3.1. Effects of particle size
Figure 3 illustrates the influence of different milling times on the particle size of the vitrified bond. It was observed that as the milling time increased, the particle size gradually decreased. The d50 values of the vitrified bond were approximately 36.2 µm, 17.8 µm, 9.2 µm, and 4.8 µm for milling times of 1 h, 2 h, 4 h, and 6 h, respectively. It shows that prolonging the ball milling time increases the number and energy of collisions between the balls and particles, which favors more particles breaking, thus reducing the particle size.
The different particle sizes of vitrified bond were mixed uniformly with the pre-prepared photosensitive mixture system (without dispersant) and ball-milled for 2 hours to obtain the vitrified bond slurry. The effect of different particle sizes on the viscosity of the slurry was compared through rheological analysis, as shown in Fig. 4. It was found that for vitrified bond with d50 values of 36.2 µm, 17.8 µm, 9.2 µm, and 4.8 µm, the viscosities of the slurry at a shear rate of 30 s− 1 were 1.42 Pa·s, 1.47 Pa·s, 1.50 Pa·s, and 1.54 Pa·s, respectively. This is because as the particle size decreases, the specific surface area of the particles increases, resulting in higher slurry viscosity. Combined with Fig. 1, although the viscosity of the slurry system was lowest when the particle size d50 was 36.2 µm for 1 h of ball milling, some of the particles exceeded the thickness of the slicing layer by 50 µm, which might damage the equipment when printing, so the particle size d50 of 17.8 µm was selected for the ball milling time of 2 h.
3.2. Effects of dispersant
Vitrified bond diamond wheels must possess sufficient strength to meet the demands of the grinding process, which requires a good dispersion of the bond and abrasive in the slurry. If the dispersion is not good, in the printing process, a certain area will produce less raw material and more resin, which will lead to defects in the area during sintering. Therefore, in order to obtain a well-dispersed slurry, we investigated the effects of five dispersants on the viscosity of the slurry, taking into account existing research reports [27–29, 34–37], and set up a control group without any dispersant. The results are shown in Fig. 5. It was found that PEG with a molecular weight of 200, OA, SA, SS, and SC (added at 5.0 wt.% relative to the powder weight) as dispersants, as well as the control group without dispersant, all exhibited shear-thinning behavior. This indicates that all five dispersants and the control group meet the minimum requirements for printing. At a shear rate of 30 s− 1, the viscosity of the slurry with the five dispersants was 0.48 Pa·s, 0.52 Pa·s, 0.92 Pa·s, 2.22 Pa·s, and 3.42 Pa·s, respectively, while the viscosity of the slurry without dispersant was 1.32 Pa·s. When PEG was used as a dispersant, the slurry exhibited the lowest viscosity, indicating its ability to effectively reduce particle collisions resulting from Brownian motion and minimize the friction between particles in the slurry, thereby reducing the slurry viscosity. On the other hand, SS and SC, when used as dispersants, the viscosity beyond that of the slurry without dispersant, indicating their tendency to promote particle aggregation and collision in the slurry without effectively reducing the slurry viscosity.
Considering the significant effect of PEG with a molecular weight of 200, OA, and SA in reducing the viscosity of the slurry, the influence of their addition at various concentrations on the slurry viscosity was compared. The concentrations tested were 1.0, 3.0, 5.0, and 7.0 wt.% (relative to the total weight of the powder), as illustrated in Fig. 6. The research revealed that with an increase in the content of these three dispersants, the viscosity of the slurry showed a decreasing trend followed by an increasing trend, the minimum value of viscosity was obtained at an addition level of 5.0 wt.%. This indicates that when the concentration of the dispersant is too low, not all vitrified bond particles in the slurry are fully adsorbed by the dispersant, leading to ineffective modification. Due to the particle collisions resulting from Brownian motion, the particles not covered by the dispersant tend to agglomerate, resulting in poor dispersion and high viscosity. At an appropriate concentration, the dispersant forms a network structure within the slurry and creates an organic protective layer on the particle surfaces, preventing particle collisions and improving dispersion, thereby reducing viscosity. However, at high dispersant concentrations, a large amount of free dispersant bridges and crosslinks between the particles. According to the DLVO theory [38], bridging can cause flocculation, increasing the internal friction of the slurry and reducing its stability, leading to an increase in viscosity. Therefore, PEG as a dispersant at a concentration of 5.0 wt.% is more suitable.
