Intelligent ultrasonic machining system and method for machining brittle and hard materials

Abstract

The invention discloses an intelligent ultrasonic processing system and method for processing brittle and hard materials, the system comprises a physical entity layer, a network connection layer, a cloud platform and data layer and an application service layer, the physical entity layer comprises an ultrasonic processing machine tool body, an ultrasonic main shaft module, a cooling device and a sensor module; the network connection layer is used for realizing data intercommunication and interconnection among the physical entity layer, the cloud platform and data layer and the application service layer, and the cloud platform and data layer is used for providing data processing, analysis and storage; the application service layer is used for displaying data collected in real time, predicting and optimizing technological parameters, controlling a tool nose point of a tool to move along a preset track, automatically adjusting cutting parameters and predicting fault occurrence time when performance deviation is detected, and generating maintenance early warning. The machining efficiency, the product quality and the equipment reliability can be improved, and meanwhile the production cost and dependence on experience of operators are reduced.

Description

Technical Field

[0001] This invention belongs to the field of precision machining technology, specifically relating to an intelligent ultrasonic machining system and method for machining hard, brittle, and difficult-to-machine materials (such as reinforced glass, ceramics, sapphire, carbides, ceramic matrix composites, etc.). Background Technology

[0002] Brittle and hard materials, due to their superior properties such as high strength, high hardness, wear resistance, corrosion resistance, and insulation, have become key basic materials in high-end manufacturing. However, their "hard and brittle" characteristics also make them typical difficult-to-machine materials: traditional metal cutting techniques easily lead to severe tool wear, workpiece chipping or surface damage, resulting in high processing costs and long cycles; non-traditional processes such as electrochemical machining (ECM) and electrical discharge machining (EDM) rely on the conductivity of the workpiece and cannot be adapted to insulating brittle and hard materials such as ceramics and sapphire; processing methods such as electron beam and laser have limitations such as large equipment investment, low processing efficiency, and harsh environmental requirements, making it difficult to meet the needs of large-scale production.

[0003] Rotary ultrasonic machining (RUM), as an advanced process combining diamond grinding and ultrasonic vibration, has become a mainstream technology for machining brittle and hard materials due to its advantages of high material removal rate (MRR), low tool wear, and excellent machining accuracy. However, existing rotary ultrasonic machining technologies still have the following core challenges: 1. Difficulty in matching process parameters: The RUM process involves multiple sets of coupled input variables such as ultrasonic amplitude, frequency, spindle speed, and feed rate. The relationship between these variables and output variables such as material removal rate, surface roughness, and tool wear is complex. Currently, parameter determination mainly relies on operator experience and trial and error, making it difficult to achieve optimal configuration, resulting in large fluctuations in processing efficiency and quality. 2. Outdated equipment maintenance model: Existing equipment mostly adopts "preventive maintenance" or "reactive maintenance" - the former is prone to ineffective downtime and excessive replacement of parts, while the latter causes production interruption due to sudden failures, resulting in high maintenance costs and poor reliability; 3. Insufficient tool motion accuracy: In five-axis machining, the motion of the rotary axis will generate nonlinear errors. Traditional control systems lack effective compensation mechanisms. Especially when machining complex curved surfaces, the actual tool trajectory deviates greatly from the preset trajectory, affecting the dimensional accuracy and surface integrity of the workpiece. 4. Low level of intelligence: It lacks real-time monitoring, data-driven parameter optimization and adaptive adjustment capabilities, and the processing relies on manual intervention, making it difficult to achieve unmanned and standardized production.

