A method and system for adjusting parameters in powder metallurgy injection molding of large components

Abstract

This application provides a method and system for adjusting parameters in large component powder metallurgy injection molding, applied in the field of powder metallurgy injection molding technology. It acquires acoustic vibration signals from the injection screw during material transport and mixing, performs frequency domain analysis, and extracts acoustic characteristic parameters reflecting material uniformity. Based on these parameters, it identifies the segregation region, type, and degree in the material flow. When the segregation degree reaches moderate or severe, it selects intervention methods for the local material state in the segregation region according to the segregation type, including introducing auxiliary fluid or applying local dynamic shear force. It monitors the acoustic vibration signals after intervention in real time, extracts acoustic characteristic parameters again, and adjusts the amount of auxiliary fluid introduced or the intensity of the dynamic shear force based on their changes. Therefore, this application has the advantage of effectively improving the uniformity and product quality of large component powder metallurgy injection molded parts.

Description

Technical Field

[0001] This application relates to the field of powder metallurgy injection molding technology, and in particular to a method and system for adjusting parameters in powder metallurgy injection molding of large components. Background Technology

[0002] In modern industrial production, powder metallurgy injection molding technology faces a core challenge in manufacturing high-performance large metal components: ensuring uniformity of material microstructure. One key issue is that the injection molding machine screw inevitably wears down during long-term operation, affecting its ability to mix materials. This leads to microscopic inhomogeneities within the material, ultimately severely impacting product quality and performance.

[0003] This microscopic inhomogeneity causes high-frequency, minute localized flow fluctuations in the mixture during injection. However, traditional injection molding machine control systems treat these signals as noise and filter them out, unable to compensate for them. For large components, the long material flow path and filling time lead to the continuous accumulation and amplification of microscopic differences, forming macroscopic density gradients and uneven stress distributions within the green body. During subsequent debinding and high-temperature sintering, the inconsistent shrinkage behavior of different density regions generates enormous internal stress, causing component deformation, warping, or even cracking, or resulting in uneven microstructure such as grain size and porosity distribution, ultimately leading to the production failure or substandard performance of the entire large component.

[0004] To address the aforementioned issues, existing technologies urgently need improvement. Summary of the Invention

[0005] In view of the shortcomings of the prior art, this application provides a method and system for adjusting parameters of powder metallurgy injection molding of large components. It can monitor the material segregation state in real time and make local intervention according to the type and degree of segregation, which has the advantages of effectively improving the uniformity and product quality of powder metallurgy injection molded parts of large components.

[0006] In a first aspect, a method for adjusting parameters in powder metallurgy injection molding of large components is provided, for adjusting the segregation state of materials during the injection molding process. The method includes the following steps: S1: Acquire acoustic vibration signals of the injection screw during material transport and mixing; S2: Perform frequency domain analysis on the acoustic vibration signal to extract acoustic characteristic parameters that reflect the uniformity of the material; S3: Based on the acoustic characteristic parameters, identify the segregation region, segregation type, and segregation degree in the material flow; S4: When the degree of segregation reaches moderate or severe segregation, a method of intervention on the local material state is selected in the segregation occurrence area according to the segregation type. The intervention method includes introducing auxiliary fluid into the material flow to adjust the local rheological properties and applying local dynamic shear force to promote powder dispersion. S5: Monitor the acoustic vibration signal after intervention in real time, and extract the acoustic feature parameters again. Based on the changes in the extracted acoustic feature parameters, adjust the amount of auxiliary fluid introduced or the intensity of the dynamic shear force.

[0007] Furthermore, step S1 includes: S11: The vibration signal generated by the operation of the injection screw is synchronously acquired by an array of piezoelectric accelerometers arranged along the outer wall of the barrel at a sampling frequency of not less than 100kHz. S12: Perform bandpass filtering preprocessing on the acquired raw signal to filter out power frequency interference and low-frequency mechanical vibration noise; S13: Perform a fast Fourier transform or wavelet transform on the preprocessed signal to extract the spectral features in the high-frequency band from 5kHz to 50kHz as the acoustic feature parameters.

[0008] Furthermore, step S2 includes: S21: Perform a short-time Fourier transform on the acoustic vibration signal to obtain the time-frequency domain energy distribution matrix; S22: Based on the time-frequency domain energy distribution matrix, calculate the energy spectral entropy, spectral peak frequency, and energy variation coefficient within at least one sensitive frequency band; S23: The energy spectral entropy is compared with the preset reference entropy value under uniform mixing state to obtain the segregation entropy difference; the acoustic characteristic parameters include the segregation entropy difference, the spectral peak frequency and the energy variation coefficient.

[0009] Furthermore, step S3 includes: S31: Divide the barrel into multiple spatial window areas along the screw axis, calculate the segregation entropy difference in each window area, and determine the window area where the segregation entropy difference exceeds the first segregation threshold as the segregation occurrence area. S32: Determine the segregation type based on the shift direction of the spectral peak frequency. The segregation type includes powder agglomeration or sedimentation type, and binder enrichment type. S33: Within the segregation region, the energy variation coefficient is compared with multiple preset degree thresholds, and the degree of segregation is classified as slight segregation, moderate segregation, or severe segregation based on the comparison results.

[0010] Furthermore, step S4 includes: S41: When the degree of segregation reaches moderate or severe segregation, and the peak frequency of the spectrum shifts down relative to the reference value, and the segregation type is determined to be powder agglomeration or sedimentation, local dynamic shear force is preferentially applied to the segregation area. S42: When the peak frequency shifts upward relative to the reference value, and the segregation type is determined to be binder enrichment type, auxiliary fluid is preferentially introduced into the material flow in the segregation region. S43: When the direction of the spectral peak frequency shift is unclear, local dynamic shear force and auxiliary fluid are alternately applied in the segregation region, and the switching is based on the rate of change of acoustic characteristic parameters monitored in real time.

[0011] Furthermore, in step S41, local dynamic shear force is preferentially applied to the segregation region, including but not limited to: S411: Activates an ultrasonic transducer array with an operating frequency of 20kHz to 40kHz to break up powder agglomerates by inducing cavitation and microjets in the material. Alternatively, a micro motor can be used to drive retractable micro-stirring blades into the material flow, providing localized mechanical shearing force.

[0012] Furthermore, in step S42, preferentially introducing auxiliary fluid into the material flow in the segregation region includes the following steps: S421: Adjust the injection volume of the auxiliary fluid injected by the micro-orifice nozzle, wherein the orifice diameter of the micro-orifice nozzle is 0.2 mm to 0.5 mm; The auxiliary fluid includes a low-viscosity polymer solution or a fatty acid ester dispersant, used to locally reduce the viscosity of the material and improve the wettability of the powder.

