Apparatus for machining a workpiece and method for controlling a machine tool

Abstract

The invention relates to an apparatus for machining a workpiece, in particular a thread cutting apparatus (1) for manufacturing a thread on the end of a metal tube (3), comprising at least one machine tool (2) for machining the workpiece, at least one control device (6) for converting control commands to the machine tool (2), at least one higher-level process control (5), at least one device for optically or tactilely measuring the machined workpiece, means for detecting and storing machining process data, and means for detecting and storing geometric data of the workpiece detected when measuring the workpiece, the higher-level process control (5) comprising at least one correlation database for storing correlations between the geometric data and the process data, and a prediction module (15) for deriving open-loop and / or closed-loop control interventions for the machine tool (2), the prediction module (15) being based on machine learning methods and / or expert systems and coupled to the control device of the machine tool (2). The invention further relates to a method for controlling a machine tool.

Description

[Technical field]

[0001] The present invention relates to an apparatus for machining a workpiece, and in particular to a thread cutting apparatus for producing threads on the end of a metal tube.

[0002] The invention further relates to a method for controlling a machine tool for machining a workpiece with at least one tool having a specific cutting geometry, and in particular to a method for controlling a machine tool for manufacturing a thread in a metal tube.

[0003] The invention further relates to a method for manufacturing a thread on at least one end of at least one metal tube by machining the metal tube in at least one CNC controlled machine tool.

[0004] High demands are placed on the threads of pressure-resistant, gas-tight and liquid-tight pipes that are screwed together and are used to transport fluids under pressure, such as natural gas or oil. In such OCTG pipes (OCTG = Oil Country Tubular Goods), as feed or riser pipes for natural gas exploration or oil exploration drilling or as conveying pipes for natural gas or oil, conical threads with undercut thread flanges are usually used. A sealing lip usually connects these threads at the end face of the pipe. Both the thread and the sealing lip must meet the highest precision requirements. In the prior art, optical measurement of threads for quality inspection of pipes is basically known.

[0005] A method and an apparatus for optically measuring the male thread profile of a tube are known, for example, from WO 2019 / 09371. The apparatus here comprises a rest for the tube to be measured and an optical measuring unit with at least one measuring device including a light source and a camera arranged in its light path for taking an image of the male thread profile, where the optical measuring unit is arranged stationary on a support element that is held pivotably about three spatial axes and further comprises at least two measuring devices with mutually intersecting light paths. The method here comprises placing the tube to be measured on the rest such that the spatial axes extend transversely to the measuring plane of the measuring unit and the male thread is arranged in the beam path between the light source and the associated camera, orienting the measuring unit such that the measuring plane is perpendicular to the spatial axes, taking an image of the male thread using the camera of at least one measuring device, and evaluating the image.

[0006] Another device for measuring threads is known, for example, from EP 3465079. The device here comprises a holder for removably receiving a tube with a thread formed at its end and a first optical measurement section with a first optical sensor, the first optical measurement section being attached to a manipulator of the device that is configured to move the first optical measurement section relative to the tube and is adjustable about a first adjustment axis relative to the thread axis of the thread, on which a second optical measurement section of the device with a second optical sensor is arranged, the first optical measurement section and the second optical measurement section as a whole forming a measurement channel for simultaneous measurements from opposite sides of the thread. The device here is particularly characterized in that the measurement channel can be tilted by means of the manipulator relative to the thread axis about at least one second adjustment axis, so that the measurement channel can be freely oriented within a predefined spatial angle interval.

[0007] The measurement data obtained by known devices and methods is usually determined by random sampling and from this knowledge can be derived about the wear of the tools cutting the threads. The measurement results are also used to record quality assurance data.

[0008] EP 3306424 A1 discloses a method for cross-machine data utilization, in which operational data of a number of machines, each having a different first data format, is detected by a number of sensors, the data format depending on each sensor, at least one abstraction module converts the operational data into abstracted data having a unified second data format and stores it in a central storage device, an analysis is performed on this abstracted data based on the operational data of at least two machines, the operational state of one of the machines is evaluated based on this analysis, and an analysis of the abstracted data is output.