[0004] Therefore, developing an intelligent ultrasonic processing system that can achieve "intelligent parameter optimization, precise motion control, and predictive equipment maintenance" has become an urgent need in the field of high-end processing of brittle and hard materials. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of existing rotary ultrasonic processing technology and provide an intelligent ultrasonic processing system and method for processing brittle and hard materials. Through the collaborative design of a four-layer architecture of "physical entity - network connection - cloud platform - application service", the processing process can be precisely controlled, dynamically optimized and intelligently managed, ultimately improving the processing efficiency, product quality and equipment reliability of brittle and hard materials, while reducing production costs and dependence on operator experience.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows: A smart ultrasonic processing system for machining brittle and hard materials, comprising a physical entity layer, a network connection layer, a cloud platform and data layer, and an application service layer, characterized in that: The physical entity layer includes the ultrasonic machining tool body, the ultrasonic spindle module, the cooling device, and the sensor module; The ultrasonic processing machine tool body includes a machine tool bed, linear guide rails, a feed device, and a power supply; The ultrasonic spindle module includes a spindle motor, a spindle, an ultrasonic transmitter, an ultrasonic transducer, an ultrasonic generator, an ultrasonic amplitude transformer, a tool holder, a tool, an ultrasonic receiver, and a servo module. The power supply provides power to the spindle motor and the ultrasonic generator. The servo module includes a servo and a servo driver for driving the spindle motor along a preset axis. The ultrasonic transmitter and ultrasonic transducer are housed within the spindle housing. The ultrasonic receiver is located outside the spindle housing, above the tool holder of the tool holder. A gap is provided between the receiver coil of the ultrasonic receiver and the transmitter coil of the ultrasonic transmitter to achieve non-contact energy and signal transmission. The ultrasonic transmitter, the ultrasonic transducer, the ultrasonic amplitude transformer, the tool holder, and the tool are sequentially installed at the end of the spindle. The ultrasonic transducer converts the high-frequency electrical energy generated by the ultrasonic generator into mechanical longitudinal vibration. The ultrasonic amplitude transformer amplifies the vibration of the ultrasonic transducer and transmits it to the tool holder and the held tool, causing the ultrasonic spindle module to operate in a resonant state, forming a standing wave. The cooling device includes a coolant reservoir, a coolant treatment tank, a high-pressure pump, a solenoid valve, a flow meter, and a pressure regulator, which are used to supply coolant to the machining area and the spindle, and the coolant is sprayed into the machining area through the inner hole of the tool. The sensor module includes a sensor and a data acquisition module, which are installed on the machine tool body and are used to collect variable data during the processing. The network connection layer is used to realize data interconnection between the physical entity layer, the cloud platform and data layer and the application service layer. The data acquisition module transmits data to the cloud platform and data layer and the application service layer through the network connection layer. The cloud platform and data layer include a cloud server, a database, and a data processing and analysis module. The cloud server provides computing and storage resources; the database stores and manages historical and real-time data collected by the data acquisition module; and the data processing and analysis module cleans, preprocesses, and extracts features from the collected real-time data. The application service layer includes a real-time monitoring module, a parameter optimization module, a control module, an adaptive control module, and a predictive maintenance module. The real-time monitoring module displays real-time data collected by the data acquisition module in a visual format and identifies anomalies and faults based on the analysis of the predictive maintenance module, issuing alarms. The parameter optimization module, based on historical data in the database, applies machine learning algorithms to establish a predictive model between input variable data and output performance variable data, aiming to maximize the material removal rate (MRR) while satisfying surface roughness and tool wear constraints. It outputs an optimized set of process parameters and sends it to the control module. The control module is used to automatically calculate and compensate for nonlinear errors caused by spindle rotation based on the pre-set workpiece contour, desired tool angle, and process parameter set, control the tool tip to move along a predetermined trajectory, and maintain a constant angle between the tool and the machined surface; the adaptive control module is used to evaluate the machining status in real time based on real-time data and predictive models, and automatically adjust the cutting parameters to eliminate the disturbance when performance deviates from expectations; the predictive maintenance module is used to analyze whether there is a performance degradation trend in the physical entity layer components based on the collected real-time data and machine learning algorithms, predict the time of potential failure and component life, and generate maintenance warnings.

[0007] Furthermore, the application service layer can be deployed as microservices on a cloud platform and data layer or edge server; the control module can also automatically compensate for changes in tool length; the network connection layer consists of industrial IoT infrastructure; and coolant is sprayed into the machining area through the inner hole of the tool.

[0008] Furthermore, the sensors include a cutting force sensor for measuring three-dimensional cutting forces, a vibration sensor, a spindle speed encoder for feeding back spindle speed and position, a coolant flow sensor, a coolant pressure sensor, and an ultrasonic amplitude and frequency monitoring sensor; wherein, the coolant flow sensor and the coolant pressure sensor are used to monitor the state of the coolant, and the vibration sensor is an accelerometer used to monitor the state of the worktable during material removal.

[0009] Furthermore, the control module is a programmable logic controller.

[0010] Furthermore, the variable data includes spindle speed, feed rate, and ultrasonic transducer power; the performance variable data includes material removal rate (MRR), surface roughness, and tool wear.

[0011] Furthermore, the prediction model is a neural network model; the machine learning algorithm includes neural networks, random forests, and / or support vector machines.

[0012] Furthermore, the real-time monitoring module can display processing status parameters, including ultrasonic transducer power, spindle vibration, spindle speed and / or feed rate.

[0013] This invention also discloses an ultrasonic machining method with tip tracking (TPT) function, the method comprising the following steps: S1, uses a prediction model trained by machine learning algorithms to output a set of process parameters for hard and brittle materials; S2, calculate and compensate for the nonlinear error caused by the spindle rotation motion based on the pre-set workpiece contour, the desired tool angle and the set of process parameters, control the tool tip to move along the predetermined trajectory and keep the angle between the tool and the machining surface constant. S3 collects sensor data in real time and uses a predictive model to evaluate the machining status in real time. When the performance deviates from the expected value, the cutting parameters are dynamically adjusted in real time. When the analysis based on sensor data and machine learning algorithms shows that the physical entity layer component has a performance degradation trend, a maintenance warning is generated.

[0014] In step S3, the specific method for adjusting the cutting parameters is as follows: real-time acquisition of cutting force, spindle vibration spectrum characteristics, spindle speed and position, and coolant state parameters. When the cutting force is found to continuously exceed the parameter set range of the prediction model, the feed rate is dynamically adjusted in real time to maintain a stable cutting state. Step S3 also includes: when an abnormal upward trend in the amplitude of a specific frequency component is detected, it is determined that there is early damage to the spindle bearing or transducer, and a maintenance warning is issued. The physical entity layer components include the spindle bearing, ultrasonic transducer, and cutting tool.

