Academia.eduAcademia.edu

Figure 21 – uploaded by Sebastian Meyers

See full PDFdownloadDownload figure
Figure 3.6: LSI temperature cycle.

Figure 3 6: LSI temperature cycle.

Related Figures (112)

5 Slurry-based Indirect Laser Sintering of Alumina Ceramics
5 Slurry-based Indirect Laser Sintering of Alumina Ceramics
Figure 1.1: The powder metallurgy process for the production of ceramic parts.
Figure 1.1: The powder metallurgy process for the production of ceramic parts.
Again, densification of boron carbide can be achieved in a cost-efficient and fast manner by using reaction bonding [9]. A porous preform consisting o boron carbide, sometimes with additional free carbon, is infiltrated with molten silicon, as is the case for reaction bonded silicon carbide. The end result is a composite material, Reaction Bonded Boron Carbide (RBBC), consisting of B4C and a ternary By2(B,C,Si)3, reaction formed 6-SiC and residual Si. A typical microstructure of RBBC is presented in figure 1.3 [10]. The reaction ormed SiC in this case can originate from two different reactions [11], the first one only taking place when there is free carbon available in the preform:
Again, densification of boron carbide can be achieved in a cost-efficient and fast manner by using reaction bonding [9]. A porous preform consisting o boron carbide, sometimes with additional free carbon, is infiltrated with molten silicon, as is the case for reaction bonded silicon carbide. The end result is a composite material, Reaction Bonded Boron Carbide (RBBC), consisting of B4C and a ternary By2(B,C,Si)3, reaction formed 6-SiC and residual Si. A typical microstructure of RBBC is presented in figure 1.3 [10]. The reaction ormed SiC in this case can originate from two different reactions [11], the first one only taking place when there is free carbon available in the preform:
Table 1.3: Material properties of boron carbide materials, modified from [9], [7].
Table 1.3: Material properties of boron carbide materials, modified from [9], [7].
Figure 1.7: Indirect powder-bed fusion AM processing steps for ceramics. Indirect powder-bed fusion of ceramics has been more widely studied than direct powder-bed fusion. Research on indirect laser sintering of ceramics started in the 1990’s with the patenting of a method for the production of high-temperature parts through low-temperature sintering [24]. The typical processing route is shown in figure 1.7.
Figure 1.7: Indirect powder-bed fusion AM processing steps for ceramics. Indirect powder-bed fusion of ceramics has been more widely studied than direct powder-bed fusion. Research on indirect laser sintering of ceramics started in the 1990’s with the patenting of a method for the production of high-temperature parts through low-temperature sintering [24]. The typical processing route is shown in figure 1.7.
Figure 1.8: Over-infiltration in reaction bonded silicon carbide parts [28].
Figure 1.8: Over-infiltration in reaction bonded silicon carbide parts [28].
Figure 1.10: Direct powder-bed fusion AM processing steps for ceramics. is that there is no need for debinding, as there is no sacrificial polymer binder involved.
Figure 1.10: Direct powder-bed fusion AM processing steps for ceramics. is that there is no need for debinding, as there is no sacrificial polymer binder involved.
respectively were molten and acted as a liquid phase during sintering to bind the alumina particles [46], [47]. Wu et al. did not add a second ceramic phase to the starting powder mix. They deposited densely packed powder layers using aerosol-assisted spray deposition of optimized alumina slurries. Laser scanning caused the smaller particles in the powder bed to melt, resulting in dense single layer structures obtained by liquid phase sintering or partial melting [48).
respectively were molten and acted as a liquid phase during sintering to bind the alumina particles [46], [47]. Wu et al. did not add a second ceramic phase to the starting powder mix. They deposited densely packed powder layers using aerosol-assisted spray deposition of optimized alumina slurries. Laser scanning caused the smaller particles in the powder bed to melt, resulting in dense single layer structures obtained by liquid phase sintering or partial melting [48).
\ second approach that is used in this dissertation is discussed in chapter 4. s vastly different from the one previously studied at KU Leuven, as shown | . schematic comparison in figure 2.2. Comparing the two powder metallurs srocesses, only the first step is similar. Even so, the powder preparatio echniques used for the original KU Leuven strategy will not be applied het nstead of using a phase separation technique to produce ceramic-polym« ‘omposite micro-spherical powders, silicon and silicon carbide or boron carbic yowders are dry- mitted, as no po blended. They are then directly laser sintered. Debinding ymer binder is involved in the laser sintering process. Instea ‘arbon is infiltrated in the porous laser sintered preforms in the form of yhenolic resin that is pyrolysed. Finally, solid state sintering is replaced t iquid silicon infil tration. This approach can only be used for carbide ceramic t also incorporates reaction forming of SiC in the process, since the infiltrate ‘arbon reacts wit orm SiC. h Si during the final liquid Si infiltration densification step 1 Figure 2.2: Previous research approach (left) compared to direct laser sintering of reaction bonded carbide ceramics (right). No polymer binder is used during laser sintering. Phenolic resin infiltration is performed to provide free carbon in the laser sintered preforms. Solid state sintering is replaced with liquid silicon infiltration in order to reach fully dense carbide ceramics.
\ second approach that is used in this dissertation is discussed in chapter 4. s vastly different from the one previously studied at KU Leuven, as shown | . schematic comparison in figure 2.2. Comparing the two powder metallurs srocesses, only the first step is similar. Even so, the powder preparatio echniques used for the original KU Leuven strategy will not be applied het nstead of using a phase separation technique to produce ceramic-polym« ‘omposite micro-spherical powders, silicon and silicon carbide or boron carbic yowders are dry- mitted, as no po blended. They are then directly laser sintered. Debinding ymer binder is involved in the laser sintering process. Instea ‘arbon is infiltrated in the porous laser sintered preforms in the form of yhenolic resin that is pyrolysed. Finally, solid state sintering is replaced t iquid silicon infil tration. This approach can only be used for carbide ceramic t also incorporates reaction forming of SiC in the process, since the infiltrate ‘arbon reacts wit orm SiC. h Si during the final liquid Si infiltration densification step 1 Figure 2.2: Previous research approach (left) compared to direct laser sintering of reaction bonded carbide ceramics (right). No polymer binder is used during laser sintering. Phenolic resin infiltration is performed to provide free carbon in the laser sintered preforms. Solid state sintering is replaced with liquid silicon infiltration in order to reach fully dense carbide ceramics.
