Materials Processing
Combustion Foaming of Porous (Al, Fe)3Ti with L12 Ordered Structure
2024 Volume 65 Issue 4 Pages 414-421
Details
2024 Volume 65 Issue 4 Pages 414-421
The combustion foaming process enables the fabrication of closed-cell porous intermetallics by utilizing the combustion reactions generating a large amount of heat. A closed-cell porous Al3Ti (D022 crystal structure) fabricated by the combustion foaming process has a lightweight, high stiffness, high melting temperature, and good oxidation resistance but exhibits brittleness due to the poor deformability of Al3Ti. It is known that the addition of third elements (X = Cr, Mn, Fe, etc.) changes the crystal structures from D022 to L12, resulting in an improvement in the deformability in the case of dense materials. In the present study, an attempt was made to fabricate the closed-cell porous (Al, Fe)3Ti with an L12-ordered crystal structure through the combustion foaming. The effect of the amount of the exothermic agent, which was added to control the reaction heat of combustion foaming, on the porosity, pore morphology, microstructure, and constituent phases was investigated. The closed-cell porous L12-(Al, Fe)3Ti without any intermediate phases was successfully fabricated when the exothermic agent was added over 5 vol%. The porosity reached the maximum at approximately 80% when the exothermic agent of 10 vol% was added. The TiB2 particles, which were formed by the reactions of the exothermic agent, were aggregated in the porous sample with the exothermic agent of 5 vol% but uniformly dispersed in the porous samples foamed with the exothermic agent over 10 vol%. Based on the results above and the temperature measurements during the combustion foaming, it is important for fabricating highly porous L12-(Al, Fe)3Ti with the uniform microstructure to control the maximum temperature just above the liquidus temperature of (Al, Fe)3Ti. This study provides new insights into the hierarchical control of porous materials with the desired pore structures and cell wall materials.
Porous structures impart lightweight, high specific stiffness to bending and torsion, high energy absorption capacity, high thermal insulation, and high sound absorption to the materials.1–4) Al3Ti is an intermetallic compound with a low density of 3400 kg·m−3, high melting temperature (1340°C), and high Young’s and shear moduli (215 and 93 GPa).5,6) Porous Al3Ti, which has characteristics of both the cellular structure and Al3Ti, can be used for lightweight structural components and heat-resistant thermal insulators.7)
A closed-cell porous Al3Ti can be fabricated by the combustion foaming process.7–10) The combustion foaming process utilizes the combustion reaction between reactant powders, generating a large amount of heat and raising the reactant/product temperature rapidly. A rapid and spontaneous rise in temperature has the advantage of efficiently producing intermetallics and ceramics with high melting temperatures. When the porous Al3Ti is produced by the combustion foaming, two combustion reactions are used: (i) 3Al + Ti → Al3Ti and (ii) 3Ti + B4C → 2TiB2 + TiC. The heat generated by reactions (i) and (ii) are approximately 146 kJ/mol Ti and 253 kJ/mol Ti at standard states. Reaction (i) is the main reaction to form Al3Ti. Reaction (ii) is used for controlling the total amount of reaction heat and the maximum temperature. Therefore, 3Ti + B4C is denoted as the exothermic agent.9) It is suggested that the bubble is generated by H2 gas release from the hydrous oxide on the Al powder and/or H atoms dissolved in Al powder.11,12)
One of the crucial problems of the porous Al3Ti is its brittleness. When the porous Al3Ti is compressed, the stress first increases but then sharply decreases due to the fracture of the Al3Ti cell wall.7) The brittleness is caused by the D022-type crystal structure of Al3Ti. The crystal structure of Al3Ti is shown in Fig. 1(a).13) In the D022 structure, the main deformation mode is $(111)[11\bar{2}]$ twin deformation, and there are four slip systems.6) This does not satisfy von Mises’ condition that five or more independent slip systems are necessary for polycrystals to deform into an arbitrary shape.14) One of the simple solutions to improve ductility is increasing the molar ratio of Al. Then, the product of reaction (i) becomes an Al/Al3Ti composite, leading to improving the deformability due to the presence of the ductile α-Al phase.7) However, the stiffness and melting temperature (heat resistance) will be degraded by introducing the α-Al phase. On the other hand, the addition of third elements including Cr, Mn, Fe, Pd, and Ag into Al3Ti changes the crystal structure from the D022 to L12 with a higher symmetry.15,16) For example, the L12 crystal structure formed by the addition of Fe is shown in Fig. 1(b).17) Al and Fe atoms preferentially occupy the in-plane sites and Ti atoms preferentially occupy the corner sites. The other third elements also occupy the in-plane site, and therefore the intermetallic is denoted as (Al, X)3Ti, where X is the third element. In the L12 structure, ⟨110⟩ slip on the (001) plane is activated,18) resulting in an improvement in ductility. Figure 1(c) shows the isothermal section at 1200°C for the Al–Fe–Ti ternary system. Although the melting temperature (solidus temperature) of (Al, Fe)3Ti has not been clarified, its phase region is confirmed even at 1200°C.17) Therefore, a high melting temperature and heat resistance will be sustained in this intermetallic. In addition, high stiffness compared to Al is also reported,5,19) suggesting that the (Al, X)3Ti with L12 crystal structure is an attractive intermetallic as a matrix of porous metal.
