New Technology for Increasing Through-Life Reliability of Ceramics Components Using Self-Crack-Healing Ability

Structural ceramics are superior to metallic materials in terms of their high-temperature strengths and critical heat proof temperatures. However, compared to metallic materials, ceramics exhibit lower fracture toughness, so they are more sensitive to flaws such as pores and cracks. The shortness considerably decreases the component reliability. To overcome the shortness, in this study, special attention is paid to structural ceramics with self-crack-healing ability.There are several advantages for using a material with self-crack-healing ability. (1) After an efficient machine operation, the materials are able to self-heal the cracks introduced by the machining. (2)Thematerials are able to self-heal the cracks introduced during service and recover the strength completely at healing temperature. However, ways of organizing the available knowledge to increase the through-life reliability of ceramics components have not been extensively studied. The authors propose a new concept and the corresponding flowchart. This new concept is a promising technique for increasing the through-life reliability of ceramics components with excellent self-crack-healing ability.


Introduction
Structural ceramics are superior to metallic materials in terms of their high-temperature strengths and critical heat-proof temperatures. Structural ceramics are leading candidate materials for high-temperature applications such as components of gas turbines. However, compared to metallic materials, ceramics exhibit lower fracture toughness, so they are more sensitive to flaws such as pores and cracks. The reliability of structural ceramics in machinery is therefore limited by the following three problems. (1) Cracks occur during machining (e.g., grinding and polishing), considerably decreasing the component reliability. In order to prevent this, precise polishing is required in the final stage, which is timeconsuming and lowers the fabrication efficiency and raises the fabrication cost. (2) In particular, cracks of depth about 10-30 m affect the reliability. Nondestructive inspection technologies for detecting cracks of depth 10-30 m have yet to be developed. (3) Cracks may occur for various reasons in components during their use at high temperatures.
To overcome the above problems, in this study, special attention is paid to structural ceramics with self-crack-healing ability. There are several advantages to using a material that can heal its own surface cracks. First of all, if the selfhealing of the surface crack is carried out after an efficient machine operation is performed, the fabrication efficiency is greatly improved and the fabrication cost is lowered. Furthermore, the healing of all surface cracks greatly improves reliability. Finally, the healing of a crack that occurs during service is highly advantageous, especially if the self-crack-healing ability allows the material to completely recover its original strength.
In this paper, a new methodology is proposed for guaranteeing the through-life reliability of ceramics components, based on studies by the authors of materials with excellent self-crack-healing ability, proof test theory, and self-crackhealing behavior during service under low oxygen partial pressure ( O 2 ) [1, 2].

