Method for identifying interference components and device for identifying interference components

Abstract

Method for identifying interference components in a control device (1) for controlling a gear machining machine which performs machining under synchronization of a tool axis and a workpiece axis, where, when applying a vibration frequency that is contained in any positional deviation and synchronization error of the tool axis and the workpiece axis during machining, as fd, and setting a sampling frequency with the time until a rotating tool fixed to the tool axis, in conjunction with the rotation of the workpiece, after machining a flank line on the workpiece, machines the same flank line again, as sampling period as Fs then, if the oscillation frequency fd is sufficiently large and the Nyquist frequency Fs / 2 exceeds the sampling frequency, A frequency Fa of a ripple occurring on each of several flank lines in the flank line direction is given by the relation formula Fa = MIN | fd − N × Fs | , where N is a natural number is determined The step length of the waviness that generates the frequency fd in the flank line direction is calculated using the above relation formula and the velocity in the feed direction of the tool axis with respect to the workpiece axis. the calculated step length of the waviness and an actually measured value of the waviness that occurs after machining in the flank line direction of the workpiece are compared, and If the calculated step length of the waviness and the actually measured value match, the result of the comparison indicates that the vibration frequency fd is a disturbance component of the waviness occurring in the flank line direction of the workpiece after machining.

Description

Background of the invention 1. Field of the invention

[0001] The present invention relates to a method for identifying interference components and a device for identifying interference components for identifying interference components in a gear processing machine. 2. Description of the state of the art

[0002] Generally, a gear machining machine performs the machining through the interaction and synchronization of multiple axes driven by motors. An example of such a gear machining machine is a rolling gear machining machine. In a rolling gear machining machine, a workpiece to be machined is fixed to a rotating workpiece axis, a tool is rotated about a tool axis, and a workpiece axis motor and a tool axis motor are controlled so that the tool axis and workpiece axis rotate synchronously.

[0003] In a machining center that performs the machining of the workpiece while synchronizing these two axes, external disturbances can cause vibrations and affect the machining accuracy. Patents JP 3 986 320 B2 and JP 5 308 404 B2 describe a measurement of the machining accuracy of the workpiece and a correction of the position of a grinding worm using the measurement result.

[0004] The object of the present invention is to reduce the disadvantages described above. Brief description of the invention

[0005] External disturbances that affect the machining accuracy of a workpiece can include components caused by positional deviation of the tool axis or workpiece axis, synchronization errors, mechanical vibrations of machine parts, and the like. The key to improving the machining accuracy of a workpiece lies in identifying the disturbance components that affect machining accuracy.

[0006] The object is thus achieved by a method for identifying interference components having the features of patent claim 1 and by a device for identifying interference components having the features of patent claim 2.

[0007] One form of the present disclosure is a method for identifying noise components in a control device for controlling a gear processing machine that performs machining with synchronization of a tool axis and a workpiece axis, wherein, when setting a vibration frequency included in any one of a positional deviation and a synchronization error of the tool axis and the workpiece axis during machining as fd, and setting a sampling frequency with the time until a rotating tool fixed to the tool axis, in conjunction with the rotation of the workpiece after machining a tooth trace on the workpiece, machines the same tooth trace again as a sampling period as Fs, when the vibration frequency fd is sufficiently large and the Nyquist frequency exceeds Fs / 2 of the sampling frequency,a frequency Fa of a ripple occurring at each of several flank lines in the flank line direction by the relationship formula, Fa=MIN|fd−N×Fs|, where N is a natural number, is determined, the step length of the ripple that generates the frequency fd in the tooth trace direction is calculated using the above relationship formula and the speed in the feed direction of the tool axis with respect to the workpiece axis, the calculated step length of the waviness and an actually measured value of the waviness that occurs in the flank line direction of the workpiece after machining are compared, and if the calculated step length of the waviness and the actually measured value match, it is determined as a result of the comparison that the vibration frequency fd is a disturbance component of the waviness occurring in the tooth trace direction of the workpiece after machining.

[0008] Another form of the present disclosure is a noise component identification device in a control device for controlling a gear cutting machine that performs machining with synchronization of a tool axis and a workpiece axis, the device comprising an axis information recording unit that records any one of a positional deviation and a synchronization error of the tool axis and the workpiece axis during machining; and a noise component identification unit that, by setting a vibration frequency included in any one of the positional deviation and the synchronization error as fd, and setting a sampling frequency with the time until a rotating tool fixed to the tool axis, in conjunction with the rotation of the workpiece, after machining a tooth trace on the workpiece, again machines the same tooth trace as a sampling period as Fs, If the oscillation frequency fd is sufficiently large and exceeds the Nyquist frequency Fs / 2 of the sampling frequency, a frequency Fa of a ripple occurring at each of several flank lines in the flank line direction can be determined by the relationship formula Fa=MIN|fd−N×Fs|, where N is a natural number, determined, the step length of the ripple that generates the frequency fd in the tooth trace direction is calculated using the above relationship formula and the speed in the feed direction of the tool axis with respect to the workpiece axis, compares the calculated step length of the waviness and an actually measured value of the waviness that occurs in the flank line direction of the workpiece after machining, and if the calculated step length of the waviness and the actually measured value match, as a result of the comparison, it is determined that the vibration frequency fd is a disturbance component of the waviness occurring in the tooth trace direction of the workpiece after machining. Simple explanation of the drawings

