Evaluation of corrosion resistance of 316L stainless steel by laser ultrasonic nondestructive testing technology

At present, the corrosion resistance of materials such as 316L stainless steel is mainly tested by electrochemical testing, which is time-consuming, laborious and environmentally unfriendly. In this paper, the laser ultrasonic detection technology, which is a rapid, non-destructive, pollution-free microstructure and corrosion resistance evaluation method is used for the detection of 316L stainless steel, and the ultrasonic signals are filtered by the wavelet threshold denoising method. The results show that the laser ultrasonic attenuation coefficient obtained by the wavelet threshold denoising signal processing method has a good correlation with the corrosion resistance of 316L stainless steel. A laser ultrasonic potential evaluation model was established based on the passivation film rupture potential and ultrasonic attenuation coefficient. The average relative error of the evaluation model for corrosion resistance of 316L stainless steel is less than 5.00%. The relative error of corrosion resistance evaluated by the laser-receiving method is lower than that by the probe-receiving method. It is feasible to nondestructive evaluation of the corrosion resistance of 316L stainless steel by laser ultrasonic technology.


Introduction
316L stainless steel is widely used in nuclear power, aviation and other fields due to its excellent corrosion resistance and mechanical properties in high temperature and corrosive environments [1][2][3].Failures due to intergranular corrosion and intergranular stress corrosion cracking in structures and components of austenitic stainless steel materials have always been a tough problem in engineering practice.Electrochemical tests [4], ultrasonic tests [5] and surface morphology observation [6] are commonly used to evaluate corrosion resistance, but the experimental process is complicated, time-consuming and destructive.
Ultrasonic detection has been widely used in the various performance detecting of stainless steel.Silva [7] observed that the change in sound velocity measured value is proportional to the change in material hardness measured value, and sound velocity can be used to characterize the embrittlement of duplex stainless steel.Mutlu [8] found that corrosion products would reduce ultrasonic velocity and increase ultrasonic attenuation, and Attenuation measurements are more sensitive than velocity measurements in characterizing corrosion products.Ultrasonic testing can be used for large-area nondestructive testing, but in harsh environments, such as high temperature, high radioactivity or narrow Spaces, conventional ultrasonic testing techniques are difficult to adapt [9].In addition, it also has the disadvantages of contact type and needs a coupling agent.
With the improvement of product performance requirements for industrial production, a non-destructive and non-contact detection technology is urgently needed [10][11][12].Laser ultrasonic detection has the advantages of a wide frequency band, is non-contact, and can be used in harsh environments [13,14].It can realize the complete inspection of the workpiece under the condition of time permits, which is a promising detection technology.If the non-contact detection of the corrosion resistance of 316L stainless steel is carried out by laser ultrasonic detection technology, it can not only save a lot of sample preparation time, but also is not destructive to 316L stainless steel.It can effectively improve the detection efficiency, realize online rapid detection, and predict the corrosion resistance of 316L stainless steel.
Grain size is an important parameter with a great effect on such properties of metal as yield strength, plasticity, toughness, fatigue strength, creep strength and corrosion resistance [15].Many researchers have used laser ultrasound to characterize the grain size of metal materials or studied the corrosion resistance of metal materials through grain size.Yin [16] used the analytical method of scattering theory to calculate the absolute value of the average grain size directly from the ultrasonic trajectory.Bai [17] found that there was an approximately linear relation between the centroid frequency ratio of the two successive echoes and the mean grain size was observed.Xu [18] found that the ultrasonic velocity seemed to be insensitive to the size of MTC.There is an approximate linear relationship between the frequency downshift of the spectrum centroid and the average size of the MTC.Yun [19] found that with the decrease in grain size, the corrosion resistance of 316L stainless steel was enhanced with the increase of passivation in 0.05M H 2 SO 4 solution.The results show that laser ultrasonic nondestructive testing is very sensitive to the average grain size of polycrystalline materials.However, few researchers have directly evaluated the corrosion resistance of metal materials using laser ultrasonic detection technology.Therefore, based on the effect of average grain size on the corrosion resistance of the material, the corrosion resistance of 316L stainless steel can be directly evaluated by laser ultrasonic detection.
In this paper, 316L stainless steel is used as the experimental material, and the Q-switched Nd:YAG nanosecond pulse laser is used to excite the ultrasonic wave.Ultrasonic signals in different 316L stainless steel samples are received by interferometer and ultrasonic probe.Then, the ultrasonic signals are filtered by the wavelet threshold denoising method to obtain attenuation coefficient characteristic values of ultrasonic signals of different 316L stainless steel samples.The characteristic value of the attenuation coefficient is coupled with the microstructure detected by SEM and the polarization curve obtained by the electrochemical method.The microstructure of each sample was measured by scanning electron microscope (SEM).The electrochemical potential polarization curves were measured by the electrochemical test.To establish a laser ultrasonic potential evaluation model and propose a new laser ultrasonic test method for the corrosion resistance of 316L stainless steel.

