Method for eccentric balancing of a main spindle of a machining center

Abstract

This invention discloses a method for balancing the spindle eccentricity of a piston machining center, comprising the following steps: S1, clamping and parameter calibration: A standard piston sample with the same specifications and material as the piston to be machined is clamped, and a hydraulic chuck is used to complete the positioning and clamping, ensuring that the clamping deviation is ≤0.002mm. Spindle radial runout, no-load and light-load torque data are collected using a laser displacement sensor and a torque sensor. The initial eccentricity e0 and initial imbalance U0 are calculated based on the radial runout data. A BP neural network algorithm is used to establish a correlation model between the eccentricity e and the spindle speed n and cutting load F to determine the target balance parameters under different working conditions. This invention employs the synergistic effect of dual-plane dynamic counterweight and CNC eccentricity compensation, combined with a BP neural network correlation model, to stably control the spindle eccentricity to ≤0.005mm, piston skirt ellipticity to ≤0.01mm, pin hole coaxiality to ≤0.01mm, and surface roughness Ra to ≤0.8μm. Compared with traditional balancing methods, the machining accuracy is improved by more than 40%, significantly increasing the product qualification rate.

Description

Technical Field

[0001] This invention belongs to the field of piston machining technology, and particularly relates to a method for eccentric balancing of the spindle of a piston machining center. Background Technology

[0002] As a core reciprocating component of an engine, the piston's machining accuracy directly determines the engine's power performance, fuel economy, and operational reliability. Spindle eccentricity balancing is a key control element in high-precision machining of piston skirt ellipticity and pin hole eccentricity. Currently, piston machining centers generally employ high-speed spindles for efficient cutting. However, the vibration, thermal deformation, and cutting trajectory deviation caused by spindle eccentricity directly lead to problems such as excessive piston skirt ellipticity, low pin hole coaxiality, and deteriorated surface roughness, becoming a core bottleneck restricting piston machining accuracy and batch consistency.

[0003] Traditional piston spindle balancing methods can only perform static correction and have poor dynamic adaptation. They cannot compensate in real time for cutting loads, thermal deformation, and clamping deviations, leading to excessive spindle eccentricity and vibration. This makes it difficult to reliably ensure high-precision piston machining. Furthermore, they lack an intelligent correlation model between eccentricity and operating conditions, and balancing parameters rely on manual experience for adjustment. Under high-speed conditions, the balancing accuracy is low and the response is slow, making it difficult to improve batch machining consistency and production efficiency. To address these issues, we propose a new spindle eccentricity balancing method for piston machining centers. Summary of the Invention

[0004] To address the problems in the prior art, the present invention proposes the following technical solution: The method for balancing the eccentricity of the spindle in a piston machining center includes the following steps: S1. Clamping and Parameter Calibration: A standard piston sample with the same specifications and material as the piston to be processed is clamped. A hydraulic chuck is used to complete the positioning and clamping to ensure that the clamping deviation is ≤0.002mm. The radial runout, no-load and light-load torque data of the spindle are collected by laser displacement sensor and torque sensor. The initial eccentricity e0 and the initial imbalance U0 are calculated by combining the radial runout data. The BP neural network algorithm is used to establish a correlation model between the eccentricity e and the spindle speed n and the cutting load F to determine the target balance parameters under different working conditions. S2. Static counterweight compensation: Based on the calibrated target balance parameters, drive the double-plane counterweight blocks located at the front and rear ends of the spindle to move to the target position to complete the static counterweight, so that the unbalance of the spindle under no-load condition is ≤20g·mm and the effective vibration value is ≤0.05mm / s. S3. Online detection and dynamic correction: During the machining process, vibration, temperature and force data of the spindle vibration, temperature rise and cutting force are collected in real time by vibration sensor, temperature sensor and force sensor. When the vibration value is ≥0.09mm / s or the temperature rise is ≥5℃ or the cutting force change amplitude is ≥50N, the position and phase of the counterweight are automatically adjusted and the CNC system is linked to compensate for the spindle eccentricity and correct the cutting trajectory. S4. Closed-loop optimization: After processing, the key dimensions of the piston are automatically measured. Based on the dimensional deviation, the counterweight parameters in the correlation model are optimized in reverse, the target balance parameters are updated, and closed-loop calibration is achieved.

