A large thick-wall head hot forming formability collaborative intelligent control method and system

Abstract

The application discloses a large thick-wall head hot forming shape property synergistic intelligent control method and system, and belongs to the technical field of head processing. The method obtains the forming load and the punch displacement in real time in the deep drawing forming process, obtains the contact area of the punch and the blank based on a die geometric model, and calculates the real-time contact stress; the resistance is compared with a preset safety range, if it is within the safety range, the parameters remain unchanged, if it exceeds the upper limit, the feeding speed is reduced, and if it is below the lower limit, the feeding speed is increased; finally, the control instruction is executed to adjust the punch feeding speed in real time. At the same time, the contact area is corrected based on infrared temperature field boundary identification, and a cumulative work control gain adjustment coefficient is introduced to realize fine grading control. The application solves the problem that the traditional process only adopts a fixed processing strategy, cannot adapt to the interference generated in the forming process, and seriously affects the forming quality and precision by realizing real-time sensing of the deformation resistance and dynamically matching the control strategy.

Description

Technical Field

[0001] This invention relates to the field of head processing technology, specifically to a collaborative intelligent control method and system for the thermoforming properties of large thick-walled heads. Background Technology

[0002] Large, thick-walled heads are critical pressure-bearing components in major equipment such as nuclear power plants, petrochemical plants, and high-pressure vessels. Their forming quality directly affects the operational safety and service life of the entire equipment. In the field of head manufacturing, deep drawing is the core process for achieving plastic deformation of large-sized, thick-walled metal sheets. This process is typically carried out on large hydraulic equipment. The downward pressure of the punch causes the sheet material to undergo radial stretching and tangential compression under the constraint of the die and blank holder, ultimately obtaining a hemispherical, elliptical, or dish-shaped structure with a predetermined curvature. For large, thick-walled heads with a thickness exceeding 50 mm and a diameter of several meters, the forming process is a typical three-dimensional unsteady plastic deformation, involving multiple nonlinear couplings in geometry, materials, and contact, resulting in extremely complex stress-strain distribution within the material.

[0003] In traditional hot drawing processes, to reduce the material's deformation resistance and improve its plasticity, the billet is typically heated to the austenitizing temperature range before processing. Existing techniques mainly rely on finite element numerical simulations to pre-set process parameters, such as initial heating temperature, die preheating temperature, and preset feed pressure curves. In actual production, workers or control systems often set a constant punch feed rate based on experience and operate the system with reference to the simulated process window.

[0004] Existing forming control methods exhibit significant lag and inconsistency when dealing with large, thick-walled parts. The main technical problem lies in the unavoidable uneven temperature distribution and uncontrolled cooling rates that occur during the heating, transport, and forming processes of thick-walled blanks, leading to deviations between the actual rheological stress within the material and the preset parameter models. Most existing control systems can only execute preset programs and cannot dynamically correct process parameters based on real-time feedback during the forming process. Especially when dealing with fluctuations in local rheological resistance caused by non-uniform temperature fields, fixed punch feed speeds and pressures are insufficient to adapt to changes in the material's resistance properties in real time.

[0005] When the material is under high resistance and the system maintains its original high-speed feed, the load will increase abnormally, easily causing stress concentration in the fillet transition zone, which can then induce shear fracture or excessive local thinning. Conversely, if the material resistance is too low and the feed pressure is insufficient, the billet is prone to instability and wrinkling in the sidewall or flange area due to stiffness mismatch. For thick-walled parts, once internal wrinkles occur, due to their enormous stiffness, subsequent processes are almost impossible to iron them out, directly leading to the scrapping of the entire expensive material. Based on this, it is necessary to study a collaborative intelligent control method and system for the hot forming properties of large thick-walled heads. Summary of the Invention

[0006] Therefore, the purpose of this invention is to provide a collaborative intelligent control method and system for the thermoforming properties of large thick-walled heads, which effectively solves the problems of existing head processing systems lacking real-time feedback during the forming process, only mechanically executing preset parameters, causing cracking or excessive thinning when the resistance is too high, and inducing unstable wrinkling that is difficult to repair when the resistance is insufficient, which seriously affects the forming quality and accuracy.

