A barrel deep hole numerical control machining self-adaptive chatter suppression method

Abstract

This invention relates to the field of alloy metal processing technology, and particularly to an adaptive chatter suppression method for deep hole CNC machining of tubes. The method comprises the following steps: S1. Collecting and statistically analyzing chatter data from the CNC machine tool involved in tube machining; S2. Performing data analysis, cleaning, and establishing a data ledger, classifying the data, and generating a leaf-shaped diagram; S3. Data processing and self-learning, transmitting the data to a cloud-based process library, classifying the tubes, and optimizing the CNC machining parameters for different tubes; S4. Acquiring and analyzing machining data in real time, performing FFT spectrum analysis during tube machining to analyze and warn of chatter precursors; S5. Two-way data interaction, exchanging cloud data with the CNC system PLC; S6. The CNC machine tool adaptively adjusts various parameters during tube machining, issuing warnings before chatter fully occurs through real-time FFT spectrum analysis, and promptly suppressing chatter through adaptive parameter adjustment.

Description

Technical Field

[0001] This invention relates to the field of alloy metal processing technology, and in particular to an adaptive chatter suppression method for CNC machining of deep holes in tubes. Background Technology

[0002] As a core component of artillery, the quality of deep-hole machining of the barrel directly determines the weapon's firing accuracy, range, and service life. The barrel is a typical slender rod-like part, with a very high length-to-diameter ratio during deep-hole machining. During CNC machining, it is highly susceptible to chatter due to various factors such as cutting forces, spindle vibration, and tool wear.

[0003] In traditional deep hole CNC machining of tubes, the machining parameters of CNC machine tools (such as spindle speed S, feed rate F, radial depth of cut, axial depth of cut, etc.) are mostly preset manually. During the machining process, the parameters cannot be adaptively adjusted according to the generation and development of chatter, and manual intervention can only be carried out by the operator's experience.

[0004] On the one hand, manual intervention is delayed, and measures can only be taken after flutter has already occurred and caused processing defects, making it impossible to achieve early warning and real-time suppression of flutter; On the other hand, the experience of operators varies greatly, making it difficult to guarantee the accuracy and timeliness of parameter adjustments. This makes it impossible to adapt to the processing needs of tubes of different specifications and materials, and also unable to cope with the dynamic changes in cutting load and vibration state during processing, resulting in poor chatter suppression. Summary of the Invention

[0005] The purpose of this invention is to solve the problem that chatter cannot be suppressed in the prior art, and that early warning and active intervention of chatter cannot be achieved. Therefore, this invention proposes an adaptive chatter suppression method for deep hole CNC machining of tubes.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: An adaptive chatter suppression method for CNC machining of deep holes in tubes is designed, and the specific steps are as follows: S1. Collect and analyze the chatter data of CNC machine tools involved in tube machining. S2. Perform data analysis and cleaning, establish a data ledger, classify the data, and generate a leaf diagram; S3, data processing and self-learning, transmits data to the cloud process library, classifies tubes, and performs optimal matching of CNC machining parameters for different tubes; S4. Processing data acquisition and real-time analysis: Perform FFT spectrum analysis in real time during the tube processing to analyze and warp precursors. S5, bidirectional data interaction, enables interaction between cloud data and the CNC system PLC; S6. During the machining process of the CNC machine tool, various parameters are adaptively adjusted to suppress chatter.

[0007] Preferably, S1 specifically includes the following data: X / Y / Z spindle vibration acceleration, vibration frequency, amplitude, machine tool operating data, spindle speed S, spindle load torque, servo current, feed speed F, radial depth of cut, and axial depth of cut.

[0008] Preferably, in step S2, data analysis and cleaning are performed to remove outliers, and the maximum and minimum values ​​of the data in step S1 are removed.

[0009] Preferably, the S3 data processing and self-learning establishes a stability prediction model for the stable leaflet diagram of the tube processing, with a stability boundary (a p =ƒ(spindle speed S, workpiece natural frequency ƒ) n Damping ζ).

[0010] Preferably, the adaptive method in S6 is based on the FFT spectrum flutter determination algorithm.

[0011] Preferably, the flutter determination algorithm for the FFT spectrum is a flutter determination rule.

[0012] Preferably, the flutter determination rule is as follows: A max This represents the maximum amplitude of the spectrum. A th The security threshold for historical data statistics in the cloud; ƒ n This is the inherent frequency of the barrel-tool system; Judgment conditions: If A max >A th ; And the spectral peak appears in ƒ n Within ±5Hz; Furthermore, there are obvious second and third harmonic components; If the conditions are met, it is determined that regenerative flutter or modal coupling flutter is triggered.