[0005] As a preferred embodiment of the above technical solution, the initial eccentricity e0 and the initial imbalance U0 are calculated using the least squares method, with a calculation error of ≤5%. The target balance parameters include target eccentricity ≤0.005mm and target imbalance ≤20g·mm.

[0006] As a preferred embodiment of the above technical solution, the dual-plane counterweights are respectively located at the front end of the spindle near the chuck and at the rear end near the bearing, which can realize radial sliding and phase synchronous adjustment, with a displacement accuracy of ≤0.01mm and a phase adjustment accuracy of ≤1°, and can be adapted to a speed range of 8000rpm-15000rpm.

[0007] As a preferred embodiment of the above technical solution, the laser displacement sensor is selected as KEYENCE IL-600, with a measurement accuracy of ≤0.001mm; the vibration sensor is selected as ACM1000, with a measurement accuracy of ±0.001mm / s; the temperature sensor is selected as PT1000, with a measurement accuracy of ±0.1℃; and the torque sensor is selected as HBM T40B, with a measurement range of 0-50N·m, an accuracy of ±0.1N·m, a data acquisition frequency of 1000Hz, and a cutting force data acquisition interval of 10ms.

[0008] As a preferred embodiment of the above technical solution, the compensation accuracy of the spindle eccentricity compensation of the CNC system linkage servo axis is ≤0.001mm. During dynamic correction, the displacement fine adjustment of the counterweight block is ≤0.01mm, the phase fine adjustment is ≤0.5°, the spindle vibration value after correction is ≤0.05mm / s, and the spindle temperature rise is ≤5℃.

[0009] As a preferred embodiment of the above technical solution, the key dimensions of the piston to be measured include the ellipticity of the skirt, the coaxiality of the pin hole, and the roundness. The associated model can update the parameters in real time according to the machining conditions.

[0010] As a preferred embodiment of the above technical solution, the counterweight is made of high-density alloy material, and the mass adjustment range of a single counterweight is 50g-200g. The combination of counterweights can be flexibly matched according to the specifications and mass of the piston to be processed.

[0011] The beneficial effects of this invention are as follows: 1. This invention employs a dual-plane dynamic counterweight and CNC eccentricity compensation working in synergy, combined with a BP neural network correlation model, to stably control the spindle eccentricity to ≤0.005mm, piston skirt ellipticity to ≤0.01mm, pin hole coaxiality to ≤0.01mm, and surface roughness Ra to ≤0.8μm. Compared with traditional balancing methods, the machining accuracy is improved by more than 40%, significantly increasing the product qualification rate. 2. This invention achieves real-time monitoring of working conditions and adaptive adjustment of balance parameters through multi-sensor fusion detection and intelligent algorithm modeling, without the need for manual intervention, thus lowering the technical threshold for operators. At the same time, it can record various data during the processing, providing data support for subsequent process optimization and further improving processing stability and accuracy. 3. This invention achieves fully automated control of the entire process through offline calibration, dynamic compensation, and closed-loop optimization. It eliminates the need for manual shutdown to adjust the counterweight, eliminates the need for secondary compensation in a single processing cycle, improves processing efficiency by more than 30%, is suitable for mass piston production, reduces labor costs, increases production cycle time, and meets the needs of large-scale production. 4. This invention is suitable for high-speed machining conditions of 8000rpm-15000rpm. Through real-time detection and dynamic compensation, it can effectively offset the eccentricity caused by factors such as dynamic changes in cutting load, thermal deformation, and clamping deviation, and control the effective value of spindle vibration within ≤0.05mm / s, reduce spindle bearing wear, extend the service life of spindle and bearing by more than 30%, and reduce equipment maintenance costs. Detailed Implementation

[0012] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below in conjunction with the embodiments.

