Method and device for wear resistance detection of gantry machine tool spindle system

Abstract

The application discloses a wear resistance detection method and equipment for a gantry machine tool spindle system, and relates to the technical field of machine tool spindle state detection. The method comprises the following steps: when the spindle is running, first, torque and rotating speed signals are collected to calculate friction power input energy; the spindle and lubricating oil characteristic sets are synchronously collected to establish heat dissipation energy, and acoustic emission energy is calculated by using an acoustic emission array; oil detection is performed to establish abrasive particle generation potential, equivalent grinding energy is calculated to obtain a first wear detection result; an ultrasonic sensor is activated to obtain a second wear detection result; and after interactive authentication, a wear resistance detection result is output. The application solves the technical problem that the wear resistance detection means of the gantry machine tool spindle system is single, the wear state cannot be reflected in real time, only partial physical quantities can be captured, and the detection data is one-sided, the accuracy and reliability are insufficient, achieves multi-dimensional real-time comprehensive detection of the wear resistance of the gantry machine tool spindle system, and improves the technical effect of comprehensiveness, accuracy and reliability of the detection data.

Description

Technical Field

[0001] This invention relates to the field of machine tool spindle condition detection technology, and in particular to a method and equipment for testing the wear resistance of a gantry machine tool spindle system. Background Technology

[0002] The spindle system of a gantry milling machine is a core functional component of the machine tool, and its wear resistance directly affects machining accuracy and equipment lifespan. Accurate testing of the spindle system's wear resistance is crucial for ensuring stable machine tool operation. Current technologies for testing the wear resistance of gantry milling machine spindle systems primarily employ single-physical quantity monitoring methods, including offline oil level testing, single-point temperature monitoring, or vibration monitoring. While these methods can be effective in simple, stable operating scenarios, their limitations become apparent in practical applications as high-end manufacturing demands increasing machine tool precision. They cannot simultaneously capture multi-dimensional information such as changes in spindle friction energy, microscopic damage, and abrasive particle formation, resulting in incomplete and inaccurate test data, failing to meet the needs for accurate assessment and timely maintenance of the gantry milling machine spindle system's wear resistance. Summary of the Invention

[0003] This application provides a method and equipment for testing the wear resistance of a gantry milling machine spindle system, which solves the technical problem that the wear resistance testing methods for gantry milling machine spindle systems are limited, cannot reflect the wear state in real time, and can only capture local physical quantities, resulting in biased, inaccurate and unreliable test data.

[0004] The first aspect of this application provides a method for testing the wear resistance of a gantry milling machine spindle system. The method includes: acquiring spindle torque and speed signals while the spindle is running; calculating frictional power input energy based on the spindle torque and speed signals; simultaneously acquiring spindle and lubricating oil feature sets to establish heat dissipation energy; utilizing an acoustic emission array to execute transient energy release signals from the friction surface and calculating acoustic emission energy; performing oil detection to establish abrasive particle generation potential; calculating equivalent wear energy based on the difference between the frictional power input energy and the heat dissipation energy, acoustic emission energy, and abrasive particle generation potential; establishing a first-path wear detection result based on the equivalent wear energy; activating an ultrasonic sensor to perform ultrasonic monitoring and identification of the spindle and establishing a second-path wear detection result; performing interactive authentication of the first-path and second-path wear detection results and outputting a wear resistance test result.

[0005] A second aspect of this application provides an electronic device comprising: a memory for storing executable instructions; and a processor for executing the executable instructions stored in the memory to implement a wear resistance testing method for a gantry milling machine spindle system.

[0006] This application proposes one or more technical solutions, which have at least the following technical effects:

[0007] This application achieves accurate wear detection of the gantry machine tool spindle system by collecting spindle torque and speed signals during spindle operation to calculate frictional power input energy, simultaneously collecting spindle and lubricating oil characteristics to establish heat dissipation energy, calculating acoustic emission energy using an acoustic emission array, and combining this with the abrasive particle generation potential established by oil detection. The equivalent wear energy is then calculated using multi-energy difference to obtain the first wear detection result. Simultaneously, an ultrasonic sensor is activated to collect and process signals and extract key features, identifying microscopic defects and wear states to obtain the second wear detection result. Through interactive authentication of the dual-path results, accurate wear resistance detection of the gantry machine tool spindle system is achieved, making the detection results more reliable. This results in multi-dimensional, real-time, and comprehensive wear resistance detection of the gantry machine tool spindle system, improving the comprehensiveness, accuracy, and reliability of the detection data. Attached Figure Description

[0008] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0009] Figure 1 This is a schematic flowchart of the wear resistance testing method for the spindle system of a gantry milling machine provided in the embodiments of this application.

[0010] Figure 2 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0011] This application provides a method and equipment for testing the wear resistance of a gantry milling machine spindle system, which solves the technical problem that the wear resistance testing methods for gantry milling machine spindle systems are limited, cannot reflect the wear state in real time, and can only capture local physical quantities, resulting in biased, inaccurate and unreliable test data.

[0012] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0013] It should be noted that any variation of the terms "comprising" and "having" is intended to cover non-exclusive inclusion, for example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or modules that are not explicitly listed or that are inherent to such processes, methods, products, or devices.

[0014] Example 1, as Figure 1 As shown, a method for testing the wear resistance of a gantry milling machine spindle system includes:

[0015] Step A100: While the spindle is running, acquire the spindle torque and speed signals, and calculate the friction power input energy based on the spindle torque and speed signals.

[0016] Specifically, when the spindle is in operation, the first step is to deploy sensors to collect spindle torque and speed signals. The torque sensor can be integrated into a key position in the spindle's drive train to ensure that it can directly sense the dynamic changes in torque during spindle operation. At the same time, the speed sensor is installed in a non-stressed area near the end of the spindle to avoid interference with speed detection caused by spindle vibration or load changes. Through the coordinated deployment of these two types of sensors, synchronous real-time capture of spindle torque and speed signals can be achieved.

[0017] After the two types of sensors are deployed, the signal acquisition process is initiated. As the spindle rotates, the torque sensor continuously records torque fluctuation data under different load conditions, while the speed sensor synchronously records the real-time rotational speed data of the spindle. These raw signals are transmitted in real time to the data processing unit in the spindle system. Due to electromagnetic interference, mechanical vibration, and other factors in the spindle's operating environment, the raw torque and speed signals transmitted to the data processing unit are usually mixed with high-frequency noise. Furthermore, affected by the temperature drift of the sensors themselves, the signals may also exhibit low-frequency drift. Directly using these raw signals would severely affect the accuracy of subsequent calculations; therefore, targeted processing is necessary.

[0018] Finally, after removing high-frequency noise and low-frequency drift from the spindle torque and speed signals, the torque in the spindle torque is multiplied by the angular velocity in the speed to calculate the instantaneous friction power. Then, the instantaneous friction power is integrated over time to construct the friction power input energy. The specific steps are explained in detail in A110-A130.

[0019] Through a series of steps including sensor deployment, real-time signal acquisition, and digital filtering preprocessing, clean and stable spindle torque and speed signals were obtained, laying a data foundation for the subsequent accurate calculation of friction power input energy.

[0020] Step A200: Synchronously acquire the spindle and lubricating oil feature sets, establish heat dissipation energy, use the acoustic emission array to execute the transient energy release signal of the friction surface, and calculate the acoustic emission energy.

[0021] In this embodiment, the acoustic emission array is a detection device composed of multiple acoustic emission sensors, used to collect transient energy release signals from the friction surfaces of the spindle system of a gantry milling machine.