[0015] Compared with the prior art, the present invention has the following significant advantages: 1. Significantly improved processing efficiency: Through data-driven optimization of process parameters and adaptive control, the material removal rate (MRR) is maximized while ensuring processing quality, reducing trial and error time and ineffective processing steps, thus greatly improving processing efficiency; 2. Precise and controllable machining quality: The tool tip tracking (TPT) function enables non-linear error compensation, ensuring the accuracy of the tool movement trajectory and the constant angle between the tool and the machined surface. Combined with real-time quality monitoring and parameter adjustment, it effectively reduces surface roughness and dimensional errors, and improves product consistency. 3. Significantly improved equipment reliability: The predictive maintenance module can provide early warnings of potential faults, avoiding losses from sudden downtime, while reducing resource waste caused by over-maintenance, and extending equipment lifespan and mean time between failures (MTBF). 4. Effective reduction in production costs: By optimizing tool wear control, reducing ineffective maintenance, and improving processing efficiency, the costs of tool consumption, equipment maintenance, and manual operation are reduced. At the same time, the reliance on operator experience is reduced, making it easier to achieve large-scale and standardized production. 5. Comprehensive upgrade in intelligence level: Integrating functions such as real-time monitoring, data-driven decision-making, adaptive control and predictive maintenance, it realizes intelligent management of the entire processing process and provides technical support for unmanned production. Attached Figure Description

[0016] Figure 1 : A schematic diagram of the four-layer architecture of the intelligent ultrasonic processing system of this invention; Figure 2 : Schematic diagram of the intelligent ultrasonic processing system of the present invention; Figure 3 : A schematic diagram of the ultrasonic spindle module of the present invention; Figure 4 : Schematic diagram of the control structure of the spindle system and the worktable during machining in this invention. Detailed Implementation

[0017] In the following description, numerous specific details are set forth in order to provide a more thorough understanding of the invention. However, it will be apparent to those skilled in the art that the invention can be practiced without one or more of these details. In other instances, certain technical features well-known in the art have not been described in detail to avoid obscuring the invention.

[0018] To further and thoroughly understand the contents of this invention, the accompanying drawings will be consulted. Figure 1-3 A detailed structural description and workflow are provided to illustrate the technical solution of this invention. Obviously, the implementation of this invention is not limited to the specific details familiar to those skilled in the art. Preferred embodiments of this invention are described in detail below; however, it should be understood that this invention can be implemented in different forms and should not be construed as limited to the embodiments presented herein. When the terms "comprising" and / or "including" are used in this specification, they indicate the presence of the stated features, integrals, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components, and / or combinations thereof.

[0019] Reference Figure 1 The intelligent ultrasonic processing system for processing brittle and hard materials according to the present invention adopts a four-layer architecture design of "physical entity - network connection - cloud platform - application service". Specifically, it includes a physical entity layer, a network connection layer, a cloud platform and data layer, and an application service layer. Each layer achieves collaborative linkage through standardized interfaces to ensure the intelligence and precision of the processing process.

[0020] (a) Physical entity layer The physical entity layer is built on a five-axis CNC machine tool and is the core of the hardware execution of the entire machining system. Its specific composition and connection relationship are as follows: The physical entity layer mainly includes the ultrasonic machining tool body, ultrasonic spindle module, cooling device, and sensor module. In addition, according to actual control requirements, it is also equipped with XY axis servo 15, three-axis motion control card 16, XY axis servo driver 17, data acquisition card 18, and industrial control computer 19. All components work together through mechanical connections or signal lines.

[0021] 1. Ultrasonic machining tool body The ultrasonic machining machine body serves as the supporting foundation, including a machine bed 111, a linear guide rail 112, a feed device 113, and a power supply 115. The machine bed 111 comprises a machine platform 1111, a fixture 1112, and a worktable 1113. The worktable 1113 has a vacuum chuck 7 at its upper end to fix the workpiece 6, and can move flexibly along the XY axis under the drive of the XY axis servo 15, meeting the position adjustment requirements for machining complex curved surfaces. The linear guide rail 112 provides guiding support for the movement of the worktable 1113 and the feed device 113, ensuring movement accuracy. The power supply 115 provides stable power to the electrical components of the entire physical entity layer (such as the spindle motor 121, ultrasonic generator 125, servo driver, etc.).

[0022] 2. Ultrasonic spindle module like Figure 1-3 As shown, the ultrasonic spindle module is the core machining execution unit of this system, including a spindle motor 121, a spindle 122, an ultrasonic transmitter 123, an ultrasonic transducer 124, an ultrasonic generator 125, an ultrasonic amplitude transformer 126, a tool holder 127, a tool 128, an ultrasonic receiver 1231, and a servo module 129. The servo module 129 includes a Z-axis servo driver 1291 and a Z-axis servo 1292. The Z-axis servo 1292 is connected to the spindle motor 121 and can drive the spindle motor 121 to move up and down along the Z-axis, thereby adjusting the machining distance between the tool 128 and the workpiece 6. (Those skilled in the art can add X-axis and Y-axis servos and corresponding drivers according to actual machining needs to expand multi-dimensional motion capabilities.) Figure 3 The embodiments shown are only typical application scenarios and do not constitute a limitation on the server module 129.

[0023] The ultrasonic spindle module adopts a built-in integrated design, which is one of the core innovations of this embodiment. Its structural layout is as follows: Figure 2As shown: the ultrasonic transmitter 123 and the ultrasonic transducer 124 are deeply integrated into the housing of the spindle 122. The ultrasonic receiver 1231 is fixedly installed on the outside of the housing of the spindle 122 and on the upper end of the tool holder of the tool holder 127. A specific gap is reserved between the receiver coil of the ultrasonic receiver 1231 and the transmitter coil of the ultrasonic transmitter 123. This gap design ensures the stability of non-contact energy and signal transmission and does not interfere with the high-speed rotation of the spindle 122.

[0024] The parameters and functions of each core component are as follows: - Ultrasonic Generator 125: Used to convert 50Hz conventional power supply into high-frequency electrical energy. Its maximum output power is 60W, the output frequency range is 20-35 kHz, the vibration amplitude adjustment range is 1-14 μm, the input voltage supports AC 220V or DC24V, and the maximum output voltage is AC 200V. The generator 125 has overvoltage and overcurrent protection functions, and can automatically track, match and tune the frequency to ensure that the ultrasonic spindle module always works in the optimal resonance state.