[0007] To achieve the above objectives, the technical solution adopted by this invention is: a collaborative intelligent control method for the thermoforming properties of large thick-walled heads, comprising the following steps:

[0008] Step 1: During the deep drawing process, the forming load and punch displacement of the blank are acquired in real time.

[0009] Step 2: Based on the punch displacement, obtain the contact area between the punch and the blank;

[0010] Step 3: Determine the real-time contact stress based on the forming load and contact area;

[0011] Step 4: Compare the real-time contact stress with the preset safety range:

[0012] If it is within a safe range, keep the current parameters unchanged;

[0013] If the upper limit of the safety range is exceeded, it is determined to be a high resistance state, and the feed speed of the punch is reduced;

[0014] If the force is below the lower limit of the safety range, it is determined to be a low resistance state, and the feed speed of the punch is increased;

[0015] Step 5: Execute the control command to adjust the feed speed of the punch in real time.

[0016] Furthermore, the initial state information of the blank is obtained before forming, including material composition, original thickness and initial average temperature; based on the initial state information, a suitable standard feed pressure curve and interface friction conditions are matched.

[0017] Furthermore, during the forming process, infrared sensing equipment is used to collect temperature field data on the upper surface of the billet, identify characteristic abrupt changes in the temperature gradient in the radial direction to locate the actual edge of the billet, and determine the boundary radial radius; the contact area correction factor is determined by the degree of deviation between the boundary radial radius and the theoretical standard radius.

[0018] Furthermore, when the radial radius of the boundary is smaller than the preset radius, the contact area is reduced; conversely, the contact area is increased.

[0019] Furthermore, multiple sampling profiles are selected along the circumference of the billet, and corresponding feature abrupt change points are identified on each profile to obtain multiple sets of original radial data; based on the preset geometric features of the end cap, the multiple sets of original radial data are averaged or weighted to obtain the final boundary radial radius.

[0020] Furthermore, based on historical real-time contact stress and contact area data, the cumulative work done by the punch during the forming process is calculated through integral calculation; the first amplitude deviation between this value and the preset energy upper limit, and the second amplitude deviation between this value and the preset energy lower limit are obtained; and the control gain adjustment coefficient used to quantify the degree of control is determined according to the ratio of the first amplitude deviation to the second amplitude deviation.

[0021] Furthermore, a preset adjustment time window is defined, where increasing the punch feed speed is positively correlated with the control gain adjustment coefficient, and decreasing the punch feed speed is negatively correlated with the control gain adjustment coefficient.

[0022] Furthermore, the forming load is obtained by converting the signal from the hydraulic system cylinder pressure transmitter or by measuring it through a force sensor; the punch displacement is monitored by a displacement sensor or a laser rangefinder; the load and displacement signals are synchronously triggered for sampling by the controller.

[0023] Furthermore, using a preset mold geometry model, the contact area between the punch and the blank is obtained based on the current punch displacement value.

[0024] A system for achieving the above-mentioned collaborative intelligent control of the hot forming properties of large thick-walled heads includes:

[0025] Sensing unit: Real-time acquisition of forming load and punch displacement during the hot pressing process of the billet;

[0026] Central processing module: It has a built-in calculation unit and logic judgment unit, which are used to calculate the real-time contact area and real-time contact stress, and output control commands after comparing the resistance with the safe range.

[0027] Execution control unit: connected to the output of the central processing module, used to receive control commands and drive the proportional servo valve group to adjust the feed flow of the hydraulic press main cylinder in real time.

[0028] The beneficial effects of the above technical solution are as follows: Existing technologies struggle to handle the complex working conditions in the hot stamping of large, thick-walled heads, given the challenges posed by fluctuations in billet temperature, changes in lubrication conditions, and differences in material properties. This leads to uneven head wall thickness, making wrinkling and cracking more likely. This invention solves the problems of sensing lag and control instability by real-time sensing of forming load and punch displacement, dynamically calibrating the safe range of real-time contact stress, and adjusting the punch feed speed based on the resistance offset. This significantly improves the uniformity of head wall thickness.

[0029] This invention acquires forming load and punch displacement signals by installing a pressure transmitter and displacement sensor on a hydraulic press. The forming load is divided by the real-time contact area to obtain the real-time contact stress, which transforms the vague material flow resistance into a quantifiable pressure scale, breaking the blindness of experience-based judgment.