[0022] Optionally, firstly, flow data and temperature signals are collected using flow sensors and inlet / outlet temperature sensors in the lubricating oil passage. The specific heat capacity and density of the lubricating oil are then substituted into the instantaneous heat power loss model to calculate the instantaneous heat power and perform time integration to obtain the first heat loss energy. Then, the surface temperature distribution of the spindle is collected using an infrared thermal imaging device and combined with the structural heat conduction model to obtain the second heat loss energy. Finally, the second heat loss energy is used to compensate for the first heat loss energy to establish the heat loss energy. The specific steps are explained in detail in A210-A250.

[0023] Next, an array of acoustic emission sensors is arranged in the friction surface area to collect transient acoustic emission signals containing amplitude, duration, and frequency characteristics. The transient energy release is calculated by time integration of the transient acoustic energy of the signal. Then, the transient energy releases of multiple measurement points are accumulated to establish the acoustic emission energy. The specific steps are explained in detail in A260-A280.

[0024] Step A300: Perform oil detection, establish abrasive particle generation potential, calculate equivalent wear energy based on the difference between the friction power input energy, heat loss energy, acoustic emission energy, and abrasive particle generation potential, and establish the first path wear detection result based on the equivalent wear energy.

[0025] In this embodiment of the application, the abrasive grain generation potential refers to a parameter established by performing oil detection, which reflects the total energy required to generate all abrasive grains during the friction process of the spindle system.

[0026] In one embodiment of this application, firstly, the online oil monitoring unit is activated to perform time-series detection on the oil to obtain an abrasive feature set containing the number, size distribution, and material composition of abrasive particles. Based on the material composition, the density and hardness parameters of the abrasive particles are read. The energy required to generate a single abrasive particle is calculated by combining the size distribution and the number of abrasive particles. Then, the abrasive particle generation potential is established based on the energy required for all abrasive particles. The specific steps are described in detail in A310-A330.

[0027] Next, before calculating the equivalent wear energy, we must first ensure the time synchronization and data validity of the four core energy parameters already acquired: frictional power input energy, heat loss energy, acoustic emission energy, and abrasive particle generation potential. This is the foundation for accurate subsequent difference calculations. A timestamp synchronization mechanism can be used to unify the acquisition and calculation cycles of the four energy parameters to the same spindle operating time period, such as a continuous 1-hour spindle operation period, avoiding energy difference distortion caused by time dimension mismatch. First, verify that the time labels of the four data points all correspond to this 1-hour period, and that each data point has been preprocessed to ensure there are no outliers or calculation errors.

[0028] Subsequently, a calculation logic for the difference in equivalent wear energy is constructed based on the principle of energy balance. During the spindle friction process, the friction power input energy is the total energy generated by friction. This energy is consumed through three pathways: part of it is dissipated as heat through lubricating oil cooling and thermal radiation from the spindle surface; part of it is released as acoustic emission energy through transient damage to the friction surface, such as microcrack propagation; and part of it is consumed as abrasive grain generation potential due to material peeling and abrasive grain formation. The remaining energy not consumed by these three pathways is the equivalent wear energy truly used for the wear of the spindle friction surface. Therefore, the formula for calculating the equivalent wear energy is: Equivalent wear energy = Friction power input energy - Heat dissipation energy - Acoustic emission energy - Abrasive grain generation potential. During the calculation, it is necessary to ensure that the units of the four energy parameters are consistent in MJ to avoid affecting the results due to unit conversion deviations. At the same time, sufficient decimal places are retained for parameters with large differences in magnitude, such as acoustic emission energy and abrasive grain generation potential, which are much smaller than the first two items, to prevent calculation errors caused by ignoring small energy terms.

[0029] Then, the calculated equivalent wear energy is verified for rationality to eliminate deviations caused by abnormal single energy parameters. During verification, the equivalent wear energy data of the same model of gantry milling machine under the same working conditions, speed and load are compared. If the current calculation result is within the normal fluctuation range of history, the result is deemed valid. If it exceeds the range, the acquisition and calculation process of each energy parameter needs to be checked back. For example, check whether the lubricating oil flow sensor is faulty, causing the heat loss energy calculation to be too low, or whether the acoustic emission sensor is interfered with, causing the acoustic emission energy calculation to be too low. After correcting the abnormal parameters, the calculation is recalculated to ensure that the equivalent wear energy can truly reflect the actual wear energy consumption state of the spindle.

[0030] Finally, historical experimental data and external calibration data of the same model are obtained to map the equivalent wear energy to the wear state. After interactive authentication, a true mapping table between the two is established. The first wear detection result is established using this mapping table and the equivalent wear energy. The specific steps are explained in detail in A340-A360.

[0031] Step A400: Activate the ultrasonic sensor to perform ultrasonic monitoring and identification of the spindle, and establish the second wear detection result.

[0032] Specifically, after activating the ultrasonic sensor, ultrasonic signals are collected to establish an ultrasonic feature set containing propagation time, amplitude, waveform characteristics, and spectral characteristics. After noise suppression and normalization, wave velocity change, echo delay, radiation intensity attenuation, and spectral center shift characteristics are extracted to form a key feature set. Then, based on the first two types of features, microcracks and local pits on the main shaft are identified, and the wear state is identified based on the latter two types of features. Finally, the second wear detection result is established by combining the two types of identification results. The specific steps are explained in detail in A410-A460.

[0033] Step A500: Perform interactive authentication of the first wear detection result and the second wear detection result, and output the wear resistance test result.

[0034] Optionally, the wear detection results of the first and second channels can be converted into a unified numerical wear index to determine whether they are within the same wear level range. If they are, the two can be used as verification results, and the current wear level can be output as the wear resistance test result. The specific steps are explained in detail in A510-A530.

[0035] Furthermore, step A100 in the method provided in this application embodiment includes:

[0036] A110: High-frequency noise and low-frequency drift are removed from the spindle torque and speed signals.

[0037] A120: Performs a product operation on the torque in the spindle torque and the angular velocity in the speed signal to calculate the instantaneous friction power.

[0038] A130: Perform the runtime integral of the instantaneous frictional power to construct the frictional power input energy.

[0039] In this embodiment, high-frequency noise refers to short-duration, high-frequency irregular fluctuations in the collected spindle torque and speed signals caused by high-frequency interference factors such as electromagnetic radiation and mechanical vibration during spindle operation. Low-frequency drift refers to a slow, continuous deviation of the reference value in the collected spindle torque and speed signals caused by factors such as sensor temperature changes, device aging, or slow fluctuations in ambient temperature during continuous spindle operation.

[0040] Specifically, the acquired spindle torque and speed signals first need to be preprocessed to eliminate the impact of signal interference on subsequent calculations. The spindle operating environment is often subject to interference such as electromagnetic radiation and mechanical vibration, which can easily cause high-frequency noise to be mixed in with the spindle torque signal. At the same time, the sensor is affected by temperature changes during long-term operation, which can cause low-frequency drift in the speed signal. These two types of interference will directly affect the accuracy of the friction power input energy calculation.

[0041] Specifically, digital filtering techniques can be used to process the raw signal: For high-frequency noise, a low-pass filter is selected, and a cutoff frequency matching the spindle operating frequency is set. For example, when the spindle's rated speed is 2000 r / min, the cutoff frequency is set to 50 Hz to filter out interference signals higher than this frequency. For low-frequency drift, a moving average filtering algorithm is used to smooth multiple sets of data on the continuously acquired speed signal, gradually correcting the slow signal shift. Taking a certain type of gantry milling machine as an example, the original noise amplitude of its spindle torque signal can reach ±3 N·m. After processing, the noise amplitude is controlled within ±0.5 N·m, and the low-frequency drift of the speed signal is reduced from ±10 r / min to ±1 r / min, effectively ensuring the purity of the signal.