[0025] - Ultrasonic transducer 124: A piezoelectric ceramic transducer is selected. Its core function is to convert the high-frequency electrical energy generated by the ultrasonic generator 125 into mechanical longitudinal vibration. Its rated power is 10kW, and its frequency and amplitude range are matched with those of the ultrasonic generator 125 (20-35 kHz, 1-14 μm). The outer shell diameter is 120mm, and it is connected to the spindle 122 using an HSK-40 tapered interface. The transducer 124 has an internal cooling channel to meet the requirements of waterproofing and dustproofing. The overall structure is compact and is an independent electromechanical integrated unit. With the non-contact transmission module, it can support the spindle 122 speed of up to 24,000 rpm.

[0026] - Ultrasonic amplitude transformer 126: One end is connected to the ultrasonic transducer 124, and the other end is connected to the tool holder 127. Its core function is to amplify the vibration amplitude output by the ultrasonic transducer 124 and then transmit it to the tool holder 127 and the tool 128 to ensure that the end of the tool 128 receives sufficient machining vibration energy.

[0027] - Tool holder 127: Used to hold diamond abrasive tools 128, supports automatic tool changing function, can hold tools with a maximum diameter of 120mm, length of 200mm and weight of 3kg, has high clamping accuracy and strong stability, and can meet the cutting force requirements for machining brittle and hard materials.

[0028] The entire ultrasonic spindle module (including ultrasonic transducer 124, ultrasonic amplitude transformer 126, tool holder 127, and tool 128, etc.) has been precisely designed and debugged to ensure that it works in a resonant state and forms a standing wave: the vibration amplitude at the end of the tool 128 is amplified to the maximum antinode (up to tens of micrometers, depending on the gain coefficient of the ultrasonic amplitude transformer), while the connection between the ultrasonic transducer 124 and the spindle 122 is the minimum node (theoretically close to zero). This design can effectively isolate the transmission of ultrasonic vibration to the spindle box and machine bed 111, avoiding the negative impact of vibration on the machining accuracy and service life of the machine tool.

[0029] 3. Cooling device The cooling device includes components such as a coolant reservoir, a coolant treatment tank 131, a pump 132, a solenoid valve, a flow meter, and a pressure regulator. Its core function is to supply coolant to the machining area (the contact area between the tool 128 and the workpiece 6) and the spindle 122. Specifically, a high-pressure cooling pump (voltage AC380V, cooling capacity not less than 3.5KW, ambient temperature operating range 5-45°C) is used to spray coolant directionally into the machining area through the inner hole of the tool 128. This effectively removes the grinding debris generated during machining, preventing secondary scratches on the surface of the workpiece 6, and reduces the cutting temperature, preventing the workpiece 6 from cracking or degrading due to high temperature. It also provides cooling protection for the spindle 122 and the ultrasonic transducer 124. The flow meter and pressure regulator can monitor the flow rate and pressure of the coolant in real time to ensure stable cooling effect.

[0030] 4. Sensor Module The sensor module includes a sensor and a data acquisition module 142 ( Figure 1 The system (marked as a data acquisition system) is fixed to the machine tool body using a dedicated mounting bracket. The installation position avoids moving parts (such as linear guide 112 and feed device 113) to prevent motion interference and ensure the accuracy of data acquisition and the safety of equipment operation.

[0031] The specific types and functions of the sensors are as follows: - Cutting force sensor: mounted on spindle module 12, used to accurately measure the three-dimensional cutting force during the machining process, providing force feedback data for machining status assessment; - Vibration sensor: An accelerometer is selected and installed on the worktable 1113 or spindle 122 housing to monitor the vibration state of the worktable 1113 and spindle 122 during the material removal process and capture the vibration spectrum characteristics; - Acoustic emission sensor: used to monitor acoustic emission signals during the material removal process, reflecting the breakage of workpiece 6 and the wear state of tool 128; - Spindle speed encoder: Connected to spindle 122, it is used to provide real-time feedback of the spindle speed and position information of spindle 122, providing position closed-loop feedback for motion control; - Coolant flow sensor and pressure sensor: installed in the pipeline of cooling device 13 respectively, to monitor the flow and pressure of coolant and ensure the normal operation of the cooling system; - Ultrasonic amplitude and frequency monitoring sensor: Installed in conjunction with ultrasonic transducer 124 or ultrasonic amplitude transformer 126, it is used to monitor the amplitude and frequency parameters of ultrasonic vibration in real time to ensure the stability of ultrasonic processing parameters.

[0032] The data acquisition module 142 can be a data acquisition board or a combination of "data acquisition board + edge calculator". Its core function is to perform preliminary summarization and format conversion of the raw data collected by each sensor, in preparation for subsequent uploading to the cloud platform and data layer through the network connection layer.

[0033] (ii) Network connectivity layer The network connectivity layer is a key hub for enabling data interconnection between different levels. It consists of Industrial Internet of Things (IIoT) infrastructure 201, including IIoT gateways, industrial Ethernet, wireless communication modules (such as 5G and WiFi), and supports wired and wireless hybrid communication modes.

