Search

ZusammenfassungDer Einsatz von Tiefen Neuronalen Netzen (Deep Learning) eröffnet neue Möglichkeiten in der digitalen Bildverarbeitung. Auch für die Auswertung von Bilddaten in der Ophthalmologie wird diese Methode erfolgreich eingesetzt und findet weite Verbreitung. In diesem Artikel wird die methodische Vorgehensweise beim Deep Learning betrachtet und der klassischen Vorgehensweise für die Entwicklung von Methoden für die digitale Bildverarbeitung gegenübergestellt. Dabei wird auf Unterschiede eingegangen und die wichtiger werdende Rolle von Trainingsdaten für die Modellbildung erklärt. Weiterhin wird die Vorgehensweise des Transfer-Lernens (Transfer Learning) für Deep Learning am Beispiel eines Datensatzes aus der kornealen Konfokalmikroskopie vorgestellt. Dabei wird auf die Vorteile der Methode und auf Besonderheiten beim Umgang mit medizinischen Mikroskopdaten eingegangen.

The use of deep neural networks ("deep learning") creates new possibilities in digital image processing. This approach has been widely applied and successfully used for the evaluation of image data in ophthalmology. In this article, the methodological approach of deep learning is examined and compared to the classical approach for digital image processing. The differences between the approaches are discussed and the increasingly important role of training data for model generation is explained. Furthermore, the approach of transfer learning for deep learning is presented with a representative data set from the field of corneal confocal microscopy. In this context, the advantages of the method and the specific problems when dealing with medical microscope data will be discussed.Der Einsatz von Tiefen Neuronalen Netzen (Deep Learning) eröffnet neue Möglichkeiten in der digitalen Bildverarbeitung. Auch für die Auswertung von Bilddaten in der Ophthalmologie wird diese Methode erfolgreich eingesetzt und findet weite Verbreitung. In diesem Artikel wird die methodische Vorgehensweise beim Deep Learning betrachtet und der klassischen Vorgehensweise für die Entwicklung von Methoden für die digitale Bildverarbeitung gegenübergestellt. Dabei wird auf Unterschiede eingegangen und die wichtiger werdende Rolle von Trainingsdaten für die Modellbildung erklärt. Weiterhin wird die Vorgehensweise des Transfer-Lernens (Transfer Learning) für Deep Learning am Beispiel eines Datensatzes aus der kornealen Konfokalmikroskopie vorgestellt. Dabei wird auf die Vorteile der Methode und auf Besonderheiten beim Umgang mit medizinischen Mikroskopdaten eingegangen.

Post frying conditions play a very important role in consumer acceptability of French fries. During post frying holding time, from when fried potatoes are removed from a fryer to when consumed, texture of the product dramatically changes. Therefore, it is important to understand how and why these changes happen. In this study, the effect of different batter formulations and post frying holding time (with and without using a heat lamp) on fat and moisture distribution, textural and thermal properties of fried potato strips was studied. Moisture content ranged from 3.90 to 5.32 (in core region), 0.54 to 1.19 (in crust region), and 1.96 to 2.89 g/g db (in whole potato strips). Fat content of the core, crust and whole potato strip regions ranged between 0.01–0.03, 0.14–0.25, and 0.07–0.15 g/g db, respectively. A slight migration of moisture from the core to crust regions was observed under heat lamp holding, but this did not affect the total moisture content. Potato samples that were not kept under heat lamp showed lower hardness than samples kept under heat lamp. Brittleness values of potato samples increased from 0 to 10 min heat lamp duration and after that it significantly decreased. It was found that glass transition temperature of the samples was also affected by batter formulation and heat lamp duration.

The effect of different batter formulations and post frying holding time (with and without using a heat lamp) on oil spots and moisture pocket redistributions on the surface crust and in the core regions of fried potato strips was studied. Potato strips were fried at 180 ± 2 °C in canola oil and then kept under a heat lamp for different time intervals. Fractal dimension was used to measure the surface roughness of the samples. The fractal values ranged between 2.63 and 2.69. Image processing technique was applied to study the oil spots on the surface and moisture pocket distributions in the core regions. The results indicated that the number of oil spots and moisture pockets were indirectly related to the area. Keeping the samples for longer time under a heat lamp reduced the number of big oil spots, whereas increased their mean area.