[0013] Example 1: Eccentric machining of aluminum piston skirt This embodiment addresses the scenario of eccentric machining of the skirt of a φ90mm aluminum piston. This aluminum piston is a special piston for automotive engines, made of 6061 aluminum alloy with a density of 2.7g / cm³. 3 The piston skirt design requires an ellipticity of ≤0.01mm and a surface roughness Ra≤0.8μm after machining. The spindle eccentricity balance control is performed using the method of this invention, and the specific steps are as follows: S1. Clamping and parameter calibration: Select a standard φ90mm aluminum piston sample that is completely consistent with the specifications and material of the aluminum piston to be processed. The sample is free from deformation and damage, and has a surface roughness Ra≤0.4μm. It is used as the reference sample for parameter calibration. The sample was positioned and clamped using a hydraulic chuck. During the clamping process, a dial indicator was used to monitor the clamping deviation in real time. The clamping pressure of the hydraulic chuck was adjusted to 0.4MPa to ensure that the clamping was firm and without deformation. The final clamping deviation was 0.0015mm, ensuring that the clamping deviation was ≤0.002mm. Start the machining center spindle and set the no-load running speed to 10,000 rpm, keeping the speed stable (speed fluctuation ≤ ±50 rpm). Use a laser displacement sensor (model KEYENCE IL-600, measurement accuracy ≤ 0.001 mm), align the sensor probe with the front end of the spindle near the chuck, and collect the spindle radial runout data in real time. Set the sampling frequency to 1000 Hz and the sampling time to 35 seconds, collecting a total of 35,000 sets of data. After data filtering and noise reduction, the calculated average spindle radial runout is 0.02 mm, the maximum value is 0.023 mm, and the minimum value is 0.017 mm. The data fluctuation meets the requirements. Meanwhile, a torque sensor (model HBM T40B, measurement range 0-50 N·m, accuracy ±0.1 N·m) was used to collect torque data of the spindle under no-load conditions. The acquisition time was 10 seconds, and the average no-load torque was 8 N·m, with torque fluctuation ≤ ±0.2 N·m. Then, a light-load cutting condition was set up, with the cutting force controlled at 150N (within the light-load range of 100N-200N). Torque data under light-load conditions were collected, with an average value of 10.5N·m. Combining the radial runout data and torque data, the initial eccentricity e0=0.025mm and the initial unbalance U0=150g·mm were accurately calculated using the least squares method. The calculation error was 4.2%, which is less than the preset error requirement of 5%, and the basic parameter calibration was completed. Subsequently, a correlation model was established, employing orthogonal experimental design. Four gradients were set for spindle speeds: 8000 rpm, 10000 rpm, 12000 rpm, and 15000 rpm; four gradients were set for cutting forces: 200 N, 400 N, 600 N, and 800 N. Three sets of data were repeatedly collected for each working condition combination, and the average of the three sets was taken as the valid data. A total of 4 × 4 × 3 = 48 sets of eccentricity data were collected to ensure the representativeness and reliability of the data. A nonlinear correlation model between eccentricity e and spindle speed n and cutting load F was constructed using a BP neural network algorithm. This neural network had an input layer of 2 neurons (corresponding to speed n and cutting load F respectively), a hidden layer of 10 neurons, and an output layer of 1 neuron (corresponding to eccentricity e). Gradient descent was used to optimize the model parameters, with 1000 training iterations. The training error converged to within 0.0001, ensuring the model's fitting accuracy.

[0014] Based on the actual requirements of machining the aluminum piston skirt in this embodiment, the optimal working conditions for machining the aluminum piston skirt of this specification are determined to be a rotational speed of 12000 rpm and a cutting force of 500 N. The corresponding target balance parameters are a target eccentricity ≤ 0.005 mm and a target imbalance ≤ 20 g·mm. The trained model parameters are imported into the CNC system of the machining center (model FANUC 0i-MF) to complete the model deployment and ensure that the CNC system can accurately call the model parameters according to the real-time working conditions.