[0042] After signal preprocessing, instantaneous friction power is calculated. First, the real-time torque value is extracted from the processed spindle torque signal, and the real-time speed data is extracted from the speed signal. Since the calculation of friction power requires angular velocity as a parameter, the speed unit is converted from r / min to rad / s using the formula ω = 2πn / 60, where n is the speed and ω is the angular velocity. For example, when the spindle speed is 1500 r / min, the corresponding angular velocity is approximately 157.08 rad / s. If the extracted torque is 160 N·m, according to the friction power calculation formula... =T×ω, where Let T be the instantaneous frictional power, ω be the torque, and ω be the angular velocity. The instantaneous frictional power at this moment can be calculated to be approximately 160 × 157.08 = 25132.8 W, or 25.13 kW. Through the above physical formulas and unit conversions, the pre-processed spindle torque and speed signals are transformed into instantaneous power data that directly reflects the frictional energy input, providing a basis for subsequent energy integration.

[0043] After acquiring the instantaneous friction power data, the friction power input energy is then constructed by integrating over the running time. Since the friction power input energy is the total energy generated by friction during the spindle's operation, the integration interval is the continuous running time of the spindle. The trapezoidal integration method is used to calculate the instantaneous friction power within this interval. Specifically, instantaneous friction power data is collected at fixed time intervals (e.g., 1 second). The power values ​​at two adjacent time points are considered as the two bases of a trapezoid, and the time interval is considered as the height of the trapezoid. The area of ​​each trapezoid is calculated sequentially and accumulated to obtain the friction power input energy for that time period. For example, if the spindle runs continuously for 20 minutes (1200 seconds), the instantaneous friction power collected during this period fluctuates between 22kW and 26kW. After trapezoidal integration, the friction power input energy is approximately (22+26) / 2×1200=28800kJ, or 28.8MJ.

[0044] Through a series of consecutive steps, including interference removal from the spindle torque and speed signals, product calculation of torque and angular velocity, and time integration of instantaneous friction power, accurate friction power input energy was obtained, providing reliable basic data for subsequent calculation of equivalent wear energy by combining heat loss energy, acoustic emission energy, and abrasive particle generation potential energy.

[0045] Furthermore, step A200 in the method provided in this application embodiment includes:

[0046] A210: Data acquisition is performed using a flow sensor and inlet / outlet temperature sensors installed in the lubricating oil passage to establish lubricating oil flow data and temperature signals.

[0047] A220: Obtain the specific heat capacity and density data of the lubricating oil, and send the specific heat capacity, density data, flow rate data, and temperature signal to the instantaneous heat power loss model to calculate the instantaneous heat power.

[0048] A230: Integrate the instantaneous thermal power over time to establish the first heat loss energy.

[0049] A240: Uses infrared thermal imaging equipment to collect the temperature distribution of the spindle surface, performs additional heat flow analysis based on the temperature distribution of the spindle surface and the structural heat conduction model, and establishes the second heat loss energy.

[0050] A250: After compensating the first heat loss energy with the second heat loss energy, heat loss energy is established.

[0051] Optionally, data acquisition can begin with the lubricating oil. To accurately obtain information about the heat carried by the lubricating oil, a flow sensor and inlet / outlet temperature sensors can be fixedly installed at key nodes in the lubricating oil passage of the spindle. The flow sensor captures the dynamic flow data of the lubricating oil in real time, while the inlet / outlet temperature sensors monitor the temperature changes of the lubricating oil as it enters and exits the spindle, thus establishing a complete lubricating oil flow data and temperature signal system. For example, in the lubricating oil passage of a certain type of gantry milling machine, a turbine-type flow sensor with a range of 0-50 L / min and an accuracy of ±1% is selected, and a PT100 platinum resistance sensor with an accuracy of ±0.5℃ is used for the temperature sensor. Both collect data at a frequency of 1 time / second to ensure that the heat transfer status of the lubricating oil can be reflected in real time.

[0052] Next, after obtaining the flow rate and temperature signal of the lubricating oil, the instantaneous thermal power is further calculated to quantify the heat carried away by the lubricating oil per unit time. First, the inherent physical parameters of the lubricating oil need to be determined, including its specific heat capacity and density. These parameters can be obtained from the oil specification manual. For example, the commonly used No. 46 spindle lubricating oil has a specific heat capacity of approximately 2.1 kJ / (kg·℃) and a density of approximately 0.89 g / cm³. 3 .

[0053] Subsequently, the specific heat capacity and density data, along with the lubricating oil flow rate data and inlet / outlet temperature signals collected in the previous steps, are input into the instantaneous heat power loss model. The inlet / outlet temperature signals need to be calculated to determine the temperature difference; for example, if the inlet temperature is 35℃ and the outlet temperature is 42℃, the temperature difference is 7℃. The instantaneous heat power loss model is based on the fundamental formula of heat transfer. The calculation is performed using ρ×Q×c×ΔT, where... Let ρ be the instantaneous heat power, ρ be the density of the lubricating oil, Q be the flow rate of the lubricating oil, c be the specific heat capacity of the lubricating oil, and ΔT be the temperature difference between the inlet and outlet of the lubricating oil. Taking a certain operating condition as an example, when the flow rate of the lubricating oil is 20 L / min, it can be converted to 0.00033 m³ / min. 3 / s, substituting into the formula, the instantaneous heat power is approximately 890 kg / m³. 3 ×0.00033m 3 / s×2100J / (kg·℃)×7℃≈4327W, that is, through this step, the physical parameters of the lubricating oil and the collected data are converted into quantifiable instantaneous thermal power data.

[0054] Furthermore, the instantaneous heat power dissipation model is constructed based on the heat transfer characteristics of lubricating oil in the spindle system. First, the core input parameters of the model are defined: the specific heat capacity, density, flow rate data, and temperature signal of the lubricating oil. Then, based on the fundamental principles of heat transfer, the volumetric flow rate data is converted to mass flow rate by combining the lubricating oil density: mass flow rate = volumetric flow rate × density. Next, the core calculation relationship of instantaneous heat power = mass flow rate × specific heat capacity × inlet and outlet temperature difference is constructed, ensuring that the model can directly quantify the heat carried away by the lubricating oil from the spindle system per unit time through the input parameters. Finally, through verification and calibration, combined with known heat power data of the same model of gantry milling machine spindle under standard operating conditions, the model calculation results are corrected to ensure that the model can accurately output instantaneous heat power matching the actual heat dissipation under different loads and speeds, thus completing the construction of the instantaneous heat power dissipation model.

[0055] Next, after obtaining the instantaneous thermal power, the total heat carried away by the lubricating oil over a period of time is calculated using time integration, thus establishing the first heat loss energy. Since the instantaneous thermal power fluctuates with changes in spindle load and speed, the integration interval must be the continuous operating time of the spindle, and the trapezoidal integration method is still used for calculation. Specifically, instantaneous thermal power data is extracted at fixed time intervals, such as 1 second. The instantaneous thermal power at two adjacent time points is considered as the two bases of a trapezoid, and the time interval is considered as the height of the trapezoid. The area of ​​each small trapezoid is calculated sequentially and accumulated to finally obtain the total heat within that interval. For example, if the spindle runs continuously for 30 minutes, or 1800 seconds, during which the instantaneous thermal power fluctuates between 4000W and 4500W, the first heat loss energy obtained after trapezoidal integration is approximately (4000W + 4500W) / 2 × 1800s = 7.65 × 10^6 J, or 7.65 MJ. This first heat loss energy only represents the heat lost through the lubricating oil and does not yet include the heat lost directly from the spindle surface.

