A new computation method to predict a transonic wing flutter under the limit-cycle oscillation (LCO) is developed using the time spectral (TS) method. The fluid and structural unknowns at a specified time interval in one cycle of the flutter are obtained by steady-state solutions of the aeroelastic TS equations, and thus higher computational efficiency is expected. Determining the frequency is the key for the flutter prediction, and we propose a method of minimizing the residuals of the fluid equation. The time histories of the plunge and pitch are reconstructed by the discrete Fourier transform, and compared with the existing time marching (TM) method. The effects of the number of harmonics in the TS and the time-step size in the TM are examined for the detailed comparison. The frequency, amplitude, and phase differences between the structural vibration modes all well agree between the two methods for a wide range of flutter speed index. Complex flutter boundaries are also predicted. The TS method is faster to reach the LCO than the TM method for the first vibration mode of the flutter. It is slower for the second vibration mode, but the increase in the computation time is much smaller than the reduction in the first mode. In general, the TS method is particularly advantageous for the prediction near flutter boundaries and is useful for aircraft design.

The generated cavitation on the NACA16-012 hydrofoil suddenly disappears at a specific angle of attack, even at a low cavitation number, where cavitation normally develops. This was named the cavitation disappearance phenomenon by the authors. It has not been clear why, once cavitation occurs, it disappears. In this study, pressure distributions on a suction surface under noncavitation and disappearance conditions were measured in the cavitation tunnel at an angle of attack at which the cavitation disappearance phenomenon occurred. Moreover, the boundary layer and pressure distribution on a suction surface under noncavitation conditions were comprehensively measured in the wind tunnel at a wide range of the angle of attack. As a result of the wind tunnel experiment, the angle of attack where the cavitation disappearance phenomenon occurred corresponded to the angle of attack where the onset of the short bubble occurred under noncavitation conditions. As the result of the cavitation tunnel experiment, the pressure distribution was found to differ between noncavitation and disappearance conditions. Under disappearance conditions, the suction peak disappeared and the low-pressure region extended to the downstream region. Additionally, the pressure distribution under the disappearance condition seemed to correspond with the distribution in a burst state, which was measured in a wind tunnel. At the same time, the boundary layer in the sheet/cloud cavitation periodically repeated the short bubble (during cloud release) and burst (during sheet cavitation) states. The disappearance of cavitation may be caused by the situation in which the burst boundary layer cannot return to a short bubble after the cloud cavity release and sheet cavitation cannot occur because the pressure is high in the burst boundary layer.