A method for collecting a milling titanium alloy tool wear dataset
By collecting and analyzing vibration, sound, cutting force, and torque signals during the milling process of titanium alloys, as well as optical measurements, a full life-cycle tool wear dataset is generated. This solves the problem of unpredictable tool wear in titanium alloy milling and enables real-time monitoring and process optimization of wear conditions.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- QILU INST OF TECH
- Filing Date
- 2024-04-30
- Publication Date
- 2026-07-07
AI Technical Summary
The high hardness, low thermal conductivity, and chemical reactivity of titanium alloys make milling complex, resulting in high cutting forces and high cutting temperatures. This leads to severe chip adhesion, abrasive wear, adhesive wear, and diffusion wear, which affect workpiece quality. Existing technologies make it difficult to understand and effectively predict tool wear in real time.
Vibration, sound, cutting force, and torque signals are collected during the milling process of titanium alloys. Combined with optical instrument measurements, a wear dataset covering the entire tool life cycle is formed, including direct and indirect measurement data, for the prediction and diagnosis of tool wear.
It enables real-time understanding and effective prediction of tool wear, provides data support throughout the entire tool lifecycle, and helps diagnose faults and optimize titanium alloy milling processes.
Smart Images
Figure CN118357780B_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of tool wear prediction technology, specifically to a method for collecting wear datasets of milling titanium alloy tools. Background Technology
[0002] Titanium alloys refer to various alloys made of titanium and other metals. Titanium alloys are characterized by high strength, good corrosion resistance, and high heat resistance. The density of titanium alloys is generally around 4.51 g / cm³, only 60% of that of steel. Some high-strength titanium alloys even surpass the strength of many alloy structural steels. Therefore, the specific strength (strength / density) of titanium alloys is much greater than that of other metallic structural materials, allowing for the manufacture of components with high strength per unit area, good rigidity, and light weight. Consequently, titanium alloys are widely used in aerospace, medical, and automotive fields due to their high strength, low density, and excellent corrosion resistance.
[0003] However, the high hardness, low thermal conductivity, and chemical reactivity of titanium alloys make their milling process complex and challenging, classifying them as typical difficult-to-machine materials. Milling titanium alloys often involves high cutting forces and temperatures, leading to severe chip adhesion to the cutting tool. This is accompanied by abrasive wear, adhesive wear, and diffusion wear, resulting in significant tool wear. Severe tool wear negatively impacts the cutting process, consequently affecting workpiece quality (especially surface finish). Therefore, real-time monitoring and effective prediction of tool wear are crucial. Summary of the Invention
[0004] In order to solve the above-mentioned technical problems, this application proposes the following technical solution:
[0005] In a first aspect, embodiments of this application provide a method for collecting wear data of milling titanium alloy cutting tools, including:
[0006] Determine the titanium alloy cutting tool to be worn and the test platform for data acquisition;
[0007] The titanium alloy tool to be worn was worn on the test platform, and vibration, sound signals, cutting force, and torque signals were collected by sensors during the wear process to obtain the first dataset.
[0008] A second dataset was obtained by labeling the tool wear measurements using optical instruments, wherein the wear measurements were taken multiple times at different wear levels of the tool throughout the entire wear process.
[0009] The first dataset and the second dataset are combined to form the tool's full lifecycle data.
[0010] In one possible implementation, the determination of the wear-prone titanium alloy cutting tool and test platform for acquiring data includes:
[0011] The experiment used titanium alloy rods (Ti-6Al-4V) with a base area of 100 cm². 2 The diameter is approximately 11.3cm and the height is 40cm. The cutting tool is a coated four-flute end mill with a diameter of 10mm. The tool body is made of cemented carbide (YG3X), with WC of 96.5%, TaC (NbC) of 0.5%, and Co of 3%. The surface is coated with TiAlN, and the tip helix angle is 55°.
[0012] The test platform can be divided into three parts: the cutting area, the signal acquisition area, and the tool measurement area. The cutting equipment is a VDF-850 vertical machining center from Dalian Machine Tool Plant in China, with a maximum spindle speed of 8000 r / min and a maximum feed rate of 1700 mm / min. The machine tool is equipped with a tool clamping device, and a force sensor is installed on the tool clamping device. A DH5922D three-axis vibration acceleration sensor and a sound sensor are installed on the machine tool spindle.
