技術論文とプレゼンテーション

Capacitive Micromachined Ultrasonic Transducers (CMUTs) fabricated on flexible polymeric substrates represent a transformative technology for conformable, non-invasive medical imaging. This work presents a rigorous computational framework for the finite element modeling, dynamic parameter extraction, and analytical validation of a single flexible CMUT cell tailored for ultrasound tomography applications. Utilizing COMSOL Multiphysics, the transducer’s electromechanical response was evaluated through coupled stationary and frequency-domain perturbation studies. To accurately simulate realistic acoustic radiation losses into a semi-infinite medium, Perfectly Matched Layers (PMLs) were integrated into the 2D axisymmetric model. To bridge this complex multiphysics data with 1D analytical representations, a numerical optimization strategy driven by the SNOPT algorithm was implemented, systematically fine-tuning the components of both the Butterworth-Van Dyke (BVD) equivalent electrical circuit and the Lumped Mechanical System (LMS). This optimized extraction methodology successfully captured critical structural and electrical nonidealities, including a quantified 6% structural softening induced by the boundary layers, the electrostatic spring softening effect under DC bias, and the dynamic shifting of the antiresonant frequency fundamentally driven by parasitic capacitances. The analytical framework was firmly validated using Mason’s equivalent circuit model: mapping the optimized mechanical variables into the electrical domain via an ideal electromechanical transformer yielded a negligible relative error of less than 2%. This profound convergence conclusively demonstrates the high fidelity of the 1D lumped-parameter framework. Ultimately, this work establishes the mathematically robust baseline necessary for future fluid-coupled vibroacoustic optimization. It paves the way for the low-temperature microfabrication of fully populated flexible arrays, capable of seamlessly adapting to curved biological tissues for high-resolution, deep-tissue diagnostic imaging.