JJOD 2270 1–10 journal of dentistry xxx (2014) xxx–xxx

Available online at www.sciencedirect.com

ScienceDirect journal homepage: www.intl.elsevierhealth.com/journals/jden 1 2 3

Effect of core ceramic grinding on fracture behaviour of bilayered lithium disilicate glass–ceramic under two loading schemes

4 5

6 7 8 9 10 11 12 13 14


Xiao-Dong Wang a,e, Yu-Tao Jian b,e, Petra C. Guess c, Michael V. Swain c, Xin-Ping Zhang d,**, Ke Zhao a,* a

Department of Prosthodontics, Guanghua School of Stomatology, Sun Yat-sen University, Guangdong Provincial Key Laboratory of Stomatology, Guangzhou, China b Institute of Stomatological Research, Guangdong Provincial Key Laboratory of Stomatology, Sun Yat-sen University, Guangzhou, China c Department of Prosthodontics, School of Dentistry, Albert-Ludwigs University, Freiburg, Germany d Department of Metallic Materials Science and Engineering, School of Materials Science and Engineering, South China University of Technology, Guangzhou, China

article info


Article history:

Objectives: The purpose of this in vitro study was to evaluate the effect of core ceramic grinding on the fracture

Received 6 January 2014

behaviour of bilayered lithium disilicate glass–ceramic (LDG) under two loading schemes.

Received in revised form 25 March 2014 Accepted 26 March 2014 Available online xxx

Methods: Interfacial surfaces of sandblasted LDG disks (A) were ground with 220 (B), 500 (C) and 1200 (D) grit silicon carbide (SiC) sandpapers, respectively. Surface roughness and topographic analysis were performed using a profilometer and a scanning electron microscopy (SEM), and then underwent retesting after veneer firing. Biaxial fracture strength (sf) and Weibull modulus (m) were calculated either with core in tension (subgroup t) or in compression (subgroup c). Failure modes were observed by SEM, and loading induced stress distribution was simulated and analyzed by finite element analysis. Statistical data analysis was performed using Kruskal–Wallis,

Keywords: Dental porcelain

one-way ANOVA, and paired test at a significance level of 0.05. Results: As the grits size of SiC increased, LDG surface roughness decreased from group A to D ( p < 0.001), which remained unchanged after veneer firing. No difference in sf ( p = 0.41 for subgroups At–Dt; p = 0.11 for subgroups Ac–

Fracture strength

Dc), m values as well as failure modes was found among four subgroups for both loading schemes. Specimens in

Finite element analysis

subgroup t showed higher sf ( p < 0.001) and m values than subgroup c. Stress distribution between loading schemes


did not differ from each other. Cracks, as the dominant failure mode initiated from bottom tensile surface. No sign of interfacial cracking or delamination was observed for all groups. Conclusions: Technician grinding changed surface topography of LDG ceramic material, but was not detrimental to the bilayered system strength after veneer application. LDG bilayered system was more sensitive to fracture when loaded with veneer porcelain in tension. Clinical significance: Within the limitations of the simulated grinding applied, it is concluded that veneer porcelain can be applied directly after technician grinding of LDG ceramic as it has no detrimental effect on the strength of bilayered structures. The connector areas of LDG fixed dental prosthesis are more sensitive to fracture compared with single crowns, and should be fabricated with more caution. # 2014 Elsevier Ltd. All rights reserved.

17 15 16 18 * Corresponding author at: Department of Prosthodontics, Hospital of Stomatology, Sun Yat-sen University, 56 Lingyuan West Road, Q2 Guangzhou 510055, China. Tel.: +86 20 83802805; fax: +86 20 83822807.

** Corresponding author. E-mail addresses: [email protected] (X.-P. Zhang), [email protected] (K. Zhao). e

Joint first authors: these authors contributed equally to this work. http://dx.doi.org/10.1016/j.jdent.2014.03.014 0300-5712/# 2014 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Wang X-D, et al. Effect of core ceramic grinding on fracture behaviour of bilayered lithium disilicate glass– ceramic under two loading schemes. Journal of Dentistry (2014), http://dx.doi.org/10.1016/j.jdent.2014.03.014

JJOD 2270 1–10


journal of dentistry xxx (2014) xxx–xxx

19 20


21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77

The increasing patients’ demand for better aesthetics and biocompatibility has been driving the development of allceramic dental materials. With modification of ceramic microstructure and mechanical properties, the fracture strength of lithium disilicate glass–ceramic (LDG) achieves almost 500 MPa,1 a value sufficient to bear normal occlusal generated stresses that are not generally higher than 120 MPa.2 The translucent properties of LDG restorations imparts relatively natural aesthetic effect for clinical applications in monolithic and bilayer configuration and thereby, it has been widely used and become of increasing research interest in recent years. Due to the limited transparency of core materials, bilayered ceramic restorations have been introduced to provide veneer support and improved aesthetics, particularly in the anterior dentition.3 Veneering porcelain has a crucial influence on the mechanical behaviour of bilayered all-ceramic structures and may result in a more complex stress distribution than that of single component systems.4 Although bilayered LDG single crowns demonstrate a comparable 5-year fracture rate with conventional metal–ceramic crowns,5 for fixed partial dentures (FPD), a higher rate of catastrophic fracture and veneer chipping incidence occurs for both anterior and posterior areas.6,7 The fracture of FPDs predominately occur due to cracking within connector areas.7 In vitro fractography observations and finite element analysis (FEA) of FPDs have shown that a peak of tensile stress concentration in veneer porcelain appears at the connector area.8,9 Further investigation has indicated that there are three microscopic fracture mechanisms in bilayered ceramic structures: cone cracking, subsurface quasi-plastic deformation, and radial cracking.10 But other research has suggested that the core–veneer interface is an important fracture initiation mechanism of bilayered ceramic structures.11 In a study on failure behaviour of ceramic FPDs, it was found that approximately 70–78% connector fracture originated from the core–veneer interface.8 Core–veneer interface is regarded as the ‘‘weakest link’’ in the design of bilayered ceramic structures and crucial to the life time of restorations.12,13 Dental ceramic restorations are subjected to a set of fabrication procedures in the laboratory, and grinding of LDG ceramic is a standardized procedure used to improve fit after divesting and finishing. Additionally, in the cut back veneering technique, dental technicians need to reduce the shape of the core ceramic to create space for additional veneer application. The processing inevitably results in a core surface with complex topography and roughness, which is an essential characteristic of bilayered structures reflecting the tortuous modality and defect distribution in the resulting porcelain after veneering. It has been argued that although a rough interface can transfer high tensile stress from core to veneer layer, it simultaneously lowers the strength of veneer porcelain and consequently promotes fracture at the interface.14 It was also found that a smooth core–veneer interface may result in a significantly increased fracture strength of bilayered specimens of In-Ceram Alumina.15 Grinding procedures of the core ceramic, therefore, appear to affect the


fracture behaviour of bilayered structures by means of changing interfacial topography, roughness and defects. The above findings, however, were only for a glassinfiltrated alumina ceramic.15,16 It is known that the microstructure of LDG is different from that of the glass infiltrated alumina ceramic, and is composed predominantly of needle like crystals in approximately 30 vol% glass phase.1 This structure is deemed to form a modified glass composition of the core ceramic surface during the veneering heat treatment, and consequently change the interfacial topography and roughness.17 The aim of this study was thereby to evaluate the effect of core ceramic grinding on fracture behaviour of a bilayered LDG system. Two different loading schemes were used to represent schematic clinical stress distributions in bilayered crowns and connector areas of FPDs, respectively. The primary null hypothesis was that grinding of LDG ceramic had no influence on the fracture behaviour of the bilayered structure.


Materials and methods


Sample preparation

78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97

184 IPS e.max Press (LT A2, Ivoclar Vivadent, Schaan, Liechtenstein) core ceramic disks with the geometry of 15.0 mm  1.4 mm (diameter  height) were fabricated according to the manufacturer’s instructions, and then were sandblasted (Silfradent SRL, Sofia, Italy) using 110 mm Al2O3 at 1.5 bar pressure for 20 s with a distance of 10 mm to remove the investment materials and reaction layers completely, at the same time to formulate a coarse surface with uniform topography. All specimens were cleaned in an ultrasonic cleaner with distilled water for 5 min prior to further treatments.

98 99 100 101 102 103 104 105 106 107 108



Treatment of core interfacial surface

Grinding with diamond burs is closer to the clinical situation, but it is hard to form a comparable flat surface by this nonuniform abrasion procedure leading to substantial differences between samples.16 We therefore used a series of silicon carbide (SiC) sandpapers for grinding on an automatic metallographic lapping machine (Tegramin-30, Struers, Ballerup, Denmark) to generate a similar core–veneer interfacial topography to that ground with diamond burs.18 The sandblasted core specimens were divided into four groups (n = 46) randomly, and the core–veneer interfacial surfaces were wet ground with SiC sandpapers of different grits to formulate a specific topography and roughness: Group A without any grinding as control; Group B ground with SiC sandpapers of 100 grit and then 220 grit (equivalent 75 mm diamond burs); Group C ground first with 100, 220 and 500 grit SiC sandpapers (equivalent 25 mm diamond burs); Group D ground successively with 100, 220, 500, 800, 1000 and 1200 grit SiC papers as well (equivalent 5 mm diamond burs). Grinding was performed at a speed of 200 rpm with a constant force of 20 N. The opposite surfaces were ground with 500 grit, and the final size of 15.0 mm  1.1 mm (diameter  height) with the

Please cite this article in press as: Wang X-D, et al. Effect of core ceramic grinding on fracture behaviour of bilayered lithium disilicate glass– ceramic under two loading schemes. Journal of Dentistry (2014), http://dx.doi.org/10.1016/j.jdent.2014.03.014

110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130

JJOD 2270 1–10


journal of dentistry xxx (2014) xxx–xxx

131 132

parallelism tolerance less than 0.02 mm in 10 mm was determined by a micrometre.



134 135 136 137 138 139 140 141 142 143 144 145 146 147 148

The interfacial surface roughness of the core ceramic after grinding was measured using an optical 3D profilometer (SMS Expert, Breitmeier, Ettlingen, Germany) with a LED white light source probe (light spot size: 6 mm, vertical resolution: 30 nm, max lateral travel: 50 mm). The measurement was performed over a scan distance of 4 mm (according to BS EN 623-4:200419) across the centre of the treated surface three times, and the average (Ra) and maximum (Ry) surface roughness values were recorded (n = 40), describing the mathematical average of irregularities measured from mean line and the maximum peak-to-valley height measured parallel to mean line, respectively.20 Three extra specimens of each group were sputter coated with gold for topography analysis with a scanning electron microscopy (SEM) (Quanta 200, FEI, Eindhoven, The Netherlands).



150 151 152 153 154 155 156 157 158

To evaluate the effect of veneer firing procedure on the topography of LDG, all specimens were heated according to the manufacturer’s recommended dentine firing parameters but omitting veneer porcelain application step. After heat treatment, Ra and Ry values of ground surface were undertaken by retesting based on the aforementioned method, and another three extra specimens of each group were observed using SEM to assess the topography variation after veneer firing simulation.



160 161 162 163 164 165 166 167 168 169 170 171 172

All specimens were ultrasonically cleaned for 5 min and air dried afterwards. Before veneer application, a thin wash layer (Build-Up Liquids mixed with Dentine porcelain powder) was coated on core ceramic and fired as the manufacturer recommends. IPS e.max Ceram Dentine (Ivoclar Vivadent, Schaan, Liechtenstein) slurry was brushed on the top of core specimens under the control of a stainless steel set-up mould, and excess slurry was removed using a sharp blade. To compensate for the about 20% linear firing shrinking of veneer porcelain, two more correction veneer applications were performed. The final size of bilayered disks was 15.0 mm  2.0 mm (diameter  height) with core to veneer thickness ratios of ca. 1.1:0.9.



174 175 176 177 178 179 180 181 182

The maximum load P (N) at fracture was measured with piston-on-three-ball biaxial flexural testing with a universal testing machine (AG-X 100, Shimadzu, Kyoto, Japan) according to ISO 6872:2008 (E) and relative references.21,22 Three austenitic steelballs with a diameter of 5 mm were positioned 1208 apart on a support circle with a diameter of 10 mm. Specimens in each group (n = 40) were divided randomly into two subgroups: subgroups At–Dt (n = 20) were loaded with the core ceramic in tension; while subgroups Ac–Dc (n = 20)

Roughness measurement

Veneer firing simulation

were tested with the core ceramic in compression (i.e., veneer porcelain in tension). A tinfoil sheet (thickness: 0.6 mm) was inserted between the loading piston and the disks in order to evenly distribute pressures on the disks. The disks were loaded at the centre with a flat piston (diameter: 1.5 mm) at a crosshead speed of 1 mm/min until fractured. The flexural strength was calculated according to Eqs. (1a)–(1d)22,23: sf ¼

PE1 ð1 þ vÞðz  z Þ 8pð1  v21 ÞD    a 1  v  c2 a2 þ 1 2  1 þ 2ln c 1þv 2a R2


ðE1 t21 =2ð1  v21 ÞÞ þ ðE2 t22 =2ð1  v22 ÞÞ þ ðE2 t1 t2 =ð1  v22 ÞÞ ðE1 t1 =ð1  v21 ÞÞ þ ðE2 t2 =ð1  v22 ÞÞ


191 192 193 194 195 196 197 198 199 200 201 202 203 204 206 205

D ¼

E2 t32 E1 t31 E2 t1 t2 ðt1 þ t2 Þ þ þ 2 1  v22 3ð1  v1 Þ 3ð1  v22 Þ

207 2

½ðE1 t21 =2ð1  v21 ÞÞ þ ðE2 t22 =2ð1  v22 ÞÞ þ ðE2 t1 t2 =ð1  v22  ÞÞ ðE1 t1 =ð1  v21 ÞÞ þ ðE2 t2 =ð1  v22 ÞÞ (1c)

Bilayered sample preparation

Biaxial loading testing and Weibull analysis


where sf is the maximum tensile stresses at fracture in bottom layers of layered specimens; P is the load to fracture; a, c and R are the radii of the supporting circle, the piston and the specimen disc, respectively; v1 and v2 are the Poisson’s ratio of bottom and top layer, respectively, here v1 = 0.26 for IPS e.max Press and v2 = 0.23 for IPS e.max Ceram24; E1 and E2 are the Young’s modulus of bottom and top materials, i.e., E1 = 95 GPa for IPS e.max Press and E2 = 65 GPa for IPS e.max Ceram1; z is core–veneer interface direction in vertical cylindrical coordinates; z*, D* and v are the position of the neutral surface, flexural rigidity and average Poisson’s ratio calculated according to the following equations. Q3 z ¼

183 184 185 186 187 188 189

209 208 v¼

v1 t1 þ v2 t2 t1 þ t2


Weibull analysis was performed based on the data flexural strength to evaluate the structural reliability bilayered ceramic structures. Weibull modulus (m) and 95% confidence interval were calculated as Eq. (2)21:   m  s P f ¼ 1  exp  s0

of of its


210 213 211 212 214 215 216 217

where Pf is the fracture probability, s is the fracture strength at a given Pf, s0 is the characteristic strength.

218 219 220



Finite element analysis

To evaluate the stress distribution in bilayered specimens, a 3D finite element analysis (FEA) was performed using ANSYS 9.0 (Houston, USA). Symmetric specimen was modelled using 8-node elements (solid 45), assuming the core–veneer interface remained perfectly bonded during computation. The model was composed of 106,472 elements and 124,451 nodes in total, while the specimen was composed of 50,512 elements and 55,437 nodes with a significantly finer mesh size of 0.16 mm (Fig. 1). An average load magnitude of about 500 N and 300 N were applied, simulating the load at fracture for subgroups At–Dt and Ac–Dc, respectively.

Please cite this article in press as: Wang X-D, et al. Effect of core ceramic grinding on fracture behaviour of bilayered lithium disilicate glass– ceramic under two loading schemes. Journal of Dentistry (2014), http://dx.doi.org/10.1016/j.jdent.2014.03.014

222 223 224 225 226 227 228 229 230 231 232

JJOD 2270 1–10


journal of dentistry xxx (2014) xxx–xxx



234 235 236 237 238 239 240

All fractured specimens of each subgroup were then sputter coated for fractographic analysis using SEM. Failure modes, initiation as well as propagation of the cracks were visualized and compared. Wake hackle markings from the pores, cone cracks, radial cracks were selected as reference points for determination of fracture modes in the bilayered ceramic structures.25



242 243 244 245 246 247 248 249 250 251

The differences in the interfacial surface roughness values of LDG among groups A-D were compared, respectively, before and after veneer firing simulation by Kruskal–Wallis and Dennett T3 test. Additionally, Wilcoxon paired test was used to compare the roughness values before and after firing for each group. Statistical differences of the fracture load and biaxial flexural strength of the four groups A–D were analyzed by one-way ANOVA and Bonferroni test. Paired t-test was used to determine the difference between two loading schemes for each subgroup (SPSS 15.0, a = 0.05).26


Analysis of failure modes

Statistical analysis




Surface roughness and topography


254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270

Ra and Ry values of treated LDG ceramic are listed in Table 1. The control group A recorded the highest Ra and Ry values, while specimens in group D showed the lowest values. Decreased Ra and Ry values were found between each of the two groups as the grit sizes of SiC papers becoming finer from group A to D ( p < 0.001). The descending tendency was not changed and differences within groups from group A to D were still significant after heat treatment ( p < 0.001). No difference was found between before (BH) and after (AH) heat treatment for all the four groups. The SEM observations of group A revealed that a nonuniform ‘‘bumpy’’ core surface with pores and grooves was produced by air abrasion. The defects were removed after grinding; the remaining groups of B–D had smoother surfaces with fewer and finer scratches (Fig. 2A–E). After heat simulation, the sharp projections decreased and became rounded to some degree by coated with a new glass-like

Fig. 1 – Symmetric configuration finite element model with a significantly finer mesh size in bilayered specimen.

coating formed during the heating procedure, whilst the surface was still coarse as the valleys and grooves generated through air abrasion or grinding failed to be filled (Fig. 2F).

271 272 273



Fracture stress and Weibull analysis

Mean load at fracture (P), biaxial flexural strength (sf), and Weibull modulus (m) of all groups are listed in Table 2. No difference in P ( p = 0.36 for subgroups At–Dt; p = 0.08 for subgroups Ac–Dc) and sf ( p = 0.41 for subgroups At–Dt; p = 0.11 for subgroups Ac–Dc) was found among the four subgroups for both loading schemes. In subgroups A–D, respectively, specimens loaded with core ceramic in tension showed higher P and sf values ( p < 0.001) than those with core ceramic in compression without exception. Again for the Weibull statistics, no statistical difference was found within subgroups for both loading schemes (Table 2). When the subgroups are compared with each other, At–Dt showed higher m values than Ac–Dc, respectively.

275 276 277 278 279 280 281 282 283 284 285 286 287



Finite element analysis

A fast load of 500 N and 300 N gave rise to catastrophic fracture in subgroups At–Dt and Ac–Dc, respectively. The critical biaxial loading was then used to calculate the corresponding stress in FEA. The stress distribution in bilayered structures was dependent upon which material was loaded in tension

Table 1 – Average, maximum surface roughness and SDs of core ceramic (mm). Groups


Average roughness (Ra) BH

A B C D Chi-square p

Sandblasted (as control) 220 grit SiC sandpaper 500 grit SiC sandpaper 1200 grit SiC sandpaper


2.15  0.32 a 2.13  0.39 a 0.73  0.04 b 0.72  0.04 b 0.39  0.03 c 0.39  0.03 c 0.12  0.03 d 0.12  0.02 d 149.18 149.12

Effect of core ceramic grinding on fracture behaviour of bilayered lithium disilicate glass-ceramic under two loading schemes.

The purpose of this in vitro study was to evaluate the effect of core ceramic grinding on the fracture behaviour of bilayered lithium disilicate glass...
4MB Sizes 0 Downloads 3 Views