Selective Laser Melting of TiC-Fe via Laser Pulse Shaping: Microstructure and Mechanical Properties

Abstract

In the present study, TiC-Fe cermets were fabricated through selective laser melting (SLM) for the first time employing pulse wave using a pulse shaping technique and regular laser pulse wave. Two samples were fabricated each with adapting pulse shaping technique and regular laser pulse wave with varied laser peak power and exposure time to obtain an optimized parameter. The pulse shaping technique proves to be an optimal method for fabrication of the TiC-Fe-based cermet. The effect of the laser peak power and pulse shaping on the microstructure development was investigated through scanning electron microscopy and X-ray diffraction analysis. Two-phased microstructures revealed the distribution of TiC and Fe. A maximum hardness and fracture toughness of 1010 ± 65 MPa and 16.3 ± 1.7 MPa m^1/2, respectively, were observed for the pulsed-shaped samples illustrating that pulse shaping can be an effective way to avoid cracking in brittle materials processed by SLM.

Introduction

Cermets are useful in applications that demand high hardness and wear resistance combined with a high toughness.^1–5 Some of their applications include high-speed metal cutting and high temperature erosive and corrosive environments.^6,7 TiC-based cermets have received renewed interest in recent years, especially bonded with iron-based alloys, where environmentally harmful and scarce metals such as W, Co, and Ni are excluded. Due to its high hardness, wear resistance, chemical, and high temperature stability, TiC proves to be a most promising alternative for conventional tungsten carbide (WC)-based cermets.^8–11

Fe-based binders are considered an alternative solution as they are abundantly available, cost-effective, and nontoxic and offer to strengthen by heat treatment. Selective laser melting (SLM) or laser-based powder bed fusion is an additive manufacturing technique based on powder bed fusion (PBF) technology that fabricates parts layer by layer by scanning the powder bed selectively and melting and fusing the subsequent layer of fine metal powder to deposit on the substrate.^12–14 The formation of melt pool during laser scan and the subsequent rapid solidification offers a wide range of advantages including refined microstructures compared with their conventional counterparts.^15–19

There is increasing demand to fabricate parts using additive manufacturing especially for rapid tooling applications due to its high density, high temperature wear resistance, hardness, and homogeneity in the microstructure.^20–22 Even though laser power has a significant influence on the processing of parts,²³ laser beam emission type (i.e., continuous or pulsed wave) plays a significant role. Laser beam may act as the heat source for melting the single powder layer, and the temporal distribution of laser power plays a vital role in achieving the density and evolution of unique microstructure in the processed parts.^24–26

Due to strong metallurgical bonding between layers and small heat-affected regions, pulsed wave lasers (PWs) have been found to be more suitable for SLM manufactured parts than continuous-wave lasers.^26–28 The process parameters influence the fabrication process and generate thermal stress and temperature gradient in the parts, and the parameters should be carefully optimized to produce parts without defects such as cracks, porosities, warpage, and delamination.^29–31 There are only a few studies on cermets fabricated using SLM, and the majority of these studies focus on the classical WC-Co-based cermets.^7,32–34

In situ TiC processing, where TiC is used as a reinforcement particle in the metal matrix, has been documented by several researchers.^35–37 However, the authors are not aware of any literature dealing with the processing of TiC-based cermets and yielding crack-free specimens. Most of these cermets are anticipated to induce cracks during the SLM process because of their inherent brittleness, which limits the processing of these materials by solidification-based additive manufacturing processes.

In this work, the pulse shaping technique is adapted as it temporarily distributes the energy within a single layer of the pulse. Pulse shaping help to control the laser material exposure by delivering controlled energy.³⁸ In particular, pulse shape and duration regulate the release of the energy onto the processed material.^38–40 By tuning the pulse with different peak powers, exposure times allow gradual heating and/or prolonged cooling effect within the powder bed that helps controlling the melt pool properties and subsequently the formation of the final part.

