Abstract
This paper reports on the process-fatigue relation of lithium disilicate glass ceramic (LDGC) using low-cycle, high-load Hertzian indentations with a rigid indenter to simulate teeth grinding/clenching of LDGC restorations with different surface asperities obtained in CAD/CAM milling, sintering, polishing and glazing. The maximum contact stresses were evaluated as functions of the number of load cycles and surface treatments using the Hertzian model. Indentation-induced surface damage was viewed using scanning electron microscopy (SEM) to understand the relationships among microstructures, surface asperities, crack morphology and propagation. Different processes and surface treatments significantly affected the maximum contact stresses of indented LDGC surfaces (ANOVA, p < 0.05), which were all significantly reduced with the number of cycles (ANOVA, p < 0.05). Quasi-plastic deformation was dominant in single-cycle indentation of all processed and treated surfaces. In higher cycle indentations, inner cone cracks were formed on all surfaces; median and transverse cracks were formed on the roughest surfaces processed by CAD/CAM milling and sintering. Ring cracks, fretting, pulverization, micro-bridges, surface smearing and wedging, and edge chippings were also propagated on all surfaces. The process-fatigue relation provides an understanding of the mechanical functions of surface asperities produced in different processes and treatments. It indicates that the rougher the surfaces obtained in CAD/CAM milling and sintering were, the more severe the induced mechanical damage was. Finally, the smoothest surfaces produced by CAD/CAM milling, polishing and sintering sustained the highest contact stresses and the least fatigue damage at higher cycles, ensuring their superior fatigue performance compared to other processed LDGC surfaces.
Keywords: CAD/CAM milling, Fatigue damage, Hertzian indentation, Lithia-based glass ceramics, Polishing, Glazing
1. Introduction
High-strength lithium disilicate glass ceramic (LDGC) is an excellent aesthetic ceramic for monolithic crowns and bridges, inlays and onlays and veneers (Beuer et al., 2012; Guess et al., 2010b; Miyazaki et al., 2013; Silva et al., 2010; 2011; Zhang and Kelly, 2017). It is suitable for not only anterior restorations but also posterior regions where concentrating stresses are greatest (Höland et al., 2000; Ma et al., 2013; Rekow and Thompson, 2005; 2007; Studart et al., 2007; Zhang et al., 2013a). LDGC restorations are generally manufactured by CAD/CAM-milling of the intermediate lithium metasilicate glass-ceramic (LMGC) and subsequent sintering of the milled workpieces for crystallization into lithium disilicate crystals (Alao et al., 2017; Denry and Holloway, 2010; Li et al., 2014; Ortiz et al., 2019; Ritzberger et al., 2010). The CAD/CAM-milling process with diamond tools inevitably initiates surface cracks in LMGC which cannot be completely healed in sintering, thereby compromising the strengths of LDGC (Alao et al., 2017; Rekow and Thompson, 2005; Rekow et al., 2011). Other post fabrication processes are also required, including polishing and glazing of exterior (occlusal, buccal and lingual) surfaces for chewing functions and aesthetics. Polishing and glazing potentially improve the surface quality but may not always diminish diamond milling-induced surface cracks (Alao et al., 2017; Rekow and Thompson, 2005; Zhang et al., 2004).
Process-induced surface defects in ceramic restorations potentially causes fatigue failures of the restorations during mastication, teeth grinding/clenching involving cyclic movements of jaws at 100–800 N loads at 1500 cycles/day (Kosmač et al., 2008; Peterson et al., 1998; Rekow and Thompson, 2007). Therefore, the understanding of the fatigue behavior of dental ceramics influenced by processing and surface treatments is critical to their clinical long-term performance. Fatigue failures involve crack initiation, nucleation, coalescence and propagation. Severe fatigue damage occurs in posterior regions in which high stress concentrations result in crack propagations in tensile fields (Zhang et al., 2013b). Traditional load-to-failure methods are amenable to long-crack tests to obtain rising (R-curve) toughness values for ceramics (Kelly et al., 2010). However, these tests do not report correct stress states; reported crack systems are not compatible with bulk clinical failures; and documented failure loads far exceed clinical acceptable ranges for mastication, swallowing and bruxism (Kelly et al., 2010). Further, long-crack tests lack the capacity of extrapolation down to the microstructural scale (Padture and Lawn, 1995; Pajares et al., 1995; Peterson et al., 1998). Consequently, they are not favourably sensitive to short-crack properties like strength and wear (Padture and Lawn, 1995).
Hertzian contact tests using a rigid spherical indenter can characterize the fatigue damage of dental ceramics because they provide a simple means to quantify failure mechanisms within the critical short-crack domain (Pajares et al., 1995; Lawn, 1998; Lawn et al., 2001). In this domain, the strength of dental ceramics is most vulnerable (Guiberteau et al., 1993; Zhang et al., 2004b). This test consists of indenting simplified flat geometries with a contacting sphere (Lawn, 1998), which is a hard steel or tungsten carbide to represent the biting of a hard object (Lawn, 1998; Lorenzoni et al., 2020; Kruzic et al., 2018). Thus, Hertzian contact testing can mimic relevant clinical conditions, including contact loads simulating occlusal loads and indenter radius mimicking cuspal curvature (Peterson et al., 1998). The loads can be static, dynamic or cyclic. Static fatigue involves applications of constant stresses or strains; dynamic fatigue involves the conditioning of constant stressing or straining rates; and cyclic fatigue represents the implementation of oscillatory stresses or strains. The failure modes and fatigue mechanisms associated with each loading system are different, among which the cyclic loads induce the most severe damage on dental ceramics due to the mechanical degradation mechanism (Zhang et al., 2013b).