[0009] The objects, features, and advantages of the present invention will become more apparent from the explanation of the following embodiment in conjunction with the accompanying drawings. In the accompanying drawings is Fig. 1 is a block diagram showing the overall structure of a processing machine with a control device according to an embodiment; is Fig. 2 a view showing an example of the construction of the processing machine when designed as a generating gear grinding machine; is Fig. 3 is an enlarged view of the environment of the workpiece in the machining center, showing the state of performing the rolling gear machining; is Fig. 4 is a view for explaining the flank line direction and the tooth profile direction on the flank line of a workpiece; is Fig. 5 is a view for explaining the trace produced by a contact point of a cutting grinding worm on the peripheral surface of the workpiece, the positional deviation, the sampling points of the waviness in the flank line direction, etc.; is Fig. 6 a diagram showing the oscillation of the tool axis along the time axis; is Fig. 7 a diagram in which the values ​​of the waviness in the flank line direction at the individual sampling points were plotted; is Fig. 8 is a view for explaining the sampling period of the ripple in the flank line direction; is Fig. 9 is a view showing the step length of the measured waviness in the flank line direction; is Fig. 10 is a view showing an example of tool axis position deviation data; is Fig. 11 is a view showing an example of workpiece axis position deviation data; is Fig. 12 is a view showing the state in which the flank line on the workpiece is rolled by teeth of the tool fixed to the tool axis; is Fig. 13 is a diagram showing an example of the waviness of a single pitch deviation; is Fig. 14 is a view for explaining the vibration of the waviness on the tooth surface of each flank line in a helical gear with a helix angle θ; is Fig. 15 is a flowchart showing the method for identifying noise components; is Fig. 16 is a flowchart showing correction method selection processing; is Fig. 17 is a flowchart of the processing for selecting a correction method; is Fig. 18 is a view showing an example of the structure of a control system of individual drive axles; and is Fig. 19 is a view showing an example of the frequency characteristics of a speed control loop of the control system. Detailed explanation

[0010] An embodiment of the present disclosure will be explained below with reference to the accompanying drawings. Throughout the drawings, corresponding structural elements are designated by common reference numerals. For ease of understanding, the scale of these drawings has been appropriately changed. The embodiment shown in the drawings is only one example of carrying out the invention; the present invention is not limited to the illustrated embodiment.

[0011] Fig. Figure 1 is a block diagram showing the overall structure of a machining machine 100 with a control device 1 according to an embodiment. The machining machine 100 is, for example, a generating gear grinding machine, a gear skiving machine, or a hob. Fig. The processing machine 100 shown in FIG. 1 includes the control device 1, n control axes CS1 to CSn, each having a motor and a drive axis, a measuring device 20 that measures the machining accuracy of a workpiece W, and a noise component identification unit 6. The control device 1 includes a position command generation unit 2, control units C1 to Cn that respectively control the control axes CS1 to CSn, an axis information recording unit 5, and a correction method selection unit 7. The noise component identification unit 6 may also be formed in the control device 1. A position command P (a command vector of n elements) for rotating each of the control axes CS1 to CSn from the position command generation unit 2 is input to the control units C1 to Cn, and the rotation of each of the control axes C1 to Cn is controlled according to the position command P.

[0012] The control device 1 may be configured as a computer including a CPU, a ROM, a RAM, and the like. In the control device 1, the control between the control units C1 to Cn and the control axes SC1 to CSn may be synchronization control, in which a synchronization ratio is applied to feedback from the master axis and a command is generated for the slave axis, but may also be command synchronization control, in which a synchronization ratio is applied to the command of the master axis and a command is generated for the slave axis. The individual functional blocks included in Fig. 1 can be implemented by special hardware or by the interaction of hardware and functions that the CPU performs through software.

[0013] In the processing machine, which is in Fig. As shown in Figure 1, the machining accuracy of a workpiece to be machined (hereinafter referred to as the workpiece) W may be affected by noise components associated with vibration. The key to accurately correcting the influence of such noise components on machining accuracy lies in identifying the noise component (the frequency of the noise component, the amplitude, etc.). By identifying the noise component, direct factors that cause the noise (for example, a positional deviation of the tool axis with a certain frequency peak) can be identified, and it becomes possible to apply an efficient correction method to improve machining accuracy.The machining apparatus 100 according to the present embodiment includes, as structural elements for identifying the disturbance component and applying an efficient correction method, the measuring device 20, the axis information recording unit 5, the disturbance component identifying unit 6, and the correction method selecting unit 7.

