Selective Laser Melting of Pure Ag and 925Ag Alloy and Their Thermal Conductivity

Abstract

Due to the high reflectivity of Ag to infrared lasers, there is little research focused on the manufacturing of Ag and Ag alloys by selective laser melting (SLM) technique. In this paper, the manufacturing characteristics, microstructure, and thermal conductivity of SLMed Ag, 925Ag, and their heat-treated parts were studied. With the suitable processing parameters, Ag and 925Ag samples with relative densities of 91.06% and 96.56%, respectively, were obtained. Due to the existence of non-molten particles inside the samples and local high energy density of the laser during the processing, a large number of irregular pores and micropores were formed in the microstructures. XRD analysis shows that no phase transition occurred in the annealed Ag and solution-treated 925Ag parts, as compared to their as-built conditions. The SLMed Ag exhibited fine equiaxed grains, while both columnar grains and elongated lath grains existed in the SLMed 925Ag parts. The annealed Ag and solution-treated 925Ag exhibited large equiaxed grains. Due to the grain growth that occurred in the microstructure, the thermal conductivity of Ag increased by 11.35% after completing the annealing treatment. However, that of 925Ag decreased by 17.14% after completing the solid solution treatment, due to the precipitation of the strengthening phase at grain boundaries. A comparison of the thermal conductivities of Ag and 925Ag shows that the influence of the materials on the obtained thermal conductivities was more pronounced than that of the porosity.

1. Introduction

Additive manufacturing (AM), which is absolutely different than the traditional equal material manufacturing and subtractive manufacturing techniques, is a new technology that manufactures parts through a layer-by-layer method [1,2,3,4,5]. Selective laser melting (SLM) is an additive manufacturing technology that selectively melts and solidifies metal powders by laser beams to form parts according to the two-dimensional part contours generated by computer-aided design (CAD). There are many metals that can be manufactured by SLM, such as common titanium alloy [6,7], aluminum alloy [8,9,10], magnesium alloy [11,12], and stainless steel [5,13]. One of the reasons for the mature technological usage of these materials by SLM is that they have a high absorption rate of laser energy. Thus, these materials can be melted well under the irradiation of laser sources.

Metal materials with high thermal conductivity used to manufacture heat dissipation devices have broad application prospects in the aerospace, automobile, information, and communication industries [14]. Among all metals and their alloys, Ag has a high thermal conductivity of approximately 428 W·m⁻¹·K⁻¹ [15]. Therefore, a radiator that is made of Ag can effectively dissipate heat. Some high-power devices and corresponding radiators have unique shapes, which are difficult to manufacture by traditional manufacturing methods. SLM technology has the potential to solve this challenge. Therefore, there is an urgent need to manufacture Ag and Ag alloys having a high thermal conductivity by the SLM method.

Due to the high thermal conductivity and reflectivity of Ag powders, it is difficult to fabricate Ag samples with a high relative density and excellent mechanical properties using the SLM. Consequently, the current research on the laser additive manufacturing of Ag is limited. Robinson et al. [16] tried to manufacture 99% pure Ag on a copper substrate and steel substrate by single melt and multilayer printing techniques. The influence of applied processing parameters on the printability of Ag was explored [17]. It is found that the laser reflectivity of metal powders is related to the particle size of powders and the applied laser wavelength. Momen et al. [18] found that under the usage of a long wavelength laser (1064 nm), the laser absorptivity of powders, irrespective of their particle sizes, is stable or becomes smaller than that of powders imposed to average wavelength lasers. By comparison, the laser absorptivity of Ag powders with small particle sizes (5–30 μm) becomes higher under the application of a short-wavelength laser (266 nm).

The yield strength, elongation, and ultimate tensile strength of casting Ag are approximately 35 MPa, 50%, and 125 MPa [15], respectively. The mechanical properties of Ag parts manufactured by SLM were studied by Robinson et al. [17]. The yield strength, elongation, and ultimate tensile strength of SLMed Ag parts could reach up to 67.02 MPa, 14.91%, and 126 MPa, respectively. Xiong et al. [19] reported that 925Ag porous skeleton parts, having a relative density of 97% used for medical applications, show a tensile strength of 95.9 MPa and a yield strength of 57.7 MPa. Due to the low printing quality and the high price of Ag, there are few papers, except the above discussed papers, focused on investigating the printing of Ag. In addition, the thermal conductivity of SLMed Ag and 925Ag samples has not been studied.