[0013] Preferably, the chatter determination result is transmitted to the CNC machine tool, and the CNC machine tool adjusts and adapts the tube machining parameters.

[0014] The adaptive chatter suppression method for deep hole CNC machining of tubes proposed in this invention has the following advantages: 1. Achieve early warning and real-time suppression of chatter: Through real-time FFT spectrum analysis, an early warning can be issued before chatter fully occurs, and chatter can be suppressed in a timely manner through adaptive parameter adjustment, which solves the problem of lag in chatter suppression in traditional methods and reduces the scrap rate of workpieces.

[0015] 2. Achieve adaptive matching of processing parameters: Combined with the self-learning function of the cloud-based process library, it can automatically match the optimal processing parameters according to the different specifications and materials of the tubes, adapt to the processing needs of various tubes, and improve processing consistency and stability.

[0016] 3. High accuracy in flutter detection: The system adopts a multi-index comprehensive judgment rule (maximum amplitude of spectrum, peak frequency, harmonic components), which effectively avoids misjudgment and omission caused by single index judgment, provides a reliable basis for parameter adaptive adjustment, and improves the accuracy of flutter suppression.

[0017] 4. Achieve closed-loop data optimization: Through bidirectional data interaction between the cloud-based process library and the CNC machine tool, the machining data and optimal parameter library are updated in real time, enabling self-learning and iteration. As machining data accumulates, the chatter warning accuracy and parameter matching accuracy are continuously improved, resulting in better long-term performance. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the adaptive chatter suppression method for deep hole CNC machining of tubes according to the present invention. Detailed Implementation

[0019] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0020] like Figure 1 As shown, this embodiment provides an adaptive chatter suppression method for CNC machining of deep holes in tubes. The equipment used is a CNC machine tool equipped with X / Y / Z three-axis vibration sensors, a spindle speed sensor, and a load torque sensor. The cloud-based process library is built using the Alibaba Cloud Industrial Internet Platform. The specific implementation steps are as follows: S1. Collection of historical data and daily processed data: During CNC machining, the following data are collected in real time using various sensors mounted on the CNC machine tool: X / Y / Z spindle vibration acceleration (unit: m / s²), vibration frequency (unit: Hz), amplitude (unit: μm), spindle speed S (unit: r / min), spindle load torque (unit: N·m), servo current (unit: A), feed rate F (unit: mm / min), radial depth of cut (unit: mm), and axial depth of cut (unit: mm). Historical machining data and current machining process data are collected continuously to complete the statistics and collection of chatter-related data.

[0021] S2. Data analysis, cleaning, and data ledger establishment: The data collected by S1 was cleaned using Python data processing tools. Outliers were removed using the 3σ criterion, and the maximum and minimum values ​​in each data set were also removed to avoid the impact of extreme data on subsequent analysis. After cleaning, the data were classified according to the spindle machining process, and a data ledger was established to record the vibration data and machine tool operating parameters under different processes. Based on the classified data, chatter lobe diagrams were plotted using MATLAB software to clearly show the stable and unstable chatter regions under different spindle speeds S and radial cutting depths. The chatter amplitude in the stable region was ≤4μm, and the chatter amplitude in the unstable region was >4μm.

[0022] S3, Data Processing and Self-Learning: The data after cleaning and sorting in S2 is transmitted to the cloud-based process library. The cloud-based process library archives data according to tube specifications and materials. Combined with historical processing data, a stability prediction model for the tube processing lobe diagram is established. The stability boundary (a) p =ƒ(spindle speed S, workpiece natural frequency ƒ) n Damping ζ), where the workpiece's natural frequency ƒ n The modal test yielded a frequency of 120 Hz and a damping ζ = 0.03. This model was used to optimally match the processing parameters for this type of tube, forming an optimal processing parameter library. Some optimal parameters are as follows: Drilling process: S = 800 r / min, F = 120 mm / min, a p =0.3mm, a e =0.5mm; Boring operation: S=600r / min, F=80mm / min, a p =0.2mm, a e =0.4mm; Reaming process: S=400r / min, F=60mm / min, a p =0.1mm, a e =0.3mm.

[0023] S4. Processing data acquisition and real-time analysis: During deep hole CNC machining, sensors collect real-time data on X / Y / Z spindle vibration acceleration, frequency, amplitude, and machine tool operating parameters. The collected vibration signals are transmitted in real-time to the data analysis module for FFT spectrum analysis. The cloud-based process library presets a chatter safety threshold A for this type of tube. th =5μm, workpiece natural frequency ƒ n =120Hz, the data analysis module combines FFT spectrum results with cloud data to perform real-time flutter precursor analysis. When the spectrum amplitude is detected to be close to A th When the spectral peak value is close to 115Hz (≥3.5μm), a flutter warning is issued.