In the present study, an attempt was made to fabricate the porous (Al, Fe)3Ti by the combustion foaming process assisted by the exothermic agent of 3Ti + B4C. The effect of exothermic agent content on the maximum temperature during the combustion foaming, pore structure, crystal structure, constituent phases, and microstructure were investigated. The results were used for discussing the optimized exothermic agent content to obtain a highly porous L12-(Al, Fe)3Ti with a uniform microstructure.
Al (purity: 99.99%, size: <45 µm, Kojundo chemical lab, Ltd.), Ti (purity: 99.9%, size: <45 µm, Kojundo chemical lab, Ltd.), Fe (purity: 99.9%, size: <53 µm, Kojundo chemical lab, Ltd.), and B4C (purity: 99%, average particle size: 0.5 µm, Kojundo chemical lab, Ltd.) powders were utilized as raw materials. The morphology of the powders was reported in works of literature.20,21) These powders were weighed so that the molar ratio of Al, Fe, and Ti was 2.6:0.3:1.1, and the volume fraction of (TiB2+TiC) (Ve) after the reaction complete was 0, 5, 10, and 15 vol% and dry mixed for 1800 s. Figure 1(d) shows the liquidus projection for Al–Ti–Fe ternary system.22) The composition of the matrix (Al:Fe:Ti = 2.6:0.3:1.1) is also plotted on the liquidus projection. If the matrix is fully melted, the primary phase at the solidification will be (Al, Fe)3Ti. The liquidus temperature is in the range of 1300–1400°C. The powder blends were cold compacted at 200 MPa to make cylindrical precursors with a diameter of 15 mm and a height of 15 mm. A hole with a diameter of 2 mm was drilled on the top surface of the precursor to insert a thermocouple.
The precursors were heated at a rate of 1.0°C·s−1 under an Ar gas atmosphere using a high-frequency induction furnace (VMF-I-I1C, DIAVAC LIMITED, Japan) while monitoring the temperature measured by a thermocouple. When the combustion reaction started around a temperature of approximately 660°C (melting temperature of Al), the temperature rose rapidly. Then, the furnace’s power was turned off, followed by the natural cooling to ambient temperature.
The porosities of foamed samples were measured by the Archimedes method. The theoretical densities of cell wall materials were set at 3740 kg·m−3 for Ve = 0 vol%, 3780 for Ve = 5 vol%, 3820 for Ve = 10 vol%, and 3860 for Ve = 15 vol%, respectively, assuming that cell wall materials did not contain intermediate phases and were composed of (Al, Fe)3Ti, (density: 3740 kg·m−3 23)), TiB2 (density: 4510 kg·m−3 24)), and TiC phases (density: 4900 kg·m−3 25)). The pore morphology was taken by X-ray computed tomography (CT) (SKYSCAN 1275, Bruker, USA) under a current of 95 µA, voltage of 103 kV, and a pixel size of 26.4 µm. The tomography images were reconstructed into two-dimensional cross-sections using software (NRECON, Bruker, USA). The pore size (equivalent circle diameter) distributions were analyzed using the two-dimensional 50 cross-sectional images in software (CTAN, Bruker, USA).