[0030] When the real-time contact stress is within a safe range, the system maintains stable operation with current parameters. When it exceeds the upper or lower limits, the corresponding control direction is output based on the high or low resistance state. At the same time, the contact area is corrected based on the boundary identification of the infrared temperature field: when the measured edge radius of the billet deviates from the theoretical value, the contact area is calibrated in real time according to the deviation rate square factor to avoid misjudgment of resistance due to area calculation errors, and further improve the control accuracy.

[0031] Based on the offset of real-time contact stress and the safety range, as well as their positive and negative polarities, the change in material deformation difficulty is quantified as the basis for control. An adjustment coefficient for the cumulative work gain is introduced, and the cumulative plastic deformation energy is obtained by integrating the entire forming process of the punch from the initial position to the current position. This cumulative energy is compared with preset upper and lower limits, and the feed rate is adjusted according to the percentage of energy already completed, achieving an upgrade from qualitative control to quantitative follow-up matching.

[0032] In summary, this invention pre-sets working parameters based on the properties of the blank, senses the forming load and punch displacement in real time, and obtains the contact stress safety range in real time during the hot forming process. Based on the resistance offset and the cumulative work control gain adjustment coefficient, it adaptively adjusts the feed speed of the electro-hydraulic proportional valve. Combined with contact area correction and energy upper and lower limit protection, it significantly improves the shape and property coordination control capability in the hot forming process of large thick-walled heads. While ensuring equipment safety, it effectively improves the dimensional accuracy and microstructure consistency of the heads. Attached Figure Description

[0033] Figure 1 This is a flowchart of the intelligent control method of the present invention;

[0034] Figure 2 This is a flowchart of the temperature field boundary correction process of the present invention;

[0035] Figure 3 This is a flowchart of the energy gain coefficient control process of the present invention;

[0036] Figure 4 This is a block diagram of the overall system structure. Detailed Implementation

[0037] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments:

[0038] Example 1: This example aims to provide a collaborative intelligent control method and system for the hot forming properties of large thick-walled heads, mainly applied to the hot stamping forming of large thick-walled heads in critical pressure-bearing equipment such as nuclear island vessels, petrochemical reactors, and power plant boilers. Addressing the problem that existing head processing systems lack real-time feedback during the hot forming process of thick-walled heads, mechanically executing preset parameters, leading to cracking or excessive thinning under excessively high resistance, and inducing irreparable instability and wrinkling under insufficient resistance, severely affecting forming quality and accuracy, this example provides a system and method capable of online calculation of real-time contact stress based on real-time forming load and punch displacement, and performing graded intelligent compensation control based on a safety range comparison.

[0039] like Figure 4 As shown, a collaborative intelligent control system for the thermoforming properties of a large thick-walled head includes a sensing unit, a central processing module, and an execution control unit.

[0040] The sensing unit collects forming load and punch displacement signals in real time through pressure transmitters and displacement sensors installed in the main cylinder of the hydraulic press. After the signals are synchronously triggered and sampled by hard wiring, they are converted into digital quantities and transmitted to the central processing module.

[0041] The central processing module has a built-in calculation unit and a logic judgment unit. On the one hand, it calculates the contact area between the punch and the blank in real time based on the punch displacement and the die geometry model, and then calculates the real-time contact stress. On the other hand, it dynamically compares the real-time resistance value with the preset safety range and outputs control commands based on the comparison results.

[0042] After receiving the command, the control unit drives the proportional servo valve group to adjust the feed flow of the main cylinder of the hydraulic press in real time, thereby changing the punch feed speed and realizing the coordinated control of the shape and microstructure properties of the thick-walled head thermoforming process.

[0043] In this embodiment, the load signal acquisition of the sensing unit is implemented in two ways. One is to read the oil pressure value through the pressure transmitter of the main cylinder of the hydraulic system and convert it into a forming load by combining it with the effective working area of ​​the cylinder piston; the other is to directly install a spoke-type force sensor between the top of the punch and the piston rod of the cylinder to acquire the load data.

