Enhanced Hardness-Toughness Balance Induced by Adaptive Adjustment of the Matrix Microstructure in In Situ Composites

With the development of high-speed and heavy-haul railway transportation, the surface failure of rail turnouts has become increasingly severe due to insufficient high hardness-toughness combination. In this work, in situ bainite steel matrix composites with WC primary reinforcement were fabricated via direct laser deposition (DLD). With the increased primary reinforcement content, the adaptive adjustments of the matrix microstructure and in situ reinforcement were obtained at the same time. Furthermore, the dependence of the adaptive adjustment of the composite microstructure on the composites’ balance of hardness and impact toughness was evaluated. During DLD, the laser induces an interaction among the primary composite powders, which leads to obvious changes in the phase composition and morphology of the composites. With the increased WC primary reinforcement content, the dominant sheaves of the lath-like bainite and the few island-like retained austenite are changed into needle-like lower bainite and plenty of block-like retained austenite in the matrix, and the final reinforcement of Fe3W3C and WC is obtained. In addition, with the increased primary reinforcement content, the microhardness of the bainite steel matrix composites increases remarkably, but the impact toughness decreases. However, compared with conventional metal matrix composites, the in situ bainite steel matrix composites manufactured via DLD possess a much better hardness-toughness balance, which can be attributed to the adaptive adjustment of the matrix microstructure. This work provides a new insight into obtaining new materials with a good combination of hardness and toughness.


Introduction
With the development of high-speed and heavy-haul railway transportation, rollingcontact fatigue crack and peeling on the surface of rail turnouts have become increasingly severe [1][2][3]. The surface failure of rail turnouts severely reduces their service life, which leads to increased operation costs and potential safety hazards [4]. It is well known that the failure of rail turnouts is closely related to the insufficient hardness and toughness of the components [5]. Hence, it is of great importance to develop a rail turnout material with high hardness and high toughness.
Compared with the conventional surface treatment methods (such as thermal spraying, plasma spraying), laser surface-treatment technologies demonstrate obvious advantages, such as a small heat effect zone, good interface bonding, high reliability and high provides new insights into obtaining new materials with a good combination of hardness and toughness.

Materials and Methods
The bainite steel matrix composite was fabricated via DLD on a U75V steel substrate (a kind of railway material). The surface of the substrate was ground and then sand blasted in order to remove the surface oxide layer. The gas-atomized Fe-based alloyed powder with a particle size of 50~70 µm in diameter was applied to construct the matrix of the composite. The chemical compositions (in wt%) of the substrate and Fe-based alloyed powder are shown in Table 1. In order to avoid the sputtering of the tungsten carbide ceramic particle from the composite powder by the laser during laser deposition, tungsten carbide coated with a Co layer was used as a primary reinforcement. The composite powder containing the Fe-based powder and WC powder were thoroughly mixed using a planetary ball mill in an argon atmosphere at a speed of 200 rpm for 2 h. Finally, the composite powders were dried in a vacuum furnace at 80 • C for 2 h. The bainite steel matrix composites were manufactured using a laser processing system (as shown in Figure 1) comprising a semiconductor laser device with a maximum output power of 2.5 kW (LDM-2500-60, Laserline, Mülheim-Kärlich, Germany), a three-axis numerical control machine controlling the laser scanning path, a powder coaxial nozzle feeding system with a shielding gas device and a stable temperature platform. The process involved the following 3 steps. Firstly, the substrates were heated to 300 ± 5 • C in the resistance furnace using argon protection and then placed on a platform with the pre-set heated temperature of 300 • C to avoid martensite transformation during laser deposition. Afterwards, the composite powder was deposited on the surface of the substrates using DLD technology. The processing parameters were as follows: laser power of 800 W, laser spot diameter of 1.5 ± 0.1 mm, overlap ratio of 40% and scanning velocity of 360 mm/min. As shown in Figure 1b, the samples had a good surface quality and no macroscopic cracks were observed. Finally, the composite samples were put into a 300 ± 3 • C salt bath for isothermal treatment for 200 min, and the final composites, air cooled to room temperature (RT), were obtained. Specimens were sectioned using electric discharge wire cutting to obtain the composite samples and characterize their microstructure and mechanical properties.