The foamed samples were cut into two pieces. The one was embedded into epoxy resin and mechanically polished using silicon carbide papers. The sample was further buffed with diamond slurries with 1 µm and 3 µm particle sizes, followed by the final polishing using colloidal silicon oxide. The microstructure of the cell wall was observed using a scanning electron microscope (SEM, JSM-6610A, JEOL, Japan). The elemental distributions were analyzed by energy-dispersive X-ray spectroscopy (EDS). The other piece was crushed into powder and embedded into epoxy resin, followed by exposing the cross-section of the powder. To investigate the constituent phases, the cross-section was irradiated with an X-ray from a Cu-Kα radiation source using an X-ray diffraction (XRD) apparatus (Ultima-IV, Rigaku, Japan) operated at 40 kV and 40 mA.
Figure 2 shows the temperature profiles measured by the thermocouple. The rapid rise in temperature occurred at around 660°C, indicating that the combustion reaction started from the melting of Al powder (Fig. 2(a)). The starting point of the combustion reaction in Al–Ti–Fe–B4C was consistent with the other systems including Al–Ti and Al–Ti–B4C systems.20,26) The maximum temperature changed from approximately 1150°C to 1500°C and increased with increasing the volume fraction of the exothermic agent. The temperature profiles surrounded by dotted rectangles in Fig. 2(a) were magnified in Fig. 2(b)–(e). When Ve was 0 vol%, the temperature decreased almost linearly after the temperature peaked (Fig. 2(b)). On the other hand, when Ve was 10 and 15 vol%, the cooling rate was slowed down at around 1340°C (Fig. 2(d), (e)). This would be caused by releasing the latent heat during the solidification of the (Al, Fe)3Ti phase. If the liquidus temperature of (Al, Fe)3Ti phase was considered as the temperature at which the cooling rate started to be slowed down, it was quantified as approximately 1340°C. The liquidus temperature is in good agreement with the liquidus projection shown in Fig. 1(d). When Ve was 5 vol%, an almost constant cooling was confirmed but the temperature peak was broadened compared to the other samples.
Temperature histories during heating and combustion foaming of (Al, Fe)3Ti foams with various volume fractions of exothermic agent (Ve). The temperature histories of (Al, Fe)3Ti foam with Ve = (b) 0, (c) 5, (d) 10, and (e) 15 vol% are shown in detail.
Figure 3 shows the change in the maximum temperature as a function of the volume fraction of the exothermic agent. The dotted line shows the liquidus temperature measured in this study (the temperature at which the cooling rate starts to be slowed down). In addition, the gray bar indicates the possible liquidus temperature range estimated from the liquidus projection (Fig. 1(d)). The liquidus temperature measured in this study was in the range of the possible liquidus temperature range. The maximum temperature was approximately 1150°C when the volume fraction of the exothermic agent was 0 vol%. The maximum temperature was close to the liquidus temperature when the volume fraction of the exothermic agent was 5 and 10 vol%. When the volume fraction of the exothermic agent was 15 vol%, the maximum temperature was approximately 1500°C and much higher than the liquidus temperature.
Change in maximum temperature of the reaction measured by the thermocouple as a function of volume fraction of exothermic agent. The dotted line and gray bar indicate the measured and reported liquidus temperature at the composition of (Al, Fe)3Ti in this study.
Figure 4 shows the photographs showing the cross-section of (Al, Fe)3Ti foams with Ve = (a) 0, (b) 5, (c) 10, and (d) 15 vol%. The outline shape of the sample with Ve = 0 vol% was not changed from the cylindrical precursor although the sample was expanded. This indicates that the sample did not experience the gas foaming. The sample with Ve over 5 vol% was shape-changed and gas-foamed. The sample with Ve = 5 vol% exhibited inhomogeneous cell structures. Large pores are observed at the upper part of the sample whereas the bottom part was relatively dense. The sample with Ve = 10 vol% had relatively homogenous cell structures composed of thin cell walls. The volume of the sample reached the maximum at Ve = 10 vol%. The sample with Ve = 15 vol% was relatively dense and its volume was shrunk compared with the sample Ve = 10 vol%.
Photographs showing the porous structures of (Al, Fe)3Ti foams with volume fractions of exothermic agent of (a) 0, (b) 5, (c) 10, and (d) 15 vol%.