powder is used in the second step, indirect laser sintering, to produce greet parts, in an approach similar to one used at the University of Texas [5][6][7][8 A laser scans the powder bed and locally melts the polymer binder, which act as a glue to bind the ceramic particles. Indirect laser sintering can therefor be seen as the shaping step, forming a ceramic-polymer composite part wit] the required, pre-determined geometry. After obtaining the green parts, th polymer binder has to be removed by debinding the parts in a high temperatur urnace. Finally, liquid silicon infiltration is performed as an alternative to solic state sintering, which is notoriously difficult for carbide ceramics [9]. Durin; LSI, the debinded parts are placed in a vacuum furnace at high temperature i1 direct contact with a silicon (Si) source. This Si melts and infiltrates the pore in the SiC part as a result of capillary forces, aided by the good wettabilit; of SiC by Si [10][11]. In the end, a dense Si-SiC part shaped by indirect lase sintering is obtained. The goals of the study described in this chapter are:
powder is used in the second step, indirect laser sintering, to produce greet parts, in an approach similar to one used at the University of Texas [5][6][7][8 A laser scans the powder bed and locally melts the polymer binder, which act as a glue to bind the ceramic particles. Indirect laser sintering can therefor be seen as the shaping step, forming a ceramic-polymer composite part wit] the required, pre-determined geometry. After obtaining the green parts, th polymer binder has to be removed by debinding the parts in a high temperatur urnace. Finally, liquid silicon infiltration is performed as an alternative to solic state sintering, which is notoriously difficult for carbide ceramics [9]. Durin; LSI, the debinded parts are placed in a vacuum furnace at high temperature i1 direct contact with a silicon (Si) source. This Si melts and infiltrates the pore in the SiC part as a result of capillary forces, aided by the good wettabilit; of SiC by Si [10][11]. In the end, a dense Si-SiC part shaped by indirect lase sintering is obtained. The goals of the study described in this chapter are:
Starting materials for powder preparation were silicon carbide powder (either a-SiC, Carborex BW F1200, Washington Mills, purity = 98.3 %, dso = 3.2 pm or a-SiC, Carsimet F1000, Keyvest, purity = 98 %, ds) = 5 um) and semi-crystalline isotactic polypropylene (PP, Sigma-Aldrich, Mw = 12000) PP was selected as the polymer binder material since it exhibits a wide ’SLS window’, i.e., a large difference between the melting and crystallization point This property makes PP exceptionally suitable as a binder material, as a wider SLS window makes it easier to control the powderbed preheating temperature during laser sintering. Two approaches were used to create composite SiC-PP powder particles, the first being thermally induced phase separation (TIPS) and the second being spray atomisation. TIPS involves mixing the polypropylene with SiC and a suitable solvent, in this case xylene, at room temperature. The mixture is continuously stirred and heated up in order to dissolve the PP in the xylene. Afterwards, the mixture is cooled down, while stirring, in order to allow the PP to precipitate from the solution. Different morphologies can be obtained depending on the PP concentration in the solution. Figure 3.2 gives an overview of the different possibilities. As can be seen, when the solution is cooled down below the binodal curve, it is no longer homogeneous and de-mixing occurs into a polymer-rich phase and a polymer-lean phase. When the PP concentration is below a certain
Starting materials for powder preparation were silicon carbide powder (either a-SiC, Carborex BW F1200, Washington Mills, purity = 98.3 %, dso = 3.2 pm or a-SiC, Carsimet F1000, Keyvest, purity = 98 %, ds) = 5 um) and semi-crystalline isotactic polypropylene (PP, Sigma-Aldrich, Mw = 12000) PP was selected as the polymer binder material since it exhibits a wide ’SLS window’, i.e., a large difference between the melting and crystallization point This property makes PP exceptionally suitable as a binder material, as a wider SLS window makes it easier to control the powderbed preheating temperature during laser sintering. Two approaches were used to create composite SiC-PP powder particles, the first being thermally induced phase separation (TIPS) and the second being spray atomisation. TIPS involves mixing the polypropylene with SiC and a suitable solvent, in this case xylene, at room temperature. The mixture is continuously stirred and heated up in order to dissolve the PP in the xylene. Afterwards, the mixture is cooled down, while stirring, in order to allow the PP to precipitate from the solution. Different morphologies can be obtained depending on the PP concentration in the solution. Figure 3.2 gives an overview of the different possibilities. As can be seen, when the solution is cooled down below the binodal curve, it is no longer homogeneous and de-mixing occurs into a polymer-rich phase and a polymer-lean phase. When the PP concentration is below a certain
For TIPS, the SiC powder was first de-agglomerated in xylene and 5 mm zirconia (ZrO) milling balls in a polyethylene bottle on a Turbula mixer for 24 hours. Afterwards, the de-agglomerated SiC and PP were brought into xylene (p-xylene, reagent grade, Sigma-Aldrich, T, = 138°C) in different ratios. The mixture was brought into a three-neck flask, mechanically stirred and heated to 133°C, above the dissolution temperature of PP in xylene, in a silicon oil bath and under a nitrogen atmosphere. The temperature was held at 133°C for 30 minutes to allow the PP to completely dissolve. Finally, the mixture was cooled to room temperature while stirring in order to induce precipitation of composite SiC-PP particles. These particles were filtered out of the xylene using a filter paper and a vacuum pump. Afterwards, they were washed three times with ethanol in order to remove residual xylene, and then dried in air at 80°C. Figure 3.3 shows the TIPS powder fabrication process. The second technique used for powder synthesis is spray atomization. For spray atomisation, the same SiC-PP-Xylene mixture was used as for TIPS. It was heated to 133°C for 30 minutes, and then poured into the container of a spray gun (Metabo FSP 600 HVLP). The spray gun was fed with hot pressurized air (2 bar, 135°C)) in order to prevent the system from being clogged with solidified polymer. The atomized SiC-PP-xylene suspension was forced through the nozzle of the spray gun and sprayed into ethanol at room temperature for
For TIPS, the SiC powder was first de-agglomerated in xylene and 5 mm zirconia (ZrO) milling balls in a polyethylene bottle on a Turbula mixer for 24 hours. Afterwards, the de-agglomerated SiC and PP were brought into xylene (p-xylene, reagent grade, Sigma-Aldrich, T, = 138°C) in different ratios. The mixture was brought into a three-neck flask, mechanically stirred and heated to 133°C, above the dissolution temperature of PP in xylene, in a silicon oil bath and under a nitrogen atmosphere. The temperature was held at 133°C for 30 minutes to allow the PP to completely dissolve. Finally, the mixture was cooled to room temperature while stirring in order to induce precipitation of composite SiC-PP particles. These particles were filtered out of the xylene using a filter paper and a vacuum pump. Afterwards, they were washed three times with ethanol in order to remove residual xylene, and then dried in air at 80°C. Figure 3.3 shows the TIPS powder fabrication process. The second technique used for powder synthesis is spray atomization. For spray atomisation, the same SiC-PP-Xylene mixture was used as for TIPS. It was heated to 133°C for 30 minutes, and then poured into the container of a spray gun (Metabo FSP 600 HVLP). The spray gun was fed with hot pressurized air (2 bar, 135°C)) in order to prevent the system from being clogged with solidified polymer. The atomized SiC-PP-xylene suspension was forced through the nozzle of the spray gun and sprayed into ethanol at room temperature for
quenching at a flow rate of 170 1/min. When exiting the nozzle, the PP was still dissolved in the xylene, and the atomized solution forms spherical droplets When quenched in ethanol, the rapid cool-down causes the PP to precipitate around the SiC particles into spherical beads, while the xylene is dissolved in the ethanol. In this way, spray atomised composite PP-SiC microspheres are obtained. After precipitation in the xylene/ethanol mixture, the particles were again filtered out using a filter paper and a vacuum pump, and washed three times with ethanol and dried in air at 80°C. Figure 3.4 presents the spray atomisation process.