[0009] The object of the present invention is to provide an apparatus for machining workpieces, in particular a thread cutting apparatus for producing threads on the ends of metal tubes, which is improved with regard to the feedback of measurement data from quality inspections.

[0010] The present invention is further based on the object of providing a method for controlling a machine tool for machining a workpiece, which allows complete documentation and control of the quality of the machined workpiece, preferably in an uninterrupted manner.

[0011] The above-mentioned problem on which the invention is based is solved by providing an apparatus having the features of claim 1 and by providing a method having the features of claim 5. Advantageous configurations of the invention result from the respective dependent claims.

[0012] An aspect of the invention relates to an apparatus for machining a workpiece, in particular a thread cutting apparatus for manufacturing a thread on an end of a metal tube, comprising at least one machine tool for machining the workpiece, at least one control device for converting control commands to the machine tool, at least one higher-level process control, at least one device for optically or tactilely measuring the machined workpiece, means for detecting and storing machining process data, and means for detecting and storing workpiece geometry data detected when measuring the workpiece, wherein the process control comprises at least one correlation database for continuously storing correlations between the geometry data and the process data, and a prediction module for deriving open-loop and / or closed-loop control interventions for the machine tool, the prediction module being based on machine learning methods and / or expert systems and coupled to the control device of the machine tool.

[0013] The machine tool according to the invention is preferably configured as a CNC-controlled machine tool. The device for measuring the machined workpiece is preferably configured as an optical measuring device for the workpiece. The device according to the invention is preferably configured in such a way that the results of the automatic part measurement or thread measurement are used for pre-adjustment of the machine tool. With the aid of the prediction module, for example, the progression of tool wear can be identified early, which allows the derivation of control commands for controlling the machine tool, for example, in order to carry out an early tool change.

[0014] In a preferred variant of the device according to the invention, the machine tool comprises a number of sensors for detecting process data when machining a workpiece, these sensors being selected from the group of sensors comprising: sensors for detecting motor data of the drives of the machine tool, sensors for measuring vibrations in the machine frame of the machine tool, sensors for measuring vibrations and / or forces in the workpiece receptacle and / or tool receptacle of the machine tool, sensors for measuring feed speeds and / or feed forces of the main spindle and / or feed spindle and / or tool spindle of the machine tool.

[0015] The complete detection of process data by various sensors during the cutting process allows both continuous and discontinuous changes in measured values ​​to be detected, which can be evaluated, for example, with respect to tool wear, tool breakage or material defects in the component. With the aid of the prediction module based on machine learning methods provided by the invention, the adjusted operating state of the device can be identified in time during the process, thus allowing correspondingly early intervention. This in particular reduces significant material defects that would otherwise occur during control by random sampling during the process.

[0016] Sensors for measuring the forces in the workpiece receptacle can be provided, for example, in the form of piezo sensors or strain gauges. Vibrations can be detected, for example, by means of ultrasonic sensors, allowing, for example, early onset of tool wear to be detected. In this way, for example, cutting structures occurring on the clamping and releasing faces of the tool can be identified, which cause changes in the geometric data of the workpiece, which can be identified online during the process with the device according to the invention.

[0017] Preferably, the device according to the invention comprises a processing station and a measuring station arranged one after the other in a process line, whereby the processing station comprises a machine tool and the measuring station comprises at least one device for optically measuring a thread cut into a workpiece formed as a metal tube.

[0018] The device is preferably specifically configured to cut internal and external threads on OCTG tubing.

[0019] As the optical measurement means, at least one measuring head having at least one optical measurement section can be provided, the measuring head being arranged on a manipulator, the manipulator being configured to move the measuring head relative to the workpiece for measuring the workpiece.