[0034] Its core working logic is as follows: the data acquisition module 142 of the sensor module collects various raw data (such as cutting force, vibration parameters, spindle speed, etc.), which is then converted through the IIoT gateway (supporting multiple industrial protocols such as Modbus / TCP, IEC 60870, DNP3, etc.) and uploaded in real time to the cloud platform and the cloud server of the data layer 30; at the same time, the control commands issued by the application service layer (such as process parameter adjustment commands, motion control commands, etc.) are also accurately transmitted through this network connection layer to the various execution components of the physical entity layer (such as the server module 129, cooling device, ultrasonic generator 125, etc.), ensuring the real-time performance and reliability of command transmission.

[0035] (III) Cloud Platform and Data Layer The cloud platform and data layer are the core of the entire system's data processing and storage, including the cloud server 301, the database, and the data processing and analysis module 302. These components work together to realize the functions of data storage, processing, and analysis. - Cloud Server 301: Adopts a high-performance industrial cloud server to provide the entire system with ample computing resources and large-capacity storage resources, supporting parallel processing and secure storage of massive historical data and real-time data, and meeting the needs of multiple devices accessing at the same time and high-concurrency data processing. - Database: A distributed database architecture is adopted, which is specifically used for classifying, storing and managing various types of data transmitted through the network connection layer, including real-time data collected by sensor modules (such as spindle speed 122, feed rate of feed device 113, cutting force, etc.), historical machining data (such as process parameter sets of different workpieces, machining quality results, etc.), and equipment operation data (such as spindle bearing temperature, ultrasonic transducer 124 working time, etc.). Data storage adopts an encrypted format to ensure data security. - Data Processing and Analysis Module 302: The data analysis engine is built using Python with TensorFlow / PyTorch framework. It cleans, denoises, preprocesses and extracts features from the collected raw data, removes invalid data and outliers, and extracts key features related to processing quality and equipment status (such as peak cutting force, vibration spectrum characteristic frequency, etc.) to provide high-quality data support for the various modules of the application service layer 40.

[0036] (iv) Application Service Layer The application service layer is the core of the system's intelligent decision-making and control. It includes a real-time monitoring module, a parameter optimization module, a control module, an adaptive control module, and a predictive maintenance module. Each module can be flexibly deployed in the form of microservices on the cloud platform, data layer, or edge server according to the actual application scenario, and has good scalability and compatibility.

[0037] 1. Real-time monitoring module The core functions of the real-time monitoring module are data visualization and anomaly alarm: On the one hand, the real-time data collected by the data acquisition module 142 (such as the power of the ultrasonic transducer 124, the vibration parameters of the spindle 122, the feed rate of the feed device 113, the flow rate and pressure of the coolant, etc.) are displayed on the Web or the screen of the industrial control computer 19 in a visual form such as dashboard, graphs, and digital displays, so that operators can keep track of the processing status in real time; on the other hand, based on the analysis results of the predictive maintenance module, the module accurately identifies abnormal situations (such as cutting force exceeding the threshold, insufficient coolant pressure, abnormal vibration of the spindle 122, etc.) and potential faults in the processing process, and promptly reminds the operators through audible and visual alarms, pop-up prompts, etc., to ensure the safety and stability of the processing process.

[0038] 2. Parameter Optimization Module The parameter optimization module uses historical machining data stored in the cloud platform and data layer 30 database to establish a predictive model between input variable data and output performance variable data using machine learning algorithms (such as neural networks, random forests, and support vector machines). The input variable data includes process parameters such as spindle speed (122), feed rate (113), and ultrasonic transducer power (124); the output performance variable data includes machining result parameters such as material removal rate (MRR), workpiece surface roughness (6), and tool wear (128).

[0039] The optimization objective of this module is to maximize the material removal rate (MRR) while satisfying the workpiece surface roughness (6) and tool wear (128) constraints. The specific workflow is as follows: Beforehand, a large amount of machining data under different combinations of process parameters is collected through Design Experiments (DOE) on the machine tool to train a neural network model for specific brittle and hard materials (such as sapphire, glass, and ceramics). During actual machining, after the user inputs the target MRR and surface roughness requirements, the model outputs a set of optimized process parameters within the constraints of ultrasonic system parameters (power ≤ 60W, frequency 20-35kHz, amplitude 1-14μm) and spindle speed (0-24,000 rpm), and sends this set to the control module to provide optimal parameter guidance for the machining process.

[0040] 3. Control Module The control module employs a highly integrated programmable logic controller (PLC). This PLC features a modular design, high reliability, ease of maintenance, and stable operation in high electrical noise environments. It complies with the IEC 60801 electromagnetic compatibility standard, and can operate in environments with temperatures up to 50°C and humidity up to 95%. It utilizes passive cooling and supports data interaction with the cloud platform and data layer 30 via multiple protocols such as Modbus / TCP, IEC 60870, and DNP3. It is equipped with an isolation converter for acquiring signals from devices such as electricity meters, providing crucial support for data uplink and downlink control.

[0041] The control module is integrated into the machine tool's CNC system. Its core advantage lies in its tip-tracking (TPT) function: allowing users to pre-set the workpiece 6's contour, desired tool 128 angle, and process parameter set via the industrial control computer 19. The control system automatically calculates and compensates for nonlinear errors caused by the spindle 122's rotational motion and the movement of the linear guide 112 and feed device, ensuring that the tool tip of the tool 128 always moves along the predetermined trajectory and maintains a constant angle between the tool 128 and the workpiece 6's machined surface. Furthermore, the control module supports automatic compensation for changes in the tool 128's length. When the tool 128's length changes due to wear, the system automatically adjusts the compensation amount based on the signal fed back from the spindle speed encoder, significantly improving program portability and machining accuracy.

