NDE Characterization of Surface Defects on Piston Rods in Shock Absorbers Using Rayleigh Waves

Abstract

In general, shock absorbers are components that can absorb shock and vibration energy caused by wheel behavior, and they provide handling stability. As a piston rod is an important component in shock absorbers, multiple processes are performed in order to guarantee its quality during manufacturing. Micro-defects can be generated on the surfaces of piston rods after processing. Because these defects can degrade the function of shock absorbers, proper non-destructive techniques are necessary to monitor the surfaces of piston rods. In this study, micro-defects were artificially machined on the surfaces of piston rods. In particular, a Rayleigh wave technique was adopted to detect defects on the surfaces of the piston rods, and Rayleigh wave behaviors were analyzed to establish beam profiles. In terms of the experimental method, defects were fabricated on the piston rods, and the optimal Rayleigh angle was determined using the pulse-echo method with ultrasonic transducers in a water tank. This was performed to evaluate the characteristics of the Rayleigh waves. In testing, regardless of the types of micro-defects on the surfaces of the pistons, it was found that the optimal inspection condition could be in the range of 5–10 mm, where ultrasonic signals were received with a high resolution. Moreover, the behaviors of the transmitted Rayleigh waves were simulated, and reflection, transmission, and scattering occurred due to defects at the interface between the water and steel. Thus, the propagation of Rayleigh waves and the optimal test conditions were implemented through FEM simulation to generate effective Rayleigh waves.

1. Introduction

A shock absorber is a part of a vehicle; it absorbs the impact energy of wheel behavior to provide a comfortable ride, and it secures traction to provide driving stability. A piston rod plays an important role in this component. To adjust the damping force of shock absorbers, lightweight parts for performance and high fuel efficiency are necessary, in addition to reducing defects and run-out defects. Automobile manufacturers need effective NDE techniques to guarantee the integrity of piston rods. Various processes are performed on the surfaces of piston rods during the manufacturing process. Currently, sampling and visual inspections are used to test integrity [1,2].

A study was conducted to visualize the crack defects of structures using laser-scanning-system-based ultrasonic imaging techniques. Such techniques were applied to evaluate structural defects, and they entailed guided ultrasonic and standing wave response imaging techniques to visualize crack defects [3]. Recently, laser imaging techniques using acoustic wavenumber spectroscopy (AWS) have been widely used to detect defects in plate structures. For example, a study was conducted to estimate the size of wall-thinning defects on thin-plate structures using finite element analysis (FEA) [4].

Additionally, small cracks, sometimes termed microcracks or microstructural defects, can themselves compromise the mechanical strength and integrity of structures [5]. Piston rods for shock absorbers in automobiles are either solid (integrated) and made by forging, or they are light weight and hollow to improve fuel efficiency and made by cutting drawn pipes or round bars [6,7]. During the manufacturing process, various types of surface defects occur on piston rods, which ultimately decrease product reliability and cause financial damage to the manufacturer. If defective products are used to manufacture cars, they can cause problems that ultimately result in casualties [8]. Moreover, a new guided wave technique was proposed to determine microstructural defects based on multiple modes [9]. Non-destructive techniques that can be used to measure defects on the surfaces of piston rods include Eddy-current testing, ultrasonic testing, and liquid penetrant testing. Rayleigh waves are widely used as an ultrasonic testing method because they are sensitive to micro-defects on sample surfaces [10,11,12,13].

Therefore, this study aimed to evaluate micro-defects on the surfaces of piston rods used in shock absorbers based on the Rayleigh wave technique. We analyzed beam profile behaviors to predict the propagation of Rayleigh waves affected by the micro-defects of the piston rods. In particular, we fabricated piston rod defects and the optimal Rayleigh angle set (the incident angle at which incident ultrasonic waves propagate by converting all of the ultrasonic energy into surface waves along the surface) with the use of the pulse-echo method with ultrasonic transducers in a water tank to evaluate the characteristics of the Rayleigh waves. Based on these Rayleigh angles, the pitch-catch technique was adopted to acquire Rayleigh wave signals, and the Rayleigh wave signals, which were affected by the micro-defects on the piston rods, were analyzed. Finally, the behavior of the Rayleigh waves and the optimal test conditions were implemented using FEM simulation to generate effective Rayleigh waves. Moreover, in this study, Rayleigh wave signals were analyzed according to the types and sizes of micro-defects on the surfaces of the piston rods, and the behaviors of the Rayleigh waves were visualized using FEM simulations. The optimal evaluation conditions were determined by analyzing the correlation between the types and sizes of the micro-class defects on the piston rods in ultrasonic testing.

