Dressing method for multi-threaded worm grinding wheels used for gear grinding

By measuring and recording parameters such as motor current during the worm gear dressing process, and combining this with machine learning algorithms to evaluate the quality of the dressing process, the unreliability of the dressing process in existing technologies is solved, enabling early detection and correction of quality problems and reducing costs.

CN116419819BActive Publication Date: 2026-05-26KAPP NILES GMBH & CO KG

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
KAPP NILES GMBH & CO KG
Filing Date
2021-10-29
Publication Date
2026-05-26

AI Technical Summary

Technical Problem

Existing technologies make it difficult to reliably monitor the dressing process of grinding worm gears, which may lead to continued use of substandard gears, increasing tooling and manufacturing costs.

Method used

By measuring and recording multiple process parameters during the dressing process of the grinding worm, such as the motor current of the dressing tool and the amount of material removed, and combining them with machine learning algorithms, the quality of the dressing process is evaluated, feature values ​​are generated to assess whether the dressing process meets the predetermined standards, and warnings are issued when necessary.

Benefits of technology

It improves the reliability of the finishing process, enables early detection and correction of quality problems, avoids the production of defective products, and reduces tool replacement and production costs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN116419819B_ABST
    Figure CN116419819B_ABST
Patent Text Reader

Abstract

This invention relates to a dressing method for a multi-threaded grinding worm (1) for gear grinding, the grinding worm (1) comprising multiple worm threads (2, 3), each worm thread being dressed by a dressing tool (4) having at least one dressing stroke (H1, H2, H3) to provide the desired profile for the grinding worm (1). To improve the method and obtain more reliable information on whether the dressing of the grinding worm is adequate by improving monitoring of the dressing, the method of the present invention includes: a) dressing the first worm thread (2), measuring and recording at least one process parameter (P1) during the dressing; b) dressing another worm thread (3), measuring and recording the process parameter (P1) during the dressing; c) after dressing all worm threads: combining the recorded process parameter (P1) data; d) comparing the combined data of the process parameters (P1) with the data stored in the machine controller, and checking whether the combined data is within a predetermined acceptable range; e) if the combined data is not within the predetermined acceptable range, outputting a signal.
Need to check novelty before this filing date? Find Prior Art

Description

[0001] The present invention relates to a dressing method for a multi-threaded grinding worm for gear grinding, wherein the grinding worm includes multiple worm threads, each worm thread being dressed by a dressing tool having at least one dressing stroke in order to provide the desired profile for the grinding worm.

[0002] Grinding of worm gears is typically used for finishing, particularly in gear production. A ground worm gear has several parallel worm threads. During grinding, these threads mesh with the teeth of the gear being ground.

[0003] The precise profile of the ground worm plays a decisive role in the quality of the gears to be produced. For ground worms containing abrasive in a dressable matrix, the geometry of the surface in contact with the gear teeth is produced through a dressing process. Therefore, in this case (unlike ground worms with a steel matrix covered with abrasive), the ground worm is dressed before the manufacturing process and during appropriate manufacturing intervals to eliminate wear-related negative variations in the ground worm profile.

[0004] Figure 1 The dressing process is illustrated schematically. The figure shows a grinding worm 1 with several worm threads 2, 3, which run parallel to each other around the axis of rotation of the grinding worm 1. Also visible is a dressing tool 4, mounted on a dressing shaft 5, driven by a motor (not shown). An unnamed dressing tool, also mounted on the dressing shaft 5 along with dressing tool 4, is shown for dressing the outer diameter (“tip circle”) of the grinding worm 1.

[0005] During the dressing process, the dressing tool 4 is guided to pass successively through each worm thread 2, 3 of the grinding worm 1 in order to mill out their contours. Figure 1 The relevant parameters are shown, such as the diameter d0 of the grinding worm 1 and the diameter d of the dressing tool 4. R The speed n of the grinding worm 1 during the dressing process S The speed n of the dressing tool 4 during the dressing process R During the dressing process, the relative moving speed v between the grinding worm 1 and the dressing tool 4 fad And the radial feed a of dressing tool 4 relative to the radial feed of grinding worm 1 ed Also note the coordinate Y, which represents the position of the dressing tool 4 relative to the grinding worm 1 during the dressing process.

