Identification Of Material Microscale Prdperties And Damage Model Parameters By Inverse Analysis

Multi-functional filming/coating materials have been widely used in many fields, such as aerospace, mechanical engineering, electrical and electronic industry, architecture and so on. As an important branch of composite materials, coatings or films not only have the advantages of high strength, toughness, temperature resistance, corrosion resistance and oxidation resistance, but also can satisfy some specific functional requirements. Their macroscopic sizes are usually at micro or nano level, although which is far less than the substrate thickness, coatings play a key role in the performance and life of coating/substrate structures. Therefore, characterization of mechanical properties of micro-scale coating material is very important for its engineering application. Besides, cohesive zone model has been widely used to simulate the interfacial delamination and degumming as well as fatigue and fracture of non-interfacial materials. Acquirement of cohesive zone model parameters is directly related to the correct description of damage evolution, which is also required for a detailed research.As for the above two types of micro size and highly nonlinear problems, the traditional test methods of mechanics performance is rather difficult to characterize. In this paper, a procedure based on inverse analysis was used to extract mechanical properties of coatings in micro scale. This procedure utilizes indirect experimental records obtained from instrumented indentation, finite element solutions, and the numerical optimization algorithm to solve the above nonlinear problem. In order to improve the efficiency and reliability of convergence results, some improved algorithms and inverse analysis schemes were proposed from different aspects, which were used for obtaining satisfied inverse analysis results. The main work of this paper was summarized as follows.(1) To determine the elastic-plastic constitutive parameters of coating meterials in micro-nano scale, an improved inverse analysis procedure based on the extended kalman filter (EKF) technique was proposed. In this procedure, the load-indentation depth curves from indentation test were utilized, in combination with finite element simulation, to determine the material parameters. The load-indentation depth response surfaces were constructed by Lagrange interpolation function based on the numerical simulation results, in order to promptly obtain the resultant response from any set of material parameters in the prescribed domains. At the same time, another inverse analysis method, combining genetic algorithm with the response surface interpolation, was used to repeat the above inverse process for comparison of the technical applicability with Kalman filter based inverse method. Estimation results show that experimental results obtained from different sizes or different geometry indenters, together used as input data to carry out the inverse analysis, will efficiently improve the convergence, accuracy and well-posedness of solutions.(2) As for utilizing the cohesive zone model to characterize the interfacial damage problem, a "two step method" was proposed to extract the model parameters for known crack path problem, depending on the traction-displacment curves obtained from double cantilever beam model (DCB) tests. However, for a mixed-mode failure case with previously unknown crack path, the extended finite element method (XFEM) was introduced to simulate the crack propagation. A new inverse analysis method was proposed, by utilizing both experimental load-displacement curve and crack path information, to obtain the cohesive zone model parameters.(3) A multi-parameter identification procedure was developed, based on the sensitivity analysis. At the same time, a weight assignment scheme of objective functions, based on the dispersion degree concept, was proposed to deal with the multi-objective problem. With comparison to the experimental observation, the above referred inverse analysis method were identified being effective.(4) In view of the defects (e.g. complicated, containing huge information and operation steps) of inversion process, a set of automatic parameter inverse analysis system was developped, by using visual basic programme language. This system integrates a variety of inverse analysis methods, which can be widely used for the mechanical property identification of material, post-processing of simulation data, output of graphic and so on. This software system facilitates the inverse analysis, improves the analysis efficiency and can let us visualize the analysis process, which is an important tool for achieving the engineering application.