EP1066510A1

EP1066510A1 - Automatic error detection method used during crack inspection according to the dye penetration method

Info

Publication number: EP1066510A1
Application number: EP00906158A
Authority: EP
Inventors: Klaus Abend
Original assignee: Tiede GmbH and Co Risprufanlagen
Current assignee: Tiede GmbH and Co Risprufanlagen
Priority date: 1999-01-22
Filing date: 2000-01-21
Publication date: 2001-01-10
Also published as: WO2000043758A1; KR20010092248A; DE19902525C2; CN1293758A; CN1187603C; DE19902525A1

Abstract

The invention relates to an automatic fault detection method used during crack inspection according to the dye penetration method. According to the inventive method, workpieces are treated with penetrating agents containing dyes during which the dyes concentrate at locations with surface faults and, after a predetermined developing duration, are recorded by at least one image recording unit. In addition, the recorded images are assessed for faults in an image processing unit by scanning and identifying areas which have a concentration of dye, and corresponding signals are output. The method comprises the following steps: Recording images of the same workpiece (10) at at least two times t1, t2 after the workpiece has been treated with penetrating agents and optionally developing, upon receipt, at least two recorded images A1, A2; comparing the recorded images A1, A2 and using the evaluation logic of the signal processing unit (22) to evaluate the comparison; outputting signals via the evaluation logic which, in corresponding areas on the recorded images, depict the changes in the concentration of penetration agent over the time period DELTA t1, t2, said changes being greater than a change threshold value for a reference time difference; and evaluating the measured workpiece-related parameters in order to produce assessment values of the cracking, such as a good/bad statement, or to conduct a fault value assessment according to a predetermined value interval or in the predetermined superficial area.

Description

Procedure for automatic error detection in crack testing after

Dye penetration process

The invention relates to a method for automatic error detection in crack testing by the dye penetration method, wherein workpieces for dye penetration testing are treated with dye-containing penetrant with an accumulation of the dyes on surface defects and after a predetermined development period are recorded by means of at least one image recording device and the images are scanned and recognized in an image processing unit from areas with dye enrichment errors are evaluated and corresponding signals are output.

Automated optical error detection in color penetration testing in production plants that continuously produce workpieces to be checked, such as, for example, continuous casting plants, wire end tests or the like is known. currently images of workpieces with dyes are already optically evaluated by so-called optical image recognition, the errors made visible by the color penetration method known per se being recognized by an optical scanning and image recognition method and compared with a stored error logic. The dye penetration test has attracted renewed interest, since recently non-ferritic light metals such as aluminum or magnesium alloys or even titanium alloys have been used for aluminum supports, light metal engine blocks etc., and ceramics are also being used to an increasing extent, such as for valve components and coatings of highly stressed parts. These are routine examinations - in-process control - for cracks in non-magnetizable workpieces, such as those made of magnesium or aluminum alloys or ceramics.

These are crack testing methods, in which workpieces for dye penetrant testing are treated in a manner known per se with test agents containing dyes with the dyes being enriched for surface defects and under illumination by an illumination device, such as UV lamps for fluorescent dyes, but also lasers or other lamps with suitably absorbent dyes. The workpieces are usually prepared for the color inspection by cleaning, possibly pickling and drying them, sprayed with a test agent containing dyes, in particular also fluorescent dyes, with enrichment of the dyes on surface defects, in particular cracks, and then of the excess dye-containing agent Test equipment, for example freed by wiping or wiping, the workpiece treated in this way, if necessary, treated with a developer and then recorded and evaluated under UV or visible light after a predetermined development time. Since the time in which the cracks are clearly visible after development is very short - often in the range of less than one minute, measurements should be carried out exactly within a certain reproducible period after development.

So far, these color penetration tests have mostly been carried out by operating personnel and evaluated by visual inspection. The anmeider has already proposed such methods for dye penetration testing, along with the associated device. Since a frequent possibility of error is the fatigue of the persons who take over this test, automated detection systems via image processing have already been proposed, for example in DE 19639020.6 or DE 19645377.1, to which reference is made in full to avoid repetitions.