quenching at a flow rate of 170 1/min. When exiting the nozzle, the PP was still dissolved in the xylene, and the atomized solution forms spherical droplets When quenched in ethanol, the rapid cool-down causes the PP to precipitate around the SiC particles into spherical beads, while the xylene is dissolved in the ethanol. In this way, spray atomised composite PP-SiC microspheres are obtained. After precipitation in the xylene/ethanol mixture, the particles were again filtered out using a filter paper and a vacuum pump, and washed three times with ethanol and dried in air at 80°C. Figure 3.4 presents the spray atomisation process.
Table 3.1: Different composite powder batches produced.
Table 3.1: Different composite powder batches produced.
Two different SiC starting powders, with different median particle sizes, were used to produce microspheres using TIPS. The starting materials are presented in figures 3.7 (Keyvest, dso = 5 wm) and 3.8 (Washington Mills, ds59 = 3.2 um). Both starting powders have an irregular, angular morphology, as expected for silicon carbide.
Two different SiC starting powders, with different median particle sizes, were used to produce microspheres using TIPS. The starting materials are presented in figures 3.7 (Keyvest, dso = 5 wm) and 3.8 (Washington Mills, ds59 = 3.2 um). Both starting powders have an irregular, angular morphology, as expected for silicon carbide.
meta-stable region (denoted by ’A’ in figure 3.13) during cooling down. This means phase separation is not spontaneous. Here, the SiC particles present in the solution act as nuclei for precipitation and the PP tends to form and grow from these nucleation sites. However, in experiment 4, the polymer concentration has increased to the point where the unstable region (denoted by ’B’ on figure 3.13) is entered during cool-down. Here, phase separation is spontaneous and the SiC particles are no longer needed as nucleation sites. Instead, the PP demixes and forms precipitates on its own, leaving loose SiC particles in the solution. This is not ideal for the purpose of laser sintering.
meta-stable region (denoted by ’A’ in figure 3.13) during cooling down. This means phase separation is not spontaneous. Here, the SiC particles present in the solution act as nuclei for precipitation and the PP tends to form and grow from these nucleation sites. However, in experiment 4, the polymer concentration has increased to the point where the unstable region (denoted by ’B’ on figure 3.13) is entered during cool-down. Here, phase separation is spontaneous and the SiC particles are no longer needed as nucleation sites. Instead, the PP demixes and forms precipitates on its own, leaving loose SiC particles in the solution. This is not ideal for the purpose of laser sintering.
Figure 3.15: Experiment 6: powders produced by spray atomisation using Washington Mills 3.2 wm SiC starting powder with a 9 wt% solution of PP in xylene at an air temperature of 135°C, an air pressure of 2 bar and a flow rate of 170 1/min, at low (a) and high (b) magnification.
Figure 3.15: Experiment 6: powders produced by spray atomisation using Washington Mills 3.2 wm SiC starting powder with a 9 wt% solution of PP in xylene at an air temperature of 135°C, an air pressure of 2 bar and a flow rate of 170 1/min, at low (a) and high (b) magnification.
Laser sintering was performed on two of the produced TIPS powders, those from experiment 1 (Keyvest SiC starting powder, dso of 5 wm, stirring rate of 320 rpm) and experiment 3 (Washington Mills SiC starting powder, dso of 3.2 um, stirring rate of 270 rpm). Layer deposition in the DTM machine was clearly better for the Washington Mills TIPS powder, as can be seen in figure 3.18. The quality of the layer and more specifically the packing density had a large influence on the green density of the produced parts. The Keyvest SiC TIPS powder yielded parts with a relative density of 42 %, while the Washington Mills TIPS powder resulted in green parts with relative densities of around 59 %, as shown in figure 3.19. SEM images of the parts revealed a clearly visible layer structure still present in the green parts produced with TIPS Keyvest SiC-PP powder, while the TIPS Washington Mills SiC-PP powder yielded a more homogeneous structure, as presented in figure 3.20.
Laser sintering was performed on two of the produced TIPS powders, those from experiment 1 (Keyvest SiC starting powder, dso of 5 wm, stirring rate of 320 rpm) and experiment 3 (Washington Mills SiC starting powder, dso of 3.2 um, stirring rate of 270 rpm). Layer deposition in the DTM machine was clearly better for the Washington Mills TIPS powder, as can be seen in figure 3.18. The quality of the layer and more specifically the packing density had a large influence on the green density of the produced parts. The Keyvest SiC TIPS powder yielded parts with a relative density of 42 %, while the Washington Mills TIPS powder resulted in green parts with relative densities of around 59 %, as shown in figure 3.19. SEM images of the parts revealed a clearly visible layer structure still present in the green parts produced with TIPS Keyvest SiC-PP powder, while the TIPS Washington Mills SiC-PP powder yielded a more homogeneous structure, as presented in figure 3.20.
Table 3.4: Optimal laser sintering parameter sets.