An optical microscope (OM, Carl Zeiss Jena Axio Vert.A1) and a field-emission scanning electron microscope (FESEM, Nova Nano SEM450) were used to characterize the microstructural features. A backscattered electron (BSE) mode of FESEM was used to distinguish the reinforcements and steel matrix. The composition of the samples was analyzed using an energy-dispersive X-ray spectrometer (EDS) equipped on the FESEM. X-ray diffraction (XRD, D8 Advance) analyses with a Cu target were conducted for phase identification.
The microhardness of the samples was measured using a Vickers microhardness tester (Duramin-40, Struers, Denmark), with a 200 g load and a 10 s dwell time. Charpy U-notched impact tests were conducted with 55 mm × 10 mm × 10 mm samples on a pendulum impact machine (PTMS4300, Suns, China) at the RT. The notch was prepared perpendicular to the laser deposition direction. The reported impact toughness of each sample was averaged from three independent tests. The fracture surfaces of the impact samples were observed via FESEM. The microhardness of the samples was measured using a Vickers microhardnes tester (Duramin-40, Struers, Denmark), with a 200 g load and a 10 s dwell time. Charp U-notched impact tests were conducted with 55 mm × 10 mm × 10 mm samples on a pen dulum impact machine (PTMS4300, Suns, China) at the RT. The notch was prepared pe pendicular to the laser deposition direction. The reported impact toughness of each sam ple was averaged from three independent tests. The fracture surfaces of the impact sam ples were observed via FESEM. Figure 2 illustrates the phase constituents and their relative contents in the bainit steel matrix composite. As shown in Figure 2a, the matrix of the composite is change from a mainly ferrite phase (α-Fe) with a little austenite phase (γ-Fe) to α-Fe with a con siderable amount of γ-Fe when WC primary reinforcements are added. Furthermore, Fe3W3C phase appears instead of a WC phase when the addition of primary reinforcemen is relatively low. The WC phase presents together with the Fe3W3C phase when the pr mary reinforcement exceeds 15 vol%. The volume fraction of different phases was furthe quantified, as shown in Figure 2b. In the case of peak overlapping (the magnified image in Figure 2a), a Pearson VII function was used for the peak separation and fitting [15,40 42]. As the volume fraction of the WC primary reinforcement is increased, the volum fraction of the α-Fe phase declines; however, that of the γ-Fe phase increases. The volum fraction of the γ-Fe phase is not higher than that of the α-Fe phase until the WC primar reinforcement content reaches 20 vol%. As for the carbides, the volume fraction of Fe3W3 increases approximately linearly with the increased primary reinforcement content, an its maximum value is about 14 vol%. However, with the increased WC primary reinforce ment content, the final WC content of the composites is maintained at zero when the con tent is less than 10%; meanwhile, when the content of primary reinforcement is highe than 10 vol%, the volume fraction of the final WC rises. When the content of primary re inforcement is 20%, the final WC reinforcement content in the composites is 4.1 vol%.  Figure 2 illustrates the phase constituents and their relative contents in the bainite steel matrix composite. As shown in Figure 2a, the matrix of the composite is changed from a mainly ferrite phase (α-Fe) with a little austenite phase (γ-Fe) to α-Fe with a considerable amount of γ-Fe when WC primary reinforcements are added. Furthermore, a Fe 3 W 3 C phase appears instead of a WC phase when the addition of primary reinforcement is relatively low. The WC phase presents together with the Fe 3 W 3 C phase when the primary reinforcement exceeds 15 vol%. The volume fraction of different phases was further quantified, as shown in Figure 2b. In the case of peak overlapping (the magnified images in Figure 2a), a Pearson VII function was used for the peak separation and fitting [15,[40][41][42]. As the volume fraction of the WC primary reinforcement is increased, the volume fraction of the α-Fe phase declines; however, that of the γ-Fe phase increases. The volume fraction of the γ-Fe phase is not higher than that of the α-Fe phase until the WC primary reinforcement content reaches 20 vol%. As for the carbides, the volume fraction of Fe 3 W 3 C increases approximately linearly with the increased primary reinforcement content, and its maximum value is about 14 vol%. However, with the increased WC primary reinforcement content, the final WC content of the composites is maintained at zero when the content is less than 10%; meanwhile, when the content of primary reinforcement is higher than 10 vol%, the volume fraction of the final WC rises. When the content of primary reinforcement is 20%, the final WC reinforcement content in the composites is 4.1 vol%. Figure 3 presents the optical micrographs of the bainite steel matrix composites with different volume fractions of WC. For the bainite steel without WC, the morphology of prior austenite grains can hardly be recognized. However, the prior austenite grains in the bainite steel matrix composites show a typical dendritic shape. Meanwhile, both the primary dendrite arm spacing (PDAS) and secondary dendrite arm spacing (SDAP) decrease with the increased volume fraction of WC. When the WC primary reinforcement content is higher than 15 vol%, plenty of white undissolved particles can be observed. The average diameter of the particles is about 47 µm, which is a little smaller than the average particle size of WC powder (about 65 µm). With the increased WC volume fraction, the bainite morphology changes from a lath shape to a needle shape and the content of blocklike retained austenite (RA) increases, which is consistent with the XRD results ( Figure 2b). Furthermore, the black network-like microstructure begins to emerge in the interdendritic region when the WC primary reinforcement volume fraction is higher than 10%.  