[0014] The axis information recording unit 5 records data of each axis during the actual machining of the workpiece W. The data of each axis includes a position deviation, which is the deviation of a detected position of each axis from a command position, and a synchronization error. When a first axis and a second axis rotate in synchronization, the synchronization error can be obtained as the difference between a converted position feedback of the first axis, for which the position feedback of the first axis has been converted into a value corresponding to the position feedback of the second axis using a synchronization ratio, and the position feedback of the second axis.The disturbance factor identification unit 6 identifies the disturbance factor affecting the machining accuracy of the workpiece W based on the data of the individual axes recorded by the axis information recording unit 5 and the machining accuracy measured by the measuring device 20. The correction method selection unit 7 selects an efficient correction method for correcting the positional deviation of the individual axes or the like from a plurality of pre-stored correction methods, depending on the disturbance factor identified by the disturbance factor identification unit 6.

[0015] Hereinafter, the machining machine 100 as a generating gear grinding machine which forms a helical gear as a workpiece W by generating gear machining is explained. Fig. Figure 2 shows a construction example when the processing machine 100 is designed as a generating gear grinding machine. As in Fig. As shown in Figure 2, the control device 1 in this case comprises a tool axis control unit C1, a workpiece axis control unit C2, and a feed axis control unit C3 as control units C1, C2, and C3. The control axes CS1, CS2, and CS3 are each designed as a tool axis CS1, a workpiece axis CS2, and a feed axis CS3 (not shown).

[0016] Fig. 3 is an enlarged view of the surroundings of the workpiece W in the machining device 100, schematically showing the state of gear grinding. A spiral cutting edge 51a is formed on the surface of a grinding worm 51, which is the tool fixed to the tool axis CS1. The workpiece W is a helical gear having oblique flank lines 61 already roughly machined by hobbing or skiving, or the like. The workpiece axis CS2 is arranged parallel to a Z axis. The grinding worm 51 rotates around the tool axis CS1, which has been positioned so as to be inclined with respect to a plane perpendicular to the Z axis, and the workpiece W is synchronously rotated with respect to the tool axis CS1 at a certain synchronization ratio.While the grinding worm 51 is in contact with the flank lines of the workpiece W, it is moved from top to bottom in the Z-axis direction by a drive of the feed axis CS3. This grinds the entire flank lines of the workpiece W. In . Fig. 3, the spiral cutting edge 51a on the grinding worm 51 and the plurality of flank lines 61 formed in the peripheral surface of the workpiece W are shown using solid lines for convenience of explanation, and only a part of the flank lines is provided with reference numerals.

[0017] The following procedures are used to identify the interference components (1) the identification of a disturbance component by the waviness in the flank line direction (flank line deviation); (2) the identification of a disturbance component by the waviness in the tooth profile direction (tooth profile deviation); (3) the identification of a disturbance component by a single pitch deviation; and (4) the identification of a disturbance component is explained by the amplitude of the position deviation.

[0018] In the processing machine 100, a waviness (tooth trace deviation) formed on the workpiece W in the tooth trace direction when vibration has occurred as a disturbance component on the tool axis CS1 is inspected. The workpiece W is assumed to be a helical gear with a helix angle θ. First, a method is shown below in which the waviness generated by the disturbance component in the tooth trace direction of the workpiece W is determined by calculation. Then, the waviness in the tooth trace direction generated in the workpiece W after machining is measured. By comparing the waviness in the tooth trace direction determined by calculation and the waviness in the tooth trace direction generated in the workpiece W after machining, the disturbance component is identified. Here, for the waviness in the tooth trace direction, as shown in Fig. 4, the unevenness in the depth direction with respect to the surface (the tooth surface) of the workpiece W in the direction of the tooth trace 61 (the tooth trace direction 71) on the peripheral surface of the workpiece W is indicated. On the other hand, the unevenness in the depth direction with respect to the surface (the tooth surface) of the workpiece W in the direction orthogonal to the tooth trace direction 71 is set as waviness (tooth profile deviation) in the tooth profile direction 72.

[0019] As in Fig. 5 shows the track L, which is the contact point T (see Fig. 3) between the grinding worm 51 and the peripheral surface of the workpiece W, when the workpiece W is considered fixed, is generated spirally. The vibration D1 caused by the disturbance component on the tool axis CS1 is in Fig. 5 as wavy continuous lines. This vibration D1 is actually an unevenness in the depth direction with respect to the tooth surface, but is in Fig. 5, for convenience, is expressed by waveforms as shown. By sampling the vibration of the tool axis CS1 along the tooth trace direction 71 using the data of the individual axes, the ripple in the tooth trace direction can be determined by calculation. It is assumed that the frequency of the vibration of the tool axis CS1 is sufficiently faster than the sampling frequency. The data showing the vibration of the tool axis CS1 can be obtained from the data of the individual axes recorded by the axis information recording unit 5 (the feedback values ​​of position sensors (encoders) provided on the individual drive axes).

