Peter Lehmann

Surface-roughness measurement based on the intensity correlation function of scattered light under speckle-pattern illumination

The statistical properties of speckle patterns generated from a rough surface under a fully developed static speckle-pattern illumination are examined. The roughness dependence of the intensity autocorrelation function is studied and utilized to characterize typical engineering surfaces with anisotropic roughness. The speckle patterns under investigation are recorded by use of a CCD technique and are then analyzed by digital image processing algorithms to obtain a parameter that describes the surface roughness. It is shown that an in-process surface inspection can be achieved by this method.

Surface roughness measurement by means of polychromatic speckle elongation

A new approach for determining the roughness of engineering surfaces that is applicable to industrial in-process measurements is introduced. Laser speckle patterns, arising from light scattered from rough surfaces that are illuminated by polychromatic laser light, are detected in the far-field region. The incoherent superposition of these light intensities and the angular dispersion cause the effect of speckle elongation. This is characterized by increasing speckle widths and leads to a radial structure of the speckle patterns. With increasing surface roughness, the elongation is replaced more and more by the decorrelation of the monochromatic speckle patterns for the different wavelengths. Such effects were detected with the CCD technique and analyzed by local autocorrelation functions of intensity fluctuations that were calculated for different areas of the speckle patterns. The results of surface-roughness determination by means of the speckle elongation effect are presented.

Requirements for the Application of Speckle Correlation Techniques to On-Line Inspection of Surface Roughness

Abstract Conventional surface roughness measurements using profilometric instruments (either mechanical or optical) are standardized and extensively applied in industry. However, these techniques are not suitable to an on-line inspection of industrial machining processes. If an optically rough surface leading to a diffuse scattering is illuminated by coherent light, a speckle pattern occurs showing a granular spatial intensity distribution of the scattered light. In general, the averaged intensity distribution depends on the distribution of the local surface slopes. However, if two speckle patterns at different laser wavelengths are compared via correlation techniques, a measure of the root-mean-square (rms) roughness (Rq) can be obtained, based on certain statistical assumptions. This paper contributes to the application of speckle correlation techniques to the inspection of machined surfaces. Surface statistics dependent on the manufacturing processes are checked. Optical arrangements of a measuring device are introduced. Furthermore, results showing the limits of speckle correlation are presented.

Aspect ratio of elongated polychromatic far-field speckles of continuous and discrete spectral distribution with respect to surface roughness characterization

Here polychromatic speckle patterns generated either by a polychromatic light source that emits at discrete frequencies or by a light source showing a continuous narrow-band spectral distribution are studied. The purpose here is the application of polychromatic speckle-pattern analysis to an in-process surface roughness characterization. To compare the coherence properties of the different polychromatic light sources, first a modified definition of the coherence length is introduced. Furthermore, the relevant optical phenomena, namely, the speckle elongation caused by the angular dispersion and the roughness-dependent speckle decorrelation, are summarized. It is shown that light sources with a continuous spectral distribution have essential advantages in comparison with discrete wavelength sources. The theoretical results are confirmed by experimental investigations based on a digital algorithm for the evaluation of CCD images of polychromatic speckle patterns, which are recorded in the Fourier plane of a Fourier-transforming lens.

Detection based on rainbow refractometry of droplet sphericity in liquid–liquid systems

The shape of droplets in liquid-liquid systems influences their mass and momentum transfer processes. The deviation from sphericity of rising droplets in liquid-liquid systems was investigated for different droplet sizes. Rainbow refractometry permits one to test, in this case, whether the use of laser-optical particle sizing will be correct or faulty. Since the assumption of spherical particle geometry is a general basis of laser-optical particle-sizing techniques such as rainbow refractometry or phase Doppler anemometry, deviation from the spherical shape results in a measuring error. A sphericity check based on rainbow refractometry is introduced.

Parametric optical surface roughness measurement by means of polychromatic speckle autocorrelation

A method for determining surface roughness of engineering surfaces that is applicable to in-process measurements under harsh circumstances of industrial production plants (e.g. vibrations, humidity) is introduced. The rough surface is illuminated with polychromatic laser light. The angular distribution of scattered light intensities, i.e. a polychromatic speckle pattern, is the result of an incoherent superposition of monochromatic speckle intensities. The angular dispersion leads to increasing speckle widths with an increasing distance to the optical axis an effect called speckle elongation. This gives rise to a radial structure of the speckle pattern. However, with increasing surface roughness the radial structure vanishes because of a decreasing similarity of the monochromatic speckle patterns of the different wavelengths. The markedness of this effect is analyzed by digital image processing algorithms, e.g. the procedure of polychromatic speckle autocorrelation. The latest approach to an in-process roughness measurement device was made by the use of singlemode fiber-pigtailed laser diodes in order to supply a trichromatic, temporally partially coherent laser beam. A brief introduction to the theoretical background is followed by the presentation of the experimental setup. The image processing algorithms for calculating an optical roughness measure from digitalized speckle patterns are explained, and first results of surface roughness determination are presented.