Figure 3 presents the optical micrographs of the bainite steel matrix com different volume fractions of WC. For the bainite steel without WC, the m prior austenite grains can hardly be recognized. However, the prior austenite bainite steel matrix composites show a typical dendritic shape. Meanwhile, mary dendrite arm spacing (PDAS) and secondary dendrite arm spacing (SD with the increased volume fraction of WC. When the WC primary reinforce is higher than 15 vol%, plenty of white undissolved particles can be observed diameter of the particles is about 47 μm, which is a little smaller than the av size of WC powder (about 65 μm). With the increased WC volume fractio morphology changes from a lath shape to a needle shape and the content retained austenite (RA) increases, which is consistent with the XRD result Furthermore, the black network-like microstructure begins to emerge in the i region when the WC primary reinforcement volume fraction is higher than

Microstructure
The fine microstructure of the bainite steel matrix composites was fu gated using the BSE mode of SEM, as shown in Figure 4. As for the bainite WC addition (Figure 4a), the bainite steel mainly consists of sheaves of lath- The fine microstructure of the bainite steel matrix composites was further investigated using the BSE mode of SEM, as shown in Figure 4. As for the bainite steel with no WC addition (Figure 4a), the bainite steel mainly consists of sheaves of lath-like bainite, a few granular bainite (GB) and island-like RA. With the increased WC reinforcement volume fraction, the lath-like bainite and GB transform into black needle-like lower bainite (LB), and the length and width of the LB needles decrease gradually (Figure 4b-e). Meanwhile, the morphology of RA also changes from an island-like to block-like shape. For the bainite steel matrix composite with a relatively high volume fraction of WC (no less than 10 vol%), the white fish-bone-shaped microstructure appears at the boundary of the prior austenite grains (Figure 4d,e). With the increased addition of WC, the area of the intergranular region occupied by the white fish-bone-shaped microstructure increases; at the same time, the prior austenite grain is refined. 10 vol%), the white fish-bone-shaped microstructure appears at the boundary of the prior austenite grains (Figure 4d,e). With the increased addition of WC, the area of the intergranular region occupied by the white fish-bone-shaped microstructure increases; at the same time, the prior austenite grain is refined.  According to the phase constituent and the microstructure of the bainite steel matrix composite with 15 vol% WC particles, EDS analysis was conducted to further identify the phase composition of the intergranular region and undissolved particles. As shown in Figure 5a, the elemental maps of the intergranular region indicate that W enriches the white fish-bone-shaped microstructure. Combining the volume fraction of the white fish-bone-shaped microstructure obtained from the SEM images with the phase analysis results from the XRD patterns, the white fish-bone-shaped phase is Fe 3 W 3 C. The elemental distributions of the partial dissolved particles and the undissolved particles are presented in Figure 5b. Much W and little Fe can be detected in the partially dissolved WC particles region. Meanwhile, in the undissolved WC particles region, the enrichment degree of W in the interior of the particles is much higher than that of the partially dissolved particles, and no Fe is detected.  According to the phase constituent and the microstructure of the bainite steel matrix composite with 15 vol% WC particles, EDS analysis was conducted to further identify the phase composition of the intergranular region and undissolved particles. As shown in Figure 5a, the elemental maps of the intergranular region indicate that W enriches the white fish-bone-shaped microstructure. Combining the volume fraction of the white fishbone-shaped microstructure obtained from the SEM images with the phase analysis results from the XRD patterns, the white fish-bone-shaped phase is Fe3W3C. The elemental distributions of the partial dissolved particles and the undissolved particles are presented in Figure 5b. Much W and little Fe can be detected in the partially dissolved WC particles region. Meanwhile, in the undissolved WC particles region, the enrichment degree of W in the interior of the particles is much higher than that of the partially dissolved particles, and no Fe is detected.