[0020] Fig. Figure 6 is a graph showing the oscillation D1 of the tool axis CS1 along the time axis. As shown in Fig. 5, the phase of the oscillation D1 shifts at the positions of the sampling points (P0, P1, P2 ... PQ) on the oscillation D1 along the flank line direction for each sampling point, but the phase at the sampling point PQ again coincides with the oscillation D1. If the minimum value of the difference between the sampling period Ts and an integer multiple of the oscillation period Td is set as Δt, the phase deviation of the oscillation D1 at the sampling points P1, P2 becomes Δt and 2Δt, respectively, and at the sampling point PQ becomes QΔt. The sampling points P1 to PQ represent one period of the ripple generated in the flank line direction. Their relationship can be arranged as follows: Δt=|sampling period(Ts)−N×oscillation period(Td)|, where N is a natural number. Oscillation frequency (fd) = 1 / (Q × Δt) Ripple period = sampling period (Ts) × Q

[0021] From the above relationship formula, the ripple frequency can be defined as below: Ripple frequency=MIN|Oscillation frequency(fd)−N×Sampling frequency(Fs)| where N is a natural number.

[0022] Fig. Figure 7 is a graph plotting the waviness values ​​in the flank direction at each sampling point (P0, P1, P2, ... PQ). The step length (length) of the waviness in the flank direction can be determined using the waviness period determined above and the flank component of the feed rate of the tool axis CS1 in the Z-axis direction as follows: Ripple step length = ripple period × (feed rate in the Z-axis direction / cosθ

[0023] Next, the sampling period (Ts) is determined by setting the rotational speed of the workpiece axis CS2 as (ω(rad / s)) and the helix angle as θ. Here, as in Fig. 8, a coordinate system that fixes the workpiece W is assumed. The sampling period Ts can be regarded as the time until the workpiece makes one revolution at the moving speed of the contact point T viewed from the flank line 61. When the workpiece radius is set as r (mm), the peripheral speed is rω (mm / s), and when the feed rate of the tool axis in the Z-axis direction is set as Vz, the moving speed in the circumferential direction of the flank line 61 viewed from the Z-axis becomes Vz × tanθ. That is, when the rotation direction is the direction of Fig. 8, the moving speed of the contact point viewed from the flank line 61 is T rω × tanθ. Based on the above, the sampling period (Ts) in the flank line direction is determined as follows: Ts=2πrrω−Vz×tanθ

[0024] If the direction of rotation is the opposite direction of Fig. 8, the sampling period (Ts) is as follows: Ts=2πrrω+Vz×tanθ

[0025] The sampling frequency (Fs) is determined as the inverse of Ts. The sampling period Ts is the time until the rotating tool fixed to the tool axis CS1, after machining a flank line on the workpiece W, machines the same flank line again in conjunction with the rotation of the workpiece W.

[0026] According to the above, it became possible to determine the waviness frequency in the flank line direction by calculating using the data of each axis. Next, the waviness in the flank line direction of the workpiece W after actual machining is measured. It is assumed that the step length of the waviness in the flank line direction is measured by the measuring device 20 as shown in Fig. 9 shown as A (mm). From the measurement result of Fig. 9 the ripple frequency is defined as follows: (Ripple frequency)=Vz / (A×cosθ)

[0027] On the other hand, since the ripple frequency has been defined according to formula (1), the following equation is obtained: MIN|fd−N×Fs|=Vz / (A×cosθ)

[0028] Next, the identification of fd, which represents the disturbance frequency, is considered. Since there are several combinations of disturbance frequencies Fd and natural numbers N that satisfy the above formula (2), the positional deviation data recorded in the axis information recording unit 5 is referred to for the identification of the disturbance frequency. Here, it is assumed that by performing a frequency analysis of the positional deviation data for the tool axis CS1, a value as shown in Fig. 10 shown peak of HA (Hz) is present in the position deviation data, and for the workpiece axis CS2 two as shown in Fig. 11 are present in the positional deviation data. Since the disturbance frequency fd in this case is one of HA (Hz), HB (Hz), and HC (Hz), these are respectively substituted into the above formula (2), and the one that satisfies formula (2) is the disturbance frequency that generates the ripple in the tooth trace direction with the step length A. For example, if the frequency peak HA (Hz) of the positional deviation of the tool axis CS1 has a natural number N that satisfies formula (2), and the frequency peaks HB (Hz) and HC (Hz) of the positional deviation of the workpiece axis CS2 have natural numbers N that do not satisfy formula (2), the disturbance component that generates the ripple in the tooth trace direction with the measured step length A is identified as vibration with the frequency peak HA (Hz) of the tool axis and becomes the subject of the correction processing described later.