Today, micro components are used in various industrial sectors such as electronics engineering and medical applications. The final quality of such parts depends on each individual step of the production chain from the manufacturing of semi-finished parts to the post-processing. In this study, magnetron sputtering is used to manufacture thin (15-30 μm) aluminium-zirconium alloy foils for the deep drawing of high strength and hardenable micro cups, which can be, for example, employed as micro valve caps. The development of a novel process chain for the production of these parts includes four different steps, beginning with the production of Al-Zr foils by magnetron sputtering. Secondly, tensile tests are performed with the foils in order to estimate their mechanical properties. Subsequently, micro deep drawing is used to produce the cup’s shape, and finally, a heat treatment in a drop-down tube furnace adjusts the cup’s hardness during fall. It is shown in particular that Al-Zr foils produced by magnetron sputtering have an attractive cold forming and hardening potential due to a microstructure consisting essentially of an oversaturated solid solution of zirconium in the aluminium matrix. This material state enables adequate formability and simplifies the heat treatment process since no solution annealing is required.

Wires with 1 mm initial diameter have been reduced between 10 and 64 percent at different temperatures and strain rates by infeed rotary swaging, which is an incremental cold forming process mainly used for rods and pipes. The volume fraction of martensite in the deformed wires has been determined by X-Ray diffraction and by magnetic induction for different processing parameters. Measurements show that for already small percentage of reduction, martensite is present in the wires and its amount changes with the strain rate and temperature. While for smaller strain rates at room temperature the formation of martensite is promoted, it is restrained for higher strain rates and higher temperatures. Results also reveal that the martensite distribution in the sample is inhomogeneous. Further investigations have been made to analyze the microstructure by optical microscopy and to determine mechanical properties by tensile testing.

Pre-fatigue damage can have drastic effects on the elastic nonlinearity of solid materials, offering the possibility of extracting spatially resolved information on the pre-fatigue state from nonlinear effects on guided waves, like harmonic generation [1]. Of particular interest are surface and wedge acoustic waves. They are sensitive to surfaces and edges and are non-dispersive in homogeneous media, at planar surfaces and — in the case of wedge waves — perfect wedge tips on the scale of the wavelength. This favors nonlinear effects. Our current investigations focus on the nonlinearity generated by micro-cracks with the goal of extracting information about their spatial and orientational distribution from the efficiency of second harmonic generation and its dependence on amplitude and frequency of the input wave [2]. The linear dispersion arising from a spatially inhomogeneous distribution of the micro-cracks is accounted for in the propagation of the second harmonic.

Elastic constants of components are usually determined by tensile tests in combination with ultrasonic experiments. However, these properties may change due to e.g. mechanical treatments or service conditions during their lifetime. Knowledge of the actual material parameters is key to the determination of quantities like residual stresses present in the medium. In this work the acoustic nonlinearity parameter (ANP) for surface acoustic waves is examined through the derivation of an evolution equation for the amplitude of the second harmonic. Given a certain depth profile of the third-order elastic constants, the dependence of the ANP with respect to the input frequency is determined and on the basis of these results, an appropriate inversion method is developed. This method is intended for the extraction of the depth dependence of the third-order elastic constants of the material from second-harmonic generation and guided wave mixing experiments, assuming that the change in the linear Rayleigh wave velocity is small. The latter assumption is supported by a 3D-FEM model study of a medium with randomly distributed micro-cracks as well as theoretical works on this topic in the literature.

Aero-engine components exposed to high mechanical stresses are made of high-strength alloys and additionally, they are surface treated by shot peening. This process introduces compressive residual stress into the material making it less sensitive to stress corrosion cracking and fatigue and therefore benefits the components performance and lifetime. Moreover cold work is induced in an amount depending on the peening parameters. To approximate the remaining lifetime, a quantitative, non-destructive method for stress assessment is required. It was shown that surface treatment of such alloys can be characterized by broadband Rayleigh wave dispersion measurements. However, the relative contributions of residual stress and cold work, respectively, remained an open point. This paper presents the determination of third order elastic constants (TOEC) for IN718 and Ti6246, providing, together with a model for the inversion of dispersion data, a quantitative access to the acoustoelastic effect. Finally, some measurements of differently treated samples are given.

International audience; In view of qualitative temperature measurement by scanning thermal microscopy, we introduce a model-based control law for a new microfabricated probe. The underlying model is the time-space two-scale electro-thermal model presented in [15], since it has the power to represent transcients of harmonic modulations. The control method accounts for an estimation of the heat source in the sample and for the delay in the lock-in filter based observation. Experiment-based model calibration is a prerequisite and is discussed in detail.

