Taewung Kim

A Novel Algorithm for Crash Detection Under General Road Scenes Using Crash Probabilities and an Interactive Multiple Model Particle Filter

Driver inattention causes the majority of vehicular crashes, and these accidents produce extensive economic and social costs, as well as injuries and fatalities. Thus, the development of imminent crash detection systems is one of the most important issues in automotive safety. Various crash detection algorithms have been proposed, but the coverage of these algorithms has been limited to one or two crash scenarios. To widen the coverage of crash detection systems to include various crash modes, driver behaviors that are dependent on road scenes and vehicle dynamics should be considered. This paper proposed an algorithm for detecting an imminent collision in general road scenes. The proposed algorithm consists of crash probability data generated from Monte Carlo simulations that consider driver behavior and vehicle dynamics, a tracking algorithm that uses an interactive multiple-model particle filter, and a threat assessment algorithm that estimates crash probabilities. To reduce nuisance and false-positive alarms, the algorithm discriminated between normal and dangerous road scenes, and a point of no return was detected using three driver models that addressed different levels of driver input. The performance of the proposed algorithm was evaluated under three scenarios, and it successfully discriminated between collision and near-miss cases, and it adjusted warning times depending on the road scenes. It is expected that the proposed algorithm would have good driver acceptability based on the results of the near-miss cases. The proposed algorithm can be used as an integrated crash detection algorithm for crash warning, avoidance, and mitigation purposes while incorporating tracking information from multiple sources.

Biofidelity evaluation of WorldSID and ES-2re under side impact conditions with and without airbag

This study evaluated the biofidelity of the WorldSID and the ES-2re under whole-body side impact conditions with and without a side airbag using the biomechanical cadaveric response data generated from 4.3m/s whole-body side impact tests. Impact forces, spinal kinematics, and chest deflections were considered in the biofidelity evaluation. Average responses and response corridors of PMHS were created using a time-alignment technique to reduce variability of the PMHS responses while maintaining the sum of the time shifts to be zero for each response. Biofidelity of the two dummies was compared using a correlation and analysis (CORA) method. The WorldSID demonstrated better biofidelity than the ES-2re in terms of CORA ratings in the conditions with airbag (0.53 vs. 0.46) and without an airbag (0.57 vs. 0.49). Lastly, the kinematic analysis of the two dummies indicated an overly compliant shoulder response of the WorldSID and excessive forward rotation of the ES-2re relative to the PMHS.

Validation of Shoulder Response of Human Body Finite-Element Model (GHBMC) Under Whole Body Lateral Impact Condition

In previous shoulder impact studies, the 50th-percentile male GHBMC human body finite-element model was shown to have good biofidelity regarding impact force, but under-predicted shoulder deflection by 80% compared to those observed in the experiment. The goal of this study was to validate the response of the GHBMC M50 model by focusing on three-dimensional shoulder kinematics under a whole-body lateral impact condition. Five modifications, focused on material properties and modeling techniques, were introduced into the model and a supplementary sensitivity analysis was done to determine the influence of each modification to the biomechanical response of the body. The modified model predicted substantially improved shoulder response and peak shoulder deflection within 10% of the observed experimental data, and showed good correlation in the scapula kinematics on sagittal and transverse planes. The improvement in the biofidelity of the shoulder region was mainly due to the modifications of material properties of muscle, the acromioclavicular joint, and the attachment region between the pectoralis major and ribs. Predictions of rib fracture and chest deflection were also improved because of these modifications.

Prediction of the structural response of the femoral shaft under dynamic loading using subject-specific finite element models

Abstract The goal of this study was to predict the structural response of the femoral shaft under dynamic loading conditions using subject-specific finite element (SS-FE) models and to evaluate the prediction accuracy of the models in relation to the model complexity. In total, SS-FE models of 31 femur specimens were developed. Using those models, dynamic three-point bending and combined loading tests (bending with four different levels of axial compression) of bare femurs were simulated, and the prediction capabilities of five different levels of model complexity were evaluated based on the impact force time histories: baseline, mass-based scaled, structure-based scaled, geometric SS-FE, and heterogenized SS-FE models. Among the five levels of model complexity, the geometric SS-FE and the heterogenized SS-FE models showed statistically significant improvement on response prediction capability compared to the other model formulations whereas the difference between two SS-FE models was negligible. This result indicated the geometric SS-FE models, containing detailed geometric information from CT images with homogeneous linear isotropic elastic material properties, would be an optimal model complexity for prediction of structural response of the femoral shafts under the dynamic loading conditions. The average and the standard deviation of the RMS errors of the geometric SS-FE models for all the 31 cases was 0.46 kN and 0.66 kN, respectively. This study highlights the contribution of geometric variability on the structural response variation of the femoral shafts subjected to dynamic loading condition and the potential of geometric SS-FE models to capture the structural response variation of the femoral shafts.