[0015] S2. Static counterweight compensation: The CNC system calls the deployed associated model, inputs the specifications (φ90mm), material (6061 aluminum alloy), and preset machining parameters (speed 12000rpm, cutting force 500N) of the aluminum piston to be processed, and calculates the target displacement of the front and rear counterweight blocks Δx1=1.2mm and Δx2=0.8mm, and the target phase angles θ1=180° and θ2=180° (the phase angle is opposite to the eccentricity direction, and the deviation is ≤1°). Start the servo drive mechanism (model SGM7J-08AFC6S) to drive the double-plane counterweights (high-density alloy material, 100g each) at the front and rear ends of the spindle to move precisely in the radial direction. The displacement adjustment accuracy is ≤0.01mm and the phase adjustment accuracy is ≤1°. After the movement is completed, keep the counterweights fixed. After static counterweight compensation, the spindle was started under no-load operation, maintaining a speed of 12000 rpm. The effective value of the spindle vibration was detected using a vibration sensor (model ACM1000, measurement accuracy ±0.001mm / s), and the spindle imbalance was detected using a dynamic balancing instrument. The test results showed: The spindle imbalance was reduced to 28 g·mm, and the effective vibration value was 0.045 mm / s, both of which met the static balance accuracy requirements preset by this invention (imbalance ≤ 20 g·mm, effective vibration value ≤ 0.05 mm / s). The static counterweight compensation was qualified, and the process could proceed to the subsequent processing stage.

[0016] S3. Online inspection and dynamic correction: Install the φ90mm aluminum piston to be processed, using the same clamping method as the standard sample, with a clamping pressure of 0.4MPa and a clamping deviation of 0.0014mm, which meets the requirements; Start the processing flow and set the processing parameters: The spindle speed is 12000 rpm, the cutting depth is 0.2 mm, the feed rate is 100 mm / min, the cutting fluid is emulsion (concentration 5%), and the skirt eccentric turning is performed using a carbide tool (model YT15).

[0017] During the processing, operating data is monitored in real time using various sensors: The vibration sensor collects the spindle vibration value in real time, maintaining it within 0.04-0.06 mm / s, and does not exceed the preset threshold of 0.09 mm / s; The temperature sensor (model PT1000, measurement accuracy ±0.1℃) monitors the spindle temperature rise and keeps it below 3℃, never exceeding the 5℃ threshold. The force sensor (model Kistler 9257B) monitors the cutting force in real time, stabilizing at around 500N with fluctuations of ≤±10N. No sudden changes in cutting force were detected (the amplitude of the change was ≥50N), so no secondary dynamic compensation was required. The machining process was smooth, with no abnormal noise or vibration.

[0018] S4. Closed-loop optimization: After the individual aluminum piston skirt is machined, the key dimensions of the piston skirt are automatically measured using an online contact probe (model Renishaw OMP40-2, measurement accuracy ≤0.001mm). The ellipticity is 0.008 mm, the roundness is 0.006 mm, and the surface roughness Ra=0.6 μm, all of which meet the preset processing standards (ellipticity ≤0.01 mm, roundness ≤0.008 mm, Ra≤0.8 μm).

[0019] To verify the accuracy and stability of the method of this invention, 50 aluminum pistons of this specification were continuously processed, and 1 out of every 10 was randomly selected for dimensional inspection, for a total of 5 pieces. The inspection results showed: The ellipticity of the skirts of the five products is between 0.007mm and 0.009mm, the roundness is between 0.005mm and 0.007mm, the surface roughness Ra is between 0.5μm and 0.7μm, and the dimensional deviation of all products is ≤0.002mm, indicating good consistency in batch processing.

[0020] Meanwhile, during continuous processing, the effective value of spindle vibration remained between 0.04 mm / s and 0.055 mm / s, and the spindle bearing temperature remained stable between 35℃ and 38℃ with no abnormal wear, fully verifying the accuracy, stability, and reliability of the method of the present invention.

[0021] Example 2: Eccentric machining of steel piston pin holes This embodiment addresses the scenario of eccentric machining of a φ100mm steel piston pin hole. This steel piston is a special piston for engineering machinery engines, made of 40Cr alloy steel with a hardness of HRC30-32 and a density of 7.85 g / cm³. 3 The pin hole diameter is φ25mm, the pin hole eccentricity is 5mm, and the coaxiality requirement of the pin hole after machining is ≤0.01mm, and the roundness requirement is ≤0.008mm. The spindle eccentricity balance control is performed using the method of this invention, and the specific steps are as follows: S1. Clamping and parameter calibration: Select a standard φ100mm steel piston sample that is completely consistent with the specifications and material of the steel piston to be processed. The sample surface is free of defects and deformation, and the pre-machining accuracy of the pin hole meets the requirements. It is used as the reference part for parameter calibration. A hydraulic chuck was used for clamping. Considering the high hardness of 40Cr alloy steel, the clamping pressure was adjusted to 0.45MPa to ensure a firm clamping without deformation. The clamping deviation was detected in real time using a dial indicator, and the final result was 0.001mm, which meets the requirement of clamping deviation ≤0.002mm. Start the spindle under no-load operation, set the speed to 15000 rpm, and control the speed fluctuation within ±50 rpm. Use a laser displacement sensor to collect the radial runout data of the spindle, with a sampling frequency of 1000 Hz and a collection time of 30 s. After data processing, the average radial runout is 0.03 mm, the maximum value is 0.032 mm, and the minimum value is 0.028 mm.

