Webex: Simulation of Variability in the Hybrid 3 Crash Test Dummy - 25 August
Crash test dummies are the primary measurement device for all legislative and consumer crash testing around the world. The current dummy design, Hybrid 3, has not changed significantly since the 1970s, and the certification requirements, which each dummy has to comply with, were aligned to loading regimes relevant at that time. With modern restraint systems, in particular advanced airbag and seat belts, these certification procedures are less representative of current dummy loadings in crash tests. Additionally, the width of the certification compliance corridors allows significant variability being observed in certification test results, and it is not clear where this comes from, or how relevant it is to real crash test results. As they are the only method for demonstrating legal compliance, and bearing in mind the effort and cost involved in designing vehicles to be safe, it is essential that dummies can be adequately characterised, and that this information is fed into the design process.
The dummies are very complex systems, and detailed F.E. models have been available for more than 10 years from, amongst others, FTSS, a manufacturer of physical and virtual dummies. During this piece of work, carried out in conjunction with FTSS, the existing Hybrid 3 / 50% frontal crash DYNA dummy model was modified to include a set of input parameters, allowing the stiffness properties of the model to be varied. This enabled the crash injury values output by the dummy to be studied, as a function of dummy stiffness characteristics. By building a model of the certification test rig, a population of dummy models complying with the certification requirements was generated, using a DOE-based parameter definition. This population was introduced into vehicle crash models, and since each dummy is valid with respect to the certification test, the spread of results in the vehicle crash models was also considered to be valid. The relationship between output in the certification test, and that from the vehicle tests allowed conclusions to be drawn as to the relevance of the certification tests, and their usefulness in characterising the dummies and validating the dummy models. Further, we determined the spread of results to be expected from nominally identical vehicle tests, as a result of using different, valid, dummies. The dummy set created provides a way of assessing the robustness of a given restraint system to this source of noise, and hence reducing the sensitivity of the design to it. A further aim is to allow the results of a specific certification test to be used to derive input parameters that match the model to the physical dummy used. This will help in the interpretation of test results, and increase confidence in decisions made on the basis of the results.
The dummy model is being used for research purposes at JLR and FTSS, and is also providing useful information for potential improvements in the model. In future it could also point at areas of the physical dummy that are heavily involved in generating the variability, thereby focussing efforts to improve the manufacturing process.
Add value to your product or company by attending this FREE webinar.
Agenda
Welcome & Introduction
Matthew Ladzinski, NAFEMS North America
Simulation of Variability in the Hybrid 3 Crash Test Dummy
Richard Brown, Jaguar Land Rover
Q & A Session