[0044] Displacement signals are monitored by a magnetostrictive displacement sensor or laser rangefinder installed between the hydraulic press frame and the movable crossbeam. To ensure the timing consistency between the load and displacement signals, the controller sends a synchronization pulse in each control cycle, triggering the pressure transmitter and displacement sensor to simultaneously latch the current value, eliminating the phase difference caused by signal transmission delay.

[0045] The acquired raw load signal contains hydraulic pump pulsation, proportional valve spool jitter, and mechanical vibration noise. A first-order low-pass filter is embedded in the front end of the central processing module, and the cutoff frequency is set to 1 / 5 to 1 / 10 of the natural frequency of the hydraulic system. The load curve after filtering is smooth and retains the true process fluctuation characteristics.

[0046] like Figure 1 As shown, the large thick-walled head thermoforming collaborative intelligent control method based on the above system provided in this embodiment includes the following steps:

[0047] Step 1: During the deep drawing process, the forming load and punch displacement of the billet during hot pressing are acquired in real time. Initial state information of the billet is acquired before forming, including material composition, original thickness, and initial average temperature. Based on this initial state information, a suitable standard feed pressure curve and interface friction conditions are matched.

[0048] Based on this initial state information, the central processing module matches the standard feed pressure curve in the process knowledge base, that is, extracts the load displacement envelope under ideal conditions according to the rheological properties and thickness-to-diameter ratio of the material, as the benchmark for subsequent safety judgment; by obtaining the interface friction conditions, the module calibrates the appropriate friction material according to the current high temperature environment and material properties, and arranges it on the surface of the billet.

[0049] Taking a large, thick-walled head as an example, the initial thickness of the blank is 180 mm. After being heated to the austenitic region temperature of 1200±10℃, it is transferred to the forming station. During the downward movement of the punch, the displacement sensor continuously records the punch position with a sampling period of 10 ms, while the pressure transmitter synchronously records the corresponding forming load. When the punch contacts the surface of the blank, the load increases from zero. The system identifies this feature point as the starting reference point for forming and sets the punch displacement value at that moment to zero. All subsequent displacements are accumulated using this zero point.

[0050] Step two: Based on the punch displacement, obtain the contact area between the punch and the blank. For head forming using a spherical punch, the preset die geometry model has known parameters, and the spherical radius R of the punch is a constant. When the punch's pressing depth is h, the contact surface between the punch and the blank is a spherical cap, and its contact radius and contact area can be calculated using the geometric model. The central processing module has this geometric analytical model built in and directly calculates the contact area based on the real-time punch displacement h. For head forming using an ellipsoidal or complex curved surface punch, the die geometry model is stored in the flash memory of the central processing module in the form of a discrete point cloud, and the corresponding contact area is found by interpolation based on the current punch displacement.

[0051] Step 3: Determine the real-time contact stress based on the forming load and contact area. Real-time contact stress is defined as the average pressure exerted by the punch on the workpiece, i.e., the load borne per unit contact area. The calculation formula is:

[0052]

[0053] In the formula, For real-time contact stress, For real-time forming load, The contact area is represented by this value. This real-time contact stress, in a physical sense, characterizes the billet's overall ability to resist plastic deformation under current temperature, strain rate, and friction conditions. Its numerical variation directly reflects the material flow behavior during the hot forming process of thick-walled heads.

[0054] Step four: Compare the real-time contact stress with the preset safety range. The lower and upper limits of the safety range are respectively... and The process is pre-set based on the initial state information of the blank. Before forming, the material composition, original thickness and initial average temperature of the blank are obtained. The process knowledge base built into the system flash memory is matched to obtain the reference curve and safety range of standard real-time contact stress as a function of punch displacement. It is usually set as ±15% of the reference curve value as the upper and lower limits.

[0055] If real-time contact stress Within a safe range, that is The central processing module determines that the current forming state is normal, maintains the current punch feed speed parameter unchanged, and the system enters the next sampling cycle of cyclic monitoring.

[0056] like The system determines this to be a high-resistance state. This state typically corresponds to excessively low billet temperature, deteriorated lubrication conditions, or inherent defects within the material. The central processing module outputs a control command to reduce the punch feed speed.