[0013] In one possible implementation, the titanium alloy tool to be worn is mounted on the test platform and subjected to wear. During the wear process, sensors are used to collect vibration, sound signals, and cutting force and torque signals to obtain a first dataset, including:
[0014] The cutting force signal collected by the force sensor is transmitted to the computer using a charge amplifier. The Kistler signal acquisition system is used to collect cutting force and torque signal data in real time during the milling process.
[0015] Vibration and sound signals are collected using a DH5922D triaxial vibration accelerometer and sound sensor. The accelerometer and sound sensor are integrated into a dynamic signal acquisition system through an interface. The collected signals are projected onto a computer via a USB interface to display the waveforms of vibration and sound.
[0016] In one possible implementation, the step of using a charge amplifier to transmit the cutting force signal acquired by the force sensor to a personal computer, and using a Kistler signal acquisition system to acquire cutting force and torque signal data in real time during the milling process, includes:
[0017] The acquired signals were renamed and saved in .csv format;
[0018] Each time, the collected signal consists of 5 columns of data. The first column is time, and the other 4 columns are the cutting force and torque signals of the X, Y, and Z axes, respectively.
[0019] After smoothing the acquired signals using Origin 2021 software, images of the corresponding X, Y, and Z axis cutting force and torque signals can be obtained.
[0020] In one possible implementation, the DH5922D triaxial vibration accelerometer and sound sensor are used to collect vibration and sound signals. The accelerometer and sound sensor are integrated into a dynamic signal acquisition system via an interface. The collected signals are then transmitted to a computer via a USB interface to display the vibration and sound waveforms.
[0021] The acquired signals were renamed and saved in .csv format;
[0022] Each time, the collected signal consists of 6 columns of data. The first 2 columns are time, and the other 4 columns are the X, Y, and Z axis vibration signals and sound signals, respectively.
[0023] After smoothing the acquired signals using Origin 2021 software, images of the corresponding X, Y, and Z axis vibration signals and sound signals can be obtained.
[0024] In one possible implementation, the step of using optical instruments to annotate tool wear measurements to obtain a second dataset includes:
[0025] After each stage of the milling process is completed, the cutting tool is removed from the spindle;
[0026] The bottom edge and peripheral edge of the cutting tool were measured using a 19Jc universal tool microscope.
[0027] In one possible implementation, the measurement of the bottom cutting edge and peripheral cutting edge of the tool using a 19Jc universal tool microscope includes:
[0028] For the bottom cutting edge, the main measurement is its wear area. Value, maximum wear width VB The wear area is measured and marked using the closed cloud line function of QMS software, and the maximum wear width is marked using the straight line annotation function.
[0029] For the peripheral blade, the main measurement is the wear area. Value, maximum wear width VB Value and wear value at 1 / 2 cutting depth VB value.
[0030] In one possible implementation, for the maximum wear width VB Value and wear value at 1 / 2 cutting depth VB The value can be obtained using the distance annotation function in QMS3D-M software; for the wear area value... The measurement steps are as follows: First, before the tool wears, measure the triangular area of the circumferential cutting edge. S 1 Then, after the tool wears down, its total area (including the worn area and the area of the triangular region) is measured.S 2 Then use the total area S 2 Subtract the area of the triangular region S 1 The area of the wear zone of the peripheral cutting edge can then be obtained. value: Measure the four cutting edges of the end mill in sequence according to the above measurement method, and fill the data into the tool wear data set.
[0031] In one possible implementation, the peripheral cutting edge of the tool plays a major role in milling. Initial wear is defined as a VB value of 0-0.1 mm at the maximum wear width or half of the cutting depth of the tool peripheral cutting edge, normal wear is defined as a VB value of 0.1-0.5 mm, and severe wear is defined as a VB value of 0.1-0.5 mm. Wear exceeding 0.5 mm is the sign of entering the severe wear stage. The tool data acquisition stops when the tool enters the severe wear stage. The tool data acquisition covers the entire process of tool wear from initial wear, normal wear to severe wear.
[0032] In one possible implementation, the four cutting edges of the end mill are measured sequentially, including:
[0033] The cutting edge corresponding to the factory mark is designated as the first edge, i.e., edge 1, and then sequentially labeled as edge 2, edge 3, and edge 4 in the clockwise rotation direction.
[0034] After measuring the wear of all cutting teeth in the given order, fill the data into the milling cutter wear statistics table and save it in xls format.
