Promoting excellence in mobility engineering

  1. FISITA Store
  2. Technical Papers

A Study on the Crash Performance Prediction Method for Dummy Injury
FISITA2016/F2016-APSC-004

Authors

Sangyoon Hwang*, Wook Jin*, Sunil Yeom*, Byeongdong Youn**, Taejin Kim**, Boseong Seo**

* Hyundai Motor Company, Hwaseong-Si Gyeonggi-Do. The Republic of Korea

** Seoul National University, Seoul. The Republic of Korea

Abstract

Research and/or Engineering Questions/Objective

Since 2010, The new car assessment program so-called NCAP has been applied to the passenger car manufactured in US. Among of injury criteria, chest deflection in 35mph Full-frontal impact is one of the most difficult injuries to predict and improve. Until now, to estimate the chest deflection, we depended on the computer simulation method such as multi-body system program. The objective of this study was to develop chest deflection prediction model by using only crash test data.

Methodology

We chose shoulder belt load and vehicle crash deceleration as the most effective 2 input factors to chest deflection(output) through paper study and engineering judgment. Cross-correlation method is a measure of similarity and time lag of two continuous functions. By using this method, cross-correlation function was calculated for each test data which is carried out under 35mph full-frontal new NCAP protocol. We found out the similarity of those functions and determined the representative cross-correlation function by normalization and averaging. Since 2 input factors were given, we obtained 2 different representative cross-correlation functions and applied inverse cross-correlation method. Through that, we could predict chest deflection with respect to new vehicle having different shoulder belt load and vehicle crash deceleration. In order to improve the accuracy of this model, the hybrid model was suggested by using weighted factor.

Results

We applied cross-correlation method to 52 vehicles tested in NHTSA from 2011 to 2013. Average accuracy of estimated chest deflection compared with test data was 85% by using the hybrid model. If it applied to only mid-size family sedan (21 vehicles), average accuracy was 90%.

Limitations of this study

An important limitation of this study is that only 2 input factors were used despite of a lot of underlying factors affecting the chest deflection. Furthermore, since normalized representative cross-correlation function was used by averaging, to predict chest deflection, we should earn maximum value of representative cross-correlation function by regression method. Therefore, this make chest deflection model having difference with test data.

What does the paper offer that is new in the field including in comparison to other work by the authors?

The study about chest deflection by using cross-correlation method in this paper is the first approach to predict chest injury in the vehicle safety field.

Conclusions

New chest injury prediction model has been developed by using cross-correlation method and it show relatively good accuracy for NHTSA 35mph full-frontal crash test. If this model is applied at the vehicle development stage, we can save time and effort to estimate chest deflection.

Key Words : NCAP 35mph Frontal, Injury Criteria, Chest Deflection, Belt Load, Crash Deceleration, Cross-Correlation

Add to basket