BACKGROUND An analysis of the corneal subbasal nerve plexus (SNP) allows an evaluation of the peripheral neuropathy in cases of degenerative diseases. In order to study the SNP structures quantitatively the automatically calculated morphological and topological parameters are required. METHODS In vivo confocal laser scanning microscopy (Heidelberg Retina Tomograph II/Rostock Cornea Module) was performed in healthy volunteers as well as patients with severe diabetic neuropathy. An adapted image processing algorithm was used to preprocess, segment and evaluate quantitatively the nerve fibers of the SNP. Data sets were analysed statistically. RESULTS The developed algorithm allows an automated detection of SNP structures. Furthermore, it allows the collection of data based on morphological and topological parameters. The main parameters that show significant differences between healthy cornea and cases of diabetic neuropathy are nerve fibre density and length, number of branching, tortuosity and number of terminal and crossing points. All parameters of the measurements can be used isolated, combined or weighted for quantification of the SNP networks. CONCLUSION The presented fully automated preprocessing eliminates a large number of motion-induced artefacts. The quality of the resulting pictures allows an automated quantification using characteristic measurements. This represents an in vivo, non-invasive technology analysing degenerative changes of SNP especially in the course of diabetes mellitus.

Two different foils out of the alloy aluminium-scandium with a thickness of about 15 μm were produced by the d.c. magnetron-sputtering process applying different substrate temperatures, i.e. S37 at the substrate temperature of 37°C and S160 at the substrate temperature of 160°C. They show different forming properties, e.g. flow stress. In this work these two different foils were used as blank material in micro deep drawing with a punch diameter of 0.75 mm to investigate the formability of these foils. A limit drawing ratio of 1.6 was achieved for both foils. Using the strip drawing test the friction coefficients between the foils and the tools were acquired experimentally, i.e. μ = 0.12 on the smooth side and μ = 0.16 on the rough side for the foil of S37 and μ = 0.15 on the smooth side and μ = 0.17 on the rough side for the foil of S160.

We present a multi-scale model of a probe for scanning thermal microscopy. The probe is built by microfabrication techniques. In active mode, it is supplied by a source of harmonic and/or continuous current and the tip temperature is measured after a lock-in amplifier. The model distinguishes two time scales and two space scales. Simulation results show the potential of the model in terms of accuracy and computation speed and they are compared to experimental results. Finally, a temperature control law constructed from this model is stated.

Certain aspects of diffusive ultrasound fields in concrete are still unknown and thus, systematic parameter studies using numerical time-domain simulations of the ultrasonic propagation process could lead to further insights into theoretical and experimental questions. In the present paper, the elastodynamic finite integration technique (EFIT) is used to simulate a diffusive reverberation measurement at a concrete specimen taking aggregates, pores, and viscoelastic damping explicitly into account. The numerical results for dissipation and diffusivity are compared with theoretical models. Moreover, the influence of air-filled pores in the cement matrix is demonstrated.

In this article we describe a novel piezoresistive cantilever technology The described cantilever can be also applied in the investigations of the thermal surface properties in all Scanning Thermal Microscopy (SThM) techniques. Batch lithography/etch patterning process combined with focused ion beam (FIB) modification allows to manufacture thermally active, resistive tips with a nanometer radius of curvature. This design makes the proposed nanoprobes especially attractive for their application in the measurement of the thermal behavior of micro- and nanoelectronic devices. Developed microcantilever is equipped with piezoresistive deflection sensor. The proposed architecture of the cantilever probe enables easy its easy integration with micro- and nanomanipulators and scanning electron microscopes.In order to approach very precisely the microcantilever near to the location to be characterized, it is mounted on a compact nanomanipulator based on a novel mobile technology. This technology allows very stable positioning, with a nanometric resolution over several centimeters which is for example useful for large samples investigations. Moreover, thanks to the vacuum-compatibility, the experiments can be carried out inside scanning electron microscopes.

