Lightweight casting forming and control process for thin-walled castings

Abstract

The present application relates to the technical field of casting quality control, and particularly relates to a thin-wall casting lightweight casting forming and control process, which comprises the following steps: after the smelting of molten iron is completed, composite rare earth passivation modification is performed, and the impurity content is limited in a preset metallurgical control interval; an initial forming control data set is constructed, and a global differentiated forming control data set suitable for the special-shaped structure of the thin-wall casting is formed; during the filling process, time sequence differentiated stream inoculation is performed, and the filling parameters are monitored and the dosing time sequence is corrected; after the casting enters the eutectic solidification stage, the sand mold cooling air volume is adjusted and fed back according to the time length deviation and the rate dispersion degree; the micro-positive pressure in the cavity is maintained according to the cavity pressure monitoring value, and the secondary migration and local enrichment of the low-melting-point solute at the solid-liquid interface are inhibited; after the casting is completely solidified, the temperature decay rate is collected throughout the process, and the cooling intensity is adjusted, so that the formed thin-wall casting is obtained. The process significantly improves the pressure resistance and pressure maintaining performance of the hydraulic valve body and the reliability of continuous full-load operation.

Description

Technical Field

[0001] This invention relates to the field of casting quality control technology, and in particular to a lightweight casting forming and control process for thin-walled castings. Background Technology

[0002] Thin-walled valve body castings made of gray cast iron and ductile iron used in hydraulic systems are common parts in the fields of engineering machinery and agricultural machinery. They mainly include conventional structural parts such as multi-way directional valve intermediate bodies, gear pump bodies, hydraulic steering gear end caps, shuttle valve bodies, and relief valve bodies. The control of their production quality mainly relies on the casting process.

[0003] Currently, these thin-walled valve body castings are generally produced using a coated sand gravity casting process. By controlling macroscopic process parameters such as pouring temperature, filling flow rate, and sand mold cooling conditions, the casting quality can be controlled. Commonly used parts produced by the above production method can basically meet the needs of conventional equipment operation and can meet the daily short-term and intermittent farming needs of agricultural machinery.

[0004] However, under harsh working conditions such as deep tillage of contiguous land by micro-tillers and continuous operation of straw return equipment, the hydraulic system needs to supply oil continuously for a long time, frequently change direction, and work under continuous load. The oil pressure is in a state of high-frequency cyclic switching and long-term continuous full load. Thin-walled castings produced by the existing casting process system often have the following hidden dangers:

[0005] On the one hand, existing solidification theory clearly states that during the eutectic solidification of gray cast iron, a solid-liquid interface temperature exists at the boundary between the solid and liquid phases. This temperature differs from both the bulk temperature of the molten iron and the thermodynamic equilibrium eutectic temperature; rather, it is a physical quantity that dynamically changes with the solidification process, falling between the two. However, in actual production, existing casting processes generally rely on overall molten iron temperature data or the equilibrium eutectic temperature as the basis for process control. This leads to discrepancies between the estimated grain growth rate and the actual solidification front state, making it difficult to achieve precise control over the eutectic solidification process.

[0006] In addition, under the existing extensive temperature control mode, the slight undercooling fluctuation in the thin-walled area will cause the global grain growth rate to become discrete, forming a directional metallographic texture. Under the continuous action of high-frequency alternating oil pressure, the metallographic anisotropy will be transformed into micro-stress difference, inducing intergranular microcracks and hidden micro-leakage, which can easily lead to excessive pressure decay under precision pressure holding conditions, resulting in common faults such as hydraulic internal leakage, unstable pressure holding, and machine power decay.

[0007] On the other hand, the valve body has its own oil circuit junction heat joint and thick and thin transition irregular structure, which will cause the heat dissipation thermal resistance of different structural areas to be inherently different. In conventional process forming, it will cause the eutectic phase transformation start-up sequence of each area to be out of sync. Low melting point impurity elements in molten iron will inevitably continue to accumulate and remain at the thick and thin transition interface. Under the long-term continuous full-load operation of agricultural machinery, the area with accumulated impurities will continuously induce micro shrinkage porosity and intergranular peeling fatigue defects, directly reducing the durability and service life of the valve body.

[0008] Therefore, it is necessary to optimize the stability of standardized thin-walled hydraulic valve body castings under harsh operating conditions such as high-frequency alternation, precision pressure holding, and long-term durability, in order to better improve the product quality and service stability of standardized thin-walled valve body castings. Summary of the Invention

[0009] To solve one of the above-mentioned technical problems, the present invention adopts the following technical solution: a lightweight casting forming and control process for thin-walled castings, comprising the following steps:

[0010] S1. After the molten iron is smelted, the molten iron is subjected to composite rare earth passivation modification treatment. The content of low melting point impurity elements in the molten iron is obtained in real time by pre-furnace spectral detection. The solute distribution is adjusted according to the detection results to limit the content of low melting point impurities in the molten iron to within the preset metallurgical control range.

[0011] S2. Receive the processed molten iron and perform pouring. Before pouring, collect the no-load heat dissipation temperature parameters of different structural areas of the casting corresponding to the mold, predict the solidification undercooling of each area and build an initial forming control dataset. After pouring, collect the actual temperature parameters, calculate the actual undercooling and correct the dataset to form a global differentiated forming control dataset that is adapted to the irregular structure of thin-walled castings.

[0012] S3. During the filling process, the initial molding control dataset is called, and the corresponding inoculant addition parameters are matched according to the predicted supercooling of each region. The time-differentiated inoculant process is executed, and the filling process parameters are monitored in real time and the inoculant addition sequence is corrected accordingly.

[0013] S4. When the casting enters the eutectic solidification stage, the duration of eutectic solidification is locked based on real-time monitoring data of cavity temperature. At the same time, the global grain growth rate is calculated and its discrete parameters are statistically analyzed. The sand mold cooling air volume is adjusted according to the time deviation and the degree of grain rate dispersion.

[0014] S5. When the casting enters the final stage of solid-liquid coexistence molding, close the cavity venting channel to form a closed environment inside. Based on the real-time monitoring results of the cavity pressure, maintain a slight positive pressure inside the cavity to suppress the secondary migration and local enrichment of low-melting-point solutes at the solid-liquid interface.

[0015] S6. After the casting has solidified, a zoned gradient slow-release cooling process is immediately implemented. The temperature decay rate of each area is collected in real time throughout the cooling process, and the cooling intensity is adjusted accordingly to finally obtain a thin-walled casting.

[0016] Preferably, step S1 is performed according to the following steps:

[0017] a1. Prepare a composite passivating agent by mixing rare earth ferrosilicon alloy and silicon-calcium alloy at a mass ratio of 3:1, wherein the particle size of the passivating agent is controlled at 1mm to 5mm, and dry it at 200℃ to 300℃ for 1h to 2h.

[0018] a2. Before tapping the molten iron, an oxygen probe is used to detect the oxygen activity of the molten iron, and the oxygen activity is controlled at 2×10⁻⁶. -6 ~8×10 -6 Within the specified range, when the oxygen activity exceeds the specified range, the oxygen activity is adjusted by adding a passivating agent.

[0019] Specifically, when the oxygen activity is high, if the sulfur content is below the lower limit of the control range, silicon-calcium alloy should be added first; if the sulfur content is above the upper limit of the control range, rare earth silicon-iron alloy should be added first; if the sulfur content is within the control range, the composite passivating agent prepared in step a1 should be added.

[0020] a3. Simultaneously, a furnace-front laser-induced breakdown spectrometer is used to detect the mass fraction of sulfur and lead low-melting-point impurities in molten iron. Each furnace is tested no less than twice. For critical batches, an online continuous monitoring mode is used with a monitoring interval of 20s to 40s.

[0021] The sulfur mass fraction is controlled at 0.05% to 0.08%, and the lead mass fraction is controlled at 0.002% to 0.005%. When the impurity content exceeds the upper limit of the control range, the composite passivating agent prepared in step a1 is added at 0.05% to 0.10% of the molten iron mass, and the amount added at one time shall not exceed 0.15% of the molten iron mass, until the impurity content falls back to the control range.

[0022] Preferably, step S2 is performed according to the following steps:

[0023] b1. K-type armored thermocouples are pre-embedded in the inner wall of the mold cavity corresponding to the hot spot area, thin-walled plane area and thick-thin transition area of ​​the casting. The distance between the thermocouple measuring point and the surface of the mold cavity is controlled between 2mm and 5mm, the response time is not greater than 0.5s, and the temperature measurement range meets the short-term temperature measurement requirements of 0℃ to 1600℃.

[0024] b2. Before casting, a blank temperature measurement process is performed to collect the no-load cooling curves of each area. Considering the non-linear decay of the thermal diffusivity of the coated sand with increasing temperature, different thermal diffusivity correction values ​​are used to correct the cooling rate in segments according to three temperature ranges: below 200℃, 200℃~600℃, and above 600℃. A correction model for the heat dissipation characteristics of the sand mold in each area is established. Based on this, the solidification undercooling and phase change initiation time of each area are predicted, and an initial molding control dataset is constructed.

[0025] After casting, temperature curves of each region are collected. Based on the heat dissipation characteristics, the model is modified to back-derive the molten iron cooling curve. The difference between the equilibrium eutectic temperature and the actual eutectic solidification initiation temperature is extracted to calculate the actual solidification undercooling. The calculated undercooling value ranges from 5K to 40K. The initial forming control dataset is corrected based on the actual undercooling.

[0026] The revised dataset is used for both the control of subsequent cooling processes in this batch and the iteration of the forming control dataset for the next batch.

[0027] b3. Subsequently, the castings are divided into three levels according to the supercooling value: high supercooling zone (25K~40K), medium supercooling zone (15K~25K), and low supercooling zone (5K~15K). The inoculant dosing acceleration rate range and cooling air volume range are set for each zone respectively, and finally a global differentiated molding control dataset is formed.

[0028] Preferably, step S3 is performed as follows:

[0029] c1. First, call the initial forming control dataset and match the molten iron mass flow rate of 0.20%~0.30%, 0.10%~0.20%, and 0.05%~0.10% of the 75% ferrosilicon inoculant addition acceleration rate for the high subcooling zone, medium subcooling zone, and low subcooling zone, respectively, with the inoculant particle size controlled between 0.2mm and 1.0mm.

[0030] c2. Because the apparent viscosity of molten iron is approximately 2 mPa·s to 6 mPa·s within the filling temperature range, and after the primary crystalline phase precipitates as the filling temperature decreases, it exhibits non-Newtonian fluid characteristics where the apparent viscosity decreases with increasing shear rate; in the turbulent filling region, the shear rate can reach 100 s⁻¹. -1 The above factors lead to a decrease in the conformability of probiotic particles;

[0031] The inoculant addition acceleration rate is adjusted in real time based on the flow velocity of the molten iron stream. Specifically, for every 0.1 m / s increase in flow velocity, the inoculant addition acceleration rate is increased by 1% to 3%.

[0032] c3. At the same time, based on the arrival time of molten iron in each area, the inoculant addition start time in the corresponding area is advanced by 0.5s to 1.0s to ensure that the inoculant and molten iron enter each area of ​​the mold cavity synchronously.

[0033] c4. After the filling process is completed, the inflection point of the eutectic phase transition initiation temperature is collected in real time by pre-embedded thermocouples in each area, and the actual phase transition initiation time of each area is calculated.

[0034] When the actual phase change start time of a certain area is found to deviate from the predicted time in the dataset by more than 5 seconds, it is determined that there is a deviation in the inoculation control parameters. The deviation data is recorded, and the inoculation agent dosing rate is increased or decreased according to the direction of the deviation. The control parameters of the corresponding area in the production database are corrected and used for the iteration of the molding control dataset for the next batch.

[0035] In addition, when the actual phase change start time deviation between adjacent regions is greater than 5s, the solidification uniformity deviation data of the whole region is recorded and stored in the production database as the basis for the iterative optimization of the next batch dataset.

[0036] Preferably, step S4 is performed as follows:

[0037] d1. The temperature curve is collected in real time by pre-embedded thermocouples in the cavity. The duration of the eutectic reglow temperature plateau is locked between 20s and 60s. When the monitoring duration is less than 20s, the cooling air volume of the sand mold surface is reduced, and when it is more than 60s, the cooling air volume is increased.

[0038] d2. Simultaneously, control the cooling rate during the eutectic stage to remain within the range of 0.5℃ / s to 2.0℃ / s;

[0039] d3. The grain growth rate is calculated using a thin-walled valve body eutectic solidification grain growth rate prediction model. The interface radius of curvature is estimated based on the current cooling rate and undercooling value. The larger the cooling rate and the larger the undercooling, the smaller the radius of curvature. The estimated value range is 1μm to 100μm. When the estimated value exceeds this range, the boundary value is substituted into the calculation.

[0040] ;

[0041] In the model, temperature determines the growth rate, curvature determines the effective undercooling, and the two together determine the grain growth rate.

[0042] In the formula, This is the calculated value of grain growth rate, in μm / s. Within the range of process parameters, the calculated value is generally 5μm / s to 40μm / s, and it is used to statistically analyze the relative discrete parameters of each point in the entire domain.

[0043] Frequency factor, unit μm / The value range is 1.0 × 10 4 μm / ~1.5×10 4 μm / The range of solidification kinetics data for gray cast iron is determined by referring to the data and combining it with on-site process experience as needed, and can be finely adjusted by those skilled in the art as needed.

[0044] The solidification activation energy is expressed in J / mol and ranges from 1.0 × 10⁻⁶. 5 J / mol ~ 1.5 × 10 5 J / mol, which is determined by referring to the data range of activation energy for eutectic phase transformation of iron-carbon alloys and combining on-site process experience as needed, and can be finely adjusted by those skilled in the art as needed;

[0045] The gas constant is expressed in J / L. Take 8.314 J / , which are internationally recognized physical constants;

[0046] This is the solid-liquid interface temperature, in K, with a value range of 1390K to 1425K.

[0047] This represents the actual undercooling during eutectic solidification, expressed in K, with a value ranging from 5K to 40K.

[0048] The Gibbs-Thomson coefficient, in K.μm, ranges from 0.1 K.μm to 0.5 K.μm. It is determined based on the solidification interface energy and latent heat of fusion parameters of iron-carbon alloys. The specific calculation formula is as follows: ;

[0049] in The solid-liquid interface energy of iron-carbon alloys. The equilibrium eutectic temperature for gray cast iron. The latent heat of fusion per unit volume;

[0050] The radius of curvature of the solidification interface is expressed in μm and ranges from 1 μm to 100 μm.

[0051] Based on the calculation results, the relative dispersion parameter of grain growth rate at each point in the entire region is statistically analyzed. When the relative deviation of the grain growth rate at each point from the average value of the entire region exceeds +15%, the sand mold purging air volume in the corresponding area is reduced by 0.1 m³. 3 / min~0.5m 3 / min;

[0052] Conversely, when the relative deviation between the grain growth rate at each point and the global average exceeds -15%, the sand mold purging airflow in the corresponding area is increased by 0.1 m³. 3 / min~0.5m 3 / min, so that the deviation of the global grain growth rate converges to within ±15% of the average rate.

[0053] Preferably, step S5 is performed according to the following steps:

[0054] e1. Lay a high-temperature resistant aluminum silicate sealing gasket on the sand mold parting surface, wherein the thickness of the sealing gasket is controlled at 2mm to 4mm and the compression is controlled at 30% to 50% to ensure that the sealing gap of the parting surface is not greater than 0.1mm.

[0055] e2. When the cavity temperature drops to 5℃~20℃ below the equilibrium eutectic temperature, it is determined that the casting has entered the final stage of solidification with solid-liquid coexistence, and all exhaust channels are closed to form a sealed environment inside the cavity.

[0056] e3. The cavity pressure is monitored in real time by the cavity pressure sensor, and the cavity micro-positive pressure is maintained at 5kPa to 15kPa. This pressure is dynamically balanced by the inert gas pressure replenishment, residual gas generation of the coated sand and cavity venting.

[0057] When the pressure is below 5 kPa, the pressure is replenished through the inert gas inlet valve; conversely, when the pressure is above 15 kPa, the pressure is released by opening the micro exhaust valve.

[0058] e4. The sealed pressure maintenance time should be controlled between 8 and 15 minutes. During this period, it is forbidden to open the sand mold or disturb the casting.

[0059] Preferably, step S6 is performed according to the following steps:

[0060] f1. First, perform gradient slow-release cooling according to the global subcooling benchmark zone calibrated in step S2. In the high subcooling zone, heat preservation slow cooling is adopted, and the cooling rate is controlled at 0.05℃ / s~0.1℃ / s.

[0061] The intermediate subcooling zone uses weak airflow cooling at a rate of 0.1℃ / s to 0.3℃ / s;

[0062] The low subcooling zone uses forced air cooling at a rate of 0.3℃ / s to 0.8℃ / s;

[0063] f2. Throughout the cooling process, temperature decay data is collected in real time by pre-embedded thermocouples in each area. Considering the non-linear increase of the elastic modulus of gray cast iron as the temperature decreases, different temperature coefficients of elastic modulus are used for three temperature ranges: below 200℃, 200℃~600℃, and above 600℃. The thermal stress caused by the temperature difference between adjacent areas is calculated in combination with the thermal expansion coefficient of gray cast iron. When the temperature difference between adjacent areas is greater than 80℃, the cooling process of the side with higher cooling intensity in the area with the temperature difference exceeding the limit is suspended. The cooling process of that side is resumed after the temperature difference drops to within 50℃.

[0064] f3. When the overall temperature of the casting drops below 200℃, the zoned cooling control is released and the casting is switched to air cooling to room temperature. The temperature of the casting after unpacking should not exceed 150℃.

[0065] Preferably, after step S6 is completed, a casting solidification stability evaluation process is also included, which is performed according to the following steps:

[0066] g1. Summarize four monitoring parameters collected throughout the molding process: undercooling deviation, maximum absolute dispersion deviation of grain growth rate, local enrichment concentration increment of low melting point impurities, and deviation of phase transformation initiation time.

[0067] g2. Subsequently, the solidification stability was calculated using the geometric mean evaluation model for solidification stability:

[0068] ;

[0069] In the formula, This is a dimensionless index for evaluating solidification stability, with a value ranging from 0 to 1.

[0070] , , , The sensitivity coefficients for each defect factor are dimensionless and range from 0.65 to 0.75. They are determined by referring to literature data on the influence of each defect factor on the service performance of gray cast iron castings and by combining field process tests.

[0071] The maximum deviation of supercooling across the entire domain, in K, with a value range of 0K to 50K;

[0072] This is a reference deviation for undercooling, in K, taken as 50K, determined based on the maximum allowable deviation for undercooling of gray cast iron;

[0073] The maximum absolute dispersion deviation of grain growth rate is expressed in μm / s, with a range of 0 μm / s to 20 μm / s.

[0074] This is the reference deviation for grain rate, in μm / s, and is set to 20 μm / s. It is determined based on the allowable dispersion range of grain growth rate in gray cast iron.

[0075] The concentration increment of low-melting-point impurities in local enrichment, in ppm, with a value range of 0ppm to 200ppm, refers to the segregation increment of impurity concentration at grain boundaries and in the last solidified region relative to the average concentration of the matrix at the end of solidification.

[0076] The reference concentration for impurities is in ppm, taken as 200 ppm. The concentration should be determined as needed based on the allowable increase in the segregation of low-melting-point impurities in gray cast iron.

[0077] The maximum deviation at the start of the phase transition, in seconds, with a range of 0s to 30s;

[0078] This is a timing reference deviation, in seconds (s), taken as 30 seconds, determined based on the allowable deviation range of the phase transformation initiation time for thin-walled castings.

[0079] The above reference values ​​can be calibrated and adjusted before production by combining process tests with blank temperature measurement results, based on the casting wall thickness and structural characteristics.

[0080] When any monitoring parameter exceeds the corresponding reference value, the solidification stability is directly judged as unqualified, and the formula calculation is no longer performed.

[0081] g3. Based on the calculation results, when It is judged as Level 1 qualified, when It was determined to be Level 2 qualified. The product is then deemed defective and transferred to the remelting process.

[0082] g4. After associating the evaluation results with the casting batch number, store them in the production database to form a traceable record of solidification quality parameters for subsequent iterative updates of the molding control dataset.

[0083] Preferably, the feedback adjustment of the sand mold cooling air volume in step S4 is executed using solidification synergistic deviation convergence control logic, and is performed according to the following steps:

[0084] h1. Real-time acquisition of three parameters: maximum deviation of global undercooling, maximum absolute discrete deviation of grain growth rate, and maximum deviation of phase transformation initiation time. These parameters are divided by their respective calibration reference values ​​for normalization, and then weighted and summed according to preset weights (the sum of the three weights is normalized to 1). The square root is then taken to obtain the current solidification coordination deviation index. The deviation index ranges from 0 to 1. The closer it is to 0, the more uniform the solidification state.

[0085] The preset weight values ​​are all in the range of 0.2 to 0.5, determined according to the degree of influence of each deviation factor on the quality of the casting, and calibrated by process test combined with blank temperature measurement results before the equipment is put into production;

[0086] The calibration reference values ​​are determined based on the upper limit of the allowable deviation of each parameter of gray cast iron, specifically: supercooling calibration reference 40K~50K, grain rate calibration reference 15μm / s~20μm / s, and phase transformation initiation time calibration reference 25s~30s.

[0087] h2. Calculate the adaptive control gain based on the low-melting-point impurity content of the molten iron detected at the furnace. The higher the impurity content, the greater the predicted tendency for agglomeration at the end of solidification, and the greater the gain. The gain value range is 0.1m. 3 / min~0.5m 3 / min, determined based on the thermal inertia of the sand mold and the air volume response lag time;

[0088] By combining the solidification synergy deviation index and the deviation change rate, and superimposing the differential advance adjustment amount, where the differential action time coefficient is taken from 0.5s to 2.0s and determined according to the cooling system response time constant, the airflow adjustment increment is calculated, with the adjustment increment value ranging from -1.0m. 3 / min~1.0m 3 / min;

[0089] In addition, for the structural thermal section region, 0m needs to be superimposed. 3 / min~0.3m 3 The feedforward compensation amount is preset based on the hot spot volume and structural characteristics of the casting;

[0090] h3. After superimposing the incremental air volume adjustment with the reference air volume, the output is constrained by the upper and lower limits and the single-step change rate and sent to the variable frequency fan to adjust the sand mold surface purging and cooling intensity.

[0091] The baseline air volume is 0.5m³. 3 / min~1.5m 3 / min, air volume output range 0m 3 / min~3.0m 3 / min, with the maximum rate of change per step controlled at 0.1m. 3 / ~0.3m 3 / ;

[0092] h4. Repeat steps h1 to h3 every 1 to 5 seconds until the casting enters the final stage of solidification with solid-liquid coexistence, then stop the control logic operation and switch to the closed pressure stabilization control in step S5.

[0093] Preferably, each step of the process is executed sequentially by an industrial computer according to a preset control logic. The computer has a built-in global differentiated forming control dataset constructed in step S2, and receives data on impurity content, cavity temperature, cavity pressure, and molten iron flow rate collected by each sensor in real time. The data acquisition and control command update is performed every 1 to 5 seconds.

[0094] A graded alarm is triggered when the monitored data exceeds the preset control range. The deviation ratio is calculated based on the width of the preset control range corresponding to the parameter.

[0095] A Level 1 alarm occurs when the parameter deviation exceeds the allowable range but does not exceed 10%. The computer will automatically fine-tune the corresponding process parameters and record the adjustment log.

[0096] A level 2 alarm occurs when the parameter deviation exceeds the allowable range by more than 10% but not more than 30%. The computer will pause the current process and display a manual confirmation interface, waiting for the operator to review the information before deciding whether to continue or terminate the process.

[0097] A Level 3 alarm occurs when the parameter deviation exceeds the allowable range by more than 30%. The computer terminates the pouring process and initiates the emergency remelting procedure, while simultaneously sending an alarm message to the management personnel.

[0098] In addition, process data for each process is automatically archived according to the casting batch number, with a retention period of no less than 3 years, supporting subsequent quality traceability and iterative optimization of the forming control dataset.

[0099] Preferably, the thin-walled casting is a thin-walled cast iron casting of a hydraulic valve body, including any one of a multi-way valve housing, a hydraulic steering gear housing, a gear pump housing, a plunger pump housing, and a hydraulic motor housing;

[0100] The dimensions of the casting blank, the layout of the oil circuit, and the forming mold all adopt the existing fixed design, and the casting structure and mold cavity are not changed during the casting process.

[0101] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0102] 1. This invention constructs a global differentiated forming control dataset by blank temperature measurement modeling and post-pouring correction. Combined with time-series differentiated in-flow incubation and eutectic solidification stage grain growth rate discrete parameter feedback adjustment of sand mold cooling air volume, the global grain growth rate deviation of the casting is brought to within ±15% of the average rate. This effectively eliminates the grain growth rate discreteness and directional metallographic texture caused by heat dissipation differences in thin-walled irregular structures, avoids the generation of grain boundary microcracks under high-frequency alternating hydraulic pressure, and significantly improves the pressure resistance and pressure holding performance of hydraulic valve body and the reliability of continuous full-load operation.

[0103] 2. This invention controls the content of low-melting-point impurities in molten iron within a preset metallurgical range by combining composite rare earth passivation modification treatment with real-time pre-furnace spectral detection. Furthermore, by maintaining a slight positive pressure within the cavity during the final stage of solidification with solid-liquid coexistence, the exhaust channel is closed to suppress the secondary migration and local enrichment of low-melting-point solutes at the solid-liquid interface. This collaborative control of solute segregation from both the smelting source and the final stage of solidification eliminates intergranular peeling and microscopic shrinkage defects caused by impurity enrichment at grain boundaries in the thick-thin transition zone, improving the density and mechanical properties of castings and extending their service life.

[0104] 3. This invention upgrades the traditional experience-based macroscopic parameter control in coated sand gravity casting to precise quantitative control based on microscopic solidification metallurgical behavior. Through real-time monitoring of cavity temperature, calculation and discrete statistics of grain growth rate, closed-loop feedback adjustment of air volume, and geometric mean evaluation model of solidification stability, it realizes data-driven control of all aspects from supercooling prediction, inoculation control, solidification adjustment to quality evaluation. This makes the forming quality of castings in each batch stable and controllable, significantly reduces the performance dispersion between batches, and effectively improves the consistency and yield of mass production.

[0105] 4. Without altering the existing structural dimensions, oil circuit layout, wall thickness design, and forming mold of the fixed hydraulic valve body casting, this invention can improve the micro-metallurgical quality simply by optimizing the casting process and controlling the process parameters. It is applicable to various standard thin-walled cast iron castings such as multi-way directional valve housings, hydraulic steering gear housings, gear pump housings, plunger pump housings, and hydraulic motor housings. No additional investment in mold modification costs is required, and process adjustments are simple and easy to implement. It has good engineering versatility and mass production feasibility. Attached Figure Description

[0106] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the accompanying drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. In all the drawings, similar elements or components are generally identified by similar reference numerals. In the drawings, the elements or components are not necessarily drawn to scale.

[0107] Figure 1 This is a scatter plot comparing the predicted supercooling from blank temperature measurement with the actual supercooling from temperature measurement in this invention.

[0108] Figure 2 This is a cloud map showing the global grain growth rate distribution of the present invention.

[0109] Figure 3 This is a comparison diagram of the grain size distribution of the present invention. Detailed Implementation

[0110] The embodiments of the technical solution of the present invention will now be described in detail with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and are therefore merely examples and should not be used to limit the scope of protection of the present invention. The specific structure of the present invention is as follows: Figures 1-3 As shown in the image.

[0111] Figure 1This is a scatter plot comparing the supercooling predicted by the blank temperature measurement model and the supercooling inverted by the thermocouple in an embodiment of the present invention. The horizontal axis represents the supercooling predicted by the model, the vertical axis represents the supercooling measured by the thermocouple, and the dashed line is the reference line where the predicted value is equal to the measured value. In the figure, ●, ▲, and ■ represent the measurement points in the hot spot region, the thin-walled plane region, and the thick-thin transition region, respectively. The background color partitions correspond to the high supercooling region, the medium supercooling region, and the low supercooling region, respectively.

[0112] Figure 2 The cloud map shows the global grain growth rate distribution for comparative and embodiment examples. The left image is a comparative example of the conventional process, and the right image is an embodiment of the process of the present invention. The color represents the grain growth rate value (unit μm / s), and the contour lines are 10μm / s, 20μm / s, and 30μm / s, respectively. The global average rate and the relative deviation range are marked below each image.

[0113] Figure 3 This is a comparison diagram of the equivalent grain diameter distribution between the comparative example and the embodiment, where the horizontal axis represents the equivalent grain diameter (size increases from left to right), and the vertical axis represents the area percentage (unit: %); gray fill represents the comparative example, and dark fill represents the embodiment.

[0114] To make the technical solution of the present invention clearer, the present invention will now be further described. The present invention also provides a lightweight casting forming and control process for thin-walled castings, including the following steps:

[0115] S1. After the molten iron is smelted, the molten iron is subjected to composite rare earth passivation modification treatment. The content of low melting point impurity elements in the molten iron is obtained in real time by pre-furnace spectral detection. The solute distribution is adjusted according to the detection results to limit the content of low melting point impurities in the molten iron to within the preset metallurgical control range.

[0116] Specifically, step S1 controls the content and distribution of low-melting-point impurities at the source. This step introduces composite rare earth passivation modification: rare earth elements (such as cerium and lanthanum) have a strong affinity for sulfur and lead, which can form high-melting-point rare earth sulfides and rare earth intermetallic compounds. The melting point of these compounds is much higher than that of molten iron. During solidification, they will not be squeezed to the grain boundary, but will be uniformly distributed in the matrix as dispersed particles, thereby passivating the originally harmful low-melting-point impurities into harmless or even beneficial dispersed phases.

[0117] Real-time acquisition of spectral data before the furnace is essential for achieving closed-loop control. The ultimate goal of this step is to provide a stable and controllable molten iron base for subsequent processes, eliminating the potential for grain boundary weakening caused by impurity segregation.

[0118] S2. Receive the processed molten iron and perform pouring. Before pouring, collect the no-load heat dissipation temperature parameters of different structural areas of the casting corresponding to the mold, predict the solidification undercooling of each area and build an initial forming control dataset. After pouring, collect the actual temperature parameters, calculate the actual undercooling and correct the dataset to form a global differentiated forming control dataset that is adapted to the irregular structure of thin-walled castings.

[0119] The irregular structure (hot spots, thin-walled planes, and thick-thin transition zones) of thin-walled castings leads to significant differences in heat dissipation conditions in different regions. These differences directly determine the solidification undercooling in each region, which is a key parameter affecting grain size, graphite morphology, and phase transformation timing.

[0120] Traditional processes typically employ uniform casting temperatures and cooling conditions, neglecting uneven heat dissipation caused by structural differences. This results in thin-walled areas cooling too quickly, producing white iron or hard spots, while hot-spot areas cooling too slowly, causing shrinkage porosity or coarse grains.

[0121] This step proposes a two-stage dataset construction method: blank temperature measurement before pouring and actual correction after pouring. Before pouring, the empty mold is heated and the no-load cooling curve is collected to establish a basic model of the heat dissipation characteristics of the sand mold. After pouring, the actual cooling curve of the molten iron is collected to infer the true solidification undercooling and correct the predicted value with the actual value.

[0122] This prediction-verification-correction closed-loop modeling approach utilizes the foresight of blank temperature measurement (which can generate initial control parameters before casting) and ensures the accuracy of the model through actual temperature measurement, thus solving the problems of large errors in predicting supercooling solely based on empirical formulas and strong lag in relying solely on actual temperature measurement.

[0123] Global differentiation lies in acknowledging and utilizing structural differences, not eliminating them. By establishing independent supercooling benchmarks for each region, subsequent processes such as incubation and cooling can be tailored to specific regions. The distinction between the initial molding control dataset and the global differentiated molding control dataset reflects the evolutionary logic of the data: the former represents predicted values ​​before casting, used to guide incubation control during the filling stage; the latter represents corrected actual values ​​after casting, used to guide cooling control during the solidification stage and iterative optimization for the next batch. This design aims to ensure that each casting becomes the basis for optimization of the next casting, forming a closed loop of continuous iterative quality improvement.

[0124] S3. During the filling process, the initial molding control dataset is called, and the corresponding inoculant addition parameters are matched according to the predicted supercooling of each region. The time-differentiated inoculant process is executed, and the filling process parameters are monitored in real time and the inoculant addition sequence is corrected accordingly.

[0125] Inoculants provide a large number of heterogeneous nucleation sites for graphite precipitation, thereby refining grains, promoting the formation of type A graphite, and inhibiting white iron formation. Traditional in-flow inoculation typically uses a fixed dosing rate and start time, neglecting the influence of differences in supercooling in different regions on the inoculation effect. Regions with high supercooling have strong nucleation driving force and require less inoculant to obtain a sufficient number of nuclei; regions with low supercooling have weak nucleation driving force and require more inoculant.

[0126] The purpose of this step is to achieve time-series differentiated inoculation: based on the predicted supercooling of each region in the initial dataset constructed in S2, different inoculant addition rates are matched for high, medium and low supercooled regions, so that the amount of inoculant added matches the actual nucleation requirements of the region, avoiding the problems of excessive inoculation in regions with high supercooling, resulting in hard spots, and insufficient inoculation in regions with low supercooling, resulting in white spots, which are common in traditional processes.

[0127] Real-time monitoring of filling process parameters and correction of inoculant addition sequence: After the primary crystalline phase precipitates during the filling and cooling process of molten iron, the apparent viscosity decreases with increasing shear rate (shear thinning), and the shear rate can reach 100 s in the turbulent region. -1 The above results in a decrease in the following properties of the inoculant particles, with some particles arriving at the mold cavity later than the molten iron flow. Therefore, it is necessary to dynamically adjust the start time of addition according to the actual flow rate to ensure that the inoculant and molten iron arrive at each area synchronously.

[0128] The timing design utilizes the initial forming control dataset instead of the global differential dataset: filling occurs immediately after pouring, before the actual post-pouring temperature measurement is completed and the global dataset has been corrected. Therefore, only the initial dataset built before pouring can be used. This step, through precise inoculation control, obtains uniform and appropriate amounts of graphite crystal nuclei throughout the casting, laying the foundation for uniform grain growth in the subsequent eutectic solidification stage and eliminating the potential for uneven grain size from the nucleation stage.

[0129] S4. When the casting enters the eutectic solidification stage, the duration of eutectic solidification is locked based on real-time monitoring data of cavity temperature. At the same time, the global grain growth rate is calculated and its discrete parameters are statistically analyzed. The sand mold cooling air volume is adjusted according to the time deviation and the degree of grain rate dispersion.

[0130] The eutectic solidification of gray cast iron in step S4 is a process of austenite and graphite co-growth. The duration of eutectic solidification directly affects the length and morphology of graphite flakes. If the solidification time is too long, the graphite flakes will be coarse and the mechanical properties will decrease. If the solidification time is too short, the graphite will not have enough time to grow, and supercooled graphite or white iron will easily be produced.

[0131] This step proposes a dual-parameter feedback control mechanism: on the one hand, the duration of the eutectic reglow temperature plateau is monitored in real time using cavity thermocouples, keeping it within a reasonable range; on the other hand, the grain growth rate at each point in the entire region is calculated using a grain growth rate model, and its dispersion is statistically analyzed. When the dispersion is too large, it indicates poor solidification uniformity across the entire region, requiring adjustment of the cooling airflow. The solidification process is monitored from both the time dimension (eutectic duration) and the spatial dimension (grain rate dispersion), forming a spatiotemporal dual-dimensional control system.

[0132] When adjusting the cooling airflow of the sand mold: increasing the cooling airflow accelerates heat dissipation from the sand mold, increases subcooling, accelerates grain growth rate, and shortens eutectic time; conversely, decreasing the cooling airflow slows heat dissipation, decreases subcooling, slows grain growth rate, and prolongs eutectic time. By adjusting the airflow, the eutectic time and grain rate can be simultaneously controlled, achieving coordinated control of two control objectives with one actuator.

[0133] Calculate the global grain growth rate instead of direct measurement: Grain growth rate cannot be directly measured during solidification (requires metallographic dissection). It is calculated in real time using temperature data and kinetic models to obtain uniform grain size and appropriate graphite morphology during the eutectic solidification stage, eliminate the microstructure differences caused by uneven solidification, and improve the consistency of mechanical properties throughout the casting.

[0134] S5. When the casting enters the final stage of solid-liquid coexistence molding, close the cavity venting channel to form a closed environment inside. Based on the real-time monitoring results of the cavity pressure, maintain a slight positive pressure inside the cavity to suppress the secondary migration and local enrichment of low-melting-point solutes at the solid-liquid interface.

[0135] In step S5, at the end of the solid-liquid coexistence solidification stage, 5%-10% of the liquid phase remains in the casting. This liquid phase is enriched with a large number of low-melting-point impurities (sulfur, lead, etc.) that have been displaced by the solid phase. As the solid phase fraction increases, the impurity concentration in the remaining liquid phase increases exponentially, eventually leading to severe segregation at the grain boundaries and in the final solidification region, resulting in grain boundary weakening and hot brittleness.

[0136] This step suppresses secondary solute migration through a closed, slightly positive pressure system: At the end of solidification, the exhaust channels are closed to create a sealed environment, and a slightly positive pressure of 5-15 kPa is maintained. This suppresses solute segregation in two ways: First, the slightly positive pressure increases the hydrostatic pressure in the liquid phase. According to the relationship between pressure and diffusion coefficient, increased pressure reduces the diffusion activation energy of atoms, thus slowing down the diffusion rate of impurity atoms. Second, the sealed environment eliminates the evaporation-condensation channel at the gas-liquid interface, reducing the possibility of low-melting-point impurities migrating through the gas phase.

[0137] To balance safety and effectiveness, a slightly positive pressure design is adopted instead of a high-pressure design: too low a pressure (below 5 kPa) will not have a significant suppression effect; too high a pressure (above 15 kPa) may cause the sand mold to expand or burst, and the coated sand will exhibit abnormal gas evolution behavior under high pressure. A closed-loop design maintained based on real-time monitoring results of the cavity pressure ensures precise pressure control and avoids pressure fluctuations caused by residual gas evolution from the coated sand.

[0138] The timing of the solid-liquid coexistence molding process at the end is relatively important: closing the venting too early will prevent gas from escaping during the filling stage, resulting in porosity; closing it too late will mean that the liquid phase has basically solidified, rendering the suppression of solute migration meaningless. The ultimate goal of this step is to artificially suppress the migration and enrichment of impurities through pressure during the final stage of solidification, when solute segregation is most severe, thereby eliminating grain boundary impurity segregation defects from the perspective of mass transfer kinetics.

[0139] S6. After the casting has solidified, a zoned gradient slow-release cooling process is immediately implemented. The temperature decay rate of each area is collected in real time throughout the cooling process, and the cooling intensity is adjusted accordingly to finally obtain a thin-walled casting.

[0140] After the casting solidifies, temperature differences exist in different areas due to varying heat dissipation conditions. These temperature differences lead to different shrinkage rates in different areas, resulting in thermal stress. When the thermal stress exceeds the material's high-temperature strength, cracks (hot or cold cracks) will occur. Even if cracks do not form, residual stress can cause deformation of the casting during subsequent processing or service. This step, based on the undercooling zone defined by S2, uses heat preservation and slow cooling for the high undercooling zone (thin-walled, fast cooling) and forced air cooling for the low undercooling zone (thick-walled, slow cooling). By adjusting the slow areas even slower and the fast areas even faster, the cooling rate of each zone is made more uniform, thereby reducing temperature differences and thermal stress.

[0141] A closed-loop design that monitors the temperature decay rate in real time and adjusts the cooling intensity ensures dynamic controllability of the cooling process: when the temperature difference between adjacent areas exceeds 80℃, cooling on the side with higher intensity is paused and resumed after the temperature difference decreases, effectively avoiding the accumulation of thermal stress. The elastic modulus increases non-linearly with temperature: the elastic modulus of gray cast iron is relatively low at high temperatures (approximately 40%-50% of that at room temperature above 600℃), resulting in relatively small thermal stress from the temperature difference; as the temperature decreases, the elastic modulus increases sharply, and the same temperature difference will generate much greater thermal stress. Therefore, more stringent temperature control is required in the low-temperature range. Through precise cooling control, residual stress in the casting is kept at a low level, preventing cracks and deformation, while simultaneously achieving a uniform microstructure.

[0142] Preferably, step S1 is performed according to the following steps:

[0143] a1. Prepare a composite passivating agent by mixing rare earth ferrosilicon alloy and silicon-calcium alloy at a mass ratio of 3:1, wherein the particle size of the passivating agent is controlled at 1mm to 5mm, and dry it at 200℃ to 300℃ for 1h to 2h.

[0144] a2. Before tapping the molten iron, an oxygen probe is used to detect the oxygen activity of the molten iron, and the oxygen activity is controlled at 2×10⁻⁶. -6 ~8×10 -6 Within the specified range, when the oxygen activity exceeds the specified range, the oxygen activity is adjusted by adding a passivating agent.

[0145] Specifically, when the oxygen activity is high, if the sulfur content is below the lower limit of the control range, silicon-calcium alloy should be added first; if the sulfur content is above the upper limit of the control range, rare earth silicon-iron alloy should be added first; if the sulfur content is within the control range, the composite passivating agent prepared in step a1 should be added.

[0146] a3. Simultaneously, a furnace-front laser-induced breakdown spectrometer is used to detect the mass fraction of sulfur and lead low-melting-point impurities in molten iron. Each furnace is tested no less than twice. For critical batches, an online continuous monitoring mode is used with a monitoring interval of 20s to 40s.

[0147] The sulfur mass fraction is controlled at 0.05% to 0.08%, and the lead mass fraction is controlled at 0.002% to 0.005%. When the impurity content exceeds the upper limit of the control range, the composite passivating agent prepared in step a1 is added at 0.05% to 0.10% of the molten iron mass, and the amount added at one time shall not exceed 0.15% of the molten iron mass, until the impurity content falls back to the control range.

[0148] This scheme further limits the composite rare earth passivation modification treatment in step S1. Its design aims to precisely control the existence form and distribution upper limit of low-melting-point impurities in molten iron from the smelting source, laying the compositional foundation for inhibiting solute segregation during subsequent solidification. Rare earth ferrosilicon alloy and silicon-calcium alloy are compounded in a 3:1 mass ratio because rare earth elements have a strong affinity for impurities such as sulfur and lead, forming high-melting-point rare earth sulfides and rare earth lead compounds which float to the surface and are removed. The calcium in the silicon-calcium alloy further deoxidizes and desulfurizes, improving the solubility of rare earth elements in molten iron. The synergistic effect of the two is superior to single rare earth treatment. Particle size is controlled between 1mm and 5mm and dried at 200℃ to 300℃ to avoid incomplete reaction due to excessively large particle size and severe burn-off due to excessively small particle size, while also preventing secondary oxidation and gas absorption in the molten iron caused by moisture adsorption.

[0149] Among them, a zirconium oxide solid electrolyte oxygen probe was used to detect the oxygen activity of molten iron and control it at 2×10. -6 Up to 8×10 -6The range is defined because oxygen activity directly affects the effective utilization rate of rare earth elements and the nucleation conditions of graphite: when oxygen activity is too high, rare earth elements preferentially react with oxygen and are consumed in impurity removal; when oxygen activity is too low, the reducing power of the molten iron is too strong, and the inoculation effect is weakened. Based on the direction of sulfur content deviation from the range, priority is given to adding silicon-calcium or rare earth ferrosilicon, achieving directional and precise feeding to control impurities and avoiding compositional fluctuations caused by blind addition. Using pre-furnace laser-induced breakdown spectroscopy to detect the mass fraction of sulfur and lead, with a requirement of at least two detections per furnace and continuous monitoring for 20 to 40 seconds for key batches, is to achieve real-time closed-loop monitoring of impurity content. When the upper limit is exceeded, composite passivating agent is added at 0.05% to 0.10% of the molten iron mass, with a single addition not exceeding 0.15%, ensuring both the effectiveness of impurity removal and preventing excessive addition that could cause a significant drop in molten iron temperature and excessive rare earth residue. The above parameters together constitute a source control system from composition detection to precise feeding, ensuring that the molten iron is in a metallurgical state with stable low-melting-point impurity content and suitable oxygen activity before entering the casting process, fundamentally reducing the total amount of solute that can participate in segregation during solidification.

[0150] Preferably, step S2 is performed according to the following steps:

[0151] b1. K-type armored thermocouples are pre-embedded in the inner wall of the mold cavity corresponding to the hot spot area, thin-walled plane area and thick-thin transition area of ​​the casting. The distance between the thermocouple measuring point and the surface of the mold cavity is controlled between 2mm and 5mm, the response time is not greater than 0.5s, and the temperature measurement range meets the short-term temperature measurement requirements of 0℃ to 1600℃.

[0152] b2. Before casting, a blank temperature measurement process is performed to collect the no-load cooling curves of each area. Considering the non-linear decay of the thermal diffusivity of the coated sand with increasing temperature, different thermal diffusivity correction values ​​are used to correct the cooling rate in segments according to three temperature ranges: below 200℃, 200℃~600℃, and above 600℃. A correction model for the heat dissipation characteristics of the sand mold in each area is established. Based on this, the solidification undercooling and phase change initiation time of each area are predicted, and an initial molding control dataset is constructed.

[0153] After casting, temperature curves of each region are collected. Based on the heat dissipation characteristics, the model is modified to back-derive the molten iron cooling curve. The difference between the equilibrium eutectic temperature and the actual eutectic solidification initiation temperature is extracted to calculate the actual solidification undercooling. The calculated undercooling value ranges from 5K to 40K. The initial forming control dataset is corrected based on the actual undercooling.

[0154] The revised dataset is used for both the control of subsequent cooling processes in this batch and the iteration of the forming control dataset for the next batch.

[0155] Based on the temperature data collected by K-type armored thermocouples pre-embedded in the inner wall of the mold cavity corresponding to the hot spot area, thin-walled plane area, and thickness transition area of ​​the casting, the solidification undercooling of each area predicted by the blank temperature measurement process based on the heat dissipation characteristic correction model is compared with the actual solidification undercooling calculated by inversion based on the actual temperature curve after casting. The results are as follows: Figure 1 As shown. By Figure 1 It can be seen that the predicted and measured undercooling at each measuring point are both distributed near the diagonal reference line, and the measuring points in different structural regions fall into the three levels of high undercooling, medium undercooling, and low undercooling, respectively. This indicates that the established sand mold heat dissipation characteristic correction model can accurately predict the solidification undercooling of each region of the irregular structure of thin-walled castings. The initial forming control dataset built based on this model has a reliable predictive basis, and after correction of the actual undercooling after casting, it can be effectively used for differentiated control of subsequent processes.

[0156] b3. Subsequently, the castings are divided into three levels according to the supercooling value: high supercooling zone (25K~40K), medium supercooling zone (15K~25K), and low supercooling zone (5K~15K). The inoculant dosing acceleration rate range and cooling air volume range are set for each zone respectively, and finally a global differentiated molding control dataset is formed.

[0157] Specifically, this scheme further defines the temperature monitoring and dataset construction in step S2. Its design aims to establish a control basis that can accurately predict and dynamically correct the undercooling in each region of the irregularly shaped thin-walled casting, addressing the inherent uneven heat dissipation conditions in different areas. This avoids the localized overcooling or undercooling caused by the uniform cooling parameters used in traditional processes. K-type armored thermocouples are pre-embedded in the hot spot region, the thin-walled planar region, and the thickness transition region, with the measuring points 2mm to 5mm from the cavity surface and a response time of no more than 0.5s. This is because these three types of regions represent typical locations where the solidification behavior of the casting differs most: the hot spot region has the slowest heat dissipation and the smallest undercooling; the thin-walled planar region has the fastest heat dissipation and the largest undercooling; and the thickness transition region has heat dissipation conditions between the two and the most complex temperature gradient. The measuring points being 2mm to 5mm from the cavity surface avoid direct erosion damage from the high-temperature molten iron while ensuring a fast response time for the temperature signal. A response time of no more than 0.5s ensures the capture of temperature inflection points in rapid phase transformation processes such as eutectic solidification. Before casting, blank temperature measurement was performed on the empty mold and the no-load cooling curve was collected. Considering the nonlinear decay of the thermal diffusivity of the coated sand with increasing temperature, the cooling rate was corrected in segments according to three temperature ranges: below 200℃, 200℃ to 600℃, and above 600℃. This is because the thermal conduction mechanism of the coated sand is different in different temperature ranges: in the low temperature range, solid phase heat conduction is the main process; in the medium temperature range, the decomposition of the resin binder leads to changes in porosity and a decrease in thermal conductivity; and in the high temperature range, the proportion of radiative heat transfer increases. If a single thermal diffusivity is used for calculation, the prediction deviation of the cooling rate in the high temperature range will exceed 20%. Based on the segmented correction heat dissipation characteristic model, the supercooling degree and phase transformation initiation time of each region are predicted, and an initial molding control dataset is constructed to provide a forward-looking control basis for subsequent inoculation and cooling processes. After casting, the actual temperature curve is collected to deduce the molten iron cooling curve. The difference between the equilibrium eutectic temperature and the actual eutectic solidification initiation temperature is extracted to calculate the actual undercooling. The initial dataset is then corrected. The corrected dataset is used for real-time adjustment of subsequent cooling processes in this heat and also serves as the basis for iterative optimization of the dataset for the next heat, forming a closed-loop data accumulation method of prediction, measurement, correction, and iteration. Undercooling is divided into three levels: high, medium, and low, corresponding to the inoculant addition acceleration rate and cooling air volume range, respectively. This allows subsequent processes to implement differentiated control parameters based on the undercooling level of each region, achieving refined forming control tailored to specific regions.

[0158] Preferably, step S3 is performed as follows:

[0159] c1. First, call the initial forming control dataset and match the molten iron mass flow rate of 0.20%~0.30%, 0.10%~0.20%, and 0.05%~0.10% of the 75% ferrosilicon inoculant addition acceleration rate for the high subcooling zone, medium subcooling zone, and low subcooling zone, respectively, with the inoculant particle size controlled between 0.2mm and 1.0mm.

[0160] c2. Because the apparent viscosity of molten iron is approximately 2 mPa·s to 6 mPa·s within the filling temperature range, and after the primary crystalline phase precipitates as the filling temperature decreases, it exhibits non-Newtonian fluid characteristics where the apparent viscosity decreases with increasing shear rate; in the turbulent filling region, the shear rate can reach 100 s⁻¹. -1 The above factors lead to a decrease in the conformability of probiotic particles;

[0161] The inoculant addition acceleration rate is adjusted in real time based on the flow velocity of the molten iron stream. Specifically, for every 0.1 m / s increase in flow velocity, the inoculant addition acceleration rate is increased by 1% to 3%.

[0162] c3. At the same time, based on the arrival time of molten iron in each area, the inoculant addition start time in the corresponding area is advanced by 0.5s to 1.0s to ensure that the inoculant and molten iron enter each area of ​​the mold cavity synchronously.

[0163] c4. After the filling process is completed, the inflection point of the eutectic phase transition initiation temperature is collected in real time by pre-embedded thermocouples in each area, and the actual phase transition initiation time of each area is calculated.

[0164] When the actual phase change start time of a certain area is found to deviate from the predicted time in the dataset by more than 5 seconds, it is determined that there is a deviation in the inoculation control parameters. The deviation data is recorded, and the inoculation agent dosing rate is increased or decreased according to the direction of the deviation. The control parameters of the corresponding area in the production database are corrected and used for the iteration of the molding control dataset for the next batch.

[0165] In addition, when the actual phase change start time deviation between adjacent regions is greater than 5s, the solidification uniformity deviation data of the whole region is recorded and stored in the production database as the basis for the iterative optimization of the next batch dataset.

[0166] Specifically, this scheme further defines the timing-differentiated inoculation process in step S3. Its design aims to precisely match the inoculant dosage and timing based on the solidification undercooling level of different areas of the casting, ensuring uniform inoculation across the entire region and avoiding insufficient inoculation in thin-walled areas or excessive inoculation in hot spots caused by traditional uniform inoculation. The undercooling zoning results from the initial forming control dataset are used to match 75% ferrosilicon inoculant dosing acceleration rates of 0.20% to 0.30%, 0.10% to 0.20%, and 0.05% to 0.10% of the molten iron mass flow rate for high, medium, and low undercooling zones, respectively. This is because areas with higher undercooling cool faster molten iron cools, resulting in shorter inoculant dissolution and diffusion time, requiring a higher dosing acceleration rate to ensure effective inoculation concentration. Conversely, hot spots with lower undercooling cool slower, allowing sufficient time for inoculant dissolution and diffusion; an excessively high dosing acceleration rate can lead to inoculation degradation and coarse graphite. The particle size is controlled between 0.2 mm and 1.0 mm to balance the dissolution rate of the inoculant and its resistance to molten iron erosion: particles that are too large will not dissolve sufficiently and are prone to forming inclusions, while particles that are too small will be rapidly burned off during the flow process. Considering the non-Newtonian fluid characteristics of molten iron after filling and cooling to precipitate the primary crystalline phase, its apparent viscosity decreases with increasing shear rate, and the shear rate in the turbulent region can reach 100 s⁻¹. -1 The above factors lead to a decrease in the follow-through of inoculant particles. To compensate for the loss in actual inoculant delivery caused by turbulent ferroic entrainment and decreased particle follow-through, the inoculant dosing rate is increased by 1% to 3% for every 0.1 m / s increase in filling flow rate. This ensures a stable and effective inoculant dosage in each region under different filling flow rates. Furthermore, the inoculant dosing start time for each region is advanced by 0.5 to 1.0 seconds based on the arrival time of ferroic iron. This compensates for the delay in inoculant particle delivery to the target region, ensuring thorough mixing of the inoculant and ferroic iron within the critical filling time window in the target region and preventing misalignment of inoculant effects due to delayed dosing. After the molding process is completed, the inflection point of the eutectic phase transformation initiation temperature is collected to calculate the actual phase transformation initiation time. If the deviation from the predicted time is greater than 5 seconds, it is determined that there is a deviation in the inoculation control parameters, which is recorded and used to correct the parameters for the next batch. If the deviation between adjacent regions is greater than 5 seconds, the deviation in the overall solidification uniformity is recorded for iteration. This design enables the post-event evaluation of the inoculation effect and the optimization of parameters across batches, avoiding the process chaos caused by forced intervention when the process is irreversible in the current batch. The above measures together ensure that the inoculant reaches the correct region at the correct time and with the correct dosage, resulting in uniform graphite nucleation conditions throughout the entire region.

[0167] Preferably, step S4 is performed as follows:

[0168] d1. The temperature curve is collected in real time by pre-embedded thermocouples in the cavity. The duration of the eutectic reglow temperature plateau is locked between 20s and 60s. When the monitoring duration is less than 20s, the cooling air volume of the sand mold surface is reduced, and when it is more than 60s, the cooling air volume is increased.

[0169] During the eutectic solidification of gray cast iron, austenite and graphite precipitate simultaneously, releasing a large amount of latent heat of crystallization. This causes the declining temperature curve to rebound or pause, forming a eutectic reglow plateau. Simply put, the eutectic reglow temperature plateau is a stable temperature segment on the cooling curve formed during the eutectic solidification of gray cast iron, where the release of latent heat of solidification and heat dissipation reach a dynamic equilibrium. Its length is a direct signal for judging whether the eutectic solidification is normal.

[0170] The duration of this platform directly reflects the progress of eutectic solidification: a platform that is too short (<20s) indicates that cooling is too rapid, eutectic solidification is completed in a short time, graphite does not have enough time to grow, and supercooled graphite and white iron are easily produced; a platform that is too long (>60s) indicates that cooling is too slow, eutectic solidification time is too long, graphite flakes are coarse, and mechanical properties are reduced. 20-60s is the preferred range for thin-walled valve body castings, and the specific value depends on the casting wall thickness and structure. The control logic for reducing / increasing the purging cooling air volume is: reduced air volume → slower heat dissipation from the sand mold → longer solidification time → increased platform duration; increased air volume → faster heat dissipation → shorter solidification time → reduced platform duration. Through closed-loop adjustment of the air volume, the platform duration is locked within the target range. Sand mold surface purging cooling refers to blowing air to the outer surface of the sand mold by a fan to forcibly remove the heat from the sand mold. The advantages of this cooling method are adjustable cooling intensity, fast response speed, and non-contact with the casting, making it suitable for online closed-loop control. By dynamically adjusting the cooling intensity through real-time monitoring of the characteristic parameters (platform duration) of the temperature curve, precise control of solidification time can be achieved.

[0171] d2. Simultaneously control the cooling rate of the eutectic stage to be maintained within the range of 0.5℃ / s to 2.0℃ / s; this sub-step is responsible for controlling the cooling rate of the eutectic stage and is executed in parallel with the plateau duration control of d1.

[0172] 0.5-2.0℃ / s is the preferred cooling rate range for obtaining uniform A-type graphite and moderate grain size in thin-walled gray cast iron parts. Simultaneous control indicates that step d1 runs in parallel with step d2, both jointly adjusting the cooling airflow—d1 focuses on the platform duration (time dimension), and d2 focuses on the cooling rate (rate dimension). When their adjustment directions conflict (e.g., a shorter platform duration requires reduced airflow, but a faster cooling rate also requires reduced airflow, resulting in a consistent direction; or a longer platform requires increased airflow, but a slower cooling rate requires reduced airflow, resulting in a conflicting direction), the control system needs to coordinate based on priority or weighted algorithms.

[0173] From the perspective of control theory, d1 is interval control (locking the duration between 20-60s), and d2 is rate control (maintaining the cooling rate between 0.5-2.0℃ / s). The combination of the two achieves comprehensive constraint on the solidification process.

[0174] d3. The grain growth rate is calculated using a thin-walled valve body eutectic solidification grain growth rate prediction model. The interface radius of curvature is estimated based on the current cooling rate and undercooling value. The larger the cooling rate and the larger the undercooling, the smaller the radius of curvature. The estimated value range is 1μm to 100μm. When the estimated value exceeds this range, the boundary value is substituted into the calculation.

[0175] ;

[0176] In the model, temperature determines the growth rate, curvature determines the effective undercooling, and the two together determine the grain growth rate.

[0177] In the formula, This is a calculated value for grain growth rate, in μm / s. Within the range of process parameters, the calculated value is generally 5μm / s to 40μm / s. It is used to statistically analyze the relative dispersion parameters of each point in the entire region. Instead of using the rate value of a single point, it statistically analyzes the dispersion of the rate at each point in the entire region (such as standard deviation and coefficient of variation) to evaluate the solidification uniformity.

[0178] Frequency factor, unit μm / The value range is 1.0 × 10 4 μm / ~1.5×10 4 μm / The range of solidification kinetics data for gray cast iron is determined by referring to the data and combining it with on-site process experience as needed. It can be finely adjusted by those skilled in the art as needed. It refers to the maximum growth rate under unit supercooling, reflecting the inherent growth capacity of the material. Those skilled in the art can make appropriate adjustments based on the specific molten iron composition, casting structure and equipment conditions.

[0179] The solidification activation energy is expressed in J / mol and ranges from 1.0 × 10⁻⁶. 5 J / mol ~ 1.5 × 10 5 J / mol, which is determined by referring to the data range of activation energy for eutectic phase transformation of iron-carbon alloys and combining on-site process experience as needed, can be finely adjusted by those skilled in the art as needed; this parameter represents the energy barrier that atoms need to overcome to diffuse from the liquid phase to the solid phase, and the unit is J / mol (joules per mole).

[0180] The gas constant is expressed in J / L. Take 8.314 J / This is an internationally recognized physical constant that is universally accepted and does not change with materials, processes, or scenarios.

[0181] The solid-liquid interface temperature, in K, ranges from 1390K to 1425K. The solid-liquid interface temperature refers to the temperature at the interface between the solid and liquid phases during solidification. It is not equal to the overall temperature of the molten iron, nor is it equal to the equilibrium eutectic temperature. Rather, it is a dynamic value between the two, and its range is determined by technicians based on the actual situation of eutectic solidification of gray cast iron.

[0182] In addition, those skilled in the art know that the solid-liquid interface temperature can be obtained by solidification heat transfer calculation using numerical simulation software for the casting process (such as ProCAST, MAGMA, etc.), or it can be estimated based on measured cooling curves combined with back-calculation thermal analysis, or approximate measurement can be performed by pre-embedding miniature thermocouples in key parts of the casting.

[0183] The actual undercooling during eutectic solidification, expressed in K, ranges from 5K to 40K; it refers to the difference between the equilibrium phase transition temperature and the actual phase transition initiation temperature, and is the core driving force of the solidification process—the greater the undercooling, the stronger the solidification driving force, and the higher the nucleation rate and growth rate. In the model, undercooling is the base value for the Gibbs-Thomson correction term. This creates an effective undercooling, which directly drives grain growth.

[0184] The Gibbs-Thomson coefficient, in K.μm, ranges from 0.1 K.μm to 0.5 K.μm. It is determined based on the solidification interface energy and latent heat of fusion parameters of iron-carbon alloys. The specific calculation formula is as follows: ;in The solid-liquid interface energy of iron-carbon alloys. The equilibrium eutectic temperature for gray cast iron. The latent heat of fusion per unit volume;

[0185] The physical meaning of the Gibbs-Thomson coefficient is the amount of melting point reduction (K) corresponding to a unit curvature (1 / μm).

[0186] The Gibbs-Thomson coefficient in the model serves to correct for supercooling using curvature—when the interface curvature radius... When it is 1μm, It can reach 0.1-0.5K, accounting for 2%-10% of the low supercooled region (5K), and has a perceptible impact on the growth rate; when When it is 100μm, The effect is negligible, ranging from only 0.001 to 0.005 K. This correction significantly improves the model's prediction accuracy in the fine-grained region.

[0187] The radius of curvature of the solidification interface, in μm, ranges from 1 μm to 100 μm. The radius of curvature is a geometric parameter describing the degree of curvature at the solid-liquid interface. During solidification, the solid-liquid interface is not an ideal plane but exhibits complex morphologies such as dendrites and cellular crystals. The radius of curvature varies at different points on the interface—smaller at the dendrite tips (1-5 μm), and larger at the dendrite trunk and eutectic junction (50-100 μm). The range of 1-100 μm covers all possible curvature scales that may occur during the eutectic solidification of gray cast iron. The radius of curvature cannot be directly measured; in this sub-step, it is estimated using the cooling rate and undercooling: the higher the cooling rate and the greater the undercooling, the more unstable the solidification interface, the more developed the dendrites, and the smaller the radius of curvature; conversely, the lower the cooling rate and the greater the undercooling, the larger the radius of curvature.

[0188] Based on the calculation results, the relative dispersion parameter of grain growth rate at each point in the entire region is statistically analyzed. When the relative deviation of the grain growth rate at each point from the average value of the entire region exceeds +15%, the sand mold purging air volume in the corresponding area is reduced by 0.1 m³. 3 / min~0.5m 3 / min;

[0189] Conversely, when the relative deviation between the grain growth rate at each point and the global average exceeds -15%, the sand mold purging airflow in the corresponding area is increased by 0.1 m³. 3 / min~0.5m 3 / min, so that the deviation of the global grain growth rate converges to within ±15% of the average rate.

[0190] This system is responsible for adjusting the cooling airflow based on discrete parameters related to grain rate. Its working principle is based on the correlation chain of cooling rate, subcooling, and growth rate: increased cooling airflow → faster heat dissipation from the sand mold → increased cooling rate → increased subcooling → faster grain growth rate; conversely, decreased airflow → slower heat dissipation → decreased subcooling → slower growth rate. Therefore, when the grain rate in a certain region is more than 15% higher than the overall average (positive deviation), it indicates that the region is cooling too fast, and the airflow should be reduced to decrease the growth rate; when the rate is more than 15% lower than the average (negative deviation), it indicates that the region is cooling too slowly, and the airflow should be increased to increase the growth rate. The ±15% control threshold is a trade-off between process accuracy and control costs—a threshold that is too strict (e.g., ±5%) will lead to frequent airflow adjustments and system instability; a threshold that is too lenient (e.g., ±30%) will result in excessive grain size variation and poor microstructure uniformity.

[0191] To visually demonstrate the control effect of the above-mentioned grain growth rate feedback regulation, a thin-walled valve body eutectic solidification grain growth rate prediction model was used. Global grain growth rate calculations were performed on a comparative example using the traditional process and an embodiment using the process of this invention. Spatial distribution cloud maps were generated based on the calculation results at each point. The results are as follows: Figure 2 As shown. By Figure 2 As can be seen, the comparative example, due to the use of uniform cooling conditions, exhibits significant differences in grain growth rates across different structural regions of the casting, resulting in a pronounced global dispersion. In contrast, the embodiment utilizes sand mold cooling airflow feedback adjustment based on the relative dispersion parameter of grain growth rate. When the relative deviation of the grain growth rate at each point from the global average exceeds +15%, the corresponding area's purging airflow is reduced; when it exceeds -15%, the corresponding area's purging airflow is increased. This brings the global grain growth rate deviation to within ±15% of the average rate, resulting in a more uniform rate distribution. These results demonstrate that the grain growth rate dispersion parameter feedback adjustment mechanism described in this invention effectively reduces the global growth rate unevenness caused by differences in heat dissipation due to the irregular structure of the casting, providing a control basis for obtaining a uniform solidification structure.

[0192] The model works by decomposing the grain growth process into two coupled physical effects: the Arrhenius effect describes the influence of temperature on the atomic diffusion rate (higher temperature, faster diffusion, and greater growth rate), and the Gibbs-Thomson effect describes the influence of interface curvature on the equilibrium melting point (greater curvature, lower equilibrium melting point, smaller effective undercooling, and slower growth rate). After coupling, the grain growth rate is jointly determined by temperature and interface curvature.

[0193] The interface radius of curvature is estimated based on the cooling rate and supercooling because the radius of curvature cannot be directly measured, but it has an empirical correlation with these two factors: the greater the cooling rate and supercooling, the more unstable the solidification interface, the sharper the dendrite tips, and the smaller the radius of curvature; conversely, the smaller the cooling rate and supercooling, the larger the radius of curvature. The value range of 1-100μm covers all cases from the fine-grained region (1μm) to the coarse-grained region (100μm). When the estimated value exceeds the range, the boundary value is substituted to avoid calculation instability caused by extreme values. The unconventional role of this model is to achieve soft measurement of grain growth rate—grain growth rate cannot be directly measured during solidification (it requires metallographic dissection after solidification), but it can be calculated in real time through temperature data and kinetic models, thus providing a basis for online control. The scenario limitation of thin-walled valve bodies allows the model parameters to be optimized for the structure and composition of the valve body casting, resulting in higher prediction accuracy than general solidification models. The ultimate goal of this step is to obtain the grain growth rate at all points in the entire domain through model calculation, providing a data foundation for subsequent discrete parameter statistics and airflow regulation.

[0194] Specifically, this scheme further limits the feedback adjustment of cooling airflow during the eutectic solidification stage in step S4. Its design aims to dynamically adjust the sand mold cooling airflow by real-time monitoring of the eutectic solidification process and grain growth rate dispersion, ensuring uniform solidification speed and grain growth conditions throughout the casting during the eutectic solidification stage. This fundamentally eliminates the polarization of grain size caused by differences in structural heat dissipation. Thermocouples are embedded in the mold cavity to collect temperature curves, and the duration of the eutectic reglow temperature plateau is locked between 20 and 60 seconds. This is because the duration of the eutectic reglow plateau directly reflects the balance between the latent heat released and heat dissipation during eutectic solidification: a duration shorter than 20 seconds indicates excessive heat dissipation and insufficient eutectic phase transformation, easily leading to undercooled graphite and a tendency for white iron; a duration longer than 60 seconds indicates excessive heat dissipation and prolonged eutectic solidification time, easily resulting in coarse graphite and grain coarsening. Reducing the airflow below 20 seconds slows cooling, while increasing the airflow above 60 seconds enhances cooling, keeping the eutectic solidification process within a suitable time window. Maintaining the cooling rate during the eutectic stage between 0.5℃ / s and 2.0℃ / s ensures the eutectic phase transformation occurs within a reasonable range of undercooling. Too low a cooling rate results in insufficient undercooling and coarse graphite, while too high a rate leads to excessive undercooling and the formation of D-type undercooled graphite. A thin-walled valve body eutectic solidification grain growth rate prediction model was used to calculate the grain growth rate at each point. This model comprehensively considers the influence of interface temperature on atomic diffusion activation energy and the Gibbs-Thomson effect of interface curvature on equilibrium melting point, enabling relatively accurate estimation of the grain interface migration rate based on real-time temperature data. The interface curvature radius was calculated based on the cooling rate and undercooling and limited to 1μm to 100μm; values ​​exceeding this range were substituted using boundary values, avoiding the difficulty of directly measuring curvature parameters. Based on the calculation results, the relative dispersion parameters of the grain growth rate at each point across the entire region were statistically analyzed. When the relative deviation exceeded +15%, the corresponding area's purging airflow was reduced by 0.1m³. 3 / min to 0.5m 3 / min, when the airflow exceeds -15%, increase the corresponding area's purging airflow by 0.1m³. 3 / min to 0.5m 3 / min, bringing the global deviation to within ±15% of the average rate, this closed-loop adjustment method transforms the unquantifiable grain uniformity target in traditional processes into a specific airflow adjustment command that can be calculated in real time and controlled by feedback. This achieves proactive and precise control of grain growth behavior during the eutectic solidification stage, providing process assurance for obtaining a uniform metallographic structure.

[0195] Preferably, step S5 is performed according to the following steps:

[0196] e1. Lay a high-temperature resistant aluminum silicate sealing gasket on the sand mold parting surface, wherein the thickness of the sealing gasket is controlled at 2mm to 4mm and the compression is controlled at 30% to 50% to ensure that the sealing gap of the parting surface is not greater than 0.1mm.

[0197] e2. When the cavity temperature drops to 5℃~20℃ below the equilibrium eutectic temperature, it is determined that the casting has entered the final stage of solidification with solid-liquid coexistence, and all exhaust channels are closed to form a sealed environment inside the cavity.

[0198] e3. The cavity pressure is monitored in real time by the cavity pressure sensor, and the cavity micro-positive pressure is maintained at 5kPa to 15kPa. This pressure is dynamically balanced by the inert gas pressure replenishment, residual gas generation of the coated sand and cavity venting.

[0199] When the pressure is below 5 kPa, the pressure is replenished through the inert gas inlet valve; conversely, when the pressure is above 15 kPa, the pressure is released by opening the micro exhaust valve.

[0200] e4. The sealed pressure maintenance time should be controlled between 8 and 15 minutes. During this period, it is forbidden to open the sand mold or disturb the casting.

[0201] Specifically, this scheme further defines the closed micro-positive pressure control in step S5. Its design purpose is to suppress the secondary migration and local enrichment of low-melting-point solutes at the solid-liquid interface by constructing a closed micro-positive pressure environment in the solid-liquid coexistence stage at the end of solidification, thereby eliminating the inherent defect of grain boundary impurity segregation that traditional processes cannot solve from the perspective of solidification kinetics.

[0202] A 2mm to 4mm thick high-temperature resistant aluminum silicate sealing gasket is laid on the sand mold parting surface, and the compression is controlled at 30% to 50%, with the sealing gap on the parting surface not exceeding 0.1mm. This is because aluminum silicate fiber material has good elasticity and heat insulation properties at high temperatures. A compression of 30% to 50% can ensure that the sealing gasket fits tightly with the parting surface to form an effective seal, without causing the sealing gasket to lose its resilience due to excessive compression. A sealing gap of not more than 0.1mm ensures that the cavity gas will not leak a large amount from the parting surface during subsequent pressurization, thus maintaining stable cavity pressure. When the cavity temperature drops to 5°C to 20°C below the equilibrium eutectic temperature, it is determined that the solid-liquid coexistence solidification stage has begun, and all venting channels are closed. The selection of this temperature window is crucial: when the temperature is 5°C above the equilibrium eutectic temperature, the casting is still in the early stage of eutectic solidification with a low solid fraction. Closing the venting channels too early will affect the discharge of residual gas and gas-generating products, easily leading to porosity defects. When the temperature is 20°C below the equilibrium eutectic temperature, the casting is close to complete solidification, and solute segregation has basically formed. At this time, closing the venting channels will not have an inhibitory effect. After closing the venting channels, inert gas is used to replenish the pressure, and the residual gas generated by the coated sand and the cavity venting are dynamically balanced to maintain a slight positive pressure of 5kPa to 15kPa in the cavity. The role of the slight positive pressure is to increase the gas partial pressure in the liquid phase at the solid-liquid interface front, inhibiting the migration driving force of low melting point impurity elements to the grain boundaries and the last solidified region due to the concentration gradient and solidification shrinkage. When the pressure is below 5kPa, the inhibitory effect is insufficient, and when it is above 15kPa, it may cause scorching or sulfur penetration defects on the casting surface. Maintaining a closed, stable pressure for 8 to 15 minutes, during which opening the sand mold or disturbing the casting is prohibited, is to ensure a consistently stable pressure environment during the most active solute migration period at the end of the solid-liquid coexistence phase, allowing the inhibitory effect to fully cover the entire process of segregation formation. From the perspective of solidification mass transfer kinetics, these measures, without altering the casting composition and structure, change the boundary conditions for solute migration through pressure environment control, effectively reducing the enrichment of low-melting-point impurities at grain boundaries in the thick-thin transition zone and hot spot region.

[0203] Preferably, step S6 is performed according to the following steps:

[0204] f1. First, perform gradient slow-release cooling according to the global subcooling benchmark zone calibrated in step S2. In the high subcooling zone, heat preservation slow cooling is adopted, and the cooling rate is controlled at 0.05℃ / s~0.1℃ / s.

[0205] The intermediate subcooling zone uses weak airflow cooling at a rate of 0.1℃ / s to 0.3℃ / s;

[0206] The low subcooling zone uses forced air cooling at a rate of 0.3℃ / s to 0.8℃ / s;

[0207] f2. Throughout the cooling process, temperature decay data is collected in real time by pre-embedded thermocouples in each area. Considering the non-linear increase of the elastic modulus of gray cast iron as the temperature decreases, different temperature coefficients of elastic modulus are used for three temperature ranges: below 200℃, 200℃~600℃, and above 600℃. The thermal stress caused by the temperature difference between adjacent areas is calculated in combination with the thermal expansion coefficient of gray cast iron. When the temperature difference between adjacent areas is greater than 80℃, the cooling process of the side with higher cooling intensity in the area with the temperature difference exceeding the limit is suspended. The cooling process of that side is resumed after the temperature difference drops to within 50℃.

[0208] f3. When the overall temperature of the casting drops below 200℃, the zoned cooling control is released and the casting is switched to air cooling to room temperature. The temperature of the casting after unpacking should not exceed 150℃.

[0209] Specifically, this scheme further defines the zoned gradient slow-release cooling in step S6. Its design purpose is to implement differentiated cooling intensity control based on the supercooling level and heat capacity differences in different regions after the casting has completely solidified, so that the temperature decays uniformly across the entire region and avoids thermal stress concentration and casting deformation and cracking caused by excessive differences in cooling rate. The high subcooling zone, medium subcooling zone, and low subcooling zone, as defined in step S2, are respectively subjected to heat preservation and slow cooling at 0.05℃ / s to 0.1℃ / s, weak air cooling at 0.1℃ / s to 0.3℃ / s, and forced air cooling at 0.3℃ / s to 0.8℃ / s. This distribution of cooling intensity is matched with the heat capacity and solidification state of each zone: the high subcooling zone usually corresponds to thin-walled areas, which have small heat capacity and rapid temperature drop after solidification. If rapid cooling continues, the temperature in this zone will be much lower than that in the thick-walled areas, resulting in a huge temperature difference and thermal stress. Therefore, heat preservation and slow cooling are used to slow down the rate of temperature drop and keep it in sync with the thick-walled areas. The low subcooling zone usually corresponds to hot thick-walled areas, which have large heat capacity, large heat storage, and slow natural cooling. Forced air cooling is used to accelerate its cooling so that its temperature catches up with the thin-walled areas as soon as possible, thereby reducing the overall temperature difference. Throughout the cooling process, temperature decay data for each region is collected in real time. Considering the non-linear increase in the elastic modulus of gray cast iron with decreasing temperature, different temperature coefficients of elastic modulus combined with the coefficient of thermal expansion are used to calculate thermal stress in three temperature ranges: below 200℃, 200℃ to 600℃, and above 600℃. This is because the magnitude of thermal stress depends not only on the temperature difference but also directly on the material's elastic modulus and coefficient of thermal expansion at that temperature: in the high-temperature range, the elastic modulus is low and the material has good plasticity, resulting in relatively small thermal stress for the same temperature difference; in the low-temperature range, the elastic modulus increases significantly and the material tends to be brittle, resulting in a substantial increase in thermal stress for the same temperature difference. Ignoring the non-linear temperature variation of the elastic modulus would lead to a severe underestimation of thermal stress in the low-temperature range. When the temperature difference between adjacent regions exceeds 80℃, cooling on the side with stronger thermal stress is paused until the temperature difference drops below 50℃ before resuming cooling. This temperature difference threshold control provides a safety margin for thermal stress, preventing local thermal stress from exceeding the material's strength limit and causing cracks. Once the overall temperature drops below 200℃, the zoned cooling is discontinued and air cooling is initiated to room temperature. The opening temperature should not exceed 150℃. This ensures that the casting completes most of its cooling and shrinkage during the higher temperature and better plasticity stage. Opening the casting after it has cooled to a low temperature avoids surface hardening and residual stress caused by external rapid cooling. The above-mentioned zoned gradient cooling strategy achieves a uniform temperature field decay of the casting from high temperature to room temperature, effectively reducing residual stress and deformation tendency.

[0210] Preferably, after step S6 is completed, a casting solidification stability evaluation process is also included, which is performed according to the following steps:

[0211] g1. Summarize four monitoring parameters collected throughout the molding process: undercooling deviation, maximum absolute dispersion deviation of grain growth rate, local enrichment concentration increment of low melting point impurities, and deviation of phase transformation initiation time.

[0212] g2. Subsequently, the solidification stability was calculated using the geometric mean evaluation model for solidification stability:

[0213] ;

[0214] In the formula, The solidification stability evaluation index is dimensionless. It is a dimensionless comprehensive quality score with a value range of 0-1. The closer the value is to 1, the better the solidification quality and the smaller the deviation in each dimension. The closer the value is to 0, the worse the solidification quality and the more serious the defects.

[0215] Using a dimensionless index of 0-1 instead of specific physical quantities (such as MPa, μm) as the evaluation output has the following advantages: First, it facilitates quality comparison between different batches and types of castings—regardless of the size or structure of the casting, the stability index is within the range of 0-1, making it comparable; Second, it facilitates setting grading thresholds—thresholds such as 0.70 and 0.50 have intuitive percentage meanings (70 points, 50 points), making them easy to understand and manage; Third, it avoids dimensional conflicts—the four input parameters have different dimensions (K, μm / s, ppm, s), making them impossible to add or average directly, but they can be comprehensively calculated after dimensionless conversion.

[0216] , , , The sensitivity coefficients for each defect factor are dimensionless and range from 0.65 to 0.75. They are determined by referring to literature data on the influence of each defect factor on the service performance of gray cast iron castings and by combining field process tests.

[0217] The sensitivity coefficient reflects the degree of influence of the corresponding defect factor on the overall stability. The larger the value, the stronger the degeneration of this factor has on stability. The four factors... All factors are set to a range of 0.65-0.75, and the range is the same. This design reflects the evaluation concept of equal weighting for each factor—undercooling, grain rate, impurities, and timing have equal importance to the casting quality and should not be favored over one another. Earlier versions used different sensitivity coefficients (e.g., undercooling 0.70, grain rate 0.65, impurities 0.75, timing 0.60), but this led to an imbalance in the dynamic range of individual factors—the degradation of some factors had a significant impact on the results, while others were ineffective. After verification, the range was unified to 0.65-0.75 to ensure a balanced dynamic range for the four factors. The specific value of 0.65-0.75 has been numerically verified: when… When a single factor reaches the reference deviation, its value is 0.25; when only one of the four factors is at its worst, the stability is... It falls within the first-level qualified range (≥0.70); when When the single-factor worst-case stability is This is also within a reasonable range. This range ensures that single-factor degradation will not directly lead to non-compliance, but simultaneous degradation of multiple factors will significantly reduce stability, which is consistent with actual quality patterns.

[0218] The maximum deviation of supercooling across the entire domain, in K, with a value range of 0K to 50K;

[0219] This is a reference deviation for undercooling, in K, taken as 50K, determined based on the maximum allowable deviation for undercooling of gray cast iron; The maximum deviation of supercooling across the entire domain, in K, with a value range of 0K to 50K; It represents the maximum deviation between the supercooling degree of all regions in the entire domain and the target value, expressed in K (Kelvin), with a value range of 0-50K. 0K indicates that the supercooling degree is completely uniform across the entire domain (ideal state), while 50K indicates that the supercooling degree deviation has reached the upper limit of the allowable range. The reference deviation of 50K is based on the maximum allowable deviation of undercooling in gray cast iron. When the undercooling deviation exceeds 50K, the microstructure differences between different regions become very significant (white iron may form in thin-walled areas, and coarse grains may form in thick-walled areas), making the casting quality unacceptable. Therefore, 50K is set as the reference deviation. When the factor reaches 50K, its value is (Approximately 0.25-0.35), which has a significant impact on stability.

[0220] The maximum absolute dispersion deviation of grain growth rate is expressed in μm / s, with a range of 0 μm / s to 20 μm / s. It represents the maximum absolute dispersion deviation of the grain growth rate at all points in the entire domain, i.e., the difference between the maximum and minimum rates (or the maximum deviation from the average rate), with units of μm / s and a range of 0-20 μm / s. 0 μm / s indicates that the grain growth rate is completely uniform across the entire domain (ideal state), while 20 μm / s indicates that the dispersion has reached the upper limit of allowable variation.

[0221] This is the reference deviation for grain rate, in μm / s, and is set to 20 μm / s. It is determined based on the allowable dispersion range of grain growth rate in gray cast iron.

[0222] The value of 20 μm / s is determined based on the allowable dispersion range of the grain growth rate in gray cast iron. Within this process parameter range, the grain growth rate is generally 5-40 μm / s. A global dispersion deviation exceeding 20 μm / s indicates excessive grain size variation (possibly with both fine and coarse grain regions), leading to severely uneven mechanical properties. When the speed reaches 20 μm / s, the factor value is This has a significant impact on stability. The expression of the maximum absolute discrepancy clarifies how the deviation is calculated—not as a relative deviation (percentage), but as an absolute deviation (μm / s), which is consistent with the physical meaning of grain rate.

[0223] The concentration increment of low-melting-point impurities in local enrichment, in ppm, with a value range of 0ppm to 200ppm, refers to the segregation increment of impurity concentration at grain boundaries and in the last solidified region relative to the average concentration of the matrix at the end of solidification. This is the increase in impurity concentration at grain boundaries and in the last solidified region relative to the average matrix concentration at the end of solidification, expressed in ppm (parts per million), ranging from 0 to 200 ppm. It is particularly important to emphasize the increase in impurity concentration rather than the absolute impurity content, because the absolute content in S1 is already controlled within a reasonable range (sulfur 0.05%-0.08%, i.e., 500-800 ppm; lead 0.002%-0.005%, i.e., 20-50 ppm). The increase in impurity concentration refers to the localized increase in concentration caused by solute redistribution during solidification, and this increase directly reflects the severity of grain boundary segregation.

[0224] The reference concentration for impurities is in ppm, taken as 200 ppm. The concentration should be determined as needed based on the allowable increase in the segregation of low-melting-point impurities in gray cast iron. The value of 200 ppm is determined based on the allowable increase in the segregation of low-melting-point impurities in gray cast iron. When the increase in the segregation of grain boundary impurities exceeds 200 ppm, the grain boundary weakening effect is significant, and the high-temperature strength and resistance to hot cracking of the casting decreases sharply.

[0225] The maximum deviation at the start of the phase transition, in seconds, with a range of 0s to 30s;

[0226] This is a timing reference deviation, in seconds (s), taken as 30 seconds, determined based on the allowable deviation range of the phase transformation initiation time for thin-walled castings.

[0227] It represents the maximum deviation in the onset time of the eutectic phase transition in all regions of the entire domain, expressed in seconds (s), with a value range of 0-30s. 0s indicates that the entire domain begins eutectic solidification simultaneously (ideal state), while 30s indicates that the phase transition timing deviation has reached the upper limit of the allowable range. The 30s value is determined based on the allowable deviation range of the phase transformation initiation time in thin-walled castings—the overall solidification time of thin-walled castings is typically only a few minutes. A deviation of more than 30s in the phase transformation initiation time between adjacent regions indicates severe asynchrony in solidification, resulting in large differences in thermal stress and shrinkage between the first-solidified and later-solidified regions, which can easily lead to hot cracks and deformation. When it reaches 30 seconds, the factor value is This has a significant impact on stability.

[0228] The above reference values ​​can be calibrated and adjusted before production by combining process tests with blank temperature measurement results, based on the casting wall thickness and structural characteristics. , , , It is not fixed, but can be calibrated and adjusted according to the wall thickness and structural characteristics of the casting.

[0229] When any monitoring parameter exceeds the corresponding reference value, the solidification stability is directly judged as unqualified, and the formula calculation is no longer performed.

[0230] g3. Based on the calculation results, when It is judged as Level 1 qualified, when It was determined to be Level 2 qualified. The product is then deemed defective and transferred to the remelting process.

[0231] This sub-procedure specifies the grading criteria for solidification stability. The threshold settings for the three grades (0.70 and 0.50) have undergone rigorous numerical verification: when all four factors are optimal (deviation = 0), stability = 1.0; when a single factor reaches the reference deviation (… Stability = Just exceeding the Level 1 pass threshold of 0.70 means that even if a single dimension reaches the upper limit of the allowable range, it can still be judged as Level 1 pass; when both factors reach the reference deviation simultaneously, the stability = The value falls precisely on the boundary between Level 2 qualified and unqualified, meaning that if two dimensions simultaneously reach the allowable upper limit, the product is on the verge of being qualified. When three or four factors reach the reference deviation, the stability values ​​are 0.354 and 0.25 respectively, both below 0.50, and the product is judged as unqualified. The design logic of this grading system is as follows: Level 1 qualified (≥0.70) corresponds to high-quality products with good performance in all dimensions, with at most one dimension approaching the upper limit; Level 2 qualified (0.50-0.70) corresponds to qualified products with certain defects but within the tolerance range; Unqualified (<0.50) corresponds to scrap products with serious defects in multiple dimensions, which are transferred to remelt. This three-level classification is more refined than the traditional binary classification of qualified / unqualified—Level 1 products can be used for high-requirement conditions (such as high-pressure hydraulic systems), Level 2 products are used for general conditions, and unqualified products are remelted, achieving refined quality management and rational utilization of resources.

[0232] g4. After associating the evaluation results with the casting batch number, store them in the production database to form a traceable record of solidification quality parameters for subsequent iterative updates of the molding control dataset.

[0233] Specifically, this scheme further defines the quantitative evaluation method for the solidification and forming quality of castings. Its design aims to establish a unified evaluation index that comprehensively reflects the deviation of multiple key parameters in the solidification process, upgrading the traditional quality judgment based on experience and visual inspection to a standardized evaluation system that is quantifiable, traceable, and gradeable. Four parameters are summarized: maximum deviation of global undercooling, maximum absolute dispersion deviation of grain growth rate, increase in local enrichment concentration of low-melting-point impurities, and maximum deviation of phase transformation initiation time. These four parameters characterize the stability and uniformity of the solidification process from four different perspectives: thermal undercooling, grain kinetics, solute segregation, and phase transformation sequence. A single parameter cannot fully reflect the forming quality; the comprehensive evaluation of these four parameters can avoid misjudgments caused by accidental fluctuations in any single parameter. A geometric mean evaluation model for solidification stability is adopted, which uses a geometric mean to calculate the normalized deviations of the four parameters. The geometric mean, rather than the arithmetic mean, is chosen because it is more sensitive to extreme deterioration of any parameter: when a parameter deviates significantly, the geometric mean drops significantly, allowing for timely detection of severe local defects; whereas the arithmetic mean might mask severe defects in a single parameter due to good performance in other parameters. The weighting coefficients are set between 0.65 and 0.75, ensuring that solidification stability remains within a reasonable range of 0.45 to 0.78 when any single parameter reaches its upper limit of the reference value, thus avoiding evaluation distortion caused by imbalances in the dynamic range of single factors. When any monitored parameter exceeds its corresponding reference value, it is directly deemed unqualified without further calculation. This design sets a hard safety baseline for the evaluation system, preventing the logical contradiction that a parameter may be severely out of bounds while the geometric mean remains within the acceptable range. The evaluation results are divided into three levels: Level 1 (qualified) with an S value of not less than 0.70, Level 2 (qualified) with an S value between 0.50 and 0.70, and Level 3 (unqualified) with an S value below 0.50, requiring remelting. This provides a clear grading standard for production quality management. Level 1 products can be used in high-reliability applications, Level 2 products can be used in general applications, and unqualified products are promptly returned to the furnace to prevent them from flowing into the next process and causing greater losses. The evaluation results are linked to the batch number and stored in the production database for quality traceability and iterative optimization of the dataset. This ensures that the evaluation data for each batch can continuously improve process parameters, forming a virtuous cycle of quality evaluation and process optimization.

[0234] Furthermore, the distribution of equivalent grain diameters was statistically analyzed using image analysis methods on the above metallographic samples after etching. The results are as follows: Figure 3 As shown. By Figure 3 It is evident that the comparative example exhibits a wider grain size distribution range, displaying a coexistence of fine and coarse grains. This corresponds to the significant differences in cooling rates and high dispersion in grain growth rates across different structural regions in the comparative example. In contrast, the example's grain size distribution range is significantly narrower, exhibiting a more concentrated single-peak distribution, indicating a more uniform grain size across the entire domain. These results are consistent with... Figure 2The distribution characteristics of the grain growth rate shown are consistent, which further verifies that the present invention can effectively achieve uniform control of the grain size across the entire domain by adjusting the cooling air volume of the sand mold through feedback of the discrete parameters of the grain growth rate during the eutectic solidification stage, and avoid the polarization of grain size caused by differences in heat dissipation of the structure.

[0235] Preferably, the feedback adjustment of the sand mold cooling air volume in step S4 is executed using solidification synergistic deviation convergence control logic, and is performed according to the following steps:

[0236] h1. Real-time acquisition of three parameters: maximum deviation of global undercooling, maximum absolute discrete deviation of grain growth rate, and maximum deviation of phase transformation initiation time. These parameters are divided by their respective calibration reference values ​​for normalization, and then weighted and summed according to preset weights (the sum of the three weights is normalized to 1). The square root is then taken to obtain the current solidification coordination deviation index. The deviation index ranges from 0 to 1. The closer it is to 0, the more uniform the solidification state.

[0237] The preset weight values ​​are all in the range of 0.2 to 0.5, determined according to the degree of influence of each deviation factor on the quality of the casting, and calibrated by process test combined with blank temperature measurement results before the equipment is put into production;

[0238] The calibration reference values ​​are determined based on the upper limit of the allowable deviation of each parameter of gray cast iron, specifically: supercooling calibration reference 40K~50K, grain rate calibration reference 15μm / s~20μm / s, and phase transformation initiation time calibration reference 25s~30s.

[0239] h2. Calculate the adaptive control gain based on the low-melting-point impurity content of the molten iron detected at the furnace. The higher the impurity content, the greater the predicted tendency for agglomeration at the end of solidification, and the greater the gain. The gain value range is 0.1m. 3 / min~0.5m 3 / min, determined based on the thermal inertia of the sand mold and the air volume response lag time;

[0240] By combining the solidification synergy deviation index and the deviation change rate, and superimposing the differential advance adjustment amount, where the differential action time coefficient is taken from 0.5s to 2.0s and determined according to the cooling system response time constant, the airflow adjustment increment is calculated, with the adjustment increment value ranging from -1.0m. 3 / min~1.0m 3 / min;

[0241] In addition, for the structural thermal section region, 0m needs to be superimposed. 3 / min~0.3m 3 The feedforward compensation amount is preset based on the hot spot volume and structural characteristics of the casting;

[0242] h3. After superimposing the incremental air volume adjustment with the reference air volume, the output is constrained by the upper and lower limits and the single-step change rate and sent to the variable frequency fan to adjust the sand mold surface purging and cooling intensity.

[0243] The baseline air volume is 0.5m³. 3 / min~1.5m 3 / min, air volume output range 0m 3 / min~3.0m 3 / min, with the maximum rate of change per step controlled at 0.1m. 3 / ~0.3m 3 / ;

[0244] h4. Repeat steps h1 to h3 every 1 to 5 seconds until the casting enters the final stage of solidification with solid-liquid coexistence, then stop the control logic operation and switch to the closed pressure stabilization control in step S5.

[0245] Specifically, this scheme further defines the solidification coordination deviation convergence control logic in the eutectic solidification stage. Its design aims to integrate three key process parameters—undercooling deviation, grain rate dispersion deviation, and phase transformation initiation time deviation—and combine them with adaptive adjustment of furnace front impurity content and differential advance control to achieve multi-variable coordinated optimization adjustment of sand mold cooling air volume, avoiding potential conflicts and adjustment lags that may occur when adjusting single parameters independently. The maximum deviation of global undercooling, the maximum absolute dispersion deviation of grain rate, and the maximum deviation of phase transformation initiation time are acquired in real time. These are then normalized by dividing by a calibration benchmark value, weighted by preset weights, and finally square-rooted to obtain the solidification synergistic deviation index. Normalization eliminates the influence of differences in the dimensions and numerical ranges of the three parameters, allowing parameters with different physical meanings to be integrated on a unified scale. The weights are set between 0.2 and 0.5, with a weight sum of 1. These weights can be determined before production through process experiments combined with blank temperature measurement calibration, based on the casting structure characteristics and quality priorities. Parameters with a greater impact on forming quality are assigned higher weights. The square-root processing makes the deviation index more sensitive to large deviations, enhancing its responsiveness to abnormal operating conditions. The adaptive control gain is calculated based on the low-melting-point impurity content of the molten iron before the furnace; the higher the impurity content, the greater the gain, with a value range of 0.1m. 3 / min to 0.5m 3The airflow rate is set at 0.5 m³ / min because molten iron with high impurity content has a greater tendency for solute segregation during solidification and is more sensitive to changes in cooling rate, requiring a larger airflow adjustment range to maintain solidification uniformity. Using the impurity content, which can be measured in real-time before the furnace, rather than the segregation increment that can only be detected after solidification, as the gain input ensures real-time adjustment. Combining the deviation index and the deviation change rate with a differential lead adjustment, and using a differential time coefficient ranging from 0.5 s to 2.0 s, the airflow adjustment can predict the deviation change trend and apply the adjustment amount in advance, overcoming the inherent lag of pure proportional adjustment. The airflow adjustment increment is superimposed on a reference airflow of 0.5 m³ / min. 3 / min to 1.5m 3 / min, with upper and lower limits from 0 to 3.0m 3 / min and single-step rate of change constraint 0.1m 3 / up to 0.3m 3 / The output is then sent to the variable frequency fan. The limiting and rate of change constraints prevent sudden changes in air volume from impacting the solidification process. The above calculation and adjustment are repeated every 1 to 5 seconds until the end of the solid-liquid coexistence period, at which point the system stops and switches to S5 closed-loop pressure stabilization control, thus achieving continuous dynamic coordinated regulation of the entire eutectic solidification process.

[0246] Preferably, each step of the process is executed sequentially by an industrial computer according to a preset control logic. The computer has a built-in global differentiated forming control dataset constructed in step S2, and receives data on impurity content, cavity temperature, cavity pressure, and molten iron flow rate collected by each sensor in real time. The data acquisition and control command update is performed every 1 to 5 seconds.

[0247] A graded alarm is triggered when the monitored data exceeds the preset control range. The deviation ratio is calculated based on the width of the preset control range corresponding to the parameter.

[0248] A Level 1 alarm occurs when the parameter deviation exceeds the allowable range but does not exceed 10%. The computer will automatically fine-tune the corresponding process parameters and record the adjustment log.

[0249] A level 2 alarm occurs when the parameter deviation exceeds the allowable range by more than 10% but not more than 30%. The computer will pause the current process and display a manual confirmation interface, waiting for the operator to review the information before deciding whether to continue or terminate the process.

[0250] A Level 3 alarm occurs when the parameter deviation exceeds the allowable range by more than 30%. The computer terminates the pouring process and initiates the emergency remelting procedure, while simultaneously sending an alarm message to the management personnel.

[0251] In addition, process data for each process is automatically archived according to the casting batch number, with a retention period of no less than 3 years, supporting subsequent quality traceability and iterative optimization of the forming control dataset.

[0252] Specifically, this solution further defines the computer control implementation and hierarchical alarm methods for the process. Its design aims to integrate parameter acquisition, logical operations, execution output, and anomaly handling of the aforementioned steps into the industrial computer's automated control system. This ensures the accuracy, repeatability, and traceability of process execution. Simultaneously, a hierarchical alarm system employs differentiated handling strategies for different degrees of process deviation, balancing production efficiency and quality safety. The industrial computer executes steps S1 to S6 sequentially according to preset logic. It has a built-in global differentiated forming control dataset and receives real-time monitoring data such as furnace impurity content, cavity temperature, cavity pressure, and molten iron flow rate. Data is collected and updated every 1 to 5 seconds. This acquisition frequency matches the dynamic change rate during the eutectic solidification stage: a frequency that is too low cannot capture parameter fluctuations during rapid phase transformation, while a frequency that is too high increases data storage and computational burden without any practical gain. The tiered alarm system calculates the deviation percentage based on the control range width. A Level 1 alarm occurs when the deviation exceeds the allowable range but does not exceed 10%. In this case, the system automatically fine-tunes relevant parameters and records the changes without manual intervention, ensuring production continuity. A Level 2 alarm occurs when the deviation exceeds 10% but does not exceed 30%. In this case, the current process is paused, and a manual confirmation interface pops up, allowing the operator to decide whether to continue or adjust parameters before resuming, preventing further deviation. A Level 3 alarm occurs when the deviation exceeds 30%. In this case, pouring is immediately terminated, and an emergency remelting procedure is initiated, along with an audible and visual alarm, preventing the continued production of defective castings and the resulting waste of raw materials and time. This three-tiered progressive alarm strategy avoids frequent production interruptions due to minor deviations while ensuring that serious deviations are promptly addressed. All process data is archived by batch and retained for at least three years, meeting regulatory requirements and industry practices for casting product quality traceability. When quality problems arise during subsequent service, the complete process data for the corresponding batch can be retrieved for cause analysis and responsibility determination. The accumulated historical data also provides a data foundation for continuous optimization of process parameters.

[0253] Preferably, the thin-walled casting is a thin-walled cast iron casting of a hydraulic valve body, including any one of a multi-way valve housing, a hydraulic steering gear housing, a gear pump housing, a plunger pump housing, and a hydraulic motor housing;

[0254] The dimensions of the casting blank, the layout of the oil circuit, and the forming mold all adopt the existing fixed design, and the casting structure and mold cavity are not changed during the casting process.

[0255] Specifically, this solution further defines the types of casting products to which the process of this invention is applicable. Its design purpose is to clarify the application objects and boundaries of this invention, emphasizing that this invention is a pure process improvement scheme for existing standardized thin-walled cast iron hydraulic valve bodies, without involving any changes to the casting structure, wall thickness, oil circuit layout, or forming mold. This clearly distinguishes it from schemes that improve performance by changing the casting structure. This invention is applicable to any type of thin-walled casting among multi-way valve housings, hydraulic steering gear housings, gear pump housings, piston pump housings, and hydraulic motor housings. These products are standard, general-purpose, standardized parts in the hydraulic systems of engineering machinery and agricultural machinery. Their structural dimensions, oil circuit layout, and wall thickness are fixed according to industry standards and overall machine requirements. Manufacturers typically use fixed drawings and standardized molds for mass production, lacking the conditions and space to arbitrarily modify the structure.

[0256] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention. For those skilled in the art, any alternative improvements or transformations made to the implementation of the present invention fall within the protection scope of the present invention.

[0257] Any aspects of this invention not described in detail are well-known to those skilled in the art.

Claims

1. A lightweight casting forming and control process for thin-walled castings, characterized in that, Includes the following steps: S1. After the molten iron is smelted, it undergoes composite rare earth passivation modification. The content of low melting point impurities is obtained by pre-furnace spectral detection. The solute distribution is adjusted according to the results to limit the impurity content to a preset metallurgical control range. S2. Accept the molten iron after processing and pour it. Before pouring, collect the no-load heat dissipation temperature of each area of ​​the mold, predict the undercooling and build an initial forming control dataset. After pouring, collect the actual temperature, calculate the actual undercooling and correct it to form a global differentiated forming control dataset that is suitable for the irregular structure of thin-walled castings. S3. During the filling process, the initial molding control dataset is called, the inoculant addition parameters are matched according to the predicted supercooling of each region, and the time-series differentiated in-flow inoculant is executed. At the same time, the filling parameters are monitored and the addition sequence is corrected. S4. After the casting enters the eutectic solidification stage, the eutectic solidification time is locked based on the cavity temperature monitoring. At the same time, the global grain growth rate is calculated and the discrete parameters are statistically analyzed. The sand mold cooling air volume is adjusted according to the time deviation and the rate dispersion. S5. When the casting enters the final stage of solid-liquid coexistence, the exhaust channel is closed to form a closed environment. The cavity pressure is maintained slightly positive according to the cavity pressure monitoring value to inhibit the secondary migration and local enrichment of low melting point solutes at the solid-liquid interface. S6. After the casting has solidified, a zoned gradient slow-release cooling process is immediately implemented. The temperature decay rate is collected throughout the process and the cooling intensity is adjusted to obtain a shaped thin-walled casting.

2. The process according to claim 1, characterized in that, Step S1 is performed as follows: a1. Prepare a composite passivating agent by mixing rare earth ferrosilicon alloy and silicon-calcium alloy at a mass ratio of 3:1, wherein the particle size of the passivating agent is controlled at 1mm to 5mm, and dry it at 200℃ to 300℃ for 1h to 2h. a2. Before tapping the molten iron, an oxygen probe is used to detect the oxygen activity of the molten iron, and the oxygen activity is controlled at 2×10⁻⁶. -6 ~8×10 -6 Within the specified range, when the oxygen activity exceeds the specified range, the oxygen activity is adjusted by adding a passivating agent. a3. Simultaneously, a furnace-front laser-induced breakdown spectrometer is used to detect the mass fraction of sulfur and lead low-melting-point impurities in molten iron. Each furnace is tested no less than twice. For critical batches, an online continuous monitoring mode is used with a monitoring interval of 20s to 40s. The sulfur mass fraction is controlled at 0.05% to 0.08%, and the lead mass fraction is controlled at 0.002% to 0.005%. When the impurity content exceeds the upper limit of the control range, the composite passivating agent prepared in step a1 is added at 0.05% to 0.10% of the molten iron mass, and the amount added at one time shall not exceed 0.15% of the molten iron mass, until the impurity content falls back to the control range.

3. The process according to claim 1, characterized in that, Step S2 is performed as follows: b1. K-type armored thermocouples are pre-embedded in the inner wall of the mold cavity corresponding to the hot spot area, thin-walled plane area and thick-thin transition area of ​​the casting. The distance between the thermocouple measuring point and the surface of the mold cavity is controlled between 2mm and 5mm, the response time is not greater than 0.5s, and the temperature measurement range meets the short-term temperature measurement requirements of 0℃ to 1600℃. b2. Before casting, perform a blank temperature measurement process to collect the no-load cooling curves of each area. The cooling rate is corrected in segments according to three temperature ranges: below 200℃, 200℃~600℃, and above 600℃. A correction model of the heat dissipation characteristics of the sand mold in each area is established. Based on this, the solidification undercooling degree and phase change initiation time of each area are predicted, and an initial molding control dataset is constructed. After casting, temperature curves of each region are collected. Based on the heat dissipation characteristics, the model is modified to back-derive the molten iron cooling curve. The difference between the equilibrium eutectic temperature and the actual eutectic solidification initiation temperature is extracted to calculate the actual solidification undercooling. The calculated undercooling value ranges from 5K to 40K. The initial forming control dataset is corrected based on the actual undercooling. The revised dataset is used for both the control of subsequent cooling processes in this batch and the iteration of the forming control dataset for the next batch. b3. Subsequently, the castings are divided into three levels according to the supercooling value: high supercooling zone (25K~40K), medium supercooling zone (15K~25K), and low supercooling zone (5K~15K). The inoculant dosing acceleration rate range and cooling air volume range are set for each zone respectively, and finally a global differentiated molding control dataset is formed.

4. The process according to claim 1, characterized in that, Step S3 is performed as follows: c1. First, call the initial forming control dataset and match the molten iron mass flow rate of 0.20%~0.30%, 0.10%~0.20%, and 0.05%~0.10% of the 75% ferrosilicon inoculant addition acceleration rate for the high subcooling zone, medium subcooling zone, and low subcooling zone, respectively, with the inoculant particle size controlled between 0.2mm and 1.0mm. c2. Adjust the inoculant dosing acceleration rate in real time according to the flow velocity of the molten iron stream. Specifically, for every 0.1 m / s increase in flow velocity, the inoculant dosing acceleration rate should be increased by 1% to 3%. c3. At the same time, based on the arrival time of molten iron in each area, the inoculant addition start time in the corresponding area is advanced by 0.5s to 1.0s to ensure that the inoculant and molten iron enter each area of ​​the mold cavity synchronously. c4. After the filling process is completed, the inflection point of the eutectic phase transition initiation temperature is collected in real time by pre-embedded thermocouples in each area, and the actual phase transition initiation time of each area is calculated. When the actual phase change start time of a certain area deviates from the predicted time in the dataset by more than 5 seconds, it is determined that there is a deviation in the inoculation control parameters. The deviation data is recorded, and the inoculation agent dosing acceleration rate is increased or decreased according to the direction of the deviation. The control parameters of the corresponding area in the production database are then corrected for the next batch molding control dataset iteration.

5. The process according to claim 1, characterized in that, Step S4 is performed as follows: d1. The temperature curve is collected in real time by pre-embedded thermocouples in the cavity. The duration of the eutectic reglow temperature plateau is locked between 20s and 60s. When the monitoring duration is less than 20s, the cooling air volume of the sand mold surface is reduced, and when it is more than 60s, the cooling air volume is increased. d2. Simultaneously, control the cooling rate during the eutectic stage to remain within the range of 0.5℃ / s to 2.0℃ / s; d3. The grain growth rate is calculated using a thin-walled valve body eutectic solidification grain growth rate prediction model. The interface radius of curvature is estimated based on the current cooling rate and undercooling value. The larger the cooling rate and the larger the undercooling, the smaller the radius of curvature. The estimated value range is 1μm to 100μm. When the estimated value exceeds this range, the boundary value is substituted into the calculation. ； In the formula, The values ​​represent the calculated grain growth rate, with a parameter range of 5 μm / s to 40 μm / s. This is a frequency factor, with a value range of 1.0 × 10⁻⁶. 4 μm / ~1.5×10 4 μm / ; The solidification activation energy has a value range of 1.0 × 10⁻⁶. 5 J / mol ~ 1.5 × 10 5 J / mol; It is the gas constant; This is the solid-liquid interface temperature, with a value ranging from 1390K to 1425K. This represents the actual undercooling during eutectic solidification, with a value ranging from 5K to 40K. These are the Gibbs-Thomson coefficients, ranging from 0.1 K.μm to 0.5 K.μm; Based on the calculation results, the relative dispersion parameter of grain growth rate at each point in the entire region is statistically analyzed. When the relative deviation of the grain growth rate at each point from the average value of the entire region exceeds +15%, the sand mold purging air volume in the corresponding area is reduced by 0.1 m³. 3 / min~0.5m 3 / min; Conversely, when the relative deviation between the grain growth rate at each point and the global average exceeds -15%, the sand mold purging airflow in the corresponding area is increased by 0.1 m³. 3 / min~0.5m 3 / min, so that the deviation of the global grain growth rate converges to within ±15% of the average rate.

6. The process according to claim 1, characterized in that, Step S5 is performed as follows: e1. Lay a high-temperature resistant aluminum silicate sealing gasket on the sand mold parting surface, wherein the thickness of the sealing gasket is controlled at 2mm to 4mm and the compression is controlled at 30% to 50% to ensure that the sealing gap of the parting surface is not greater than 0.1mm. e2. When the cavity temperature drops to 5℃~20℃ below the equilibrium eutectic temperature, it is determined that the casting has entered the final stage of solidification with solid-liquid coexistence, and all exhaust channels are closed to form a sealed environment inside the cavity. e3. The cavity pressure is monitored in real time by the cavity pressure sensor, and the cavity micro-positive pressure is maintained at 5kPa to 15kPa. This pressure is dynamically balanced by the inert gas pressure replenishment, residual gas generation of the coated sand and cavity venting. When the pressure is below 5 kPa, the pressure is replenished through the inert gas inlet valve; conversely, when the pressure is above 15 kPa, the pressure is released by opening the micro exhaust valve. e4. The sealed pressure maintenance time should be controlled between 8 and 15 minutes. During this period, it is forbidden to open the sand mold or disturb the casting.

7. The process according to claim 1, characterized in that, Step S6 is performed as follows: f1. First, perform gradient slow-release cooling according to the global subcooling benchmark zone calibrated in step S2. In the high subcooling zone, heat preservation slow cooling is adopted, and the cooling rate is controlled at 0.05℃ / s~0.1℃ / s. The intermediate subcooling zone uses weak airflow cooling at a rate of 0.1℃ / s to 0.3℃ / s; The low subcooling zone uses forced air cooling at a rate of 0.3℃ / s to 0.8℃ / s; f2. Throughout the cooling process, temperature decay data is collected in real time by pre-embedded thermocouples in each area. When the temperature difference between adjacent areas is greater than 80℃, the cooling process on the side with higher cooling intensity in the area with the excessive temperature difference is suspended. The cooling process on that side is resumed after the temperature difference drops to within 50℃. f3. When the overall temperature of the casting drops below 200℃, the zoned cooling control is released and the casting is switched to air cooling to room temperature. The temperature of the casting after unpacking should not exceed 150℃.

8. The process according to claim 1, characterized in that, After step S6 is completed, the solidification stability evaluation process of the casting is also included, which is performed according to the following steps: g1. Summarize four monitoring parameters collected throughout the molding process: undercooling deviation, maximum absolute dispersion deviation of grain growth rate, local enrichment concentration increment of low melting point impurities, and deviation of phase transformation initiation time. g2. Subsequently, the solidification stability was calculated using the geometric mean evaluation model for solidification stability: ； In the formula, This is a solidification stability evaluation index, with a value ranging from 0 to 1; , , , The sensitivity coefficients for each defect factor range from 0.65 to 0.

75. This represents the maximum deviation of supercooling across the entire domain, with a value ranging from 0K to 50K. This is the reference deviation for supercooling. The maximum absolute dispersion deviation of the grain growth rate is denoted as 0 μm / s to 20 μm / s. The reference deviation for grain rate is set to 20 μm / s; This represents the localized concentration increase of low-melting-point impurities, with a value ranging from 0 ppm to 200 ppm. The reference concentration for impurities is 200 ppm. The maximum deviation at the start of the phase transition is represented by a value ranging from 0s to 30s. This is the timing reference deviation; The above reference values ​​can be calibrated and adjusted before production by combining process tests with blank temperature measurement results, based on the casting wall thickness and structural characteristics. When any monitoring parameter exceeds the corresponding reference value, the solidification stability is directly judged as unqualified, and the formula calculation is no longer performed. g3. Based on the calculation results, when Determined to be Level 1 qualified, when It was determined to be Level 2 qualified. The product is then deemed defective and transferred to the remelting process. g4. After associating the evaluation results with the casting batch number, store them in the production database to form a traceable record of solidification quality parameters for subsequent iterative updates of the molding control dataset.

9. The process according to claim 1, characterized in that, The feedback adjustment of the sand mold cooling air volume in step S4 is executed using solidification synergistic deviation convergence control logic, and is performed according to the following steps: h1. Real-time acquisition of three parameters: maximum deviation of global undercooling, maximum absolute discrete deviation of grain growth rate, and maximum deviation of phase transformation initiation time. These parameters are divided by their respective calibration reference values ​​for normalization, and then weighted and summed according to preset weights. The square root is then taken to obtain the current solidification coordination deviation index. The deviation index ranges from 0 to 1. The closer it is to 0, the more uniform the solidification state is. The preset weight values ​​are all in the range of 0.2 to 0.5, determined according to the degree of influence of each deviation factor on the quality of the casting, and calibrated by process test combined with blank temperature measurement results before the equipment is put into production; The calibration reference values ​​are determined based on the upper limit of the allowable deviation of each parameter of gray cast iron, specifically: supercooling calibration reference 40K~50K, grain rate calibration reference 15μm / s~20μm / s, and phase transformation initiation time calibration reference 25s~30s. h2. Calculate the adaptive control gain based on the low-melting-point impurity content of the molten iron detected at the furnace front. The gain value range is 0.1m. 3 / min~0.5m 3 / min, determined based on the thermal inertia of the sand mold and the air volume response lag time; By combining the solidification synergy deviation index and the deviation change rate, and superimposing the differential advance adjustment amount, where the differential action time coefficient is taken from 0.5s to 2.0s and determined according to the cooling system response time constant, the airflow adjustment increment is calculated, with the adjustment increment value ranging from -1.0m. 3 / min~1.0m 3 / min; h3. After superimposing the incremental air volume adjustment with the reference air volume, the output is constrained by the upper and lower limits and the single-step change rate and sent to the variable frequency fan to adjust the sand mold surface purging and cooling intensity. h4. Repeat steps h1 to h3 every 1 to 5 seconds until the casting enters the final stage of solidification with solid-liquid coexistence, then stop the control logic operation and switch to the closed pressure stabilization control in step S5.

Images