2022 年 88 巻 913 号 p. 22-00178
Thermoplastic resins, which can be used for matrix materials in fiber reinforced thermoplastics (FRTP), have distinctive nonlinear characteristics that depend on temperature and strain rate compared to conventional materials such as metals and thermosets. Therefore, in order to use FRTP as a structural material, it is necessary to properly measure its material behavior, adopt an appropriate constitutive law, and accurately identify the material parameters. Conventionally, although there have been reports on the nonlinear behavior of thermoplastic resins, most of them employ a single constitutive law such as plasticity or creep using data measured under specific conditions. In this study, adopting a viscoelastic-viscoplastic-damage constitutive law to represent the material behavior of thermoplastic resin, we propose a hybrid identification method that can determine the material properties of by the combined use of the results of dynamic mechanical analysis and uniaxial cyclic loading-unloading test. Specifically, material parameters for viscoelasticity are identified from the complex moduli of elasticity obtained from the dynamic mechanical analysis, and those for viscoplasticity as well as the damage and time-temperature shift factor in low temperature range are identified from the uniaxial cyclic loading-unloading test. Differential Evolution, one of the optimization algorithm, is adopted to identify the parameters. Conducting some material tests, we confirm the effectiveness of the proposed method in reproducing the nonlinear material behavior of three typical thermoplastic resins: polycarbonate, acrylic resin and thermoplastic epoxy resin. In addition, the identification accuracy is significantly improved by the modification of the time-temperature shift factor, which has been determined by dynamic mechanical analysis, using the results of uniaxial cyclic loading-unloading tests in the low temperature range.
