-
-
-
-
Capability
-
-
Special Feature on New Technologies: Full-Focusing Ultrasonic Testing of Main Pipeline Welds with Stainless Steel Surfacing
Release Date:
2021/09/11
The primary circuit piping in a pressurized water reactor nuclear power plant connects the main coolant pump, steam generator, and reactor pressure vessel, and is classified as a Class 1 nuclear safety system. The base material of this piping is medium- to low-alloy steel; to meet specific requirements such as corrosion resistance, both the base material and the inner surfaces of the welds are typically overlay-welded with a protective layer of austenitic stainless steel of a specified thickness.
The primary-loop piping of a pressurized-water reactor nuclear power plant connects the main coolant pump, steam generator, and reactor pressure vessel, and is classified as a Class 1 nuclear safety system. The base material is medium- to low-alloy steel; to meet stringent requirements such as corrosion resistance, a protective overlay of austenitic stainless steel of specified thickness is typically deposited by surfacing on both the base metal and the inner surfaces of the welds. In the 1970s, the United States first detected reheat cracking in the heat-affected zone beneath the stainless-steel surfacing layer inside nuclear pressure vessels, which subsequently drew widespread attention within the nuclear industry to pre-service and in-service nondestructive testing of surfacing-layer structures. Conventional ultrasonic testing combined with radiographic inspection can be used to examine the welds of main pipelines equipped with stainless-steel surfacing layers; however, radiography is relatively insensitive to area-type defects such as highly hazardous cracks and cannot provide quantitative depth measurements for defects oriented along the wall thickness. Ultrasonic testing, by contrast, is highly sensitive to area-type defects, exhibits small positioning and quantification errors, and has been extensively employed for weld inspection in nuclear power plants. The surfacing layer, being made of austenitic stainless steel, features coarse grain structure and pronounced elastic anisotropy, which lead to attenuation and scattering of ultrasonic waves during propagation. This results in “grass-like” echo patterns and a low signal-to-noise ratio, thereby hindering the identification and quantification of defects.
Currently, the total focusing method (TFM) based on full-matrix capture (FMC) data is widely used in ultrasonic testing. By applying time-delayed summation to the array’s A-scan signals, TFM achieves point-by-point focusing of acoustic energy in the inspection region, thereby enhancing the signal-to-noise ratio of the inspection. Building on this approach, CAMACHO et al. further proposed the phase-coherent imaging (PCI) technique for suppressing structural noise. This method leverages the phase coherence of the array signals to construct a phase-coherence factor and applies weighted processing to the total-focusing image, further improving both the signal-to-noise ratio and imaging quality.
Technical Principle >
Full-Focus Method
TFM is a signal post-processing method based on FMC data from phased-array transducers. The underlying principle is as follows: for a phased-array transducer with N elements, each element sequentially transmits an ultrasonic pulse, while all elements simultaneously receive and store A-scan signals, resulting in a complete full-matrix dataset comprising N² time-domain signals. Subsequently, the imaging region is discretized into a grid and a coordinate system is established, with each transducer element represented as a single point. Under conditions of normal incidence of the ultrasound wave, the position of transducer element i is denoted by (xi, 0). Let the coordinates of an arbitrary focal point Q be (xref, zref); based on the acoustic path lengths from each transducer element to point Q, a delay rule is derived, and the total amplitude response ITFM(xref, zref) of all signals passing through this point is expressed as:
In the equation: yij(t) denotes the ultrasonic signal transmitted by array element i (xi, 0) and received by array element j (xj, 0); tij(xref, zref) represents the propagation time of the signal emitted by array element i and received by array element j as it passes through point Q.
By applying time-domain array signals through delayed superposition and focusing at each grid point, TFM imaging and defect detection of the inspected area can be achieved.
Phase-coherent imaging method
Structural noise induced by austenitic stainless steel overlay welding on the inner surface of the main pipeline can interfere with defect identification. TFM performs time-delayed signal averaging solely based on amplitude information, which, while enhancing the amplitude of defect signals, also amplifies the noise amplitude. In contrast, the PCI method employs adaptive weighted imaging based on statistical characteristics of the phase distribution of ultrasonic signals, leveraging the fact that defect echoes exhibit a consistent phase distribution in array A-scan data, whereas noise exhibits a scattered phase distribution. By constructing a coherence factor that characterizes the phase distribution of the signal based on phase coherence, the noise amplitude in TFM-reconstructed images can be reduced. The phase-coherence factor C(xref, zref) constructed during the imaging process is as follows:
In the equation: φ denotes the signal phase angle; var denotes the variance calculation.
Real-time point-by-point weighting of the full-focus image is performed using the following equation to enhance the detection signal-to-noise ratio.
In the equation, IPCI(xref, zref) represents the total amplitude of the response of the signal after weighted processing.
Metallographic Analysis
The appearance of the alloy steel butt-welded pipe section with an austenitic stainless steel surfacing layer is shown in Figure 1(a). The wall thickness of the pipe section is 74 mm, and the surfacing-layer test specimen was taken from the weld zone. The dimensions of the test specimen are 21 mm in length, 14.5 mm in width, and 11.5 mm in height, as shown in Figure 1(b).
Figure 1: Appearance of the alloy steel butt-welded pipe section with an austenitic stainless steel surfacing layer and the surfacing-layer test specimen
The sample was ground and polished using water-based sandpaper, followed by etching of the specimen surface with a potassium dichromate cold-acid etchant. The macro- and microstructural features of the surfacing layer specimens were then examined using a Zeiss MEF-4 metallographic microscope.
The macroscopic metallographic photograph of the austenitic stainless steel surfacing layer specimen is shown in Figure 2. It can be seen that the grains within the surfacing layer are columnar in distribution, with grain widths ranging from 200 to 300 μm and lengths from 3,000 to 4,000 μm. The grain orientations are relatively random, with growth directions spanning −60° to +30°. The austenitic grains grow along the temperature gradient and their growth is governed by the epitaxial growth mechanism. During the solidification of the surfacing layer during welding, rapid heat dissipation occurs at the weld bead edges, resulting in the formation of a columnar grain structure and complex grain orientations in the surfacing layer. This, in turn, leads to pronounced elastic anisotropy in the austenitic stainless steel surfacing layer.
Figure 2 Macroscopic metallographic photograph of the austenitic stainless steel surfacing layer specimen
Further microstructural metallographic examination at 500× magnification was conducted on three distinct regions within the surfacing layer shown in Figure 2, as illustrated in Figure 3. The microstructure of the surfacing layer consists of austenite and ferrite, with a relatively high ferrite content (approximately 40%). In Figure 3, the white matrix regions correspond to austenite, while the fishbone-shaped black regions represent high-temperature ferrite. Significant variations are observed in the distribution of the fishbone-shaped ferrite across different regions, with both its content and orientation undergoing continuous changes, thereby further enhancing the microstructural heterogeneity of the surfacing layer.
Figure 3 Microstructural metallographic photographs of different regions of the austenitic stainless steel surfacing layer specimen
Full-Focus Imaging Experiment
A side-drilled hole with a depth of 65 mm and a diameter of 2 mm was machined at the center of the overlay weld in the butt-joint pipe section shown in Figure 1(a). Ultrasonic testing was performed using a Multi X++ ultrasonic flaw detector equipped with the Multi 2000 operating software, employing a 64-element phased-array linear transducer with a nominal frequency of 3.5 MHz. FMC data were acquired using 16, 32, 48, and 64 array elements, followed by TFM imaging; the grid resolution in the target region was 0.5 mm. The TFM images obtained with different numbers of array elements are shown in Figure 4, demonstrating that the defect imaging results are dependent on the number of array elements. When the number of array elements is 16, the image quality is poor: the defect appears as an elongated elliptical shape with low amplitude, which hinders accurate identification. In contrast, when the number of array elements is 48 or 64, the defect image quality improves significantly, with energy concentrated in the defect area and increased amplitude.
Figure 4: TFM imaging results for different numbers of array elements
To evaluate the results of ultrasonic imaging, quantitative analysis is performed using the signal-to-noise ratio (SNR) and the array performance index (API), which are defined as follows:
SNR = 20 log(Imax/Iaverage) (4)
In the equation: Imax is the amplitude of the defect signal; Iaverage is the average amplitude of the noise signal.
API = A - 6 dB/λ² (5)
In the equation, A-6dB is the cross-sectional area corresponding to a 6-dB reduction in the amplitude of a point reflector; λ is the wavelength.
The signal-to-noise ratio characterizes the relative strength of defect signals versus noise signals and thus reflects the quality of ultrasonic imaging and the effectiveness of defect detection. The array performance index is a dimensionless parameter that indicates the resolvability of defects in the image; a lower value signifies a narrower ultrasonic beam, which in turn results in higher defect imaging quality (since a wider beam can easily cause signal overlap between adjacent defects in the image).
The signal-to-noise ratios and array performance indices for TFM imaging before and after applying the PCI method at different element counts are shown in Figure 5. As the number of elements increases, the signal-to-noise ratio continuously improves; the SNR for a 64-element array is 5.3 dB higher than that for a 16-element array. Conversely, the array performance index steadily decreases, with the 64-element array exhibiting a 35% reduction compared with the 16-element array. This is because, as the number of elements grows, the FMC data contain an increasing number of A-scan signals, thereby expanding the beam coverage and enabling each transducer element to capture more characteristic information about both the material and any defects. After time-delayed summation processing, the energy at defect locations is enhanced relative to that at non-defect locations, leading to progressively improved imaging quality. When the array comprises 64 elements, the TFM images can accurately depict the features of the weld seam in the austenitic stainless steel overlay layer, effectively suppressing artifacts caused by beam distortion and scattering; however, noticeable structural noise interference still remains.
Figure 5 Signal-to-noise ratio and array performance index of TFM imaging before and after applying the PCI method for different numbers of array elements
The phase information of each A-scan signal in the FMC data is extracted, and a phase-coherence factor matrix is constructed using Equation (2). Subsequently, adaptive weighting is applied to each TFM image shown in Figure 4, yielding the results presented in Figure 6.
Figure 6 TMF imaging results using the PCI method with different numbers of array elements
As shown by comparing Figure 4 with Figure 6, TFM imaging combined with PCI effectively preserves defect features while suppressing structural noise in the surfacing layer, thereby facilitating defect identification.
Conclusion
1. The weld overlay layer of austenitic stainless steel exhibits a microstructure characterized by coarse, columnar grains with complex grain orientation, a high volume fraction of fishbone-shaped ferrite, and significant variations in the orientation of the fishbone ferrite across different regions.
2. TFM helps improve the quality of imaging-based inspection of weld defects in stainless steel overlay-welded joints; moreover, the greater the number of array elements used for imaging, the higher the image signal-to-noise ratio and the lower the array performance index.
3. Extracting the signal phase information from FMC data and applying adaptive weighting can further suppress structural noise, thereby improving the image’s signal-to-noise ratio and enhancing defect detectability.
Related Blog
To further carry forward the spirit of Lei Feng and promote the volunteer ethos of “dedication, friendship, mutual assistance, and progress,” in accordance with the spirit of the Henan Provincial Association for Science and Technology’s “Notice on Launching a Centralized ‘i-Science Popularization’ Science and Technology Volunteer Service Campaign in the Spirit of Lei Feng,” Luoyang Weicheng Engineering Testing Co., Ltd. responded enthusiastically to the call and organized a science popularization event on the safe operation of forklifts. On March 4, our company meticulously planned and conducted an emergency response drill for special equipment—forklift accidents—and invited several nearby companies that use forklifts to observe the exercise. The purpose of this drill was to simulate real-world working conditions, specifically the sudden occurrence of an accident during the process of using a forklift to handle goods.
2024-03-05
On the afternoon of December 29, 2023, Weicheng Testing Co., Ltd. held a grand gratitude celebration—“Our Ten Years”—inviting distinguished guests, including Mr. Jia Hong, Director and Vice Chairman of the Titanium-Zirconium Industry Development and Application Center of the China Equipment Management Association, as well as Mr. Lu Fusheng, Executive Deputy Director, Vice Chairman, and Secretary-General, along with clients and partners, to jointly witness this milestone occasion. The first segment of the event: A Decade of Development The event officially commenced at 2:00 p.m., with the company welcoming all attendees
2024-01-11
On October 11, Luoyang Weicheng Engineering Testing Co., Ltd. held a large-scale training and competition—specifically, a γ-ray radiographic inspection radiation emergency response drill—under the theme “Everyone Talks About Safety, Everyone Knows How to Respond to Emergencies.” The event invited leaders from the provincial, municipal, and district levels to attend and observe. Prior to the start of the emergency response drill, Cui Songbiao, General Manager of Luoyang Weicheng Engineering Testing Co., Ltd., and Gong Aifeng, Deputy General Manager, extended a warm welcome to the visiting provincial, municipal, and district leaders and pledged that Weicheng Testing would consistently focus on enhancing its professional technical expertise and emergency response capabilities as key drivers for corporate safety management.
2023-10-19
Contact Us
Address: Room 104, Building 9, Futong Intelligent Manufacturing Industrial Park, Jinye 1st Road, Yiyang County, Luoyang City, Henan Province
Phone:+86-379-66915081 +86-379-66915005
Email:lywcjc@lywcjc.cn
Fax:+86-379-66919852
Website:www.lywcjc.cn
Copyright © 2025 Luoyang Weicheng Engineering Testing Co., Ltd. www.300.cn Privacy Policy