[0056] Next, to compensate for heat loss from the spindle surface, an infrared thermal imaging device is used to collect the temperature distribution on the spindle surface and establish a second heat loss energy. An infrared thermal imager with a resolution of at least 640×512 can be selected to collect the temperature of the entire spindle surface, especially areas prone to heat generation such as bearings and the connection between the spindle and the tool, and obtain real-time temperature values ​​for different areas, such as 45℃ for the bearing area, 38℃ for the middle of the spindle, and 25℃ for the ambient temperature.

[0057] The temperature distribution data is then substituted into the spindle structure heat conduction model. This model considers the spindle material properties, including its specific material type and thermal conductivity; the spindle structural dimensions, such as diameter and length; and the environmental heat exchange coefficient. It then calculates the additional heat flux dissipated from the spindle surface to the surrounding environment through heat conduction and radiation. Integrating this additional heat flux over operating time yields the second heat loss energy. For example, under a certain operating condition, the calculated second heat loss energy is approximately 0.85 MJ. This energy represents the heat lost by the spindle directly to the environment without the lubricating oil.

[0058] The construction of the spindle structure heat conduction model is based on the actual structure and heat transfer characteristics of the gantry milling machine spindle. The specific steps are as follows:

[0059] Step a: First, obtain the basic parameters of the spindle. Determine its structural dimensions through the spindle design drawings, including spindle diameter, length, bearing installation position and fitting clearance, etc.; geometric shape, such as cylindrical spindle body, stepped connecting section, etc., and look up the thermal properties of the material used in the spindle, including thermal conductivity, specific heat capacity, density, etc. For example, the thermal conductivity of 40Cr steel is about 48W / (m·K)). At the same time, record the spindle surface temperature distribution data collected by the infrared thermal imaging equipment and use it as the boundary temperature input of the model.

[0060] Step b: Based on Fourier's law of heat conduction and combined with the three-dimensional structural characteristics of the main shaft, establish the mathematical equation for heat conduction inside the main shaft, and clarify the heat conduction path and calculation relationship in the radial direction of the main shaft from the shaft center to the surface and along the length of the main shaft.

[0061] Step c: Introduce heat transfer boundary conditions, including the convective heat transfer coefficient between the spindle surface and the surrounding environment. This can be determined by considering factors such as wind speed and temperature in the machine tool's working environment. For example, the convective heat transfer coefficient in a windless environment is approximately 5-10 W / (m²). 2 ·K); The surface thermal radiation coefficient can be determined based on the emissivity of the spindle surface material, such as the emissivity of a metal surface, which is approximately 0.1-0.3.

[0062] Step d: Substitute the above boundary conditions and basic parameters of the spindle into the mathematical equation for heat conduction inside the spindle, and then use the finite element numerical method to discretize the spindle structure, dividing the spindle into multiple grid units, and solve the heat flow distribution of each grid unit through numerical calculation.

[0063] Step e: Using known heat loss data of the same model spindle under standard working conditions, calibrate and correct the model calculation results to ensure that the model can accurately calculate the additional heat flow dissipated by the spindle to the environment through heat conduction, heat convection and heat radiation based on the input spindle surface temperature distribution, and finally complete the construction of the spindle structure heat conduction model.

[0064] Finally, the first heat loss energy is compensated by the second heat loss energy. Since the first heat loss energy only reflects the heat carried away by the lubricating oil, while the second heat loss energy reflects the heat directly lost from the spindle surface, both together constitute the total heat loss generated by friction during spindle operation. Therefore, adding the second heat loss energy to the first heat loss energy completes the compensation process. For example, adding the 7.65 MJ of the first heat loss energy in the previous example to the 0.85 MJ of the second heat loss energy yields a final heat loss energy of 8.5 MJ, ensuring that this energy data comprehensively reflects the heat loss generated by spindle friction.

[0065] By collecting data from lubricating oil passage sensors, calculating instantaneous heat power, and integrating over time to obtain the first heat loss energy, and combining this with infrared thermal imaging and a structural heat conduction model to obtain the second heat loss energy and perform compensation, an accurate and comprehensive heat loss energy was obtained. This provides reliable data support for eliminating non-wear energy interference when calculating the equivalent wear energy in the subsequent process.

[0066] Furthermore, step A200 in the method provided in this application embodiment includes:

[0067] A260: An acoustic emission sensor array is arranged in the friction surface area to collect transient acoustic emission signals, which include amplitude, duration and frequency characteristics.

[0068] A270: Perform time integration of the transient acoustic energy on the acoustic emission signal to calculate the transient energy release.

[0069] A280: Performs the accumulation of transient energy release at multiple measurement points to establish acoustic emission energy.

[0070] Specifically, the first step is to rationally arrange and acquire signals from the acoustic emission sensor array. For the characteristics of the spindle friction surfaces, such as the mating friction area between the spindle and the bearing, and the connecting friction area between the spindle and the tool, piezoelectric acoustic emission sensors are usually selected to construct the array. These sensors are highly sensitive to transient elastic waves generated by mechanical vibration, and their frequency response range can cover 100kHz-1MHz. They can accurately capture transient signals generated by the propagation of microcracks on the friction surface and the adhesion and peeling of materials during the wear process.

[0071] Next, multiple piezoelectric acoustic emission sensors can be evenly distributed along the circumference or axis of the friction surface during setup. For example, four sensors can be arranged at 90° intervals around the circumference of the spindle bearing friction area to ensure that the detection range of each sensor covers a local area of ​​the friction surface, avoiding signal omissions due to single-point acquisition. After acquisition is started, the piezoelectric acoustic emission sensors will synchronously capture transient acoustic emission signals at a sampling rate of not less than 10MS / s, and record the signal amplitude, which reflects the energy release intensity, usually in the range of 50-500μV; the duration, which reflects the time span of energy release, generally 10-100μs; and the frequency characteristics, where different wear forms correspond to different frequencies. For example, the frequency of adhesive wear signals is mostly concentrated in 200-300kHz, while that of abrasive wear is in 300-500kHz, thus forming a complete set of transient acoustic emission signals.

[0072] Then, the transient energy release at a single measuring point is calculated using time integration. Since the amplitude of the acoustic emission signal changes dynamically with time, directly reading the amplitude cannot accurately quantify the energy. The original signal can be preprocessed to remove low-frequency noise caused by environmental electromagnetic interference. Then, based on the principle of acoustic energy calculation and combined with the sensor sensitivity, the voltage amplitude of the transient acoustic emission signal is converted into a sound pressure signal. For example, the sensitivity of a commonly used sensor is 10mV / Pa.

[0073] The square of the sound pressure signal is then integrated over the signal duration, and the acoustic impedance parameters of the medium are substituted, such as the acoustic impedance of a steel spindle being approximately 45 × 10^6 kg / (m). 2 The energy release of a single transient signal at a single measuring point is obtained by calculating the energy release of the signal at a specific measuring point. Taking a certain measuring point as an example, the amplitude of the preprocessed acoustic emission signal is 200 μV and the duration is 50 μs. The transient energy release is obtained by combining the sensor sensitivity and the acoustic impedance of the steel, which accurately reflects the energy scale of a single micro-wear event on the friction surface at that measuring point.

[0074] Finally, the transient energy release values ​​at all measuring points are summed to establish an acoustic emission energy covering the entire friction surface. Because wear events may occur simultaneously in different areas of the spindle friction surface, a single measuring point can only capture localized energy release. Relying solely on single-point data would underestimate the overall acoustic emission energy, leading to biased subsequent wear assessments. In practice, the transient energy release values ​​at all measuring points are summarized at a fixed time period, such as 1 second. The energy values ​​at each measuring point are then summed sequentially to obtain the total acoustic emission energy of the friction surface within that period, comprehensively reflecting the total transient energy release of the entire friction surface during that time period.

