Search Results

This paper presents the analysis of a series of frontal crash tests conducted by the Insurance Institute of Highway Safety that are commonly referred to as Small Overlap Impacts (SOI). The occurrence and severity of such frontal impacts in the real world were estimated using two different methods. Both methods used the National Automotive Sampling Scheme (NASS), which is a stratified sample of crashes in the US. The first method utilized an algorithm commonly known as Frontal Impact Taxonomy (FIT). The second method was based on comparison of deformation patterns of vehicles involved in frontal crashes in the NASS data files with those produced in tests conducted by the IIHS. FIT analysis of the data indicate that approximately 7.5% of all 11-1 o'clock frontal crashes in NASS are represented by the IIHS SOI test condition and they account for 6.1% of all serious-to-fatal injuries to front seat occupants restrained by seat belts and airbags.

The influence of cell geometries on the quasi-static crush behaviors of aluminum honeycombs is explored by experiments. Aluminum 5052-H38 honeycomb specimens with different in-plane orientation angles, cell wall thicknesses and cell sizes were tested under compression dominant combined loads. The load histories of these specimens were obtained. A quadratic and a linear phenomenological yield criteria are used to fit the obtained experimental normal crush and shear strengths for three types of honeycomb specimens under compression dominant combined loads. The quadratic yield criterion is used to fit the experimental results for two types of honeycomb specimens with low relative densities. The linear yield criterion is used to fit the experimental results for one type of honeycomb specimens with a high relative density.

The quasi-static crush behavior of aluminum 5052-H38 honeycomb specimens under non-proportional compression dominant combined loads is investigated by experiments. Compression dominant combined loads and pure compressive loads were applied in different sequences to induce non-proportional combined loads. The experimental results show that the normal crush and shear strengths in combined loading regions and the normal crush strengths in pure compressive loading regions of the non-proportional combined loads are quite consistent with the existing phenomenological yield criterion based on the experimental normal crush and shear strengths under proportional combined loads. The experimental results indicate that the sequence of loading paths for the non-proportional combined loads does not affect the crush strengths of honeycomb specimens.

Effects of impact velocity on the crush behavior of aluminum 5052-H38 honeycomb specimens are investigated by experiments. An impact test machine using pressurized nitrogen was designed to perform dynamic crush tests. A test fixture was designed such that inclined loads can be applied to honeycomb specimens in dynamic crush tests. The results of dynamic crush tests indicate that the effects of impact velocity on the normal and inclined crush strengths are significant. The trends of the inclined crush strengths for specimens with different in-plane orientation angles as functions of impact velocity are very similar to that of the normal crush strength. Experimental results show similar progressive folding mechanisms for honeycomb specimens under pure compressive and inclined loads. Under inclined loads, the inclined stacking patterns were observed. The inclined stacking patterns are due to the asymmetric locations of the horizontal plastic hinge lines.

The crush strength of aluminum 5052-H38 honeycomb materials under combined compressive and shear loads are investigated here. The experimental results indicate that both the peak and crush strengths under combined compressive and shear loads are lower than those under pure compressive loads. A yield function is suggested for honeycomb materials under the combined loads based on a phenomenological plasticity theory. The microscopic crush mechanism under the combined loads is also investigated. A microscopic crush model based on the experimental observations is developed. The crush model includes the assumptions of the asymmetric location of horizontal plastic hinge line and the ruptures of aluminum cell walls so that the kinematic requirement can be satisfied. In the calculation of the crush strength, two correction factors due to non-associated plastic flow and different rupture modes are considered.

A general failure criterion for spot welds is proposed with consideration of the plastic anisotropy and the separation speed for crash applications. A lower bound limit load analysis is conducted to account for the failure loads of spot welds under combinations of three forces and three moments. Based on the limit load solution and the experimental results, an engineering failure criterion is proposed with correction factors determined by different spot weld tests. The engineering failure criterion can be used to characterize the failure loads of spot welds with consideration of the effects of the plastic anisotropy, separation speed, sheet thickness, nugget radius and combinations of loads. Spot weld failure loads under uniaxial and biaxial opening loads and those under combined shear and twisting loads from experiments are shown to be characterized well by the engineering failure criterion.

