Project has been carried out with partner, Institute for Physical and Information Technologies (ITEFI), Spanish National Research Council (CSIC).
The mechanical and chemical properties, temperature, stresses and structural integrity of materials can be studied using ultrasound due to its direct interaction with the physical properties of the material under study. Ultrasound examination uses probing signals, the reflection amplitude or delay time of which determines the parameter under investigation. Usually there are several reflections, each of them occupies a certain time, therefore they influence each other, complicating the interpretation of the obtained results. Applying decomposition allows you to break down a signal into components more suitable for imaging, interpretation, or measurement. This method requires a backup copy of the transmitted signal, which is used to determine the amplitude and arrival time needed to perform the iterative deconvolution. The application of the method is complicated by the fact that the probing signal in the material is attenuated differently depending on its frequency. Therefore, the shape and spectral content of the signal changes. For this reason, the quality of decomposition deteriorates or becomes impossible. It is new that it is proposed to adapt this support signal based on the residual decomposition energy. Attenuation and reflections from the structure degrade the signal-to-noise ratio. As a result, amplitude and propagation time determination errors increase, frequency components can be drowned in noise. This worsens the adaptivity of the signal and the quality of the decomposition. If a pulse is used for probing, the signal energy can be increased by increasing the signal amplitude. However, the amplitude is limited by the converter and excitation electronics. What is new is that it is proposed to use binary spread spectrum signals to change the situation: not only the signal-to-noise ratio is improved, but also the frequency band is widened, the mode of sending and receiving degrees is facilitated, it is possible to use smaller-sized, more economical research equipment. Adjacent signals increase position and amplitude detection errors, which degrades the quality of the decomposition. It is new that measures are taken to reduce these errors by applying iterative time and frequency domain decomposition methods. The improvements mentioned above provide a new quality of decomposition when receiving an opportunity. The project is intended to investigate the application of the method in the ultrasonic measurements of polymeric and composite materials.
Project funding:
KTU R&D&I Fund
Project results:
The equipment for generating and receiving ultrasonic signals was designed and prepared, supplementing the existing system with new excitation and reception cascades, since a generator that minimally distorts the mathematically described signal was needed to optimize the excitation signals. It is new that the topology of a bipolar two-stroke cascade with a floating controller is proposed, with a differential capacitive decoupling scheme for transmitting the control signals. An excitation cascade power supply unit with galvanic decoupling is designed, manufactured and tested. In the protection circuits of the receiver, it is proposed to use a combination of reverse connection of isolated gate field transistors with an induced channel and a diode limiter. The proposed methodology for complex evaluation of system parameters. The asymmetry of the excitation signal front was found to be reduced from 15ns to 3.5ns; receiver recovery time reduced from 11us to 3 µs.
Research was carried out to create an optimal set of excitation signals, evaluating the properties of the transducer and the material.
An iterative decomposition algorithm is proposed for the separation of received ultrasonic reflections, supplemented by a reiterative algorithm. By applying iterative decomposition, the near-zone time estimation errors have been reduced from 50ns to 25ns. After completing the reiterative algorithm, the errors are reduced to 5ns, and the close zone is reduced from 600ns to 50ns.
It is new that the iterative decomposition algorithm is proposed to be supplemented with the adaptation of the support signal, using a digital filter of freely selectable unbounded impulse response. In the investigated case, the residual decomposition energy was reduced to -28dB (-10dB without adaptation).
Two papers in WoS journals were published, 4 conference presentations and publications.
Period of project implementation: 2015-04-03 - 2015-12-31
Project partners: Institute of Information and Physics Technologies, Spanish National Research Council