Mechanical Properties
As shown in Figure 6, the microhardness of the bainite steel is about 330 HV0 which is lower than that of the U75V steel substrate (375 HV0.2). The microhardness the DLD manufactured bainite steel matrix composites is much higher than that of t

Mechanical Properties
As shown in Figure 6, the microhardness of the bainite steel is about 330 HV0.2, which is lower than that of the U75V steel substrate (375 HV0.2). The microhardness of the DLD manufactured bainite steel matrix composites is much higher than that of the U75V steel substrate. With the increased primary reinforcement content, the microhardness of bainite steel matrix composites increases remarkably. The microhardness increases rapidly to 461 HV0.2 when the WC primary reinforcement volume fraction is only 5 vol%, which is approximately 40% higher than that of the bainite steel. Moreover, the microhardness of composites with 20 vol% primary reinforcement is increased to 561 HV0.2. In contrast, the impact toughness of the bainite steel matrix composite decreases with the increased volume fraction of WC primary reinforcement ( Figure 6). However, the impact toughness of the composite when the primary reinforcement volume fraction is less than 10% is still higher than that of the U75V steel substrate (26 J), which can satisfy the demand of the rail turnout service.   Figure 7 indicates the impact fracture surface morphology of the U75V steel substrate and bainite steel matrix composites with different volume fractions of WC primary reinforcement. The U75V steel substrate shows a typical feature of cleavage fracture, which consists of cleavage steps and river patterns (Figure 7a). In contrast, dimples and tearing ridges are observed in the fracture of the bainite steel (Figure 7b), which indicates that the facture mechanism occurring is microvoids coalescence ductile fracture. Figure 7c presents the fracture of the bainite steel matrix composite with 5 vol% primary reinforcement. Both the features of ductile fracture (dimples and tearing ridges) and cleavage fracture (cleavage steps and river patterns) are evident on the fracture surface. This suggests that the fracture mechanism occurring is quasi-cleavage fracture in the composites. The crystal sugar fracture (Figure 7d) indicates that the bainite steel matrix composite with 15 vol% primary reinforcement follows the brittle intergranular fracture mechanism.

Formation Mechanism of the Reinforcement during Laser Deposition
Information about the phases in the bainite steel matrix composite that was revealed by XRD ( Figure 2) indicates that the in situ Fe 3 W 3 C reinforcement is the main reinforcement in the composite when the volume fraction of primary reinforcement WC is not higher than 10%. When the primary reinforcement volume fraction increases from 10% to 20%, the final reinforcement in the composites includes Fe 3 W 3 C in situ reinforcement and WC primary reinforcement. Moreover, the content of Fe 3 W 3 C is much higher than that of WC. Therefore, it can be deduced that the decomposed WC during laser deposition has dissolved in the bainite steel matrix and participated in the formation of Fe 3 W 3 C in the matrix. This is proved by the morphology of the Fe 3 W 3 C phase adjacent to the WC particles ( Figure 8). The Fe 3 W 3 C phase in the vicinity of the partially dissolved WC particle presents the feature of a continuously fish-bone-shaped microstructure. The content of the Fe 3 W 3 C phase decreases with the increased distance from the partially dissolved WC particles. In contrast, compared with the Fe 3 W 3 C phase near the partially dissolved WC particle, the Fe 3 W 3 C phase near the undissolved WC particle has a much lower content and the carbides are distributed more uniformly.