[0006] Dressing typically occurs over several dressing strokes, beginning with a coarse stroke to produce the possible shape of the single worm thread flank, followed by a fine stroke to produce the final profile of the worm thread flank. Each dressing stroke can be performed as follows: one flank of the worm thread is dressed in an outward stroke, and the other flank is dressed in a subsequent return stroke. Between the outward and return strokes, the dressing tool can move in the Y-direction (positive movement during the outward stroke and negative movement during the return stroke).

[0007] To monitor the dressing process, it is well known that the current of the motor driving the dressing shaft must be closely monitored. This current is recorded, and its variation over time is observed. An allowable range is defined for this process—an envelope curve with upper and lower limits—within which the current must lie to allow for the setting of an appropriate dressing procedure.

[0008] Due to the continuous grinding and dressing of the worm gear, its diameter decreases. Changes in diameter or rotational speed are also reflected in the current consumption of the dressing motor. Therefore, even if the dressing process remains completely normal, it is possible to deviate from the aforementioned envelope curve. Thus, it is known to take this into account in a changing envelope curve (“adaptive envelope curve”). That is, the envelope curve is also adaptive in various dressing processes.

[0009] One problem is that, in some erroneous astronomical charts, the quality of the dressing process cannot be reliably summarized using previously known methods. This is especially true in the case of partial cutting in grinding worm gears. Problematic dressing processes in previously known methods sometimes lead to conclusions here, where the ground worm gear may still be usable, but with certain limitations. This results in higher tooling costs, and further, higher manufacturing costs due to tool replacements.

[0010] DE 102018131915 A1 relates to the purpose of determining the quality of a dressable grinding worm. For this purpose, after a dressing process, a dressing tool contacts the grinding worm to be dressed, and when said contact exists between the dressing tool and the grinding worm, at least one signal is measured and evaluated. Specifically, it specifies that the tip region of the worm thread is moved during contact with the area of ​​the dressing tool. In this case, the dressing tool contacts the tip of the worm thread of the grinding worm in an area on the outer diameter side. According to this previously known solution, after the dressing process, the dressed grinding worm needs to be inspected.

[0011] As described in WO 2020 / 193228 A1, process control during workpiece grinding is achieved through the grinding worm gear. Therefore, during workpiece machining, at least one measured variable is monitored, specifying a warning indicator for an unacceptable process deviation determined from at least one monitored measured variable. In particular, the detection of grinding wheel cutting is considered here.

[0012] This invention is based on the aim of further developing a general method that, through improved monitoring of the dressing process, provides more reliable information as to whether the dressing process of the grinding worm is still sufficient for use, or whether further measures (further dressing or worm replacement) are necessary to ensure the production of good parts. Therefore, one objective of this invention is to monitor the dressing process in a process-safe manner.

[0013] The present invention provides a solution to this objective by including the following steps:

[0014] a) The first worm thread is dressed using a dressing tool through at least one dressing stroke, wherein at least one process parameter is measured and detected during the dressing process;

[0015] b) Dress another worm thread, which is different from the first worm thread, using a dressing tool through at least one dressing stroke, wherein process parameters are measured and detected during the dressing operation;

[0016] c) After dressing all the worm threads of the ground worm: combine the data of at least one process parameter that was measured and recorded;

[0017] d) Compare the combined data of at least one process parameter with the data stored in the machine controller and check whether the combined data is within a predetermined acceptable range;

[0018] e) If the combined data is not within the predetermined acceptable range, then output a signal.

[0019] Dressing is preferably performed in several dressing strokes, with data measurement and acquisition only performed on selected dressing strokes, and preferably only on the last dressing stroke executed.

[0020] It is preferable to measure and record at least one process parameter along the path of the dressing stroke. Although it is also possible to record over time, recording along the path of the dressing stroke can improve the allocation of the location of problem areas on the grinding worm.

[0021] The preferred process parameter is the amperage of the drive motor used to rotate the dressing shaft carrying the dressing tool.