[0057] like The system determines the condition to be a low-resistance state. This state may correspond to excessively high billet temperature, localized overheating, or abnormally increased material fluidity. The system outputs a control command to increase the punch feed speed.

[0058] During implementation, if the pressure is below the lower limit of the safety range, it is determined to be a low resistance state, and the real-time feedback pressure of the pressure ring is further obtained: if the pressure fluctuation does not exceed the preset threshold, the punch feed speed is appropriately increased; if the pressure fluctuation exceeds the preset threshold, the increase in feed speed is limited and only a slight increase is made.

[0059] Step 5: Execute the control command to adjust the punch feed speed in real time. After receiving the target speed and feed pressure values ​​from the central processing module, the control unit linearly maps the speed command to the PWM drive duty cycle of the proportional servo flow valve through its built-in signal conversion circuit. During this process, based on the compression characteristics of the hydraulic oil and the cylinder volume effect, the dynamic response adjustment of the proportional servo valve allows the punch feed speed to smoothly transition from the current value to the target value, thereby effectively mitigating the impact load on the hydraulic system caused by sudden changes in control parameters.

[0060] This embodiment senses the forming load and punch displacement in real time, calculates the real-time contact stress online and compares it dynamically with the safe range, and performs intelligent graded control based on the comparison results. This ensures that the punch feed speed matches the performance of the blank, effectively solving the problems of delayed resistance sensing and blind control in traditional processes. While ensuring equipment safety, it significantly improves the forming quality and process stability of large thick-walled heads.

[0061] Example 2

[0062] Based on Example 1, this embodiment further provides a contact area correction method based on infrared temperature field boundary recognition, aiming to eliminate the influence of factors such as blank positioning deviation, local softening differences caused by uneven temperature field, and mold wear on the theoretical calculation value of contact area, and improve the calculation accuracy of real-time contact stress.

[0063] In step two of Example 1, the theoretical calculation of the contact area is based on the mold geometry model and the punch displacement. However, during the actual forming process, the edge of the blank may shift due to thermal expansion, positioning eccentricity, or local uneven deformation, resulting in a deviation between the actual contact area and the theoretical value. To correct this deviation, this example introduces the following correction mechanism during the forming process: During the forming process, infrared sensing equipment is used to collect temperature field data on the upper surface of the blank, and characteristic abrupt changes in the temperature gradient in the radial direction are identified to locate the actual edge of the blank and determine the boundary radial radius; the contact area correction factor is determined by the degree of deviation between the boundary radial radius and the theoretical standard radius.

[0064] like Figure 2 As shown, in practical implementation, eight infrared thermal imagers installed around the press frame collect temperature field data of the upper surface of the billet. Due to the significant temperature difference between the edge region of the billet and the ambient background, the temperature gradient exhibits a steep drop from the high-temperature region to the low-temperature region in the radial direction. By performing radial temperature gradient calculation on each frame of the temperature field image, the pixel position with the largest absolute gradient value is identified as the feature abrupt change point, which is the actual edge of the billet.

[0065] Multiple sampling profiles are selected along the circumference of the billet, and the corresponding feature abrupt change points on each profile are identified to obtain multiple sets of raw radial data. In this embodiment, a profile is selected at every 45 degrees, and the coordinates of the feature abrupt change points on a total of eight profiles are obtained, thereby determining the boundary radial radius corresponding to each profile. .

[0066] Based on the pre-defined geometric features of the end cap (typically circular or elliptical), the multiple sets of original radial data are averaged or weighted to obtain the final boundary radial radius. For circular end caps, an arithmetic mean is used for calculation. For elliptical end caps, the cross-sectional data along the major and minor axes are weighted separately.

[0067] The measured average radius The theoretical standard radius corresponding to the current punch displacement in the mold geometry model Compare and define the deviation rate :

[0068]

[0069] The system presets the startup threshold. In this embodiment, =±3%. When At that time, contact area correction is initiated.

[0070] Correction factor The deviation rate is determined based on its sign and magnitude. For a circular head, the contact area is proportional to the square of the radius. In this embodiment, the following calculation method is used to quantify the correction factor. ,

[0071]

[0072] When the boundary radial radius is smaller than the preset radius, the contact area is reduced; conversely, the contact area is increased. Specifically, when the boundary radial radius is smaller than the theoretical radius, This indicates that the edges of the billet shrink inward, and the actual contact area is smaller than the theoretically calculated value. The contact area is reduced according to this factor.