[0022] Meanwhile, a torque sensor was used to collect the spindle's no-load torque, with an average value of 12 N·m. The average torque under light load (cutting force 180 N) was 15.2 N·m. Combined with the radial runout data, the initial eccentricity e0 = 0.03 mm and the initial unbalance U0 = 180 g·mm were calculated using the least squares method. The calculation error was 3.8%, which meets the error requirement of ≤5%, and the parameter calibration was completed. A correlation model was established, and an orthogonal experimental design was adopted. The spindle speed gradient was set to 8000rpm, 10000rpm, 12000rpm, and 15000rpm, and the cutting force gradient was set to 200N, 400N, 600N, and 800N. Three sets of data were collected for each working condition combination, and the average value was taken as the valid data. A total of 48 sets of eccentricity data were collected. Considering the characteristics of steel piston material, such as high hardness, large cutting load, and easy sudden changes in cutting force during processing, the BP neural network model was optimized by increasing the number of hidden layer neurons to 12 and extending the number of training iterations to 1200 times. This ensures that the model can accurately adapt to dynamic working condition changes, and the model training error converges to within 0.00008.

[0023] Based on the actual requirements of machining the steel piston pin hole in this embodiment, the optimal machining conditions are determined to be a rotational speed of 15000 rpm and a cutting force of 700 N. The corresponding target balance parameters are a target eccentricity ≤ 0.005 mm and a target imbalance ≤ 20 g·mm. The optimized model parameters are then imported into the CNC system of the machining center (model FANUC 0i-MF) to complete the model deployment and ensure that the model response speed is ≤ 10 ms.

[0024] S2. Static counterweight compensation: Start the static counterweight process, the CNC system calls the associated model, inputs the specifications (φ100mm), material (40Cr alloy steel) of the steel piston to be processed and the preset processing parameters (speed 15000rpm, cutting force 700N), and calculates the target displacement of the front and rear counterweight blocks Δx1=1.5mm, Δx2=1.0mm, and the target phase angle θ1=175°, θ2=175°; A servo drive mechanism is used to move the counterweight to the target position with precision. The displacement adjustment accuracy is ≤0.01mm and the phase adjustment accuracy is ≤1°. After the counterweight is completed, the spindle is started under no-load operation at a speed of 15,000 rpm. The spindle imbalance is measured to be 25 g·mm using a dynamic balancing tester, and the effective value of the spindle vibration is measured to be 0.05 mm / s using a vibration sensor. Both of these values ​​meet the static balance accuracy requirements, and the static counterweight is qualified. The spindle can then proceed to the machining stage.

[0025] S3. Online detection and dynamic correction: Install the φ100mm steel piston to be processed, clamping pressure 0.45MPa, clamping deviation 0.0012mm, which meets the requirements; Start the processing flow and set the processing parameters: The cutting depth is 0.3 mm, the feed rate is 80 mm / min, the cutting fluid used is cutting oil (model L-AN46), and the eccentric boring of the pin hole is performed using a carbide tool (model YW2). In the initial stage of processing, due to the high hardness of the steel piston material, the cutting force suddenly changes at the moment the tool contacts the workpiece, with the peak value rising to 750N, exceeding the preset cutting force of 700N. This causes the spindle vibration value to rise rapidly to 0.12mm / s, exceeding the preset threshold of 0.09mm / s, and the system automatically triggers the dynamic correction mechanism. The CNC system uses real-time data from vibration and force sensors, combined with an associated model, to quickly calculate the fine-tuning amount of the counterweight. The displacements of the front and rear counterweights were fine-tuned to 1.6mm and 1.1mm respectively, and the phase was fine-tuned to 174.5°. The fine-tuning amounts were all within the preset range of displacement fine-tuning amount ≤0.01mm and phase fine-tuning amount ≤0.5°. Simultaneously, the servo axes of the CNC system are linked to compensate for the spindle eccentricity of 0.02mm, and the tool cutting trajectory is corrected in real time with a correction response time of 8ms. After the correction was completed, real-time monitoring data showed that the spindle vibration value dropped to 0.048 mm / s, the cutting force stabilized at around 700 N with fluctuations of ≤ ±15 N, the spindle temperature rise was controlled within 4.5℃, the machining process returned to stability, and there were no abnormal working conditions in subsequent machining.

