The fracture toughness determination of fine-grained zirconia ceramics using the chevron notched beam method (CNB) was investigated to assess the feasibility of this method for quality assurance and material characterization.
CNB tests were performed using four different yttria-stabilized zirconia ceramics under various testing modes and conditions, including displacement-controlled and load-rate-controlled four point bending to assess the influence of slow crack growth and identify most suitable test parameters. For comparison, tests using single-edge V-notch beams (SEVNB) were conducted.
It was observed that the CNB method yields well-reproducible results. However, slow crack growth effects significantly affect the measured K IC values, especially when slow loading rates are used. To minimize the effect of slow crack growth, the application of high loading rates is recommended.
Despite a certain effort needed for setting up a sample preparation routine, the CNB method is considered to be very useful for measuring and controlling the fracture toughness of zirconia ceramics.
Due to their excellent mechanical properties and inertness, zirconia-based ceramics have been developed successfully and are broadly used for dental restauration and orthopedics . In particular, very high fracture toughness and mechanical strength values can be realized by tailoring the microstructure and composition of zirconia. However, the effect of transformation toughening mechanism that enables excellent strength properties may also be weakened in presence of permanent exposure to corrosive environment and mechanical stresses . Zirconia-based ceramics with enhanced resistance against stress crack corrosion and low-temperature degradation have been developed .
The fracture toughness K IC is one of the most important properties of ceramic materials exposed to cyclic loading, like it is the case in the field of e.g., dental applications. Therefore, the K IC is strongly recommended to be evaluated wherever mechanical stability of materials and components is of interest. Moreover, the fracture toughness has become a key property to evaluate and compare material quality.
To measure fracture toughness, the Single-Edge V-Notch Beam (SEVNB) method has become the most widely used standard . Here a V-notched beam is used which is prepared by a sharp blade in combination with a diamond suspension. This method was proved to be practical, reliable and accurate enough to yield reproducible results . However, it has been doubted that this standard is applicable for very fine-grained zirconia ceramics, due to the fact that even by using very sharp blades, the notch tip diameter would certainly be much larger than the average grain size, and small differences of the blade quality and type of diamond suspension might have a strong influence on the measured toughness value causing systematic deviations . A particularly critical drawback of the SEVNB method is that the lower the notch quality the higher will be the obtained K IC value, since lower notch qualities generally have increased notch tip diameters. It has been supposed that in case of low-quality notches the SEVNB technique may overestimate the K IC value of modern fine-grained zirconia ceramics by a factor of 1.5 .
As an alternative testing method, the Chevron Notch Beam (CNB) test has been identified, which is also a standardized method to evaluate the fracture toughness of ceramic materials and was previously reported to yield well-reproducible results with low standard deviation also in glass . By accurately following the prescribed preparation steps of this standard, a pre-crack should be produced that will at first grow stably upon bending. The corresponding Stress-Intensity Factors (SIF) as a function of crack length are known , and from the maximum load measured during the bending test, the K IC value can be easily determined. The main advantage of this method is that it is independent of the notch tip diameter, since due to stable crack growth, the K IC of a real (natural) crack is measured.
Despite the expected benefits for measuring the K IC of fine-grained zirconia ceramics, there have been some drawbacks of the chevron notch tests: first, the stadium of stable crack growth is not reached in every test. Therefore, there is always a considerable number of tests that do not yield a valid result. This difficulty has been counteracted by two ways:
Inducing crack formation in compression as recommended in DIN EN 14425-3
Slow loading rates
However, while an appropriate pre-cracking step significantly enhances the rate of valid test results, the use of slow load rates leads to a significant influence of subcritical crack growth effects, which result in lower measured K IC values .
Aim of the present work was to assess the feasibility of CNB test method for the measurement of the K IC of zirconia ceramics. In particular, it was surveyed if this test may be a preferred method for the quality control of zirconia ceramics manufacture. Four zirconia materials were selected and tested using different test parameters to identify suitable conditions and parameters for the K IC determination.
The specimens were prepared out of pre-sintered dental milling blanks via high pressure waterjet-cutting followed by grinding to the desired size on a surface grinding machine. The sintering was performed according to supplier recommendations. The final specimen size according EN14425-3 was about 3.0 × 4.0 × 45.0 mm 3 .
Two samples were mounted in a special sample holder ( Fig. 1 ) with antidromic angulation of 26°30′. The chevron notch was then cut with a precision diamond saw (Buehler ISOMET 4000) and a metal bound wafering blade with low diamond content (Buehler series 15LC; Ø 76 mm × 0.20 mm). Revolution and cutting speed were 5000 rpm and 4 mm/min respectively. After the first cut, the position of the two samples was exchanged, ensuring a final chevron notch geometry in compliance with EN14425-3. Since small variations of the notch geometry – even within the allowed geometry range of the EN14425-3 – have a significant influence on the obtained K IC values, the geometry parameters of the Chevron notches ( Fig. 1 b) should be close to the upper limits but avoid overcuts ( a 0 : 0.80 ± 0.08 and a 1 : min. 3.80 to max. W ).
After cutting, samples were cleaned for 5 min with acetone in an ultrasonic bath (Bandelin electronic RK 510) and afterwards dried for 30 min at 110 °C in a drying cabinet (Heraeus UT 6060).
The CNB tests were conducted according to DIN EN 14425-3, sect. 7. After measuring the dimensions of the samples and the notches, the samples were precisely inserted into the 4-point-test assembly. The tests were conducted at room temperature in a lab environment (max. 50% humidity). According to sect. 7.2 (comment 2), the rate of valid test results can be increased by loading the notch in compression: this was done by turning the notched bend bar upside-down and loading it three times with a load of 200–220 N (which was three times the maximum fracture load of the notched samples) ( Fig. 2 ). This preloading step in compression induces the formation of a small pre-crack in the notch.
Directly after the preloading step, the CNB test was performed in the usual 4 point bending configuration ( Fig. 3 ). The DIN standard allows a free selection of the cross-head speed lower than 0.05 mm/min, while the fracture should occur within 2–5 min. As will be shown later, the speed limit of 0.05 mm/min will not yield plausible results. It was therefore decided to perform further tests with constant loading rates between 1 and 10 N/s, which resulted in test durations between only 10 and 20 s. Some additional tests were conducted in an inert atmosphere using silicone oil to prevent subcritical crack growth to adversely affect the test results. These tests were performed at a loading rate of 1 N/s and a displacement rate of 0.05 mm/min.
After the test the notch geometry was measured. From the valid tests, the load maxima were determined, and the resulting K IC values were calculated according to the following formula (acc. to DIN EN 14425-3) :
K I , CNB = Y min * [ P max ( S 0 − S i ) B W 3 / 2 ]
Y min * = ( 3.08 + 5.00 α 0 + 8.33 α 0 2 ) ( 1 + 0.007 ( S o S i W ) 1 / 2 ) ( α 1 − α 0 1 − α 0 )
α 0 = a 0 W and α 1 = a 1 W .