**2. Cast Irons**

Cast irons comprise a large family of solid ferrous alloys. The main difference with steels is the higher carbon (C) and silicon (Si) content, with the richer carbon phase being critical in the microstructural transformations of the matrix. Additionally, significant amounts of manganese (Mn), phosphorus (P), and sulfur (S), as well as minor contents of molybdenum (Mo), chromium (Cr), and nickel (Ni), can be found in these castings. The most common classification for cast irons is based on microstructural features. The first feature is graphite morphology, where cast irons can be classified mainly as lamellar (gray) or nodular (spheroidal or ductile). On the other hand, the matrix phase can be used to identify cast irons as ferritic, pearlitic, martensitic, austenitic, or bainitic [5].

GIs are a broad class of casting alloys characterized by a lamellar (flake) graphite disposal, usually in a ferrous matrix. They are traditionally used in industrial applications, due to their good castability, flexible mechanical properties, and low cost compared to steel [6,7]. A variety of fatigue and crack growth models have been performed [8–10] to predict the fatigue life of GIs in standard applications, while other tribological studies [11–16] have identified the main wear mechanisms under dry and lubricated conditions, emphasizing the high variability in wear rates in different tribo-systems. Graphite plays a

major lubricating role that helps reduce the wear coefficient. However, a cracking tendency is observed under high loads, and adhesion is the main wear mechanism in lubricated environments.

On the other hand, DI is a ferrous material containing graphite in a small and rounded shape, due to the nodulizing effect of magnesium, and it is present in higher concentrations. In comparison, DI exhibits higher strength and ductility, which is the main reason for its growth in industrial applications [5]. The effects of cooling rate and matrix type (i.e., pearlitic/ferritic) on the mechanical properties of DIs have been explored both experimentally and numerically [17–22], figuring that an increment of the ferritic or pearlitic phase can lead to very dissimilar properties. Thus, an appropriate balance of the alloying elements and the initial microstructure is needed to achieve the desired performance. In this regard, studies on crack growth have been conducted [23,24] that show a comparable behavior to that of GIs. However, under similar conditions, DI presents better fatigue resistance due to the higher stress intensity factor and the nodular shape of graphite that blunts the crack, while the lamellar shape of GI exhibits an anisotropic behavior that leads to stress concentrations at the interface between the matrix and the flake tips. Furthermore, wear analyses of DIs have been carried out to understand the underlying wear mechanisms and to compare the removal rates with those of industrial steels [4,25–28]. In mild wear environments, plowing is the predominant wear mechanism, and as the load applied increases, sliding wear becomes more severe, due to enhancement of the plastic deformation and initiation of cracks in the subsurface. Moreover, the influence of both bulk and matrix microhardness in the wear rates of ferrous alloys has been addressed, revealing that some initial cast irons microstructures could lead to tribological performances in the range of soft industrial steels, e.g., 52,100 and 1070.

Over the last decades, austempered ductile cast iron (ADI) has emerged as a novel option for achieving a good combination of mechanical properties. Because of its low cost and good machinability, ADIs have gained more interest in automotive and agricultural applications [2,3,29]. The characteristic ausferritic matrix is achieved by a pre-heating cycle, whose quality is primarily influenced by the processing window [30], i.e., the time interval between the ausferritic transformation and the carbide precipitation (Figure 1). Additional modifications of the austempering process have been explored to enhance the mechanical properties of ADIs [31,32].

**Figure 1.** Phase transformations and variables of a conventional austempering heat treatment (Reprinted with permission from ref. [30]. 2015 Springer Nature).