[0026] S4. Closed-loop optimization: After the machining of a single steel piston pin hole is completed, the key dimensions of the pin hole are measured using an online measuring device (model Renishaw OMP40-2, measurement accuracy ≤0.001mm). The coaxiality of the pin hole is 0.009mm and the roundness is 0.007mm, both of which meet the preset processing standards (coaxiality ≤ 0.01mm, roundness ≤ 0.008mm).

[0027] To verify the adaptability of the method of this invention to dynamic loads, 30 steel pistons of this specification were continuously processed, and one out of every five was randomly selected for testing, for a total of six pistons. The test results showed: The coaxiality of the pin holes of the six products is between 0.008 and 0.01 mm, the roundness is between 0.006 and 0.008 mm, and the dimensional deviation fluctuation is ≤0.002 mm.

[0028] Meanwhile, during continuous machining, the spindle vibration value is maintained at 0.045-0.055 mm / s, the spindle bearing temperature is stable at 38-41℃, there is no abnormal wear, and the system can quickly trigger dynamic correction when the cutting force changes suddenly. After correction, the working condition is stable, which fully verifies the adaptability, response speed and stability of the method of the present invention to dynamic loads, and can meet the eccentric machining requirements of pistons with high hardness materials.

[0029] Comparative Example 1: Processing a φ90mm aluminum piston skirt using traditional manual static counterweight method To compare the effects of the method of this invention with the traditional balancing method, the eccentric machining of the same specification φ90mm aluminum piston skirt was performed using the traditional manual static counterweight method. The machining conditions, equipment, and workpiece material were completely consistent with those in Example 1, except for the balancing method. The specific steps are as follows: Clamping: The standard φ90mm aluminum piston sample and the piston to be processed are clamped using the same hydraulic chuck as in Example 1. The clamping pressure is 0.4MPa, and the clamping deviation is controlled between 0.001-0.0015mm, which is consistent with the clamping accuracy in Example 1.

[0030] Manual static counterweighting: By manually observing the spindle vibration, a fixed counterweight is manually installed near the chuck at the front end of the spindle. The counterweight is made of ordinary cast iron and weighs 100g. By repeatedly adjusting the position and number of the counterweight, the effective value of the vibration of the spindle under no-load (12000rpm) state is reduced to about 0.08mm / s and the imbalance is reduced to 80g·mm. The manual static counterweighting is completed. The entire counterweighting process takes about 25 minutes and requires repeated manual stops and adjustments.

[0031] Machining: The skirt was eccentrically turned using the same machining parameters as in Example 1 (spindle speed 12000rpm, depth of cut 0.2mm, feed rate 100mm / min, cutting fluid emulsion concentration 5%, tool type YT15). No dynamic detection or correction was performed during the machining process, and the balance was maintained solely by the initial manual counterweight.

[0032] Testing and Verification: Fifty aluminum pistons of this specification were continuously machined. One out of every ten was randomly selected for dimensional inspection. Simultaneously, spindle vibration, bearing temperature, and machining efficiency were monitored during the machining process. The test results are as follows: (1) Machining accuracy: The ellipticity of the skirt of the 5 sampled products was between 0.012mm and 0.018mm, the roundness was between 0.009mm and 0.015mm, and the surface roughness Ra was between 0.9μm and 1.3μm. All of these exceeded the machining standards of Example 1, and the dimensional deviation fluctuated by 0.006mm to 0.008mm, indicating poor batch consistency. (2) Operational stability: During the processing, the effective value of the spindle vibration was maintained between 0.07 mm / s and 0.11 mm / s, and repeatedly approached or exceeded the critical threshold of 0.09 mm / s. The spindle bearing temperature was stable at 38℃-42℃, which is 3℃-4℃ higher than that of Example 1, and the bearing wear rate was accelerated. (3) Processing efficiency: It takes 25 minutes to manually counterweight once. During the process, the machine needs to be stopped and the counterweight needs to be readjusted 3 times due to abnormal vibration. The total downtime is 40 minutes. The total processing time for 50 products is 65 minutes longer than that of Example 1, and the processing efficiency is reduced by about 35%.