Flowchart of Through-Life Reliability
Managements of Ceramics Components with Excellent Self-Crack-Healing Ability organizing the available knowledge to increase the throughlife reliability of ceramics components have not been extensively studied. The authors propose a new concept, and the corresponding flowchart is shown in Figure 1. This new concept is a promising technique for increasing the through-life reliability of ceramics components with excellent self-crackhealing ability.
Heavily Machined Components. This is related to process no.
(1) in Figure 1. After the sintering of ceramics, machining such as grinding is carried out to give the required form. In general, long and deep cracks can be introduced by machining, as shown in Figure 2 [3]. These cracks considerably reduce the strength and reliability of the components. In the case of traditional technology, after the high-performance machining process, final machining processes such as polishing and lapping are performed to remove unacceptable flaws. Although these processes are generally very costly and timeconsuming, they cannot eliminate minute cracks completely. It is anticipated that substituting the crack-healing process for the final machining processes will lead to economical manufacturing of ceramics components with high reliability.
Service Temperature. This is related to process no. (2) in Figure 1.
In Figure 1, is the service temperature, and csch is the critical lower temperature for self-crack-healing. If is higher than csch , the material is able to self-heal cracks introduced during service and completely recover its strength, as shown in Figure 4 [4]. Consequently, the fatigue limit is almost equal to the minimum fracture strength, as shown in Figure 11 [5].
If is lower than csch , the material is not able to self-heal cracks introduced during service. To increase the reliability of ceramics components, compressive residual stress should therefore be introduced by shot peening (SP); this method is very useful, as shown in Figures 13,14,and 15 [6].
Lapping. This is related to process no. (3) in Figure 1. Polishing and lapping are performed as final machining processes. These processes have two purposes: (a) reducing the roughness induced by heavy machining and (b) removing unacceptable flaws. Although these processes are generally very costly, they cannot eliminate minute cracks, as mentioned in Section 4.1.
When is higher than csch , if necessary, slight lapping should be carried out to reduce the roughness. When is lower than csch , heavy lapping should be carried out to eliminate cracks introduced by machining and also to prevent chipping by SP.
Shot Peening. This is related to process no. (4) in Figure 1. SP is a common procedure used to increase the near-surface strengths of metals. In the SP process, a stream of small, hard spheres is shot at a treated surface. Even in the case of ceramics, compressive residual stress is generated by localized plastic deformation or microcrack formation on the surface layer [7,8]. Considering the results obtained in previous Heavily machined ceramics component (1) Lapping (3) No Shot peening (4) Self-crack-healing (5) Proof test (6) Service (7) Yes Service (10) If necessary lapping (3) Self-crack-healing (8) Proof test (9) T s ≥ T csch (2) T s : service temperature T csth : critical lower temperature for self-crack-healing studies, the SP of ceramics is a promising technique for increasing reliability in the surface region.
A high compressive residual stress will increase the fracture strength, fatigue strength, apparent fracture toughness, and contact strength of ceramics components such as bearings and turbine blades, in which higher reliability is desirable.
Self-Crack-Healing. This is related to processes nos. (5) and (8) in Figure 1. This process should be carried out under optimum conditions. In the case of process no. (5), if surface cracks are unintentionally induced or grown during the machining and SP processes, the reliability of the ceramics components will be decreased. To prevent this, the authors propose the use of a combination of SP and crack healing [6]. As a result of this combination, the surface strength and reliability of the ceramic will be improved simultaneously.
In the case of process no. (8), there is a high risk of introducing long and deep cracks by heavy machining, so a selfcrack-healing process is necessary. By this process, surface cracks introduced by machining can be healed completely, as shown in Section 4.3.
Proof Test. This is related to processes nos. (6) and (9) in Figure 1. Oxygen is necessary for crack healing, and thus embedded flaws such as pores and abnormally large grains cannot be healed. These flaws may reduce the reliability of ceramics components. It is necessary to conduct a proof test  to prevent the use of a low-reliability component that contains unacceptable flaws. A proof test is therefore important for achieving higher reliability. Using the proof test theory [9], the minimum fracture stress at any temperature can be evaluated easily and with high accuracy, as shown in Section 6.
Service. This is related to processes nos. (7) and (10) in Figure 1. In the case of process no. (7), is lower than csch thus the materials are unable to self-heal the cracks introduced during service. Consequently, periodic heating to a higher temperature than csch is a useful method of maintaining higher-level reliability of ceramics components. In contrast, in the case of process no. (10), is higher than csch , thus the materials are able to self-heal cracks introduced during service. Therefore, a higher level of reliability can be maintained by the self-crack-healing ability.

Ceramics Materials for Service
3.1. Nanocomposite Materials. The nanocomposite materials developed by Ando's group are described as follows [1,2]. (a) SiC nanoparticles are added to the matrix, preventing grain growth in the matrix during the sintering process and improving its strength through grain refinement. For example, in the case of alumina, the bending strength can be increased from about 400 MPa to 700-1000 MPa by the inclusion of SiC nanoparticles [10]. (b) Composites containing 15-30 vol% nanosized SiC particles exhibit self-crack-healing ability. (c) These SiC particles are distributed mainly in the grain boundaries. In the case of alumina, some grains are located in grains and increase the heat-proof temperature by about 300 K. The developed ceramics materials are Al 2 O 3 / SiC, mullite/SiC, Si 3 N 4 /SiC, SiC, ZrO 2 /SiC, and AlN/SiC [1,2].

Multicomposite Materials.
A multicomposite material is a composite containing SiC nanoparticles and SiC whiskers at a total volume fraction of 25-30%. These materials have excellent self-crack-healing ability, fracture toughness, and hightemperature strength [11]. The nanosized SiC particles mainly contribute to increasing the strength and self-crack-healing rate, and the SiC whiskers mainly contribute to increasing fracture toughness and high-temperature strength.
The proportion of SiC particles to SiC whiskers is determined by taking into consideration the self-crack-healing ability and fracture toughness requirements. The developed ceramics materials are Al 2 O 3 /SiC and mullite/SiC [1,2].
The self-crack-healing materials described in this paper have the following three attributes. (1) The material itself detects cracks in order to begin the crack-healing process. (2) A crack that reduces the strength of the material by 50-90% can be completely healed. (3) The strength of a crack-healed area is equivalent to or higher than that of a matrix area up to about 1673 K.