Experimental materials
The experimental material is a 316L sheet with a thickness of 2 mm and the chemical composition is shown in table 1.
In order to obtain 316L stainless steel samples with different corrosion resistance, 14 samples were heat treated, and the heat treatment process was adopted as shown in table 2. The thickness of the sample after heat treatment was measured by a spiral micrometer.The sample thickness is listed in table 2. Then the samples were cut from the sheets and then mechanically polished, and etched for 45s with a corrosive reagent (HNO 3 : HCl: H 2 O = 1: 1: 1), the grain size was measured by SU5000 (SEM Hitachi, Tokyo, Japan).

Laser ultrasonic testing system
The laser ultrasonic detection system is shown in figure 1.The pulse laser used in the test is a YAG Laser Q-switched neodymium-doped yttrium aluminum garnet Laser (Q-Switched Nd:YAG Laser, Continuum, San Jose, CA, USA).The pulse width of the laser is 12 ns, the wavelength is 1064 nm, and the pulse energy is 200mj.
Because the attenuation of ultrasound is related to its frequency, in order to improve the accuracy, the evaluation shows that two receiving systems are used to receive ultrasonic signals in this experiment.One is the TWM-532 interferometer [figure 1(a)] and the other is the 10 MHz ultrasonic probe [figure 1(b)].In order to eliminate the ultrasonic signal acquisition error, each sample was measured 9 times, and each signal averaged 128 times.

Electrochemical test
The 316L stainless steel samples were subjected to standardized electrochemical corrosion tests using a CorrTest 350H electrochemical workstation to check for trends in corrosion resistance.A copper wire was connected to the test piece by soldering and then filled with an insulating epoxy resin.A test area of 1 cm 2 was exposed and polished with 2000 grit water sandpaper, then polished with 1 μm adamantine plaster, and finally polished with a colloidal silica mirror.The sample was then placed into a 3.

Microstructure test results
The typical microstructure of the sample obtained by SEM is shown in figure 2, and the grain distribution of the sample is shown in figure 3.
To evaluate the grain-size distribution effects on the ultrasonic response, there are a series of studies [20,21] that have shown that a log-normal distribution function closely represents realistic polycrystalline microstructures.Fit the grain size distribution data by using the Levenberg-Marquardt nonlinear least squares algorithm.The probability density function for a log-normal grain size distribution reads as: Where P(D) is the probability density function of grain size distribution, D is grain size, μ is the mean of grain size distribution, and σ d is the standard deviation of grain size distribution.Turner et al [15] combined with the lognormal distribution of particle size, gave an analytical formula for ultrasonic attenuation evaluation.The granularity D σ estimated by the Turner fitting model is defined as: Dates for each sample in the SEM investigation are presented in table 3, including the mean grain size D calculated by the cut-line method, the distribution width of grain sizes and the number of grains, σ d is the standard deviation of grain size distribution.Cr is the main determinant element of the corrosion resistance of stainless steel.The higher the Cr concentration in the passivation film formed on the surface of the stainless steel, the better the corrosion resistance [22].The mean grain size has a great influence on the density of the grain boundaries.When the density of the grain boundary is relatively high, it can provide more paths for the diffusion of Cr, so that the Cr element is enriched in the passivation film [23].Therefore, if the surface of 316L stainless steel undergoes a stable passivation process in the solution, refining the grains can improve the corrosion resistance.At the same time, the high-density grain boundary will also become a rapid diffusion channel for the corrosive medium [24].Therefore, when the surface of stainless steel is actively dissolved in the solution, the grain refinement will accelerate the corrosion rate and reduce the corrosion resistance of the material.The electrochemical environment is a non-passivating environment.It can be found from figure 4 that the intergranular corrosion resistance of the 316L stainless steel sample increases with the increase of the mean grain size.
The solid solution temperature of all samples is higher than 1050 °C, which will promote the precipitation in the original plate to melt back into the substrate, and the air-cooling method avoids chromium carbide.Therefore, it can be considered that the mean grain size of the sample has a major influence on corrosion resistance.As shown in figure 5, the sample with a mean grain size of 34.66 μm has the worst corrosion resistance.When the mean grain size of the sample is greater than the value, as the mean grain size continues to increase, the corrosion resistance of the sample increases rapidly.

