strong> TEM) can <strong>in</strong>clude precipitate overlap, which<br />
can result both <strong>in</strong> an overestimation <strong>of</strong> the average length<br />
and also <strong>in</strong> difficulties <strong>in</strong> resolv<strong>in</strong>g thicknesses at the<br />
nanometer-scale, lead<strong>in</strong>g to large uncerta<strong>in</strong>ties <strong>in</strong> these<br />
measurements. From Figs. 11 and 12, it is clear that the<br />
thickness values from our simulations are <strong>in</strong> closer<br />
agreement with experimental results than are the values<br />
<strong>of</strong> precipitate length. We assert that the source <strong>of</strong> the<br />
discrepancy <strong>in</strong> the case <strong>of</strong> precipitate length is due to the<br />
pronounced <strong>in</strong>terfacial mobility anisotropy and therefore,<br />
the much longer non-equilibrium lengths <strong>in</strong> experimental<br />
measurements as compared to the equilibrium<br />
values predicted by the simulation. The average aspect<br />
ratio <strong>of</strong> h 0 precipitates <strong>in</strong> the 2%<strong>Cu</strong> alloy predicted from<br />
our simulation (see Fig. 13) agrees well with the measurements<br />
<strong>of</strong> Merle and Fouquet [38], who obta<strong>in</strong>ed data<br />
after long ag<strong>in</strong>g times (a few hundred to a thousand<br />
hours). The maximum value <strong>of</strong> equilibrium aspect ratio<br />
from the model predictions does not exceed 30, whereas<br />
large non-equilibrium values are observed dur<strong>in</strong>g ag<strong>in</strong>g<br />
experiments, presumably <strong>in</strong>duced by the <strong>in</strong>terfacial mobility<br />
anisotropy.<br />
The multiscale model can also be used to study and<br />
compare the precipitation k<strong>in</strong>etics <strong>of</strong> alloys with different<br />
precipitate volume fractions. As an example, the<br />
precipitate morphology and aspect ratios are compared<br />
for three different alloy compositions, 2%, 3% and 4%<br />
<strong>Cu</strong>. The smallest composition studied (2%) is the closest<br />
to the <strong>Cu</strong> compositions <strong>in</strong> many b<strong>in</strong>ary and multicomponent<br />
<strong>Al</strong>–<strong>Cu</strong>-based alloys <strong>of</strong> practical <strong>in</strong>terest.<br />
<strong>Al</strong>though the higher volume fractions might be somewhat<br />
unrealistic, they are important, as they provide<br />
Fig. 14. Precipitate evolution as a function <strong>of</strong> dimensionless time for (a) <strong>Al</strong>–3%<strong>Cu</strong> and (b) <strong>Al</strong>–4%<strong>Cu</strong> alloy. t ¼ 900 corresponds to a real time <strong>of</strong> 10<br />
and 7.5 h for a 3% and 4% alloy, respectively, while for the 2% alloy <strong>in</strong> Fig. 10, it corresponds to 30 h.