Table 3.4: Optimal laser sintering parameter sets.
The laser sintered parts, i.e. the green parts, were first infiltrated with a carbon suspension and then debinded, as previously explained in section 3.2. The debinding was either done in a tube furnace or in a high temperature vacuum furnace. Afterwards, that same high temperature vacuum furnace was used tc achieve final densification by means of liquid Si infiltration. The two resulting final part fabrication routes are presented schematically in figure 3.22, and will be referred to as route 1 (tube furnace debinding) and route 2 (vacuum furnace debinding) from here on.
The laser sintered parts, i.e. the green parts, were first infiltrated with a carbon suspension and then debinded, as previously explained in section 3.2. The debinding was either done in a tube furnace or in a high temperature vacuum furnace. Afterwards, that same high temperature vacuum furnace was used tc achieve final densification by means of liquid Si infiltration. The two resulting final part fabrication routes are presented schematically in figure 3.22, and will be referred to as route 1 (tube furnace debinding) and route 2 (vacuum furnace debinding) from here on.
A thermogravimetric analysis was performed on the SiC-PP powder under argon atmosphere in order to assess the pyrolysis temperatures and behaviour of the polypropylene. The debinding started around 400°C and was completed at 470°C. The weight loss was 29%, corresponding to the weight percentage of PP in the composite SiC-PP powder (40:60 volume ratio or 70:30 weight ratio). This means that above 460°C, all of the polypropylene will have debinded and outgassed from the laser sintered part. The TGA curve is presented in figure 3.23. Debinding in a tube furnace, i.e. route 1, was done at a very slow heating rate of 0.1°C/min. However, severe crack formation and part distortion occurred as shown in dense, and t Complete de urther, and figure 3.24. The green parts, i.e. before debinding, are only 60% his 60% consists of 40% SiC and 60% PP, the latter being a very high amount. This means that after debinding, the SiC only takes up 24% o the original volume of the part, leading to loss of strength and severe distortion binding should therefore not be considered unless the green densit} of the laser sintered parts is enhanced or the polymer content is decreased. Duc to the bad part quality, debinding in the tube furnace was not investigatec the components resulting from preliminary testing were not suitable for transfer to the vacuum furnace for liquid Si infiltration due to poor shape« retention and excessive damage.
A thermogravimetric analysis was performed on the SiC-PP powder under argon atmosphere in order to assess the pyrolysis temperatures and behaviour of the polypropylene. The debinding started around 400°C and was completed at 470°C. The weight loss was 29%, corresponding to the weight percentage of PP in the composite SiC-PP powder (40:60 volume ratio or 70:30 weight ratio). This means that above 460°C, all of the polypropylene will have debinded and outgassed from the laser sintered part. The TGA curve is presented in figure 3.23. Debinding in a tube furnace, i.e. route 1, was done at a very slow heating rate of 0.1°C/min. However, severe crack formation and part distortion occurred as shown in dense, and t Complete de urther, and figure 3.24. The green parts, i.e. before debinding, are only 60% his 60% consists of 40% SiC and 60% PP, the latter being a very high amount. This means that after debinding, the SiC only takes up 24% o the original volume of the part, leading to loss of strength and severe distortion binding should therefore not be considered unless the green densit} of the laser sintered parts is enhanced or the polymer content is decreased. Duc to the bad part quality, debinding in the tube furnace was not investigatec the components resulting from preliminary testing were not suitable for transfer to the vacuum furnace for liquid Si infiltration due to poor shape« retention and excessive damage.
ceramic was residual Si, which is quite a lot. The residual Si content should be decreased in order to improve mechanical and thermal properties. Nevertheless, fully dense Si-SiC parts were obtained. SEM images of the polished cross- sectioned microstructure of the final ceramic are shown in figure 3.26.
ceramic was residual Si, which is quite a lot. The residual Si content should be decreased in order to improve mechanical and thermal properties. Nevertheless, fully dense Si-SiC parts were obtained. SEM images of the polished cross- sectioned microstructure of the final ceramic are shown in figure 3.26.
Table 3.5: Composite SiC-PP powders used for indirect laser sintering. Indirect laser sintering was successfully used to produce green SiC-PP parts with both powders. However, powder 2 clearly resulted in higher green densities, with an average of 59% compared to 42% for powder 1. Part densities were studied as a function of the laser energy density. At higher laser energy densities, the heat input resulted in polymer burning, degradation and consequently lower densities. Optimal parameter sets were defined for both density and dimensional accuracy.
Table 3.5: Composite SiC-PP powders used for indirect laser sintering. Indirect laser sintering was successfully used to produce green SiC-PP parts with both powders. However, powder 2 clearly resulted in higher green densities, with an average of 59% compared to 42% for powder 1. Part densities were studied as a function of the laser energy density. At higher laser energy densities, the heat input resulted in polymer burning, degradation and consequently lower densities. Optimal parameter sets were defined for both density and dimensional accuracy.
Silicon powder (Si, Simet 985, Keyvest, purity = 98%, dso = 45 um) was dry blended with silicon carbide powder (a-SiC, Carborex BW F320, Washington Mills, purity = 99.2%, dso = 29 wm) in a 40:60 vol% ratio in a pan mixing device (Eirich 1-Liter Mixer EL1, Maschinenfabrik Gustav Eirich GmbH & Co KG). A mixing speed of 2400 rpm and time of 5 minutes were used. The morphology of the starting powders is shown in figures 4.3 (Si) and 4.4 (SiC). Figure 4.3: SEM image of the Si starting powder (dso = 45m).
Silicon powder (Si, Simet 985, Keyvest, purity = 98%, dso = 45 um) was dry blended with silicon carbide powder (a-SiC, Carborex BW F320, Washington Mills, purity = 99.2%, dso = 29 wm) in a 40:60 vol% ratio in a pan mixing device (Eirich 1-Liter Mixer EL1, Maschinenfabrik Gustav Eirich GmbH & Co KG). A mixing speed of 2400 rpm and time of 5 minutes were used. The morphology of the starting powders is shown in figures 4.3 (Si) and 4.4 (SiC). Figure 4.3: SEM image of the Si starting powder (dso = 45m).
After gaining knowledge on the morphology and size of the powders, the powder flowability was measured. This was done using the dynamic angle of repose, a generally accepted characterisation technique to measure flowability of powders [15]. The in-house test set-up uses a rotating cylinder (3 rpm, 2 min) that is filled with 12.5 ml of powder. The angle of repose is measured as the angle of the rotating powder mass relative to the horizontal plane. This angle is measured frame by frame using a camera and a matlab script, see figure 4.8, as a function of rotation time.