Figure 5 shows the change in the porosity as a function of Ve. The porosity increased with increasing Ve in the range of 0–10 vol% and then decreased at Ve = 15 vol%. The results were consistent with the volume of the sample shown in Fig. 4. These results indicated that the optimized Ve for fabricating porous (Al, Fe)3Ti with high porosity was around 10 vol%. The pore size of (Al, Fe)3Ti with Ve = 10 vol% was characterized and shown in Fig. 6. The pores with the size of 6–8 mm had the highest area fraction, and the average pore diameter was 7.8 mm. The pore size and distribution of the porous (Al, Fe)3Ti were larger than those of previously reported porous Al3Ti/Al composites.8) Although the pore morphology needs to be further controlled, the highly porous sample can be fabricated by the combustion reaction of the Al–Ti–Fe–B4C powder mixture.
Change in porosity of (Al, Fe)3Ti foams as a function of volume fraction of exothermic agent.
Histogram of pore diameter of (Al, Fe)3Ti foam with Ve = 10 vol%.
Figure 7 presents the XRD profiles of the samples with Ve = (a) 0 vol%, (b) 5 vol%, (c) 10 vol%, and (d) 15 vol%. The peaks derived from the L12 crystal structure were detected. Not only the diffractions derived from the face-centered cubic structure (all the miller indexes (h, k, and l) are even or odd) but also the superlattice diffractions (h, k, and l are combinations of even and odd) were clearly detected, indicating the existence of L12-ordered phase. The lattice constant of the L12-ordered phase in the samples was quantified as 0.3947 nm, which was close to the previously reported lattice constant (0.3944 nm) of the L12-(Al, Fe)3Ti phase.17) Thus, the L12-(Al, Fe)3Ti phase was successfully formed by the combustion foaming process. In the sample with Ve = 0 vol%, the peaks of the Al2FeTi phase27) were also detected (Fig. 7(a)), suggesting that the combustion reaction was not completed, and intermediate products were retained. No peaks of the intermediate phases were detected in the samples with Ve = 5, 10, and 15 vol% although the peaks of TiB228) which are the reaction products of the exothermic agent were observed. Therefore, the reaction would be completed in the sample with Ve over 5 vol%. Clear peaks of TiC phase were not confirmed in this study (In Fig. 7(d), the peak is overlapped with that of TiB2 phase) presumably due to the fine morphology and/or low volume fraction of TiC phases.
XRD profiles of (Al, Fe)3Ti foams with Ve = (a) 0, (b) 5, (c) 10, and (d) 15 vol%.
Figure 8 shows the (a) low- and (b)–(c) high-magnification backscattered electron images (BEIs) showing the pore structure and microstructure in porous (Al, Fe)3Ti with Ve = (a), (b) 0 vol%, (c) 5 vol%, (d) 10 vol%, and (e) 15 vol%. Although the pores were not clearly observed in the photograph shown in Fig. 4(a), irregular-shaped pores with approximately 100 µm in size were observed in the sample with Ve = 0 vol%. The area fraction of pores was approximately 46%, which was close to the porosity in Fig. 5. The pore structure seems to be open-cell (connected) rather than closed-cell and is frequently observed in the materials fabricated by combustion reactions.29) In the cell wall, the bright phase was observed in the darker matrix. In addition, the variation in contrast was also observed in the matrix. These would be related to the intermediate phases. In the samples with Ve over 5 vol%, a needle-like phase was observed, which is TiB2 according to the previous study.20) The TiB2 phase was aggregated to form a cell structure in the sample with Ve = 5 vol%, whereas it was uniformly dispersed in the samples with Ve = 10 vol% and 15 vol%. The presence of TiC phase was not identified in all the samples under the SEM resolution in this study. To identify the TiC phase, minute observations using transmission electron microscopy (TEM) might be required.
(a) Low- and (b)–(e) high-magnification backscattered electron images (BEIs) showing the microstructures of (Al, Fe)3Ti foams with Ve = (a), (b) 0, (c) 5, (d) 10, and (e) 15 vol%.