[0042] like Figure 3 As shown, the hardware connection logic of the control module is as follows: the industrial control computer 19 has a built-in core program for the control module; the data acquisition card 18 reads the data transmitted by the data acquisition module 142 and inputs it to the industrial control computer 19; the industrial control computer 19 sends control commands to the XY-axis servo driver 17 and the Z-axis servo driver 1291 through the three-axis motion control card 16, driving the XY-axis servo driver 15 and the Z-axis servo driver 1292 respectively, thereby driving the worktable 1113 to move along the XY-axis and the spindle 122 to move along the Z-axis, realizing the relative motion control between the tool 128 and the workpiece 6. In practical applications, this control module utilizes the feedback signals from the PLC and the high-precision encoder of the spindle 122 to significantly reduce contour errors and improve machining accuracy when machining complex curved surfaces.

[0043] 4. Adaptive Control Module The adaptive control module evaluates the machining status in real time based on real-time process monitoring data and using predictive models (such as tool wear prediction models and online surface roughness estimation models) established by the parameter optimization module. When it detects that the performance deviates from expectations due to tool 128 wear, temperature changes in the machining area, or other external disturbances (such as cutting force continuously exceeding the predicted range or online surface roughness estimation exceeding the standard), the module sends control commands to the PLC through the industrial control computer 19 to dynamically adjust cutting parameters (such as the feed rate of the feed device 113, the speed of the spindle motor 121, and the power of the ultrasonic transducer 124) to maintain an optimized machining status and effectively counteract the negative impact of disturbances. For example, when the real-time collected cutting force data continuously exceeds the model prediction range, it indicates that the tool 128 may have worn. The system will automatically dynamically adjust the feed rate within the range of 1-20,000 mm / min to ensure the stability of the cutting status.

[0044] 5. Predictive Maintenance Module The predictive maintenance module continuously analyzes the performance degradation trends of key components (such as spindle bearings, ultrasonic transducers 124, and tool wear-related signals) in the physical entity layer 10 based on real-time data collected by sensor 141 (such as vibration spectrum characteristics of spindle 122, operating parameters of ultrasonic transducers 124, and wear-related signals of tool 128) and machine learning algorithms (such as neural networks and random forest algorithms). When an abnormal upward trend in the amplitude of a specific frequency component is detected (such as an increase in the characteristic frequency amplitude corresponding to bearing wear in the vibration spectrum of spindle 122, or a decrease in the stability of vibration amplitude of ultrasonic transducers 124), it is determined that the corresponding component may have early damage. The system will issue a maintenance warning in advance and accurately predict the occurrence time of potential failures and the remaining lifespan of components, guiding operators to perform condition-based maintenance, avoiding resource waste caused by unplanned downtime and over-maintenance, and extending the mean time between failures (MTBF) of the equipment.

[0045] II. Specific Execution Steps of Ultrasonic Processing This embodiment also proposes an ultrasonic machining method with tool tip tracking (TPT) function. This method is applied to the above-mentioned intelligent ultrasonic machining system and specifically includes the following steps: S1: Output optimized process parameter set The parameter optimization module of the application service layer 40 uses historical processing data stored in the cloud platform and the database of the data layer 30 to train a prediction model for hard and brittle materials (such as sapphire, ceramics, and tempered glass) using machine learning algorithms (such as neural network algorithms). The model aims to meet the surface roughness of the workpiece 6 and the wear constraints of the tool 128, and maximize the material removal rate (MRR). It outputs an optimized set of process parameters (including the spindle speed 122, the feed rate of the feed device, the power of the ultrasonic transducer 124, the ultrasonic vibration frequency and amplitude, etc.) and sends the parameter set to the control module.

[0046] S2: Precise control of tool movement trajectory The control module receives the set of process parameters output by the parameter optimization module through the three-axis motion control card 16. Combined with the workpiece 6 contour and the desired tool 128 angle set in advance in the industrial control computer 19, it automatically calculates and compensates for the nonlinear errors generated during the rotation of the spindle 122 and the movement of the worktable 1113 driven by the XY axis servo 15. At the same time, the Z-axis servo 1292 is controlled by the Z-axis servo driver 1291 to drive the spindle motor 121 to drive the spindle 122 to move along the Z-axis. This controls the tool tip of the tool 128 to always move along the predetermined trajectory and maintains a constant angle between the tool 128 and the workpiece 6 machining surface, ensuring that the machining accuracy meets the design requirements.

[0047] S3: Real-time monitoring and dynamic adjustment Sensors 141 in sensor module 14 (including cutting force sensor, vibration sensor, spindle speed encoder, coolant flow sensor, coolant pressure sensor, etc.) collect various data in real time during the machining process. This data is transmitted to industrial control computer 19 via data acquisition module 142 and data acquisition card 18, and simultaneously uploaded to cloud platform and data layer 30 for processing. The adaptive control module in application service layer 40 uses a predictive model to analyze the real-time data and evaluate the machining status in real time. When performance deviates from expectations (such as cutting force exceeding the threshold or surface roughness exceeding the standard), the cutting parameters are dynamically adjusted in real time. Specifically, this adjustment involves real-time acquisition of cutting force and spindle 122 vibration frequency. Based on spectral characteristics, spindle speed and position, and coolant status parameters (flow rate, pressure), when the cutting force is found to continuously exceed the parameter set range of the prediction model, the feed rate of the feed device 113 is dynamically adjusted in real time by the PLC to maintain a stable cutting state. At the same time, the predictive maintenance module continuously analyzes whether there is a performance degradation trend in key components (spindle bearing, ultrasonic transducer 124, tool 128, etc.) of the physical entity layer 10 based on the data collected by the sensor 141 and machine learning algorithms. When an abnormal upward trend in the amplitude of a specific frequency component is found, it is determined that there is early damage to the spindle bearing or ultrasonic transducer 124, and a maintenance warning is issued in a timely manner to guide the operator to perform targeted maintenance.