The main limitations of ADI are its lower machinability and ductility when compared to as-cast ductile iron, due to higher austenitic content and the absence of pro-eutectoid ferrite, responsible for the work hardening of the material [29]. Nonetheless, tribological studies have demonstrated a superior ADI wear resistance in comparison to both GI and DI [33]. This difference is attributed to the higher toughness achieved by the bainitic microstructure of ADI, as it leads to severe plastic deformation in the direction of sliding with minimum material loss. Moreover, wear resistance can be further enhanced in ADIs based in the pre-austempered iron matrix, as graphite morphology plays a significant role in the control of crack propagation and thermal conductivity of the material and, therefore, of its thermal fatigue resistance.

Alternatively, the improvement of mechanical properties to fulfill industrial requirements has been achieved by applying conventional heat treatments on as-cast iron samples that can modify their entire primary microstructure. The most common treatments consist of normalizing, annealing, or quench-tempering, where furnaces are operated to provide uniform heating through the metallic piece until a critical temperature is reached, since, in the absence of chromium, iron carbides can be dissociated into austenite and graphite at annealing temperatures [34]. The microstructural transformations are also influenced by the cooling rate, which varies with the selected treatment.

The main results of normalizing are the enhancement of different properties, such as hardness, impact toughness, and tensile strength, and can be optimized with selected temperature, holding time, and cooling rate [35]. If the goal is to achieve high ductility and machinability at the expense of strength, then full annealing is conducted. The matrix transformations include the removal of cementite and decomposition of pearlite into graphite and ferrite [36]. Quench-tempering is used to mostly improve the wear resistance and strength of cast irons. The resulting microstructure of GIs consists of a retained austenite matrix with carbon-saturated martensite plates that becomes finer with higher alloying elements content [37], while a primary martensitic structure with small soft ferrite regions adjacent to the graphite nodules can be achieved in DIs [38].

### **3. Laser Processing of Cast Irons**

As reviewed in the previous section, conventional heat treating of cast irons has been extensively used to enhance their performance to higher standards. However, it has been acknowledged that altering the properties of the entire bulk, e.g., to achieve a hardening effect, can be detrimental to the machinability advantages of these materials. Therefore, LST has appeared as a major opportunity to efficiently improve the overall performance of cast irons.

This technique is characterized by a highly localized, chemically clean, automatable, and fast thermal process, where only controlled regions of the material are hardened during the treatment. Steen and Mazumder [39] extensively detail the theoretical basis of LST, in which a laser beam is defocused or oscillated to irradiate a small area with high power density, and the relative motion between the beam and the sample, usually provided by a CNC table, allows users to precisely cover complex geometries (Figure 2). The laser heat supply raises the temperature at the surface of the specimen, which is transferred into the metallic body by thermal conduction.

**Figure 2.** Schematic representation of LST in rectangular samples.

The objective of LST is to achieve temperatures above the point of critical transformation, which is the austenitizing temperature (solid-state transformation) in the case of laser surface hardening (LSH), or the melting temperature in the case of laser surface melting (LSM). Then a rapid self-quenching process takes place, because of the large volume of adjacent and unaffected material, resulting in a controlled low distortion of the workpiece. Since the heating rates are very high, the initial transformations take place under conditions far from equilibrium, while phase transformations on cooling can be addressed with the help of suitable continuous cooling transformation (CCT) diagrams. Depending on the cooling and solidification rates to which they are exposed during the treatment, the final microstructure directly influences the new mechanical properties of the material.

Laser surface modification of metallic samples, including cast irons, has been performed for over fifty years. As one of the first available tools, CO2 lasers have been widely employed to achieve both LSH and LSM, usually aided by the application of a thin coating, in order to ensure higher absorption and thermal efficiency. However, modern breakthroughs have allowed the use of more efficient lasers for these purposes, such as Nd:YAG and diode lasers, where LST has been carried out with and without coatings with successful results. As shown in Table 1, other LST features include setting an appropriate scanning velocity, the laser power distribution, spot geometry and size, the number of scanned tracks (with or without overlapping), and the use of shielding gas to prevent oxidation. As some authors define, the linear energy density (or heat input) is a global parameter that includes the effect of both laser power and scanning speed by the following expression:

$$
\overline{E} = \frac{P}{\overline{v}'} \tag{1}
$$

where *E* is the heat input, *P* is the laser power, and *v* is the scanning velocity. In circular spots, this definition can be modified to consider the geometry of the irradiation zone and the energy distribution.


