1 Introduction

Ceramic materials are widely used in industry due to their high hardness, thermal shock resistance, wear resistance and corrosion resistance [1]. Initial tiny cracks are not easy to detect, but as the cracks expand, they will quickly lead to structural fracture, causing catastrophic accidents. In engineering practice, commonly used non-destructive crack detection technologies include magnetic particle testing, ultrasonic testing, X-ray testing, and penetration testing [2]. They have high sensitivity, but the accuracy can only reach 0.3 to 0.4 mm. Although the penetration method can achieve a detection accuracy of 1 μm, it cannot directly describe the width of the crack. The fiber optic acoustic emission detection technology [3], the potential method [4] and the CCD monitoring of crack expansion [5] used in the laboratory to test metal cracks can achieve an accuracy of micrometers. However, these methods generally have obvious external interference, complex processes and subsequent processing techniques, resulting in unstable test results.

Fluorescent materials are easy to make coatings or platings, and are suitable for the distribution detection of large-area strain fields. In addition, fluorescence is not easily disturbed by environmental factors such as vibration and noise [6]. Compared with traditional fluorescent dyes, semiconductor fluorescent quantum dots have a narrow spectrum and a wide excitation spectrum. Their fluorescence intensity and stability are about 100 times that of ordinary fluorescent dyes, and there is almost no light fading phenomenon [7]. These characteristics meet the requirements of crack detection.

At present, the research on fluorescence crack detection at home and abroad is mainly based on rare earth elements. In 2003, Kim's research group [8] mixed rare earth luminescent elements with ceramics, vividly depicted the crack extension of ceramic materials, and explained the stress fluorescence properties of rare earth luminescent elements. In 2008, Xu's research group [9] combined rare earth luminescent elements with metals to construct a stress distribution image visualization system, which vividly described the distribution of stress in metal materials. Subsequently, Chandra and Xu et al. [10] jointly reported the stress luminescence properties of ZnS∶ Mn. They found that the fluorescence intensity of ZnS∶ Mn film first increased and then weakened by loading a glass substrate coated with ZnS∶ Mn film.

At present, there is no report on the method of using quantum dots to induce fluorescence changes to test crack extension and stress distribution. Although there have been preliminary studies on rare earth elements to detect crack extension and stress distribution, they are limited to a few research groups and the analysis of the mechanism is unclear. This paper uses quantum dots to coat ceramics for fatigue tensile tests to detect changes in fluorescence when cracks appear, and uses ANSYS software to simulate stress distribution to establish a method for quantum dots to detect microcracks in real time.

2 Experiments

The experiment used a standard compact tensile specimen of 304 stainless steel and alumina ceramic stacked on top of each other (as shown in Figure 1), with the thickness ratio of ceramic to stainless steel being 1:2.

The CdS/ZnS quantum dots used in the experiment were homemade in the laboratory [11]. 1 mL of CdS/ZnS quantum dot stock solution (concentration of 22.2 mg/mL) was mixed with 3 mL of acetone and centrifuged three times. After washing, it was dissolved in 1 mL of chloroform for later use. 2 mL of 6002 epoxy resin was mixed with 0.5 mL of curing agent, added to 1 mL of chloroform in which CdS/ZnS quantum dots were dissolved after washing, and stirred evenly. The mixed solution was evenly coated on the surface of the ceramic part of the composite CT sample. After vacuum drying for 5 h, the thickness of the film was measured to be about 0.18 mm.

 The CT specimens were stretched on a high-frequency fatigue testing machine (Changchun Testing Machine Institute, GPS50). The alternating load was 5.24 kN and the average load was 6.4 kN. Sine wave cross-section loading was used and the loading was stopped after obvious cracks were generated.

The experiment used a confocal microscope (Nikon A1Ｒ) to observe and analyze the quantum dot epoxy resin film, and a portable spectrometer (OceanOptic QE65Pro) was used to test the fluorescence spectra of the cracks and non-cracks on the quantum dot epoxy resin film.

3 Results and discussion

3.1 Detection of fluorescence intensity changes using confocal microscopy

The change of sample fluorescence intensity was observed by confocal microscopy. On the side where fatigue load was applied, the crack edge had obvious fluorescence enhancement relative to the non-crack area, while on the side where preload load was applied, i.e. constant load, there was no obvious fluorescence enhancement relative to the non-crack area, as shown in Figure 2. It can be measured that the crack width is about 35 μm. The actual crack width was observed by optical microscopy, as shown in Figure 3, and the actual crack width was about 37 μm, with a difference of 2%.

3.2 Fluorescence spectra of cracks and non-cracks tested by spectrometer

In order to find out the law of fluorescence intensity change, points were taken along the longitudinal and transverse directions of the crack. Two points D and E were taken on the side of the quantum dot epoxy resin film subjected to preload, where point D is the crack edge, and three points C, B, and A were taken on the side subjected to fatigue load, where point C is the crack edge and point B is between points A and C. Five points were tested, and the spectrum of each point is shown in Figure 4. The fluorescence peak of the quantum dots is shown in Table 1. It can be seen that there is no obvious fluorescence enhancement on the side subjected to preload, and the peak values of points D and E differ by 149, while there is obvious fluorescence enhancement on the side subjected to fatigue load, and the peak values between points CB and BA decrease by 3267 and 2496 respectively.

Five points A', B', C', D', and E' were taken along the crack edge on the fatigue load side of the quantum dot film, where point A' was close to the initial crack position and point E' was close to the crack tip. The five points were tested, and the spectrum of each point is shown in Figure 5. The fluorescence peak of the quantum dots is shown in Table 2. It was found that the fluorescence intensity increased regionally from the initial crack position to the crack tip, and the peak values between the five points decreased by 8218, 370, 2806, and 246, respectively.

The results of the experiment show that there is no obvious fluorescence enhancement on the side perpendicular to the crack direction, which is subject to preload, while on the side subject to fatigue load, the fluorescence intensity of the outer side of the specimen toward the crack edge is enhanced, and the fluorescence intensity of the outer edge is equivalent to that of the side subject to preload. Along the crack direction, the fluorescence intensity is regionally enhanced from the initial position of the crack to the crack tip. The fluorescence intensity of the D' and E' regions is close, with an average of 51841, and the fluorescence intensity of the B' and C' regions is also close, with an average of 54955. The average fluorescence intensity of the A' region is 63358. Compared with the B'-C' region and the D'-E' region, the fluorescence regional intensity growth rate of the A' region is 6% and 15.3%, respectively.

4 ANSYS simulation of stress distribution after fatigue cracking

In 2008, Xu found that the fluorescence distribution diagram and stress distribution diagram basically matched when he combined rare earth luminescent elements with metals. In order to investigate the relationship between the fluorescence intensity of quantum dots and the stress magnitude, the stress distribution after fatigue cracking was simulated using ANSYS software. The life and death of a unit in ANSYS means that some units in the analysis process can be set to exist or disappear. Not all unit types support the unit life and death option. If you want to use the unit life and death function, you should select a unit type that supports unit life and death when modeling. The PLANE82 unit used in this paper meets the above requirements [12].

4.1 Pre-treatment

First, the model is built and the material properties are defined, with elastic modulus E = 310 GPa, Poisson's ratio μ = 0.2, and density ρ = 3.97 g/cm3. Free meshing is performed using PLANE182.

4.2 Defining the load step and solving

The “Pressure” option is used to load the left and right holes, and the values are σleft = 24.49 MPa and σright = 21.39sint + 24.49 MPa respectively.

4.3 Solution analysis of crack propagation process

When the loading is determined, the equivalent stress diagram of the specimen is obtained by solving the problem as shown in Figure 6a.

As shown in Figure 6a, the equivalent stress reaches the maximum value at the initial crack tip; therefore, crack extension occurs first from here, and the kill/activate command in ANSYS is used to simulate the fracture of the specimen. In order to determine the next crack extension direction, the solution operation is performed again under the condition of killing the first unit, and then the selection function is used to select the non-activated unit from the selection set. The Von Mises equivalent stress after the specimen fracture is shown in Figure 6b. According to this analysis method, the Von Mises equivalent stress distribution diagram after eighty steps of fracture is obtained as shown in Figure 7.

As shown in Figure 7, the maximum stress at the crack tip reaches 286 MPa, the stress on the side subjected to constant load (left side) is almost zero, and the stress on the side subjected to fatigue load (right side) has a regional distribution, and the stress increases regionally from the initial position of the crack to the tip position.