The thermal history of the laser emission is considerably linked with the heating and cooling rates, which influence both the microstructure and in turn their mechanical properties. The difference in the laser scanning strategy plays a vital role in the melt pool formation due to the nonuniform heating and cooling effect in the melt pool and eventually inducing differences in the microstructure.³¹ Accordingly, for the first time, this article deals with the fabrication of TiC-based cermets with Fe as binder through an additive manufacturing process by adapting a point exposure and in the y-direction, scanning strategy with different pulse shaping to avoid possible cracking in the cermet during solidification.

Experimental Details

Powder preparation

The details of both TiC and Fe powder used in the present work are furnished in Table 1. Feedstock powder (200 g) for SLM with 60 wt.% TiC and 40 wt.% Fe was prepared by mechanical mixing with the help of zirconia balls (five in number with the following dimensions: 2 mm in diameter and five in number) for 24 h at 10 rpm. Stainless steel vial with a volume of 500 mL, dimensions of 80 mm in diameter and 130 mm in height was used.⁴¹

Table 1.

The Size, Purity, and Source of the Powders Used in This Study

Material	Average particle size (μm)	Purity (%)	Source
TiC	2–3	99.9	Pacific Particulate Materials Ltd.
Fe	45	99.8	TLS Technik GmbH

Figure 1 shows the scanning electron microscopy (SEM) images of the initial and the mixed powders. It can be observed from Figure 1a that the Fe particles are spherical in shape with quite significant distribution in size. However, the TiC particles are irregular in shape with a narrow size distribution (Fig. 1b). The cermet powder after mechanical mixing shows the presence of both spherical and irregularly shaped powder particles with a very wider distribution in their size (Fig. 1c). However, they are <45 μm in size. The X-ray diffraction (XRD) pattern of the mixed cermet shows the presence of the peaks of both TiC and Fe phases (Fig. 1d). No other additional phases are observed indicating that the mechanical milling process does not form any new phases.

FIG. 1.

SEM images of the starting powders: (a) Fe and (b) TiC and mechanically mixed cermet (c) TiC-Fe mixture. (d) The X-ray diffraction pattern of the mixed cermet TiC-Fe powder. SEM, scanning electron microscopy.

Selective laser melting

The pulsed laser was used for the fabrication of TiC-Fe-based cermets using Realizer 50 SLM system equipped with Nd:YAG laser with a maximum power of 100 W. Samples with both standard and shaped pulse laser were used for fabrication. A schematic diagram showing the scanning strategy adapted for a single pulse to fabricate the samples with both types of pulses is shown in Figure 2. The focal diameter of the laser beam spot on the work piece surface is fixed at 0.0134 mm. Stainless steel substrate was used to build the samples in an argon atmosphere to protect the melt pool from oxidation. With the standard pulse laser, the scan direction was along the y-axis with a hatch distance of 0.5 μm and an exposure time of 200 μs.

FIG. 2.

Schematics illustrating the different pulse shaping used in the present study: (a) hatching along the y-axis (conventional hatching) and (b) laser point exposure. Plots showing the two different pulsing strategies used to produce the TiC-Fe cermets (c) regular pulse (used to fabricate samples A and B) and (d) pulse shaping (used to fabricate samples C and D).

Samples A and B were printed with the same exposure time and with different laser powers (60 and 72 W, respectively, Fig. 2c). However, for the samples, C and D pulse shaping technique was adapted with different peak laser powers with the employment of pulse to fabricate the cermets, as shown in Figure 2d. With the employment of pulsing, the laser exposure was targeted to an exposure point and the same pulse is repeated to each laser exposure point. The samples were built in the vertical direction with a layer thickness of 0.035 mm and a point distance of 0.5 μm. Table 2 represents the SLM process parameter adapted for the fabrication of cermets.

Table 2.