Evaluation of biofidelity of THUMS pedestrian model under a whole-body impact conditions with a generic sedan buck

ABSTRACT Objective: The goal of this study was to evaluate the biofidelity of the Total Human Model for Safety (THUMS; Ver. 4.01) pedestrian finite element models (PFEM) in a whole-body pedestrian impact condition using a well-characterized generic pedestrian buck model. Methods: The biofidelity of THUMS PFEM was evaluated with respect to data from 3 full-scale postmortem human subject (PMHS) pedestrian impact tests, in which a pedestrian buck laterally struck the subjects using a pedestrian buck at 40 km/h. The pedestrian model was scaled to match the anthropometry of the target subjects and then positioned to match the pre-impact postures of the target subjects based on the 3-dimensional motion tracking data obtained during the experiments. An objective rating method was employed to quantitatively evaluate the correlation between the responses of the models and the PMHS. Injuries in the models were predicted both probabilistically and deterministically using empirical injury risk functions and strain measures, respectively, and compared with those of the target PMHS. Results: In general, the model exhibited biofidelic kinematic responses (in the Y–Z plane) regarding trajectories (International Organization for Standardization [ISO] ratings: Y = 0.90 ± 0.11, Z = 0.89 ± 0.09), linear resultant velocities (ISO ratings: 0.83 ± 0.07), accelerations (ISO ratings: Y = 0.58 ± 0.11, Z = 0.52 ± 0.12), and angular velocities (ISO ratings: X = 0.48 ± 0.13) but exhibited stiffer leg responses and delayed head responses compared to those of the PMHS. This indicates potential biofidelity issues with the PFEM for regions below the knee and in the neck. The model also demonstrated comparable reaction forces at the buck front-end regions to those from the PMHS tests. The PFEM generally predicted the injuries that the PMHS sustained but overestimated injuries in the ankle and leg regions. Conclusions: Based on the data considered, the THUMS PFEM was considered to be biofidelic for this pedestrian impact condition and vehicle. Given the capability of the model to reproduce biomechanical responses, it shows potential as a valuable tool for developing novel pedestrian safety systems.

Improvement of lateral shoulder impact response of a multi-body pedestrian model

ABSTRACT The interaction of the shoulder complex of the pedestrian and the striking vehicle strongly influences the responses and injury risk of the head, neck and torso during vehicle-to-pedestrian impacts. While the current MADYMO facet pedestrian model met the shoulder force corridor provided by ISO9790, the kinematics of its shoulder complex during a lateral blunt impact has not been evaluated. In this study, this model was evaluated relative to more detailed and newly available cadaveric responses under lateral shoulder impact, and exhibited much higher shoulder impact force and displacement. To improve the biofidelity of the shoulder complex, a component level validation was performed on its upper arm model based on component-level upper arm compression test data by maximising ISO rating scores between the cadaveric and the model responses. After the improvement of the arm model, the updated pedestrian model showed improved biofidelity based on ISO rating scores on shoulder impact force, displacement and shoulder deflection under lateral shoulder impact conditions. Finally, under a 20 km/h sedan-to-pedestrian lateral impact, the 15% higher head relative impact velocity was observed for the updated pedestrian model, which demonstrated the strong influence of the shoulder complex on the pedestrian head response.

Predicting Pedestrian Injury Metrics Based on Vehicle Front-End Design

Simulations provide vehicle designers with the capability to evaluate the safety of their designs in a wide variety of scenarios. However, the high-fidelity simulations required for safety assessment carry significant computational costs. As such, the engineering team must carefully select automotive designs to simulate, and use the results obtained to accurately predict the performance of new designs over a wide range of metrics. This paper describes the modeling of automotive simulation outputs to accurately predict a large number of widely used pedestrian injury metrics given the vehicle front-end design. The models in this paper allow the vehicle designer to identify and focus on the variables that most affect the different injury metrics, and determine which variables are most important to the overall safety performance of the vehicle.

