Method and device for near-surface non-destructive testing of a rotationally symmetrical workpiece with sectionally changing diameter using ultrasound

Abstract

Method for near-surface non-destructive testing of a rotationally symmetrical workpiece (1) with sectionally changing diameter by means of ultrasound, comprising the following method steps: a. Impinging an ultrasonic test pulse into the workpiece (1) at a coupling point E at a defined angle of incidence Theta, b. Recording an ultrasonic echo signal from the workpiece (1) preferably at the angle of incidence Theta, c. Selection of a runtime interval I depending on the sound path W of the ultrasonic test pulse in the workpiece (1), wherein the selected runtime interval I corresponds to a near-surface region ROI of the workpiece (1), wherein a start of the runtime interval I is defined based on the highest possible sound speed in the workpiece (1) and an end of the runtime interval I is defined based on the lowest possible sound speed in the workpiece (1), d. Generating an echo value G by analyzing the ultrasonic echo signal in the selected transit time interval I, and e. Generating a representation (50) of the surface of the workpiece (1), wherein the echo value G is reproduced in spatial resolution in the representation (50).

Description

[0001] The present invention relates to a method and a device for near-surface, non-destructive ultrasonic testing of a rotationally symmetrical workpiece with a sectionally varying diameter. The workpiece has no internal bore. The method and device are particularly suitable for testing a workpiece with an anisotropic speed of sound. An anisotropic speed of sound is frequently observed, for example, in forged parts, such as forged solid shafts, which can be used in rail vehicles.

[0002] Rail vehicle wheelsets typically comprise a pair of wheels mounted on a rigid solid or hollow shaft. The shafts used often have varying outer diameters in certain sections, for example, defined areas for accommodating functional components such as the wheels, rolling bearings, or brake discs. It is obvious that the shafts of rail vehicle wheelsets are safety-relevant components that are subject to natural wear over the long service life of rail vehicles. For this reason, non-destructive testing methods must be used not only to ensure the freedom from defects during manufacture of rail vehicle wheelsets. Rather, regular inspection of all components, in particular the wheels and the shaft used, is required throughout the entire service life of a wheelset.In practice, the most common wear phenomenon observed in the shafts of rail vehicle wheelsets is the occurrence of incipient cracks, i.e., crack-like fatigue fractures originating from the surface of the respective shaft. Every rail vehicle operator must therefore maintain suitable testing procedures and equipment to regularly check the integrity of rail vehicle wheelsets for defects.

[0003] From the article by Wüstenberg, H.: "New Approaches to Ultrasonic Testing of Railway Axles with Array Probes," DGZfP Annual Conference, Berlin, 2001, pages 1-11, a method and device for near-surface non-destructive testing of a rotationally symmetrical workpiece in the shape of a railway shaft are described. For this purpose, ultrasonic test pulses are transmitted into the workpiece at a specific launch location at a defined angle of incidence. The echo signal of this ultrasonic test pulse is recorded. Taking into account a propagation time interval dependent on the sound path traveled, an echo value is determined and can then be visualized two-dimensionally.

[0004] To date, the ultrasonic testing of rotationally symmetric workpieces with varying diameters in sections, especially solid shafts of rail vehicles, has been a particularly challenging task. This is due, on the one hand, to the fact that rotationally symmetric workpieces with varying diameters in sections generally have only a few surfaces suitable for coupling ultrasound. Furthermore, such workpieces are often forged pieces, which generally exhibit increased anisotropy of the speed of sound for ultrasound.

[0005] Furthermore, the inspection of a rail vehicle's wheelset often requires the vehicle to be stationary, which directly results in high downtime costs due to the vehicle's failure. To minimize these costs, it would be desirable to be able to inspect a fully assembled wheelset, i.e., a wheelset with mounted bearings and / or brake discs. Once these are mounted, the current testing methods do not allow for sound irradiation of the shaft shank or the end face (e.g., with a conical probe).

[0006] Finally, the generation of an easily interpretable representation of the results of an ultrasonic test obtained on a rotationally symmetrical workpiece represents, to the applicant's knowledge, a problem that has not yet been solved.

[0007] It is therefore an object of the present invention to provide a method and a device for near-surface non-destructive ultrasonic testing of a rotationally symmetrical workpiece with a sectionally varying diameter, wherein the workpiece has no internal recess. The results of the ultrasonic testing should be particularly easy to interpret.

[0008] This object is achieved by a method having the features of claim 1 and by a device having the features of claim 15. The subclaims disclose advantageous developments of both the method and the device according to the invention, whereby the subclaims may be freely combined with one another within the scope of what is technically reasonable. However, such a combination is not mandatory.

[0009] The method according to the invention is used for near-surface, non-destructive ultrasonic testing of a rotationally symmetrical workpiece with a sectionally varying diameter. The workpiece does not have a rotationally symmetrical internal recess. For simplicity, such a workpiece is also referred to as a "solid shaft." In its simplest form, the method comprises the following steps: a. Impingement of an ultrasonic test pulse into the workpiece at a coupling point E under a defined angle of incidence Theta, b. Recording an ultrasonic echo signal from the workpiece preferably at the angle of incidence Theta, c. Selection of a runtime interval I depending on the sound path W of the ultrasonic test pulse in the workpiece, whereby the selected runtime interval I corresponds to a preselected near-surface region ROI (ROI: “region of interest” = defect expectation area) of the workpiece, d. Generating an echo value G by analyzing the ultrasonic echo signal in the selected travel time interval I, and e. Generating a representation of the surface of the workpiece, whereby the echo value G is reproduced in a spatially resolved manner in the representation.

[0010] The sound path W of the ultrasonic test pulse in the workpiece is generally dependent on the workpiece geometry, the coupling location E, the insonification angle Theta, the insonification direction Phi (see below) and the acoustic properties of both the workpiece and the ultrasonic test head used to generate the ultrasonic test pulse.