Mechanism of Crack
Healing. The crack healing of the nanocomposite and multicomposite ceramics described in this paper is caused by the following oxidation reaction of SiC: From (1), it can easily be understood that a monolithic mullite or a monolithic alumina does not have a crack-healing ability, since oxidation is saturated in these materials. A schematic diagram of the crack-healing mechanism is shown in Figure 3. To achieve complete strength recovery by selfcrack-healing, the following three conditions are necessary. (A) The healing substance has to fill the crack completely. This condition is achieved by an increase in volume of about 80% when SiC is converted to SiO 2 . (B) The healing substance must be completely welded to the base material. This condition is satisfied by the exothermic oxidation reaction, which generates a large heat release of 943 kJ per mole of SiC. The heat melts the matrix and the healing substance to form the healed mixture. (C) The strength of the crack-healed zone has to exhibit equivalent or higher strength than that of the base material up to about 1673 K. There is a glass phase and a crystalline phase in SiO 2 in (1). If the crack-healing material is crystalline SiO 2 , then the crack-healed sample exhibits high bending strength, even at an elevated temperature. However, if the crack-healing material is glassy SiO 2 , then the crackhealed sample exhibits low bending strength at an elevated temperature. A key point in crack-healing technology is how many crystalline phases are deposited in a crack-healing substance.
Three-point bending test specimens were used to evaluate the bending strength ( ) of a crack-healed specimen. One crack was introduced into the central part of the test specimen. The crack was semielliptical in shape, with a surface length of 100 m and a depth of 45 m (hereafter called a standard crack). The authors used a shorter bending span of 16 mm than that typically used of 30 mm in order to evaluate the strengths of the crack-healed zones and to reduce the elastic energy at the fracture and prevent the specimen from breaking into small pieces.
Self-crack-healing behaviors are dependent on the material, temperature, oxygen partial pressure, and crack size (mainly depth). Figure 4 shows the effect of the atmosphere on self-crack-healing behavior [4]. The of an as-received test specimen was about 650 MPa. When a standard crack was introduced into the specimen, the was reduced to 180 MPa, that is, a decrease of about 73%. However, when the healing of a precracked specimen was carried out in air at 1573 K for 1 h, the improved to about 800 MPa. The of a crack-healed specimen is larger than the of an as-received specimen. This is because even a smooth specimen may have minute cracks on the surface that reduce its bending strength, compared with a crack-healed specimen, in which even preexisting cracks are healed completely. However, when the precracked specimens were heated in vacuum and under N 2 and argon gas atmospheres, the value recovered to a maximum of 350 MPa. Under these conditions, the recovery is insufficient. The slight increase in achieved using this heat treatment is the result of removal of the tensile residual stress of the crack tip. Similar crack-healing and strength-recovery behaviors were reported for mullite [4], Al 2 O 3 [12], and Si 3 N 4 [13,14]. Figure 5 shows specimens after the bending tests, with Figure 5(a) showing a precracked specimen [4]. The precracked specimens were fractured by cracks introduced using the Vickers indentation method. In the case of Figure 5(b), because the crack was healed completely, the fracture was initiated from outside the crack-healed zone.   The influence of temperature on crack-healing behavior is shown in Figure 6 [2]. When the crack-healing times were 10 h and 300 h, the temperatures corresponding to complete healing were about 1473 K and 1273 K, respectively. The systematic investigations have been made on various materials to evaluate the shortest time for complete healing, , at a certain temperature . The relationship between and can be expressed by the Arrhenius equation as follows [15]: The activation energy ( ) and proportional modulus ( 0 ) of each material were calculated, and the results are shown in Table 1 [15]. The data in Table 1 can be used to determine the relationship between and the crack-healing temperature for a standard crack. However, (2) is applicable for a standard crack with = 1-300 h and under ambient air. For a crack that is larger than the standard size and an oxygen partial pressure that is lower than that of air, the necessary time for complete crack healing is longer.