Formation Mechanism of the Reinforcement during Laser Deposition
Information about the phases in the bainite steel matrix composite that was revealed by XRD ( Figure 2) indicates that the in situ Fe3W3C reinforcement is the main reinforcement in the composite when the volume fraction of primary reinforcement WC is not higher than 10%. When the primary reinforcement volume fraction increases from 10% to 20%, the final reinforcement in the composites includes Fe3W3C in situ reinforcement and WC primary reinforcement. Moreover, the content of Fe3W3C is much higher than that of WC. Therefore, it can be deduced that the decomposed WC during laser deposition has dissolved in the bainite steel matrix and participated in the formation of Fe3W3C in the matrix. This is proved by the morphology of the Fe3W3C phase adjacent to the WC particles ( Figure 8). The Fe3W3C phase in the vicinity of the partially dissolved WC particle presents the feature of a continuously fish-bone-shaped microstructure. The content of the Fe3W3C phase decreases with the increased distance from the partially dissolved WC particles. In contrast, compared with the Fe3W3C phase near the partially dissolved WC particle, the Fe3W3C phase near the undissolved WC particle has a much lower content and the carbides are distributed more uniformly. For the preparation of WC reinforced Fe-based matrix composites, the temperature of the material during direct laser deposition is always higher than that during traditional powder metallurgy and casting process [43][44][45]. Under the irradiation of a laser beam with an extremely high-energy density, the maximum temperature of the metal molten pool can exceed 3000 °C, which is higher than the decomposition temperature of WC (1250 °C) [46]. Then, the WC primary reinforcement will react with the steel matrix. The reaction is as follows [47,48]: Additionally, the high temperature gradient of the molten pool during the DLD process brings the obvious convection of liquid metal, which is beneficial to a successful reaction [49]. When the addition of WC is low (less than 10 vol% in this work), all WC particles participate in the formation of Fe3W3C in the bainite steel matrix. In contrast, when WC is excessive, such as the 15 vol% WC in this work, both the Fe3W3C in situ reinforcement and WC primary reinforcement exist in the matrix at the same time. Free W atoms, which join to form Fe3W3C, come from the dissolved WC particle. Therefore, the concentration of W atoms around the partially dissolved WC particles is higher than that of the matrix far from the WC particles. Accordingly, the in situ Fe3W3C content near the partially dissolved WC particles is high.

Mechanism of the Adaptive Adjustment of the Matrix Microstructure with Primary Reinforcement
With the increased volume fraction of WC primary reinforcement, not only the constituent and morphology of the final reinforcement in the composites are changed, but also the microstructure of the bainite steel matrix is significantly altered (Figures 3 and 4), which is attributed to the change in the solute W and C content in the bainite steel matrix. Figure 9 presents the constituent of C and W derived from the WC primary reinforcement. It can be concluded that the formation of the Fe3W3C phase and WC phase does not con- For the preparation of WC reinforced Fe-based matrix composites, the temperature of the material during direct laser deposition is always higher than that during traditional powder metallurgy and casting process [43][44][45]. Under the irradiation of a laser beam with an extremely high-energy density, the maximum temperature of the metal molten pool can exceed 3000 • C, which is higher than the decomposition temperature of WC (1250 • C) [46]. Then, the WC primary reinforcement will react with the steel matrix. The reaction is as follows [47,48]: Additionally, the high temperature gradient of the molten pool during the DLD process brings the obvious convection of liquid metal, which is beneficial to a successful reaction [49]. When the addition of WC is low (less than 10 vol% in this work), all WC particles participate in the formation of Fe 3 W 3 C in the bainite steel matrix. In contrast, when WC is excessive, such as the 15 vol% WC in this work, both the Fe 3 W 3 C in situ reinforcement and WC primary reinforcement exist in the matrix at the same time. Free W atoms, which join to form Fe 3 W 3 C, come from the dissolved WC particle. Therefore, the concentration of W atoms around the partially dissolved WC particles is higher than that of the matrix far from the WC particles. Accordingly, the in situ Fe 3 W 3 C content near the partially dissolved WC particles is high.

Mechanism of the Adaptive Adjustment of the Matrix Microstructure with Primary Reinforcement
With the increased volume fraction of WC primary reinforcement, not only the constituent and morphology of the final reinforcement in the composites are changed, but also the microstructure of the bainite steel matrix is significantly altered (Figures 3 and 4), which is attributed to the change in the solute W and C content in the bainite steel matrix. Figure 9 presents the constituent of C and W derived from the WC primary reinforcement. It can be concluded that the formation of the Fe 3 W 3 C phase and WC phase does not consume all the C and W derived from the WC primary reinforcement and the remaining C and W dissolve in the bainite steel matrix, which significantly retards the transformation of undercooled austenite [50,51].