[0075] By arranging an array of acoustic emission sensors in the friction surface area to collect multi-parameter transient signals, performing time integration on the signals to calculate the transient energy release at a single point, and accumulating the energy from multiple measurement points, an accurate acoustic emission energy covering the entire friction surface was obtained. This provides key data support for subsequent calculations of equivalent wear energy by combining friction power input energy and heat loss energy, reflecting the dynamic wear state of the friction surface.

[0076] Furthermore, step A300 in the method provided in this application embodiment includes:

[0077] A310: Activate the online oil monitoring unit, perform time-series detection of the oil, and obtain the abrasive feature set of the lubricating oil, which includes the number of abrasive particles, size distribution, and material composition.

[0078] A320: Read the density and hardness parameters of the abrasive material based on the material composition, and calculate the energy required to generate a single abrasive grain based on the reading results, size distribution, and number of abrasive grains.

[0079] A330: Establish the abrasive generation potential based on the energy requirements of all collected abrasive particles.

[0080] Specifically, the first step is to activate the online oil monitoring unit and conduct time-series detection. The online oil monitoring unit typically integrates a laser particle counting and spectral analysis modules. The former is used to capture the number and size distribution of abrasive particles, while the latter is used to identify the material composition of the abrasive particles. The online oil monitoring unit extracts oil samples from the spindle lubrication oil circulation path at fixed intervals, such as every 5 minutes. The laser particle counting module records the number of abrasive particles in different size ranges through the laser beam blocking effect. For example, in a certain period, there are 60 abrasive particles of 0.5-2μm, 30 abrasive particles of 2-5μm, and 10 abrasive particles of 5-10μm. The spectral analysis module uses atomic emission spectroscopy to detect the elemental composition of the abrasive particles in the oil. If elements such as iron and chromium are detected, it can be determined that the abrasive particles originated from the spindle or bearing. Finally, an abrasive feature set containing the number, size distribution, and material composition of abrasive particles is formed.

[0081] Next, the energy required to generate a single abrasive grain is calculated. First, based on the material composition determined by the spectral analysis module, the density and hardness parameters of the corresponding abrasive material are retrieved from the existing material database. For example, when the abrasive grain is 40Cr steel, the density is approximately 7.8 g / cm³. 3 The Vickers hardness is approximately 250 HV. Then, considering the abrasive grain size distribution, the volume is calculated using the simplified model of approximately spherical abrasive grains, with the formula V = 4 / 3πr. 3 Where r is the abrasive grain radius. Based on the principle that the energy required to generate abrasive grains is positively correlated with the material's hardness and volume, the energy is calculated using the formula E=H×V, where H is the hardness and V is the volume. Substituting the data, the energy required to generate a single abrasive grain of that size can be obtained. Similarly, the energy required to generate a single abrasive grain of different sizes can be derived.

[0082] Finally, the abrasive grain generation potential is established by summing the energy requirements of all collected abrasive grains. Since the time-series detection covers a specific time period of spindle operation, it is necessary to calculate the total number and corresponding size of abrasive grains in all detection cycles within this time period, and then sum the energy requirements for generating each abrasive grain sequentially. For example, if 200 abrasive grains are collected in 1 hour, with the energy of each grain ranging from 1×10^-10 J to 5×10^-10 J, the summation yields the abrasive grain generation potential, which accurately reflects the total energy consumed by the spindle friction to generate abrasive grains.

[0083] By activating the online oil monitoring unit to obtain the abrasive feature set, combining material parameters to calculate the energy required for the generation of a single abrasive, and accumulating the energy of all abrasive particles, an accurate abrasive generation potential was obtained. This provides key data support for subsequent calculations of equivalent wear energy by combining friction power input energy, heat loss energy, and acoustic emission energy, reflecting the energy consumption of abrasive generation.

[0084] Furthermore, step A300 in the method provided in this application embodiment includes:

[0085] A340: Acquire historical experimental data and external calibration data of the same model, wherein the historical experimental data and the external calibration data of the same model are a mapping dataset of equivalent wear energy and wear state.

[0086] A350: After performing interactive authentication of the historical experimental data and the external calibration data of the same model, a true mapping table of equivalent wear energy and wear state is established.

[0087] A360: The first wear detection result is established using the real mapping table and the equivalent wear energy.

[0088] Specifically, firstly, historical experimental data and external calibration data of the same model are acquired. These two types of data together constitute a mapping dataset of equivalent wear energy and wear state. The historical experimental data usually comes from the test records of the same gantry milling machine spindle in different operating cycles. For example, through the previous 10 complete running experiments, the equivalent wear energy was recorded as 18MJ, 22MJ, and 28MJ. The corresponding spindle wear state was determined to be slight wear, moderate wear, and heavy wear after disassembly and testing, forming multiple sets of equivalent wear energy-wear state mapping data. The external calibration data of the same model comes from the industry standard test or third-party laboratory test results of the same model spindle. For example, in the calibration data of the same model spindle provided by a machine tool manufacturer, the equivalent wear energy of 15-20MJ corresponds to slight wear, 20-30MJ corresponds to moderate wear, and above 30MJ corresponds to heavy wear. The combination of the two types of data can cover the mapping relationship under more working conditions and avoid the limitations of a single data source.

[0089] After obtaining the two types of mapping datasets mentioned above, interactive authentication is then performed to ensure data reliability, thereby establishing a true mapping table. Since historical experimental data may be affected by equipment aging and fluctuations in experimental conditions, and external calibration data of the same model may have differences in testing environments, such as different temperatures and loads, direct merging can easily lead to mapping deviations. Interactive authentication can be performed using consistency checks and cross-validation. The specific steps and examples are as follows:

[0090] First, compare the consistency of the equivalent wear energy-wear state correspondence between the two types of data. For example, filter out the overlapping parts of the energy range, such as 18-28MJ. If 22MJ corresponds to moderate wear in historical experimental data and 22MJ also corresponds to moderate wear in external calibration data of the same model, then the data set is considered consistent. If there are differences, such as 25MJ corresponding to moderate wear in historical experimental data and 25MJ corresponding to moderate to heavy wear in external calibration data of the same model, then it is necessary to backtrack and analyze the operating conditions. If the operating conditions of historical experimental data are 2000 r / min speed and 150 N·m load, and the operating conditions of external calibration data of the same model are 3000 r / min speed and 200 N·m load, then the two types of data need to be standardized according to the operating conditions, such as converting to equivalent energy under standard operating conditions, correcting and then verifying consistency. After all data passes interactive authentication, it is organized according to the equivalent wear energy range to establish a clear real mapping table. For example, 15-20MJ: slight wear, 20-30MJ: moderate wear, and above 30MJ: heavy wear.

[0091] Finally, by using the actual mapping table and the currently calculated equivalent wear energy, the first wear detection result can be determined. Specifically, first, the equivalent wear energy calculated in the current spindle detection is extracted, assuming it is 24MJ. Then, the range to which this energy belongs is found in the actual mapping table. According to the actual mapping table in the example above, 24MJ falls within the 20-30MJ range, corresponding to a moderate wear state. Thus, the first wear detection result is directly established. If the equivalent wear energy is 20MJ and falls within the range boundary, then the majority of the boundary energy states in historical experimental data are considered. For example, if 20MJ has been judged as slightly to moderate wear multiple times in historical experimental data, the result is refined to slightly to moderate wear to ensure the accuracy of the detection result.

