GB2377023A

GB2377023A - Diagnosis of movement precision in a multi-axis system

Info

Publication number: GB2377023A
Application number: GB0214279A
Authority: GB
Inventors: Matthias Stupp; Jens Bohm
Original assignee: DaimlerChrysler AG
Current assignee: Daimler AG
Priority date: 2001-06-28
Filing date: 2002-06-20
Publication date: 2002-12-31
Anticipated expiration: 2022-06-20
Also published as: DE10131408A1; GB0214279D0; GB2377023B; DE10131408C2

Abstract

A method for diagnosis of the mechanical movement precision of a multi-axis rotary pivot head 1 comprises dynamically measuring the change in spacing between a first reference point P1 at the head 1 and a second, stationary reference point P2 in dependence on the rotational angles of different axes A, C controlled in drive. For this purpose, for example, a length-measuring instrument 2 is rotatably connected at one end with a rigid support 3 and at the other end with the rotary pivot head 1, the first reference point P1 lying on an extension of one rotational axis C of the head at the point of intersection with the centre axis of the length-measuring instrument 2. On pivotation, rotation or combined movement of the head 1, the signals of the instrument 2 are picked up in dependence on a predetermined drive control of the axes (pivot angle and rotational angle). Respective characteristic measurement curves result for different kinds of movement. The course of the curves departs in characteristic manner due to errors (bearing play, etc.) in relation to a computerised ideal form and thus enables error diagnosis.

Description

DIAGNOSIS OF MOVEMENT PRECISION IN A MULTI-AXIS SYSTEM The present invention relates to a method of diagnosing movement precision in a multiaxis system, especially a diagnostic system by which, in the case of a machine with a rotary pivot head, the precision of the rotary or pivot motions can be detected by measuring. In particular, the method concerns systems in which checking of the deviations in the motion of a tool mount (spindle, chuck, clamping device) from the desired ideal movement is to be carried out with several degrees of freedom which are present (linear axes, rotational axes), for example in 5-axis milling machines.

Deviations in the movement course can arise through, for example, play in rotary mounts, step-down transmissions, etc. In the result, the tool, for example a milling device, inserted in the tool mount then does not exactly assimilate the positions in space predetermined by the drive control or else slightly departs from the predetermined track in the space, so that - according to the respective requirement for maintenance of dimensions of the workpiece - expensive reprocessing can be required. For this reason it is necessary to check the accuracy of the guidance of the tool mount at specific intervals. This checking must take place under the same conditions as in operation, i. e. the dynamic track behaviour of the tool mount for simple motions (for example rotation about one axis) and compound motions (for example, simultaneous rotational and pivot motion) has to be measured. In that case, the requisite measuring accuracy can lie in the region of a few micrometres and below, depending on the respective machine type.

Various diagnostic systems for checking dynamic track behaviour are known, thus: a) Circularity tests, corner travels and planar free-form tests can be carried out by a cross-grating measuring device with accuracy in the region of about 2 vim. This measuring device consists of a measuring plate with a cross-grating division, which is mounted on the machine table. The scanning head clamped to the tool mount is guided in contactless manner over the measuring plate. Due to the construction, the measurements are confined to movements in the plane parallel to the machine table. b) In the case of a laser interferometer, a reflector for reflecting the beam emitted by a laser is mounted at the tool mount A measuring accuracy of 1 11m can be

achieved with this device. Only characteristics of individual axes (linear axes) can be determined, and not the interaction of several axes (exception : rectangularity).

Adjustment is very complicated with this method and the costs for the measuring device are also relatively high. c) Another method for measuring the movement of the tool mount in one plane is the "ball-bar"test. In this method, a rod-shaped length-measuring instrument is rotatably mounted by one end at a holder at a spacing from the machine plate.

The other end is guided by the tool mount at the rotary pivot head so as to be similarly freely rotatable. With this arrangement a circularity test is usually carried out in that the rotary pivot head describes a circle about the holder in a plane parallel to the machine table and the signals of the length-measuring instrument are recorded. Deviations of the machine movement from the ideal circular path (= radius constant) are thus recognisable as change in the spacing between holder and tool mount. This method, too, is suitable only for checking linear axes, since only the movement of one plane is detected.

