Abstract
Far side impact trauma has been demonstrated as a significant portion of the total trauma in side impacts. The objective of the study was to assess the potential usefulness of countermeasures and assess the trade-offs associated with generic countermeasure design. Because the WorldSID dummy has demonstrated promise as a potential far side impact dummy, it was chosen to assess countermeasures in this mode. A unique far side impact buck was designed for a sled test system that included, as a standard configuration, a center console and outboard three-point belt system. This configuration assumed a left side driver with a right side impact. The buck allowed for additional options of generic restraints including shoulder or thorax plates or an inboard shoulder belt. The entire buck could be mounted on the sled in either a 90-degree (3-o’clock PDOF) or a 60-degree (2-o’clock PDOF) orientation. A total of 19 WorldSID tests were completed. The inboard shoulder belt configuration produced high shear forces in the lower neck (2430 N) when the belt position was placed over the mid portion of the neck. Shear forces were reduced and of opposite sign when the inboard belt position was horizontal and over the shoulder; forces were similar to the standard outboard belt configuration (830 – 1100 N). A shoulder or thorax restraint was effective in limiting the head excursion, but each caused significant displacement at the corresponding region on the dummy. A shoulder restraint resulted in shoulder displacements of 30 – 43 mm. A thorax restraint caused thorax deflections of 39 – 64 mm. Inboard restraints for far side impacts can be effective in reducing head excursion but the specific design and placement of these restraints determine their overall injury mitigating characteristics.
INTRODUCTION
Side impact crashes are second only to frontal impacts in frequency. It has been demonstrated however, that the severity of injury in side impacts is greater than frontal impacts [Banglmaier, et al, 2003; Frampton, et al, 1998; Franklyn, et al, 1998; Haland, et al, 1990; Yoganandan, et al, 2000]. When a side impact occurs to the opposite side of an occupant’s seating location it is termed a far side crash or a non-struck side crash. Far side impact trauma has been demonstrated as a significant portion of the total trauma in side impacts. In general the most common trauma occurs to the head, chest, and abdomen of the occupant [Digges, et al, 2005; Gabler, et al, 2005]. Far side direct 90 degree crashes (3 o’clock) and 60 degree impacts (2 o’clock) are both prominent crash directions [Gabler, et al, 2005]. It was determined in a preliminary experimental investigation that a common cause for head injury is contact with the opposite side door or B-pillar [Fildes, et al, 2002]. Chest trauma occurs commonly to the internal organs such as liver and spleen and has been largely attributed to belt loading [Augenstein, et al, 2000; Yoganandan, et al, 2000]. Although current belt systems were not designed for protection in far side crashes, observations from real-world crashes indicate that the occupant will slip out of the shoulder belt approximately 35% of the time [Mackay, et al, 1991]. Because of newer belt technologies such as pretensioners and belt positioning systems, there may be inherent protection to far side crash occupants. Countermeasures designed specifically for far side impact are few. Belt positioning and belt geometry as well as limiting thoracic excursion may be methods of enhancing the protection to far side crash occupants. The objective of the present study, therefore, was to assess the potential usefulness of countermeasures and assess the trade-offs associated with generic countermeasure design. Because the WorldSID dummy has demonstrated some promise as a potential far side impact dummy [Fildes, et al, 2002], it was chosen to assess and to evaluate countermeasures in this mode.
METHODS
A unique far side impact buck was designed for a sled test system that included, as a standard configuration, a center console and outboard three-point belt system. This configuration assumed a left side driver with a right side impact. The buck allowed for additional options of generic restraints including shoulder or thorax plates or an inboard shoulder belt. The entire buck could be mounted on the sled in either a 90-degree (3-o’clock PDOF) or a 60-degree (2-o’clock PDOF) orientation. The center console plate was designed such that the entire hip engaged the plate; the top of the plate was slightly higher than the iliac crest of a 50th percentile male. The for-aft dimension was determined as the dividing point on the Hybrid-III dummy between the hip and thigh junction when the dummy sat in the seat. The lower belt anchor point was determined by using the Hybrid-III 50th male dummy and positioning the lap belt such that belt traversed a 45-degree angle from pelvis to anchor point. Upper belt anchor points were adjustable as described later. The geometry and dimensions are shown in Figure 1. The console plate was padded with 25 mm of 207 kPa (30 psi) paper honeycomb. If either a thorax plate or shoulder plate was used, the padding was 25 mm of 103 kPa (15 psi) paper honeycomb. The dimensions of the shoulder or thorax plates were 100 mm in height and 460 mm long. Twelve different far side test conditions were evaluated (Figure 2) including reverse belt geometry (inboard shoulder belt), and shoulder or thorax restraints (Table 1). All tests were conducted with a lap belt and a center console. The shoulder and lap belts were low elongation standard belts (6% elongation at 11.1 kN). The shoulder belt could be configured such that the D-ring anchor point was horizontal with the top of the shoulder (low position), 90 mm above the shoulder (mid position), or 150 mm above the shoulder (high position). All of these D-ring locations were approximately 120 mm behind the mid point of the shoulder. As a realistic worst case configuration, the shoulder belt D-ring was positioned in the mid position vertically, and forward (30 mm behind shoulder instead of 120 mm behind) of the usual anchor location (tests FSWS132-133). Tests were conducted at either a direct 90 degree impact or an oblique 60 degree direction and low speed (11 km/h) or high speed (30 km/h) delta-v. The high speed test condition was a 100 ms square wave sled pulse with 8.8 g average acceleration. The load wall was instrumented with tri-axial load cells: three for the leg plate, two for the pelvis plate, two for the center console plate, and two for the thorax or abdomen plates if used. Seat belt force transducers were used and sled acceleration in three directions was recorded. The coordinate system followed the SAE-j211 standard sign convention. A 50th percentile WorldSID production model was used in the tests with the half arms in the horizontal, or “up” position, to simulate driver arm position. The WorldSID instrumentation included head linear and angular accelerations, upper/lower neck loads, chest accelerations/deflections, lumbar spine loads, and pelvic loads.