Materials 2023, 16, x FOR PEER REVIEW Figure 9. Constituent of C and W derived from primary reinforcement in the comp JMatPro software version 7.0 was applied to investigate the austen transformation kinetics of the matrix of the composite using the general ( Figure 10). For the bainite steel (Figure 10a), the nose temperature of the formation is 607 °C, with an incubation period of about 3 min; the nose tem bainite transformation is 397 °C, with an incubation period of about 0.5 mi the martensite transformation starts at 247 °C . Owing to the rapid cooling r JMatPro software version 7.0 was applied to investigate the austenite isothermal transformation kinetics of the matrix of the composite using the general steel database ( Figure 10). For the bainite steel (Figure 10a), the nose temperature of the pearlite transformation is 607 • C, with an incubation period of about 3 min; the nose temperature of the bainite transformation is 397 • C, with an incubation period of about 0.5 min. Meanwhile, the martensite transformation starts at 247 • C. Owing to the rapid cooling rate of the DLD process (10 3~1 0 4 • C/s), the pearlite transformation is completely prevented during the DLD process. In order to obtain the lower bainite microstructure, the transformation of martensite must be avoided the during cooling process and the following isothermal temperature has to be lower than the nose temperature of the bainite transformation. Hence, the preheating and isothermal temperature is set as 300 • C. According to the bainite transformation time at 300 • C, the isothermal treatment time was set to 200 min in this work. Figure 9. Constituent of C and W derived from primary reinforcement in the composites.
JMatPro software version 7.0 was applied to investigate the austenite isotherma transformation kinetics of the matrix of the composite using the general steel databas (Figure 10). For the bainite steel (Figure 10a), the nose temperature of the pearlite trans formation is 607 °C, with an incubation period of about 3 min; the nose temperature of th bainite transformation is 397 °C, with an incubation period of about 0.5 min. Meanwhile the martensite transformation starts at 247 °C . Owing to the rapid cooling rate of the DLD process (10 3~1 0 4 °C /s), the pearlite transformation is completely prevented during th DLD process. In order to obtain the lower bainite microstructure, the transformation o martensite must be avoided the during cooling process and the following isothermal tem perature has to be lower than the nose temperature of the bainite transformation. Hence the preheating and isothermal temperature is set as 300 °C . According to the bainite trans formation time at 300 °C , the isothermal treatment time was set to 200 min in this work. The nose temperature of the bainite transformation and the martensite transfor mation temperature of the bainite steel matrix in the composite were also calculated using JMatPro software, as well as the bainite transformation time (Figure 10b). With the in creased volume fraction of primary reinforcement in the composite, both the nose tem perature of bainite transformation and the martensite transformation temperature de crease. Furthermore, both the incubation period and completion time of the bainite trans formation increase significantly when the primary reinforcement content in the composit is increased. The lowest bainite transformation nose temperature of the composite matrix The nose temperature of the bainite transformation and the martensite transformation temperature of the bainite steel matrix in the composite were also calculated using JMatPro software, as well as the bainite transformation time (Figure 10b). With the increased volume fraction of primary reinforcement in the composite, both the nose temperature of bainite transformation and the martensite transformation temperature decrease. Furthermore, both the incubation period and completion time of the bainite transformation increase significantly when the primary reinforcement content in the composite is increased. The lowest bainite transformation nose temperature of the composite matrix is about 300 • C, which ensures that the bainite structure obtained via isothermal treatment in the matrix is needle-like lower bainite. When WC primary reinforcement is added into the bainite steel matrix, the start temperature of martensite transformation (M s ) in the matrix is still higher than room temperature, while the finish temperature of martensite transformation (M f ) in the matrix is much lower than room temperature. Meanwhile, the time required for the matrix to complete bainite transformation is much longer than the isothermal treatment time (200 min) for the composites. As a result, a small part of the untransformed undercooled austenite transforms into high-carbon cryptocrystalline martensite, while most of the untransformed undercooled austenite maintains its original crystal structure, which leads to the formation of a large amount of RA structure in the matrix. Moreover, the content of RA increases with the increased solute W and C content in the bainite steel matrix, which has a proportional relationship with the volume fraction of primary reinforcement in the composites (Figure 2b).