[0022] However, it is also possible that optional or additional process parameters are dressing work related to the removal of material from the grinding worm (in J / mm). 3 (in mm), this is necessary for at least one defined portion of the grinding worm during the dressing stroke. The volume of material removed from the grinding worm by the dressing tool in the corresponding worm thread (in mm) can be determined by the recorded path covered by the dressing tool relative to the grinding worm during the dressing process and the radial feed of the dressing tool relative to the grinding worm. 3 (in units). Accordingly, the energy (in J) can be determined by the current consumption of the motor driving the dressing shaft and the corresponding time for covering the distance, and the quotient (in J / mm) can be derived from there. 3 (in units).

[0023] Therefore, the power or current consumption of the dressing axis can be used to determine the energy absorbed over a specified time, which is necessary for cutting and grinding worm gear materials. Taking into account the geometry of the dressing tool and the profile of the worm thread, as well as process parameters (especially the feed of the dressing tool to the grinding worm), the corresponding machining volume can be determined. With the aforementioned parameters of cutting energy per unit volume, transferability between different dressing processes and different grinding worm gears is facilitated.

[0024] In addition, optionally or additionally, the process parameters are structural sound or acoustic emission, which are detected by structural sound or acoustic sensors during the trimming process.

[0025] In the simplest case, the signal is added to a combination of data from at least one process parameter that has been measured and recorded above.

[0026] The comparison of the aforementioned combined data may include a comparison with at least one pre-determined and stored maximum or minimum value. In this case, the comparison is particularly intended for a pre-determined portion of the grinding worm.

[0027] It is also possible to consider several process parameters; examples of this have been mentioned above.

[0028] From at least one process parameter, a characteristic value can be determined and output, which characterizes the dressing process of the grinding worm. In this respect, a characteristic value can be defined for each dressed grinding worm, which provides information about the dressing process that has been performed.

[0029] In this respect, it is an index-based process monitoring. The process can be evaluated by a numerical value (range: such as 0 to 100); "0" means: "no abnormality", and "100" means: "there is a problem with the process - the process must be stopped".

[0030] These indices can be calculated using a pre-defined algorithm. The algorithm's input can then be project data from the current machining operation and various drive signals from machine control, such as current consumption, power signals, and speeds, particularly for the dressing and tool axes. Additionally, signals from installed structural acoustic sensor systems and permanently installed accelerometer systems can be used. The sampling rate is preferably in the range of several hundred hertz. These signals can be adaptively filtered.

[0031] When we mention the motor's current consumption (I) here, we are of course referring to the motor's power absorption (P), which can be calculated using the following relationship.

[0032]

[0033] When a voltage U is applied, it can be calculated.

[0034] Evaluation can also be performed by recovering stored data (within machine control), thereby learning from previous trimming processes. This makes it easier to detect process deviations (abnormalities) so that machine operators can be alerted or the trimming process can be stopped.

[0035] During the manufacturing process, as an example, some signals inside the control system (i.e., signals present in the machine control system) and signals outside the control system (e.g., signals recorded by sensors that collect structural noise, such as from the machine tool or hall floor) can be recorded and considered.

[0036] In addition, data from within the machine (such as the set correction, dressing tool and grinding worm diameter, and the paths generated by the mutual guidance of the dressing tool and grinding worm) can be used to evaluate the process, and, if necessary, to adaptively filter and classify them (e.g., to divide the entire dressing process into different strokes).

[0037] To ensure that the measured data has universal validity and thus transferability across different trimming processes, the aforementioned characteristic values ​​can be standardized.

[0038] With the help of algorithms in statistics, and especially in the field of machine learning, these data are evaluated to determine the quality of the editing process. Well-known algorithms in machine learning include supervised and unsupervised learning, deep learning, and reinforcement learning.

[0039] This makes it possible to improve the monitoring quality of the trimming process, thereby stabilizing it. It can identify faults in the trimming process in an improved way, including the type of fault. This requires that such errors or similar errors have already been taught in the stored data (within machine control).

[0040] Therefore, the recommended method is used to monitor the dressing process and evaluate it based on an index. This means that for each dressing process, an index, i.e., a numerical value, is calculated to assess the quality of the dressing process. These index values ​​can then be provided to the machine operator as a progress chart, allowing the operator to easily obtain an overview of the dressing process occurring in each case. Several characteristic values ​​can be used to determine the aforementioned index. These characteristic values ​​then evaluate the dressing process in different ways. Therefore, errors can be detected at an early stage, which would otherwise be delayed or undetectable. These errors particularly include the cutting pattern of the grinding worm gear.