Pelvic restraint cushion sled test evaluation of pelvic forward motion

ABSTRACT Objective: This study was designed to evaluate the performance of a pelvic restraint cushion (PRC), a submarining countermeasure that deploys under the thighs when a crash is detected in order to block the forward motion of the pelvis. Methods: Sled tests approximating low- and high-speed frontal impacts were conducted with 4 female postmortem human subjects (PMHS) restrained by a lap and shoulder belt in the right front passenger seat. The subjects were tested with and without a PRC. Results: The PRC is effective in reducing forward motion of the PMHS pelvis and reduces the risk of injury due to lap belt loading in a high-speed frontal crash. Conclusions: Although small sample size limits the utility of the studys findings, the results suggest that the PRC can limit pelvic forward motion and that pelvic injury due to PRC deployment is not likely.

Injury risk functions based on population-based finite element model responses: Application to femurs under dynamic three-point bending

ABSTRACT Objective: The goal of this study was to explore a framework for developing injury risk functions (IRFs) in a bottom-up approach based on responses of parametrically variable finite element (FE) models representing exemplar populations. Methods: First, a parametric femur modeling tool was developed and validated using a subject-specific (SS)-FE modeling approach. Second, principal component analysis and regression were used to identify parametric geometric descriptors of the human femur and the distribution of those factors for 3 target occupant sizes (5th, 50th, and 95th percentile males). Third, distributions of material parameters of cortical bone were obtained from the literature for 3 target occupant ages (25, 50, and 75 years) using regression analysis. A Monte Carlo method was then implemented to generate populations of FE models of the femur for target occupants, using a parametric femur modeling tool. Simulations were conducted with each of these models under 3-point dynamic bending. Finally, model-based IRFs were developed using logistic regression analysis, based on the moment at fracture observed in the FE simulation. In total, 100 femur FE models incorporating the variation in the population of interest were generated, and 500,000 moments at fracture were observed (applying 5,000 ultimate strains for each synthesized 100 femur FE models) for each target occupant characteristics. Results: Using the proposed framework on this study, the model-based IRFs for 3 target male occupant sizes (5th, 50th, and 95th percentiles) and ages (25, 50, and 75 years) were developed. The model-based IRF was located in the 95% confidence interval of the test-based IRF for the range of 15 to 70% injury risks. The 95% confidence interval of the developed IRF was almost in line with the mean curve due to a large number of data points. Conclusions: The framework proposed in this study would be beneficial for developing the IRFs in a bottom-up manner, whose range of variabilities is informed by the population-based FE model responses. Specifically, this method mitigates the uncertainties in applying empirical scaling and may improve IRF fidelity when a limited number of experimental specimens are available.

Identification of characteristics and frequent scenarios of single-vehicle rollover crashes during pre-ballistic phase; part 1 – A descriptive study

This study aimed to identify common patterns of pre-ballistic vehicle kinematics and roadway characteristics of real-world rollover crashes. Rollover crashes that were enrolled in the National Automotive Sampling System-Crashworthiness Data System (NASS-CDS) between the years 2000 and 2010 were analyzed. A descriptive analysis was performed to understand the characteristics of the pre-ballistic phase. Also, a frequency based pattern analysis was performed using a selection of NASS-CDS variables describing the pre-ballistic vehicle kinematics and roadway characteristics to rank common pathways of rollover crashes. Most case vehicles departed the road due to a loss of control/traction (LOC) (61%). The road departure with LOC was found to be 13.4 times more likely to occur with slippery road conditions compared to dry conditions. The vehicle was typically laterally skidding with yawing prior to a rollover (66%). Most case vehicles tripped over (82%) mostly at roadside/median (69%). The tripping force was applied to the wheels/tires (82%) from the ground (79%). The combination of these six most frequent attributes resulted in the most common scenario, which accounted for 26% of the entire cases. Large proportion of road departure with LOC (61%) implies electronic stability control (ESC) systems being an effective countermeasure for preventing single-vehicle rollover crashes. Furthermore, the correlation between the road departure with LOC and the reduced friction limit suggests the necessity of the performance evaluation of ESC under compromised road surface condition.