Several failure modes in Hertzian contacts have been reported to identify clinical failures of ceramic crowns and bridges (Lawn et al. 2001). The near-surface damage in a thick ceramic structure induced by cyclic blunt contacts was used to mimic occlusal contact cracks of outer, inner and partial cones as well as quasi-plastic deformation (Lawn et al. 2001; Peterson et al., 1998; Zhang et al., 2013b). The outer cone cracks were initiated from the top surface outside the contact circle where the Hertzian tensile stresses were maximum and grew steadily with time under static loads or cyclic load via the slow crack growth mechanism, leading to tooth chipping (Deng et al., 2002a; Zhang et al., 2013b). The inner cones appeared after a prolonged cyclic load in a wet environment and were driven by hydraulic pumping of fluid into surface micro-cracks (Zhang et al., 2013b). In contrast, partial cones were developed in sliding contacts by the tangential load component, which skewed tensile stress fields at trailing contact edges (Zhang et al., 2013b). The quasi-plastic deformation was generally initiated below the contact when the maximum shear stress in the Hertzian near field exceeded a half of the yield stress (Deng et al., 2002a). It led to damage accumulation, micro-crack coalescence and accelerated wear. In cyclic loadings, median cracks evolved beneath accumulated plastic damage zones (Zhang et al., 2013b).
Further, contact fatigue tests on bilayer structures, in which a ceramic was bonded unto a compliant substrate, analogously simulated the fatigue behavior of a dental ceramic crown bonded onto dentine (Deng et al. 2002a; Lawn et al., 2002a; 2004; Zhang and Lawn, 2004; 2005; Zhang et al., 2004a; 2004b; 2013b). On the cementation surface beneath the contact, radial cracks popped in spontaneously from starting flaws when maximum tensile stresses equaled the flexural strength (Deng et al. 2002a; Lawn et al., 2002a; 2004; Zhang and Lawn, 2004; 2005; Zhang et al., 2004a; 2004b; 2013b). They were highly deleterious in strength losses and barely undetectable in opaque ceramics, leading to catastrophic bulk failures (Lawn et al., 2002b; 2004).
The developed failure modes for dental ceramics were influenced by loads, load types, environment, indenter radius, and material types (Zhang et al., 2005). For monolayer structures, most failure modes were developed based on Hertzian indentations of polished surfaces (Zhang et al., 2005). Only few tests were performed on abraded surfaces with 600-mesh size silicon carbide abrasives to introduce controlled flaws for cone crack initiations (Bhowmick et al., 2005; Zhang et al., 2005). For bilayer systems, only some cementation surfaces have been ground, polished or sandblasted to introduce controlled flaws for radial cracking initiations for simulation of clinically relevant surface treatments (Guess et al., 2010a; Zhang and Lawn, 2005; Zhang et al., 2004a). In addition, most processed and treated surfaces were not quantitatively assessed with respect to roughness. Further, little work has been reported to prevail fatigue mechanisms for LDGC surfaces produced by CAD/CAM milling of LMGC and subsequent sintering, polishing, and glazing. These clinical processing and treatments introduce surface asperities and flaws to LDGC restorations. Therefore, the understanding the fatigue behavior of CAD/CAM milled and sintered, and treated LDGC surfaces is critical to their clinical successes (Rekow et al., 2011).
This paper aims to investigate the fatigue behavior of processed and treated LDGC surfaces in low-cycle, high-load cyclic Hertzian indentations to mimic teeth clenching or grinding. Using LDGC surfaces produced in diamond milling, sintering, polishing and glazing to simulate various clinical surfaces (Alao et al., 2017), the Hertzian maximum contact stresses were evaluated as functions of the number of load cycles and surface treatments. Fatigue damages in different processed and treated LDGC surfaces after cyclic indentations were examined using scanning electron microscopy (SEM) to understand the relationships among microstructures, surface asperities and crack propagation. Finally, the fatigue mechanism of LDGC conditioned to different asperities was proposed.
2. Experimental procedures
2.1. Materials
LMGC blocks of 14.5 mm × 12.4 mm × 18 mm (IPS e.max CAD, Ivoclar Vivadent, Liechtenstein) were selected for this investigation. The material contained approximately 40 vol. % needle-like lithium metasilicate crystals with thickness and lengths of 0.2–0.5 μm 0.5–1.5 μm, respectively (El-Meliegy and van Noort, 2012; Ortiz et al., 2019). It had a low strength of 130 ± 30 MPa, designed as a chairside CAD/CAM material (Bühler-Zemp and Völkel, 2005). The LMGC blocks were subject to CAD/CAM milling, sintering and surface treatments to simulate various surface conditions in dental clinical practice, detailed by Alao et al., 2017. Briefly, these processes and treatments were designated as CAD/CAMed-sintered, CAD/CAMed-polished-sintered, CAD/CAMed-sintered-polished, CAD/CAMed-sintered-glazed and CAD/CAMed-polished-sintered-glazed (Alao et al., 2017). These processed and treated surfaces had different asperities providing different residual stress distributions. Table 1 shows the arithmetic mean surface roughness (Ra) and the maximum surface roughness (Rz) for different treated LDGC surfaces (Alao et al., 2017). It shows that the CAD/CAMed-polished-sintered surfaces achieved the smoothest quality with least roughness values with Ra = 0.12 ± 0.08 μm and Rz = 0.89 ± 0.26 μm and the CAD/CAMed-polished-sintered-glazed surfaces were the second best with Ra = 0.25 ± 0.01 μm and Rz = 1.55 ± 0.08 μm (Alao et al., 2017). CAD/CAMed-sintered-polished and CAD/CAMed-sintered surfaces had intermediate roughness with Ra = 0.42 ± 0.12 μm and Rz = 2.54 ± 0.88 μm, and with Ra = 0.58 ± 0.07 μm and Rz = 3.50 ± 0.08 μm, respectively (Alao et al., 2017). CAD/CAMed-sintered-glazed surfaces were the roughest with Ra = 0.70 ± 0.10 μm and Rz = 3.75 ± 0.09 μm (Alao et al., 2017). Under each processing and treatment condition, three repeated specimens were prepared for Hertzian fatigue testing.
Table 1.
Fabrication process | Surface roughness parameters | |
---|---|---|
Ra (μm) | Rz (μm) | |
CAD/CAMed-sintered-glazed | 0.70 ± 0.10 | 3.75 ± 0.09 |
CAD/CAMed-sintered | 0.58 ± 0.07 | 3.50 ± 0.08 |
CAD/CAMed-sintered-polished | 0.42 ± 0.12 | 2.54 ± 0.88 |
CAD/CAMed-polished-sintered-glazed | 0.25 ± 0.01 | 1.55 ± 0.08 |
CAD/CAMed-polished-sintered | 0.12 ± 0.08 | 0.89 ± 0.26 |
The mechanical properties of LDGC include the Vickers hardness, Hv of 5.8 ± 0.1 GPa, the Young’s modulus, E of 95 ± 5 GPa, the fracture toughness, and KIC of 2.25 ± 0.25 MPa m1/2 (Bühler-Zemp and Völkel, 2005), the biaxial strength of 480 MPa ( Zhang and Kelly, 2017), and the Poisson ratio ν of 0.23 (Albakry et al., 2003).
