Precision for practice
Quality assurance is not possible without measurements. Networked systems as well as technological sticking points make measurement work no less easy. For example, one topic discussed at the "6th International Symposium on Production Measurement Technology" at the Interstate University of Applied Sciences Buchs (NTB) revolved around "Determining measurement uncertainty in practice".
Around 190 experts gathered at the NTB to discuss new trends in addition to the usual problems of production measurement technology. The first of the nine specialist presentations dealt with the complex traceability of measurement results. Rudolf Thalmann of the Swiss Federal Institute of Metrology (METAS) gave an overview before differentiating between individual influencing variables in measurements.
Unit transfer not easy
"When we talk about traceability of measurement results, we are basically talking about suitable procedures for calibration", says Rudolf Thalmann, the METAS Deputy (Physics and Chemistry Department) and Division Manager (Length, Optics and Time Division). Nevertheless, the traceability of a measurement value lies in the detail. For calibrations, trained technicians must pay attention to various coordinates before they are accredited by an official institute. Despite standardized geometries, permanent measuring procedures, and highly sensitive and precise sensor devices, there are always deviations in length measurements. So-called measurement uncertainties resonate with the factors "workpiece", "environment", "person", "procedure". Experts speak of traceability when several essential points, certainly the SI units, serve as a reference.
The measurement uncertainty (benefit, effort, methods)
Be it prototypes, be it workpieces in series: measurement uncertainties are present at many levels. For this reason, experts focus on individual process steps and note the underlying parameters so that an overall uncertainty can be defined. Depending on the application, recalibrations are carried out at appropriate intervals. However, "there is a considerable difference between inspecting workpieces in a sterile laboratory and in an industrial hall", emphasized Heimo Masser in his presentation on the work in multisensor coordinate measuring technology at Julius Blum GmbH, which specializes in hinge and furniture systems.
In the third presentation, "Determining measurement uncertainty in practice - benefits, effort and methods", Michael Hernla went into further detail about the difficulty of uncertainties. "The more measurement applications or comparable re-measurements can be made, the better", Michael Hernla, engineer and certifier, explained. He presented ISO-approved measurement uncertainty balances and newer strategies.
In his presentation, Hernla described test methods for more accurate coordinate measuring machines in the same way as Thalmann. However, he noted that not every measuring device is suitable for determining "a task-specific measurement uncertainty of any inspection characteristics". The measurement uncertainty could "deviate significantly from the number and position of the measuring points on the surface of the workpiece to be inspected, which is characterised by form deviations".
According to the expert, new methods for the correct determination of coordinate measurements have only been implemented in the recent past. In the so-called "virtual CMM", a model calculation is aimed at with calibrated workpieces and estimated individual factors. This method is the subject of the guideline VDI/VDE 2617; Sheet 11 Measurement Uncertainty Balances. In conclusion, the speaker said that correct documentation, for example in accordance with ISO/TS 23165, or the so-called "Monte Carlo Simulation", promote recognized measurement methods. Hernla pointed out that such measurement methods are very complex and costly. There were also some changes of opinion on this after the actual technical presentations, which indicated a large number of new regulating laboratory standards as well as increased production requirements.
It is becoming more difficult to distinguish irregularities from actual errors.
In this context, Markus Rauhut from the Fraunhofer Institute for Industrial Mathematics ITWM also met with a great response from the conference participants with his topic "Cost savings through process-integrated precision measurements".
Low effort and accuracy?
Markus Rauhut from the Fraunhofer Institute demonstrated what it means to maintain specific personnel expenditure and yet the highest accuracy by means of a surface inspection system for the quality control of engine components. Nevertheless, measurements vary. For example, when specialists have to trace back measurements on complex objects such as engine components. For example, they have to take into account factors such as gravity, temperature fluctuations or air turbulence. "The smallest cracks or impact points can affect the reliability and service life of the component ", however, inspectors are also prone to error in processes that are always repeating themselves. "Especially in safety-critical applications like in the EU Clean Sky project," says Rauhut, ITWM department head (image processing), "manual recalibrations could be saved. Likewise, a customer like the Clean Sky project could define stricter measurement tolerances.
Recently, image processing methods have been used to measure curved engine components. Various cameras take pictures of the object surface from different angles. These are then evaluated by software. Rauhut: "The entire spectrum of possible inspection tasks cannot be covered with standard solutions. The surface to be examined could change depending on the camera angle and illumination." The researchers at the ITWM are therefore working with "MASC - Modular Algorithms for Surface InspeCtion" to close technical gaps in measurement methods.
First, the workpiece surface is illuminated and scanned. "This is important in order to detect impact points or cracks that are only visible from one side of the turbine," explains Rauhut. In the case of free-form surfaces, regions covered by curvatures or corners are also detected in this way. The more complex the geometry, the more cameras are usually needed. The MASC basic version can thus combine more than 300 algorithms. For example, one algorithm is programmed to find edges or certain colour points in the image, the researcher explains. However, to reduce the measurement effort, they would focus on areas where primarily irregularities can be elicited and analyzed. "Thus, defective components could be weeded out right away," says the image processing researcher. The surfaces
Perception of aesthetic defects on reflective surfaces as a challenge.
However, analyse also helps the goods inspection with specific inspections with bullet holes or breaks already marked by the software.
Clearly traceable parameters
Ultimately, no component, not even a gemstone sphere, is truly homogeneous. Every product, according to the speakers, has deviations. They are not visible to the naked eye. "With projects that go from resolution into the microscopic range," said Sebastian Höfer, Fraunhofer- Institut für Optronik (Systemtechnik und Bildauswertung), "it becomes increasingly difficult to distinguish irregularities from actual defects." The consequence: perfect components are sorted out as defective. That is why measurement specialists from universities support important industrial companies in order to avoid pseudo-defects. Höfer, but also his predecessor Rainer Tutsch, Technical, University of Applied Sciences Braunschweig, specialized in microscopic lenses, therefore looks for specific test parameters for sensitive products before they have to be classified into quality classes. Here, objective evaluation measures are first introduced, "which are based on a model of human perception," says Sebastian Höfer, researching at the Fraunhofer Institute to define threshold values for defects that industrial technicians have not yet been able to register. Nevertheless, the perception of aesthetic defects on reflective surfaces is a challenge.
In the afternoon of the conference, speakers devoted themselves to deflectometry in the thermal infrared. The infrared spectrum enables the inspection of diffusely reflecting surfaces such as raw sheet metal. "In practice, however," said other researchers, "the number of measurements and the time required for measurements must be in reasonable proportion." The research project of Alex Schöch, who himself is investigating "Precise measurements on hot parts" at the NTB, therefore also received a lot of interest. Difficult forming processes, for example high-speed laser triangulation at maximum temperatures, could thus be optimized. Finally, the presentation by Ernst Ammon, Schaeffler Technologies GmbH & Co. KG, dealt with "Precise measurement results through unambiguous specifications".
Ammon has been working on clear design guidelines for decades. In his presentation, he dealt in depth with the withdrawal of DIN 7167 and the "relationship between dimensional, form and parallelism tolerances". Technical drawings in particular are now defined by ISO 14405-1. Ernst Ammon and his colleagues demonstrated the highest level of competence with regard to documentable technologies and standards.
Further information about the "6th Symposium on Production Metrology for Practice" can be found on the website of the NTB Interstate University of Applied Sciences Buchs (www.ntb.ch).