摘要：
自由振动试验法(FreeV法)用于平行分离双幅桥气动导数识别时,由于人工初始激励不同,节段模型产生的气动干扰不同,进而导致气动导数识别结果规律性差,识别数据离散性大.为此,提出了混合强迫振动试验法(MFV法),并对比分析了MFV法和FreeV法识别出的气动导数,以及基于MFV法和FreeV法的二维颤振分析结果与节段模型测试结果.研究结果表明:无论是上游桥还是下游桥,MFV法识别的气动导数均离散性较小,随折算风速的变化规律明确;而FreeV法识别的气动导数离散性较大,随折算风速的变化规律较差,尤其是受气动干扰效应影响较大的下游桥;基于MFV法预测的颤振临界风速与节段模型测试结果较为接近,而基于FreeV法预测的颤振临界风速与节段模型测试结果的偏差较大,尤其是下游桥,偏差更为显著;相比于FreeV法,MFV法能更为准确地模拟平行分离双幅桥的气动干扰效应,其颤振预测结果更准确.%When identifying the aerodynamic derivatives of parallel separated double-deck bridges using free-vibration testing method (FreeV method),the different initial manual excitations produce different aerodynamic interference between two sectional models,which cause the identification results are always significant discrete and irregular.Therefore,the mixed forced-vibration testing method (MFV method) is proposed in this article.The aerodynamic derivatives identified by FreeV method and MFV method are compared and analysed.The flutter analysis results based on the FreeV and MFV method are also in contrast with the test results.Research results show that,no matter the windward deck or the leeward deck,the aerodynamic derivatives identified by MFV method put up to small dispersions and good laws of variation with reduced wind speed.However,the aerodynamic derivatives obtained by FreeV method show larger dispersions and worse variation laws,especially those of the leeward deck which suffered from aerodynamic interference seriously.Furthermore,the flutter critical wind speeds calculated based on the MFV method are rather close to the test results,whereas those based on the FreeV method have a big discrepancies with the test results,especially those of leeward bridge.In contrast with the FreeV method,the MFV methodcould simulate the aerodynamic interference effect rationally and the flutter prediction results are more accurate.