[0029] Through the method explained above, the noise component causing waviness in the tooth trace direction can be identified from the measured value of the tooth trace deviation and the actual machining data.

[0030] Next, the identification of the disturbance component by the waviness in the tooth profile direction (hereinafter referred to as tooth profile deviation) is explained. The condition is assumed that, as in Fig. 12, a flank line 161 is rolled on the workpiece W by the teeth of a tool 151 fixed to the tool axis CS1. Here, the engagement ratio of the workpiece is set as x. The engagement ratio is the number of teeth of the workpiece W that engage with the tool on average. In this case, the tooth profile is generated in the time t from the engagement to the disengagement of the teeth of the workpiece; and if the machining speed of the workpiece W is set as ω c(rpm) and the number of teeth of the workpiece W is taken as z, this time t is given by the following formula. t=(ωc60)−1×xz=60xωcz(sec)

[0031] If it has been measured that undulations occur at the tooth profile deviation a during this time t, the frequency fd of the disturbance component which is the cause of this is given as follows. fd=at=awcz60x(Hz)

[0032] Next, as in the case of undulations in the tooth trace direction described above, it is checked whether or not a frequency peak consistent with the noise component fd exists in the result of a frequency analysis of the position deviation data of each axis according to the above formula (3). If a frequency peak consistent with the noise component fd is found in the position deviation data of each axis, this frequency peak can be identified as the noise component.

[0033] Next, the identification of the disturbance component by a single pitch deviation is explained. The single pitch deviation is the difference between the actual pitch on the pitch circle of adjacent tooth surfaces on the same side and the theoretical pitch. For example, it is assumed that, as in Fig. As shown in Figure 13, the ripple of a single pitch deviation has been generated, and K peaks exist in the single pitch deviation within one revolution of the workpiece. In this case, the disturbance frequency causing this single pitch deviation is simply the K-fold vibration component in one revolution of the workpiece. Therefore, the disturbance frequency is obtained by the following formula (4). Disturbance frequency = (rotation speed of the workpiece) × K

[0034] In this case, if a noise component having a frequency peak consistent with the noise frequency obtained by the formula (4) is present in the position deviation data of each axis recorded in the axis information recording unit 5 during actual machining, this noise component can be identified as the cause of the single pitch deviation.

[0035] Next, the identification of the disturbance component by the amplitude of the positional deviation is explained. From the amplitude of the positional deviation recorded in the axis information recording unit 5 or the synchronization error of the tool axis CS1 and the workpiece axis CS2, the magnitude of the waviness formed on the tooth surface can be determined. Here, it is assumed that an amplitude A1 (deg) has been calculated as the waviness in the workpiece circumferential direction on the gear pitch circle. In this case, an amplitude x of the waviness on the tooth surface of each tooth trace at a value as shown in Fig. 14 shown helical gear with a helix angle θ is given by the following formula (5). x=r×A1180π×cosθ

[0036] Assume that vibration with an amplitude A1 (deg) and a frequency B (Hz) occurs as the positional deviation of the data of each axis. The amplitude x of the waviness generated by this vibration component on the tooth surface is calculated by the above formula (5). Then, this amplitude x is compared with the amplitude of the waviness of the tooth surface obtained by measuring the workpiece after actual machining. If the two values ​​are close to each other, it can be determined that there is a high probability that the above vibration component is the cause of the tooth trace deviation.However, in order to more accurately identify the disturbance component causing the waviness of the tooth trace deviation, it is necessary to calculate the step length of the waviness in the tooth trace direction from the frequency component B1 (Hz) of the position deviation by the method described above, and check whether it is consistent with the step length of the waviness in the tooth trace direction obtained by measuring the workpiece after actual machining.

[0037] The procedures explained above for identifying the interference component can be used as an operational procedure as in Fig. 15. This processing is performed by the noise component identification unit 6. Initially, the gear precision (helix deviation, profile deviation, pitch single deviation, amplitude, etc.) of an actually machined gear is measured (step S11). Then, the data of each axis recorded in the axis information recording unit 5 is referred to (step S12). Subsequently, using the gear precision measurement results and the data of each axis, the noise component that influences the gear precision (frequency, amplitude) is identified (step S13). It may also be that multiple noise components are identified.

[0038] Next, the processing for selecting the type of efficient correction with respect to the noise components identified by each method explained above is explained. Fig. Figure 16 shows the entire flow of the processing for selecting the correction method. This processing is performed by the correction method selecting unit 7. Initially, in step S101, the correction method obtained by the processing for identifying the noise component ( Fig. 15) is set to a variable d(i) (i = 1 to n) (step S101). Then, i is initialized (step S102), and during the period in which the condition that i is at most n is satisfied, the processing of steps S104 to S107 is performed. In step S104, it is determined whether or not the noise component d(i) is an integer multiple of the rotation frequency of the tool axis CS1 or the workpiece axis CS2. If the noise component is an integer multiple of the rotation frequency of the tool axis CS1 or the workpiece axis CS2, the correction method 1 (learning control) appropriate for this noise component is selected from a plurality of pre-stored types of correction methods 1 to 4 (step S105).