After gaining knowledge on the morphology and size of the powders, the powder flowability was measured. This was done using the dynamic angle of repose, a generally accepted characterisation technique to measure flowability of powders [15]. The in-house test set-up uses a rotating cylinder (3 rpm, 2 min) that is filled with 12.5 ml of powder. The angle of repose is measured as the angle of the rotating powder mass relative to the horizontal plane. This angle is measured frame by frame using a camera and a matlab script, see figure 4.8, as a function of rotation time.
Figure 4.10: A set of laser sintered cubes, with a high power, high scanning speed set (left, notice the poor part quality) and a low power, low scanning speed set (right). Base area of one cube is 10 by 10 mm?.
Figure 4.10: A set of laser sintered cubes, with a high power, high scanning speed set (left, notice the poor part quality) and a low power, low scanning speed set (right). Base area of one cube is 10 by 10 mm?.
with P the laser power in W and v the scanning speed in mm/s. The laser energy therefore increases with increasing laser power P and decreasing scanning speed v. Cubes made with a laser power below 13 W were not consolidated, except at the lowest scanning speed of 50 mm/s. This is caused by a lack-of-fusion issue due to insufficient laser energy being delivered to the powder bed to melt the silicon. Cubes made with laser powers above 21 W crumbled and lost their geometrical accuracy. Some of the silicon in the powder bed likely started to evaporate at these higher laser energy densities and the SiC in the powder bed decomposed into Si and C. Good results were obtained with laser powers
with P the laser power in W and v the scanning speed in mm/s. The laser energy therefore increases with increasing laser power P and decreasing scanning speed v. Cubes made with a laser power below 13 W were not consolidated, except at the lowest scanning speed of 50 mm/s. This is caused by a lack-of-fusion issue due to insufficient laser energy being delivered to the powder bed to melt the silicon. Cubes made with laser powers above 21 W crumbled and lost their geometrical accuracy. Some of the silicon in the powder bed likely started to evaporate at these higher laser energy densities and the SiC in the powder bed decomposed into Si and C. Good results were obtained with laser powers
Table 4.1: Relative density of preforms with different laser powers and scanning speeds in the optimal laser processing window, as shown in figure 4.15.
Table 4.1: Relative density of preforms with different laser powers and scanning speeds in the optimal laser processing window, as shown in figure 4.15.
of 15 W and a scanning speed of 100 mm/s. This scanning speed is used as a trade-off between post-SLS SiC content and production speed, as scanning speeds of 50 mm/s lead to very long build times. Figure 4.17 presents a heat sink component with cooling ribs after SLS and after liquid silicon infiltration. Figure 4.18 presents a laser sintered and a liquid Si infiltrated gear.
of 15 W and a scanning speed of 100 mm/s. This scanning speed is used as a trade-off between post-SLS SiC content and production speed, as scanning speeds of 50 mm/s lead to very long build times. Figure 4.17 presents a heat sink component with cooling ribs after SLS and after liquid silicon infiltration. Figure 4.18 presents a laser sintered and a liquid Si infiltrated gear.
The phenolic resin was infiltrated into the laser sintered preforms and pyrolysed to yield porous carbon. In order to gain information on the temperature range for the pyrolysis of the resin, a thermogravimetric analysis was performed, as shown in figure 4.19. A first weight loss occurred around 100°C and corresponds to water evaporation from the resin. The subsequent steady weight loss virtually levels off around 600°C’. The carbon yield at that point amounts to about 6( wt% of the original weight of the resin. This leads to high amounts of porous carbon being present in the impregnated Si-SiC preform, as can be seen on the SEM pictures of impregnated preforms (treatment I2) in figures 4.20 and 4.21 This is important for the formation of reaction-formed SiC during the liquid silicon infiltration. The reaction taking place during liquid silicon infiltration is:
The phenolic resin was infiltrated into the laser sintered preforms and pyrolysed to yield porous carbon. In order to gain information on the temperature range for the pyrolysis of the resin, a thermogravimetric analysis was performed, as shown in figure 4.19. A first weight loss occurred around 100°C and corresponds to water evaporation from the resin. The subsequent steady weight loss virtually levels off around 600°C’. The carbon yield at that point amounts to about 6( wt% of the original weight of the resin. This leads to high amounts of porous carbon being present in the impregnated Si-SiC preform, as can be seen on the SEM pictures of impregnated preforms (treatment I2) in figures 4.20 and 4.21 This is important for the formation of reaction-formed SiC during the liquid silicon infiltration. The reaction taking place during liquid silicon infiltration is:
was made concerning the expected final amount of SiC in the part, assuming that the added carbon is fully converted to 6-SiC. The weight values in the table are based on a preform (and final part) volume of 1 cm*. Values in bold are measured data for the I1 and I2 impregnation cycles.
was made concerning the expected final amount of SiC in the part, assuming that the added carbon is fully converted to 6-SiC. The weight values in the table are based on a preform (and final part) volume of 1 cm*. Values in bold are measured data for the I1 and I2 impregnation cycles.
according to phenolic impregnation according to phenolic impregnation treatment [2, with a high amount of treatment [2, with a high amount of
according to phenolic impregnation according to phenolic impregnation treatment [2, with a high amount of treatment [2, with a high amount of
Figure 4.22: Light optical micrograph of a dense Si-SiC part after Si infiltration without phenolic resin impregnation. Liquid silicon infiltration leads to fully dense reaction bonded silicon carbide material, as illustrated in figures 4.22 and 4.23 for parts that were scanned with a laser power of 15 W, a scanning speed of 100 mm/s, a scan spacing of 77 wm and a powder layer thickness of 30 um. Image analysis techniques were used to study light microscopy images and analyse the amount of silicon carbide in the final structure. If the phenolic resin impregnation step is omitted, ne 6-SiC is formed since there is no carbon for the Si melt to react with. This leads to microstructures as shown in figure 4.22, with darker grey a-SiC grains embedded in a light grey Si matrix. The structure is dense, but the SiC content is only about 35 vol%.