[0024] S5, Two-way data interaction: The cloud-based process library will store the optimal machining parameters for the tube and the chatter safety threshold A. th The natural frequency of the workpiece ƒ n Data such as these are transmitted in real time to the PLC control system of the CNC machine tool. At the same time, the CNC machine tool transmits the real-time collected machining data and vibration data back to the cloud process library. The cloud process library updates the optimal parameter library and stability prediction model based on the new data, realizing real-time data synchronization and self-learning iteration.

[0025] S6. Adaptive parameter adjustment and flutter suppression: During processing, the data analysis module executes the FFT spectrum flutter detection algorithm in real time, and determines flutter according to the following rules: If A max >4μm (A) th If the frequency peak appears at 115Hz±5Hz (110Hz~120Hz) and there are obvious harmonic components of the second harmonic (around 240Hz) and the third harmonic (around 360Hz), then flutter is determined to be triggered. When a precursor to chatter is detected in a tube during the boring process (A) max =3.8μm, peak frequency =113Hz, and when harmonic components of 2nd harmonic 226Hz and 3rd harmonic 334Hz appear, the system issues an early warning. The PLC control system immediately receives the chatter judgment result and adaptively adjusts the machining parameters according to the optimal parameters matched by the cloud-based process library: the spindle speed S is adjusted from 600r / min to 580r / min, the feed rate F is adjusted from 80mm / min to 75mm / min, and the radial depth of cut a p The axial cutting depth a was adjusted from 0.2mm to 0.18mm. e Keep the diameter constant at 0.4 mm; after adjustment, monitor vibration data in real time. A max The wavelength rapidly decreased to 3.2 μm, the spectral peak deviated from 115 Hz, the second and third harmonic components disappeared, the flutter was effectively suppressed, and the processing returned to stability.

[0026] This embodiment takes the semi-precision boring deep hole machining process of a certain type of 155mm howitzer barrel as an example. The barrel blank is made of 30CrNi2MoV high-strength alloy structural steel, with a blank inner diameter of φ120mm, a finished semi-precision boring inner diameter of φ152mm, a total barrel length of 4200mm, and an aspect ratio of L / D=35. It is machined using a CK61160 CNC deep hole boring machine. A three-axis piezoelectric vibration sensor (sampling frequency 10kHz, measurement accuracy ±0.1m / s²), a spindle photoelectric encoder (speed measurement accuracy ±1r / min), and a spindle torque sensor (measurement accuracy ±0.1N) are installed at the front end of the machine tool spindle. The cloud-based process library, built on the Alibaba Cloud Industrial Internet platform, has pre-stored over 1,560 sets of historical processing data for tubes of this material and specification.

[0027] Before machining, the cloud-based process library automatically matches the optimal initial machining parameters based on the material, specifications, machining process, and tool type (carbide indexable deep hole boring tool): spindle speed S=600r / min, feed rate F=80mm / min, radial depth of cut ap=0.2mm, axial depth of cut ae=0.4mm; at the same time, the chatter safety threshold Ath=5μm, the natural frequency of the tube-tool system ƒn=120Hz, and the system damping ζ=0.03 of this model of tube are simultaneously sent to the machine tool PLC control system.

[0028] After machining begins, various sensors collect real-time data at a frequency of 10kHz, including X / Y / Z axis spindle vibration acceleration, vibration frequency, amplitude, spindle load torque, servo current, and actual feed rate. The data undergoes preliminary filtering and preprocessing via the machine tool edge computing gateway before being synchronously transmitted to the local real-time data analysis module and the cloud-based process library. The local data analysis module performs a Fast Fourier Transform (FFT) spectrum analysis every 100ms to extract the amplitude, frequency, and harmonic characteristics of the vibration signal in real time.

[0029] When machining reached a depth of 2250mm along the tube axis, the data analysis module detected a significant anomaly in the vibration signal: the maximum amplitude of the spectrum, Amax, rapidly increased to 3.8μm (exceeding the preset warning threshold of 3.5μm), and the main peak of the spectrum appeared at 113Hz (within the 110~120Hz range of the natural frequency ƒn±5Hz). Simultaneously, clear second and third harmonic components appeared at 226Hz and 334Hz, respectively, fully meeting the criteria for regenerative chatter precursors. The system immediately issued a yellow warning signal to the machine tool PLC control system and the cloud-based process database, and recorded all current machining parameters and vibration data.