Figure 9 shows the (a) BEIs and corresponding EDS element maps of (b) Al, (c) Ti, and (d) Fe in the sample with Ve = 0 vol%. Fe was enriched in the brightest phase, whereas Ti was enriched in slightly brighter regions in the matrix. The EDS point analysis was carried out at points 1–3 in Fig. 9(a). The composition at point 2 (in the matrix) was close to the composition of the (Al, Fe)3Ti phase. Compared with the XRD profile (Fig. 7(a)), the phase was identified as L12-(Al, Fe)3Ti phase. At point 1 (in the brightest phase), the molar ratio of Al:Ti:Fe was close to 2:1:1, and the phase was identified as the Al2FeTi phase according to the XRD profile. At point 3 (in the brighter region), the molar ratio of Al:Ti was close to 2:1. The phase would be the Al2Ti phase although further characterizations are required. The combustion reaction starts from the melting of Al powder, and then solid/liquid interfacial reactions occur around the Ti and Fe powders. The Fe-enriched Al2FeTi and Ti-enriched Al2Ti phases would be retained as intermediate phases around originally located Fe and Ti powders.
(a) BEIs and corresponding EDS element maps of (b) Al, (c) Ti, and (d) Fe of (Al, Fe)3Ti foam with Ve = 0 vol%. (e) EDS analyzed concentrations of each element obtained at points shown in (a).
Figure 10 shows the (a) BEIs and corresponding EDS element maps of (b) Al, (c) Ti, and (d) Fe in the sample with Ve = 10 vol%. The Ti enrichment in the TiB2 phase was clearly observed. The distribution of Ti and Al was almost uniform in the matrix (L12-(Al, Fe)3Ti phase). Although the concentration of the Fe phase was slightly varied inside the matrix, intermediate phases were not observed in the SEM resolution.
(a) BEI and (b)–(d) EDS element maps of (b) Al, (c) Ti, and (d) Fe of (Al, Fe)3Ti foams with Ve = 10 vol%.
When Ve was 10 vol%, porosity reached the maximum (approximately 80%, Fig. 5), and intermediate phases existing in the sample with Ve = 0 vol% (Fig. 7–9) were not detected and observed (Fig. 7 and 10). In addition, the TiB2 formed by the reaction of the exothermic agent was uniformly dispersed in the samples with Ve = 10 and 15 vol%, whereas the TiB2 was aggregated in the sample with Ve = 5 vol%. The formation sequence of highly porous L12-(Al, Fe)3Ti with uniformly dispersed TiB2 particles is discussed here.
The maximum temperature during the combustion foaming strongly depended on Ve and reached a slightly higher temperature than the liquidus temperature when Ve was 10 vol% (Fig. 3). In fact, the temperature plateau caused by releasing the latent heat during the solidification was detected (Fig. 2(c)). Therefore, the (Al, Fe)3Ti phase formed in the samples with Ve = 10 and 15 vol% would be completely melted during the combustion reactions. The sample with Ve = 0 vol% showed a maximum temperature of approximately 1150°C and had (Al, Fe)3Ti phase as a main phase and Al2FeTi/Al2Ti phases as intermediate phases. The phase region of solid (Al, Fe)3Ti phase was confirmed even at 1200°C (Fig. 1(d)), indicating that the solidus temperature is higher than 1200°C. Therefore, the (Al, Fe)3Ti phase formed in the sample with Ve = 0 vol% would be never melted during the combustion reactions. The maximum temperature of the sample with the Ve = 5 vol% reached just below the liquidus temperature. In addition, the temperature peak of this sample was broadened (Fig. 2(c)). If the sample is partially melted, the solidification latent heat would start to be released at the same time the temperature peaked. Then, the cooling rate from the peak temperature would be slowed down, resulting in a broadened temperature peak. Therefore, the (Al, Fe)3Ti phase formed in the sample with Ve = 5 vol% would be partially melted.
The combustion foaming mechanism is basically H2 gas bubble generation from the hydroxides (Al2O3·3H2O) on and/or solute H in Al powders.11,12) For the gas foaming, materials need to be partially or completely melted. In fact, the sample with Ve = 0 vol%, which would be never melted, did not exhibit a significant shape change (Fig. 4) and had an open cell rather than a gas-foamed closed-cell, whereas the samples with Ve over 5 vol%, which were partially or completely melted, exhibited significant shape changes and had closed-cell. The volume of the gas-foamed sample first expands by gas generation and/or expansion and then shrinks by escaping gas from the sample.20) The sample with Ve = 5 vol%, which would be partially melted, had a high resistance to foaming (high apparent viscosity due to the partially melted state) as well as solidified the earliest among the foamed samples due to the lowest maximum temperature. The sample with Ve = 15 vol%, which was completely melted, would be largely expanded once but shrunk due to the latest start of the solidification, until which the temperature needs to be reduced from 1500°C (maximum temperature) to 1340°C (liquidus temperature). The sample with Ve = 10 vol%, which would be also completely melted, reached the temperature just above the liquidus temperature. The sample would be largely expanded and solidified soon without experiencing a significant shrinkage. As a result, the porosity reached the maximum at Ve = 10 vol% (Fig. 5).
