Search

Wearing helmets while riding motorcycles is compulsory in Vietnam since 2007. The goal of this study is to build finite element models of two of the most common type of helmet to be worn by Vietnamese. Based-on observation, half helmet makes up the majority followed by open face helmet. The viability of the two helmet is studied using finite element method. The open face helmet and the half helmet was designed by the Solidwork CAD tool and modeled by LS- Prepost with different thickness of the impact absorbing liner. Among with the modified V-dummy, these helmets have been conducted drop test following the European helmet standard ECE R22.05 at four impact position. The result has pointed out that the helmet models created a significant decrease of multiple injury index comparing with non-helmet model. Nevertheless, even with a thicker liner, the half helmet failed more tests than the open face helmet due to ECE 22-05. Moreover, the study also points out that traditional helmets have limited effect on rotational-based brain injury.

Besides cyclist and pedestrian, motorcyclist also is vulnerable road users (VRU) because of lack of safety factors. Considering the traffic situation at present of developing countries which represent the number of motorbikes, the dangerous traffic accidents regarding motorcyclist always higher than others. Particularly, motorbike, which main vehicle in Vietnam, combine with the increase of the number of cars has created many problems about the crash between car and motorcyclist and the level of user's injury. Hence, the main aim of this study is that researching how car crashes influence Vietnamese human body in a traffic accidence, in particular, human head because it is the most important part in the whole body. Human Body Model is Dummy Hybrid III, but it is scaled down to average size of Vietnamese male named V-Dummy while the finite element car models used in this research are already published. Furthermore, motorbike finite element model is completely built in new. Then, V-Dummy is changed the posture from pedestrian to driver on motorbike to conduct the crash simulation model in different cases (velocity and impact angle) via LS-Dyna software. The results are taken out via LS- prepost and processed to predict easier the head injury of motorcyclist. Based on the databases, the dummy head was frequently hit by the car's windshield during the primary impact and the HIC show the acceleration of the head corresponds to the speed of impactor. At 0° of impact angle, the motorcyclist suffers higher HIC than 30o of impact angle. The velocity of motorbike does not cause a significant effect at low velocities from 20 km/h to 40 km/h while there is a considerable higher HIC at 50 km/h of car's velocity when motorcyclist go at 30 km/h of velocity instead of 0 km/h of velocity.

In this research presents the vehicle body surface design and analysis of vehicle aerodynamic using flow simulation. The purpose of this paper is to develop a method for surface design of vehicle body model in SOLIDWORK software by using 2D real car images. Then this body structure is utilized for simulation and aerodynamic analysis using flow air. In this article, flow simulation/SOLIDWORK was used to analyze and to make the vehicle model to achieve more realistic simulation results. The model used in the simulation of this air flow has the dimensions as well as the mass, materials and properties that are almost equivalent to real life to evaluate the results of the air currents acting on the vehicle, at each other level of speed from 40km/h to 120 km/h. This paper is intended to provide some experience and familiarity with flow analysis and simulation.

Nowadays, the quality of classification systems depends on the presentation of the dataset, a process that takes time to use in-depth knowledge to produce specific characteristics. Meanwhile, deep learning can extract features from a dataset, without having to design feature extractors. Convolutional Neural Network (CNN) is a special type of deep learning that achieves many accomplishments in speech recognition, image recognition and classification. In this paper, we use CNN to classify and recognize breast cancer images from public BreakHis dataset. This dataset includes 7,909 breast cancer (BC) histopathology images with four benign subclasses and four malignant subclasses. Our new task with this dataset is the automated multi-classification of these breast cancer images in eight classes, which can help reducing death rates and saving people’s lives in the world. Our method based on the resizing original images for building CNN model and classifying breast cancer classes.