A series of mitomycin C (MMC) analogues, namely cyclopentanthraquinone derivatives, were synthesized via Diels-Alder cyclization of naphthoquinone with 1-vinylcyclopent-1-enes. These new compounds are planar structures, like MMC, and bear an aziridine ring and a methyl carbamate side chain. After bioreduction, they are anticipated to be capable of intercalating into double-stranded DNA and bind covalently. Structure-activity relationships were studied. Of these compounds, 2,3-aziridino-4-[[(methylamino)carbonyl]methyl] cyclopent[alpha]anthracene-6,11-dione (4) was shown to have inhibitory activity against several leukemic and solid tumor cell lines. Mice (BDF1) bearing Lewis lung adenocarcinoma were treated with 4 and MMC (i.p., QD x 5). At a dose of 30.0 mg/kg, compound 4 was as effective as MMC (0.8 mg/kg). Compound 4 appears to be less toxic than MMC. DNA unwinding assay indicated that 4 is able to intercalate into DNA double strands and is also a topoisomerase II inhibitor.

A new series of high power diode laser line sources is reported. The modules are designed for the industrial materials processing market and include both fiber coupled and direct beam configurations. Typical applications include welding, hardening and semiconductor processing. The biggest challenge in delivering line sources lies in the variety of application specific requirements. This problem is approached with modular concepts that allow for power scaling and custom beam shaping. All modules are available either as an OEM laser head or as a turn-key solution including power supply and chiller. Fiber coupled diode laser modules are available at power levels ranging from 600W to 4kW at various wavelengths. New developments include a 1kW module with a single wavelength and 400um / 0.22NA fiber and a 2kW module based on two wavelengths. Dilas offers up to 5kW from a 1mm / 0.22NA fiber with a single wavelength. At 200 μm fiber diameter, power levels of 850W are available with a single wavelength. While fiber coupled modules allow for easy power scaling, free space systems are capable of even higher overall electrooptical efficiencies and lower cost. Based on modular building blocks, Dilas provides customized solutions that are optimized for individual applications. Two modules will be described in detail. The first module is a 600W line source with line dimensions of 10.5mm x 350 μm at a working distance of 160mm. The second module operates at 3kW output power and creates a homogenized line with dimensions of 9mm x 1.3mm at a working distance of 200mm. Optical design trade-offs will be discussed and concepts for the modules described above are shown. Experimental results are presented.

PURPOSE. To overcome the anterior corneal mosaic (ACM) phenomenon in in vivo confocal laser scanning microscopy (CLSM) and to reconstruct undistorted images of the subbasal nerve plexus (SNP), facilitating morphometric analysis in the presence of ACM ridges. METHODS. CLSM was performed in five healthy volunteers. An original image processing algorithm based on phase correlation was used to analyze and reduce motion distortions in volume scan image sequences. Three-dimensional tracing of the SNP was performed to reconstruct images containing only the SNP layer, with nerve fibers clearly visible even in ACM areas. RESULTS. Real-time mapping of the SNP revealed the presence of ridges with K-structures underneath them in all cases. The occurrence of K-structures correlated directly with development of ACM observed by slit lamp and resulted in massive deformation at the level of Bowman’s membrane, seriously interfering with examination of SNP structures. The average elevation of ACM ridges was 20.6 m (range, 8.7‐34.0 m). The novel method presented permitted reconstruction of the SNP layer in regions of ACM. CONCLUSIONS. The described method allows the precise analysis and elimination of motion artifacts in CLSM volume scans, in conjunction with the capability to reconstruct SNP structures even in the presence of severe ACM. The robustness and automation of the described algorithms require ongoing development, but this will provide a sound basis for extended studies of corneal nerve regeneration or degeneration and for use in clinical practice. (Invest Ophthalmol Vis Sci. 2011;52: 5022‐5028) DOI:10.1167/iovs.10-6065

A series of mitomycin C (MMC) analogues, namely cyclopentanthraquinone derivatives, were synthesized via Diels−Alder cyclization of naphthoquinone with 1-vinylcyclopent-1-enes. These new compounds are planar structures, like MMC, and bear an aziridine ring and a methyl carbamate side chain. After bioreduction, they are anticipated to be capable of intercalating into double-stranded DNA and bind covalently. Structure−activity relationships were studied. Of these compounds, 2,3-aziridino-4-[[(methylamino)carbonyl]methyl]cyclopent[a]anthracene-6,11-dione (4) was shown to have inhibitory activity against several leukemic and solid tumor cell lines. Mice (BDF1) bearing Lewis lung adenocarcinoma were treated with 4 and MMC (ip, QD × 5). At a dose of 30.0 mg/kg, compound 4 was as effective as MMC (0.8 mg/kg). Compound 4 appears to be less toxic than MMC. DNA unwinding assay indicated that 4 is able to intercalate into DNA double strands and is also a topoisomerase II inhibitor.

