A motorcycle shell high-pressure casting mold defect detection system and method

Abstract

The present invention relates to the field of defect detection technology, specifically a defect detection system and method for high-pressure casting molds for motorcycle shells, which enables high-precision detection of cold shut defects in thin-walled motorcycle shell molds. The system comprises: acquiring and processing product data of the product to be inspected to obtain static defect data; deploying multiple sensors in different areas to collect material data; analyzing the material data using an abnormal region detection model to identify abnormal material regions and mark the severity of the abnormality; screening out a first abnormal region based on a preset material abnormality threshold; performing dynamic defect detection by calculating a dynamic defect score for the first abnormal region to obtain dynamic defect data; and performing comprehensive defect analysis by comparing static and dynamic defect data.

Description

Technical Field

[0001] The present invention relates to the technical field of defect detection, and in particular to a system and method for detecting defects in a high-pressure casting die for a motorcycle housing. Background Art

[0002] Defect detection for high-pressure casting molds primarily focuses on the following aspects: Physically, testing focuses on hardness, density, microstructure, grain size, and the distribution of internal defects (such as pores and cracks); chemically, attention is paid to material composition, alloying element distribution, oxidation level, and impurity content. Commonly used detection methods include X-ray imaging, ultrasonic testing, scanning electron microscopy (SEM), energy dispersive spectroscopy (EDS), and spectral analysis. While these methods have achieved breakthroughs in detection accuracy and application scope, they still have certain limitations, especially when dealing with complex defect types.

[0003] Traditional defect detection relies primarily on manual visual inspection or single detection methods, such as ultrasonic testing, X-ray testing, and metallographic analysis. These methods suffer from low detection efficiency, high missed detection rates, and insufficient real-time performance, making them unable to meet the modern manufacturing industry's demand for efficient and intelligent detection. Furthermore, because the casting process involves multiple process parameters, such as temperature, flow rate, pressure, and solidification time, defects are often caused by the coupling of multiple factors. The lack of in-depth analysis and traceability of the causes of defects limits the scope for process optimization and quality improvement. Therefore, there is an urgent need for a mold defect detection system that can achieve high-precision detection of cold shut defects in thin-walled motorcycle housing molds, quantitatively evaluate the metal flow optimization effect, and provide a scientific basis for mold design and process parameter optimization.

[0004] Therefore, a system and method for detecting defects in a high-pressure casting mold for a motorcycle housing are proposed. Summary of the Invention

[0005] The present invention aims to provide a defect detection system and method for high-pressure casting molds for motorcycle shells, enabling high-precision and high-efficiency cold shut defect detection during the manufacturing process of thin-walled motorcycle shell molds. The system comprises a product data acquisition module, a static defect detection module, a material data acquisition module, an abnormal region detection module, a dynamic defect detection module, and a comprehensive defect analysis module. The method comprises: acquiring product data of the product to be inspected; processing the product data to obtain static defect data; placing multiple sensors in a first region, a second region, and a third region to collect material data; analyzing the material data using an abnormal region detection model to identify abnormal material regions and mark the severity of the material abnormality; screening out the first abnormal region based on a preset material abnormality threshold; performing dynamic defect detection by calculating a dynamic defect score for the first abnormal region to obtain dynamic defect data; and performing comprehensive defect analysis by comparing the static defect data with the dynamic defect data.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] A motorcycle shell high-pressure casting mold defect detection system, comprising:

[0008] A product data acquisition module, used to acquire product data of the product to be tested; the product data includes: hardness, color, shape and ultrasonic data;

[0009] A static defect detection module, configured to process the product data to obtain static defect data;

[0010] A material data acquisition module is configured to arrange a plurality of sensors in the first area, the second area, and the third area to collect material data; the material data includes temperature, pressure, flow rate, and solidification time;

[0011] An abnormal region detection module is used to analyze the material data using an abnormal region detection model, identify the material abnormal region, and mark the severity of the material abnormality; and screen out the first abnormal region among the material abnormal regions according to a preset material abnormality threshold;

[0012] a dynamic defect detection module, configured to perform dynamic defect detection by calculating a dynamic defect score of the first abnormal area to obtain dynamic defect data;

[0013] The defect comprehensive analysis module is used to perform defect comprehensive analysis by comparing the static defect data with the dynamic defect data.

[0014] Preferably, the static defect detection module includes: a product data preprocessing unit, a hardness detection unit, a color detection unit, a shape detection unit, an ultrasonic detection unit and a static defect data acquisition unit;

[0015] The product data preprocessing unit preprocesses the product data; the preprocessing includes denoising and normalization;

[0016] The hardness detection unit identifies local hardness defects by analyzing the hardness;

[0017] The color detection unit identifies color abnormal areas by calculating the color distribution histogram and comparing it with the product standard color;

[0018] The shape detection unit marks the deformed area by comparing the shape with the standard shape of the product;

[0019] The ultrasonic detection unit processes the ultrasonic data through signal processing technology to identify internal defects of the product;

[0020] The static defect data acquisition unit comprehensively identifies the local hardness defect, the color abnormality area, the deformation area and the internal defect, identifies the static defect of the product according to the preset static defect judgment standard, and generates the static defect data, including the static defect type and the static defect location.

[0021] Preferably, the first area is a gate area; the second area is a thin-wall area; and the third area is an end filling area.

[0022] The various sensors include: thermal imaging sensors, ultrasonic sensors and pressure sensors.

[0023] Preferably, the abnormal region detection model includes: a material data preprocessing unit, an abnormal feature recognition unit, an abnormal region marking unit and a first abnormal region screening unit;

[0024] Wherein, the material data preprocessing unit preprocesses the material data;

[0025] The abnormal feature recognition unit analyzes the pre-processed material data to identify abnormal features; the abnormal features include: flow obstruction, turbulence enhancement and uneven filling;

[0026] The abnormal region marking unit identifies an abnormal pattern in the material data according to the abnormal feature and marks the abnormal region;

[0027] The first abnormal region screening unit performs a graded evaluation on the abnormal region based on the preset material abnormality threshold, and screens the abnormal region according to a result of the graded evaluation to obtain the first abnormal region.

[0028] Preferably, the formula for the dynamic defect score is:

[0029] ;

[0030] in, Scoring dynamic defects; is the temperature term weight; is the temperature regulation coefficient; is the temperature of the material in the abnormal area; is the weight of the velocity change rate term; is the flow rate change rate adjustment coefficient; is the velocity change rate in the abnormal area; is the weight of filling pressure item; is the local filling pressure in the abnormal area; is the simulated filling pressure; is the filling pressure adjustment coefficient; is the weight of solidification time item; is the coagulation time of the abnormal area; is the simulated solidification time; is the coagulation time adjustment coefficient;

[0031] The dynamic defect is identified based on the dynamic defect score according to a preset dynamic defect threshold to generate the dynamic defect data; the dynamic defect data includes: dynamic defect severity and dynamic defect location.

[0032] Preferably, the defect comprehensive analysis module includes: a defect position comparison unit, a defect influencing factor correlation analysis unit and a defect cause comprehensive analysis unit;

[0033] The defect position comparison unit compares the static defect position in the static defect data with the dynamic defect position in the dynamic defect data, identifies the overlapping area between the two, and marks it as the cold shut defect position;

[0034] The defect influencing factor correlation analysis unit calculates the correlation between the product data and the material data in the cold shut defect location and the defect;

[0035] The defect cause comprehensive analysis unit identifies the most likely cause of the defect based on the correlation and generates a detailed defect cause analysis report.

[0036] Preferably, a method for detecting defects in a high-pressure casting mold for a motorcycle housing comprises:

[0037] Obtain product data of the product to be tested; the product data includes: hardness, color, shape and ultrasonic data;

[0038] Processing the product data to obtain static defect data;

[0039] Arrange a variety of sensors in the first area, the second area, and the third area to collect material data; the material data includes: temperature, pressure, flow rate, and solidification time;

[0040] Analyzing the material data using an abnormal region detection model, identifying material abnormal regions, and marking the severity of the material abnormality; screening out a first abnormal region among the material abnormal regions according to a preset material abnormality threshold;

[0041] Performing dynamic defect detection by calculating a dynamic defect score of the first abnormal area to obtain dynamic defect data;

[0042] A comprehensive defect analysis is performed by comparing the static defect data and the dynamic defect data.

[0043] Compared with the prior art, the present invention has the following beneficial effects:

[0044] 1. The present invention uses a static defect detection module to conduct a comprehensive analysis of the hardness, color, shape and ultrasonic data of the product, and can accurately identify initial defects on the surface and inside of the product, such as local hardness anomalies, color differences, deformations, and internal pores or inclusions. The acquisition of static defect data can not only provide a basis for the overall quality assessment of the product, but also provide reference data for subsequent dynamic defect detection. When dynamic defects are subsequently analyzed, static defect data can be used to compare defect locations and defect types, further enhancing the accuracy of defect identification. In addition, static defect detection can be completed as soon as the product is formed, which helps to quickly screen out products with obvious defects, reduce unnecessary subsequent testing, and improve production efficiency.

[0045] 2. The present invention proposes an abnormal area detection model that analyzes the changes in material data during the high-pressure casting process to identify areas of abnormal material flow, including flow obstruction, increased turbulence, and uneven filling, and marks the severity of the abnormal area. The model can predict potential defect areas in the casting process in advance, provide precise target areas for subsequent dynamic defect detection, and make detection more efficient. At the same time, the division of abnormal areas can help to optimize the casting process in the future, such as adjusting the pouring temperature, flow rate, and mold design to reduce the occurrence of defects. By screening the abnormal areas and performing in-depth dynamic defect detection only on the first abnormal area, the detection efficiency and the accuracy of subsequent dynamic defect detection can be effectively improved.

[0046] 3. The dynamic defect detection method of the present invention identifies possible cold shut defect areas by calculating dynamic defect scores, while static defect detection provides data such as surface deformation and hardness change, which can serve as additional verification. By comparing the dynamic defect locations obtained by dynamic defect detection with the static defect locations in the static defect data, the cold shut defect can be accurately located, reducing the false positive rate. Furthermore, this combined dynamic and static method improves the automation level of cold shut defect detection, avoiding the potential limitations of a single detection method, making defect analysis more comprehensive and accurate, and providing a reliable theoretical basis for subsequent optimization of production processes and reducing the occurrence of cold shut defects. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 A structural diagram of a defect detection system for a motorcycle housing high-pressure casting mold provided by an embodiment of the present invention;

[0048] Figure 2 A flowchart of a method for detecting defects in a high-pressure casting mold for a motorcycle housing provided by an embodiment of the present invention;

[0049] Figure 3 A diagram showing the working principle of a static defect detection module provided by an embodiment of the present invention;

[0050] Figure 4 A diagram illustrating the working principle of the abnormal region detection model provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0051] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0052] As a critical external structural component, the production quality of motorcycle housings directly impacts the performance and safety of the entire vehicle. Due to the thin wall thickness of motorcycle housings, typically only 2-3mm, improper mold flow path design can lead to defects such as cold shut during the manufacturing process. Cold shut defects not only significantly degrade the material's mechanical properties but can also affect safety and reduce product reliability. Therefore, detecting and preventing cold shut defects is crucial to improving motorcycle housing quality and production efficiency.

[0053] This invention proposes a system and method for detecting defects in high-pressure casting molds for motorcycle housings, achieving high-precision and high-efficiency detection of cold shut defects in thin-walled motorcycle housing molds. To demonstrate the effectiveness of the present method in accurately and efficiently detecting cold shut defects in thin-walled motorcycle housing molds, the following two examples illustrate the effectiveness of the present invention.

[0054] Example 1

[0055] In the embodiment of this application, the method proposed in this invention is used to detect the cold shut defects of thin-walled motorcycle shell molds with high precision and high efficiency, and the process of providing a scientific basis for mold design and process parameter optimization is described in detail. This embodiment of the application is aimed at the defect detection of the high-pressure casting mold of the motorcycle shell of a type A motorcycle of a certain motorcycle technology company. Figure 1 The content describes in detail the defect detection process of high pressure casting mold for this type of motorcycle shell; Figure 1 This is a specific structural diagram of the system proposed in the present invention. The system includes: a product data acquisition module, a static defect detection module, a material data acquisition module, an abnormal area detection module, a dynamic defect detection module and a comprehensive defect analysis module. Figure 2 This is a specific flow chart of the method proposed in the present invention, which includes: obtaining product data of the product to be tested; processing the product data to obtain static defect data; arranging multiple sensors in the first area, the second area and the third area to collect material data; analyzing the material data using the abnormal area detection model, identifying the material abnormal area, and marking the severity of the material abnormality; screening out the first abnormal area according to the preset material abnormality threshold; performing dynamic defect detection by calculating the dynamic defect score of the first abnormal area to obtain dynamic defect data; and performing comprehensive defect analysis by comparing the static defect data and the dynamic defect data. Figure 1 and Figure 2 The following describes the contents:

[0056] A motorcycle shell high-pressure casting die defect detection system includes:

[0057] A product data acquisition module, used to acquire product data of the product to be tested; the product data includes: hardness, color, shape and ultrasonic data;

[0058] Specifically, a digital hardness tester is used to perform multi-point hardness measurement on multiple key positions of the motorcycle shell to obtain hardness data; the key positions include: thin-walled areas, edge areas and connection parts;

[0059] A high-precision industrial camera combined with a spectrum analyzer is used to capture the surface color of the motorcycle shell, and the RGB value, HSV value, and spectral reflectance of the color are obtained to obtain color data;

[0060] Using a 3D laser scanner or a structured light scanning system to obtain 3D point cloud data of the motorcycle shell and generate a high-precision 3D model; using edge detection and / or morphological processing algorithms to analyze the contour of the motorcycle shell based on the high-precision 3D model to obtain shape data;

[0061] Ultrasonic testing equipment is used to perform non-destructive testing on the interior of the motorcycle shell to obtain ultrasonic data.

[0062] Preferably, the static defect detection module is used to process the product data to obtain static defect data; Figure 3 ;

[0063] The static defect detection module includes: a product data preprocessing unit, a hardness detection unit, a color detection unit, a shape detection unit, an ultrasonic detection unit and a static defect data acquisition unit;

[0064] The product data preprocessing unit preprocesses the product data; the preprocessing includes denoising and normalization;

[0065] The hardness detection unit identifies local hardness defects by analyzing the hardness;

[0066] The color detection unit identifies color abnormal areas by calculating the color distribution histogram and comparing it with the product standard color;

[0067] The shape detection unit marks the deformed area by comparing the shape with the standard shape of the product;

[0068] The ultrasonic detection unit processes the ultrasonic data through signal processing technology to identify internal defects of the product;

[0069] The static defect data acquisition unit comprehensively identifies the local hardness defect, the color abnormality area, the deformation area and the internal defect, identifies the static defect of the product according to the preset static defect judgment standard, and generates the static defect data, including the static defect type and the static defect location.

[0070] Specifically, the denoising includes: using a wavelet transform algorithm to perform multi-scale decomposition on the hardness data and the ultrasonic data to remove environmental interference noise; using a median filter algorithm to remove uneven illumination or external impurities generated by the imaging process in the color data and the shape data;

[0071] The normalization adopts the Min-Max normalization method to standardize the hardness, color, shape, ultrasonic and other data to a unified range.

[0072] The hardness detection unit calculates the hardness mean and standard deviation based on the measured hardness data to evaluate the local hardness uniformity; if the local hardness uniformity of a certain area is lower than a preset hardness uniformity threshold, it is determined to be a local hardness defect and the local hardness defect area is marked; in this embodiment, the preset hardness uniformity threshold is 0.9;

[0073] The color detection unit calculates a color distribution histogram based on the color RGB value, HSV value and spectral reflectance, and compares it with the standard product color model; calculates the color deviation through Euclidean distance, and marks the area as a color abnormal area if the color deviation exceeds a preset color deviation threshold; in this embodiment, the preset color deviation threshold is 5;

[0074] The shape detection unit compares the shape data with the standard shape of the product to calculate the local shape deviation; sets a shape deviation threshold, and determines that if the shape deviation is less than the shape deviation threshold, it is determined to be a normal shape; if the shape deviation exceeds the shape deviation threshold, it is determined to be the deformation area;

[0075] The ultrasonic detection unit performs time-frequency domain analysis on the ultrasonic data, calculates signal attenuation and reflection, and extracts ultrasonic features; uses the ultrasonic features to identify internal defects of the product and mark internal defect areas;

[0076] The static defect data acquisition unit integrates the local hardness defect, the color abnormality area, the deformation area and the internal defect, and calculates the static defect score. The specific formula is:

[0077] ;

[0078] in, Scoring static defects; is the hardness defect weight; is the standard hardness value; is the measured hardness data; is the color defect weight; is color deviation; is the preset color deviation threshold; is the shape defect weight; It is a standard shape; is the measured shape data; is the internal defect weight; is the internal defect area; is the total area of ​​the detection area;

[0079] The static defect classification is performed on the total static defect score according to the preset static defect judgment standard, and the preset static defect judgment standard is obtained through multiple experiments; specifically, the static defect classification is as follows: color abnormality - minor defect, deformation exceeding the standard - medium defect, hardness abnormality - severe defect, internal crack - major defect.

[0080] Table 1 shows the detection results of static defects.

[0081] Table 1 Static defect detection results

[0082]

[0083] The embodiment of the present application uses a product data acquisition module to accurately collect hardness, color, shape and ultrasonic data, and combines the static defect detection module to pre-process and analyze the product data, thereby achieving comprehensive detection of surface and internal defects of the product. The hardness detection unit can identify local hardness anomalies, especially for accurate measurement of areas prone to cold shut defects; the color detection unit can effectively identify defects such as color anomalies, cracks and scratches caused by materials or processes by comparing the product standard color based on the color distribution histogram; the shape detection unit uses deformation comparison technology to quickly detect deformation areas and detect product contour deformation that may be caused by cold shut defects, thereby ensuring product dimensional accuracy; the ultrasonic detection unit uses signal processing technology to identify internal defects and detect internal cracks, fractures and other defects that may be caused by cold shut defects, thereby improving the depth and reliability of defect detection. Finally, the static defect data acquisition unit conducts a comprehensive analysis of the above-mentioned types of defects, providing static defect data support for the subsequent precise positioning and high-precision detection of cold shut defects by combining dynamic defect data and comprehensive defect analysis, thereby improving the efficiency and accuracy of subsequent cold shut defect detection.

[0084] Preferably, the material data acquisition module is used to arrange multiple sensors in the first area, the second area and the third area respectively to collect material data; the material data includes: temperature, pressure, flow rate and solidification time;

[0085] The first area is the gate area; the second area is the thin-wall area; and the third area is the end filling area.

[0086] The various sensors include: thermal imaging sensors, ultrasonic sensors and pressure sensors.

[0087] The embodiment of the present application can collect key material data in real time and accurately by rationally arranging thermal imaging sensors, ultrasonic sensors and pressure sensors in the gate area, thin-wall area and end filling area. This multi-area, multi-parameter detection method ensures comprehensive monitoring of material flow and solidification state during the casting process, and improves the accuracy and timeliness of the data. Monitoring of the gate area helps to analyze the initial filling state of the molten metal and ensure flow uniformity; detection of the thin-wall area can identify problems such as flow obstruction and insufficient filling, and prevent the occurrence of cold shut or shrinkage defects; data collection in the end filling area can determine the final filling condition of the molten metal, optimize the solidification sequence, and avoid the occurrence of defects such as pores and shrinkage. Through the collaborative work of multiple sensors, a reliable basis is provided for abnormal area detection and dynamic defect analysis, which improves the accuracy of subsequent dynamic defect detection, thereby improving production stability and reducing defect and scrap rates.

[0088] Preferably, the abnormal region detection module is used to analyze the material data using the abnormal region detection model, identify the material abnormal region, and mark the severity of the material abnormality; filter out the first abnormal region in the material abnormal region according to the preset material abnormality threshold; the abnormal region detection model includes: a material data preprocessing unit, an abnormal feature recognition unit, an abnormal region marking unit and a first abnormal region screening unit; refer to Figure 4 ;

[0089] Among them, the material data preprocessing unit preprocesses the material data; the abnormal feature identification unit analyzes the preprocessed material data and identifies abnormal features; the abnormal features include: flow obstruction, enhanced turbulence and uneven filling; the abnormal area marking unit identifies abnormal patterns in the material data according to the abnormal features and marks the abnormal area; the first abnormal area screening unit performs a graded evaluation on the abnormal area based on the preset material abnormality threshold, and screens the abnormal area according to the result of the graded evaluation to obtain the first abnormal area.

[0090] Specifically, the preprocessing includes denoising, data interpolation and completion, and normalization processing; the denoising eliminates high-frequency noise that may be generated during the acquisition process through wavelet transform or filtering algorithm; the data interpolation and completion uses interpolation method to interpolate and complete missing data points; the normalization processing converts the collected material data to a unified scale.

[0091] The flow obstruction is detected by detecting a sudden drop in flow velocity and / or a flow velocity in a local area that is significantly lower than the normal flow pattern, thereby identifying the obstruction of molten metal filling. The turbulence enhancement is determined by analyzing the flow velocity fluctuation and calculating the flow velocity standard deviation. If the flow velocity standard deviation exceeds the flow velocity standard deviation threshold, it is determined that there is excessive turbulence. The flow velocity standard deviation threshold is obtained based on industry standard data. The filling unevenness is determined by calculating the difference in filling pressure in different areas to determine whether there is insufficient filling and / or filling delay.

[0092] The abnormal area marking unit divides the boundary of the abnormal area and marks the abnormal area according to the abnormal features using edge detection or cluster analysis methods; the first abnormal area screening unit quantifies the severity of the abnormal area according to the degree of deviation of the abnormal features and divides it into slight abnormality, moderate abnormality and severe abnormality; the scores of the abnormal areas are compared, and if the degree of abnormality exceeds the industry standard safety threshold and the abnormal area is close to the thin-wall area and / or the end filling area, it is preferentially marked as the first abnormal area.

[0093] Table 2 shows the performance improvement effect of the abnormal area detection model on dynamic defect detection.

[0094] Table 2 Performance improvement of abnormal area detection model for dynamic defect detection

[0095]

[0096] The embodiment of the present application accurately identifies abnormal areas of materials in the casting process through multi-sensor data fusion, and marks abnormal areas and their severity based on abnormal features such as flow obstruction, turbulence enhancement and uneven filling. Through wavelet transform denoising, data interpolation and normalization processing, the stability and accuracy of the data are improved, thereby enhancing the reliability of abnormality detection. The abnormal areas are further graded and evaluated through preset industry standard abnormality thresholds to ensure that only the first abnormal areas with a greater impact on product quality are screened out, especially high-risk defects near thin-walled areas and / or end filling areas, thereby improving the pertinence and accuracy of screening. Not only does it improve the accuracy of abnormality detection in the casting process, reduce misjudgments and missed judgments, but it can also provide an accurate reference range of abnormal areas for subsequent accurate cold shut defect detection, thereby improving the efficiency and accuracy of subsequent cold shut defect detection.

[0097] Preferably, the dynamic defect detection module is configured to perform dynamic defect detection by calculating a dynamic defect score of the first abnormal area to obtain dynamic defect data; the formula for the dynamic defect score is:

[0098] ;

[0099] in, Scoring dynamic defects; is the temperature term weight; is the temperature regulation coefficient; is the temperature of the material in the abnormal area; is the weight of the velocity change rate term; is the flow rate change rate adjustment coefficient; is the velocity change rate in the abnormal area; is the weight of filling pressure item; is the local filling pressure in the abnormal area; is the simulated filling pressure; is the filling pressure adjustment coefficient; is the weight of solidification time item; is the coagulation time of the abnormal area; is the simulated solidification time; is the coagulation time adjustment coefficient;

[0100] The dynamic defect is identified based on the dynamic defect score according to a preset dynamic defect threshold to generate the dynamic defect data; the dynamic defect data includes: dynamic defect severity and dynamic defect location.

[0101] Specifically, the dynamic defect is a cold shut defect; the preset dynamic defect threshold is set comprehensively based on industry standards and expert experience.

[0102] Table 3 shows the detection results of dynamic defects.

[0103] Table 3 Dynamic defect detection results

[0104]

[0105] The embodiment of the present application calculates a dynamic defect score for the first abnormal region, enabling precise detection of dynamic defects based on the first abnormal region, thereby improving the accuracy and efficiency of defect identification during the casting process. This dynamic defect scoring formula comprehensively considers key parameters such as temperature, flow rate change rate, filling pressure, and solidification time, thereby enhancing the reliability of defect detection. Furthermore, it classifies defects of different types and degrees based on dynamic defect thresholds and provides defect location data, providing an important basis for subsequent analysis of defect causes and optimization of processes and quality control.

[0106] Preferably, a defect comprehensive analysis module is used to perform defect comprehensive analysis by comparing the static defect data and the dynamic defect data; the defect comprehensive analysis module includes: a defect position comparison unit, a defect influencing factor correlation analysis unit and a defect cause comprehensive analysis unit;

[0107] Among them, the defect position comparison unit compares the static defect position in the static defect data with the dynamic defect position in the dynamic defect data, identifies the overlapping area between the two, and marks it as the cold shut defect position; the defect influencing factor correlation analysis unit calculates the correlation between the product data and the material data in the cold shut defect position and the cold shut defect; the defect cause comprehensive analysis unit identifies the most likely cause of the cold shut defect based on the correlation, and generates a detailed cold shut defect cause analysis report.

[0108] Specifically, the defect position comparison unit uses a spatial matching algorithm, such as KD-Tree search or a matching algorithm based on Euclidean distance, to compare the positions of the two types of defect data and find the overlapping areas between the two. The overlapping parts are marked as "precise defect positions"; for cold shut defect areas that are partially offset but correlated, the morphological expansion method is used to perform expansion matching to ensure maximum coverage of potential defect areas.

[0109] The defect influencing factor association analysis unit adopts a correlation calculation method based on mutual information to evaluate the correlation between each parameter in the product data and the material data and the occurrence of cold shut defects, and obtains a correlation ranking table; sets a correlation threshold to screen out factors that have a greater impact on cold shut defects.

[0110] The defect cause comprehensive analysis unit infers the most likely cause of the cold shut defect based on the dynamic defect severity and historical defect data, and generates a detailed cold shut defect cause analysis report, including the location of the cold shut defect, influencing factors, the most likely cause and improvement suggestions, providing engineers with actionable guidance.

[0111] The embodiment of the present application improves the accuracy of defect detection by fusing static defect data and dynamic defect data, can effectively identify the precise location of cold shut defects, and calculate the key influencing factors of defect occurrence in combination with product data and material data. The spatial matching algorithm and morphological expansion method are used to ensure the accuracy of defect position comparison. At the same time, the correlation calculation method based on mutual information is used to screen out key parameters that have a greater impact on cold shut defects, providing a scientific basis for the analysis of the causes of cold shut defects. Combined with dynamic defect types and historical cold shut defect data, the system can infer the most likely cause of the defect and generate a detailed cold shut defect cause analysis report to provide engineers with improvement suggestions, which will help optimize production processes, reduce defect rates, and improve product quality and production efficiency.

[0112] The embodiment of the present application proposes a defect detection system for high-pressure casting molds for motorcycle shells. Through the product data acquisition module and the material data acquisition module, it realizes comprehensive monitoring of product quality and casting process, ensuring the comprehensiveness and accuracy of the detection data. The static defect detection module can identify defects caused by abnormal hardness, color, shape or internal structure, while the abnormal area detection module accurately identifies abnormal areas of the material based on data such as temperature, pressure, flow rate and solidification time, and screens out key areas that may cause defects, thereby improving the reliability of abnormal detection. The dynamic defect detection module further identifies defects caused by abnormal material fluidity, filling pressure or solidification characteristics by calculating dynamic defect scores, thereby enhancing the dynamic adaptability of defect detection. Finally, the comprehensive defect analysis module combines static and dynamic defect data to accurately locate the defect position, analyze the cause of the defect, and provide targeted improvement suggestions to improve product quality, optimize the casting process, reduce production losses, and improve production efficiency and product qualification rate.

[0113] Example 2

[0114] In Example 1, the method proposed in this application successfully achieved high-precision and high-efficiency detection of cold shut defects in thin-walled motorcycle housing molds, providing a scientific basis for mold design and process parameter optimization. To further verify the effectiveness of this invention, defect detection was also performed on a high-pressure casting mold for a Type A motorcycle housing produced by a motorcycle technology company.

[0115] A method for detecting defects in a high-pressure casting mold for a motorcycle housing, comprising:

[0116] Acquire product data of the product to be tested; the product data includes: hardness, color, shape and ultrasonic data.

[0117] Preferably, the product data is processed to obtain static defect data; the static defect detection module includes: a product data preprocessing unit, a hardness detection unit, a color detection unit, a shape detection unit, an ultrasonic detection unit and a static defect data acquisition unit; the product data preprocessing unit preprocesses the product data; the preprocessing includes denoising and normalization; the hardness detection unit identifies local hardness defects by analyzing the hardness; the color detection unit identifies color abnormality areas by calculating a color distribution histogram and comparing it with the product standard color; the shape detection unit marks the deformation area by comparing the shape with the product standard shape; the ultrasonic detection unit processes the ultrasonic data through signal processing technology to identify internal defects of the product; the static defect data acquisition unit comprehensively identifies the local hardness defects, the color abnormality areas, the deformation areas and the internal defects, identifies static defects of the product according to preset static defect judgment standards, and generates the static defect data, including static defect type and static defect location.

[0118] Preferably, multiple sensors are arranged in the first area, the second area and the third area respectively to collect material data; the material data include: temperature, pressure, flow rate and solidification time; the first area is the gate area; the second area is the thin-wall area; the third area is the end filling area; the multiple sensors include: thermal imaging sensors, ultrasonic sensors and pressure sensors.

[0119] Preferably, the material data is analyzed using an abnormal area detection model to identify material abnormal areas and mark the severity of the material abnormality; the first abnormal area in the material abnormal area is screened out according to a preset material abnormality threshold; the abnormal area detection model includes: a material data preprocessing unit, an abnormal feature recognition unit, an abnormal area marking unit and a first abnormal area screening unit; wherein, the material data preprocessing unit preprocesses the molten metal flow data; the abnormal feature recognition unit analyzes the preprocessed material data and identifies abnormal features; the abnormal features include: obstructed flow, enhanced turbulence and uneven filling; the abnormal area marking unit identifies abnormal patterns in the material data according to the abnormal features and marks the abnormal area; the first abnormal area screening unit performs a graded evaluation on the abnormal area based on the preset material abnormality threshold, and screens the abnormal area according to the result of the graded evaluation to obtain the first abnormal area.

[0120] Preferably, dynamic defect detection is performed by calculating a dynamic defect score of the first abnormal area to obtain dynamic defect data; the formula for the dynamic defect score is:

[0121] ;

[0122] in, Scoring dynamic defects; is the temperature term weight; is the temperature regulation coefficient; is the temperature of the material in the abnormal area; is the weight of the velocity change rate term; is the flow rate change rate adjustment coefficient; is the velocity change rate in the abnormal area; is the weight of filling pressure item; is the local filling pressure in the abnormal area; is the simulated filling pressure; is the filling pressure adjustment coefficient; is the weight of solidification time item; is the coagulation time of the abnormal area; is the simulated solidification time; is the coagulation time adjustment coefficient;

[0123] The dynamic defect is identified based on the dynamic defect score according to a preset dynamic defect threshold to generate the dynamic defect data; the dynamic defect data includes: dynamic defect severity and dynamic defect location.

[0124] Preferably, a comprehensive defect analysis is performed by comparing the static defect data and the dynamic defect data; the comprehensive defect analysis module includes: a defect position comparison unit, a defect influencing factor correlation analysis unit and a defect cause comprehensive analysis unit; wherein, the defect position comparison unit compares the static defect position in the static defect data with the dynamic defect position in the dynamic defect data, identifies the overlapping area between the two, and marks it as a cold shut defect position; the defect influencing factor correlation analysis unit calculates the correlation between the product data and the material data in the cold shut defect position and the cold shut defect; the comprehensive defect cause analysis unit identifies the most likely cause of the cold shut defect based on the correlation, and generates a detailed cold shut defect cause analysis report.

[0125] Table 4 gives the analysis report on the causes of cold shut defects.

[0126] Table 4 Cold shut defect cause analysis report

[0127]

[0128] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A motorcycle shell high pressure casting mold defect detection system, characterized in that: include: A product data acquisition module, used to obtain product data of the product to be tested; Product data includes: hardness, color, shape and ultrasonic data; A static defect detection module is used to process product data to obtain static defect data; the static defect data includes static defect type and static defect location; The material data acquisition module is used to arrange multiple sensors in the first, second, and third areas to collect material data. The first area is the gate area; the second area is the thin-wall area; and the third area is the end filling area. The material data includes: temperature, pressure, flow rate, and solidification time. An abnormal region detection module is used to analyze material data using an abnormal region detection model, identify material abnormal regions, and mark the severity of the material abnormality; and screen out the first abnormal region among the material abnormal regions according to a preset material abnormality threshold; The dynamic defect detection module is configured to perform dynamic defect detection by calculating a dynamic defect score of the first abnormal area to obtain dynamic defect data; the dynamic defect data includes: dynamic defect severity and dynamic defect location; the dynamic defect score formula is: ； in, Scoring dynamic defects; is the temperature term weight; is the temperature regulation coefficient; is the temperature of the material in the abnormal area; is the weight of the velocity change rate term; is the flow rate change rate adjustment coefficient; is the velocity change rate in the abnormal area; is the weight of filling pressure item; is the local filling pressure in the abnormal area; is the simulated filling pressure; is the filling pressure adjustment coefficient; is the weight of solidification time item; is the coagulation time of the abnormal area; is the simulated solidification time; is the coagulation time adjustment coefficient; The defect comprehensive analysis module includes a defect position comparison unit, a defect influencing factor correlation analysis unit and a defect cause comprehensive analysis unit, and is used to perform defect comprehensive analysis by comparing static defect data and dynamic defect data; the defect position comparison unit compares the static defect position in the static defect data with the dynamic defect position in the dynamic defect data, identifies the overlapping area between the two, and marks it as a cold shut defect position; the defect influencing factor correlation analysis unit calculates the correlation between the product data and the material data in the cold shut defect position and the cold shut defect; the defect cause comprehensive analysis unit identifies the cause of the cold shut defect based on the correlation and generates a detailed cold shut defect cause analysis report.

2. A motorcycle shell high pressure casting mold defect detection system according to claim 1, characterized in that: The static defect detection module includes: a product data preprocessing unit, a hardness detection unit, a color detection unit, a shape detection unit, an ultrasonic detection unit and a static defect data acquisition unit; the product data preprocessing unit preprocesses the product data; the preprocessing includes denoising and normalization; the hardness detection unit identifies local hardness defects by analyzing the hardness; the color detection unit identifies color abnormality areas by calculating the color distribution histogram and comparing it with the product standard color; the shape detection unit marks the deformation area by comparing the shape with the product standard shape; the ultrasonic detection unit processes the ultrasonic data through signal processing technology to identify internal defects of the product; the static defect data acquisition unit comprehensively considers the local hardness defects, the color abnormality areas, the deformation areas and the internal defects, identifies static defects of the product according to preset static defect judgment standards, and generates the static defect data.

3. A motorcycle shell high pressure casting mold defect detection system according to claim 1, characterized in that: The various sensors include: thermal imaging sensors, ultrasonic sensors and pressure sensors.

4. A motorcycle shell high pressure casting mold defect detection system according to claim 1, characterized in that: The abnormal area detection model includes: a material data preprocessing unit, an abnormal feature identification unit, an abnormal area marking unit and a first abnormal area screening unit; wherein, the material data preprocessing unit preprocesses the material data; the abnormal feature identification unit analyzes the preprocessed material data and identifies abnormal features; the abnormal features include: flow obstruction, enhanced turbulence and uneven filling; the abnormal area marking unit identifies abnormal patterns in the material data according to the abnormal features and marks the abnormal area; the first abnormal area screening unit performs a graded evaluation on the abnormal area based on the preset material abnormality threshold, and screens the abnormal area according to the result of the graded evaluation to obtain the first abnormal area.

5. The motorcycle shell high pressure casting mold defect detection system according to claim 1, characterized in that: The dynamic defect is identified based on the dynamic defect score according to a preset dynamic defect threshold to generate the dynamic defect data.

6. A method for detecting defects in a high-pressure casting mold for a motorcycle housing, using a defect detection system for a high-pressure casting mold for a motorcycle housing according to any one of claims 1 to 5, characterized in that: include: Obtain product data of the product to be tested; The product data includes: hardness, color, shape and ultrasonic data; Processing the product data to obtain static defect data; Arrange a variety of sensors in the first area, the second area, and the third area to collect material data; the material data includes: temperature, pressure, flow rate, and solidification time; Analyzing the material data using an abnormal region detection model, identifying material abnormal regions, and marking the severity of the material abnormality; screening out a first abnormal region among the material abnormal regions according to a preset material abnormality threshold; Performing dynamic defect detection by calculating a dynamic defect score of the first abnormal area to obtain dynamic defect data; A comprehensive defect analysis is performed by comparing the static defect data and the dynamic defect data.

Images