[0041] The data recorded during the actual dressing process is compared with the data stored in the database. In this way, for example, certain process signals can be assigned to certain areas of the grinding worm; correspondingly, for example, signals can be assigned to certain worm diameters and associated rotational speeds of the grinding worm.

[0042] The accompanying drawings illustrate embodiments of the present invention.

[0043] Figure 1 The diagram schematically illustrates a grinding worm gear being dressed with a dressing tool.

[0044] Figure 2 The process parameters are schematically illustrated as the amount of dressing work per volume of material removed during three dressing strokes of the grinding worm, with the process shown during the dressing of the first worm thread.

[0045] Figure 3 It schematically shows the following based on Figure 2 The process parameters are described in the two dressing strokes, with the process shown in the dressing of the second worm thread.

[0046] Figure 4 It schematically shows the following based on Figure 2 and Figure 3 The process parameters during the dressing stroke are shown, where the sum of the values ​​of the process parameters is shown, and an acceptable dressing process is still performed on the grinding worm.

[0047] Figure 5 The process parameters are schematically shown in the form of the dressing work per volume of material removed during two dressing strokes of the grinding worm, with the dressing process of the first worm thread (with long lines) and the dressing process of the second worm thread (with dashed lines) shown.

[0048] Figure 6 It schematically shows the following based on Figure 5The process parameters during the dressing stroke are shown, where the sum of the values ​​of the process parameters is shown, and no further acceptable dressing operations are performed on the grinding worm.

[0049] Figure 2 The diagram schematically illustrates the process of process parameter P1 along the paths of several dressing strokes H1, H2, and H3, which is the result of dressing the first worm thread 2 (see...). Figure 1 ).

[0050] Therefore, the figure shows the process of dressing the first worm thread 2 with dressing tool 4 (see Figure 2). Figure 1 During the trimming process, process parameter P1 is measured and recorded.

[0051] The process parameters here are related to the dressing work (in units of J / mm) concerning the material removed during the grinding of the worm gear 1. 3 This is necessary, at least for the defined part Y of the grinding worm 1, for dressing strokes H1, H2, and H3. For clarity, it should be noted that the dressing process for the different dressing strokes H1, H2, and H3 is plotted here on coordinate Y, although usually each dressing stroke is provided with forward and backward motion (outward stroke, backward stroke), or a retracing of coordinate Y.

[0052] It should be mentioned that, for a given radial feed of dressing tool 4 relative to grinding worm 1, at least at a constant feed rate v fad See below (see below) Figure 1 There is a proportional relationship between the repair work and the current absorbed by the motor of the repair shaft 5 or the power of the motor.

[0053] from Figure 2 It can be seen that the value of process parameter P1 is within the expected range.

[0054] Figure 3 The diagram schematically illustrates the process of process parameter P1 along the paths of the two dressing strokes H1 and H2, which is the result of dressing the second worm thread 3 (see...). Figure 1 ).

[0055] Therefore, the figure shows the dressing of another worm thread 3 using dressing tool 4, wherein process parameter P1 is measured and recorded again during the dressing process.

[0056] As we can now see, the conditions here are clearly no longer optimal, because the process parameter P1 shows a depression approximately in the middle of the upper platform, which is usually located on the upper platform during dressing, indicating that there is cutting on the grinding worm 1.

[0057] After all the worm threads have been dressed, the measured and recorded data for process parameter P1 are combined. This is in Figure 4 As shown in the figure. In this case, the total value of process parameter P1 is used, where the dressing strokes of each worm thread 2 and 3 can be simply allocated by simultaneously recording the coordinate Y.

[0058] Figure 4 This shows a relatively strong indentation in the middle of the dressing stroke (see...) Figure 3 It is still noticeable, but not so noticeable.

[0059] Now compare the combined data of process parameter P1 stored in the machine controller (not shown) and check whether the combined data is within the predetermined allowable range.

[0060] Therefore, in Figure 4 The diagram shows coordinates Y1 and Y2, and the evaluation is performed between these two. Therefore, observations occur within a defined region that is easily determined automatically (e.g., ...). Figure 4 The curve shown rises by a certain gradient at the beginning of the dressing stroke (coordinate Y1), and falls by a certain gradient at the end of the dressing stroke (coordinate Y2).