[0013] Furthermore, step S5 includes: S51: Real-time monitoring of acoustic vibration signals after intervention, extraction of acoustic characteristic parameters again, and calculation of the first rate of change of segregation entropy difference before and after intervention, as well as the second rate of change of energy variation coefficient; S52: When both the first rate of change and the second rate of change are negative, if the absolute value of the first rate of change is greater than or equal to a preset first threshold and the absolute value of the second rate of change is greater than or equal to a preset second threshold, the current intervention parameters remain unchanged; otherwise, the fuzzy PID control algorithm is used to dynamically adjust the amount of auxiliary fluid introduced or the intensity of dynamic shear force. S53: When both the first rate of change and the second rate of change are positive, if the first rate of change is less than a preset first threshold or the second rate of change is less than a preset second threshold, the fuzzy PID control algorithm is used to dynamically adjust the amount of auxiliary fluid introduced or the intensity of dynamic shear force; otherwise, the intervention is stopped and an alarm message is issued. S54: When one of the first rate of change and the second rate of change is positive and the other is negative, a fuzzy PID control algorithm is used to dynamically adjust the amount of auxiliary fluid introduced or the intensity of dynamic shear force, and the calculated adjustment amount is multiplied by 50% to limit the magnitude of the adjustment amount.

[0014] Furthermore, in steps S52 and S53, the dynamic adjustment of the amount of auxiliary fluid introduced or the intensity of dynamic shear force using a fuzzy PID control algorithm includes the following steps: S55: The difference e between the segregation entropy difference after intervention and the reference entropy value is used as the first input variable of the PID control algorithm, and the instantaneous change rate ec of the segregation entropy difference is used as the second input variable of the PID control algorithm. e and ec are fuzzified into linguistic variables through the membership function respectively. S56: Based on the linguistic variables and the preset fuzzy rule table, output the proportional coefficient increment ΔKp, integral coefficient increment ΔKi, and differential coefficient increment ΔKd; the fuzzy rule table is formulated based on the following logic: when |e|>0.20, increase Kp to speed up the response, while limiting Ki to prevent integral saturation; when 0.05<|e|≤0.20, appropriately decrease Kp and introduce Kd to suppress overshoot; when |e|≤0.05, increase Ki to eliminate steady-state segregation residue; S57: Add ΔKp, ΔKi, and ΔKd to the initial PID parameters Kp0, Ki0, and Kd0 to obtain the PID parameters Kp, Ki, and Kd at the current time. S58: According to the fuzzy PID control algorithm When u is positive, it increases the amount of auxiliary fluid introduced or the intensity of dynamic shear force; when u is negative, it decreases the amount of auxiliary fluid introduced or the intensity of dynamic shear force.

[0015] Secondly, a parameter adjustment system for powder metallurgy injection molding of large components, the system being used to implement the steps of any of the methods described above, the system comprising: Acquisition module: Acquires acoustic vibration signals of the injection screw during material transport and mixing; Extraction module: Performs frequency domain analysis on the acoustic vibration signal to extract acoustic feature parameters that reflect the uniformity of the material; Identification module: Based on the acoustic feature parameters, it identifies the segregation region, segregation type, and degree of segregation in the material flow; Intervention module: When the segregation degree reaches moderate or severe segregation, the module selects a method to intervene in the local material state in the segregation occurrence area according to the segregation type. The intervention method includes introducing auxiliary fluid into the material flow to adjust the local rheological properties and applying local dynamic shear force to promote powder dispersion. Adjustment module: Real-time monitoring of acoustic vibration signals after intervention, and extraction of acoustic characteristic parameters again. Based on the changes in the extracted acoustic characteristic parameters, the amount of auxiliary fluid introduced or the intensity of the dynamic shear force is adjusted.

[0016] Beneficial Effects: This application proposes a method and system for adjusting parameters in large component powder metallurgy injection molding. By acquiring the acoustic vibration signals of the injection screw during material transport and mixing, and performing frequency domain analysis, acoustic characteristic parameters reflecting material uniformity are extracted. Based on these parameters, the segregation region, type, and degree in the material flow are identified. When the segregation degree reaches moderate or severe, intervention methods are selected in the segregation region according to the segregation type, including introducing auxiliary fluid or applying local dynamic shear force. The acoustic vibration signals after intervention are monitored in real time, and acoustic characteristic parameters are extracted again. The amount of auxiliary fluid introduced or the intensity of the dynamic shear force are adjusted according to their changes. Therefore, this application can monitor the material segregation state in real time and perform local intervention according to the type and degree of segregation, effectively improving the uniformity and product quality of large component powder metallurgy injection molded parts. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating a method for adjusting parameters in powder metallurgy injection molding of large components, as proposed in this application.

[0018] Figure 2 This is a structural diagram of a powder metallurgy injection molding parameter adjustment system for large components proposed in this application.

[0019] Figure 3 This is a schematic diagram of a powder metallurgy injection molding parameter adjustment system for large components proposed in this application.

[0020] Labeling Explanation: 201, Acquisition Module; 202, Extraction Module; 203, Identification Module; 204, Intervention Module; 205, Adjustment Module. Detailed Implementation

[0021] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. The components of the embodiments of this application described and marked in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely represents selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0022] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0023] Please refer to Figure 1 This application proposes a method for adjusting parameters in powder metallurgy injection molding of large components, used to regulate the segregation state of materials during the injection molding process. The method includes the following steps: S1: Acquire acoustic vibration signals of the injection screw during material transport and mixing; S2: Perform frequency domain analysis on acoustic vibration signals to extract acoustic characteristic parameters that reflect the uniformity of materials; S3: Based on acoustic characteristic parameters, identify the segregation area, segregation type and segregation degree in the material flow; S4: When the degree of segregation reaches moderate or severe segregation, the method of intervention on the local material state in the segregation area is selected according to the segregation type. The intervention method includes introducing auxiliary fluid into the material flow to adjust the local rheological properties and applying local dynamic shear force to promote powder dispersion. S5: Monitor the acoustic vibration signal after intervention in real time, and extract the acoustic characteristic parameters again. Adjust the amount of auxiliary fluid introduced or the intensity of dynamic shear force based on the changes in the extracted acoustic characteristic parameters.

[0024] The working principle of this method is based on real-time sensing and immediate intervention of the material state inside the injection molding machine screw. When the injection molding machine screw wears down during long-term operation, its shearing and mixing capabilities for powder and binder mixtures gradually decrease.

[0025] This method involves deploying a highly sensitive array of acoustic sensors outside the screw barrel to continuously monitor the acoustic vibration signals generated during screw operation. These signals, especially the traditionally ignored high-frequency components, actually contain microscopic information about the friction, shearing, and mixing between the screw and the material, and between the material and the barrel. When micro-segregation occurs in the material, its flow characteristics undergo subtle changes, such as changes in local viscosity or resistance, which trigger specific high-frequency acoustic signals. By analyzing these signals in real time, the mixing uniformity of the material inside the screw can be non-invasively diagnosed, and the specific location and type of segregation can be identified.

