The Analysis And Experiments Of Dynamic Data Driven Method For Thermal Protection System

During the service process of the thermal protection system,unreasonable results are often obtained by the calculation model due to the uncertainty of the materials and structures.The reason is that the calculation model is not competent to reflect the actual physical behavior of structures.Some simulation conditions such as the initial and boundary conditions and loads can only be obtai ned accurately until the system is actually in service.At present,the simulation and the experiment are carried in quiescent condition and serialization and have not been effectively combined for the research of thermal protection.Dynamic data driven me thod which can update the structure state and participate in task decision making by dynamic data connection between simulation and actual test system and On-line simulation and evaluation of dynamic test data is a promising trend to achieve breakthroughs in thermal protection level.In this paper,the concept of Dynamic Data Driven Application System(DDDAS)is introduced to the analysis and evaluation of thermal protection system.The typical thermal protection system is used as the research object.The framework design and implementation mechanism of dynamic data driven between simulation and test is explored,and the method of on-line analysis and evaluation under high temperature environment is carried out.The significance of the DDDAS theory and broad application prospects reveal the influence of dynamic data driven introduction on thermal response of structures.First of all,a framework of dynamic data driven application system were designed in this paper.An overall data-driven logical framework of "sensor data-online model-test system" was constructed;An on-line fast calculation model was established to address the problem of high-temperature nonlinearity of material properties.The Kirchhoff transform was used to transform it into a linear probl em.Numerical calculations were performed using the implicit finite difference method and the validity of the model was verified;Build an overall test platform and design control algorithms for system operation.Secondly,experimental verification and accuracy analysis of dynamic data-driven analysis methods for structural thermal response are carried out.The thermal response tests under different load conditions were carried out,and the effects of grid model accuracy,number of sensors,data collection i ntervals,and forecast time length on the accuracy of the forecast were analyzed.The applicability of the model to different heat flow loads is verified,embodying the advantages of thermal response analysis and forecast accuracy in the dynamic data-driven model.Finally,the thermal response test of the insulation tile under the two types of accidental overload and cyclic load conditions were carried out,to achieve the online decision and feedback control of the system in the test process.For the former,the system is able to control the measurement behavior through the forecast results including change the number of sensors,interval of data collection and model grid accuracy.For the last,the test load is controlled when the prediction result exceeds the safe range,the load is reduced in advance to ensure the structural reliability.The efficient and stable dynamic data-driven implementation mechanism is adjusted and explored by the algorithm improvement and adjustment of data transmission strategy.