**Table 1.** Typical parameters and characterization tests used in LST of cast irons.

Table 2 summarizes the main differences and challenges that LST poses for every reviewed as-cast iron. It is apparent that each material presents different limitations to the extent of LST; therefore, the laser parameters and conditions must be carefully selected to fulfill the technical requirements. In the following sections, a detailed review of the characterization and results of LST in cast irons is performed to delve into the differences and advantages of the technique as a function of the initial composition.


**Table 2.** Comparison of LST challenges for each iron casting.

### **4. Laser Surface Treatment of Gray Cast Irons**

LST on GIs date back to the beginnings of this technology. CO2 lasers are the most used for this purpose, but it is frequently necessary to apply thin coatings to increase absorptivity. The characterization of this technique is carried out mainly in terms of microstructure, hardness, and wear/erosion resistance, and the relevance of each work resides in the variation of some inputs of LST or the exploration of new approaches and tests for further understanding of the process.

Regarding the microstructural changes, when the laser parameters are sufficient to ensure reaching the solid-state transformation temperature without surpassing the melting temperature, it is reported that a martensitic structure is obtained from the fast cooling of the austenitic matrix. Hwang et al. [40] used a 5 kW CO2 laser with fixed parameters, ranging from 1 to 4.5 kW and from 1 to 13 m/min, to modify the properties of as-cast GI samples with pearlite matrix and small amounts of cementite and steadite (iron phosphite, cementite, and ferrite). The purpose of their work was the improvement of the performance of GI piston rings used in marine diesel engines by LSH, and from the experimental results, the authors determined a proper heat input window from 30 to 45 J/mm to achieve the desired hardening effect.

This conclusion is supported by Liu and Previtali [41], as the authors agree that the feasibility window for hardening without melting (LSH) is very narrow since the melting point of cast irons is lower than in pure iron or steels. In this work, a Gaussian distribution diode laser with higher absorptivity was used, and two constant temperature levels were set on the surface of the samples (1000 and 1100 ◦C), varying the laser power using a proportional–derivative–integrative (PID) controller, to avoid the melting of the material during heating.

Furthermore, Wang et al. [42,43] compared the effects of LSH on five types of GIs: untreated (GI), quench-tempered (QTGI), austempered (AGI), quench-tempered and laserhardened (LHQTGI), and laser-hardened austempered (LHAGI) samples. A continuouswave Nd: YAG laser with Gaussian energy profile and 8 ms pulse duration was used, considering a 2 mm spacing between adjacent laser spots. Three similar zones were observed in the laser-treated cases: a central laser-hardened zone, where graphite flakes completely dissolve during heating and transform to a ledeburitic matrix; a heat-affected zone (HAZ) consisting of martensite, due to self-quenching, without changes in graphite morphology; and a typical substrate microstructure of tempered martensite below the HAZ. Figure 3 presents the microstructural transformations of LHQTGI as a function of the tempering temperature.

**Figure 3.** Microstructure of LHQTGI with different tempering temperatures: (**a**) laser hardened zone (Zone 1) at 316 ◦C, (**b**) Zone 1 at 399 ◦C, (**c**) Zone 1 at 482 ◦C, (**d**) Zone 1 at 552 ◦C, (**e**) interface between Zone 1 and Zone 2 (heat-affected zone) at 316 ◦C, (**f**) interface between Zone 1 and Zone 2 at 399 ◦C, (**g**) interface between Zone 1 and Zone 2 at 482 ◦C, (**h**) interface between Zone 1 and Zone 2 at 552 ◦C, (**i**) interface between Zone 2 and Zone 3 (substrate) at 316 ◦C, (**j**) interface between Zone 2 and Zone 3 at 399 ◦C, (**k**) interface between Zone 2 and Zone 3 at 482 ◦C, (**l**) interface between Zone 2 and Zone 3 at 552 ◦C, (**m**) Zone 3 at 316 ◦C, (**n**) Zone 3 at 399 ◦C, (**o**) Zone 3 at 482 ◦C, (**p**) Zone 3 at 552 ◦C (Reprinted with permission from ref. [43]. 2020 Elsevier).

