The optical micrographs shown in the Figs. 4 and 5 depict the porosities in horizontal and vertical orientations in as-printed samples, respectively. The as-printed samples showed high porosity levels in Figs. 4, 5 and 8, 9. The formation of such pores may be due to the gas entrapment causing the creation of spherical pores, which is also evident in micrographs (Figs. 4 and 5) for as-printed samples in both horizontal and horizontal and vertical conditions. Similar results were also revealed where the round shape pores developed during the laser interaction where gas was entrapped and released during the solidification [21, 27]. Moreover, the keyhole pores are also depicted as round-shaped spherical holes concentrated at the molten pool bottom due to a layer's over melting during the LPBF process [28]. The irregular shape pores observed in Figs. 4 and 5 are considered to lack fusion pores due to inconsistent melting between the layers. A similar observation could be seen in LPBF produced 316L stainless steel [29]. However, the porosity level for CHT samples decreased, as shown in the Figs. 6–9 in both horizontal and vertical directions.
Furthermore, in this study, the pore size classification is carried out by considering the size of the inter-micropores and the super-micropores, as shown in Fig. 9. The pore size classification helps understand the pore closure phenomenon in CHT and its effects on the mechanical properties [9, 30]. In addition to this, the box chart for number of porosities is shown in Fig. 10 where it is clearly evident that the super-micropores are reduced to inter-micropores and some of the the inter -micropores are closed owing to the sintering phenomenon after CHT.
The selection of optimum parameters in the LPBF process helps reduce the porosities in as-printed samples. The laser power, scanning speed and laser spot size play a significant role in dictating the densification of a material. The selection of an optimal laser source for vaporization reduction is expected to be based on the energy density thresholds indicated during laser-matter interaction [31]. The energy thresholds were calculated by taking into account the evolution of the melt-pool penetration and the normalized enthalpy ratio between input laser energy (ΔH) and melt enthalpy (hs) [32]. The laser energy input (ΔH) and the energy required to melt the material during fusion (hs) are given in equations (1) and (2).
ΔH = \(\frac{A.P}{\sqrt{\pi \alpha V{r}^{3}}}\) ……. (1)
where A is the coefficient of absorption of powder, P is laser power (W), α is thermal diffusivity of a material (m2/s), V is scanning speed (m/s), and r is laser spot size (m).
h s = δCp (Tm-T0) ……. (2)
where δ is the density of a material (Kg/m3), Cp is the specific heat capacity of a material (J/Kg K), Tm is the melting temperature of a material (K), and T0 is the ambient temperature of a powder bed (K).
If the ratio of ΔH/ hs around 10 results in full surface melting, greater than 25 leads to keyhole defects [33]. However, Fabbro et al. [34] obtained an analytical model where threshold (ΔH/ hs) = 16–20 forms the keyhole mode defects. In the present study, the ΔH/hs for as-printed Ti6Al4V is around 13.2, which signifies the nearly full surface melting of the layers during the LPBF process. However, the observed irregular shape pores could be due to the improper bonding between the layers, and spherical porosities arise due to the gas entrapment under the optimised conditions, as observed by other researchers [27, 28].
In the LPBF process, inevitable porosities during the part fabrication are addressed via the HIP technique with carefully optimised parameters for end-use parts with specific geometries. However, the HIP process is expensive and time-consuming, and other factors limit its use in many practical cases. A study conducted by Zhang et al. [18] reported that the effect of post-heat treatment on the hot isostatic pressed LPBF-Ti6Al4V alloy aggravates the regrowth of pores, and sharp-edged, and slit-shaped pores were formed. Further, they exhibited a detrimental effect on plain fatigue resistance. Moreover, components due to heat treatment at elevated temperatures reduced the sharpness and size of a pore due to the diffusion process.
Another study of selective electron beam melted Ti6Al4V conducted by Tammas et al. [35] revealed that the heat treatment after the HIP process facilitates the regrowth of spherical shape pores that were less in size than the defects formed during the as-printed condition. However, there was no change in the size of a lack of fusion pores after heat treatment. Only those gas pores that are identified as spherical can appear to regrow after annealing.