[0026] Once a problem is diagnosed, intervention measures are taken immediately. In the screw section where segregation is detected, the microfluidic jetting unit integrated into the barrel wall precisely injects a small amount of auxiliary fluid, such as a low-viscosity binder or dispersant, to locally adjust the viscosity of the material or improve the wettability of the powder; or it uses a micro dynamic shearing auxiliary device, such as an ultrasonic transducer, to apply additional local shear force to the material in that area, helping to break up powder agglomerates and promote the uniform redispersibility of particles.

[0027] These interventions are all immediate corrections made at the source, before the material leaves the screw. This approach transforms screw wear, traditionally considered a pure defect, into a state that can be managed and compensated for in real time. It proactively maintains the uniformity of the material without affecting production continuity, avoiding the limitations of post-event remediation in traditional solutions.

[0028] Specifically, a piezoelectric accelerometer array arranged along the outer wall of the barrel synchronously acquires vibration signals generated by the injection screw at a sampling frequency of at least 100 kHz. To achieve real-time monitoring of the material state inside the screw, a high-sensitivity piezoelectric accelerometer, such as a Kistler 8763A500 sensor, can be installed approximately every 20 cm along the conveying and mixing sections of the screw outside the injection molding machine barrel. These sensors are fixed to the outer wall of the barrel by threads or magnetic attachment, ensuring good acoustic contact with the barrel. Each sensor is connected to a multi-channel high-speed data acquisition device, such as a National Instruments cDAQ-9178 chassis with an NI 9234 module. A sampling frequency of at least 100 kHz, preferably 200 kHz, is chosen because localized flow fluctuations caused by micro-region segregation of the material are extremely transient and high-frequency events; a high sampling rate is a prerequisite for capturing these fleeting microscopic physical phenomena.

[0029] The Kistler 8763A500 sensor is an early product of Kistler, and its existence can be confirmed at the 2003 China International Process Control and Production Automation Exhibition. However, this model was later updated, and the updated model still sold on the Kistler website is the 8763B series.

[0030] The examples cited in this application are earlier models. Those skilled in the art can understand the type, installation method and performance characteristics of the sensor through these examples, and implement the technical solution of this application.

[0031] The raw vibration signals collected inevitably contain noise from the injection molding machine itself, such as power frequency interference from the motor and low-frequency mechanical vibration noise from the movement of large mechanical structures. This noise can mask the useful signals that truly reflect the material's state. Therefore, bandpass filtering preprocessing is necessary to remove power frequency interference and low-frequency mechanical vibration noise from the raw signals. For example, a bandpass filter can be used to remove frequency components below 1kHz and above 80kHz, resulting in a pure signal with a high signal-to-noise ratio that accurately reflects the material's flow state.

[0032] The preprocessed signal is a time-domain signal. In order to reveal its inherent frequency composition, it is necessary to perform a fast Fourier transform on the preprocessed signal to extract the spectral features in the high-frequency band from 5 kHz to 50 kHz as acoustic feature parameters.

[0033] The Fast Fourier Transform (FFT) can convert time-domain vibration signals into frequency-domain spectrograms, visually displaying the energy distribution of the signal at different frequencies. Microscopic events within materials, such as friction, collision, and breakage of powder agglomerates, or viscous flow in binder-rich regions, produce unique energy responses in specific high-frequency bands, particularly between 5 kHz and 50 kHz. Therefore, extracting the spectral characteristics within this high-frequency band is akin to extracting the fingerprint of the material's state, providing crucial information for subsequent segregation identification.

[0034] Furthermore, to extract acoustic characteristic parameters that accurately reflect the uniformity of materials, the frequency domain analysis process of the acoustic vibration signal was further refined. First, a short-time Fourier transform (SFT) was performed on the acoustic vibration signal to obtain a time-frequency domain energy distribution matrix. Traditional fast Fourier transforms provide the overall frequency distribution of the signal over a period of time, failing to reflect the changes in frequency components over time. However, the flow of material within the screw is a dynamic process, and the occurrence and development of segregation are time-varying. The SFT, through piecewise windowing of the signal, can generate a two-dimensional time-frequency domain energy distribution matrix, demonstrating how the energy of each frequency component changes during the brief time the material flows past the sensor location, thus capturing the instantaneous characteristics of segregation.

[0035] Based on the obtained time-frequency domain energy distribution matrix, the energy spectral entropy, peak frequency, and energy variation coefficient within at least one sensitive frequency band are calculated. These three parameters characterize the homogeneity of the material from different dimensions. Energy spectral entropy is a concept in information theory used to measure the degree of disorder or randomness of a system or signal. In a homogeneous mixed material flow, the microscopic motion of countless powder particles is random and complex, and the resulting acoustic signal has a relatively broad and flat energy distribution in the spectrum, exhibiting a high energy spectral entropy.

[0036] Conversely, if large-scale powder agglomeration or binder enrichment occurs in the material, the material flow will become more orderly or exhibit regular impacts, resulting in the spectral energy being concentrated on a few frequencies, manifested as a lower energy spectral entropy.

[0037] Entropy of the energy spectrum is a physical quantity that measures the complexity of the frequency domain distribution of an acoustic signal. When the material is uniformly mixed, the frictional vibration energy generated by the powder and binder is relatively flat and random in the frequency spectrum, and the corresponding entropy value is at a high level.

[0038] The baseline entropy value is the average entropy value of the acoustic signal during the stable phase, obtained by conducting multiple sets of repeatable injection molding experiments with standard feed ratios under wear-free screw conditions.

[0039] The peak frequency refers to the frequency point with the highest energy in the spectrum, which can be understood as the dominant tone emitted by the material flow. The frequency of this dominant tone will shift depending on the type of segregation. The coefficient of variation of energy is used to measure the degree of energy fluctuation over time within the sensitive frequency band. The energy fluctuation of a uniform material flow is relatively stable, while the energy output of a segregated material flow fluctuates greatly due to the drastic changes in local flow resistance, thus increasing the coefficient of variation of energy.

[0040] To establish a quantitative evaluation standard, the calculated energy spectrum entropy is compared with the preset baseline entropy value under uniform mixing conditions to obtain the segregation entropy difference. By subtracting the real-time calculated energy spectrum entropy from this baseline entropy value, the resulting segregation entropy difference can quantify the degree to which the current material flow deviates from the ideal mixing state. The larger the segregation entropy difference, the greater the internal order of the material, indicating the occurrence of localized aggregation of powder or binder.

[0041] This baseline entropy value can be collected and calibrated when using a brand-new screw in an injection molding machine and processing standard feed under ideal process parameters. It represents the acoustic characteristics of the material in its most uniform state. Segregation entropy difference directly reflects the degree of deviation between the current material state and the ideal state. Ultimately, the segregation entropy difference, spectral peak frequency, and energy variation coefficient are used together as comprehensive acoustic characteristic parameters to provide a comprehensive and sensitive basis for subsequent segregation identification.