From DE 39 07 732 a method for monitoring a device for evaluating surface cracks by means of the dye penetration method has already become known, in which the lamp intensity and the quality of the test material are monitored and, if the results are unsatisfactory, the system is switched off and the quality of the test object is checked by means of cameras. The monitoring signals are only used there to switch off the system - there is no provision for readjustment of the test equipment content or the lamp intensity, let alone documentation of the data on the system behavior. This known system is therefore only able to switch off the system.

From 19645377.1 it is already known to check and document the safety and verifiability of the system. According to 19645377.1 it is proposed to automatically check the change in the setting of image recording devices, such as the focus or else the geometric arrangement of the recording device for the test object, which are easily changed; likewise other parameters which strongly influence the test, such as the quality of the cleaning agent, the test liquid, the pickling agent and the temperature.

In the known method, both the test method and its limits, test failure and its handling, performance limitation, tolerance specifications, etc., which are desired today, are monitored and the recording or documentation of the results and the reproducibility of the results - including the verification of the results Functional maintenance of the inspection system itself guaranteed. This provides additional security if, in particular in automatic test systems over a longer operating period, there are points of view in order to reduce costs or increase the security that workpieces which are to be classified as incorrect could be assessed more reliably.

A regular run through so-called test specimens with predetermined test errors can determine whether these were still correctly recognized - however, this method could only determine that the test specimen was not recognized, but not why not. Since no documentation was created, it was also not understandable when the system no longer worked satisfactorily and why.

The measurements are carried out with penetrating agent that creeps into depressions and other surface changes such as defects, cavities, pores, depressions due to surface tension phenomena. Creeps - changes in the penetrant over time, such as occur over time due to changes in concentration due to evaporation of the solvent of the penetrant, mixing with components of the workpiece (residual fat content, contamination, etc.), so far a measurement has been carried out within a relatively short period of time the treatment of the workpiece with Penethermittel / developer - then the error indications change - this is called "blooming" of the error. This means that the sharp accumulation of the penetrant dye on / in the error subsides and the dye migrates out of the error again and so the contrast is getting worse. The error evaluation according to the color powder method is therefore subject to dynamic changes that strongly influence the measurement result. These changes in the error display over time could not be taken into account, even due to the high self-checking properties of the known systems. Due to the dynamics of the error display, errors often occurred because the time between the application of penetrant / developer and the taking of the test object was not exactly observed by an image processing unit. According to the prior art, incorrect evaluations frequently occurred due to the dynamic behavior of the error displays, since some errors were overrated and others were not recognized due to rapid "blooming".

Accordingly, the safety and effectiveness of crack detection systems using the dye penetration method had already been improved - but the problem still remained that depending on the workpiece, the penetration liquid accumulates very differently. Both the surface quality of the workpieces and the surface tension of the penetration liquid that is formed on them differ from material to material - the error display blooms depending on the type of error - i.e. depending on the depth of the error, porosity of the material or smoothness of the surface at different speeds.

It is therefore an object of the invention to provide a method for improved detection of errors with the penetrant test.

According to the invention, the object is achieved by a generic method with the steps:

Taking recordings (A1, A2) of the same workpiece at least two times (t1, t2) after treatment with penetrant,

Comparing the recordings (A1, A2) produced at these different times (t1, t2) and evaluating the comparison using the evaluation logic of the optical image processing unit, and - Output of signals by the evaluation logic, which represent the changes in the penetration agent enrichment over the period (Δ t1, t2) in the corresponding areas on the recordings (A1.A2), which lie above a change threshold value for a reference time difference; and

- Evaluation of the signals output taking into account workpiece-related parameters and test facility-related operating variables for the production of evaluation variables for crack formation, such as good / bad information, error size assessment according to a specified size interval or in a specified surface area.

Advantageous further developments result from the dependent claims.

Because the dynamic behavior of the dyes enriched on surface discontinuities is now reliably recorded and evaluated, a completely new evaluation of the surface defects is possible. Recording the temporal behavior of the error display by means of image processing methods by recording at intervals and calculating the differences in the contrasts enables data processing to automatically classify and evaluate errors by automatically evaluating the differences in the images recorded at different times, and accordingly to display errors with certain types to spend.

Compared to the most frequently performed evaluation by humans, the method according to the invention has the advantage that the human errors which unavoidably occur when looking at images that are always similar over a longer period of time can be avoided since cameras cannot show any signs of fatigue.