[0035] In this embodiment, vibration signals, sound signals, cutting force signals, and torque signals were collected during the milling process of titanium alloys, and the tool wear condition was measured and labeled, forming a full life cycle dataset covering the tool from initial wear, normal wear to severe wear. This dataset can be used to predict tool wear behavior and diagnose tool failures, providing data support for intelligent cutting processes of titanium alloys. Attached Figure Description
[0036] Figure 1 A flowchart illustrating a method for collecting wear data of milling titanium alloy cutting tools provided in this application embodiment;
[0037] Figure 2 A schematic diagram of the test platform provided in the embodiments of this application;
[0038] Figure 3 This is a schematic diagram of the cutting force and torque signal waveforms provided in the embodiments of this application;
[0039] Figure 4 This is a schematic diagram of the vibration signal and sound signal waveforms provided in the embodiments of this application;
[0040] Figure 5 This is a schematic diagram of bottom edge wear measurement provided in an embodiment of this application;
[0041] Figure 6 This is a schematic diagram of peripheral wear measurement provided in an embodiment of this application;
[0042] Figure 7 This is a schematic diagram illustrating the sequence of cutting edges of a cutting tool provided in an embodiment of this application;
[0043] Figure 8 This is a schematic diagram illustrating the trend of tool wear value changes provided in an embodiment of this application. Detailed Implementation
[0044] The present solution will now be described in conjunction with the accompanying drawings and specific embodiments.
[0045] See Figure 1 The method for collecting milling titanium alloy tool wear datasets in this embodiment includes:
[0046] S101, determine the titanium alloy cutting tool to be worn and the test platform for data acquisition.
[0047] The experimental material in this embodiment is titanium alloy (Ti-6Al-4V), mainly a rod-shaped material with a base area of 100 cm². 2 With a diameter of approximately 11.3 cm and a height of 40 cm, this material, a typical example of difficult-to-machine material, possesses excellent physical and mechanical properties, as detailed in Table 1. The cutting tool used was a coated four-flute end mill with a diameter of 10 mm. The tool body was made of cemented carbide (YG3X), comprising 96.5% WC, 0.5% TaC (NbC), and 3% Co, coated with a TiAlN layer. The helix angle at the tool tip was 55°.
[0048] Table 1 Physical and mechanical properties of Ti-6Al-4V titanium alloy materials
[0049]
[0050] like Figure 2 As shown, the test platform can be divided into three parts: the cutting machining area, the signal acquisition area, and the tool measurement area. Specifically: ① Cutting machining: The machining equipment used is a VDF-850 vertical machining center from Dalian Machine Tool Plant, China, with a maximum spindle speed of 8000 r / min and a maximum feed rate of 1700 mm / min. A Kistler 9170B251 force sensor is installed on the tool holder, and the acquired cutting force signal is amplified using a charge amplifier. A DH5922D three-axis vibration acceleration sensor and a sound sensor are installed on the machine tool spindle to collect vibration and sound signals.
[0051] Experimental design was conducted using cutting parameters suitable for finishing titanium alloys. The spindle speed was set at 1430 r / min, the feed rate at 130 mm / min (or 0.023 mm / z), and the milling area was 100 cm². 2 The circular area was milled to a depth of 0.2 mm and a width of 8 mm, using a reverse milling method. During the experiment, the vibration, sound, and force signals (including torque signals) were all acquired at a frequency of 10 kHz, and the signal acquisition time (which is also the milling process time) was 8 minutes and 36 seconds. Specific experimental parameters are shown in Table 2 below.
[0052] Table 2 Test conditions
[0053]
[0054] S102, the titanium alloy tool to be worn is worn on the test platform, and vibration, sound signals and cutting force and torque signals are collected by sensors during the wear process to obtain the first dataset.
[0055] Tool wear measurement methods include direct measurement and indirect measurement, resulting in two types of tool wear data. Direct measurement uses visual or optical sensors and instruments to measure tool wear. However, optical instrument measurement typically requires machine shutdown and cannot be used for online monitoring, thus this method is generally used in laboratory environments. Indirect measurement, on the other hand, monitors tool condition indirectly by collecting one or more sensor signals related to tool wear or breakage. Compared to direct measurement, indirect measurement does not affect the machining process and is more widely used in actual cutting operations. This dataset contains data required for both direct and indirect measurement methods, namely tool wear values observed by optical instruments and tool vibration, sound, and force signals collected by sensors. Furthermore, different sensors were used to collect tool-related wear signals, making this dataset massive, multi-source, and heterogeneous.