[0042] Based on the extracted acoustic feature parameters, the identification of segregation occurrence areas, segregation types, and segregation degrees in the material flow is achieved through a structured judgment process. First, to accurately locate the physical location of segregation, the barrel is divided into multiple spatial window zones along the screw axis, each corresponding to one or a group of acoustic sensors. The segregation entropy difference obtained from the acquired signals within each window zone is calculated, and window zones with segregation entropy differences exceeding a first segregation threshold are identified as segregation occurrence areas. For example, if the first segregation threshold is set to 0.1, and the calculated segregation entropy difference for the third spatial window zone is 0.18, then the corresponding physical segment of the screw is determined to be a segregation occurrence area. These thresholds are determined by reverse derivation based on the quality index of density uniformity for large components. The first segregation threshold serves as the red line for determining segregation occurrence, and its setting references the minimum material non-uniformity that prevents macroscopic defects in the sintered part, typically set between 15% and 20% of the baseline entropy value.

[0043] After determining the location of segregation, further diagnosis of the specific type of segregation is needed. This is determined based on the direction of the shift in spectral peak frequency. Segregation types mainly include powder agglomeration or sedimentation type and binder enrichment type. The underlying physical logic is that different types of segregation alter the local rheological properties of the material, thereby producing different acoustic responses. When large, hard powder agglomerates form in the material, or when powder settles at the bottom of the screw channel due to gravity, forming a high-concentration layer, these high-density areas generate stronger friction and a lower-pitched impact sound with the barrel wall and screw edges as the screw rotates. This phenomenon is similar to scraping with a rough object, with more acoustic energy concentrated in a relatively low frequency range, causing the spectral peak frequency to shift downward relative to the reference value.

[0044] Conversely, when binder-rich regions appear in the material, the local viscosity of these regions decreases significantly, making the material slipperier. When the screw shears these low-viscosity regions at high speed, local slippage or a high-frequency whistling sound similar to that of fluid passing through a slit may occur, with the acoustic energy concentrating in the high-frequency region, causing the spectral peak frequency to shift upward relative to the reference value.

[0045] Finally, within the segregation region where the location and type of segregation have been determined, the severity of segregation is quantified based on the energy coefficient of variation (ECV). The calculated ECV is compared with several preset severity thresholds, and the segregation severity is classified as slight, moderate, or severe based on the comparison results. For example, two thresholds can be preset: an ECV less than 0.05 indicates slight segregation, between 0.05 and 0.15 indicates moderate segregation, and greater than 0.15 indicates severe segregation. The severity thresholds are then used to classify segregation into different levels based on the magnitude of the ECV, which reflects the intensity of resistance fluctuations during material flow. The threshold range corresponding to slight segregation typically does not trigger automated intervention, while the threshold points for moderate and severe segregation are determined based on the sensitivity curve of injection pressure fluctuations to green compact density, ensuring that intervention is initiated before density deviations transform into irreversible sintering deformation. This classification helps determine whether intervention measures are needed and the intensity of the intervention.

[0046] When moderate or severe segregation is identified, the most appropriate local material state intervention method will be selected and activated within the segregation area based on the diagnosed segregation type. When the segregation level reaches moderate or severe and the spectral peak frequency shifts downward relative to the reference value, indicating that the segregation type is powder agglomeration or sedimentation, local dynamic shear force will be preferentially applied to the segregation area. Because the root cause of this type of segregation is physical particle aggregation, the most direct and effective intervention method is to apply additional mechanical energy to break up these agglomerates.

[0047] When the spectral peak frequency shifts upward relative to the reference value, indicating that the segregation type is binder enrichment, auxiliary fluids are preferentially introduced into the material flow in the segregation region. The root cause of this type of segregation lies in chemical or rheological imbalances, i.e., uneven binder distribution. By introducing specific auxiliary fluids, the rheological properties of the material can be locally altered, such as reducing local viscosity or improving powder wettability, thereby promoting the redistribution of the enriched binder.

[0048] In certain complex operating conditions, the direction of spectral peak frequency shift may be unclear, or multiple segregation phenomena may coexist. To address this, local dynamic shear force and auxiliary fluid are alternately applied in the segregation region, with the rate of change of acoustic characteristic parameters monitored in real time serving as the switching criterion. This is an adaptive, tentative intervention strategy. For example, a short-duration ultrasonic shear force can be applied first, followed immediately by monitoring the rate of change of the segregation entropy difference. If the segregation entropy difference decreases rapidly, indicating that the shear force is effective, it is continued; if the segregation entropy difference does not change or even increases, the shear force is immediately stopped, and the auxiliary fluid is injected, with its effect monitored again. In this way, even in cases of unclear diagnosis, the most effective intervention can be found through real-time feedback.

[0049] When determining whether the segregation type is powder agglomeration or sedimentation and selecting to apply local dynamic shear force, there are multiple ways to achieve this.

[0050] In one embodiment, localized dynamic shear force is applied by activating an array of ultrasonic transducers operating at frequencies from 20 kHz to 40 kHz. Multiple piezoelectric ceramic ultrasonic transducers are pre-integrated at corresponding locations on the outer wall of the barrel. When the controller issues a command, a high-frequency power supply drives these transducers to generate high-frequency vibrations, which penetrate the barrel wall and are transmitted into the internal material flow. In the binder phase of the material, the high-intensity ultrasound induces cavitation, i.e., the generation, growth, and instantaneous collapse of tiny bubbles. Upon collapse, these bubbles generate powerful shock waves and high-speed microjets with extremely high local pressure and shear force. These microjets, like countless miniature hammers, efficiently break up and disperse hard powder agglomerates from the inside and outside without causing excessive macroscopic disturbance to the overall material flow.

[0051] In another embodiment, localized dynamic shear force is applied by using a micro-motor-driven retractable micro-stirring blade to enter the material flow, providing localized mechanical shear force. In this design, tiny sealed channels are formed in the barrel wall, housing stirring blades driven by micro-stepping motors or servo motors. When intervention is needed, the motor-driven blades extend into the material flow within the screw channel and rotate at a set speed. This direct mechanical stirring provides strong, concentrated shear force, forcibly breaking up powder agglomerates and promoting localized mixing of the material. After intervention, the blades can retract into the channels, avoiding interference with normal material transport. Both methods—one a non-contact intervention based on acoustic energy, and the other a contact intervention based on mechanical energy—effectively provide localized dynamic shear force to address powder agglomeration or sedimentation-type segregation.