[0092] By acquiring the equivalent wear energy-wear state mapping dataset, performing interactive authentication to establish a real mapping table, and matching the current equivalent energy to determine the wear state, the first wear detection result based on energy parameters was obtained, providing a reliable quantitative basis for subsequent interactive authentication with the second ultrasonic monitoring result.

[0093] Furthermore, step A400 in the method provided in this application embodiment includes:

[0094] A410: After activating the ultrasonic sensor, perform ultrasonic signal acquisition and establish an ultrasonic feature set, which includes propagation time, amplitude, waveform characteristics, and spectral characteristics.

[0095] A420: Perform noise suppression and normalization on the ultrasonic feature set.

[0096] A430: Extract key feature parameters from the normalized ultrasonic feature set to establish a key feature set, which includes wave velocity variation features, echo delay features, radiation intensity attenuation features, and spectrum center shift features.

[0097] A440: Based on the wave velocity variation characteristics and echo delay characteristics in the key feature set, identify microcracks and local pitting of the main shaft, and establish the first identification result.

[0098] A450: Based on the radiation intensity attenuation characteristics and spectrum center shift characteristics in the key feature set, wear status is identified, and a second identification result is established.

[0099] A460: Use the first identification result and the second identification result to establish a second wear detection result.

[0100] In one embodiment, sensor deployment and ultrasonic signal acquisition are first performed. A piezoelectric ultrasonic sensor with a frequency response range of 0.5-10MHz is selected, suitable for the detection of metal components. It is placed at key locations along the axial and radial directions of the spindle, such as bearing mating sections and easily worn areas like the spindle journal, to ensure that the ultrasonic waves can penetrate critical parts of the spindle. After the ultrasonic sensor is activated, ultrasonic waves are emitted to the spindle through a pulse emission circuit. Simultaneously, echo signals propagating inside the spindle and reflected from defects are received. The propagation time of the signal, i.e., the time from emission to reception, amplitude, waveform characteristics (i.e., the wave pattern of the signal), are recorded. When there are no defects, the waveform is regular; when there are defects, noise appears. The spectral characteristics, i.e., the frequency distribution of the signal, are also recorded. Finally, these parameters are integrated to form an ultrasonic feature set.

[0101] After acquiring the ultrasonic feature set, noise suppression and normalization are performed to eliminate interference and unify the data magnitude. Mechanical vibration and electromagnetic interference during spindle operation can cause noise to be mixed in the feature set. Wavelet transform denoising is used to decompose the signal into different frequency scales and remove high-frequency noise components, making the waveform smoother and the amplitude more stable. Normalization transforms the feature parameters of different dimensions in the ultrasonic feature set to a unified range of 0-1. For example, the linear normalization formula X'=(X-Xmin) / (Xmax-Xmin) is used. For example, for the propagation time of 68-72μs, the propagation time of 68μs is converted to 0 and 72μs is converted to 1, so as to avoid the impact of parameter magnitude differences on the accuracy of subsequent feature extraction. After processing, a clean and standardized ultrasonic feature set is obtained.

[0102] Next, key feature parameters are extracted from the normalized ultrasonic feature set. For propagation time, the wave velocity variation feature is obtained by calculating the difference between the actual propagation time and the standard propagation time when the spindle is defect-free. Wave velocity = spindle thickness / propagation time. If the propagation time increases, the wave velocity decreases accordingly, forming the wave velocity variation feature. Based on the delay time of the echo signal relative to the normal echo, for example, the normal echo time is 10μs, the actual time is 12μs, and the delay is 2μs, an echo delay feature is established. By comparing the initial ultrasonic amplitude with the echo amplitude after propagation, the amplitude attenuation ratio is calculated. For example, if the initial amplitude is 5V and the echo amplitude is 1V, the attenuation is 80%, resulting in the radiation intensity attenuation feature. Fourier transform is performed on the spectral features to analyze the shift of the peak frequency of the spectrum. For example, the center frequency of the spectrum is 2MHz when there is no defect, and it shifts to 1.8MHz after wear, establishing the spectral center shift feature. Finally, these four types of parameters are integrated to form a key feature set.

[0103] Then, based on the wave velocity variation characteristics and echo delay characteristics in the key feature set, microcracks and localized spalling pits in the spindle can be identified. When microcracks or localized pits exist inside the spindle, ultrasonic waves will be reflected and refracted at the defect, resulting in a decrease in wave velocity and an extension of the echo propagation path. By setting thresholds: if the wave velocity variation exceeds 50 m / s and the echo delay time exceeds 1.5 μs, it is determined that there is a microcrack in the spindle; if the echo delay shows a local abrupt change, such as an echo delay of 3 μs in a certain area while other areas are normal, it is determined that there is a localized spalling pit. Combining these judgment results, the type and location of spindle defects are clarified, and the first identification result is established. For example, there is a microcrack of about 0.5 mm in length in the bearing mating section of the spindle, without obvious localized pits.

[0104] Subsequently, based on the radiation intensity attenuation characteristics and spectral center shift characteristics in the key feature set, the wear condition of the spindle can be identified. Wear on the surface or inside the spindle increases the propagation resistance of ultrasonic waves, leading to a greater attenuation of radiation intensity. At the same time, wear changes the uniformity of the local material of the spindle, causing the center frequency of the ultrasonic spectrum to shift. For example, when the radiation intensity attenuation ratio exceeds 25% (e.g., initial amplitude 5V, echo amplitude below 3.75V), and the spectral center frequency shift exceeds 0.15MHz (e.g., standard 2MHz, actual below 1.85MHz), it is judged as mild wear. If the attenuation ratio exceeds 40% and the spectral shift exceeds 0.3MHz, it is judged as moderate wear. Through this kind of quantitative judgment, the degree of wear of the spindle is clarified, and a second identification result is established. For example, the overall wear condition of the spindle is moderate wear, and the wear at the bearing mating area is more obvious.

[0105] Finally, the first identification result of defect identification and the second identification result of wear status are fused together to remove contradictory information between the two types of results. If the first identification result shows no defect but the second identification result shows severe wear, the feature data needs to be reviewed. After confirming that there are no abnormalities, the wear status is taken as the standard, and consistent judgment conclusions are integrated to finally establish the second wear detection result.

[0106] Through a series of steps including ultrasonic signal acquisition and feature set establishment, noise suppression and normalization, key feature parameter extraction, and separate identification and fusion of defects and wear states, a second wear detection result containing spindle defects and wear states was obtained. This provides a precise basis for microscopic state determination for subsequent interactive authentication with the first wear detection result and for improving the overall reliability of wear resistance detection.

[0107] Furthermore, step A500 in the method provided in this application embodiment includes:

[0108] A510: Converts the first and second wear detection results into a unified numerical wear index.

[0109] A520: Determine whether the first and second wear detection results after conversion are within the same wear level range.

[0110] A530: If the wear levels are within the same range, the first wear test result and the second wear test result will be used as verification results for each other, and the current wear level will be output as the wear resistance test result.

[0111] Optionally, interactive authentication of the first and second wear detection results can be performed. The first step is to convert the two types of results into a unified numerical wear index. This is because the two detection methods have different judgment dimensions. The first method is based on equivalent wear energy and outputs the wear state related to macroscopic energy consumption. The second method is based on the identification of microscopic defects and surface conditions by ultrasonic monitoring and outputs a wear description containing local defect information. Direct comparison is prone to judgment bias due to the different expression forms.