The Selective Laser Melting Process Parameters Used in This Study to Build the TiC-Fe Cermets

Sample designation	Pulse type	Laser power (W)	Exposure time (μs)
A	Standard pulse	60	200
B	Standard pulse	72	200
C	Pulse shaping	12-24-60-19.2-14.4	100-200-200-500-200
D	Pulse shaping	12-24-72-19.2-14.4	100-200-200-500-200

Phase identification and microstructure observation

The printed cermets were ground and polished for the microstructural characterization. The longitudinal section of the cermets was examined by SEM Zeiss EVO MA15 equipped with the X-ray spectroscopy (EDS) system INCA for microstructural and elemental composition analysis, respectively. The structural characterization (in terms of phase identification) was carried out by the Rigaku SmartLab X-Ray Diffractometer.

Thermo-Calc^® software with TCFE9 database was used to study the possible phase formation during the SLM process (Fig. 3a). Since Ti, C, and Fe are the elements involved, the following observations are made using Thermo-Calc: TiC phase is stable over a wide range of compositions. No new intermetallic phase can be observed. Iron titanium solid solution can be seen along with graphite formation at lower temperatures. The volume of the phases present and particle size were measured from the SEM images using the ImageJ software.

FIG. 3.

(a) Ti-C-Fe phase diagram calculated using Thermo-Calc^® software and TCFE9 database. (b) The XRD patterns of the SLM fabricated cermets fabricated as function of laser pulsing. SLM, selective laser melting; XRD, X-ray diffraction.

Mechanical characterization

Cuboidal shaped TiC-Fe samples with near theoretical density were fabricated through SLM with different pulse shaping parameters. Vickers Hardness tests (longitudinal section) were performed using a microhardness testing machine Indentec 5030KV (Indentec Hardness Testing Machines Limited, West Midlands, UK) with a 30 kg load and dwell time of 10 s. The fracture toughness values of the samples were measured through the indentation method with the same load as used for the hardness test. The following equation⁴² was used to calculate the fracture toughness.

$$K_{IC}=0.15\sqrt{\frac{H{V}_{30}}{\sum l}},$$ (1)

where K_IC is the fracture toughness, $\sum l$ is the sum of crack lengths of each indentation in mm, and HV is the hardness value in N·mm⁻². All the reported mechanical testing results are an average of three samples.

Results and Discussion

Phase identification

Figure 3b shows the XRD results obtained from the longitudinal section of the SLM fabricated cermets. The patterns reveal the presence of high intense sharp peaks corresponding to the following phases: TiC and α-Fe, where TiC belongs to a simple cubic system with the space group Fm-3m (225) (line NaCl) and α-Fe belongs to body-centered cubic system (bcc) with a space group Im-3m (229). No intermetallic and carbide phases appear in the diffraction peak. It may be safe to say that laser pulsing did not have a significant influence on phase formation.

Microstructure

Figure 4 shows the surface morphologies of the SLM processed samples through (regular) conventional pulse mode without shaping but as a function of energy density. It shows the presence of the typical laser scan tracks existing on the surface of the sample. Two distinct zones with different TiC phases can be observed in the microstructure (Fig. 4b).

FIG. 4.

SEM images (longitudinal section) of the as-built cermets through regular pulsing but with varying energy densities: (a–c) sample A, E_v 137 J·mm⁻³; (d–f) sample B, E_v 165 J·mm⁻³.

The presence of unmelted TiC particles in the melt pool due to the inhomogeneous powder distribution in the feedstock powder and its poor flowability can be found as near-spherical particles with light gray areas. The dark gray area represents TiC dendritic phase after melting/dissolution. However, the bright areas correspond to the bcc α-Fe phase (Fig. 4b, c). The scan speed for samples A and B was kept constant at 250 mm·s⁻¹. However, the change in the laser power varies their energy density. The equation used to calculate the volumetric energy density is as follows⁴³:

$$E_v=\frac{P\cdot ET}{PD\cdot HS\cdot h},$$ (2)

where P represents the laser power in Watt, ET represents exposure time in mm·s⁻¹, PD represents point distance in mm, HS represents hatch space in mm, and h represents layer thickness in mm. The energy density for samples A and B was found to be 137 and 165 J·mm⁻³, respectively. In the melt pool, the dendritic TiC phase can be found due to melting and subsequent solidification. With increasing energy density, relatively finer dendritic TiC is observed (Fig. 4d–f).

