Human Body Model and Passive Safety of Vehicles: Analysis of biomechanical results and study of injuries on abdominal organs.

Car accidents are frequent in the world and, consequently, safety in the car becomes an important requirement and one of the more requested functionalities by people when buying a car. For this reason, crash safety tests were born to understand the security level of the vehicle, to discover which improvements can be introduced to increase security. The principal problems of crash tests are their price, their inability to represent many people and their inaccuracy to evaluate the types of injuries. For these reasons, especially the last two, companies have begun to look to make safety systems more accurate than previous. Consequently, the application of a crash test virtual model based on the Finite Element Method (FEM) was introduced.  In this model, the car test is modelled in FE using HBM (Human Body Computational) to simulate human behaviour. The human model used in this thesis is the THUMS (Total Human Model for Safety). It helps to understand how the human body reacts during the impacts and to better predict the types and location of injuries with respect to classic rigid dummies tests.   Even if THUMS is a modern tool, it can be still improved. The purpose of this thesis is to modify it to have a more realistic one. In this way, it is possible to have a model which simulates better the properties and the behaviour of the human body. Before making these changes, it is important to understand if the original model is reliable or not. For this reason, the work is tested with corridors, experimental data, obtained by cadavers. These curves are compared against the model’s answer to see if it is acceptable or not. Studying this comparison, a different behaviour between cadavers and human model is noted both in literature and in our study. This distinction appears mostly in the thorax. To improve this aspect, the ribs’ cortical bone is analysed. Based on the literature, its thickness and material are modified. In particular, it is given a greater variability in the thickness, from a uniform value in all ribs to a more specific value for some parts of the ribs. For the material, instead, new parameters are proposed to simulate it. To discover how these changes can modify the thorax’s answer in the model, after every change, the validation tests are re-made.   Once the changes have given acceptable results, the original THUMS is modified and a new one is obtained. This new model is used in the virtual crash test to recreate a real accident and to analyse the THUMS’s behaviour. At this point, based on the model’s answer, an injury criterion is obtained: the Peak Virtual Power method (PVP). Through the PVP method, the Abbreviated Injury Scale level (AIS) is calculated. This value is then compared to the real one obtained in a real accident, to see if the PVP can be used or not as a new criterion for two abdominal inner organs (liver and spleen).