2.2. Hertzian indentation fatigue tests
Hertzian fatigue testing was conducted using a tungsten carbide spherical indenter against each processed and treated LDGC surface. Placed inside an aluminum sleeve with a steel base, the indenter had a curvature radius of 2.5 mm in the range of the cuspal radii (2–4 mm) for posterior teeth (Krejci et al., 1999; Peterson et al., 1998). Each specimen surface and the indenter were coupled together and mounted on a universal testing machine (United STM–50KN Testing System). Cycling between 20 N and 800 N was conducted on each specimen surface at a frequency of 0.1 Hz and 3000 cycles, which conformed to mastication and bruxism conditions (Yin et al., 2013).
2.3. Surface characterization
Optical microscopy was used to view cyclic indentation-induced residual imprints. The maximum contact stress, σmax, of each indented surface was calculated using the Hertzian equation for an elastic contact of a rigid sphere on a flat surface (Budinski and Budinski, 2010) as:
(1) |
where Pmax is the maximum load and a is the contact radius. The contact radii of the indented imprints were measured from the optical images (Guiberteau et al., 1993). Three repeated measurements were conducted on three specimens for each surface condition to obtain their mean maximum stress and their standard deviation. A two-factor analysis of variance (ANOVA) with replication was applied at 5% confidence interval to examine significant effects of surface treatments and the number of cycles on maximum contact stresses of indented LDGC surfaces.
After 1, 10, 100, 1000 and 3000 indentations, each treated LDGC specimen surface was gold-coated and examined using scanning electron microscopy (SEM) (Jeol JSM5410LV, Tokyo, Japan) to study damage evolution and morphology.
3. Results
3.1. Maximum contact stresses
Fig. 1 shows the maximum contact stresses for different treated LDGC surfaces versus the number of cycles. All stresses decreased with the increase in the number of cycles, indicating that the contact areas in all treated surfaces became larger when increasing the number of cycles. At 1 cycle, the roughest CAD/CAMed-sintered-glazed surfaces had the lowest maximum contact stress of 3.02 GPa. In comparison, smoother CAD/CAMed-polished-sintered, CAD/CAMed-sintered-polished, and CAD/CAMed-polished-sintered-glazed surfaces sustained significantly higher contact stresses of larger than 4.42 GPa. At 10 cycles, rougher CAD/CAMed-sintered-glazed and CAD/CAMed-sintered surfaces also had lower maximum stresses of 1.86–2.20 GPa while smoother CAD/CAMed-polished-sintered, CAD/CAMed-sintered-polished, and CAD/CAMed-polished-sintered-glazed surfaces had higher stresses of 2.29–3.53 GPa. At 100 cycles, the smoothest CAD/CAMed-polished-sintered surfaces had the highest maximum contact stresses with mean 2.27 GPa in contrast to low stresses of 1.70–1.92 GPa in rough CAD/CAMed-sintered-glazed and CAD/CAMed-sintered surfaces. At 1000 cycles, the smoothest CAD/CAMed-polished-sintered surfaces had the highest maximum contact stresses of mean 2.23 GPa, significantly higher than 1.73–1.64 GPa in rough CAD/CAMed-sintered-glazed and CAD/CAMed-sintered surfaces. Further, 3000 cyclic indentations yielded 2.12 GPa maximum contact stresses in the smoothest CAD/CAMed-polished-sintered surfaces, at 40% higher than those in rough CAD/CAMed-sintered-glazed and CAD/CAMed-sintered surfaces. As indicated in Table 2 on ANOVA results, surface treatments and the number of cycles significantly affected maximum contact stresses in LDGC surfaces with different roughness values (p < 0.05).
Table 2.
Source of Variation | SS | Df | MS | F | P-value | F crit |
---|---|---|---|---|---|---|
Surface treatments | 57.96 | 4 | 14.49 | 76.39 | 0.00 | 2.56 |
Number of cycles | 10.55 | 4 | 2.64 | 13.90 | 0.00 | 2.57 |
Interaction | 7.47 | 16 | 0.47 | 2.46 | 0.01 | 1.85 |
Within | 9.49 | 50 | 0.19 | |||
Total | 85.46 | 74 |
3.2. Indention fatigue morphologies
Figs. 2-6 reveal SEM Indentation fatigue morphologies in different treated LDGC surfaces at 1, 10, 100, 1000, and 3000 cycles, respectively. Fig. 2 shows 1-cyclic indentation morphologies for differently treated LDGC surfaces. Fig. 2(a) reveals shallow Hertzian ring cracks in the roughest and pulverized CAD/CAMed-sintered-glazed surface. Fig. 2(b) details meandering microcracks with widths of approximately 1 μm and micro-pulverized debris in the surface. Fig. 2(c) reveals shallow Hertzian ring cracks in the rough CAD/CAMed-sintered surface. Fig. 2(d) details fretting and pulverized debris in the wedged area with crack widths up to 11 μm in the surface. Fig. 2(e) reveals shallow Hertzian ring cracks in the CAD/CAMed-sintered-polished surface. Fig. 2(f) details micro-bridges with widths up to 3 μm and enlarged smeared fracture in the ring cracks of the surface. Fig. 2(g) reveals shallow Hertzian ring cracks and indentation-induced smeared areas inside the ring in the smooth CAD/CAMed-polished-sintered-glazed surface. Fig. 2(h) details meandering microcracks with widths of approximately 0.8 μm and peeled and smeared debris in the surface. Fig. 2(i) reveals shallow Hertzian ring cracks in the smoothest CAD/CAMed-polished-sintered surface. Fig. 2(j) details meandering microcracks with widths of approximately 0.6 μm in the surface.