Figure 4.22: Light optical micrograph of a dense Si-SiC part after Si infiltration without phenolic resin impregnation. Liquid silicon infiltration leads to fully dense reaction bonded silicon carbide material, as illustrated in figures 4.22 and 4.23 for parts that were scanned with a laser power of 15 W, a scanning speed of 100 mm/s, a scan spacing of 77 wm and a powder layer thickness of 30 um. Image analysis techniques were used to study light microscopy images and analyse the amount of silicon carbide in the final structure. If the phenolic resin impregnation step is omitted, ne 6-SiC is formed since there is no carbon for the Si melt to react with. This leads to microstructures as shown in figure 4.22, with darker grey a-SiC grains embedded in a light grey Si matrix. The structure is dense, but the SiC content is only about 35 vol%.
An XRD measurement was performed on cube 3 to confirm that the phases seen in light microscopy were indeed Si, a-SiC and 6-SiC, as presented in figure 4.24. Peaks for Si are visible, as well as coinciding peaks for a-SiC and 6-SiC. The presence of these phases was quantified using Rietveld analysis, yielding 27.5 vol% a-SiC, 56.1 vol% 8-SiC and 16.4 vol% Si. This points to a SiC content of 83.6 vol%, which is very close to the 84.9 + 1.3 vol% measured by light microscopy image analysis on the same sample and the composition predicted in table 4.5. Table 4.6: A comparison between Si-SiC cubes made with different effect. A comparison between 3 cubes with different post-processing cycles is made in table 4.6. These cubes were all laser sintered with a scan spacing of 77 wm and a layer thickness of 30 um. Cube 1 (figure 4.22) was not subjected to carbon impregnation, cube 2 was subjected to phenolic resin impregnation treatment I1 and cube 3 (figure 4.23) underwent phenolic resin impregnation treatment 12. It is clear that the additional impregnation and pyrolysis results in the formation of an additional 21 vol% SiC after Si infiltration.
An XRD measurement was performed on cube 3 to confirm that the phases seen in light microscopy were indeed Si, a-SiC and 6-SiC, as presented in figure 4.24. Peaks for Si are visible, as well as coinciding peaks for a-SiC and 6-SiC. The presence of these phases was quantified using Rietveld analysis, yielding 27.5 vol% a-SiC, 56.1 vol% 8-SiC and 16.4 vol% Si. This points to a SiC content of 83.6 vol%, which is very close to the 84.9 + 1.3 vol% measured by light microscopy image analysis on the same sample and the composition predicted in table 4.5. Table 4.6: A comparison between Si-SiC cubes made with different effect. A comparison between 3 cubes with different post-processing cycles is made in table 4.6. These cubes were all laser sintered with a scan spacing of 77 wm and a layer thickness of 30 um. Cube 1 (figure 4.22) was not subjected to carbon impregnation, cube 2 was subjected to phenolic resin impregnation treatment I1 and cube 3 (figure 4.23) underwent phenolic resin impregnation treatment 12. It is clear that the additional impregnation and pyrolysis results in the formation of an additional 21 vol% SiC after Si infiltration.
Material properties investigated include electrical resistivity, Young’s modulus and flexural strength. The tests were done on 8 3x4x45 mm® beams. These beams were laser sintered with a laser power of 15 W, a scanning speed of 100 mm/s, a scan spacing of 77 wm and a powder layer thickness of 30 wm. The resistivity of the beams was 2.0 + 0.7 * 10-4Qm. The conductivity o, calculated as the inverse of the resistivity, is 5.3 + 0.9 * 10° S/m. These values point to a moderate electrical conductivity.
Material properties investigated include electrical resistivity, Young’s modulus and flexural strength. The tests were done on 8 3x4x45 mm® beams. These beams were laser sintered with a laser power of 15 W, a scanning speed of 100 mm/s, a scan spacing of 77 wm and a powder layer thickness of 30 wm. The resistivity of the beams was 2.0 + 0.7 * 10-4Qm. The conductivity o, calculated as the inverse of the resistivity, is 5.3 + 0.9 * 10° S/m. These values point to a moderate electrical conductivity.
Table 4.8: Optimal laser sintering parameter sets for RBSC and RBBC. Direct laser sintering was used as a shaping step to form porous preform: for subsequent infiltration towards fully dense reaction bonded silicon carbid (RBSC) and reaction bonded boron carbide (RBBC) material. For the RBSC two different optimized sets of laser sintering parameters were obtained, one fo: maximum density and one for optimal dimensional accuracy. For the RBBC, : parameters set was used that yielded good parts, but this set was not optimizec to the same extent as was the case for RBSC. The parameter sets are summarizec in table 4.8. Parameters used for further research were 15 W and 100 mm/: or the RBSC, as a good compromise between density and accuracy, and 30 W and 50 mm/s for RBBC.
Table 4.8: Optimal laser sintering parameter sets for RBSC and RBBC. Direct laser sintering was used as a shaping step to form porous preform: for subsequent infiltration towards fully dense reaction bonded silicon carbid (RBSC) and reaction bonded boron carbide (RBBC) material. For the RBSC two different optimized sets of laser sintering parameters were obtained, one fo: maximum density and one for optimal dimensional accuracy. For the RBBC, : parameters set was used that yielded good parts, but this set was not optimizec to the same extent as was the case for RBSC. The parameter sets are summarizec in table 4.8. Parameters used for further research were 15 W and 100 mm/: or the RBSC, as a good compromise between density and accuracy, and 30 W and 50 mm/s for RBBC.
Table 4.9: Properties of laser sintered Si-SiC and Si-B4C compared to commercially available Si-SiC by Schunk GmbH. The objectives set out at the start of this chapter have been accomplished:
Table 4.9: Properties of laser sintered Si-SiC and Si-B4C compared to commercially available Si-SiC by Schunk GmbH. The objectives set out at the start of this chapter have been accomplished:
that allows an even slurry supply over a certain width on the baseplate of the machine. The deposited slurry is then spread and paved into a thin layer by a conventional coater system. After layer deposition and during laser scanning, the coater is submerged in a water bath in order to prevent the remaining slurry from drying out and forming hard powder residues, as these initially caused a decrease in the quality of subsequent layers. The process chamber of the refurbished slurry-based laser sintering machine is shown in figure 5.3.
that allows an even slurry supply over a certain width on the baseplate of the machine. The deposited slurry is then spread and paved into a thin layer by a conventional coater system. After layer deposition and during laser scanning, the coater is submerged in a water bath in order to prevent the remaining slurry from drying out and forming hard powder residues, as these initially caused a decrease in the quality of subsequent layers. The process chamber of the refurbished slurry-based laser sintering machine is shown in figure 5.3.
