《卫星力控振动试验的模型简化原理分析》

航天器振动环境试验的首要目标，是为了发现航天器在结构动力学设计方面存在的问题，防止在发射过程中由于严酷的振动环境导致发射任务失败。在常规航天器的振动测试中，振动台台面的加速度输入是按照加速度规范进行控制的，将飞行环境中实际测量得到的加速度峰值的进行包络，这种测试方法会在卫星或飞船的固有频率处产生较大的过试验现象。为此，NASA从1993年以来，推行“力限振动试验（force limited vibration testing）”技术，即采用振动台加速度和界面力进行“双控”，以降低过试验的危害。但是，由于真实飞行时的星箭界面力无法通过实际测量得到，因此，通过对星箭界面的计算模型、实验获得的模态参数、以及从实测界面加速度等条件下获得界面力数据的研究成为FLVT技术研究的核心内容。本文通过星箭耦合系统的动力学模型，利用子结构模态综合法推导出界面动力学响应以及界面力与界面加速度的关系式。然后根据子结构各阶主模态在特定频率区间内的特性，对动力学模型进行简化，并对NASA文献中的“复杂二自由度模型”进行了理论证明和误差分析。

Analysis of Model Simplification Method in Force Limited Vibration Testing

In conventional vibration test of space hardware, the acceleration input at the interface between the test article and the mounting structure is controlled to a specification which is the envelope of the acceleration peaks in the flight environment. This traditional approach of testing is known to produce great overtesting to the test article at its own resonance frequencies. An improved approach called forcelimited vibration （FLV） testing was developed and implemented at NASA Jet Propulsion Laboratory （JPL）, in the 1990s. In addition to controlling the input acceleration, the FLV testing measures and limits the reaction forces between the test article and the shaker. This approach can reduce the over-testing. However, it is very difficult to derive the force data of the interface during the flight, which is supposed to be used to define the force specifications of the FLV test. So it becomes very critical to find a way to get the force specifications with the acceleration data and the dynamic model of the spacecraft. This paper derived the expression of the relationship between the interface force and acceleration using component modal synthesis method with a dynamic model of the spacecraft system. Then according to the contribution of every mode of the system to the interface responses in a certain range of frequency, we reduce the dynamic model into a complex two-degree-of-freedom system which has already been introduced in many literatures. Through this reduction process we theoretically find how the Complex TDFS method works with its simplification model. And then we analyzed the error of the method with a numerical example.