Fig. 3 shows 10-cyclic indentation morphologies for differently treated surfaces. Fig. 3(a) reveals overall Hertzian cracks and wedged areas in the roughest and pulverized CAD/CAMed-sintered-glazed surface. Fig. 3(b) details large wedged areas of the surface, comprising fretting and pulverizations, and wedged cracks with widths up to 50 μm. Fretting and pulverized debris scaled up to 1 μm in the wedged area. Fig. 3(c) reveals overall Hertzian ring cracks in the rough CAD/CAMed-sintered surface. Fig. 3(d) details meandering crack paths with widths of approximately 1 μm and micro-pulverized debris in the wedged area of the surface. Fig. 3(e) reveals overall Hertzian cracks with internal smears in the CAD/CAMed-sintered-polished surface. Fig. 3(f) details concentric microcrack paths with maximum crack widths of approximately 2.6 μm and pulverized debris in the surface. Fig. 3(g) reveals overall Hertzian ring cracks in the smoother CAD/CAMed-polished-sintered-glazed surface. Fig. 3(h) details meandering microcrack paths with crack widths of approximately 1 μm and peeled smeared debris in the surface. Fig. 3(i) reveals overall Hertzian cracks of the smoothest CAD/CAMed-polished-sintered surface. Fig. 3(d) details meandering micro-paths with widths of approximately 1 μm, debris and smears in the surface.
Fig. 4 shows 100-cyclic indentation morphologies for differently treated surfaces. Fig. 4(a) reveals overall Hertzian ring cracks and wedged area on the roughest and pulverized CAD/CAMed-sintered-glazed surface. Fig. 4(b) details continuous and discontinuous ring cracks, fragmented areas, fretting debris and median cracks extending outside ring cracks in the surface. The cumulative effect of all these cracks resulted in the formation of a large-scale fractured surface. Fig. 4(c) reveals overall Hertzian ring cracks, inner edge chippings and wedged areas in the rough CAD/CAMed-sintered surface. Fig. 4(d) details continued and discontinued ring cracks and chipped edge resulting into large scale surface fractures with a maximum width of approximately 200 μm in the surface. Fig. 4(e) reveals overall ring cracks, fragmentations, smearing and debris in the CAD/CAMed-sintered-polished surface. Fig. 4(f) details propagations of ring cracks, smeared areas and fretting debris in wedged areas of approximately 21 μm in width in the surface. Fig. 4(g) reveals overall Hertzian ring cracks in the smooth CAD/CAMed-polished-sintered-glazed surface. Fig. 4(h) details concentric ring cracks and fretting debris in wedged areas in the surface. Fig. 4(i) reveals overall ring cracks, internal smears and wedged areas in the smoothest CAD/CAMed-polished-sintered surface. Fig. 4(j) details ring crack propagations, fretting debris inside wedged areas with a maximum width of approximately 21 μm in the surface.
Fig. 5 shows 1000-cyclic indentation morphologies for differently treated surfaces. Fig. 5(a) reveals overall Hertzian ring cracks with inner and outer edge chippings in the roughest and pulverized CAD/CAMed-sintered-glazed surface. Fig. 5(b) details wedged areas and sub-surface penetrations of transverse cracks in the surface. Fig. 5(c) reveals overall Hertzian ring cracks, large-scale edge chipping, and median cracks in the rough CAD/CAMed-sintered surface. Fig. 5(d) details large-scale chipped edges of 300 μm wide, and micro-craters in the surface. Fig. 5(e) reveals overall Hertzian ring cracks, fragmentations, wedged areas in the CAD/CAMed-sintered-polished surface. Fig. 5(f) details large-scale fractures and fretting debris in the wedged area of the surface. Fig. 5(g) reveals overall Hertzian continued and discontinued ring cracks, fragmentations and debris in the smooth CAD/CAMed-polished-sintered-glazed surface. Fig. 5(h) shows large-scale chipped edges, and micro-craters resulting from wedged areas in the surface. Fig. 5(i) reveals overall Hertzian ring cracks, and inner edge chipping in the smoothest CAD/CAMed-polished-sintered surface. Fig. 5(h) details large-scale chipped edges and micro-craters in the surface.
Fig. 6 shows 3000-cyclic indentation morphologies for differently treated surfaces. Fig. 6(a) reveals overall Hertzian ring cracks with inner and outer edge chippings in the roughest and pulverized CAD/CAMed-sintered-glazed LDGC. Fig. 6(b) details large-scale inner chipped edges of approximately 250 μm in width penetrating into the material, leading to large-scale micro-craters and material removal in the surface. Fig. 6(c) reveals overall Hertzian ring cracks with large-scale outer edge chippings in the CAD/CAMed-sintered surface. Fig. 6(d) details a large-scale chipped edge of approximately 400 μm wide, material removal and micro-craters in the surface. Fig. 6(e) reveals overall Hertzian ring cracks with large-scale inner and outer edge chippings and wedged areas in the CAD/CAMed-sintered-polished surface. Fig. 6(f) details large-scale chipped edges, material removal, micro-craters from wedged areas of approximately 150 μm wide in the surface. Fig. 6(g) reveals overall Hertzian ring cracks with large-scale inner edge chipping and wedged areas in the smooth CAD/CAMed-polished-sintered-glazed surface. Fig. 6(h) details large-scale fractures with transverse cracks in edge chippings and wedged areas in the surface. Fig. 6(i) reveals overall Hertzian ring cracks with large-scale outer edge chipping on the smoothest CAD/CAMed-polished-sintered surface. Fig. 6(j) shows large-scale fretting, fractures and pulverization in edge-chipped areas of the surface.
4. Discussion
This study presents results of low-cycle, high-load fatigue behavior of high-strength LDGC material with different surface roughness values using the Hertzian indentations. Low-cycle, high-load fatigue tests enable the prediction of the fatigue property within the short-crack domain in which the strength of dental ceramics is most vulnerable (Guiberteau et al., 1993; Yin et al., 2013). The attempt was also made to understand indentation damage evolutions in LDGC surfaces from the perspective of clinically relevant surface treatments. The analysis of indentation damage morphologies in differently treated LDGC surfaces assisted the uncovering of interrelationships among microstructures, surface asperities and propagation of cracks.