[0011] The method according to the invention provides a testing method which allows the result of an ultrasonic test of a rotationally symmetrical workpiece for defects close to the surface to be represented in a manner which is particularly easy for an inspector to grasp. The representation of the surface of the workpiece generated according to the invention is two-dimensional, for example in the form of a C-image, or spatial, with the spatial, i.e. three-dimensional representation being preferred. It is obvious to a person skilled in the art that the generation of a graphical representation of the workpiece described here comprises both the generation of a dataset which represents a representation of the workpiece, for example in the sense of a CAD model, and the actual display of a graphical representation of the workpiece on a suitable display unit, for example on a suitable display which can be connected, for example, to a device according to the invention.

[0012] In a preferred embodiment of the method, a point w on the surface of the representation of the workpiece is assigned to the sound path W of an ultrasonic test pulse in the workpiece. For example, the location of the first impact of the ultrasonic test pulse on the inner surface of the workpiece can be used for the point w. In the graphic representation of the surface of the workpiece, the assigned echo value G of the ultrasonic test pulse is then reproduced at this point w in a suitable manner, e.g., by local color or brightness coding. This will be described in more detail below as an example within the framework of the exemplary embodiment.

[0013] If the largest amplitude of the ultrasonic echo signal occurring in the selected transit time interval I is used as the echo value G, a particularly preferred embodiment of the test method according to the invention is obtained due to its simple technical implementation.

[0014] In a further advantageous development of the method according to the invention, the ultrasonic echo signal is subjected to a propagation time-dependent and / or angle-dependent amplification at least in the selected propagation time interval I. This allows sound-attenuating effects, e.g., caused by the geometric expansion of the sound field along the propagation direction, its attenuation in the workpiece, e.g., due to scattering at anisotropies, as well as any angular dependence of the reflection of the ultrasonic test pulse at an inner boundary surface of the workpiece, to be compensated. Consequently, defects of the same size and orientation generate echo signals of approximately the same magnitude regardless of their position in the workpiece, which in turn further improves the interpretability of the test method results.

[0015] It is obvious to those skilled in the art that the recorded ultrasonic echo signal, particularly after digitization, can be subjected to suitable signal processing, for example, to improve the signal-to-noise ratio. Various methods for this are known from the prior art.

[0016] The signal-to-noise ratio can also be significantly improved if the process steps a to d are repeated several times for a fixed coupling location E and a fixed angle of incidence Theta and an average value <g>of the generated echo values G. This average <g>is then displayed in step e. with spatial resolution.

[0017] In a particularly advantageous development of the method according to the invention, a plurality of successive ultrasonic pulses are transmitted into the workpiece at different angles of incidence θ. It is possible to vary the angle of incidence θ from pulse to pulse, but it is also possible to vary the angle of incidence θ for a subsequent pulse sequence only after a finite sequence of pulses at the same angle of incidence θ. Thus, averaging the error signals to be evaluated across a plurality of echo signals resulting from a plurality of ultrasonic test pulses coupled at the same angle of incidence θ improves the signal-to-noise ratio. The method according to the invention is preferably carried out for each ultrasonic test pulse transmitted into the workpiece.In a further preferred embodiment of the method, the position of the coupling location E on the surface of the workpiece is kept substantially constant relative to its axis of symmetry S. "Kept substantially constant" in this context means in particular that the position X of an ultrasonic probe, which comprises an ultrasonic transducer for generating the ultrasonic test pulses, is kept constant relative to the axis of symmetry S of the workpiece. In the ultrasonic probes commonly used in practice for angled sounding with a variable beam angle, in which the ultrasonic transducer is arranged on a wedge-shaped forward beam, for example, the actual coupling location changes slightly when the beam angle changes. This effect can be neglected here.

[0018] So-called "phased array" ultrasonic probes, which are known in the art and whose use in connection with the present invention will be discussed in more detail below, allow electronic tuning of the insonification angle Theta over a wide angular range. Particularly in connection with the preferred embodiments of the method according to the invention, in which the insonification angle Theta is varied between different ultrasonic test pulses of a test pulse sequence, the use of such "phased array" probes with electronically tunable insonification angle Theta has proven particularly advantageous. Particular advantages arise when probes according to the teaching of the WO 2010 / 130 819 A9 family are also used, which teaching is incorporated by reference into the disclosure of the present application.The use of such probes allows the curvature of the coupling surface to be taken into account in the axial and radial directions, which is particularly advantageous for shafts of wheel sets, whose shaft geometries can be partially or completely curved in the longitudinal direction, so that - at least with components mounted on the shaft such as impellers, bearings or brake discs - there is no purely cylindrical area with a constant diameter for the ultrasonic coupling.

[0019] Preferably, the one or more ultrasonic test pulses are transmitted into the workpiece in such a way that the sound path W of the ultrasonic test pulse(s) in the workpiece and the symmetry axis S of the rotationally symmetric workpiece span a common plane, i.e., the sound path W of the ultrasonic test pulse(s) intersects the symmetry axis S of the rotationally symmetric workpiece. This common plane is also referred to below as the insonification plane P. The preselected near-surface region (ROI) is generally located after this intersection point.

[0020] In a further preferred development of the method according to the invention, the relative position of probe position X and workpiece is not changed during the execution of a first method section of the testing method, in which the insonification angle Theta is continuously changed. This means that the above-mentioned condition is met for all ultrasonic test pulses coupled into the workpiece in this first method section.

[0021] Since the workpiece is rotationally symmetrical, effective ultrasound transmission through the workpiece can be achieved in particular by a relative rotation of the workpiece and ultrasonic test head around the workpiece's axis of symmetry, characterized by a rotation angle beta. In a further preferred development of the method according to the invention, a relative rotational movement of the test head and workpiece of the type mentioned is therefore realized in a second method section, wherein the rotation angle beta is preferably at least 360°. Preferably, the above-mentioned condition is also met, according to which the sound path W of each ultrasonic test pulse coupled into the workpiece intersects the workpiece's axis of symmetry S.The easiest way to achieve a relative rotational movement of the test head and the workpiece around the symmetry axis S of the workpiece is to rotate the workpiece around its symmetry axis S under the test head which is held in its position X.

