Formula selection method for improving hardness of gray cast iron brake drum

Abstract

A formula selection method for improving the hardness of a gray cast iron brake drum is characterized by comprising the following steps that S1, the size range of a gray cast iron brake drum product is selected; s2, selecting elements suitable for addition or adjustment and a content range; s3, preferably selecting 3-5 specific elements from the elements in the S2 according to a cost factor; s4, according to the content of elements to be added after optimization in the step S3, carrying out actual production test on the product in the step S1, and recording data of component, size and hardness change; and S5, arranging the data recorded in the step S4, and processing the data by using an artificial neural network. According to the invention, aiming at the iron castings such as the brake drum, a rapid reference for adding (or adjusting) each element in production to improve the performance of the iron castings is given, and a proper component proportion of each element can be obtained in a short time, so that the performance indexes such as hardness and the like of the iron castings are improved, and repeated test and exploration are not needed.

Description

Technical Field

[0001] The present invention relates to the field of casting, and particularly relates to a method for selecting a formula to improve the hardness of gray cast iron brake drums. Background Art

[0002] Due to its good wear resistance, thermal conductivity, and casting performance, gray cast iron has a relatively low manufacturing cost while meeting the usage requirements of brake drums. Therefore, it has long been the preferred material for automotive brake drums.

[0003] In recent years, the scale of the logistics industry has been continuously expanding. As an important part of transportation, automotive transportation has received increasing attention, and its load capacity and power have gradually increased.

[0004] From the perspective of safety, the increase in operating power places higher requirements on the braking performance of automobiles. As a key component of the braking system of large trucks, the performance indicators of brake drums are related to the braking performance of the entire vehicle. However, the current gray cast iron brake drums can no longer meet the relevant requirements.

[0005] There are mainly three ways to improve the performance of gray cast iron brake drums: composition improvement, subsequent heat treatment, and structural improvement. For composition improvement, adjusting the contents of C and Si in gray cast iron and adding Cr, Ni, Cu, Sn, V, Nb, Mo, Ti, and rare earth elements can improve its properties such as hardness. For the heat treatment process, after casting, it is cooled with the mold to a specific temperature and then rapidly cooled, and then subjected to low-temperature tempering treatment to obtain gray cast iron parts with excellent performance.

[0006] The invention patent with the publication number CN 1718825 A proposes a method for obtaining high-strength gray cast iron materials through microalloying. Specifically, it involves optimizing the alloy composition design and adding elements such as V and N for microalloying treatment to finally obtain high-strength gray cast iron with excellent mechanical properties. The invention patent with the publication number CN 102747267 A proposes a design method for microalloyed ultra-high-strength gray cast iron materials with high carbon equivalent. Specifically, by optimizing the alloy composition design and adding trace amounts of Zr, Ti, V, and N elements, the strength of gray cast iron with high carbon equivalent is significantly improved. The invention patent with the publication number CN 103074538 A also proposes a preparation method for microalloyed ultra-high-strength gray cast iron with high carbon equivalent. The invention patent with the publication number CN 111673045 A proposes a high-carbon equivalent high-strength gray cast iron part and its lost foam casting process. The specific composition of the high-carbon equivalent high-strength gray cast iron includes: C: 3.9% - 4.2%, Si: 1.6% - 1.8%, Mn: 1.9% - 2.3%, Cu: 0.5% - 0.7%, and the balance is iron. The invention patent with the publication number CN 109898014 A proposes a gray cast iron material with high strength and reduced defects, including: carbon (C) in an amount of about 3.10 to 3.50 wt%, silicon (Si) in an amount of about 2.10 to 2.40 wt%, manganese (Mn) in an amount of about 0.50 to 0.80 wt%, phosphorus (P) in an amount less than or equal to about 0.10 wt% (excluding 0%), sulfur (S) in an amount less than or equal to about 0.10 wt% (excluding 0%), chromium (Cr) in an amount of about 0.25 to 0.45 wt%, copper (Cu) in an amount of about 1.00 to 1.40 wt%, nickel (Ni) in an amount less than or equal to about 0.20 wt% (excluding 0%), and the balance of iron (Fe). The invention patent with the publication number CN 108315633 A proposes a high thermal conductivity high-strength gray cast iron and its preparation method. The content of each element in this gray cast iron is 3.3 - 3.8 wt% C, 1.2 - 1.8 wt% Si, 0.4 - 0.8 wt% Mn, 0.1 - 0.5 wt% Mo, 0.4 - 0.8 wt% Cu, 0.05 - 0.15 wt% Sn, 0.01 - 0.04 wt% S, 0.004 - 0.02 wt% Sr, P < 0.03 wt%, Ni < 0.5 wt%, and the balance is Fe and inevitable impurities; this gray cast iron is obtained through melting, inoculation, and pouring. The tensile strength of this gray cast iron is 280 - 350 MPa, the hardness is 210 - 240 HB, and the thermal conductivity is 50 - 60 W / (m·K).The invention patent with the publication number CN 107119221 A proposes a microalloyed high-strength gray iron casting and its melting method. The gray iron casting contains the following components in weight percentage: C 2.92 - 3.48%, Si 1.52 - 2.36%, Mn 0.22 - 0.78%, Cr 0.15 - 0.5%, Cu 0.3 - 0.78%, Nb 0.011 - 0.098%, V 0.02 - 0.1%, Ti 0.015 - 0.025%, and the rest is iron and other trace inevitable impurities. This invention strengthens the casting matrix through microalloying, improves the strength and stiffness of the casting, adds an efficient inoculant, can effectively improve the graphite morphology, hardness, tissue uniformity, and reduce the sensitivity, enhances the mechanical properties of the gray iron casting. By adopting the in-mold inoculation method, it can improve the utilization rate of the inoculant, improve the inoculation effect, and reduce costs.

[0007] The utility model patent with the publication number CN 210686782 U proposes a high-strength gray iron automotive brake drum. Specifically, a number of heat dissipation holes are opened in the transition connection part of the traditional brake drum body, and an air collecting cavity is welded at the corresponding positions of the heat dissipation holes. The two ends of the air collecting cavity adopt the design of streamlined windward surfaces, which greatly avoids the air collecting cavity generating heat again due to friction with air resistance. At the same time, the structural design of the air collecting cavity can not only take away the frictional heat generated by the sudden braking of the brake drum body through the heat dissipation grilles on both sides of it, but also use the air to continuously blow the inner cavity surface of the air collecting cavity to reduce its temperature, achieving the function of heat dissipation and temperature reduction, thereby prolonging the service life of the brake drum and improving the safety factor of the brake drum. The utility model patent with the publication number CN 208935224 U also proposes a high-strength gray iron automotive brake drum. Specifically, through the combined action of a braking mechanism, a first composite layer, a first brake shoe, a second brake shoe, a cavity, and a second composite layer, the wear of the first brake shoe and the second brake shoe caused by friction is reduced by using the first wear-resistant layer and the second wear-resistant layer. The heat generated by friction is dissipated by using heat dissipation holes and cooling water, and the noise generated by friction is absorbed and isolated by using the first sound-absorbing layer, the first vacuum layer, the second vacuum layer, and the second sound-absorbing layer. The utility model patent with the publication number CN 207848262 U also proposes a high-strength gray iron automotive brake drum. Specifically, through the heat dissipation fins and sound-absorbing cotton installed in the heat dissipation slots, the braking frictional heat and frictional noise generated during braking on the brake drum can be quickly dissipated and absorbed, reducing the noise during the use of the automotive brake drum, effectively improving the heat dissipation performance of the automotive brake drum. The frictional abrasives generated by the braking friction of the brake drum can be discharged from the chip removal groove of the brake drum, effectively improving the braking stability of the automotive brake drum.

[0008] All of the above three approaches have the problem of excessively high costs. The costs of precious metals such as Cr, Ni, Cu, Sn, V, Nb, Mo, Ti, and rare earth elements are relatively high, and it will also cause waste of some rare metal resources. Moreover, the equipment and processes required for subsequent heat treatment of castings also involve large investments. Therefore, adopting these approaches will reduce the advantage of low cost of gray iron brake drums.

[0009] And through experiments, it is found that adding the same content of alloying elements brings different performance improvements to brake drums of different sizes. In order to reduce waste in costs, it is necessary to clarify the improvement effect of adding a specific content of alloying elements on the hardness of brake drum products of specific sizes. At the same time, the relationship between the improvement of hardness and the added elements and sizes is not a simple linear correspondence relationship, and there are multiple independent variables, and a suitable expression formula cannot be obtained through simple fitting. Summary of the Invention

[0010] The object of the present invention is to overcome the defects and deficiencies of the prior art. For cast iron parts such as brake drums, a quick reference for adding (or adjusting) various elements in production to improve their performance is given, which can obtain appropriate component ratios of various elements in a short time, thereby improving performance indicators such as its hardness, and a method for selecting the formula of the hardness of gray iron brake drums that does not require repeated experimentation to explore.

[0011] Artificial neural network is a branch of machine learning models, constructed based on the principle of neuron organization found in the biological neural networks that make up the animal brain. The nodes in an artificial neural network are called artificial neurons, and the connections between them can transmit signals to other artificial neurons like synapses in the brain. Each artificial neuron receives signals, processes the signals, and then sends signals to the artificial neurons connected to it. In addition, each artificial neuron has a threshold, and only when the threshold is reached will it send a signal, as Figure 1 shown as the first neural computing model that emerged, called the McCulloch-Pitts model. Among them, x is the input, which can also be the output of other neurons, and is connected to multiple neurons through weighting; y is the output of the neuron, which can also be used as the input of other neurons; w is the connection weight; θ is the threshold of the neuron; u is the value after aggregating the inputs, equal to each input multiplied by the weight and then subtracting the threshold, ;f(u i ) is called the activation function; y is the output of the neuron, which can also be used as the input of other neurons, .

[0012] In an artificial neural network composed of multiple neurons, the strength of synaptic information transmission is variable, and its transmission effect can be enhanced, weakened, and saturated, that is, it has learning, forgetting, and fatigue effects. Therefore, artificial neural networks have many advantages for data and information processing. For example, they can approximate any non-linear function with arbitrary precision and have the ability to map non-linear functions; they can store information distributively and process information in parallel, with a very fast data processing speed; they can extract regular relationships from input and output data and have generalization ability, with strong self-adaptability and generalization ability. With these characteristics, it is very convenient and efficient to use artificial neural networks to process the data of the influence of various factors on the hardness of gray cast iron brake drums and extract the laws of the influence of various factors on hardness changes.

[0013] To achieve the above object, the technical solution of the present invention is: a method for selecting a formula to improve the hardness of gray cast iron brake drums, characterized by including the following steps: S1. Select gray cast iron brake drum products according to three dimensional parameters of diameter D, height h, and wall thickness t; S2. Select graphitizing elements that can meet the requirements of thermal fatigue performance, strengthening elements that can refine pearlite structure or form hard phases, and determine the content ranges of each element according to practical experience and reference documents; S3. Optimize and select specific 3-5 elements from the elements in S2 according to cost factors; S4. Conduct actual production tests on the products in S1 according to the content of the elements to be added after optimization in S3, and record the data of composition, dimensions, and hardness changes; S5. Sort out the data recorded in S4 and process it using an artificial neural network. The regularity between the output and input obtained after processing is where: D, h, and t are the numerical values of the diameter, height, and average thickness of the brake drum in mm respectively; The actual contents of Si, Ti, and B in the brake drum after adding alloy elements are w(Si), w(Ti), and w(B) respectively; w1 = 100*w(Si), w2 = 100*w(Ti), w3 = 100*w(B); y is the increased value of the hardness of the brake drums of each dimension after adding alloy elements relative to that of general gray cast iron brake drums in HB units.

[0014] In the step S1, the dimension range of the diameter D is: 260-460 mm, the dimension range of the height h is: 150-300 mm, and the dimension range of the wall thickness t is: 8-15 mm.

[0015] In the step S2, the elements selected for addition or adjustment are C, Si, Cr, Ni, Cu, Sn, Mn, B, V, Nb, Mo, Ti and rare earth elements; The content ranges mentioned refer to the appropriate ranges for each element added or adjusted separately to improve the hardness of gray cast iron. The specific mass percentage content ranges are as follows: C: 3.2 - 3.4%, Si: 1.7 - 2.2%, Cr: 0.2 - 0.4%, Ni: 0.5 - 1.0%, Cu: 0.9 - 1.1%, Sn: 0 - 0.03%, Mn: 0.5 - 1.4%, B: 0 - 0.06%, V: 0 - 0.5%, Nb: 0 - 0.1%, Mo: 0 - 0.1%, Ti: 0 - 0.1%, rare earth elements: 0 - 0.03%.

[0016] In the step S3, the cost factors include raw material prices, processing costs, additional working hours and processes, waste gas and waste treatment, etc.; the selected preferred elements are specifically: Si, Ti, B.

[0017] In the step S4, the composition refers to the composition of the gray cast iron brake drum product. The contents of C and Si elements are measured by a carbon-sulfur analyzer from the white-mouth sample taken in the intermediate frequency furnace, and Ti, B and other elements are measured by taking infrared spectrum samples from the product after casting; the dimensions refer to the diameter, height and wall thickness of the gray cast iron brake drum product. The diameter is measured by measuring the maximum outer contour diameter of the brake drum product, the height is measured by the distance from the bottom of the brake drum product to the maximum cross-section, and the wall thickness is measured by measuring 3 wall thicknesses of the rotating part of the brake drum product from bottom to top axially and taking their average value; the hardness change refers to the hardness increase of the test group products added with Ti and B elements or adjusted with the Si element content compared to the control group products of the original process without modification in the products cast in the same furnace. It is obtained by measuring the hardness of the products after casting. For each product, 3 point specimens are taken from the upper part and the side part for Brinell hardness measurement and the average value is taken, and then the hardness value of each test group product is subtracted from the hardness value of the control group to obtain the hardness change.

[0018] In step S5, the processing procedure is as follows: First, the data recorded in S4 is organized into a data set in the form of row vectors, that is, each row vector corresponds to the data of a brake drum workpiece sample. Then, 80% of the data in this data set is randomly selected as training samples, 10% as test samples, and 10% as verification samples. Then, an artificial neural network model is constructed using industrial software such as PyTorch or TensorFlow frameworks based on the Python programming language or Matlab, and the organized data set is used to train it. Among them, the neurons in the input layer each receive the diameter D, height h, wall thickness t, and the contents of Si, Ti, and B elements of the corresponding gray cast iron brake drum product. The neurons in the output layer output the values of the hardness change of the gray cast iron brake drum product under the influence of these factors. The neurons in the hidden layer receive the inputs of the neurons in the input layer. Through the information transmission between the neurons, the regularity between the output and the input is obtained, that is, the brake drum hardness prediction model.

[0019] In step S5, in general gray cast iron brake drums, w(C)=3.4%, w(Si)=1.7%, w(Ti)=0, w(B)=0.

[0020] The present invention provides a quick reference for adding (or adjusting) various elements in production to improve the performance of cast iron parts such as brake drums, and can obtain appropriate component ratios of various elements in a short time, thereby improving performance indicators such as its hardness, without the need to repeatedly conduct experiments to explore. The present invention uses the method of artificial neural networks (ANN) to train the data set of brake drum products obtained from experiments, and obtains the relationship between the addition amounts of different sizes and different alloy elements and the hardness improvement of gray cast iron brake drum products, which is more convenient and efficient compared to other analysis methods. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It is a schematic diagram of the McCulloch-Pitts model.

[0022] Figure 2 It is a schematic diagram of the external structure of a gray cast iron automotive brake drum.

[0023] Figure 3 It is the process of constructing the brake drum hardness prediction model in the present invention.

[0024] Figure 4 It is a schematic diagram of the artificial neural network model in the present invention. EMBODIMENTS

[0025] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments: A method for selecting a formula to improve the hardness of gray cast iron brake drums, characterized by comprising the following steps: S1. Select the product size range of gray cast iron brake drums; The gray cast iron brake drum products of an enterprise have similar compositions and various styles. They are classified by diameter, height, and wall thickness in terms of size. The gray cast iron brake drum is for large trucks, and its structure is as Figure 2 shown, which is a large-diameter cylindrical part similar to a drum shape. The content of each element in its composition is: C: 3.2 - 3.4%, Si: 1.7 - 2.2%, Mn: 0.6 - 1.0%, Cr: 0.3 - 0.5%, Fe: the balance, other elements: less than 0.1%; the size refers to three dimensional parameters of its representative diameter D, height h, and wall thickness t; the size range refers to: diameter: 260 - 460 mm, height: 150 - 300 mm, wall thickness: 8 - 15 mm.

[0026] S2. Select the elements suitable for addition or adjustment and their content ranges; In existing public standards, generally there are no requirements for the chemical composition of gray cast iron or only approximate restrictions are imposed on the contents of C, Si, Mn, S, and P elements in the compositions of different grades of gray cast iron. For example, taking the commonly used HT250 for brake drum products with a wall thickness less than 30 mm as an example, generally it is required that C: 3.2 - 3.4%, Si: 1.7 - 2.0%, Mn: 0.8 - 1.0%, P: less than 0.2%, S: less than 0.12%; therefore, elements such as P and S that are usually harmful to the performance of gray cast iron are excluded, and then graphitizing elements such as C, Si, Cu, and Ni are selected in combination with the high requirements for thermal fatigue resistance of brake drum products during operation. Finally, strengthening elements such as Sn, Mn, B, V, Nb, Mo, Ti, and rare earth elements that can refine the pearlite structure or form hard phases are selected, and the elements for addition or adjustment obtained are C, Si, Cr, Ni, Cu, Sn, Mn, B, V, Nb, Mo, Ti, and rare earth elements; then the appropriate addition content of each of the above elements is determined by referring to manuals and literature, and in combination with the experience accumulated in production practice and the composition and performance of existing brake drum products as a reference, the obtained content range refers to the appropriate range of addition or adjustment of each element alone for improving the hardness of gray cast iron, and the specific mass percentage content range is: C: 3.2 - 3.4%, Si: 1.7 - 2.2%, Cr: 0.2 - 0.4%, Ni: 0.5 - 1.0%, Cu: 0.9 - 1.1%, Sn: 0 - 0.03%, Mn: 0.5 - 1.4%, B: 0 - 0.06%, V: 0 - 0.5%, Nb: 0 - 0.1%, Mo: 0 - 0.1%, Ti: 0 - 0.1%, rare earth elements: 0 - 0.03%.

[0027] S3. Select specifically 3 to 5 elements from the elements in S2 according to cost factors; The cost factors mentioned above include raw material prices, processing costs, additional working hours and processes, waste gas and waste treatment, etc. Since low cost is the main advantage of gray cast iron compared with other materials for brake drum manufacturing, it is necessary to pay attention to the cost factors in the production process. From the perspective of raw materials, the prices of ferrovanadium alloy and ferroniobium alloy used in smelting are much higher than those of ferro titanium alloy and ferroboron alloy. Adding the same mass of elements such as V and Nb is 2 - 3 times the cost of adding Ti and B elements. The existing Mn and Cr elements in the current gray cast iron brake drum are introduced from pig iron and scrap steel in the raw materials, and their contents are within a reasonable range. Combining with the existing process flow, the specific selected preferred elements are: Si, Ti, B.

[0028] S4. Conduct actual production tests on the product in S1 according to the element contents to be added after optimization in S3, and record the data of composition, dimensions, and hardness changes; The composition mentioned above refers to the composition of the gray cast iron brake drum product. The contents of C and Si elements are measured by a carbon - sulfur analyzer from white - mouth samples taken in the intermediate - frequency furnace, and Ti, B, and other elements are measured by taking infrared spectrum samples from the product after casting. The dimensions refer to the diameter, height, and wall thickness of the gray cast iron brake drum product. The diameter is measured by measuring the maximum outer - contour diameter of the brake drum product, the height is measured by the distance from the bottom of the brake drum product to the maximum cross - section, and the wall thickness is measured by measuring 3 wall thicknesses of the rotating part of the brake drum product from bottom to top axially and taking their average value. The hardness change refers to the hardness increase of the test - group products with added Ti and B elements or adjusted Si element content compared to the control - group products with the original unmodified process among the products cast in the same furnace. It is obtained by measuring the hardness of the product after casting. For each product, 3 - point specimens are taken from the upper and side parts for Brinell hardness measurement and the average value is taken, and then the hardness value of each test - group product is subtracted from the hardness value of the control group to obtain the hardness change value.

[0029] S5. Organize the data recorded in S4 and process it using an artificial neural network; The specific processing process is as Figure 3 shown. First, organize the data recorded in S4 into a data set in the form of row vectors, that is, each row vector corresponds to the data of a brake - drum workpiece specimen. Then randomly divide 80% of the data in this data set as training samples, 10% as test samples, and 10% as verification samples. Then use industrial software such as PyTorch or TensorFlow framework based on the Python programming language or Matlab to construct a structure as Figure 4The artificial neural network model shown is trained using the organized dataset. Each neuron in the input layer receives the diameter D, height h, wall thickness t of the gray cast iron brake drum product, and the contents of Si, Ti, and B elements. The neuron in the output layer outputs the value of the hardness change of the gray cast iron brake drum product under the influence of these factors. The neurons in the hidden layer receive the inputs from the neurons in the input layer. Through the information transmission between neurons, the regularity between the output and the input is obtained, that is, the brake drum hardness prediction model. At this time, only a set of information containing the brake drum size and the content of added elements needs to be submitted as input to this model, and the corresponding hardness prediction will be generated as output. The processing process of the entire model is opaque to the outside, equivalent to a black box. Therefore, in order to make the prediction model more convenient to use, mathematical methods are used to simplify and fit the implicit relationship of the model, and an explicit mathematical expression that ensures high accuracy and is not too complex is obtained. The algebraic relationship expression between the obtained output and input (hereinafter referred to as the empirical formula) is as follows: Where: , D, h, and t are the numerical values of the diameter, height, and average thickness of the brake drum in mm respectively; The actual contents of Si, Ti, and B in the brake drum after adding alloy elements are w(Si), w(Ti), and w(B) respectively; w1 = 100 * w(Si), w2 = 100 * w(Ti), w3 = 100 * w(B); y is the value of the hardness increase of the brake drum of each size after adding alloy elements when = 3.40%, = 1.70%, = 0, = 0 in HB units.

[0030] The empirical formula is generally used in the preliminary screening stage before actual experiments to judge the approximate appropriate range, reduce the number of experiments, and thus reduce the labor and material costs. The following are the specific method steps for applying the empirical formula: S1. Determine the current size of the gray cast iron brake drum product: Measure the size of the gray cast iron brake drum product, including the diameter, height, and wall thickness. The diameter D is taken as the maximum outer diameter of 260 mm of the brake drum product; the height h is taken as the distance from the bottom of the brake drum product to the maximum cross-section of 260 mm; the wall thickness t is measured at 3 places of the rotating part of the brake drum product from bottom to top axially, which are 9.2 mm, 9.5 mm, and 9.5 mm, and the average value of 9.4 mm is taken.

[0031] S2. Determine the composition of the current gray cast iron product: During the production process, control the C content in the furnace to be about 3.40% and the Si content to be about 1.7%. Randomly take 3 - 5 workpieces cast from one ladle of molten iron. Here, 3 workpieces are selected. Cut a 15*15*10 mm sample from the bottom of each of them using a cutting machine. After grinding and polishing, make an infrared spectroscopy specimen, and conduct composition detection on it. The average content of components other than C and Si is as follows: Mn 0.80%, P 0.05%, S 0.08%, Cr 0.30%, Ni 0.02%, Cu 0.07%, V 0.03%, and the balance is Fe and trace components and impurities less than 0.01%.

[0032] S3. Measure the hardness of the current gray cast iron product: Select 3 points on the upper and side parts of the 3 workpieces selected in S2 respectively, use a Brinell hardness tester to measure their hardness and take the average value. The average hardness of the three workpieces is as follows: 172HB, 165HB, 173HB.

[0033] S4. According to the required hardness of the product, calculate the types and contents of alloying elements to be added according to the empirical formula obtained in the present invention: Substitute the diameter D = 260mm, height h = 260mm, wall thickness t = 9.4mm in S1 and w(Si)=1.7% in S2 into the above empirical formula, and also substitute the value of the expected increase in product hardness y = 20HB into the empirical formula, and the relationship between w(Ti) and w(B) to be added can be obtained. From experience, the addition amount of w(Ti) should be less than 0.1%, and the addition amount of w(B) should be less than 0.06%. At the same time, considering cost and process, as much Ti element as possible should be added. Therefore, here w(Ti)=0.1%, and substituting it into the formula can obtain w(B)=0.011%.

[0034] S5. Add alloying elements according to the results calculated in S4 for production tests and measure the hardness of the samples: Select 3 positions on the upper and lower parts of the sample respectively, use a Brinell hardness tester to measure the hardness and take the average value. The average hardness of the samples is 190HB, 188HB, 195HB respectively, and the increased hardness values are 18HB, 23HB, 22HB respectively. The errors from the results calculated by the above empirical relationship formula are 10%, 15%, 10% respectively.

Claims

1. A method for selecting a formula to improve the hardness of gray cast iron brake drums, characterized in that, It includes the following steps: S1. Select gray iron brake drum products according to three dimensional parameters of diameter D, height h, and wall thickness t; S2. Select graphitizing elements that can meet the requirements of thermal fatigue performance, strengthening elements that can refine pearlite structure or form hard phases, and determine the content ranges of various elements according to practical experience and references; S3. Optimize and select specific 3 - 5 elements from the elements in S2 according to cost factors; S4. Conduct actual production tests on the products in S1 according to the element contents to be added after optimization in S3, and record the data of composition, dimension, and hardness changes; S5. Organize the data recorded in S4 and process it using an artificial neural network. The regularity between the output and input after processing is Wherein: , when D, h, and t are in millimeters, they are the values of the diameter, height, and average thickness of the brake drum respectively; The actual contents of Si, Ti, and B in the brake drum after adding alloy elements are w(Si), w(Ti), and w(B) respectively; w1 = 100 * w(Si), w2 = 100 * w(Ti), w3 = 100 * w(B); y is the increased value of the hardness of brake drums with various dimensions after adding alloy elements relative to that of general gray iron brake drums in units of HB.

2. The method for selecting a formula to improve the hardness of gray cast iron brake drums according to claim 1 above, characterized in that: In the step S1, the dimension range of the diameter D is: 260 - 460 mm, the dimension range of the height h is: 150 - 300 mm, and the dimension range of the wall thickness t is: 8 - 15 mm.

3. The method for selecting a formula to improve the hardness of gray cast iron brake drums according to claim 1 above, characterized in that: In the step S2, the selected elements to be added or adjusted are C, Si, Cr, Ni, Cu, Sn, Mn, B, V, Nb, Mo, Ti, and rare earth elements; The content range refers to the appropriate range of addition or adjustment of each element alone for improving the hardness of gray iron. The specific mass percentage content range is: C: 3.2 - 3.4%, Si: 1.7 - 2.2%, Cr: 0.2 - 0.4%, Ni: 0.5 - 1.0%, Cu: 0.9 - 1.1%, Sn: 0 - 0.03%, Mn: 0.5 - 1.4%, B: 0 - 0.06%, V: 0 - 0.5%, Nb: 0 - 0.1%, Mo: 0 - 0.1%, Ti: 0 - 0.1%, rare earth elements: 0 - 0.03%.

4. The method for selecting a formula to improve the hardness of gray cast iron brake drums according to claim 1 above, characterized in that: In the step S3, the cost factors include raw material prices, processing costs, additional working hours and processes, waste gas and waste treatment, etc.; the selected preferred elements are specifically: Si, Ti, B.

5. The method for selecting a formula to improve the hardness of gray cast iron brake drums according to claim 1 above, characterized in that: In step S4, the components refer to the components of the gray cast iron brake drum product. The contents of C and Si elements are measured by a carbon-sulfur analyzer from a white iron sample taken from an intermediate frequency furnace, and Ti, B, and other elements are measured by taking an infrared spectrum sample from the product after casting. The dimensions refer to the diameter, height, and wall thickness of the gray cast iron brake drum product. The diameter is measured by measuring the maximum outer contour diameter of the brake drum product, the height is measured by the distance from the bottom of the brake drum product to the maximum cross-section, and the wall thickness is measured by successively measuring 3 wall thicknesses of the rotating part of the brake drum product from bottom to top along the axis and taking their average value. The hardness change refers to the hardness increase of the test group products with added Ti and B elements or adjusted Si element content compared to the control group products with the original process unchanged in the products cast in the same furnace. It is obtained by measuring the hardness of the products after casting. For each product, 3 point specimens on the upper and side parts are taken for Brinell hardness measurement and the average value is taken. Then, the hardness value of each test group product is subtracted from the hardness value of the control group to obtain the hardness change.

6. The method for selecting a formula to improve the hardness of gray cast iron brake drums according to claim 1 above, characterized in that, In step S5, the processing procedure is as follows: First, the data recorded in S4 is organized into a data set in the form of a row vector, that is, each row vector corresponds to the data of a brake drum workpiece specimen. Then, 80% of the data in this data set is randomly selected as the training sample, 10% as the test sample, and 10% as the verification sample. Then, an artificial neural network model is constructed using industrial software such as PyTorch or TensorFlow framework based on the Python programming language or Matlab, and the organized data set is used to train it. The neurons in the input layer each receive the corresponding diameter D, height h, wall thickness t of the gray cast iron brake drum product, and the contents of Si, Ti, and B elements. The neurons in the output layer output the value of the hardness change of the gray cast iron brake drum product under the influence of these factors. The neurons in the hidden layer receive the inputs from the neurons in the input layer. Through the information transfer between neurons, the regularity between the output and the input is obtained, that is, the brake drum hardness prediction model.

7. The method for selecting a formula to improve the hardness of gray cast iron brake drums according to claim 1 above, characterized in that: In step S5, in general gray cast iron brake drums, w(C) = 3.4%, w(Si) = 1.7%, w(Ti) = 0, w(B) = 0.

Images