[0033] The comparison results show that the traditional manual static counterweight method cannot achieve precise balance. Its processing accuracy, operational stability and processing efficiency are far lower than those of the method of this invention. In addition, the manual labor intensity is high and it cannot meet the needs of batch high-precision processing.

[0034] Comparative Example 2: Machining a φ100mm steel piston pin hole using the traditional general dynamic balancing method To further verify the superiority of the method of the present invention, a traditional general dynamic balancing method (not optimized for piston machining scenarios, without BP neural network modeling, and only using a single-plane dynamic counterweight) was used to perform eccentric machining on the same specification φ100mm steel piston pin hole. The machining conditions, equipment, and workpiece material were completely consistent with those in Example 2. The specific steps are as follows: Clamping: The standard φ100mm steel piston sample and the piston to be processed are clamped using the same hydraulic chuck as in Example 2. The clamping pressure is 0.45MPa, and the clamping deviation is controlled between 0.001mm and 0.0012mm, which is consistent with the clamping accuracy in Example 2.

[0035] Universal dynamic counterweight: A single-plane dynamic counterweight method is adopted, with a counterweight block placed only at the front end of the spindle near the chuck. The vibration of the spindle is monitored by a vibration sensor, and the counterweight parameters are calculated using a universal dynamic balance algorithm. There is no need to establish an associated model or consider the influence of piston material and specifications on the balance parameters. After counterweighting, the effective vibration value of the spindle under no-load (15000rpm) state is 0.07mm / s, the unbalance is 60g·mm, and the counterweighting takes about 18 minutes.

[0036] Machining: The same machining parameters as in Example 2 (spindle speed 15000rpm, depth of cut 0.3mm, feed rate 80mm / min, cutting fluid L-AN46 cutting oil, tool type YW2) were used for eccentric boring of the pin hole. During the machining process, only the vibration value was monitored. When the vibration value exceeded 0.12mm / s, the counterweight adjustment was manually triggered. There was no automatic dynamic correction function.

[0037] Testing and Verification: Thirty steel pistons of this specification were continuously machined. One out of every five pistons was randomly selected for dimensional inspection. Simultaneously, spindle vibration, bearing temperature, dynamic response speed, and machining efficiency were monitored during the machining process. The test results are as follows: (1) Machining accuracy: The coaxiality of the pin holes of the 6 sampled products was between 0.011mm and 0.016mm, and the roundness was between 0.009mm and 0.014mm. All of them exceeded the machining standards of Example 2, and the dimensional deviation fluctuated between 0.005mm and 0.007mm, indicating poor batch consistency. (2) Operational stability: When the cutting force suddenly changes during the initial stage of processing, the spindle vibration value rises rapidly to 0.14 mm / s. Since there is no automatic dynamic correction function, manual shutdown is required to adjust the counterweight. Each adjustment takes about 8 minutes, and a total of 4 shutdowns are required for adjustment. During the processing, the effective value of spindle vibration is maintained between 0.065 mm / s and 0.13 mm / s, and the spindle bearing temperature is stable at 42℃-45℃, which is 4℃-6℃ higher than that of Example 2, and the bearing wear is severe. (3) Dynamic response and efficiency: The response to sudden changes in cutting force is entirely dependent on manual operation, with a response time of ≥8 minutes, which is much longer than the 8ms in Example 2. The total processing time for 30 products increased by 32 minutes compared to Example 2, and the processing efficiency decreased by about 30%. (4) Adaptability: Due to the lack of optimization of the balancing strategy for the high hardness and large cutting load of the steel piston material, the counterweight failure occurred multiple times during the processing, making it unable to stably adapt to the processing conditions of high speed and large load.