The maximum contact stresses of different treated LDGC surfaces in Fig. 1 were extracted based on the assumption of the elastic contact property of the indenter. To determine whether this elasticity condition in the maximum contact stress measurements was accommodated, the following condition should be valid (Guiberteau et al., 1993):
(2) |
where H is the Vickers hardness of the tungsten carbide indenter, which is 20 GPa (Guiberteau et al., 1993). Consequently, the measured maximum contact stresses should be less than 8 GPa to avoid the permanent deformation of the indenter. Fig. 1 reveals that all maximum contact stresses of indented LDGC surfaces at all cycles were less than 8 GPa, indicating the validity of the indenter’s elasticity condition. Further, tungsten carbide indenters had similar veneered crown fatigue reliability and failure modes to more compliant steatite indenters, reconfirming the minimum influence of the carbide indenters’ elastic moduli on the fatigue behaviour of dental ceramics (Lorenzoni et al., 2020).
Although all maximum contact stresses of treated LDGC surfaces reduced by at least 50% after 3000 cyclic indentations, CAD/CAMed-polished-sintered surfaces with the lowest roughness (Alao et al., 2017) sustained the highest maximum contact stress at higher cycles of 100–3000. This indicates that a better surface treatment with low roughness results in the better fatigue property at high cycles.
Hertzian ring cracks produced on all treated LDGC surfaces at 1 indentation in Fig. 2 indicate the precursor incipient cone cracking or quasi-plasticity. To determine the responsible mechanisms for ring cracks, a critical load for cone cracks and quasi-plastic deformation is required since the simultaneous occurrence of both modes is impossible (Peterson et al., 1998). Thus, the critical load, PC, for cone cracks can be predicted by the following relation (Lawn et al., 2001):
(3) |
where KIC is the fracture toughness and E is the Young’s modulus of the indented material, r is the radius of the indenter and A is a dimensionless constant expressed as (Lawn 1998):
(4) |
The value of A has been established as 8600 (Lawn et al., 2001). By using the properties of LDGC, the PC for cone crack of LDGC was 1146 N, which was higher than the maximum applied maximum indentation load of 800 N. This means that well-developed cone cracks were not able to form underneath the ring cracks.
Further, the occurrence of quasi-plastic deformation can be predicted at a critical yielding load, PY, expressed as (Rhee et al., 2001):
(5) |
where Hv is the Vickers hardness of the indented material, E and r are defined previously, and D is a dimensionless constant expressed by (Rhee et al., 2001):
(6) |
The value of D was quantitatively established as 0.85 (Rhee et al., 2001). By applying the properties of LDGC, the PY for the quasi-plastic deformation was 115 N, which was less than the applied maximum load of 800 N. This indicates that the quasi-plastic deformation have occurred below Hertzian stress fields. Therefore, the fatigue mechanisms of all treated LDGC surfaces at 1 indentation were mechanical in origin caused by attrition of frictional tractions at micro-cracks interfaces similar to that observed in mica glass-ceramic (Cai et al., 1994; Peterson et al., 1998). This is consistent with the quasi-plasticity formation before cone cracks in a single-cycle indentation of LDGC (Coldea et al. (2014). Thus, the downward propagation of ring cracks into cone cracks could be suppressed by deflections along weak frictional tractions away from tensile stress trajectories into compressive stress states (Padture and Lawn, 1995; Pajares et al., 1995).
The critical load for cone cracks, PC, in Eq. (3), can be greatly reduced during cyclic fatigue loading by a time-cumulative slow crack growth mechanism (Lawn et al., 2007; Zhang et al., 2005; 2009). Similarly, the quasi-plastic deformation critical load, PY, in Eq. (5) can be immensely reduced in cyclic loadings due to the action of mechanical driving forces which arise from residual stresses within a precursor quasi-plasticity zone forming median cracks (Zhang et al., 2005). Thus, the induced median cracks accelerate material removal by the rapid coalescence of adjacent faults, leading eventually to the strength loss (Deng et al., 2002a; Jung et al., 2000; Lawn et al., 2001; 2002b; 2004; 2007). This might explain why all maximum contact stresses of treated LDGC surfaces in Fig. 1 significantly reduced with the number of cycles.
At higher cycles of 10–3000, surface damage occurred predominantly in annular regions delineated by inner and outer contact circles at the minimum and maximum contact loads as shown in Figs. 2-6. These segmented concentric ring cracks have been attributed to the formation of multiple inner cone cracks developed by cyclic loads, skewing frictionally assisted tensile stress fields at expanding contact edges (Guiberteau et al., 1993; Kim et al., 2007; Rekow et al., 2011; Yin et al., 2013; Zhang et al., 2013b). This indicates that cone cracks were in fact formed at higher cycles. In addition, median cracks were induced at 100 indentations in the roughest CAD/CAMed-sintered-glazed surfaces (Fig. 4(b)). Further, it is worth noting that the roughest surfaces were induced with transverse cracks at 1000 indentations as revealed in Fig. 5(b).
Surface asperities played a critical role in the fatigue damage of treated LDGC surfaces. Figs. 2-6 have shown that smooth surfaces produced less damage than rough surfaces. At 1 indentation, fatigue damages in rough CAD/CAMed-sintered-glazed (Figs. 2(a) and 2(b)), CAD/CAMed-sintered (Figs. 2(c) and 2(d)) and CAD/CAMed-sintered-polished (Figs. 2(e) and 2(f)) surfaces included ring cracks, fretting, pulverization and micro-bridges. In comparison, damages in smooth CAD/CAMed-polished-sintered-glazed (Figs. 2(g) and 2(h)) and smoothest CAD/CAMed-polished-sintered (Figs. 2(i) and 2(j)) comprised ring cracks of lower widths, with the smoothest surfaces producing lowest crack widths (Fig. 2(j)).
At 10 indentations in rough CAD/CAMed-sintered-glazed (Figs. 3(a) and 3(b)), CAD/CAMed-sintered (Figs. 3(c) and 3(d)), and CAD/CAMed-sintered-polished (Figs. 3(e) and 3(f)) surfaces, more fatigue damages were formed with the formation micro-pulverization, fretting debris and large-scale wedging. In contrast, fatigue damages were restricted to concentric ring cracks and pulverization, surface smearing and peeling in smooth CAD/CAMed-polished-sintered-glazed (Figs. 3(g) and 3(h)) and smoothest CAD/CAMed-polished-sintered (Figs. 3(i) and 3(j)) surfaces, with lowest crack widths in the smoothest surface (Figs. 3(i) and 3(j)).