[0052] When determining that the segregation type is binder enrichment and selecting to introduce an auxiliary fluid, the specific implementation steps are as follows. First, intervention is achieved by adjusting the injection volume of the auxiliary fluid through the micro-orifice nozzles. Multiple micro-orifice nozzles, controlled by high-precision solenoid valves or piezoelectric valves, are integrated at corresponding positions on the barrel wall. The orifice diameter of these nozzles is typically designed between 0.2 mm and 0.5 mm to ensure the ejection of fine droplets or liquid streams, thereby achieving precise control over the amount of auxiliary fluid introduced.

[0053] The selected auxiliary fluids typically include low-viscosity polymer solutions or fatty acid ester dispersants. Low-viscosity polymer solutions, such as low-molecular-weight polyethylene wax solutions, act similarly to diluents. When sprayed onto areas rich in binder, they can rapidly mix and blend with the high concentration of binder, thereby locally and controllably reducing the viscosity of the material in that area, making its rheological properties closer to those of the surrounding normal material, and improving flowability.

[0054] Fatty acid ester dispersants are surfactants with molecular structures that are both powder- and binder-dependent. When sprayed into materials, they can adsorb onto the surface of powder particles, reducing van der Waals forces between particles and improving the wettability between powder particles and binders. This allows the enriched binder liquid phase to spread over a larger powder surface area instead of agglomerating.

[0055] The entire intervention process is not a one-time open-loop operation, but a continuously optimized closed-loop control process. After the intervention is implemented, the acoustic vibration signal is monitored in real time, and the acoustic characteristic parameters are extracted again. Based on the changes in the extracted acoustic characteristic parameters, the amount of auxiliary fluid introduced or the intensity of dynamic shear force is dynamically adjusted.

[0056] Specifically, firstly, the acoustic vibration signals after the intervention are monitored in real time, and the acoustic characteristic parameters are extracted again. Then, the first rate of change of the segregation entropy difference before and after the intervention, and the second rate of change of the energy coefficient of variation are calculated. These two rates of change directly quantify the immediate effect of the intervention measures, whether it is a positive improvement or a negative deterioration, and the speed of improvement or deterioration.

[0057] Subsequently, different control strategies are implemented based on the sign and magnitude of the two rates of change. When both the first and second rates of change are negative, it indicates that the segregation entropy difference and energy variation coefficient are decreasing, the material uniformity is improving, and the intervention is effective. At this point, if the absolute value of the first rate of change is greater than or equal to a preset first threshold, and the absolute value of the second rate of change is greater than or equal to a preset second threshold, it indicates that the improvement speed is rapid and the expected effect has been achieved. In this case, the current intervention parameters should be kept unchanged to avoid over-adjustment and the introduction of new fluctuations. Otherwise, if the improvement speed is slow, and either rate of change fails to meet the threshold requirement, a fuzzy PID control algorithm is used to dynamically adjust the amount of auxiliary fluid introduced or the intensity of dynamic shear force to accelerate the improvement speed.

[0058] The first threshold is used to determine whether the rate of change of segregation entropy difference has reached an effective improvement rate, and the second threshold is used to determine whether the rate of change of energy coefficient of variation has reached an effective improvement rate. Both are pre-calibrated based on a standard segregation perturbation experiment. The calibration process is as follows: Under a uniform mixing state, a standard segregation perturbation is actively introduced through a micro-orifice nozzle. The 30th percentiles of the rates of change of segregation entropy difference and energy coefficient of variation within 3 seconds after the perturbation are collected, and this is repeated 5 times, with the average value corresponding to the first and second thresholds. If calibration is not available, the recommended values ​​can be preferred: the first threshold is 0.10 s^-1, and the second threshold is 0.15 s^-1. These recommended values ​​correspond to the minimum rate at which a perceptible improvement in the segregation state occurs. When the absolute values ​​of the corresponding rates of change are lower than the above recommended values, it indicates that the intervention effect is weak and active adjustment is required; when they are higher than the above recommended values, it indicates that the intervention is effective and the current parameters can be maintained.

[0059] Specifically, when both the first and second rates of change are positive, it indicates that the material uniformity is deteriorating, and the current intervention measures may be inappropriate or insufficient. If the first rate of change is less than a preset first threshold or the second rate of change is less than a preset second threshold, the rate of deterioration is still within a controllable range. In this case, a fuzzy PID control algorithm should be used to dynamically adjust the amount of auxiliary fluid introduced or the intensity of dynamic shear force to try to find better intervention parameters. However, if the rate of deterioration is very rapid, and one or both rates of change exceed the threshold, this may indicate a more serious process problem. In this case, intervention should be stopped immediately and an alarm should be issued to prompt the operator to conduct a manual inspection.

[0060] When one of the first and second rates of change is positive and the other is negative, it indicates that the intervention effect is uncertain; it may improve one aspect but worsen another. In this complex situation, the fuzzy PID control algorithm is also used to dynamically adjust the amount of auxiliary fluid introduced or the intensity of dynamic shear force. However, to prevent the system from oscillating violently under uncertain conditions, the calculated adjustment amount is multiplied by a coefficient less than 1, such as 50%, to limit the magnitude of the adjustment and make more cautious fine adjustments.

[0061] In the above control strategy, a fuzzy PID control algorithm is used to dynamically adjust the amount of auxiliary fluid introduced or the intensity of dynamic shear force. The specific implementation steps are as follows: First, the difference *e* between the segregation entropy difference after intervention and the reference entropy value is used as the first input variable of the PID control algorithm. This *e* intuitively represents the deviation between the current state and the target state. Simultaneously, the instantaneous rate of change *ec* of the segregation entropy difference is used as the second input variable of the PID control algorithm. *ec* represents the trend of deviation change, and *ec* = *de* / dt. Then, the precise values ​​*e* and *ec* are fuzzified using membership functions, transforming them into fuzzy linguistic variables such as large deviation, medium deviation, small deviation, positive rate of change of deviation, and zero rate of change of deviation.

[0062] Fuzzy processing is the process of mapping precise deviation values ​​to fuzzy sets. Taking the difference in segregation entropy, e, as an example, its range is set to -1 to +1. Using a triangular membership function, e is divided into five linguistic variables: negative large, negative medium, zero, positive medium, and positive large. Similarly, ec is also divided into five linguistic variables: negative large, negative medium, zero, positive medium, and positive large. This processing method allows the control logic to no longer rely on precise mathematical formulas, but rather simulates human experience to make fuzzy judgments about the severity of segregation. When the segregation entropy difference increases rapidly, the control output exhibits a non-linear, rapid response, while when the deviation is small, the output tends to be smoother, thus ensuring the adjustment speed while avoiding violent oscillations in the internal pressure field of the screw.

[0063] Next, based on these linguistic variables and the pre-defined fuzzy rule table, fuzzy inference is performed, outputting the proportional coefficient increment ΔKp, integral coefficient increment ΔKi, and differential coefficient increment ΔKd. Specifically, the fuzzy rule table is shown in Table 1 below: Table 1

[0064] In the table, the triples in each cell represent the fuzzy linguistic values ​​of (ΔKp, ΔKi, ΔKd) in sequence, with NB = negative large, NM = negative medium, NS = negative small, ZO = zero, PS = positive small, PM = positive medium, and PB = positive large.