[0039] With learning control, it is possible to reduce the position deviation with a period of the operating pattern, such as approximately one revolution of the tool axis, as the learning period, which is effective in suppressing the noise component generated at a frequency that is an integer multiple of the rotation frequency of the tool axis CS1 or the workpiece axis CS2. As the learning control, a known learning control in which the correction amount is repeatedly determined through learning until the position deviation converges to near zero can be used.For example, in learning control, learning is repeated until the position deviation becomes close to zero. An example of such learning control in a machine tool is described in Japanese Patent Application No. 4 043 996 B2.

[0040] Next, the axis to which the correction is applied is set as the axis identified by the above-described noise component identification (step S107). This allows an efficient correction method to be applied to the axis that is the main cause of the noise in the next machining operation. In step S108, i is incremented to perform the processing from step S103 to step S107 for the next noise component.

[0041] On the other hand, if the disturbance component d(i) is not an integer multiple of the rotation frequency of the tool axis CS1 or the workpiece axis CS2, the selection of a correction method 2 to 4 is made in step S106. Fig. Figure 17 is a flowchart of the processing for selecting the correction method 2 to 4. First, in step S201, it is determined whether the cause of the disturbance component d(i) is a machine resonance or not. Now, the method for determining whether the cause of the disturbance component d(i) is a machine resonance or not will be explained. It is assumed that the control system of each drive axle consists, for example, of a system as shown in Fig. 18, which comprises a speed control unit 201 that generates a current command following a speed command and an actual speed detected by a speed sensor 205, a current control unit 202 that generates a voltage command according to the current command, a two-phase-to-three-phase conversion unit 203 that performs a two-phase-to-three-phase conversion according to the current command, and the speed sensor 205. It is assumed that by measuring the frequency characteristics of a speed control loop as in Fig. 19 can be obtained. If the frequency characteristics obtained as a result of the measurement have a value as shown in Fig. If the increase in gain at A2 (Hz) resulting from machine resonance shown in Figure 19 is present, oscillation of A2 (Hz) may be generated during machining. In step S201, it is determined whether or not the disturbance frequency d(i) coincides with the previously determined machine resonance frequency A2 (Hz). If the disturbance frequency d(i) coincides with the previously determined machine resonance frequency A2 (Hz) as a result of the determination, it is determined that the disturbance frequency d(i) is a machine resonance (S201: YES), and the correction method 2 (notch filter) effective for the machine resonance is selected.

[0042] As a method for preliminarily determining the machine resonance frequency, there is also the method of examining by conducting a hammer impact test on the structure of the processing machine or the like. For example, the resonance frequencies revealed by the hammer impact test are assumed to be A3 (Hz), B3 (Hz), and C3 (Hz). In this case, if the disturbance component identifying unit 6 has identified one of A3, B3, and C3, it is determined in step S201 that this disturbance frequency d(i) is a machine resonance (S201: YES), and the correction method 2 (notch filter) is selected (step S202).

[0043] Here, the notch filter is a filter that eliminates frequency components within a specific band range. If a notch filter is selected in step S202, a notch filter with filtering characteristics that eliminate the peak of the machine resonance frequency (for example, A3 (Hz)) identified in step S201 is incorporated into the control system at the control unit Cn that applies the notch filter.

[0044] On the other hand, if it is determined in step S201 that the disturbance component d(i) is not a machine resonance (S201: NO), it is determined whether the frequency of the disturbance component d(i) is at most a reference value of 1 (step S203). Here, the reference value of 1 may be a value, whereby it can be determined that the disturbance is a comparatively low frequency. If the frequency of the disturbance component d(i) is at most the reference value of 1 (S203: YES), the correction method 3 effective for correcting the disturbance with the comparatively low low frequency is selected.

[0045] As the correction method 3, various types of correction methods known in the related art that are effective for correcting a disturbance with a comparatively low low frequency can be used. An example of the correction method 3 is a procedure in which feedback synchronization, which is a method in which a synchronization ratio is applied to feedback from a master axis (tool axis) and a command is designed for the slave axis (workpiece axis), and command synchronization, which is a method in which a synchronization ratio is applied to the command of the master axis and a command is designed for the slave axis, are unified (hereinafter referred to as a unified function).Such a combined function can be implemented, for example, by a procedure in which a position feedback of the master axis is converted into a position feedback of the slave axis using a synchronization ratio, a synchronization error is calculated from the difference between the converted position feedback of the master axis and the position feedback of the slave axis, the difference between the calculated synchronization error and the position deviation of the slave axis is subjected to filter processing, and the position deviation of the slave axis is corrected using the filter output. By using such a combined function, a phenomenon such as a decrease in accuracy at the time of machining start can also be improved. An example of the combined function is described in Japanese Patent Application Publication No. S58-15784 B2.Since the correction method 3 (combined function) is effective for disturbances with a comparatively low frequency, the reference value 1 can be set to a value of about 150 Hz, for example, as a servo-controllable range.

