1. Introduction
In recent years, metal additive manufacturing (M-AM) has gained increasing relevance in many industrial fields due to its design freedom and the ability to manufacture complex metal geometries, which enable optimization of the material distribution, the reducing of mass while maintaining the mechanical and other performance requirements of the part, the combination of multiple components, a reduction in both the risk and cost in manufacturing multiple components, a decrease in potential failure modes across joints, and a reduction in lead time and associated costs [
1,
2].
For these reasons, M-AM processes are very attractive for companies for rapid tooling (RT), which consists in the use of additive manufacturing technologies to produce tools and molds [
3]. Rapid tooling involves both direct and indirect tooling processes. As far as direct tooling is concerned, the mold is made directly in one step using one of the additive manufacturing technologies. In contrast, indirect tooling process allows a mold or tool to be obtained by means of more than one intermediary step by using additive manufacturing model as the first step [
4].
Traditional tooling is quite time consuming and expensive. As a matter of fact, conventional mold and tool production depends on machining operations to generate the desired mold cavity and shape, which is becoming less and less competitive. Often, successive iterations, modifications, and redesigns are necessary, and the typical waste of material in subtractive technologies cannot be avoided.
In order to overcome such drawbacks, many industrial and academic researchers are concentrating on the use of additive manufacturing techniques to obtain metallic tools and molds, since additive technologies allow for significant improvements in parts development such as on-demand, low-cost, quick functional prototyping; a simpler supply chain for effective low-volume production; free geometric complexity based on “design for use” rather than “design for manufacturing”; innovation in tool design, allowing the creation of desired cooling channels and other complex features that are difficult or impossible to achieve with conventional tooling processes; the lightweighting of components; parts consolidation, with the consequent removal of the need for assembly operations in many cases; the realization of local manufacturing, parts repair, and refurbishment; and health and humanitarian benefits [
5,
6].
Technologies that are dominant in the market of metal additive manufacturing include those based on powder consolidation, such as powder bed fusion (PBF-LB/M) [
7,
8,
9], direct energy deposition (DED) [
10,
11], and binder jetting (BJ) [
12].
Among the metal additive manufacturing technologies, BJ involves several steps to produce the first part, typically called a “green” part, which is made from powder particles that are bonded together by a binder. Production of the green part includes printing, curing, and depowdering. BJ is a powder-bed-based AM technology in which the powdered material is spread into a thin layer and selectively joined into the desired layer shape with a binder, which is typically a polymeric liquid [
13]. As the build progresses layer by layer, the layers are bonded together, resulting in a box of powder with binder arranged in the 3D shape of the desired part geometry. Then, the printed green parts are removed from the powder bed by means of a depowdering process. The green part is then postprocessed using heat treatments that involve debonding and sintering to obtain the final dense metal part with the desired properties [
5,
12,
14,
15,
16].
Compared with metal additive manufacturing processes that use a powerful source of energy, such as a laser or electron beam (i.e., PBF-LB/M and DED), binder-jetting-printed parts do not exhibit thermally induced stress and distortion due to large thermal gradients. Therefore, BJ technology allows for more complex geometries to be obtained, which are also characterized by thin walls, without the use of support structures, compared with both metal AM processes that use thermal energy to melt raw materials [
10] and metal injection molding (MIM) processes [
13,
14,
15,
16,
17,
18,
19,
20,
21]. Furthermore, binder-jetting-printed components have significantly lower costs—up to 60% lower—compared with PBF technologies, with strong potential for high-speed production [
22].
The main technologies used in rapid tooling are those that are laser-based [
23,
24] due to their higher part density and strength, wider material selection, better surface quality, and higher precision and accuracy. Then, in a direct rapid tooling approach, additive manufacturing processes are directly used to print real tools and molds in the end material [
3].
In this context, the binder jetting process could be used in rapid tooling, since it ensures high-speed and low-cost manufacturing with respect to the other metal technologies based on the thermal energy selectively used to fuse regions of a powder bed. Based on the results obtained by scientific research analyzing the dimensional accuracy and surface quality of binder-jetting-printed parts [
21,
25,
26], the BJ technology might not only produce tools and molds with mechanical properties comparable to those obtained through traditional methods but could also improve the efficiency of the production process. Furthermore, BJ could offer several advantages in terms of tool productivity, as it has both a higher efficiency and faster build speeds than additive manufacturing processes based on powerful sources of energy. It is a more cost-effective process, avoids postprocessing operations to remove support structures, allows a high level of recyclability of the unused powder, and has a large build volume [
27,
28,
29]. The effectiveness of this process depends on the selection of appropriate parameters, such as the control and qualification of powder characteristics, binder selection, the binder deposition method and its compatibility with printing, the powder–binder interaction, stability, and burnout characteristics; also, the printing process parameters (such as printing speed, printing orientation angle, layer thickness, drying time, roller speed, and rate of debinding) and specifications, the correlation of densification kinetics with complex geometry, and the required post-processing procedures (such as sintering, infiltration, post-heat treatment and surface finish) play a key role in the properties of the final product [
12,
30]. Radhakrishnan et al. [
31] and Emanuelli et al. [
32] studied the effect of H900 and H1150 heat treatment on the mechanical properties and microstructure of 17-4 PH printed by binder jetting. Nezhadfar et al. [
33], Kaletsch et al. [
34], and Huber et al. [
35] investigated the effect of hot isostatic pressing in combination with heat treatment on the mechanical tensile behavior of 17-4 PH binder-jet-printed parts and its microstructure. Favi et al. [
36] and Mueller et al. [
37] studied the 3D-printing technology based on bed fusion and its applications on the rapid production of tooling.