[0061] For an evaluation range defined in this way, the maximum value "max" and minimum value "min" of the process parameter curve can be automatically determined.

[0062] Figure 4 The limit value GW is also shown, which in this case marks the minimum permissible value of the stored process parameter P1.

[0063] Therefore, the software in the machine control can easily identify that the value is not below the limit value GW anywhere (between coordinates Y1 and Y2), so the cutting-dressing process on the grinding worm can still be judged as appropriate.

[0064] For example, based on the difference between the "maximum" and "minimum" values ​​and their relative position to the limit value GW, an eigenvalue can be calculated, providing information about the quality of the trimming process. For instance, the eigenvalue "100" can be assigned to a value similar to... Figure 2 The summary process takes the form of "optimal conditions during the adjustment process", and when the value "minimum" approaches the limit value GW, it is subtracted from the value "100", indicating that it has left the optimal range.

[0065] Figure 5 Similar examples are shown, where the respective dressing strokes in the worm threads 2 and 3 (two dressing strokes H1 and H2 are shown) result in the curve of the process parameter P1 shown. The dressing process of the first worm thread 2 is represented by a long line, and the dressing process of the second worm thread 3 is represented by a dashed line.

[0066] It can now be seen that in the two worm threads 2 and 3, the process parameter P1 decreases at approximately the middle of the upper platform, which in turn indicates that cutting has occurred in the grinding of the screw.

[0067] If we add the process parameter P1 values ​​recorded in segments 2 and 3 of a single worm screw thread, the result is... Figure 6 The curve shown here. It can be seen that in the middle region of the upper platform, the sum of process parameters P1 drops sharply; therefore, a similar evaluation can be performed on this curve, such as... Figure 4 As shown.

[0068] There is now a maximum value "max" and a minimum value "min" here. Although the minimum value "min" is now lower than the limit value GW, this indicates that the trimming process is no longer proceeding normally.

[0069] Therefore, a warning can now be issued to the machine operator stating that the dressed grinding worm no longer has the proper profile and is unusable. This signals that the combination data is no longer within the specified allowable range.

[0070] Of course, the characteristic value can be redefined, which quantifies the error that has occurred, so that machine control or machine operator can assess whether or should the production process continue, despite the less-than-ideal conditions during the repair period.

[0071] This procedure allows for the characterization of the dressing process: if the damage occurs only on a single worm thread, the impact on the overall process is relatively small. This is then reflected in the combined signals of all worm threads (according to...). Figure 4 or Figure 6 Through signal combination, several dressing strokes of a worm thread or several worm threads of a ground worm are combined by a described calculation.

[0072] Therefore, the failure modes can be identified in a more distinct way: a cyclic failure on several worm threads can be distinguished from a localized burst within a single worm thread. If several worm threads are damaged, this will also affect the combined signal. This can be taken into account in the exponential calculation of the process.

[0073] In the implementation scheme, as previously described, the dressing work related to the removal of material from the grinding worm 1 is used as process parameter P1. However, it is also possible to consider other or different process parameters in a similar manner. All parameters suitable for characterizing the dressing process are appropriate.

[0074] In particular, eigenvalues ​​sensitive to the trimming process can be used, for which the above explanation can be referenced (e.g., recording structural sound, which also allows for description of the trimming process). As shown in the figure, only a portion of the recorded process parameters (in this case: the values ​​between coordinates Y1 and Y2) can be used.

[0075] As mentioned above, different weights can be assigned to the dressing stroke of the worm thread. The weight of the first roughing stroke can usually be lower than that of the final roughing stroke, especially the finishing stroke that produces the quality-critical surface of the ground worm.

[0076] The machine's rotary encoder and linear scale can be used to assign the value of process parameter P1 to the current position of the dressing tool relative to the grinding worm. In this way, any faults on the grinding worm can be easily observed as a function of different dressing strokes and worm speeds.

[0077] After the trimming process is characterized by the calculated index (eigenvalue), various messages can be sent to the machine operator, or a response can be given regarding how to handle the situation if the limit value GW is exceeded.

[0078] According to the first option, the location Y of the damaged axis can be output to the machine operator first. In addition, the machine can automatically move the grinding worm with the damaged area to a defined position (e.g., in front of the dressing device) so that the operator can perform a quick and efficient visual inspection.