Surprisingly, it is therefore now possible according to the invention to classify errors over time using the dynamic behavior of the error display. Because initial values are stored in a system of this type and stored, the systems can be set up for a wide variety of test objects and test liquids. By evaluating the error displays via recording devices, it is now also possible to create a documented test report of the monitored company size. It is advantageous if the optical image processing by setting windows and scanning the window by the image recording unit, the selection and evaluation and the crack error display are automatically linked to the test sequence (time cycle) and the data obtained therefrom is processed in a computer.

It can be provided that a recording device produces recordings at a predetermined time interval. By providing a single recording unit that produces at least two recordings at a fixed time interval, the size of the crack testing system and its costs can be kept low and also problems that are avoided by using multiple recording units that do not work completely the same.

However, it is also possible to place the workpiece to be checked in the same spatial manner on at least two holding devices K1, K2 Kn arranged at a distance from one another

Demonstrate orientation by a conveyor, so that from the various receiving devices K1, K2 ... Kn. Recordings A1, A2 ... on the workpiece with the same spatial orientation, but at different times after treatment with penetrant, the recordings A1, A2 ... at the different recording devices can be compared with each other using evaluation logic and the differences in the recordings to form signals as a function of the time intervals that have elapsed between the recordings, which signals are then significant for the type of errors or their dynamic behavior. The use of a conveyor and several pick-up devices has the advantage of being able to check many parts very quickly.

It makes sense to store target data Δ A1, A2 as well as data on the time difference Δtn, tn + 1 between the respective periods of time that elapsed between the recordings in the evaluation logic and have the evaluation logic compare whether the measured difference values lie within the specified target values. This makes it possible to select only errors that are displayed within a certain time interval. If no fixed time interval between two measurements is set, this can be replaced by measuring the time difference tn, tn + 1 between two recordings An, An + 1 of the recording device and assigning this time period Δ tn, tn-1 to the determined change in contrast during this time . This becomes necessary, for example, if parts cannot be guided past the receiving devices at predetermined times.

In any case, it is advisable to monitor system components in predetermined periods by monitoring units and to have monitoring signals emitted which are checked by the measured value processing unit and signals are correspondingly emitted. The geometrical arrangement, focus and also function of the at least one recording device; and / or the functionality of the liquids used in the method: the test liquid and / or the developer liquid and / or the mordant liquid and / or the cleaning agent and / or bath data, such as the bath temperature / s; Fill levels; Pollution can be checked by monitoring devices. These monitoring signals can be used to control the system, for example readjustment units.

In order to demonstrate the functionality of the systems and the accuracy of the quality control by the method according to the invention, it is usually necessary and essential that the monitoring signals and / or the signals of the measured value processing unit be recorded on a medium. For this purpose, it is usually also necessary to measure and, if necessary, record workpiece-related parameters such as part identification numbers, number of pieces.

The lighting intensity and / or the sensor sensitivity of the lighting monitoring sensors and / or the test agent concentration and amount and / or the cleaning agent concentration and amount and / or the cleaning agent and / or the mordant concentration and amount and / or settings of the image recording unit / s can be determined via the monitoring signals how the geometric arrangement of the focus or the sensitivity can be adjusted. As an "integral test" of the system, test pieces with reference errors can be automatically passed through and the functionality of the entire system can be checked by measuring them.

Of course, in the method according to the invention, the already n; Checking the irradiation devices and the radiation measuring devices for functionality can now monitor the function of the test system and keep it at the same level. The fact that the function of the inspection system and its individual components can now also take place automatically at predetermined intervals ensures the following advantages:

It is now possible for the first time to automatically differentiate between faults on a test object and thus significantly improve the accuracy and informative value of the method, whereby the behavior of the system can also be checked over the entire operating time, including documentation thereof.

Preferred embodiments of the invention are explained in more detail below with reference to the schematic drawing, but the invention is in no way limited to this embodiment, but any further embodiments are familiar to the person skilled in the art. In this shows:

Fig. 1 is a block diagram of a dye penetration test method

Figure 2 is a schematic representation of a crack testing system according to a first embodiment of the invention. and

Fig. 3 is a schematic representation of a crack testing system according to a further embodiment of the invention with several receiving devices.