On the other hand, Trafford et al. [44] made the first approach to LSM in GIs with different sets of fixed laser power and scanning velocity, established to ensure three different cases on the surface: hardening without fusion, with an incipient degree of fusion, and one with complete fusion on the laser path. In the case of LSM, three differentiated regions can be seen along the cross-section of the laser path: a fusion zone close to the surface, characterized by a fine ledeburite structure, which is caused by the rapid rate solidification of graphite sheets completely dissolved during heating; a deeper transition region, formed by a matrix composed of retained austenite and thick martensite plates, in which there is a partial degree of undissolved graphite sheets surrounded by a layer of ledeburite; and a base zone, where the action of the laser is not enough to produce the phase transformation. The authors also observed that cracks formed on the surface, with increased number and size according to the degree of fusion.

De Oliveira [45] also implemented a computer interface to control the temperature at the surface of GI samples, to ensure the transformation to austenite without melting, by linearly changing the heat input of a 2 kW Nd:YAG laser. However, this could not be fully achieved near the center of the laser spot, as three zones were identified by using microscopic observation: a melting zone next to the surface, with an arrange of needle-like martensite plates; a transformed region, in which graphite flakes remain unaltered and

the pearlite matrix transforms to austenite during heating, and then to larger martensite plates after the cooling process; and the substrate material zone, where no transformation was observed. The author also highlighted the difference between the shape of martensite plates, due to the high thermal conductivity of the nearby graphite flakes and the amount of diffused carbon into austenite. Moreover, a continuous heating transformation (CHT) diagram was described to explain these transformations as a function of temperature and heating rates.

In terms of mechanical properties, the hardening effect is mostly measured with Vickers microhardness tests. Hwang et al. [40] reported an increment in the hardness from about 300 to 800–950 HV0.1 in the heat-affected zone. This satisfied the piston rings requirement of the minimum hardness of 450 HV0.1 in an effective depth of 300 μm. A tempering effect was also identified in overlapping zones, where hardness dropped to 470 HV0.1, still fulfilling the desired condition.

These results are in agreemen<sup>t</sup> with the work of Liu and Previtali [41]. A region of maximum hardness between 800 and 900 HV0.3 was generated in the non-overlapping area, while in the overlapping region, there was a tempering effect which reduced the hardness to 300–400 HV0.3 (Figure 4). Since overlapping is used to create a homogeneous modified surface in the case of a Gaussian laser, this unwanted effect is unavoidable. In this sense, the authors determined that an overlapping degree between 1 and 1.25 mm was acceptable to comply with industrial requirements, and this, based on their research, equaled a transformed zone with a minimum depth of 0.25 mm and hardness greater than 700 HV0.3. Wang et al. [42,43] also obtained similar hardness profiles with a Rockwell test (HRC), since the 2 mm spacing between adjacent laser tracks ensured avoiding the back-tempering effect.

**Figure 4.** Microhardness profile as a function of overlapping size (Reprinted with permission from ref. [41]. 2010 Elsevier).

In the case of LSM, the improvement in hardness is comparable to LSH, but with a different magnitude. Trafford et al. [44] measured a maximum hardness of 850 HV0.3 for laser surface hardening without melting, and 950 HV0.3 for the LSM treatment. These values were reached in regions close to the surface, where the transformation is homogeneous, while hardness rapidly decays to the base value when the heat-affected zone (HAZ) is exceeded. De Oliveira [45] found a similar trend but with higher dispersion. The average hardness increased from 250 to about 600–900 HV, and the high variation was attributed to the randomness of the indentation location. Moreover, the higher hardness values were related to a higher content of martensite, as proved by X-ray diffraction (XRD) tests.

