A melting speed prediction model for VAR furnace titanium alloy ingot

Abstract

The application discloses a melting speed prediction model for VAR furnace titanium alloy ingot, which is based on the internal correlation between the electric parameters and the melting speed in the VAR furnace smelting process, defines core process parameters, constructs a smelting voltage-current linear sub-model and an output power-melting speed linear sub-model, and derives a complete melting speed prediction formula by combining the electric power physical formula P=UI. Through multiple parallel test verifications and coefficient calibrations, the prediction deviation is controlled within ±3%. The application realizes accurate and real-time prediction of the melting speed, can be embedded into a VAR furnace smelting control system to realize closed-loop regulation and control, is suitable for smelting of various types of titanium alloy ingots, significantly improves the ingot quality consistency and production efficiency, reduces the operation threshold, and has important industrial application value.

Description

Technical Field

[0001] This invention relates to the field of vacuum arc remelting technology, specifically to a melting rate prediction model for titanium alloy ingots in a VAR furnace. Background Technology

[0002] Vacuum arc remelting (VAR) technology, as a core technology for preparing high-quality titanium alloy ingots, holds an irreplaceable position in key fields such as aerospace, marine engineering, and high-end equipment manufacturing. Its core working principle is as follows: In a vacuum environment, high temperatures are released through an electric arc discharge between the electrode and the ingot, melting the electrode raw material into a molten metal. This molten metal gradually solidifies in a water-cooled copper crucible, ultimately forming a titanium alloy ingot with uniform composition and high density. In this complex physicochemical process, the melting rate of the titanium alloy raw material is a key process parameter determining the final quality of the ingot and the economics of production; its control precision directly affects the internal quality of the ingot and production efficiency.

[0003] When the melting rate is too fast, the molten metal does not remain in the crucible long enough, and the solidification process cannot proceed fully, easily leading to internal defects such as shrinkage cavities and porosity. Rapid solidification can also cause uneven distribution of alloying elements, affecting the consistency of the ingot's mechanical properties. Conversely, when the melting rate is too slow, it not only significantly extends the production cycle, increases energy consumption and production costs, but may also cause problems such as compositional segregation and coarse grains in the ingot due to localized overheating, similarly reducing the product yield. Therefore, achieving precise control of the melting rate is a key technical challenge in the production of titanium alloy ingots using VAR furnaces.

[0004] In existing technologies, the melting rate control of titanium alloy ingots in VAR furnaces mainly relies on the operator's experience and judgment. This involves manually adjusting melting parameters based on visual phenomena such as the arc state and voltage / current fluctuations during the melting process. This experience-based control method has the following significant problems: First, low prediction accuracy; differences in operator experience lead to inconsistent melting rate control and large quality fluctuations between different batches of ingots. Second, lag in response; manual adjustments cannot match changes in melting conditions in real time, easily causing significant fluctuations in melting rate. Third, low production efficiency; to avoid quality defects, conservative melting parameters are usually used, extending the production cycle and increasing energy costs. Fourth, high learning costs; the training period for skilled operators is long, making it difficult to meet the needs of large-scale production.

[0005] While existing technologies attempt to use electrical parameters to aid in determining melting rate, the lack of a quantitative mathematical correlation between electrical parameters and melting rate prevents accurate prediction. Therefore, there is an urgent need to establish a mathematical model for melting rate prediction based on key electrical parameters. This model would enable scientific prediction and precise control of melting rate, addressing the shortcomings of existing empirical control methods and improving the production quality and efficiency of VAR furnace titanium alloy ingots. Summary of the Invention

[0006] The present invention aims to provide a melting rate prediction model for titanium alloy ingots in a VAR furnace. By constructing a quantitative correlation between melting electrical parameters and melting rate, the model achieves accurate prediction of melting rate, providing a basis for the scientific control of melting process.

[0007] To achieve the above objectives, the first aspect of the present invention provides a method for establishing a melting rate prediction model for titanium alloy ingots in a VAR furnace, comprising the following steps: S1. Define the core parameters of VAR furnace melting: melting current I, melting voltage U, output power P, and melting speed υ; S2. Establish a linear sub-model of smelting voltage-current: Based on the fitting of VAR furnace arc discharge characteristics and smelting test data, construct U=aI+b, where a is the current coefficient and b is the reference voltage constant. S3. Establish a linear sub-model of output power-melting rate: Based on the quantitative correlation between smelting energy input and titanium alloy melting rate, construct P=cυ+d, where c is the rate coefficient and d is the basic energy consumption constant; S4. Construct a complete prediction model: Combining P=U×I, substitute the voltage-current sub-model from step S2, and then combine it with the power-melting rate sub-model from step S3 to derive the melting rate prediction model υ=[(aI+b)×Id] / c.

[0008] The optimization also includes model verification and calibration steps: through multiple sets of parallel melting tests under different current conditions, the deviation between the predicted melting rate and the actual melting rate is compared, and the coefficients a, b, c, and d are fine-tuned.

[0009] Furthermore, in step S1, the melting current I is 20~80kA, the corresponding melting voltage U is 35~65V, and the melting speed υ is 5~30kg / min.

[0010] Better yet, in the voltage-current sub-model, a=0.5 and b=25; in the power-melting rate sub-model, c=35 and d=120, and the melting rate prediction model is υ=[(0.5I+25)×I-120] / 35.

[0011] The second aspect of the present invention provides a prediction model obtained by the method established in the first aspect.

[0012] Furthermore, the prediction model is integrated into the VAR furnace smelting automation control system. By collecting smelting current signals in real time, it dynamically outputs the predicted value of melting rate, thereby realizing closed-loop control of the smelting process.

[0013] Furthermore, the prediction model is applicable to TA series and TC series titanium alloys, as well as other titanium alloy ingots prepared by VAR furnace melting.

[0014] Working principle and beneficial effects of the present invention: Compared with the prior art, the present invention has the following beneficial effects: 1. High prediction accuracy: The melting rate is accurately predicted through a quantitative mathematical model, with the prediction deviation controlled within ±3%, replacing traditional experience-based judgment and greatly improving the reliability of prediction. 2. Real-time control: The model takes the melting current as input and can dynamically predict the melting rate by collecting the current signal in real time, providing data support for the real-time closed-loop control of the melting process and avoiding large fluctuations in the melting rate; 3. Improve quality consistency: Model-based scientific control reduces process fluctuations caused by differences in human experience, ensures stable melting rates for different batches of ingots, and improves the uniformity and density of ingot composition; 4. Optimize production efficiency: The required melting current can be calculated by working backward from the target melting rate, and process parameters can be preset in advance to avoid prolonged production cycles caused by conservative operation and reduce energy consumption costs; 5. High versatility: Applicable to VAR furnace melting processes of different types of titanium alloys (TA series, TC series, etc.), and can be adapted to different specifications of VAR furnaces through simple calibration, with a wide range of applications; 6. Lowering the operational threshold: The model simplifies the judgment logic of melting speed, allowing ordinary operators to adjust the process based on the model, thus shortening the personnel training cycle. Detailed Implementation

[0015] The following detailed description illustrates the specific implementation method: Example 1: Model accuracy verification test, to verify the prediction accuracy of the prediction model of the present invention (coefficients a=0.5, b=25, c=35, d=120, prediction formula υ=[(0.5I+25)×I-120] / 35), and ensure that the deviation meets the process requirement of ±3%.

[0016] Five sets of smelting current conditions covering the commonly used range of 20~80kA were selected, namely 25kA, 30kA, 45kA, 60kA, and 75kA. Three parallel smelting tests were conducted for each set of conditions. The electric arc was started under vacuum conditions of ≤10Pa according to the standard smelting process of VAR furnace. After stable operation, the smelting current I and smelting voltage U were collected in real time. The mass of the electrode melted per unit time was measured by weighing method, and the actual melting rate υ1 of each test group was calculated. The average value of three parallel tests was taken as the actual melting rate under this condition. The melting current I of each test group was substituted into the prediction formula to calculate the predicted melting rate υ2. The relative deviation was calculated according to the formula δ=|υ2-υ1| / υ1×100% to verify whether it meets the allowable range of ±3%.

[0017] The experimental results are shown in Table 1 below:

[0018] The relative deviations for all operating conditions are ≤±3%, with the deviation for operating condition 1, which is within the commonly used industrial range (5~30kg / min), being only 0.04%. This demonstrates that the model has excellent accuracy in core application scenarios and fully meets the process requirements for VAR furnace titanium alloy ingot smelting. Operating conditions outside the commonly used range still maintain low deviations, reflecting the model's wide-range adaptability.

[0019] Example 2: Verify the industrial application effect of accurately controlling the target melting speed by back-deriving melting parameters through the model, and provide a basis for setting actual production processes.

[0020] Experimental materials: TC4 titanium alloy electrode, the same as in Example 1; experimental equipment: industrial-grade VAR furnace, the same as in Example 1, equipped with an automated parameter control module; target melting rate υtarget = 20 kg / min (core value of commonly used process range).

[0021] Parameter back-calculation: Substituting the target value υ = 20 kg / min into the prediction formula υ = [(0.5I + 25) × I - 120] / 35, and solving for the positive root, we get I ≈ 22.3 kA; Based on the calculation results, we set the melting current to 22.3 kA, and configured other parameters according to the standard VAR furnace configuration, and started the melting process; During the melting process, we collected current and voltage signals in real time, and dynamically monitored the actual melting speed by weighing; We recorded the average actual melting speed during the stable operation phase (60 minutes) and calculated the deviation from the target value.

[0022] The average actual melting rate during the stabilization phase was 19.95 kg / min; the relative deviation from the target value of 20 kg / min was |19.95-20| / 20×100%=0.25%, which is far below the allowable range of ±3%.

[0023] Example 3: Model Validation and Calibration Experiments Based on Real Industrial Smelting Records The applicability of the model under real production conditions was verified using complete original smelting records from industrial production sites. For non-standard current conditions, the model's adaptability through coefficient calibration was verified. Tables 2 and 3 show two sets of original VAR furnace smelting records from industrial production sites.

[0024] Table 2

[0025] Table 3

[0026] Abnormal data from the arc initiation, arc adjustment, and arc termination stages were removed from Table 2. Only data from the stable melting stage, where the arc was stable and current fluctuations were ≤ ±0.5kA, were retained. The melting current I was 13.0kA (constant), exceeding the commonly used range of 20~80kA, verifying the adaptability of non-standard operating conditions. The actual melting rate υ1 = 9.65kg / min (mean of 12 valid data sets). Valid data: 9.12, 9.90, 9.23, 9.48, 9.14, 9.82, 9.20, 9.65, 9.68, 10.03, 9.72, 9.57kg / min.

[0027] Abnormal data from the arc initiation, arc adjustment, and arc termination stages were removed from Table 3. Only data from the stable melting stage, where the arc was stable and current fluctuations were ≤ ±0.5kA, were retained. The melting current I was 18.0kA, exceeding the commonly used range of 20~80kA, thus verifying the adaptability of non-standard operating conditions. The actual melting rate υ1 = 13.05kg / min (mean of 15 valid data sets). Valid data: 12.66, 14.02, 13.76, 13.63, 13.35, 13.02, 12.87, 12.77, 12.64, 13.16, 13.33, 13.26, 13.24, 13.21, 13.11kg / min.

[0028] Using the fixed coefficients (a=0.5, b=25, c=35, d=120) from Example 1, and substituting the stable currents from the two sets of records, the predicted melting rate υ2 and the initial relative deviation were calculated. The model validation results are shown in Table 4 below: Table 4

[0029] Because the current recorded in Figure 1 exceeded the commonly used range, and the actual voltage value was higher than the theoretical value, the voltage sub-model coefficients a=0.52 (originally 0.5) and b=24.8 (originally 25) were fine-tuned, while the values ​​of c and d remained unchanged. The predicted melting rate was: υ2=[(0.52×13+24.8)×13-120] / 35≈9.63kg / min. After only fine-tuning a and b, the deviation between the predicted melting rate and the actual value (9.65) was reduced to 0.21%, which fully met the accuracy requirements. This proves that the power-melting rate sub-model (c=35, d=120) is still applicable under this non-standard working condition and does not require adjustment.

[0030] It should be noted that those skilled in the art can make various modifications and improvements without departing from the technical solution of this invention. These modifications and improvements should also be considered within the scope of protection of this invention, and will not affect the effectiveness of the invention or the practicality of the patent. The scope of protection claimed in this application shall be determined by the content of its claims, and the specific embodiments described in the specification can be used to interpret the content of the claims.

Claims

1. A method for establishing a model to predict the melting rate of titanium alloy ingots in a VAR furnace, characterized in that, Includes the following steps: S1. Define the core parameters of VAR furnace melting: melting current I, melting voltage U, output power P, and melting speed υ; S2. Establish a linear sub-model of smelting voltage-current: Based on the fitting of VAR furnace arc discharge characteristics and smelting test data, construct U=aI+b, where a is the current coefficient and b is the reference voltage constant. S3. Establish a linear sub-model of output power-melting rate: Based on the quantitative correlation between smelting energy input and titanium alloy melting rate, construct P=cυ+d, where c is the rate coefficient and d is the basic energy consumption constant; S4. Construct a complete prediction model: Combining P=U×I, substitute the voltage-current sub-model from step S2, and then combine it with the power-melting rate sub-model from step S3 to derive the melting rate prediction model υ=[(aI+b)×Id] / c.

2. The prediction model establishment method according to claim 1, characterized in that, It also includes model verification and calibration steps: through multiple sets of parallel melting tests under different current conditions, the deviation between the predicted melting rate and the actual melting rate is compared, and the coefficients a, b, c, and d are fine-tuned.

3. The prediction model establishment method according to claim 2, characterized in that, In step S1, the melting current I is 20~80kA, the corresponding melting voltage U is 35~65V, and the melting speed υ is 5~30kg / min.

4. The prediction model establishment method according to claim 3, characterized in that, In the voltage-current sub-model, a=0.5 and b=25; in the power-melting rate sub-model, c=35 and d=120, and the melting rate prediction model is υ=[(0.5I+25)×I-120] / 35.

5. A prediction model, obtained by the method according to any one of claims 1 to 4.

6. The prediction model according to claim 5, characterized in that, The prediction model is integrated into the VAR furnace smelting automation control system. By collecting smelting current signals in real time, it dynamically outputs the predicted value of melting rate, thereby realizing closed-loop control of the smelting process.

7. The prediction model according to claim 6, characterized in that, The prediction model is applicable to TA series and TC series titanium alloys, as well as other titanium alloy ingots prepared by VAR furnace melting.

Images