As can be seen from FIG. 1, a - usually non-ferritic - test part is pre-cleaned, if necessary pickled and dried, and then treated with the test agent - also referred to as a dye penetrant - in the crack test method using the dye penetration method. The excess dye penetrant is removed after a certain period of time, the workpiece is cleaned and then treated with a developer solution. After the development time, the workpiece may be dried and inspected at different times, and then, based on the different recordings at different times, a statement is made about the defectiveness of the workpiece, which may also be documented.

As can be seen from FIG. 1, a developed workpiece 10 is guided as a test specimen into a test station, in which the application of the color penetrant from a color penetrant tank 12 is represented schematically by spray nozzles - the test specimen actually passes through several stations, in which it is provided with cleaning and pickling solutions as well as developer solutions and color solutions, which are not shown here. In the line to the spray heads, a test means checking and replenishing system 17 is provided in line, preferably one according to DE-A-4438510.2.

There the test equipment is checked for functionality and if necessary dye or the like can be replenished into the tank 12 if this is necessary. In this embodiment, which works with fluorescent dye, the test specimen is irradiated by means of a UV lamp 11, which in turn is monitored in a manner known per se and the current of which can be readjusted accordingly.

From a storage container 12 (in simpler embodiments by means of a spray), which is connected to a circulating pump, test liquid 13a, which serves to mark the surface defects, is fed via a supply line by means of spray heads 13 to a spraying system and atomized over the surface of the workpiece 10. The test liquid is now distributed on the workpiece, whereby the dye particles - as is generally known as a physical phenomenon - concentrate on cracks due to the surface tension. An increased particle concentration is then found at these points. The unnecessary test liquid is removed, for example by wiping. The test specimen is then processed with a developer liquid. After a development time, which is to be determined experimentally for each test arrangement and test specimens, has elapsed, the surface of the workpiece 10 is then irradiated by a lamp 11, thereby causing the test liquid particles to fluoresce or to absorb and which are located in the area of the surface cracks Enriching dye particles are recorded by a camera 16 and this recording is stored in the image processing system 22. After a time interval of approximately 20-150 seconds, a second picture is taken, which is also stored in the image processing system 22. These two recordings are then compared with one another by an evaluation logic of the image processing unit and the time interval is assigned to the comparison value. Possibly. other recordings can also be recorded and processed at other times. The calculated comparison values are then compared in the evaluation logic with a stored target value table and it is determined in this way whether the image change values are within a predetermined range or above a predetermined threshold value. Correspondingly, an error display can then be output by the evaluation logic, which can lead to the classification or also to the rejection of the measured part. The functional reliability of the system preferably includes a self-checking device for checking or self-checking the associated working parameters, ie compliance with the respective operating parameters within the prescribed value interval. Such a self-inspection system can readjust within certain limits if the inspection values are outside a desired measurement range - this can result in unnecessary waste of material, such as that caused by premature replacement of the marking agent or by routine routine replacement of the lighting, such as a UV lamp or the like. occurs. This considerably increases the service life of the test system, it can run uninterrupted longer and the associated operating costs as well as those for material and energy are also reduced as a result. The self-checking device 14 is preferably connected to a documentation device 30, in which it creates test reports by means of which the functionality of the system can be verified.

A further embodiment of a plant for carrying out a method according to the invention is shown schematically in FIG. 3. Measuring unit groups 16, 16 ', 16 "can output their images, which are fed to the respective input of an image processing unit 22. In this case, at least two images of each workpiece 10 are taken at different times and the differences between the two images are determined, for example by forming a difference, These differences can NEN, for example, can be fed into a display device 20 but also into a downstream sorting device, which automatically separates out parts classified as bad.

Instead of using a camera, the data flow based on brightness values can advantageously also be recorded by a diode cell or other suitable means, as are known to the person skilled in the art. Of course, the documentation on remote data transmission can also be created and stored remotely from the device.

Because the kinetic behavior of test equipment on the surfaces of workpieces is now being evaluated for the first time, it is now surprisingly possible to classify errors and thus to make more precise distinctions between rejects and usable parts and, if necessary, quality structuring of the parts, for example in A- and B- Quality - to be provided.