Evaluation of corrosion resistance by laser ultrasonics 4.1. Laser ultrasonic signal analysis and attenuation feature extraction
In the process of ultrasonic wave propagation, it will encounter the interface composed of different acoustic impedance media.Ultrasonic waves are scattered and reflected, and the sound energy is dispersed.Therefore, ultrasonic amplitude attenuation can be used to evaluate the average grain size of the material.There is a lot of high-frequency noise in the ultrasonic signal obtained by ultrasonic detection.The ultrasonic signal is processed by smoothing and wavelet threshold denoising.The ultrasonic signal received by the laser and probe is shown in figure 6.
The baseline of the ultrasonic signal is the fitted curve of the area without any waveform.This is due to the thermal strain caused by the local temperature rise of the sample, the low-frequency vibration of the sample and the low-frequency vibration of the experimental platform.The misalignment of the ultrasonic signal baseline will affect the calculation accuracy of the attenuation coefficient.
As can be seen from figure 6, the baseline misalignment of the signal is controlled by smoothing denoising.So we use the signal after smoothing and denoising to calculate the coefficient of attenuation in laser ultrasonic detection.
The frequency of the signal received by the probe and laser is 10 MHz and 2 ∼ 25 MHz, and the central frequency of the laser reception signal is around 11 MHz, so the main operative mechanism is the Rayleigh mechanism in the samples.The attenuation coefficient calculation method adopts the amplitude method.Select the amplitude of the primary and secondary echoes of the ultrasonic signal in figure 6 to calculate the attenuation coefficient of the ultrasonic.Substituting the extracted amplitude information into the formula, the attenuation  coefficient of each sample can be calculated: In the formula, α is the ultrasonic attenuation coefficient, L is the thickness of the sample, and A is the amplitude of ultrasonic.The calculation results of the signal before and after wavelet threshold denoising of each sample are shown in table 5.

316L corrosion evaluation by laser ultrasonic
Ultrasonic testing is mainly used for nondestructive testing of materials and is one of the main methods of internal testing.The attenuation of ultrasonic waves is highly sensitive to the mean grain size of polycrystalline materials, and the mean grain size of 316L stainless steel has a great influence on corrosion resistance.Therefore, laser ultrasonic testing can be used for the quality inspection of 316L plate products and to evaluate the corrosion resistance of the product.The relationship between the passivation film rupture potential, the attenuation coefficients of ultrasonic signals of the two receiving methods and the average grain size of the sample is shown in figure 7.
It can be found from figure 7 that the trend of the potential laser ultrasonic evaluation model was similar to figure 5.The potential calculates model of laser ultrasonic is shown in (5) (laser reception) and (6) (probe reception).Where p is the potential calculated by the laser ultrasonic model and α is the ultrasonic attenuation coefficient.The ultrasonic attenuation coefficient was substituted into the model to evaluate the potential.The measured values of experimental polarization curves are used as the true values to calculate the relative errors of laser ultrasonic evaluation results.The potential calculated by experimental polarization curve and laser ultrasonic detection were shown in figure 8.
The potential was calculated by the established laser ultrasonic potential evaluation model of 316L stainless steel.The evaluation error of the signals received by the probe receiving method was 0.37%-9.59%,and the average evaluation error was 4.71%.The evaluation error of signals received by the laser interferometer is 0.26%-8.72%,and the average evaluation error is 3.32%.It can be found from figure 8 that the use of laser ultrasonic nondestructive testing technology to evaluate the corrosion resistance of 316L stainless steel is feasible.However, due to the many factors affecting the corrosion resistance of 316L stainless steel, as well as the influence of ultrasonic testing errors and other factors, the establishment error of the relationship is larger.How to establish the relationship between other corrosion factors and ultrasonic attenuation, then improved the evaluation accuracy by combining the influence of mean grain size influencing the corrosion resistance on the ultrasonic attenuation is important.