[0073]

[0074] Conversely, when the radial radius of the boundary is greater than the theoretical radius, This indicates that the billet has expanded beyond expectations, increasing the contact area.

[0075] The corrected contact area replaces the original theoretical value in the real-time contact stress calculation, that is, the calculation formula in step three is updated.

[0076] This embodiment eliminates systematic errors caused by uneven temperature field, positioning deviation, or mold wear, ensuring the accuracy of real-time contact stress calculation and avoiding potential equipment overload in the future.

[0077] Example 3

[0078] Based on Example 1, this embodiment further provides a control gain adjustment coefficient based on the cumulative work done, which adjusts the control amplitude by controlling the gain adjustment coefficient. This aims to solve the problem of overshoot or undershoot caused by different resistance deviations during the forming process of large thick-walled heads, and to achieve compensation of process parameters.

[0079] like Figure 3 As shown, this embodiment calculates the cumulative work done by the punch during the forming process based on historical real-time contact stress and contact area data through integral calculation. During the hot drawing process of large, thick-walled heads, the integral calculation unit built into the central processing module continuously retrieves historical real-time contact stress and contact area data. Based on the integral formula, the system performs continuous integral calculations throughout the entire forming process, starting from the initial position where the punch contacts the blank, to obtain the cumulative work done by the punch in real time. This value physically represents the total work done by plastic deformation currently input into the blank, used to quantify the forming state.

[0080] For end caps with specific materials, thicknesses, and initial temperatures, the system presets upper and lower limits for work done that match their formability. When quantifying the current progress using the upper and lower limits, the allowable work error is determined. When the work is close to the upper limit, it indicates that the accumulated work is too much and the overall feed is too much. When it is close to the lower limit, it indicates that the accumulated work is too little and the overall feed is too little.

[0081] The system obtains the first deviation of the current cumulative work done from the preset energy upper limit and the second deviation from the preset energy lower limit. Based on the ratio of the first deviation to the second deviation, it determines the control gain adjustment coefficient used to quantify the degree of control. When executing the control logic in step four, the central processing module first obtains the first deviation of the current cumulative work done from the energy upper limit and the second deviation from the energy lower limit. Based on the ratio of the first deviation to the second deviation, the system calculates the control gain adjustment coefficient used to quantify the degree of control.

[0082] A preset adjustment time window is defined. Increasing the punch feed speed is positively correlated with the control gain adjustment coefficient, while decreasing the punch feed speed is negatively correlated with the control gain adjustment coefficient. In specific adjustments, the system performs dynamic amplitude adjustment within the preset adjustment time window. When a high-resistance state is detected, due to the risk of material over-deformation, the system reduces the feed speed, which is negatively correlated with the control gain adjustment coefficient. That is, the closer the accumulated work is to the upper limit, the greater the deceleration, in order to control and reduce the feed input. Conversely, when a low-resistance state is detected, the magnitude of the parameter increase is positively correlated with the control gain adjustment coefficient. That is, the closer the energy reserve is to the lower limit, the greater the speed increase and pressure boost, in order to compensate for the feed.

[0083] This embodiment provides an amplitude adjustment mechanism based on energy accumulation state, which realizes precise quantification of the amplitude of the control command, enabling the working changes of the punch to be dynamically adjusted as the material deformation work accumulates. By correlating real-time resistance fluctuations with the overall forming energy state, process deviations can be smoothly compensated, and high-quality finished products with dense structure and uniform wall thickness distribution can be obtained while ensuring that the thick-walled head does not wrinkle or crack.

[0084] The embodiments of the present invention described above do not constitute a limitation on the scope of protection of the present invention. The basic concept of the present invention is to: acquire the forming load and punch displacement in real time during the hot forming process of thick-walled heads, calculate the real-time contact stress based on the contact area, compare it with a preset safety range, and dynamically adjust the punch feed speed according to the comparison result, thereby realizing real-time response and graded control of the resistance state. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the claims of the present invention.