Mass Transfer Processes in Liquid‐Liquid Systems with Surfactants

The aim of this study is the investigation of mass transfer processes in the liquid-liquid system toluene/water. The transfer of different organic solvents and the influence of surfactants on the transfer are investigated. The results are compared with data reported in literature and with ray-tracing computations. In most experimental work concerning mass transfer in liquid-liquid systems a gradient of concentration outside the droplet is neglected. For the experimental determination of such gradients digital holographic interferometry will be introduced.

Features of a combined FFT and Hilbert transform for phase Doppler signal processing

Phase Doppler anemometry (PDA) is a laser optical method to determine particle diameters and velocities within two-phase or multiphase flows simultaneously and with high spatial and temporal resolution. The measured phase difference between two Doppler bursts is related to the particle diameter. The proposed method for signal processing is based on a combined Hilbert transform and Fast Fourier transform (FFT) phase Doppler burst analysis. The numerical determination of the burst envelopes gives an estimation of the time delay between two related signals or, more generally, of the burst maximum position within the record time. This estimate is completed by the conventional FFT based signal analysis which is used to estimate frequency and phase difference. By this two-step estimation the restriction to the interval resulting from conventional signal processing can be avoided. The feasibility of the method in terms of an on-line determination of absolute phase differences is investigated by Monte Carlo simulations and demonstrated by means of a standard PDA arrangement with ball lenses rotating as spherical particles through the measurement volume. Phase differences up to were determined reliably. In addition, algorithms to calculate Doppler frequencies and Doppler intensities, burst lengths and further characteristic parameters taking the signal quality into account are introduced and studied by Monte Carlo simulations. Based on these results validation strategies to reduce the influence of maltriggered bursts, trajectory or Gaussian beam effects on measurement results can be developed. Furthermore, by using the information of burst length and maximum position, optimized signal processing algorithms can be realized in order to achieve maximum accuracy in frequency, phase difference and signal-to-noise ratio (SNR) estimation.

Time-resolved laser Doppler and phase Doppler signal processing

Signal processing in Laser Doppler Anemometry (LDA) and Phase Anemometry (PDA) is carried out in both, the time domain, for instance by counter processors, and the frequency domain by processors based on the Fast Fourier Transform (FFT). FFT based signal processing allows a frequency and phase determination even for rather low signal to noise ratios. Nevertheless, the time-varying amplitude, frequency, and phase difference of Doppler bursts contain relevant information which is generally not utilized by conventional signal analysis. By determining the bust envelopes of PDA bursts it is possible to improve the resolution of particle sizing which is restricted to a corresponding phase difference from 0 to 360 degrees when conventional FFT processing is used. Time-resolved phase difference and frequency determination allow an improved signal validation since effects like trajectory effects, particle coincidences within the measuring volume, droplet oscillations or non-spherical particle characteristics can be detected. In this contribution signal processing methods for determining such effects are introduced. Thee methods will offer improved signal processing facilities with respect to the above mentioned topics. Practical applications are demonstrated by simulation and experimental results.

Measuring roughness with dichromatic speckle correlation

The most established way to inspect machined surfaces is measuring with profilometers. Roughness measurements by these methods take plenty of time and an on-line integration in production facilities is not feasible due to different reasons. Over two decades ago, Goodman and Parry discovered the correlation between laser speckles of different light wavelengths and surface roughness. Since then several authors have been working on this topic. Nevertheless, industrial sensors based on the speckle correlation phenomenon measuring a roughness value are not available yet. In this article a new method is described, which is suitable for measuring the roughness of surfaces online under production conditions. The surface which is to be inspected is illuminated by a dichromatic laser beam. The scattered light is converted by lenses, and the far field speckles are detected in the Fourier plane of the lenses by a separate CCD-array for each wavelength. The CCD-data are captured by a frame grabber and stored for evaluation. A digital computer processes the data by calculating 2D-cross correlation functions of the two related speckle patterns for the different light wavelengths. The maximum values of the cross correlation function represents the correlation coefficient, which can then be translated into the root mean square of the surface heights of the workpiece. The main benefit of the method described is the feasibility of measuring roughness during processing. This is achieved by a time synchronous detection of speckle patterns of two different wavelengths, which are used to determine the roughness of a surface. The basic measuring time is determined by the shutter speed of the cameras, which is in the pilot project 1/10000 s. Therefore, only vibration frequencies above 1 kHz disturb the measuring results. A further reduction of the measuring time will be no major problem. Thus any realtime roughness measuring problem can be solved this way. This article describes the state of the art and the state of the experimental investigations measuring roughness by means of this new measuring process.