Table of contents

This paper is devoted to the Karman vortex shedding flowmeter—its physical fundamentals, research, design, optimization and applications. The flowmeter is currently in a stage of rapid development. Many valuable discoveries concerning the applied Karman vortex street phenomenon have been made. In this paper, various aspects of the problems concerning the vortex meter are described. Fundamental problems and their solutions are presented based on a description of the principles of operation. The most important methods of investigating the phenomena that appear in the vortex meter are described: measured signal analysis, flow field investigations using a hot-wire anemometer, flow visualization supported by image processing and numerical modelling. The most powerful achievements in these fields are also described.

The standard approach in digital particle image velocimetry (DPIV) data processing is to use fast Fourier transforms to obtain the cross-correlation of two single exposure subregions, where the location of the cross-correlation peak is representative of the most probable particle displacement across the subregion. This standard DPIV processing technique is analogous to matched spatial filtering, a technique commonly used in optical correlators to perform the cross-correlation operation. Phase only filtering is a well-known variation of matched spatial filtering, which when used to process DPIV image data yields correlation peaks which are narrower and up to an order of magnitude larger than those obtained using traditional DPIV processing. In addition to possessing desirable correlation plane features, phase only filters also provide superior performance in the presence of dc noise in the correlation subregion. When DPIV image subregions contaminated with surface flare light or high background noise levels are processed using phase only filters, the correlation peak pertaining only to the particle displacement is readily detected above any signal stemming from the dc objects. Tedious image masking or background image subtraction is not required. Both theoretical and experimental analyses of the signal-to-noise ratio performance of the filter functions are presented. In addition, a new symmetric phase only filtering (SPOF) technique, which is a variation on the traditional phase only filtering technique, is described and demonstrated. The SPOF technique exceeds the performance of the traditionally accepted phase only filtering techniques and is easily implemented in standard DPIV FFT-based correlation processing with no significant computational performance penalty. The SPOF-based optical correlation processing approach is presented as a new paradigm for more robust cross-correlation processing of low signal-to-noise ratio DPIV image data.

Using low numerical aperture lenses to achieve a large field of view when carrying out micron resolution particle image velocimetry (micro-PIV) experiments may result in the out-of-plane resolution being a significant fraction of the overall channel depth. A method to estimate the effect of out-of-plane resolution on micro-PIV velocity measurements is applied to two microchannel flows: a two-dimensional developed flow in a straight channel and a three-dimensional periodic flow in a ribbed channel. The method combines numerical simulation based on computational fluid dynamics (CFD) with an approximation for the contribution to the correlation function arising from partially defocused particles. The flows are then investigated experimentally with measurements obtained on a number of evenly spaced planes. The dominating factor in the comparison between the micro-PIV results and CFD simulations is not the spatial resolution of the experimental data, but instead the precision with which the geometrical parameters can be determined. A methodology is also presented for using micro-PIV results to measure the depth of microfluidic devices. Parabolic fitting of flow profiles allows the top and bottom surfaces of the channel to be located to within 0.2 µm.

We have investigated the application of a laser Doppler profile sensor for simultaneous measurement of position and velocity on moving rough surfaces. It is shown that, with this technique, the shape of rotating workpieces and components, e.g., turbine blades or turning parts, can be measured absolutely and in-process with only one single sensor. Measurements on different surfaces with defined shape and roughness are presented. The obtained minimum uncertainty in the position is about 250 nm in the centre of the measurement volume. For the velocity, a relative statistical error of 0.02% was obtained. Furthermore, shading effects as occurring for example at triangulation are reduced since illumination and signal detection can be coaxial. Because the measurement occurs contactless and a high temporal resolution is achievable, this sensor can open up new perspectives in the field of real-time production metrology, for example controlling the turning and the grinding process or at tip-clearance measurements in gas turbines.

Frequency domain analysis of dispersed strain signals in pressure bars has been an active area of study for many years. Over the last two decades, methods have been developed for the correction of dispersed signals, by adjustment of the phase angles of the Fourier components of the signal. Recent work has shown that, theoretically, dispersion correction should be significantly improved by the application of additional correction factors to the amplitude of the Fourier components. This paper describes a study using experimental and related numerical work to further test these proposals.

This paper reports on a novel method to characterize the mechanical properties, such as the elastic modulus, of circular microfabricated elastomeric membranes by measuring their load–displacement curve (compliance). An ultra-precision instrument was developed to simultaneously measure both the applied forces and the resultant displacements of the membrane under a central load. During all the deformation experiments, a high-resolution optical microscope system was used for viewing the side-view of the deformed membrane and thereby the complete membrane deflection profile was obtained. In the current experiment, measurements were performed on a set of thin, circular silicon rubber membranes with a thickness of 180 µm. A theoretical linear elastic solution was applied to quantitatively correlate the elasticity to the measured membrane compliance under such a central point load. Viscoelastic response of the membranes used was not manifested; therefore, it had no effect on the determination of Young's modulus of the membranes in the current work. The good agreement between the experimental results and the theoretical analyses facilitates the determination of Young's modulus of the membrane.

A nuclear magnetic resonance (NMR) spin–spin (T₂) relaxation technique has been described for detecting post-damage microstructural changes in human cortical bone tissue. The technique is applied to quantify apparent changes in bone porosity resulting from cyclic loading induced microdamage in cortical bone. Overall bone porosity is determined using the calibrated NMR fluid volume from the proton relaxation data divided by the overall bone volume. The NMR porosities obtained from cortical bone specimens pre- and post-damage are compared with the currently available but destructive histomorphometrically determined porosity. The advantages of using NMR T₂ relaxation techniques for bone microdamage are illustrated. The T₂ relaxation data can be inverted to T₂ relaxation distribution. The inversion T₂ relaxation distribution can then be transformed to a pore-size distribution with the longer relaxation times corresponding to larger pores if the surface relaxivity constant is known. It is shown that by using NMR 2 MHz or 27 MHz proton resonance, similar surface relaxivity constants are obtained. It is also demonstrated that the NMR T₂ relaxation data are sensitive to changes resulting from the creation of microdamage in cortical bone, which can be interpreted as an effective increase in bone porosity. These results indicate that the detection of cortical bone microdamage is possible by this technique.

The problem of extraction of information from the capacitance measurement between a Gaussian rough surface and a controlled electrode is studied with 2D numerical simulations. It is shown that with a plane electrode it is possible to obtain information on the standard deviation of the height, but not on the correlation length. With a corrugated, periodic electrode, on the other hand, information on the correlation length can be obtained without much difficulty, but the determination of the rms value of the height variations of the surface is limited due to poor depth resolution. Considerations of the results of numerical calculations of Gaussian randomly rough surfaces with a plane electrode and with a corrugated electrode suggest a simple method of obtaining the statistics of this type of surface.

Generally, the extended Kalman filter (EKF) is used for sensor fusion in a land vehicle navigation system. However, defects of the first-order linearization of the nonlinear model in the EKF can introduce large estimated errors, and may lead to sub-optimal performance. In order to yield higher accuracy of navigation, in this paper, a novel particle filter (PF) for sensor fusion is proposed and the sampling importance resampling particle filter (SIR-PF) is applied to address the nonlinear measurement model and it shows better performances when compared with the EKF. The basic theories and application of the general PF and the SIR-PF for a global position system/dead reckoning (GPS/DR) integrated navigation system are discussed.

This article revisits the proposal that the mass of any particle or atom can be defined absolutely as its rest-frame de Broglie frequency—a definition which requires no arbitrarily chosen mass standard, which has the physical basis that this frequency does indeed display the property of inertia, and which also makes electric charge dimensionless. The mass of any macroscopic object in kilogram could then be specified as a defined multiple of its total absolute mass, which is sufficiently close to the conventional mass for all present-day practical purposes, without having to wait for the determination of Avogadro's number or Planck's constant with few-parts-per-billion precision.

In situ thin film thickness monitoring and profilometering was implemented for a spray pyrolysis deposition technique, based on measuring the reflection of a laser beam from the film–substrate system; the interference pattern obtained during film deposition was used to calculate the film thickness. Recording in situ the reflection pattern for points separated 0.5 cm along the film allowed the film profile to be obtained. The growth of pyrolitic ZnO film on glass was studied by this method. It was found that the in situ measured thickness is in good agreement with that measured by ex situ profilometering or scanning electron microscopy (SEM). SEM micrographs of the films showed that the surface is formed by elongated-shaped particles, which become rounded and larger in size as the film becomes thicker. This fact was correlated with an increment of the diffuse component of the optical spectral reflectance and transmittance and with the trend of the surface roughness factor, obtained in situ.

We experimentally investigate the use of an arrayed waveguide grating (AWG) to interrogate fibre Bragg grating (FBG) sensors. A broadband light source is used to illuminate the FBG sensors. Reflected spectral information is directed to the AWG containing integral photodetectors providing 40 electrical outputs. Three methods are described to interrogate FBG sensors. The first technique makes use of the wavelength-dependent transmission profile of an AWG channel passband, giving a usable range of 500 με and a dynamic strain resolution of 96 nε Hz^−1/2 at 13 Hz. The second approach utilizes wide gratings larger than the channel spacing of the AWG; by monitoring the intensity present in several neighbouring AWG channels an improved range of 1890 με was achieved. The third method improves the dynamic range by utilizing a heterodyne approach based on interferometric wavelength shift detection, providing an improved dynamic strain resolution of 17 nε Hz^−1/2 at 30 Hz.

A cylindrical capacitive sensor (CCS) is developed and applied to several rotating machinery applications due to its merit of accurately measuring the spindle error motion without a significant effort. So far, analysis and design were performed using a linearized approximate model of the CCS. This paper presents a nonlinear analysis of the CCS. First, a nonlinear mathematical model of the measuring process of the CCS is derived, and it is shown that the nonlinearity is affected not only by the rotor position but also by the angular size of the CCS. Second, odd harmonic errors are found to exist in the measured rotor displacement through a power series form derivation of the measuring process. Among various harmonic errors of the nonlinear gain, the third harmonic error is found to be largest one and this error can be removed by using an eight-segment CCS with 120° sensor angular size. Third, a nonlinear analysis of the effect of geometric errors is performed and the analysis result shows that the harmonic geometric error of a rotor is transmitted into different harmonic errors and its transmitability increases as the normalized eccentricity increases. The nonlinear analysis in this paper will give us useful information for the design and application of the CCS.

To study the effect of the fuel concentration gradient on the auto-ignition phenomenon in a cylinder, a rapid compression machine (RCM), whose axial direction is set horizontally, is employed. The fuel concentration gradient is controlled by adjusting the duration of evaporation and diffusion from the time of injection of the fuel into the bottom of the cylinder. To achieve the desired concentration distribution, the history of the fuel concentration distribution in the RCM is measured by an infrared laser absorption method. There is a possibility that the initial concentration distribution will be altered by gas movement such as roll-up vortices, so the resistance wire CT method is proposed and applied. This method can measure a two-dimensional temperature distribution generated by a fuel concentration gradient during the compression process with the histories of wire resistance variations and computer tomography (CT) algorithm. The auto-ignitions took place with various concentration distributions in the cylinder. Various patterns of pressure history and direct photographs of the auto-ignition process are recorded. The Livengood–Wu integral method is applied using two-dimensional temperature and concentration history and the results of knocking intensity and tendency coincide qualitatively well with the experimental data.

Data reduction and calibration procedures are introduced for a novel rotating-analyser-type spectral imaging ellipsometer. Using a monaxial power spectrograph, we developed a unique spectral imaging ellipsometer. It combines one-dimensional imaging ellipsometry with spectroscopic ellipsometry, and enables real-time measurement of the optical parameters and dimensional structures of patterned or multilayered thin film. It also has more calibration factors than conventional ellipsometers. We therefore present a method using Jones matrices for describing the polarization sensitivity of the spectrograph and random noise of the CCD array. For a patterned SiO₂ layer on a silicon wafer, we used the spectral imaging ellipsometer to obtain a one-dimensional thickness profile.

By utilizing arc discharge, we have constructed an argon arc radiation source for ultraviolet (UV) radiometry, and have studied how to initiate the source using laser-induced gas breakdown (LIGB) without any contamination or degradation of its components. Investigation of the source of the arc discharge characteristics and spectral radiance characteristics of the arc yielded the optimum operating conditions. The spectral radiance of the arc source was stable to within 0.3%, and could be adjusted by controlling the arc current and argon pressure during normal operation.

Methods for real-time in situ analysis are needed for the monitoring and detection of pollutants and nutrients in water bodies. In situ instruments reduce the analysis time and decrease the chance of contamination or altering of the sample if taken to the laboratory. For this purpose, a mobile total analysis system is being developed for real-time environmental monitoring. The sensing platform, based on either an autonomous underwater vehicle or remotely operated vehicle, consists of an integrated flow-through channel with a chemically sensitive strip on one side of the channel and an optoelectronic detector opposite the strip. The strip contains an analyte-specific reagent that changes colour upon reaction with the analyte of interest introduced through the flow channel. In determining the feasibility of the design, the detection of copper ions was used as the case study. The interpretation strategy, referencing and calibration from the lab experiments including the effects of flow are shown.

In-flight measurements are made of the translational accelerations and attitude motion of a hand-thrown flying disc using miniaturized accelerometers and other sensors and a microcontroller data acquisition system. The experiments explore the capabilities and limitations of sensors on a rapidly rotating platform moving in air, and illustrate several of the complex gyrodynamic aspects of Frisbee flight. The data give insight into the biomechanics of Frisbee launch, and indicate lift, drag and pitch moment coefficients consistent with previous wind-tunnel measurements. The experiments constitute an instructive exercise in aerospace vehicle systems integration and in attitude reconstruction, and open the way to guided disc wings using control surfaces actuated during specific spin phases determined by onboard sensors.

This paper presents a novel technique for measuring the local axial, radial and azimuthal velocity components of the gas in bubbly gas–liquid flows using a local four-sensor conductance probe. A mathematical model is presented showing how the velocity vector of a gas bubble can be calculated from seven time intervals taken from the output signals from each of the four conductance sensors located within the probe. The paper goes on to describe the construction of a local four-sensor probe and the associated electronic measurement circuitry. Results are presented showing the distributions of the mean local axial, radial and azimuthal gas velocity components in vertical, bubbly gas–liquid flows, both with and without swirl. These results were obtained using the four-sensor probe in a vertical 80 mm diameter pipe into which a swirl generator could be installed. Additional results are presented showing the local gas volume fraction distribution, also obtained from the probe, in bubbly gas–liquid flows with and without swirl. It was found, as expected, that the presence of swirl caused a significant increase in the magnitude of the measured azimuthal velocity of the gas, particularly at the pipe walls. It was also found that, at a comparatively high water flow rate, the presence of swirl caused the gas bubbles to preferentially accumulate at the centre of the pipe.

The aim of the present work is to demonstrate the feasibility of a porous pressure-sensitive paint (PSP) for time-resolved surface pressure measurements in unsteady high-speed flows. The porous PSP was composed of bathophenanthroline ruthenium(II) complex, Ru(Ph₂-phen) and a silica-gel thin-layer chromatography aluminium plate. The dynamic response of the porous PSP was characterized by a point-wise luminescence intensity measurement conducted in a shock tube facility. The result showed that the time constant of the porous PSP was 13.6 µs. The porous PSP was then applied to the surface pressure distribution imaging of an unsteady flow induced in a two-dimensional Laval nozzle by using a fast-framing complementary metal oxide semiconductor camera. It was clearly shown that the porous PSP well captured the shock-wave motion of the order of kilohertz during the starting process of the supersonic nozzle in a qualitative manner.

The analytical solution for the frequency response of a constant temperature hot-wire system at given turbulent velocity fluctuations is used to investigate the effects on the cut-off frequency of the hot-wire system at different operating conditions. It is found that the optimal frequency response of the hot-wire system determined by the traditional square-wave test, under certain circumstances, cannot guarantee the stability of the hot-wire system from velocity perturbations. It is also found that the cut-off frequency depends not only on the hot-wire length to diameter ratio, the mean velocity, the hot-wire materials and the overheat ratio as previously found, but also on the gain bandwidth product of the amplifiers in the feedback circuit, the resistances used in the Wheatstone bridge and the gain of the amplifier in the feedback loop. Hot-wire systems with preferred operating conditions are recommended.

This study deals with two aspects of measurement science: the forward problem of optimal calibration and the inverse problem of optimal restoration of input data based on the output of the measurement system. Within the frame of the forward problem a general methodology is developed and introduced for the choice of the optimal fitting set points to result with the minimal uncertainty of a calibration curve of any order. Examples of the application of this methodology to the optimal determination of calibration polynomials up to the seventh degree are provided. Within the frame of the inverse problem, a methodology is developed for the optimal restoration of input data given the response function of the measurement system and taking into account a priori information regarding statistical features of the input data. The methodology is illustrated by three examples of its application to practical systems. It is implied that establishing a unified methodology for the optimization of both the forward and the inverse problems is feasible.

The paradigm set forth by Pereira et al (2000 Exp. Fluids29 S78–84) was an important milestone in capturing the optical geometry of a three-dimensional defocusing digital particle image velocimetry (DDPIV) design within a set of systematic equations. However, the opportunity to improve upon their pseudo-three-dimensional conceptual implementation of the two-dimensional equations exists by revisiting the derivations of these equations and revising some of their assumptions in order to define a modified set of equations for a true full-three-dimensional derivation. This paper introduces this newly revised set of equations that will explicitly and more accurately represent the three-dimensional DDPIV measurement system. A three-dimensional geometric uncertainty model has also been established through uncertainty analysis. Finally, a discussion of the differences and benefits of the new system of equations is presented.

Image analysis is currently used to study the structure of materials. In this paper, classical techniques of signal analysis such as power spectrum and autocorrelation functions are applied for this purpose. These techniques allow the detection of hidden features in complex images such as fracture surfaces, before highlighting them using spatial filtering. For that, the power spectrum allows us to locate the range of scales of peculiar texture, present but invisible to the naked eye. Thus, this study deals with the homogeneity of epoxy networks. Another way to use the method concerns aggregation phenomena. Indeed, the method allows us to assimilate several images of different magnifications, in order to characterize the aggregation on a large range of scales from the size of the particle or fibre to the full size of the sample. Then, the peculiar morphology of thermoplastic composites made from commingled fibres can be studied. The autocorrelation function method is also used to highlight features in the images of such composite materials.

We demonstrated a simple interrogation system for multiplexed fibre Bragg grating (FBG) sensors in a high-frequency range. A tunable fibre Fabry–Perot (FFP) filter with narrow free spectral range (FSR) was used to simplify the multiplexing demodulator for FBG vibration sensors. A stabilization-controlling unit was also developed for the maintenance of maximum sensitivity of the sensors. In order to verify the performance of the stabilization control unit, we measured the sensitivity of the FBG sensor by changing environmental temperature, and the system showed an average sensitivity of 2.5 nε_RMS Hz^−1/2 for a stabilization-controlled case. Finally, multi-point vibration tests using in-line FBG sensors were conducted to validate the multiplexing performance of the FBG system.

A new way of evaluating the ratio between a reference wavelength radiation and an unknown wavelength radiation in a two-beam interferometer is proposed here. The advantage of two-beam interferometry is the sinusoidal fringe signal for which precise phase detection algorithms exist. Modern algorithms can cope with different sources of errors, and correct them. We recall the principle of the Michelson-type lambdameter using temporal interference and we introduce the Young-type lambdameter using spatial interference. The Young-type lambdameter is based on the acquisition of the interference pattern from two point sources (e.g. two ends of monomode fibres projected onto a CCD camera). The measurement of an unknown wavelength can be achieved by comparing with a reference wavelength. Accurate interference phase maps can be calculated using spatial phase shifting. In this way, each small group of contiguous pixels acts as a single interferometer, and the whole set of pixels corresponds to many hundreds or thousands of interferometric measurement system units. The analysis of uncertainties shows that resolutions better than 10⁻⁷ can be achieved. An advantage of the fibre wavelength metre described here is the measurement velocity that takes only a few seconds.

The influence of a non-circular cross section of a preform on the measurement of the refractive index was investigated. The deviation of the refractive index value from the value for a circular cross section can be both positive and negative depending on the rotational orientation of the preform. Therefore, the measurement is done at many rotational angles and the refractive index is calculated by averaging over the angles. The best solution was to define a number of layers and calculate the thickness and index for each layer in the preform. This was done for each angle and then the average of all index values for all angles was calculated. With this approach, the obtained refractive index is almost independent of the ovality for a variation of the diameter of ±10%.

In order to further improve the rotation accuracy of ultraprecision roundness measuring instruments, a new single-step rotation error separation technique (SEST) is proposed to accurately separate instrument spindle rotation error and workpiece roundness error. This is done by first selecting an appropriate rotation angle and rotating the workpiece through a small angle with respect to the instrument spindle, and then measuring mixed errors A(n) and B(n) including workpiece error g(θ) and spindle rotation error z(θ) before and after rotation, and finally achieving accurate separation of z(θ) and g(θ) through Fast Fourier Transformation and harmonic analysis. Theoretical analysis and simulation results indicate that the signals in the harmonic range 1–100 upr can be totally separated from the wide range of harmonics measured when rotation angle error Δα < 0.01° is achieved by optimizing the rotation angle α. In comparison with the multi-step separation technique commonly used in roundness measuring instruments, SEST can be used to totally eliminate any harmonics singularity in the range of 1–100 upr and to make the error separation system very simple, shorten the separation process and reduce the separation time.

Ultrasonic sensors together with a fast data acquisition system have been used to monitor the barrel thickness and barrel/screw separation during low-density polyethylene as well as high-density polyethylene extrusion in 30 mm and 50 mm twin-screw extruders. The sensors include sol–gel sprayed high temperature (HT) piezoelectric thick ceramic film ultrasonic transducers (UTs), stand-alone HTUTs and air-cooled buffer rod type sensors consisting of a room temperature UT and a non-clad or clad buffer rod to which the room temperature UT is attached. The installation and use of these sensors are non-intrusive to the extruder and non-destructive to the polymers being processed. This study has demonstrated the capability of appropriately designed ultrasonic sensors in monitoring the barrel and screw integrity at the melting, mixing and pumping zones of the extruder via barrel or flange. The merits and limitations of these sensors are discussed. The measurement speed and analysis of the sensitivity for quantitative wear measurements are also presented.

The motion of a single particle in a fluidized bed has been followed with high temporal and spatial resolution using an ECAT EXACT HR+ PET camera. An account is given of the analysis of the output from the camera, and the calculation of the particle position. The particle position was determined with a precision of 1 mm once per ms. The scatter in the data was calculated using two methods, the variate method and a linear fit method. A comparison of the still bed with the fluidized bed did not reveal any fast particle movement in the latter. It is shown that optimal smoothing of the time series, without losing information about the particle movement, is obtained if the arithmetic averages of 10–50 data points are plotted against time.

A new low temperature, ultrahigh vacuum cryostat design has been developed for atomic force and scanning tunnelling microscopy measurements. A microscope can be operated at 5 K in ultrahigh vacuum. The microscope body is thermally connected to a reverse pendulum and completely surrounded by a radiation shield. The design allows in situ dosing and irradiation of the sample as well as for easy access of tip and sample. The temperature performance and the vibrational properties of the reverse pendulum design are demonstrated in detail. A brief overview of low temperature instrumentation in scanning probe microscopy is given.

The present study proposes an approach for robust estimation of the instantaneous motion field from a time-lagged pair of PIV images. The method is based on phase correlation, where the phase of the Fourier components is used for motion parameter estimation. Unlike the cross correlation-based techniques, this technique uses 'whitening' FIR filters to sharpen the cross correlation maxima, thereby improving the accuracy of identification of the peak. The proposed method also combines the advantages of the phase correlation and the cross correlation techniques in determining the reliability of the estimates, thus providing a method of filtering out a significantly large number of spurious vectors. This reliability metric helps reduce the possibility of over-smoothing the flow field when performing data validation. With regard to the efficiency of the technique, both phase correlation and cross correlation are derived from the Fourier components of the same image region. Each of the estimates can thus be obtained in parallel, without increasing the computational complexity of the system. Unlike many region-based methods that are currently available, the entire motion is decomposed as a global and a local motion field, which helps accurately obtain high interrogation resolution estimates for the local motion field.

New Abel inversion methods were developed for real-time calculations of emissions from circularly and elliptically symmetric radiation sources. Using the geometrical relationship between the plasma and the measuring device, an upper triangular area matrix was introduced to determine the local plasma emissivity from the measured irradiation value. Fast calculations of Abel inversion could be achieved using this area matrix since there was no need to fit the data for inversion. Inversion correctness was checked by the use of a given test function. This fast calculation algorithm could also be applied for the radiation source with the elliptical symmetry. This technique was applied to the measurement of arc plasma intensity in flat and V-grooved arc welding.

Although the laser welding process has many advantages over other techniques, especially for the welding of complex parts, it requires a precise focusing of the laser beam onto the workpiece to achieve the proper penetration over the entire weld seam. However, the unavoidable set-up tolerances and thermal distortion of the workpiece can result in focus errors that should be minimized. In this paper, an optoelectronic device for real-time measurement and control of the focal position is presented. It is based on the non-intrusive capture of the light emitted by the welding process by means of an optical fibre inside the laser head, and the estimation and correction of the focal error from the analysis of the light at two different spectral bands. The reported system has been optimized for use in a real environment: it is robust, compact, easy to operate and able to adjust itself to different welding conditions. Details of the design, figures of performance obtained from lab testing and results from recent field trials on complex aerospace parts are provided.

The use of a single implantable microprobe to detect multiple, bi-directional action potential velocities is demonstrated. By using multiple electrode sites along the direction of propagation to record the same action potentials, the velocities of action potentials from both sensory and motor neurons can be determined.