Volume 92, Issue 5 - Special Issue: XXXVIII. Messtechnisches Symposium des AHMT 2024

The article reviews existing traceable small force-displacement measurement methods and presents a modified version of the force-displacement measurement device used to calibrate the stiffnesses of MEMS and cantilevers, as well as the forces they generate in the range from 10 nN to 2 mN. Separately, improved results for force constants and soft cantilever stiffness calibration (force up to 300 nN) with an expanded relative uncertainty of 0.46 % are shown. In conclusion, the prospects for the development of the force-displacement measurement system are discussed.

For reliable structures in power electronics, it is crucial to guarantee the proper quality of their interconnection layers. At the moment, most of these joints are made by a solder process. Like every joining process, irregularities such as cracks or voids can impair their thermal and structural resilience. In the context of power electronics, one of the most important factors is the heat path from semiconductor devices into the substrates they are soldered onto. A disturbed heat path can, in the worst case, lead to failure of an entire (safety-relevant) system. Recent investigations have shown that the Laser Speckle Photometry (LSP) is a promising method to detect such impurities. Therefore, the specimen is illuminated by coherent monochromatic light. The interfering effects can then be recorded and analysed, since their behaviour can be correlated with the sample’s state. Additionally, due to its easy setup and its non-contact, non-destructive manner, the LSP could be well suited even for inline applications. The previous publication presented a simulation that showed the underlying physical processes and the transmission of these into a real setup, testing on synthetically provoked voids. This paper will concentrate on improving the system to analyse real samples, optimize measurement times and provide an outlook for further development steps.

A sensor network has been developed consisting of wireless sensor nodes, that are interconnected by mechanical distance sensors. This approach makes it possible to precisely measure strain and shape of an object. The total number of probes and their positions will vary depending on the shape of the individual object that shall be measured. We developed a design procedure of optimally distributing the network on the surface with respect to the minimum number of sensors to be used taking into account the mechanical boundary conditions. Furthermore, two possible evaluations are shown to determine the measurement uncertainty of the sensor network. This can be used to select the network with the minimum uncertainty. Both methods developed will be illustrated using an artificial surface. Although the example was chosen specifically to achieve the greatest possible clarity, the results can be regarded as generally applicable.

Coherence scanning interferometry is an optical, non-contact measurement technique employed for the high-resolution three-dimensional characterization of surface topography. In order to enhance the lateral resolution of an interference microscope, a Linnik interferometer with immersion objectives of high numerical aperture was set up and used to measure technical surfaces. This study demonstrates the impact of an immersion medium to improve the lateral resolution and compares immersion and non-immersion systems in terms of signal formation and transfer characteristics in the three-dimensional spatial frequency domain. The measurement results of transparent mr-NIL210 gratings show a significant improvement in signal quality using immersion systems compared to the non-immersion system. Furthermore, measurement results of hemispheres sub-structured by substrate conformal imprint lithography are presented. Finally, measurements are presented on various line pairs whose period lengths are close to and below the Abbe limit and it is explained how these can still be resolved.

High-speed tensile tests are widely used to generate materials data for crash simulation purposes. The measurement of strain is of particular interest in these experiments. Laser-Doppler technology enables accurate strain measurements without contact, but the laser speckle effect causes an increased noise level and so-called signal dropouts. Signal diversity reduces the negative effects of these laser speckles. We implemented polarization diversity, where two photodetectors detect the light of two orthogonal polarization states, in a laser-Doppler extensometer. In this paper, we build a mathematical model to describe the carrier-noise-ratio under the laser speckle effect. Our findings show that polarization diversity leads to greater improvement with two diversity channels instead of the other two diversities (angular and wavelength diversity). Furthermore, we introduce that our sensor can measure the strain in the high-speed tensile testing, where velocities of over 30 m/s occur. Finally, we estimate the measurement uncertainty of our laser-Doppler extensometer. The measurement data obtained shows that our sensor can accurately measure such fast tearing processes.

In this paper, we evaluate the effect of lossy compression on luminance-based area measurements in images. On a set of 1,800 images, we vary three parameters: compression quality, edge width, and edge gradient, of the luminance edge bounding the area. As a result, when using the JPEG codec, the root-mean-square error after compression approximately doubles when decreasing the edge width from 16 to 4 pixel while keeping the gradient fixed. Conversely, for a fix width, the root-mean-square error doubles when halving the gradient. We also compared JPEG with the H.264 codec using default settings. H.264, based on inter-frame compression, requires four to six times the bitrate to achieve the results of JPEG. We conclude that for area measurements like these, among lossy codecs, intra-frame compression like JPEG is the best choice. Whether the influence of the compression is relevant depends on individual requirements, and more on the edge gradient than the edge width. Generally, we found that for steep gradients and reasonable JPEG compressions of Q = 85 or higher, our results were acceptable.