vehicle model representing a vehicle class or a non-existing car. Several ... time is exceeding the limit to perform a large set of simulations within one night,.
ICrash 2006, Athens Greece, 4th-7th July 2006
Development of a Generic Vehicle Modelling Procedure R R P Rodenburg*, C D van der Zweep*, G Kellendonk*, H G Mooi*, W J Witteman** *
TNO Science and Industry, Business Unit Automotive, Safety Department, Schoemakerstraat 97, Delft, The Netherlands **
Eindhoven University of Technology, Den Dolech 2, 5612 AZ, Eindhoven, The Netherlands
Abstract: Fleet wide vehicle crash research requires a wide range of numerical vehicle models to simulate the crash behaviour of a real life vehicle fleet in crashes. Representative simulations of a vehicle fleet may require 10,000 till 100,000 crash simulations. This requires relatively simple vehicle crash models in order to obtain acceptable CPU times, in the order of minutes. The current used multi body vehicle models have a relative short CPU time (one hour) with a good representation of the frontal kinematics and global dummy responses including airbag and belt interaction, however there is a need for even less time consuming models with an acceptable accuracy for (frontal) impact. This paper describes a method to construct simple vehicle (frontal) crash models by an automated process with the use of vehicle measurement and test data which are available for other research purposes. Structural elements are selected which determine the crash behaviour of the vehicle. An algorithm is created which uses a geometry database to build specific vehicle types. These vehicle models are fitted and validated using crash test data. A midsize European vehicle model is built using this new methodology which shows promising results for usage in different crash research activities.
NOTATION DoE DoF FE FWDB
Design of Experiments Degree if Freedom Finite Element Full Width Deformable Barrier
GCM MB ODB PDB
Generic Car Model Multibody Offset Deformable Barrier Progressive Deformable Barrier
INTRODUCTION With ever increasing requirements regarding vehicle performance and shortening of design cycles, the role of computer simulation techniques increased rapidly over the last decade. Computer simulations allow researchers to augment test data by simulating crashes over a wide range of conditions. Within most crashworthiness research, simulations tend to focus more on the characteristics of a vehicle model or class in a vehicle fleet. The objective of this study is to develop accurate vehicle models with a limited calculation time to estimate the consequences of regulatory changes on the frequency and distribution of injury levels within a fleet. Various tools are available for fleet wide crash research, namely lumped massspring or multibody (MB) and finite element (FE) vehicle models. The type of modelling techniques used depends on the purpose of the models [1,2]. Commonly used FE vehicle models typically have CPU times in the order of 24 hours and MB vehicle models in the order of one hour [6, 13]. The current models, e.g. FE models of car manufactures, are created based on specific car types, where in research studies and in the conceptual phase a need is for a vehicle model representing a vehicle class or a non-existing car. Several frontal car models were developed [5, 6] to study general trends in car to car impacts, however these models were specific for a car design. To be able to predict behaviour of a larger range of vehicle models, generic vehicle models are developed. Generic in these models is the structure of the front-end and the characteristics of these structures applied and furthermore, the overall measures of the vehicles are average values of the vehicle class they belong to. The level of detail in a generic model needs to be as generic as possible and therefore the amount of information that can be withdrawn from the model will reduce. The most relevant parameters (crash characteristics) are specified and based on research and engineering knowledge. The aim is to achieve an accurate vehicle model with fast calculation time.
ICrash 2006, Athens Greece, 4th-7th July 2006
National Highway Traffic Safety Administration (NHTSA) [1] proposed to develop vehicle models of late model year vehicles, for each of the five following weight categories in the US:
