Comparative Three-Point Bending Flexural Test of Dental Ceramics Using Standard and Self-Adjustable Specimen Support Fixtures

Abstract

The flexural strength of dental ceramic specimens was comparatively measured by using the standard fixture and the self-adjustable specimen support fixture which carefully prepared for this work. One hundred and twenty dental ceramic specimen bars fabricated in accordance with the BS EN ISO 6872 standard were ground, polished and classified into 6 groups with opposing surface parallelism differences of 0.00, 0.01, 0.02, 0.03, 0.04 and 0.05 mm. Each group contained 20 specimens which ten specimens were tested by using the standard fixture and the other ten specimens were tested by using the self-adjustable specimen support fixture. The flexural strength and Weibull modulus of both specimen groups were analyzed using two way analysis of variance (ANOVA) and Weibull analysis. The fracture surfaces of both specimen groups were investigated under the scanning electron microscope and analyzed for their fracture behavior. Results from the study found that the self-adjustable specimen support fixture for the three-point flexural test was able to improve the force distribution on the surface of specimens under test. As a result, the actual flexural strength of materials under test with the higher value and reliability was achieved. The flexural strength from this study was found to be comparable to that measured with the bi-axial test. The strength values increased significantly from 79.1 ± 5.1 MPa to be 90.1 ± 3.4 MPa when tested with the standard fixture and the self-adjustable specimen support fixture respectively. The 10% failure stress and Weibull modulus of specimens tested with the self-adjustable specimen support fixture also improved obviously. The fracture surface characteristics of specimens tested with the standard fixture indicated most crack origin occurred at the specimen edge. In the specimens tested with the self-adjustable specimen support, the crack origin was observed as the appearance of the arrest lines or Wallner lines at the inner area beyond the specimen edge.

Fig. 3 The flexural strengths of specimen groups tested using the standard un-adjustable and self-adjustable specimen supports.

1. Introduction

Ceramics have been used for dental restorations such as crowns, bridges, inlays etc. Prior to use ceramics for dental restorations, the appropriate mechanical properties and specific characteristics, particularly ceramic strength, need to be evaluated. The strength of materials has usually been investigated using three-point (3PF) or four-point (4PF),^1–4) bi-axial (BI) flexural tests^1,2,5) and also fracture toughness tests.^6,7) Dental cements, ceramics and resin composites, which are brittle materials, have been comparatively tested for their failure stress using 4PF, BI and diametral tensile tests and the mean strength of all specimens tested by the bi-axial flexure test are higher and more reliable than the other methods.¹⁾ Although BI test results have been found more reliable for these brittle materials, preparing specimens for, and conducting, these tests are more complicated. Therefore, uniaxial testing methods are still more commonly used. In the uniaxial flexural strength testing, dental ceramics specimens are prepared, in accordance with ISO 6872, in to a flat rectangular bar shape with the surface parallelism difference less than 0.05 mm.⁸⁾ During testing, impact bar moves downward to be in a linear contact with the specimen surface, resulting in the origin of fracture easily occurring at the specimen edge. The surface roughness of all-ceramic crown materials has been found to reduce the reliability of the flexure strength data.⁹⁾ Flaws at ceramic edges, surfaces and in the interior of ceramic body affecting the strength of dental ceramics have been discussed.^2,10) Weibull statistics have been widely used for analysis various effects on the strength of various materials including dental ceramics such as influence of test metnod,^1,2) surface roughness,^5,9) shape parameters¹¹⁾ and flaw containing materials.¹²⁾ Elimination of edge flaws can enhance the 4PF strength and improve the Weibull modulus of the feldspathic porcelain.³⁾ Rounding edges of specimen under test is recommended for a better uniaxial flexural strength data. However, the size of the rounded edge specimen may not be correct and give rise to the error of the measured strength.

Because of those problems of specimen preparation in the uniaxial flexural measurement, the self-adjusting specimen support bars were prepared for the 3PF test in order to reduce the specimen edge effect and the effect from the specimen surface parallelism differences. The self-adjusting specimen support bars were replaceable and comparable with the standard specimen support fixture. This research work aimed to reduce the variance and to enhance the reliability of flexural strength values of dental porcelain from using the new specimen support bars under the 3PF test. The flexural strength data was analyzed using two way analysis of variance (ANOVA) and Weibull analysis and the actual flexural strength was expected to be achieved with a higher reliability. The fracture specimens were investigated under a scanning electron microscope and the fracture surfaces were analyzed.

2. Experimental Procedure

From the three point bending test of large engineering materials according to the ASTM designation C674-88, the specimen support has been assigned to be adjustable.¹³⁾ While, the specimen support for the three point bending test of small dental ceramics in accordance with the BS EN ISO 6872 standard is un-adjustable with the opposing surface parallelism difference of lower than 0.05 mm. In this work, the specimen support in the lower part of the test fixture was prepared to be self adjustable and was able to be replaced in to the Instron, DS4843 D-01 standard test fixture. The 410 stainless steel was used for preparation of the whole self adjustable specimen support test fixture which was carefully prepared using the wire cut machine. The surface roughness of these parts was closely similar to the polished surface of metal with 3 micron alumina polishing powder. The adjustable specimen support bars were prepared in to the semi-circular disk shape with a thickness of 2 mm and a radius of 8 mm. The semi-circular disks were coated with thin lubricant grease and slightly loose fitted into the half circular disk holes as the schematic draw shown in Fig. 1(a) and with the specimen under test in Fig. 1(b). All three points of the bars in contact with the specimen under test were similar to the standard test fixture, except that the two specimen support bars in this work were rotatable with the same rotating axis which was perpendicular to the downward travelling direction of the impact bar. As a result, the self adjustable specimen support bars were able to make a close contact to the rectangular specimen surface although the opposing surfaces were differently parallel or the specimen was slightly twisted.

Fig. 1

The schematic draws showing (a) the self adjustable specimen support at the lower part of the test fixture for three-point bending flexural test (b) with a specimen under test.

The ceramic specimens under test in these experiments were prepared, in accordance with the BS EN ISO 6872 standard, from a VITA VMK 95 body porcelain into a rectangular shape with a total number of 120 specimens. The specimens were prepared by casting with the stainless steel mold. The specimens were prepared using the standard method of the company by pre-heating at 600°C for 6 min, sintering at 930°C for 1 min and annealing in vacuum for 6 min. After cooling down, all specimens were polished to the final size of 1.2 × 4.0 × 23 mm³ with the opposing surface parallelism difference, Δt. In this work, the specimens were carefully polished and measured for the surface parallelism differences along the specimen length using the micrometer with a resolution of 0.002 mm. The specimen polishing and measurement were repeated until the assigned surface parallelism difference was achieved. The schematic draw of ceramic specimen indicating Δt is shown in Fig. 2(a) and the well polished specimen is shown in Fig. 2(b). The specimens were classified in to six groups with the opposing surface parallelism differences of 0.00, 0.01, 0.02, 0.03, 0.04 and 0.05 mm which were corresponding to the specimens groups 1 to 6 respectively. Each group contained 20 specimens which ten specimens were tested by using the standard fixture (groups 1a–6a) and the other ten specimens were tested by using the self adjustable specimen support fixture (groups 1b–6b).

Fig. 2

Dental ceramic specimens (a) the schematic draw showing the opposing surface parallelism difference (b) the actual specimen after sintering and polishing.

The specimens under test were placed on the universal testing machine (UTM, Instron 5566, MA, USA) with the standard fixture (Instron, DS4843 D-01) and with the self-adjustable specimen support fixture. Prior to setting the continuous cross head speed of the moving impact bar, the force of the impact bar was set to the specimen protection mode with a constant impact force of 1 N and the surface in-contact of specimens under test with the lower bar was checked. The cross head speed of these experiments was 1 mm/min according to the BS EN ISO 6872:1999 and the equivalent rotation speed of the semicircular specimen support at any opposing surface parallelism differences of specimens from 0.01–0.05 mm was 0.5 rad/min for the 4 mm specimen width. The flexural strength was analyzed statistically by two-way ANOVA and Weibull analysis.

The fracture surfaces of specimens after testing were investigated using the scanning electron microscope (JEOL, JSM = 5410LV, Tokyo, Japan) and the fracture micrographs were analyzed.

3. Results and Discussion

The flexural strengths of all specimen groups are shown in Fig. 3. The average values of the flexural strength of specimens tested using the self-adjustable specimen support are clearly higher than those tested using the standard un-adjustable specimen support and the deviation of flexural strength is also lower in every specimen group. The average flexural strengths are 79.1 ± 5.1 MPa and 90.1 ± 3.4 MPa for the specimens tested using the standard un-adjustable specimen support and the self-adjustable specimen support respectively. The results from testing with the self-adjustable specimen support are almost constant with the high average value and low deviation. While, the flexural strength of specimens tested using the standard test fixture seem to decrease as the specimen off-surface parallelism difference increased with the low average value and high deviation.