The miniaturization of micro- and nanoelectronic components requires new methods for the inspection of buried inner structures at the nanoscale. We used the atomic force acoustic microscopy technique (AFAM) to image subsurface defects. This technique combines high lateral resolution with the capability to determine local elastic properties of materials near the surface. As the structures buried near the surface change the effective tip-sample contact stiffness it is possible to detect them. For the verification of the detection capabilities of AFAM we fabricated well-defined buried void structures with different geometries and dimensions. Large, thin, plate like structures of silicon nitride with a local filling were our first test samples. Then, sets of nine small, square, thin plates with thicknesses increasing stepwise from 30 to 270 nm were etched in a thinned silicon wafer. The last two samples contained wedge structures of widths varying between 1.6 and 10 μm. Our results showed that it was possible to detect buried void structures at depths between 180 and 900 nm. We also observed that the depths at which the buried defects can be detected by the use of the AFAM method depend on the defect dimensions and geometry, and on the mismatch in the elastic properties of the sample and the defects. The experimental results obtained for the groups of small, thin plates were verified by quantitative analysis via finite element method (FEM) simulations.

In recent years x-ray imaging techniques have dramatically expanded their range of applications. However, there are still specific problems where x-ray techniques fail due to fundamental physical reasons. These “no-go” applications can be divided into three main categories, (i) applications in which the x-ray absorption is too large, so that no significant transmission signal can be detected (e.g. pores in high density materials like tungsten carbide), (ii) applications in which the die gap of a planar delamination or a micro-crack is so much smaller than the transmitted thickness that no significant change in X-ray transmission can be observed (e.g. micro-cracks with nanometer die gaps), (iii) applications in which the absorption contrast between adjacent materials is too small to be detected (e.g. silicon chip/mold compound combinations). Most of the above mentioned applications can be successfully treated by using high-frequency scanning acoustic microscopy (SAM). These ultrasonic microscopes are typically working in the frequency range between 10 and 400 MHz and are using differences in acoustic impedance for imaging purposes.

We investigated the influence of thin oxide layers present on the surface of polycrystalline copper on the effective stiffness of the sample surface. The thicknesses of the oxide layers were in the range of single to few tens of nanometers. In order to detect the presence of such thin films on a sample surface, the so called atomic force acoustic microscopy (AFAM) technique was used. The investigated sample was polycrystalline oriented copper. The AFAM results obtained from heat treated sample were compared to these obtained for a reference sample with a thinner native oxide layer. The study revealed the ability of the AFAM technique to detect the presence of nanometer thin films. Our preliminary results suggest that AFAM technique may have a potential application for microelectronic industry as a tool for thin film characterization.

We studied nanoparticles by several high resolution microscopic methods as scanning electron microscopy (SEM), transmission electron microscopy (TEM) and scanning probe techniques especially atomic force microscopy (AFM) in contact and non-contact mode. While AFM in non-contact mode gives reliable information for 100 nm range nanoparticles it fails for smaller particles, showing lack of reproducibility. TEM and SEM prove to be reliable. By SEM imaging the agglomeration behavior and the structure of agglomerates are discussed in detail.

We have formed nanoparticles of Si, C, Ti, and Fe by pulsed laser ablation and have explored the feasibility of passivating these particles during the synthesis process in order to minimize particle coarsening. Passivation was done by forming the nanoparticles in an Ar/N 2 background, which allowed the formation of a thin nitride layer on the nanoparticle surface. In this paper we describe the synthesis procedure for synthesizing and passivating the nanoparticles. The characterization results indicate the nanoparticle synthesis in N 2 does not significantly affect agglomeration.

A numerical model of an acoustic microscope based on the elastodynamic finite integration technique (EFIT) is presented. It allows time-domain simulations of elastic wave propagation in both, fluids and solids, and includes focusing of the incident wave field as well as scattering at defects and the fluid-solid interface taking mode converted echoes and leaky Rayleigh waves into account. The simulations can be performed for different frequencies and materials and can be used for the continuous and time-resolved mode as well as for transmission and reflection microscopy. The simulation results can be represented by time-domain signals and wave front snapshots. The formation of V(r,z) curves is also possible. In the present paper the simulations are applied to the problem of vertical cracks and spherical inclusions in a solid substrate as well as for subsurface characterization of thin coatings.