Claims

1. A collaborative intelligent control method for the thermoforming properties of large thick-walled heads, characterized in that, Includes the following steps: Step 1: During the deep drawing process, the forming load and punch displacement of the blank are acquired in real time. Step 2: Based on the punch displacement, obtain the contact area between the punch and the blank; Step 3: Determine the real-time contact stress based on the forming load and contact area; Step 4: Compare the real-time contact stress with the preset safety range: If it is within a safe range, keep the current parameters unchanged; If the upper limit of the safety range is exceeded, it is determined to be a high resistance state, and the feed speed of the punch is reduced; If the force is below the lower limit of the safety range, it is determined to be a low resistance state, and the feed speed of the punch is increased; Step 5: Execute the control command to adjust the feed speed of the punch in real time; in: During the forming process, infrared sensing equipment is used to collect temperature field data on the upper surface of the billet, identify characteristic abrupt changes in the temperature gradient in the radial direction to locate the actual edge of the billet, and determine the boundary radial radius; the contact area correction factor is determined by the degree of deviation between the boundary radial radius and the theoretical standard radius. The measured average radius The theoretical standard radius corresponding to the current punch displacement in the mold geometry model Compare and define the deviation rate : The system presets the startup threshold. ,when At that time, contact area correction is initiated; Contact area correction factor The deviation rate is determined by its sign and magnitude; for a circular head, the contact area is proportional to the square of the radius. When the radial radius of the boundary is smaller than the preset radius, the contact area is reduced; otherwise, the contact area is increased.

2. The method for coordinated intelligent control of the thermoforming properties of large thick-walled heads according to claim 1, characterized in that, Before forming, the initial state information of the blank is obtained, including material composition, original thickness and initial average temperature; based on the initial state information, a suitable standard feed pressure curve and interface friction conditions are matched.

3. The method for collaborative intelligent control of the thermoforming properties of large thick-walled heads according to claim 1, characterized in that: Multiple sampling profiles are selected along the circumference of the billet, and corresponding feature abrupt change points are identified on each profile to obtain multiple sets of original radial data. Based on the preset geometric features of the end cap, the multiple sets of original radial data are averaged or weighted to obtain the final boundary radial radius.

4. The method for collaborative intelligent control of the thermoforming properties of large thick-walled heads according to claim 1, characterized in that: Based on historical real-time contact stress and contact area data, the cumulative work done by the punch during the forming process is calculated through integral calculation; the first amplitude deviation between this value and the preset energy upper limit, and the second amplitude deviation between this value and the preset energy lower limit are obtained; the control gain adjustment coefficient used to quantify the degree of control is determined according to the ratio of the first amplitude deviation to the second amplitude deviation; wherein, the preset energy refers to the upper and lower limits of work that the system presets to match the formability of the end cap part with a specific material, thickness and initial temperature.

5. The method for collaborative intelligent control of the thermoforming properties of large thick-walled heads according to claim 4, characterized in that: A preset adjustment time window is defined. Increasing the punch feed speed is positively correlated with the control gain adjustment coefficient, while decreasing the punch feed speed is negatively correlated with the control gain adjustment coefficient.

6. The method for collaborative intelligent control of the thermoforming properties of large thick-walled heads according to claim 1, characterized in that: The forming load is obtained by converting the signal from the hydraulic system cylinder pressure transmitter or by measuring it through a force sensor; the punch displacement is monitored by a displacement sensor or a laser rangefinder; the load and displacement signals are synchronously triggered for sampling by the controller.

7. The method for collaborative intelligent control of the thermoforming properties of large thick-walled heads according to claim 1, characterized in that: Using a preset mold geometry model, the contact area between the punch and the blank is obtained based on the current punch displacement value.

8. A collaborative intelligent control system for the thermoforming properties of large thick-walled heads implementing the method of any one of claims 1 to 7, characterized in that, include: Sensing unit: Real-time acquisition of forming load and punch displacement during the hot pressing process of the billet; Central processing module: It has a built-in calculation unit and logic judgment unit, which is used to calculate the real-time contact area and real-time contact stress, and output control commands after comparing the resistance with the safe range. Execution control unit: connected to the output of the central processing module, used to receive control commands and drive the proportional servo valve group to adjust the feed flow of the hydraulic press main cylinder in real time.

Images