High-Temperature Strength of Crack-Healed Specimen.
The temperature dependence on the of crack-healed specimens is shown in Figure 7 [1, 2]. Each test specimen was Journal of Powder Technology  crack-healed under optimum conditions after a standard crack was introduced. The critical heat-proof temperature for the bending strength of Al 2 O 3 /SiC is about 300 K higher than that of monolithic Al 2 O 3 ; this impressive increase in the critical heat-proof temperature is caused by the inclusion of SiC nanoparticles, as mentioned in Section 2. The critical heat-proof temperature at the crack-healed zone of a commercial SiC composite is 873 K, which is much lower than that of the matrix [16]. As shown in the figure, a newly developed SiC composite exhibits an excellent critical heat-proof temperature of about 1673 K in the crack-healed zone, and we examined its use [17]. The critical heat-proof temperature of a silicon nitride composite, with Al 2 O 3 added to it, was about 1273 K. The critical heat-proof temperature of the new silicon nitride, which has an improved composition system, is about 1673 K, which is higher than the critical heatproof temperature of the silicon nitride with added Al 2 O 3 . The large improvement in the critical heat-proof temperature was attained by the crystallization of a grain boundary phase  and a crack-healing substance. It should also be noted that the various types of composites, except the commercial SiC composite, fractured from a site other than the crack-healed zone within the temperature range below the critical heatproof temperature. This implied that the crack-healing zone had sufficient bending strength even at high temperatures.

Crack-Healing Behavior of Heavily Machined Component.
The crack-healing behaviors of heavily machined specimens have been investigated by several researchers [3,18,19]. The experimental results showing the effects of crack healing in increasing the reliability of machined components have been explained [3]. Sintered plates of Al 2 O 3 /20 vol% SiC whiskers were cut into 3 mm × 4 mm × 22 mm rectangular bar specimens. A semicircular groove was made at the center of the smooth specimens using a diamond-coated ball drill, as shown in Figure 8. The cut depth from one pass ( ) during machining of the semicircular groove was = 5-15 m. The total cut depth was 0.5 mm. The fracture tests were conducted at room temperature. The local fracture stresses ( ) related to the stress concentration factor of the semicircular groove ( = 1.2) were evaluated. Figure 8 shows the strength of the machined specimens with a semicircular groove as a function of the cut depth by one pass ( ). The open triangle symbols show the local fracture stresses ( ) of the asmachined specimen. The of the as-machined specimens decreased with increasing . This behavior was caused by the fact that the crack depth caused by machining increased with increasing . Figure 2 shows scanning electron microscope images of the fracture surface of the as-machined specimen. The dashed lines in Figure 2 indicate the crack front. As shown in this figure, these surface cracks acted as fracture initiators. Thus, the values of the as-machined specimens decreased compared with those of the healed smooth specimens. The open square symbols in Figure 8 show the of the machined specimen healed at 1673 K for 10 h in air. The of these specimens increased significantly as a result of crack healing. The values of the machined specimens were almost equal to the fracture stress of the healed smooth specimens (solid circles). Thus, the surface cracks introduced by the heavy machining process were healed. It was therefore concluded that crack healing could be an effective method for improving the structural integrity of machined ceramics with self-crack-healing abilities and reducing machining costs.

Crack-Healing Behavior during Service
Ceramics components are often exposed to continuous constant or cyclic loading at elevated temperatures under low oxygen partial pressures. If a crack is initiated during service, the component's reliability will be considerably reduced. If the crack can be healed under service conditions and the components recover their strength completely, the reliability and lifetime of ceramics components can be increased. Crackhealing behavior under constant or cyclic stress has been systematically studied for Si 3 N 4 /SiC [5,20], mullite/SiC [21,22], Al 2 O 3 /SiC [23], and SiC [17]. Standard precracks were healed, and the samples completely recovered their strength at the healing temperature, even under constant or cyclic stress.
To take advantage of a material's crack-healing ability during service, it is essential to determine its threshold stress for crack healing, under which a crack can be completely healed. In this section, the critical stress conditions for crack healing in Si 3 N 4 /SiC and other ceramics are shown.