[0056] In this embodiment, a charge amplifier is used to transmit the cutting force signal collected by the force sensor to a computer, and a Kistler signal acquisition system is used to collect cutting force and torque signal data in real time during the milling process.
[0057] The Kistler signal acquisition system integrates cutting force and torque acquisition functions, enabling real-time acquisition of cutting force and torque signals during milling operations. These signals are then displayed online on a PC via a corresponding interface. The acquired signals are renamed and saved in .csv format. Each acquired signal consists of five columns: the first column is time, and the other four columns represent the X, Y, and Z-axis cutting force and sound signals, respectively. After smoothing the acquired signals using Origin 2021 software, the corresponding X, Y, and Z-axis cutting force and sound signal images are obtained, as shown below. Figure 3 As shown (taking the data from the 68th collection as an example).
[0058] Vibration and sound signals are collected using a DH5922D triaxial vibration accelerometer and sound sensor. The accelerometer and sound sensor are integrated into a dynamic signal acquisition system through an interface. The collected signals are projected onto a computer via a USB interface to display the waveforms of vibration and sound.
[0059] The CR DH5922D vibration signal acquisition system integrates vibration and sound acquisition functions. It can acquire vibration and sound signals in real time during milling operations and display them online on a PC via a corresponding interface. The acquired signals are renamed and saved in .csv format. Each acquired signal consists of 6 columns of data: the first two columns represent time, and the other four columns represent the X, Y, and Z-axis vibration and sound signals, respectively. After smoothing the acquired signals using Origin 2021 software, the corresponding X, Y, and Z-axis vibration and sound signal images are obtained, as shown below. Figure 4 As shown (taking the data from the 68th collection as an example).
[0060] S103, the tool wear measurement is labeled using optical instruments to obtain a second dataset, wherein the wear measurement is performed multiple times during the entire wear process at different wear levels of the tool.
[0061] In this embodiment, after each stage of the milling process is completed, the tool is removed from the spindle, and the bottom edge and peripheral edge of the tool are measured using a 19Jc universal tool microscope.
[0062] The wear condition of the cutting tools was measured using a 19Jc universal tool microscope, focusing on the bottom and peripheral cutting edges. The specific measurement method was as follows: ① For the bottom cutting edge, the wear area was measured. Value, maximum wear width VB The wear area was measured and annotated using the closed cloud line function of QMS software, and the maximum wear width was annotated using the straight line annotation function. Specific details are as follows: Figure 5 As shown (taking the first bottom cutting edge as an example). ② For the peripheral cutting edge, the main measurement is the wear area. Value, maximum wear width VB Value and wear value at 1 / 2 cutting depth VB (i.e. 1 / 2) a p Wear value at the location VB There are three values. For the latter two, the distance annotation function in QMS3D-M software can be used to obtain them; for the wear area value... The measurement steps are as follows: First, before the tool wears out, that is, before using a new tool, measure the triangular area of its circumferential cutting edge, such as... Figure 6 (a) The area enclosed by the blue line (taking the first edge as an example) S 1 Then, after the tool wears down, its total area (including the worn area and the area of the triangular region) is measured. S 2 ,like Figure 6 (b) The area enclosed by the blue region, then use the total area. S 2 Subtract the area of the triangular region S 1 The area of the wear zone of the peripheral cutting edge can then be obtained. The value is the wear area calculated according to formula (1). value:
[0063] (1)
[0064] Measure the four cutting edges of the end mill sequentially using the above measurement method, and fill the data into the tool wear data set.
[0065] In this embodiment, the four cutting edges of the end mill are measured sequentially, including:
[0066] The cutting edge corresponding to the factory marking is designated as the first cutting edge, i.e., cutting edge 1. Following the clockwise rotation direction, they are sequentially labeled as cutting edge 2, cutting edge 3, and cutting edge 4. Figure 7 As shown. After measuring the wear of all cutting teeth in the given order, fill the data into the "Milling Cutter Wear Statistics Table" and save it in .xls format.