Experiments to obtain the failure loads of spot welds are first reviewed under combined opening and shear loading conditions. A failure criterion is then presented for spot welds under combined opening and shear loading conditions based on the results from the experiments and a lower bound limit load analysis. In order to account for spot welds under more complex loading conditions, another lower bound limit load solution is presented to characterize the failure loads of spot welds under combinations of three forces and three moments. Based on the limit load solution, an engineering failure criterion is proposed with correction factors determined by different spot weld tests. The engineering failure criterion can be used to characterize the failure loads of spot welds with consideration of the effects of sheet thickness, nugget radius and combinations of loads.

Failure behavior of spot welds is investigated under impact loading conditions. Three different impact speeds were selected to test both HSLA steel and mild steel specimens under combined opening and shear loading conditions. A test fixture was designed and used to obtain the failure loads of spot weld specimens of different thicknesses under a range of combined opening and shear loads with different impact speeds. Accelerometers were installed on the fixtures and the specimens for investigation of the inertia effects. Optical micrographs of the cross sections of failed spot welds were obtained to understand the failure processes in both HSLA steel and mild steel specimens under different combined impact loads. The experimental results indicate that the failure mechanisms of spot welds are very similar for both HSLA steel and mild steel specimens with the same sheet thickness. These micrographs show that the sheet thickness can affect the failure mechanisms.

We conduct static experiments to investigate the influence of shear stress on the crush behavior of honeycomb materials. The aluminum honeycomb materials selected in this investigation are orthotropic due to their manufacturing processes. A test fixture and honeycomb specimens are designed such that combined compressive and shear loads along the strongest material symmetry axis can be controlled and applied accurately. The experimental results indicate that both the peak and crush strengths under combined compressive and shear loads are lower than those under pure compressive loads. A yield function is suggested for honeycomb materials under the combined loads based on a phenomenological plasticity theory. The microscopic crush mechanism under the combined loads is also investigated. A microscopic crush model based on the experimental observations is developed. The crush model includes the rupture of aluminum cell walls so that the kinematic requirement can be satisfied.

This study examines the biofidelity, repeatability, and reproducibility of various anthropomorphic devices in rear impacts. The Hybrid III, the Hybrid III with the RID neck, and the TAD-50 were tested in a rigid bench condition in rear impacts with ΔVs of 16 and 24 kph. The results of the tests were then compared to the data of Mertz and Patrick[1]. At a AV of 16 kph, all three anthropomorphic devices showed general agreement with Mertz and Patrick's data [1]. At a AV of 24 kph, the RID neck tended to exhibit larger discrepancies than the other two anthropomorphic devices. Also, two different RID necks produced significantly different moments at the occipital condyles under similar test conditions. The Hybrid III and the Hybrid III with the RID neck were also tested on standard production seats in rear impacts for a AV of 8 kph. Both the kinematics and the occupant responses of the Hybrid III and the Hybrid III with the RID neck differed from each other.

Recent design characteristic changes in a small segment of production passenger car front seats have focused attention on the influence of seat back strength on occupant kinematics and potentially injurious loads placed on occupants during rear end collisions. The National Accident Sampling Study database from the years of 1980 to 1993 was interrogated to determine the relationship between vehicle change in velocity, and the nature and severity of injuries sustained by passengers occupying those seats in rear end collisions. The results of the NASS data analysis show that the yielding seats in most current automobiles perform well as a passive restraint system. When the yielding passenger car seats are compared to the stiffer seat/cab, the passenger car seats offered improved protection. Additionally, the data indicate that the three point restraint system provides protection and restraint for front seat occupants in rear impact.

The development of the Head Injury Risk Curve (HIRC) and the Skull Fracture Risk Curves (SFRC) that were proposed by Prasad and Mertz for the adult driving population are reviewed, and the problem with using the Maximum Likelihood method to analyze the cadaver 15ms HIC data is discussed. The cadaver data base for skull fracture is expanded by including non-fracture 15ms HIC values for a number of cadavers which had skull fractures at higher impact severities and by including cadaver test results of Ono and Tarriere. This expanded data set was analyzed using the Mertz/Weber method and a new method called “Certainty Grouping”. An updated version of the Skull Fracture Risk Curve (SFRC) is proposed. The efficacy of this revised curve is demonstrated by comparing its predictions to the results of simulated fracture impacts using a finite element model of the head. The HIRC was not changed since no additional brain damage data were analyzed.