[0020] The manipulator can be configured, for example, as an industrial robot with a jointed arm having multiple degrees of freedom. At the free end of the manipulator, a measuring head can be arranged so that it can be freely oriented in space in at least three spatial axes. The measuring head can be arranged on a support and guided at the support so that it can be adjusted in multiple degrees of freedom relative to it. The measuring head can be configured, for example, so that it can be moved relative to the line of the support and / or so that it can be tilted about at least one spatial axis.

[0021] An optical measurement section in the sense of the present invention can be an optical detection system with an optical sensor capable of optically measuring an object. The optical measurement section can include at least one light source and a camera and / or an optical scanner sensor. The optical measurement section can include, for example, a telecentric optical system that images a parallel beam path on the object side onto the optical sensor. In a preferred variant of the method, at least one measuring head is used that is configured to measure a conical thread with an undercut thread flange. Such a measuring head can include, for example, a first optical measurement section and a second optical measurement section that together form a measuring channel for simultaneously measuring the thread from mutually opposite sides. Such a measuring head is described, for example, in EP 3465079.

[0022] Another essential aspect of the invention is a method for controlling a machine tool for machining a workpiece with at least one tool having a particular cutting geometry, in particular for manufacturing a thread in a metal tube, comprising the following method steps: - detecting and storing machining process data in a host process control unit; - optically and / or tactilely detecting and storing the geometric data of the workpiece following the machining process; - correlating process data of the machine tool with geometric data of the tool and storing the correlation data in a correlation database of the process control unit; providing the correlation data to a predictive model or algorithm to derive open-loop control instructions and / or closed-loop control interventions for process control of the machine tool; The present invention relates to a method comprising the steps of:

[0023] By position-dependent evaluation of the workpiece geometric data and correlation of said geometric data with position-dependent measured process data in a preceding cutting or machining process, a correlation database is built up according to the invention, preferably using an expert system, artificial intelligence or fuzzy logic to evaluate the correlations between the position-dependent recorded process data and the time-continuous or discontinuous changes with local features or local defects in the part. After a learning phase of the predictive model or algorithm, the calculated correlations can be used to make predictions of the quality of the workpiece or of the cut thread by directly evaluating the data recorded during the machining process.

[0024] Preferably, the process data of the machine tool are detected continuously or discontinuously during operation of the machine tool with respect to the mutual position data of the workpiece and the tool, respectively, so that the geometric shape data of the workpiece and the process data of the machine tool can be correlated.

[0025] In this way, a complete documentation and control of the quality of the workpiece is guaranteed. The method predicts a direct description of the quality of the processed workpiece and allows the derivation of operating requests to the operator and / or control commands for the device or machine tool in order to avoid errors or to correct them later if possible, thereby reducing or avoiding defects in the workpiece that occur depending on the situation.

[0026] In a preferred variant of the method, the predictive model is based on machine learning methods, in particular artificial neural networks, deep artificial neural networks, decision trees, ensemble methods based on decision trees, linear or non-linear regression models with or without regularization, support vector machines with linear or polynomial or other kernel functions, etc.

[0027] A variant of the method is characterized in that the process data are detected during machining using a number of sensors provided on the machine tool.

[0028] The process data are preferably selected from a group of process data comprising motor data of the drive of the machine tool, vibrations measured on the machine frame of the machine tool, forces and / or vibrations measured on the workpiece holder and / or on the tool holder of the machine tool, feed rates and / or feed forces of the main spindle and / or feed spindle and / or tool spindle of the machine tool.

[0029] Preferably, actual workpiece geometry data is measured and compared to target workpiece geometry data, and data derived from this comparison is correlated with measured process data.

[0030] At least some control commands of the machine tool can be configured to be derived from predefined action tables stored in a higher-level process controller depending on the predictions generated by the predictive model.

[0031] Furthermore, the invention relates to a method for manufacturing a thread on at least one end of a metal tube by machining the metal tube in at least one CNC controlled machine tool, comprising optical measurement of the thread during the thread cutting process and further comprising controlling the machine tool according to the method described above.

[0032] The machine tool of the device or the machine tool to be controlled can be configured as a lathe, a turning and milling centre, a thread cutting machine or a sleeve cutting machine.

[0033] The machine tool may include at least one rotatable clamping chuck for clamping the metal tube and a tool holder including at least one tool, fixable and positionable relative to the clamping chuck. For example, the machine tool may include at least one tool holder with a number of different tools, preferably multiple tool holders formed as a turret head.

[0034] The invention will now be described by way of example with reference to the accompanying drawings, in which: [Brief description of the drawings]

[0035] [Figure 1] FIG. 2 is a schematic diagram showing the layout of a thread cutting device according to the present invention; [Diagram 2] FIG. 2 shows a schematic perspective view of a measuring station of a thread cutting device according to the invention; [Diagram 3] FIG. 2 is a perspective view of a measuring head according to the present invention during thread measurement; [Figure 4] FIG. 1 is a schematic diagram showing closed loop control of a thread cutting device. [Diagram 5] 2 is a flow chart illustrating a method according to the present invention.

[0036] In figure 1 the layout of a thread cutting device 1 according to the invention is shown. The thread cutting device 1 comprises a machine tool 2 for cutting one end of a metal tube 3 which can be formed, for example, as a feed pipe or riser pipe for natural gas production drilling or oil production drilling. Such metal tubes 3 are screwed together to form a pressure-resistant, liquid-tight and gas-tight connection between a plug end with an external thread 4 and a sleeve end with an internal thread. The method relates to the manufacture of both the external thread 4 and the internal thread of such metal tubes. The example will be described in particular with respect to the thread cutting process for the external thread.

[0037] The metal tube 3 is provided with a conical external thread 4 in the machine tool 2. For this purpose, the metal tube 3 to be machined is first clamped in a rotatable clamping chuck of the machine tool 2. For producing the thread, the machine tool 2 preferably has two tool turrets with tools, which are arranged relative to the metal tube 3 rotated by the clamping chuck or rotated at a predefined machining speed, in which case the tools are respectively guided in engagement with the end of the metal tube so as to be able to cut. The thread profile referred to here is a target profile stored in a process control 5 supervised by the machine tool 2.

[0038] The method according to the invention comprises the cutting of a number of metal tubes 3 in a process line and an optical measurement of the external threads 4 of the metal tubes 3 in the process line. As is evident from the installation layout shown in Figure 1, a measuring station 7 is arranged in the process line downstream of the machine tool 2, which is configured as a CNC-controlled machine tool 2.

[0039] The measuring station 7 is shown diagrammatically in Fig. 2. It comprises a robot 8 with a rotatable and pivotable robot arm 9, preferably with five degrees of freedom, at the free end of which a measuring head 10 is arranged. The measuring head 10 comprises a support 11, on which optical measuring means are provided for optically measuring the external thread 4. After the thread cutting process, the metal tube 3 is fed to the measuring station 7 via a roll bed 12 and placed in a defined measuring position. The measuring position can be determined, for example, by a positioning roller (diabolo roller) 13 with a central constriction, as shown diagrammatically in Fig. 3, by means of which the position of the metal tube 3 is determined. Alternatively, lateral stops can also be provided, which determine the position of the metal tube 3 to be measured. In the measurement position of the metal tube 3 , the support 11 of the measuring head 10 is moved to a measurement position in which the measuring head 10 is oriented relative to the metal tube 3 , possibly after a diameter calibration of the metal tube 3 has been performed.

[0040] The diameter calibration of the measuring head 10 serves to position the measuring means of the measuring head 10 relative to the support 11 in such a way that the metal tube 3 is placed between them and the measuring head 10 does not collide with the metal tube 3 during pre-orientation. For this purpose, limit gauges 18 are arranged in the measuring station 7 as reference parts, on the basis of which the measuring head 10 can be calibrated before the measurement process is carried out. To position or pre-orient the measuring head 10, at least one position sensor can be provided, for example as a laser scanner sensor, by means of which the position of the measuring head 10 relative to the metal tube 3 placed in the measuring position can be checked and possibly corrected. The method can include both a pre-orientation of the measuring head 10 by correspondingly driving the robot arm 9 and a fine alignment of the measuring head 10 by moving the measuring head 10 relative to the support 11. The fine alignment includes an alignment of at least one measurement section relative to the tube axis of the metal tube 3.

