Quality detection method in functionally graded material preparation process

Abstract

The invention discloses a quality detection method in a functionally graded material preparation process, and relates to the technical field of functionally graded materials.The quality detection method comprises the following steps that the temperature of a green body in the functionally graded material preparation process is collected, and green body temperature distribution data is obtained; performing real-time thickness measurement on the green body according to the green body temperature distribution data to obtain a green body thickness change curve; identifying an abnormal fluctuation section in the thickness change curve of the green body, and carrying out phase detection on a physical region with the abnormal fluctuation section of the green body to obtain regional phase composition data; performing internal defect scanning on the green body according to the regional phase composition data to obtain an internal defect map of the green body; performing quality evaluation on the green body based on the green body internal defect map to obtain a quality grade judgment result; according to the method, the ultrasonic flaw detection scanning range is guided based on the phase abnormal region, and high-resolution internal flaw detection can be performed on the high-risk region of the functionally graded material in a targeted manner.

Description

Technical Field

[0001] This invention relates to the field of functionally graded materials technology, and more specifically to a quality inspection method in the preparation process of functionally graded materials. Background Technology

[0002] Functionally graded materials (FGMs), with their spatially continuously varying composition and properties, can effectively alleviate thermal stress and mechanical mismatch at material interfaces, showing broad application prospects in aerospace, biomedicine, nuclear energy, and electronic devices. With the development of manufacturing technologies, processes such as powder metallurgy, plasma spraying, laser cladding, and 3D printing are widely used in the preparation of FGMs. However, because their multilayered heterogeneous structures are built layer by layer under high temperature and non-equilibrium conditions, they are prone to problems such as uneven temperature fields, poor interlayer bonding, residual stress concentration, and abnormal microstructure evolution during the forming process, seriously affecting the reliability and service life of the final product. Therefore, real-time and accurate quality monitoring during the preparation process is particularly crucial.

[0003] In existing technologies, quality inspection of functionally graded materials (FJTs) primarily employs offline, single-point, and post-processing analysis methods, mainly including: 1. Manual sampling metallographic inspection: This requires cutting samples after preparation and observing the microstructure under a microscope, which damages the preform and cannot provide real-time feedback; 2. Fixed-point thermocouple temperature measurement: This can only obtain the temperature at a limited number of points, failing to comprehensively reflect the temperature field distribution of the preform and making it difficult to correlate with the spatial changes in the material gradient; 3. Independent non-destructive testing: Such as using ultrasonic testing for full-area scanning or only performing X-ray sampling on the finished product. These methods for quality inspection of FJTs suffer from isolated data, low efficiency, and insensitivity to early anomalies (such as minor phase deviations) occurring during preparation, failing to achieve accurate diagnosis through "anomaly-guided detection." Therefore, existing inspection methods have inherent defects such as a disconnect between process and quality, limited data dimensions, high degree of blindness in inspection, and insufficient early warning capabilities for potential defects, making it difficult to meet the online monitoring requirements for high-quality preparation of FJTs. Summary of the Invention

[0004] To address the problems existing in the prior art, this invention provides a quality inspection method in the preparation process of functionally graded materials, which realizes ultrasonic flaw detection scanning range guided by the phase anomaly region, and can perform high-resolution internal defect detection in high-risk areas of functionally graded materials.

[0005] The technical solution of the present invention is as follows: In a first aspect of the present invention, a quality inspection method is provided during the preparation process of functionally graded materials, comprising the following steps: The temperature of the billet was collected during the preparation of functionally graded materials to obtain the temperature distribution data of the billet; Real-time thickness measurement of the billet is performed based on the billet temperature distribution data to obtain the billet thickness variation curve; Identify abnormal fluctuation segments in the billet thickness variation curve, perform phase detection on the physical regions of the billet with abnormal fluctuation segments, and obtain regional phase composition data; Based on the regional phase composition data, the internal defects of the billet are scanned to obtain the internal defect map of the billet. The quality of the billet is assessed based on the internal defect map, and the quality grade is determined.

[0006] In some embodiments of the present invention, the temperature of the billet is collected using an infrared thermometer, and the temperature distribution data of the billet is obtained in the following manner: The outer surface and cross-section of the billet in the process of preparing functionally graded materials are divided into regions using a mesh generation tool to obtain a billet temperature measurement mesh. The billet temperature measurement mesh is then numbered according to the gradient region type to obtain a numbered temperature measurement mesh. Based on a numbered temperature measurement grid, an infrared thermometer is used to collect the temperature of the billet at fixed points to obtain the temperature distribution data of the billet.

[0007] In some embodiments of the present invention, a laser thickness gauge is used to measure the thickness of the billet in real time, and the thickness variation curve of the billet is obtained in the following way: Based on the temperature distribution data of the billet, a temperature gradient analysis is performed on the billet to obtain a temperature gradient map. Based on the temperature gradient map, the billet is divided into key areas to obtain key monitoring areas. In the key monitoring area, a laser thickness gauge is used to measure the thickness of the billet in real time to obtain real-time thickness data. The real-time thickness data is then time-aligned to obtain time-series thickness data. The change trend of the time-series thickness data is calculated to obtain the thickness change characteristics, and the thickness change curve of the billet is generated based on the thickness change characteristics.

[0008] In some embodiments of the present invention, an X-ray diffractometer is used to detect the phase composition of the physical region in which the billet exhibits abnormal fluctuations. The phase composition data of the region is obtained in the following manner: The spatial location of the billet corresponding to the abnormal fluctuation section is marked by a coordinate mapping tool to obtain the area to be tested, and the area to be tested is sampled to obtain a phase detection sample. The phase detection sample was scanned by X-ray diffraction to obtain the diffraction intensity pattern, and the positions of the characteristic peaks in the diffraction intensity pattern were identified to obtain the coordinate values ​​of the characteristic peaks. The characteristic peak coordinates are matched with known phase standard diffraction data using a standard card comparison tool to obtain phase matching results. The phase matching results are then sorted by content percentage to obtain regional phase composition data.

[0009] In some embodiments of the present invention, the spatial location of the billet corresponding to the abnormal fluctuation segment is marked using a coordinate mapping tool to obtain the area to be detected, specifically including: The abnormal fluctuation segment in the billet thickness variation curve is segmented to obtain fluctuation interval data. The measurement time points in the fluctuation interval data are matched with the scanning positions of the laser thickness gauge to obtain a fluctuation point location table. Based on the fluctuation point table, the billet is divided into grids and marked to obtain a detection grid map. The regions in the detection grid map that match the fluctuation point table are then expanded to obtain the regions to be detected.

[0010] In some embodiments of the present invention, X-ray diffractometer is used to perform diffraction scanning on the phase detection sample to obtain a diffraction intensity spectrum, specifically including: The sample for phase detection was continuously scanned using an X-ray diffractometer to obtain raw diffraction data. The raw diffraction data was then processed by background subtraction to obtain net diffraction data. Peak enhancement processing is performed on the net diffraction data to obtain enhanced diffraction data, and coordinate transformation is performed on the enhanced diffraction data to obtain the diffraction intensity spectrum.

[0011] In some embodiments of the present invention, coordinate transformation is performed on the enhanced diffraction data to obtain a diffraction intensity spectrum, specifically including: The diffraction angle value and diffraction intensity value in the enhanced diffraction data are extracted in pairs to obtain diffraction data pairs. The horizontal and vertical axis parameters of the diffraction data pairs are set by the coordinate definition tool to obtain the spectrum coordinate system. Based on the spectral coordinate system, the diffraction data pairs are mapped to obtain diffraction coordinate points, and the diffraction coordinate points are continuously connected to obtain the diffraction intensity spectrum.

