Fig. 5 presents the SEM micrographs

Fig. 5 presents the SEM micrographs of original substrate and the HZ obtained by LTH with different parameters. The element contents of P1-P3 detected by EDS are listed in Table 3. The microstructure of original substrate was tempered sorbite which was composed of whiteblocky ferrite with a smooth grain boundary and a great number of spheroidal carbide particles dispersed on boundary of the ferrite. For the laser transformation hardened samples, the microstructure was composed of uniform martensite with distinct boundary. However,there were no pores observed, while some cracks and other defects can be found. It was worth noting that the microstructure in the HZ consisted of plate martensite (PM), lath martensite (LM), retained austenite (RA) and some carbide particles. The structure that possessed the morphological feature of lenticular shape but the slices did not parallel to each other and intersected into a certain angle (60° or 120°) [19] was identified as PM, while structure that had the morphological features of being nearly parallel, almost equal length and arranged in bunches[20,21] was LM. For the sample with 1.7 kW 3 mm/s (Fig. 5(c e)), the microstructure was uniform and fine with almost the same amount of PM and LM, and some blocky RA were distributed widely. However, many carbide particles could still be observed on the boundary of martensite as shown in Fig. 5(e), which illustrated that the temperature on the surface was not high enough with lower laser power (namely,lower heat input) to dissolve more carbides during the heating process.As a result, the carbon content in RA surrounded by martensite of the sample with 1.7 kW 3 mm/s only accounted for 1.2% (P1). As the laser power increased to 1.8 kW 3 mm/s (Fig. 5(f-h)), the microstructure was coarser obviously, the number of PM was increased while the blocky RA was reduced simultaneously, and almost no carbide particles could be observed on boundary of the martensite (Fig. 5(h)). This suggested that higher laser power led to higher heat input and thus higher surface temperature, which would improve the diffusing ability of carbon atoms. More carbon atoms dissolved into the austenite during the austenitizing process, but they could not diffuse out of the grains in time during the rapid cooling process, thus leading to the formation of high-carbon martensite in the end. Meanwhile, due to the higherhardening temperature, the coarse martensite which transformed earlier bounded more austenite and prevented them from expanding and transforming into BCC structures, resulting in an increasing in number of RA surrounded by martensite at room temperature. Therefore, the carbon content in RA surrounded by martensite of the sample with 1.8 kW 3 mm/s was as high as 1.8% (P2), and more alloying elements such as Cr had been dissolved. For the sample with 1.8 kW 4 mm/s(Fig. 5(i–k)), due to the higher laser power, carbide particles couldhardly be found (Fig. 5(k)) as same as the sample with 1.8 kW 3 mm/s,and the blocky RA was less than the sample with 1.7 kW 3 mm/s.However, the lowest liner energy made it consist of more LM but less PM, this was because faster scanning velocity led to shorter interactiontime and faster cooling rate, some of carbon atoms had not yet dissolved into austenite before the process entering the cooling stage, so the carbon content in its RA surrounded by martensite was 1.5% (P3).