According to the reaction sequence during the combustion foaming from Al–Ti–B4C powders, Al3Ti is initially formed by the interfacial reaction between liquid Al and solid Ti powder.7) The growth of the Al3Ti phase leads to the aggregation of the B4C phase by pushing the B4C particles to boundaries between adjacent Al3Ti grains.20) The aggregated B4C reacts with the Al3Ti phase to form the TiB2 phase. The formation of the TiB2 phase raises the sample temperature to the melting temperature of the Al3Ti phase. The melting of the Al3Ti matrix facilitates the dissolution of elements from the intermediate phases (e.g., Ti3Al) around Ti particles and the decomposition of the intermediate phase. The Al–Ti liquid phase is formed once and then solidified as the Al3Ti after the foaming. Under the condition that Al3Ti is partially melted, the TiB2 particles are likely to be aggregated by taking over their distribution before foaming.20) The sample with Ve = 0 vol% contained intermediate phases. This would be because the maximum temperature did not reach the solidus temperature of (Al, Fe)3Ti phase, and the sample was never melted. The samples with Ve over 5 vol% did not contain the intermediate phase because the (Al, Fe)3Ti were partially or completely melted during the combustion foaming. The melting of (Al, Fe)3Ti phase and the formation of Al–Ti–Fe liquid facilitated the decomposition of intermediate phases, and the subsequent solidification generated the (Al, Fe)3Ti phase with L12 crystal structure. The sample with Ve = 5 vol% exhibited the aggregated TiB2 particles, whereas the samples with Ve = 10 and 15 vol% exhibited the uniform dispersion of TiB2 particles. The aggregation or dispersion would be attributed to the maximum temperature below or above the liquidus temperature. When the sample was completely melted, assuming a good wettability between TiB2 particles and Al-rich Al–Ti–Fe liquid at a temperature of around 1340°C, TiB2 particles would be dispersed in the liquid Al-rich phase through liquid flow. It is reported that the contact angle between TiB2 and molten Al was approximately 60° at 1100°C and tended to decrease with increasing temperature,30) supporting a good wettability between TiB2 and Al-rich liquid at around 1340°C. When the sample was partially melted, TiB2 particles could not disperse inside the regions that did not experience the melting. The formation of TiB2 particles promotes an increase in the temperature of the sample, and the temperature increase should be more pronounced where TiB2 is formed. Therefore, the temperature rise and melting occur mainly in the regions where B4C was aggregated. It is assumed that the melting would not occur in the region where the B4C was poor, and TiB2 in the foam would be aggregated, taking over the aggregation of B4C before foaming.
Thus, the reason why a highly porous L12-(Al, Fe)3Ti with uniformly dispersed TiB2 was obtained under Ve = 10 vol% would be maximum temperature controlled just above the liquidus temperature of the (Al, Fe)3Ti phase. To further clarify the sequences of the decomposition of intermediate phases and dispersion of TiB2 particles, the reaction sequence needs to be investigated by fabricating a sample with a graded reaction progress through quenching the sample during the self-propagation high-temperature synthesis (SHS) mode combustion reaction.20) In addition, to clarify the reason why the maximum temperature was optimally controlled at Ve = 10 vol%, the adiabatic combustion temperature,20) assuming the completed reaction and reaction heat were only used for raising the sample temperature, is effective. Unfortunately, the thermal properties of (Al, Fe)3Ti (specific heat, melting enthalpy, and solidus temperature) have not been reported and, therefore need to be quantified to consider the adiabatic combustion temperature. A highly porous L12-(Al, Fe)3Ti was obtained in this study under an optimized Ve = 10 vol%. The mechanical properties of the porous L12-(Al, Fe)3Ti need to be evaluated in future studies.
In this study, the combustion foaming of porous L12-(Al, Fe)3Ti was attempted. The effect of the content of exothermic agents on the porosity, constituent phases, and microstructure was investigated. The main conclusions are listed as follows:
This research was supported by JSPS KAKENHI, Grant Number JP21K14424.