They can be explained by the origin of the gas particles that make up these pores. Although argon gas employed during the LPBF process cannot diffuse out of the material during the HIP process, it remains in the sample within the collapsed pores. This high internal pressure will cause the regrowth of these pores during the heat treatment process [36]. Furthermore, the surface roughness after the HIP process increases due to the depressions and dimples formed on the material's surface undergoing high pressure and heat treatment. The HIP process does not improve the fatigue strength if surface roughness is not improved; since the treatment only affected the internal porosity, it did not affect the surface finish's fatigue strength [13]. A study conducted by Uzan et al. [37] revealed that stress-relieving heat treatment provides superior fatigue strength to the stress-relieving and HIP processes.
Moreover, the surface roughness of a material can be reduced through CHT owing to the melting of sharp edges at elevated temperatures. Therefore, the pores can be reduced due to the reduction of surface roughness through CHT [38]. In addition, HIP cannot remove surface-connected porosity. It can still expose subsurface pores because pressurized gas penetrates the porous surface during HIP and pushes against the subsurface pores [39]. Majeed et al. [40] reported that the solution treatment and artificial ageing reduced the porosity levels of thin materials to less than 2.5 mm. However, the relative density was lowered for the specimens with a thickness between 2.5 mm and 5 mm. Indeed, there is an improvement in the density of thin sections after solution treatment and artificial ageing; still, there are no significant and proposed explanations for the reported differences in porosities.
The inter-micropores' closure at elevated temperatures can also be attributed to the sintering process. The study conducted by Tascioglu et al. [24] showed similar results as observed in the present study. The 316L stainless steel fabricated through the LPBF process was examined for porosity level in as-printed and heat-treated conditions. The heat-treated samples exhibited a significantly lower porosity than as-printed samples. The reduction in porosity was attributed to the formation of a homogeneous microstructure [24]. A group of researchers led a study on the heat treatment effect on the porosity measurement of thin-walled AlSi10Mg manufactured using the LPBF and showed that densification was achieved using heat treatments. This process alters the grain structure and microstructures of the part to improve its densification [40]. In another study [41], the heat-treated AlSi10Mg showed less porosity than the as-printed AlSi10Mg due to microstructural refinement. The current study evaluated the microstructures of as-printed and CHT samples. A homogeneous microstructure originated for the CHT samples discussed in the section (see Fig. 11). The homogenization of the specimen makes it hard to maintain its natural porosity. This is because the high-temperature homogenization produces a uniform structure within the specimen.
Interesting, the positive effects of heat treatment on the densification behaviour are reported in the thermal spray coating process, in which powder feedstock is melted (similar to the LPBF process) and deposited on a part surface with a high velocity. The high nickel thermal spray coating produces a deleterious porosity level to the fatigue life. The pores in thermal spray coatings are due to insufficient melt source and unmelting of powder, identical to the pores produced in the LPBF process [42]. These developed pores during the thermal spray coating were reduced through the heat treatment. Through a process known as sintering, the porosity of a thermal spray coating can be reduced by high-temperature treatment. This process involves raising the metal or ceramic particles to an elevated temperature, reducing the particles' spaces [43]. Another study on 316L stainless steel thermal spray coating conducted by Mangour et al. [44] showed the beneficial effect of heat treatment on reducing porosity levels. The decrease in porosity caused by increasing annealing temperature in N2-sprayed coatings can be attributed to the usual sintering models. However, it is also possible that the inter-microporosity observed in He and N2-treated samples is due to incomplete inter-particle inter-sintering.
Contrastingly, the research work on metallic glass foams had revealed the advantage of the presence of critical pore size resulting in better plasticity. The pores of sizes 20 µm – 30 µm enable force shear band proliferation compared to the non-porous samples, enhancing compressive plasticity [45, 46].
In the present study, the reduction in porosities in CHT samples is attributed to the sintering process and closure of the pores at elevated temperatures. However, the porosity level in vertically built samples is slightly higher than in horizontal built samples, attributed to many layers in vertically built samples [47]. The higher the number of layers in the material, the more likely inevitable pores will develop. The possible mechanism for reducing porosity due to CHT is shown in Fig. 14., where the pore gap gets close to mitigate the void. Due to high temperature, the study revealed that only inter-micropores are closed after CHT. The high surface energy of inter-micropores makes them reduce as much as possible but not in the case of super-micropores [48]. This phenomenon is known as diffusion, which occurs at high temperatures, and it is also considered a self-healing of the porosities [49].
Moreover, in inter-micropores, the high surface energy is the driving force for the diffusion at high temperatures, resulting in the closure of pores. Further, elevated temperatures can cause morphological changes in the pores. For instance, pores may spheroidize the surface to minimize the free surface energy. According to Young Laplace, Eq. (3) [18] is a simple calculation that shows the driving force for the closure of a pore.
Δp = \(\frac{\gamma }{R}\) …… (3)
where Δp is the pressure, γ is the surface energy, and R is the radii of the pore. The radius of an inter-micropore is less than the radius of a super-micropore. It increases the driving force for the closure of a pore. Therefore, the super-micropores with low surface energy prevent these pores from getting close effectively [50]. The solid-state sintering effect facilitates the closure of inter-micropores. The main challenge in solid-state sintering is to reduce the surface energy. It can be achieved through various means, such as transporting materials and reducing the surface area [51]. There are three stages in solid-state sintering: (1) particles keep growing as they contact the neck. As they do so, neck growth occurs, and the pore shapes are also irregular; (2) with sufficient neck growth, the enlarged pore channels become cylindrical in this stage. The resulting curvature gradient is high enough for faster sintering; (3) in the final stage, the closure of the pore channel can occur when the porous region is isolated and the inter-connected portions are no longer necessary [52]. Moreover, reducing super-micropores by converting them to inter-micropores at high-temperature CHT may enhance fatigue resistance. The porosities act as stress concentration sites where the crack initiates and degrades the fatigue life. However, there is no correlation between the fatigue life and defect size of a material; Muhammad Shamir et al. [53] showed that large defect sizes reduce the fatigue life of wire arc additive manufactured (WAAM) Ti6Al4V material where porosity ranges from 40–220 µm. It was observed that a 20% decrease in defect size enhances the 100% fatigue life of a material. Nevertheless, this trend was not true for all the samples tested at different defect sizes. Furthermore, the location of pores also plays a key role, along with the pore size, in determining fatigue resistance [53]. Generally, surface defects are more prone to initiate cracks. A study conducted by Williams et al. [54] showed the effect of pore size on fatigue crack initiation of electron beam melted Ti6Al4V alloy. The pore size ranging from 20 µm to 170 µm was observed, and the study revealed that the size of the pores affected the fatigue lives. The researchers noted that the larger the pores, the less fatigue they exhibited. Moreover, Murakami [55] showed that defects found in pores are comparable to cracks in the same stress intensity. Murakami’s equation exhibited the maximum stress intensity (Kmax) caused due to the pores is given below as Eq. (4):
………
Where σ∞ is the global applied stress, and An is the pore area normal to the applied stress. According to Murakami’s equation, the constant 0.5 increases to 0.65 when the defect is on the surface, indicating that greater stress intensity arises due to the surface defect [54]. Furthermore, the fractographic images of fatigue failed components revealed that 90% of porosities were super-micropores with more than 10 µm in size. It is important to emphasise that CHT has reduced the present study's super-micropores and inter-micropores. Therefore, institutively, it is presumed that CHT can offer superior fatigue resistance to LPBF Ti6Al4V. However, detailed fatigue testing and fractographic analysis are required to confirm this. In addition, due to the low surface energy associated with super-micropores, they can still convert to inter-micropores and improve the fatigue life.
The high cooling rates in the LPBF process produce a fine needled shape α’ martensite, as observed in Figs. 11 (a and b). Furthermore, the dark bands on the melt pool are observed in as-printed Ti6Al4V in both horizontal and vertical directions, as indicated in Figs. 11 (a and b). The possible reason for these unique features visible in microstructures is the segregation of aluminium due to rapid solidification in the LPBF process, which promotes the formation of Ti3Al precipitates [6]. Due to the high conductive heat transfer rate and the short interaction times [6]. The α’ martensite produced through the LPBF process increases the strength of Ti6Al4V significantly. However, the enhanced strength of Ti6Al4V in the LPBF process comes with the expense of loss in ductility, which is detrimental to high cycle fatigue behaviour [8]. Therefore, CHT is required to convert α’ martensite to α + β, which provides the balanced combination of strength and ductility necessary for the fatigue strength of a material. It is revealed from Figs. 11 (c and d). that the α’ martensite is converted to α + β with increased α lath thickness, and a reasonable ductility is expected.
Moreover, the CHT process produces a homogenous microstructure. Due to microstructure refinement, a homogenous microstructure produced at high temperatures also reduced the porosity level. Furthermore, it is also evident from the XRD results (see Fig. 13.) that after CHT, the β phase is also observed, which is favourable to enhancing the ductility.