[0041] As already mentioned above, the measuring head 10 can move linearly relative to the support and can preferably be pivoted about at least one axis. The linear movement can be effected, for example, by at least one driven ball spindle or by at least one rack and pinion. The measuring means are respectively arranged on legs 19 of the measuring head 10. The legs 19 of the measuring head 10 are linearly movable at a distance from one another. The legs 19 of the measuring head 10 form a U-shaped enclosure of the metal tube 3. The two legs 19 can be designed to be movable independently of one another relative to one another. In the exemplary embodiment described, one leg 19 of the measuring head 10 is arranged in a fixed position, whereas the other leg 19 of the measuring head 10 is arranged to be movable relative to the opposite leg 19 of the measuring head 10.

[0042] In the superordinate process control 5, an evaluation of the measurement data of the male thread profile takes place by comparison of the actual measurement data from the measuring station with the setpoint geometric data from the structure and further control commands are derived for controlling the machine tool 2, in particular in the event of deviations between the setpoint profile and the actual profile obtained by the measurement data. The respective setpoint profile may be freely selectable, for example in a user interface (HMI) of the process control 5, from a catalog of different thread types.

[0043] The control commands can be, for example, readjusting the tool position, selecting a tool, the rotation speed, the torque applied to the clamping chuck of the machine tool 2 and the metal tube 3, performing a tool change, changing the cycle time of the machine tool 2, etc.

[0044] As already mentioned above, the closed-loop control of the machine tool can be implemented as a self-learning closed-loop control (AI) and / or as a closed-loop control assisted by an expert system. The process control 5 comprises a computer-implemented prediction module 15, which has access to a correlation database in which the geometric data detected during the measurement of the male thread 4 are cross-correlated with a number of process data detected by sensors 14 of the machine tool 2. The sensors here include, for example, the current consumption I(t) of the drive of the machine tool 2, the feed force F(t) of the spindle and / or tool receptacle of the machine tool 2, the feed speed v(t) of the spindle and / or tool receptacle of the machine tool 2 and the rotation speed n(t) of the spindle of the machine tool 2. The process data here are given by way of example only; many other process data can be detected within the framework of the present invention.

[0045] Process data of the machine tool 2, with regard to the mutual position data of the male thread 4 and the tool, respectively, are continuously detected during operation of the machine tool 2. The measured geometric data of the male thread are correlated with the corresponding process data of the machine tool 2. Permanent or discontinuous changes in the process data are evaluated by an expert system and / or a self-learning closed-loop control, taking into account the detected correlations. With the aid of set and / or learned algorithms, corresponding control commands and / or actions are derived from the prediction module 15, which are transmitted from the process control 5 to the control device 6.

[0046] The measured data calculated for the tube are used according to the invention not only for the feedback connection to the machine tool 2 and its control, but also for the assurance and tracking of quality data.

[0047] The method according to the invention is preferably arranged so that the optical measurement of the male thread 4 in the process line is carried out approximately 3-4 cycles after the thread cutting process. [Explanation of symbols]

[0048] 1 Thread cutting device 2 Machine tools 3 metal tube 4 External thread 5 Process Control Section 6. Control device 7 Measuring Station 8. Robot 9. Robot Arm 10 Measuring Head 11 Support 12 Roll Bed 13 Positioning roller 14 Sensors 15 Prediction Module 16 Not set 17 Not set 18 Limit Gauge 19 Legs

Claims

1. A machine for machining a workpiece, in particular a thread cutting machine (1) for producing a thread on the end of a metal tube (3), comprising: At least one machine tool (2) for machining the workpiece; At least one control device (6) for converting control commands to the machine tool (2); At least one upper process control unit (5); at least one device for optically or tactilely measuring the machined workpiece; A means for detecting and storing machining process data; means for detecting and storing geometric data of the workpiece detected when measuring the workpiece; Equipped with said higher-level process control (5) comprising at least one correlation database for storing correlations between said geometric data and said process data and a prediction module (15) for deriving open-loop control measures and / or closed-loop control interventions for said machine tool (2), the prediction module (15) being based on machine learning methods and / or expert systems and coupled to the control device of the machine tool (2), Device (1).

