*3.1. Nucleation and Growth of Reversed Austenite*

Figure 1a shows that reversed austenite started to nucleate when the temperature reached around 640 ◦C. Due to the limitation of magnification, only nucleation at the grain boundaries was observed clearly; the nucleation inside the grain cannot be presented. Theoretically, the pre-existing retained austenite were the preferable sites for austenite nucleation and growth [16]. In this investigated steel, the 5~10% retained austenite played a beneficial role in improving nucleation number. Therefore, after completion of reversed austenite transformation, the grain size was small. As seen in Figure 1b, the grain size was around 3~10 μm at the temperature of 850 ◦C.

**Figure 1.** Reversed austenite at different temperatures on heating during thermal cycle: (**a**) 642.3 ◦C, (**b**) 748.1 ◦C, (**c**) 810.4 ◦C.

After completion of reversed austenite transformation on heating, austenite started to grow, as shown in Figure 2. The austenite grain size at different temperatures was measured on the CLSM samples using the circular-intercept method. The austenite growth rate was plotted in Figure 3. It reveals that the grain size growth rate was small below the temperature of 1100 ◦C on continuous heating. There was a sharp increase in austenite growth from 1100 ◦C up to 1250 ◦C. When the peak temperature of 1308 ◦C was attained, the maximum grain size of ~52 μm was observed and almost no growth was found during the following continuous cooling until it reached austenite decomposition temperature.

**Figure 2.** Austenite growth at different temperatures on heating during thermal cycle: (**a**) 1172 ◦C, (**b**) 1224.5 ◦C, (**c**) 1308.3 ◦C.

**Figure 3.** Austenite grain size VS temperature in the heating and cooling stage.

The reversed austenite generally nucleated at grain boundaries, as seen in Figure 1. Especially, the pre-existing retained austenite took a positive role in improving austenite nucleation sites. The filmy retained austenite grew directly in one dimension, which implies a larger nucleation number in the present investigated steel. The large initial nucleation number caused the slow grain growth. On the other hand, the grain grew in the boundary migration way, which is controlled by the relatively short distance of carbon diffusion. Therefore, the grain growth rate was small below 1100 ◦C.

During the temperature increases, the element diffusion coefficient became lager. The large diffusion of both carbon and Mn changed the method of grain growth from grain boundary migration to grain annexation. The sharp increase of grain growth rate was present during continuous heating from 1100 ◦C up to 1250 ◦C.

In low-alloyed steel, the austenite grain continuously grows up during following cooling process above peak temperature [2], which is referred to as "hot-inertia" [17]. Interestingly, in this present steel, there was almost no growth of austenite measured, as shown in Figure 3. It is reported that Mn atoms are likely to segregate toward grain boundaries during continuous cooling [18]. Mn segregation at the grain boundary was assumed to pin the grain boundary migration and hinder the growth of austenite grain. Further study is needed to figure out the Mn segregation behavior and its effect on grain growth.
