采用离散傅里叶变换方法对颤振时间响应历程计算结果进行分析，提出了一种从多个模态分支计算结果中归纳耦合主颤振分支的方法。分别从频率移动理论和能量变化观点出发，提出了针对颤振边界的临界颤振动压判定频率移动判据和能量因子判据。通过对国际标准跨声速颤振算例AGARD445�6机翼的时域颤振计算结果进行分析，验证了方法在耦合模态分支确定、临界颤振动压判定、主颤振分支判定3个方面的高有效性。此外，将该方法应用于高速中等展弦比气动翼面的跨声速颤振特性研究，成功地根据较复杂的响应曲线判断出了颤振边界，表明该方法具有良好的工程应用前景。%This paper puts forward a new method which is used to pick out coupling modes from muti-modes and to estimate flutter boundary. The means of the research was Discrete Fourier Transform( DFT) , by which the time do-main result was transformed into amplitude-frequency plot. After confirming the stability of DFT program written by us, we chose time domain flutter response of Wing AGARD445�6 at Ma=0�9 to be the research object. Amplitude-frequency plot of general displacement showed the main frequency of each mode. Based on frequency superposition principle, the two modes with main frequencies close to each other are coupling modes. With coupling modes known, there are two ways to obtain flutter boundary, Q-f plot and energy factor estimation. In Q-f plot, the fre-quency superposition point is flutter point and the flutter velocity and flutter frequency can be found. The calculation of energy factor follows. First, calculate the general oscillation mechanical energy. Second, find the linear fit ( or other fitting means) function of general oscillation mechanical energy. Define its slope ( or other parameter) as en-ergy factor. At last, draw the plot showing the energy factor changing with velocity ( or dynamic pressure) , and the point of curve on the x-axis is flutter point. Each coupling mode has its own flutter point, and the lower one is the real flutter point and its corresponding mode is the main flutter mode. Then, to verify the overall prediction preci-sion , all experimental points of Wing AGARD445�6 were calculated and the deviation was between 2% and 6% ex-cept the result at Mach number 1�141. This means the deviation is acceptable. To test the practicality of this means, apply it to a project of predicting the flutter boundary of a rocket tail. The results and their analysis show preliminarily that this means can find effectively coupling modes, main flutter modes, flutter velocity and flutter fre-quency. The preliminarily conclusions are, (1) by the means of Discrete Fourier Transform(DFT), coupling modes can be found effectively; ( 2 ) frequency superposition criterion can predict flutter frequency and flutter velocity;(3) energy factor criterion can provide flutter velocity and main flutter mode. Although both criterias can predict flutter velocity, the result given by energy factor criterion has higher precision.
展开▼
