Method for calculating the fluidity characteristics of molten metal in casting molds

Abstract

This invention provides a configuration for calculating an index value representing the fluidity of molten metal within the flow path 4 of a mold, which is referenced in the prior prediction of the filling performance of molten metal within the flow path 4 of the mold in a casting method. [Solution] A method for calculating a fluidity feature, which is an index value of the fluidity of molten metal in a flow path within a casting mold, includes the steps of: detecting the flow length L, which is the distance from the inlet 2 to the outlet 3 of the molten metal in the flow path 4; detecting the flow volume V, which is the volume of the molten metal from the inlet to the outlet in the flow path; and calculating the value obtained by dividing the flow length by the flow volume as the fluidity feature.

Description

Technical Field

[0001] The present invention relates to a method for casting a product, and more particularly to a method for calculating a fluidity characteristic value, which is an index value of the fluidity of molten metal in the flow path of a casting mold.

Background Art

[0002] For the production of metal products such as aluminum alloy products with complex shapes, casting methods are used in which molten metal is injected into a flow path (cavity) in a casting mold, such as die casting, and solidified. Various configurations have been proposed to improve such metal product casting methods. For example, Patent Document 1 proposes a configuration that enables accurate prediction of casting defects by predicting defect information of a product shape from shape features obtained from a newly generated product shape using a prediction formula derived by machine learning that correlates shape features, including the design dimensions of the product shape, the mold volume, the ratio of the volume of the design product shape to the mold, the dimensions of the product shape after cooling cast in the mold, and the gate distance (the distance from the gate where molten metal is injected into the mold to the measurement point), within a spherical range with a predetermined diameter larger than a predetermined pitch, centered on measurement points placed at a predetermined pitch on the surface of the product shape generated by the CPU from a 3D model, with defect information including the number of defects in the product shape, the defect level, the results of temperature simulation by CAE, and predicted values ​​by CAE. Patent Document 2 proposes creating a mold model for CAE analysis by dividing the cavity of the mold used to obtain the casting into multiple elements, in order to accurately predict the mechanical properties of each part of a die-cast product. Using this mold model, molten metal flow analysis and solidification analysis are performed under predetermined casting conditions to calculate factors related to the growth of the solidification structure, factors related to the cleanliness of the molten metal, and factors related to porosity defects for each element. The mechanical properties of each part of the casting are then determined by a regression equation using multiple regression analysis, with the mechanical properties of the casting as the dependent variable and the factors listed above as independent variables. Patent Document 3 proposes a mold polymer liquid crystal flow analysis apparatus and method that can simulate the flow of polymer liquid crystal filled in a mold. In this apparatus, the temperature field and physical properties of the polymer liquid crystal at a reference time are set, and the orientation field of the polymer liquid crystal at a reference time is calculated from the set temperature field and physical properties.Patent Document 4 proposes a configuration for evaluating the flow analysis of an injection molding die, which quantitatively evaluates the resin flow path dimensions based on scientific calculations in order to prevent burning, a surface defect that occurs during injection molding. This configuration involves performing a resin filling analysis with parameters such as resin flow rate and gate diameter to calculate the distribution of pressure, temperature, etc. Based on the analysis results, index values ​​for each part under initial conditions of sprue diameter, runner diameter, and gate diameter are calculated. Based on these initial analysis results, the change in the index value of each part by changing the diameter of each part is calculated. An appropriate index value is found where the amount of change is less than or equal to a predetermined amount, and this is replaced with the index value of each part from the initial analysis results. The obtained appropriate index values ​​for each part are then displayed graphically. [Prior art documents] [Patent Documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2023-49823 [Patent Document 2] Japanese Patent Publication No. 2019-105592 [Patent Document 3] Japanese Patent Publication No. 2006-213015 [Patent Document 4] Japanese Patent Publication No. 5-329905 [Overview of the Initiative] [Problems that the invention aims to solve]

[0004] In thin-walled casting methods such as die casting, in addition to heat-related defects such as seizing at specific points within the flow path of the mold, as covered by the technology described in Patent Document 1, there are also many defects caused by the "fillability" of the molten metal within the flow path. A defect caused by poor filling occurs when the molten metal loses heat as it flows through the flow path, solidifying prematurely before it can fill the entire flow path as intended, resulting in the molten metal not being able to fill the entire flow path. Therefore, in order to ensure the quality of cast products in thin-walled casting methods, it is preferable to predict in advance whether the molten metal can flow well into the entire flow path of the mold, that is, the fillability of the molten metal within the flow path of the mold, when considering the use of a particular mold.

[0005] While the above-mentioned prediction of molten metal fillability is generally performed using prediction technologies such as CAE (Computer-Aided Engineering), CAE requires significant man-hours and lead time, limiting the number of evaluations that can be performed. Therefore, in recent years, attempts have been made to develop technologies that utilize AI-based surrogate technology to instantly predict the fillability of molten metal within the mold's flow path, thereby significantly reducing the design review man-hours. However, ensuring sufficient accuracy in surrogate technology requires a vast amount of training data, and in many cases, it has been difficult to collect the necessary training data to raise the accuracy to a practical level, preventing practical application. Therefore, a configuration that can ensure accuracy even with a small amount of training data would be advantageous. To achieve such a configuration, since molten metal fillability is related to the ease of flow (fluidity) of the molten metal within the flow path, it is conceivable to calculate an index value defined to represent the fluidity of the molten metal within the mold's flow path in the casting method and refer to it when predicting the fillability of molten metal within the mold's flow path in the casting method.

[0006] In view of the above circumstances, the main object of the present invention is to provide a configuration for calculating an index value representing the fluidity of molten metal in the flow path of a mold, which is referenced in advance in predicting the filling performance of molten metal in the flow path of a mold in a casting method. [Means for solving the problem]

[0007] According to one aspect of the present invention, the above problem is solved by a method for calculating a fluidity characteristic quantity, which is an index value of the fluidity of molten metal in a flow path within a casting mold, A process for detecting the flow length, which is the distance along the flow path from the inlet to the outlet of the molten metal in the aforementioned flow path, A process for detecting the flow volume, which is the volume of the flow path from the inlet to the outlet of the molten metal in the aforementioned flow path, This is achieved by a method that includes the process of calculating the value obtained by dividing the flow length by the flow volume as the flow characteristic quantity.

[0008] In the above configuration, the "casting mold" may be a mold used in a casting system that forms a casting shaped like a flow channel (cavity) from molten metal using any casting method such as vacuum die casting or die casting, and may be made of materials commonly used in this field. The mold is provided with an inlet into which molten metal is injected into the flow channel and an outlet from which air inside the flow channel is pushed out, as described above. The casting mold to be evaluated in the above invention may be a model virtually constructed in a computer, and the flow length and flow volume may be detected in the model constructed in the computer.

[0009] In the above configuration of the present invention, the flow length and flow volume of the molten metal from the inlet to the outlet in the mold's flow path are detected, and a fluidity characteristic quantity that serves as an index value representing the fluidity of the molten metal is Fluidity characteristic = (flow length) / (flow volume) ... (1) It is calculated by [formula]. The larger the value of this fluidity characteristic, the more difficult it becomes for the molten metal to pass through the channel, resulting in poorer packing performance.

[0010] In utilizing fluidity features, for example, if the fluidity features are excessive in a mold whose filling properties are to be predicted, the filling properties of the molten metal will deteriorate, so measures such as changing the design of the flow path in the mold may be taken.

[0011] Furthermore, the fluidity feature of the present invention may be used as one of the input data in predicting the fillability of molten metal in a mold channel using AI-based surrogate technology. In short, predictive calculation models that use the shape and characteristics of the mold channel, the characteristics of the molten metal, and the flow velocity during molten metal injection as input data, and an index value of the fillability of the molten metal in the channel (for example, the solid fraction in the channel) as output data, are constructed using training data which consists of output data calculated using CAE for various input data as ground truth data, and the model is trained to output ground truth data for the input data. The fluidity feature of the present invention may be used as one of the input data in such predictive calculation models for prediction using surrogate technology.

[0012] In the configuration of the present invention described above, the flow volume may be detected by any method. In one embodiment, the flow volume may be obtained by setting up a plurality of spheres along the flow path of the mold to be evaluated so as not to overlap with each other, detecting the volume of the flow path portion in each sphere, and calculating the sum of the volumes of the flow path portions in each sphere as the flow volume.

[0013] Furthermore, in the above-described configuration of the present invention, the value obtained by multiplying the fluidity feature by a correction coefficient may be calculated as the corrected fluidity feature. That is, the corrected fluidity feature is Corrected fluidity feature = (flow length) / (flow volume) × (correction coefficient) …(2) It may be given by the above. When sufficient accuracy cannot be obtained as an index value representing the fluidity of the molten metal using the fluidity feature, a corrected fluidity feature obtained using an appropriately set correction coefficient as described above may be used as an index value representing the fluidity of the molten metal with higher accuracy.

[0014] More specifically, the correction coefficient may be set to a value that eliminates the difference between the packing index value representing the packing of molten metal in the training channel (the training channel for the predictive calculation model for packing prediction using the surrogate technique described above) determined using CAE technology, and the packing index value representing the packing of molten metal in the training channel determined using the surrogate technique described above, with the corrected fluidity feature used as one of the input data. This will improve the accuracy of predicting the packing of molten metal in the channel using the surrogate technique. [Effects of the Invention]

[0015] Thus, according to the above-described configuration of the present invention, a fluidity feature, which is a novel index value representing the fluidity of molten metal in the flow path of a mold and referenced in the prior prediction of the fillability of molten metal in the flow path of a mold in a casting method, is calculated. As already described, the fluidity feature according to the present invention can be used as one of the input data in the prediction of the fillability of molten metal in a flow path using surrogate technology, and in that case, it is expected that the accuracy of the fillability prediction result in a machine learning model with a smaller number of training data will be improved. That is, if the fluidity feature according to the present invention is used as one of the input data for a model, it is expected that the number of training data required for the model to output a fillability prediction result with sufficient accuracy can be reduced.

[0016] Other objects and advantages of the present invention will become apparent from the following description of preferred embodiments of the present invention. [Brief explanation of the drawing]

[0017] [Figure 1] Figure 1 is a schematic plan view of a product manufactured by the casting method, showing the positions of the molten metal inlet and outlet during casting. [Figure 2] Figure 2 is a schematic diagram of the flow path within the mold for which the fluidity characteristics are calculated by the method of this embodiment. [Figure 3]FIG. 3 is a schematic view showing a sphere set along a flow path when detecting a flow volume in the method of the present embodiment. [Figure 4] FIG. 4 is a diagram showing the relationship among the fluidity characteristic value, the flow length, and the flow volume of the flow path calculated by the method of the present embodiment. [Figure 5] FIG. 5 is a diagram schematically showing changes in the index value of the melt filling property determined by using CAE for the flow path of a certain mold and the index value of the melt filling property determined by the surrogate technology using the fluidity characteristic value for the same flow path with respect to the melt flow velocity.

Explanation of Reference Numerals

[0018] 1... casting product, 2... flow path inlet, 3... flow path outlet, 4... flow path, 10... sphere

Best Mode for Carrying Out the Invention

[0019] Calculation and use of liquidity features In the method according to the present embodiment, a fluidity characteristic value, which is an index value of the fluidity of the melt in the flow path of the mold used in any casting method such as the vacuum die casting method or the die casting method, is calculated The product manufactured by the casting method targeted in the present embodiment may be an article as indicated by reference numeral 1 in FIG. 1, and the shape of such an article 1 is formed in a cavity in a mold not shown The cavity is provided with an inlet for injecting the melt and an outlet for the air pushed out as the melt enters the cavity, and the space between such an inlet and an outlet becomes a flow path through which the melt flows Usually, in a cavity corresponding to one article 1, as shown in FIG. 1, a plurality of inlets 2 and outlets 3 may be provided, and in this case, the flow path from one inlet 2 to one outlet 3 is appropriately determined during the design of the cavity.

[0020] The fluidity characteristics of this embodiment are calculated as follows from the flow length L, which is the distance from inlet 2 to outlet 3 along the flow path 4 of molten metal formed between one inlet 2 and one outlet 3 in the cavity of a mold shaped like the article 1 described above, as schematically depicted in Figure 2, and the flow volume V, which is the volume within the flow path 4. Liquidity feature = L / V …(3) Here, the flow length L and the flow volume V may be detected in the actually manufactured mold, but are typically detected in a computer-designed model of the mold. The flow length L may be the distance along the flow path from inlet 2 to outlet 3 in any manner, but is typically the distance of the path connecting the bending points in the flow path, as shown in the figure. The flow volume V may be detected in any manner; for example, as schematically depicted in Figure 3, multiple spheres of radius r may be set along the flow path 4 without overlapping, the volume v of the flow path portion within each sphere may be detected, and the sum Σv of these volumes may be taken as the flow volume V.

[0021] The ease with which molten metal flows within a channel worsens with increasing flow length and improves with increasing flow volume V. On the other hand, the fluidity characteristic value calculated as described above increases with increasing flow length L and decreases with increasing flow volume V, as shown in Figure 4. In other words, the fluidity characteristic value increases as the ease of flow deteriorates, making it advantageous to use as an indicator of fluidity.

[0022] In utilizing fluidity features, as mentioned in the section on the summary of the invention, in one embodiment, fluidity features may be used to evaluate the fluidity of a flow channel. For example, if the fluidity features for a flow channel in a certain mold are excessive, a change in the mold design may be proposed.

[0023] Furthermore, in another aspect of the use of fluidity features, as described in the summary of the invention, fluidity features may be used as one of the input data for a predictive calculation model for predicting the fillability of molten metal in the flow path of a mold using AI-based surrogate technology. Research by the inventors of this embodiment has shown that by using fluidity features along with the shape and characteristics of the flow path of the mold, the characteristics of the molten metal, and the flow velocity during molten metal injection as input data for a model that calculates an index value of the fillability of molten metal in the flow path, it is possible to predict the fillability of molten metal with greater accuracy by learning from a smaller number of training data.

[0024] Calculation of corrected liquidity features In this embodiment, a corrected fluidity feature may be calculated by correcting the above-mentioned fluidity feature using the following formula. Corrected liquidity feature = L / V × α …(4) Here, α is a correction coefficient. The above-mentioned corrected fluidity feature may be used when the fluidity feature given by equation (3) does not provide sufficient accuracy for evaluating fluidity or predicting the fillability of molten metal in the flow path of a mold using surrogate technology. For example, as schematically shown in Figure 5, in the training of a prediction calculation model for predicting the fillability of molten metal in the flow path of a mold using surrogate technology, if there is a discrepancy Δ between the prediction result using the fluidity feature as one of the input data for the training flow path used for verification during the training process and the prediction result by CAE, the corrected fluidity feature may be used as input to the prediction calculation model, and the value of the correction coefficient α may be adjusted so that the output of the prediction calculation model matches the prediction result by CAE. Specifically, the ease of heat transfer in a fluid changes when the flow velocity changes or when the shape changes relative to a constant flow. This leads to changes in the ease of solidification, causing errors in the accuracy of predicting poor filling. In particular, the influence of flow velocity on the ease of heat transfer in a fluid is difficult to represent solely by the shape of the flow path. Therefore, it is conceivable to reflect such circumstances in the fluidity feature by adjusting the correction coefficient α. The correction coefficient α may be calculated using machine learning.

[0025] Procedure for calculating liquidity features or corrected liquidity features In the implementation, the liquidity feature (or corrected liquidity feature) may be calculated using the following procedure. (1) Creation of product shape ↓ (2) Discretization of the shape (dividing into FEM or difference mesh) ↓ (3) Specify the molten metal inlet, outlet, and flow path. ↓ (4) A sphere of radius r for calculating the flow volume is set along the flow path. ↓ (5) Detection of the volume of the flow channel portion for each sphere ↓ (6) Detection of flow length ↓ (7) Calculate the liquidity feature or corrected liquidity feature using the calculation formula.

[0026] While the above description is made in relation to embodiments of the present invention, many modifications and changes are readily possible for those skilled in the art, and it will be clear that the present invention is not limited to the embodiments illustrated above, but can be applied to various devices without departing from the concept of the present invention.

Claims

1. A method for calculating a fluidity characteristic quantity, which is an index value of the fluidity of molten metal in a flow path within a casting mold, A process for detecting the flow length, which is the distance along the flow path from the inlet to the outlet of the molten metal in the aforementioned flow path, A process for detecting the flow volume, which is the volume of the flow path from the inlet to the outlet of the molten metal in the aforementioned flow path, A method comprising the step of calculating the value obtained by dividing the flow length by the flow volume as the flow characteristic quantity.

2. A method according to claim 1, comprising the steps of: setting up a plurality of spheres along the flow path so as not to overlap with each other, detecting the volume of the flow path portion in each of the spheres, and calculating the sum of the volumes of the flow path portions in each of the spheres as the flow volume.

3. A method according to claim 1, comprising the step of calculating a corrected fluidity feature by multiplying the fluidity feature by a correction coefficient.

4. The method according to claim 3, wherein the correction coefficient is a value set such that the difference between a filling index value representing the filling ability of molten metal in a learning channel determined using CAE technology and a filling index value representing the filling ability of molten metal in the learning channel determined using surrogate technology that uses the corrected fluidity feature quantity as a parameter is eliminated.

Images