There is accordingly a need for an improved method for diagnosis of the movement precision of a rotary pivot head of a multi-axis machine tool, in particular to achieve a high measuring accuracy, enable simple and correct measurements and largely overcome the stated disadvantages with respect to measurement restriction to individual axes.

According to the present invention there is provided a method for diagnosis of movement precision in multi-axis machine systems with a rotary pivot head, wherein the movement path of the rotary pivot head is detected by continuous measurement of the change in the spacing between a first reference point at the rotary pivot head and a second, stationary reference point in dependence on the drive control (target value presetting) of the rotary pivot head, characterised in that the change in the spacing is measured in dependence on the drive control (target value presetting) of at least two different axes of the rotary pivot head.

Preferably, the first reference point lies on an extension of one axis of rotation of the rotary pivot head and preferably the second reference point lies on an extension of another such axis of rotation. For preference, a length-measuring device rotatably mounted at both reference points is used for measuring the change in the spacing.

Expediently, a comparison is undertaken between the theoretical anticipated values of the change in spacing and the measured values and deviations between the theoretical and the measured values of the change in spacing are detected as machine errors. The comparison can be carried out by way of a computer system. The computer system can store data which represents characteristic changes in spacing in the case of typical machine errors and the comparison can include an identification of the machine error present.

An example of the method according to the invention will now be more particularly described with reference to the accompanying drawings, in which : Fig. 1 is a schematic view of a system for measuring the movement precision of a rotary pivot head; Fig. 2 is a diagram illustrating measurement curves in the case of an error-free rotation about a pivot axis A and a rotational axis C (ideal state); Fig. 3 is a diagram illustrating the rotary pivot head and corresponding curves in the case of tipping of the rotational axis C; Fig. 4 is a diagram illustrating the rotary pivot head and corresponding curves in the case of tipping of the pivot axis A; Fig. 5 is a diagram illustrating the rotary pivot head and corresponding curves in the case of tipping not only of the rotational axis C, but also of the pivot axis A; and Fig. 6 is a diagram illustrating the influence of reversal play on the curves obtained in the case of rotation about the rotational axis C or pivot axis A.

In the described example, the method uses the measurement possibilities of the"ball-bar" test in expanded form : as in the case of the "ball-bar" test, in the event of movements of the rotary pivot head, changes in the spacing between a tool mount and a fixed reference

point in space are detected by a length-measuring device, wherein, however, the interaction of different axes is taken into consideration.

Fig. 1 shows the preferred measuring system with a length-measuring device 2, which in this example is illustrated as a rod-shaped "ball-bar" length-measuring device, but can also be another form of high resolution distance-measuring apparatus, for example, a laser. At one end, the length-measuring device 2 is rotatably connected, at the first reference point P1, with a rotary pivot head 1 in the region of a tool mount. The other end of the device 2 is rotatably mounted at a second reference point P2, which is disposed in stationary position at a holder 3. The holder 3 can, as illustrated in Fig. 1, be fastened to a machine table 4 or be rigidly fixed relative to the rotary pivot head 1 in another manner. For preference, the first reference point P1 lies as exactly as possible on a first axis C of the rotary pivot head 1 and the second reference point P2 on a second axis A of the head.

This yields specific measurement curves for different movements of the head (Figs. 2 to 5). Another choice of the position of the reference point is equally possible, in which case other spacing changes and correspondingly changed measurement curves result.

For detection of the movement precision of the rotary pivot head 1 different axes are controlled in drive and the changes in spacing between the two reference points P1 and P2 are continuously measured in dependence on a target value presetting (rotational angle, tilt angle, etc. ). For a diagnosis, a corresponding graphic illustration of the spacing change in dependence on one of the target presettings, for example, on the angle of rotation about the C axis, can be subsequently produced, for example, in polar co-ordinate illustration. In addition, the theoretical graphs of an ideal, error-free machine can be calculated for each movement with known co-ordinates of the reference points P1 and P2.

A comparison between ideal and actually measured curves enables statements to be made about the quality of the tested machine.