Although the invention has been explained on the basis of a preferred exemplary embodiment, modifications which are within the scope of protection of the claims are familiar to the person skilled in the art. The invention is therefore in no way limited to the described embodiment.

Claims

claims

1. A method for automatic detection of defects in the crack inspection according to the dye penetration method, wherein workpieces for the dye penetration inspection are treated with dye-containing penetrant while the dyes are enriched with surface defects, and after a predetermined development period are recorded by means of at least one image recording device and the images are scanned and recognized in an image processing unit from areas with dye enrichment errors are assessed and corresponding signals are output, characterized by

Comparing the recordings (A1, A2) produced at these different times (t1, t2) and evaluating the comparison using the evaluation logic of the optical image processing unit, and

- Output of signals by the evaluation logic, which represent the changes in the penetration agent enrichment over the period (Δ t1, t2) in the corresponding areas on the recordings (A1.A2), which lie above a change threshold value for a reference time difference; and

- Evaluation of the output signals taking into account workpiece-related parameters and test facility-related operating variables for the production of evaluation variables for crack formation, such as good / bad information, error size assessment according to a specified size interval or in a specified surface area.

2. The method according to claim 1, characterized in that the optical image processing is implemented by setting windows and scanning the windows by the image recording device, the selection and evaluation and the crack error display is automatically linked to the timing of the test sequence and the processing of the resultant Data is done in a computer.

3. The method according to claim 1 or 2, characterized in that the image recording device produces recordings at a predetermined time interval.

4. The method according to claim 1 or 2, characterized in that

- The workpiece is guided past at least two spaced-apart image recording devices (K1, «2 Kn) in the same spatial orientation by a conveyor device, so that recordings (A1, A2 ..) of the different image recording devices (K1, K2 ... Kn) On) the workpiece in the same spatial orientation, but at different times after treatment with penetrant, and

- The recordings (A1, A2 ... An) of the different image recording devices (K1, K2 .... Kn) compared with each other by an evaluation logic and from the differences of the recordings (A1, A2 .... An) depending on the signals are formed between the elapsed time intervals.

5. The method according to any one of the preceding claims, characterized in that in the memory of the evaluation logic target data for the image changes (Δ A1, A2) and data on the time difference (Δ tn, tn + 1) between the respective between the recordings (A1, A2 ... are stored on elapsed periods, and the evaluation logic compares whether the measured difference values lie within the predetermined threshold values and accordingly signals are emitted which only represent the errors within a predetermined time interval.

6. The method according to claim 2, characterized in that at an unspecified timing of the test sequence, the time interval by measuring the time difference (Δtn, tn + 1) between two recordings (An, An + 1) of the image recording device and assignment of this period (Δ tn , tn + 1) to the determined change in contrast is defined in this time.

7. The method according to any one of the preceding claims, characterized in that system components and parameters are monitored by monitoring units in predetermined periods and monitoring signals are emitted, which are checked by the measured value processing unit and corresponding signals are emitted,

8. The method according to claim 7, characterized in that the system components and parameters to be monitored, the geometric arrangement, the focus and also the function of the at least one image recording device; the function of the lighting device and / or the functionality of the liquids used in the process.

9. The method according to claim 8, characterized in that test liquid, developer liquid and mordant liquid are implemented in the process.

10. The method according to claim 8, characterized in that the respective bath temperature, the fill level and the contamination are monitored as bath data.

11. The method according to claim 7, characterized in that the monitoring signals for controlling the system and / or its readjustment units are used.

12. The method according to any one of claims 1 to 11, characterized in that the monitoring signals / and or the signals of the measured value processing unit are recorded on a medium.

13. The method according to any one of the preceding claims, characterized in that workpiece-related parameters are measured and, if necessary, recorded.

14. The method according to any one of claims 7 to 12, characterized in that the monitoring signals, the lighting intensity and / or the sensor sensitivity of the lighting monitoring sensors and / or the test agent concentration and amount and / or the detergent concentration and amount and / or the detergent and / or the mordant concentration and amount and / or settings of the image recording device (s), such as the geometric arrangement of the focus or the sensitivity is readjusted.

15. The method according to any one of the preceding claims, characterized in that test pieces with reference errors are automatically passed through and the functionality of the entire system is checked by their measurement