[0048] III. Actual Implementation Examples The following describes in detail the complete workflow of this intelligent ultrasonic machining system through a specific scenario of machining a sapphire watch backplate (i.e., workpiece 6), demonstrating the collaborative working process of each component module: 1. Process Parameter Optimization Stage: The parameter optimization module of the application service layer 40, based on the historical data of sapphire processing and the neural network model in the cloud platform and the database of the data layer 30, aims to maximize the material removal rate (MRR) while satisfying the surface roughness Ra of the workpiece 6 ≤ 0.1μm. It outputs an optimized set of process parameters: spindle 122 speed 20,000rpm, feed rate of feed device 113 120mm / min, power of ultrasonic transducer 124 100W, ultrasonic vibration frequency 30kHz, amplitude 8μm. This set of parameters is transmitted to the control module through the network connection layer 20.

[0049] 2. Motion Control Stage: The operator inputs the three-dimensional contour data of the sapphire watch backplate (workpiece 6) and the desired tool 128 angle through the industrial control computer 19. The control module activates the tip tracking (TPT) function, automatically calculating and compensating for nonlinear errors caused by the rotation of the spindle 122 and the movement of the worktable 1113 during the five-axis linkage process. At the same time, the XY axis servo driver 17 drives the XY axis servo 15, which drives the worktable 1113 (with the workpiece 6 fixed on it) to move along the XY axis. The Z axis servo driver 1291 drives the Z axis servo 1292, which drives the spindle 122 to move along the Z axis, ensuring that the tip of the tool 128 moves accurately along the predetermined contour trajectory.

[0050] 3. Process Monitoring and Adjustment Stage: During processing, the high-pressure pump 132 of the cooling device 13 is activated, and the coolant is continuously sprayed into the processing area through the inner hole of the tool 128. The coolant flow sensor and pressure sensor monitor the coolant status in real time. The cutting force sensor, vibration sensor, and acoustic emission sensor of the sensor module 14 continuously collect data such as cutting force, spindle 122 vibration, and acoustic emission, and upload them to the cloud platform and data layer 30 for processing through the data acquisition module 142. When the adaptive control module detects that the cutting force abnormally rises from the initial 50N to 80N (exceeding the predicted range), it determines that the tool 128 has experienced slight wear. It immediately adjusts the feed rate of the feed device 113 from 120mm / min to 80mm / min through the PLC to maintain a stable cutting state. At the same time, the predictive maintenance module, based on the vibration spectrum analysis of the spindle 122, finds that the characteristic frequency amplitude corresponding to the wear of the tool 128 is continuously rising, issues a tool replacement warning to the operator 15 minutes in advance, and evaluates the remaining life of the spindle bearing through historical data fitting to be approximately 200 hours.

[0051] 4. Processing Completion Stage: After 30 minutes of continuous processing, a high-quality sapphire watch backplate (workpiece 6) with a surface roughness Ra of 0.08-0.095μm was successfully obtained. The processing dimensional error was controlled within ±0.005mm, which fully verified the advantages of this system in high precision, high efficiency and high stability in the processing of brittle and hard materials.

[0052] The intelligent ultrasonic processing system and method provided by this invention, by integrating core intelligent functions such as real-time process monitoring, intelligent process parameter decision-making and optimization, tool tip tracking (TPT) function, adaptive control and predictive maintenance, can significantly improve the processing efficiency, product quality and equipment reliability of brittle and hard materials, while reducing production costs and dependence on operator experience, and provides reliable technical support for the large-scale, standardized and unmanned processing of brittle and hard materials.

[0053] The above-described embodiments are merely preferred embodiments provided to fully illustrate the present invention, and the scope of protection of the present invention is not limited thereto. Equivalent substitutions or modifications made by those skilled in the art based on the present invention are all within the scope of protection of the present invention. The scope of protection of the present invention is defined by the claims.