2. The machine tool (2) includes a plurality of sensors (14) for detecting process data when machining the workpiece; The sensor (14) A sensor for detecting motor data of a drive unit of the machine tool (2); a sensor for measuring vibrations on a machine frame of the machine tool (2); sensors for measuring vibrations and / or forces in the workpiece receptacle and / or the tool receptacle of the machine tool (2); a sensor for measuring the feed rate and / or the feed force of the main spindle and / or the feed spindle and / or the tool spindle of the machine tool (2); 10. The apparatus of claim 1, wherein the sensor is selected from a group including:

3. The apparatus (1) comprises at least a processing station and a measuring station (7) arranged successively in a process line, wherein: the machining station comprises the machine tool (2) and the measuring station (7) comprises at least one device for optically measuring a thread cut in the workpiece formed as a metal tube (3); 3. The device according to claim 1 or 2.

4. As optical measuring means, at least one measuring head (10) having at least one optical measuring section is provided, said measuring head (10) being arranged on a manipulator, said manipulator being configured to move said measuring head (10) relative to said workpiece for measuring said workpiece.

4. Apparatus according to any one of claims 1 to 3.

5. A method for controlling a machine tool (2) for machining a workpiece with at least one tool having a specific cutting geometry, in particular for manufacturing a thread in a metal tube (3), comprising the steps of: - detecting and storing machining process data in a higher-level process control unit (5); - optically and / or tactilely detecting and storing the geometric data of the workpiece following the machining process; - correlating process data of the machine tool with geometric data of the workpiece and storing the correlation data in a correlation database of the process control unit (5); - feeding said correlation data to a predictive model to derive open-loop control commands and / or closed-loop control interventions for controlling said machine tool (2); A method comprising:

6. 6. The method according to claim 5, wherein the process data of the machine tool, with respect to the mutual position data of the workpiece and the tool, respectively, are detected continuously or discontinuously during the operation of the machine tool (2).

7. 7. The method according to claim 5 or 6, wherein the predictive models are based on machine learning methods and / or expert systems, in particular based on artificial neural networks, artificial deep neural networks, decision trees, ensemble methods based on decision trees, linear or non-linear regression models with or without regularization, support vector machines with linear, polynomial or other kernel functions, etc.

8. 8. The method according to claim 5, wherein the process data is detected during the machining using a plurality of sensors (14) provided on the machine tool (2).

9. The process data includes: Motor data of the drive unit of the machine tool (2); Vibrations measured on the machine frame of the machine tool (2), - forces and / or vibrations measured in the workpiece receptacle and / or the tool receptacle of the machine tool (2), the feed rate and / or feed force of the main spindle and / or the feed spindle and / or the tool spindle of the machine tool (2), The method of claim 5 , wherein the process data is selected from a group comprising:

10. 10. The method of claim 5, wherein actual geometry data of the workpiece is measured and compared to target geometry data of the workpiece, and data derived from the comparison is correlated with measured process data.

11. 10. The method according to claim 5, wherein at least some control commands of the machine tool (2) are derived from a predefined action table of the process controller in dependence on predictions generated by the predictive model.

12. A method for manufacturing a thread on at least one end of at least one metal tube (3) by machining the metal tube in at least one CNC-controlled machine tool (2), comprising the steps of: The method comprises performing an optical measurement of the thread during the thread cutting process and further comprises controlling the machine tool (2) according to a method according to any one of claims 5 to 11. method.

Images