Multichannel Photorefractive Laser Ultrasonic Sensor

The use of photorefractive laser ultrasonic sensor for laser ultrasound is now well established. The holographic principle on which it is based allows to register the speckled wavefront issued from the target with highly scattering surfaces and to give it back to create a local oscillator with an adapted wavefront. The imaging properties of holography gives the possibility to work with several illuminated points each imaged on its detector. We can then easily transform the system in a multichannel laser ultrasonic sensor. We present here the experimental implementation and characterization of such a sensor, showing that an uniform detection without crosstalk can be realized between the measurement points, if a properly designed set-up is used. Introduction: The basic principle of the Photorefractive Laser Ultrasonic Sensor [1] is a homodyne detection in which we create a reference beam that is identical to the signal beam whatever its wavefront structure. We send the phase modulated signal beam on the photorefractive crystal with a coherent pump beam (Fig. 1). These beams interfere and create in the material a hologram of the signal beam structure. This hologram is a stationary hologram, as the photorefractive crystal has a rather slow response time compared to the time evolution of the phase modulation. The pump beam diffracts on this hologram as a wave that will propagate in the same direction and with the same wavefront structure than the transmitted beam, but without its phase modulation. This diffracted beam thus acts like a local oscillator in a reference beam interferometer. This mechanism is illustrated in the following expression : Es x,t ( )= e 2 Es t ( ) e −1 ( )+ Es t ( ) [ ] (1) in which the first term of the right member of the equation is the amplitude of the diffracted beam and the second term is the transmitted signal beam. The diffracted beam has the same spatial structure than the transmitted signal beam ES(t) but is sensitive only to the mean value of the amplitude (indicated by brakets that means an average over the response time of the holographic media). Thus the written hologram is stationnary and is read by a stationnary pump beam what gives a stationnary diffracted beam which amplitude depends on the diffraction efficiency of the hologram characterized by a photorefractive gain γ (the absorption α characterizes the losses of the media of thickness x). Diffracted pump beam and signal beam will then interfere (optimally in quadrature as in an homodyne detection) and give an intensity modulated signal [1]: Is x,t ( )= e Is0 e ′ γ x − 2e ′ γ x sin ′ ′ γ x φ t ( ) [ ] (2) To obtain this expression we consider a phase modulated beam with a phase modulation φ(t) created by an ultrasonic vibration of low amplitude (δ(t)<<λ/2). The photorefractive gain is a complex parameter (γ=γ'+iγ") that renders the spatial displacement of the hologram compared to the illumination pattern (an index modulation in phase with illumination would give a purely imaginary gain). In order to have a linear demodulation the imaginary part γ" of the gain should be non zero, what is equivalent to the quadrature condition in a classical homodyne detection. In usual photorefractive crystal, the gain is purely real (π/2 phase shifted index grating), unless an external DC field is applied to the crystal [2]. We can note that the use a dynamic holographic material (such as photorefractive crystals) allows the set-up to adapt to fluctuations of the signal beam slower than the crystal response time, writing a new adapted hologram almost instantaneously. This properties gives the Photorefractive Laser Ultrasonic Sensor a high pass frequency response [3], with a cut-off frequency that is related to the response time of the dynamic holographic material. In the case of the photorefractive materials this response time is controlled by the pump beam power and cut-off frequency around 10kHz can be obtained giving a system insensitive to the low frequency acoustic vibrations that are usually encountered in industrial environment [4]. This system is now well characterized [5] and is commercially available. Fig. 1. Schematic principle of the Photorefractive Laser Ultrasonic Sensor The detection system is usually used with a single detector (even if often in a differential configuration) i.e. that only one measurement point is tested at each laser shot. In usual experimental configurations, the illuminated point on the tested sample is imaged in the photorefractive crystal and imaged again on the photodiode. If we now replace the photodiode by an array or a line of detectors, several detection points can be defined on the target, imaged on the array of detectors and the ultrasonic signal can be measured simultaneously on these multiple points using a single optical system and a single photorefractive crystal [6]. The principle of the multichannel laser ultrasonic sensor thus derives naturally from the Photorefractive Laser Ultrasonic Sensor. We will present here the implementation of the multichannel laser ultrasonic sensor and an experimental characterization of its performances, regarding the uniformity of the response and the possible crosstalk between the different channels. Implementation of the laser ultrasonic sensor: The first precaution to take, in the implementation of the multichannel sensor, is the illumination of the target. The main parameter that governs the sensitivity of the sensor is the light power collected from the target for a given point [7]. If the laser beam illuminates a large area of the target, the spatial resolution and the position of the tested point on the target will be defined by the detector array. If the detector matrix has a low filling factor (total area of the sensitive zones compared to the total surface of the sensor) a lot of light will be lost by illumination of useless area. The best implementation of such a multichannel sensor will thus require the use of a structured illumination, in the form for example of an array of points of small width and separated by the desired period. This array of illumination points will be then imaged on the detector array, each photodiode receiving light from a well defined area on the target. The resolution is then defined by the size of the illuminating spots, and no more by the photodiode array, and all the light of a single point will be sent on its corresponding photodiode only (in the limit of low inter-spot scattering level). To generate this array of points we use a diffractive spot array generator (Dammann grating), that generates a 5x5 spot array (Fig. 2A) with a 1mm spacing when placed at the object focal plane of a 63.5mm focal length Fourier lens. The typical diffraction efficiency in each spot was about 3% what gives a total diffraction efficiency around 70%.