**5. Effect of UFH on Microstructure**

The effect of heating rate on the microstructure was clearly reported in [43] for a process involving a heating rate of 778 ◦C/s in comparison to a standard 10 ◦C/s heating rate.

The recrystallized volume for a low-carbon steel (0.14 wt. % C), decreased from 77.4% to 27.3%, the fraction of recovered ferrite increased from 8.6% to 54.5%, and the martensite + pearlite fraction increased from 14% to 21.8% (Figure 11).

**Figure 11.** Microstructural evolution for CA (HR: 10 ◦C/s) and UFH (HR: 778 ◦C/s), low-carbon steel (0.14% C) (data from [43]).

This result is in contrast with that in [38,39], where an increase in Ac1 temperature was reported with increasing HR; therefore, less martensite was expected in the microstructure of the UFH sample than that in the CA sample.

The difference can be explained by the lack of ferritic recrystallization in the UFH sample.

A possible explanation of the above results can be that the peak temperature of the specimen treated with CA was lower than that processed by UFH, and that, in the latter treatment, some austenite was formed. In the case of CA, it is reasonable to assume that not only was the peak temperature lower, but also the treatment time at high temperature was longer, so that the recrystallized fraction could be higher than that of the steel treated with UFH. When the steel was cooled, the austenite transformed into martensite.

From the comparison between CA + soaking and UFH prior to Q and P heat treatment [38], for a carbon steel (0.25 wt. % C), the fraction of ferrite increased for higher heating rates from 0% at 10 ◦C/s + soak. to 25% at 1000 ◦C/s. This was due to the very short heating time and consequently the absence of complete austenitization. The most important difference from 500 to 1000 ◦C was the fraction on fresh martensite (15% → 2%) that was replaced by ferrite (5% → 25%) (Figure 12).

Therefore, for this class of steels, UFH heating rates of up to about 500 ◦C/s are still suitable to achieve a similar microstructure to those of Q and P grades. Instead, for heating rates above 500 ◦C/s, complex microstructures are produced where ferrite coexists with retained austenite.

The dependence of microconstituent fraction formation as a function of a combination of HR, peak temperature, and steel chemical composition is reported in Figures 13–16 (data from [57]). The comparison is between a low-carbon steel (0.2 wt. % C) and stabilized low-carbon steel (0.2 wt. % C + Mo + Ti + Nb), and the results show that:


**Figure 12.** Microstructural evolution for different heating rates (data from [40]; steel chemical composition: 0.25% C + 1.2% Si + 3.0% Mn).

**Figure 13.** Microstructural evolution at different HRs for peak temperature at 902 ◦C (data from [57]; steel chemical composition: 0.2% C).

**Figure 14.** Microstructural evolution at different HRs for peak temperature at 915 ◦C (data from [57]; steel chemical composition: 0.2% C + Mo + Ti + Nb).

**Figure 15.** Microstructural evolution at different HRs for peak temperature at Ac3: 950 ◦C (data from [57]; steel chemical composition: 0.2% C).

**Figure 16.** Microstructural evolution at different HRs for peak temperature at Ac3: 950 ◦C (data from [57]; steel chemical composition: 0.2% C + Mo + Ti + Nb).

The effect of an increased heating rate in reducing the final fraction of martensite at the end of the treatment was apparent at 100 ◦C/s. Possible reasons to explain why the martensite fraction was generally lower in the low-alloyed steel compared to the alloyed one are, on one hand, the higher critical transformation temperatures and, on the other hand, the lower hardenability. In the former case, the amount of austenite transformed during the process was lower; in the latter case, a lower fraction of the austenite formed at high temperature is converted to martensite during cooling.

#### **6. Effect of UFH on the Textural Evolution and Magnetic Properties of NGO Steels**

Nonoriented grain (NGO) electrical steels can be classified as low-C steels, and regarding the microstructural evolution, mechanical properties, and phase transformation temperatures, the effect of UFH can be considered to be similar. In this section, the effect of UFH on the textural evolution and magnetic properties of NGO steels is reported.

For NGO electrical steels, the role of ultrafast annealing on textural evolution is evident, since an increase in the occurrence of Goss component is observed, which leads to an improvement of magnetic properties [61–63].

The magnetic properties of Fe–Si steels strictly depend on textural evolution [64–70], and consequently on the starting microstructure [71–73], annealing conditions [74], and cold-rolling settings [75–80]. Core losses and eddy current losses are strongly dependent on recrystallization and grain size morphology after cold rolling. Grain size and the absence of impurities such as dislocations, grain boundaries, or precipitates are the most important factors determining power losses. The most relevant textural components in electrical steels are the Goss component (110) [001], the rotated Cube component (100) [011], and the γ-fiber [111]//ND [78–82]. After cold rolling, the texture is characterized by very strong Goss and cube component orientations that, in principle, are very positive in terms of magnetic features; in fact, the matrix is highly deformed. However, the effect of the subsequent recrystallization annealing treatment promotes the development of the γ-fiber, the typical recrystallization texture of ferritic steels, and the consequent reduction in deformation components. In conventional industrial lines, the nucleation of textural components starts during the (slow) heating stage and interacts with the concurrent recovery process. Therefore, since the Goss and Cube components nucleate in high-energy zones of the material, the UFH treatment could represent an opportunity to reduce the undesired effect of recovery, and to produce NGO steels with better textural properties.

Wang et al. [64] explored different heat treatment approaches to optimize the magnetic properties of electrical steels. A cold-rolled NGO steel was subjected to an extremely short annealing cycle in the range of 3–30 s, with heating rates from 15 to 300 ◦C/s and peak temperature from 880 to 980 ◦C. The presence of silicon (no less than 1%) increases the critical temperatures of the steel, thus permitting the use of peak temperatures higher than those of the corresponding low-carbon steels. In the case of fast heating, a very strong {110}<001> (Goss) texture develops [64,72]. Its intensity increases with increasing heating rate, but decreases as the annealing time increases. Moreover, in coarse-grained specimens, the Goss component is significantly strengthened, and the {111}<112> component is slightly weakened with increasing heating rate. On the other hand, in a fine-grained specimen, the intensity of the Goss component showed only a slight increase, but the {111}<112> component was greatly reduced.

Generally speaking, as the heating rate increases, the typical recrystallization texture of ferrite (γ-fiber) weakens significantly due to the very short heating treatment, whereas the Goss component is extremely favored. Since the nucleation of the Goss component occurs in regions of the deformed matrix with higher stored energy (namely, shear bands), higher heating rates reduce the negative effect of recovery, thus promoting their nucleation and further growth [65–68,71]. This result is in agreement with the work of Hutchinson [83], who reported a schematic ranking of the nucleation rate of different textural components as a function of time during annealing. In fact, from a kinetic viewpoint, the {110} orientations started nucleating earlier than {111} did, thus exploiting a larger kinetic advantage as the heating rate increased.

Regarding the magnetic properties of NGO electrical steels as a function of heating rate, the following results emerged from the experiments [63,64]:

