Manufacturing method and system of large-size shell castings based on double-station shell pattern line

Abstract

The present application relates to the technical field of coated sand shell casting, in particular to a large-size shell casting manufacturing method and system based on a double-station shell line, comprising the following steps: S1, real-time monitoring of double-station environmental parameter fluctuation; S2, real-time monitoring of shell deformation state and data correction; S3, pre-calibration and deviation correction of split shell reference alignment, fixation; S4, joint gap and port sealing covering, adjustment of sealing measures to maintain the closed and isolated state of the cavity; S5, layered flow limiting pouring according to the structure characteristics of the casting, real-time control of pouring process parameters; S6, completion of casting solidification control; S7, completion of shell disassembly, casting finishing and quality detection, rotation of division of labor, and maintenance of line continuous circulation. The present application realizes stable improvement of double-station timing adaptation degree, improvement of line production efficiency, and significant improvement of workpiece transfer smoothness, which can effectively adapt to the large-size casting scale and continuous production demand, and the process coordination and line operation efficiency are improved.

Description

Technical Field

[0001] This invention relates to the field of coated sand shell casting technology, and in particular to a method and system for manufacturing large-size shell castings based on a dual-station shell profile. Background Technology

[0002] Traditional high-pressure axial piston pump closed-cell housings are mostly produced using a single-station coated sand molding process. Although a few production lines have introduced dual-station shell molding lines to achieve parallel operation of shell making and casting processes, the existing dual-station architecture mostly adopts the fixed cycle mode of small and medium-sized castings. When machining large-sized hydraulic housings with closed-cell cavities and multiple high-pressure oil passages, the following shortcomings exist:

[0003] On the one hand, existing dual-station shell molding lines only use a fixed process cycle to arrange the operation sequence, which cannot be dynamically matched to the fluctuations in workshop temperature and humidity. After demolding, closed box-shaped shells have non-uniform shrinkage characteristics in the warp and weft directions. Traditional processes use a uniform resting time and a fixed linear shrinkage coefficient to predict deformation, which cannot adapt to the differentiated shrinkage patterns of different directions of the box-shaped shell. This can easily cause uneven gaps between shell joint surfaces and misalignment of reference points. Subsequent pouring can easily lead to oil channel forming deviation and assembly dimension deviation. At the same time, the lack of staggered peak avoidance between the two stations means that the shells after shell making are left to wait for pouring for a long time, which can further induce secondary deformation in thin-walled areas.

[0004] On the other hand, the existing process only uses a simple sealing method to seal the gaps between the box and shell, and subsequent visual leak detection cannot identify tiny hidden leaks, resulting in a high failure rate of high-pressure shells during pressure resistance tests.

[0005] Furthermore, most castings adopt a uniform heat dissipation mode after filling. However, when this method is applied to the high-pressure oil passage of a box-shaped shell, shrinkage and loose structure defects are prone to occur in the hot spot area, which cannot meet the high density and high pressure resistance sealing requirements of hydraulic castings.

[0006] Therefore, it is necessary to develop a dual-station shell profile manufacturing method and system that is compatible with large-size closed-type hydraulic housings. Summary of the Invention

[0007] To solve one of the above-mentioned technical problems, the present invention adopts the following technical solution: a method for manufacturing large-size shell castings based on a dual-station shell profile, comprising the following steps: Step S1, collecting and removing anomalies from the on-site environmental parameters of the casting workshop to form an effective environmental monitoring dataset; completing the dual-station layout division, timing operation arrangement parameter calibration, and workpiece transfer path planning between stations, and monitoring the fluctuation of dual-station environmental parameters in real time.

[0008] Step S2: Demolding of large-size shells, completing multi-dimensional deformation detection and data recording of the shells; placing the shells in sections for static placement, real-time monitoring of shell deformation status and data correction.

[0009] Step S3: Perform pre-calibration and deviation correction of the split shell reference alignment, complete the pre-tightening and locking of the shell assembly, check the assembly accuracy and form and position parameters, and achieve overall rigid fixation of the shell.

[0010] Step S4: Seal and cover the gaps and ports of the shell splicing, set up monitoring points for the internal cavity environment, collect internal cavity environment parameters in real time, and adjust the sealing measures to maintain the sealed and isolated state of the cavity.

[0011] Step S5: After the inner cavity of the shell is sealed, a pouring rhythm is generated, and layered flow-limited pouring is implemented according to the structural characteristics of the casting, while the pouring process parameters are controlled in real time.

[0012] Step S6: After the molten metal is filled into the mold, the wall thickness distribution of the casting is detected and the solidification control area is divided. Gradient heat dissipation parameters are set according to the area to complete the solidification control of the casting.

[0013] Step S7: After confirming that the casting has completely solidified and solidified, complete the shell disassembly, casting trimming and quality inspection, rotate the dual-station process division, and call the cycle calibration parameters to maintain the continuous flow of the production line.

[0014] As a preferred option, the specific steps for environmental parameter collection and screening, and dual-station layout control in step S1 are as follows:

[0015] Collect ambient temperature, relative humidity, dust concentration and air pressure parameters in the workshop. Collect no less than 10 sets of data continuously, remove abnormal data, and select effective data with ambient temperature of 15℃~40℃, relative humidity of 38%~75%, dust concentration ≤50mg / m³, and air pressure of 80kPa~105kPa to form an effective environmental monitoring dataset.

[0016] Divide the work into two independent work units, with the unit spacing controlled at 3m to 5m;

[0017] Dual-station timing job layout parameters are calibrated using a timing adaptation coupling calculation formula; dual-station timing adaptation degree is calculated. Adjust the timing operation scheduling parameters based on the calculation results to ensure... ≥0.85;

[0018] Survey the layout of workstations and equipment distribution, and delineate directional transfer paths for workpieces between workstations with a width of 1.5m to 2.0m to avoid blind spots in equipment operation;

[0019] Monitor the fluctuations of environmental parameters in the two sets of work units in real time at a frequency of no less than once every 5 minutes to ensure that the timing parameters are compatible.

[0020] The timing adaptation coupling calculation formula is as follows:

[0021] ;

[0022] In the formula: For dual-station timing adaptation; This is the correction factor for stable operating conditions; The average process time for a single batch at the shell-making station; The average process time per batch at the pouring and cooling station; The difference in ambient temperature between the two work units; The relative humidity difference between the two sets of work units; This refers to the dynamic coupling factor of environmental parameters. This is the timing lag correction factor; This is a correction factor for the thermal field superposition at the workstation.

[0023] As a preferred embodiment, the specific steps of shell demolding, deformation detection, and static control in step S2 are as follows:

[0024] The large-sized shell molded by solidification is demolded using a parallel lifting method, with the lifting speed controlled at 0.1m / s to 0.2m / s, and the lifting points are symmetrically distributed at the edge of the shell.

[0025] Test points are set up in stress concentration areas such as the flange connection of the shell and the area around the oil passage, with a density of 4 to 8 points per square meter. A dial indicator is used to test the linear deformation values ​​in the length, width and height directions of the shell, with the test accuracy controlled at 0.01 mm. The test data are recorded to form a deformation parameter dataset.

[0026] According to the production rhythm, the shell-shaped areas are neatly arranged and placed in stillness. The ambient temperature of the stillness area is controlled at 18℃~35℃ and the relative humidity is controlled at 40%~70%. The stillness area is far away from heat sources and areas with airflow disturbance.

[0027] During the static setting process, the shell deformation is detected every 10 minutes to generate a real-time deformation monitoring dataset.

[0028] The shell-shaped directional shrinkage amount is calculated using a multi-factor linkage correction formula, as follows:

[0029] ;

[0030] In the formula: This refers to the unidirectional linear shrinkage of the shell-shaped structure. For different structural directions, the linear contraction coefficient is denoted as . This refers to the shell-shaped settling time; This represents the percentage of resin adhesion on the coated sand. The coefficient representing the influence of the resin's room temperature crosslinking reaction. For shell-type surface curing degree; This is the curing degree coupling correction coefficient; It is the resin crosslinking degree attenuation factor; This is a correction factor for ambient temperature linkage. Real-time ambient temperature of the static area; The reference temperature for mold storage; These are dynamic parameters of the resin crosslinking reaction; This is the nonlinear shrinkage correction coefficient.

[0031] As a preferred embodiment, the specific steps for fixing the shell assembly in step S3 are as follows:

[0032] Laser positioning equipment is used to perform benchmark alignment pre-calibration on the split shell type. The laser positioning accuracy is controlled to 0.005mm, and the calibration deviation is controlled within 0.05mm.

[0033] Combined with the calibration ambient temperature, a laser calibration deviation compensation factor is used. Correcting calibration deviations, The value range is 0.97 to 1.03;

[0034] A hydraulic pressure equalization assembly device is used to synchronously pre-tighten the shell-shaped splicing end faces, with the pre-tightening force increasing in increments of 25%.

[0035] After pre-tightening, perform uniform pressure locking and use a feeler gauge to check the gap between the shell-shaped splicing end faces to ensure that the gap value is not greater than 0.08mm;

[0036] A coordinate measuring machine is used to detect the shape and position parameters of the assembled shell, and the shape and position deviation is controlled within 0.3mm to complete the overall rigid fixation of the shell.

[0037] As a preferred embodiment, the specific steps for monitoring the shell-type sealing and airtightness in step S4 are as follows:

[0038] A process of alternating refractory mortar and high-temperature sealing tape is used to seal the gaps and openings of the shell joints in layers. When the gap width is 0.2mm to 3.5mm, 3 to 4 layers are used, and when the gap width is <0.2mm, 2 layers are used. The thickness of each layer is controlled to be 0.5mm to 1.0mm, and the edge of the covering layer extends at least 5mm beyond the edge of the gap.

[0039] Distributed sensor points are arranged in the inner cavity of the shell at a density of 3 to 5 per cubic meter. The sensor points avoid the dead corners of the inner cavity and the key forming areas of the casting. Miniature sensors are used to collect the air pressure, temperature and humidity parameters of the inner cavity in real time at a frequency of 1 time / 2 minutes to form an inner cavity environment monitoring dataset.

[0040] The probability of maintaining a sealed steady state within the shell cavity is calculated using a Bayesian decision formula, as follows:

[0041] ;

[0042] In the formula: The probability of maintaining a sealed steady state within the shell cavity is controlled by a threshold value ≥ 0.95. The conditional probability of a perfectly sealed interior with a stable internal pressure differential; The prior probability of the sealed structure working properly; The conditional probability of a leaking seal failure but a randomly stable internal pressure differential. This represents the prior probability of the sealed structure failing. The internal air pressure fluctuation coefficient; The aging correction factor for the sealing structure is 0.90-0.98, determined according to the standard for high-temperature aging characteristics of refractory sealing materials. This is the aging rate factor for sealing tape, with a value of 0.92-0.99, determined according to the industry standard for aging rate of high-temperature resistant sealing tape.

[0043] As a preferred embodiment, the specific steps for controlling the pouring of molten metal in step S5 are as follows:

[0044] The probability of maintaining a sealed steady state within the housing cavity was confirmed by collecting parameters from distributed sensing points inside the housing. ≥0.95;

[0045] Call the dual-station time-series iterative operation model, input the effective environmental monitoring dataset and dual-station operation parameters, and generate the initial staggered peak avoidance arrangement rhythm;

[0046] Measure the temperature of the molten metal, controlling it between 1500℃ and 1650℃, and calculate the viscosity correction factor of the molten metal based on the temperature value. , The value range is 0.86 to 0.97;

[0047] The pouring flow rate is set according to the wall thickness of the casting. When the wall thickness is 4mm to 6mm, the flow rate is controlled at 0.6m / s to 0.8m / s; when the wall thickness is 7mm to 12mm, the flow rate is controlled at 0.4m / s to 0.6m / s; and when the wall thickness is 13mm to 18mm, the flow rate is controlled at 0.3m / s to 0.4m / s.

[0048] A layered, flow-limited, and progressive pouring operation is implemented using a pouring ladle. A flow controller is used to control the pouring flow rate in real time with a response time of no more than 0.5 steps (S). At the same time, a pressure sensor is used to monitor the internal pressure of the cavity in real time and control the pressure between 0.1 MPa and 0.3 MPa.

[0049] As a preferred embodiment, the specific steps for controlling the solidification of the casting in step S6 are as follows:

[0050] After the molten metal is filled into the mold, an ultrasonic testing device is used to test the wall thickness parameters of the casting at a speed of 0.5 m / s. The testing accuracy is controlled to be 0.02 mm. Based on the wall thickness parameters, the casting is divided into three solidification control areas: thin-walled area, medium-thickness area, and thick-walled area.

[0051] The overall thermal resistance at the shell-to-casting interface is calculated using a dynamic interfacial thermal resistance calculation formula. ,according to The values ​​represent gradient heat dissipation parameters set for three solidification control zones: the heat dissipation rate is controlled at 0.8℃ / min~1.2℃ / min in the thin-walled zone, 0.5℃ / min~0.8℃ / min in the medium-thick zone, and 0.3℃ / min~0.5℃ / min in the thick-walled zone, thus completing the solidification control of the casting; the formula for calculating the dynamic thermal resistance of the interface is as follows:

[0052]

[0053] In the formula: The overall thermal resistance at the interface between the shell and the casting; Thermal resistance at the contact point between the mold and the metal base; The interfacial air film interference coefficient; The interfacial air film gap thickness; This is the solidification time coupling factor; This is the temperature gradient correction factor; This refers to the temperature difference between the inner and outer surfaces of the casting. This refers to the solidification time of the casting; This is a correction factor for the distribution of thermal points in castings.

[0054] As a preferred embodiment, the specific steps of the casting post-processing and dual-station cyclic operation in step S7 are as follows:

[0055] Thermocouple monitoring data confirmed that the casting had completely solidified and solidified, meaning that the surface temperature of the casting had dropped to room temperature ±5℃.

[0056] The shell-shaped structure is disassembled step by step, proceeding from non-critical parts to critical parts.

[0057] After disassembly, special finishing tools are used to finely finish the flash and residual sand on the casting. The finishing accuracy of the flash is controlled at 0.01mm to ensure that the residual sand is cleaned up and there is no residue. After finishing, the surface roughness of the casting Ra≤1.6μm.

[0058] The dimensions and surface quality parameters of the finished castings were inspected using 3D inspection equipment, and the surface roughness data of the castings were collected using a roughness meter to calculate the surface roughness influencing factor. , The value range is 0.96 to 1.04;

[0059] Include dimensional parameters, surface quality parameters, and Substitute the data into the multidimensional detection model to calculate the overall dimensional deviation of the casting, ensuring that the deviation is ≤ ±0.1mm;

[0060] Set the process division and rotation interval between the two sets of work units to 40min to 90min, and rotate and replace the process division of the two sets of work units;

[0061] Collect production process parameters and quality inspection data for the first n-1 batches to form a batch dataset, set batch deviation weighting factors, calculate the quality fluctuation coefficient for the first n-1 batches, and calculate the batch stability correction factor based on the quality fluctuation coefficient. , The value range is 0.97 to 1.03;

[0062] Batch deviation weighting factor, Substitute multiple batches of data into the iterative optimization model, obtain cyclic calibration parameters, call the cyclic calibration parameters to adjust the workstation operation parameters, and continuously carry out production line flow operations.

[0063] As a preferred embodiment, the invocation step of the dual-station time-series iterative job model in step S5 is as follows:

[0064] Extract the dual-station timing fit obtained above And a valid environmental monitoring dataset, combined with the molten metal temperature and molten metal viscosity correction factor detected in step S5. Input a dual-station time-series iterative operation model;

[0065] Model by The control requirement is ≥0.85. The staggered arrangement rhythm is dynamically adjusted to avoid peak periods and the pouring start lag time is corrected. The lag time is controlled to be 1min to 3min.

[0066] Simultaneously, the pouring flow rate corresponding to the casting wall thickness is combined with the dual-station transfer connection time to be calibrated. The transfer connection time is controlled to be 2 min to 5 min to ensure that the pouring operation and the dual-station sequence are accurately matched and to avoid process interruption.

[0067] As a preferred approach, a multi-process parameter coupling optimization step is also included, specifically:

[0068] Extracting the shell-shaped linear shrinkage amount in step S2 The probability of maintaining a sealed steady state within the shell cavity in step S4. Metal liquid viscosity correction factor in step S5 and the overall thermal resistance of the interface in step S6 A multi-process parameter coupling optimization model is constructed, and coupling weight coefficients are set, where... The weights range from 0.22 to 0.28. The weights range from 0.30 to 0.38. The weights are 0.18 to 0.25. The weights range from 0.12 to 0.18.

[0069] Calculate the coupling fit of multi-process parameters ,when When the value is less than 0.90, adjust the parameters of each process in order of weight priority, prioritizing the adjustment of the closed-state parameters and the pouring flow rate parameters, until... ≥0.90;

[0070] The formula for calculating the multi-process parameter coupling fit degree is as follows:

[0071] ;

[0072] In the formula: To improve the coupling adaptability of multi-process parameters; This represents the standard value for unidirectional linear shrinkage of the shell-shaped structure. This represents the actual linear shrinkage of the shell-shaped structure in one direction. The probability of maintaining a sealed steady state within the shell cavity; This is a viscosity correction factor for molten metal. This is the standard value for the combined thermal resistance at the interface between the shell and the casting. The actual combined thermal resistance at the interface between the shell and the casting is given; the weighting coefficients are set according to the degree of influence of each process on the quality of the casting, and the sum is 1.

[0073] The present invention also provides a large-size shell casting manufacturing system based on a dual-station shell profile line, wherein the system stores a computer program, and the computer program implements the above-described method when executed by a processor.

[0074] Terminology Explanation: Dual-station shell casting line is a common term in the casting industry, referring to a shell casting production line with two independent workstations that can operate in parallel, used for the continuous production of large-size shell castings; where shell casting line refers to the shell casting production line.

[0075] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0076] 1. This invention, through collaborative management and control via dual-station environmental parameter screening and dynamic scheduling of a time-series iterative operation model, overcomes the technical challenges of traditional dual-station shell molding lines that simply replicate workstations and suffer from disjointed process sequences. It achieves precise staggered peak avoidance and seamless connection of the entire process, including shell making, casting, cooling, and transfer, eliminating workstation interference and connection bottlenecks. This results in a stable improvement in dual-station time sequence adaptability, increased production line efficiency, and significantly improved workpiece transfer smoothness. It can effectively adapt to the large-scale and continuous production needs of large-size castings, and improves process synergy and production line operating efficiency.

[0077] 2. This invention achieves full-chain dimensional accuracy control by constructing a multi-factor linkage shrinkage calculation, laser alignment compensation assembly, and layered alternating sealing coating. It incorporates multiple factors such as shell material characteristics, ambient temperature, stress relaxation, assembly deviation, and sealing tightness into quantitative control, replacing the traditional experience-based control mode. This effectively solves the problems of large shrinkage deviation, excessive assembly gaps, and sealing leakage in large-size shells. The method of this invention can achieve precise control of the linear shrinkage of the shell, stabilize the dimensional accuracy of the casting, improve the assembly gap qualification rate, and eliminate quality defects such as dimensional deviation, air ingress during pouring, and leakage.

[0078] 3. This invention employs a precise molding control process that utilizes wall thickness-zoned flow-limited casting, dynamic interface thermal resistance quantification, and zoned gradient heat dissipation solidification. This effectively solves the shortcomings of traditional fixed casting and solidification parameters and uneven heat dissipation. Based on the casting structural characteristics and interface heat transfer mechanism, the pouring flow rate and heat dissipation rate are dynamically matched to precisely suppress thermal defects and balance the solidification sequence, resulting in more stable metal filling and more balanced heat dissipation. This reduces the rate of typical molding defects such as shrinkage porosity, gas holes, and cracks in the casting by more than 85%, and significantly improves the uniformity of the internal structure and mechanical properties of the casting, thus better meeting the internal quality requirements of castings for heavy components, engineering machinery, wind power equipment, and other equipment.

[0079] 4. The Bayesian closed-loop steady-state probability determination and adaptive linkage adjustment mechanism of the present invention upgrades the closed-loop state from manual qualitative judgment to quantitative and precise judgment. At the same time, it realizes real-time identification, quantitative correction and collaborative correction of key process parameter anomalies, effectively resists the production risks caused by environmental fluctuations, process drift and equipment deviations, so that the probability of the closed-loop steady-state of the shell cavity remains stable at ≥0.95, the response time of abnormal working conditions is shortened by 90%, the overall yield of castings is increased to 98.5%, and the production stability and operational safety are comprehensively enhanced.

[0080] 5. This invention integrates process parameters such as shell shrinkage, sealed state, casting viscosity, and interfacial thermal resistance, and achieves global process optimization control through coupling adaptability calculation. At the same time, it relies on post-processing quality inspection data to complete the iterative update of process parameters, eliminating the need for frequent manual adjustments. It can adaptively match the production of large-size castings of different specifications and precision levels, such as gearbox housings, excavator booms, and wind turbine bearing seats. The production line operation and maintenance costs are reduced, and the process versatility and scalability are significantly better than traditional casting processes. Attached Figure Description

[0081] 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.

[0082] Figure 1 This is a process flow diagram of the method for manufacturing large-size shell castings according to the present invention.

[0083] Figure 2 This is a flowchart illustrating the steps of controlling the molten metal pouring of the present invention. Detailed Implementation

[0084] 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 1-2 As shown in the image.

[0085] A method for manufacturing large-size shell castings based on a dual-station shell profile includes the following steps: S1, Environmental parameter collection and screening and dual-station layout control: Collect and eliminate abnormal environmental parameters in the foundry workshop to form an effective environmental monitoring dataset; complete the dual-station layout division, time sequence operation arrangement parameter calibration and workpiece transfer path planning between stations, and monitor the fluctuation of dual-station environmental parameters in real time.

[0086] S2. Shell demolding, deformation detection and static management: Demolding operation is carried out on large-sized shells that have been solidified, and multi-dimensional deformation detection and data recording of the shells are completed; the shells are placed in static areas according to specifications, and the deformation status of the shells is monitored in real time and the data is corrected.

[0087] S3. Shell assembly and fixing: The split shell is pre-calibrated and the deviation is corrected based on the benchmark. The shell is pre-tightened and locked by the pressure equalization assembly process. The assembly accuracy and shape and position parameters are tested to achieve the overall rigid fixing of the shell.

[0088] S4. Shell sealing and airtightness monitoring: Seal and cover the gaps and ports of the shell splicing, set up monitoring points for the internal cavity environment, collect internal cavity environment parameters in real time, and adjust the airtightness measures to maintain the airtight and isolated state of the cavity.

[0089] S5. Molten Metal Pouring Control: After confirming that the inner cavity of the shell is sealed, a pouring rhythm is generated based on the dual-station timing characteristics. Layered flow-limited pouring is implemented according to the structural characteristics of the casting, and the pouring process parameters are controlled in real time.

[0090] S6. Casting solidification control: After the molten metal is filled into the mold, the wall thickness distribution of the casting is detected and the solidification control area is divided. Based on the interface thermal resistance calculation results, gradient heat dissipation parameters are set according to the area to complete the casting solidification control.

[0091] S7. Casting post-processing and dual-station cyclic operation: After confirming that the casting has completely solidified and solidified, complete the shell disassembly, casting trimming and quality inspection, rotate the dual-station process division, and call the cyclic calibration parameters to maintain the continuous flow of the production line.

[0092] In this solution, step S1 first completes the basic control of the environment and workstations, eliminating external interference and workstation conflicts; steps S2 to S4 focus on the deformation, assembly, and sealing of the shell mold, ensuring the accuracy and airtightness of the cavity structure; steps S5 and S6 ensure the internal and external quality of the casting through precise pouring and controlled solidification; step S7 realizes post-processing and process rotation, allowing the dual-station production line to form a closed loop for sustainable operation. The static setting and deformation correction after shell mold demolding can release internal stress and reduce assembly deviations; pouring can be started after the airtight state is met, which can significantly reduce the risk of forming defects; layered flow restriction adapts to the differences in wall thickness of large castings; gradient heat dissipation solves the problem of uneven solidification; post-processing and process rotation allow the dual-station to operate efficiently alternately, and the cyclical calibration of parameters continuously optimizes the process. The entire solution fits the actual needs of dual-station production of large-size castings. The overall solution first establishes a stable production foundation through environmental data collection and workstation planning, then conducts full-process precision control on the shell shape, followed by core molding operations such as sealing, pouring, and solidification, and finally achieves efficient production line operation through post-processing and cyclical operations.

[0093] It should be explained that the process model of the solution is divided into four levels: basic control, shell treatment, forming control, and cycle optimization. The input-output relationship of each level is clear and sequential: the basic control level takes the workshop environment as input and outputs stable environmental parameters and workstation layout parameters; the shell treatment level takes the stabilized shell as input and outputs a high-precision, high-rigidity sealed cavity; the forming control level takes a qualified cavity as input and outputs a casting with qualified forming quality; the cycle optimization level takes the status of the casting and production line as input and outputs optimized process parameters and feeds them back to the previous level.

[0094] The stability of environmental parameters in this technical solution reduces the probability of shell deformation, precise deformation correction improves assembly accuracy, reliable assembly ensures sealing effect, airtightness supports stable casting, layered casting combined with gradient solidification reduces defects, and cyclic optimization continuously improves process level; it solves the problems of disconnection of dual-station production processes, parameter mismatch, and large quality fluctuations, and the synergy of each step achieves simultaneous improvement in production efficiency, molding accuracy and product quality.

[0095] Optionally, in step S1, the specific steps for environmental parameter collection and screening, and dual-station layout control are as follows:

[0096] Temperature and humidity sensors, dust detectors, and barometers were used to collect ambient temperature, relative humidity, dust concentration, and air pressure parameters in the workshop. The data collection frequency was once every 3 minutes, and no less than 10 sets of data were collected continuously. Abnormal data were removed by the 3σ criterion, and effective data with ambient temperature of 15℃~40℃, relative humidity of 38%~75%, dust concentration ≤50mg / m³, and air pressure of 80kPa~105kPa were selected to form an effective environmental monitoring dataset.

[0097] Two independent work units were divided, with a unit spacing of 3m to 5m. The dual-station sequential work arrangement parameters were calibrated using a time-adaptive coupling calculation formula. The ambient temperature difference ΔT and relative humidity difference ΔRH between the two work units were extracted, with ΔT controlled within the range of 2℃ to 12℃ and ΔRH controlled within the range of 5% to 22%. The average process time per batch at the shell-making station and the casting and cooling station was collected and recorded as follows: and The values ​​range from 45 min to 165 min and from 90 min to 360 min, respectively; calculate the dynamic coupling factor of environmental parameters. Timing lag correction factor and workstation thermal field superposition correction factor Set the stability correction coefficient for operating conditions ;

[0098] Substitute the above parameters into the timing adaptation coupling calculation formula to calculate the dual-station timing adaptation degree. Adjust the timing operation scheduling parameters based on the calculation results to ensure... ≥0.85; Survey the layout of workstations and equipment distribution, delineate workpiece orientation transfer paths between workstations with a width of 1.5m to 2.0m, and avoid blind spots in equipment operation; Monitor the fluctuations of environmental parameters of the two sets of work units in real time at a frequency of no less than once per 5 minutes to ensure the adaptation of timing parameters;

[0099] The timing adaptation coupling calculation formula is as follows:

[0100] ;

[0101] In the formula: The timing adaptability of the dual-station production line is set with a value of 0-1 and a control threshold of ≥0.85, based on the industry standard for timing coordination stability of the dual-station casting production line. The value of the correction factor for stable operating conditions is 0.03-0.15, and it is determined based on the known law of correction for fluctuations in the operating conditions of the foundry workshop. The average process time for a single batch at the shell-making station is in minutes, ranging from 45 to 165 minutes, and is set according to the standard cycle of the shell-making process for large-size shells. The average process time for a single batch at the pouring and cooling station is in minutes, ranging from 90 to 360 minutes, and is set according to the standard cycle of pouring and cooling process for large-size castings. The difference in ambient temperature between the two work units is expressed in °C and ranges from 2 to 12 °C, determined according to the control specifications for ambient temperature differences in dual-station work units. The relative humidity difference between the two work units is expressed as a percentage, ranging from 5% to 22%, and is determined according to the environmental humidity difference control specifications for dual-station work units. The dynamic coupling factor for environmental parameters ranges from 0.85 to 0.98 and is derived from industry data on the coupling effects of environmental temperature, humidity, and air pressure. This is the timing lag correction factor, ranging from 0.02 to 0.08, determined according to the timing lag correction standard for dual-station processes. The value of the correction factor for the superposition of thermal field at the workstation is 0.82 to 0.95, and is determined according to the correction specification for superposition interference of thermal field at the casting workstation.

[0102] In the production of large-sized castings in a dual-station shell molding line, the stability of the workshop environmental parameters directly determines the solidification and dimensional stability of the shell mold. The layout spacing, timing coordination, and transfer path of the two stations directly affect the operating efficiency and safety of the production line. Traditional production line construction simply divides the work into two work stations without standardizing the collection and anomaly removal of workshop environmental parameters, nor does it use quantitative models to calibrate the timing of the two stations. The transfer path is planned arbitrarily, which easily leads to problems such as shell mold deformation and cracking caused by environmental fluctuations, process conflicts and congestion at the two stations, and workpiece collisions and scratches during transfer. This has been a long-standing problem that has plagued the large-scale production of dual-station shell molding lines. Through long-term production line planning, process debugging, and R&D iteration, we have analyzed a large amount of production data and failure cases and found that clear environmental control and workstation layout are the core conditions for the stable operation of dual-workstation production lines. Therefore, we have formed a systematic layout control approach. The entire operation process first completes the collection of environmental data and the elimination of anomalies to ensure the stability of the basic production environment. Then, it divides the work into independent work units to avoid interference between workstations. Next, it uses time-series adaptation and coupled calculation to quantitatively calibrate the dual-workstation operation rhythm, plans dedicated transfer paths, and finally monitors environmental fluctuations in real time to ensure continuous time-series adaptation. Each step is matched with the actual planned operation of the dual-workstation production line.

[0103] In addition, temperature, humidity, dust, and air pressure are the core environmental factors affecting the curing of coated sand shell molds. The 3σ criterion can effectively avoid the interference of accidental environmental fluctuations by eliminating abnormal data, and the effective data after screening can ensure the stability of the shell mold production environment. The distance between the independent working units of the two stations is controlled at 3-5m, which can not only avoid the interference of the thermal radiation of the casting station on the sand mold curing of the shell making station, but also save workshop space utilization. Among them, the timing adaptation coupling calculation integrates five core influencing factors: process time ratio, difference of ambient temperature and humidity, dynamic coupling factor, timing lag, and thermal interference. Through the combination logic of exponential correction and logarithmic operation, it accurately quantifies the timing adaptation of the two stations, replaces the traditional manual experience in arranging the timing, and avoids process conflicts. The 1.5-2.0m wide directional transfer path avoids the blind spot of equipment operation and can ensure smooth and safe transfer of workpieces. It should be noted that the temperature and humidity sensors, dust detectors, and barometers used in this step are all standard testing equipment in the foundry workshop, and all parameter values ​​are determined in accordance with the environmental control standards for foundry workshops, the layout design specifications for dual-station production lines, and the timing scheduling standards for industrial production lines.

[0104] This solution organically couples process time, environmental differences, dynamic coupling, timing lag, and the impact of thermal interference. It employs a composite mathematical framework combining exponential correction and logarithmic operations, matched with the physical mechanism of timing coordination in a dual-station production line. This framework plays a crucial role in accurately calibrating the timing of dual-station operations and ensuring efficient collaborative operation of the production line. Furthermore, the working principle of this solution perfectly aligns with the actual production logic of a dual-station shell molding line. First, it collects basic environmental data from multiple sensors, forming a valid dataset after anomaly filtering. Then, it calculates the process time and environmental differences between the two stations, determines various correction factors based on industry standards, and substitutes them into the timing adaptation coupling calculation formula to complete quantitative accounting. Finally, it adjusts the work arrangement parameters based on the timing adaptation degree, allowing seamless connection and collaborative operation of the shell molding, casting, cooling, and transfer processes in the dual-station line. The entire process does not require complex artificial intelligence model training. It only performs calculations based on well-known industrial scheduling algorithms and casting industry standards. Moreover, the design of the above formulas in this invention is not a simple combination of mathematical formulas, but a precise design of the timing arrangement of dual workstations based on the collaborative implementation of dual workstation production lines. The three-level mathematical accounting model built based on industrial production line scheduling logic has a very clear hierarchical structure, divided into three clear levels: input layer, calculation layer, and output layer.

[0105] The overall design of this time-series adaptation coupling calculation formula adopts a superposition structure of dynamic coupling of working condition correction, working hour ratio, environmental index correction, and logarithmic lag correction. Compared with the existing technology, this formula integrates multi-dimensional parameters to achieve accurate quantitative calculation of time-series adaptation degree and constructs a systematic layout that couples environmental control and time-series calibration.

[0106] This solution quantifies and calibrates the timing of dual-station operations through a time-adaptive coupling calculation, replacing traditional manual experience-based layout. This resolves issues of process conflicts and bottlenecks in dual-station operations, significantly improving production line efficiency. Furthermore, this solution plans dedicated directional transport paths to avoid equipment blind spots, eliminating safety hazards such as workpiece collisions and scratches during transport, and enhancing production line safety. A real-time environmental parameter monitoring mechanism is established for the copper pots, rapidly responding to fluctuations and adjusting timing parameters to ensure the long-term, continuous, and stable operation of the dual-station production line. The overall technical concept and implementation results are significantly superior to traditional production line layouts.

[0107] Optionally, in step S2, the specific steps for shell demolding, deformation detection, and static control are as follows:

[0108] The large-sized shell molded by solidification is demolded using a parallel lifting method, with the lifting speed controlled at 0.1m / s to 0.2m / s, and the lifting points are symmetrically distributed at the edge of the shell.

[0109] Test points are set up in stress concentration areas such as the flange connection of the shell and the area around the oil passage, with a density of 4 to 8 points per square meter. A dial indicator is used to test the linear deformation values ​​in the length, width and height directions of the shell, with the test accuracy controlled at 0.01 mm. The test data are recorded to form a deformation parameter dataset.

[0110] According to the production rhythm, the shell-shaped areas are neatly arranged and placed in stillness. The ambient temperature of the stillness area is controlled at 18℃~35℃ and the relative humidity is controlled at 40%~70%. The stillness area is far away from heat sources and areas with airflow disturbance.

[0111] During the static setting process, the shell deformation is detected every 10 minutes to generate a real-time deformation monitoring dataset.

[0112] The directional shrinkage of the shell is calculated using a multi-factor linkage correction formula, taking into account the surface curing degree after demolding. and percentage of resin adhesion to coated sand ,in The value range is 0.8 to 0.95. The value range is 2.2% to 3.8%; the real-time ambient temperature of the static area is detected. and mold storage reference temperature Calculate the ambient temperature difference; determine the dynamic parameters of the resin crosslinking reaction. Resin crosslinking degree attenuation factor Linear contraction coefficient in different structural directions Influence coefficient of resin room temperature crosslinking reaction Curing degree coupling correction coefficient Ambient temperature linkage correction coefficient and nonlinear shrinkage correction coefficient Record the shell type and its static setting time. Substituting the above parameters into the multi-factor linkage correction calculation formula, the unidirectional linear shrinkage of the shell is calculated. ,make sure Within the control range of 0.15mm to 1.20mm; the multi-factor linkage correction calculation formula is as follows:

[0113] .

[0114] In the formula: This is the unidirectional linear shrinkage amount of the shell, in mm, ranging from 0.15 to 1.20 mm, set according to the assembly accuracy requirements and actual shrinkage characteristics of large-size coated sand shell molds; The linear contraction coefficient is given in different structural directions, in °C⁻¹, and is 2.8 × 10⁻ 6 -7.5×10⁻ 6 ℃⁻¹, determined according to the standard for thermal expansion and contraction in coated sand; This is the shell-type static time, in minutes, ranging from 25 to 110 minutes, set according to the industry's standard requirements for stress release cycles of large-size shell types. The percentage of resin adhesion on the coated sand is expressed as a percentage, ranging from 2.2% to 3.8%, and is set according to the product standard for coated sand used in casting. The coefficient representing the influence of the room temperature crosslinking reaction of the resin is 0.25-0.65, determined based on the well-known laws of the room temperature crosslinking kinetics of phenolic resin. The degree of curing of the shell-type surface layer is 0.8-0.95, set according to the quality standard of the coating sand thermosetting process; This is the curing degree coupling correction coefficient, with a value ranging from 0.18 to 0.55, derived based on the correlation characteristics between curing degree and shrinkage. The resin crosslinking degree decay factor has a value of 0.88-0.96, determined based on industry data on the resin's later crosslinking decay rate. This is the ambient temperature linkage correction coefficient, in mm / ℃, with a value ranging from 0.02 to 0.09 mm / ℃, determined according to the specifications on the influence of ambient temperature in the foundry on shell shrinkage; The real-time ambient temperature of the static area is expressed in °C, with a range of 18-35°C, set according to the conventional environmental control range of the foundry workshop. This is the reference temperature for mold storage, in °C, and ranges from 22 to 28 °C, set according to the general standard temperature for mold curing. This is a dynamic parameter for the resin crosslinking reaction, with a value ranging from 0.9 to 1.1, determined based on a known calculation method for the real-time crosslinking state of the resin. This is a nonlinear shrinkage correction factor, with a value ranging from 0.01 to 0.03, determined based on the nonlinear shrinkage law of shells after long-term static placement.

[0115] In the actual scenario of mass production of large-size castings on a dual-station shell molding line, the demolding safety and static shrinkage control of the shell mold are the core factors determining the final dimensional accuracy of the casting. Due to their large size, uneven wall thickness, and numerous stress concentration areas, large-size shell molds are prone to cracking and deformation when demolded using traditional single-point lifting. Traditional deformation detection only samples non-critical areas, and static shrinkage is estimated using only a single coefficient, completely ignoring the coupled effects of multiple factors such as resin content, degree of curing, ambient temperature, and stress relaxation. This leads to significant deviations in shrinkage calculation, resulting in excessive gaps in subsequent shell mold assembly, and consequently, a series of quality defects such as dimensional deviations, sealing failures, and pouring leakage. We have developed a systematic control approach: parallel demolding to prevent cracking—precise detection of stress zones—zoned static stabilization—multi-factor shrinkage calculation. The entire operation process strictly follows the production sequence. First, parallel lifting ensures demolding safety; then, precise deformation detection is performed on stress concentration areas; next, zoned static stabilization releases stress; and finally, shrinkage is calculated using a multi-factor formula. Each step is matched to the actual on-site operation.

[0116] Among these measures, controlling parallel lifting, symmetrical placement, and low-speed operation ensures uniform stress on the shell during demolding, preventing stress concentration and cracking. High-density placement of detection points in stress concentration areas accurately captures deformation in critical parts of the shell. A multi-factor linkage correction calculation integrates resin content, degree of cure, ambient temperature, settling time, and cross-linking attenuation variables to accurately calculate shrinkage, providing reliable data for subsequent assembly deviation compensation. The lifting equipment, dial indicator, and temperature and humidity sensors used in this step are all standard equipment in the foundry workshop.

[0117] The multi-factor linkage correction calculation formula of this invention differs from the conventional method of single-factor linear estimation in existing technologies. It couples and superimposes three core influencing factors: material properties, process parameters, and environmental conditions. It is divided into three parts: material-dominant term, environmental correction term, and nonlinear compensation term, and matches them with the physical mechanism of shell shrinkage to achieve accurate calculation of shrinkage and ensure shell dimensional accuracy. The working principle of this scheme is completely consistent with actual production logic. First, basic data such as shell curing degree, resin content, ambient temperature, and settling time are collected through dedicated testing equipment. These data directly reflect the shell's own state and external working conditions. Then, based on the known principles of resin crosslinking kinetics and thermal expansion and contraction, various correction coefficients are determined. After substituting all parameters into the formula, the linear shrinkage of the shell is accurately calculated by performing separate calculations and then superimposing and integrating them.

[0118] It should be explained that the multi-factor linkage correction model adopted in this scheme is a mathematical calculation model based on the mechanism of casting materials science. It has a clear hierarchical structure, divided into three levels: input layer, calculation layer, and output layer. The design concept of this multi-factor linkage correction calculation formula revolves around the multi-factor coupling characteristics of shrinkage in large-size coated sand shell molds. It adopts a three-stage superposition structure, organically integrating material, process, and environmental factors. First, it collects basic parameters of the shell mold and environment, then determines the correction coefficients, and finally calculates the shrinkage amount. Each step of the scheme works synergistically. Resin content and curing degree determine the shrinkage basis; environmental temperature corrects for external interference; resting time and cross-linking parameters reflect the stress release state; nonlinear coefficients compensate for offset errors; algorithm characteristics and process characteristics are deeply coupled; formula calculation provides a quantitative basis for process adjustment; and on-site parameters provide precise input for formula calculation, with the two mutually supporting each other. This design approach is not a simple superposition of conventional technologies, but a design based on the physical mechanism of shell mold shrinkage. The synergistic effect of each step significantly improves the dimensional stability of the shell mold, providing reliable support for subsequent assembly processes and is a key link in achieving stable operation of a dual-station production line.

[0119] This solution achieves accurate quantitative calculation of shrinkage in large-size shell molds during the shell shrinkage calculation stage. By coupling multiple parameters to match the actual shrinkage mechanism of the shell mold, it reduces calculation errors. This solution establishes a collaborative calculation system integrating material properties, process parameters, and environmental conditions, consolidating disparate influencing factors into a unified calculation model, simplifying on-site operation procedures, and improving process control efficiency. Furthermore, this solution provides a reliable compensation basis for subsequent shell mold assembly through accurate shrinkage calculation, reducing assembly gaps and dimensional deviations, and effectively lowering the probability of dimensional defects in castings.

[0120] Optionally, in step S3, the specific steps for assembling and fixing the shell are as follows:

[0121] Laser positioning equipment is used to perform benchmark alignment pre-calibration on the split shell type. The laser positioning accuracy is controlled to 0.005mm, and the calibration deviation is controlled within 0.05mm.

[0122] Combined with the calibration ambient temperature, a laser calibration deviation compensation factor is used. Correcting calibration deviations, The value range is 0.97 to 1.03;

[0123] A hydraulic pressure equalization assembly device is used to synchronously pre-tighten the shell-shaped splicing end face. The pre-tightening force is increased in a 25% gradient, with the initial pre-tightening force controlled between 5kN and 8kN and the final pre-tightening force controlled between 15kN and 20kN.

[0124] After pre-tightening, perform uniform pressure locking and use a feeler gauge to check the gap between the shell-shaped splicing end faces to ensure that the gap value is not greater than 0.08mm;

[0125] A coordinate measuring machine is used to detect the shape and position parameters of the assembled shell, and the shape and position deviation is controlled within 0.3mm to complete the overall rigid fixation of the shell.

[0126] The assembly accuracy of the split shell is the core factor that determines the subsequent sealing effect, casting quality, and even the final dimensional accuracy of the casting. Due to the large size and end face span of large-size shells, traditional assembly processes generally rely on manual visual alignment, single-point pressing, or asymmetrical force application. This not only fails to ensure uniform force on the spliced ​​end faces, but also fails to consider the systematic deviation caused by changes in ambient temperature on the laser calibration accuracy. As a result, problems such as local shell cracking, excessive splicing gaps, and excessive dimensional and positional deviations are very likely to occur.

[0127] This solution utilizes laser positioning equipment to provide a micron-level high-precision reference for the split shell, completely solving the problem of insufficient accuracy in manual alignment. Simultaneously, a laser calibration deviation compensation factor is introduced for dynamic correction, avoiding slight deviations in the laser path caused by ambient temperature fluctuations. It also prevents deformation due to thermal expansion and contraction of the shell's own sand-like material caused by temperature changes. Temperature compensation ensures the accuracy of the calibration results, improving subsequent assembly precision. However, directly applying the final preload to the joint face of large-sized shells can cause excessively high concentrated stress in localized areas, easily leading to shell cracking or end-face warping. A 25% gradient preload loading method allows the shell to gradually adapt to the stress state, ensuring uniform fit across the entire joint face and effectively avoiding defects caused by localized stress concentration. This solution achieves precise and rigid shell fixing through the coordinated use of laser alignment, temperature compensation, gradient preload, and precision testing.

[0128] This solution employs a gradient-increasing synchronous pre-tightening process during the shell assembly and fixing stage. This gradually evens out the stress on the joint end faces, preventing localized stress concentration that could cause shell cracking or end face deformation, and significantly improving the structural integrity of the shell assembly. Furthermore, this solution establishes a dual-dimensional detection mechanism for joint gaps and geometric parameters, incorporating assembly quality into quantitative control and reducing the cascading defects such as sealing failures and injection leakage caused by assembly deviations.

[0129] Optionally, in step S4, the specific steps for monitoring the shell-type sealing and airtightness are as follows:

[0130] A process of alternating refractory mortar and high-temperature sealing tape is used to seal the gaps and openings of the shell joints in layers. When the gap width is 0.2mm to 3.5mm, 3 to 4 layers are used, and when the gap width is <0.2mm, 2 layers are used. The thickness of each layer is controlled to be 0.5mm to 1.0mm, and the edge of the covering layer extends at least 5mm beyond the edge of the gap.

[0131] A thickness gauge was used to check the uniformity of the coating thickness, and the coating force was adjusted as needed according to the quality control standards for the density of the sealing coating.

[0132] Distributed sensor points are arranged in the inner cavity of the shell at a density of 3 to 5 per cubic meter. The sensor points avoid the dead corners of the inner cavity and the key forming areas of the casting. Miniature sensors are used to collect the air pressure, temperature and humidity parameters of the inner cavity in real time at a frequency of 1 time / 2 minutes to form an inner cavity environment monitoring dataset.

[0133] The probability of maintaining a sealed steady state within the shell cavity is calculated using a Bayesian decision formula.

[0134] Set the conditional probability of a perfect seal and a stable internal pressure difference. The prior probability of the sealing structure working properly The conditional probability of a leaky seal but a randomly stable internal pressure differential. and the prior probability of sealing structure failure ;

[0135] The internal cavity pressure fluctuation coefficient is calculated based on the real-time collected internal cavity air pressure data. ; Determine the aging correction coefficient for the sealing structure by considering the covering time and environmental humidity. and sealing tape aging rate factor Substitute the above parameters into the Bayesian decision formula to calculate the probability of maintaining a sealed steady state within the shell cavity. ,when When the value is ≥0.95, the internal cavity is considered to be in a stable sealed state. The sealing measures are adjusted to maintain the cavity's sealed isolation, controlling the internal air pressure between 30Pa and 85Pa, the temperature between 20℃ and 30℃, and the relative humidity ≤60%. The Bayesian determination formula is as follows:

[0136] ;

[0137] In the formula: The control threshold for maintaining a sealed steady state within the shell cavity is ≥0.95, set according to the industry standard for the sealing safety of large-size shell casting. The conditional probability of a good seal and stable internal pressure difference is set at 0.96-0.99, determined according to the reliability standard for high-temperature sealing structures used in casting. The prior probability of the sealing structure working normally is set to 0.72-0.94, based on the probability of the high-temperature resistant sealing material in normal working condition. The conditional probability of a leaky seal but a randomly stable internal pressure difference is taken as 0.18-0.25, determined according to the known conditional probability calculation standard in probability theory. The prior probability of sealing structure failure is 0.06-0.28, determined based on industry statistics on the probability of sealing material aging failure. The internal cavity air pressure fluctuation coefficient is 0.92-0.99, derived based on the general standard for stable control of casting cavity air pressure. The aging correction factor for the sealing structure is 0.90-0.98, determined according to the standard for high-temperature aging characteristics of refractory sealing materials. This is the aging rate factor for sealing tape, with a value of 0.92-0.99, determined according to the industry standard for aging rate of high-temperature resistant sealing tape.

[0138] The sealing effect of the shell mold directly determines the safety of the casting process and the internal quality of the casting. Large-sized shell molds are much more difficult to seal than conventional small parts due to their large size, many joint gaps, and large open area at the ends. Traditional sealing processes generally use a single sealing material to simply apply or wrap it. There is no design for different number of coating layers according to the gap width, no quantitative control of the density of the sealing layer, and no real-time monitoring and quantitative judgment of the sealing status of the cavity. The judgment of whether the seal is qualified is entirely based on the experience of the operators. Problems such as cracking of the sealing layer, micropore leakage, and air intake into the cavity are often not detected in time, which leads to serious quality defects such as molten metal leakage, casting porosity, and slag inclusions. First, we match the corresponding number of covering layers according to the actual width of the shell splicing gap to complete the physical sealing. Then, we use a thickness gauge to detect the uniformity of the covering layer and adjust the covering force in combination with the density factor. Next, we set up sensor points in the shell cavity according to the specifications to collect real-time environmental data. Finally, we substitute multiple parameters into the probability formula to calculate the probability of maintaining a tight and steady state. The quantitative value is used as the admission condition for the casting process. Each step is matched with the actual sealing operation on site.

[0139] Among these technologies, refractory mortar can be used to seal tiny pores in the shell-shaped gaps, while high-temperature resistant sealing tape possesses excellent sealing properties and structural strength. The alternating wrapping process combines the advantages of both materials to effectively seal gaps of different widths. Distributed sensing points are arranged according to the cavity volume and avoid critical forming areas of the casting, enabling comprehensive monitoring of the internal environment without interfering with the normal forming of the casting. High-frequency data acquisition can reflect the dynamic changes in the sealing state in real time. In addition, the Bayesian decision calculation integrates multiple influencing factors such as the prior reliability of the sealing structure, real-time air pressure fluctuations, and the aging rate of the sealing material to achieve quantitative and accurate determination, ensuring that the pouring process is only initiated when the sealing state is completely up to standard.

[0140] The Bayesian decision calculation formula of this invention couples three core influencing factors: the prior performance of the sealing structure, the real-time internal cavity air pressure fluctuation, and the aging rate of the sealing material. Based on the well-known Bayesian probability theory, it constructs a shell cavity sealing steady-state quantitative accounting model to accurately determine the shell sealing state and strictly control the entry conditions of the casting process.

[0141] This solution first achieves physical sealing of the shell assembly gaps and ports by alternately wrapping with refractory mortar and sealing tape. Then, the quality of the sealing layer is adjusted in real time to eliminate potential microscopic leakage. Next, distributed micro-sensors collect real-time data on air pressure, temperature, and humidity within the shell cavity, extracting the air pressure fluctuation coefficient that reflects the sealed state. Combined with parameters such as the prior probability of the sealing structure, aging correction coefficient, and aging rate factor, these parameters are substituted into a Bayesian formula to calculate the probability of maintaining a sealed steady state. Finally, a quantified probability value is used to determine whether the cavity's sealed state meets the casting requirements. This solution achieves precise quantitative assessment of the sealed steady state through Bayesian probability determination, establishing clear entry standards for the casting process and preventing defects such as casting leakage and casting porosity caused by unqualified sealing.

[0142] Optionally, the specific steps for controlling the pouring of molten metal in step S5 are as follows:

[0143] The probability of maintaining a sealed steady state within the housing cavity was confirmed by collecting parameters from distributed sensing points inside the housing. ≥0.95; Call the dual-station time-series iterative operation model, input the valid environmental monitoring dataset and dual-station operation parameters, and generate the initial staggered peak avoidance arrangement rhythm;

[0144] Thermocouples were used to detect the temperature of the molten metal, which was controlled between 1500℃ and 1650℃. The viscosity correction factor of the molten metal was calculated based on the temperature values. , The value range is 0.86 to 0.97, and the staggered arrangement rhythm is adjusted.

[0145] The pouring flow rate is set according to the wall thickness of the casting. When the wall thickness is 4mm to 6mm, the flow rate is controlled at 0.6m / s to 0.8m / s; when the wall thickness is 7mm to 12mm, the flow rate is controlled at 0.4m / s to 0.6m / s; and when the wall thickness is 13mm to 18mm, the flow rate is controlled at 0.3m / s to 0.4m / s.

[0146] A layered, flow-limited, and progressive pouring operation is implemented using a pouring ladle. A flow controller is used to control the pouring flow rate in real time with a response time of no more than 0.5 steps (S). At the same time, a pressure sensor is used to monitor the internal pressure of the cavity in real time and control the pressure between 0.1 MPa and 0.3 MPa.

[0147] In the production of large-size castings in batches on a dual-station shell molding line, molten metal pouring is the core process that determines the quality of the casting. Large-size castings have a wide range of wall thicknesses and complex cavity structures. Dual-station production lines also have problems such as conflicting process sequences and uncoordinated work rhythms. Traditional pouring processes use fixed temperature, flow rate, and pressure parameters, which cannot adapt to the work sequence of dual stations, do not take into account the viscosity changes caused by fluctuations in molten metal temperature, and cannot match differentiated pouring flow rates according to different wall thickness areas of the casting. This easily leads to problems such as insufficient filling, turbulent air entrapment, cavity sand flushing, and interference between dual-station processes. We have developed a systematic casting control system that includes: closed-loop state verification, dual-station timing matching, molten metal viscosity correction, wall thickness zone flow restriction, and high-pressure closed-loop control. This system first verifies the closed-loop state of the shell to ensure casting safety, then matches the dual-station operation sequence to avoid process conflicts, then adjusts the viscosity according to the molten metal temperature to adapt to the casting rhythm, then sets the flow rate according to the casting wall thickness to ensure filling quality, and finally maintains stable working conditions of the cavity through real-time closed-loop control of flow rate and pressure.

[0148] The filling requirements of large-sized castings vary significantly depending on the wall thickness. In thin-walled areas, excessively low flow rates can easily lead to incomplete filling defects, while in thick-walled areas, excessively high flow rates can easily cause turbulent air entrapment. Zoned flow restriction can precisely match the filling characteristics of different areas. The rapid response of the flow controller, combined with the real-time monitoring of the pressure sensor, forms a dual closed-loop control of flow rate and pressure, ensuring that the pressure inside the cavity is always maintained within a reasonable range, which guarantees full filling without damaging the sand mold due to excessive pressure.

[0149] The working principle of this solution is completely consistent with the actual production logic of dual-station casting of large-size castings. First, the safety conditions for casting are confirmed by the pre-sequence closed state detection. Then, the dual-station timing model is called to generate the staggered operation rhythm, the temperature of the molten metal is detected in real time and the viscosity parameters are corrected. Differentiated flow rates are set according to the casting wall thickness and the regions are divided. Finally, stable filling is achieved through rapid closed-loop control of flow rate and pressure.

[0150] This solution dynamically adjusts the temperature and viscosity of the molten metal to match the pouring rhythm with the actual flow state of the molten metal in real time, eliminating the interference of temperature fluctuations on the filling process and improving the stability and consistency of the pouring process. Furthermore, this solution employs a casting wall thickness-zoned flow-limiting pouring process, matching differentiated flow rates to different wall thickness areas. This avoids forming defects such as incomplete filling, turbulent air entrapment, and shrinkage cavities from the filling mechanism perspective, significantly improving the internal quality of the casting; while ensuring full filling, it also avoids the risk of sand mold erosion.

[0151] Optionally, in step S6, the specific steps for controlling the solidification of the casting are as follows:

[0152] After the molten metal is filled into the mold, an ultrasonic testing device is used to test the wall thickness parameters of the casting at a speed of 0.5 m / s. The testing accuracy is controlled at 0.02 mm. Based on the wall thickness parameters, the casting is divided into three solidification control zones: thin-walled zone (4 mm to 6 mm), medium-thickness zone (7 mm to 12 mm), and thick-walled zone (13 mm to 18 mm).

[0153] The distribution of hot spots in the casting was detected using an infrared thermometer, and the area ratio of hot spots was calculated to obtain a correction factor for the distribution of hot spots in the casting. ;

[0154] Setting the contact thermal resistance between the mold and the metal base Interfacial air film interference coefficient Temperature gradient correction factor ;

[0155] The thickness of the interfacial gas film gap during the solidification process of the casting is monitored in real time using thermocouples. Temperature difference between the inner and outer surfaces of the casting and solidification time Calculate the solidification time coupling factor ;

[0156] The overall thermal resistance at the shell-to-casting interface is calculated using a dynamic interface thermal resistance formula. Substituting the aforementioned parameters into the formula yields the overall interface thermal resistance. ,according to The values ​​are set with gradient heat dissipation parameters for the three solidification control zones: the heat dissipation rate in the thin-walled zone is controlled at 0.8℃ / min~1.2℃ / min, the rate in the medium-thick zone is controlled at 0.5℃ / min~0.8℃ / min, and the rate in the thick-walled zone is controlled at 0.3℃ / min~0.5℃ / min, thus completing the solidification control of the casting.

[0157] The formula for calculating the dynamic thermal resistance of the interface is as follows:

[0158] .

[0159] In the formula: The value is the combined thermal resistance at the interface between the shell and the casting, expressed in m²·℃ / W, and determined according to the general design specifications for heat transfer at the casting interface. The contact thermal resistance between the mold and the metal base is expressed in m²·℃ / W, with a value of 0.12 to 0.14, set according to the JB / T6047 standard for casting heat exchange base parameters. The interfacial gas film interference coefficient is taken as 0.65 to 0.72, and is determined based on the known laws of the heat insulation mechanism of the solidified gas film in the casting. The thickness of the interfacial gas film gap, in mm, ranges from 0.04 to 0.22, and is set according to the industry standard range for gas film gaps formed by solidification shrinkage of large-sized castings. The solidification time coupling factor has a value of 0.88 to 0.96 and is derived based on the correlation standard between the solidification time of the casting and the heat transfer efficiency. This is the temperature gradient correction coefficient, in m²·℃ / W, with a value ranging from 0.01 to 0.03, determined according to the casting temperature gradient heat dissipation control specifications. This is the temperature difference between the inner and outer surfaces of the casting, in °C, and the value ranges from 50 to 150, set according to the industry control range of solidification temperature field distribution for large-size castings. The solidification time of the casting is expressed in minutes and ranges from 60 to 300, determined according to the standard for the complete solidification cycle of large-size shell castings. The value of the heat transfer correction factor for the distribution of hot spots in the casting ranges from 0.90 to 1.00 and is determined according to the industry-standard heat transfer correction corresponding to the proportion of hot spot area in the casting.

[0160] In the actual working conditions of mass production of large-size castings on a dual-station shell molding line, the solidification process is the core process that determines the internal structure of the casting and eliminates shrinkage porosity and crack defects. Traditional processes generally adopt a uniform heat dissipation mode throughout the entire area, which completely ignores the objective physical characteristics of large-size castings, such as uneven wall thickness, dispersed heat points, gas film formation during solidification, and dynamic changes in interface heat transfer. This process cannot adapt to the solidification requirements of different areas and is prone to fatal defects such as uneven heat dissipation, excessive shrinkage stress, shrinkage porosity in heat points, and cracks.

[0161] Most failure defects in large-sized castings are concentrated at hot spots and in areas where thickness transitions occur. Because the interfacial heat transfer resistance is not quantified and heat dissipation is not zoned, a single heat dissipation rhythm inevitably leads to localized solidification that is too fast or too slow, resulting in shrinkage stress and internal porosity. After the casting is filled, the internal molten metal begins to gradually crystallize and solidify. Due to shrinkage, tiny air film gaps form at the interface between the shell and the casting. These air films impede heat transfer. Simultaneously, a natural temperature difference exists between the inside and outside of the casting, and hot spots are high-risk areas for solidification defects. These factors collectively determine the actual heat transfer capacity of the interface.

[0162] This solution first uses ultrasonic waves to precisely scan the entire wall thickness to divide the solidification area. Then, infrared thermometry is used to locate the distribution of hot spots and quantify the hot spot correction coefficient. Simultaneously, dynamic parameters such as air film gap, temperature difference, and solidification time are collected in real time. These parameters are then substituted into the dynamic thermal resistance formula to accurately calculate the overall interface thermal resistance. Finally, based on the thermal resistance value, the gradient heat dissipation parameters of the corresponding area are matched so that each area matches the heat dissipation rhythm according to its own solidification characteristics.

[0163] The interface dynamic thermal resistance calculation formula of this invention integrates basic contact thermal resistance, air film insulation interference, solidification time coupling, temperature gradient change, and thermal distribution correction factors into one, which fully conforms to the dynamic heat transfer mechanism of the solidification process of large-size castings. In the overall invention scheme, it plays the role of quantifying interface heat transfer capacity and supporting the control of zoned gradient heat dissipation. Through formula coupling calculation, a comprehensive thermal resistance value that can truly reflect the actual working conditions is obtained, and the quantified thermal resistance is used as the core basis for setting heat dissipation parameters.

[0164] This dynamic thermal resistance calculation formula adopts a segmented structure that highly matches the actual solidification heat transfer process, incorporating inherent thermal resistance, gas film interference correction, and temperature gradient thermal point coupling. It takes into account the coupling effects of static basic heat transfer, dynamic gas film barrier, duration, temperature gradient, and thermal points, establishing a zoned gradient solidification control system centered on dynamic interface thermal resistance calculation. The various steps of the scheme work synergistically. Wall thickness zoning provides a classification basis for differentiated heat dissipation, thermal point correction optimizes areas with high defect incidence, gas film and temperature parameters reflect the dynamic working conditions of the interface in real time, and the thermal resistance formula couples multi-dimensional parameters into a unified quantitative index. Gradient heat dissipation then adapts to the solidification needs of each region. Through multi-factor coupling, it accurately recreates the actual heat transfer state of the entire solidification process, making heat dissipation control data-driven. Furthermore, this scheme establishes a solidification control mode that links wall thickness zoning and thermal point correction, setting differentiated heat dissipation rhythms for different structural regions and areas with high defect incidence, balancing the overall solidification rate of the casting from the perspective of heat transfer mechanism. In addition, this scheme incorporates dynamic variables such as air film gap, temperature gradient, and solidification time into a unified accounting system to achieve coordinated control of all operating parameters during the solidification process, avoiding the limitations of single parameter control.

[0165] Optionally, in step S7, the specific steps of the casting post-processing and dual-station cyclic operation are as follows:

[0166] Thermocouple monitoring data confirmed that the casting had completely solidified and settled, meaning the surface temperature of the casting had dropped to room temperature ±5℃ (room temperature 20℃~25℃).

[0167] A step-by-step disassembly method is adopted, using disassembly tools to gradually disassemble the shell structure, with the disassembly speed controlled at 0.05m / s to 0.1m / s, and the disassembly sequence proceeding from non-critical parts to critical parts;

[0168] After disassembly, special finishing tools are used to finely finish the flash and residual sand on the casting. The finishing accuracy of the flash is controlled at 0.01mm to ensure that the residual sand is cleaned up and there is no residue. After finishing, the surface roughness of the casting Ra≤1.6μm.

[0169] The dimensions and surface quality parameters of the finished castings were inspected using 3D inspection equipment, and the surface roughness data of the castings were collected using a roughness meter to calculate the surface roughness influencing factor. , The value range is 0.96 to 1.04;

[0170] Include dimensional parameters, surface quality parameters, and Substitute the data into the multidimensional detection model to calculate the overall dimensional deviation of the casting, ensuring that the deviation is ≤ ±0.1mm;

[0171] Set the process division and rotation interval between the two sets of work units to 40min to 90min, and rotate and replace the process division of the two sets of work units;

[0172] Collect production process parameters and quality inspection data for the first n-1 batches to form a batch dataset, set batch deviation weighting factors, calculate the quality fluctuation coefficient for the first n-1 batches, and calculate the batch stability correction factor based on the quality fluctuation coefficient. , The value range is 0.97 to 1.03;

[0173] Batch deviation weighting factor, Substitute multiple batches of data into the iterative optimization model, obtain cyclic calibration parameters, call the cyclic calibration parameters to adjust the workstation operation parameters, and continuously carry out production line flow operations.

[0174] As production batches increase, tooling wear, minor environmental changes, and slight drifts in material properties can all cause the original process parameters to gradually deviate from optimal conditions, resulting in batch-by-batch fluctuations in casting quality. Traditional production models cannot automatically identify this drift and can only rely on periodic manual adjustments, which is not only labor-intensive but also prone to quality control delays. This has long been a pain point in the industry for dual-station continuous production lines. We first complete physical disassembly and surface finishing, then use a 3D scanner for full-area quantitative inspection, extract batch quality deviation data to calculate the fluctuation coefficient, combine it with the production line load status to determine the process rotation rhythm, and finally use historical parameters from multiple batches to substitute into the cyclic calibration formula to calculate the parameter correction amount and update the process baseline of all preceding processes in reverse.

[0175] In addition, mechanical vibration disassembly can achieve rapid shell separation without damaging the casting body, and grinding and finishing can eliminate surface forming defects; three-dimensional full-domain scanning replaces traditional sampling inspection, which can accurately capture the size and shape deviation of the entire batch of castings and objectively reflect the stability of the current process conditions; the batch quality fluctuation coefficient quantitatively characterizes the degree of quality dispersion and serves as the core basis for process rotation and process correction; the production line load correction factor adapts to different production conditions such as peak, off-peak, and low-peak, dynamically adjusts the dual-station rotation interval, and avoids station idleness or process congestion; the cyclic calibration attenuation calculation formula integrates multiple factors such as quality fluctuation, production line load, historical batch parameters, and natural process attenuation to quantify the process parameter iterative correction amount, realizing the transformation from passive manual parameter adjustment to active closed-loop optimization.

[0176] This solution first accumulates batch deviation data through post-processing full-domain quality inspection, calculates the quality fluctuation coefficient to characterize the current process stability, then determines the load correction factor by combining the production line operating load, retrieves the original process baseline parameters of multiple consecutive batches for average statistics, superimposes the process natural decay correction coefficient, and finally calculates a reasonable parameter iteration correction amount. It then synchronously updates the basic parameters of all processes from step S1 to step S6 in reverse, so that the entire manufacturing process is always maintained in the optimal operating range.

[0177] This solution enables comprehensive quantitative inspection of the post-processing quality of large-size castings, accurately capturing dimensional and positional deviations across the entire batch, providing objective and reliable data support for process optimization. Secondly, this solution establishes a dynamic rotation mechanism for dual-station processes, adjusting the rotation interval in real time based on quality fluctuations and production line load to avoid station congestion and process bottlenecks, continuously improving the smoothness of dual-station production line operation.

[0178] In addition, this solution constructs a multi-factor coupled process parameter cyclic calibration system to quantitatively compensate for parameter drift caused by natural process decay and operating condition fluctuations, realize automatic iterative updates of process benchmarks, adapt to long-term continuous batch production of dual-station shell molding lines, and reduce manual operation and maintenance costs while stabilizing casting quality.

[0179] Optionally, in step S5, the calling step of the dual-station time-series iterative job model is as follows:

[0180] Extract the dual-station timing fit obtained above And a valid environmental monitoring dataset, combined with the molten metal temperature and molten metal viscosity correction factor detected in step S5. Input a dual-station time-series iterative operation model;

[0181] Model by To meet the control requirement of ≥0.85, we dynamically adjust the staggered scheduling rhythm and correct the lag time of pouring start, controlling the lag time to 1min to 3min. At the same time, we combine the pouring flow rate corresponding to the casting wall thickness and synchronously calibrate the transfer connection time of the dual station, controlling the transfer connection time to 2min to 5min, to ensure that the pouring operation and the dual station sequence are accurately matched and to avoid process connection blockage.

[0182] In the collaborative operation of a dual-station shell molding line, the casting process, as the core forming step, requires precise matching of its start timing and rhythm with the entire process of shell making, transfer, cooling, and post-processing. Traditional production lines only have a fixed casting sequence, which easily leads to problems such as casting starting too early, resulting in the shell not being ready, starting too late, causing the molten metal to cool down, and mismatched transfer connection times causing process delays. These issues directly affect the casting quality and production line efficiency. This solution first extracts the previously calculated timing compatibility and environmental data, then inputs the current molten metal state and casting process parameters. Through dynamic adjustment of the staggered rhythm using a model, it corrects casting lags and transfer connection times, ultimately achieving precise timing coordination between casting and the entire dual-station process, matching the on-site model call operation. In addition, the dual-station timing adaptation serves as a fundamental constraint for model invocation, ensuring that the overall timing does not deviate from the optimal range; the current production environment status is reflected through an effective environmental monitoring dataset, eliminating the interference of environmental fluctuations on the timing; the molten metal temperature and viscosity correction factor directly determine the fluidity of the molten metal and the pouring window period, and dynamically adjusting the pouring rhythm can match the real-time status of the molten metal; the pouring start lag time reserves a window for the final shell inspection and equipment preparation, ensuring safety while avoiding waiting waste.

[0183] Optionally, it also includes a multi-process parameter coupling optimization step, specifically:

[0184] Extracting the shell-shaped linear shrinkage amount in step S2 The probability of maintaining a sealed steady state within the shell cavity in step S4. Metal liquid viscosity correction factor in step S5 and the overall thermal resistance of the interface in step S6 A multi-process parameter coupling optimization model is constructed, and coupling weight coefficients are set, where... The weights range from 0.22 to 0.28. The weights range from 0.30 to 0.38. The weights are 0.18 to 0.25. The weights range from 0.12 to 0.18.

[0185] Calculate the coupling fit of multi-process parameters ,when When the value is less than 0.90, adjust the parameters of each process in order of weight priority, prioritizing the adjustment of the closed-state parameters and the pouring flow rate parameters, until... ≥0.90;

[0186] The formula for calculating the multi-process parameter coupling fit degree is as follows:

[0187] .

[0188] In the formula: To ensure the compatibility of multi-process parameter coupling, the control threshold is ≥0.90, set according to the industry standard for multi-process collaborative stability. This is the standard value for the unidirectional linear shrinkage of the shell shape, in mm, with a value of 0.65 mm, set according to the general standard for the dimensional accuracy of large-size shell shapes; The actual linear shrinkage in one direction of the shell is measured in mm, ranging from 0.15 to 1.20 mm, and the calculation range is the same as in step S2. The probability of maintaining a sealed steady state within the shell cavity is ≥0.95, and the control threshold in step S4 is used as before. The viscosity correction factor for molten metal is set to 0.86–0.97, following the range specified in step S5. This is the standard value for the combined thermal resistance at the interface between the shell and the casting, in m²・℃ / W, with a value of 0.13 m²・℃ / W, set according to the heat transfer standard for the casting interface. The actual combined thermal resistance at the interface between the shell and the casting is expressed in m²·℃ / W, and the calculation range is the same as in step S6. The weighting coefficients are set according to the degree of influence of each process on the casting quality, and the sum is 1.

[0189] In the full-process production of the dual-station shell molding line, step S2 shell shrinkage, step S4 sealing state, step S5 pouring viscosity, and step S6 interface thermal resistance are the core process parameters that determine the final quality of the casting. The four process parameters achieve global coordination through strong coupling. Therefore, a multi-process parameter coupling optimization step is added to form an optimized process and set weights according to the degree of quality impact. The fit is calculated through the coupling formula, and the parameters are adjusted according to the weight priority until the fit meets the standard, ensuring that each step matches the process optimization operation.

[0190] In addition, the coupling fit of multi-process parameters is the core indicator for quantifying the synergy of the entire process. The weight coefficient reflects the degree of contribution of each process to the quality. Among them, the sealing state and casting parameters are given the highest weight and are adjusted first to quickly improve the fit. Shell shrinkage and interface thermal resistance are the basic process parameters, and subsequent fine-tuning can achieve global synergy. The global coupling fit is obtained by weighted summation, which intuitively reflects the synergy status of the entire process.

[0191] This solution couples and superimposes the four core process parameters according to their quality impact weights, adopting a weighted summation and ratio correction architecture. In the overall invention, it plays a role in global process coupling optimization and ensuring the stability of casting quality. By first extracting the core measured parameters of each process, introducing industry standard values ​​for deviation correction, and obtaining the coupling fit degree by weighted summation, the deviation process parameters are adjusted according to the weight priority to achieve global process synergy optimization. The scattered process parameters are integrated into global synergy quantitative indicators. Each part corresponds to the degree of deviation of a single process and the global contribution, which is highly consistent with the overall process synergy mechanism.

[0192] This solution achieves global coupling and quantitative optimization of core process parameters throughout the entire dual-station process. Parameters are adjusted according to weight and priority, with priority given to key processes, which greatly improves the efficiency of process optimization, quickly achieves global collaboration, continuously ensures the stability of process collaboration, reduces casting quality fluctuations, and is suitable for continuous batch production in dual-station environments.

[0193] Optionally, it also includes an adaptive adjustment step for abnormal operating conditions, specifically:

[0194] Key parameters of each process are collected in real time, and abnormal thresholds are set. The abnormal threshold for ambient temperature in step S1 is <15℃ or >40℃, the abnormal threshold for internal air pressure in step S4 is <30Pa or >85Pa, the abnormal threshold for pouring pressure in step S5 is <0.1MPa or >0.3MPa, and the abnormal threshold for heat dissipation rate in step S6 is ±0.2℃ / min.

[0195] When a process parameter is detected to exceed the abnormal threshold, an adaptive adjustment mechanism is activated. Based on the type and degree of deviation of the abnormal parameter, the correction parameters of the corresponding process are called to adjust the relevant process operations, and the parameters of related processes are coordinated for correction.

[0196] After the anomaly is eliminated, continuous monitoring is performed for 30 minutes to ensure that the parameters stabilize within the normal range and that the coupling and compatibility of multi-process parameters are good. Normal production can only resume when the deviation is ≥0.90; among which, the abnormal deviation correction amount The calculation formula is as follows:

[0197] ;

[0198] In the formula: This is the correction amount for abnormal parameters, with the unit consistent with the corresponding abnormal parameter, and is determined according to the abnormal deviation control specifications. This is the anomaly correction coefficient, ranging from 0.95 to 1.05, and is dynamically adjusted according to the anomaly type of different processes. These are the actual parameter values ​​for abnormal processes, and are real-time data collected on-site. For abnormal process parameters, the standard values ​​should be used for each process. This is the linkage correction coefficient for related processes, with a value ranging from 0.92 to 0.98, determined based on the standard for the strength of coupling between processes.

[0199] In the continuous production process of a dual-station shell molding line, key parameters such as ambient temperature, cavity air pressure, pouring pressure, and heat dissipation rate are easily affected by factors such as equipment fluctuations, environmental changes, and material differences, which can lead to abnormalities. If only the abnormal process is adjusted after an abnormality is detected, it can easily cause deviations in parameters of related processes, resulting in production interruption, scrapped castings, and significant losses to the production line.

[0200] Therefore, an adaptive adjustment step for abnormal operating conditions is added, forming a real-time control process. This process strictly follows the abnormal handling sequence, collects key parameters in real time, compares the status with abnormal thresholds to determine the condition, initiates adaptive adjustment, links and corrects related processes, and resumes production after continuous monitoring and confirmation that the standards have been met. Furthermore, by overlaying abnormality types, deviation degrees, and process coupling factors, and employing a mathematical framework of absolute value deviation, ratio normalization, and coefficient correction, the adjustment range for abnormalities is precisely quantified. This framework plays a crucial role in the overall invention in rapidly adaptively correcting abnormal operating conditions and ensuring continuous production line operation. This solution first collects parameters in real time to determine abnormalities, quantifies the correction amount using formulas, adapts correction coefficients according to process type, links and adjusts related processes in a coordinated manner, and finally verifies that the standards have been met before resuming production.

[0201] This solution strictly follows the anomaly handling sequence, with data collection and judgment preceding calculation and adjustment, and monitoring and recovery following. The process connections are closely aligned with actual production, eliminating the need to build complex modules. Technical personnel can directly build the model and implement adjustments based on the recorded parameter ranges and calculation logic.

[0202] This formula for calculating abnormal deviation correction is based on the real-time control characteristics of abnormal industrial operating conditions. Each part corresponds to the degree of abnormality, the adjustment range, and the process correlation characteristics, which is highly consistent with the abnormal control mechanism to achieve adaptive and precise correction. Addressing the problems of delayed abnormal response, blind adjustment, and uncontrolled correlation in existing dual-station production lines, an adaptive linkage adjustment system for abnormal operating conditions is constructed. Real-time data acquisition enables early detection of abnormalities, threshold judgment ensures accurate identification, quantitative correction guarantees precise adjustment, linkage correction avoids correlation deviation, and continuous monitoring ensures stability.

[0203] This solution enables real-time monitoring and precise anomaly identification of key process parameters on the production line, proactively mitigating the risk of anomaly escalation and reducing production losses. By calculating correction amounts using quantitative formulas, it achieves adaptive and precise adjustments under abnormal operating conditions, improving response speed and adjustment accuracy. Furthermore, this solution employs a collaborative correction mechanism for related processes to avoid cascading deviations caused by adjustments to a single process, ensuring the stable coordination of the entire process. It ensures production resumes only after the anomaly is completely eliminated, preventing recurrence and better adapting to the needs of continuous automated dual-station production.

[0204] The present invention also provides a large-size shell casting manufacturing system based on a dual-station shell profile line, wherein the system stores a computer program, and the computer program implements the above-described method when executed by a processor.

[0205] The following is a detailed explanation using specific embodiments:

[0206] Example 1: Taking the dual-station coated sand mold casting production of heavy-duty vehicle gearbox housing as an example.

[0207] This embodiment is for a heavy-duty vehicle gearbox housing with external dimensions of 1750mm × 1050mm × 620mm. The steps are as follows:

[0208] Step S1: Environmental Parameter Collection and Screening, and Dual-Station Layout Control: Data was collected using temperature and humidity sensors, a dust detector, and a barometer. The collection frequency was once every 3 minutes, and after 10 consecutive sets of data, outliers were removed using the 3σ criterion. The valid environmental parameters were determined as follows: temperature 24℃, relative humidity 52%, dust concentration 32mg / m³, and air pressure 101.3kPa. The distance between the two work units was 4.0m, with a temperature difference ΔT = 5℃ and a humidity difference ΔRH = 10%. The shell-making time was also specified. =90min, pouring time =180min; Substitute into the timing adaptation coupling calculation formula, and take =0.08、 =0.92、 =0.05、 =0.88, calculated as follows =0.92≥0.85; the width of the transfer path is defined as 1.8m, and environmental fluctuations are monitored every 5 minutes.

[0209] Step S2, Shell Demolding, Deformation Detection and Static Control: Parallel lifting demolding speed 0.15m / s, stress concentration zones arranged 6 per m. 2 The testing point was set with a dial indicator accuracy of 0.01 mm; the ambient temperature was 24℃ and the humidity was 52% for a settling time of 60 minutes; data was collected. =0.90、 =3.0%, =24℃ =25℃, substituting into the multi-factor linkage correction calculation formula, the result is... =0.65mm, which falls within the control range of 0.15mm~1.20mm.

[0210] Step S3, Shell assembly and fixing: Laser positioning accuracy 0.005mm, calibration deviation 0.03mm. =1.00; the hydraulic preload increases in increments of 25%, starting at 6kN and ending at 18kN; the splicing gap is 0.06mm, the form and position deviation is 0.22mm, and rigid fixing is completed.

[0211] Step S4, Shell sealing and airtightness monitoring: Gap width 2.0mm, 3 layers of alternating wrapping, single layer thickness 0.8mm; =0.95, with 4 sensor points arranged inside the cavity; substituting into the Bayesian decision formula, the result is... =0.97≥0.95, control the internal air pressure 50Pa, temperature 24℃, and humidity 48%.

[0212] Step S5, Molten Metal Pouring Control: Molten metal temperature 1580℃, =0.91; flow velocity in thin-walled region 0.7m / s, medium-thick region 0.5m / s, thick-walled region 0.35m / s, flow response 0.4 steps / s, cavity pressure 0.2MPa; corrected casting lag time 2min, transfer connection time 3min.

[0213] Step S6, Casting solidification control: Ultrasonic thickness measurement accuracy 0.02mm, dividing into thin-walled / medium-thick / thick-walled zones; =0.95, substituting into the formula for calculating the dynamic thermal resistance of the interface, we get... =0.57m²・℃ / W; set the gradient heat dissipation rates to 1.0℃ / min, 0.6℃ / min, and 0.4℃ / min respectively.

[0214] Step S7, Post-processing and Cyclic Operation: After the casting is cooled to 25℃, it is disassembled, and the finishing accuracy is 0.01mm, the roughness Ra=1.2μm; the dimensional deviation is 0.08mm, the process rotation interval is 60min, and the process parameters are iteratively optimized.

[0215] Multi-process coupling optimization: Extract core parameters and substitute them into the calculation formula to calculate the coupling fit. =0.94≥0.90, no abnormal operating conditions throughout the process.

[0216] Example 2: Taking the automated production of the boom shell of an engineering machinery excavator in a real-world scenario as an example, with two workstations on the shell profile line.

[0217] This embodiment focuses on the production method of an excavator boom housing with external dimensions of 2400mm × 1200mm × 780mm, based on the method described above:

[0218] Step S1: Screening valid environmental parameters: temperature 27℃, humidity 58%, dust 38mg / m³, air pressure 99.8kPa; distance between dual workstations 5.0m, ΔT=7℃, ΔRH=14%. =120min =240min, calculated =0.90≥0.85.

[0219] Step S2: Demolding speed 0.1 m / s, 8 detection points / m 2 The temperature was set at 28℃ and the humidity at 58% for 90 minutes. =0.92、 =3.5%, calculated as follows =0.82mm.

[0220] Step S3, laser calibration deviation 0.04mm, =1.02, initial preload 7kN, final preload 20kN; splicing gap 0.07mm, form and position deviation 0.25mm.

[0221] Step S4, gap width 3.0mm, 4 layers of wrapping, =0.96; the calculated probability of closure is... =0.96≥0.95, internal air pressure 60Pa, temperature 27℃, humidity 55%.

[0222] Step S5, the temperature of the molten metal is 1620℃. =0.94; zone flow rate matches wall thickness, cavity pressure 0.25MPa, casting lag 3min, transfer connection 4min.

[0223] Step S6, calculation results =0.62m²・℃ / W, with heat dissipation rates set at 1.1℃ / min, 0.7℃ / min, and 0.35℃ / min.

[0224] Step S7: Casting roughness Ra = 1.4 μm, dimensional deviation 0.09 mm, rotation interval 80 min; coupling fit. =0.92≥0.90; If the heat dissipation rate deviates from 0.25℃ / min during production, initiate an abnormal adjustment and calculate the correction amount using the formula. =0.22, the parameter stabilized 30 minutes after linkage adjustment.

[0225] Example 3: Taking the high-precision production of dual-station shell profiles in the actual scenario of wind turbine yaw bearing housing as an example.

[0226] This embodiment is for a wind turbine yaw bearing housing with external dimensions of 3100mm×1700mm×950mm, manufactured according to the above method:

[0227] Step S1: Screening valid environmental parameters: temperature 22℃, humidity 46%, dust 28mg / m³, air pressure 102.1kPa; distance between dual workstations 4.5m, ΔT=4℃, ΔRH=8%. =150min =300min, calculated =0.93≥0.85.

[0228] Step S2: Demolding speed 0.12 m / s, 7 detection points / m 2 The temperature was set at 22℃ and the humidity at 46% for 100 minutes. =0.94、 =3.2%, calculated as follows =0.95mm.

[0229] Step S3, laser calibration deviation 0.03mm, =0.99, initial preload 8kN, final preload 20kN; splicing gap 0.06mm, form and position deviation 0.24mm.

[0230] Step S4, gap width 3.2mm, 4 layers of wrapping, =0.97; the calculated probability of closure is... =0.98≥0.95, internal air pressure 70Pa, temperature 22℃, humidity 42%.

[0231] Step S5, the temperature of the molten metal is 1650℃. =0.96; zone flow rate matches wall thickness, cavity pressure 0.3MPa, casting lag 3min, transfer connection 5min.

[0232] Step S6, calculation results =0.68m²・℃ / W, with heat dissipation rates set at 1.2℃ / min, 0.8℃ / min, and 0.5℃ / min.

[0233] Step S7: Casting roughness Ra = 1.6 μm, dimensional deviation 0.10 mm, rotation interval 90 min; coupling fit. =0.95≥0.90, no abnormalities were found throughout the process.

[0234] As can be seen from the above embodiments, the three embodiments are respectively adapted to three typical large-size castings: gearbox housing, excavator boom, and wind turbine bearing housing. The dimensional accuracy of the castings can be stably controlled within ±0.1mm, the shell assembly clearance qualification rate reaches 99.2%, the internal shrinkage porosity and gas porosity defect rate of the castings is reduced by more than 85%, the production efficiency of the dual-station production line is increased by 40%, and the overall finished product rate of the castings is increased from 82% to 98.5%. At the same time, the multi-process coupling optimization and abnormal adaptive adjustment function effectively solves the problems of timing conflicts, excessive shell deformation, and sealing failure in traditional dual-station production lines, and can meet the needs of large-scale and stable production of large-size shell castings with different specifications and different precision requirements.

[0235] 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.

[0236] Any aspects of this invention not described in detail are well-known to those skilled in the art.

Claims

1. A method for manufacturing large-size shell castings based on a dual-station shell profile, characterized in that, Includes the following steps: S1. Collect and remove anomalies from the on-site environmental parameters of the foundry workshop to form an effective environmental monitoring dataset; complete the layout division of the dual workstations, calibration of the time sequence operation arrangement parameters, and planning of the workpiece transfer path between workstations, and monitor the fluctuation of environmental parameters of the dual workstations in real time. S2. Demolding of large-size shells, completing multi-dimensional deformation detection and data recording of the shell; static placement of shell sections, real-time monitoring of shell deformation status and data correction; S3. Perform pre-calibration and deviation correction of the split shell reference alignment, complete the pre-tightening and locking of the shell assembly, detect the assembly accuracy and form and position parameters, and achieve overall rigid fixation of the shell. S4. Seal and cover the gaps and ports of the shell splicing, set up monitoring points for the internal environment, collect the internal environment parameters in real time, and adjust the sealing measures to maintain the sealed and isolated state of the cavity. S5. After the sealing of the inner cavity of the shell meets the standard, the pouring rhythm is generated, and layered flow-limited pouring is implemented according to the structural characteristics of the casting, and the pouring process parameters are controlled in real time. S6. After the molten metal is filled into the mold, the wall thickness distribution of the casting is detected and the solidification control area is divided. Gradient heat dissipation parameters are set according to the area to complete the solidification control of the casting. S7. After confirming that the casting has completely solidified and solidified, complete the shell disassembly, casting repair and quality inspection, rotate the work of the two workstations, and call the cycle calibration parameters to maintain the continuous flow of the production line.

2. The method for manufacturing large-size shell castings based on a dual-station shell profile according to claim 1, characterized in that, In step S1, the specific steps for environmental parameter collection and screening, and dual-station layout control are as follows: Collect ambient temperature, relative humidity, dust concentration and air pressure parameters in the workshop. Collect no less than 10 sets of data continuously, remove abnormal data, and select effective data with ambient temperature of 15℃~40℃, relative humidity of 38%~75%, dust concentration ≤50mg / m³, and air pressure of 80kPa~105kPa to form an effective environmental monitoring dataset. Divide the work into two independent work units, with the unit spacing controlled at 3m to 5m; Dual-station timing job layout parameters are calibrated using a timing adaptation coupling calculation formula; dual-station timing adaptation degree is calculated. Adjust the timing operation scheduling parameters based on the calculation results to ensure... ; Survey the layout of workstations and equipment distribution, and delineate directional transfer paths for workpieces between workstations with a width of 1.5m to 2.0m to avoid blind spots in equipment operation; Monitor the fluctuations of environmental parameters in the two sets of work units in real time at a frequency of no less than once every 5 minutes to ensure that the timing parameters are compatible. The timing adaptation coupling calculation formula is as follows: ； In the formula: For dual-station timing adaptation; This is the correction factor for stable operating conditions; The average process time for a single batch at the shell-making station; The average process time per batch at the pouring and cooling station; The difference in ambient temperature between the two work units; The relative humidity difference between the two sets of work units; This refers to the dynamic coupling factor of environmental parameters. This is the timing lag correction factor; This is a correction factor for the thermal field superposition at the workstation.

3. The method for manufacturing large-size shell castings based on a dual-station shell profile according to claim 1, characterized in that, In step S2, the specific steps for shell demolding, deformation detection, and static control are as follows: The large-sized shell molded by solidification is demolded using a parallel lifting method, with the lifting speed controlled at 0.1m / s to 0.2m / s, and the lifting points are symmetrically distributed at the edge of the shell. Test points are set up in stress concentration areas such as the flange connection of the shell and the area around the oil passage, with a density of 4 to 8 points per square meter. A dial indicator is used to test the linear deformation values ​​in the length, width and height directions of the shell, with the test accuracy controlled at 0.01 mm. The test data are recorded to form a deformation parameter dataset. According to the production rhythm, the shell-shaped areas are neatly arranged and placed in stillness. The ambient temperature of the stillness area is controlled at 18℃~35℃ and the relative humidity is controlled at 40%~70%. The stillness area is far away from heat sources and areas with airflow disturbance. During the static setting process, the shell deformation is detected every 10 minutes to generate a real-time deformation monitoring dataset. The shell-shaped directional shrinkage amount is calculated using a multi-factor linkage correction formula, as follows: ； In the formula: This refers to the unidirectional linear shrinkage of the shell-shaped structure. For different structural directions, the linear contraction coefficient is denoted as . This refers to the shell-shaped settling time; This represents the percentage of resin adhesion on the coated sand. The coefficient representing the influence of the resin's room temperature crosslinking reaction. For shell-type surface curing degree; This is the curing degree coupling correction coefficient; It is the resin crosslinking degree attenuation factor; This is a correction factor for ambient temperature linkage. Real-time ambient temperature of the static area; The reference temperature for mold storage; These are dynamic parameters of the resin crosslinking reaction; This is the nonlinear shrinkage correction coefficient.

4. The method for manufacturing large-size shell castings based on a dual-station shell profile according to claim 1, characterized in that, In step S3, the specific steps for assembling and fixing the shell are as follows: Laser positioning equipment is used to perform benchmark alignment pre-calibration on the split shell type. The laser positioning accuracy is controlled to 0.005mm, and the calibration deviation is controlled within 0.05mm. Combined with the calibration ambient temperature, a laser calibration deviation compensation factor is used. Correcting calibration deviations, The value range is 0.97 to 1.03; A hydraulic pressure equalization assembly device is used to synchronously pre-tighten the shell-shaped splicing end faces, with the pre-tightening force increasing in increments of 25%. After pre-tightening, perform uniform pressure locking and use a feeler gauge to check the gap between the shell-shaped splicing end faces to ensure that the gap value is not greater than 0.08mm; A coordinate measuring machine is used to detect the shape and position parameters of the assembled shell, and the shape and position deviation is controlled within 0.3mm to complete the overall rigid fixation of the shell.

5. The method for manufacturing large-size shell castings based on a dual-station shell profile according to claim 1, characterized in that, In step S4, the specific steps for monitoring the shell seal and the airtight state are as follows: A process of alternating refractory mortar and high-temperature sealing tape is used to seal the gaps and openings of the shell joints in layers. When the gap width is 0.2mm to 3.5mm, 3 to 4 layers are used, and when the gap width is <0.2mm, 2 layers are used. The thickness of each layer is controlled to be 0.5mm to 1.0mm, and the edge of the covering layer extends at least 5mm beyond the edge of the gap. Distributed sensor points are arranged in the inner cavity of the shell at a density of 3 to 5 per cubic meter. The sensor points avoid the dead corners of the inner cavity and the key forming areas of the casting. Miniature sensors are used to collect the air pressure, temperature and humidity parameters of the inner cavity in real time at a frequency of 1 time / 2 minutes to form an inner cavity environment monitoring dataset. The probability of maintaining a sealed steady state within the shell cavity is calculated using a Bayesian decision formula, as follows: ； In the formula: The probability of maintaining a sealed steady state within the shell cavity is controlled by a threshold value ≥ 0.

95. The conditional probability of a perfectly sealed interior with a stable internal pressure differential; The prior probability of the sealed structure working properly; The conditional probability of a leaking seal failure but a randomly stable internal pressure differential. This represents the prior probability of the sealed structure failing. The internal air pressure fluctuation coefficient; The aging correction factor for the sealing structure is 0.90-0.98, determined according to the standard for high-temperature aging characteristics of refractory sealing materials. This is the aging rate factor for sealing tape, with a value of 0.92-0.99, determined according to the industry standard for aging rate of high-temperature resistant sealing tape.

6. The method for manufacturing large-size shell castings based on a dual-station shell profile according to claim 1, characterized in that, In step S5, the specific steps for controlling the pouring of molten metal are as follows: The probability of maintaining a sealed steady state within the housing cavity was confirmed by collecting parameters from distributed sensing points inside the housing. ≥0.95; Call the dual-station time-series iterative operation model, input the effective environmental monitoring dataset and dual-station operation parameters, and generate the initial staggered peak avoidance arrangement rhythm; Measure the temperature of the molten metal, controlling it between 1500℃ and 1650℃, and calculate the viscosity correction factor of the molten metal based on the temperature value. , The value range is 0.86 to 0.97; The pouring flow rate is set according to the wall thickness of the casting. When the wall thickness is 4mm to 6mm, the flow rate is controlled at 0.6m / s to 0.8m / s; when the wall thickness is 7mm to 12mm, the flow rate is controlled at 0.4m / s to 0.6m / s; and when the wall thickness is 13mm to 18mm, the flow rate is controlled at 0.3m / s to 0.4m / s. A layered, flow-limited, and progressive pouring operation is implemented using a pouring ladle. A flow controller is used to control the pouring flow rate in real time with a response time of no more than 0.5 steps (S). At the same time, a pressure sensor is used to monitor the internal pressure of the cavity in real time and control the pressure between 0.1 MPa and 0.3 MPa.

7. The method for manufacturing large-size shell castings based on a dual-station shell profile according to claim 1, characterized in that, In step S6, the specific steps for controlling the solidification of castings are as follows: After the molten metal is filled into the mold, an ultrasonic testing device is used to test the wall thickness parameters of the casting at a speed of 0.5 m / s. The testing accuracy is controlled to be 0.02 mm. Based on the wall thickness parameters, the casting is divided into three solidification control areas: thin-walled area, medium-thickness area, and thick-walled area. The overall thermal resistance at the interface between the shell and the casting is calculated using a dynamic interface thermal resistance calculation formula. ,according to The values ​​represent gradient heat dissipation parameters set for three solidification control zones: the heat dissipation rate is controlled at 0.8℃ / min~1.2℃ / min in the thin-walled zone, 0.5℃ / min~0.8℃ / min in the medium-thick zone, and 0.3℃ / min~0.5℃ / min in the thick-walled zone, thus completing the solidification control of the casting; the formula for calculating the dynamic thermal resistance of the interface is as follows: ； In the formula: The overall thermal resistance at the interface between the shell and the casting; Thermal resistance at the contact point between the mold and the metal base; The interfacial air film interference coefficient; The interfacial air film gap thickness; This is the solidification time coupling factor; This is the temperature gradient correction factor; This refers to the temperature difference between the inner and outer surfaces of the casting. This refers to the solidification time of the casting; This is a correction factor for the distribution of thermal points in castings.

8. The method for manufacturing large-size shell castings based on a dual-station shell profile according to claim 1, characterized in that, In step S7, the specific steps of casting post-processing and dual-station cyclic operation are as follows: Thermocouple monitoring data confirmed that the casting had completely solidified and solidified, meaning that the surface temperature of the casting had dropped to room temperature ±5℃. The shell-shaped structure is disassembled step by step, proceeding from non-critical parts to critical parts. After disassembly, special finishing tools are used to finely finish the flash and residual sand on the casting. The finishing accuracy of the flash is controlled at 0.01mm to ensure that the residual sand is cleaned up and there is no residue. After finishing, the surface roughness of the casting Ra≤1.6μm. The dimensions and surface quality parameters of the finished castings were inspected using 3D inspection equipment, and the surface roughness data of the castings were collected using a roughness meter to calculate the surface roughness influencing factor. , The value range is 0.96 to 1.04; Include dimensional parameters, surface quality parameters, and Substitute the data into the multidimensional detection model to calculate the overall dimensional deviation of the casting, ensuring that the deviation is ≤ ±0.1mm; Set the process division and rotation interval between the two sets of work units to 40min to 90min, and rotate and replace the process division of the two sets of work units; Collect production process parameters and quality inspection data for the first n-1 batches to form a batch dataset, set batch deviation weighting factors, calculate the quality fluctuation coefficient for the first n-1 batches, and calculate the batch stability correction factor based on the quality fluctuation coefficient. , The value range is 0.97 to 1.03; Batch deviation weighting factor, Substitute multiple batches of data into the iterative optimization model, obtain cyclic calibration parameters, call the cyclic calibration parameters to adjust the workstation operation parameters, and continuously carry out production line flow operations.

9. The method for manufacturing large-size shell castings based on a dual-station shell profile according to claim 8, characterized in that, In step S5, the calling steps for the dual-station time-series iterative job model are as follows: Extract the dual-station timing fit obtained above And a valid environmental monitoring dataset, combined with the molten metal temperature and molten metal viscosity correction factor detected in step S5. Input a dual-station time-series iterative operation model; Model by ≥ The control requirement of 0.85 is to dynamically adjust the staggered and avoidance arrangement rhythm and correct the pouring start lag time, with the lag time controlled to be 1min to 3min; Simultaneously, the pouring flow rate corresponding to the casting wall thickness is combined with the dual-station transfer connection time to be calibrated. The transfer connection time is controlled to be 2 min to 5 min to ensure that the pouring operation and the dual-station sequence are accurately matched and to avoid process interruption.

10. A large-size shell casting manufacturing system based on a dual-station shell molding line, wherein the system stores a computer program, characterized in that: When the computer program is executed by a processor, it implements the method as described in any one of claims 1-9.

Images