Simulating Sensors and Detectors
Creating reliable models of acoustic probes requires reliable materials characteristics and sophisticated software systems.
One of the major driving forces for developing new sensors and detectors is in medical applications. This includes the integration of fiber optic sensors, smart sensors, silicon micromachined sensors (MEMS or microelectromechanical systems) and thin-film devices. Smart sensors are devices that incorporate electronic logic, control or signal processing functions and therefore offer enhanced measurement capabilities, information quality and functional performance.
The continuing trends for designing new iterations of these devices are in their increased complexity, miniaturization and the use of specialized materials. To design these small complex devices and meet the demands for quick optimized design turnarounds, researchers are increasingly employing multiphysics-based finite element modeling (FEM) and simulation systems. Researchers in the R&D Dept., Esaote S.p.A., Florence, Italy, for example used multiphysics FEM software from Comsol to develop medical ultrasound imaging probes. They also used a KLM (Krimholtz, Leedom, Matthaei) circuit model to find the starting values for the principal parameters involved in this model—the Comsol model was used for the complete modeling. With a given input specification for the probe (such as sensitivity and bandwidth), it was possible for the developers to run KLM simulations and calculate the optimized thicknesses for the four matching layers in the transducer design (in the multiphysics model) and the acoustic impedance, which was transferred to the Young’s modulus and Poisson coefficient. With these simulation models, the researchers were able to investigate both changes in materials performance and changes in the operational characteristics of the devices.
Ultrasound imaging transducers operate by generating a pressure field in the human body. Differences in the acoustic properties of different types of tissues allow the ultrasound scanner to generate an image. The quality of this image is directly related to the technology level of the materials used in manufacturing the transducer and an understanding of their interactions. The Esaote designers based their probe design on a 144-element array probe with 0.245 pitch, 5-MHz central frequency and a wide frequency range of 2 to 11 MHz. The transducer design was created from a special piezo-composite material, a hard rubber backing substrate four acoustic matching layers and a silicon rubber lens.
Developing an optimized transducer involves looking at a variety of specialized materials and manufacturing processes on specialized machines and with materials layers as thin as 30 microns. The software allowed the researchers to investigate changes in both the materials and in the operation. The use of multiphysics software models reduced the time needed to evaluate the different models from one year for physical models to just a month for the virtual device iterations.
In their first multiphysics model, the developers used perfectly matched layers to reduce computational time and computer system memory. For the transducers, it was important to study pressure in the far field region. The simulated far field pressure was compared to actual measurements taken with a membrane hydrophone placed at a 30-cm depth in a water tank. Biological (human) tissue has similar acoustic properties to water, which is often used as a standard for testing purposes.
Final results for the far field sound pressure levels revealed good agreement between measured and simulated transducer performance, with less than 3 dB of amplitude differences throughout the bandwidth operating range.
Development of the multiphysics model did not result in any unexpected results according to Lorenzo Spicci, the materials engineer and piezo/transducer designer for Esatote. “We did, however, find that we had a number of complications for optimizing the materials parameters that had to be supplied for the multiphysics model. It’s never easy to get all the data needed to complete the materials characterization from available literature or actual measurements.”
Follow-up multiphysics evaluations on these devices included a thermal analysis of the piezo-disk ultrasound probe. The virtual models created and evaluated with the multiphysics software have all resulted in the creation and validation of actual production devices.
The Esaote researchers revealed that for development of these ultrasound imaging transducers using multiphysics FEM software that:
- Elasticity, piezoelectric and dielectric permitivity matrices needed to be specified to build the model.
- Manufacturer's materials data are often incomplete and should be checked for the particular operating condition of the piezoelectric material.
- Physical insight should be used as the starting point for the model.
- Optimization procedures should be used.