[0012] In some embodiments of the present invention, an ultrasonic flaw detector is used to scan the internal defects of the billet, and the internal defect map is obtained by the following method: The differences in phase distribution in the regional phase composition data are analyzed to mark areas where the phase composition does not conform to the gradient design, thus obtaining key inspection areas. The key inspection areas are then divided into grid cells to obtain the flaw detection grid. The ultrasonic flaw detector collects ultrasonic reflection signals from each cell in the flaw detection grid to obtain cell reflection signals. Abnormal reflection waves in the cell reflection signals are then identified to obtain defect signal data. The defect signal data is classified according to the defect signal intensity to obtain classified defect data. Based on the positional correspondence between the classified defect data and the flaw detection grid, an internal defect map of the billet containing the defect location and intensity level is drawn.

[0013] In some embodiments of the present invention, ultrasonic reflection signals are collected from each cell in the flaw detection grid using an ultrasonic flaw detector to obtain cell reflection signals, specifically including: Based on the cell size and arrangement of the flaw detection grid, the scanning path of the ultrasonic flaw detector probe is planned to obtain the grid adaptation path. The probe is then driven to move according to the grid adaptation path to perform the first signal acquisition on each cell in the flaw detection grid and obtain the initial reflection signal. The grid cells in the initial reflection signal that did not acquire a valid signal are marked to obtain a blind zone cell table. Based on the blind zone cell table, the probe angle and transmission power are adjusted, and a second signal acquisition is performed on the blind zone cells in the flaw detection grid to obtain the complete reflection signal. The initial reflection signal and the complete reflection signal are integrated according to the grid cell number to obtain the cell reflection signal.

[0014] In some embodiments of the present invention, the quality grade determination result is obtained in the following manner: The defect location and strength level are extracted from the defect map of the billet to obtain defect feature data. The defect feature data are then classified and statistically analyzed according to the preset defect type to obtain defect classification and statistical results. The defect classification statistics are quantitatively scored using defect rating standards to obtain individual defect scores. A weighted sum is then calculated based on these individual defect scores to obtain a comprehensive quality score. Finally, the comprehensive quality score is compared with a preset quality level threshold to obtain the quality level determination result.

[0015] One or more technical solutions of the present invention have the following beneficial effects: (1) The quality inspection method provided by the present invention evaluates the quality of the billet based on the internal defect map of the billet and obtains the quality grade judgment result. It realizes the ultrasonic flaw detection scanning range guided by the abnormal phase area, and can perform high-resolution internal defect detection in a targeted manner in high-risk areas to obtain the defect map. It avoids the inefficiency of full-area scanning and improves the detection rate and positioning accuracy of defects such as micro-cracks, pores, and delamination.

[0016] (2) The quality inspection method provided by the present invention can obtain the internal defect spectrum of the billet by sequentially detecting the temperature, thickness and phase of the billet, and realize the quality assessment of the billet by obtaining the internal defect spectrum of the billet based on the phase. The entire inspection process adopts a multi-level and progressive inspection process, and establishes the intrinsic relationship between process parameters and final quality. When the thickness curve is abnormal, it can automatically trigger subsequent phase and defect detection, and accurately focus the inspection resources on high-risk areas, overcome the blindness and inefficiency of the existing full-area scanning, and realize the transformation from "general inspection" to "precision inspection".

[0017] (3) The quality inspection method provided by the present invention, in the process of inspection, adopts key area division based on temperature gradient and precise coordinate mapping based on thickness anomaly, so that the ultrasonic flaw detector can directly perform high-resolution scanning on the phase anomaly area. This precise inspection strategy, combined with the grid partitioning and secondary signal completion mechanism in ultrasonic flaw detection, significantly enhances the sensitivity of identification and spatial positioning accuracy of potential defects such as micro cracks, pores and interface delamination.

[0018] (4) The quality inspection method provided by the present invention establishes the final quality assessment on the quantitative analysis of the internal defect map. Through defect feature extraction, classification statistics and weighted scoring, it outputs objective quality level judgment results, which changes the traditional subjective judgment mode that relies on human experience and provides a data-driven and standardized decision basis for process optimization and quality traceability. Attached Figure Description

[0019] Figure 1 This is a flowchart of the quality inspection method in the preparation process of the functionally graded materials of the present invention. Detailed Implementation

[0020] It should be noted that the following detailed description is illustrative and intended to provide further explanation of the invention. Unless otherwise specified, all technical and scientific terms used in this invention have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0021] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention. As used herein, unless otherwise expressly indicated by the invention, the singular form is intended to include the plural form as well. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0022] Example 1 In a typical embodiment of the present invention, a quality inspection method is proposed during the preparation process of functionally graded materials, such as... Figure 1As shown, it includes the following steps: S1. Collect the temperature of the billet during the preparation of functionally graded materials and obtain the temperature distribution data of the billet. S2. Based on the billet temperature distribution data, perform real-time thickness measurement of the billet to obtain the billet thickness variation curve; S3. Identify abnormal fluctuation segments in the billet thickness variation curve, perform phase detection on the physical regions of the billet with abnormal fluctuation segments, and obtain regional phase composition data. S4. Based on the regional phase composition data, perform an internal defect scan on the billet to obtain an internal defect map of the billet. S5. Based on the internal defect map of the billet, the quality of the billet is evaluated to obtain the quality grade determination result.

[0023] Each step is explained in detail below.

[0024] S1. Collect the temperature of the billet during the preparation of functionally graded materials to obtain the temperature distribution data of the billet.

[0025] Specifically, the temperature of the billet is collected using an infrared thermometer. The method is as follows: During the layer-by-layer forming process of functionally graded materials using laser cladding or plasma spraying, the infrared thermometer is fixedly installed at an appropriate position on the forming equipment, ensuring that its temperature measurement field of view completely covers the surface area of ​​the billet currently being deposited. The optical axis of the thermometer is also ensured to be perpendicular to the measured surface of the billet to reduce measurement errors. The infrared thermometer continuously receives infrared radiation signals emitted from the surface of the billet and converts the radiation intensity into corresponding temperature values ​​according to preset material emissivity parameters. The acquisition frequency must match the material deposition rate to ensure that the time resolution of the data is sufficient to reflect instantaneous temperature changes. The acquired temperature values ​​at each spatial point are transmitted and recorded in real time through a data acquisition system, forming two-dimensional or three-dimensional temperature field data that varies with time and space, i.e., billet temperature distribution data. This data reflects the thermal history of the billet during each stage of heating, melting, solidification, and cooling.

[0026] For example, when preparing functionally graded titanium alloy / ceramic materials, if the interface region where the titanium alloy layer transitions to the ceramic layer experiences a sudden change in thermal conductivity, resulting in abnormal local heating or cooling rates, an infrared thermometer can detect the temperature deviation in this region. This provides accurate temporal and spatial information for subsequent thickness measurement using a laser thickness gauge based on the temperature distribution data, ensuring the continuity and synergy of the entire detection process.

[0027] The billet temperature distribution data were obtained in the following way: 1. Using a mesh generation tool, the outer surface and cross-section of the billet in the functionally graded material preparation process are divided into regions to obtain a billet temperature measurement mesh. The billet temperature measurement mesh is then numbered according to the gradient region type to obtain a numbered temperature measurement mesh.

[0028] The specific implementation method is as follows: In the process of forming functionally graded materials using layer-by-layer deposition processes such as laser cladding or plasma spraying, firstly, based on the design structural parameters of the blank and the gradient distribution law of the material composition, the geometric model of the blank is spatially discretized using a mesh generation tool pre-integrated in the control system. Based on the actual three-dimensional size information of the blank, the tool divides its outer surface into several rectangular or triangular units with the same area or adaptively adjusted according to the curvature. At the same time, on the cross section perpendicular to the deposition direction, layered regions corresponding to the material gradient layers are divided along the thickness direction. Each layered region corresponds to a specific material composition ratio range.

