Prediction and Validation of Unknown GISSMO Properties Using by Regression Analysis of Experimental Damage Tests

2019-01-1426

03/25/2019

Event
Asia-Pacific Automotive Engineering Conference
Authors Abstract
Content
A seat, as one of the main automobile components is closely related to the passenger safety. It plays an important role on a protection of passengers from a sudden movement of a car and external front/rear/side crash. In order to achieve these purposes, strength, rigidity and durability of a seat have to be satisfied which are regulated by the law. Therefore, a prediction of the fracture time and the behavior of the seat structure by Finite Element analysis are very important. However, the fracture prediction method from axial tensile strength test has limits to present the behavior which is obtained in multiple loading cases. For this reason, a new analysis method for the fracture prediction considering multiple loading cases has to be established.
In this research, the phenomenon of a seat fracture is implemented and the simulation is performed to predict the behavior using GISSMO Damage model in LS-DYNA. In order to find a fracture strain in a diverse stress mode, deformation tests of shear into a multiple direction, bi-axial tensile tests, notch impact test and plane deformation tests were performed. By introducing DIC(Digital Image correlation), fracture local strain is measured and at the same time, parameters for GISSMO Damage Model which are suitable to the force-displacement curve measured by DIC are calculated by the Reverse Engineering method using LS-OPT. In addition, according to the analysis of the correlation of strain-triaxial stress by materials and thickness, the formula is derived that predicts the fracture time of material in which its fracture property is unknown. In order to validate the fomular by the correlation analysis, strength test of a single rail pendulum is performed and by comparing simulation and experiment results, reliability of the product is obtained.
Meta TagsDetails
DOI
https://doi.org/10.4271/2019-01-1426
Citation
Jun, T., "Prediction and Validation of Unknown GISSMO Properties Using by Regression Analysis of Experimental Damage Tests," SAE Technical Paper 2019-01-1426, 2019, https://doi.org/10.4271/2019-01-1426.
Additional Details
Publisher
Published
Mar 25, 2019
Product Code
2019-01-1426
Content Type
Technical Paper
Language
English