To determine the effect of molecular weight on the viscosity of the vitrified bond slurry, PEG with molecular weights of 200, 400, 800, 1000, and 2000 were investigated, as shown in Fig. 7. The study revealed that as the molecular weight increased, the viscosity of the slurry also increased. At a shear rate of 30 s− 1, the viscosities were measured to be 0.48, 0.60, 1.26, and 2.07 Pa·s for PEG with molecular weights of 200, 400, 800, and 1000, respectively. This can be attributed to the fact that as the molecular weight of the dispersant increases, the molecules become longer and heavier, leading to a more complex spatial arrangement between the molecules. The intermolecular forces, such as van der Waals forces and electrostatic attraction, among these molecules, increase the tendency of the dispersant molecules to aggregate within the slurry. This aggregation increases the intermolecular friction within the slurry, resulting in an increase in viscosity. Therefore, PEG with a molecular weight of 200 is more suitable as a dispersant.
The stability of vitrified bond slurry can be primarily measured through sedimentation tests. The prepared slurry is placed in a 10 ml graduated cylinder, and the height of the supernatant liquid (H1) and the initial height of the slurry (H0) are recorded every 12 h. The value of H1/H0 is then calculated, and a lower value of H1/H0 indicates higher slurry stability. In order to obtain well-stabilized vitrified bond slurry, the effects of three dispersants that significantly reduced the viscosity of the slurry, PEG with molecular weight of 200, OA, and SA (added at 5.0 wt.% relative to the total weight of the powders), were compared with the effects of a blank control group with no dispersant on the stability of the slurry, as shown in Fig. 8. Figure 8(a), 8(b), 8(c), and 8(d) represent the sedimentation behavior of slurry samples with PEG, OA, SA, and no dispersant, respectively, after 120 h of sedimentation, while Fig. 8(e) shows the H1/H0 values. The study revealed that after 120 hours of sedimentation, the heights of the supernatant liquid (H1) for the four dispersants were 1.6, 3.7, 2.9, and 7.5 ml, respectively. PEG, as a dispersant, exhibited the smallest H1/H0 value, which maintained the best stability of the slurry. This can be attributed to the excellent coverage of the PEG molecular chains on the surface of the vitrified bond particles, effectively reducing the contact and collision between the particles, resulting in slower sedimentation of the slurry. The slurry system without a dispersant showed the maximum height of the supernatant liquid in a short period, indicating severe aggregation and flocculation within the slurry system, leading to strong sedimentation and a faster sedimentation rate, making the slurry least stable.
3.3. Effects of solid content
On the premise of guaranteeing the solid content of the vitrified bond slurry, lower viscosity is conducive to improving the dispersion of the slurry to meet the requirements of UV-curing printing. And in order to meet the requirements of the rheological properties of the slurry under the premise of raising the solid content, it helps to improve the printing accuracy and is conducive to reducing the proportion of organic matter. In the later degreasing and sintering processes, it is conducive to avoiding defects such as cracking of raw blanks and ensuring that the densification of the wheel embryo body and grinding processing performance.
