Ultrasonic wave propagation in polycrystalline materials is strongly influenced by microstructural heterogeneity, leading to attenuation, dispersion, and scattering effects. Establishing quantitative links between ultrasonic measurements and microstructural features such as grain size remains a key challenge. In this work, a digital twin of ultrasound experiments on polycrystaline specimens is presented based on an FFT-based elastodynamic solver[1]. Polycrystalline representative volume elements with controlled grain size distributions are generated to enable systematic parametric studies. Different excitation signals representative of ultrasonic transducers—including broadband pulses, tone bursts, and continuous waves—are applied over frequency ranges relevant to practical inspection. Time- and frequencydomain metrics such as time of arrival, peak amplitude, spectral content, and signal energy are extracted at multiple receiver locations. The simulations reveal clear trends linking grain size, excitation frequency, and signal attenuation, highlighting the transition between different scattering regimes. These results demonstrate the capability of the proposed framework as a virtual ultrasonic testing tool and provide a basis for correlating numerical predictions with forthcoming experimental measurements.