[0046] On the other hand, if it is determined in step S203 that the frequency of the noise component d(i) is not at most the reference value 1 (S203: NO), it is determined whether or not the frequency of the noise component d(i) is at least a reference value 2 (step S205). Here, the reference value 2 may be a value, whereby it can be determined that the noise frequency is a relatively high frequency. If it is determined that the frequency of the noise component d(i) is at least the reference value 2 (S205: YES), the correction method 4, which is effective in the case of a relatively high noise frequency, is selected.

[0047] A method known in the field can be used for vibration damping control. An example of vibration damping control is control in which, in the case of machining with two axes—the tool axis and the workpiece axis—synchronized at a certain speed ratio, a correction amount to reduce vibration is calculated using the speed difference between the two axes, and the torque command is corrected using this correction amount, thereby reducing vibration caused by interference between the two axes. An example of such vibration damping control is described in Japanese Patent Application Publication No. 5 863 860 B2.Since the vibration damping control is effective for suppressing resonance frequencies having a comparatively high frequency (for example, 300 Hz) or the like, the reference value 2 can be set to 200 Hz as an example.

[0048] In particular, if the frequency of the interference component d(i) is not at least the reference value 2 (S205: NO), no selection of a correction method is required. If the processing for selecting the correction method is Fig. 17 ends, the processing goes to step S107 in Fig. 16 over.

[0049] As explained above, the present embodiment makes it possible to identify disturbance components that affect the accuracy of the workpiece after machining. Furthermore, it makes it possible to accurately select the type of effective correction to improve the accuracy of the workpiece after machining.

[0050] While an embodiment of the present disclosure has been described above, it will be understood by those skilled in the art that various improvements and modifications may be made without departing from the scope of the disclosure of the claims set forth below.

[0051] In the above-described embodiment, when identifying a noise component due to a tooth trace deviation, position deviation data (or synchronization error data) recorded in the axis information recording unit 5 is used as frequency information and used to calculate the ripple frequency. However, time may also be added to the information recorded by the axis information recording unit 5. In this case, the axis information recording unit 5 records the machining start time as the reference time, and the machining time and the position deviation or synchronization error are correlated.In this case, it happens that the shape of the deviation in the tooth line direction, which is formed, for example, by a positional deviation on the individual tooth surfaces, can be calculated, and not only the step length of the waviness but also the phase of the waviness or the like can be calculated.

[0052] Regarding noise components, it is conceivable that, depending on the machining position and time, the amplitude may change or multiple noise components may occur, and the shape of the waviness in the tooth trace direction occurring on individual tooth surfaces may also change accordingly. By recording a reference time of the position deviation, the precise shape of the waviness in the tooth trace direction corresponding to this phenomenon can be calculated, enabling even more detailed identification of the noise component.

[0053] The above-described embodiment was primarily explained for performing generating gear machining in which a tool axis and a workpiece axis rotate in synchronization, but the machining machine according to the present embodiment can also be used for gear profiling or gear profiling grinding. For example, in gear profiling grinding, a cutting tool with the same shape as the gear gaps is used, and machining is performed with a milling machine. Since the tool axis is stopped and the gear gaps are individually shaped, it can also be assumed that the tool axis and the workpiece axis rotate at a synchronization ratio of 1:0.

[0054] To achieve the object of the present disclosure, the following various forms and their results can be provided. The numbers in parentheses in the text of the following explanation of the forms correspond to the reference numerals in the drawings of the present disclosure.

[0055] A first form of the present disclosure is a method for identifying noise components in a control device for controlling a gear processing machine that performs machining with synchronization of a tool axis and a workpiece axis, wherein, when setting a vibration frequency included in any one of a positional deviation and a synchronization error of the tool axis and the workpiece axis during machining as fd, and setting a sampling frequency with the time until a rotating tool fixed to the tool axis, in conjunction with the rotation of the workpiece, after machining a tooth trace on the workpiece, machines the same tooth trace again as a sampling period as Fs, then, If the oscillation frequency fd is sufficiently large and exceeds the Nyquist frequency Fs / 2 of the sampling frequency, a frequency Fa of a ripple occurring at each of several flank lines in the flank line direction can be determined by the relationship formula Fa=MIN|fd−N×Fs|, where N is a natural number, is determined, the step length of the ripple that generates the frequency fd in the tooth trace direction is calculated using the above relationship formula and the speed in the feed direction of the tool axis with respect to the workpiece axis, the calculated step length of the waviness and an actually measured value of the waviness that occurs in the flank line direction of the workpiece after machining are compared, and if the calculated step length of the waviness and the actually measured value match, it is determined as a result of the comparison that the vibration frequency fd is a disturbance component of the waviness occurring in the tooth trace direction of the workpiece after machining.