To obtain a vitrified bond slurry with high solid content and low viscosity, we investigated the rheological performance of the slurry with PEG of molecular weight 200 (added at 5.0 wt.% relative to the total weight of the powder) as a dispersant and different solid contents, as shown in Fig. 9. Figure 9(a) illustrates the variation of slurry viscosity with shear rate. It was found that the slurry viscosity gradually decreased as the shear rate increased from 1 s− 1 to 200 s− 1. This indicates that the slurry with different solid contents exhibits shear-thinning behavior and good dispersion. Figure 9(b) presents the scatter plot of slurry viscosity at a shear rate of 30 s− 1 for different solid contents, ranging from 40 wt.% to 65 wt.%. The viscosities of the slurry were 0.48, 0.77, 0.90, 1.65, 2.73, and 5.53 pa·s for solid contents of 40, 45, 50, 55, 60, and 65 wt.%, respectively. The red curve represents the Krieger-Dougherty model curve, which reflects the relationship between slurry viscosity and solid content [39]. As the solid content increases, the slurry viscosity increases significantly, and the trend of viscosity change aligns well with the model curve. The viscosity increase is most pronounced when the solid content increases from 60 wt.% to 65 wt.%, indicating that at higher solid contents, more severe clustering occurs within the slurry. This clustering, along with sedimentation effects, prevents the interlayer flow of the slurry even at high shear rates, resulting in a sharp increase in viscosity. Therefore, adding 5 wt.% PEG with a molecular weight of 200 and a solid content of 60 wt.% is most suitable.
3.4. Effects of Debinding and Sintering
In order to obtain defect-free vitrified bond diamond grinding wheels after sintering, a two-step process was employed. Firstly, debinding was performed in a vacuum environment to ensure the slow pyrolysis of organic materials and avoid the generation of defects. After debinding, oxygen was introduced for sintering in an ambient air environment to react with the residual carbon and eliminate impurities to the maximum extent. Previous studies have reported the pyrolysis temperatures of the photopolymer resins HDDA and TMPTA to be approximately 290 ℃ and 200 ℃ [40, 41], while the pyrolysis temperature of PEG is approximately 300 ℃. Based on the thermogravimetric analysis of the preforms (40 wt.% solids, 5.0 wt.% molecular weight 200 PEG), a slow debinding process was established as shown in Fig. 10. During the debinding stage, the temperature was held at 200, 300, 370, and 550 ℃ for 120 minutes at a rate of 1 ℃/minute and at 370–550 ℃ at a rate of 2 ℃/minute. During the sintering stage, a heating rate of 2 ℃/min was applied, followed by holding temperatures of 600 ℃, 630 ℃, 650 ℃, 680 ℃, and 700 ℃ for 150 min, as depicted in Fig. 10b. After cooling to room temperature, defect-free vitrified bond diamond grinding wheels were obtained.
The density, porosity, microhardness, and flexural strength of vitrified bond samples for different sintering temperatures are shown in Fig. 11. Flexural strength, microhardness, and density were discovered to display a tendency of initially rising and subsequently decreasing with the rise in sintering temperature. At a sintering temperature of 680 ℃, the greatest values were attained, reaching 44.24 MPa, 660 MPa, and 2.36 g/cm3, respectively. At first, the porosity declined, then it grew until it reached a minimum of 35.2% at 680 ℃. The data are further explained by the microstructure of the vitrified bond samples at various sintering temperatures, which is depicted in Fig. 12. At temperatures below the optimal sintering temperature, the samples exhibited lower density and higher porosity, indicating a weaker occurrence of solid-state sintering reactions and weaker interactions among the components of the bond. This resulted in lower strength. When the sintering temperature exceeded the optimum, the liquid phase in the bond increased, leading to increased fluidity at high temperatures. The trapped pores inside the bond could not be expelled, resulting in over-sintering and foaming, leading to an increase in porosity and a decrease in strength. Therefore, a sintering temperature of 680°C was found to be more suitable. Due to the low solid content, the overall performance of the vitrified bond was relatively low. Enhancing the solid content becomes a key factor for the future development of vitrified bond diamond grinding wheels using UV-curing technology.