In this framework, the present work aims to study the applicability of the BJ technology for the rapid production of tools and molds in 17-4 PH stainless steel. To this purpose, the effect of printing orientation angle and postprocessing heat treatment on the mechanical properties of binder-jetting-printed parts was investigated. Tensile specimens were obtained at different printing orientation angles (0°, 45°, and 90°). Half of the specimens were left in the as-sintered condition after the 3D-printing operation, while the other half of the specimens was subjected to postprocessing H900 heat treatment. Experimental tensile tests, hardness tests, and metallographic analysis were carried out to evaluate the mechanical properties and microstructure of as-sintered and thermal-treated specimens. Finally, the fracture surfaces and microscopical defects of the binder-jetting-printed parts as a function of the different printing orientation angles and postprocessing thermal treatments were observed by means of scanning electron microscopy with energy-dispersive X-ray spectroscopy analysis.
The paper is organized as follows: after the Introduction,
Section 2 describes the materials and the experimental procedures used in the present research;
Section 3 reports the results and the discussion, considering the scientific literature outcomes; then the main conclusions are drawn.
3. Results and Discussion
In order to evaluate the effect of the printing orientation angle and heat treatment on the mechanical properties of binder-jetting-printed parts in 17-4 PH and to assess the optimal conditions for rapid tooling, tensile specimens printed at different printing orientation angles were tested at room temperature. Furthermore, the effect of H900 heat treatment was investigated.
Figure 5 shows as-sintered specimens before and after tensile tests.
Figure 6 shows the mean stress–strain curves of the printed tensile specimens in 17-4 PH, obtained using the binder jetting process at different printing orientation angles, in the as-sintered condition.
Table 4 summarizes the mechanical properties of the as-sintered specimens and their standard deviation values at the different printing orientation angles investigated. Irrespective of the printing orientation angle investigated, the tensile stress linearly increases with strain in the elastic region until yielding. Then, the stress rises with a non-linear behavior up to a peak value corresponding to the onset of necking. Finally, the stress decreases with increasing strain until failure.
As far as the effect of the printing orientation angle is concerned, it can be observed that the stress–strain curves obtained by testing the specimens in the D0°, D45°, and D90° conditions exhibit negligible discrepancies; in particular, in terms of the E, UTS, and strain at the necking values. Such results can be attributed to the low influence of the binder deposition direction with respect to the subsequent sintering process on the microstructure of the specimens.
However, the specimen in the D0° condition is characterized by the lowest E, YS, and EF values. On the contrary, similar ductility levels were obtained by testing tensile specimens in the D45° and D90° conditions. These results are reflected in the different elongations to failure achieved by the tensile specimens, as can be seen in
Figure 5.
Figure 7 shows the mean stress–strain curves obtained after tensile tests of the H900 heat-treated specimens.
Table 5 reports the values of the mechanical properties resulting from the tensile test after the H900 heat treatment, as well as their standard deviations.
Linear growth of the stress with strain can be observed until the YS is reached. In the plastic region, tensile stress increases with strain until necking, with a non-linear trend. After necking, a decrease in stress appears, and failure rapidly occurs.
The comparison between the stress–strain curves obtained by the as-sintered (
Figure 6) and H900 heat-treated tensile tests (
Figure 7) highlights that the H900 heat treatment allows significant improvement of the tensile strength levels of 17-4 PH stainless steel, regardless of the printing orientation angle, both in the elastic and plastic regions. To this purpose,
Figure 8 shows the effect of the H900 heat treatment on the mechanical properties obtained after tensile tests on the binder-jetting-printed specimens. Irrespective of the printing orientation angle investigated, the H900 treatment leads to an increase in the elastic modulus, with an average increase of 7% (
Figure 8a). Furthermore, the H900 heat-treated specimens exhibit a significant improvement in terms of the YS and UTS values as compared with the as-sintered ones (
Figure 8b,c); in particular, the YS and UTS tend to rise, with an average increase, respectively, of about 40% and 22%. In addition, an increase in the strain-hardening coefficient (n) was observed after the heat treatment (
Figure 8d). This leads to a delay in the onset of necking. As a matter of fact, while necking occurs at approximately 50% of the total deformation in the as-sintered specimens, it occurred at approximately 80% of the total deformation in the H900 heat-treated specimens; this result can explain the low stress reduction after the necking in the H900 heat-treated specimens.