2. Principle of Ultrasonic Waves and FEM Simulation

2.1. Measurement of Rayleigh Waves

Rayleigh waves are generated when ultrasonic waves reach the Rayleigh angle (the incident angle at which incident ultrasonic waves propagate by converting all of the ultrasonic energy into surface waves along the surface) at the liquid–solid interface as shown in Figure 1, and their energy is concentrated and propagated on a solid surface [1,5]. This study used Rayleigh waves with a high resolution to analyze the types of surface defects on piston rods. The energy of the Rayleigh waves was propagated on the solid surfaces in the direction of the angle of incidence through mode conversion.

Figure 1 shows how the waves are reflected by the law of reflection when ultrasonic waves reach the Rayleigh angle, how the reflection path is shifted by Schoch displacement, and how a null field is formed due to the offset interference of Rayleigh waves and geometrically reflected waves [10,11]. When two different media come into contact with each other, ultrasonic behaviors, such as refraction, transmission, reflection, and diffraction, occur depending on the angle of incidence, and this is called the mode conversion of ultrasonic waves. The Rayleigh wave used in this study is classified as an ultrasonic wave, and it was generated through the mode conversion of ultrasonic waves. Here, the Rayleigh wave is leaked as a longitudinal wave with directionality through mode conversion in the incident direction.

The Rayleigh wave speed of the sample is obtained using Equation (1) below [1,13].

$$C_R=\frac{C_i}{sin{\theta }_i }.$$

(1)

Here, $C_R$. denotes the Rayleigh wave speed of the piston rod, $C_i$ is the ultrasonic speed in the water, and ${\theta }_i$ is the Rayleigh angle. The Rayleigh angle was 29.1°, and the speed of the surface waves obtained using Equation (1) was 3040 m/s.

In order to perform the FEM simulation program, COMSOL Multiphysics 4.3 b was utilized; therefore, an acoustic–solid structure model was adopted to implement the ultrasonic mode conversion generated at the interface between the water and the solid surface. The propagation behavior of the Rayleigh wave was analyzed under the time domain. In the water area, except for the interface with the steel, the infinite plane was assumed by setting the plane wave radiation as the boundary condition. The behavior of the Rayleigh wave could be confirmed by simulating the ultrasonic signal as in Equation (2). Moreover, the center frequency was 10 MHz, the mesh size of the water was 0.02 mm, and the mesh size of the steel was 0.01 mm.

2.2. Simulation of Ultrasonic Wave Propagation

In this study, the critical angle of the ultrasonic transducer was determined in order to generate Rayleigh waves on the piston rods. For this purpose, a finite element method (FEM) was performed based on 2D simulation, and the propagation of ultrasonic waves was analyzed on the surfaces of the piston rods. This study also used FEM tools to visualize the ultrasonic waves in order to analyze the propagation of the ultrasonic waves on the surfaces of the piston rods. As shown in Figure 2, a finite element (FE) model was built for simulation under the assumption of an infinite plane in the water region, and the plane wave radiation was set as the boundary condition, but not for the interface with steel. The reflection of the ultrasonic waves was minimized by setting a low reflecting boundary everywhere in the steel region, except for at the interface with water to simulate ultrasonic waves from the transducer as shown in Equation (2), and the propagation of Rayleigh waves was confirmed [10,11].

$$p\left(t\right)=\left[1-\cos \left(\frac{2\pi \times f\times t}{2}\right)\right]\cos \left(2\pi \times f\times t\right)$$

(2)

Here, f and t denote frequency and time, respectively; the center frequency was set to 10 MHz, the mesh of the water was set to 0.1 mm, the mesh of the sample was set to 0.02 mm, and the depth of the Circum-type defect was set to 700 μm by considering the specifications of the FEM simulation workstation. Here, the minimum detectable size was approximately 152 μm. Additionally, the resolution of ultrasonic waves is strongly related to the frequency, and for use of bulk waves, the equation of the wave length is usually considered to be λ/2, but for surface micro-defects, the wave length of Rayleigh waves can be obtained to be approximately 50 μm through previous studies [11].

Figure 3 shows the propagation of the Rayleigh waves over time. Here red dotted circles are mainly directions in wave propagations. Figure 3a shows a simulation of the ultrasonic waves at the Rayleigh angle, and Figure 3b shows the moment when the ultrasonic waves are converted into Rayleigh waves at the interface between the water and steel, where reflection, transmission, and scattering occurred due to the defects. Figure 3c shows the behavior of the transmitted Rayleigh waves. Here, it was confirmed whether the Rayleigh angle value obtained through Equation (1) is an actual value. In addition, in order to obtain the optimal values of the Rayleigh angle, FEM simulation was implemented with the changing of several parameters, and the behavior of the ultrasonic waves was visualized as shown in Figure 3.

3. Experiment System and Measurement

3.1. Measurement System

A piston rod, a key component of shock absorbers, absorbs the impact and vibration energy from wheels. As surface defects on piston roads lead to a deterioration in performance, it is necessary to detect and evaluate any micro-defects on their surfaces. The outer diameter of the piston rods used in this study was 18 mm, and the final length was 390 mm. Figure 4 shows SM45C round bars, which were used for piston rods with micro-defects on the surface. Figure 4b shows a sample with a “Circum (circumferential)” crack on the surface. Figure 5a shows a sample with no cracks, and Figure 5b–d shows different types of defects that have the same length (5 mm) and width (500 μm) but different depths (100 μm, 300 μm, and 700 μm, respectively). Figure 5b shows an Axial crack, Figure 5c shows a Circum (circumferential) crack, and Figure 5d shows a Shear crack that is inclined 45 degrees from the axial direction.

3.2. Measurement of Ultrasonic Testing

Figure 6a shows the ultrasonic inspection system set up in this study. We also designed a defect evaluation system that moves in six axes to develop inspection algorithms for various surface defects. In each direction, the X, Y, and Z axes were constructed symmetrically from left to right. The inspection areas of the X, Y, and Z axes were about 300 mm, 420 mm, and 200 mm, respectively. The system allowed us to adjust the angle of the rotating module by 0.1° up to 240°. Using the system, we produced Rayleigh waves by introducing ultrasonic waves at a theoretical Rayleigh angle. The error of this theoretical angle was minimized by slightly changing the angle to confirm the angle with the largest signal reflected from the surface. The tests were performed by applying the pitch-catch technique using the ultrasonic inspection system. The transmitting and receiving transducers were fixed to maintain the Rayleigh angle, and the receiving transducer scanned the defect while moving in 0.5 mm increments toward the defect. The tests were conducted in the same way as those in the FEM simulation setup. Figure 6b shows the setup and the measurement method used to generate the Rayleigh waves. The transmitting transducer was fixed at the location of −5 mm. Both the transmitting transducer and receiving transducer were almost touching in the immersion tank. Then, the receiving transducer was moved to the right in 0.5 mm increments to measure the peak-to-peak amplitude. The optimal angle of incidence (θ) was set to 29.1° according to Snell’s law, and the size and frequency of the ultrasonic transducer were set to 6.35 mm and 10 MHz, respectively. The Rayleigh angles of the ultrasonic waves were changed depending on the difference in the ultrasonic impedance of the specimens, the properties of the materials, and the presence or absence of surface treatment. Here, experiments on ultrasonic waves were performed with specimens used in commercial vehicles, and the reproducibility of the ultrasonic testing results was shown to be very high. This study shows a correlation between the types of defects and the transmission/reception locations of the transducers. In this system, the use of an algorithm could allow for the sizing and evaluation of defect signals. If the optimal inspection location is selected based on the research results and defects are analyzed through the selected inspection location, it is expected that a system can be established to determine the number of defects.

Figure 7 shows the typical peak-to-peak amplitudes used to evaluate the characteristics after machining defects were fabricated on the surfaces of the piston rods. Here red dotted boxes mean mainly amount of amplitudes. They are examples of measuring the amplitude by measuring each type of defect on the piston rods. The figure shows the A-scan signals of the maximum peak points for “Non-crack”, “Circum”, “Axial”, and “Shear” defect samples with a defect depth of 700 μm. Figure 7a shows the peak-to-peak amplitude of the sample with no defects, and Figure 7b–d shows peak-to-peak amplitudes of the different defect types. According to the values in Figure 7, the amplitude is largest when there are no defects, while the other three defect types show slight differences. By measuring the peak-to-peak amplitude, we can implement a Rayleigh wave profile to evaluate the defect behavior characteristics.

4. Discussion and Results

4.1. Optimization Measurement Technique of Rayleigh Angles

The Rayleigh angle must be determined before evaluating the surface defects on the piston rods. Firstly, as shown in Figure 8, ultrasonic signals were transmitted based on the pulse-echo method, and two ultrasonic transducers were utilized to obtain the signals reflected from the crack to perform the Rayleigh wave test. The distance was set to 5 mm between the crack sample and the point where the ultrasonic waves propagated the sample surface.

Although the Rayleigh angle is generally determined using Snell’s law, the peak-to-peak amplitude of the ultrasonic waves received at the optimal incident angle (29.1°) was measured first since each material has different properties, [1,13]. Secondly, the peak-to-peak amplitude at a random small angle θ₁ was measured. Finally, the peak-to-peak amplitude received at a random small angle θ₂, where random small angles θ can be repeated multiple times, was measured. The largest peak-to-peak amplitude of the three values became the Rayleigh angle used to perform the Rayleigh wave test. Figure 9 shows the Rayleigh wave angles determined using the pulse-echo technique to obtain the maximum peak-to-peak value. The results in Figure 9 were obtained in tests where the pulse-echo technique was applied using the ultrasonic inspection system, and the transmitting and receiving transducers were fixed and scanned while maintaining the Rayleigh angle. The incident point of the incident beam was fixed to 5 mm before the defect was fabricated. Figure 9a shows the peak-to-peak amplitude received after adjusting the random small angle θ₁ to 28.6°, and Figure 9b shows the peak-to-peak amplitude received after adjusting the angle to 29.1°. Figure 9c shows the peak-to-peak amplitude received after adjusting the random small angle θ₂ to 29.6°. Figure 9d compares the three peak-to-peak amplitudes according to the Rayleigh angle. Therefore, the Rayleigh angle was set to 29.1°, which had the largest peak-to-peak amplitude.

4.2. Evaluation of Beam Profiles on the Defect Types

Figure 10 shows the size of the amplitude according to the distance between the transmitting and receiving transducers for the defect depth of the “Circum” defect type.

To compare the amplitudes of the ultrasonic defect signals according to the defect depth, we performed a standardization process for 100 μm, 300 μm, and 700 μm based on the amplitude when there was no defect. The Rayleigh wave scan section (−5–+3 mm), as shown in Figure 6b, was where the ultrasonic signals were received. This section was expected to be a null field, as shown in Figure 1 because there was no correlation according to the size of the defect. Moreover, the amplitude of the ultrasonic signals received beyond 3 mm becomes larger as the size of the defect decreases. We can also see an increase in the ultrasonic waves leaking into the water as the distance between the two transducers increases.

Figure 11 shows the size of the amplitude according to the distance between the transmitting and receiving transducers for the defect depth of the “Axial” defect type. The same standardization process as that shown in Figure 10 was performed.

In terms of the “Axial” defect type, the amplitude of the ultrasonic defect signals also decreases as the distance between the two transducers increases. In particular, the amplitude of the ultrasonic signal decreases slowly around 2–3 mm and then decreases rapidly after 5 mm. This is because the defect length is in the same direction as the propagation of the ultrasonic waves, and the ultrasonic wave propagation characteristics change according to the defect type.

Figure 12 shows the size of the amplitude according to the distance between the transmitting and receiving transducers for the defect depth of the “Shear” defect type. In the case of the “Shear” defect type, ultrasonic signals of 300 μm and 700 μm appear at similar locations. This is because the ultrasonic waves and the defect intersect at an angle of 45°. Thus, the ultrasonic waves scatter in various forms, such as diffraction, reflection, and transmission, when the defect and ultrasonic waves come into contact.

Based on Figure 10, Figure 11 and Figure 12, we chose the 700 μm crack with the highest resolution of Rayleigh waves and compared the normalized peak-to-peak amplitudes of the non-crack and three crack types. Figure 13 shows the correlation of the defect types when the defect depth is 700 μm. When the defect depth is 700 μm, the ultrasonic waves transmit to a certain extent in the “Axial” and “Shear” defect types. However, most of the ultrasonic waves cannot transmit in the “Circum” defect type. It is also difficult to distinguish the defect types from the amplitude of the ultrasonic waves received from about 8 mm. We can see a difference depending on the crack type in the vicinity of 5 mm on the X axis. The peak-to-peak amplitude is very low because the “Circum”-type crack can block a large number of Rayleigh waves. As expected, the “Axial”-type crack shows a higher peak-to-peak amplitude than the “Circum” crack does.

Figure 14 shows the peak-to-peak amplitudes of the Rayleigh waves according to the crack type and depth of the piston rod. The figure shows a correlation of the defect types according to crack depth when the distance of the receiving transducer is 6.5 mm from the crack location. In all of the defect types, the amplitude of the received signals tends to decrease as the crack depth increases. The amplitude of the ultrasonic signal is the smallest in the “Circum”-type defect, where the crack is perpendicular to the propagation of the Rayleigh waves. In the case of the “Axial”-type defect, where the crack is in the same direction as the propagation of the Rayleigh waves, the amplitude is about 1.8–2.4 times that of the “Circum”-type defect. In the case of the “Shear”-type defect, where the Rayleigh waves and the crack location intersect at an angle of 45°, the peak-to-peak amplitude is in the middle of the “Circum” and “Axial” defects depending on the crack depth. However, even though the amplitude values of the “Axial” and “Shear” defects overlap when the crack depth is 100 μm, this does not significantly affect the defect type. The amplitude of the Rayleigh waves decreases as the crack depth increases, which shows that there is a close correlation with the defect type. The test results also show that the receiving transducer receives signals with a high reliability in the range of about 5–10 mm. However, the amplitude of the Rayleigh waves is expected to decrease rapidly beyond this section.

5. Conclusions

In this study, regarding the application and utilization of UT in a non-destructive examination of defects on piston rods, a measurement technique was established. Thus, ultrasonic defect signals were analyzed according to the types and depths of the defects on piston rods for automobiles in order to determine how to select the optimal inspection conditions.

(1)

To evaluate ultrasonic beam behavior in defects on piston rods, the Rayleigh angle was determined using the optimal peak-to-peak amplitude based on the pulse-echo technique after machining the defects on the surfaces of the piston rods.

(2)

Regardless of the types of micro-defects on the surfaces, the peak-to-peak amplitude of the transmitted ultrasonic waves decreased as the size of the defect increased. The received ultrasonic signals decreased linearly as the distance increased between the transmitting and receiving transducers. Accordingly, the optimal inspection condition could be suggested to be in the range of 5–10 mm, where ultrasonic signals with a high resolution were received.

(3)

When the surface defect occurred in the direction of the propagation of the Rayleigh waves, the amplitude of the Rayleigh waves was high. In contrast, the lowest number of ultrasonic waves was transmitted when the crack was perpendicular to the propagation of the Rayleigh waves. Although the defect detection resolution is proportional to the wavelength (λ) (usually λ/2) in ultrasonic inspection, a resolution of λ/3 was found despite the loss of Rayleigh waves on the curved surface as a result of using piston rod samples in this study.

(4)

Additional studies are required in the future to analyze the effects of different defect types and heat treatment characteristics using defects with more diverse depths in order to improve the reliability of piston rods in automobiles.