At 100 indentations, edge chipping, large-scale fractures, micro-craters, fragmentations and wedging were formed in rough CAD/CAMed-sintered-glazed (Figs. 4(a) and 4(b)), CAD/CAMed-sintered (Figs. 4(c) and 4(d)), and CAD/CAMed-sintered-polished (Figs. 4(e) and 4(f)) surfaces. In comparison, damages with wedging and pulverization were observed in smooth CAD/CAMed-polished-sintered-glazed (Figs. 4(g) and 4(h)) and smoothest CAD/CAMed-polished-sintered (Figs. 4(i) and 4(j)) surfaces.
At 1000 indentations, fatigue damages metamorphosed into full-blown chippings in all treated LDGC surfaces, with the highest surface roughness yielding the most severe damage. Large-scale fracture and transverse cracks penetrated into rough CAD/CAMed-sintered-glazed (Figs. 5(a) and 5(b)) and CAD/CAMed-sintered (Figs. 5(c) and 5(d)) surfaces. In contrast, wedging occurred in relatively smooth CAD/CAMed-sintered (Figs. 5(e) and 5(f)) and CAD/CAMed-polished-sintered-glazed (Figs. 5(g) and 5(h)) surfaces. Localized fracture was observed on the smoothest CAD/CAMed-polished-sintered surfaces.
At 3000 indentations, rough CAD/CAMed-sintered-glazed (Figs. 6(a) and 6(b)) and CAD/CAMed-sintered (Figs. 6(c) and 6(d)) surfaces were severely damaged with large-scale edge chippings, fractures and micro-craters penetrated to their subsurfaces. Relatively smooth CAD/CAMed-sintered-polished (Figs. 6(e) and 6(f)), CAD/CAMed-polished-sintered-glazed (Figs. 6(g) and 6(h)), and CAD/CAMed-polished-sintered (Figs. (i) and (j)) surfaces had less edge chipping damage and fractures. Although transverse cracks were propagated through the fractured surfaces, less cracking penetrations occurred into their subsurfaces. This indicates that a better fatigue resistance in smooth surfaces than in rough surfaces.
The underlying fatigue mechanism in low-cycle, high-load Hertzian cyclic indentations of treated LDGC surfaces is a mechanically assisted growth of their asperities. First, cyclic fatigue enhances cone cracking by slow crack extensions, leading to edge chipping (Peterson et al., 1998; Zhang et al., 2013a; 2013b). Edge chipping damages at 3000 cycles were found in all treated surfaces in Fig. 6. Next, quasi-plastic deformation was also enhanced in cyclic loadings by rapid coalescences of microcracks into median cracks as an evident in the roughest CAD/CAMed-sintered-glazed surfaces in Fig. 4b, resulting in mechanical fatigue and accelerated wear (Figs. 3-6). Therefore, the proposed fatigue mechanism is schematically illustrated in Fig. 7. A single cycle indentation of a treated LDGC surface leaves a deformed residual surface area in the form of ring cracks with a contact diameter, 2a1. Subsequent indentations enlarge the deformed surface area with a larger contact diameter, 2a2, with the formation of more concentric ring cracks, leading to stress concentrations along the edges. Further indentations across these stress concentrating sites produce weak cleavage planes, which in turn lead to mechanical damage depending on the surface roughness. The rougher the surface is, the heavier will be the induced mechanical damage.
Within the limitation of this study, the smoothest CAD/CADed-polished-sintered surfaces were least damaged in low-cycle, high-load spherical cyclic indentations in comparison to other treated surfaces. This reinforced our proposed fabrication selection, i.e., CAD/CAM milling-polishing-sintering as the optimized process, for dental restorations to obtain optimized occlusal functions and bacterial plaque retention (Alao et al., 2017).
5. Conclusions
The paper investigated the low-cycle, high-load Hertzian cyclic indentation behavior of LDGC surfaces with clinically relevant surface preparations in a short-crack domain to which the material strength is most vulnerable. It addressed the interrelationships among microstructures, surface asperities and crack propagation of clinically relevant LDGC surfaces. The increased number of indentations remarkably weakened all surfaces with reduced maximum contact stresses. Surface processes and treatments with different roughness also significantly affected the maximum contact stresses. The smoothest CAD/CAMed-polished-sintered surfaces sustained the highest maximum contact stresses at high cycles. Single indentations dominantly induced quasi-plastic deformation and higher cycle indentations induced inner cone cracks detailed with concentric ring cracks, fretting, pulverization, micro-bridges, surface smearing and wedging, and edge chippings in all surfaces. Fatigue damage severities increased with the number of indentations and roughness of treated LDGC surfaces. Transverse median cracks were formed in the roughest CAD/CAM-sintered-glazed surfaces and less cracks were found on smoother surfaces. Finally, a fatigue mechanism has been established based on the mechanically assisted growth of surface flaws for all treated LDGC surfaces in cyclic indentations. The study results highlight the importance of process and treatment selection and support that the smoothest CAD/CAMed-polished-sintered surfaces achieved the superior fatigue resistance. This augmented and reinforced our proposed optimized manufacturing process route for LDGC restorations (Alao et al., 2017).
Acknowledgements
The authors thank Drs. Shane Askew and Kevin Blake of the Advanced Analytical Center and Mr. Kevin Chong of College of Medicine & Dentistry at James Cook University for experimental assistance; Mr. Jim Ruddy of Ivoclar Vivadent, Australia for providing e.max CAD blocks; and Mr. Matthew Batty of Sirona Australia for offering CAD/CAM technical support. The work was supported by the PhD IPRS Scholarship and the Collaboration Grants Scheme from James Cook University; the Australia-China Science and Research Fund from the Department of Industry, Innovation, Climate Change, Science, Research and Tertiary Education, Australia (Grant No. ACSRF GMB 12029); and the United States National Institutes of Health, National Institute of Dental and Craniofacial Research Grants (R01DE026279 and R01DE026772).