[0056] Following the first form described above, it becomes possible to identify disturbance components that influence the tooth trace accuracy of a gear after machining. Furthermore, it becomes possible to precisely select the type of effective correction to improve the tooth trace accuracy of the gear.

[0057] Furthermore, a second form of the present disclosure is an apparatus for identifying disturbance components in a control apparatus for controlling a gear machining machine that performs machining under synchronization of a tool axis and a workpiece axis, the apparatus comprising an axis information recording unit (5) that records any one of a positional deviation and a synchronization error of the tool axis and the workpiece axis during machining;and a noise component identification unit (6) which, when setting a vibration frequency included in any one of the positional deviation and the synchronization error as fd, and setting a sampling frequency with the time until a rotating tool fixed to the tool axis machines the same flank line again in conjunction with the rotation of the workpiece after machining a flank line on the workpiece, as a sampling period as Fs then,; If the oscillation frequency fd is sufficiently large and exceeds the Nyquist frequency Fs / 2 of the sampling frequency, a frequency Fa of a ripple occurring at each of several flank lines in the flank line direction can be determined by the relationship formula Fa=MIN|fd−N×Fs|, where N is a natural number, determined, the step length of the ripple that generates the frequency fd in the tooth trace direction is calculated using the above relationship formula and the speed in the feed direction of the tool axis with respect to the workpiece axis, compares the calculated step length of the waviness and an actually measured value of the waviness that occurs in the flank line direction of the workpiece after machining, and if the calculated step length of the waviness and the actually measured value match, as a result of the comparison, it is determined that the vibration frequency fd is a disturbance component of the waviness occurring in the tooth trace direction of the workpiece after machining.

[0058] According to the second form described above, it becomes possible to identify disturbance components that influence the tooth trace accuracy of a gear after machining. Furthermore, it becomes possible to precisely select the type of effective correction to improve the tooth trace accuracy of the gear.

Claims

[1] Method for identifying interference components in a control device (1) for controlling a gear processing machine which carries out machining under synchronization of a tool axis and a workpiece axis, where, when applying a vibration frequency contained in any one of a position deviation and a synchronization error of the tool axis and the workpiece axis during machining, as fd, and setting a sampling frequency with the time until a rotating tool fixed to the tool axis in conjunction with the rotation of the workpiece after machining a flank line on the workpiece, machines the same flank line again, as a sampling period as Fs then, if the oscillation frequency fd is sufficiently large and the Nyquist frequency exceeds Fs / 2 of the sampling frequency, a frequency Fa of a ripple occurring at each of several flank lines in the flank line direction by the relationship formula Fa=MIN|fd−N×Fs|, where N is a natural number, is determined, the step length of the ripple that generates the frequency fd in the tooth trace direction is calculated using the above relationship formula and the speed in the feed direction of the tool axis with respect to the workpiece axis, the calculated step length of the waviness and an actually measured value of the waviness that occurs in the flank line direction of the workpiece after machining are compared, and if the calculated step length of the waviness and the actually measured value match, it is determined as a result of the comparison that the vibration frequency fd is a disturbance component of the waviness occurring in the tooth trace direction of the workpiece after machining. [2] Device for identifying interference components in a control device (1) for controlling a gear processing machine which carries out machining with synchronization of a tool axis and a workpiece axis, comprising: an axis information recording unit (5) that records any of a position deviation and a synchronization error of the tool axis and the workpiece axis during machining; and a disturbance component identification unit (6) which, when setting a vibration frequency included in any one of the positional deviation and the synchronization error as fd, and setting a sampling frequency with the time until a rotating tool fixed to the tool axis, in conjunction with the rotation of the workpiece, after machining a flank line on the workpiece, machines the same flank line again as a sampling period as Fs then, if the oscillation frequency fd is sufficiently large and the Nyquist frequency exceeds Fs / 2 of the sampling frequency, a frequency Fa of a ripple occurring at each of several flank lines in the flank line direction by the relationship formula Fa=MIN|fd−N×Fs|, where N is a natural number, determined, the step length of the ripple that generates the frequency fd in the tooth trace direction is calculated using the above relationship formula and the speed in the feed direction of the tool axis with respect to the workpiece axis, compares the calculated step length of the waviness and an actually measured value of the waviness that occurs in the flank line direction of the workpiece after machining, and If the calculated step length of the waviness and the actually measured value match as a result of the comparison, it is determined that the vibration frequency fd is a disturbance component of the waviness occurring in the tooth trace direction of the workpiece after machining.

Images