Crack-Healing Behavior of Si 3 N 4 /SiC under Stress and
Low Oxygen Pressure. The material studied was hot-pressed Si 3 N 4 , which contained 20 wt% SiC particles and 8 wt% Y 2 O 3 as a sintering additive [24]. A semicircular surface crack of surface length 100 m (standard crack) was introduced at the  center of the tensile surface of the bending test specimens. After introducing the precracks, crack-healing tests under cyclic or constant stress were carried out. Figure 9 shows the crack-healing process [24].
Crack healing was carried out under cyclic ( max,ap ) or constant ( ap ) bending stresses at the healing temperature ing process, the bending strengths of the specimens were measured at room temperature or at the healing temperature in air. Figure 10 shows the results of the bending tests on specimens crack-healed at 1473 K for 5 h under constant stress with O 2 = 500 Pa [24]. The solid triangles represent the bending strengths of the crack-healed specimens. Open triangles indicate the bending strengths of the precracked specimens. The asterisks indicate specimens that fractured outside the crack-healed zone, suggesting that the precracks were healed completely. The threshold stress for crack healing was defined as the maximum stress below which the crackhealed specimens recovered their bending strength, and below which most specimens fractured outside the crackhealed zone.
From Figure 10, the threshold stress for crack healing in air was determined to be 200 MPa. A similar crack-healing test under cyclic stress was carried out using Si 3 N 4 /SiC, Al 2 O 3 /SiC(p), Al 2 O 3 /SiC(w), and mullite/SiC, and the threshold stress was found to be 100-300 MPa. All of the materials exhibited higher threshold stress under cyclic testing than under constant stress.
The above test results indicate that the Si 3 N 4 /SiC composite is able to recover its strength completely, even under service conditions, that is, with an applied stress below ap = 200 MPa and an oxygen partial pressure above 500 Pa.
As to the effect of O 2 on the threshold stress, if O 2 is higher than 500 Pa, the threshold stress is independent of O 2 . However, below O 2 = 50 Pa, the threshold stress has a slightly lower value. The specimens crack-healed under a constant stress of 200 MPa showed comparable bending strengths to those of specimens crack-healed under no stress.
The crack-healing behavior under stress is shown in Figure 9 [5]. The changes in the crack length during crack healing are divided into three cases. In case (I), slow crack growth from the tip of the precrack occurs, and the cracks lead to specimen failure because the crack-growth rate is higher than the crack-healing rate. In case (II), slow crack growth from the tip of the precrack also occurs. The crackhealing process starts at high temperature, causing a decrease in the crack-growth rate, as a result of crack healing, and finally the cracks are healed. In case (III), slow crack growth from the precrack does not occur, and the precracks are easily healed completely.
The threshold stresses (constant or cyclic) for crack healing as a function of the bending strengths of the precracked specimens were tested systematically using the above mentioned materials [11]. All the data have proportional relationships, in spite of differences in crack-healing conditions. For precracked specimens under constant stress, the threshold stress is 64% of the bending strength, and for precracked specimens under cyclic stress, the threshold stress is 76% of the bending strength.

Strength of Crack-Healed Sample at Healed Temperature and Fatigue
Limit. The test material was Si 3 N 4 /SiC(p), and a standard crack was introduced in the sample. The crack was healed at a temperature between 1173 K and 1473 K under cyclic stress [5]. The cyclic stress conditions during crack healing were as follows: maximum stress max = 210 MPa, stress ratio = 0.2, and frequency = 5 Hz. The bending strength ( ) and the cyclic fatigue limit ( 0 ) at the healing temperature are shown in Figure 11 by solid circles and open squares, respectively. Furthermore, the bending strength of a run out specimen in a fatigue test at the healing temperature is shown in Figure 11 by solid triangles. An asterisk indicates that fracture occurred from outside the crack-healed zone. Similar tests have been made above mentioned Al 2 O 3 /SiC, mullite/SiC, Si 3 N 4 /SiC, and SiC. From Figure 11 and these tests results, the following three conclusions can be obtained.
(1) The strength of the crack-healed zone was the same as or higher than that of the base material, even at the crackhealed temperature, because many samples fractured from base material. (2) The fatigue limit is almost equal to the bending strength. This indicates that cracks initiated during the fatigue test, are healed during the fatigue test and the material recovers its strength completely. (3) The properties described in (1) and (2) are very useful for the through-life reliability of ceramics components.

Proof Test Theory
As mentioned above, oxygen is necessary for crack healing, and thus embedded flaws such as pores and abnormally large grains cannot be healed.
So, for higher reliability, a proof test is necessary. Generally, a proof test is conducted at room temperature, for economic reasons. If the proof-tested components are used at where 0 and are the strength of the plain specimen (intrinsic bending strength) and the proof test stress at room temperature, respectively, and is the plane strain fracture toughness. In (3), the superscripts and to , 0 , and indicate the value at room temperature and the value at elevated temperature , respectively.
The proof test theory was applied to Si 3 N 4 [9], Al 2 O 3 / SiC(p) [25], and a coiled spring made of Si 3 N 4 [26]. The test results are shown in Figure 12; the measured minimum fracture stresses ( min ) are plotted as a function of the evaluated .
in the figure denotes the number of samples used to obtain min , and is 9-15. The standard is 15.
The four solid diamonds indicate the data on the ceramic coiled spring made of silicon nitride. All the min values showed good agreement with , regardless of the test temperature. From this figure, it can be concluded that (3) can be applied to Al 2 O 3 /SiC, Si 3 N 4 , and even to a coiled spring.

