L. Gomez-Ullate

An ultrasonic imaging system based on a new SAFT approach and a GPU beamformer

The design of newer ultrasonic imaging systems attempts to obtain low-cost, small-sized devices with reduced power consumption that are capable of reaching high frame rates with high image quality. In this regard, synthetic aperture techniques have been very useful. They reduce hardware requirements and accelerate information capture. However, the beamforming process is still very slow, limiting the overall speed of the system. Recently, general-purpose computing on graphics processing unit techniques have been proposed as a way to accelerate image composition. They provide excellent computing power with which a very large volume of data can easily and quickly be processed. This paper describes a new system architecture that merges both principles. Thus, using a minimum-redundancy synthetic aperture technique to acquire the signals (2R-SAFT), and a graphics processing unit as a beamformer, we have developed a new scanner with full dynamic focusing, both on emission and reception, that attains real-time imaging with very few resources.

A computational method to calculate the longitudinal wave evolution caused by interfaces between isotropic media

This paper presents a computational method to calculate the reflected and transmitted ultrasonic fields at interfaces of complex geometry. The method is performed in two steps. As first step, the velocity potential impulse response from an arbitrary aperture is determined at the interface using the Rayleigh integral and considering the reflection and transmission coefficients. In a second step, the simulated fields are calculated by applying the Rayleigh-Sommerfeld integral to the whole, extended interface. In order to validate the method, some experimental cases as, for instance, plane and cylindrical concave surfaces between two media (water-acrylic) were tested. The experimental ultrasonic fields are in good agreement with those provided by the model. Furthermore, in the work, the compromise between the accuracy of the method and the computation time is studied.

P3R-3 Design and Characterization of Air Coupled Ultrasonic Transducers Based on MUMPs

This paper deals with the design and characterization of several capacitive micromachined ultrasonic transducer (cMUT) cells for the future design of an air-coupled transducer for non destructive testing (NDT). Each design was manufactured using the multi-user MEMS process (MUMPs). Special boundary conditions were used to obtain high efficiency: two opposite sides of the 150 mum square shaped membranes are free while the other sides are anchored to a fixed layer of the process. Using these designs a large displacement of the membrane and a higher surface contact is obtained in spite of their mechanical losses

Double frequency piezoelectric transducer design for harmonic imaging purposes in NDT

Harmonic imaging (HI) has emerged as a very promising tool for medical imaging, although there has been little published work using this technique in ultrasonic non-destructive testing (NDT). The core of the technique, which uses nonlinear propagation effects arising in the medium due to the microstructure or the existence of defects, is the ability to design transducers capable of emitting at one frequency and receiving at twice this frequency. The transducers that have been used so far are usually double crystal configurations with coaxial geometry, and commonly using a disc surrounded by a ring. Such a geometry permits the design of broadband transducers if each transducer element is adapted to the medium with its corresponding matching layers. Nevertheless, the different geometry of the emission and reception apertures creates difficulties when resolving the images. In this work, a new transducer design with different emission and reception apertures is presented. It makes use of the traditional construction procedures used to make piezocomposite transducers and the well-known theory of the mode coupling in piezoelectric resonators when the lateral dimensions are comparable with the thickness of the piezoceramic. In this work the design, construction, and characterization of a prototype to be used in NDT of metallic materials is presented. The acoustic field is calculated using water as a propagation medium, and these theoretical predictions then are compared with the experimental measurements. The predicted acoustic performances for the case of propagation in stainless steel are shown.

P4M-2 A Linear CMUT Air-Coupled Array For NDE Based on MUMPS

In this paper, the design and characterization of a linear cMUT array based on MUMPS technology is presented. The array has 33 elements having each element 68times2 cMUT cells electrically connected in parallel. The single cell consists of a 130 mum square shaped polysilicon membrane, a gold top electrode, silicon nitride as the isolation layer between conductors and heavily doped silicon which works as the bottom electrode. The central frequency is 720 kHz. Electrical and optical measurements were performed and compared with FEM calculations. These results were used to study the mechanical and electrical behaviour of the device.

A computational method to calculate the transmit-receive mode echo responses from targets of complex geometry [ultrasonic nondestructive testing applications]

In this paper a 3D computational method is proposed to calculate the echo responses from finite-sized targets of complex geometry. It can predict the form of the echo response from the target, considering mode conversion at the reflector surface. The model was compared to other modeling tools, and excellent agreement between our results and those of the others were found. Moreover, experimental results obtained using broadband transducers together with, for instance, plane and cylindrical concave rectangular reflectors in a water medium were compared to the theoretical results. The method can predict the measured echoes accurately.

Method and architecture to accelerate multi-element synthetic aperture imaging

Multi-element synthetic aperture focusing (n-SAF) methods have been proposed as a suitable way to reduce cost and size for complex ultrasonic imaging systems. In this method, the larger the sub-aperture, the better the image contrast. However, when the number n of elements in the sub-aperture increases, some problems arise due to the great number of signals involved in the beamforming process, demanding a more complex electronic architecture, a higher bandwidth to manage the signals, and a greater computational power to compose the images in real time. This paper presents the n-SAF method in combination with a data reduction algorithm that reduces the number of signals intervening in the image data processing. The proposed method, called nR-SAF, simplifies the beamforming process; the hardware adds all signals in phase from identical coarray elements (similar spatial frequencies) thus a reduction of approximately 2/n of the data intervening in the dynamical focusing process is attained. The phase errors due to the simplified algorithms are also analyzed. We concluded that there are some limitations for sub-aperture size when image points are very near the transducer. Finally, an electronic architecture is presented, which is able to implement high velocity images from nR-SAF methods.

Design of Curvilinear Array Apertures for 3D Ultrasonic Imaging

The development of ultrasonic volumetric imaging is closely linked to the development of systems that are able to operate bidimensional array transducers. These arrays are useful for ultrasonic volumetric imaging, because they produce steered and focused beams throughout a volume of interest. Typical 2-D arrays are based on a Squared Matrix (SM) configuration, where the array elements are the matrix cells. Their performance is determined by their width in terms of wavelenght. Resolution and the dynamic range are determined by wavelenght/aperture diameter ratio and number of elements and the wavelenght/ interelement distance ratio respectively (Smith et al. (1991)). In SM apertures, since element distribution is uniform, the composition of the secondary lobes is determined by interelement distance. All elements contribute to its formation. These lobes are known as grating lobes and produce image artifacts that can reduce the signal-to-noise ratio. Nowadays avoiding image artifacts is key to array design. For matrix and linear arrays the composition of these lobes can only be avoided by limiting the interelement distance to λ/2. In practice, it means that for 1o of lateral resolution a 60λ diameter aperture with 14400 elements is needed. Then, several problems can be identified:

Piezoelectric sectorial 2D array for 3D acoustical imaging

3D acoustical imaging for both NDT and Echography purposes needs of 2D array transducers. Among them, the sectorial design gives good performances maintaining enough resolution. A piezoelectric transducer array having a piezocomposite as substrate is designed, simulated, and tested following time and frequency measurement techniques. Simulations were performed with a multilayer scheme based on the KLM circuit. Time response, inter-element cross-coupling and acoustic aperture are measured.