[0112] Specifically, a unified conversion is achieved based on a preset wear characteristic-numerical mapping rule. For example, an industry-standard numerical system is established: slight wear corresponds to 1-3, moderate wear corresponds to 4-6, and heavy wear corresponds to 7-9. At the same time, the correspondence is refined by combining the core parameters of the two types of detection. In the first path, the equivalent wear energy of 20-30MJ corresponds to the value 5 for moderate wear; in the second path, the judgment result of moderate wear directly corresponds to the value 5. Through this rule, the wear detection results of the two paths are converted into directly comparable quantitative indicators, avoiding interference from differences in expression in the certification, and enabling both to enter a unified numerical system.

[0113] After completing the numerical conversion, it is determined whether the two wear detection results fall within the same wear level range. The division of the wear level range needs to be determined in conjunction with the industry operating standards of the gantry milling machine spindle, historical detection data of the same model of equipment, and safe wear thresholds. For example, the following can be set: the range for slight wear level is 1-3, corresponding to an equivalent wear energy of 15-20MJ, no obvious defects detected by ultrasonic monitoring, and wear amount <0.1mm; the range for moderate wear level is 4-6, corresponding to an equivalent wear energy of 20-30MJ, possible slight micro-cracks detected by ultrasonic monitoring, and wear amount 0.1-0.3mm; the range for severe wear level is 7-9, corresponding to an equivalent wear energy >30MJ, obvious peeling or cracks detected by ultrasonic monitoring, and wear amount >0.3mm. During the judgment, the numerical indicators after conversion of the two wear detection results are extracted, and their respective ranges are checked. If they both fall within the same range, the result consistency requirement is met; if they belong to different ranges, for example, the first wear detection result value of 3 indicates slight wear, and the second wear detection result value of 4 indicates moderate wear, then a result deviation is determined. For example, in the aforementioned tests, the first wear test result value 4 and the second wear test result value 5 are both within the moderate wear level range of 4-6, and are therefore determined to be within the same wear level range.

[0114] Next, when the two wear detection results fall within the same wear level range, they need to be used as verification results to output the final wear resistance test result. The first detection is based on the energy balance principle, reflecting the macroscopic energy consumption accumulation state of spindle wear; the second detection is based on ultrasonic propagation characteristics, reflecting the microscopic structural changes of spindle wear. The two types of detection characterize the wear situation from different physical dimensions. Consistent results mean that the wear state at the macroscopic and microscopic levels matches each other, eliminating misjudgments caused by interference from a single detection, including energy calculation errors and ultrasonic signal noise. In specific operation, after confirming that the wear levels are consistent, the evidence from the two wear detection results is integrated to form a complete conclusion. For example, combining the equivalent wear energy of 23MJ with the ultrasonic monitoring results, the spindle has no serious defects, and the overall wear level is moderate. The two wear detection results are used to verify each other, and the moderate wear is finally output as the wear resistance test result.

[0115] By converting two different wear detection results into a unified numerical index, determining whether they are within the same wear level range, and verifying each other when the results are consistent to output a conclusion, the limitations of single detection are effectively avoided, and the reliability and accuracy of wear resistance test results for the gantry milling machine spindle system are improved.

[0116] Furthermore, step A520 in the method provided in this application embodiment includes:

[0117] A521: If they are not within the same wear level range, an anomaly analysis command will be triggered.

[0118] A522: After executing the redundancy verification acquisition according to the anomaly analysis instruction, the redundancy verification acquisition results, the first wear detection results, and the second wear detection results are weighted and certified to reconstruct the wear resistance test results.

[0119] Optionally, when determining whether the first and second channel wear detection results after conversion are within the same wear level range, if they belong to different levels, the spindle system will automatically trigger an anomaly analysis command. The core function of this command is to mark the current detection results as contradictory, avoiding direct output of potentially biased conclusions, and simultaneously initiating subsequent redundant verification processes to investigate the cause of the contradiction. The anomaly analysis command will synchronously record the contradiction points between the two channel results. For example, the first channel may determine slight wear based on energy, while the second channel may determine moderate wear based on ultrasonic defects, providing a targeted direction for subsequent redundant acquisition and ensuring that the verification process can accurately supplement key data rather than blindly repeating the detection. For instance, in a certain detection scenario, the first channel's equivalent wear energy of 19MJ is converted to a value of 3, outputting slight wear, while the second channel's ultrasonic monitoring detects a wear amount of 0.12mm, which is converted to a value of 4, outputting moderate wear. The two levels are inconsistent, and an anomaly analysis command is triggered to record the contradiction point where there is a deviation between the energy calculation and the wear measurement.

[0120] Next, redundant verification data acquisition is performed according to the anomaly analysis instructions. Key parameters that can supplement or correct the original data are selected for secondary testing to address contradictions, ensuring the relevance and reliability of the redundant data. The redundant acquisition methods include: re-acquiring spindle torque and speed signals to verify the friction power input energy, which forms the basis for energy calculation in verifying the first-path wear detection results; extracting the lubricating oil abrasive particle feature set to confirm the abrasive particle generation potential, which can help correct the equivalent wear energy; and adjusting the ultrasonic sensor sampling frequency to re-acquire signals to verify the wear amount and defect state, used to verify the microscopic identification basis of the second-path wear detection results. For example, in the aforementioned contradictory scenario, the equivalent wear energy was recalculated during redundant acquisition. A 2MJ error was found in the original calculation of heat loss energy; after correction, the equivalent wear energy was 21MJ, converted to a value of 4 for moderate wear. Simultaneously, secondary ultrasonic testing confirmed a wear amount of 0.11mm, also converted to a value of 4 for moderate wear. Both sets of redundant data point to a moderate wear level, providing a reliable basis for subsequent weighted authentication.

[0121] After completing the redundant verification data acquisition, the redundant verification data acquisition results need to be weighted and certified with the original two-channel test results to finally reconstruct the wear resistance test results. The core of the weighted certification is to assign weights based on the historical accuracy of different test methods and their adaptability to the current test scenario. Specifically, the redundant verification data acquisition results, because they are specifically designed for testing the contradiction points, have the highest accuracy and are assigned the highest weight of 0.4. The weights of the first and second channel wear test results are determined based on their average accuracy in the testing of the same type of spindle. For example, if the first channel has an accuracy of 88%, the weight is 0.3; and the second channel has an accuracy of 85%, the weight is also 0.3. Then, the weighted sum is calculated using the formula: Weighted value = First channel index × First channel weight + Second channel index × Second channel weight + Redundant result index × Redundant weight. Finally, the final result is determined according to the range of the weighted value. For example, the original first-path index is 3×0.3=0.9, the original second-path index is 4×0.3=1.2, the redundant result index is 4×0.4=1.6, and the weighted sum is 3.7, which falls within the reasonable calculation deviation of the moderate wear level range 4-6. Since the numerical index is an integer, the weighted value is allowed to have a deviation of ±0.5. The final reconstructed wear resistance test result is moderate wear.

[0122] By triggering anomaly analysis commands when the two detection results are inconsistent, performing targeted redundant verification collection, and combining weights to perform weighted authentication on multiple sets of results, the problem of being unable to determine a single contradictory result is effectively solved. In the end, a more accurate and reliable wear resistance test result is reconstructed, ensuring the accuracy of wear condition determination of the gantry milling machine spindle system.

[0123] In summary, the wear resistance testing method for gantry milling machine spindle systems provided in this application has the following technical effects:

[0124] This application acquires torque and speed signals during the operation of a gantry milling machine spindle. After high-frequency noise and low-frequency drift removal, torque-angular velocity product calculation, and instantaneous friction power time integration, the friction power input energy is obtained. Simultaneously, the spindle and lubricating oil feature sets are acquired to establish heat dissipation energy, acoustic emission energy is calculated using an acoustic emission array, and abrasive particle generation potential is established through oil detection. The equivalent wear energy is then calculated based on the difference between the friction power input energy and the aforementioned energy parameters. Combined with ultrasonic monitoring, two-channel wear detection results are established and cross-certified. This accurately detects the wear resistance of the gantry milling machine spindle system, making the wear resistance test results more precise and reliable. It achieves multi-dimensional, real-time, and comprehensive detection of the wear resistance of the gantry milling machine spindle system, improving the comprehensiveness, accuracy, and reliability of the detection data.

[0125] Example 2, as Figure 2 As shown, based on the same inventive concept as in Embodiment 1 above, this application provides an electronic device, the electronic device comprising:

[0126] The memory 303 is used to store executable instructions; the processor 302 is used to execute the executable instructions stored in the memory 303 to implement a wear resistance testing method for the spindle system of a gantry milling machine.

[0127] Figure 2 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention, showing a block diagram of an exemplary electronic device suitable for implementing the embodiments of the present invention. Figure 2 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments of the present invention. This electronic device is in the form of a general-purpose computing device, and its components may include, but are not limited to, an input device 301, a processor 302, a memory 303, and an output device 304. The processor 302 may be one or more; the memory 303 may include a computer-readable medium and at least one program product having a set (at least one) of program modules configured to perform the functions of the embodiments of this application.

[0128] The memory 303 shown in this embodiment of the invention can be any combination of one or more computer-readable media. The computer-readable storage medium can be, but is not limited to, infrared, semiconductor systems, devices or components, or any combination thereof, used to store software programs, computer-executable programs and modules, such as the program instructions / modules corresponding to the wear resistance testing method for the spindle system of a gantry milling machine in this embodiment of the invention. The processor 302 executes various functional applications and data processing of the computer device by running the software programs, instructions and modules stored in the memory 303, thereby realizing the above-mentioned wear resistance testing method for the spindle system of a gantry milling machine.

[0129] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0130] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of this application and its equivalents, this application also intends to include such modifications and variations.

Claims

1. A method for testing the wear resistance of a gantry milling machine spindle system, characterized in that, The method includes: While the spindle is running, the spindle torque and speed signals are acquired, and the friction power input energy is calculated based on the spindle torque and speed signals. Synchronously acquire spindle and lubricating oil feature sets, establish heat dissipation energy, use acoustic emission array to execute transient energy release signals of friction surfaces, and calculate acoustic emission energy; Perform oil level detection, establish abrasive particle generation potential, calculate equivalent wear energy based on the difference between the friction power input energy, heat loss energy, acoustic emission energy, and abrasive particle generation potential, and establish the first path wear detection result based on the equivalent wear energy; Activate the ultrasonic sensor to perform ultrasonic monitoring and identification of the spindle, and establish a second path of wear detection results; Perform interactive authentication of the first wear detection result and the second wear detection result, and output the wear resistance test result; Synchronously acquire spindle and lubricating oil feature sets to establish heat dissipation energy, including: Data acquisition is performed using flow sensors and inlet / outlet temperature sensors installed in the lubricating oil passage to establish lubricating oil flow data and temperature signals; Acquire the specific heat capacity and density data of the lubricating oil, and send the specific heat capacity, density data, flow rate data, and temperature signal to the instantaneous heat power dissipation model to calculate the instantaneous heat power; Integrate the instantaneous thermal power over time to establish the first heat loss energy; Infrared thermal imaging equipment is used to collect the temperature distribution on the spindle surface. Based on the temperature distribution on the spindle surface and the structural heat conduction model, additional heat flow analysis is performed to establish the second heat loss energy. After compensating the first heat loss energy with the second heat loss energy, a heat loss energy is established; The transient energy release signal of the friction surface is executed using an acoustic emission array, and the acoustic emission energy is calculated, including: An array of acoustic emission sensors is arranged in the friction surface region, and transient acoustic emission signals are collected using the acoustic emission sensor array. The acoustic emission signals include amplitude, duration and frequency characteristics. The transient acoustic energy of the acoustic emission signal is integrated over time to calculate the amount of transient energy released; Accumulate the transient energy release at multiple measurement points to establish acoustic emission energy; Perform oil fluid testing to establish abrasive generation potential, including: Activate the online oil monitoring unit to perform time-series detection of the oil and obtain the abrasive feature set of the lubricating oil, which includes the number of abrasive particles, size distribution, and material composition. Based on the material composition, the density and hardness parameters of the abrasive material are read, and the energy required to generate a single abrasive grain is calculated based on the reading results, size distribution, and number of abrasive grains. The abrasive generation potential is established based on the energy requirements of all collected abrasive particles. The first path of wear detection results is established based on the equivalent wear energy, including: Acquire historical experimental data and external calibration data of the same model, wherein the historical experimental data and the external calibration data of the same model are mapping datasets of equivalent wear energy and wear state; After performing interactive authentication of the historical experimental data and the external calibration data of the same model, a true mapping table of equivalent wear energy and wear state is established. The first wear detection result is established using the real mapping table and the equivalent wear energy.

2. The wear resistance testing method for a gantry milling machine spindle system as described in claim 1, characterized in that, The activated ultrasonic sensor performs ultrasonic monitoring and identification of the spindle, establishing a second path of wear detection results, including: After activating the ultrasonic sensor, ultrasonic signal acquisition is performed to establish an ultrasonic feature set, which includes propagation time, amplitude, waveform characteristics, and spectral characteristics. The ultrasonic feature set is subjected to noise suppression and normalization. The normalized ultrasonic feature set is processed by extracting key feature parameters to establish a key feature set, which includes wave velocity variation features, echo delay features, radiation intensity attenuation features, and spectrum center shift features. Based on the wave velocity variation characteristics and echo delay characteristics in the key feature set, microcracks and localized spalling pits on the main shaft are identified, and a first identification result is established. Wear status is identified based on the radiation intensity attenuation characteristics and spectral center shift characteristics in the key feature set, and a second identification result is established. A second wear detection result is established using the first identification result and the second identification result.

3. The wear resistance testing method for a gantry milling machine spindle system as described in claim 1, characterized in that, The calculation of friction power input energy based on the spindle torque and speed signals includes: High-frequency noise and low-frequency drift are removed from the spindle torque and speed signals; The instantaneous friction power is calculated by multiplying the torque in the spindle torque signal and the angular velocity in the speed signal. Perform the runtime integral of the instantaneous frictional power to construct the frictional power input energy.

4. The wear resistance testing method for a gantry milling machine spindle system as described in claim 1, characterized in that, The process of performing interactive authentication of the first wear detection result and the second wear detection result, and outputting the wear resistance test result, includes: The first and second wear detection results are converted into a unified numerical wear index. Determine whether the first and second wear detection results after conversion are within the same wear level range; If they are within the same wear level range, the first wear test result and the second wear test result will be used as verification results for each other, and the current wear level will be output as the wear resistance test result.

5. The wear resistance testing method for a gantry milling machine spindle system as described in claim 4, characterized in that, The determination of whether the converted first-path wear detection results and second-path wear detection results are within the same wear level range includes: If they are not within the same wear level range, an anomaly analysis command will be triggered; After executing the redundancy verification acquisition according to the anomaly analysis command, the redundancy verification acquisition results, the first wear detection results, and the second wear detection results are weighted and certified to reconstruct the wear resistance test results.

6. An electronic device, characterized in that, The electronic device includes: Memory, used to store executable instructions; The processor, when executing executable instructions stored in the memory, implements the wear resistance testing method for a gantry milling machine spindle system as described in any one of claims 1 to 5.

Images