References
- Alao A-R, 2019. Elasticity, plasticity and analytical machinability prediction of lithium metasilicate/disilicate glass ceramics. J. Mech. Behav. Biomed. Mater 96, 9–19. [DOI] [PubMed] [Google Scholar]
- Alao A-R, Stoll R, Song X-F, Abott JR, Zhang Y, Yin L, 2017. Fracture, roughness and phase transformation in CAD/CAM milling and subsequent surface treatments of lithium metasilicate/disilicate glass-ceramics. J. Mech. Behav. Biomed. Mater 74, 251–260. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Alao A-R, Yin L, 2015. Nano-mechanical behavior of lithium metasilicate glass-ceramic. J. Mech. Behav. Biomed. Mater 49, 162–174. [DOI] [PubMed] [Google Scholar]
- Albakry M, Guazzato M, Swain MV, 2003. Fracture toughness and hardness evaluation of three pressable all-ceramic dental materials. J. Dent 31, 181–188. [DOI] [PubMed] [Google Scholar]
- Beuer F, Schweiger J, Edelhoff D, 2008. Digital dentistry: an overview of recent developments for CAD/CAM generated restorations. Brit. Dent. J 204, 505–511. [DOI] [PubMed] [Google Scholar]
- Bhowmick S, Zhang Y, Lawn BR, 2005. Competing fracture modes in brittle materials subject to cyclic loading in liquid environments: Bilayer structures. J. Mater. Res 20, 2792–2800. [Google Scholar]
- Bühler-Zemp P, Völkel T, 2005. IPS e.max CAD Scientific documentation. Ivoclar Vivadent. 1–16. [Google Scholar]
- Budinski KG, Budinski MK, 2010. Engineering Materials: Properties and Selection, 9th ed. Pearson, New Jersey. [Google Scholar]
- Cai H, Kalceff MAS, Hooks BM, Lawn BR, 1994. Cyclic fatigue of a mica-containing glass-ceramic at Hertzian contacts. J. Mater. Res 9, 2654–2661. [Google Scholar]
- Coldea A, Swain MV, Thiel N, 2014. Hertzian contact response and damage tolerance of dental ceramics. J. Mech. Behav. Biomed. Mater 34, 124–133. [DOI] [PubMed] [Google Scholar]
- Deng Y, Lawn BR, Lloyd IK, 2002. Characterization of damage modes in dental ceramic bilayer structures. J. Biomed. Mater. Res. (Appl. Biomater.) 63, 137–145. [DOI] [PubMed] [Google Scholar]
- Denry I, Holloway JA, 2010. Ceramics for dental application: A review. Materials 3, 351–368. [Google Scholar]
- El-Meliegy E, van Noort R, 2012. Glasses and glass ceramics for medical applications. Springer, New York. [Google Scholar]
- Guess PC, Zhang Y, Kim J-W, Rekow ED, Thompson VP, 2010a. Damage and reliability of Y-TZP after cementation surface treatment. J. Dent. Res 89, 592–596. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Guess PC, Zavenalli RA, Silva NRFA, Bonfante EA, Coelho PG, Thompson VP, 2010b. Monolithic CAD/CAM lithium disilicate versus veneered Y-TZP crowns: Comparison of failure modes and reliability after fatigue. Int. J. Prosthodont 23, 434–442. [PubMed] [Google Scholar]
- Guiberteau F, Padture N, Cai H, Lawn BR, 1993. Indentation fatigue: A simple cyclic Hertzian test for measuring damage accumulation in polycrystalline ceramics. Philos. Mag. A 68, 1003–1016. [Google Scholar]
- Höland W, Schweiger M, Frank M, Rheinberger V, 2000. A comparison of the microstructure and properties of the IPS Empress®2 and IPS Empress®glass-ceramics. J. Biomed. Mater. Res. Part B: Appl. Biomater 53, 297–303. [DOI] [PubMed] [Google Scholar]
- Jung Y-G, Peterson IM, Kim DK, Lawn BR 2000. Lifetime-limiting strength degradation from contact fatigue in dental ceramics. J. Dent. Res 79, 722–731. [DOI] [PubMed] [Google Scholar]
- Kim J-W, Kim J-H, Thompson VP, Zhang Y, 2007. Sliding contact fatigue damage in layered ceramic structure. J. Dent. Res 86, 1046–1050. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kosmač T, Oblak C, Marion L, 2008. The effects of dental grinding and sandblasting on ageing and fatigue behaviour of dental zirconia (Y-TZP) ceramics. J. Eur. Ceram. Soc 28, 1085–1090. [Google Scholar]
- Krejci I, Albert P, Lutz F, 1999. The influence of antagonist standardization on wear. J Dent Res 78, 713–719. [DOI] [PubMed] [Google Scholar]
- Kruzic JJ, Arsecularatine JA, Tanaka CA, Hoffman MJ, Cesar PF 2018. Recent advances in understanding the fatigue and wear behavior of dental composites and ceramics. J. Mech. Behav. Biomed. Mater 88, 504–533. [DOI] [PubMed] [Google Scholar]
- Lawn BR, 1998. Indentation of ceramics with spheres: A century after Hertz. J. Am. Ceram. Soc 81, 1977–1994. [Google Scholar]
- Lawn BR, Bhowmick S, Bush MB, Qasim T, Rekow ED, Zhang Y, 2007. Failure modes in ceramic-based layer structures: A basis for materials design of dental crowns. J. Am. Ceram. Soc 90, 1671–1683. [Google Scholar]
- Lawn BR, Deng Y, Lloyd IK, Janal MN, Rekow ED, Thompson VP, 2002a. Materials design of ceramic-based layer structures for crowns. J. Dent. Res 81, 433–438. [DOI] [PubMed] [Google Scholar]
- Lawn BR, Deng Y, Miranda P, Pajares A, Chai H, Kim DK, 2002b. Overview: Damage in brittle layer structures from concentrated loads. J. Mater. Res 17, 3019–3036. [Google Scholar]
- Lawn BR, Deng Y, Thompson VP, 2001. Use of contact testing in the characterization and design of all-ceramic crownlike layer structures: A review. J. Prosthet. Dent 86, 495–510. [DOI] [PubMed] [Google Scholar]
- Lawn BR, Pajares A, Zhang Y, Deng Y, Polack MA, Lloyd IK, Rekow ED, Thompson VP, 2004. Materials design in the performance of all-ceramic crowns. Biomaterials 25, 2885–2892. [DOI] [PubMed] [Google Scholar]
- Li RWK, Chow TW, Matinlinna JP, 2014. Ceramic dental biomaterials and CAD/CAM technology: State of the art. J. Prosthodont. Res 58, 208–216. [DOI] [PubMed] [Google Scholar]
- Lorenzoni FC, Bonfante EA, Valverde GB, Coelho PG, Bonfante G, Thompson VP, Silva NRFA, 2020. Effect of indenter material on reliability of all-ceramic crowns. J. Mech. Behav. Biomed. Mater 108, 103831. [DOI] [PubMed] [Google Scholar]
- Ma L, Guess PC, Zhang Y, 2013. Load-Bearing Properties of Minimal-Invasive Monolithic Lithium Disilicate and Zirconia Occlusal Onlays: Finite Element and Theoretical Analyses. Dent. Mater 29, 742–751. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Miyazaki T, Nakamura T, Matsumura H, Ban S, Kobayashi T 2013. Current status of zirconia restoration. J. Prosthodont. Res 57, 236–261. [DOI] [PubMed] [Google Scholar]
- Ortiz AL, Borrero-López, Guiberteau F, Zhang Y 2019. Microstructural development during heat treatment of a commercially available dental-grade lithium disilicate glass-ceramic. Dent. Mater 35, 697–708. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Padture NP, Lawn BR, 1995. Contact fatigue of a silicon carbide with a heterogeneous grain structure. J. Am. Ceram. Soc 78, 1431–1438. [Google Scholar]
- Pajares A, Wei L, Lawn BR, 1995. Damage accumulation and cyclic fatigue in Mg-PSZ at Hertzian contacts. J. Mater. Res 10, 2613–2625. [Google Scholar]
- Peterson IM, Pajares A, Lawn BR, Thompson VP, Rekow ED, 1998. Mechanical characterization of dental ceramics by Hertzian contacts. J. Dent. Res 77, 589–602. [DOI] [PubMed] [Google Scholar]
- Rekow ED, Silva NRFA, Coelho PG, Zhang Y, Guess P, Thompson VP, 2011. Performance of dental ceramics: Challenges for improvements. J. Dent. Res 90, 937–952. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rekow D, Thompson VP, 2005. Near-surface damage-a persistent problem in crowns obtained by computer-aided design and manufacturing. Proc. Inst. Mech. Eng. H 219, 233–243. [DOI] [PubMed] [Google Scholar]
- Rekow D, Thompson VP, 2007. Engineering long term clinical success of advanced ceramic prostheses. J. Mater. Sci:-Mater. Med 18, 47–56. [DOI] [PubMed] [Google Scholar]
- Rhee Y-W, Kim H-W, Deng Y, Lawn BR, 2001. Contact-induced damage in ceramic coatings on compliant substrates: fracture mechanics and design. J. Am. Ceram. Soc 84, 1066–1072. [Google Scholar]
- Ritzberger C, Apel E, Höland W, Peschke A, Rheinberger VM, 2010. Properties and clinical application of three types of types of dental glass-ceramics and ceramics for CAD-CAM technologies. Materials 3, 3700–3713. [Google Scholar]
- Silva NRFA, Sailer I, Zhang Y, Coelho PG, Guess PC, Zembic A, Kohal RJ, 2010. Performance of Zirconia for dental health. Materials 3, 863–896. [Google Scholar]
- Silva NRFA, Thompson VP, Valverde GB, Coelho PG, Powers JM, Farah JW, Esquivel-Upshaw J, 2011. Comparative reliability analyses of zirconium oxide and lithium disilicate restorations in vitro and in vivo. J. Am. Dent. Assoc 142, 4s–9s. [DOI] [PubMed] [Google Scholar]
- Studart AR, Filser F, Kocher P, Gauckler LJ, 2007. In vitro lifetime of dental ceramics under cyclic loading in water. Biomaterials 28, 2695–2705. [DOI] [PubMed] [Google Scholar]
- Yin L, Lymer R, Billiau N, Peng Z, Stoll R, 2013. Damage morphology produced in low-cycle high-load indentations of feldspar porcelain and leucite ceramic. J. Mater. Sci 48, 7902–7912. [Google Scholar]
- Zhang Y, Bhowmick S, Lawn BR 2005. Competing fracture modes in brittle materials subject to concentrated cyclic loading in liquid environments: Monoliths. J. Mater. Res 20, 2021–2029. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zhang Y, Kelly JR, 2017. Dental ceramics for restoration and metal-veneering. Dent. Clin. N. Am 61, 797–819. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zhang Y, Kim J-W, Bhowmick S, Thompson VP, Rekow ED, 2009. Competition of fracture mechanisms in monolithic dental ceramics: Flat model system. J. Biomed. Mater. Res. Part B: Appl. Biomater 88B, 402–411. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zhang Y, Lawn BR, 2004. Long-term strength of ceramics for biomedical applications. J. Biomed. Mater. Res. Part B: Appl. Biomater 69B, 166–172. [DOI] [PubMed] [Google Scholar]
- Zhang Y, Lawn BR 2005. Fatigue sensitivity of Y-TZP to microscale sharp-contact flaws. J. Biomed. Mater. Res. Part B: Appl. Biomater 72B, 388–392. [DOI] [PubMed] [Google Scholar]
- Zhang Y, Lawn BR, Rekow ED, Thompson VP, 2004a. Effect of sandblasting on the long-term performance of dental ceramics. J. Biomed. Mater. Res. Part B: Appl. Biomater 71B, 381–386. [DOI] [PubMed] [Google Scholar]
- Zhang Y, Lee JJ-W, Srikanth R, Lawn BR, 2013a. Edge chipping and flexural resistance of monolithic ceramics. Dent. Mater, 29, 1201–1208. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zhang Y, Pajares A, Lawn BR 2004b. Fatigue and damage tolerance of Y-TZP ceramics in layered biomechanical systems. J. Biomed. Mater. Res. Part B: Appl. Biomater 71B, 166–171. [DOI] [PubMed] [Google Scholar]
- Zhang Y, Sailer I, Lawn BR, 2013b. Fatigue of dental ceramics. J. Dent 41, 1135–1147. [DOI] [PMC free article] [PubMed] [Google Scholar]