The average values of the fluorescence intensity in different regions are compared with the average values in the corresponding stress distribution regions. The fitting results are shown in FIG8 . It can be seen that the stress magnitude obtained by fitting is basically linearly related to the actually measured fluorescence intensity.

From the comparison results, it can be seen that perpendicular to the crack direction, the fluorescence intensity and stress on the side subjected to preload are basically unchanged, while the fluorescence intensity and stress on the side subjected to fatigue load show obvious changes, and gradually decrease from the crack edge to the outer edge of the specimen. Along the crack direction, from the initial position to the tip of the crack, the fluorescence intensity and stress both show regional growth, and the change of fluorescence intensity and stress magnitude approximately satisfies the linear change relationship. Further quantitative results need to be verified by more experimental data.

 5 Conclusion

The paper explored the relationship between the crack width detected by fluorescence and the actual crack width observed by microscope, and found that the difference between the two was 2%, which achieved real-time monitoring of microcracks through quantum dots. The relationship between fluorescence intensity and stress when quantum dots detect fatigue cracking of ceramic materials was explored from both experimental and simulation aspects.

(1) After the ceramic cracks under loading, the fluorescence intensity on the side subjected to preload is almost unchanged, while the fluorescence intensity on the side subjected to fatigue load changes significantly, gradually decreasing from the crack edge to the outer edge of the specimen. Along the longitudinal direction of the crack, from the initial position of the crack to the tip, the fluorescence intensity increases regionally;

(2) After the crack is generated, the stress on the side subjected to the preload force has almost no change, while the stress on the side subjected to the fatigue load has obvious changes, gradually decreasing from the crack edge to the outer edge of the specimen. Along the longitudinal direction of the crack, from the initial position of the crack to the tip, the stress increases regionally, which is close to the linear law of the fluorescence intensity value measured in the experiment. The results of this study show that the fluorescence characteristics of quantum dots can be used to detect fatigue cracking of ceramic materials. In practical applications, it is only necessary to irradiate the quantum dot coating with ultraviolet light and determine the location of the crack by whether a bright line is generated. This method is simple and reliable.

references

［1］ Fan Dewei, Zhang Weiru, Liu Juncheng． Research status of thermal conductivity of β-Si3N4 ceramics［J］． Bulletin of the Chinese Ceramic Society，2011，30(5):1105-1109．

［2］ Liu Guimin, Ma Lili． Nondestructive Testing Technology［M］． Beijing: National Defense Industry Press, 2010:12．

［3］ Zhao Jianghai, Liu Xia． Weak current detection based on fiber optic acoustic emission sensor［J］． Electronic Components and Applications, 2006, 8(12):36-38．

［4］ WANG Liang, DING Chuanfu． Application of DC potential method to measure crack length at high temperature［J］． Journal of Aeronautical Materials，2006，26(3):359-360．

［5］ Mo Guoying. Basic research on the application of CCD image recognition technology in fatigue crack detection［D］. Nanjing: Master's degree thesis of Nanjing University of Aeronautics and Astronautics, 2008.

［6］ Yue Junxin, Zhang Weiwei. Stress measurement by fluorescence method［J］． Failure Analysis and Maintenance，2012，7(1):63-68．

［7］ Zou Mingqiang, Yang Rui, Li Jinfeng, et al. Optical characteristics of quantum dots and their applications in life sciences［J］. Journal of Analysis and Measurement, 2005, 24(6):133-137.

［8］ Kim JS，Kwon YN，Sohn KS．Dynamic visualization of crack propagation and bridging stress using the mechano-luminescence of SrAl2O4 :(Eu，Dy，Nd)［J］．Acta Materialia，2003，51:6437-6442

［9］ Li CS，Xu CN．Dynamic visualization of stress distribution on metalbymechanoluminescence images［J］．Journal of Visualization，2008，11(4):329-335．

［10］ Chandra BP，Xu CN，Yamada H，et al．Luminescence induced by elastic deformation of ZnS∶ Mn nanoparticles［J］．Journal of Luminescence，2010，130:442-450．

［11］ Wan Z，Luan WL，Tu SD．Size controlled synthesis of blue emitting core / shell nanocrystals via microreaction［J］．The journal of Physical Chemistry，2011，115(5):1569-1575．

［12］ Zhao Xuecai. Study on the fracture characteristics of Q235 under ultra-low cycle conditions［D］. Lanzhou: Master's degree thesis of Lanzhou University of Technology, 2007.