Fig. 3

The flexural strengths of specimen groups tested using the standard un-adjustable and self-adjustable specimen supports.

Furthermore, results from the test of the 4a and 4b specimen groups with 0.03 mm opposing surface parallelism difference are found to have very close similar values. From theoretical expectation, the flexural strength of both specimen groups with no surface parallelism difference should be similar and it should be observed in the 1(a) and 1(b) specimen groups. In this study, the 4a specimen group was re-measured and tested using the standard un-adjustable specimen support under the assumption that the test fixture possesses the basic surface parallelism difference of about 0.03 mm. Prior to testing, the specimens were placed in to the direction in which the basic surface parallelism difference was compensated with the appearance surface parallelism difference close to 0.00 mm. Result from re-measurement found that the average flexural strength increased from 82.7 ± 5.6 MPa to be 85.9 ± 3.5 MPa which is much closer to the measured value using the self-adjustable specimen support with a lower deviation. This is believed that the standard test fixture itself possesses the basic surface parallelism difference of about 0.03 mm and gives rise to the higher deviation of the test results. The flexural strength from testing with the standard test fixture was found to be similar to the company reference strength of 85 MPa. While, the value from testing with the self-adjustable specimen support fixture was clearly shown the better strength value up to 90 MPa which was closely similar to the more reliable value tested using the Bi-axial flexural test.

The results from two way ANOVA for flexural strength of both specimen groups, 1(a)–6(a) and 1(b)–6(b), with a total number of each group of 60 and with various opposing surface differences are shown in Table 1 and it is clearly seen that the strength of each group are different significantly (P-value = 5.91E-28). However, the effects of the opposing surface difference of specimens are found no significant difference (P-value = 0.024904).

Table 1 Results of two way ANOVA for flexural strength.

The flexural strength of both specimen groups were also analyzed using Weibull analysis. The graphs indicated the relationship between probability of survival and flexural strengths of specimen tested using standard test fixture and using the self adjustable specimen support are shown in Fig. 4(a) and 4(b) and the 10% failure stress are found to be 72.6 MPa and 85.8 MPa respectively.

Fig. 4

Probability of survival of specimens tested using (a) standard test fixture (b) self-adjustable specimen support.

The graphs showing the relationship between ln (strength) and ln ln (1/(1 − Pf)), where Pf is the probability of failure, are indicated in Fig. 5. Results from testing with standard test fixture and the self-adjustable specimen support are shown in Fig. 5(a) and 5(b). The Weibull modulus of 17.9 and 30.3 and the 95% Confidence limit of 16.5–19.2 and 28.1–32.8 are respectively achieved with the similar R² value of about 0.92. The results from Weibull analysis are closed similar to the comparison values reported by D. Hudak, Murat Tiryakioğlu.¹¹⁾ However, the high ratio of Weibull modulus of about 1.69 from this study indicate that the test using the self adjustable specimen support are clearly better and more reliable than those using the standard fixture and they agree well with the strength values as indicated in Fig. 3 and also with those from two way ANOVA analysis. From the details reported by Xu and et al.,²⁾ the comparative measurement of flexural strength of dental ceramics investigated using three different test methods and the results indicated that the flexural strength obtained from the bi-axial test was higher and more reliable than those from the three point and four point bending flexural tests. These were similar to the results in this study in which the 3PF strength of ceramic specimens tested with the self-adjustable specimen support was higher than that tested with the standard test fixture. The increase of the strength value was comparable to that tested with the bi-axial test as reported by Xu and et al.²⁾

Fig. 5

Relationship between ln (Strength) and ln ln (1/(1 − Pf)) of flexural strength of specimens tested under (a) standard test fixture (b) self-adjustable specimen support.

All characteristics values from the study are summarized in Table 2. The flexural strength, 10% failure stress, Weibull modulus and 95% confidence limit of specimens tested using the self-adjustable specimen support are higher than those using the standard test fixture.

Table 2 Summary of characteristics values of dental ceramics from 3PF strength measurements and Weibull analysis.