[0079] According to the second option, the machine operator can choose whether trimming should continue until the error is eliminated (i.e., until the index reaches a predetermined value, meaning it no longer exceeds or falls below the limit). The maximum number of further trimming cycles can be limited to prevent unnecessary further trimming in the event of a serious error.

[0080] According to the third option, the machine operator can choose whether to skip the damaged area (i.e., extending along the Y-axis) when further grinding the gear.

[0081] According to another option, the machine operator can choose whether to continue normal grinding, thus ignoring the error message.

[0082] By using existing data (the displacement amount set during the grinding process), it is possible to trace which workpiece was processed at which displacement position on the grinding worm. This allows information to be sent to the machine operator indicating which previously ground workpieces have been ground into damaged areas, and therefore may have experienced malfunctions.

[0083] List of reference numerals in the attached diagram:

[0084] 1. Grinding worm gear

[0085] 2. Worm thread

[0086] 3. Worm thread

[0087] 4. Trimming tools

[0088] 5. Dressing the shaft

[0089] P1 Process Parameters

[0090] H1 Dressing Stroke

[0091] H2 Dressing Stroke

[0092] H3 Dressing Stroke

[0093] Y-path of the trim stroke

[0094] GW Limit: The minimum permissible value of a process parameter.

Claims

1. A dressing method for a multi-threaded grinding worm (1) for gear grinding, wherein the grinding worm (1) comprises a plurality of worm threads (2, 3), each worm thread being dressed by a dressing tool (4) having at least one dressing stroke (H1, H2, H3) to provide the grinding worm (1) with a desired profile. in, The method includes the following steps: a) The first worm thread (2) is dressed using the dressing tool (4) through at least one dressing stroke (H1, H2, H3), wherein at least one process parameter (P1) is measured and detected during the dressing process. b) Using the dressing tool (4) to dress another worm thread (3) through at least one dressing stroke (H1, H2, H3), the worm thread (3) being different from the first worm thread (2), wherein at least one process parameter (P1) is measured and detected during the dressing operation. c) After dressing all the worm threads of the ground worm (1): combine the data of at least one process parameter (P1) in step a) and at least one process parameter (P1) in step b) that were measured and recorded; d) Compare the combined data of at least one process parameter (P1) from step a) and at least one process parameter (P1) from step b) with the data stored in the machine controller, and check whether the combined data is within a predetermined acceptable range; e) If the combined data is not within the predetermined acceptable range, then output a signal.

2. The method according to claim 1, wherein the dressing is performed in several dressing strokes (H1, H2, H3), wherein data measurement and data acquisition are performed only in selected dressing strokes.

3. The method according to claim 1, wherein at least one process parameter (P1) of step a) and at least one process parameter (P1) of step b) are measured and recorded on the path (Y) of the trimming stroke (H1, H2, H3).

4. The method according to claim 1, wherein at least one process parameter (P1) in step a) and at least one process parameter (P1) in step b) is the current of a drive motor, and the drive motor drives the trimming shaft (5) carrying the trimming tool (4) to rotate.

5. The method according to claim 1, wherein at least one process parameter (P1) in step a) and at least one process parameter (P1) in step b) are dressing work related to the removal of material by the grinding worm (1), the dressing work being required for a portion of the grinding worm (1) during the dressing strokes (H1, H2, H3).

6. The method according to claim 1, wherein at least one process parameter (P1) in step a) and at least one process parameter (P1) in step b) are structural sound or acoustic emissions detected by a structural sound or acoustic sensor during the dressing stroke.

7. The method of claim 1, wherein combining the data of at least one process parameter (P1) of step a) and at least one process parameter (P1) of step b) of claim 1 in step c) is to add the signals together.

8. The method of claim 1, wherein the comparison of the combined data in step d) of claim 1 includes a comparison with at least one predetermined and stored maximum or minimum value (GW).

9. The method according to claim 8, wherein the comparison is performed on a portion of the grinding worm (1).

10. The method of claim 1, wherein several process parameters (P1) are taken into consideration.

11. The method according to claim 1, wherein characteristic values ​​characterizing the dressing process of the grinding worm (1) are determined from at least one process parameter (P1) in step a) and at least one process parameter (P1) in step b).