Effect of the Adaptive Adjustment of the Matrix Microstructure on the Mechanical Properties of Composites
Since reinforcement particles in MMCs are usually hard and brittle, the increased volume fraction of the reinforcement particles results in an enhancement in the strength and hardness of MMCs and a reduction in the ductility and toughness of MMCs (i.e., hardness-toughness trade-off). Figure 11 presents the hardness-toughness trade-off caused by the reinforcement particles' volume fraction in MMCs [52][53][54][55][56][57][58][59]. In order to facilitate a comparison between the mechanical properties of different MMCs, the microhardness and impact toughness are converted into the increase in microhardness and the decrease in impact toughness, respectively. Because the data point is close to the top right corner, the degree of hardness-toughness trade-off in the materials is low, and the hardness-toughness balance is good. The degree of hardness-toughness trade-off in the composite in this work is lower than that of most conventional MMCs, which can be attributed to the change in the matrix microstructure induced by the increased volume fraction of primary reinforcement. In conventional MMCs, the phase constituent and morphology of the matrix are almost not changed with the increased volume fraction of reinforcement, but the matrix microstructure of the composite in this work is altered with the increased primary reinforcement content. The residual austenite content in the steel matrix increases with the increased volume fraction of WC primary reinforcement, which is beneficial to improve the toughness of the steel matrix in composites [60]. The high volume fraction of reinforcement improves the hardness of the composite, while the adaptive adjustment of the matrix microstructure offsets a part of the reduction in impact toughness caused by the increase in the reinforcement volume fraction. As a result, the bainite steel matrix composite in this work demonstrates a better balance of hardness and impact toughness compared with most conventional MMCs.
itate a comparison between the mechanical properties of differen ness and impact toughness are converted into the increase in mic crease in impact toughness, respectively. Because the data point corner, the degree of hardness-toughness trade-off in the materia ness-toughness balance is good. The degree of hardness-toughne posite in this work is lower than that of most conventional MMCs, to the change in the matrix microstructure induced by the incre primary reinforcement. In conventional MMCs, the phase constitu the matrix are almost not changed with the increased volume fr but the matrix microstructure of the composite in this work is alt primary reinforcement content. The residual austenite content in t with the increased volume fraction of WC primary reinforcemen improve the toughness of the steel matrix in composites [60]. The reinforcement improves the hardness of the composite, while the the matrix microstructure offsets a part of the reduction in impa the increase in the reinforcement volume fraction. As a result, the b posite in this work demonstrates a better balance of hardness and pared with most conventional MMCs.

Conclusions
The in situ bainite steel matrix composites with WC primary reinforcement were manufactured using direct laser deposition. The effect of the primary reinforcement volume fraction on the composite microstructure and its mechanical properties were investigated. With the increased primary reinforcement content, the adaptive adjustment of the matrix microstructure was obtained. Furthermore, the dependence of the adaptive adjustment of the matrix on the combination of hardness and impact toughness in the composites was evaluated. The main conclusions are as follows: (1) The interaction of the primary composite powder irradiated by laser during DLD leads to significant changes in the phase constituent and morphology of the reinforcement and matrix of the composites at the same time. With the increased volume fraction of primary reinforcement, the main phase in the matrix changes from dominant α-Fe to a mixture of γ-Fe and α-Fe. Specifically, with the increased primary reinforcement content, the matrix microstructure is changed from lath-like bainite, granular bainite and few islandlike retained austenite into needle-like lower bainite and plenty of block-like retained austenite, and the final reinforcement is changed from Fe 3 W 3 C into Fe 3 W 3 C and WC.
(2) The microhardness increases rapidly to 461 HV0.2 when the WC primary reinforcement volume fraction is 5 vol%, which is approximately 40% higher than that of the bainite steel. Moreover, the microhardness of composites with 20 vol% primary reinforcement is increased to 561 HV0.2. In contrast, the impact toughness of the bainite steel matrix composite decreases with the increased primary reinforcement volume fraction. However, the impact toughness of the composite when the primary reinforcement volume fraction is less than 10% is still higher than that of the U75V steel substrate, which can satisfy the demand of rail turnout service.
(3) Compared with the conventional metal matrix composites, the bainite steel matrix composites manufactured via DLD possess a much lower degree of hardness-toughness trade-off. The better combination of microhardness and impact toughness can be attributed to the adaptive adjustment of the matrix microstructure with the increased volume fraction of primary reinforcement, which provides new insights into obtaining new materials with a good combination of hardness and toughness.