Coarser TiC phase can be observed along the metal pool boundaries, where minimum temperature gradient (G) is observed due to the Gaussian distribution of laser energy (where the maximum temperature gradient is observed at the center of the melt pool and minimum value at its boundaries). The change in the grain size from the melt pool to the boundary can be explained using a thermal gradient (G), solidification rate (R), and cooling rate (T) by fineness factor (T = G × R) and shape factor (G/R).^40–47 The maximum temperature in the melt pool for the Gaussian distributed laser source is proportional to the $P∕\surd v$ ratio.⁴⁸

$$T_{max}\propto \frac{P}{\surd v},$$ (3)

where P is the laser power in Watt and v is the scanning speed in mm·s⁻¹. In the present study, the scanning speed is kept constant; however, an increase in the laser power increases the temperature of the melt pool, which means higher G, and apparently higher G/R ratio resulting in a finer microstructure. However, surrounding the melt pool boundaries, these G/R ratios will be low due to the Gaussian distribution of laser energy, and hence, coarser equiaxed grains were formed.

On the surface of as-built samples, cracks can be observed perpendicular to the direction of the melt pool/laser tracks. The large thermal residual stress, which is inherent to SLM samples in the as-built condition due to rapid melting and cooling rates under a high-intensity laser input, leads to crack formation.^29,49–51 The length and number of cracks increase with the increase in the cooling rate and increasing cooling rate increase the degree of residual stresses.^52–54 These multidirectional cracks initiated from the sample's edge and spread across its entire surface.

Due to surface and internal defects in as-built samples, the fabrication of TiC-based cermet using regular pulsing without shaping was not considered optimal. The exhibition of cracks hampers the mechanical performance of the samples and shortens the lifetime during constant or cyclic loadings and leads to premature failure of the as-built cermet parts. Consequently, it is very important to suppress these cracks and produce dense parts for real-time applications. Therefore, for the first time, pulse-shaping technology is adapted for the fabrication of TiC-Fe-based cermets to control the energy distribution of the laser power during fabrication, and significantly, these distributions in energy affect the melting and cooling rate of the process.

Figure 5 shows the characteristic morphology of the TiC and Fe phase fabricated with the application of pulse shaping using SLM. It can be noticed that changes in the laser power significantly influence the morphology of the TiC phase, similar to the counterpart without laser shaping. Using a laser power of 60 W (Fig. 5a, b), columnar dendritic microstructure can be observed near the zone of laser exposure surrounded by coarse TiC phase. There exists a gradient microstructure in terms of TiC size as we move from the core of the melt pool toward the overlapping boundaries, which may be attributed to the laser shape (pulse shaping).

FIG. 5.

SEM images (longitudinal section) of the as-built cermets with pulse shaping showing different morphologies at different laser powers: (a, b) sample C, 60 W laser power; (c, d) sample D, 72 W laser power.

However, increasing the laser power to 72 W, a high cooling/solidification rate will be observed, leading to finer microstructure similar to the ones observed for the samples A and B fabricated without pulse shaping. Samples produced with 72 W laser power shows more uniform finer dendrites along with the core of the melt pool and even in the overlapping regions; coarse TiC particles (significantly reduced in size compared with the samples fabricated with 60 W laser power) are observed. A gradient in the size of the TiC dendrites is observed moving from the center of the melt pool toward the overlapping regions due to the inherent laser beam geometry (Fig. 5). This might also be attributed to the difference in the solidification rates observed along the different sides of the melt pool.^55–57

The most interesting aspect to be noticed is that with the introduction of pulse shaping, the SLM fabricated cermets do not show signs of cracking, and in fact, no visible cracks are observed in these microstructures (Fig. 4). In addition, no unmelted TiC particles are present suggesting a complete melting of the initial powder particles. The results suggest that the fabrication of cermets through the pulse shaping technique can eliminate the formation of cracks and may be considered an optimal way to fabricate TiC-Fe-based cermets.