[0067] Directly measured tool wear data is of great significance, as it allows for a direct observation of the tool wear evolution. For example, the data from the first 16 measurements of the tool's peripheral and bottom cutting edges can be compiled and plotted as follows: Figure 8 As shown in the figure, the maximum wear width is represented by... M ax -VB This indicates that the wear width at 1 / 2 of the cutting depth is represented by... 1 / 2a p -VB Indicates the wear area using S VBexpress.
[0068] Depend on Figure 8 It can be observed that with the increase of machining measurements, almost all wear values gradually increase. Further observation reveals that the wear value of the bottom cutting edge is generally smaller than that of the peripheral cutting edge. This is mainly because the peripheral cutting edge plays a major role in milling, removing workpiece material. Throughout the milling process, the peripheral cutting edge has a large contact area with the workpiece material, resulting in high contact stress and significant chip impact. Furthermore, milling is an intermittent machining process, with the peripheral cutting edge constantly subjected to alternating loads. Additionally, the cutting temperature is high but heat transfer efficiency is low during titanium alloy milling. Under the combined effects of force and heat, the peripheral cutting edge experiences severe wear, even leading to tool tip chipping, creating a relatively harsh working environment. In contrast, the bottom cutting edge mainly bears the springback effect of the workpiece, resulting in a more stable working environment and less severe wear compared to the peripheral cutting edge. Even when the peripheral cutting edge enters a stage of severe wear, the bottom cutting edge only experiences slight wear. Furthermore, due to… Figure 8 (b) It can be found that the wear width value is much larger than the wear area value, that is, during the milling of titanium alloys, the peripheral cutting edge wears faster along the linear direction than the area wear.
[0069] S104, combine the first dataset and the second dataset to form the tool's full life cycle data.
[0070] Since the peripheral cutting edge plays a major role in milling, the initial wear is defined as a maximum wear width (or half the depth of cut) VB value of the peripheral cutting edge within the range of 0-0.1 mm, normal wear within the range of 0.1-0.5 mm, and a wear exceeding 0.5 mm, indicating the entry into the severe wear stage. The specific wear stages of the tool are shown in Table 3. Based on the stage division in Table 3, the end mill milling test on titanium alloy was conducted 68 times until the tool entered the severe wear stage. Tool data acquisition and tool measurement were also conducted 68 times. The dataset covers the entire process of tool wear from initial wear, normal wear to severe wear, and can be considered a full life cycle dataset of the tool.
[0071] Table 3. Tool Wear Stage Classification
[0072]
[0073] The dataset collected in this embodiment describes the entire lifecycle data of a coated four-flute end mill milling titanium alloy, including X, Y, and Z-axis vibration signals, sound signals, cutting force signals and torque signals in the X, Y, and Z directions, as well as the maximum wear width and wear area of the bottom cutting edge, the maximum wear width of the peripheral cutting edge, the wear width and wear area at 1 / 2 depth of cut, etc. The dataset is information-rich and multi-dimensional, and its multi-source and heterogeneous nature also presents certain challenges for data processing. This dataset is of significant value to current industrial front-line processing and production, and future research directions may include, but are not limited to:
[0074] (1) This dataset records the full life cycle data of carbide end mills from normal to fault state. The multi-source, heterogeneous and massive data are processed using software such as Matlab, Python, Origin, and SPSS to extract data features, which can be used to analyze and explore tool degradation behavior.
[0075] (2) Data can be processed using artificial neural networks such as convolutional neural networks, which can be used to construct a nonlinear mapping relationship model of vibration (sound) - cutting force (torque) - tool wear value, that is, to establish a certain data model to explore the evolution and distribution law of tool wear and the tool wear location.
[0076] (3) Based on the above two points, we can use the knowledge of materials mechanics and elasticity to assist software such as Ansys and Abaqus in constructing a tool wear mechanism (physical law) model. We can further seek to establish a tool wear model based on mechanism-data dual drive, i.e., a fusion model, which can monitor tool wear online and predict the remaining life.
[0077] (4) Use software such as SolidWorks, Photoshop, 3ds Max, Unity 3D to build high-fidelity models of CNC machine tools, cutting tools, etc., construct a virtual space for cutting and machining, and further use digital twin technology to construct a titanium alloy milling process system. Embed the fusion model into the process system and establish a titanium alloy milling process optimization model that takes into account tool wear.
[0078] (5) Based on the above points, we can further combine new generation information technology, integrated circuit technology and semiconductor technology to create a dedicated interface, build an online quality monitoring and optimization process system for titanium alloy milling, and explore a dedicated test platform suitable for intelligent cutting of titanium alloys.