Figure 5.4: Schematic view of the deposition nozzle, with the slurry in blue (left) and the real component (right) [14].
Figure 5.4: Schematic view of the deposition nozzle, with the slurry in blue (left) and the real component (right) [14].
In the first iteration of the slurry-based laser sintering process, a techniqu called ’pre-scanning’ was used. This is defined as scanning the deposited slurr layer at a moderate laser power and a high scanning speed, so a low energ density, in order to dry the slurry layers and evaporate the water more quickl without melting the polymer binder in the slurry. After pre-scanning a larg square area of the newly deposited layer, smaller square sample areas were lase scanned and sintered. This is schematically shown in figure 5.11.
In the first iteration of the slurry-based laser sintering process, a techniqu called ’pre-scanning’ was used. This is defined as scanning the deposited slurr layer at a moderate laser power and a high scanning speed, so a low energ density, in order to dry the slurry layers and evaporate the water more quickl without melting the polymer binder in the slurry. After pre-scanning a larg square area of the newly deposited layer, smaller square sample areas were lase scanned and sintered. This is schematically shown in figure 5.11.
The average density of the samples is 73%, and is hardly influenced by the upplied laser energy density. A typical optical microstructure is shown in figure 5.13. Dense areas can be observed, but they are interrupted by highly porous areas in a characteristic raster pattern. This pattern is likely caused by the laser pre-scanning. In order to confirm this, a benchmark part was ana After leaving the slurry deposit in the machine to naturally dry out comp an unscanned region of the alumina-PVA deposit (see figure 5.11, blue was debinded and furnace sintered. The density of this benchmark part ysed. etely, part) was 97%, which points to the interaction of the laser with the deposited layer as the cause for the porosity.
The average density of the samples is 73%, and is hardly influenced by the upplied laser energy density. A typical optical microstructure is shown in figure 5.13. Dense areas can be observed, but they are interrupted by highly porous areas in a characteristic raster pattern. This pattern is likely caused by the laser pre-scanning. In order to confirm this, a benchmark part was ana After leaving the slurry deposit in the machine to naturally dry out comp an unscanned region of the alumina-PVA deposit (see figure 5.11, blue was debinded and furnace sintered. The density of this benchmark part ysed. etely, part) was 97%, which points to the interaction of the laser with the deposited layer as the cause for the porosity.
Figure 5.13: Optical micrograph of a polished top surface of an alumina part laser sintered with a power of 8 W, a scanning speed of 400 mm/s, a scan spacing of 70 wm and a layer thickness of 30 wm. Pre-scanning was done with a laser power of 15 W and a scanning speed of 2000 mm/s. A characteristic porosity pattern can be observed.
Figure 5.13: Optical micrograph of a polished top surface of an alumina part laser sintered with a power of 8 W, a scanning speed of 400 mm/s, a scan spacing of 70 wm and a layer thickness of 30 wm. Pre-scanning was done with a laser power of 15 W and a scanning speed of 2000 mm/s. A characteristic porosity pattern can be observed.
Laser pre-scanning to evaporate the water from the slurry layers resulted in part: with an average density of 73%. A characteristic porosity pattern was identified Its origin was attributed to the laser pre-scanning resulting in a too intenss water evaporation and a concomitant particle flow, lead alumina-poor areas in the green part. This leads to in and solid-state sintering, yielding fully dense areas inter ing to alumina-rich anc homogeneous debinding rupted by highly porous areas in the final alumina parts. However, unscanned layers were also debindec and solid-state sintered in a furnace, and were conver ted into closed porosity alumina parts with a density of 97%, further supporting the hypothesis that the laser pre-scanning caused the unusual porosity distribution in the laser scannec parts.
Laser pre-scanning to evaporate the water from the slurry layers resulted in part: with an average density of 73%. A characteristic porosity pattern was identified Its origin was attributed to the laser pre-scanning resulting in a too intenss water evaporation and a concomitant particle flow, lead alumina-poor areas in the green part. This leads to in and solid-state sintering, yielding fully dense areas inter ing to alumina-rich anc homogeneous debinding rupted by highly porous areas in the final alumina parts. However, unscanned layers were also debindec and solid-state sintered in a furnace, and were conver ted into closed porosity alumina parts with a density of 97%, further supporting the hypothesis that the laser pre-scanning caused the unusual porosity distribution in the laser scannec parts.
Figure 5.17: Light microscopy images of two fragile square components produced by vacuum-draining assisted laser sintering, showing cracks. Parts produced with a laser power of 3 W and a scanning speed of 100 mm/s (left) and a laser power of 8 W and a scanning speed of 300 mm/s (right).
Figure 5.17: Light microscopy images of two fragile square components produced by vacuum-draining assisted laser sintering, showing cracks. Parts produced with a laser power of 3 W and a scanning speed of 100 mm/s (left) and a laser power of 8 W and a scanning speed of 300 mm/s (right).
Table 6.1: Optimal laser sintering parameter sets for RBSC and RBBC. Table 6.1: Optimal laser sintering parameter sets for RBSC and RBBC. The silicon carbide content in these parts was increased through a phenolic resin infiltration and pyrolysis process, which yielded carbon that could be converted to SiC during liquid silicon infiltration. Table 6.2 summarises the mechanical properties of the Si-infiltrated carbide ceramics produced through direct laser sintering and compares them to a commercially available alternative.
Table 6.1: Optimal laser sintering parameter sets for RBSC and RBBC. Table 6.1: Optimal laser sintering parameter sets for RBSC and RBBC. The silicon carbide content in these parts was increased through a phenolic resin infiltration and pyrolysis process, which yielded carbon that could be converted to SiC during liquid silicon infiltration. Table 6.2 summarises the mechanical properties of the Si-infiltrated carbide ceramics produced through direct laser sintering and compares them to a commercially available alternative.
For the SiC materials, it is clear that a decrease in residual Si content leads to an increase in Young’s modulus and flexural strength. The B4C ceramic has an even higher Young’s modulus, but a lower flexural strength. The commercially available IntrinSiC material, which is a binder jetted SiC produced by Schunk GmbH, has properties that are comparable to those of the high SiC-content RBSC produced via direct laser sintering. Both processes have their merits. Binder jetting works with coarser starting powder and higher layer thicknesses, speeding up the processing time. Laser sintering, on the other hand, uses a lower layer thickness, finer starting powder and a small laser spot size, yielding a slower but more precise process. This leads to the conclusion that both processes are complementary: binder jetting can be used for large parts, whereas laser sintering is more suited for the production of intricate parts with fine features.
