A method and system for controlling the solidification quality of castings
By identifying thin-walled and thick-walled regions in the 3D model of the casting, calculating the solidification time, generating and optimizing the pouring sequence, and adjusting the solidification process in real time, the problem of uneven solidification of the casting was solved, and the efficiency of casting quality control and yield were improved.
Patent Information
- Application Number
- CN202611581771.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-10-13
- Publication Date
- 2026-11-17
AI Technical Summary
Existing technologies cannot accurately match the solidification characteristics of different areas during the solidification process of castings, resulting in low efficiency in solidification quality control and an inability to respond in real time to changes in solidification state, which easily leads to uneven solidification and casting defects.
By acquiring a 3D model of the casting, identifying thin-walled and thick-walled regions, calculating the solidification time of each region, generating multiple candidate pouring sequence schemes, selecting the optimal pouring sequence based on the solidification matching index, and adjusting the sorting of unpouring regions in real time during the solidification process, and optimizing the cooling intensity by combining casting flipping operations to achieve precise control.
It improves the efficiency of solidification quality control of castings, reduces the casting defect rate, increases the casting yield, and realizes real-time response and adaptive control of solidification state.
Smart Images

Figure CN123102976A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of casting manufacturing and quality control technology, specifically to a method and system for controlling the solidification quality of castings. Background Technology
[0002] In the casting production process, the solidification quality of castings directly affects the mechanical properties and service life of the product. Solidification quality control is of paramount importance. Large and complex castings have diverse structures and large thickness differences. Solidification quality control systems are often used to monitor and regulate the solidification process of castings and minimize casting defects. This is especially true in the manufacture of high-quality copper castings, high-quality titanium castings, and high-quality magnesium alloy castings, where the solidification characteristics of the materials are complex and the quality requirements are extremely high. Precise control of the solidification process is particularly crucial.
[0003] Existing technologies employ a preset fixed pouring sequence or rely on experience for pouring. However, the solidification behavior of castings is affected by various factors such as thickness distribution, heat dissipation path, and pouring sequence, leading to misjudgments in existing technologies. Differences in solidification time in different areas of the casting affect the overall solidification quality. The inability to accurately match the solidification characteristics of each area and dynamically adjust the pouring sequence results in numerous casting defects. During the solidification process, existing technologies mainly use fixed cooling parameters, experience-based adjustments, or post-inspection for quality control. This can lead to unfavorable situations such as the inability to respond in real time to changes in solidification state and uneven solidification in thin and thick wall areas, resulting in low efficiency in controlling the solidification quality of castings.
[0004] Therefore, it is necessary to design a casting solidification quality control method and system that can more accurately control the solidification matching degree of different areas of the casting and improve the efficiency of solidification quality control. Summary of the Invention
[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a method for controlling the solidification quality of castings, the method comprising the following steps: Obtain a three-dimensional model of the casting to be poured, identify the thin-walled region and the thick-walled region in the three-dimensional model, and calculate the thin-walled solidification time corresponding to the thin-walled region and the thick-walled solidification time corresponding to the thick-walled region respectively. Multiple candidate pouring sequence schemes are generated. Each candidate pouring sequence scheme includes different arrangements and combinations of the thin-walled region and the thick-walled region. For each candidate pouring sequence scheme, a solidification matching index is calculated based on the solidification time and pouring interval time of each region. The candidate pouring sequence scheme with the best solidification matching index is selected as the target pouring sequence. According to the target pouring sequence, molten metal is poured into each area in sequence. During the pouring process, the surface temperature distribution of the poured area is collected to calculate the actual liquid phase fraction. When the deviation between the actual liquid phase fraction and the preset liquid phase fraction exceeds the preset deviation threshold, the pouring order of the unpoured areas in the target pouring sequence is adjusted. Before performing the casting flipping operation, it is determined whether the liquid phase fraction of the thin-walled region and the thick-walled region is less than the preset flipping liquid phase threshold. If not, the cooling intensity is adjusted until the preset flipping liquid phase threshold is met before performing the casting flipping operation.
[0006] According to the above technical solution, a three-dimensional model of the casting to be poured is obtained, and thin-walled and thick-walled regions in the three-dimensional model are identified. The solidification time of the thin-walled region and the solidification time of the thick-walled region are calculated respectively, including: The coverage area of the controllable area of the nozzle is determined according to the nozzle diameter of the pouring nozzle, and the three-dimensional model is divided into several evaluation units according to the coverage area of the controllable area of the nozzle. For each evaluation unit, its geometric shape type is determined. When the evaluation unit is a regular trapezoidal structure, the wall thickness is sampled proportionally along the height direction of the trapezoid and the wall thickness value of each sampling position is marked. When the evaluation unit is an irregular structure, wall thickness sampling points are arranged in the evaluation unit according to preset mechanical constraint points and dimensional key points and the wall thickness value of each wall thickness sampling point is marked. Calculate the unit average wall thickness of each evaluation unit, identify the evaluation unit with the unit average wall thickness less than the preset thin wall threshold as the thin wall region, and identify the evaluation unit with the unit average wall thickness greater than the preset thick wall threshold as the thick wall region; Feature annotation is performed on local locations within the thin-walled region and the thick-walled region that deviate from the unit average wall thickness by more than a preset deviation threshold; When calculating the thin-wall solidification time corresponding to the thin-walled region and the thick-wall solidification time corresponding to the thick-walled region, the feature modulus of each region is calculated based on the unit average wall thickness, and the thin-wall solidification time and the thick-wall solidification time are corrected based on the degree of wall thickness deviation at the feature marking position.
[0007] According to the above technical solution, multiple candidate pouring sequence schemes are generated, each candidate pouring sequence scheme including different arrangements and combinations of the thin-walled region and the thick-walled region, including: Obtain the spatial coordinates of the thin-walled region and the thick-walled region, and construct a region adjacency graph based on the geometric distance and connectivity between each region. The region adjacency graph uses each region as a node and the geometric connection channel between adjacent regions as an edge. Identify the solidification time difference characteristics of each node in the region adjacency graph, compare the solidification time difference between adjacent nodes with multiple preset difference level thresholds, and label each edge with the corresponding difference level according to the threshold range in which the solidification time difference is located. In the adjacency graph of the region, multiple traversal paths are generated starting from the initial pouring position. Each traversal path passes through edges of different difference levels in sequence to reach the final pouring position. Each traversal path corresponds to a candidate pouring sequence scheme.
[0008] According to the above technical solution, for each candidate pouring sequence scheme, a solidification matching index is calculated based on the solidification time and pouring interval time of each region. The candidate pouring sequence scheme with the optimal solidification matching index is selected as the target pouring sequence, including: The spatial coordinates of adjacent pouring area pairs in a single candidate pouring sequence scheme are extracted sequentially, the nozzle movement path length between adjacent area pairs is calculated, and the expected transfer time between each adjacent area pair is determined according to the preset nozzle movement speed. The expected transfer time is compared with the solidification time difference of the adjacent region pair. When the expected transfer time exceeds the solidification time of the preceding region or is greater than the solidification start time of the subsequent region, it is marked as a failed transfer. The number of failed transfers in each of the candidate casting sequence schemes is counted and the candidate casting sequence schemes with the number exceeding the preset failure threshold are removed. Extract the solidification time difference sequence of adjacent pouring area pairs in the remaining candidate pouring sequence schemes, perform segmented clustering on the solidification time difference sequence to identify the inflection point position of the change trend of the solidification time difference, and divide the candidate pouring sequence scheme into multiple solidification stages according to the inflection point position. Calculate the fluctuation range of the solidification time difference in each region within each solidification stage, evaluate the coordination matching degree between the fluctuation range and the expected transfer time fluctuation range of the corresponding solidification stage, construct the solidification matching degree index based on the coordination matching degree evaluation result, and take the candidate pouring sequence scheme corresponding to the solidification matching degree index with the largest value as the target pouring sequence.
[0009] According to the above technical solution, molten metal is poured into each area sequentially according to the target pouring sequence. During the pouring process, the surface temperature distribution of the poured areas is collected to calculate the actual liquid phase fraction, including: According to the target pouring sequence, molten metal is poured into each area in sequence. After the pouring of a single area is completed, the surface temperature distribution data of the poured area is collected. The temperature gradient direction and gradient magnitude in the surface temperature distribution data are identified. The main heat dissipation path of the area is determined according to the temperature gradient direction. The thickness distribution features of the region are extracted along the main heat dissipation path, and the temperature values of each sampling point in the surface temperature distribution data are correlated and mapped with the thickness values at the corresponding locations to construct a temperature-thickness correlation matrix. Based on the correspondence between surface temperature and thickness in the temperature-thickness correlation matrix, the temperature distribution state of each depth layer in the region is calculated in combination with the preset temperature attenuation coefficient. The depth layer with a temperature value higher than the liquidus temperature is extracted as the liquidus layer, and the depth layer with a temperature value between the liquidus temperature and the solidus temperature is extracted as the solid-liquid coexistence layer. The distribution ratio of the liquid phase layer and the solid-liquid coexistence layer at each thickness position in the region is statistically analyzed, and the actual liquid phase fraction of the cast area is calculated based on the distribution ratio.
[0010] According to the above technical solution, when the deviation between the actual liquid phase fraction and the preset liquid phase fraction exceeds a preset deviation threshold, the pouring order of the unpoured areas in the target pouring sequence is adjusted, including: The difference between the actual liquid phase fraction of the poured area and the corresponding preset liquid phase fraction is calculated. When the absolute value of the difference exceeds the preset deviation threshold, the thickness distribution characteristics of the poured area and the pouring time interval of the adjacent poured areas are extracted. The influencing factors that cause the liquid phase fraction deviation are identified based on the thickness distribution characteristics and the pouring time interval. In the uncast areas of the target pouring sequence, the heat conduction path length between each uncast area and the cast area is calculated. Uncast areas with heat conduction path lengths less than a preset path threshold are selected as sensitive areas associated with the influencing factor. The solidification time coupling degree between each sensitive area and the cast area is calculated. The sensitive areas are prioritized according to the solidification time coupling degree. For the highest priority sensitive area, the position of the sensitive area in the target pouring sequence is adjusted according to the deviation direction of the actual liquid fraction. The adjusted pouring sequence is substituted into the calculation process of the solidification matching index for iterative verification. The pouring sequence that makes the deviation of the actual liquid fraction converge to within the preset deviation threshold is extracted as the updated target pouring sequence.
[0011] According to the above technical solution, before performing the casting flipping operation, it is determined whether the liquid phase fractions of both the thin-walled region and the thick-walled region are less than a preset flipping liquid phase threshold. If not, the cooling intensity is adjusted until the preset flipping liquid phase threshold is met before performing the casting flipping operation, including: Before performing the casting flipping operation, the flipping direction and flipping angular velocity of the casting to be performed are obtained. Based on the casting flipping direction, the spatial position relationship of each region of the casting relative to the flipping axis is identified, and the distribution of inertial force borne by the thin-walled region and the thick-walled region during the flipping process is calculated. The liquid phase stability coefficient of each region is determined based on the inertial force distribution and thickness distribution characteristics of each region. The liquid phase stability coefficient is coupled with the preset liquid phase inversion threshold to obtain the actual liquid phase inversion threshold of each region. The surface temperature distribution of the thin-walled region and the thick-walled region is collected to calculate the liquid phase fraction of each region. Regions with liquid phase fraction exceeding the actual inverted liquid phase threshold are identified as regions to be cooled more effectively. The target cooling medium flow rate increase is determined based on the excess liquid phase fraction and inertial force distribution characteristics of the regions to be cooled more effectively. Adjust the cooling medium flow rate of the corresponding cooling nozzle in the area to be cooled according to the target cooling medium flow rate increase, continuously monitor the change in liquid phase fraction in the area to be cooled, and perform the casting flipping operation when the liquid phase fraction of the thin-walled area and the thick-walled area both meet their respective actual flipping liquid phase thresholds.
[0012] A second aspect of this application provides a casting solidification quality control system, the system comprising: The three-dimensional model module is used to obtain a three-dimensional model of the casting to be poured, identify the thin-walled region and the thick-walled region in the three-dimensional model, and calculate the thin-walled solidification time corresponding to the thin-walled region and the thick-walled solidification time corresponding to the thick-walled region. The sequential pouring module is used to generate multiple candidate pouring sequence schemes. Each candidate pouring sequence scheme includes different arrangements and combinations of the thin-walled region and the thick-walled region. For each candidate pouring sequence scheme, a solidification matching index is calculated based on the solidification time and pouring interval time of each region. The candidate pouring sequence scheme with the best solidification matching index is selected as the target pouring sequence. The sorting adjustment module is used to sequentially pour molten metal into each area according to the target pouring order. During the pouring process, the surface temperature distribution of the poured area is collected to calculate the actual liquid phase fraction. When the deviation between the actual liquid phase fraction and the preset liquid phase fraction exceeds the preset deviation threshold, the pouring order of the unpoured areas in the target pouring order is adjusted. The flipping confirmation module is used to determine whether the liquid phase fraction of the thin-walled region and the thick-walled region is less than the preset flipping liquid phase threshold before performing the casting flipping operation. If not, the cooling intensity is adjusted until the preset flipping liquid phase threshold is met before performing the casting flipping operation.
[0013] A third aspect of this application provides an electronic device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, it causes the electronic device to perform the method described in the first aspect of this application.
[0014] A fourth aspect of this application provides a computer-readable storage medium for storing a computer program that, when run on a computer, causes the computer to perform the method described in the first aspect of this application.
[0015] Compared with existing technologies, the beneficial effects achieved by this invention are as follows: This invention acquires casting model data and identifies thin-walled and thick-walled regions, calculates the solidification time of different thickness regions, characterizes the differences in solidification characteristics of each region, provides a data basis for optimizing the pouring sequence, avoids solidification imbalance caused by a fixed pouring sequence, generates candidate pouring sequence schemes, and calculates the solidification matching index based on the solidification time and pouring interval time of each region. The scheme with the optimal index is selected as the target pouring sequence, achieving solidification time matching of each region, reducing casting defects caused by regional solidification differences, and collecting surface temperature distribution data of the pouring region to predict the solidification time of each region. The actual liquid phase fraction is calculated, and the pouring sequence of the unpouring area is dynamically adjusted when the deviation between the actual liquid phase fraction and the preset liquid phase fraction exceeds a threshold. This achieves real-time response and adaptive control of the solidification state, improves the efficiency of solidification quality control, and determines whether the liquid phase fraction in the thin-walled and thick-walled areas meets the flipping conditions before the casting is flipped. The actual flipping liquid phase threshold under the influence of the inertial force of the liquid metal is determined in combination with the flipping direction. The flipping timing is reasonably determined, reducing casting defects caused by the flow of liquid metal during the flipping process. This achieves precise control of the casting solidification process, reduces the casting defect rate, and improves the efficiency of solidification quality control and the casting yield. Attached Figure Description
[0016] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a flowchart of the steps of a casting solidification quality control method according to an embodiment of this application; Figure 2 This is a schematic diagram of the system modules for controlling the solidification quality of castings according to an embodiment of this application; Figure 3 This is a schematic diagram of an electronic device according to an embodiment of this application. Detailed Implementation
[0017] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.
[0018] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.
[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0020] Please see Figure 1 The present invention provides a technical solution: a method for controlling the solidification quality of castings, comprising: Obtain the three-dimensional model of the casting to be poured, identify the thin-walled region and the thick-walled region in the three-dimensional model, and calculate the thin-walled solidification time corresponding to the thin-walled region and the thick-walled solidification time corresponding to the thick-walled region. Multiple candidate pouring sequence schemes are generated. Each candidate pouring sequence scheme includes different arrangements and combinations of thin-walled and thick-walled regions. For each candidate pouring sequence scheme, the solidification matching index is calculated based on the solidification time and pouring interval time of each region. The candidate pouring sequence scheme with the best solidification matching index is selected as the target pouring sequence. According to the target pouring sequence, molten metal is poured into each area in sequence. During the pouring process, the surface temperature distribution of the poured area is collected to calculate the actual liquid phase fraction. When the deviation between the actual liquid phase fraction and the preset liquid phase fraction exceeds the preset deviation threshold, the pouring order of the unpoured areas in the target pouring sequence is adjusted. Before performing the casting flipping operation, determine whether the liquid phase fraction in both the thin-walled and thick-walled regions is less than the preset flipping liquid phase threshold. If not, adjust the cooling intensity until the preset flipping liquid phase threshold is met before performing the casting flipping operation.
[0021] In some preferred embodiments, a three-dimensional model of the casting to be poured is obtained, thin-walled regions and thick-walled regions in the three-dimensional model are identified, and the solidification time of the thin-walled region and the solidification time of the thick-walled region are calculated respectively, including: The coverage area of the controllable area of the nozzle is determined according to the nozzle diameter, and the three-dimensional model is divided into several evaluation units according to the coverage area of the controllable area of the nozzle. For each evaluation unit, its geometric shape type is determined. When the evaluation unit is a regular trapezoidal structure, the wall thickness is sampled proportionally along the height direction of the trapezoid and the wall thickness value of each sampling position is marked. When the evaluation unit is an irregular structure, wall thickness sampling points are arranged in the evaluation unit according to the preset mechanical constraint points and dimensional key points and the wall thickness value of each sampling point is marked. Calculate the average wall thickness per unit for each evaluation unit, identify evaluation units with an average wall thickness per unit less than a preset thin-wall threshold as thin-walled areas, and identify evaluation units with an average wall thickness per unit greater than a preset thick-wall threshold as thick-walled areas. Feature annotation is performed on local locations in thin-walled and thick-walled regions that deviate from the unit average wall thickness by more than a preset deviation threshold; When calculating the thin-walled solidification time corresponding to the thin-walled region and the thick-walled solidification time corresponding to the thick-walled region, the characteristic modulus of each region is calculated based on the average wall thickness per unit, and the thin-walled solidification time and thick-walled solidification time are corrected according to the degree of wall thickness deviation at the feature marking position.
[0022] For example, when controlling the solidification quality of a casting to be poured, the structural features of the casting are accurately identified and analyzed. After obtaining the three-dimensional model of the casting, considering the physical limitations of the nozzle control capability during the actual pouring process, the coverage range of the controllable area of the nozzle is determined based on the nozzle diameter. The coverage range of the controllable area refers to the spatial range within which a single pouring nozzle can achieve uniform molten metal distribution and effective temperature control during the pouring process; this range is usually positively correlated with the nozzle diameter. The three-dimensional model is divided into several evaluation units according to this coverage range, so that each evaluation unit corresponds to an independently controllable pouring area, thereby providing a reasonable spatial unit basis for subsequent pouring sequence optimization and solidification time calculation.
[0023] Due to the complexity of casting structures, different evaluation units exhibit different geometric morphological characteristics. For each evaluation unit, its geometric morphology is determined. When the evaluation unit is a regular trapezoidal structure, this type of structure has a clear height direction and regular thickness variation characteristics. Wall thickness sampling is performed proportionally along the height direction of the trapezoid, and the wall thickness value at each sampling location is marked. The sampling interval is determined based on the trapezoidal height and the gradient of wall thickness variation, typically set to 5-10 sampling points. When the evaluation unit is an irregular structure, this type of structure lacks regular geometric characteristics, and its wall thickness distribution is significantly affected by mechanical constraint points and dimensional critical points. Mechanical constraint points refer to the locations where the casting bears the main load or stress concentration during use, while dimensional critical points refer to locations that have a significant impact on the functional dimensions or assembly accuracy of the casting. Based on the preset mechanical constraint points and dimensional critical points, wall thickness sampling points are arranged within the evaluation unit, and the wall thickness value at each sampling point is marked to ensure that the sampling points can reflect the true wall thickness distribution characteristics of the evaluation unit.
[0024] After acquiring the wall thickness sampling data for each evaluation unit, the unit average wall thickness of each evaluation unit is calculated. The unit average wall thickness is the arithmetic mean of all wall thickness sampling points within that evaluation unit. Evaluation units with a unit average wall thickness less than a preset thin-wall threshold are identified as thin-walled regions, and those with a unit average wall thickness greater than a preset thick-wall threshold are identified as thick-walled regions. The preset thin-wall threshold is determined based on the casting material type and solidification characteristics. For copper castings, it is typically set to 8mm-12mm; for titanium castings, it is typically set to 6mm-10mm; and for magnesium alloy castings, it is typically set to 5mm-8mm. The preset thick-wall threshold is typically set to 2-3 times the preset thin-wall threshold to distinguish between thick-walled regions and transition regions. Evaluation units with a unit average wall thickness between the preset thin-wall threshold and the preset thick-wall threshold are identified as transition regions. In subsequent solidification time calculations, the solidification characteristics of both thin-walled and thick-walled regions need to be considered in these regions.
[0025] Considering the potential for abrupt changes in wall thickness or abnormal structures within the evaluation unit, local locations in both thin-walled and thick-walled regions that deviate from the unit average wall thickness by more than a preset deviation threshold are marked with features. The preset deviation threshold is set to 30%-50% of the unit average wall thickness. When the absolute value of the difference between the wall thickness value at a given sampling point and the unit average wall thickness exceeds this threshold, that location is marked as a wall thickness anomaly feature point. These marked locations typically correspond to structural features of the casting, such as reinforcing ribs, bosses, and hole edges, whose solidification behavior differs significantly from the overall solidification behavior of the evaluation unit.
[0026] After identifying the thin-walled and thick-walled regions, the solidification time for the thin-walled region and the solidification time for the thick-walled region are calculated. The characteristic modulus of each region is calculated based on the average wall thickness per unit area. The characteristic modulus reflects the heat dissipation capacity of a local area of the casting and is calculated as the ratio of the region's volume to its heat dissipation surface area. For the evaluation unit, the volume can be directly obtained from the 3D model, while the heat dissipation surface area needs to consider the contact area between the evaluation unit and the cavity, adjacent evaluation units, and the external environment. According to Chvorinov's solidification law, the solidification time is positively correlated with the square of the characteristic modulus. The solidification time t of the thin-walled region is calculated as follows. i The thick-walled region's solidification time t c This can be expressed as: t = K·M n Where t is the solidification time, M is the characteristic modulus, and K is the solidification coefficient, which is related to the thermophysical properties of the casting material, the pouring temperature, and the cavity temperature. For copper castings, K is typically taken as 1.2-1.8 min / cm². 2 For titanium castings, the value is 1.5-2.2 min / cm. 2 For magnesium alloy castings, the value is 0.8-1.3 min / cm. 2 n is the solidification index, which is usually taken as 1.8-2.2.
[0027] Since the degree of wall thickness deviation at the feature annotation location affects the actual solidification time of that area, the solidification time of thin-walled and thick-walled sections is corrected based on the degree of wall thickness deviation at the feature annotation location. The wall thickness deviation of all abnormal feature points within each evaluation unit is extracted, and the wall thickness deviation coefficient of that evaluation unit is calculated. The wall thickness deviation coefficient is the weighted average of all wall thickness deviations, with the weights determined by the proportion of the local volume corresponding to each feature point to the total volume of the evaluation unit. When the wall thickness value of an abnormal feature point is greater than the unit average wall thickness, the solidification time at that location is prolonged, and the wall thickness deviation coefficient is used as an amplification factor to correct the solidification time. Conversely, when the wall thickness value of an abnormal feature point is less than the unit average wall thickness, the solidification time at that location is shortened, and the wall thickness deviation coefficient is used as a reduction factor to correct the solidification time. The corrected solidification time more accurately reflects the actual solidification behavior of each evaluation unit, providing reliable input data for subsequently generating multiple candidate casting sequence schemes and calculating solidification matching indices.
[0028] The above technical solution enables refined regional division of the three-dimensional model of the casting and accurate calculation of solidification time, solving the technical problem of insufficient accuracy in solidification time prediction in the existing technology, and laying the foundation for achieving optimal matching of solidification time sequence in each region of the casting.
[0029] In some preferred embodiments, multiple candidate pouring sequence schemes are generated, each candidate pouring sequence scheme including different arrangements and combinations of thin-walled and thick-walled regions, including: Obtain the spatial coordinates of the thin-walled and thick-walled regions, and construct a region adjacency graph based on the geometric distance and connectivity between each region. The region adjacency graph uses each region as a node and the geometric connectivity channels between adjacent regions as edges. Identify the solidification time difference characteristics of each node in the region adjacency graph, compare the solidification time difference between adjacent nodes with multiple preset difference level thresholds, and label the corresponding difference level of each edge according to the threshold range in which the solidification time difference is located. In the region adjacency graph, multiple traversal paths are generated starting from the initial pouring position. Each traversal path passes through edges of different difference levels in sequence to reach the final pouring position. Each traversal path corresponds to a candidate pouring order scheme.
[0030] For example, the spatial coordinates of the thin-walled and thick-walled regions are obtained. These coordinates are determined by the geometric center points of each evaluation unit in the 3D model, reflecting the relative positional relationship of each region within the overall casting structure. A region adjacency graph is constructed based on the geometric distance and connectivity between regions. The geometric distance is the straight-line distance between the geometric center points of adjacent evaluation units, while connectivity considers whether there are physical connections or molten metal flow channels between adjacent evaluation units. The region adjacency graph uses each region as a node and the geometrically connected channels between adjacent regions as edges. A geometrically connected channel refers to a physical path that allows molten metal to flow from one region to another during the pouring process; the existence and width of this path directly affect the feasibility of the pouring sequence.
[0031] During the construction of the region adjacency graph, it is necessary to exclude region pairs that lack actual connectivity. For region pairs with a spatial distance less than a preset connectivity threshold but no physical connection, it is determined whether there are cavity channels or gating system drainage channels between them. The preset connectivity threshold is typically set to 1.5-2 times the coverage area of the nozzle controllable region. This threshold is used to limit the maximum spatial interval through which continuous flow of molten metal can be achieved by gating control. Adjacency relationships are established only between region pairs that meet the connectivity conditions to ensure that the region adjacency graph can accurately reflect the castable topology of the casting.
[0032] The solidification time difference characteristics of each node in the adjacency graph are identified, reflecting the degree of difference in solidification rate between adjacent regions. The solidification time difference between adjacent nodes is compared with multiple preset difference level thresholds. The solidification time difference is the absolute value of the difference in solidification time between two adjacent regions. Multiple preset difference level thresholds are set to classify solidification time differences into different levels. The first difference level threshold is set to 20%-30% of the solidification time of a smaller region, used to identify regions with similar solidification times. These regions offer high flexibility in casting sequence arrangement. The second difference level threshold is set to 50%-70% of the solidification time of a smaller region, used to identify regions with moderate solidification time differences. The casting sequence arrangement of these regions has a significant impact on the overall solidification quality. The third difference level threshold is set to more than 100% of the solidification time of a smaller region, used to identify regions with significant solidification time differences. These regions typically correspond to the adjacency relationship between thin-walled and thick-walled regions, and their casting sequence directly determines whether the casting will exhibit solidification defects such as shrinkage cavities and porosity.
[0033] The edge is labeled with a corresponding difference level based on the threshold range of the solidification time difference. When the solidification time difference is less than the first difference level threshold, the edge is labeled as low difference level, indicating that the solidification behavior of adjacent areas is similar, and the pouring sequence has little impact on the solidification quality. When the solidification time difference is between the first and second difference level thresholds, the edge is labeled as medium difference level, indicating that the solidification behavior of adjacent areas has some differences, and the pouring time interval between the two areas needs to be reasonably controlled in the pouring sequence arrangement. When the solidification time difference is greater than the second difference level threshold, the edge is labeled as high difference level, indicating that the solidification behavior of adjacent areas differs significantly, and the rationality of the pouring sequence is crucial to avoiding solidification defects. The labeling of difference levels provides a quantitative basis for the path selection strategy when generating traversal paths.
[0034] Multiple traversal paths are generated in the region adjacency graph starting from the initial pouring position. The initial pouring position is determined based on the gating system layout of the casting, typically choosing a region directly connected to the gate or ingate. Each traversal path sequentially passes through edges of different difference levels to reach the final pouring position, which is usually chosen as the region farthest from the gating system or with the longest solidification time to ensure the sequential solidification characteristics of the entire casting. When generating traversal paths, a path search strategy based on difference level priority is adopted, prioritizing paths that pass through edges with high difference levels, as the regions connected by high difference level edges have the greatest impact on solidification quality and offer the largest optimization space for the pouring sequence. For nodes with multiple selectable paths, the path selection direction is determined comprehensively based on the solidification time of downstream nodes, the heat-affected zone of already poured areas, and the spatial distribution of the remaining unpoured areas.
[0035] Multiple traversal paths are generated, forming different candidate pouring sequence schemes, with each traversal path corresponding to one candidate pouring sequence scheme. The number of candidate pouring sequence schemes is determined based on the topological complexity of the region adjacency graph. For castings with 10-20 nodes, 5-10 candidate pouring sequence schemes are typically generated; for complex castings with more than 20 nodes, 10-15 candidate pouring sequence schemes can be generated. Each candidate pouring sequence scheme includes different arrangements and combinations of thin-walled and thick-walled regions. By changing the relative positions of the thin-walled and thick-walled regions in the pouring sequence, different solidification strategies are explored. For example, one candidate scheme may employ a strategy of pouring the thick-walled region first and then the thin-walled region, utilizing the heat from the thick-walled region to compensate for the solidification shrinkage of the thin-walled region. Another candidate scheme may employ a strategy of pouring the thin-walled region first and then the thick-walled region, avoiding premature solidification of the thin-walled region that could block the feeding channels in the thick-walled region.
[0036] To ensure the practical feasibility of the generated candidate pouring sequence schemes, pouring continuity verification is performed on each traversal path. This involves checking whether the flow path of molten metal between adjacent nodes in the path is unobstructed, and whether the time interval between the completion time of the pouring of the current region and the start time of the pouring of the next region meets the molten metal flowability requirements. The molten metal flowability requirements are related to the pouring temperature, molten metal viscosity, and flow path length. For traversal paths that do not meet the pouring continuity conditions, corrections are made by adjusting the visit order of some nodes or inserting intermediate nodes to ensure that all candidate pouring sequence schemes can be implemented in actual pouring operations.
[0037] The above technical solution enables the generation of multiple candidate pouring sequence schemes based on regional adjacency graphs and solidification time differences. This solves the technical problem in existing technologies that cannot systematically explore different pouring strategies, and provides a diverse set of candidate schemes for subsequent calculation of solidification matching index and selection of the optimal pouring sequence.
[0038] In some preferred embodiments, for each candidate pouring sequence scheme, a solidification matching index is calculated based on the solidification time and pouring interval time of each region, and the candidate pouring sequence scheme with the optimal solidification matching index is selected as the target pouring sequence, including: Extract the spatial coordinates of adjacent pouring area pairs in a single candidate pouring sequence scheme, calculate the nozzle movement path length between adjacent area pairs, and determine the expected transfer time between each adjacent area pair based on the preset nozzle movement speed. The expected transfer time is compared with the solidification time difference of the adjacent region pair. When the expected transfer time exceeds the solidification time of the preceding region or is greater than the solidification start time of the subsequent region, it is marked as a failed transfer. The number of failed transfers in each candidate pouring sequence scheme is counted and candidate pouring sequence schemes with a number exceeding the preset failure threshold are removed. Extract the solidification time difference sequence of adjacent pouring area pairs in the remaining candidate pouring sequence schemes, perform segmented clustering on the solidification time difference sequence to identify the inflection point position of the change trend of solidification time difference, and divide the candidate pouring sequence schemes into multiple solidification stages according to the inflection point position. Calculate the fluctuation range of the solidification time difference in each region within each solidification stage, evaluate the degree of coordination between the fluctuation range and the expected transfer time fluctuation range of the corresponding solidification stage, construct a solidification matching degree index based on the evaluation results, and take the candidate pouring sequence scheme corresponding to the solidification matching degree index with the largest value as the target pouring sequence.
[0039] For example, the spatial coordinates of adjacent pouring region pairs in a single candidate pouring sequence are extracted sequentially. An adjacent pouring region pair refers to two regions that are connected sequentially in the pouring order of the candidate pouring sequence, and their spatial coordinates are determined when constructing the region adjacency diagram. The nozzle movement path length between adjacent region pairs is calculated. This nozzle movement path length considers the actual movement trajectory of the pouring equipment, including the spatial distance traversed by the pouring nozzle from the pouring position of the preceding region to the pouring position of the following region. This distance is affected by the casting structure, the arm length limitation of the pouring equipment, and obstacle avoidance requirements. The expected transfer time between each adjacent region pair is determined based on the preset nozzle movement speed. The preset nozzle movement speed is determined according to the mechanical performance of the pouring equipment; for robotic arm pouring equipment, it is typically set to 0.5 m / s-1.2 m / s, and for gantry pouring equipment, it is typically set to 0.8 m / s-1.5 m / s. The expected transfer time is the ratio of the nozzle movement path length to the preset nozzle movement speed, reflecting the time interval between the completion of the pouring of the preceding region and the start of the pouring of the following region during the pouring process.
[0040] The expected transfer time is compared with the solidification time difference of adjacent regions. The solidification time difference is the difference between the solidification time of the subsequent region and the solidification time of the preceding region. When the expected transfer time exceeds the solidification time of the preceding region, it means that when the pouring nozzle moves to the pouring position of the subsequent region after completing the pouring of the preceding region, the preceding region has already completely solidified or is close to completely solidified. At this time, the preceding region cannot provide effective heat compensation to the subsequent region, and the adjacent region pair is marked as a failed transfer. When the expected transfer time is greater than the solidification start time of the subsequent region, the solidification start time of the subsequent region refers to the time when the subsequent region begins to precipitate solid phase under the thermal influence of the adjacent poured region, which is usually 10%-20% of the solidification time of the subsequent region. If the expected transfer time exceeds this time, it means that the subsequent region has started the solidification process in advance due to heat conduction from the adjacent region before the start of pouring, resulting in a shortened actual pourable time window for the subsequent region. The adjacent region pair is also marked as a failed transfer. The number of failure transfers in each candidate pouring sequence is counted. This number reflects the risk level of solidification discontinuity or feeding failure that may occur during the actual pouring operation. Candidate pouring sequence schemes with more than a preset failure threshold are eliminated. The preset failure threshold is determined based on the quality requirements of the casting. For high-quality castings, it is usually set to 5%-10% of the total number of adjacent regions in the candidate pouring sequence scheme, while for general castings, it can be set to 15%-20%.
[0041] The solidification time difference sequence of adjacent pouring region pairs in the remaining candidate pouring sequence schemes is extracted. This sequence, arranged sequentially according to the pouring order, reflects the pattern of solidification time variation at each stage of the pouring process. The solidification time difference sequence is then segmented and clustered to identify inflection points in the trend of solidification time difference variation. Inflection points are defined as locations where significant jumps or changes in the trend of solidification time difference occur. A sliding window method is used to calculate the local variance of the solidification time difference sequence, with a window length of 3-5 adjacent region pairs. A location is identified as an inflection point when its local variance exceeds 1.5-2 times the overall sequence variance. Based on these inflection point locations, the candidate pouring sequence schemes are divided into multiple solidification stages. Each solidification stage corresponds to a group of regions with similar solidification time variation characteristics during the pouring process. This division into solidification stages makes it possible to apply differentiated solidification control strategies to different stages.
[0042] The fluctuation range of the setting time difference within each solidification stage is calculated. This fluctuation range is obtained by calculating the standard deviation of all setting time differences within that stage. A smaller fluctuation range indicates more balanced solidification behavior across different areas within that stage, while a larger fluctuation range indicates uneven distribution of setting time. Simultaneously, the fluctuation range of the expected transfer time within each solidification stage is calculated. This expected transfer time fluctuation range reflects the stability of the nozzle movement time of the casting equipment within that stage. The compatibility and matching degree between the setting time difference fluctuation range and the expected transfer time fluctuation range of the corresponding solidification stage is evaluated. This compatibility and matching degree reflects the degree of coordination between changes in setting time and changes in casting operation time. When the setting time difference fluctuation range is large and the expected transfer time fluctuation range is small in a certain solidification stage, it indicates that although the solidification characteristics of different areas vary significantly within that stage, the casting operation can maintain a stable time rhythm, indicating good compatibility and matching. When both the setting time difference fluctuation range and the expected transfer time fluctuation range are large in a certain solidification stage, it indicates that the differences in solidification characteristics and the fluctuations in casting operation are superimposed within that stage, resulting in poor compatibility and matching.
[0043] Based on the synergy matching degree evaluation results, a solidification matching degree index is constructed. This index comprehensively considers the synergy matching degree of each solidification stage and the uniformity of the solidification time distribution across the entire candidate casting sequence. For a candidate casting sequence containing multiple solidification stages, its solidification matching degree index Q can be expressed as: Where N is the number of solidification stages, w i σ is the weighting coefficient for the i-th solidification stage, which is determined based on the number of regions included in this solidification stage and the proportion of solidification time to the entire casting process. t,i σ represents the fluctuation range of the solidification time difference in the i-th solidification stage. m,iLet α be the expected transfer time fluctuation range for the i-th solidification stage, and α be the coordination adjustment coefficient, typically ranging from 0.3 to 0.7, used to adjust the relative influence of solidification time fluctuation and transfer time fluctuation on the matching degree index. The larger the solidification matching degree index value, the more balanced the solidification time distribution and the better the time coordination of the pouring operation of the candidate pouring sequence scheme.
[0044] The candidate pouring sequence corresponding to the highest solidification matching index was selected as the target pouring sequence. This target pouring sequence achieved optimal coordination between solidification time distribution and pouring operation time among all candidate sequences, minimizing casting defects such as shrinkage cavities and porosity caused by uneven solidification, while ensuring the feasibility and stability of the pouring operation. For candidate pouring sequence sequences with multiple solidification matching index values close together, the pouring sequence relationship between the thin-walled and thick-walled regions was further compared, and the sequence that enables effective heat compensation from the thick-walled region to the thin-walled region was prioritized as the target pouring sequence.
[0045] The above technical solution enables quantitative evaluation and optimization of multiple candidate pouring sequence schemes based on solidification matching index, solves the technical problem of not being able to accurately match the solidification characteristics of each region in the existing technology, and obtains the target pouring sequence that can achieve the best overall solidification quality of the casting.
[0046] In some preferred embodiments, molten metal is poured into each region sequentially according to the target pouring order. During the pouring process, the surface temperature distribution of the poured regions is collected to calculate the actual liquid phase fraction, including: According to the target pouring sequence, molten metal is poured into each area in sequence. After the pouring of a single area is completed, the surface temperature distribution data of the poured area is collected. The temperature gradient direction and gradient magnitude in the surface temperature distribution data are identified. The main heat dissipation path of the area is determined based on the temperature gradient direction. The thickness distribution features of the region are extracted along the main heat dissipation path. The temperature values of each sampling point in the surface temperature distribution data are correlated and mapped with the thickness values at the corresponding locations to construct a temperature-thickness correlation matrix. Based on the correspondence between surface temperature and thickness in the temperature-thickness correlation matrix, the temperature distribution of each depth layer in the region is calculated by combining the preset temperature attenuation coefficient. The depth layer with a temperature value higher than the liquidus temperature is extracted as the liquidus layer, and the depth layer with a temperature value between the liquidus temperature and the solidus temperature is extracted as the solid-liquid coexistence layer. The distribution ratio of the liquid phase layer and the solid-liquid coexistence layer at various thicknesses in the region is statistically analyzed, and the actual liquid phase fraction of the cast area is calculated based on the distribution ratio.
[0047] For example, molten metal is poured into each area sequentially according to the target pouring order, and surface temperature distribution data for that area is collected immediately after pouring is completed. Taking a snowflake-shaped casting as an example, the pouring operation is performed in the order of first pouring the lower half thin-walled hollow area and recessed part, and then pouring the upper half thick-walled radiating arm. 30-45 seconds after pouring a certain hollow area in the lower half, an infrared thermal imager is used to scan the surface temperature of that area to obtain surface temperature distribution data. This temperature distribution data includes the temperature values of multiple sampling points in that area, with the sampling point spacing set to 5mm-10mm to ensure that the spatial variation characteristics of the temperature distribution can be reflected.
[0048] Identify the direction and magnitude of the temperature gradient in the surface temperature distribution data. The temperature gradient direction refers to the spatial direction of the fastest temperature decrease, reflecting the main heat transfer path within the region. For the cast hollowed-out area, extract the temperature difference between each sampling point and its adjacent sampling points to calculate the temperature gradient vector at each location. At the center of the hollowed-out area, the surface temperature is 620℃-640℃, while at the edge of the hollowed-out area in contact with the mold, the surface temperature is 580℃-600℃, a temperature difference of approximately 40℃. The temperature gradient direction points from the center to the edge, indicating that the main heat dissipation path in this area is through the edge surface in contact with the mold, dissipating heat outwards. For the thick-walled radiating arm region, the surface temperature in the middle is 650℃-670℃, and the surface temperature of the protruding structures at the end corners is 610℃-630℃. The temperature gradient direction points from the middle to the end. Simultaneously, the internal temperature is higher than the surface temperature along the wall thickness direction, with the temperature gradient direction pointing from the inside to the surface. This region exhibits two main heat dissipation paths: heat dissipation along the length of the radiating arm towards the end and heat dissipation along the wall thickness direction towards the surface.
[0049] After determining the main heat dissipation path of the region based on the temperature gradient direction, the thickness distribution characteristics of the region are extracted along the main heat dissipation path. For the hollow area, the wall thickness data at each position along the heat dissipation path from the center to the edge is extracted: the wall thickness at the center is 5mm, the wall thickness at 20mm from the center is 4mm, and the wall thickness at the edge position 40mm from the center is 3mm. For the thick-walled radiating arm region, the wall thickness at the center is 12mm, the wall thickness at 50mm from the center is 10mm, and the wall thickness at the end is 8mm. The temperature values of each sampling point in the surface temperature distribution data are correlated and mapped with the corresponding thickness values to construct a temperature-thickness correlation matrix. Each row of this matrix corresponds to a sampling point, containing the spatial coordinates of the sampling point, the surface temperature value, and the wall thickness value at that position. This matrix establishes the correspondence between surface temperature and local thickness.
[0050] Based on the correlation between surface temperature and thickness in the temperature-thickness relationship matrix, the temperature distribution at each depth within the region is calculated using a preset temperature decay coefficient. The temperature decay coefficient reflects the temperature decrease from the inside of the casting to the surface, and this coefficient is related to the thermal conductivity of the material, the cooling intensity of the mold, and the time elapsed after pouring. For aluminum alloy castings, the temperature decay coefficient is set to a temperature increment of 2℃-3℃ per millimeter of depth during the 30-45 seconds after pouring. For the center of the hollowed-out area, with a surface temperature of 630℃, a wall thickness of 5mm, and an internal center layer depth of 2.5mm from the surface, the internal center layer temperature is calculated to be 630 + 2.5 × 2.5 = 636.25℃ based on the temperature decay coefficient and the 2.5℃ temperature increment per millimeter of depth. Since this calculation does not consider the temperature accumulation effect during the initial pouring stage, the actual internal center layer temperature should be between 650℃ and 670℃. For the middle part of the thick-walled radiant arm, the surface temperature is 660℃, the wall thickness is 12mm, and the depth of the inner central layer from the surface is 6mm. According to the same method, the temperature of the inner central layer is approximately 690℃-710℃.
[0051] The temperature distribution of each depth layer within the region was calculated using the method described above, with the depth layer interval set at 0.5mm-1mm. Depth layers with temperatures above the liquidus temperature were identified as liquid phase layers, while those between the liquidus and solidus temperatures were identified as solid-liquid coexistence layers. For the aluminum alloy material used, the liquidus temperature was 615℃ and the solidus temperature was 580℃. At the center of the hollowed-out area, the internal center layer temperature was 650℃-670℃, above the liquidus temperature of 615℃, and was identified as a liquid phase layer. Temperatures at depths of 1mm-2mm from the surface were 600℃-620℃, partially between the liquidus and solidus temperatures, and were identified as solid-liquid coexistence layers. Temperatures at depths of 0mm-1mm from the surface were 580℃-600℃, and were identified as either solid-liquid coexistence layers or completely solidified solid layers. For the middle part of the thick-walled radiant arm, the internal central layer temperature is 690℃-710℃, which is identified as a liquid phase layer. The temperature at a depth of 2mm-4mm from the surface is 600℃-650℃, which is partially identified as a liquid phase layer and partially as a solid-liquid coexistence layer. The temperature at a depth of 0mm-2mm from the surface is 580℃-600℃, which is identified as a solid-liquid coexistence layer or a solid phase layer.
[0052] The distribution ratios of the liquid phase layer and the solid-liquid coexistence layer at various thicknesses within the region are statistically analyzed, and the actual liquid phase fraction of the cast area is calculated based on these ratios. The calculation of the actual liquid phase fraction comprehensively considers the volume ratio of the pure liquid phase layer and the volume ratio of the liquid phase in the solid-liquid coexistence layer. The volume ratio of the liquid phase in the solid-liquid coexistence layer is approximated as 50% of the solid-liquid coexistence layer volume, following a linear temperature distribution within the layer. For the center of the hollowed-out area, with a wall thickness of 5mm, the pure liquid phase layer thickness is approximately 1mm-1.5mm, the solid-liquid coexistence layer thickness is approximately 2mm-2.5mm, and the cured layer thickness is approximately 1mm-1.5mm, resulting in an actual liquid phase fraction of approximately 40%-50% at this location. For the middle of the thick-walled radiant arm, with a wall thickness of 12mm, the pure liquid phase layer thickness is approximately 3mm-4mm, the solid-liquid coexistence layer thickness is approximately 4mm-5mm, and the cured layer thickness is approximately 3mm-4mm, resulting in an actual liquid phase fraction of approximately 45%-55% at this location.
[0053] The actual liquid phase fraction of all sampling points within the region is weighted and averaged, with the weights determined based on the volume represented by each sampling point, to obtain the overall actual liquid phase fraction of the region. This actual liquid phase fraction serves as the basis for judging the solidification state of the region and is compared with a preset liquid phase fraction. The preset liquid phase fraction is determined based on the ideal solidification state that the region should achieve at the corresponding pouring time in the target pouring sequence, and is typically set to 30%-40%. This range ensures that the region has formed a sufficient solid phase framework to support subsequent pouring operations, while avoiding over-solidification that would prevent subsequent regions from receiving effective heat compensation. When the deviation between the actual liquid phase fraction and the preset liquid phase fraction exceeds a preset deviation threshold, it indicates an abnormal solidification rate during the current pouring sequence, requiring adjustment of the pouring order of unpollinated areas in the target pouring sequence. The preset deviation threshold is set to 8%-12%, determined based on the temperature fluctuation range during the casting solidification process and the error range of the liquid phase fraction calculation. Exceeding this threshold will significantly affect the solidification quality of subsequent pouring areas due to solidification state deviation.
[0054] In some preferred embodiments, when the deviation between the actual liquid phase fraction and the preset liquid phase fraction exceeds a preset deviation threshold, the pouring order of the unpoured areas in the target pouring sequence is adjusted, including: The difference between the actual liquid phase fraction of the poured area and the corresponding preset liquid phase fraction is calculated. When the absolute value of the difference exceeds the preset deviation threshold, the thickness distribution characteristics of the poured area and the pouring time interval of the adjacent poured areas are extracted. Based on the thickness distribution characteristics and the pouring time interval, the influencing factors that cause the liquid phase fraction deviation are identified. In the uncast areas of the target pouring sequence, calculate the heat conduction path length between each uncast area and the cast area. Select uncast areas with heat conduction path lengths less than a preset path threshold as sensitive areas associated with the influencing factor. Calculate the solidification time coupling degree between each sensitive area and the cast area. Prioritize the sensitive areas according to the solidification time coupling degree. For the highest priority sensitive area, the position of the sensitive area in the target pouring sequence is adjusted according to the deviation direction of the actual liquid fraction. The adjusted pouring sequence is substituted into the calculation process of the solidification matching degree index for iterative verification. The pouring sequence that makes the actual liquid fraction deviation converge to within the preset deviation threshold is extracted as the updated target pouring sequence.
[0055] For example, after pouring a single area, surface temperature distribution data of the poured area is collected. This data is acquired using an infrared thermal imaging device at a frequency of 1-5 frames per second, covering all poured areas to ensure real-time capture of surface temperature changes in each area. The direction and magnitude of the temperature gradient in the surface temperature distribution data are identified. The temperature gradient direction reflects the dominant direction of heat transfer in that area, while the magnitude reflects the strength of the heat dissipation rate. The temperature gradient is obtained by calculating the ratio of the temperature difference between adjacent temperature sampling points to their spatial distance, with the gradient direction pointing in the direction of the fastest temperature decrease. Based on the temperature gradient direction, the main heat dissipation path for that area is determined. This path refers to the dominant channel through which heat is transferred from the inside of the casting to the external environment. This path is typically perpendicular to the casting surface or extends along the cavity wall; in castings with complex structures, it may also extend along the contact surface between the molten metal and the core.
[0056] Thickness distribution features are extracted along the main heat dissipation path, including local thickness values and thickness change rates along the heat dissipation path. The temperature values of each sampling point in the surface temperature distribution data are correlated with the corresponding thickness values. This correlation mapping is achieved by matching the spatial coordinates of the surface temperature sampling points with the thickness sampling points in the 3D model. For cases where the surface temperature and thickness sampling points do not completely overlap, the nearest neighbor interpolation method is used to establish the correspondence. A temperature-thickness correlation matrix is constructed, where the rows correspond to different surface locations in the cast area, and the columns correspond to the surface temperature value, local thickness value, and depth coordinates along the heat dissipation path. This temperature-thickness correlation matrix establishes a quantitative relationship between observable surface temperature and unobservable internal structural features.
[0057] Based on the correlation between surface temperature and thickness in the temperature-thickness correlation matrix, the temperature distribution of each depth layer within the region is calculated using a preset temperature attenuation coefficient. The temperature attenuation coefficient reflects the attenuation of temperature as it travels along the thickness direction from the surface to the interior. This coefficient is related to the thermal conductivity, specific heat capacity, and current solidification stage of the casting material. For copper castings, the temperature attenuation coefficient is typically set to 0.15℃ / mm-0.25℃ / mm; for titanium castings, it is set to 0.20℃ / mm-0.35℃ / mm; and for magnesium alloy castings, it is set to 0.10℃ / mm-0.20℃ / mm. The temperature of each depth layer is obtained by subtracting the distance between that depth and the surface from the surface temperature and multiplying by the temperature attenuation coefficient. This calculation method is based on the steady-state heat conduction assumption and is suitable for stages where temperature changes are relatively slow during solidification. For the initial pouring stage where temperature changes rapidly, the temperature attenuation coefficient needs to be dynamically corrected based on the time interval between the current moment and the completion of pouring. The correction coefficient decreases as the time interval increases, reflecting the physical process of the temperature field gradually stabilizing.
[0058] The depth layer with a temperature above the liquidus temperature is extracted as the liquid phase layer. The metal in this liquid phase layer is completely liquid and possesses good fluidity and feeding ability. The depth layer with a temperature between the liquidus and solidus temperatures is extracted as the solid-liquid coexistence layer. In this layer, the liquid and solid phases coexist, and its fluidity and feeding ability decrease as the liquid phase fraction decreases. The liquidus and solidus temperatures are determined based on the phase diagram of the casting material. For pure metals, the liquidus and solidus temperatures are the same; for alloys, there is a temperature range between them. For example, the liquidus temperature of copper alloys is typically 1050℃-1100℃, and the solidus temperature is 980℃-1030℃; for titanium alloys, the liquidus temperature is 1650℃-1680℃, and the solidus temperature is 1600℃-1640℃. The depth layer with a temperature below the solidus temperature is identified as a fully solidified layer, which has lost its fluidity and feeding ability.
[0059] The distribution ratio of the liquid phase layer and the solid-liquid coexistence layer at various thicknesses within the region is statistically analyzed. This ratio is obtained by calculating the ratio of the total thickness of the liquid phase layer and the solid-liquid coexistence layer to the total thickness of the casting at that location. Since the liquid phase fraction in the solid-liquid coexistence layer exhibits a gradient distribution along the thickness direction, the local liquid phase fraction needs to be estimated based on the relative position of the temperature between the liquidus and solidus temperatures. The relationship between temperature and liquid phase fraction is established using the Scheil model or the lever law. For a location in the solid-liquid coexistence layer with temperature T, the local liquid phase fraction f is calculated. L It can be represented as , among which, T l T is the liquidus temperature. sLet T be the solidus temperature, and T be the current temperature at that location. The actual liquid phase fraction of the cast area is calculated based on the distribution ratio. The actual liquid phase fraction is the ratio of the sum of the liquid phase layer thickness at all locations within the area plus the weighted thickness of the liquid phase portion in the solid-liquid coexistence layer to the total volume of the area. For irregularly shaped areas, a volume integral method is used to calculate the actual liquid phase fraction. The area is divided into multiple thin-layer units along the heat dissipation path, the liquid phase volume of each thin-layer unit is calculated and summed, and then divided by the total volume of the area to obtain the actual liquid phase fraction.
[0060] The calculated actual liquid fraction reflects the current solidification process of the cast area. This parameter directly relates to whether the area still has the ability to provide heat compensation to adjacent uncast areas or subsequently cast areas. When the actual liquid fraction is higher than 60%, the area is in the early stage of solidification, possessing strong fluidity and compensation capacity. When the actual liquid fraction decreases to between 30% and 60%, the area is in the middle stage of solidification, with weakened compensation capacity but still able to provide limited heat transfer. When the actual liquid fraction is lower than 30%, the area is nearing the end of solidification, essentially losing its compensation capacity. At this point, if adjacent areas have not yet been cast, it may lead to blockage of the compensation channels, resulting in shrinkage cavities.
[0061] The above technical solution realizes a method for calculating the actual liquid phase fraction of the poured area based on surface temperature distribution data, effectively solving the technical problem of not being able to obtain the solidification state inside the casting in real time, and providing accurate state monitoring data for dynamically adjusting the pouring sequence according to the deviation between the actual liquid phase fraction and the preset liquid phase fraction.
[0062] In some preferred embodiments, before performing the casting flipping operation, it is determined whether the liquid phase fraction in both the thin-walled and thick-walled regions is less than a preset flipping liquid phase threshold. If not, the cooling intensity is adjusted until the preset flipping liquid phase threshold is met before performing the casting flipping operation, including: Before performing the casting flipping operation, the flipping direction and flipping angular velocity of the casting to be performed are obtained. Based on the casting flipping direction, the spatial position relationship of each region of the casting relative to the flipping axis is identified, and the distribution of inertial forces borne by the thin-walled region and the thick-walled region during the flipping process is calculated. The liquid phase stability coefficient of each region is determined based on the inertial force distribution and thickness distribution characteristics of each region. The actual liquid phase inversion threshold of each region is obtained by coupling the liquid phase stability coefficient with the preset inversion liquid phase threshold. The surface temperature distribution of thin-walled and thick-walled regions is collected to calculate the liquid phase fraction in each region. Regions with liquid phase fraction exceeding the actual inverted liquid phase threshold are identified as regions to be cooled. The target cooling medium flow rate increase is determined based on the excess liquid phase fraction and inertial force distribution characteristics of the regions to be cooled. Adjust the cooling medium flow rate of the corresponding cooling nozzle in the area to be cooled according to the increase in the target cooling medium flow rate, continuously monitor the change in liquid phase fraction in the area to be cooled, and perform casting flipping operation when the liquid phase fraction in both the thin-walled and thick-walled areas meets their respective actual flipping liquid phase thresholds.
[0063] For example, before performing the casting flipping operation, the flipping direction and flipping angular velocity of the casting to be performed are obtained. The casting flipping direction is determined by the structural characteristics and process requirements of the casting, and is usually flipped along the axis of symmetry or the axis of minimum moment of inertia of the casting. The flipping angular velocity is set according to the control capability of the flipping equipment, generally 0.5 rad / s-2 rad / s. Based on the casting flipping direction, the spatial positional relationship of each region of the casting relative to the flipping axis is identified. The spatial positional relationship includes the vertical distance from the geometric center point of each region to the flipping axis and the azimuth angle of the region relative to the flipping axis. These parameters directly affect the magnitude of the centrifugal inertial force and tangential inertial force experienced by each region during the flipping process. The distribution of inertial forces experienced by the thin-walled and thick-walled regions during the flipping process is calculated. The inertial forces include the tangential inertial force generated by the flipping angular acceleration and the centrifugal inertial force generated by the flipping angular velocity. The tangential inertial force is proportional to the distance from the region to the flipping axis and the flipping angular acceleration, while the centrifugal inertial force is proportional to the distance from the region to the flipping axis and the square of the flipping angular velocity. For areas farther from the flip axis, the inertial force they experience is significantly greater than that of areas closer to the flip axis. In these areas, the liquid metal is more likely to flow or separate from the cavity wall under the action of inertial force.
[0064] The liquid phase stability coefficient for each region is determined based on the distribution of inertial forces and thickness in each region. The liquid phase stability coefficient reflects the ability of the liquid metal within that region to maintain a stable state under inertial forces. For thick-walled regions, due to the larger volume of the internal liquid metal and the weaker constraint from the surrounding solid metal, inertial forces more easily induce macroscopic flow of the liquid metal, resulting in a lower liquid phase stability coefficient. For thin-walled regions, although the absolute value of the inertial force may be smaller, the thinner wall thickness results in stronger constraint from the cavity wall, leading to a higher liquid phase stability coefficient. The liquid phase stability coefficient is calculated by comparing the inertial force of each region with its characteristic modulus. Thick-walled regions with larger characteristic moduli have lower liquid phase stability coefficients under the same inertial force, while thin-walled regions with smaller characteristic moduli have higher liquid phase stability coefficients under the same inertial force. The actual liquid phase threshold for each region is calculated by coupling the liquid phase stability coefficient with a preset liquid phase reversal threshold. The preset liquid phase reversal threshold is a general liquid phase fraction limit determined based on experience or experimentation, typically set at 20%-35%, indicating that the casting can be safely reversed when the liquid phase fraction is below this value. The actual liquid phase reversal threshold is adjusted based on the liquid phase stability coefficient of each region. For regions with low liquid phase stability coefficients, the actual liquid phase reversal threshold should be lower than the preset threshold to provide a greater safety margin. For regions with high liquid phase stability coefficients, the actual liquid phase reversal threshold can be appropriately higher than the preset threshold to avoid excessively prolonging the solidification time.
[0065] The surface temperature distribution of thin-walled and thick-walled regions is collected to calculate the liquid phase fraction in each region. The calculation method for the liquid phase fraction is the same as the aforementioned real-time monitoring process. The internal temperature distribution is inferred from the surface temperature, and the distribution ratio of the liquid phase layer and the solid-liquid coexistence layer is identified. Regions where the liquid phase fraction exceeds the actual liquid phase reversal threshold are identified as areas requiring enhanced cooling. These regions need to accelerate their solidification process by increasing the cooling intensity to reduce their liquid phase fraction below the actual liquid phase reversal threshold, thereby meeting the safety conditions for the reversal operation. The target cooling medium flow rate increase is determined based on the excess liquid phase fraction and inertial force distribution characteristics of the areas requiring enhanced cooling. The excess liquid phase fraction is the difference between the current liquid phase fraction in that region and the actual liquid phase reversal threshold. The larger the difference, the more heat needs to be removed, and the greater the required increase in cooling medium flow rate. The inertial force distribution characteristics affect the spatial allocation of the cooling medium flow rate increase. For locations subjected to greater inertial forces, even if the excess liquid phase fraction is the same, a larger increase in cooling medium flow rate is required to ensure that the liquid phase fraction at that location is reduced preferentially.
[0066] Determining the target cooling medium flow rate increase includes the following steps: First, calculate the heat value that needs to be removed from the area to be cooled. This heat value is calculated based on the excess liquid fraction, the volume of the area, the metal density, and the latent heat of solidification. The calculation formula is as follows: Where R is the heat to be removed, V is the volume of the region to be cooled, ρ is the density of the molten metal, and L... f The latent heat of solidification of the metal, Δf L The excess liquid phase fraction is then calculated; subsequently, the heat that can be removed per unit flow rate of the cooling medium is calculated. This calculation is based on the specific heat capacity of the cooling medium, its flow rate, and the heat transfer coefficient of the casting surface. The calculation formula is as follows: Where q is the heat that can be removed per unit flow rate of cooling medium per unit time. c is the mass flow rate of the cooling medium. p Let η be the specific heat capacity of the cooling medium, ΔT be the temperature difference between the inlet and outlet of the cooling medium, and η be the heat transfer efficiency coefficient. The heat transfer efficiency coefficient depends on the heat transfer coefficient of the casting surface, the distance between the nozzle and the casting surface, and the flow state of the cooling medium, and is typically taken in the range of 0.4-0.8. Finally, the required increase in cooling medium flow rate is calculated based on the amount of heat to be removed and the amount of heat that can be removed per unit flow rate of the cooling medium. The calculation formula is as follows: ,in For the increase in cooling medium flow rate, t co For the expected duration of enhanced cooling, the target increase in cooling medium flow rate is the ratio of the increase in cooling medium flow rate to the base flow rate. When determining the target increase in cooling medium flow rate for each area to be enhanced, it is also necessary to make corrections based on the characteristics of inertial force distribution. For areas subjected to greater inertial force, the calculated target increase in cooling medium flow rate is multiplied by an inertial force correction factor. The inertial force correction factor is determined based on the ratio of the inertial force experienced by the area to the average inertial force, and is typically set between 1.1 and 1.5 to ensure that areas subjected to greater inertial force receive higher cooling priority.
[0067] The target increase in cooling medium flow rate is determined by establishing a correlation model between the excess liquid phase fraction, the magnitude of inertial force, and the cooling medium flow rate. This model takes into account the specific heat capacity, flow rate, and heat transfer coefficient of the casting surface of the cooling medium. For water cooling, the increase in cooling medium flow rate is usually set to 30%-80% of the base flow rate, and for air cooling, the increase in cooling medium flow rate is usually set to 50%-150% of the base flow rate.
[0068] The cooling medium flow rate of the corresponding cooling nozzles in the area to be cooled is adjusted according to the increase in the target cooling medium flow rate. The cooling nozzles are arranged on the surface of the casting or the inner wall of the cavity. The cooling medium flow rate is precisely adjusted by controlling the valve opening or the pumping speed. The change in liquid phase fraction in the area to be cooled is continuously monitored, with a monitoring cycle of 10s-30s to ensure that the downward trend of liquid phase fraction can be captured in time. After the cooling intensity is enhanced, the surface temperature of the area to be cooled decreases, the temperature gradient increases along the heat dissipation path inward, the thickness of the liquid phase layer and the solid-liquid coexistence layer gradually decreases, and the liquid phase fraction decreases accordingly. For areas where the rate of liquid phase fraction decrease is lower than expected, the cooling medium flow rate is further increased or the spray angle of the cooling nozzles is adjusted so that the cooling medium can more effectively cover the high-temperature position in the area. When the liquid phase fraction in both the thin-walled and thick-walled areas meets their respective actual flipping liquid phase thresholds, the liquid metal in each area has basically solidified or only a small amount of liquid phase remains. This small amount of liquid phase is effectively constrained by the surrounding solid phase network and will not undergo macroscopic flow or segregation under the inertial force of the flipping operation. At this time, the casting flipping operation is performed.
[0069] The casting flipping operation requires coordination with the solidification control system. Before starting the flipping equipment, the current liquid phase fraction and surface temperature distribution of each area are recorded as a baseline before flipping. During the flipping process, the cooling system must be kept running normally to avoid interruption of the cooling medium supply or deviation of the spray direction from the target area due to the flipping. After the flipping is completed, the surface temperature distribution of each area is re-acquired, and the temperature changes before and after the flipping are compared to assess the impact of the flipping operation on the solidification process. For areas where the temperature distribution is abnormal or the liquid phase fraction rises due to the flipping operation, cooling parameters or subsequent pouring sequence should be adjusted in a timely manner to prevent quality defects.
[0070] Differential flipping thresholds are determined based on the distribution of inertial forces and liquid phase stability in each region, and the cooling intensity is dynamically adjusted to reduce the flipping quality risk caused by regional solidification differences, thereby improving flipping safety and the accuracy of solidification quality control.
[0071] Specifically, the application of this invention is illustrated using a snowflake-shaped decorative casting as an example. The casting is snowflake-shaped with six main radiating arms symmetrically distributed. Each radiating arm has a raised corner structure at its end for suspension and fixation. The area between the radiating arms forms a recessed portion for placement on a support member at a specific location. The upper half of the casting features a thicker decorative texture and reinforcing rib structure with a wall thickness of 8mm-12mm, while the lower half has a thin-walled hollow structure with a wall thickness of 3mm-5mm, creating a top-heavy, bottom-light mass distribution. This mass distribution helps the casting maintain a stable posture when suspended. Due to the asymmetrical geometry of the casting and the significant difference in wall thickness between the upper and lower parts, a slanted flipping method is used for pouring and solidification control. The flipping axis is set at a 45° angle to the horizontal centerline of the casting. °The oblique axis of the angle allows the casting to maintain the continuity of the feeding channel during the flipping process and reduces the tendency of shrinkage cavities and porosity in the upper thick-walled area.
[0072] Molten metal was poured into each area sequentially according to the target pouring sequence, following the order of pouring the lower half of the thin-walled hollow area and the recessed part first, and then pouring the upper half of the thick-walled radiating arm. 35 seconds after the lower half of the recessed part was poured, an infrared thermal imager was used to scan the surface temperature of this area to obtain surface temperature distribution data, with a sampling point spacing of 8 mm. The direction and magnitude of the temperature gradient in the surface temperature distribution data were identified. The surface temperature at the center of the recessed part was 605℃-625℃, while the surface temperature at the edge in contact with the mold was 565℃-585℃, a temperature difference of approximately 40℃. The temperature gradient direction was from the center to the edge, indicating that the main heat dissipation path in this area was through the edge surface to the mold. The surface temperature in the middle of the thick-walled radiating arm area was 640℃-665℃, and the surface temperature of the protruding structure at the end corners was 600℃-620℃. This area exhibited two main heat dissipation paths: along the length of the radiating arm and along the wall thickness.
[0073] After determining the main heat dissipation path based on the temperature gradient direction, the thickness distribution features are extracted along the heat dissipation path. The wall thickness at the center of the recessed part is 4 mm, the wall thickness at 15 mm from the center is 3.5 mm, and the wall thickness at the edge is 3 mm. The temperature values of each sampling point in the surface temperature distribution data are correlated and mapped with the corresponding thickness values to construct a temperature-thickness correlation matrix. Based on this matrix and a preset temperature decay coefficient, the temperature distribution state of each depth layer inside is calculated. For snowflake-shaped castings, the temperature decay coefficient is set to a temperature increment of 3℃-4℃ per millimeter of depth 35 seconds after pouring. For the center of the recessed part, the surface temperature is 615℃, the wall thickness is 4 mm, the depth of the inner center layer from the surface is 2 mm, and the temperature decay coefficient is taken as 3.5℃ / mm. According to the internal temperature calculation method, the temperature of the inner center layer is 615 + 2.0 × 3.5 = 622℃. Considering the temperature accumulation in the early stage of pouring, the actual temperature should be between 635℃ and 655℃. The depth layer with a temperature above the liquidus temperature (615°C) is designated as the liquid phase layer, and the depth layer with a temperature between the liquidus and solidus temperatures (580°C) is designated as the solid-liquid coexistence layer. The actual liquid phase fraction is calculated by statistically analyzing the distribution ratio of the liquid phase layer and the solid-liquid coexistence layer; the liquid phase volume ratio in the solid-liquid coexistence layer is approximately 50%. For the recessed portion, the pure liquid phase layer thickness is approximately 0.8mm to 1.2mm, and the solid-liquid coexistence layer thickness is approximately 1.8mm to 2.2mm, resulting in an actual liquid phase fraction of approximately 35% to 45%. In the middle of the thick-walled radiating arm, with a wall thickness of 12mm, the pure liquid phase layer thickness is approximately 4mm to 5mm, and the solid-liquid coexistence layer thickness is approximately 4.5mm to 5.5mm, resulting in an actual liquid phase fraction of approximately 50% to 60%.
[0074] The weighted average of the actual liquid fractions at all sampling points within the area yielded an overall actual liquid fraction of approximately 38% for the recessed portion. This was compared to the preset liquid fraction, which was set at 28%-35% based on the designed solidification state of the area. This range ensures that a sufficient solid skeleton has formed in the area. The deviation between the actual liquid fraction of 38% in the recessed portion and the upper limit of the preset liquid fraction (35%) was 3 percentage points, which did not exceed the preset deviation threshold of 10%, indicating that the solidification state was within the normal range. Fifty seconds after continuing to pour the thick-walled radiant arm area, the actual liquid fraction in the middle of the radiant arm was measured to be approximately 55%, while the preset liquid fraction was set at 35%-42%. This deviation reached 13 percentage points, exceeding the preset deviation threshold of 10%, indicating that the solidification rate in this area was lower than expected. Therefore, the pouring sequence of the remaining unpoured end corner protrusions in the target pouring sequence needs to be adjusted, and their pouring time should be appropriately delayed to avoid excessive heat concentration.
[0075] Before performing the casting flipping operation, obtain the casting flipping direction and angular velocity to be performed. The casting flipping direction is along a path that passes through the geometric center of the casting and forms a 45° angle with the horizontal plane. ° The tilting angle is along an oblique axis, which ensures a symmetrical spatial distribution of the thick-walled radial arms in the upper half of the casting and the thin-walled hollowed-out areas in the lower half relative to the tilting axis. The tilting angular velocity is set to 1.2 rad / s, which ensures both smooth tilting and allows for a 180° tilt within a reasonable time. ° Flipping operation. Based on the casting flipping direction, the spatial positional relationship of each region of the casting relative to the flipping axis is identified. In the upper half, among the six thick-walled radial arms, the end protrusions of three radial arms are 120mm-150mm vertically from the flipping axis, while the ends of the other three radial arms are 80mm-100mm vertically from the flipping axis. In the lower half, the thin-walled hollowed-out area is 50mm-90mm vertically from the flipping axis, and the deepest point of the recessed part is only 30mm from the flipping axis. The distribution of inertial forces borne by the thin-walled and thick-walled areas during the flipping process is calculated. For the end of the thick-walled radial arm 150mm from the flipping axis, the centrifugal inertial force generated at a flipping angular velocity of 1.2 rad / s is the product of the mass of that area and the square of 1.44 times the angular velocity plus the distance. This inertial force is significantly greater than that of the recessed part, which is only 30mm from the flipping axis. The upper thick-walled radiating arm, due to its greater wall thickness and greater distance from the rotation axis, bears a total inertial force that is 3-5 times greater than that of the lower thin-walled hollow area.
[0076] The liquid phase stability coefficient of each region is determined based on the distribution of inertial force and thickness in each region. The upper, thick-walled radiating arm has a characteristic modulus of 8mm-12mm and bears a large inertial force. In this region, the liquid metal volume is large, and the solid-phase network is not yet fully formed. The inertial force easily causes the liquid metal to flow inside the radiating arm or concentrate towards the end corner protrusions, leading to shrinkage cavities or component segregation. Its liquid phase stability coefficient is calculated as the ratio of inertial force to characteristic modulus; a smaller value indicates poor stability. The lower, thin-walled hollow region has a characteristic modulus of 3mm-5mm. Although the absolute value of the inertial force is small, the thin wall thickness provides strong constraint on the liquid metal from the cavity wall, and the connection points between the hollow structures have formed a relatively continuous solid-phase skeleton, resulting in a higher liquid phase stability coefficient. The recessed area has the highest liquid phase stability coefficient because it is closest to the flipping axis and bears the smallest inertial force. The actual liquid phase threshold for each region is calculated by coupling the liquid phase stability coefficient with a preset liquid phase reversal threshold. The preset liquid phase reversal threshold is set to 25%, indicating that the liquid phase fraction of the casting should be below 25% before reversal can be performed under the condition of no special inertial force. The liquid phase stability coefficient of the upper thick-walled radial arm is 0.6-0.7 times the preset value, and its actual liquid phase reversal threshold is corrected to 15%-18%. This threshold is lower than the preset value to compensate for the adverse effect of inertial force on liquid phase stability. The liquid phase stability coefficient of the lower thin-walled hollow area is 1.3-1.5 times the preset value, and its actual liquid phase reversal threshold is corrected to 32%-38%. This threshold is higher than the preset value to avoid excessively prolonging the solidification time. The liquid phase stability coefficient of the recessed part is 1.8 times the preset value, and its actual liquid phase reversal threshold is corrected to 45%.
[0077] The surface temperature distribution of the thin-walled and thick-walled regions was collected to calculate the liquid phase fraction in each region. 600 seconds after casting, the surface temperature distribution of the protruding structures at the corners of the six thick-walled radiating arms in the upper half was collected, with temperatures ranging from 580℃ to 620℃. The surface temperature in the middle of the radiating arms was 620℃ to 650℃. The internal temperature was estimated from the surface temperatures, and combined with the liquidus temperature (615℃) and solidus temperature (580℃) of the aluminum alloy, the liquid phase fraction at the ends of the thick-walled radiating arms was calculated to be 22%-28%, and the liquid phase fraction in the middle of the radiating arms was 30%-35%. The surface temperature of the thin-walled hollowed-out area in the lower half was 550℃-580℃, with a calculated liquid phase fraction of 8%-15%. The surface temperature of the recessed area was 530℃-560℃, with a liquid phase fraction of 5%-10%. Regions with liquid phase fractions exceeding the actual liquid phase reversal threshold were identified as areas requiring enhanced cooling. In the upper half, the liquid phase fractions of all six thick-walled radiant arms exceeded their actual liquid phase reversal thresholds by 15%-18%, with the largest excess occurring in the central region of the radiant arms, reaching 12-17 percentage points. The excess was 4-10 percentage points in the protruding structures at the ends of the radiant arms. These areas were all identified as requiring enhanced cooling. The liquid phase fractions of the thin-walled hollowed-out areas and recessed sections in the lower half were below their respective actual liquid phase reversal thresholds and did not require enhanced cooling.
[0078] The target cooling medium flow rate increment is determined based on the excess liquid phase fraction and inertial force distribution characteristics of the area to be cooled. The excess liquid phase fraction in the middle region of the radiant arm is 12-17 percentage points, and this region experiences significant inertial force; therefore, its liquid phase fraction needs to be reduced first, and the target cooling medium flow rate increment is set to 70% of the base flow rate. The excess liquid phase fraction in the protruding structures at the corners of the radiant arm is 4-10 percentage points, but this region is farthest from the flipping axis and experiences the greatest centrifugal inertial force; therefore, the target cooling medium flow rate increment is set to 55% of the base flow rate. For different locations with the same excess liquid phase fraction, locations experiencing greater inertial force are allocated a higher cooling medium flow rate increment to ensure that the liquid phase fraction at these locations is preferentially reduced to below the actual flipping liquid phase threshold.
[0079] The cooling medium flow rate of the corresponding cooling nozzles in the area to be enhanced cooling is adjusted according to the target cooling medium flow rate increase. The original cooling medium flow rate of the nozzles located in the middle region of the radiant arm was 15 L / min, which was adjusted to 25.5 L / min based on a 70% flow rate increase. The original cooling medium flow rate of the nozzles located at the corner protrusions at the end of the radiant arm was 12 L / min, which was adjusted to 18.6 L / min based on a 55% flow rate increase. The liquid phase fraction change in the area to be enhanced cooling is continuously monitored, with a monitoring cycle set to 15 seconds. Sixty seconds after the enhanced cooling starts, the surface temperature in the middle region of the radiant arm decreases to 590℃-610℃, and the liquid phase fraction decreases to 18%-22%. However, the liquid phase fraction in some locations is still slightly higher than the actual inversion threshold, so enhanced cooling continues. 120 seconds after the enhanced cooling was initiated, the liquid phase fraction in the middle region of the radiant arm decreased to 12% to 16%, and the liquid phase fraction in the protruding structures at the ends of the radiant arm decreased to 10% to 14%. The liquid phase fractions of all six thick-walled radiant arms met their respective actual liquid phase inversion thresholds of 15% to 18%. During this process, the liquid phase fraction in the thin-walled hollowed-out area and the recessed part of the lower half further decreased to 3% to 8%, also meeting their respective actual liquid phase inversion thresholds.
[0080] When the liquid phase fractions in both the thin-walled and thick-walled regions meet their respective actual inversion liquid phase thresholds, the inversion equipment is activated to perform the casting inversion operation. The casting rotates 180 degrees along the oblique axis at an angular velocity of 1.2 rad / s. ° The flipping process lasted approximately 2.6 seconds. After flipping, the original upper thick-walled radiant arm was positioned below, and the original lower thin-walled hollow area was positioned above. This position facilitates rapid cooling of the upper thin-walled area and sequential solidification and feeding of the lower thick-walled area. During the flipping process, because the liquid phase fraction in each region had decreased to a safe range, the small amount of liquid metal remaining in the thick-walled radiant arm was effectively constrained by the surrounding solid network, preventing macroscopic flow or concentration towards the corner protrusions, thus avoiding the generation of shrinkage cavities and compositional segregation defects. After flipping, the surface temperature distribution of each region was re-acquired. The surface temperature of the upper thin-walled hollow area was 510℃-540℃, and the surface temperature of the lower thick-walled radiant arm was 580℃-600℃. The temperature distribution was normal, and no local overheating or uneven cooling caused by the flipping operation was observed.
[0081] The above application examples demonstrate that, considering the top-heavy and bottom-light mass distribution of snowflake-shaped castings and the process characteristics of oblique flipping, the inertial force distribution is calculated based on the spatial positional relationship of each region relative to the oblique flipping axis. Combined with the thickness distribution characteristics, differentiated actual flipping liquid phase thresholds are determined. Thick-walled radiant arm regions with liquid phase fractions exceeding the threshold are subjected to enhanced cooling to ensure that the flipping operation is carried out under safe conditions. This effectively avoids quality defects caused by liquid metal flow during the flipping process and guarantees the solidification quality and dimensional accuracy of the casting.
[0082] Please see Figure 2 Using the same inventive concept as the above embodiments, this application also provides a casting solidification quality control system, comprising: The 3D model module is used to obtain the 3D model of the casting to be poured, identify the thin-walled area and the thick-walled area in the 3D model, and calculate the thin-walled solidification time corresponding to the thin-walled area and the thick-walled solidification time corresponding to the thick-walled area. The sequential casting module is used to generate multiple candidate casting sequence schemes. Each candidate casting sequence scheme includes different arrangements and combinations of thin-walled and thick-walled regions. For each candidate casting sequence scheme, the solidification matching index is calculated based on the solidification time and casting interval time of each region. The candidate casting sequence scheme with the best solidification matching index is selected as the target casting sequence. The sorting adjustment module is used to pour molten metal into each area in sequence according to the target pouring order. During the pouring process, the surface temperature distribution of the poured area is collected to calculate the actual liquid phase fraction. When the deviation between the actual liquid phase fraction and the preset liquid phase fraction exceeds the preset deviation threshold, the pouring order of the unpoured areas in the target pouring order is adjusted. The flipping confirmation module is used to determine whether the liquid phase fraction in both the thin-walled and thick-walled regions is less than the preset flipping liquid phase threshold before performing the casting flipping operation. If not, the cooling intensity is adjusted until the preset flipping liquid phase threshold is met before performing the casting flipping operation.
[0083] It should be noted that although several units or sub-units of the device have been mentioned in the detailed description above, this division is merely exemplary and not mandatory. In fact, according to embodiments of this application, the features and functions of two or more units described above can be embodied in one unit. Conversely, the features and functions of one unit described above can be further divided and embodied by multiple units.
[0084] Based on the same inventive concept as the above method embodiments, this application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it enables the electronic device to implement the control method described in the above embodiments.
[0085] In one embodiment, the electronic device may be a server, and in this embodiment, the structure of the electronic device may be as follows: Figure 3 As shown, it includes a memory 2001, a communication module 2003, and one or more processors 2002.
[0086] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.
[0087] Finally, it should be noted that the above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
Claims
1. A method for controlling the solidification quality of castings, characterized in that: The method includes the following steps: Obtain a three-dimensional model of the casting to be poured, identify the thin-walled region and the thick-walled region in the three-dimensional model, and calculate the thin-walled solidification time corresponding to the thin-walled region and the thick-walled solidification time corresponding to the thick-walled region respectively. Multiple candidate pouring sequence schemes are generated. Each candidate pouring sequence scheme includes different arrangements and combinations of the thin-walled region and the thick-walled region. For each candidate pouring sequence scheme, a solidification matching index is calculated based on the solidification time and pouring interval time of each region. The candidate pouring sequence scheme with the best solidification matching index is selected as the target pouring sequence. According to the target pouring sequence, molten metal is poured into each area in sequence. During the pouring process, the surface temperature distribution of the poured area is collected to calculate the actual liquid phase fraction. When the deviation between the actual liquid phase fraction and the preset liquid phase fraction exceeds the preset deviation threshold, the pouring order of the unpoured areas in the target pouring sequence is adjusted. Before performing the casting flipping operation, it is determined whether the liquid phase fraction of the thin-walled region and the thick-walled region is less than the preset flipping liquid phase threshold. If not, the cooling intensity is adjusted until the preset flipping liquid phase threshold is met before performing the casting flipping operation.
2. The method for controlling the solidification quality of castings according to claim 1, characterized in that: Obtain a three-dimensional model of the casting to be poured, identify thin-walled and thick-walled regions in the three-dimensional model, and calculate the thin-walled solidification time corresponding to the thin-walled region and the thick-walled solidification time corresponding to the thick-walled region, including: The coverage area of the controllable area of the nozzle is determined according to the nozzle diameter of the pouring nozzle, and the three-dimensional model is divided into several evaluation units according to the coverage area of the controllable area of the nozzle. For each evaluation unit, its geometric shape type is determined. When the evaluation unit is a regular trapezoidal structure, the wall thickness is sampled proportionally along the height direction of the trapezoid and the wall thickness value of each sampling position is marked. When the evaluation unit is an irregular structure, wall thickness sampling points are arranged in the evaluation unit according to preset mechanical constraint points and dimensional key points and the wall thickness value of each wall thickness sampling point is marked. Calculate the unit average wall thickness of each evaluation unit, identify the evaluation unit with the unit average wall thickness less than the preset thin wall threshold as the thin wall region, and identify the evaluation unit with the unit average wall thickness greater than the preset thick wall threshold as the thick wall region; Feature annotation is performed on local locations within the thin-walled region and the thick-walled region that deviate from the unit average wall thickness by more than a preset deviation threshold; When calculating the thin-wall solidification time corresponding to the thin-walled region and the thick-wall solidification time corresponding to the thick-walled region, the feature modulus of each region is calculated based on the unit average wall thickness, and the thin-wall solidification time and the thick-wall solidification time are corrected based on the degree of wall thickness deviation at the feature marking position.
3. The method for controlling the solidification quality of castings according to claim 1, characterized in that: Multiple candidate pouring sequence schemes are generated, each of which includes different arrangements and combinations of the thin-walled region and the thick-walled region, including: Obtain the spatial coordinates of the thin-walled region and the thick-walled region, and construct a region adjacency graph based on the geometric distance and connectivity between each region. The region adjacency graph uses each region as a node and the geometric connection channel between adjacent regions as an edge. Identify the solidification time difference characteristics of each node in the region adjacency graph, compare the solidification time difference between adjacent nodes with multiple preset difference level thresholds, and label each edge with the corresponding difference level according to the threshold range in which the solidification time difference is located. In the adjacency graph of the region, multiple traversal paths are generated starting from the initial pouring position. Each traversal path passes through edges of different difference levels in sequence to reach the final pouring position. Each traversal path corresponds to a candidate pouring sequence scheme.
4. The method for controlling the solidification quality of castings according to claim 1, characterized in that: For each candidate pouring sequence scheme, a solidification matching index is calculated based on the solidification time and pouring interval time of each region. The candidate pouring sequence scheme with the optimal solidification matching index is selected as the target pouring sequence, including: The spatial coordinates of adjacent pouring area pairs in a single candidate pouring sequence scheme are extracted sequentially, the nozzle movement path length between adjacent area pairs is calculated, and the expected transfer time between each adjacent area pair is determined according to the preset nozzle movement speed. The expected transfer time is compared with the solidification time difference of the adjacent region pair. When the expected transfer time exceeds the solidification time of the preceding region or is greater than the solidification start time of the subsequent region, it is marked as a failed transfer. The number of failed transfers in each of the candidate casting sequence schemes is counted and the candidate casting sequence schemes with the number exceeding the preset failure threshold are removed. Extract the solidification time difference sequence of adjacent pouring area pairs in the remaining candidate pouring sequence schemes, perform segmented clustering on the solidification time difference sequence to identify the inflection point position of the change trend of the solidification time difference, and divide the candidate pouring sequence scheme into multiple solidification stages according to the inflection point position. Calculate the fluctuation range of the solidification time difference in each region within each solidification stage, evaluate the coordination matching degree between the fluctuation range and the expected transfer time fluctuation range of the corresponding solidification stage, construct the solidification matching degree index based on the coordination matching degree evaluation result, and take the candidate pouring sequence scheme corresponding to the solidification matching degree index with the largest value as the target pouring sequence.
5. The method for controlling the solidification quality of castings according to claim 1, characterized in that: Molten metal is poured into each area sequentially according to the target pouring sequence. During the pouring process, the surface temperature distribution of the poured areas is collected to calculate the actual liquid phase fraction, including: According to the target pouring sequence, molten metal is poured into each area in sequence. After the pouring of a single area is completed, the surface temperature distribution data of the poured area is collected. The temperature gradient direction and gradient magnitude in the surface temperature distribution data are identified. The main heat dissipation path of the area is determined according to the temperature gradient direction. The thickness distribution features of the region are extracted along the main heat dissipation path, and the temperature values of each sampling point in the surface temperature distribution data are correlated and mapped with the thickness values at the corresponding locations to construct a temperature-thickness correlation matrix. Based on the correspondence between surface temperature and thickness in the temperature-thickness correlation matrix, the temperature distribution state of each depth layer in the region is calculated in combination with the preset temperature attenuation coefficient. The depth layer with a temperature value higher than the liquidus temperature is extracted as the liquidus layer, and the depth layer with a temperature value between the liquidus temperature and the solidus temperature is extracted as the solid-liquid coexistence layer. The distribution ratio of the liquid phase layer and the solid-liquid coexistence layer at each thickness position in the region is statistically analyzed, and the actual liquid phase fraction of the cast area is calculated based on the distribution ratio.
6. The method for controlling the solidification quality of castings according to claim 1, characterized in that: When the deviation between the actual liquid phase fraction and the preset liquid phase fraction exceeds a preset deviation threshold, the pouring order of the unpoured areas in the target pouring sequence is adjusted, including: The difference between the actual liquid phase fraction of the poured area and the corresponding preset liquid phase fraction is calculated. When the absolute value of the difference exceeds the preset deviation threshold, the thickness distribution characteristics of the poured area and the pouring time interval of the adjacent poured areas are extracted. The influencing factors that cause the liquid phase fraction deviation are identified based on the thickness distribution characteristics and the pouring time interval. In the uncast areas of the target pouring sequence, the heat conduction path length between each uncast area and the cast area is calculated. Uncast areas with heat conduction path lengths less than a preset path threshold are selected as sensitive areas associated with the influencing factor. The solidification time coupling degree between each sensitive area and the cast area is calculated. The sensitive areas are prioritized according to the solidification time coupling degree. For the highest priority sensitive area, the position of the sensitive area in the target pouring sequence is adjusted according to the deviation direction of the actual liquid fraction. The adjusted pouring sequence is substituted into the calculation process of the solidification matching index for iterative verification. The pouring sequence that makes the deviation of the actual liquid fraction converge to within the preset deviation threshold is extracted as the updated target pouring sequence.
7. The method for controlling the solidification quality of castings according to claim 1, characterized in that: Before performing the casting flipping operation, it is determined whether the liquid phase fractions in both the thin-walled and thick-walled regions are less than a preset flipping liquid phase threshold. If not, the cooling intensity is adjusted until the preset flipping liquid phase threshold is met before performing the casting flipping operation, including: Before performing the casting flipping operation, the flipping direction and flipping angular velocity of the casting to be performed are obtained. Based on the casting flipping direction, the spatial position relationship of each region of the casting relative to the flipping axis is identified, and the distribution of inertial force borne by the thin-walled region and the thick-walled region during the flipping process is calculated. The liquid phase stability coefficient of each region is determined based on the inertial force distribution and thickness distribution characteristics of each region. The liquid phase stability coefficient is coupled with the preset liquid phase inversion threshold to obtain the actual liquid phase inversion threshold of each region. The surface temperature distribution of the thin-walled region and the thick-walled region is collected to calculate the liquid phase fraction of each region. Regions with liquid phase fraction exceeding the actual inverted liquid phase threshold are identified as regions to be cooled more effectively. The target cooling medium flow rate increase is determined based on the excess liquid phase fraction and inertial force distribution characteristics of the regions to be cooled more effectively. Adjust the cooling medium flow rate of the corresponding cooling nozzle in the area to be cooled according to the target cooling medium flow rate increase, continuously monitor the change in liquid phase fraction in the area to be cooled, and perform the casting flipping operation when the liquid phase fraction of the thin-walled area and the thick-walled area both meet their respective actual flipping liquid phase thresholds.
8. A casting solidification quality control system, using the casting solidification quality control method as described in any one of claims 1-7, characterized in that, Includes the following modules: The three-dimensional model module is used to obtain a three-dimensional model of the casting to be poured, identify the thin-walled region and the thick-walled region in the three-dimensional model, and calculate the thin-walled solidification time corresponding to the thin-walled region and the thick-walled solidification time corresponding to the thick-walled region. The sequential pouring module is used to generate multiple candidate pouring sequence schemes. Each candidate pouring sequence scheme includes different arrangements and combinations of the thin-walled region and the thick-walled region. For each candidate pouring sequence scheme, a solidification matching index is calculated based on the solidification time and pouring interval time of each region. The candidate pouring sequence scheme with the best solidification matching index is selected as the target pouring sequence. The sorting adjustment module is used to sequentially pour molten metal into each area according to the target pouring order. During the pouring process, the surface temperature distribution of the poured area is collected to calculate the actual liquid phase fraction. When the deviation between the actual liquid phase fraction and the preset liquid phase fraction exceeds the preset deviation threshold, the pouring order of the unpoured areas in the target pouring order is adjusted. The flipping confirmation module is used to determine whether the liquid phase fraction of the thin-walled region and the thick-walled region is less than the preset flipping liquid phase threshold before performing the casting flipping operation. If not, the cooling intensity is adjusted until the preset flipping liquid phase threshold is met before performing the casting flipping operation.
9. An electronic device, characterized in that, include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the steps of the method as described in any one of claims 1 to 7.
10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer program instructions that, when executed by a processor, implement the steps of the method as described in any one of claims 1 to 7.