[0079] Intelligent manufacturing is a development trend now and for a long time to come, and industrial big data is one of the fundamental and necessary conditions for achieving and doing intelligent manufacturing well. Based on this, this application conducted information collection on tool wear data related to titanium alloy milling to address the intelligent cutting problem in the field of titanium alloy machining, and the conclusions are as follows:
[0080] (1) Wear data of titanium alloy milling tools can be roughly divided into two parts: one part is indirect measurement data, such as vibration signals, sound signals, force signals, torque signals, etc., and the other part is direct measurement data, including the wear area of the flank face. S VB The dataset contains two parts of data, including the maximum wear width VB value and the wear width VB value at 1 / 2 depth of cut. It covers the entire process of tool wear from initial wear, normal wear to severe wear, and can accurately analyze and evaluate tool wear conditions.
[0081] (2) As can be seen from the wear value trend chart, both the wear area and the wear width increase with the number of machining operations. This is because the peripheral cutting edge plays a major role in the actual cutting process. When cutting titanium alloy materials, milling is an intermittent cutting process, and the chips have a large impact on the peripheral cutting edge, which causes the peripheral cutting edge to bear a large alternating load. The bottom cutting edge mainly bears the springback effect of the workpiece, and its working environment is relatively better. This results in the wear of the peripheral cutting edge being greater than that of the bottom cutting edge.
[0082] (3) This dataset helps promote technological research and engineering practice in the field of AI-based tool fault diagnosis and health management. It helps to further develop intelligent monitoring systems for tool wear using deep learning methods, and analyze the relationship between tool wear and workpiece machining quality, thereby assisting in the realization of predictive maintenance functions for intelligent machine tools.
[0083] As can be seen from the above embodiments, in response to the problem of limited data on tool wear during the milling of aerospace titanium alloy (Ti6Al4V), making it difficult to comprehensively evaluate the tool's operating condition, a titanium alloy milling experiment was conducted. The maximum tool wear width of 0.5 mm was used as a tool failure indicator. Vibration signals, sound signals, cutting force signals, and torque signals were collected during the titanium alloy milling process, and the tool wear condition was measured and labeled, forming a maximum wear width including the bottom cutting edge and the peripheral cutting edge. VB Value, wear area S VB Value, wear width at 1 / 2 depth of cut of the circumferential cutting edge. VB The dataset, which includes values, vibration, sound, cutting force, and torque signals, covers the entire lifecycle of cutting tools from initial wear, normal wear to severe wear, and has been made publicly available to scholars worldwide. This dataset can be used to predict tool wear behavior, diagnose tool failures, and provide data support for intelligent cutting processes of titanium alloys.
[0084] In this application embodiment, "at least one" refers to one or more, and "more than one" refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent the existence of A alone, the simultaneous existence of A and B, or the existence of B alone. A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects have an "or" relationship. "At least one of the following" and similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one of a, b, and c can represent: a, b, c, ab, ac, bc, or abc, where a, b, and c can be single or multiple.
[0085] The above description is merely a specific embodiment of this application. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the protection scope of this application. The protection scope of this application should be determined by the protection scope of the claims.