Concerning the improvement in wear resistance of GIs, Hwang et al. [40] first applied a pin-on-disc test under ASTM G99 conditions to evaluate the wear loss in both the pin (using LSH and untreated GI samples) and a conventional cast-iron disc. It was determined that LSH can double the wear life of the piston rings, which was supported with SEM observation to address the nature of the wear mechanisms. In untreated piston rings, adhesive wear features, such as plastic flow and tearing damage, were noted, while in

laser-modified samples, only mild wear was observed around the remaining graphite flakes that acted as stress raisers.

Instead, Wang et al. [42,43] conducted ball-on-plate reciprocating sliding wear tests with a hard 4 mm–diameter alumina ball (45 HRC and surface roughness of 10 nm). Both laser-hardened specimens exhibited significantly lower mass loss, since they bared harder surfaces, with LHAGI displaying better performance (Figure 5). The worn surface of lasertreated cast irons revealed a polishing effect of the ceramic ball, while the severe damage in untreated specimens was related to the stress concentration around graphite flakes and the low fracture toughness of tempered martensite (zone 3), as in Hwang et al. [40]. The main wear mechanism was crack formation (Figure 6), which surged around the edge of graphite flakes on the surface or subsurface. Then, these cracks propagated along with the graphite flakes, conceiving either small-scale pits or large-scale spalls. From these observations, the authors suggested the possibility of using LHQTGI as a replacement for AGI, since its overall performance was enhanced, and the cost and energy consumption could be substantially reduced.

**Figure 5.** Mass loss of GI specimens after complete sliding wear test (Reprinted from ref. [42]).

**Figure 6.** Worn surfaces of LHAGI and LHQTGI samples with different tempering temperatures: (**a**) 232 ◦C and (**b**) 316 ◦C (Reprinted from ref. [42]).

Trafford et al. [44] also compared the enhancement in wear resistance by LSH and LSM, using friction wear tests with a silicon carbide (SiC) platform, at a frequency of 2.5 Hz and a load of 5 kg, without lubrication. Wear rates, expressed as the amount of lost volume as a function of the total sliding distance and applied load, revealed that both LSH- and LSM-treated samples had higher resistance compared to the as-cast state, with less wear damage in the ledeburitic structure obtained by LSM.

As a summarizing work, Paczkowska [46] performed LST treatments by using various sets of laser powers and scanning velocities to determine the ranges of the surface energy density and interaction time that allowed the researcher to achieve different applications of the laser, such as tempering, hardening from the solid-state (LSH), fusion (LSM), or alloying. From the microscopic observation, as well as the measurements of temperature and hardness on the surface of the samples, the author defined the desired ranges based on a continuous wave CO2 laser, as reported in Figure 7, highlighting how limited the processing windows are and the importance of finding the optimal laser parameters for each case.

**Figure 7.** Energy density ranges for different heat treatments with CO2 lasers on GI (Reprinted with permission from ref. [46]. 2016 Elsevier).

### **5. Laser Surface Treatment of Ductile Irons**

LST on DIs has been also studied over the last decades to a similar extent, sharing the same transition between CO2 to solid-state lasers, as well as most of the characterization tools with gray irons. However, the results indicate a strong dependence on the initial metallic matrix (e.g., pearlite or ferrite), which is responsible for the magnitude of diffusion times and cooling rates that play a major role in achieving LSH or LSM for a fixed laser input. Moreover, as Steen and Mazumder [39] proposed, many authors have verified the difficulties of achieving LSH on DIs due to the lowering of the melting point around the graphite nodules as carbon diffuses away from the graphite during the process.