Fig. 2 shows, for a multi-axis machine without defects (ideal state), the measurement results, which are to be expected with the preferred measuring arrangement (Fig. 1), for rotations of the rotary pivot head about the A axis or C axis.

There is illustrated in each case the measured spacing D between the reference points P1 and P2 in dependence on the rotational angle (p of the respective axis (polar co-ordinate illustration). Due to the positions of the reference points P1 and P2 on the notional

extension of the rotational axes (A, C), no change of the spacing D is measured when only one axis (A or C) is controlled in drive (circular measurement curves).

In Fig. 3 there is schematically illustrated the behaviour in the case of a (faulty) tilting of the C axis. Rotation about the A axis delivers, in corresponding illustration, an elliptical curve form, whilst rotation about the C axis delivers a circle without change.

The effect of a (faulty) tilted A axis is reproduced in Fig. 4. In this case, rotation about the A axis again yields an elliptical curve form, whilst rotation about the C axis yields an unchanged, approximately elliptical curve form.

Fig. 5 illustrates the situation with a (faulty) tilting of not only the A axis, but also the C axis. The measurement curves arising in this case represent in specific manner a superimposition of the deviations, as they occur in maladjustment of the individual axes.

Fig. 6 shows the influence of a reversal play, which is present, of the A axis: a controlled 180 curvature about the A axis is carried out, reduced by the component of the reversal play. A rotation about the C axis remains uninfluenced by that.

All evaluations can be undertaken by means of a computer, such as a personal computer, which, for example, on the one hand receives the signals, delivered by the length measuring device 2, by way of an analog-to-digital converter in digital form and on the other hand takes over the drive control (target value presetting) of different linear and rotational axes of the rotary pivot head. A particular advantage of the computer-assisted evaluation is the possibility of storing corresponding graphs of spacing changes in the case of typical machine errors (bearing defect, reverse play, etc.) as a database and utilising the graphs for error identification. Thus, after performance of a measurement under predetermined conditions (defined rotational tilt process, defined positions of the reference points P1, P2) not only the presence and the amount of a path deviation, which is present, of the rotary pivot head can be established, but also an error association can be indicated by identification of characteristic path deviations. This error identification is not restricted to a specific multi-axis machine system: individual data sets of the respective characteristic error influences can be used for different machine systems.

By way of an automated evaluation of that kind, diagnoses of the movement precision can be produced in a very short time. In that case the effort is restricted almost exclusively to the mounting of the length-measuring device 2 and a single adjustment thereof; everything else (performance of different movement programs, data detection, graphical illustration and evaluation) is carried out subsequently by the computer, wherein recognised machine errors are indicated directly.

The preferred method can be realised in appropriately economic manner. Apart from a commercially available computer system (PC), there is required only an appropriate highresolution, length-measuring device, which is available in various versions.

The method is particularly suitable for mobile use at different locations, since the requisite apparatus elements (PC, length measuring device) can be readily transported. In that case, different multi-axis machine systems can be tested by one and the same method, wherein only the respective individual drive control and evaluating programs (with corresponding data relating to characteristic error influences) have to be provided in the computer system.

Claims

CLAIMS 1. A method of diagnosing movement precision in a multi-axis system, comprising the steps of detecting the movement path of a system component controllable for rotational movement about at least two different axes by continuous measurement of the change in spacing between a first reference point at the component and a stationary second reference point during such movement, the change in spacing being measured in dependence on preset control values for movement of the component about the at least two axes.
2. A method as claimed in claim 1, wherein the first reference point is disposed substantially on an extension of one of the axes.
3. A method as claimed in claim 2, wherein the second reference point is disposed substantially on an extension of the other or another one of the axes.
4. A method as claimed in any one of the preceding claims, wherein the measuring is carried out by measuring means pivotably mounted at the first and second reference points.
5. A method as claimed in any one of the preceding claims, comprising the steps of comparing measured values of the change in spacing with anticipated values for the change and recognising error in the system in the event of a difference between the compared values.
6. A method as claimed in claim 5, wherein the comparison is carried out by computing means.
7. A method as claimed in claim 6, comprising the step of comparing data representing the recognised error with stored data representing changes in spacing characteristic of typical system errors and identifying the recognised error on the basis of the data comparison.