[0065] The fuzzy rule table can be formulated based on the following logic: when the absolute value of the deviation e is greater than 0.20, that is, when the deviation is very large, the proportional coefficient Kp should be increased to speed up the response and allow the system to quickly approach the target value. At the same time, the integral coefficient Ki needs to be limited to prevent integral saturation caused by the large deviation over a long period of time.

[0066] When the absolute value of the deviation e is between 0.05 and 0.20, i.e. the deviation is moderate, Kp should be appropriately reduced to prevent overshoot, and a differential coefficient Kd should be introduced to predict deviation changes and suppress system oscillations.

[0067] When the absolute value of the deviation e is less than or equal to 0.05, that is, when the deviation is very small, the system is close to steady state. At this time, Ki should be increased to eliminate steady-state segregation residue and improve control accuracy.

[0068] The precise incremental values ​​of ΔKp, ΔKi, and ΔKd are obtained by defuzzifying them using the centroid method. These values ​​are then added to the initial PID parameters Kp0, Ki0, and Kd0 to obtain the PID parameters Kp, Ki, and Kd at the current time.

[0069] Finally, according to the output formula of the fuzzy PID control algorithm The current control output quantity *u* is calculated. When *u* is positive, it indicates a need for enhanced intervention, i.e., increasing the amount of auxiliary fluid introduced or the intensity of dynamic shear force; when *u* is negative, it indicates a need for weakened intervention, i.e., decreasing the amount of auxiliary fluid introduced or the intensity of dynamic shear force; when *u* = 0, the amount of auxiliary fluid introduced or the intensity of dynamic shear force remains unchanged. Through this series of steps, intelligent, adaptive closed-loop adjustment of the intervention measures is achieved. It is important to emphasize that *u* does not refer to the amount of auxiliary fluid introduced or the intensity of dynamic shear force itself, but is a dimensionless signal whose value range is limited to [−1, 1]. *u* > 0 indicates a need to enhance the intensity of the current intervention, and *u* < 0 indicates a need to weaken the intensity of the current intervention. If the current intervention method is applying local dynamic shear force, *u* is linearly mapped to the output power of the ultrasonic transducer or the rotational speed of the micro-stirring blade. If the current intervention method is introducing auxiliary fluid, *u* is linearly mapped to the injection pressure of the micro-orifice nozzle.

[0070] Specifically, the linear mapping of actual physical quantities is calculated using the following formula: , u≥0; , u < 0.

[0071] in, The initial settings for the current intervention method (e.g., the initial amplitude of the ultrasonic transducer, the initial rotational speed of the micro-stirring blade, or the initial injection pressure of the micro-orifice nozzle). and These represent the upper and lower limits of the actuator's safe operating range, respectively.

[0072] When the intervention method is to apply local dynamic shear force This represents the output power of the ultrasonic transducer, or the rotational speed of the micro-stirring blades; when the intervention method is the introduction of auxiliary fluid, This represents the injection pressure of the micro-orifice nozzle.

[0073] Through the above linear mapping, the normalized control commands can be unambiguously converted into the physical control quantities actually needed by each actuator, thereby achieving continuous and precise adjustment of the intervention intensity.

[0074] In one specific implementation, the injection molding screw needs to process high-density tungsten-nickel-iron (W-Ni-Fe) heavy alloy powder during the production of guide vanes for small and medium-sized gas turbines. The cross-sectional height of guide vanes for small and medium-sized gas turbines is typically between 10mm and 200mm, and the axial width is approximately 80mm to 120mm, making them typical precision and complex structural components. Located between the combustion chamber outlet and the turbine rotor, the guide vanes withstand the impact of high-temperature airflow reaching 700°C to 1000°C, placing stringent requirements on the material's high-temperature strength, oxidation resistance, and dimensional accuracy.

[0075] In injection molding (MIM), the density of tungsten powder is as high as 19.3 g / cm³, while the density of the binder system is typically around 1.0 g / cm³. This significant density difference leads to powder enrichment due to gravity settling during transport through a barrel several meters long, resulting in segregation. This phenomenon is consistent with segregation behavior observed in W-Ni-Fe heavy alloy injection molding studies, where the density difference between the powder and binder causes localized changes in their proportions during mold filling. MIM effectively avoids the unavoidable density gradient problem inherent in traditional powder metallurgy pressing processes.

[0076] For this scenario involving high-density, easily settling materials, an acoustic sensor array positioned in the middle of the feed cylinder captured a significant downward shift in the spectral frequency. The peak frequency shifted from the baseline of 25 kHz to approximately 18 kHz, and the energy spectral entropy decreased significantly, with the segregation entropy difference exceeding the first threshold. This downward shift in the spectral frequency indicates the sedimentation and agglomeration of heavy powder particles. High-density tungsten particles settle to the bottom of the screw channel under gravity, altering the sound wave propagation characteristics within the material. At this point, the system identifies it as severe powder sedimentation-type segregation, with the segregation degree reaching a severe level.

[0077] To address the aggregation of this heavy powder, the intervention module did not rely solely on fluid injection, but instead prioritized activating a micro-motor located in the segregation region. This motor drives retractable micro-stirring blades to perform localized shearing at 500 rpm against the direction of screw rotation, forcibly agitating the heavy powder deposited at the bottom of the screw channel and resuspending it in the binder.

[0078] Applying localized dynamic shear force can reduce the apparent viscosity of the feed in the segregation region, promoting the reintegration of deposited powder into the mainstream material; simultaneously, the shearing force against the screw rotation direction can also disrupt the formed powder agglomeration structure, forcing the heavy particles to redisperse. The micro-stirring blades are treated with a wear-resistant coating to cope with the high abrasiveness of tungsten powder.

[0079] During the operation of the stirring blades, sensors collect vibration signals in real time after intervention. It was found that due to the additional high-frequency mechanical friction introduced by mechanical stirring, the high-frequency energy in the spectrum began to rise, but at the same time, signs of localized temperature increases appeared. To prevent localized overheating caused by mechanical stirring, the adjustment module dynamically reduces the motor speed using a fuzzy PID algorithm based on the real-time monitored rate of change of the energy variation coefficient. This achieves a precise repair solution for severe sedimentation and segregation through mechanical shearing.