The European safety regulation plan regarding frontal barrier offset impact calls for 30° angular impact protection in 1995 and a perpendicular 40% offset deformable barrier impact protection in the 1998 time frame. However, various other governmental and private agencies are looking at alternative test conditions. The Auto Motor and Sport Magazine and other insurance agencies have been conducting rigid barrier front impact tests at 40 and 50% offsets. In this study various test conditions were examined analytically. Detailed finite element models were developed to understand the implications of these impact conditions. The models provided insight into energy management mechanism, load transfer and vehicle deformation patterns due to offset impacts on to perpendicular and angular barriers. Several potential offset conditions were simulated using the FEA models.

Due to the sparsity of cadaver lumbar shear stiffness data, tests on functional lumbar spinal units and a complete lumbar section (T12-L5) were done in both the anterior and posterior directions. Similar tests were performed on the Hybrid III lumbar spine for comparison. Sixteen lumber motion segments were tested quasi-statically for their viscoelastic properties in a multi-directional (5-axis) spine machine. A hydraulic testing machine was used to carry out dynamic tests including cyclic tests at several rates of deformation (0.5 - 50 mm/sec) and relaxation tests (300 sec) to determine the associated viscoelastic properties in constrained and unconstrained modes. The specimens were then loaded to sufficient displacement to cause hard or soft tissue failures. In the quasi-static tests the shear response was linear and the anterior stiffness (155 ± 90 N/mm) was found to be higher than posterior stiffness (104 ± 38 N/mm).

This report describes an experimental validation of an ellipsoid-to-foam contact model. A series of static foam tests was conducted using Side Impact Dummy rib cage, pelvis, upper leg, and wooden ellipsoids as impactors to validate a theoretical foam contact model previously developed. Predicted results of contact forces, calculated using the uni-axial stress-strain relationship and contact areas, yield good correlation with the test data. These studies used CFC foams and were conducted prior to switching to water-blown foam material development. The ellipsoid-to-foam contact model is being integrated into a MADYMO side impact model. The MADYMO/foam simulation model can then be used to help evaluate design variable tradeoffs (e.g., door thickness vs. body side structures and foam padding requirement vs. interior package) thereby reducing the current dependency on testing, bolster development time, and cost.

A review of the existing mathematical models of a car occupant in a rear-end crash reveals that existing models inadequately describe the kinematics of the occupant and cannot demonstrate the injury mechanisms involved. Most models concentrate on head and neck motion and have neglected to study the interaction of the occupant with the seat back, seat cushion, and restraint systems. Major deficiencies are the inability to simulate the torso sliding up the seat back and the absence of the thoracic and lumbar spine as deformable, load transmitting members. The paper shows the results of a 78 degree-of-freedom model of the spine, head, and pelvis which has already been validated in +Gz and -Gx acceleration directions. It considers automotive-type restraint systems, seat back, and seat cushions, and the torso is free to slide up the seat back.

The biodynamic response of cadaver torsos subjected to -Gx impact acceleration is discussed in this paper, with particular emphasis on the response of the vertebral column. The existence of an axial force along the spine and its manifestation as a load on the seat pan are reported. Spinal curvature appears to be an important factor in the generation of this spine load. In anthropometric dummies, the spine load does not exist. Details of the testing and results are given, and the development of a mathematical model is shown.

The head acceleration pulses obtained from monkey concussion, cadaver skull fracture (t = 0.002 sec), and football helmet experiments (0.006< t< 0.011 sec) have been subjected to injury hazard assessment by the Severity Index method. Although not directly applicable, the method correlates well with degree of monkey concussion. The range of Severity Indices for acceleration pulses obtained during impact to nine cadavers, all of which produced a linear fracture, was 540-1760 (1000 is danger to life) with a median value of 910. The helmet experiments showed good correlation between the Severity Index and the Wayne State University tolerance curve. These helmet tests also showed that a kinematics chart with curves of velocity change, stopping distance, average head acceleration, and time, with a superimposed Wayne State tolerance curve, can be useful in injury assessment.