Figure 6 shows the EDS elemental maps of the longitudinal section of the printed sample, where the elemental distribution of Ti, C, and Fe can be observed. It may be observed from Figure 6 that the SLM samples show the nonuniform distribution of phases in terms of their sizes, where the presence of both fine and coarse features are observed, typical for SLM samples.¹²

FIG. 6.

Energy-dispersive spectroscopy mapping of the cermet samples fabricated through pulse shaping: (a) sample C, (b) sample D.

Mechanical properties

The mechanical properties of the pulse-shaped as-built SLM samples were investigated via hardness and fracture toughness measurements. Since the samples fabricated without laser pulse shaping do show the presence of cracks all over theirs surface, the SLM samples produced without pulse-shaping was not utilized for any hardness and toughness measurements. The fluctuation in the hardness values (large error bars) can be observed due to the significant difference in the hardness values of TiC (>2000 HV) and Fe (<200 HV).

Hence, the hardness fluctuation depends on where the indenter is placed as in on the TiC phase or on the Fe phase or intermediate to these two phases. Both samples C and D show a reduction in their hardness values with the increase in laser power from 1010 ± 65 to 730 ± 100 HV, respectively. The gradual decrease in the hardness by increasing the laser peak power is attributed to the decreased formation of the gradient TiC phase, and thus, a reduction in the volume fraction of TiC and at the expense of an increase in the volume of the α-Fe phase (Fig. 5).

First, the fracture toughness values observed for these SLM processed TiC-Fe-based cermets are in the range of other TiC-based cermets, suggesting that the SLM processed samples are sound without any appreciable defects present within them.^58–60 Fracture toughness for samples C and D is 16.3 ± 1.7 and 13.8 ± 1.3 MPa·m^1/2, respectively. The results show that depending on the laser peak strength, the fracture toughness value varies significantly. To better understand the fracture toughness, the crack propagation mechanism was investigated. Figure 7 shows the crack propagation in the pulsed shaped samples. The propagation of the crack follows a zigzag path corroborating crack deflection during propagation.

FIG. 7.

Scanning electron micrographs showing the crack propagation in the SLM prepared: (a) sample C, (b) sample D.

Due to different mechanical features and interfacial mismatches, a strain field is observed at the interface of TiC and α-Fe. When the crack tip reaches this interface, the accumulation of stress concentration occurs due to which the propagation of the crack tends to lower the fracture energy at the interface and causes a transgranular fracture. The total length of the crack increases due to deflection, which consumes the fracture energy and reduces the driving grains providing a restraining force for the crack growth due to which intergranular fracture occur and it consumed fracture energy sharply during crack propagation and improves the toughness of the material.^61–63

Moreover, grain refinement reduces the crack propagation length and improves the strength and toughness of the material. Therefore, higher fracture toughness of the sample C was ascribed by these fracture energies consuming mechanisms.

Pulse shaping mechanism

This article aims to investigate the effect of the laser emission regimen on the formation of TiC-based cermets. The temporal profile of laser power given to the material has a significant effect on the thermal process of SLM. A PW produces bursts of energy having a specific amount of energy for a predetermined length of time (Fig. 8a). It can also be characterized as a change in laser power that alters the form of the output pulse and, as a result, the heat distribution within the pulse. The laser temporal emission mode effect on temperature fields of the melt pool influences the size, stability, and densification behavior in the melt pool. A pulse wave emission has been proved to have a beneficial effect in minimizing thermal distortions and increasing process resolution.⁶⁴

FIG. 8.