[0022] In a particularly preferred embodiment of the method according to the invention, while maintaining the position of the ultrasonic test head relative to the axis of symmetry S of the workpiece, a sequence of ultrasonic test pulses is transmitted into the workpiece while simultaneously varying the angle of incidence theta and the angle of rotation beta. In this case, for example, after traversing a predetermined interval for the angle of incidence theta, a step-by-step relative rotational movement of the ultrasonic test head and workpiece about the axis of symmetry S of the workpiece is carried out. For example, electronic tuning of the angle of incidence theta over an angular range of at least 30° to 60°, preferably of at least 20° to 75° is possible. Subsequently, a relative rotational movement of the test head and workpiece about the axis of symmetry S of the workpiece is carried out by, for example, a maximum of 5°, preferably a maximum of 1° and particularly preferably a maximum of 0.5°.For this new relative position of the probe and the workpiece, a sequence of ultrasonic test pulses is then transmitted into the workpiece at a varying angle of incidence Theta. This is followed by another relative rotation of the probe and the workpiece, and so on. Overall, the relative angle of rotation Beta of the probe and the workpiece around the symmetry axis S of the workpiece should be at least 360° over a complete test cycle; preferably, it is 360° or an integer multiple of 360°.

[0023] In an alternative preferred development of the method according to the invention, the insonification angle Theta and the relative rotation angle Beta of the workpiece and ultrasonic test head about the symmetry axis S of the workpiece are varied simultaneously, wherein the rotation speed, for example of the workpiece about its own symmetry axis S, is preferably selected to be so low that a sufficient geometric coverage of the ultrasonic test pulses in the ROI in the workpiece is still achieved.

[0024] In a further preferred development of the method according to the invention, two groups of ultrasonic test pulses are transmitted into the workpiece. The first group of ultrasonic test pulses has a direction of travel that includes a component in the positive direction of the symmetry axis S of the workpiece. The second group of ultrasonic test pulses, in contrast, has a direction of travel that includes a component in the negative direction of the symmetry axis S of the workpiece. Preferably, the first and second groups of ultrasonic test pulses are coupled into the workpiece at essentially the same location. For this purpose, it is particularly possible to integrate two ultrasonic transducers into a test head, which emit the first and second groups of ultrasonic test pulses.By means of this advantageous development of the method according to the invention, it is possible to practically double the tunable angular range and thus the spatial area of the workpiece to be detected by a test head position X (relative to the symmetry axis S of the workpiece), which enables a further increased efficiency in the execution of the method.

[0025] Advantageously, the testing method according to the invention is repeatedly performed for various probe positions X on the surface of the workpiece. This third method step serves to capture the largest possible (near-surface) volume of the workpiece. As a rule, performing the method at a few discrete probe positions X is sufficient to capture the entire (near-surface) volume of the workpiece, even with jagged workpiece geometries.

[0026] If all three process steps are completed for a workpiece, the entire near-surface volume of the workpiece can be scanned and tested with ultrasonic test pulses for most workpiece geometries. The resulting representation of the workpiece surface according to the invention thus contains complete information about the result of the ultrasonic testing of the entire near-surface volume of the workpiece. The graphic representation of the result of the ultrasonic testing method according to the invention described here is particularly informative because, during the testing of the workpiece, a complete relative rotation of the test head and workpiece by 360° or an integer multiple thereof around the workpiece's axis of symmetry has occurred.Since the method is still carried out from various probe positions X, the entire volume of the near-surface region of the workpiece is scanned and subsequently graphically displayed – assuming a suitable geometry of the rotationally symmetrical workpiece. A particularly advantageous development of the method according to the invention allows the process duration to be reduced by effectively reducing the amount of data to be analyzed. This is achieved by limiting the evaluation of the recorded ultrasonic echo signals, which correspond to sometimes very long paths of the test pulse in the workpiece, which occur primarily at large angles of incidence θ, to those echo signals that result from a preselected near-surface region of the workpiece to be examined. This preselected near-surface region is also referred to as ROI (= "region of interest") in the context of the present invention.Typically, the ROI to be used during the procedure is determined by the inspector based on the material properties and the geometry of the workpiece. Typically, the ROI is selected adjacent to the inner surface of the workpiece where the first reflection of the ultrasonic test pulse occurs within the workpiece.

[0027] The ROI can, for example, be limited to the spatial area of the workpiece that extends radially inwards from the surface of the workpiece by a few to a few tens of millimeters, for example by 30 to 60 millimeters, preferably by 40 millimeters.

[0028] The ROI can also be defined differently in sections along the symmetry axis of the workpiece, e.g. in areas with changing shaft diameters it can have a larger extent than in areas with constant diameter.

[0029] In this way, ROI can also be deliberately selected to be larger in certain sections, e.g. to display indications from a wheel, bearing or brake seat that may be formed on the solid shaft.

[0030] Due to the existing uncertainty regarding the speed of sound in a forged workpiece, it is advantageous to limit the ROI not only to the point where the ultrasonic test pulse hits the inner surface of the workpiece, but also to extend it a certain time beyond that. This means that total internal reflection may occur at the surface of the workpiece within the ROI. However, at least the ultrasonic test pulse reaches the inner surface of the workpiece with a very high degree of certainty.