Unfortunately, the H900-treated specimens are characterized by a poor post-necking deformation as compared with the as-sintered ones, in which the heat treatment causes a reduced ductility. Specifically, the ductility of specimens at the D90° and D45° conditions is reduced by about 25% as compared with the as-sintered specimens, whilst a negligible effect on the ductility of the D0° specimens can be observed.
As far as the hardness measurement is concerned, Rockwell tests were performed on the binder-jetting-printed specimens at different printing orientation angles, both in the as-sintered condition and the H900 heat-treatment one. To this purpose,
Table 6 summarizes the results obtained by the hardness tests.
Irrespective of the heat treatment investigated, in the as-sintered condition, a negligible effect of the orientation printing angle on hardness can be observed. A mean hardness value of about 30 HRC was measured. This level of hardness is typical of a 17-4 PH steel that has not been heat treated, since the material is predominantly austenitic or slightly martensitic after sintering.
As the H900 heat treatment is performed, a significant increase in hardness can be observed (
Figure 9).
Such results can be attributed to the aging treatment, which changes the microstructure. The microstructure of the as-sintered specimens before the tensile tests, observed by optical microscopy, is a combination of ferrite and martensite (
Figure 10a). This is a typical microstructure of martensitic steels with a high chromium content, such as 17-4 PH. Delta ferrite is formed due to carbon diffusion and depends on the sintering temperature and dwell time. Such a result is consistent with what has already been observed by other researchers [
41,
42,
43]. After heat treatment (
Figure 10b), the predominant phase is lath-type martensite, while a decrease in delta ferrite can be observed. Additionally, the SEM-EDSX analysis revealed the presence of Cu precipitates located at the grain boundaries (
Figure 11), which was also observed by Di Pompeo et al. in [
44].
These precipitates reinforce the martensitic matrix, increasing the hardness and mechanical strength of the material. The printing orientation angle affects the hardness increase; in particular, the highest rise in hardness occurs as the D0° specimen condition is considered. This could be due to the different defects’ distribution or to the density of hardening particles in the different orientations [
45,
46,
47].
In order to assess the fracture mechanisms, scanning electron microscopy (SEM) analysis was conducted on the fracture surfaces of binder-jetting-printed specimens at different printing orientation angles, both in the as-sintering and H900 heat-treated conditions. High-resolution SEM images at various magnifications were obtained to examine the morphology and to identify potential defects or features, such as dimples, voids, or cracks, that provide insights into the failure behavior.
As far as the as-sintered specimens are concerned (
Figure 12A–C), it can be observed that the fracture surfaces are characterized by a porous and rough texture with a noticeable presence of microvoids and regions with incomplete sintering. The lack of a well-consolidated grain structure after the sintering process leads to a low material density, which contributes to reducing the mechanical strength and stiffness. On the contrary, the H900 heat-treated specimens (
Figure 12D–F) exhibit fracture surfaces characterized by a more uniform and consolidated structure. As can be observed by the SEM images realized at 1000× magnification in
Figure 13D–F, a significant reduction in porosity appears with respect to the as-sintered specimens (
Figure 13A–C); furthermore, the microstructural features are smoother than in the as-sintered condition. Irrespective of the printing orientation angle, the heat treatment enhances the bonding between particles and densification, leading to improved mechanical behavior, particularly in terms of toughness and strength. The results obtained by the SEM analysis explain the different mechanical behavior of the as-sintered (
Figure 6) and H900 heat-treated specimens (
Figure 7).
Considering the results in the previous literature review, it is possible to compare those with the results obtained in this study. In particular, these results demonstrate that the printing orientation does not influence the tensile behavior of parts obtained by binder jetting technology. The scanning electron microscopy and mechanical properties obtained from tensile tests indicate that BJ technology can supply similar performances for parts produced in different binder deposition directions, as compared with other metal additive manufacturing processes, such as selective laser melting (PBF-LB/M), which employ intense heat sources to fuse raw materials, generating anisotropic components. As a matter of fact, as demonstrated by Challis et al. in [
48], selective laser melting induces anisotropic properties in the investigated material due to the preferential grain orientations that form during the process. Consequently, different mechanical properties, such as the Young’s modulus, are obtained as a function of the build orientation.
It is therefore possible to use binder jetting to rapidly produce tools and molds with advantages over the PBF-LB/M technology. Specifically, compared with PBF-LB/M, binder jetting allows the attainment of lower residual stresses, lower energy consumption, and higher material utilization with, consequently, more efficient recycling of unused powder and the realization of complex geometries with fewer constraints related to thermal management [
19,
31,
32,
33,
34,
35,
49,
50].