Shot Peening
SP was applied to Si 3 N 4 /SiC with excellent self-crack-healing ability. A maximum compressive residual stress of 880 MPa was observed at the surface. The compressive residual stress decreased in the depth direction. The compressive residual stress was induced up to a depth of approximately 40 m. Figure 13 shows the residual stress at the surface as a function of the heat-treatment temperature [6]. The compressive residual stress decreased with increasing heat-treatment temperature. However, the specimen heat treated at 800-1100 ∘ C showed compressive residual stresses of 300-400 MPa. In Si 3 N 4 /SiC, crack healing occurred above 800 ∘ C.
The effect of SP on the apparent fracture toughness was evaluated using the indentation fracture (IF) method. The average value of the fracture toughness ( ) for non-SP specimens, measured using the standard test method, was 4.9 MP am 1/2 . Thus, the values of the apparent fracture toughness ( ) of the non-SP specimens, measured using the IF method ( = 4.1 MP am 1/2 ), were close to the of the material. The of the SP specimens was about 10 MP am 1/2 , as shown in Figure 14. The increased 2.3-fold compared with that of the non-SP specimen. Figure 14 shows the relationship between the heat-treatment temperature and [6]. After heat treatment, the for the SP specimens decreased. However, if the heat-treatment temperature was 800-1100 ∘ C, at which crack healing is possible, the after heat treatment was more than twice as large.
The effect of SP on the contact strength was also investigated systematically [6]. The critical load, , for crack initiation of non-SP specimens was 0.5 kN, whereas the for SP specimens was more than 4.5 kN, which was the maximum load of the testing system. Thus, the increased more than ninefold as a result of SP. The for the SP + heat-treated specimens decreased with increasing heat-treatment temperature. The tendency of to decrease with increasing heattreatment temperature was similar to that of . This is because both and are closely related to the compressive residual stress. The author noted that the values of the SP + heat-treated specimens were four times higher than those of the non-SP specimens if the heat-treatment temperature was 1100 ∘ C, at which crack healing was possible. From the above test results, it can be concluded that the combination of SP and self-crack-healing is a useful technology for improving the surface strength and reliability of ceramics components.
For measurement of bending strength, non-SP and SP specimens of Si 3 N 4 /SiC were prepared [6]. A Vickers indentation was made on the center of the polished face of the non-SP and SP specimens at loads of 19.6 to 98.0 N to simulate the surface cracks introduced by shot peening or other contact events. These specimens are called "non-SP + cracked"  Figure 12: Relationship between measured minimum fracture stress of proof-tested sample and minimum fracture stress evaluated using (2). Heat-treated temperature ( ∘ C) Heat-treated condition for 1 h in air Figure 13: Compressive residual stress at surface as a function of heat-treatment temperature. Compressive residual stress was induced by shot peening at room temperature. and "SP + cracked" specimens, respectively. The specimens were subsequently tested in three-point bending at a crosshead speed of 0.5 mm/min, using a fixture with a span of 16 mm. Figure 15 shows the bending strength as a function of Vickers indentation load. The bending strength of non-SP + cracked specimens (solid triangles) decreased steeply with increasing Vickers indentation loads, although the decrease rates of the bending strength of SP + cracked specimens (solid inverse triangles) were smaller than that of non-SP + cracked specimens. The main reason for this is that the compressive residual stress restrained initiation and propagation of crack in SP + cracked specimens. These results indicate that the shot peening increased the strength at the subsurface of the test material.

Conclusion
Structural ceramics are generally brittle, which seriously affects the reliability of ceramics components. Five methods are typically used to overcome this weakness: (a) fiber reinforcement or microstructural control to toughen the ceramic, (b) high-level nondestructive inspection and repair of dangerous flaws, (c) proof testing to select components with high reliability, (d) using materials with excellent self-crack-healing ability, and (e) introducing compressive residual stress on   the surface layer. It is impossible to increase the through-life reliability of ceramics components using only one of the above technologies. Thus, it is important to combine these technologies. There are very few studies on how best to combine these technologies to increase the through-life reliability of ceramics components. In this paper, special attention was paid to methods (a), (c), (d), and (e), and a new methodology was proposed, as shown in Figure 1. Moreover, each process was explained in detail.