Schematics illustrating the difference between laser pulsing (a) without and (b) with shaping.

However, pulse shaping is a technique for distributing energy temporally within a single laser pulse. Figure 8 reveals the schematic diagram illustrating the working of pulse shaping versus a regular pulse mode. In the case of the regular pulse, contour and hatching scanning modes were used (Fig. 8a), and the hatch distance was predefined by the CAD design. The direction of the laser scan is along the y-axis; therefore, the melt pool/laser scan area is defined and narrow compared with the pulse shaping melt/scan area. The direction of the crack generation and propagation of the thermal residual stresses will be perpendicular to the laser scan direction.

However, in the case of the pulse shaping, point scanning strategies were adapted, and for each exposure point, a number of pulse raster's on the one single exposure point is defined as an event here (Fig. 8b). Each event has a varied power and exposure time, which aids in controlling the laser energy distribution. Events (1,2) and (4,5) in a series of pulse shaping have modest laser power and varying exposure times. However, event 3 has a high laser power, the peak power. Event (1,2) can be considered a pre-pulsing event that helps to preheat the feedstock powder with low laser energy density, whereas event (4,5) is a post-pulsing event with varying laser energy densities.

Due to the low laser intensity and prolonged laser exposure duration, post-pulsing serves as a controlled cooling of the melt pool. This energy distribution, together with a controlled heating/cooling rate, aids in the reduction of thermal residual stress and the eradication of cracks. The pre-pulsing events (events 1 and 2) work similar to the raster heating event adopted in the electron beam melting process, which is used to heat the powder bed.^65,66 Because of the high power and regulated laser exposure time, a larger melt pool area is observed, and thus, the melt pool area is strongly reliant on the parameters of the predefined laser exposure parameters.

So in general, the laser shaping leads to preheating of the powder bed (events 1 and 2), actual melting of the powder bed (event 3), and controlled solidification/post-processing of the powder bed (events 4 and 5). The preheating event leads to the elimination of cracks that form due to sudden expansion/contraction in the cermets and the post-processing events lead to partial annealing of the thermal residual stresses, avoids the formation of solidification cracks.

Hence, the present work for the first time demonstrates the effective fabrication of W- and Ni-free TiC-Fe-based cermets without any cracks using the laser pulse shaping technique. The pulsed shaping offers the combined advantage of preheating, melting, and controlled solidification/postheating events that may be suitable for processing brittle materials using the SLM, additive manufacturing process.

Conclusions

In this study, the TiC-Fe cermet was fabricated successfully for the first time through SLM, by adapting pulse-shaping technology. The samples show the presence of TiC and Fe phases without the formation of new additional phases in the as-built condition. The samples processed through regular pulse waves lead to specimens with cracks. However, the pulse shaping technique helps in eliminating the crack by controlled the energy distribution of laser pulse, which controls the melting and cooling rate in the melt pool. The as-built cermets show a maximum hardness of 1010 ± 65 MPa and a fracture toughness of ∼16.3 MPa m^1/2.

Mixed events involving intergranular, transgranular, crack deflection, and crack bridging control the crack propagation and in turn the fracture toughness mechanisms. The results illustrate that effective employment of laser pulse shaping can have a controlled heat input, melting, and dissipation effect on the powder bed, leading to crack-free cermets.

Authors' Contributions

Conceptualization: H.S.M., L.K., and M.T. Methodology: H.S.M., L.K., and M.T. Validation: H.S.M., K.J., F.S., and K.G.P. Formal analysis: H.S.M, L.K., and M.T. Investigation, H.S.M, L.K., and M.T. Resources: K.J., F.S., and K.G.P. Writing—original draft preparation: H.S.M. Writing—reviewing and editing: H.S.M., L.K., and K.G.P. Funding acquisition: M.T. and K.G.P. All authors have read and agreed to the published version of the article.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

This work was supported by Estonian Research Council grants PRG1145 and European Regional Development grant ASTRA6-6.