Quantitative Identification Of Three-dimensional Subsurface Defect Based On The Fuzzy Inference Of Thermal Process And Experimental Study

Defects that exist in the industrial equipment itself or formed during operation may cause major equipment failures or even safety accidents.The subsurface defects of the device have strong concealment,which increases the difficulty of defect detection and the probability of missed detection.The in-depth study of quantitative identification methods for subsurface defects of equipment has important scientific and engineering significance.In some occasions,via the device's own heat or external thermal excitation,infrared thermal image of the surface of the device can be used to reflect the subsurface defect information,and the identification of defect information can be transformed into a kind of inverse geometry heat transfer problems.Exploring effective inverse methods is the main bottleneck to the identification of subsurface defects in equipment based on heat transfer processes.In this paper,the three-dimensional quantitative identification problem of subsurface defects of equipment based on inverse heat transfer method is studied.The related work includes two basic aspects: the three-dimensional quantitative inverse method of equipment subsurface defects and the quantitative identification experiment of equipment subsurface defects.The main research contents and results are as follows:(1)The physical and mathematical model of heat transfer process of equipment with subsurface defects is established,and the steady temperature field of equipment which has subsurface defects is simulated by the finite element method.The effects of subsurface defect shape,subsurface defect size,subsurface defect location and defect media characteristics on the observed surface temperature field are discussed by numerical simulation experiments.These analyses of the effects of subsurface defect information on the surface temperature field lay the necessary foundation for the quantitative identification of subsurface defects based on the inverse heat transfer method.(2)Identifying subsurface defects of equipment based on surface temperature field is a typical inverse heat transfer problem of distributed parameters.Aiming at the basic characteristics of the inversion problem,a decentralized fuzzy inference(DFI)method for quantitative identification of subsurface defects is proposed in this paper.In the DFI method,according to the deviation field between the surface temperature's measurement information and the calculation information,the fuzzy inference components can be obtained through a set of corresponding decentralized fuzzy inference units.Temperature sensitivity matrix for defects characteristic parameters is established under the guidance of the device heat transfer model.Then the matrix synthesizes the fuzzy inference components,and obtains the compensation amount of the defect parameters to realize the estimation from the surface temperature field to the subsurface defect information of the device.(3)The DFI method described above has been applied to study the quantitative identification of subsurface defect parameters of the devices.The influence of these parameters on the identification results,such as the shape and size of the defect,the initial guess value of the defect parameter,the temperature measurement error,the number of temperature measurement points,the thermal properties of the internal medium of the defect and the external thermal boundary conditions of the test piece,are analyzed by numerical experiments,and compared with identification results of the Levenberg-Marquardt method(L-MM).The results show that,the DFI method can reduce the dependence on the initial guess value and the number of measurement points,enhance the anti-interference ability to measurement error,have higher computational efficiency and higher identification accuracy.In addition,the corresponding coupled defect decentralized fuzzy inference system is established for the qualitative identification problem of coupled defect parameters.And the DFI method has been proven to be effective for quantitative identification of coupled defect parameters by numerical simulation test.(4)Combined with experiments,the problem of quantitative identification of subsurface defects of CNG cylinder composites based on inverse heat transfer method has been studied.Aiming at the key problems in the detection of defects in CNG cylinders for vehicles,a defect detection scheme based on the surface thermal image during the steam washing process inside the cylinder is proposed.An infrared detection experiment system for subsurface defects of gas cylinders was built to obtain the infrared thermal image of the cylinder surface corresponding to defects of different sizes and depths,which provide a basis for the identification of subsurface defect parameters of gas cylinder composites by DFI method.(5)During the internal steam flushing process of CNG cylinders,the inner wall heat flux distribution is the key information for the identification of subsurface defect parameters of gas cylinder composites.And the heat flux distribution on the inner wall surface is difficult to directly measure during the detection of gas cylinder defects,so a decentralized fuzzy inference system is established.The system realizes the inversion of the heat flux distribution on the inner wall surface of the gas cylinder according to the temperature measurement information of the normal area of the outer surface of the gas cylinder,and lays a foundation for the identification of the subsurface defect parameters of the gas cylinder composite material.(6)According to the infrared thermal image of the surface of the CNG cylinder,the quantitative identification of the subsurface defects of the CNG cylinder composite material was realized by using the DFI method,and the influence of the defect size,defect depth,initial guess value and the number of temperature measurement points on the identification result was discussed,and the inversion results of the defect parameters were compared with their actual values.The results show that the DFI method can effectively identify the subsurface defect parameters of the gas cylinder composite based on the infrared thermal image of the cylinder surface,which shows a new way for the quantitative detection of subsurface defects of CNG cylinder composites.