3.5. Effects of grit ratio
The formulation of the grinding wheel is an important feature of the wheel. The grit-to-bond ratio is the mass or volume ratio between the abrasive and the bond. In the case of vitrified bond diamond grinding wheels, the main abrasive is diamond. To achieve sharp cutting edges for machining brittle materials and meet the requirements of UV-curing, extensive experimental exploration was conducted. It was found that when the diamond content exceeds 40 wt.%, it becomes difficult to shape the wheel due to reduced light transmission caused by the darker color of the diamond particles. The average particle size of diamond was chosen to be 10 µm. Considering grinding efficiency and performance, the concentration of diamond abrasive was determined to be 125%, which translates to a mass ratio of approximately 40%. Additionally, white corundum with an average particle size of 7 µm was selected as the auxiliary abrasive. The auxiliary abrasive serves to improve the grit-to-bond ratio, while the micron scale white corundum also enhances the strength of the grinding tool. To investigate the influence of the grit-to-bond ratio on vitrified bond diamond grinding wheels, the formulated compositions are presented in Table 4.
The wear resistance of grinding wheels directly reflects their durability; however, wear resistance is generally not directly measurable. Therefore, flexural strength is used as an indirect measure of the wear resistance of wheels. Figure 13 illustrates the flexural strength of grinding wheel samples with different grit-to-bond ratios at a sintering temperature of 680°C. The study reveals that with an increase in the content of vitrified bond, the flexural strength of the grinding wheel samples initially increases and then decreases, reaching a maximum value of 27.93 MPa at a vitrified bond content of 30 wt.%. This indicates that when the content of the vitrified bond is too low, the bonding force between the abrasive and the bond is insufficient, resulting in a lower flexural strength. Once the addition of vitrified bond reaches its peak, excessive bond does not effectively infiltrate the abrasive, leading to foaming or even deformation between the bond and the abrasive, further reducing the strength. Solely relying on flexural strength is insufficient to test the grinding performance of grinding wheels; therefore, subsequent grinding experiments are necessary.
Table 4
Formulation of vitrified bond diamond grinding wheels.
|
Diamond
|
white corundum
|
Vitrified bond
|
R1
|
40 wt.%
|
35 wt.%
|
25 wt.%
|
R2
|
40 wt.%
|
32 wt.%
|
28 wt.%
|
R3
|
40 wt.%
|
30 wt.%
|
30 wt.%
|
R4
|
40 wt.%
|
28 wt.%
|
32 wt.%
|
R5
|
40 wt.%
|
25 wt.%
|
35 wt.%
|
Figure 14 shows the X-ray diffraction pattern of the grinding wheel at a sintering temperature of 680°C. The study reveals the presence of a glass phase in the phase composition of the grinding wheel, indicating the amorphous structure of the vitrified bond. The main component of white corundum is Al2O3. In the spectrum, diamond and Al2O3 represent the primary abrasive and auxiliary abrasive of the grinding wheel, respectively. The high-temperature sintering of diamond occurs without any occurrence of graphitization when adequately encapsulated by the vitrified bond.
3.6. Grinding experiment
Figure 15 presents a vitrified bond diamond grinding wheel prepared through UV-curing. Table 5 shows the sintered shrinkage of the grinding wheels. Due to the lower solid content, the grinding wheels exhibit significant overall post-sintering shrinkage, with the maximum shrinkage observed in the thickness direction at 24.5%. This is attributed to the layer-by-layer stacking during the printing process. The grinding wheels with lower strength are prone to accidents when grinding, so four groups of wheels with four formulations, R2, R3, R4, and R5, which have higher flexural strength, were subjected to grinding experiments, compared, and analyzed, and then the optimal formulation was selected.
Table 5
Grinding wheel sintering shrinkage.