Ag and 925Ag samples fabricated by SLM are studied in this paper. The effects of printing parameters, such as laser power, laser scanning speed, and hatch spacing on the manufacturing characteristics of Ag are studied. The effects of heat treatments on the microstructure of Ag and 925Ag parts are investigated. The thermal conductivity of Ag and 925Ag samples, manufactured with the suitable process parameters, is tested and discussed. This study provides a basis for the application of SLM to manufacture Ag and 925Ag parts and can be used as a reference for the additive manufacturing and performance testing of other materials having a high thermal conductivity.

2. Experiments

2.1. Materials and Applied Printing Parameters

In this study, Ag and 925Ag powders, manufactured by the gas atomization method, were used as the experimental material. Unlike pure Ag, which contains about 99.9% elemental Ag, the 925Ag alloy, known as sterling silver, consists of 92.5 wt% Ag and 7.5 wt% Cu. The powder morphology and particle size range of Ag and 925Ag powders are shown in Figure 1. The experimental substrate, used to print the powders on, was made of 45 steel with a thickness of 10 mm. The front surface of the substrate was polished on a grinder, sandblasted, and wiped clean with alcohol.

The Di-Metal 100H metal laser selective melting machine of the South China University of Technology was used to prepare the test samples. The key components of this system include an Yb: YAG fiber laser (maximum laser power: 500 W, spot size: 50–70 μm, wavelength: 1075 nm) and an f-θ lens. High-purity argon gas was introduced during the manufacturing process of parts to reduce the oxygen level existing at atmospheric pressure.

To reduce the residual stresses generated by the rapid solidification that occurred in the printing process, the scanning strategy, shown in Figure 2a, with scanning angles of 135° and 225°, was applied for the Ag and 925Ag parts, respectively. The suitable processing parameters for SLM processing of 925Ag have been studied in [19], which serves as a reference for SLM processing of the 925Ag parts fabricated in this paper. The suitable processing parameters to print the Ag are, however, searched in this paper and the specific studied experimental parameters are shown in

Table 1. Printing parameters of Ag samples.

Serial Number	Laser Power (w)	Laser Scanning Speed (mm/s)	Layer Thickness (mm)	Hatch Spacing (mm)
1	400	200	0.025	0.100
2	400	300	0.025	0.100
3	400	400	0.025	0.100
4	430	300	0.025	0.130
5	430	350	0.025	0.130
6	430	400	0.025	0.130
7	400	200	0.025	0.130
8	400	300	0.025	0.130
9	400	400	0.025	0.130
10	430	400	0.030	0.080

. The parametric identification method was based on a single-loop development method in which the suitable process parameters were selected without performing further repetition. This type of development is appropriate for this study. The employed volumetric energy density $E_d$ [20] to print the parts was calculated using the below equation:

$$E_d=\frac{P}{v·d·t },$$

(1)

where P(W) is the laser power, $v$ (mm/s) is the scanning speed, d (mm) is the hatch spacing, and t (mm) is the layer thickness.

The densities of parts processed at different process parameters were identified on DHAUS analytical balance (AX224ZH) by Archimedes drainage method. The relative density (${\rho }_r$) of samples was calculated using the below equation:

$${\rho }_r=\frac{{\rho }_m}{{\rho }_s} \times 100\%,$$

(2)

where ${\rho }_m$ (g/cm³) is the density of samples and ${\rho }_s$ (g/cm³) is their theoretical density. To reduce the measurement errors, the density (${\rho }_m$) of each sample was measured three times and the average results were used in the equation. The theoretical densities of Ag and 925Ag are 10.49 g/cm³ and 10.4 g/cm³, respectively.

2.2. Heat Treatment

Due to the rapid melting and cooling of SLM process [3,6,7,8], the grains of the as-built samples do not have enough time to grow and are consequently fine in size. Annealing treatment often leads to a significant increase of the grain size of metals. The maximum solid solubility of Cu in Ag is about 1% at room temperature [21]. However, 925Ag contains about 7.5 wt% Cu in its composition. High-temperature heating can increase the dissolution of Cu in Ag. Ag–Cu supersaturated solid solution or precipitated phase can be obtained by fast cools applied during the solution treatment. Parameter setting of heat treatment procedures were performed based on that reported in literature [22]. The samples were heat-treated by a Nabertherm muffle furnace. The applied heat treatment schemes for each of the parts are explained in

Table 2. Heat treatment processes of Ag and 925Ag parts processed by SLM.

Material	Heat Treatment Mode	Heat Treatment Situation
Ag	Annealing	700 °C/2 h/furnace cooling (FC)
925Ag	Solution treatment	700 °C/2 h/water cooling (WC)

. Specifically, the Ag samples were placed into the furnace at room temperature and then heated, with a heating rate of 10 °C/min, to approach the holding temperature of 700 °C. These parts were held at this temperature for 2 h and were then cooled within the furnace. The 925Ag samples were treated by a solid-solution treatment. This consisted of placing the samples into the furnace at room temperature and then heating the parts, with a heating rate of 10 °C/min, to the holding temperature of 700 °C. The samples were held at this temperature for 2 h and were then quenched in water.

2.3. Thermal Conductivity and Microstructural Characterizations

A Germany NETZSCH LFA 427 laser thermal conductivity meter was used to test the thermal conductivity of the samples with the size shown in Figure 2b. An X’Pert Powder multi-position automatic sampling X-ray diffractometer was used to analyze the phase structures of powders and consolidated samples. The microstructure of the samples was observed by the optical microscopy (OM, LEICA DMi8) and scanning electron microscopy (SEM, NOVA NANOSEM 430) machines. The size of samples used in these tests is shown in Figure 2c. The surfaces of these samples were polished by 400#, 800#, 1000#, and 3000# sandpapers on a metallographic automatic grinding and polishing machine. The surfaces were roughly and finely polished by alumina polishing solution with particle sizes of W0.5 and W0.03, respectively. Then, the surfaces were etched with the prepared fresh etching solution (composed of 75% NH₃·H₂O and 25% H₂O₂ by volume).

3. Results and Discussion

3.1. Manufacturing Characteristics

The relationship between the density (relative density) of the manufactured Ag parts and the applied volumetric energy density is shown in Figure 3. In this study, the range of the applied volumetric energy density was varied between 307 and 800 J/mm³. When the volumetric energy density was 447.92 J/mm³, the density of the Ag parts reached the maximum value obtained in this study. Considering this, the process parameters resulting in this volumetric energy density were considered as the suitable process parameters to print the Ag parts. The suitable process parameters of the SLMed Ag and 925Ag parts are shown in

Table 3. Suitable SLM processing parameters used to print the Ag and 925Ag parts.

Material	Laser Power (w)	Scanning Speed (mm/s)	Layer Thickness (mm)	Scanning Spacing (mm)	Energy Density (J/mm3)	Density (g/mm3)	Relative Density
Ag	430	400	0.030	0.080	447.922	9.552 ± 0.079	91.06%
925Ag	180	600	0.030	0.060	333.333	10.043 ± 0.091	96.56%

, where the relative density of SLMed Ag and 925Ag were 91.06% and 96.56%. The relative density of 925Ag parts is equivalent to that reported in literature [19]. However, the relative density of pure Ag parts is lower than the results reported in [16,17], which may be due to the irregular shape of the Ag powders and the greater number of satellite particles of powders used in these studies [16,17]. It has been confirmed that powders with an irregular shape have a higher absorption rate of lasers, and the large number of satellite particles reduces the reflective properties of the powders [18].

When the experiment in [19] was replicated, the maximum density was not obtained under the parameters provided. Compared with the process parameters of highest density of 925Ag parts (100 W, 400 mm/s) in [19], our suitable parameter values are larger (180 W, 600 mm/s). The reason for this phenomenon may be the difference of the printing machine and the laser spot. In our study, microcracks appeared in the 925Ag samples, which may be related to the excessively large parameter values we selected, because Xiong et al., pointed out that excessive laser energy input led to thermal microcracks. Microcracks are also associated with scanning strategy. Using a laser scan track parallel to the XY axis, rather than 45°, the length of the laser scan track will be the same, which will help reduce the formation of inhomogeneity [3].

OM images, obtained perpendicular to the building direction (BD), of the Ag and 925Ag parts processed with the suitable process parameters are shown in Figure 4. The surface of Ag parts showed elongated and shiny spherical protrusions which were distributed along the direction of laser scanning. Short protrusions, formed between the adjacent melting tracks, are connected to form a network structure. Melting tracks are clearly seen on the surface of 925Ag parts, as shown by the yellow lines in Figure 4f. A small number of shiny spots existed on the surface of 925Ag parts. These spots increased the overall surface roughness of these parts. Due to the high reflectivity of Ag to the infrared lasers [23], a narrow effective window is observed in Figure 3 to process high-density Ag parts. Near the proper processing window, while a 3.28% change was observed in the density of parts, the volumetric energy density was noted to change by just 1.5%. However, the reflection of the laser is reduced in 925Ag as compared to Ag; this is because 925Ag contains approximately 7.5 wt% Cu in its composition. Consequently, the absorption rate of the infrared laser by 925Ag was higher than that of Ag during the SLM processes. As a result, the SLMed 925Ag parts showed a higher relative density (Table 3) and a better surface quality than Ag ones.

3.2. XRD Analysis

The XRD patterns of the Ag powder and SLMed Ag parts are shown in Figure 5. The diffraction peaks observed in the annealed and as-built SLM parts were attributed to the Ag phase and were consistent with those of the Ag powder. This means that the SLM manufacturing and annealing treatments did not lead to the phase transformation of the Ag phase in the used powder, as the peaks existing in the powder did not disappear and no new peaks appeared in the XRD patterns of these parts. XRD shows an obvious (111) preferred orientation in the patterns of these samples, and the lattice constants of the Ag powder, SLMed Ag, and annealed Ag were a = 4.0811 Å, a = 4.08414 Å, and a = 4.08396 Å, respectively.

Figure 5b is a partially enlarged view of the strongest diffraction peak observed in these samples. A shift to the right was observed for the diffraction peak of annealed Ag parts, as compared to that of the as-built ones, and for the as-built samples compared to that of the used powder. A similar increasing trend was also observed for the full width at half maxima (FWHM) of diffraction peaks of these parts as compared to those of the used powder. The possible reasons for these observations are that the high-temperature gradient generated during the SLM printing process combined with the high thermal conductivity of Ag, leading to high cooling rates, led to the generation of residual stress in the printed parts. On the other hand, the annealing time, applied during the annealing process, was too short, leading to the generation of coherent strains in the microstructure of treated parts [24,25]. Compared with the Ag powder, the diffraction peaks of the as-built and annealed Ag were more intensified. These indicate that, while the annealing treatment did not affect the crystalline state of the as-built Ag parts, the crystallinity of the Ag phase in the SLM processed parts was more improved than that of the phase in the used Ag powder.

The XRD patterns of the 925Ag powder and the SLMed 925Ag parts are shown in Figure 6. The strongest diffraction peaks observed in the solution-treated, as-built, and 925Ag powder were (111) peaks, showing the same preferential orientation as those of Ag samples. The lattice constants of 925Ag powder, as-built 925Ag, and solution-treated 925Ag were a = 4.07486 Å, a = 4.07825 Å, and a = 4.07081 Å, respectively. As compared with the 925Ag powder, the (400) peak almost did not exist in the XRD pattern of the as-built 925Ag parts. This is due to the supersaturated solid solution generated by the rapid heating melting and cooling solidification in the SLM process. Similar results were observed in 925Ag with different heat-treatment systems, in which the (400) peak reappeared again after disappearance in the as-built sample [22].

Figure 6b is a partially enlarged view of the strongest diffraction peak of these samples. The residual stresses, generated in part due to application of the solid-solution treatment, right-shifted the diffraction peaks of the as-built samples. This is because the residual stresses, generated during the applied solid-solution treatment, acts on the grain boundaries, reducing the lattice constant [26]. The diffraction peaks of the SLMed 925Ag parts were broadened as compared to the used powder. Similar peak broadening increases also appear in high-Cu–Ag alloys, which is related to the coherent strain during the solution treatment [21]. Figure 6 shows the same diffraction peaks as that of Figure 5, and no other phase other than Ag was observed in these parts.

3.3. Microstructural Characterization

Optical microscopy images of the SLMed Ag and 925Ag parts are shown in Figure 7. As supported by the relative density results, the Ag and 925Ag parts all showed irregular pores and micron-sized pores in their microstructures [27].

It is thought that the large-sized irregular pores were formed due to the local high energy density of the laser during the processing. The fluidity of the molten metal is enhanced as the molten pool temperature is increased. Excessive local temperatures, generated during the laser irradiation, produce a large number of splashed particles in the melting process [28]. This resulted in the generation of irregular pores in the microstructure of parts processed in this study. On the other hand, it is thought that a small number of powders failed to reach the melting point and could not be melted. The insufficient local energy restricted the flow of molten metal, resulting in the formation of micropores in the microstructures. In essence, the micropores were generated by the insufficient local laser energy [29]. Irregular pores and micron-sized pores existing in the SLMed Ag and 925Ag parts show that the heat distribution is uneven in the process of additive manufacturing [30]. This explains why the SLM manufacturing process window of Ag and its alloys is very narrow (Figure 3).

There are microcracks observed along the manufacturing direction of the SLMed 925Ag samples (Figure 7e). These are related to the thermal stresses generated by the short-term high energy input and the rapid cooling mechanism in the manufacturing process [31]. The existence of pores, both micropores and irregular ones, increases the probability of microcracks generation and propagation in the printed parts. During the rapid solidification of parts, the metal shrinks along the grain boundaries of lath grains being parallel to the construction direction (Figure 8c). The insufficient liquid flow cannot fill the discussed gaps formed between the solidified grains. These lead to crack propagation along the grain boundary and in parallel to the construction direction.

Irregular macropores and micropores still existed in the microstructure of annealed Ag (Figure 7c) and solution-treated 925Ag parts (Figure 7f). This phenomenon indicates that the applied heat treatments could not reduce the porosity of SLMed parts processed in this study. It might be that the thermal stress and atomic diffusion at the applied high temperatures were not enough to promote the fusion of pores.

Figure 8 shows the scanning electron microscopy images of the as-built and heat-treated SLMed Ag and 925Ag parts. While the as-built 925Ag parts were dominated by lath grains formed in parallel to the building direction, as shown in Figure 8c, the as-built Ag parts exhibited fine equiaxed grains, as shown in Figure 8a. The high thermal conductivity of Ag leads to isotropic solidification during the process. The high thermal conductivity led to a rapid heat dissipation during the solidification of these parts, preventing the grains from growing. The low-infrared laser absorption rate of Ag resulted in the lack of overlap between the remelted melting tracks. This also led to the production of equiaxed grains in the microstructures of these parts. The thermal conductivity of as-cast Ag parts is 429 W/(m · K), while that of as-cast Cu parts is 401 W/(m · K) [15]. The addition of Cu reduces the thermal conductivity of the powder. The loss of the laser energy in the direction perpendicular to the construction direction is faster than that along the construction direction [32]. These led to the formation of thin lath grains in the microstructure of 925Ag parts. Despite the addition of Cu, the content of Ag in 925Ag is still high (approximately 92.5 wt%). This made some areas show a microstructure with the same characteristics as that of Ag parts having equiaxed grains.

After the conduction of heat treatments, the grains of Ag and 925Ag were obviously enlarged and turned into the equiaxed grains (Figure 8b,d). The temperature applied during the heat treatment was at ca. 0.7 $T_m$ ($T_m$ is the melting temperature of Ag, which is 963 °C). This applied high temperature and relative long-term holding time (2 h) caused grain boundary rupture, grain growth, grain boundary fusion, and new grain boundary regeneration. Lath grains disappeared as the sample was heated uniformly in all directions during the heat treatment. The fine lath crystals and equiaxed crystals were transformed into the coarse equiaxed ones. The obvious strengthening phase precipitates at grain boundaries of 925Ag, which is significantly different from the microstructure of SLM aging treatment in [25] and the annealing twins of casting solution treatment in [33].

3.4. Thermal Conductivity

Figure 9 compares the thermal conductivity of the as-built SLMed Ag and 925Ag parts with those of heat-treated ones and also the as-cast parts reported in [15]. The thermal conductivity of SLM-processed Ag parts increased by 11.35% after completing the annealing treatment. However, the thermal conductivity of 925Ag parts decreased by 17.14% after completing the solution treatment.

Table 4. Thermal properties of as-cast, as-built, and heat-treated Ag and 925Ag parts.

	Density (g/cm3)	Thermal Diffusion Coefficient (mm2/s)	Specific Heat Capacity (J/(g·K))	Thermal Conductivity (W/(m·K))
As-cast Ag	10.490	176.270	0.232	429.000
As-built Ag	9.552 ± 0.079	41.336 ± 0.162	0.255 ± 0.011	100.879 ± 0.011
Annealed Ag		56.319 ± 0.797	0.209 ± 0.007	112.329 ± 1.589
As-cast 925Ag	10.400	93.656	0.232	228.000
As-built 925Ag	10.042 ± 0.052	21.224 ± 0.024	0.261 ± 0.009	55.708 ± 0.063
Solution-treated 925Ag		20.235 ± 0.205	0.227 ± 0.002	46.159 ± 0.469

provides a comparison of the thermal properties of as-cast, as-built, and heat-treated Ag and 925Ag parts. The specific heat capacity and density of as-cast and additive manufactured samples had no significant differences together and the specific heat capacity of the as-built SLMed parts were even larger than those of as-cast ones. The slight reduction in the densities of SLMed parts, as compared to the as-cast ones, is due to the presence of pores formed on the surface and inside these samples. While the heat transfer within the grains and between the grain boundaries can be seen as heat conduction, the heat transfer within the pores is due to the convection heat exchange. The existence of pores in the microstructure of SLMed parts deteriorated their thermal diffusion coefficients as compared to those of as-cast ones. The thermal diffusion coefficient describes the rate of the heat flow that occurs from one side of the tested samples to the other side. The reason for the low thermal conductivity of the SLMed parts was that their thermal diffusion coefficients were significantly lower than those of as-cast parts.

Metals transfer heat mainly through lattice vibration and electron thermal motion. The grains of as-built Ag and 925Ag parts are fine, causing the generation of a large grain boundary area in these parts. The average grain size of the as-cast 925Ag parts reported in [33,34,35] is approximately 20 μm. Those of SLMed parts processed in this study were only approximately 1–4 μm, leading to a large amount of heat loss during the vibration transmission between lattices. Electrons move irregularly when heated, and scatter at grain boundaries [36,37]. The larger the grain boundaries areas are, the more significant the scattering effect is [38]. Therefore, the thermal conductivities and thermal diffusion coefficients of SLMed Ag and 925Ag parts were much lower than that of the as-cast parts. While the as-cast parts are almost completely dense, a large number of pores existed within the microstructure of SLMed Ag and 925Ag parts. Lattice vibration and electron thermal motion are blocked at the pores, deteriorating the thermal diffusion coefficient and thermal conductivity of SLMed parts [39].

Comparing the thermal conductivity of Ag with that of 925Ag, whether as-cast or as-built by SLM, it is noted that the thermal conductivity of 925Ag is approximately 50% lower than that of Ag. The main reason is that the addition of Cu and other alloying elements in the Ag composition results in a supersaturated solution. Although the porosity of Ag was greater than that of 925Ag, the thermal conductivity of Ag was higher than that of 925Ag. This shows that the influence of materials’ nature on the thermal conductivity is greater than that of porosity existing in the microstructure of SLMed parts processed in this study [40,41]. The reason for the decrease of thermal conductivity of 925Ag in solution treatment is that Cu in Ag strengthens grain boundaries. This results in hindering the transmission of inter-lattice vibration and the directional transmission of moving electrons passing the heat.

4. Conclusions

In this paper, the SLM manufacturing characteristics of Ag and 925Ag were studied, and the microstructure and thermal conductivity change mechanism of as-built and heat-treated Ag and 925Ag parts were investigated.

• When the applied energy densities were 447.92 J/mm³ and 333.33 J/mm³, respectively, the obtained relative densities of SLM-processed Ag (91.06%) and 925Ag (96.56%) parts were the highest. As-built Ag and 925Ag parts showed irregular pores and micron-sized pores formed by non-molten particles, which is the direct reason for their low relative density.
• The SLMed Ag exhibited fine equiaxed grains, while both equiaxed grains and elongated lath grains existed in the SLMed 925Ag parts. The annealed Ag and solution-treated 925Ag exhibited large equiaxed grains. Irregular macropores and micropores still existed in the heat-treated parts, indicating that the applied heat treatments could not reduce the porosity of SLMed parts.
• The existence of pores hinders the lattice vibration. The electron is heated to create thermal motion, though this motion is scattered at the grain boundaries. Pores and grain size of parts processed in this study are the main reasons for the observation of their low thermal conductivities, as compared to as-cast parts. The comparison between the thermal conductivities of Ag and 925Ag shows that the material nature has a greater influence on the thermal conductivity than the porosity level of parts processed in this study.