[0080] Please refer to Figure 2 , Figure 3 This application also proposes a parameter adjustment system for powder metallurgy injection molding of large components. The system is used to implement the steps of any of the above methods, and the system includes: Acquisition module 201: Acquires acoustic vibration signals of the injection screw during material transport and mixing; Extraction module 202: Performs frequency domain analysis on acoustic vibration signals to extract acoustic characteristic parameters that reflect the uniformity of materials; Identification module 203: Based on acoustic feature parameters, it identifies the segregation area, segregation type, and segregation degree in the material flow; Intervention module 204: When the degree of segregation reaches moderate or severe segregation, the method of intervention on the local material state in the segregation occurrence area is selected according to the segregation type. The intervention method includes introducing auxiliary fluid into the material flow to adjust the local rheological properties and applying local dynamic shear force to promote powder dispersion. Adjustment module 205: Real-time monitoring of acoustic vibration signals after intervention, and extraction of acoustic characteristic parameters again. Based on the changes in the extracted acoustic characteristic parameters, the amount of auxiliary fluid introduced or the intensity of dynamic shear force is adjusted.

[0081] Specifically, the acquisition module 201 can consist of a series of highly sensitive sensors and a data acquisition unit. These sensors are strategically positioned outside the injection molding machine barrel to capture the weak acoustic vibrations generated during screw operation. For example, a MEMS-based miniature microphone array or fiber optic acoustic sensor can be used, which can convert mechanical vibrations or sound waves into electrical signals and transmit them to the data acquisition unit via a high-speed data bus. The data acquisition unit is responsible for performing analog-to-digital conversion on these analog signals and for initial signal amplification and filtering to ensure data quality and integrity.

[0082] The extraction module 202 can consist of one or more digital signal processors (DSPs) or high-performance microcontrollers, with various frequency domain analysis algorithms pre-built into it. This module receives the raw digital signal from the acquisition module and performs operations such as Fourier transform and wavelet analysis to reveal the signal's spectral characteristics. For example, parameters such as energy density, harmonic component intensity, or spectral entropy within a specific frequency band can be calculated as acoustic characteristic parameters reflecting material uniformity. The calculation logic for these parameters can be programmed and embedded in the DSP for efficient real-time processing.

[0083] The recognition module 203 can be an embedded processor or an industrial PC running a recognition algorithm based on machine learning or an expert system. This module receives the acoustic feature parameters output by the extraction module and compares them with a pre-established segregation model. For example, a trained neural network model can output the specific location of the segregation, whether the segregation is powder agglomeration, sedimentation, or binder enrichment, and the severity of the segregation (e.g., slight, moderate, severe) based on the pattern of the input feature parameters. This module may also include a decision logic unit to trigger corresponding intervention commands based on the recognition results.

[0084] The intervention module 204 can consist of a series of actuators and their drive circuits. For example, when auxiliary fluid needs to be introduced, this module can control the opening and flow rate of a high-precision micro-pump and micro-orifice nozzle to precisely inject the auxiliary fluid into the segregation region. When local dynamic shear force needs to be applied, this module can drive a micro-vibrator, electromagnetic actuator, or small mechanical stirring device to generate high-frequency vibration or mechanical shearing in a specific area. The selection and arrangement of these actuators should ensure that they can have a localized and controllable influence on the material flow without interfering with the overall molding process.

[0085] The adjustment module 205 can be a standalone control unit, such as a programmable logic controller (PLC) or an industrial-grade computer, integrating a feedback control algorithm. This module continuously receives real-time acoustic characteristic parameters from the acquisition and extraction modules and compares them with the target uniformity state. Based on the comparison results, the adjustment module calculates adjustment commands for the auxiliary fluid introduction rate or dynamic shear force intensity. For example, a proportional-integral-derivative (PID) controller or a fuzzy logic controller can be used to dynamically adjust the output of the intervention module according to the degree and trend of segregation to minimize the degree of material segregation. This module should also have fault diagnosis and alarm functions to handle abnormal situations.

[0086] The core innovation of the powder metallurgy injection molding parameter adjustment system for large components proposed in this application, compared to existing technologies, lies in its real-time, non-invasive monitoring and intelligent, localized intervention of material segregation during the injection molding process. Traditional injection molding machine control systems often treat high-frequency, minute fluctuations generated by the internal flow of materials as noise and filter them out directly, resulting in an inability to perceive the microscopic non-uniformity of the material, let alone provide effective compensation. This makes it difficult to detect and resolve microscopic segregation problems during the molding of large components, ultimately accumulating and amplifying into macroscopic defects.

[0087] This system, through the refined acquisition and analysis of acoustic vibration signals by the acquisition module 201 and the extraction module 202, successfully transforms these traditionally ignored noises into key information reflecting the uniformity of materials, thereby achieving a thorough perception of the segregation state within the material. The introduction of the identification module 203 enables the system to accurately locate segregation areas and determine the type and degree of segregation, providing a basis for subsequent precise intervention. The collaborative work of the intervention module 204 and the adjustment module 205 allows for targeted measures, such as introducing auxiliary fluids or applying dynamic shear forces, in localized areas based on the specific segregation situation, with real-time feedback adjustments. This intelligent adjustment mechanism avoids the problems of over-intervention or under-intervention that may arise from traditional global parameter adjustments, significantly improving the forming quality and performance stability of large powder metallurgy components.

[0088] Therefore, this system effectively solves the limitations of traditional methods in controlling material uniformity, provides reliable technical support for the manufacturing of high-performance large components, and has significant progressive and practical value.

[0089] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for adjusting parameters in powder metallurgy injection molding of large components, used to adjust the segregation state of materials during injection molding, characterized in that, The method includes the following steps: S1: Acquire acoustic vibration signals of the injection screw during material transport and mixing; S2: Perform frequency domain analysis on the acoustic vibration signal to extract acoustic characteristic parameters that reflect the uniformity of the material; S3: Based on the acoustic characteristic parameters, identify the segregation region, segregation type, and segregation degree in the material flow; S4: When the degree of segregation reaches moderate or severe segregation, a method of intervention on the local material state is selected in the segregation occurrence area according to the segregation type. The intervention method includes introducing auxiliary fluid into the material flow to adjust the local rheological properties and applying local dynamic shear force to promote powder dispersion. S5: Monitor the acoustic vibration signal after intervention in real time, and extract the acoustic feature parameters again. Based on the changes in the extracted acoustic feature parameters, adjust the amount of auxiliary fluid introduced or the intensity of the dynamic shear force.

2. The method for adjusting parameters in powder metallurgy injection molding of large components according to claim 1, characterized in that, Step S1 includes: S11: The vibration signal generated by the operation of the injection screw is synchronously acquired by an array of piezoelectric accelerometers arranged along the outer wall of the barrel at a sampling frequency of not less than 100kHz. S12: Perform bandpass filtering preprocessing on the acquired raw signal to filter out power frequency interference and low-frequency mechanical vibration noise; S13: Perform a fast Fourier transform or wavelet transform on the preprocessed signal to extract the spectral features in the high-frequency band from 5kHz to 50kHz as the acoustic feature parameters.