Claims
1. A method for collecting wear datasets of milling titanium alloy cutting tools, characterized in that, include: The titanium alloy cutting tool to be worn and the test platform for data acquisition were determined, including: The experiment used titanium alloy rod-shaped material - Ti-6Al-4V, with a base area of 100 cm². 2 The diameter is approximately 11.3cm and the height is 40cm. The cutting tool is a coated four-flute end mill with a diameter of 10mm. The tool body is cemented carbide-YG3X, of which WC is 96.5%, TaC (NbC) is 0.5%, Co is 3%, and the surface is coated with TiAlN. The tip helix angle is 55°. The test platform is divided into three parts: the cutting area, the signal acquisition area, and the tool measurement area. The cutting equipment is a VDF-850 vertical machining center from Dalian Machine Tool Plant in China, with a maximum spindle speed of 8000 r / min and a maximum feed rate of 1700 mm / min. The machine tool is equipped with a tool clamping device, and a force sensor is installed on the tool clamping device. A DH5922D three-axis vibration acceleration sensor and a sound sensor are installed on the machine tool spindle. The titanium alloy cutting tool to be worn was mounted on the test platform and subjected to wear. During the wear process, vibration, sound signals, cutting force, and torque signals were collected using sensors to obtain a first dataset, including: The cutting force signal collected by the force sensor is transmitted to the computer using a charge amplifier. The Kistler signal acquisition system is used to collect cutting force and torque signal data in real time during the milling process. Vibration and sound signals are collected using a DH5922D triaxial vibration accelerometer and sound sensor. The accelerometer and sound sensor are integrated into a dynamic signal acquisition system through an interface. The collected signals are projected to a computer via a USB interface to display the waveforms of vibration and sound. A second dataset was obtained by labeling the tool wear measurements using optical instruments, wherein the wear measurements were taken multiple times at different wear levels of the tool throughout the entire wear process. The method of using optical instruments to annotate tool wear measurements to obtain a second dataset includes: After each stage of the milling process is completed, the cutting tool is removed from the spindle; The bottom edge and peripheral edge of the cutting tool were measured using a 19Jc universal tool microscope. The method of using a 19Jc universal tool microscope to measure the bottom edge and peripheral edge of the tool includes: For the bottom cutting edge, the main measurement is its wear area. Value, maximum wear width VB The wear area is measured and marked using the closed cloud line function of QMS software, and the maximum wear width is marked using the straight line annotation function. For the peripheral blade, the main measurement is the wear area. Value, maximum wear width VB Value and wear value at 1 / 2 cutting depth VB value; For the maximum wear width VB Value and wear value at 1 / 2 cutting depth VB The value can be obtained using the distance annotation function in QMS3D-M software; for the wear area value... The measurement steps are as follows: First, before the tool wears, measure the area of the triangular region at the tip of the circumferential cutting edge. S 1 Then, after the tool wears down, its total area is measured. S 2 Total area S 2 This includes the area of wear and tear and the area of the triangular region, then the total area is used. S 2 Subtract the area of the triangular region S 1 The area of the wear zone of the peripheral cutting edge can then be obtained. value: Measure the four cutting edges of the end mill in sequence according to the above measurement method, and fill the data into the tool wear data set; The first dataset and the second dataset are combined to form the tool's full life cycle data; In milling, the peripheral cutting edge of the tool plays a major role. Initial wear is defined as a VB value of 0-0.1 mm at the maximum wear width or half of the cutting depth of the tool peripheral cutting edge. Normal wear is defined as a value of 0.1-0.5 mm. Wear exceeding 0.5 mm is the sign of entering the severe wear stage. The tool wear stops when the tool enters the severe wear stage. Tool data acquisition covers the entire process of tool wear from initial wear, normal wear to severe wear.
2. The method for collecting milling titanium alloy tool wear datasets according to claim 1, characterized in that, The process involves using a charge amplifier to transmit the cutting force signal collected by the force gauge sensor to a personal computer, and using a Kistler signal acquisition system to collect cutting force and torque signal data in real time during the milling process, including: The acquired signals were renamed and saved in .csv format; Each time, the collected signal consists of 5 columns of data. The first column is time, and the other 4 columns are the X, Y, and Z axis cutting force signals and torque signals, respectively. After smoothing the acquired signals using Origin 2021 software, images of the corresponding X, Y, and Z axis cutting force and torque signals can be obtained.
3. The method for collecting milling titanium alloy tool wear datasets according to claim 1, characterized in that, The system utilizes a DH5922D triaxial vibration accelerometer and sound sensor to collect vibration and sound signals. The accelerometer and sound sensor are integrated into a dynamic signal acquisition system via an interface. The acquired signals are then projected onto a computer via a USB interface to display the vibration and sound waveforms. This includes: The acquired signals were renamed and saved in .csv format; Each time, the collected signal consists of 6 columns of data. The first 2 columns are time, and the other 4 columns are the X, Y, and Z axis vibration signals, as well as the sound signal. After smoothing the acquired signals using Origin 2021 software, images of the corresponding X, Y, and Z axis vibration signals and sound signals can be obtained.
4. The method for collecting milling titanium alloy tool wear datasets according to claim 1, characterized in that, Measure the four cutting edges of the end mill in sequence, including: The cutting edge corresponding to the factory mark is designated as the first edge, i.e., edge 1, and then sequentially labeled as edge 2, edge 3, and edge 4 in the clockwise rotation direction. After measuring the wear of all cutting teeth in the given order, fill the data into the milling cutter wear statistics table and save it in xls format.