Several work performed at the Fraunhofer Institute IZFP Dresden on photo and particle acoustic methods is presented. It includes both, modeling activities by an explicit numerical method (CEFIT) and experimental work. The given examples of applied excitations are photons (Laser acoustics) and electrons (Scanning Electron Acoustic Microscopy, SEAM). Both, time resolved measurements by pulse excitation as well as monofrequent measurements by periodic excitation together with signal recovery (lock-in technique) are discussed.

We report on a diode laser system, which is based on tapered diode laser bars and provides a combination of high power and high beam quality comparable to high power lamp pumped solid-state-rod lasers. Until now diode laser systems with output powers in the kW-range are based on broad area diode lasers. However, the output of these kilowatt laser systems usually is characterized by a strongly asymmetric beam profile, which is a consequence of the asymmetric beam parameter product (BPP) of broad area diode lasers with regard to the slow- and the fast-axis direction. Apparently the output of such a laser system can not be coupled efficiently into a fiber, which is required for a variety of applications. The symmetrization of the BPP of such a laser system requires complicated and expensive beam shaping systems. In contrast tapered diode laser bars allow the design of high power laser systems with a symmetric beam profile without the necessity of using sophisticated beam shaping systems. Power scaling is realized with different incoherent coupling principles, including spatial multiplexing, polarization multiplexing and wavelength multiplexing. The total output power of the tapered diode laser system was 3230 W at a current of 75 A. Fiber coupling yielded 2380 W at 75 A for a fiber with a core diameter of 800 μm (NA 0.22) and 1650 W at 60 A for a 600 μm fiber (NA 0.22), respectively. Focusing with an objective with a focal length of 62 mm led to a beam diameter of 0.52 mm in the focal plane. Taking into account the total power of 2380 W behind the fiber the resulting intensity in the focal plane was 1.1 MW/cm2.

NDE activities at the Laboratory for Acoustic Diagnosis and Quality Assurance (EADQ) Dresden are outlined. The applied methods comprise acoustic, thermal, optical and X-ray ones. Additionally, scanning probe methods (SPM) and scanning electron microscopy (SEM) are used. Combinations of different methods are especially effective. This is demonstrated for the coupling of an acoustic approach with SEM. For NDE on a micro- and nano-meter scale, preparation of appropriate test flaws and the verification of the NDE results turn out to be a challenge. To meet this challenge, we propose an approach based on focused ion beam technique.

In this work high frequency scanning acoustic microscopy was employed to nondestructively characterize subsurface defects in chromium containing DLC (Cr‐DLC) coatings. Subsurface defects as small as one micron were successfully detected in a flat Cr‐DLC coated steel coupon. Depth of the imaged subsurface defects was estimated using a simple geometrical acoustics model. The nature of the subsurface defects was investigated by using FIB/SEM technique. Curved Cr‐DLC coated components including a roller and gear tooth were also imaged, and the encountered challenges were addressed.

A large number of microgrippers has been developed in industry and academia. Although the importance of hybrid integration techniques and hence the demand for assembly tools grows continuously a large part of these developments has not yet been used in industrial production. The first grippers developed for microassembly were basically vacuum grippers and downscaled tweezers. Due to increasingly complex assembly tasks more and more functionality such as sensing or additional functions such as adhesive dispensing has been integrated into gripper systems over the last years. Most of these gripper systems are incompatible since there exists no standard interface to the assembly machine and no standard for the internal modules and interfaces. Thus these tools are not easily interchangeable between assembly machines and not easily adaptable to assembly tasks. In order to alleviate this situation a construction kit for modular microgrippers is being developed. It is composed of modules with well defined interfaces that can be combined to build task specific grippers. An abstract model of a microgripper is proposed as a tool to structure the development of the construction kit. The modular concept is illustrated with prototypes.

Hot embossing is an inexpensive method for the reproduction of microstructures. It has matured during the last few years from a pure laboratory method to an industrially applicable process. As an example, the technology for the production of movable metallic microstructures by aligned to embossing and reactive ion etching is presented, which has been developed at the Forschungszentrum Karlsruhe/IMT.

Elastic wave propagation in a random particulate composite is a very complex phenomenon. In such a material, the application of ultrasonic nondestructive testing methods is difficult due to the multiple scattering processes, the strong backscattering from the aggregates, the frequency dependent attenuation and dispersion of the coherent wave fields and the mode conversions between pressure, shear and Rayleigh waves. Therefore, the interpretation of the received signals is complicated and the signal to noise ratio is low. In order to improve the applicability of pulse-echo and impact-echo testing methods and to optimize inverse reconstruction techniques, it is necessary to study the process of wave propagation systematically.