[0031] The transit time interval to be selected, corresponding to the ROI, refers to the response time between the coupling of the ultrasonic test pulse into the workpiece and the arrival of ultrasonic echo signals. The workpiece geometry is assumed to be known, as are the acoustic properties of the workpiece. Furthermore, the coupling location E of the ultrasonic pulse, the angle of incidence theta and the direction of incidence are known. The direction of incidence can be defined, for example, via the inclination angle phi of the sound propagation direction against the plane defined by the axis of symmetry S and the coupling location E. In preferred embodiments of the method according to the invention, the inclination angle phi is zero, i.e. the sound path W and the axis of symmetry S of the workpiece span a common plane P. The coupling location E is directly linked to the probe position X on the surface of the workpiece and to the angle of incidence theta.From this, the sound path W of the ultrasonic test pulse in the workpiece can be determined, which, for a given workpiece geometry and given workpiece properties, is generally a function of the probe position X, the angle of incidence Theta and the angle of inclination phi. In particular, the propagation time tROI_ON can be determined after which the ultrasonic test pulse enters the ROI predefined by the inspector. Furthermore, a propagation time tROI_OFF can be specified after which the first reflection of the ultrasonic test pulse has occurred on an inner surface of the workpiece. For a given angle of incidence Theta, the ROI can be defined via this propagation time interval I, i.e. every echo signal that is recorded after a response time tResponse with 2tROI_ON ≤ tResponse ≤ 2tROI_OFF results from an ultrasonic reflector (e.g. a local anisotropy in the material structure of the workpiece, the local workpiece geometry, a defect) in the ROI.It is obvious that the travel time interval I usually depends on the given angle of incidence Theta.

[0032] The ROI is defined via the travel time interval I selected for a given beam angle Theta (and thus to be analyzed). This is based on the speed of sound for the ultrasonic test pulse in the workpiece, which can only be specified precisely up to a certain degree of uncertainty. The start of the travel time interval I is defined by the time 2tROI_ON, at which the ultrasonic test pulse first hits the inner surface at the earliest, i.e. the highest possible speed of sound is assumed. The end of the travel time interval I is defined by the time 2tROI_OFF, at which the ultrasonic test pulse first hits the inner surface at the latest, i.e. the lowest possible speed of sound is assumed. This ensures that the ultrasonic test pulse reliably hits the inner surface of the workpiece in the selected travel time interval, i.e. the inner surface is always within the ROI.

[0033] If necessary, the time of flight interval I to be analyzed and thus also the ROI can be additionally increased by a defined "surcharge" (e.g. ±5%, ±10%, ±15%) to the maximum or minimum assumed speed of sound. This represents an advantageous development of the above-mentioned preferred embodiment. This can ensure that a near-surface area with a defined, in particular constant thickness of, for example, 30 to 60 mm, preferably 40 mm and more, is always examined.

[0034] According to the particularly preferred development of the method according to the invention, the analysis of the ultrasonic echo signal recorded from the workpiece at the angle Theta for error signals Fi is limited to the selected runtime interval I, which corresponds to the near-surface region of the workpiece.

[0035] In its advantageous further development, the invention provides a practical method for effective data reduction to an ROI that can be individually defined by the user for the respective inspection task. This effective data reduction allows the use of very high pulse repetition rates in the range of up to several kHz and the highest temporal resolution in the analysis of the ultrasonic echo signals. Furthermore, using the method and device, near-surface defects in the workpiece can be reliably detected, even with jagged workpiece geometry and possibly additional components mounted on the workpiece surface. The method and device can be used so effectively that excessively long inspection times are avoided.

[0036] In an alternative approach, which is also intended to be encompassed by the invention, the time-resolved echo signal is digitized essentially over a transit time interval I from entry into the workpiece to twice the transit time until the first impact on the inner surface of the workpiece on the side opposite the test head, thereby generating a large amount of raw data. This is reduced to a subset of data points to be analyzed by selecting only those data points whose origin lies in the previously defined ROI. Both approaches ultimately select the same subset of data points / echo signals to be analyzed. They can therefore be considered equivalent in their results.

[0037] A device according to the invention is provided for near-surface, non-destructive ultrasonic testing of a rotationally symmetrical workpiece with a sectionally varying diameter, wherein the workpiece does not have a rotationally symmetrical internal recess. It is particularly suitable for testing forged solid shafts of rail vehicle wheelsets. A device according to the invention has at least the following features: a) a test head (40) for transmitting an ultrasonic test pulse into the workpiece (1) at a defined angle of incidence Theta and for recording an ultrasonic echo signal from the workpiece (1), b) a control unit (20) which is arranged to i. to control the test head (40) to transmit an ultrasonic test pulse into the workpiece (1) at the defined angle of incidence Theta, ii. to record an ultrasonic echo signal from the workpiece (1) by means of the probe (40), preferably at the angle theta, iii. to select a travel time interval I depending on the sound path W of the ultrasonic test pulse in the workpiece (1), wherein the selected travel time interval I corresponds to a near-surface region ROI of the workpiece (1), and iv. to generate an echo value G by analyzing the recorded ultrasonic echo signal in the selected propagation time interval I, and v. to generate a representation (50) of the surface of the workpiece (1), wherein the echo value G is reproduced in spatial resolution in the representation (50).

[0038] A device according to the invention is particularly suitable for carrying out the method according to the invention. In advantageous developments of the device, the above-described advantageous embodiments of the method according to the invention are implemented in the control unit. These various embodiments therefore allow, in particular, the realization of those advantages that have already been discussed in connection with the method according to the invention, to which reference is made here.

[0039] In a further preferred embodiment, the testing device comprises a guide device configured to align the test head relative to the symmetry axis S of the workpiece such that the sound path W of the ultrasonic test pulse in the workpiece and the symmetry axis S span a common plane, the insonification plane P. This means that the direction of travel of the ultrasonic test pulses transmitted from the test head into the workpiece has a component in the direction of the symmetry axis of the workpiece. By ensuring the above-described direction of travel of the ultrasonic test pulses transmitted into the workpiece by means of the guide device, a particularly simple sound field is created in the workpiece. This simplifies subsequent signal processing and evaluation.

[0040] In a particularly preferred development of the testing device according to the invention, the probe has an ultrasonic transducer divided into a plurality of individually controllable transducer segments. Such probes are known from the prior art; they are referred to as "phased array" probes and, with suitable electronic control of the individual transducer segments, allow, for example, the angle of incidence of the ultrasonic pulses generated by the ultrasonic probe into the workpiece to be electronically controlled. Ultrasonic probes according to the teachings of the WO 2010 / 130 819 A9 family are particularly preferably used. Furthermore, in a preferred embodiment, the control unit is configured to control a phased array-type probe in the aforementioned manner, so that the angle of incidence Theta into the workpiece can be electronically adjusted.Furthermore, the control unit is designed to transmit a sequence of ultrasonic test pulses into the workpiece at different angles of incidence Theta by means of the test head.

[0041] In a further preferred development of the testing device according to the invention, it further comprises a rotating device. The rotating device is designed to generate a relative movement between the test head and the workpiece in such a way that the workpiece is rotated about its axis of symmetry S beneath the test head. The rotating device preferably comprises a means for detecting the angle of rotation Beta of the relative movement, e.g. an encoder. Furthermore, it is preferably connected to the control unit of the testing device in such a way that the detected angle of rotation Beta of the rotary movement can be transmitted to the control unit. In a simplified embodiment of this device, the angle of the relative movement between the test head and the workpiece actually applied by the rotating device is not detected and transmitted by the rotating device to the control unit.Rather, the control unit is configured to control the rotary device in such a way that it generates a relative movement of the test head and the workpiece by a rotation angle beta specified by the control unit. Detection of the angle of the actual rotational movement is not necessary here, meaning that, for example, an encoder can be omitted.

[0042] In a further preferred embodiment of the testing device according to the invention, the test head of the testing device has two ultrasonic transducers. These are characterized in that the propagation direction of a first part of the pulses has a component oriented in the direction of the symmetry axis S of the workpiece, and the propagation direction of a second part of the pulses has a component oriented opposite to the direction of the symmetry axis S. A particularly compact design is achieved if the two ultrasonic transducers are mounted on a common forward body, which can be made of polystyrene, polycarbonate, or Plexiglas, for example, and arranged in a common test head housing.

[0043] Finally, in a further advantageous embodiment, the device according to the invention comprises a display unit, e.g., an LCD, connected to the control unit. The control unit is then configured to generate a graphic representation of the workpiece on the display unit.

[0044] Further advantages and features emerge from the subclaims and the following exemplary embodiments. These exemplary embodiments are not to be understood as limiting; they serve to assist those skilled in the art in understanding the invention generally described above. The exemplary embodiments are explained in more detail with reference to the drawings, which show: Fig. 1: a side view of a typical solid shaft of a railway vehicle wheelset, Fig. 2: a schematic representation of a test head and a control unit according to a first embodiment of a test device according to the invention, Fig. 3: a partial sectional view through the solid shaft from Fig. 1 to illustrate the sound paths of the ultrasonic test pulses in the workpiece and the ROI, Fig. 4: a diagram showing the data reduction achieved by introducing ROI, Fig. 5: a C-scan image taken at the full wave section according to Fig. 3, and Fig. 6: a three-dimensional representation of the solid wave section from Fig. 3 with detected error signals Fi drawn in.

[0045] Fig. 1 shows a side view of a typical solid shaft 1 of a rail vehicle wheelset. It is a rotationally symmetrical forging with a sectionally varying diameter, as can be seen from Fig. 1. The shaft 1 has, in particular, various sections of constant diameter, which are provided for accommodating the wheel hubs, the rolling bearings with which the solid shaft is rotatably mounted on the rail vehicle, and a centrally arranged brake disc. As a forged part, a solid shaft according to Fig. 1 typically exhibits a certain anisotropy of the speed of sound for ultrasound, which is caused by local structural changes caused by the forging process. The solid shaft 1 is rotationally symmetrical to the indicated rotation axis S.

[0046] Fig. Figure 2 shows a first embodiment of a testing device 10 according to the invention, which comprises a control unit 20 and a test head 40 connected thereto. The test head 40 comprises a segmented ultrasonic transducer 42 of the phased array type. It therefore comprises a plurality of individually controllable transducer elements (not shown). The segmented ultrasonic transducer 42 is arranged on a forward-moving body 44, which in turn consists of a material suitable for angled sound beams into a forged steel workpiece. The forward-moving body 44 is often made of polystyrene, polycarbonate, or Plexiglas®. As a rule, both the forward-moving body 44 and the segmented transducer 42 are arranged in a common test head housing (not shown) for shielding against environmental influences. Fig. 2, the test head 40 is shown placed on the cylindrical surface of a rotationally symmetrical workpiece 1, which is, for example, the workpiece Fig. 1 can be the solid shaft 100. The contact surface formed by the advance body 44, with which the test head is placed on the surface of the workpiece 1, therefore also has a hollow cylindrical shape, the inner diameter of which is adapted to the outer diameter of the workpiece 1. As already explained in the introduction, there are many different shaft geometries which can also be completely curved in the longitudinal direction, i.e. it is possible for the workpiece to be examined not to have a purely cylindrical region with a constant diameter. Using the technical teaching known from WO 2010 / 130 819 A9, the use of test heads is also possible whose advance bodies are adapted to the cross-section of the workpiece both in the longitudinal direction and in the transverse direction. This adaptation is generally carried out locally, i.e. for a predetermined X-position relative to the axis of symmetry S of the workpiece.

[0047] The control unit 20 is configured to control the test head 40 in such a way that it generates an ultrasonic test pulse that is coupled into the workpiece 1 at a defined angle of incidence Theta. The control unit 20 is further configured to adjust the angle of incidence Theta in a controlled manner. Fig. 2 shows exemplary three sound paths of three ultrasonic test pulses which are coupled into the workpiece 1 at different angles of incidence Theta 1, Theta 2 and Theta 3. While the angles of incidence Theta 1, Theta 2 and Theta 3 can be controlled with very good accuracy by the control unit 20, the angles of incidence Gamma 1, Gamma 2 and Gamma 3 resulting in the workpiece 1 are subject to a certain uncertainty, which is directly linked to the above-mentioned anisotropy of the speed of sound for ultrasound in the forged solid shaft 100. Likewise, Fig. 2 it is immediately apparent that with a constant probe position X and a variation in the angle of incidence Theta due to the refraction when entering the workpiece the coupling point E changes slightly, i.e. with a constant position X there is a different coupling point E1, 2, 3 for each angle of incidence Theta 1, 2, 3. If the requirements for the accuracy of the test are not too high this effect can be neglected, e.g. when determining the position w at which the sound path W strikes the inner surface of the workpiece for a given angle of incidence Theta and a given probe position X. If higher requirements for accuracy are involved it can be taken into account mathematically, e.g. when determining the position w.

[0048] The control unit 20 is further configured to record an ultrasonic echo signal from the workpiece 1 in a time-resolved manner, preferably at the angle Theta, by means of the test head 40 and subsequently to digitize it in a selected propagation time interval I. In this context, the control unit 20 is configured to select a propagation time interval I depending on the sound path W of the ultrasonic test pulse in the workpiece 1, wherein this selected propagation time interval corresponds to a near-surface region of the workpiece 1. As already mentioned in the introduction, the sound path of the ultrasonic test pulse in the workpiece is generally dependent on the workpiece geometry, the position X of the test head, the angle of incidence Theta and the angle of inclination Phi (which was defined in the introductory part and is preferably zero), and on the acoustic properties of the workpiece.In particular, the control unit 20 can be configured to allow the operator to independently define the near-surface region depending on the workpiece geometry. The specifically selected probe position can also be taken into account in this process.

[0049] The ROI is preferably defined via the travel time interval selected (and thus to be analyzed) for a given angle of incidence Theta. This is based on the speed of sound for the ultrasonic test pulse in the workpiece, which can only be specified precisely to a certain extent. The start of the travel time interval I is defined by the time 2tROI_ON, at which the ultrasonic test pulse first strikes the inner surface at the earliest, i.e. the highest possible speed of sound is generally assumed here. The end of the travel time interval I is defined by the time 2tROI_OFF, at which the ultrasonic test pulse first strikes the inner surface at the latest, i.e. the lowest possible speed of sound is generally assumed here. In individual cases, deviations may arise due to workpiece geometry and changes in travel paths W due to changes in the angle of incidence Gamma (cf. Fig. 2) when the speed of sound is varied. This ensures that the ultrasonic test pulse reliably hits the inner surface of the workpiece within the selected transit time interval I, i.e., the inner surface is always within the ROI.

[0050] If necessary, the transit time interval I to be analyzed and thus also the ROI can be additionally increased by a defined "surcharge" (e.g. ± 5%, ± 10%, ± 15%) to the maximum or minimum assumed speed of sound. This represents an advantageous development of the above-mentioned preferred embodiment. This makes it possible to always examine a near-surface area with a defined, in particular constant thickness of, for example, 30 to 60 mm, preferably 40 mm and more.

[0051] As mentioned, the control unit 20 is configured to select a "near-surface" travel time interval I. Subsequently, the control unit 20 digitizes and analyzes the recorded ultrasonic echo signals in the selected "near-surface" travel time interval I for error signals Fi, i.e., ultrasonic echo signals that indicate near-surface errors in the workpiece 1, such as incipient cracks or near-surface defects. In the simplest case, only a maximum echo amplitude in the travel time interval I is determined, and no evaluation of the echo amplitude as an "error signal Fi" or "no error signal" is performed. Rather, the echo amplitude (or a similar, but differentiated value obtained) is itself viewed as an error signal Fi, i.e., for each probe position X, each beam angle theta, and each rotation angle beta (see below), at least one error value Fi is available.

[0052] By means of Fig. 3, which is a partial sectional view of the solid shaft 100 from Fig. 1, the inventive concept of selecting a near-surface area, the region of interest, is illustrated. Fig. Figure 3 shows the sound paths W of a plurality of ultrasonic test pulses, which are coupled into the workpiece 1 at a substantially constant coupling location E by means of the stationary test head 40 arranged at position X on the surface of the workpiece 1. From one ultrasonic test pulse to the next, the angle of incidence Theta is successively varied between preset limits, which are typically between 20 and 75°. In this way, an extended section of the inner surface of the solid wave 100 opposite the test head position X or the coupling location E is scanned by means of the ultrasonic test pulses. For each ultrasonic test pulse introduced into the solid wave 100 at a specific angle of incidence Theta, the test head 40 detects in a time-resolved manner the echo signal returning from the solid wave 100 at the angle Theta. If the ROI has previously been defined as a function of the geometry of the workpiece 1 to be tested, as shown in Fig. 3 is indicated by lines 11 and 12, it is possible to determine the propagation time tROI_EIN for each angle of incidence Theta set by the control unit 20 with a known coupling location until the ultrasonic test pulse transmitted into the workpiece 1 at angle Theta reaches the ROI. This propagation time tROI_EIN corresponds to a propagation length LROI_EIN in the workpiece due to the known sound velocity in the material of the workpiece 1, as can be seen from Fig. 4 becomes clear.

[0053] Fig. 4 now shows for the Fig. 3 The ROI defined by lines 11 and 12 represents the value range I of the response time or the run length L in the solid wave 100 that must be analyzed for relevant error signals at a given insonification angle θ in order to detect errors located within the ROI. The run lengths LROI_IN (=inlet ROI) and LROI_OUT (=outlet ROI) are shown as examples for an insonification angle θ = 35°.

[0054] Here, it is possible to record the echo signal in time resolution at a given probe position X for any desired angle of incidence Theta after coupling in an ultrasonic test pulse for a specified time period I. The time period I is selected such that for the selected range of the angle of incidence Theta, the selected probe position X as well as the geometry and material properties of the workpiece, it is ensured that echo signals from the ROI are always recorded in time. This means that for each point within the Fig. 4, which lies between lines 13 and 14, a digitized echo signal is present. According to the invention, only these echo signals from the ROI are examined for error signals Fi. The echo signals to be evaluated are thus limited by selecting echo signals that are identical to echo signals that originate in the ROI. Therefore, a translation of lines 11 and 12 from Fig. 3 based on the physical laws into lines 13 and 14 in Fig. 4. The set of measuring points that lie within these two boundary lines in Fig. 4, then forms a subset of data points to be analyzed, selected according to the invention. This subset is obtained according to the approach of the present invention by digitizing and analyzing the echo signal present for a long propagation time interval only in a small time window I.

[0055] So you get the knowledge that comes from Fig. 4, already during the ultrasonic testing. For a workpiece with known material properties and known geometry, an ROI is defined analogous to the representation in Fig. 3. For a given probe position, the Fig. The relationship between the insonification angle Theta and the response time or travel distance in the workpiece, as shown in Figure 4, is used to determine, for each insonification angle Theta, the response time interval I in which signals attributable to defects in the ROI are to be expected. For a given probe position X, the ultrasonic test is then limited to the aforementioned response time interval I for each electronically adjusted insonification angle Theta.

[0056] Subsequently, within the scope of the method according to the invention, the correspondingly configured evaluation unit 20 analyzes those echo signals for indications of errors that can be traced back to the ROI. Such an error analysis can be based, for example, on the amplitude of echo signals. As already mentioned in the general section, all methods known from the prior art for signal evaluation and, if necessary, signal enhancement, e.g., to increase the signal-to-noise ratio, can be used.

[0057] Within the scope of the exemplary embodiment, an echo value G is determined during the error analysis, which echo value is assigned to a point w on the surface of the workpiece 1, e.g. the location of the first impact of the ultrasonic test pulse on the inner surface of the workpiece 1. In this case, the uncertainty in the angle of incidence gamma resulting from the uncertainty regarding the local speed of sound and the resulting uncertainty regarding the location of the first occurrence on the inner surface of the workpiece 1 are preferably neglected. This echo value G is determined by determining the echo signal with the highest amplitude in the selected travel time interval I. This maximum amplitude value is then assigned to the above-mentioned point (location of the first impact) on the surface of the workpiece 1.

[0058] Fig. 6 illustrates the actual conditions in a test task on a rotationally symmetrical workpiece 1. This is Fig. 6 to a three-dimensional representation 50 of the Fig. 3 visible shaft section of the solid shaft 100 from Fig. 1. A guide device (not shown) is provided with which the test head 40 can be guided on the surface of the solid shaft 100 while maintaining the position X (X position in Fig. 6) relative to the axis of symmetry S and the orientation of the probe 40 (characterized by the angle of inclination Phi relative to the insonification plane P).

[0059] During the test procedure, the solid shaft 100 is rotated by 360° or an integer multiple thereof around its axis of symmetry S by means of a rotating device (not shown), which is Fig. 6 coincides with the X-axis. The angle of rotation of the solid shaft around its axis of symmetry is designated Beta and is recorded using a suitable angle encoder (not shown). For each angle of rotation Beta, with the X-position of the probe 40 held fixed, the entire range of the insonification angle Theta accessible with the probe 40 is electronically tuned by the control unit 20. For each individual insonification angle Theta, the echo signal is recorded in a time-resolved manner and digitized in the selected transit time interval I (Theta). The data points thus obtained can be displayed in a diagram according to Fig. 4. This means that from the total number of recorded echo signals, those corresponding to the selected ROI are selected in terms of time. These echo signals are then digitized, i.e., a set of data points for analysis is generated for a given angle of incidence Theta.

[0060] Each individual point w on the surface of the solid shaft 100 is assigned an echo value G(w), which corresponds to the maximum echo signal from the ROI assigned to this point. If this procedure is carried out for a plurality of rotation angles Beta, which can, for example, be traversed stepwise at intervals of 0.5 or 1°, up to a total rotation angle Beta of at least 360°, it is possible to enter the obtained signal values into a so-called C-image. In a C-image, the signal value assigned to an echo signal is entered into a diagram according to Fig. 5, in which, for example, the angle of incidence Theta is used as the abscissa and the angle of rotation Beta of the solid wave 100 as the ordinate. The echo value G can be coded, for example, using brightness values or color. Fig. 5, a three-stage scale was used. If an echo value G remains below a limit that must be recorded, the C-scan is Fig. 5 this point is marked brightly. If it exceeds a recording limit, but is not yet assigned to a defect size considered critical, it is coded with a second (e.g. darker, e.g. orange) color value. If the echo value G finally exceeds a value assigned to a critical defect size, it is coded with a third color value, e.g. in the signal color red. A diagram resulting in this way according to Fig. 5 already has a high informative value for a skilled user of a device according to the invention.

[0061] The interpretability of the result is further improved according to Fig. 5, if the x-position (position relative to the workpiece's axis of symmetry S) of the point w on the workpiece surface to which the ROI is assigned is not used as the abscissa. The resulting representation essentially corresponds to the representation according to Fig. 5, but is for a direct transfer to the three-dimensional representation of the examined solid wave 100 generated according to the invention according to Fig. 6. The color-coded signal values are then plotted on the surface of the three-dimensionally displayed workpiece as a function of the angle of rotation Beta of the shaft 100 and the position on the symmetry axis of the shaft 100 (position on the X-axis). This results in the Fig. 6, which has a tremendously improved interpretability compared to the visualization methods previously known from the state of the art.