Discussion
In this section, a smooth denoising method is used to control the baseline misalignment of signals.The highfrequency noise and low-frequency vibration generated in the detection process are effectively filtered by the wavelet threshold denoising method.It can effectively improve the signal-to-noise ratio of ultrasonic signals.
The laser ultrasonic potential evaluation model was constructed, and the rupture potential of the passivation film was evaluated with high accuracy.Therefore, the corrosion resistance of 316L can be evaluated by laser ultrasonic technology testing technology.However, there are many other factors affecting the corrosion resistance of 316L besides the grain size, so it is still necessary to construct a more accurate corrosion resistance prediction model combined with other factors.

Conclusion
In this paper, scanning electron microscopy, laser ultrasonic and electrochemical tests were carried out on 316L stainless steel samples with different average grain sizes and distributions.And the conclusions are as follows:  (1) The corrosion resistance of 316L stainless steel has a good correlation with the ultrasonic attenuation coefficient, and the corrosion resistance evaluation model of 316L stainless steel is established by the passivation film rupture potential and ultrasonic attenuation coefficient.
(2) The average error of the laser ultrasonic potential evaluation model for the rupture potential of 316L stainless steel passivated film is lower than 5.00%.The accuracy of the ultrasonic signal received by laser is better than that received by the ultrasonic probe in the corrosion resistance evaluation of 316L stainless steel.
(3) It is feasible to evaluate the corrosion resistance of 316L stainless steel by using laser ultrasonic nondestructive testing technology.Compared with electrochemical testing and evaluation methods, laser ultrasonic testing technology can realize rapid nondestructive testing.
The next step is to take the other factors which have a great influence on the corrosion behavior of materials and the attenuation of ultrasonic into account, to impart more practical value to the ultrasonic evaluation model.
5 wt% NaCl solution.Platinum and saturated calomel electrodes are used as opposite electrodes and reference electrodes respectively.A thermostatic water bath was used to control the temperature of 25 °C during the experiment.The reference potential of the electrochemical alternating current impedance test (EIS) was used as the open circuit potential.The scan rate of the potentiometric polarization test was 2 mV s −1 .The initial voltage of the test was set as the relative open circuit potential −0.5 V, and the termination voltage was set as the relative open circuit potential 1 V.

Figure 1 .
Figure 1.Laser ultrasonic testing system and the signal of samples: (a) Signal received by the laser; (b) signal received by the probe.

Figure 3 .
Figure 3. Grain size distribution of samples.

Figure 5 .
Figure 5.The influence of the mean grain size of the sample on the resistance to intergranular corrosion.

Figure 7 .
Figure 7. Laser ultrasonic evaluation corrosion resistance: (a) signal received by the laser; (b) signal received by the probe.

Figure 8 .
Figure 8.Comparison of potential evaluation results and error using different signal received methods.

Table 2 .
Heating temperature and holding time and the thickness of samples.

Table 3 .
The mean grain size of the sample was calculated by the Turner model.Electrochemical test results and analysis The electrochemical potential polarization curves of samples are shown in figure4.It can be found that the potentiodynamic polarization curves of different samples are not the same.Except for samples No.6, 8 and 12, the other samples have obvious passivation effects, The calculation results of each sample are shown in table4.

Table 4 .
The mean grain size and Corrosion resistance of samples.

Table 5 .
The results of the laser ultrasonic test.