3. The method for adjusting parameters in powder metallurgy injection molding of large components according to claim 1, characterized in that, Step S2 includes: S21: Perform a short-time Fourier transform on the acoustic vibration signal to obtain the time-frequency domain energy distribution matrix; S22: Based on the time-frequency domain energy distribution matrix, calculate the energy spectral entropy, spectral peak frequency, and energy variation coefficient within at least one sensitive frequency band; S23: The energy spectral entropy is compared with the preset reference entropy value under uniform mixing state to obtain the segregation entropy difference; the acoustic characteristic parameters include the segregation entropy difference, the spectral peak frequency and the energy variation coefficient.

4. The method for adjusting parameters in powder metallurgy injection molding of large components according to claim 3, characterized in that, Step S3 includes: S31: Divide the barrel into multiple spatial window areas along the screw axis, calculate the segregation entropy difference in each window area, and determine the window area where the segregation entropy difference exceeds the first segregation threshold as the segregation occurrence area. S32: Determine the segregation type based on the shift direction of the spectral peak frequency. The segregation type includes powder agglomeration or sedimentation type, and binder enrichment type. S33: Within the segregation region, the energy variation coefficient is compared with multiple preset degree thresholds, and the degree of segregation is classified as slight segregation, moderate segregation, or severe segregation based on the comparison results.

5. The method for adjusting parameters in powder metallurgy injection molding of large components according to claim 4, characterized in that, Step S4 includes: S41: When the degree of segregation reaches moderate or severe segregation, and the peak frequency of the spectrum shifts down relative to the reference value, and the segregation type is determined to be powder agglomeration or sedimentation, local dynamic shear force is preferentially applied to the segregation area. S42: When the peak frequency shifts upward relative to the reference value, and the segregation type is determined to be binder enrichment type, auxiliary fluid is preferentially introduced into the material flow in the segregation region. S43: When the direction of the spectral peak frequency shift is unclear, local dynamic shear force and auxiliary fluid are alternately applied in the segregation region, and the switching is based on the rate of change of acoustic characteristic parameters monitored in real time.

6. The method for adjusting parameters in powder metallurgy injection molding of large components according to claim 5, characterized in that, In step S41, local dynamic shear force is preferentially applied to the segregation region, including but not limited to: S411: Activates an ultrasonic transducer array with an operating frequency of 20kHz to 40kHz to break up powder agglomerates by inducing cavitation and microjets in the material. Alternatively, a micro motor can be used to drive retractable micro-stirring blades into the material flow, providing localized mechanical shearing force.

7. The method for adjusting parameters in powder metallurgy injection molding of large components according to claim 5, characterized in that, In step S42, preferentially introducing auxiliary fluid into the material flow in the segregation region includes the following steps: S421: Adjust the injection volume of the auxiliary fluid injected by the micro-orifice nozzle, wherein the orifice diameter of the micro-orifice nozzle is 0.2 mm to 0.5 mm; The auxiliary fluid includes a low-viscosity polymer solution or a fatty acid ester dispersant, used to locally reduce the viscosity of the material and improve the wettability of the powder.

8. The method for adjusting parameters in powder metallurgy injection molding of large components according to claim 1, characterized in that, Step S5 includes: S51: Real-time monitoring of acoustic vibration signals after intervention, extraction of acoustic characteristic parameters again, and calculation of the first rate of change of segregation entropy difference before and after intervention, as well as the second rate of change of energy variation coefficient; S52: When both the first rate of change and the second rate of change are negative, if the absolute value of the first rate of change is greater than or equal to a preset first threshold and the absolute value of the second rate of change is greater than or equal to a preset second threshold, the current intervention parameters remain unchanged; otherwise, the fuzzy PID control algorithm is used to dynamically adjust the amount of auxiliary fluid introduced or the intensity of dynamic shear force. S53: When both the first rate of change and the second rate of change are positive, if the first rate of change is less than a preset first threshold or the second rate of change is less than a preset second threshold, the fuzzy PID control algorithm is used to dynamically adjust the amount of auxiliary fluid introduced or the intensity of dynamic shear force; otherwise, the intervention is stopped and an alarm message is issued. S54: When one of the first rate of change and the second rate of change is positive and the other is negative, a fuzzy PID control algorithm is used to dynamically adjust the amount of auxiliary fluid introduced or the intensity of dynamic shear force, and the calculated adjustment amount is multiplied by 50% to limit the magnitude of the adjustment amount.

9. The method for adjusting parameters in powder metallurgy injection molding of large components according to claim 8, characterized in that, In steps S52 and S53, the dynamic adjustment of the amount of auxiliary fluid introduced or the intensity of dynamic shear force using a fuzzy PID control algorithm includes the following steps: S55: The difference e between the segregation entropy difference after intervention and the reference entropy value is used as the first input variable of the PID control algorithm, and the instantaneous change rate ec of the segregation entropy difference is used as the second input variable of the PID control algorithm. e and ec are fuzzified into linguistic variables through the membership function respectively. S56: Based on the linguistic variables and the preset fuzzy rule table, output the proportional coefficient increment ΔKp, integral coefficient increment ΔKi, and differential coefficient increment ΔKd; the fuzzy rule table is formulated based on the following logic: when |e|> 0.20, increase Kp to speed up the response, while limiting Ki to prevent integral saturation; when 0.05 < |e| ≤ 0.20 is moderate, appropriately decrease Kp and introduce Kd to suppress overshoot; when |e| ≤ 0.05, increase Ki to eliminate steady-state segregation residue; S57: Add ΔKp, ΔKi, and ΔKd to the initial PID parameters Kp0, Ki0, and Kd0 to obtain the PID parameters Kp, Ki, and Kd at the current time. S58: According to the fuzzy PID control algorithm When u is positive, it increases the amount of auxiliary fluid introduced or the intensity of dynamic shear force; when u is negative, it decreases the amount of auxiliary fluid introduced or the intensity of dynamic shear force.

10. A parameter adjustment system for powder metallurgy injection molding of large components, characterized in that, The system is used to implement the steps of the method according to any one of claims 1-9, and the system includes: Acquisition module: Acquires acoustic vibration signals of the injection screw during material transport and mixing; Extraction module: Performs frequency domain analysis on the acoustic vibration signal to extract acoustic feature parameters that reflect the uniformity of the material; Identification module: Based on the acoustic feature parameters, it identifies the segregation region, segregation type, and degree of segregation in the material flow; Intervention module: When the segregation degree reaches moderate or severe segregation, the module selects a method to intervene in the local material state in the segregation occurrence area according to the segregation type. The intervention method includes introducing auxiliary fluid into the material flow to adjust the local rheological properties and applying local dynamic shear force to promote powder dispersion. Adjustment module: Real-time monitoring of acoustic vibration signals after intervention, and extraction of acoustic characteristic parameters again. Based on the changes in the extracted acoustic characteristic parameters, the amount of auxiliary fluid introduced or the intensity of the dynamic shear force is adjusted.

Images