The SEM micrographs in Fig. 6 show the fracture surface taken from the 6(a) and 6(b) specimen groups which were the specimens with the 0.05 mm opposing surface parallelism difference. The fracture surface of the specimen tested with the standard un-adjustable support bars is shown in Fig. 6(a) and the origin of fracture is clearly observed, as indicated with the arrow, at the corner of the tension side of specimen. The fracture micrographs of the specimen tested with the self-adjustable specimen support are shown in Fig. 6(b)–Fig. 6(d). The arrest lines have been observed in the tension side of the specimen where the origin of fracture occurred as indicated in Fig. 6(b). The compression curls are observed in the area close to the compression side as shown in Fig. 6(c) and the Wallner lines indicated the origin of fracture is also observed in the tension side at low magnification as shown in Fig. 6(d). This appearance is similar to the fracture surface of an indented Vitadur Alpha specimen as reported by Taskonak and et al.¹⁴⁾ Result from fracture surface analysis showed that the edge effect including the effect of the surface parallelism difference decreased when tested with self-adjustable specimen support. As a result, the higher strength value was obtained with a lower deviation which indicated the higher reliability of measurement.

Fig. 6

SEM fracture micrographs of specimens tested using (a) the standard fixture (b), (c), (d) the self-adjustable specimen support fixture.

4. Conclusions

The self-adjustable specimen support fixture for the 3PF test prepared in this work was able to perform well after installation instead of the lower part of the standard test fixture of the Instron, DS4843 D-01. Results from the 3PF strength measurements of dental ceramics including the analysis of the fracture surface clearly indicated that testing with the self-adjustable specimen support fixture was able to reduce the specimen edge effect and the effect from the opposing surface parallelism difference of specimen under test. The obtained flexural strength with the high value and low deviation from testing with the self-adjustable specimen support fixture indicated the well improvement of the force distribution on the specimen surface including the higher reliability of the measurement. The increase of the measured strength of ceramic specimens was found to be comparable with those tested with the bi-axial flexure test. From the study, the 10% failure stress, Weibull modulus, and confidence limit also increased significantly. The results were found to agree well with the strength values and also with results from two way ANOVA analysis.

Acknowledgement

The authors gratefully acknowledge the research grants from the Faculty of Dentistry, Chiang Mai University.

REFERENCES

• 1)   S.  Ban and  K.J.  Anusavice: J. Dent. Res. 69 (1990) 1791–1799.
• 2)   Y.  Xu,  J.  Han,  H.  Lin and  L.  An: Regen. Biomater. 2 (2015) 239–244.
• 3)   G.D.  Quinn,  K.  Hoffaman and  J.B.  Quinn: Dent. Mater. 28 (2012) 502–511.
• 4)   M.R.  Edwards,  P.H.  Jacobsen and  G.J.  Williams: J. Dent. Res. 62 (1983) 1086–1088.
• 5)   G.J.P.  Fleming,  S.F.A.  El-Lakwah,  J.J.  Harris and  P.M.  Marquis: Dent. Mater. 20 (2004) 142–149.
• 6)   S.S.  Scherrer,  I.L.  Denery and  H.W.A.  Wiskott: Dent. Mater. 14 (1998) 246–255.
• 7)   M.  Guazzato,  M.  Albakry,  S.P.  Ringer and  M.V.  Swain: Dent. Mater. 20 (2004) 441–448.
• 8)  BS EN ISO 6872: Dental ceramic, (1999).
• 9)   G.  Bhamra,  W.M.  Palin and  G.J.P.  Fleming: J. Dent. 30 (2002) 153–160.
• 10)   P.F.  Messer,  V.  Piddock and  C.H.  Lloyd: J. Dent. 19 (1991) 51–55.
• 11)   D.  Hudak and  M.  Tiryakioğlu: Mater. Sci. Eng. A 528 (2011) 8028–8030.
• 12)   M.  Tiryakioğlu and  D.  Hudak: Metall. Mater. Trans. B 42 (2011) 1130–1135.
• 13)  ASTM Designation: C674-88 American standard test methods for flexural properties of ceramic whiteware material, Annual book of ASTM standards. Glass; Ceramic whiteware; 15.02, (Phiadelphia, PA, 1993) pp. 205–208.
• 14)   B.  Taskonak,  J.A.  Griggs,  J.J.  Mecholsky, Jr. and  J.H.  Yan: Dent. Mater. 24 (2008) 700–707.