[0030] After receiving the warning signal, the PLC control system starts the adaptive parameter adjustment program within 200ms. Based on the optimal parameter range of the stable lobe diagram under the current working conditions pushed in real time by the cloud process library, the machining parameters are adjusted linearly without stopping: the spindle speed S is uniformly reduced from 600r / min to 580r / min, the feed rate F is reduced from 80mm / min to 75mm / min, the radial depth of cut ap is finely adjusted from 0.2mm to 0.18mm, and the axial depth of cut ae remains unchanged at 0.4mm.

[0031] After parameter adjustments, the system continuously monitored vibration changes: 500ms after adjustment, the maximum amplitude of the spectrum, Amax, dropped to 3.2μm; 1.2s after adjustment, the main peak of the spectrum shifted to 97Hz, and the harmonic components at 226Hz and 334Hz completely disappeared. The vibration signal returned to a stable processing state, and the system automatically lifted the warning. During subsequent processing up to the full length of 4200mm, no further chatter precursors or chatter phenomena occurred.

[0032] After machining, the inner bore of the gun barrel underwent quality inspection: the surface roughness Ra=0.76μm, roundness error ≤0.012mm, straightness error ≤0.018mm / m, and there were no machining defects such as chatter marks or vibration marks, fully meeting the quality requirements of the semi-precision boring process for artillery barrels. Simultaneously, the machine tool transmitted the complete machining data (including chatter precursor characteristic data, parameter adjustment process data, and adjusted stable machining data) back to the cloud-based process library. The cloud-based process library automatically updated the stable lobe diagram prediction model and optimal parameter library for this type of gun barrel, completing a data closed-loop self-learning iteration. After verification through machining 12 subsequent batches of gun barrels of the same specification, the updated model's chatter warning accuracy improved to 98.7%, the average response time for parameter adjustment was shortened to less than 160ms, and the workpiece scrap rate was reduced by 92% compared to traditional manual intervention methods.

[0033] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. An adaptive chatter suppression method for CNC machining of deep holes in tubes, characterized in that, The specific steps are as follows: S1. Collect and analyze the chatter data of CNC machine tools involved in tube machining. S2. Perform data analysis and cleaning, establish a data ledger, classify the data, and generate a leaf diagram; S3, data processing and self-learning, transmits data to the cloud process library, classifies tubes, and performs optimal matching of CNC machining parameters for different tubes; S4. Processing data acquisition and real-time analysis: Perform FFT spectrum analysis in real time during the tube processing to analyze and warp precursors. S5, bidirectional data interaction, enables interaction between cloud data and the CNC system PLC; S6. During the machining process of the CNC machine tool, various parameters are adaptively adjusted to suppress chatter.

2. The adaptive chatter suppression method for deep hole CNC machining of tubes according to claim 1, characterized in that, Specifically, S1 includes the following data: X / Y / Z spindle vibration acceleration, vibration frequency, amplitude, machine tool operation data, spindle speed S, spindle load torque, servo current, feed speed F, radial depth of cut, and axial depth of cut.

3. The adaptive chatter suppression method for deep hole CNC machining of tubes according to claim 1, characterized in that, In step S2, data analysis and cleaning are performed to remove outliers, and the maximum and minimum values ​​of the data in step S1 are removed.

4. The adaptive chatter suppression method for deep hole CNC machining of tubes according to claim 1, characterized in that, The S3 data processing and self-learning establish a stability prediction model for the stable leaflet diagram of the tube processing, with a stability boundary (a p =ƒ(spindle speed S, workpiece natural frequency ƒ) n Damping ζ).

5. The adaptive chatter suppression method for deep hole CNC machining of tubes according to claim 1, characterized in that, The adaptive method in S6 is a flutter determination algorithm based on the FFT spectrum.

6. The adaptive chatter suppression method for deep hole CNC machining of tubes according to claim 5, characterized in that, The flutter determination algorithm for the FFT spectrum is a flutter determination rule.

7. The adaptive chatter suppression method for deep hole CNC machining of tubes according to claim 6, characterized in that, The flutter determination rule is as follows: A max This represents the maximum amplitude of the spectrum. A th The security threshold for historical data statistics in the cloud; ƒ n This is the inherent frequency of the barrel-tool system; Judgment conditions: Young A max >A th ; And the spectral peak appears in ƒ n Within ±5Hz; Furthermore, there are obvious second and third harmonic components; If the conditions are met, it is determined that regenerative flutter or modal coupling flutter is triggered.

8. The adaptive chatter suppression method for deep hole CNC machining of tubes according to claim 7, characterized in that, The chatter determination result is transmitted to the CNC machine tool, which then adjusts and adapts the tube machining parameters.

Images