·Table of Contents
·Methods and Instrumentation
The Integrated Acousto-Optic Sensor for Quality Control of Layered StructuresM. WEVERS and P. DE MEESTER
Katholieke Universiteit Leuven,
Department of Metallurgy and Materials Engineering,
W. de Croylaan 2,
Katholieke Universiteit Leuven, Campus Kortrijk
Interdisciplinairy Research Centre,
E. Sabbelaan 56
|Fig 1: A scheme of the experimental set-up for the acousto-optic technique using the integrated device.|
This sensor is composed of a 3,5 MHz high-power ultrasonic transducer allowed to rotate in a fork and this fork can move vertically (Z-axis). One leg of this fork contains a laser diode (30 mWatt) and the light of the laser diode is deflected on a mirror so that it crosses the plane of ultrasound perpendicular. Once the laser light is modulated by the ultrasound the detection is done in the other leg, also after deflection on a mirror, using a silicon photo diode.
Fig 2: The acousto-optic sensor (side view and front view)
The specimen investigated is placed in a Plexiglas water tank perpendicular to the plane of the ultrasound and parallel to the plane of the laser. The 3,5 MHz transducer is powered by a sine wave generator (type SG4511 pulse/function generator from IWATSU) to produce continuous ultrasound. The amplified signal of the photo diode is read by a 500 MHz Waveform Digitizer board (DA500A of Signatec). This board is triggered by the sine wave generator. The phase difference between this optical signal and the reference signal coming from the sine wave generator is calculated by the computer for each measurement point and is used for analysis.
The X- and Y-axes to move the specimen have a resolution of 1µm and the resolution of the rotation of the transducer is 0,001°. The movement of the fork in the Z-axis is controlled by hand with an accuracy of 0,5 mm.
With the Vickers hardness indentation technique a diamond square = 136° is used to indent the metal. The Vickers hardness, a technological property, is defined by dividing the load P (1 up to 120 kg depending on the material) by the surface area of the indent.
The Vickers hardness values in the middle of the bars starting at about 1 mm from the bottom are given in figure 3 for the rectangular bar and in figure 4 for the Jominy bar. The distance between two measurements is about 1 mm over a total distance of 30 mm for the rectangular bar and 19 mm for the Jominy bar in the y-direction. The load used for indentation was 10 kg.
|Fig 3: Vickers hardness profile for the rectangular bar||Fig 4: Vickers hardness profile for the Jominy bar|
Also the phase variation over the same line has been measured with the new acousto-optic device. In first instance the Rayleigh angle of the substrate was automatically searched for and next for each point of the scan, the measurement of the phase of ultrasound in the big lobe of the reflected ultrasound was performed. The results are presented in figure 5 for the rectangular bar and in figure 6 for the Jominy bar. For the latter also the result obtained with the old set-up of the acousto-optic technique is included in figure 7. (4)
|Fig 5: Acousto-optic line scan of rectangular bar, profile of phase variation||Fig 6: Acousto-optic line scan of Jominy bar, profile of phase variation|
The obtained hardness values and phase variation along the lines were attributed to the microstructure and the residual stresses within the material.
It is obvious that the reason for a certain hardness value is found in the microstructure of the material, going from martensite towards ferrite/pearlite the hardness drops. (5,6) This information is also reflected in the phase measurements which have been done "non" destructively. In fact an astonishing good correlation can be found.
|Fig 7: Old acousto-optic line scan of Jominy bar, profile of phase variation||Fig 8: Phase shifts between the reflected and incident bounded ultrasonic wave due to changes in density.|
To explain the acousto-optic result theoretically and to reveal the relation microstructure-hardness-phase shift the influence of different parameters on the phase shift was studied. Since the modulus of elasticity hardly changes going from one microstructure to the other, the density remains as a possible influencing parameter. The density will change as a consequence of the distortion of the crystallographic structure during quenching. To study the influence of the density on the phase measurements theoretical calculations were carried out. For a steel substrate the influence of the density variation on the phase shift between the reflected and incident bounded ultrasonic wave was calculated numerically for an incident angle equal to the Rayleigh angle of steel (figure 8). According to literature values the transition of ferrite-pearlite to martensite can account for a 4% change in density. A clear phase shift of 80° can herewith be associated according to the calculations. However the influence of the residual stresses is not yet accounted for in this calculation and this explains the difference with the experiment.
Aluminium plate coated with paint
An aluminium plate of 100x60 mm and a thickness of 1mm was spray coated over 60% of the length of the specimen. Going from the aluminium towards the paint coating the new acousto-optic sensor device was tested and the ultrasonic transducer was positioned in the most outspoken Lamb angle of the aluminium plate (automatic search).
Three line scans with a step size in the Y-direction of 5µm and 0,2 mm in the X-direction were taken. One result is presented in figure 9. The variations in the coating thickness can be identified. This again proofs the opportunities of this new NDT technology focusing on the phase of ultrasound instead of the amplitude and retrieving the information using the interaction between ultrasound and laser light.
|Fig 9: Acousto-optic scan of aluminium plate coated with paint|
|© AIPnD , created by NDT.net|||Home| |Top||