Claims

1. An intelligent ultrasonic machining system for processing brittle and hard materials, the system comprising a physical entity layer, a network connection layer, a cloud platform and data layer, and an application service layer, characterized in that: the physical entity layer comprises an ultrasonic machining machine tool body, an ultrasonic spindle module, a cooling device, and a sensor module; the ultrasonic machining machine tool body comprises a machine tool bed, a linear guide rail, a feeding device, and a power supply; the ultrasonic spindle module comprises a spindle motor, a spindle, an ultrasonic transmitter, an ultrasonic transducer, an ultrasonic generator, an ultrasonic amplitude horn, a tool holder, a tool, an ultrasonic receiver, and a servo module, the power supply supplies power to the spindle motor and the ultrasonic generator, the servo module comprises a servo and a servo driver, and is used to drive the spindle motor to move along a preset axis; the ultrasonic transmitter and the ultrasonic transducer are built into the housing of the spindle, the ultrasonic receiver is arranged on the upper end of the tool fixing frame of the tool holder outside the spindle housing, a gap is arranged between the receiver coil of the ultrasonic receiver and the transmitter coil of the ultrasonic transmitter for realizing non-contact energy and signal transmission; the ultrasonic transmitter, the ultrasonic transducer, the ultrasonic amplitude horn, the tool holder, and the tool are sequentially installed at the end of the spindle, the ultrasonic transducer converts high-frequency electric energy generated by the ultrasonic generator into mechanical longitudinal vibration, the ultrasonic amplitude horn amplifies the vibration of the ultrasonic transducer and then transmits it to the tool holder and the clamped tool, so that the ultrasonic spindle module works in a resonant state to form a standing wave; the cooling device comprises a cooling liquid storage tank, a cooling liquid treatment tank, a high-pressure pump, a solenoid valve, a flow meter, and a pressure regulator, and is used to supply cooling liquid to the machining area and the spindle, and the cooling liquid is sprayed to the machining area through the inner hole of the tool; the sensor module comprises a sensor and a data acquisition module, and is installed on the machine tool body to acquire variable data during the machining process; the network connection layer is used to realize data interconnection between the physical entity layer, the cloud platform and data layer, and the application service layer, and the data acquisition module transmits data to the cloud platform and data layer and the application service layer through the network connection layer; the cloud platform and data layer comprise a cloud server, a database, and a data processing and analysis module, the cloud server is used to provide computing and storage resources; the database is used to store and manage historical data and real-time data collected by the data acquisition module; the data processing and analysis module is used to clean, pretreat, and extract features from the collected real-time data; the application service layer comprises a real-time monitoring module, a parameter optimization module, a control module, a self-adaptive control module, and a predictive maintenance module. The real-time monitoring module is configured to display the real-time data collected by the data acquisition module in a visualized form, and identify abnormalities and faults based on the analysis of the predictive maintenance module, and alarm; the parameter optimization module is configured to establish a prediction model between input variable data and output performance variable data based on historical data in a database and a machine learning algorithm, so as to maximize material removal rate (MRR) under the constraints of surface roughness and tool wear, output an optimized process parameter set, and send the optimized process parameter set to the control module; The control module is configured to automatically calculate and compensate for nonlinear errors caused by the rotation of the spindle based on the pre-set workpiece profile, the desired tool angle, and the process parameter set, control the tool tip point to move along the predetermined trajectory, and keep the tool angle with the machining surface constant; the adaptive control module is configured to evaluate the machining state in real time based on the real-time data and the prediction model, and automatically adjust the cutting parameters when the performance deviates from the expectation, so as to eliminate the disturbance effect; the predictive maintenance module is configured to analyze whether there is a performance degradation trend in the physical entity layer based on the collected real-time data and the machine learning algorithm, predict the occurrence time of potential faults and the service life of the component, and generate a maintenance warning.

2. The intelligent ultrasonic machining system of claim 1, wherein: The application service layer can be deployed in the form of microservices on the cloud platform and the data layer or the edge server; the control module can also automatically compensate for the change in the tool length; the network connection layer is composed of an industrial Internet of Things infrastructure; the cooling liquid is sprayed into the machining area through the inner hole of the tool.

3. The intelligent ultrasonic machining system of claim 1, wherein: The sensors include a cutting force sensor for measuring three-directional cutting forces, a vibration sensor, a spindle speed encoder for feeding back the spindle speed and position, a cooling liquid flow sensor, a cooling liquid pressure sensor, and an ultrasonic amplitude and frequency monitoring sensor; the cooling liquid flow sensor and the cooling liquid pressure sensor are used to monitor the state of the cooling liquid, and the vibration sensor is an accelerometer used to monitor the state of the workbench during material removal.

4. The intelligent ultrasonic machining system of claim 1, wherein: The control module is a programmable logic controller.

5. The intelligent ultrasonic machining system of claim 1, wherein: The variable data includes the spindle speed, the feed rate, and the ultrasonic transducer power; the performance variable data includes the material removal rate (MRR), the surface roughness, and the tool wear.

6. The intelligent ultrasonic machining system of claim 1, wherein: The prediction model is a neural network model; the machine learning algorithm includes a neural network, a random forest, and / or a support vector machine.

7. The intelligent ultrasonic machining system of claim 1, wherein: The real-time monitoring module can display machining state parameters, including the ultrasonic transducer power, the spindle vibration, the spindle speed, and / or the feed rate.

8. An ultrasonic machining method with tool tip point tracking (TPT) function, the method is applied to the intelligent ultrasonic machining system of any one of claims 17, characterized in that, The method comprises the following steps: S1, outputting a process parameter set using a prediction model for hard and brittle materials trained by a machine learning algorithm; S2, calculating and compensating for nonlinear errors caused by the rotation of the spindle based on the pre-set workpiece profile, the desired tool angle, and the process parameter set, controlling the tool tip point to move along the predetermined trajectory, and keeping the tool angle with the machining surface constant; S3, real-time acquisition of sensor data, real-time evaluation of machining state by using prediction model, real-time dynamic adjustment of cutting parameters when performance deviates from expectation, generation of maintenance warning when performance degradation trend of physical entity layer component is analyzed based on sensor data and machine learning algorithm.

9. The ultrasonic machining method according to claim 8, characterized by: In step S3, the specific way of adjusting the cutting parameters is: real-time acquisition of cutting force, spindle vibration frequency spectrum characteristics, spindle speed and position, and cooling liquid state parameters; when it is found that the cutting force continuously exceeds the parameter set range of the prediction model, the feed rate is dynamically adjusted in real time to maintain the stability of the cutting state; step S3 also includes: when it is found that the amplitude of a specific frequency component has an abnormal rising trend, it is judged that the spindle bearing or transducer has early damage, and a maintenance warning is issued.

10. The ultrasonic machining method according to claim 8, characterized by: The physical entity layer component includes a spindle bearing, an ultrasonic transducer, and a tool.

Images