Suspension induction melting suction casting device and method for refractory high-activity multi-element complex alloy

Abstract

The invention relates to a refractory high-activity multi-element complex alloy suspension induction smelting suction casting device and method, and aims to improve the alloy smelting and casting efficiency and quality. The device comprises a casting controller, a temperature controller, an ultrasonic controller and other modules, and the melting condition of raw materials in a smelting crucible and the filling condition in a mold are monitored in real time through an ultrasonic probe. And feature extraction and state classification are performed on the ultrasonic signals by using a deep learning model, so that optimal control of the casting process is realized. The method comprises the steps of smelting starting, ultrasonic detection preparation, data acquisition and preprocessing, feature extraction, deep learning model analysis, alloy filling, mold ultrasonic detection, state judgment and adjustment and the like. By accurately controlling the flow of the inert gas and maintaining the protective atmosphere, alloy oxidation is effectively prevented, the smelting quality is improved, it is ensured that alloy components are uniform, the quality and consistency of castings are improved, the production cost is reduced, and the production efficiency is improved.

Description

Technical Field

[0001] The invention relates to the field of alloy casting, and in particular to a device and method for suspended induction melting and suction casting of a refractory high-activity multi-component complex alloy. Background Art

[0002] In the production of refractory, highly active, multi-component complex alloys, smelting and casting technology is a key link. Traditional smelting methods often have problems such as uneven alloy composition, easy oxidation during smelting, and unstable casting quality. With the development of materials science and intelligent manufacturing technology, higher requirements are placed on the precise control and real-time monitoring of the smelting process. Suspended induction smelting suction casting technology has attracted attention due to its high efficiency, energy saving and environmental friendliness.

[0003] However, the traditional suspension induction melting and suction casting technology lacks detection of the melting effect of the molten metal liquid and the filling effect during casting, resulting in a low casting success rate. The existing technology still has deficiencies in real-time monitoring and control during the melting process, especially in the collection, processing and analysis of ultrasonic detection data. There is a lack of efficient and accurate solutions. Therefore, the development of a suspension induction melting and suction casting device for refractory, high-activity, multi-component complex alloys that integrates advanced ultrasonic detection technology is of great significance for improving the alloy melting quality and optimizing the casting process.

[0004] The invention realizes the melting of the alloy by induction heating, uses ultrasonic technology to carry out real-time monitoring of the melting state, and realizes the precise filling of the alloy liquid by a precisely controlled negative pressure feeding system. Summary of the invention

[0005] In order to solve the above problems, the present invention provides a refractory high-activity multi-component complex alloy suspension induction melting suction casting device, including a casting controller, a temperature controller, an ultrasonic controller, a drive controller, a pressure controller and an analysis module;

[0006] The casting controller is connected to the temperature controller, the ultrasonic controller, the drive controller, the pressure controller and the analysis module;

[0007] It also includes a smelting chamber and a casting chamber. The casting chamber is arranged above the smelting chamber. The smelting chamber is provided with a smelting crucible. A feeding chute is arranged on the side of the smelting chamber. The casting chamber is provided with a mold. A negative pressure feeding pipe is arranged below the casting chamber. The negative pressure feeding pipe is connected to the mold. The smelting chamber and the casting chamber are sealed.

[0008] The mold and the melting crucible are both provided with heating coils and temperature sensors outside, and the heating coils are connected to a heating power supply; the casting cavity and the melting chamber are both connected to a vacuum pump; and the melting chamber is externally connected to an inert gas tank;

[0009] A lifting platform is arranged at the bottom of the melting crucible, and the lifting platform is controlled by a driving motor; ultrasonic probes are arranged on both sides of the melting crucible, the mold and the lifting platform;

[0010] The temperature controller is connected to the electromagnetic heater and the temperature sensor, the ultrasonic controller is connected to the ultrasonic probe, the drive controller is connected to the drive motor, and the pressure controller is connected to the inert gas tank and the vacuum pump;

[0011] During the casting process, the ultrasonic controller uses an ultrasonic probe to emit ultrasonic waves to detect ultrasonic detection data in the melting crucible and the mold, and sends the detection data to the casting controller. The casting controller uses an analysis module to analyze the collected ultrasonic data to obtain the melting condition of the raw materials in the melting crucible and the filling condition in the mold.

[0012] The melting chamber is made of high temperature resistant material, and the inner surface is coated with anti-adhesion coating to reduce the adhesion of alloy during melting. The bottom of the melting crucible is designed as a hemisphere, and the lifting platform covers the hemispherical part of the melting crucible at the connection position between the melting crucible and the lifting platform.

[0013] The outlet end of the feed chute is designed to be angle-adjustable to suit different casting requirements;

[0014] A cooling water channel is designed inside the casting cavity to quickly cool the alloy during the casting process; the negative pressure feed pipe below the casting cavity is made of high temperature resistant material, and an adjustable valve is set inside the negative pressure feed pipe to control the speed and amount of the alloy flowing into the mold;

[0015] The vacuum pumps connected to the melting chamber and casting chamber adopt a two-stage vacuum system to achieve high vacuum degree, ensuring that the melting and casting processes are carried out in an oxygen-free environment; the inert gas tank connected to the outside of the melting chamber accurately controls the gas flow through a pressure controller to protect the alloy from oxidation during the melting process.

[0016] During the smelting process of refractory, highly active, multi-component complex alloys, the corresponding ultrasonic probes in the melting crucible and the lifting platform synchronously emit ultrasonic waves and collect the transmission signals of the ultrasonic waves. The collected signals are the propagation characteristics of the ultrasonic waves in the melting crucible.

[0017] During the smelting process of refractory, highly active, multi-component complex alloys, when the metal in the smelting crucible is completely melted, the molten metal liquid is fed into the mold through a negative pressure feed pipe. The ultrasonic probe corresponding to the mold emits ultrasonic waves and collects the transmission signals of the ultrasonic waves. The collected signals are the propagation characteristics of the ultrasonic waves in the mold.

[0018] The collected signal is sent to the casting controller, and the casting controller pre-processes the collected signal and then sends it to the analysis module.

[0019] Data preprocessing includes: noise filtering, removing noise through mean filtering, median filtering or Gaussian filtering to improve the signal-to-noise ratio;

[0020] Signal conditioning is performed through correction of baseline drift and synchronization of pulse repetition frequency;

[0021] The signal is decomposed into a superposition of wavelet functions of different scales, and then the wavelet coefficients are thresholded and finally the signal is reconstructed to remove noise.

[0022] Perform feature extraction on the preprocessed signal:

[0023] Extract wavelet packet energy spectrum:

[0024] Select the wavelet function and decomposition level, select the 'db4' wavelet basis function and decomposition level in the Daubechies wavelet system;

[0025] Perform wavelet transform, use wavelet decomposition function to decompose the signal into approximate coefficients and detail coefficients;

[0026] Calculate the energy of wavelet coefficients. Calculate the energy of each wavelet coefficient. The energy is calculated by the sum of the squares of the wavelet coefficients.

[0027] Calculate the standard deviation and mean of the wavelet coefficients, calculate the standard deviation and mean of each wavelet coefficient, and use these statistical features as part of the feature vector;

[0028] Combined feature vector, combining the calculated energy, standard deviation and mean into a feature vector for deep learning;

[0029] Extract signal envelope features:

[0030] Eliminate the DC component to obtain a more accurate envelope;

[0031] Use Hilbert transform to process the signal to obtain the analytical signal, and then take the modulus of the analytical signal to obtain the envelope of the signal;

[0032] Superimpose the DC component and add the DC component back to the envelope to restore the envelope of the original signal;

[0033] The structure of the deep learning model is:

[0034] Input layer: normalizes or standardizes the extracted features and receives preprocessed ultrasonic signal data in the form of time series;

[0035] Convolutional layer: The first and second layers are convolutional layers, which are used to extract the spatial features of the signal. Each convolutional layer is followed by a maximum pooling layer to reduce the feature dimension. The number of convolution kernels is 32 in the first layer and 64 in the second layer. The convolution kernel size is 3x3. The activation function is ReLU.

[0036] Recurrent layer: two layers of long short-term memory network layers, used to process the time series characteristics of the signal; number of LSTM units: 128 in the first layer, 64 in the second layer, activation function: tanh;

[0037] Fully connected layers: Two LSTM layers followed by two fully connected layers for further feature extraction and classification; the number of neurons in the first fully connected layer: 256, the number of neurons in the second fully connected layer: 128; activation function: ReLU;

[0038] Output layer: uses the softmax activation function to classify the melting state and filling state of the metal liquid;

[0039] A Dropout layer is added between the fully connected layers to prevent overfitting; a BatchNormalization layer is added after the convolutional layer and the fully connected layer to speed up the training process and improve the generalization ability of the model.

[0040] A method for suspended induction melting and suction casting of a refractory high-activity multi-component complex alloy, using the refractory high-activity multi-component complex alloy suspended induction melting and suction casting device, comprises the following steps:

[0041] a. Start the smelting process: add the mixed metal raw materials into the smelting crucible, start the electromagnetic heater to heat the smelting crucible, and adjust the heating power through the temperature controller until the alloy is completely melted;

[0042] b. Preparation for ultrasonic testing: During the alloy melting process, the ultrasonic controller controls the ultrasonic probe to emit ultrasonic waves to detect the melting status of the raw materials in the melting crucible in real time;

[0043] c. Data acquisition and preprocessing: Collect the ultrasonic detection data detected by the ultrasonic probe and send the data to the casting controller for preprocessing steps such as noise filtering, baseline drift correction, and pulse repetition frequency synchronization;

[0044] d. Feature extraction: perform wavelet transform on the preprocessed signal, extract the wavelet packet energy spectrum, calculate the energy, standard deviation and mean of the wavelet coefficients, and extract the signal envelope features;

[0045] e. Deep learning model analysis: The extracted feature vectors are input into the deep learning model, which performs feature learning and state classification through convolutional layers, recurrent layers, and fully connected layers;

[0046] f. Alloy filling: After confirming that the alloy is completely melted, open the valve of the negative pressure feed pipe, and use the negative pressure to input the molten metal liquid in the melting crucible into the mold in the casting cavity; at the same time, open the inert gas tank to pressurize the melting chamber to ensure that the molten metal liquid enters the mold in the casting cavity;

[0047] g. Ultrasonic testing of molds: When the metal liquid fills the mold, the ultrasonic probe corresponding to the mold emits ultrasonic waves to collect the propagation characteristics of the metal liquid in the mold;

[0048] h. Status judgment and adjustment: The ultrasonic inspection data of the mold is input into the deep learning model, and the model outputs the filling status of the metal liquid. According to the judgment result of the model, the height of the lifting platform and the valve of the negative pressure feed pipe are adjusted through the driving controller to optimize the filling effect;

[0049] i. Completion of suction casting: When the metal liquid completely fills the mold and reaches the desired cooling effect, the negative pressure of the negative pressure feed pipe is stopped to complete the suction casting process, and then the cooling water channel in the casting cavity is opened for rapid cooling;

[0050] j. Post-processing: After cooling is completed, open the casting cavity, take out the formed alloy casting, and carry out subsequent cleaning and inspection work.

[0051] Before step a, the method further comprises the following steps:

[0052] a1. Preparation of inert gas protective atmosphere: Before starting the smelting process, the pressure controller controls the inert gas tank to transport inert gas into the smelting chamber to replace the air in the smelting chamber and form a protective atmosphere;

[0053] a2. Accurate control of inert gas flow: Use a pressure controller to accurately control the flow of inert gas to ensure that the gas environment in the melting chamber meets the protection requirements required for alloy melting and prevent the alloy from being oxidized or contaminated at high temperatures;

[0054] a3.Maintenance of inert gas protective atmosphere: Inert gas is continuously delivered throughout the entire smelting and casting process to maintain a stable protective atmosphere, ensuring that the alloy raw materials in the melting crucible and the metal liquid in the mold are surrounded and protected by inert gas throughout the process;

[0055] a4. Detection of inert gas replacement effect: The gas composition in the melting chamber is monitored by the analysis module to ensure that the inert gas has completely replaced the air in the melting chamber to achieve the required protection effect and provide an oxygen-free environment for the melting and casting of the alloy.

[0056] During the process of metal liquid filling the mold, the ultrasonic probe corresponding to the mold intermittently transmits and receives ultrasonic waves to collect the propagation characteristic data of the metal liquid in the mold.

[0057] Intermittent frequency: The intermittent frequency of the ultrasonic probe's intermittent transmission is set to 10Hz to 50Hz to meet the real-time monitoring needs of the metal liquid during the rapid filling process in the mold, ensuring the continuity and effectiveness of signal acquisition;

[0058] Duty cycle: To ensure that ultrasound does not affect casting, the duty cycle of the ultrasonic probe is set to 20% or lower to reduce interference with the casting process;

[0059] Ultrasonic frequency: The ultrasonic frequency is between 0.5MHz and 10MHz to reduce interference with the casting process.

[0060] The beneficial effects of the present invention are:

[0061] The refractory, high-activity, multi-component complex alloy suspension induction melting and suction casting device provided by the present invention realizes comprehensive control and monitoring of the melting and casting process by integrating a casting controller, a temperature controller, an ultrasonic controller, a drive controller, a pressure controller and an analysis module. The device uses an ultrasonic probe to detect the melting of raw materials in the melting crucible and the filling of the mold in real time, and analyzes the ultrasonic signal through a deep learning model to accurately determine the melting state and filling state of the alloy, thereby realizing optimized control of the casting process. Compared with traditional technologies:

[0062] 1) Improve the safety and stability of the smelting process, and effectively prevent alloy oxidation by precisely controlling the inert gas flow and maintaining a protective atmosphere;

[0063] 2) The smelting state is monitored in real time through ultrasonic detection technology, which improves the smelting quality and ensures uniform alloy composition;

[0064] 3) Using deep learning models to extract features and classify states of ultrasonic signals improves the accuracy and response speed of state judgment;

[0065] 4) Through precise control of the negative pressure feeding system, accurate filling of the alloy liquid is achieved, improving the quality and consistency of the castings;

[0066] 5) The device has a high degree of automation and intelligence, which reduces manual intervention, reduces production costs and improves production efficiency.

[0067] In general, the present invention realizes the intelligent control of the melting and casting process of refractory, highly active, multi-component complex alloys at the technical level, providing an efficient, precise and reliable solution for the alloy production field. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0069] Attached Figure 1 It is a schematic diagram of the structure of the device of the present invention;

[0070] Attached Figure 2 Schematic diagram of the device architecture of the present invention.

[0071] in:

[0072] Melting chamber 1, casting cavity 2, mold 3, negative pressure feeding pipe 4, vacuum pump 5, inert gas tank 6, lifting platform 7, melting crucible 8, ultrasonic probe 9, feeding chute 10. DETAILED DESCRIPTION

[0073] Embodiment 1:

[0074] See also Figure 1 and Figure 2 The present invention provides a refractory high-activity multi-component complex alloy suspension induction melting suction casting device, including a casting controller, a temperature controller, an ultrasonic controller, a drive controller, a pressure controller and an analysis module;

[0075] The casting controller is connected to the temperature controller, the ultrasonic controller, the drive controller, the pressure controller and the analysis module;

[0076] It also includes a smelting chamber 1 and a casting chamber 2. The casting chamber 2 is arranged above the smelting chamber 1. The smelting chamber 1 is provided with a smelting crucible 8. A feeding chute 10 is arranged on the side of the smelting chamber 1. The casting chamber 2 is provided with a mold 3. A negative pressure feeding pipe 4 is arranged below the casting chamber 2. The negative pressure feeding pipe 4 is connected to the mold 3. The smelting chamber 1 and the casting chamber 2 are sealed.

[0077] The mold 3 and the melting crucible 8 are both provided with a heating coil and a temperature sensor, and the heating coil is connected to a heating power supply; the casting cavity 2 and the melting chamber 1 are both connected to a vacuum pump 5; the melting chamber 1 is externally connected to an inert gas tank 6;

[0078] A lifting platform 7 is provided at the bottom of the melting crucible 8, and the lifting platform 7 is controlled by a driving motor; ultrasonic probes 9 are provided on both sides of the melting crucible 8, the mold 3 and the lifting platform 7;

[0079] The temperature controller is connected to the electromagnetic heater and the temperature sensor, the ultrasonic controller is connected to the ultrasonic probe 9, the drive controller is connected to the drive motor, and the pressure controller is connected to the inert gas tank 6 and the vacuum pump 5;

[0080] During the casting process, the ultrasonic controller transmits ultrasonic waves through the ultrasonic probe 9 to detect ultrasonic detection data in the melting crucible 8 and the mold 3, and sends the detection data to the casting controller. The casting controller analyzes the collected ultrasonic data through the analysis module to obtain the melting condition of the raw materials in the melting crucible 8 and the filling condition in the mold 3.

[0081] The melting chamber 1 is made of high temperature resistant material, and the inner surface is coated with an anti-adhesion coating to reduce the adhesion of the alloy during the melting process; the bottom of the melting crucible 8 is designed to be a hemisphere, and the lifting platform 7 covers the hemispherical part of the melting crucible 8 at the position where the melting crucible 8 is connected to the lifting platform 7;

[0082] The outlet end of the feed chute 10 is designed to be angle-adjustable to meet different casting requirements;

[0083] A cooling water channel is designed inside the casting cavity 2 to quickly cool the alloy during the casting process; the negative pressure feed pipe 4 below the casting cavity 2 is made of high temperature resistant material, and an adjustable valve is provided inside the negative pressure feed pipe 4 to control the speed and amount of the alloy flowing into the mold 3;

[0084] The vacuum pump 5 connected to the melting chamber 1 and the casting chamber 2 adopts a two-stage vacuum system to achieve a high vacuum degree and ensure that the melting and casting processes are carried out in an oxygen-free environment; the inert gas tank 6 connected to the outside of the melting chamber 1 accurately controls the gas flow through a pressure controller to protect the alloy from oxidation during the melting process.

[0085] During the smelting process of refractory, highly active, multi-component complex alloys, the corresponding ultrasonic probes in the melting crucible and the lifting platform synchronously emit ultrasonic waves and collect the transmission signals of the ultrasonic waves. The collected signals are the propagation characteristics of the ultrasonic waves in the melting crucible.

[0086] During the smelting process of refractory, highly active, multi-component complex alloys, when the metal in the smelting crucible is completely melted, the molten metal liquid is fed into the mold through a negative pressure feed pipe. The ultrasonic probe corresponding to the mold emits ultrasonic waves and collects the transmission signals of the ultrasonic waves. The collected signals are the propagation characteristics of the ultrasonic waves in the mold.

[0087] The collected signal is sent to the casting controller, and the casting controller pre-processes the collected signal and then sends it to the analysis module.

[0088] Data preprocessing includes: noise filtering, removing noise through mean filtering, median filtering or Gaussian filtering to improve the signal-to-noise ratio;

[0089] Signal conditioning is performed through correction of baseline drift and synchronization of pulse repetition frequency;

[0090] The signal is decomposed into a superposition of wavelet functions of different scales, and then the wavelet coefficients are thresholded and finally the signal is reconstructed to remove noise.

[0091] Perform feature extraction on the preprocessed signal:

[0092] Extract wavelet packet energy spectrum:

[0093] Select the wavelet function and decomposition level, select the 'db4' wavelet basis function and decomposition level in the Daubechies wavelet system;

[0094] Perform wavelet transform, use wavelet decomposition function to decompose the signal into approximate coefficients and detail coefficients;

[0095] Calculate the energy of wavelet coefficients. Calculate the energy of each wavelet coefficient. The energy is calculated by the sum of the squares of the wavelet coefficients.

[0096] Calculate the standard deviation and mean of the wavelet coefficients, calculate the standard deviation and mean of each wavelet coefficient, and use these statistical features as part of the feature vector;

[0097] Combined feature vector, combining the calculated energy, standard deviation and mean into a feature vector for deep learning;

[0098] Extract signal envelope features:

[0099] Eliminate the DC component to obtain a more accurate envelope;

[0100] Use Hilbert transform to process the signal to obtain the analytical signal, and then take the modulus of the analytical signal to obtain the envelope of the signal;

[0101] Superimpose the DC component and add the DC component back to the envelope to restore the envelope of the original signal;

[0102] The structure of the deep learning model is:

[0103] Input layer: normalizes or standardizes the extracted features and receives preprocessed ultrasonic signal data in the form of time series;

[0104] Convolutional layer: The first and second layers are convolutional layers, which are used to extract the spatial features of the signal. Each convolutional layer is followed by a maximum pooling layer to reduce the feature dimension. The number of convolution kernels is 32 in the first layer and 64 in the second layer. The convolution kernel size is 3x3. The activation function is ReLU.

[0105] Recurrent layer: two layers of long short-term memory network layers, used to process the time series characteristics of the signal; number of LSTM units: 128 in the first layer, 64 in the second layer, activation function: tanh;

[0106] Fully connected layers: Two LSTM layers followed by two fully connected layers for further feature extraction and classification; the number of neurons in the first fully connected layer: 256, the number of neurons in the second fully connected layer: 128; activation function: ReLU;

[0107] Output layer: uses the softmax activation function to classify the melting state and filling state of the metal liquid;

[0108] A Dropout layer is added between the fully connected layers to prevent overfitting; a BatchNormalization layer is added after the convolutional layer and the fully connected layer to speed up the training process and improve the generalization ability of the model.

[0109] Embodiment 2:

[0110] A method for suspended induction melting and suction casting of a refractory high-activity multi-component complex alloy, using the refractory high-activity multi-component complex alloy suspended induction melting and suction casting device, comprises the following steps:

[0111] a. Start the smelting process: add the mixed metal raw materials to the smelting crucible, start the electromagnetic heater to heat the smelting crucible 8, and adjust the heating power by the temperature controller until the alloy is completely melted;

[0112] b. Preparation for ultrasonic testing: During the alloy melting process, the ultrasonic controller controls the ultrasonic probe 9 to emit ultrasonic waves to detect the melting of the raw materials in the melting crucible 8 in real time;

[0113] c. Data acquisition and preprocessing: Collect the ultrasonic detection data detected by the ultrasonic probe 9 and send the data to the casting controller for preprocessing steps such as noise filtering, baseline drift correction, and pulse repetition frequency synchronization;

[0114] d. Feature extraction: perform wavelet transform on the preprocessed signal, extract the wavelet packet energy spectrum, calculate the energy, standard deviation and mean of the wavelet coefficients, and extract the signal envelope features;

[0115] e. Deep learning model analysis: The extracted feature vectors are input into the deep learning model, which performs feature learning and state classification through convolutional layers, recurrent layers, and fully connected layers;

[0116] f. Alloy filling: After confirming that the alloy is completely melted, open the valve of the negative pressure feed pipe 4, and use the negative pressure to input the molten metal liquid in the melting crucible 8 into the mold 3 in the casting cavity 2; at the same time, open the inert gas tank to pressurize the melting chamber to ensure that the molten metal liquid enters the mold in the casting cavity;

[0117] g. Ultrasonic testing of molds: During the process of metal liquid filling the mold 3, the ultrasonic probe corresponding to the mold emits ultrasonic waves to collect the propagation characteristics of the metal liquid in the mold;

[0118] h. Status judgment and adjustment: The ultrasonic detection data of the mold is input into the deep learning model, and the model outputs the filling status of the metal liquid. According to the judgment result of the model, the height of the lifting platform 7 and the valve of the negative pressure feed pipe 4 are adjusted by the driving controller to optimize the filling effect;

[0119] i. Completion of suction casting: When the metal liquid completely fills the mold 3 and reaches the desired cooling effect, the negative pressure of the negative pressure feed pipe 4 is stopped to complete the suction casting process, and then the cooling water channel in the casting cavity 2 is opened for rapid cooling;

[0120] j. Post-processing: After cooling is completed, the casting cavity 2 is opened, the formed alloy casting is taken out, and subsequent cleaning and inspection work is performed.

[0121] Before step a, the method further comprises the following steps:

[0122] a1. Preparation of inert gas protective atmosphere: Before starting the smelting process, the inert gas tank 6 is controlled by the pressure controller to deliver inert gas to the inside of the smelting chamber 1 to replace the air in the smelting chamber to form a protective atmosphere;

[0123] a2. Accurate control of inert gas flow: Use a pressure controller to accurately control the flow of inert gas to ensure that the gas environment in the melting chamber meets the protection requirements required for alloy melting and prevent the alloy from being oxidized or contaminated at high temperatures;

[0124] a3. Maintenance of inert gas protective atmosphere: Inert gas is continuously delivered throughout the smelting and casting process to maintain a stable protective atmosphere to ensure that the alloy raw material in the melting crucible 8 and the metal liquid in the mold 3 are surrounded and protected by inert gas throughout the process;

[0125] a4. Detection of inert gas replacement effect: The gas composition in the melting chamber is monitored by the analysis module to ensure that the inert gas has completely replaced the air in the melting chamber to achieve the required protection effect and provide an oxygen-free environment for the melting and casting of the alloy.

[0126] When the metal liquid fills the mold 3, the ultrasonic probe 9 corresponding to the mold intermittently transmits and receives ultrasonic waves to collect propagation characteristic data of the metal liquid in the mold.

[0127] Intermittent frequency: The intermittent frequency of the intermittent transmission of the ultrasonic probe 9 is set to 10 Hz to 50 Hz to meet the real-time monitoring requirements during the rapid filling process of the metal liquid in the mold 3 and ensure the continuity and effectiveness of signal acquisition;

[0128] Duty cycle: To ensure that ultrasound does not affect casting, the duty cycle of the ultrasonic probe 9 is set to 20% or lower to reduce interference with the casting process;

[0129] Ultrasonic frequency: The ultrasonic frequency is between 0.5MHz and 10MHz to reduce interference with the casting process.

[0130] Thus far, the description of the above-described embodiments is provided for the purpose of illustration and description. It is not intended to be exhaustive or to limit the present disclosure. The individual elements or features of a specific embodiment are generally not subject to the limitations of the specific embodiment, but when applicable, they can be interchanged and used for selected embodiments even if not specifically shown or described. In many respects, the same element or feature also can be changed. This variation is not considered to be a departure from the present disclosure, and all such modifications are intended to be included in the scope of the present disclosure.

[0131] Example embodiments are provided so that the present disclosure will be thorough and the scope will be fully conveyed to those skilled in the art. In order to thoroughly understand the embodiments of the present disclosure, numerous details are set forth, such as examples of specific parts, devices, and methods. Obviously, for those skilled in the art, it is not necessary to use specific details, the example embodiments can be implemented in many different forms, and neither should be construed as limiting the scope of the present disclosure. In certain example embodiments, well-known processes, well-known device structures, and well-known technologies are not described in detail.

[0132] Here, specialized vocabulary is used only for the purpose of describing specific example embodiments, and is not intended to be limiting. Unless the context clearly indicates otherwise, the singular forms "a" and "the" used herein may be intended to include plural forms as well. The terms "including" and "having" are inclusive, and therefore specify the presence of the claimed features, wholes, steps, operations, elements and / or components, but do not exclude the presence or additional presence of one or more other features, wholes, steps, operations, elements, components and / or combinations thereof. Unless the order of execution is clearly indicated, the method steps, processing and operations described herein are not interpreted as necessarily needing to be performed in the specific order discussed and shown. It should also be understood that additional or optional steps may be adopted.

Claims

1. A refractory high-activity multi-component complex alloy suspension induction melting suction casting device, comprising a casting controller, a temperature controller, an ultrasonic controller, a drive controller, a pressure controller and an analysis module; characterized in that: The casting controller is connected to the temperature controller, the ultrasonic controller, the drive controller, the pressure controller and the analysis module; It also comprises a smelting chamber (1) and a casting chamber (2), wherein the casting chamber (2) is arranged above the smelting chamber (1), the smelting chamber (1) is provided with a smelting crucible (8), a feeding chute (10) is arranged on the side of the smelting chamber (1), the casting chamber (2) is provided with a mold (3), a negative pressure feeding pipe (4) is arranged below the casting chamber (2), and the negative pressure feeding pipe (4) and the mold (3) are connected; the smelting chamber (1) and the casting chamber (2) are sealed and connected; the mold (3) and the smelting crucible (8) are both provided with a heating coil and a temperature sensor outside, and the heating coil is connected to a heating power supply; the casting chamber (2) and the smelting chamber (1) are both connected to a vacuum pump (5); and the outside of the smelting chamber (1) is connected to an inert gas tank (6); A lifting platform (7) is arranged at the bottom of the melting crucible (8), and the lifting platform (7) is controlled to move up and down by a driving motor; ultrasonic probes (9) are arranged on both sides of the melting crucible (8), the mold (3) and the lifting platform (7); The temperature controller is connected to the electromagnetic heater and the temperature sensor, the ultrasonic controller is connected to the ultrasonic probe (9), the drive controller is connected to the drive motor, and the pressure controller is connected to the inert gas tank (6) and the vacuum pump (5); during the casting process, the ultrasonic controller emits ultrasonic waves through the ultrasonic probe (9) to detect ultrasonic detection data in the melting crucible (8) and the mold (3), and sends the detection data to the casting controller, and the casting controller analyzes the collected ultrasonic data through the analysis module to obtain the melting condition of the raw materials in the melting crucible (8) and the filling condition in the mold (3).

2. The refractory high-activity multi-component complex alloy suspension induction melting suction casting device according to claim 1 is characterized by: The smelting chamber (1) is made of a high-temperature resistant material, and the inner surface is coated with an anti-adhesion coating to reduce the adhesion of the alloy during the smelting process; the bottom of the smelting crucible (8) is designed to be a hemisphere, and at the position where the smelting crucible (8) is connected to the lifting platform (7), the lifting platform (7) wraps the hemispherical part of the smelting crucible (8); the outlet end of the feed chute (10) is designed to be adjustable in angle to meet different casting requirements; a cooling water channel is designed inside the casting cavity (2) to quickly cool the alloy during the casting process; the negative pressure feed pipe (4) below the casting cavity (2) is made of a high-temperature resistant material, and an adjustable valve is arranged inside the negative pressure feed pipe (4) to control the speed and amount of the alloy flowing into the mold (3); The vacuum pump (5) connected to the melting chamber (1) and the casting chamber (2) adopts a double-stage vacuum system to achieve a high vacuum degree, ensuring that the melting and casting processes are carried out in an oxygen-free environment; the inert gas tank (6) connected to the outside of the melting chamber (1) accurately controls the gas flow through a pressure controller to protect the alloy from being oxidized during the melting process.

3. The refractory high-activity multi-component complex alloy suspension induction melting suction casting device according to claim 1 is characterized in that: During the smelting process of refractory, highly active, multi-component complex alloys, the corresponding ultrasonic probes in the melting crucible and the lifting platform synchronously emit ultrasonic waves and collect the transmission signals of the ultrasonic waves. The collected signals are the propagation characteristics of the ultrasonic waves in the melting crucible. During the smelting process of refractory and highly active multi-component complex alloys, when the metal in the smelting crucible is completely melted, the molten metal liquid is input into the mold through a negative pressure feed pipe. The ultrasonic probe corresponding to the mold emits ultrasonic waves and collects the transmission signal of the ultrasonic waves. The collected signal is the propagation characteristic of the ultrasonic wave in the mold. The collected signal is sent to the casting controller, which pre-processes the collected signal and sends it to the analysis module.

4. The refractory high-activity multi-component complex alloy suspension induction melting suction casting device according to claim 3 is characterized by: Data preprocessing includes: noise filtering, removing noise through mean filtering, median filtering or Gaussian filtering to improve the signal-to-noise ratio; Signal conditioning is performed through correction of baseline drift and synchronization of pulse repetition frequency; The signal is decomposed into a superposition of wavelet functions of different scales, and then the wavelet coefficients are thresholded and finally the signal is reconstructed to remove noise.

5. The refractory high-activity multi-component complex alloy suspension induction melting suction casting device according to claim 4 is characterized by: Perform feature extraction on the preprocessed signal: Extract wavelet packet energy spectrum: Select the wavelet function and decomposition level, select the 'db4' wavelet basis function and decomposition level in the Daubechies wavelet system; Perform wavelet transform, use wavelet decomposition function to decompose the signal into approximate coefficients and detail coefficients; Calculate the energy of wavelet coefficients. Calculate the energy of each wavelet coefficient. The energy is calculated by the sum of the squares of the wavelet coefficients. Calculate the standard deviation and mean of the wavelet coefficients, calculate the standard deviation and mean of each wavelet coefficient, and use these statistical features as part of the feature vector; Combined feature vector, combining the calculated energy, standard deviation and mean into a feature vector for deep learning; Extract signal envelope features: Eliminate the DC component to obtain a more accurate envelope; Use Hilbert transform to process the signal to obtain the analytical signal, and then take the modulus of the analytical signal to obtain the envelope of the signal; The DC component is superimposed and added back to the envelope to restore the envelope of the original signal.

6. The refractory high-activity multi-component complex alloy suspension induction melting suction casting device according to claim 5 is characterized by: The structure of the deep learning model is: Input layer: normalizes or standardizes the extracted features and receives preprocessed ultrasonic signal data in the form of time series; Convolutional layer: The first and second layers are convolutional layers, which are used to extract the spatial features of the signal. Each convolutional layer is followed by a maximum pooling layer to reduce the feature dimension. The number of convolution kernels is 32 in the first layer and 64 in the second layer. The convolution kernel size is 3x3. The activation function is ReLU. Recurrent layer: two layers of long short-term memory network layers, used to process the time series characteristics of the signal; number of LSTM units: 128 in the first layer, 64 in the second layer, activation function: tanh; Fully connected layers: Two LSTM layers followed by two fully connected layers for further feature extraction and classification; the number of neurons in the first fully connected layer: 256, the number of neurons in the second fully connected layer: 128; activation function: ReLU; Output layer: The softmax activation function is used to classify the melting state and filling state of the metal liquid; a Dropout layer is added between the fully connected layers to prevent overfitting; a Batch Normalization layer is added after the convolutional layer and the fully connected layer to speed up the training process and improve the generalization ability of the model.

7. A method for suspended induction melting and suction casting of a refractory high-activity multi-component complex alloy, using the refractory high-activity multi-component complex alloy suspended induction melting and suction casting device according to any one of claims 1 to 6, characterized in that The steps include: a. Start the smelting process: adding the mixed metal raw materials to the smelting crucible, starting the electromagnetic heater to heat the smelting crucible (8), adjusting the heating power by the temperature controller until the alloy is completely melted; b. Preparation for ultrasonic testing: During the alloy melting process, the ultrasonic controller controls the ultrasonic probe (9) to emit ultrasonic waves to detect the melting of the raw materials in the melting crucible (8) in real time; c. Data acquisition and preprocessing: collecting ultrasonic detection data detected by the ultrasonic probe (9) and sending the data to the casting controller for preprocessing steps such as noise filtering, baseline drift correction, and pulse repetition frequency synchronization; d. Feature extraction: perform wavelet transform on the preprocessed signal, extract the wavelet packet energy spectrum, calculate the energy, standard deviation and mean of the wavelet coefficients, and extract the signal envelope features; e. Deep learning model analysis: The extracted feature vectors are input into the deep learning model, which performs feature learning and state classification through convolutional layers, recurrent layers, and fully connected layers; f. Alloy filling: After confirming that the alloy is completely melted, the valve of the negative pressure feed pipe (4) is opened, and the molten metal liquid in the melting crucible (8) is input into the mold (3) in the casting cavity (2) by using the negative pressure; at the same time, the inert gas tank is opened to pressurize the melting chamber to ensure that the molten metal liquid enters the mold in the casting cavity; g. Ultrasonic testing of molds: During the process of metal liquid filling the mold (3), the ultrasonic probe corresponding to the mold emits ultrasonic waves to collect the propagation characteristics of the metal liquid in the mold; h. Status judgment and adjustment: The ultrasonic detection data of the mold is input into the deep learning model, and the model outputs the filling status of the metal liquid. According to the judgment result of the model, the height of the lifting platform (7) and the valve of the negative pressure feed pipe (4) are adjusted by the driving controller to optimize the filling effect; i. Completion of suction casting: When the metal liquid completely fills the mold (3) and reaches the desired cooling effect, the negative pressure of the negative pressure feed pipe (4) is stopped to complete the suction casting process, and then the cooling water channel in the casting cavity (2) is opened for rapid cooling; j. Post-processing: After cooling is completed, the casting cavity (2) is opened, the formed alloy casting is taken out, and subsequent cleaning and inspection work is carried out.

8. The method for suspension induction melting and suction casting of a refractory, high-activity, multi-component complex alloy according to claim 7, characterized in that: Before step a, the method further comprises the following steps: a1. Preparation of an inert gas protective atmosphere: Before starting the smelting process, the inert gas tank (6) is controlled by a pressure controller to deliver an inert gas to the inside of the smelting chamber (1) to replace the air in the smelting chamber to form a protective atmosphere; a2. Accurate control of inert gas flow: Use a pressure controller to accurately control the flow of inert gas to ensure that the gas environment in the melting chamber meets the protection requirements required for alloy melting and prevent the alloy from being oxidized or contaminated at high temperatures; a3. Maintenance of inert gas protective atmosphere: Inert gas is continuously delivered throughout the smelting and casting process to maintain a stable protective atmosphere to ensure that the alloy raw material in the melting crucible (8) and the liquid metal in the mold (3) are surrounded and protected by inert gas throughout the process; a4. Detection of inert gas replacement effect: The gas composition in the melting chamber is monitored by the analysis module to ensure that the inert gas has completely replaced the air in the melting chamber to achieve the required protection effect and provide an oxygen-free environment for the melting and casting of the alloy.

9. A suspension induction melting and suction casting method for a refractory, high-activity, multi-component complex alloy according to claim 8, characterized in that: During the process of the metal liquid filling the mold (3), the ultrasonic probe (9) corresponding to the mold intermittently transmits and receives ultrasonic waves to collect propagation characteristic data of the metal liquid in the mold.

10. The method for suspension induction melting and suction casting of a refractory, high-activity, multi-component complex alloy according to claim 9, characterized in that: Intermittent frequency: the intermittent frequency of the intermittent emission of the ultrasonic probe (9) is set to 10 Hz to 50 Hz to meet the real-time monitoring requirements during the rapid filling process of the metal liquid in the mold (3) and ensure the continuity and effectiveness of signal acquisition; Duty cycle: In order to ensure that ultrasound does not affect casting, the duty cycle of the ultrasonic probe (9) is set to 20% or lower to reduce interference with the casting process; Ultrasonic frequency: The ultrasonic frequency is between 0.5MHz and 10MHz to reduce interference with the casting process.

Images