[0062] Particular advantages arise in particular when the representation is carried out in accordance with Fig. 6 is designed such that a rotation of the shaft 100 about its axis of symmetry S can be represented. This is possible, for example, in a CAD model of the solid shaft 100 with the spatially resolved echo values G (w) plotted on its surface. A moving representation of the rotation of the solid shaft 100 about the angle of rotation Beta as a sequence of individual images combined to form a film is also conceivable and claimed.< / g> < / g>

Claims

[1] Method for near-surface non-destructive testing of a rotationally symmetrical workpiece (1) with sectionally changing diameter by means of ultrasound, comprising the following method steps: a. Impinging an ultrasonic test pulse into the workpiece (1) at a coupling point E at a defined angle of incidence Theta, b. Recording an ultrasonic echo signal from the workpiece (1) preferably at the angle of incidence Theta, c. Selection of a runtime interval I depending on the sound path W of the ultrasonic test pulse in the workpiece (1), wherein the selected runtime interval I corresponds to a near-surface region ROI of the workpiece (1), wherein a start of the runtime interval I is defined based on the highest possible sound speed in the workpiece (1) and an end of the runtime interval I is defined based on the lowest possible sound speed in the workpiece (1), d. Generating an echo value G by analyzing the ultrasonic echo signal in the selected travel time interval I, and e. generating a representation (50) of the surface of the workpiece (1), wherein the echo value G is reproduced in spatial resolution in the representation (50). [2] Method according to claim 1, characterized by that the representation (50) of the surface of the workpiece (1) is two-dimensional or spatial. [3] Method according to claim 1 or 2, characterized by that a sound path W of an ultrasonic test pulse in the workpiece (1) is assigned a point w on the surface of the representation (50) of the workpiece (1), at which point an echo value G is reproduced. [4] Method according to claim 1, characterized by that the largest amplitude of the ultrasonic echo signal occurring in the selected transit time interval I is used as the echo value G. [5] Method according to claim 1, characterized bythat the ultrasonic echo signal is subjected to a transit time-dependent and / or beam angle-dependent amplification at least in the selected transit time interval I. [6] Method according to claim 1, characterized by that the method steps a to d are repeated several times for a fixed coupling location E and a fixed insonification angle Theta and an average value (G) of the generated echo values G is formed, which is reproduced in step e. in spatially resolved form in the representation (50). [7] Method according to claim 1, characterized by that the runtime interval I is selected so that the ultrasonic test pulse reaches the surface of the workpiece (1) within the runtime interval I. [8] Method according to claim 1, wherein ultrasonic test pulses are transmitted into the workpiece (1) by means of a test head (40) which is arranged at a test head position X relative to the axis of symmetry of the workpiece (1) on the surface of the workpiece (1), characterized by that at the test head position X, several ultrasonic test pulses are transmitted into the workpiece (1) at different angles of incidence Theta, wherein for each ultrasonic test pulse a method according to one or more of the preceding claims is carried out. [9] Method according to one or more of the preceding claims, characterized by that the sound irradiation is carried out in such a way that the sound path W of the ultrasonic test pulse in the workpiece (1) and the axis of symmetry S of the rotationally symmetrical workpiece (1) span a common irradiation plane P. [10] Method according to claim 9, characterized by that several ultrasonic test pulses are irradiated into the workpiece (1) in such a way that the irradiation plane P rotates around the symmetry axis S of the rotationally symmetrical workpiece (1). [11] Method according to claim 10, characterized by that the insonification plane P is rotated by an integer multiple of 360°. [12] Method according to claim 1, characterized by that a plurality of ultrasonic test pulses are transmitted into the workpiece (1) in such a way that the direction of travel of a first part of the pulses has a component in the direction of the axis of symmetry S of the workpiece (1) and the direction of travel of a second part of the pulses has a component which is oriented opposite to the direction of the axis of symmetry S of the workpiece (1). [13] Method according to claim 1, characterized by that the workpiece (1) has an anisotropic sound velocity for ultrasound. [14] Method according to claim 1, characterized by that the workpiece (1) is a forged piece, in particular a forged solid shaft. [15] Testing device (10) for near-surface non-destructive testing of a rotationally symmetrical workpiece (1), with sectionally changing diameter by means of ultrasound, having the following features: a test head (40) for transmitting an ultrasonic test pulse into the workpiece (1) at a defined angle of incidence Theta and for recording an ultrasonic echo signal from the workpiece (1), b. a control unit (20) which is designed to i. to control the test head (40) to transmit an ultrasonic test pulse into the workpiece (1) at the defined angle of incidence Theta, ii. to record an ultrasonic echo signal from the workpiece (1) by means of the test head (40), preferably at the angle theta, iii. to select a travel time interval I depending on the sound path W of the ultrasonic test pulse in the workpiece (1), wherein the selected travel time interval I corresponds to a near-surface region ROI of the workpiece (1), and wherein a start of the travel time interval I is defined based on the highest possible sound speed in the workpiece (1) and an end of the travel time interval I is defined based on the lowest possible sound speed in the workpiece (1), iv. to generate an echo value G by analyzing the recorded ultrasonic echo signal in the selected travel time interval I, and v. to generate a representation (50) of the surface of the workpiece (1), wherein the echo value G is reproduced in spatial resolution in the representation (50). [16] Test device (10) according to claim 15, characterized by that the ultrasonic test pulse reaches the surface of the workpiece (1) within the selected runtime interval I. [17] Test device (10) according to claim 15, characterized by that a guide device is provided which is designed to align the test head (40) relative to the axis of symmetry S of the workpiece (1) in such a way that the sound path W of the ultrasonic test pulse in the workpiece (1) and the axis of symmetry S of the rotationally symmetrical workpiece (1) span a common plane, the sound plane P. [18] Test device (10) according to claim 15, characterized by in that the test head (40) has an ultrasonic transducer (42) which is divided into a plurality of individually controllable transducer segments, and the control unit (20) is designed to transmit a sequence of ultrasonic test pulses at different incidence angles Theta into the workpiece (1) by means of the test head (40). [19] Test device (10) according to claim 15, characterized bythat a rotating device is further provided which is designed to generate a relative movement of the test head (40) and the workpiece (1) in such a way that the workpiece (1) is rotated about its axis of symmetry S under the test head (40). [20] Test device (10) according to claim 15, characterized by that the test head (40) has two ultrasonic transducers (42) and the direction of travel of the ultrasonic test pulses introduced into the workpiece (1) by means of the first ultrasonic transducer (42) is oriented, with respect to the axis of symmetry of the workpiece (1), opposite to the direction of travel of the ultrasonic test pulses introduced into the workpiece (1) by means of the second ultrasonic transducer (42).

Images