For the SiC materials, it is clear that a decrease in residual Si content leads to an increase in Young’s modulus and flexural strength. The B4C ceramic has an even higher Young’s modulus, but a lower flexural strength. The commercially available IntrinSiC material, which is a binder jetted SiC produced by Schunk GmbH, has properties that are comparable to those of the high SiC-content RBSC produced via direct laser sintering. Both processes have their merits. Binder jetting works with coarser starting powder and higher layer thicknesses, speeding up the processing time. Laser sintering, on the other hand, uses a lower layer thickness, finer starting powder and a small laser spot size, yielding a slower but more precise process. This leads to the conclusion that both processes are complementary: binder jetting can be used for large parts, whereas laser sintering is more suited for the production of intricate parts with fine features.
Figure 6.2: The powder metallurgy process for indirect slurry-based laser sintering of alumina ceramics. Most research efforts worldwide with respect to laser sintering of alumina ceramics have focused on a dry powder starting material consisting of a polymer binder and alumina particles, combined with indirect laser sintering, de-binding and solid-state sintering in a furnace. However, the literature study pointed to some inherent weaknesses in this process chain, which mainly stem from the requirement for high amounts of polymer binder, i.e. up to 60 vol%, caused by the low powder packing density. These lead to low green and sintered densities. In order to solve this, a slurry-based approach was explored in this dissertation, allowing for lower amounts of polymer binder due to the more densely packed layers yielded by slurry deposition and solvent removal. The investigated powder metallurgy process is shown in figure 6.2. metallurgy process is shown in figure 6.2.
Figure 6.2: The powder metallurgy process for indirect slurry-based laser sintering of alumina ceramics. Most research efforts worldwide with respect to laser sintering of alumina ceramics have focused on a dry powder starting material consisting of a polymer binder and alumina particles, combined with indirect laser sintering, de-binding and solid-state sintering in a furnace. However, the literature study pointed to some inherent weaknesses in this process chain, which mainly stem from the requirement for high amounts of polymer binder, i.e. up to 60 vol%, caused by the low powder packing density. These lead to low green and sintered densities. In order to solve this, a slurry-based approach was explored in this dissertation, allowing for lower amounts of polymer binder due to the more densely packed layers yielded by slurry deposition and solvent removal. The investigated powder metallurgy process is shown in figure 6.2. metallurgy process is shown in figure 6.2.
Table 6.3: Slurry-based laser sintering experiments with different solvent removal methods. Resistive baseplate pre-heating was also used for both laser pre-scanning and top-down pre-heating experiments, and was kept constant at 50°C. Solvent removal prior to laser scanning proved to be a key factor influencing the quality of the final parts. Evaporation through heating, either by laser scanning or infrared top-down pre-heating, resulted in parts with fully dense areas interrupted by pores or cracks. However, components with closed porosity, i.e. densities over 95%, were produced with infrared top-down pre-heating. Vacuum draining, on the other hand, while promising on paper, did not deliver the expected results. The solvent could not be sufficiently drained through the porous substrate, resulting in fragile parts that broke upon removal from the powder cake. More effort should be spent on optimizing the set-up in order to obtain acceptable solvent removal from the slurry layers.
Table 6.3: Slurry-based laser sintering experiments with different solvent removal methods. Resistive baseplate pre-heating was also used for both laser pre-scanning and top-down pre-heating experiments, and was kept constant at 50°C. Solvent removal prior to laser scanning proved to be a key factor influencing the quality of the final parts. Evaporation through heating, either by laser scanning or infrared top-down pre-heating, resulted in parts with fully dense areas interrupted by pores or cracks. However, components with closed porosity, i.e. densities over 95%, were produced with infrared top-down pre-heating. Vacuum draining, on the other hand, while promising on paper, did not deliver the expected results. The solvent could not be sufficiently drained through the porous substrate, resulting in fragile parts that broke upon removal from the powder cake. More effort should be spent on optimizing the set-up in order to obtain acceptable solvent removal from the slurry layers.
Figure A.1: The three stages of granulation [4].
Figure A.1: The three stages of granulation [4].
The results of these two experiments are presented in figure A.5.
The results of these two experiments are presented in figure A.5.
laser sintering. Particle size analysis was performed on the agglomerates produced by addition of 50 ml of binder solution at a speed of 1000 rpm and a subsequent increase in rotation speed to 2000 rpm for 5 minutes. The particle size distribution presented in figure A.6, shows that most formed agglomerates are smaller thar 10 wm in diameter, which is too small for laser sintering. However, a small secondary peak can be observed between 30 and 40 wm. Sieving out the smal. fractions or further optimization of the granulation parameters could yield powders with a mean diameter around 30 um which would be more suitable for aser sintering. Angle of repose measurements were carried out on the produced agglomerates and the results were compared to that of a standard mixed (but not agglomerated) Si-SiC powder, as used in Chapter 4. The angle of repose curves are presented in figure A.7.
laser sintering. Particle size analysis was performed on the agglomerates produced by addition of 50 ml of binder solution at a speed of 1000 rpm and a subsequent increase in rotation speed to 2000 rpm for 5 minutes. The particle size distribution presented in figure A.6, shows that most formed agglomerates are smaller thar 10 wm in diameter, which is too small for laser sintering. However, a small secondary peak can be observed between 30 and 40 wm. Sieving out the smal. fractions or further optimization of the granulation parameters could yield powders with a mean diameter around 30 um which would be more suitable for aser sintering. Angle of repose measurements were carried out on the produced agglomerates and the results were compared to that of a standard mixed (but not agglomerated) Si-SiC powder, as used in Chapter 4. The angle of repose curves are presented in figure A.7.
While the mean angle of repose of the agglomerates is lower than that of the standard powder mix, the avalanche angles are higher and the variation of this angle is much more pronounced, ranging from as low as 10 to as high as 70 degrees. This points to a very poor flowability and packing density, as was also observed during layer deposition in the laser sintering machine.
While the mean angle of repose of the agglomerates is lower than that of the standard powder mix, the avalanche angles are higher and the variation of this angle is much more pronounced, ranging from as low as 10 to as high as 70 degrees. This points to a very poor flowability and packing density, as was also observed during layer deposition in the laser sintering machine.
Related topics:
Materials ScienceSelective Laser Sintering
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved