Long-service-life elastic titanium-copper alloy for VCM elastic piece and preparation method of long-service-life elastic titanium-copper alloy

Abstract

The invention relates to the technical field of alloy processing, in particular to a long-service-life elastic titanium-copper alloy for a VCM elastic piece and a preparation method thereof.The preparation method comprises the following steps that firstly, raw materials including 55%-65% of titanium, 30%-25% of copper, 10%-8% of aluminum, 3%-1% of zinc and 2%-1% of magnesium are weighed according to the mass percent; 2, the raw materials are put into a vacuum furnace to be melted, and a liquid melt is obtained; thirdly, the melt is poured into a casting mold for casting, and a blank is obtained; according to the method, the change of the state of the blank can be responded in real time by calculating the influence of the rolling times and the thickness change of the blank on the rolling control parameters, so that the rolling process can be flexibly adjusted according to the actual situation, the adaptability during machining is improved, and the rolling speed is improved when the rolling times are gradually reduced along with the reduction of the rolling times of the blank. The rolling temperature, ratio, speed and pressure are reduced, excessive deformation and damage of blanks can be effectively prevented, and the quality of final products, namely the elastic life of the alloy, is improved.

Description

Technical Field

[0001] The invention relates to the technical field of alloy processing, and in particular to a long-life elastic titanium-copper alloy for VCM shrapnel and a preparation method thereof. Background Art

[0002] In modern mechanical manufacturing and electronic equipment, shrapnel, as an important elastic element, is widely used in the connection, support and buffering of various devices. Especially in the fields of electrical connectors, switches and micro motors, the performance of shrapnel directly affects the reliability and service life of the product. However, although traditional shrapnel materials, such as stainless steel and aluminum alloy, meet the performance requirements to a certain extent, they are prone to fatigue damage, deformation and failure during use, thus affecting the overall performance of the equipment.

[0003] After searching, Chinese patent number CN202111126464.3 discloses a high-strength titanium-copper strip, including 2.0-3.5% titanium Ti by mass, and one or more of the third elements Fe, Co, Ni, Si, Cr, P, Mg, and Mn, and the rest is composed of copper and inevitable impurities, so as to refine the crystal grains. This scheme is based on the improvement of processing technology, following the sequence of ingot smelting, surface cleaning, homogenization treatment / hot rolling, annealing, surface oxide layer cleaning and repeated cold rolling and aging treatment. It has made progress in improving characteristics through the adjustment of cold rolling deformation, annealing temperature / speed, and aging treatment temperature / time.

[0004] In the above scheme, although the material strength can be improved while taking into account the bending performance by improving the process parameters in the processing technology; however, in the processing process, the remaining number of times the material can be rolled during the rolling process is not taken into account, which makes it easy for the material to reach the fatigue limit during the rolling process, and further makes the quality of the final product unable to be guaranteed; in addition, the stress introduced by the material during the rolling process cannot be considered, which further leads to uneven performance of different batches of materials, affecting the final production quality.

[0005] Therefore, a long-life elastic titanium-copper alloy for VCM shrapnel and a preparation method thereof are proposed to solve the above-mentioned problems. Summary of the invention

[0006] Technical issues solved

[0007] In view of the above-mentioned shortcomings of the prior art, the present invention provides a long-life elastic titanium-copper alloy for VCM springs and a preparation method thereof, which can effectively solve the problem in the prior art that it is impossible to set reasonable processing parameters according to the stress introduced into the billet during rolling and the material fatigue limit when rolling the billet.

[0008] Technical Solution

[0009] To achieve the above objectives, the present invention is implemented through the following technical solutions:

[0010] The present invention provides a method for preparing a long-life elastic titanium-copper alloy for a VCM shrapnel, comprising the following steps:

[0011] Step 1: Weigh the raw materials by mass percentage, including 55%-65% titanium, 30%-25% copper, 10%-8% aluminum, 3%-1% zinc, and 2%-1% magnesium;

[0012] Step 2: placing the raw materials into a vacuum furnace for melting to obtain a liquid melt;

[0013] Step 3: pouring the melt into a casting mold for casting to obtain a blank;

[0014] Step 4: heat treat the blank, and build an oil temperature control model based on the temperature of the blank during heat treatment to calculate the initial preheating temperature and real-time adjustment temperature of the quenching oil;

[0015] Step 5: Cold processing the heat-treated blank, including:

[0016] Collect the initial thickness of the blank and rolling control parameters, based on the initial thickness of the billet The parameter adjustment model is constructed with the rolling control parameters to calculate the adjusted control parameters; the adjusted control parameters are input into the pre-trained stress damage model to calculate the stress damage value D of the billet during the rolling process, and the stress damage value D is returned to the parameter adjustment model to obtain the corrected rolling times. ;

[0017] Step 6: Repeat the operation of step 5 until the billet can no longer be rolled or the target alloy specification is formed, and the rolling process of the billet is completed.

[0018] Furthermore, the melting process in step 2 is as follows:

[0019] The internal air pressure of the vacuum furnace was reduced to below 1Pa, and the heating temperature was set to 1500°C. When the temperature reached 1450°C in the vacuum furnace, the stirring device was started, and the stirring rate was set to 200-300RPM. The stirring time was 30min. After the temperature reached 1500°C in the vacuum furnace, the temperature was maintained for 2 hours.

[0020] Furthermore, the casting method in step three is:

[0021] Preheat the mold to 50-80°C, pour the melt into the mold at a pouring rate of 5 cm / s, and stir the melt in the mold at the same time. The stirring method is spiral stirring, the stirring rate is set to 200-100 RPM, and the stirring time is 5 minutes;

[0022] After the melt is completely poured into the casting mold, the melt is cooled by autothermal heating at a cooling rate of 5°C / min until the melt is cooled to room temperature to form a billet.

[0023] Furthermore, the method of heat treating the blank includes:

[0024] Put the billet into an electric furnace, set the heating temperature of the electric furnace to 900℃, and the heating rate to 50-100℃ / min. When the temperature reaches 900℃ in the electric furnace, keep it for one hour and then take out the billet. Put the billet into preheated oil for quenching until its temperature drops to 200℃. Send the billet to the annealing furnace for annealing. The temperature of the annealing furnace is set to 400-500℃, and the heating rate is set to 50-100℃ / min. After reaching 500℃ in the annealing furnace, keep the temperature for 1 hour. After the insulation is completed, the billet is naturally cooled to room temperature.

[0025] Furthermore, the method of constructing the oil temperature control model includes:

[0026] Collect the surface temperature of the blank when it is taken out , where x represents the position of the billet surface, t represents the acquisition time; based on the surface temperature of the billet , calculate the average temperature at different locations on the billet surface , the calculation formula is:

[0027] ; Where N is the number of measurement points, represents the position of the i-th measurement point;

[0028] Based on average temperature Calculate the initial preheat temperature of the oil , the calculation formula is:

[0029] ; Where K is the correction coefficient; The time required for the blank to cool to 200°C in oil; is room temperature;

[0030] Based on the initial preheat temperature of the oil Calculate the real-time adjustment temperature of the oil , the calculation formula is:

[0031] ; In the formula, is the adjustment factor.

[0032] Furthermore, the rolling control parameters in step 5 include: current rolling temperature , Current rolling ratio , Current rolling speed , Current rolling pressure ; Set the target thickness of billet rolling ;

[0033] Ways to build a parameter tuning model include:

[0034] Based on initial thickness The rolling times S of the billet can be calculated based on the rolling control parameters. The calculation formula is: ,in is a constant coefficient;

[0035] The rolling control parameters are adjusted based on the rolling times S of the billet, and the adjusted control parameters are calculated. The adjustment formula is:

[0036] ; In the formula, To adjust the rolling temperature; is the preset temperature; is the temperature control limit value; is the temperature regulation coefficient; is the influence coefficient of thickness change on temperature after each rolling;

[0037] ; In the formula, is the adjusted rolling ratio, is the influence coefficient of thickness change after each rolling on the rolling ratio;

[0038] ; In the formula, is the adjusted rolling speed; is the preset rolling speed; is the speed adjustment coefficient; is the influence coefficient of thickness change after each rolling on rolling speed;

[0039] ; In the formula, is the adjusted rolling pressure; is the preset rolling pressure; is the rolling pressure adjustment coefficient; is the pressure control limit value; is the influence coefficient of thickness change after each rolling on rolling pressure.

[0040] Furthermore, the calculation formula of the stress damage value D is:

[0041] , A is the force area of ​​the billet during each rolling;

[0042] Rolling times after correction The calculation formula is: ;

[0043] In the formula, is the stress damage correction factor, ,in, is the fatigue limit of the material; is the selection factor.

[0044] Furthermore, the training method of the stress damage model includes:

[0045] Define the long short-term memory network model as the basic structure of the stress damage model, which includes an input layer, a hidden layer and an output layer;

[0046] Introduce the parameter vector time step d, and transform the vector time step d, stress damage value , The number of rolling times after correction Combined with the adjusted control parameters, the fused feature vector is obtained ; Where s is the number of vector time steps d;

[0047] Collect the adjusted control parameters, stress damage value D and corrected rolling times of the target billet in the past fixed time , and the corresponding fusion feature vector is constructed, recorded as the historical fusion feature vector ; Fusion of history feature vector Expand according to the vector time step to form sequence data as the training input of the stress damage model. Correspondingly, the sequence data is labeled according to the vector time step d. The labels are the historical adjusted control parameters, the historical stress damage value D, and the historical corrected rolling times. ;

[0048] Initialize the network parameters of the stress damage model; the network parameters include the weight matrix from the input layer to the hidden layer , the recurrent weight matrix of the hidden layer , the bias vector of the hidden layer , the weight matrix from hidden layer to output layer and the bias vector of the output layer ;

[0049] Defining the loss function for the stress damage model ;

[0050] In the formula, S is the length of the sequence data; For the The labels of the vector time steps are the historical adjusted control parameters, the historical stress damage value D and the historical corrected rolling times. ; It is The prediction results of the vector time step are as follows: the predicted rolling control parameters, stress damage value D and the corrected rolling times ;

[0051] The sequence data is passed to the input layer, and then passes through the hidden layer and the output layer in turn to output the prediction result, and the value of the loss function of the corresponding prediction result is calculated, and the error gradient is back-propagated from the output layer to the input layer; according to the error gradient, the network parameters are updated using the optimization algorithm to reduce the value of the loss function; the sequence data is repeatedly trained until the stress damage model converges or reaches the preset number of iterations, which indicates that the stress damage model training is completed.

[0052] Furthermore, the input layer of the basic structure is used to receive the fused feature vector As input; the hidden layer consists of y state vectors, and the calculation formula for each state vector is: ; In the formula, is the hidden state of the current vector at time step d, is the activation function, is the weight matrix from the input layer to the hidden layer, is the recurrent weight matrix of the hidden layer, is the bias vector of the hidden layer;

[0053] The output layer calculates the output result of the stress damage model based on the output of the hidden layer. The calculation formula is: ; Where Y is the output result of the output layer, that is, the vector set of the predicted adjusted control parameters, stress damage value D and the corrected rolling times, is the activation function of the output layer, is the weight matrix from the hidden layer to the output layer, is the bias vector of the output layer.

[0054] A long-life elastic titanium-copper alloy for VCM shrapnel prepared by any of the above methods.

[0055] Beneficial Effects

[0056] Compared with the prior art, the technical solution provided by the present invention has the following beneficial effects:

[0057] During the heat treatment process of the blank, the present invention dynamically adjusts the quenching temperature of the oil based on the real-time temperature change of the blank surface through real-time data feedback, which helps to maintain a suitable cooling rate during the quenching process and avoids too fast or too slow cooling; at the same time, the introduction of the adjustment factor can smooth the temperature change, reduce the instability caused by instantaneous temperature adjustment, and help reduce the deformation and cracks of the blank;

[0058] The invention can respond to the change of the state of the billet in real time by calculating the influence of the rolling times and thickness change of the billet on the rolling control parameters, so that the rolling process can be flexibly adjusted according to the actual situation, thereby improving the adaptability during processing. As the rolling times of the billet decrease, the rolling temperature, ratio, speed and pressure are gradually reduced when the rolling times are small, which can effectively prevent the excessive deformation and damage of the billet and improve the quality of the final product, that is, the elastic life of the alloy;

[0059] The present invention also updates the number of times the billet can be rolled based on the stress damage value of the billet, so that the stress damage model can more comprehensively consider the actual performance changes of the billet during the rolling process. Whether it is a linear correction or a nonlinear correction, the prediction of the number of times that can be rolled can be dynamically adjusted according to different stress damage values, so that the dynamic adjustment result of the rolling control number is more in line with the actual situation, especially when the billet is close to the fatigue limit, it can effectively avoid material damage caused by excessive rolling. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] 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 prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0061] Figure 1 The figure is a schematic flow chart of a method for preparing a long-life elastic titanium-copper alloy in Embodiment 1 of the present invention. DETAILED DESCRIPTION

[0062] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0063] The present invention will be further described below in conjunction with the embodiments.

[0064] Embodiment 1:

[0065] See attached Figure 1 , this case proposes a method for preparing a long-life elastic titanium-copper alloy for VCM shrapnel, comprising the following steps:

[0066] Step 1: Weigh the raw materials by mass percentage, including 55%-65% titanium, 30%-25% copper, 10%-8% aluminum, 3%-1% zinc, and 2%-1% magnesium;

[0067] Step 2: placing the raw materials into a vacuum furnace for melting to obtain a liquid melt;

[0068] Step 3: pouring the melt into a casting mold for casting to obtain a blank;

[0069] Step 4: heat treat the blank, and build an oil temperature control model based on the temperature of the blank during heat treatment to calculate the initial preheating temperature and real-time adjustment temperature of the quenching oil;

[0070] Step 5: Cold processing the heat-treated blank, including:

[0071] Collect the initial thickness of the blank and rolling control parameters, based on the initial thickness of the billet The parameter adjustment model is constructed with the rolling control parameters to calculate the adjusted control parameters; the adjusted control parameters are input into the pre-trained stress damage model to calculate the stress damage value D of the billet during the rolling process, and the stress damage value D is returned to the parameter adjustment model to obtain the corrected rolling times. ;

[0072] Step 6: Repeat the operation of step 5 until the billet can no longer be rolled or the target alloy specification is formed, and the rolling process of the billet is completed.

[0073] Specifically, in this scheme, the method of melting the raw materials in the vacuum furnace in step 2 is: put the raw materials weighed in proportion into the vacuum furnace (make sure that the purity of the raw materials is greater than 99% and they are clean and dry), start the vacuum equipment, and evacuate the inside of the vacuum furnace to a vacuum state below 1Pa to prevent oxygen and moisture from reacting with the alloy during the smelting process to cause oxidation; while evacuating the vacuum, the furnace can be heated to promote gas overflow; the heating temperature of the vacuum furnace is set to 1500°C, and thermal stress caused by excessive temperature rise can be avoided by gradually increasing the temperature; after reaching the target temperature in the vacuum furnace, maintain the temperature for 2 hours to ensure that the alloy components are fully mixed;

[0074] At the same time, when the temperature reaches 1450°C in the vacuum furnace, the stirring device is started after all alloy elements are completely melted. The stirring rate is set to 200-300RPM and the stirring time is 30 minutes to ensure that all elements in the melt are fully mixed without introducing too many bubbles, while eliminating possible precipitates and irregular material distribution; at the same time, the stirring method adopts axial flow stirring design to achieve uniform flow and promote thorough mixing of components in the melt;

[0075] Through the above-mentioned specific smelting parameter design, the heating process of the titanium-copper alloy can be effectively controlled, avoiding thermal stress caused by rapid temperature changes, forming a homogenous melt, ensuring the uniform dispersion of each element in the alloy, reducing defects in the subsequent cooling process, and laying a good foundation for subsequent casting and processing.

[0076] More specifically, in this scheme, the method of casting the melt in step three is: select a suitable casting mold according to the target size, and select an aluminum or graphite casting mold for the casting mold. Both materials have excellent thermal conductivity and high temperature resistance, and can withstand the high temperature of the melt without deformation or melting; at the same time, before use, preheat the mold to 50°C-80°C to reduce the thermal shock when the melt is poured in, and prevent the mold from breaking due to excessive temperature difference; and before the melt is poured into the mold, it is necessary to ensure that the mold area is clean to prevent impurities and dust from entering the mold to reduce the formation of pores or debris;

[0077] Before pouring the melt into the casting mold, preheat the casting mold to effectively reduce the temperature gradient and thermal stress during the casting process; pour the liquid melt slowly and evenly into the casting mold, with a pouring flow rate of 5 cm / s, to ensure uniform flow and filling of the melt in the casting mold, reduce the generation of defects and bubbles, and help improve the quality of the billet, while protecting the mold and improving the mechanical properties of the material; setting the pouring flow rate not exceeding 5 cm / s helps to ensure that the casting quality meets the expected standards while meeting production efficiency; and stirring the melt in the casting mold while pouring the melt, the stirring method is spiral stirring, which can effectively stir the melt evenly in the casting mold, so that it can generate an upward flow when rotating, promote the flow and mixing of the melt, and ensure that the melt is evenly distributed throughout the casting mold; the stirring rate is set to 200-100 RPM, which is relatively low, aiming to avoid the generation of too many bubbles and ensure the fluidity of the melt; the stirring time is 5 minutes to ensure that the melt is fully mixed in the casting mold to avoid local unevenness caused by temperature changes;

[0078] After the melt is completely poured into the casting mold, the melt begins to cool down. The melt is allowed to cool down naturally in the mold to avoid forced cooling, so as to reduce the stress and deformation caused by too fast cooling speed. The temperature of the cooling environment is controlled to ensure that the cooling speed of the melt is maintained at 5°C / min, so that the melt can crystallize evenly and reduce internal defects. When the melt is cooled to room temperature, an alloy billet is obtained.

[0079] Furthermore, in this solution, the method of heat treating the blank in step 4 includes:

[0080] Place the billet in an electric furnace to provide a stable heating environment for heat treatment; set the heating temperature of the electric furnace to 900°C, and control the heating rate to 50-100°C / min during the heating process to avoid thermal stress caused by rapid temperature changes; when the temperature in the electric furnace reaches 900°C, keep the billet at this temperature for one hour to ensure that all elements in the alloy are fully dissolved to form a uniform solid solution; this process helps to eliminate segregation and unevenness that may be formed during the casting process;

[0081] After the insulation is completed, the billet is immediately taken out of the electric furnace and placed in the preheated oil for quenching. The preheating temperature of the oil is 60-80℃ to reduce the thermal stress and deformation during the quenching process; the billet is completely immersed in the oil until its temperature drops to 200℃; after quenching, in order to prevent the internal stress that may be generated in the billet, the billet is sent to the annealing furnace for stress relief annealing; the temperature of the annealing furnace is set to 400-500℃, which is the best choice for eliminating internal stress and restoring material toughness; the heating rate of the annealing furnace is set to 50-100℃ / min to reduce the thermal stress caused by too fast temperature change; after reaching the target temperature in the annealing furnace, the billet is kept at this temperature for one hour. This process will promote phase transformation and grain rearrangement to ensure that the stress inside the billet is fully released; after the billet is finished with insulation, the power supply of the annealing furnace is turned off, the temperature in the annealing furnace is allowed to drop naturally, and the billet is allowed to slowly cool to room temperature. Slow cooling helps prevent the re-initiation of internal stress caused by too fast cooling rate.

[0082] It is worth noting that when the billet is taken out of the electric furnace and placed in the oil for quenching, the preheating temperature of the oil needs to be adjusted according to the temperature of the billet when it is taken out, so as to avoid the billet from cooling down too quickly during the quenching process, which may increase brittleness, form microcracks or notches, and reduce the toughness and impact resistance of the material; at the same time, the temperature of the oil needs to be dynamically adjusted according to the temperature change of the billet during the cooling process. When quenching in superheated oil, bubbles are generated when the billet surface contacts the oil. These bubbles can cause uneven local cooling of the billet surface to form a temperature gradient, which will cause the concentration of internal stress and further lead to cracks or defects; by measuring the surface temperature of the billet when it is taken out of the electric furnace, and inputting the temperature data into the pre-built oil temperature control model, the initial preheating temperature of the oil is obtained. When the billet enters the oil for quenching, the cooling temperature of the billet surface is collected in real time, and the cooling temperature is returned to the oil temperature control model to obtain the real-time control temperature of the oil, thereby more accurately controlling the cooling rate of the billet;

[0083] The construction methods of the oil temperature control model include:

[0084] Collect the surface temperature of the blank when it is taken out , where x represents the position of the billet surface (such as front, back, left, right, etc.), and t represents the acquisition time; based on the surface temperature of the billet , calculate the average temperature at different locations on the billet surface , the calculation formula is:

[0085] ; Where N is the number of measurement points, represents the position of the i-th measurement point;

[0086] Based on average temperature Calculate the initial preheat temperature of the oil , the calculation formula is:

[0087] ; In the formula, K is the correction coefficient, which is obtained based on experimental data; The time required for the blank to cool to 200°C in oil; is room temperature;

[0088] Based on the initial preheat temperature of the oil Calculate the real-time adjustment temperature of the oil , the calculation formula is:

[0089] ; In the formula, It is a regulating factor between 0 and 1. The oil controls the adjustment speed of the oil temperature to avoid unstable cooling process caused by rapid temperature changes, minimize the risk of deformation and cracks in the billet, and improve the quality and performance of the casting;

[0090] In summary, the oil temperature control model is based on the real-time temperature changes of the billet surface. It dynamically adjusts the quenching temperature of the oil through real-time data feedback, which helps to maintain a suitable cooling rate during the quenching process and avoid too fast or too slow cooling. At the same time, the adjustment factor , which can smooth temperature changes, reduce the instability caused by instantaneous temperature adjustment, and help reduce billet deformation and cracks.

[0091] Furthermore, in this case, the method of cold working the blank in step 4 includes:

[0092] The billet is sent to the cold rolling mill for rolling, and the rolling temperature is set to 20-25°C. Within this temperature range, the degradation of material properties caused by high temperature can be effectively avoided; the initial rolling ratio is 30%, and the thickness is gradually reduced to the required thickness to obtain a qualified target alloy.

[0093] It is worth mentioning that during the billet rolling process, stress damage is an important factor affecting the quality of the alloy. After the billet is deformed and pressurized during the rolling process, a certain amount of stress damage will accumulate inside. When this stress damage reaches or exceeds the fatigue limit of the material, the billet will no longer be able to withstand further processing. When the billet enters the cold rolling mill for the first rolling, the initial thickness of the billet is collected. and rolling control parameters, the rolling control parameters include: current rolling temperature , Current rolling ratio , Current rolling speed , Current rolling pressure ; Set the target thickness of billet rolling ; Based on the above parameters, the rolling times S of the billet can be calculated, and the calculation formula is: ,in It is a constant coefficient, usually obtained through analysis of experimental data or historical data, and represents the durability or rolling performance of the material;

[0094] Based on the initial thickness of the blank The parameter adjustment model is constructed by combining the rolling control parameters and calculating the adjusted control parameters. The construction method includes:

[0095] During the rolling process, the temperature of the billet has an important influence on the plasticity of the billet and the quality of the final product. Therefore, the temperature of the billet during the rolling process needs to be dynamically adjusted according to the number of times the billet can be rolled S. The adjustment formula is: ; In the formula, To adjust the post-rolling temperature, which is the target temperature that needs to be adjusted during the rolling process based on real-time monitoring data and calculation results. This value can ensure that the billet is kept within the most suitable temperature range during the rolling process, thereby improving processing quality and efficiency; The preset temperature is the ideal temperature set during the rolling process, usually set based on material properties to achieve optimal plastic deformation and product quality. The temperature control limit value is a safety threshold, usually the minimum and maximum temperature range allowed by the system. When the temperature exceeds this range, it will cause potential impact on material properties, such as decreased hardness, increased plasticity or material damage. It is critical to ensure that the temperature is within this range. is the temperature adjustment coefficient, which is used to control the sensitivity of temperature adjustment. A larger value means a stronger response and faster temperature adjustment, while a smaller value means a stronger response and faster temperature adjustment. The value makes the adjustment process smoother. Usually set according to the equipment and material characteristics; is the influence coefficient of thickness change on temperature after each rolling;

[0096] The rolling ratio reflects the thickness change of the billet during the rolling process and directly affects the deformation degree of the material. Therefore, it needs to be dynamically adjusted according to the number of times the billet can be rolled during the rolling process. The adjustment formula is: ; In the formula, is the adjusted rolling ratio, is the influence coefficient of thickness change after each rolling on the rolling ratio;

[0097] During the rolling process, the rolling speed affects the processing efficiency and stress state of the billet. An appropriate speed can improve the production quality. During the rolling process, it can be dynamically adjusted based on the temperature and the number of times the billet can be rolled S. The adjustment formula is: ; In the formula, The adjusted rolling speed is the target rolling speed that needs to be adjusted according to the real-time monitoring and calculation results during the rolling process. This value can ensure that the material is kept in the appropriate speed range during the rolling process to optimize the processing quality and efficiency. The preset rolling speed is the ideal rolling speed set during the rolling process design phase. It is usually set based on material properties and processing requirements. It is the speed at which the machine is expected to run under ideal conditions to process the material evenly and efficiently. is the speed adjustment coefficient, which is used to control the adjustment sensitivity of the rolling speed; is the influence coefficient of thickness change after each rolling on rolling speed;

[0098] During the rolling process, the rolling pressure is the force applied to the billet, which affects the deformation and processing quality of the material. During the rolling process, it can be dynamically adjusted based on the number of times the billet can be rolled S. The adjustment formula is: ; In the formula, The adjusted rolling pressure is the target pressure that needs to be adjusted during the rolling process according to the real-time monitoring and calculation results. This value can ensure that the pressure applied to the material during the rolling process is optimal, thereby optimizing the deformation effect of the material and the quality of the final product. The preset rolling pressure is the ideal pressure set in the rolling process design stage. It is often set based on the material properties and process requirements. The preset value is intended to achieve the best processing effect and ensure the ideal deformation of the material during the rolling process. is the rolling pressure adjustment coefficient, which is used to control the sensitivity of pressure adjustment; It is the pressure control limit value, which is a safety threshold, usually the minimum and maximum pressure range allowed by the system; is the influence coefficient of thickness change on rolling pressure after each rolling;

[0099] In this case, by calculating the impact of the number of times the billet can be rolled and the thickness change on the rolling control parameters, it is possible to respond to changes in the billet state in real time, so that the rolling process can be flexibly adjusted according to actual conditions, thereby improving the adaptability during processing. As the number of times the billet can be rolled decreases, the rolling temperature, ratio, speed and pressure are gradually reduced. When the number of rolling times is small, the excessive deformation and damage of the billet can be effectively prevented, and the quality of the final product, that is, the elastic life of the alloy, can be improved.

[0100] Based on the adjusted control parameters, including rolling pressure , Adjusted rolling speed , Adjusted temperature , Adjusted rolling ratio and current rolling ratio Input into the pre-trained stress damage model to calculate the stress damage value of the billet during the rolling process ;Wherein, A is the force area of ​​the billet during each rolling;

[0101] Then, based on the stress damage value D of the billet during the rolling process, the number of times the billet can be rolled S is updated, and the update formula is: ; In the formula, is the number of rolling times after correction; is the stress damage correction factor, which reflects the influence of the stress damage value D on the number of rolling times. ,in, is the fatigue limit of the material; Assume that the relationship between the stress damage value D and the number of times the material can be rolled S is linear, that is, the increase in the stress damage value D causes the number of times S to decrease at a fixed ratio. , which is applicable when the blank is in a smaller stress range and stress damage will not have a significant impact on the performance of the material. The linear correction can better describe the behavior of the material. When the rolling times S exceeds the fatigue limit When , the model may be unreasonable, indicating that the material cannot continue to be stressed, so rolling needs to be stopped. Nonlinear correction is suitable for situations where stress damage is large and close to the fatigue limit, and can better predict whether the material can continue to be rolled;

[0102] The number of rolling times after correction Return to the stress damage model and repeat the above process until the billet can no longer be rolled, thus completing the rolling process of the billet;

[0103] The rolling times S of the billet are updated based on the stress damage value D of the billet, so that the stress damage model can more comprehensively consider the actual performance changes of the billet during the rolling process. Whether it is a linear correction or a nonlinear correction, the prediction of the rolling times S can be dynamically adjusted according to different stress damage values ​​D, so that the dynamic adjustment result of the rolling control times is more in line with the actual situation, especially when the billet is close to the fatigue limit, which can effectively avoid material damage caused by excessive rolling.

[0104] The training methods of stress damage model include:

[0105] A long short-term memory network model is defined as the basic structure of the stress damage model, and the basic structure includes an input layer, a hidden layer, and an output layer;

[0106] Introduce the parameter vector time step d, and transform the vector time step d, stress damage value , The number of rolling times after correction Combined with the adjusted control parameters, the fused feature vector is obtained ; Where s is the number of vector time steps d;

[0107] Collect the adjusted control parameters, stress damage value D and corrected rolling times of the target billet in the past fixed time , and the corresponding fusion feature vector is constructed, recorded as the historical fusion feature vector ; Fusion of history feature vector Expand according to the vector time step to form sequence data as the training input of the stress damage model. Correspondingly, the sequence data is labeled according to the vector time step d. The labels are the historical adjusted control parameters, the historical stress damage value D, and the historical corrected rolling times. ;

[0108] The input layer is used to receive the fused feature vector As input; the hidden layer consists of y state vectors, and the calculation formula for each state vector is: ; In the formula, is the hidden state of the current vector at time step d, is the activation function, is the weight matrix from the input layer to the hidden layer, is the recurrent weight matrix of the hidden layer, is the bias vector of the hidden layer;

[0109] The output layer calculates the output result of the stress damage model based on the output of the hidden layer. The calculation formula is: ; Where Y is the output result of the output layer, that is, the vector set of the predicted adjusted control parameters, stress damage value D and the corrected rolling times, is the activation function of the output layer, is the weight matrix from the hidden layer to the output layer, is the bias vector of the output layer;

[0110] Initialize the network parameters of the stress damage model; the network parameters include the weight matrix from the input layer to the hidden layer , the recurrent weight matrix of the hidden layer , the bias vector of the hidden layer , the weight matrix from hidden layer to output layer and the bias vector of the output layer ;

[0111] Defining the loss function for the stress damage model ;

[0112] In the formula, S is the length of the sequence data; For the The labels of the vector time steps are the historical adjusted control parameters, the historical stress damage value D and the historical corrected rolling times. ; It is The prediction results of the vector time step are as follows: the predicted rolling control parameters, stress damage value D and the corrected rolling times ;

[0113] The sequence data is passed to the input layer, and then passes through the hidden layer and the output layer in turn to output the prediction result, and the value of the loss function of the corresponding prediction result is calculated, and the error gradient is back-propagated from the output layer to the input layer; according to the error gradient, the network parameters are updated using the optimization algorithm to reduce the value of the loss function; the sequence data is repeatedly trained until the stress damage model converges (that is, the value of the damage function no longer changes) or the preset number of iterations is reached, which indicates that the stress damage model training is completed.

[0114] Embodiment 2:

[0115] Based on Example 1, this case proposes a long-life elastic titanium-copper alloy for VCM shrapnel, which is prepared by the preparation method of a long-life elastic titanium-copper alloy for VCM shrapnel in Example 1.

[0116] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements will not cause the essence of the corresponding technical solutions to deviate from the protection scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for preparing a long-life elastic titanium-copper alloy for VCM shrapnel, characterized in that: The following steps are involved: Step 1: Weigh the raw materials by mass percentage, including 55%-65% titanium, 30%-25% copper, 10%-8% aluminum, 3%-1% zinc, and 2%-1% magnesium; Step 2: placing the raw materials into a vacuum furnace for melting to obtain a liquid melt; Step 3: pouring the melt into a casting mold for casting to obtain a blank; Step 4: heat treat the blank, and build an oil temperature control model based on the temperature of the blank during heat treatment to calculate the initial preheating temperature and real-time adjustment temperature of the quenching oil; Step 5: Cold processing the heat-treated blank, including: Collect the initial thickness of the blank and rolling control parameters, based on the initial thickness of the billet A parameter adjustment model is constructed with rolling control parameters to calculate and obtain the adjusted control parameters; The adjusted control parameters are input into the pre-trained stress damage model to calculate the stress damage value D of the billet during the rolling process, and the stress damage value D is returned to the parameter adjustment model to obtain the corrected rolling times. ; Step 6: Repeat the operation of step 5 until the billet can no longer be rolled or the target alloy specification is formed, and the rolling process of the billet is completed.

2. The method for preparing a long-life elastic titanium-copper alloy for VCM shrapnel according to claim 1, characterized in that: The method for melting in step 2 is: The internal air pressure of the vacuum furnace was reduced to below 1Pa, and the heating temperature was set to 1500°C. When the temperature reached 1450°C in the vacuum furnace, the stirring device was started, and the stirring rate was set to 200-300RPM. The stirring time was 30min. After the temperature reached 1500°C in the vacuum furnace, the temperature was maintained for 2 hours.

3. The method for preparing a long-life elastic titanium-copper alloy for VCM shrapnel according to claim 1, characterized in that: The method of casting in step 3 is: Preheat the mold to 50-80°C, pour the melt into the mold at a pouring rate of 5 cm / s, and stir the melt in the mold at the same time. The stirring method is spiral stirring, the stirring rate is set to 200-100 RPM, and the stirring time is 5 minutes; After the melt is completely poured into the casting mold, the melt is cooled by autothermal heating at a cooling rate of 5°C / min until the melt is cooled to room temperature to form a billet.

4. The method for preparing a long-life elastic titanium-copper alloy for VCM shrapnel according to claim 1, characterized in that: The method of heat treating the blank comprises: Put the billet into an electric furnace, set the heating temperature of the electric furnace to 900℃, and the heating rate to 50-100℃ / min. When the temperature reaches 900℃ in the electric furnace, keep it for one hour and then take out the billet. Put the billet into preheated oil for quenching until its temperature drops to 200℃. Send the billet to the annealing furnace for annealing. The temperature of the annealing furnace is set to 400-500℃, and the heating rate is set to 50-100℃ / min. After reaching 500℃ in the annealing furnace, keep the temperature for 1 hour. After the insulation is completed, the billet is naturally cooled to room temperature.

5. The method for preparing a long-life elastic titanium-copper alloy for VCM shrapnel according to claim 4, characterized in that: The method of constructing the oil temperature control model includes: Collect the surface temperature of the blank when it is taken out , where x represents the position of the billet surface, t represents the acquisition time; based on the surface temperature of the billet , calculate the average temperature at different locations on the billet surface , the calculation formula is: ; Where N is the number of measurement points, represents the position of the i-th measurement point; Based on average temperature Calculate the initial preheat temperature of the oil , the calculation formula is: ; Where K is the correction coefficient; The time required for the blank to cool to 200°C in oil; is room temperature; Based on the initial preheat temperature of the oil Calculate the real-time adjustment temperature of the oil , the calculation formula is: ; In the formula, is the adjustment factor.

6. The method for preparing a long-life elastic titanium-copper alloy for VCM shrapnel according to claim 1, characterized in that: The rolling control parameters in step 5 include: current rolling temperature , Current rolling ratio , Current rolling speed , Current rolling pressure ; Set the target thickness of billet rolling ; Ways to build a parameter tuning model include: Based on initial thickness The rolling times S of the billet can be calculated based on the rolling control parameters. The calculation formula is: ,in is a constant coefficient; The rolling control parameters are adjusted based on the rolling times S of the billet, and the adjusted control parameters are calculated. The adjustment formula is: ; In the formula, To adjust the rolling temperature; is the preset temperature; is the temperature control limit value; is the temperature regulation coefficient; is the influence coefficient of thickness change on temperature after each rolling; ; In the formula, is the adjusted rolling ratio, is the influence coefficient of thickness change after each rolling on the rolling ratio; ; In the formula, is the adjusted rolling speed; is the preset rolling speed; is the speed adjustment coefficient; is the influence coefficient of thickness change after each rolling on rolling speed; ; In the formula, is the adjusted rolling pressure; is the preset rolling pressure; is the rolling pressure adjustment coefficient; is the pressure control limit value; is the influence coefficient of thickness change after each rolling on rolling pressure.

7. The method for preparing a long-life elastic titanium-copper alloy for VCM shrapnel according to claim 6, characterized in that: The calculation formula of the stress damage value D is: , A is the force area of ​​the billet during each rolling; Rolling times after correction The calculation formula is: ; In the formula, is the stress damage correction factor, ,in, is the fatigue limit of the material; is the selection factor.

8. The method for preparing a long-life elastic titanium-copper alloy for VCM shrapnel according to claim 7, characterized in that: The training method of the stress damage model includes: Define the long short-term memory network model as the basic structure of the stress damage model, which includes an input layer, a hidden layer and an output layer; Introduce the parameter vector time step d, and transform the vector time step d, stress damage value , The number of rolling times after correction Combined with the adjusted control parameters, the fused feature vector is obtained ; Where s is the number of vector time steps d; Collect the adjusted control parameters, stress damage value D and corrected rolling times of the target billet in the past fixed time , and the corresponding fusion feature vector is constructed, recorded as the historical fusion feature vector ; Fusion of history feature vector Expand according to the vector time step to form sequence data as the training input of the stress damage model. Correspondingly, the sequence data is labeled according to the vector time step d. The labels are the historical adjusted control parameters, the historical stress damage value D, and the historical corrected rolling times. ; Initialize the network parameters of the stress damage model; the network parameters include the weight matrix from the input layer to the hidden layer , the recurrent weight matrix of the hidden layer , the bias vector of the hidden layer , the weight matrix from hidden layer to output layer and the bias vector of the output layer ; Defining the loss function for the stress damage model ; In the formula, S is the length of the sequence data; For the The labels of the vector time steps are the historical adjusted control parameters, the historical stress damage value D and the historical corrected rolling times. ; It is The prediction results of the vector time step are as follows: the predicted rolling control parameters, stress damage value D and the corrected rolling times ; The sequence data is passed to the input layer, and then passes through the hidden layer and the output layer in turn to output the prediction result, and the value of the loss function of the corresponding prediction result is calculated, and the error gradient is back-propagated from the output layer to the input layer; according to the error gradient, the network parameters are updated using the optimization algorithm to reduce the value of the loss function; the sequence data is repeatedly trained until the stress damage model converges or reaches the preset number of iterations, which indicates that the stress damage model training is completed.

9. The method for preparing a long-life elastic titanium-copper alloy for VCM shrapnel according to claim 1, characterized in that: The input layer of the basic structure is used to receive the fused feature vector As input; the hidden layer consists of y state vectors, and the calculation formula for each state vector is: ; In the formula, is the hidden state of the current vector at time step d, is the activation function, is the weight matrix from the input layer to the hidden layer, is the recurrent weight matrix of the hidden layer, is the bias vector of the hidden layer; The output layer calculates the output result of the stress damage model based on the output of the hidden layer. The calculation formula is: ; Where Y is the output result of the output layer, that is, the vector set of the predicted adjusted control parameters, stress damage value D and the corrected rolling times, is the activation function of the output layer, is the weight matrix from the hidden layer to the output layer, is the bias vector of the output layer.

10. A long-life elastic titanium-copper alloy for VCM shrapnel prepared according to any one of claims 1 to 9.

Images