[0038] The comparative results show that the traditional general dynamic balancing method is not suitable for piston machining scenarios. It has slow dynamic response, low balancing accuracy, and poor adaptability, and cannot meet the accuracy and efficiency requirements of high-hardness, high-speed piston eccentric machining. However, the method of this invention has significant advantages in all aspects through targeted design.

[0039] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it.

Claims

1. A method for balancing the eccentricity of the spindle in a piston machining center, characterized in that, Includes the following steps: S1. Clamping and Parameter Calibration: A standard piston sample with the same specifications and material as the piston to be processed is clamped. A hydraulic chuck is used to complete the positioning and clamping to ensure that the clamping deviation is ≤0.002mm. The radial runout, no-load and light-load torque data of the spindle are collected by laser displacement sensor and torque sensor. The initial eccentricity e0 and the initial imbalance U0 are calculated by combining the radial runout data. The BP neural network algorithm is used to establish a correlation model between the eccentricity e and the spindle speed n and the cutting load F to determine the target balance parameters under different working conditions. S2. Static counterweight compensation: Based on the calibrated target balance parameters, drive the double-plane counterweight blocks located at the front and rear ends of the spindle to move to the target position to complete the static counterweight, so that the unbalance of the spindle under no-load condition is ≤20g·mm and the effective vibration value is ≤0.05mm / s. S3. Online detection and dynamic correction: During the machining process, vibration, temperature and force data of the spindle vibration, temperature rise and cutting force are collected in real time by vibration sensor, temperature sensor and force sensor. When the vibration value is ≥0.09mm / s or the temperature rise is ≥5℃ or the cutting force change amplitude is ≥50N, the position and phase of the counterweight are automatically adjusted and the CNC system is linked to compensate for the spindle eccentricity and correct the cutting trajectory. S4. Closed-loop optimization: After processing, the key dimensions of the piston are automatically measured. Based on the dimensional deviation, the counterweight parameters in the correlation model are optimized in reverse, the target balance parameters are updated, and closed-loop calibration is achieved.

2. The method for eccentric balancing of the spindle of a piston machining center according to claim 1, characterized in that, The initial eccentricity e0 and the initial imbalance U0 are calculated using the least squares method, with a calculation error of ≤5%. The target balance parameters include target eccentricity ≤0.005mm and target imbalance ≤20g·mm.

3. The method for balancing the eccentricity of the spindle in a piston machining center according to claim 1, characterized in that, The dual-plane counterweights are located at the front end of the spindle near the chuck and at the rear end near the bearing, respectively, enabling radial sliding and synchronous phase adjustment. The displacement accuracy is ≤0.01mm and the phase adjustment accuracy is ≤1°, and it can be adapted to a speed range of 8000rpm-15000rpm.

4. The method for eccentric balancing of the spindle of a piston machining center according to claim 1, characterized in that, The laser displacement sensor is KEYENCE IL-600 with a measurement accuracy of ≤0.001mm; the vibration sensor is ACM1000 with a measurement accuracy of ±0.001mm / s; the temperature sensor is PT1000 with a measurement accuracy of ±0.1℃; and the torque sensor is HBM T40B with a measurement range of 0-50N·m and an accuracy of ±0.1N·m. The data acquisition frequency is 1000Hz, and the cutting force data acquisition interval is 10ms.

5. The method for eccentric balancing of the spindle of a piston machining center according to claim 1, characterized in that, The compensation accuracy of the spindle eccentricity compensation of the CNC system linkage servo axis is ≤0.001mm. During dynamic correction, the displacement fine adjustment of the counterweight block is ≤0.01mm, the phase fine adjustment is ≤0.5°, the spindle vibration value after correction is ≤0.05mm / s, and the spindle temperature rise is ≤5℃.

6. The method for eccentric balancing of the spindle of a piston machining center according to claim 1, characterized in that, The key dimensions of the piston measured include skirt ellipticity, pin hole coaxiality, and roundness. The associated model can update parameters in real time according to the machining conditions.

7. The method for eccentric balancing of the spindle of a piston machining center according to claim 1, characterized in that, The counterweights are made of high-density alloy material, and the mass of a single counterweight can be adjusted from 50g to 200g. The combination of counterweights can be flexibly matched according to the specifications and mass of the piston to be processed.