|
Outer diameter
|
Inner diameter
|
Thickness
|
pre-sintering
|
100 mm
|
80 mm
|
10 mm
|
post-sintering
|
77.60 mm
|
62.44 mm
|
7.55mm
|
Shrinkage
|
22.40%
|
21.95%
|
24.5%
|
The grinding wheel was fixed on a customized substrate to grind silicon carbide ceramics in a vertical universal friction and wear testing machine. The grinding time was set to 500 s, and the friction and wear behavior of four different formulations of diamond wheels were tested. Table 6 presents the experimental data from the grinding tests. It was found that the material removal rate and grinding rate of R3 wheel were maximum 5.08 mg/s and 9.78 among the four groups of wheels, respectively. This is because, when the vitrified bond content is too low (less than 30 wt.%), the bond has a lower holding power on the abrasive, which causes more severe abrasive shedding when grinding, resulting in the most severe wheel loss and the smallest grinding ratio and material removal rate. On the other hand, when the vitrified bond content was too high (greater than 30 wt.%), the bonding force between the abrasive and the bond increased, making it difficult for the worn abrasives to detach from the wheel surface. As a result, the self-sharpening and sharpness of the wheel deteriorated. Although the wheel wear was reduced, the material removal rate and grinding ratio on the workpiece were also decreased. Therefore, optimal grinding performance was achieved when the diamond content was 40 wt.%, the white corundum content was 30 wt.%, and the vitrified bond content was 30 wt.%, with a grit-to-bond ratio of 7:3 between the abrasive and the bond.
Figure 16 depicts the relationship between the friction coefficient and grinding time for four groups of grinding wheels used to grind silicon carbide ceramics. The investigation revealed distinct behaviors for each group. The R2 grinding wheel exhibited an initially stable friction coefficient around 0.5 during the early stages of grinding. However, after 100 seconds, the friction coefficient increased and became more variable, leading to reduced stability until the end of the grinding process. This can be attributed to the low initial friction, allowing the bond to effectively hold the diamond grits and maintain a stable friction coefficient. However, as grinding time increased, the friction force also increased, but the limited bond failed to firmly retain the diamond grits, resulting in more unworn grits falling off and causing larger fluctuations in the friction coefficient. In contrast, the R3 grinding wheel maintained a stable friction coefficient around 0.5 with minimal fluctuations. This stability can be attributed to its high strength, which allowed the diamond grits to quickly dislodge after becoming dull, ensuring a stable grinding process and excellent grinding efficiency. The R4 grinding wheel displayed a gradual increase in the friction coefficient from 0.3 to 0.6 as grinding time increased, followed by stability. This behavior occurred because, before 200 seconds of grinding, the sharp diamond grits produced more debris adhering to the grinding wheel and workpiece surfaces. Consequently, the grits were unable to dislodge promptly, leading to an increase in the friction force and the friction coefficient. However, after 200 seconds of grinding, the friction force increased further, but the dulled diamond grits could now dislodge promptly, resulting in improved grinding process stability. Regarding the R5 grinding wheel, significant changes in the friction coefficient were observed after 120 seconds of grinding, with the most dramatic fluctuations occurring in the last 50 seconds. The excessive bond content led to slower dislodging of the diamond grits after wear, causing increased friction during workpiece grinding, and consequently resulting in larger fluctuations in the friction coefficient and poorer grinding performance.
Table 6
|
Wheel wear (g)
|
Workpiece removal(g)
|
Removal rate(mg/s)
|
Grinding ratio
|
R2
|
0.28
|
1.25
|
2.50
|
4.46
|
R3
|
0.26
|
2.54
|
5.08
|
9.78
|
R4
|
0.24
|
2.12
|
4.24
|
8.83
|
R5
|
0.22
|
1.45
|
2.90
|
6.59
|
The surface morphology of the silicon carbide workpieces after grinding with the four sets of grinding wheels was examined using a white light interferometer, as shown in Fig. 17. In the image, the red color represents surface protrusions, the blue color represents surface depressions, and the green color represents relatively flat areas. The study found that the average surface roughness Sa of the workpieces after grinding with the four sets of grinding wheels was 2.144 µm, 1.767 µm, 2.047 µm, and 2.284 µm, respectively. The surface roughness depends mainly determined by the abrasive grain size. Since the diamond particle size used in the four sets of grinding wheels was the same, the difference in surface roughness was not significant. However, a detailed analysis revealed that the R3 grinding wheel exhibited the lowest surface roughness and the best surface flatness. It had fewer depressions and protrusions. This can be attributed to the higher flexural strength and sharper cutting edges of the wheel, resulting in better self-sharpening. When grinding hard and brittle materials, a sharper wheel tends to produce lower surface roughness on the workpiece.