[0029] For example, a continuous transition region from a pure metal layer to a metal-ceramic hybrid layer and then to a pure ceramic layer forms a two-dimensional or three-dimensional grid system covering the overall spatial structure of the billet, i.e., a billet temperature measurement grid. Subsequently, based on the gradient region type in which each grid unit is located, such as a titanium alloy enrichment region, a transitional hybrid region, or a ceramic enrichment region, each grid unit is assigned a unique number. This number contains a region type code and spatial coordinate information. For example, "T1-L3-C2" represents a temperature measurement unit in the first region, third layer, and second column of the titanium alloy type. Finally, a numbered temperature measurement grid with clear spatial positioning and material property labels is generated.

[0030] 2. Based on the numbered temperature measurement grid, the temperature of the billet is collected at fixed points using an infrared thermometer to obtain the temperature distribution data of the billet.

[0031] The specific implementation method is as follows: An infrared thermometer is mounted on a programmable motion platform, and its temperature measurement field of view is aligned with the currently exposed outer surface area of ​​the billet. The control system maps the unit number in the numbered temperature measurement grid to its corresponding spatial coordinates, and guides the probe of the infrared thermometer to move sequentially to the center position above each numbered unit according to a preset scanning order, maintaining vertical alignment and stable focus. Temperature acquisition is initiated; the infrared thermometer receives the infrared radiation energy emitted by the surface of the unit and converts it into the corresponding temperature value according to the preset material emissivity correction parameter. Simultaneously, the temperature value, its corresponding unit number, and the acquisition timestamp are recorded. For temperature acquisition in the cross-sectional area, the preparation process is paused after several layers of deposition, and the internal cross-section is exposed using online cutting or tomographic imaging. Rapid temperature measurement of each numbered grid unit on this cross-section is completed within a short time. All acquired temperature data are categorized and organized according to unit number, forming a structured dataset containing spatial location, material region type, and real-time temperature value—that is, billet temperature distribution data.

[0032] For example, when preparing titanium alloy / ceramic functionally graded materials, when depositing to the third transition layer, the control system, according to the numbered temperature measurement grid, commands the infrared thermometer to prioritize the temperature of the mixed region numbered "T1-L3-C2" and "M2-L3-C4". It was found that the temperature of this region was about 80°C higher than that of the adjacent metal region, indicating that there was a local overheating phenomenon in this region. This abnormal data will serve as a key input for guiding the laser thickness gauge to perform key thickness measurements.

[0033] S2. Based on the temperature distribution data of the billet, perform real-time thickness measurement on the billet to obtain the thickness change curve of the billet.

[0034] Specifically, a laser thickness gauge is used to measure the thickness of the billet in real time. The specific implementation method is as follows: After the infrared thermometer completes the temperature acquisition of the billet surface during the preparation of functionally graded materials and forms continuous billet temperature distribution data, the timestamp and spatial coordinate information of the data are synchronously transmitted to the control system of the laser thickness gauge. Using the location of the current high-temperature active area or the location of significant temperature gradient change reflected in the temperature distribution data as a reference, the measurement starting point and scanning path of the laser thickness gauge are dynamically adjusted so that the probe of the laser thickness gauge is aligned with the billet area where material deposition or phase transformation is taking place. The laser thickness gauge emits a focused laser beam to irradiate the billet surface and receives the laser signal reflected from the surface. The distance change between the laser emission point and the reflection point is calculated by the principle of trigonometric method or interferometry, and then the thickness value at the measurement point is obtained. As the billet moves on the preparation platform according to the set trajectory, the laser thickness gauge continuously collects thickness data at multiple locations at a frequency synchronized with the deposition process. These data are integrated and processed in combination with time series, and finally a continuous curve reflecting the thickness evolution trend of the billet as it accumulates layer by layer during the forming process is formed, that is, the billet thickness change curve.

[0035] For example, in the preparation of functionally graded titanium alloy / ceramic materials, when the thickness of the deposited layer suddenly increases or decreases due to differences in material density and melting characteristics in the interface region where the titanium alloy layer transitions to the ceramic layer, a laser thickness gauge can capture this abnormal fluctuation segment in real time, thus providing a direct basis for subsequent X-ray diffraction to detect the phase in the corresponding region.

[0036] Furthermore, the thickness variation curve of the billet was obtained in the following way: 1. Based on the temperature distribution data of the billet, perform temperature gradient analysis on the billet to obtain a temperature gradient map, and divide the billet into key areas based on the temperature gradient map to obtain key monitoring areas.

[0037] The specific implementation method is as follows: After the infrared thermometer completes the fixed-point temperature acquisition of the outer surface and cross-section of the blank during the preparation of functionally graded materials and generates blank temperature distribution data containing the temperature values ​​of each numbered temperature measurement grid unit, the data is imported into the temperature analysis module. The temperature difference and spatial distance between adjacent grid units are calculated using the finite difference method or interpolation algorithm to obtain the temperature change direction and rate at each location, forming a spatial temperature gradient distribution image, i.e., a temperature gradient map, presented in the form of vector arrows or pseudo-color cloud maps. The high gradient region is represented by a color change band or a dense isotherm region. Then, based on the range of regions in the temperature gradient map where the gradient amplitude exceeds the preset process threshold (such as a temperature difference greater than 50°C per millimeter), combined with the design sequence structure information of the functionally graded materials, the regions located at the material component transition interface and exhibiting severe thermal response are identified.

[0038] For example, in titanium alloy / ceramic functional graded materials, if a steep temperature gradient caused by local high temperature accumulation or rapid cooling occurs at the junction of the third to fifth layers from the pure titanium layer to the TiC ceramic layer, the system marks this area as a potentially unstable region with continuous spatial coordinates, significant material type changes, and abnormal thermal history. This completes the division of the key areas of the billet and generates a set of key monitoring targets, i.e., the key monitoring area, which includes the start and end layer numbers, grid numbers, and three-dimensional coordinates.

[0039] 2. Use a laser thickness gauge to measure the thickness of the billet in real time in the key monitoring area to obtain real-time thickness data. Then, perform time alignment processing on the real-time thickness data to obtain time-series thickness data.

[0040] The specific implementation method is as follows: The control system transmits the spatial coordinate information of the key monitoring area to the motion control unit of the laser thickness gauge, drives the probe of the laser thickness gauge to move along a predetermined path to the surface position corresponding to each numbered grid in the area, maintains vertical incidence and emits a focused laser beam with a fixed sampling frequency, receives the light spot signal reflected from the surface of the billet, calculates the height difference of the current point through the triangular displacement sensing principle, and then derives the actual deposition thickness at the measuring point, forming a real-time thickness data stream that is continuously output as the preparation process progresses. At the same time, each set of thickness values ​​is accompanied by a timestamp generated synchronously by the central clock, and then matched with the time acquisition time recorded by the infrared thermometer, the ready status of the X-ray diffractometer and other multi-source signals on the time axis. Linear interpolation or delay compensation algorithms are used to eliminate the deviation caused by the asynchronous response of the equipment, ensuring that all data correspond to the same physical state under the same time reference, and finally integrates into a complete data sequence arranged in chronological order and accurately associated with the process time sequence, namely the chronological thickness data.

[0041] 3. Calculate the trend of change in the time-series thickness data to obtain the thickness change characteristics, and generate the thickness change curve of the billet based on the thickness change characteristics.

[0042] The specific implementation method is as follows: The filtered time-series thickness data is input into the trend analysis module. The local growth rate and fluctuation intensity are calculated using the least squares fitting or moving window standard deviation algorithm. The intervals in which the thickness deviates from the designed growth slope by more than ±10% are identified, or the non-monotonic sawtooth pattern is observed for three or more consecutive sampling points. Such data segments are judged to have thickness change characteristics with significant dynamic variability. Based on this, the system extracts parameters such as start and end time, peak deviation, and duration period, and plots a continuous line graph with time as the horizontal axis and thickness as the vertical axis. In the graph, the normal deposition trend is represented by solid lines, and the abnormal fluctuation segments are marked by dashed lines or red highlights. Finally, a visual curve that can intuitively reflect the geometric morphology evolution law of the key monitoring area throughout the entire preparation process is formed, namely the billet thickness change curve.

[0043] For example, when preparing titanium alloy / ceramic functionally graded materials, when the system detects a sharp increase in temperature gradient in the critical monitoring area numbered "M2-L3-C4" around the 180th second, it immediately guides the laser thickness gauge to intensify the scanning of this area and finds that its time-series thickness data decreases by 12% during this period. After trend calculation, it is confirmed as an abnormal thickness change feature. This result is directly reflected in the thickness change curve of the billet and becomes a key input condition for triggering subsequent phase detection by X-ray diffractometer.

[0044] S3. Identify abnormal fluctuation segments in the billet thickness variation curve, perform phase detection on the physical regions of the billet with abnormal fluctuation segments, and obtain the number of phase compositions in the regions.

[0045] Specifically, X-ray diffractometers are used to detect the phase composition of physical regions in the billet exhibiting abnormal fluctuations. The process involves: after a laser thickness gauge measures the real-time thickness of the billet during the functionally graded material preparation process and generates a continuous thickness variation curve, this curve data is input into an analysis system. The analysis system identifies abnormal fluctuation segments that deviate from the normal growth trend by setting a threshold range or using a sliding window algorithm. These abnormal fluctuation segments correspond to a thickness abrupt change region at a specific time and spatial location. The timestamp and spatial coordinates of these abnormal fluctuation segments are then transmitted to the positioning and control system of the X-ray diffractometer. These coordinates guide the X-ray diffractometer's probe to precisely align with the physical region on the billet surface corresponding to the abnormal fluctuation segment. The angle and scanning range of the incident X-rays are adjusted, X-ray irradiation is initiated, and diffraction signals generated by reflections from the internal crystal planes of the material are received. The diffraction angle and intensity distribution are recorded, and the interplanar spacing is calculated using the Bragg equation. By comparing with standard PDF cards, the types and relative contents of crystalline phases present in this region are identified, thus obtaining local phase information related to the abnormal thickness variation, i.e., regional phase composition data.

[0046] For example, in the preparation of functionally graded titanium alloy / ceramic materials, if there is an abnormal fluctuation in the thickness of the transition layer from titanium alloy to ceramic, the X-ray diffractometer will scan the interface region corresponding to this segment. It may detect the unexpected Ti2AlN phase or the incompletely reacted Al3Ti phase, indicating that a chemical reaction outside the design path or insufficient sintering has occurred in this region. The phase composition data of the region obtained in this way will serve as the basis for subsequent defect scanning by ultrasonic flaw detector.

[0047] Furthermore, the regional phase composition data were obtained in the following manner: 1. Using a coordinate mapping tool, the spatial position of the billet corresponding to the abnormal fluctuation segment is marked to obtain the area to be tested, and samples are taken from the area to be tested to obtain phase detection samples.

[0048] The specific implementation method is as follows: After the laser thickness gauge completes the real-time thickness measurement of the billet during the preparation of functionally graded materials and generates the billet thickness change curve, when the system identifies an abnormal fluctuation segment in the curve that exceeds the preset fluctuation threshold, the time interval and spatial coordinate information corresponding to the abnormal fluctuation segment are input into the coordinate mapping tool. Based on the motion trajectory record of the preparation equipment and the layer thickness accumulation model, the tool reversely calculates the physical deposition position corresponding to the abnormal fluctuation segment, that is, the specific layer number, grid number and geometric range in the three-dimensional structure of the billet. The area is then visually marked on the control interface by highlighting or color marking to form a clear area to be detected. Subsequently, the automated sampling device is started. The device moves to the area to be detected according to the marked coordinate information and uses micro-drilling or laser cutting to accurately cut a block sample with a size of about 5mm×5mm×2mm from the surface of the billet, ensuring that the sampling process does not introduce additional thermal effects or mechanical damage. The obtained sample is then polished and used as a phase detection sample for phase analysis.

[0049] Furthermore, by using coordinate mapping tools to mark the spatial location of the billet corresponding to the abnormal fluctuation segment, the area to be detected is obtained, specifically including: (1) The abnormal fluctuation segment in the thickness change curve of the billet is segmented to obtain the fluctuation interval data, and the measurement time point in the fluctuation interval data is matched with the scanning position of the laser thickness gauge to obtain the fluctuation point location table.

[0050] The specific implementation method is as follows: After generating the thickness variation curve of the billet, the system identifies abnormal fluctuation segments in the curve that deviate from the normal growth trend by setting upper and lower threshold bands or using a sliding standard deviation algorithm. These abnormal fluctuation segments are then cut out from the overall curve to form independent data segments, i.e., fluctuation interval data. This data includes the start time point, the end time point, and the thickness values ​​at all sampling times within the interval. Subsequently, each measurement time point in the fluctuation interval data is time-synchronized and matched with the scanning trajectory log recorded by the laser thickness gauge during the preparation process. Using a unified clock reference in the central control system, the spatial coordinates of the laser thickness gauge probe at the same timestamp are found, including the X, Y, and Z three-dimensional coordinates and scanning angle information, forming an ordered set of data pairs consisting of "time point - coordinate value," i.e., a fluctuation point table. Each row in this table records the acquisition time of an abnormal measurement point and its corresponding physical location. For example, thickness decreases were detected at 185 seconds, 188 seconds, and 191 seconds, and their corresponding positions were determined to be (X=32.4mm, Y=15.6mm, Z=2.3mm), (X=33.1mm, Y=15.8mm, Z=2.3mm), and (X=33.8mm, Y=16.0mm, Z=2.3mm) after matching.

[0051] (2) Based on the fluctuation point table, the billet is divided into grids and marked to obtain the detection grid map. The area in the detection grid map that matches the fluctuation point table is expanded to obtain the area to be detected.

[0052] The specific implementation method is as follows: A pre-established three-dimensional geometric model of the billet is called, and rectangular unit grids of fixed size (e.g., 2mm × 2mm) are divided on its outer surface and deposition layer. Each grid unit is assigned a unique number and its center coordinates are recorded, forming a partitioned system covering the entire billet structure, i.e., a detection grid map. Then, the coordinate values ​​in the fluctuation point table are spatially matched with the center coordinates of the grid units in the detection grid map. The nearest neighbor algorithm or coordinate range inclusion judgment method is used to determine the grid unit where each fluctuation point is located, and these grids are marked as "abnormal associated grids." Then, starting from each abnormal associated grid, one or more grid units are extended upwards, downwards, left, right, and to the adjacent layers in the vertical deposition direction. The extension range is set according to the empirical radius of the material's heat-affected zone. For example, setting the extension radius to 3mm ensures that the surrounding area potentially affected by thermal disturbances is also included in the detection range. Finally, all original abnormal grids and their extended neighboring grids are merged into a continuous spatial region, which is the region to be detected. For example, when preparing functionally graded titanium alloy / ceramic materials, if the fluctuation point table shows that multiple abnormal points are concentrated in the third transition zone (X=32.4–33.8mm, Y=15.6–16.0mm) between the 185th and 191st seconds, the system locates the grids numbered G3-12, G3-13, and G4-12 as abnormally associated grids in the detection grid map. Then, it expands a ring of grids around them to form a rectangular area containing nine grids from G2-11 to G4-14 as the detection area. This area will be used as the spatial marker result output by the subsequent coordinate mapping tool to guide the sampling device to accurately detect the sample.

[0053] 2. The phase detection sample is scanned by X-ray diffraction to obtain the diffraction intensity spectrum. The positions of the characteristic peaks in the diffraction intensity spectrum are identified and the coordinate values ​​of the characteristic peaks are obtained.

[0054] The specific implementation method is as follows: the phase detection sample is fixed on the sample stage of the X-ray diffractometer, its surface is adjusted to be perpendicular to the incident X-ray beam, the scanning angle is set to a 2θ angle ranging from 10° to 90°, and Cu-Kα radiation source is used to irradiate it in continuous scanning mode. The detector synchronously records the X-ray intensity signal at different diffraction angles, generating the original spectrum with the diffraction angle as the horizontal axis and the intensity value as the vertical axis, i.e., the diffraction intensity spectrum. Then, the peak recognition algorithm is used to smooth and filter the spectrum and perform baseline correction, detect all local maxima points that are higher than three standard deviations above the background noise, and extract their corresponding diffraction angle values ​​as characteristic peak coordinate values.

[0055] For example, distinct peaks appear at 2θ = 36.2°, 42.3°, and 61.4°. The characteristic peak coordinates are matched with known phase standard diffraction data using a standard card matching tool to obtain phase-matching results. These results are then sorted by content percentage to obtain regional phase composition data. The specific execution method is as follows: the extracted characteristic peak coordinates are input into the standard card matching tool, which calls PDF (Powder Diffraction)... The standard diffraction data of titanium alloys, ceramics and their compounds in the File database were compared one by one using the Hannawalt search method or the FOM value sorting method to find all candidate phases with a deviation of less than 2θ±0.2° from the measured peak position. For example, it was found that 36.2° corresponds to the (111) plane of TiC, 42.3° corresponds to the (101) plane of α-Ti, and 61.4° corresponds to the (002) plane of Ti2AlC. The system matched the products based on the number of matching peaks, intensity matching degree and crystal plane coverage. Then, it combined the Rietveld full spectrum fitting method to quantitatively calculate the relative content of each matching phase in the sample. Finally, the results were sorted from high to low content, such as TiC accounting for 58%, α-Ti accounting for 32%, and Ti2AlC accounting for 10%. This sorting result is the regional phase composition data.

[0056] For example, when preparing titanium alloy / ceramic functionally graded materials, if the abnormal fluctuation segment appears in the third layer transition region, the above process will detect the presence of an unexpected Ti3Al phase in the regional phase composition data, accounting for 15%, indicating that a non-equilibrium reaction has occurred in the region to be tested. This data will serve as the direct basis for subsequent ultrasonic flaw detector scanning of internal defects.

[0057] Furthermore, the phase detection sample is scanned using an X-ray diffractometer to obtain a diffraction intensity spectrum, specifically including: (1) The phase detection sample is continuously scanned by an X-ray diffractometer to obtain the original diffraction data, and the original diffraction data is subjected to background subtraction to obtain the net diffraction data.

[0058] The specific implementation method is as follows: A phase detection sample, cut from the preform during the functionally graded material preparation process and surface-polished, is fixed in the sample holder of an X-ray diffractometer. The sample stage is adjusted so that its detection surface is perpendicular to the incident X-ray beam. The scanning mode is set to θ-2θ continuous scanning. The X-ray source is Cu-Kα radiation (wavelength λ=1.5406 Å), the tube voltage is set to 40 kV, and the tube current is set to 30 mA. The detector collects diffraction signals point by point within the 2θ angle range of 10° to 90° in a step size of 0.02° / step, with a dwell time of 0.5 seconds per step. The X-ray count intensity at the corresponding angle is recorded, forming a set of raw data sequence with the 2θ angle as the abscissa and the intensity value as the ordinate, i.e., raw diffraction data. This data contains diffraction signals from the sample crystal structure and continuous background signals caused by scattering from the sample holder, air scattering, and detector background noise. Subsequently, the raw diffraction data is input into the background subtraction module, and the background trend is modeled using a polynomial fitting or iterative smoothing algorithm. For example, a quadratic or cubic polynomial function can be used to fit the non-periodic background distribution, or a Savitzky-Golay filter can be used to identify and separate low-frequency background components. The fitted background curve can be subtracted point by point from the original diffraction data, retaining only the signal part generated by crystal diffraction, thus forming net diffraction data with systematic interference removed.

[0059] (2) Peak enhancement processing is performed on the net diffraction data to obtain enhanced diffraction data, and coordinate transformation is performed on the enhanced diffraction data to obtain the diffraction intensity spectrum.

[0060] The specific implementation method is as follows: The net diffraction data is input into the peak enhancement module, and the data is sharpened by using gradient operators or high-pass filtering algorithms to enhance the contrast between local maxima and neighboring data points. For example, the edge features of the peak are highlighted by using first-order derivative zero-crossing detection or Laplacian filter. At the same time, a signal-to-noise ratio weighted amplification strategy is applied to amplify the peaks above a set threshold (such as three standard deviations) and weaken fluctuations close to the noise level, thereby improving the clarity and recognizability of obvious diffraction peaks and outputting the enhanced data sequence, i.e. enhanced diffraction data. The enhanced diffraction data was then subjected to coordinate transformation. The original data structure, which was indexed by the instrument step number or the original channel number, was rearranged according to the 2θ angle value corresponding to each step. This was converted into a standard coordinate system with the precise diffraction angle (unit: °2θ) as the horizontal axis and the calibrated intensity value (unit: counts) as the vertical axis. At the same time, the data format was converted to the common XY text or CSV format, and metadata such as sample number, scanning parameters and timestamps were embedded. Finally, a visual chart file, namely the diffraction intensity spectrum, was generated that can be used for subsequent characteristic peak identification and standard card comparison.

[0061] For example, in the preparation of functionally graded titanium alloy / ceramic materials, the phase detection sample taken from the grid region numbered G3-12 was scanned to obtain the raw diffraction data. After background subtraction, the net diffraction data was obtained, showing potential peak positions at 2θ=36.2°, 42.3°, and 61.4°. Peak enhancement processing was then applied to make these peaks more prominent. Finally, a standard format diffraction intensity spectrum was generated through coordinate transformation. The spectrum clearly shows multiple significant diffraction peaks, providing a reliable data basis for subsequent identification of TiC, α-Ti, and possible intermetallic compounds.

[0062] Furthermore, coordinate transformation is performed on the enhanced diffraction data to obtain a diffraction intensity spectrum, specifically including: I. Extract diffraction angle values ​​and diffraction intensity values ​​from the enhanced diffraction data in pairs to obtain diffraction data pairs. Then, use the coordinate definition tool to set the horizontal and vertical axis parameters of the diffraction data pairs to obtain the spectral coordinate system.

[0063] The specific implementation method is as follows: After completing the peak enhancement processing of the net diffraction data and generating enhanced diffraction data, the system performs structured parsing on the data sequence, reads the numerical information in each record line by line, and combines the corresponding 2θ angle values ​​of each group with the enhanced intensity values ​​in the form of ordered pairs. For example, the angle 36.2° is combined with its corresponding intensity value of 842 counts to form (36.2, 842), and 42.3° is combined with 1520 counts to form (42.3, 1520), and so on, forming an ordered set containing multiple (2θ, I) combinations, i.e., diffraction data pairs. Then, the diffraction data pairs are input into the coordinate definition tool. According to the standard expression specification of X-ray diffraction patterns, the tool sets the horizontal axis parameter to "diffraction angle (°2θ)", the measurement range covers from the starting scanning angle of 10° to the ending angle of 90°, and the minimum scale unit is 0.5°. The vertical axis parameter is set to "diffraction intensity (counts)", and its range is dynamically adjusted according to the maximum intensity value in the enhanced diffraction data. For example, when the highest intensity is 2000 counts, the upper limit of the vertical axis is set to 2200 counts to retain an appropriate margin. The display attributes such as coordinate axis font, line type, and grid line density are uniformly configured using preset templates. Finally, a two-dimensional rectangular coordinate system that conforms to international standards is constructed, i.e., the spectral coordinate system.

[0064] II. Based on the spectral coordinate system, coordinate points are mapped onto the diffraction data pairs to obtain diffraction coordinate points. Then, continuous lines are connected to the diffraction coordinate points to obtain the diffraction intensity spectrum.

[0065] The specific implementation is as follows: The 2θ values ​​in each diffraction data pair are mapped to the horizontal axis of the spectral coordinate system, and the intensity values ​​are mapped to the vertical axis. The corresponding physical display points are determined in the coordinate system. For example, (36.2, 842) is positioned at the intersection of the horizontal coordinate 36.2 and the vertical coordinate 842, forming discrete graphic markers, which are the diffraction coordinate points. The system uses a vector drawing engine to label all diffraction coordinate points in the spectral coordinate system as small dots or short vertical lines. Then, according to the 2θ angle from smallest to largest, a broken line connection algorithm is used to connect adjacent diffraction coordinate points sequentially with straight line segments, forming a continuously undulating curve. This curve truly reflects the intensity response trend of the material at different diffraction angles. For regions with obvious peaks, the curve shows a sharp bulge, while for flat background regions, it maintains low-amplitude fluctuations. Finally, a visual image file containing complete coordinate axes, data point connections, legend labels, and title information is generated, namely, the diffraction intensity spectrum.

[0066] For example, in the preparation of functionally graded titanium alloy / ceramic materials, the phase detection sample taken from the third transition zone was analyzed. The system extracted multiple sets of diffraction data pairs, including (36.2, 842), (42.3, 1520), and (61.4, 1180), from the enhanced diffraction data. After establishing a spectral coordinate system with "diffraction angle (°2θ)" on the horizontal axis and "diffraction intensity (counts)" on the vertical axis using a coordinate definition tool, each point was mapped to a diffraction coordinate point. A clear diffraction intensity spectrum was generated by continuous line processing. A significant peak appeared near 36.2° in the graph, providing an intuitive basis for subsequent identification of the TiC phase by comparison with standard cards.

[0067] S4. Based on the regional phase composition data, perform an internal defect scan on the billet to obtain an internal defect map of the billet.

[0068] Specifically, an ultrasonic flaw detector is used to scan the internal defects of the billet. The process involves: after X-ray diffraction is used to detect the phase composition of the region corresponding to the abnormal fluctuation segment during the preparation of functionally graded materials and obtaining the regional phase composition data, the location and spatial coordinates of the abnormal phases identified in this data are transmitted to the scanning control system of the ultrasonic flaw detector. This coordinate information is used to determine the internal regions of the billet that require focused inspection. The probe position of the ultrasonic flaw detector is adjusted so that it faces the surface projection position of the region. The ultrasonic emission frequency, pulse period, and gain parameters are preset according to the acoustic characteristics of the billet material. The ultrasonic probe is then activated to emit high-frequency acoustic signals. During propagation within the billet, the sound waves are reflected or scattered when they encounter interfaces with acoustic impedance differences, such as pores, cracks, or interface debonding. The probe receives the echo signals and records their return time, amplitude, and waveform characteristics. The defect depth is calculated through time-domain analysis, and two-dimensional or three-dimensional imaging data is constructed by combining the displacement information of the scanning path. After signal filtering and image reconstruction algorithms, a visual image reflecting the type, size, location, and distribution density of internal defects in the billet—that is, an internal defect map—is generated.

[0069] For example, when preparing functionally graded titanium alloy / ceramic materials, if the regional phase composition data indicates the presence of a large number of brittle intermetallic compounds in the transition layer between the titanium alloy and the ceramic, an ultrasonic flaw detector can perform a high-resolution scan of this region. This may identify microcrack clusters or local pore zones distributed along the interface, and mark their spatial distribution range as highlighted areas in the defect map, providing a direct basis for subsequent quality assessment based on this map.

[0070] Furthermore, the internal defect map is obtained in the following way: 1. Analyze the differences in phase distribution in the regional phase composition data to mark areas where the phase composition does not conform to the gradient design, obtain key inspection areas, and divide the key inspection areas into grid cells to obtain the flaw detection grid.

[0071] The specific implementation method is as follows: After the X-ray diffractometer completes the analysis of the phase detection samples during the preparation of functionally graded materials and generates regional phase composition data, the phase types and their content ratios at each detection point in the data are compared layer by layer with the preset gradient design model. The design model stipulates that the transition region from the titanium alloy layer to the ceramic layer should successively show a continuous change of Ti, Ti+TiC mixed phase, and TiC as the main phase. The system identifies the regions where the actual phase composition deviates from the design path through numerical comparison algorithms. For example, if the α-Ti content exceeds 15% or an unexpected Ti2AlN phase is detected in the third layer, which should be dominated by TiC, such regions are identified as phase abnormalities. These regions are marked with their spatial coordinates on the billet, forming physical regions with clear boundaries, i.e., key detection areas. Subsequently, the key detection area is locally meshed in the three-dimensional model interface of the control system. The size of the mesh unit is set according to the effective detection diameter of the ultrasonic probe. For example, a 2mm×2mm square unit is used for division. Each unit is assigned a unique number and its center coordinates and the number of layers are recorded to form a discrete detection framework covering the entire key detection area, namely the flaw detection mesh.

[0072] 2. Ultrasonic reflection signals are collected from each unit in the flaw detection grid using an ultrasonic flaw detector to obtain the unit reflection signal. Abnormal reflection waves in the unit reflection signal are then identified to obtain defect signal data.

[0073] The specific implementation method is as follows: The ultrasonic flaw detector probe is mounted on a programmable motion platform. The control system guides the probe to move directly above each grid cell according to the numbering order of the flaw detection grid, maintaining vertical incidence and applying a coupling agent to ensure effective sound wave transmission. The transmission pulse frequency is set to 5MHz and the pulse period is 1μs. The transmitting circuit is activated to generate a high-frequency voltage to excite the piezoelectric crystal and emit an ultrasonic beam into the blank. When the sound wave encounters an interface with discontinuous acoustic impedance such as pores, cracks, or interface debonding during propagation, it is reflected. The probe receives the echo signal and converts it into an electrical signal. After amplification and analog-to-digital conversion, it forms the original waveform data containing time, amplitude, and waveform characteristics, i.e., the unit reflection signal. Subsequently, time-domain analysis is performed on each group of unit reflection signals to identify additional reflected waves located before the bottom wave and with an amplitude exceeding twice the standard deviation of the baseline noise. It is determined whether they have crack characteristic waveforms (such as double peaks or sawtooth wavefronts) or pore characteristics (wide pulses, low frequencies). The reflected waves that meet the defect characteristics are extracted and marked with their occurrence time, location number, and peak intensity to form a structured dataset, i.e., defect signal data.

[0074] Furthermore, ultrasonic reflection signals are acquired from each cell in the flaw detection grid using an ultrasonic flaw detector to obtain the cell reflection signal, including: (1) Based on the unit size and arrangement of the flaw detection grid, the scanning path of the ultrasonic flaw detector is planned to obtain the grid adaptation path, and the probe is driven to move according to the grid adaptation path to perform the first signal acquisition of each unit in the flaw detection grid and obtain the initial reflection signal.

[0075] The specific implementation is as follows: After dividing the key inspection areas and generating the flaw detection grid during the functionally graded material preparation process, the system reads the geometric parameters of each grid cell, including cell size (e.g., 2mm × 2mm), arrangement order (arranged continuously in rows or columns), and spacing between adjacent cells. Combined with the effective beam coverage diameter of the ultrasonic flaw detector probe (typically 1.8–2.2mm), a path optimization algorithm is used to plan a continuous scanning trajectory with no repetition, no omissions, and the shortest travel distance—the grid adaptation path. This path starts from the initial cell and passes through the center points of all grid cells sequentially, ensuring the probe stays above each cell for sufficient time to complete signal acquisition. The control system then moves the grid adaptation path... The motion commands are converted to drive the ultrasonic probe mounted on the multi-axis robotic arm to move point by point along a predetermined trajectory. When it reaches the center of each grid cell, the probe automatically positions itself and keeps vertically aligned with the surface of the blank. An appropriate amount of coupling agent is applied to ensure effective transmission of the sound wave. Then, the transmission parameters are set to a frequency of 5MHz, a pulse width of 1μs, and an excitation voltage of 100V. The transmitting circuit is started to generate ultrasonic pulses and the receiving channel is opened simultaneously to record the echo signals reflected back from the internal interface of the material, including bottom waves, interface waves, and possible defect reflection waves. A set of time-amplitude waveform data corresponding to the grid cell is formed, which is the initial reflection signal. The initial reflection signals of all cells are stored in order of their numbers and are labeled with positions.

[0076] (2) Mark the grid cells in the initial reflection signal that have not been effectively acquired to obtain a blind zone cell table. Based on the blind zone cell table, adjust the angle and transmission power of the probe and perform secondary signal acquisition on the blind zone cells in the flaw detection grid to obtain the complete reflection signal. Integrate the initial reflection signal and the complete reflection signal according to the grid cell number to obtain the cell reflection signal.

[0077] The specific implementation is as follows: After completing the initial scan, the system performs validity judgment on each group of initial reflected signals. The judgment criteria include whether there is an identifiable bottom wave signal, whether the signal-to-noise ratio is greater than a preset threshold (e.g., 10dB), and whether the waveform is severely distorted or completely flat without fluctuations. If the signal of a certain unit does not meet the above conditions, it is determined that no valid signal has been acquired, and its number is recorded in a structured list, namely the blind zone unit table. For example, G3-13 and G4-12 are marked as blind zone units. Subsequently, the system calls the blind zone unit table, reconfigures the detection parameters for each unit, adjusts the probe tilt angle (e.g., from 0° to 5°–10° oblique incidence) to change the sound wave propagation path, avoids the sound beam deflection caused by surface curvature or roughness, and increases the transmission power to 150V to enhance the penetration capability. At the same time, the signal acquisition time window is extended, and the probe is driven to move directly above the blind zone unit again to perform secondary signal acquisition, obtain new echo data, that is, complete the reflected signal. Finally, all completed reflection signals are matched with their corresponding grid cell numbers and integrated with the original initial reflection signals. Valid signals acquired initially are retained first, and invalid data segments are replaced only with completed signals. This ensures that each group of cell reflection signals contains complete and reliable information, ultimately forming a unified dataset covering the entire flaw detection grid without missing data—the cell reflection signal. For example, in the preparation of titanium alloy / ceramic functionally graded materials, when the initial scan revealed that cell G3-13 had severe signal attenuation due to a local surface oxide layer and no effective echo, the system included it in the blind zone cell table. After adjusting the probe angle to 7° and increasing the transmission power, a clear reflection wave was successfully acquired. After completing the data, it was integrated into a complete cell reflection signal, providing reliable input for subsequent defect identification.

[0078] 3. Classify the defect signal data according to the defect signal intensity to obtain classified defect data, and draw the internal defect map of the billet containing the defect location and intensity level based on the positional correspondence between the classified defect data and the flaw detection grid.

[0079] The specific implementation method is as follows: the peak intensity value in the defect signal data is divided into three levels according to a preset threshold. For example, 0-30% of the full scale is level one (slight), 30%-60% is level two (moderate), and above 60% is level three (severe). The system assigns each defect signal to the corresponding level according to its intensity, generating graded defect data containing grid number, defect type, intensity level, and depth information. Then, the image generation module is called to construct a two-dimensional planar map based on the layout of the flaw detection grid. The horizontal and vertical axes correspond to the XY positions of the grid. Different colors or filling patterns are used to represent different intensity levels. For example, green dots represent level one, yellow triangles represent level two, and red squares represent level three. Finally, a visual image that intuitively reflects the distribution of defects in each location within the key inspection area is output, namely, the defect map of the billet's interior.

[0080] For example, when preparing functionally graded titanium alloy / ceramic materials, if the regional phase composition data shows an abnormal Ti3Al phase in the third transition zone, the system marks it as a key detection area and divides it into 9 flaw detection grids. After scanning, a strong reflected wave with an amplitude of 75% was found in unit G3-13, which was identified as a through crack signal and classified as a level 3 defect. Its location was marked with a red square in the defect map inside the billet, providing a direct basis for subsequent quality assessment.

[0081] S5. Based on the internal defect map of the billet, the quality of the billet is evaluated to obtain the quality grade determination result.

[0082] The specific implementation method is as follows: After the ultrasonic flaw detector completes the scanning of internal defects of the billet during the preparation of functionally graded materials and generates an internal defect map, the map data is imported into the quality assessment system. The system segments and extracts features from the abnormal areas in the map according to a preset defect identification algorithm, including parameters such as the area ratio of defects, maximum equivalent size, spatial distribution density, and whether they cross critical interface layers. At the same time, it sets multi-level judgment thresholds in conjunction with the design specifications of functionally graded materials. For example, defects with an area ratio of less than 1% and not connected to an interface are judged as minor defects; defects with an area ratio between 1% and 3% or with local dense distribution are judged as moderate defects; and defects with an area ratio exceeding 3% or with through-cracks are judged as severe defects. The system compares the extracted defect features with the thresholds at each level, calculates the overall quality score by weighting various indicators, and maps it to the corresponding quality level, such as Level 1 (qualified), Level 2 (downgraded for use), or Level 3 (scrapped). Finally, a structured report containing the defect location, level classification, and judgment basis is output, which is the quality level judgment result.

[0083] For example, when preparing functionally graded titanium alloy / ceramic materials, if the defect map of the billet shows that there are discrete pores accounting for 2.5% of the total in the third to fifth layer interface region where the titanium alloy transitions to ceramic, and there are no through cracks, the system judges it as a medium defect and classifies it into a second-level quality grade. This indicates that the billet needs to undergo subsequent hot isostatic pressing treatment before it can be put into use, thus completing the quality judgment process based on a complete detection chain.

[0084] Furthermore, the quality grade determination result is obtained in the following way: 1. Extract the defect location and strength level from the defect map inside the billet to obtain defect feature data, and classify and statistically analyze the defect feature data according to the preset defect type to obtain defect classification statistics results.

[0085] The specific implementation method is as follows: After the ultrasonic flaw detector completes the scanning of internal defects in key inspection areas during the preparation of functionally graded materials and generates an internal defect map of the billet containing information on defect location and strength level, the system performs image analysis processing on the map, identifies defect points marked with different colors or symbols in the image, reads the grid number, spatial coordinates (X, Y, Z), depth position, and strength level (e.g., level 1, level 2, level 3) corresponding to each defect point, and integrates this information into a set of structured data, namely defect feature data. Subsequently, according to the preset defect type classification rules, the defects are classified into types such as porosity, cracks, inclusions, and interface debonding according to their morphological characteristics and distribution patterns. The classification criteria include the spatial continuity of the defect signal (e.g., linear distribution is judged as cracks), depth concentration (e.g., dense point defects near the surface are judged as porosity), and whether it crosses the material gradient interface (e.g., defects penetrating the titanium alloy and ceramic layers are judged as interface debonding). The system performs type matching and labeling on each record in the defect feature data, and performs quantity statistics and area ratio calculation according to type. For example, a total of 3 crack-type defects were found, accounting for 2.1% of the total projected area of ​​the key inspection area, and a total of 7 porosity defects were found, accounting for 1.5% of the total area. Finally, a summary data table containing information on the number, distribution density, maximum equivalent size and proportion of each type of defect was generated, which is the defect classification statistics result.

[0086] 2. Quantify and score the statistical results of defect classification using defect rating standards to obtain individual defect scores. Then, perform weighted summation based on the individual defect scores to obtain a comprehensive quality score. Finally, compare the comprehensive quality score with the preset quality level threshold to obtain the quality level determination result.

[0087] The specific implementation method is as follows: A pre-defined defect rating standard table is invoked. This table specifies the scoring rules for various types of defects. For example, cracks are penalized 5 points per millimeter of length, porosity is penalized 2 points per 0.5% of area, and interface debonding is penalized 8 points per occurrence. The system calculates the corresponding score based on the parameters in the defect classification statistics, generating individual defect scores such as crack score and porosity score. Then, a weighted sum is calculated based on the weight coefficients of each defect's impact on material properties. For example, the weight for cracks is set to 0.4, for porosity to 0.3, and for interface debonding to 0.3. The calculation formula is: Overall Quality Score = Crack Individual Score × 0.4 + Porosity Individual Score × 0.3 + The interface de-adhesion score is multiplied by 0.3 to obtain the final score. Then, the overall quality score is compared with the preset quality level threshold. The threshold range is set as follows: a score ≥90 is Level 1 (qualified), 80–89 is Level 2 (downgraded for use), and <80 is Level 3 (scrapped). The system automatically matches the range and outputs the corresponding conclusion, i.e., the quality level judgment result.

[0088] For example, when preparing titanium alloy / ceramic functional graded materials, if the defect classification statistics show that there are 2 microcracks (deduct 10 points), 5 pores (deduct 6 points), and 1 interface debonding (deduct 8 points), the weighted calculation results in a comprehensive quality score of 82 points, which falls within the 80-89 point range. The system judges it to be of secondary quality level, indicating that the billet needs to be repaired before it can be put into use.

[0089] While the specific embodiments of the present invention have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of the present invention are still within the scope of protection of the present invention.

Claims

1. A quality inspection method in the preparation process of functionally graded materials, characterized in that, Includes the following steps: The temperature of the billet was collected during the preparation of functionally graded materials to obtain the temperature distribution data of the billet; Real-time thickness measurement of the billet is performed based on the billet temperature distribution data to obtain the billet thickness variation curve; Identify abnormal fluctuation segments in the billet thickness variation curve, perform phase detection on the physical regions of the billet with abnormal fluctuation segments, and obtain regional phase composition data; Based on the regional phase composition data, the internal defects of the billet are scanned to obtain the internal defect map of the billet. The quality of the billet is assessed based on the internal defect map, and the quality grade is determined.

2. The quality inspection method in the preparation process of functionally graded materials as described in claim 1, characterized in that, The temperature of the billet was collected using an infrared thermometer, and the temperature distribution data of the billet was obtained in the following way: The outer surface and cross-section of the billet in the process of preparing functionally graded materials are divided into regions using a mesh generation tool to obtain a billet temperature measurement mesh. The billet temperature measurement mesh is then numbered according to the gradient region type to obtain a numbered temperature measurement mesh. Based on a numbered temperature measurement grid, an infrared thermometer is used to collect the temperature of the billet at fixed points to obtain the temperature distribution data of the billet.

3. The quality inspection method in the preparation process of functionally graded materials as described in claim 1, characterized in that, The thickness of the billet was measured in real time using a laser thickness gauge, and the thickness variation curve of the billet was obtained in the following way: Based on the temperature distribution data of the billet, a temperature gradient analysis is performed on the billet to obtain a temperature gradient map. Based on the temperature gradient map, the billet is divided into key areas to obtain key monitoring areas. In the key monitoring area, a laser thickness gauge is used to measure the thickness of the billet in real time to obtain real-time thickness data. The real-time thickness data is then time-aligned to obtain time-series thickness data. The change trend of the time-series thickness data is calculated to obtain the thickness change characteristics, and the thickness change curve of the billet is generated based on the thickness change characteristics.

4. The quality inspection method in the preparation process of functionally graded materials as described in claim 1, characterized in that, X-ray diffraction was used to detect the phase composition of the physical region in the billet where abnormal fluctuations occurred. The phase composition data of the region was obtained in the following way: The spatial location of the billet corresponding to the abnormal fluctuation section is marked by a coordinate mapping tool to obtain the area to be tested, and the area to be tested is sampled to obtain a phase detection sample. The phase detection sample was scanned by X-ray diffraction to obtain the diffraction intensity pattern, and the positions of the characteristic peaks in the diffraction intensity pattern were identified to obtain the coordinate values ​​of the characteristic peaks. The characteristic peak coordinates are matched with known phase standard diffraction data using a standard card comparison tool to obtain phase matching results. The phase matching results are then sorted by content percentage to obtain regional phase composition data.

5. The quality inspection method in the preparation process of functionally graded materials as described in claim 4, characterized in that, The spatial location of the billet corresponding to the abnormal fluctuation segment is marked using a coordinate mapping tool to obtain the area to be detected, specifically including: The abnormal fluctuation segment in the billet thickness variation curve is segmented to obtain fluctuation interval data. The measurement time points in the fluctuation interval data are matched with the scanning positions of the laser thickness gauge to obtain a fluctuation point location table. Based on the fluctuation point table, the billet is divided into grids and marked to obtain a detection grid map. The regions in the detection grid map that match the fluctuation point table are then expanded to obtain the regions to be detected.

6. The quality inspection method in the preparation process of functionally graded materials as described in claim 4, characterized in that, X-ray diffraction was used to perform diffraction scanning on the phase detection sample to obtain the diffraction intensity pattern, specifically including: The sample for phase detection was continuously scanned using an X-ray diffractometer to obtain raw diffraction data. The raw diffraction data was then processed by background subtraction to obtain net diffraction data. Peak enhancement processing is performed on the net diffraction data to obtain enhanced diffraction data, and coordinate transformation is performed on the enhanced diffraction data to obtain the diffraction intensity spectrum.

7. The quality inspection method in the preparation process of functionally graded materials as described in claim 6, characterized in that, The enhanced diffraction data is subjected to coordinate transformation to obtain the diffraction intensity pattern, specifically including: The diffraction angle value and diffraction intensity value in the enhanced diffraction data are extracted in pairs to obtain diffraction data pairs. The horizontal and vertical axis parameters of the diffraction data pairs are set by the coordinate definition tool to obtain the spectrum coordinate system. Based on the spectral coordinate system, the diffraction data pairs are mapped to obtain diffraction coordinate points, and the diffraction coordinate points are continuously connected to obtain the diffraction intensity spectrum.

8. The quality inspection method in the preparation process of functionally graded materials according to claim 1, characterized in that, The internal defect map of the billet is obtained by scanning the billet with an ultrasonic flaw detector in the following way: The differences in phase distribution in the regional phase composition data are analyzed to mark areas where the phase composition does not conform to the gradient design, thus obtaining key inspection areas. The key inspection areas are then divided into grid cells to obtain the flaw detection grid. The ultrasonic flaw detector collects ultrasonic reflection signals from each cell in the flaw detection grid to obtain cell reflection signals. Abnormal reflection waves in the cell reflection signals are then identified to obtain defect signal data. The defect signal data is classified according to the defect signal intensity to obtain classified defect data. Based on the positional correspondence between the classified defect data and the flaw detection grid, an internal defect map of the billet containing the defect location and intensity level is drawn.

9. The quality inspection method in the preparation process of functionally graded materials according to claim 8, characterized in that, The ultrasonic reflection signal of each cell in the flaw detection grid is acquired using an ultrasonic flaw detector, and the cell reflection signal is obtained, specifically including: Based on the cell size and arrangement of the flaw detection grid, the scanning path of the ultrasonic flaw detector probe is planned to obtain the grid adaptation path. The probe is then driven to move according to the grid adaptation path to perform the first signal acquisition on each cell in the flaw detection grid and obtain the initial reflection signal. The grid cells in the initial reflection signal that did not acquire a valid signal are marked to obtain a blind zone cell table. Based on the blind zone cell table, the probe angle and transmission power are adjusted, and a second signal acquisition is performed on the blind zone cells in the flaw detection grid to obtain the complete reflection signal. The initial reflection signal and the complete reflection signal are integrated according to the grid cell number to obtain the cell reflection signal.

10. The quality inspection method in the preparation process of functionally graded materials according to claim 1, characterized in that, The quality grade determination result is obtained in the following way: The defect location and strength level are extracted from the defect map of the billet to obtain defect feature data. The defect feature data are then classified and statistically analyzed according to the preset defect type to obtain defect classification and statistical results. The defect classification statistics are quantitatively scored using defect rating standards to obtain individual defect scores. A weighted sum is then calculated based on these individual defect scores to obtain a comprehensive quality score. Finally, the comprehensive quality score is compared with a preset quality level threshold to obtain the quality level determination result.

Images