ObjectiveTo study the way of improving the validity of the aircraft structure ground test.MethodsThe finite element models of the specimen structure in flight and ground test conditions were calculated. Then the test model of the specimen structure in ground testing was measured to verify the validity of the finite element model. The vibration response of feature points in flight and ground testing was calculated. The mapping relation model of the feature points was calculated by machine learning methods. Then the response of each point in ground testing was determined by flight point response (or telemetry data) and their equivalent load was ob-tained by load reverse method. Finally, the load imposed on the test system was determined.ResultsWith a slender body structure as an example, the maximum relative error between the finite element model and the measured model of the test system was 6.76%. The vi-bration response of the feature points in flight state could be predicted via mapping relation model. The vibration response of the cor- responding ground testing points could be determined. The load to be applied was set at 60 N by reverse deliberation according to re-sponse points of ground test system.ConclusionBy the unified method of the flight-ground numerical calculation and the ground test-ing, the load to be applied in the ground test can be determined without deliberately simulating the boundary conditions in flight state, which improves the validity of the aircraft ground testing.%目的：研究提高飞行器结构地面试验有效性的途径。方法计算同一被试件结构在飞行状态和地面试验状态下的有限元模型，测量地面试验状态下的模态以验证有限元模型的正确性；计算各特征点（也可以是遥测点）在天地状态下的响应，用机器学习法获取各特征点的映射关系模型；基于该模型由飞行点响应（或遥测数据）确定出地面试验件对应点的响应，并用载荷反求法得到它们的等效载荷；最终确定施加在试验系统上的载荷。结果以细长体结构为例，所得到由其组成试验系统的有限元模型与实测模型之间的固有频率最大相对误差为6.76%，利用映射关系模型预测出对应点在飞行状态下的振动响应。确定了飞行状态下结构响应的特征点，由地面试验系统所对应的响应点反推出应施加的载荷为60 N。结论利用天地数值计算-地面试验验证联合法，无需在地面试验状态下刻意模拟飞行状态的边界条件，确定出所需要施加的载荷，从而提高了飞行器地面试验的有效性。
展开▼