Figure 1.
Schematic diagram, with dimensions, of far side sled buck design viewed from the side (top) and from the front (bottom). The angle of the seat bottom with respect to horizontal was 15 degrees. If thorax or shoulder plates were used they were adjustable up/down/in/out; dimensions were 100 mm by 460 mm.
Figure 2a.
Illustrations of test configurations demonstrating shoulder belt orientation and presence of a shoulder or thorax load plate.
Table 1.
Test configurations. All test configurations included a center console and lapbelt.
| Test ID | Delta-V km/h | Shoulder Belt | Plate restraint | Impact angle | ||
|---|---|---|---|---|---|---|
| In / Out | Position | Tension | ||||
| FSWS107 | 11 | Inboard | low | no | shoulder | 90-deg |
| FSWS108 | 30 | Inboard | low | no | shoulder | 90-deg |
| FSWS109 | 30 | Outboard | mid | yes | shoulder | 90-deg |
| FSWS110 | 11 | Inboard | low | No | Thorax | 90-deg |
| FSWS113 | 30 | Inboard | low | No | Thorax | 90-deg |
| FSWS114 | 30 | Outboard | Mid | Yes | Thorax | 90-deg |
| FSWS115 | 11 | Inboard | High | No | 90-deg | |
| FSWS118 | 30 | Inboard | High | No | 90-deg | |
| FSWS119 | 30 | Outboard | Mid | Yes | 90-deg | |
| FSWS121 | 30 | Inboard | Low | Yes | 90-deg | |
| FSWS122 | 11 | Inboard | low | Yes | Thorax | 60-deg |
| FSWS123 | 30 | Inboard | low | Yes | Thorax | 60-deg |
| FSWS124 | 11 | Inboard | High | No | 60-deg | |
| FSWS126 | 30 | Inboard | High | No | 60-deg | |
| FSWS127 | 30 | Inboard | Low | Yes | 60-deg | |
| FSWS129 | 11 | Outboard | Mid | No | 60-deg | |
| FSWS130 | 30 | Outboard | Mid | No | 60-deg | |
| FSWS132 | 11 | Outboard | Mid-for | No | 90-deg | |
| FSWS133 | 30 | Outboard | Mid-for | No | 90-deg | |
Retroreflective targets were placed on the head, at T1, T12, and pelvis. Reference targets were fixed to the sled and buck. A three-dimensional, 1000 f/s motion tracking system (Vicon Motion Systems, Centennial, CO) was used to quantify occupant kinematics.
RESULTS
A total of 19 WorldSID tests were completed. The low speed tests were used mainly to ensure proper recording by onboard instrumentation systems and assess dummy response at a presumed non-injurious speed [Gabler, et al, 2005]. The trends in the low speed tests were similar albeit of lower magnitudes than equivalent high speed tests. The following results highlight the differences by restraint configuration in the high speed tests. Since the configuration of the shoulder belt was considered a large contributor to response, comparisons are divided into tests using an inboard belt configuration and tests using a more standard outboard belt configuration. Tests at 60-degree orientation were evaluated separately. Response plots are given in the Appendix.
Appendix A.
Figures comparing various measured responses. The first eight plots are from inboard belt tests; the next eight plots are from outboard belt tests; the last eight plots are from 60-degree tests.
Overall, the pelvis Y-loads were all fairly close in magnitude. Because the initial test condition was such that the dummy’s pelvis was in contact with the pelvic plate, pelvic loads peaked early. Outboard belt tests demonstrated later peak shoulder belt loads than inboard belt tests. Shoulder belt load peaks were generally earlier than lap belt loads. Tests with either a thorax or shoulder load plate produced lower magnitude belt loads. The inboard shoulder belt configuration produced high shear forces in the lower neck (2430 N) when the belt position was placed over the mid portion of the neck (high-position) (Appendix). Shear forces were reduced and of opposite sign when the inboard belt position was horizontal and over the shoulder; forces were similar to the standard outboard belt configuration (830 – 1100 N). Interestingly, the lower neck lateral-flexion moments were not appreciably affected by belt position (121 – 154 Nm).
A shoulder or thorax restraint was effective in limiting the head excursion, but each caused significant displacement at the corresponding region on the dummy. A shoulder restraint resulted in shoulder displacements of 30 – 43 mm. A thorax restraint caused thorax deflections of 39 – 64 mm (Table 2). For inboard belt tests, the center console caused abdomen deflections of no more than 27 mm. The shoulder belt still seemed to modulate the abdomen loads in tests where no thorax or shoulder plates were present. This can be seen as a double peak in the abdomen displacements (Appendix).
For outboard belt tests, belt position was influential in resulting console interactions and abdomen loads. An outboard shoulder belt limited center console interaction but still caused up to 22 mm of displacement. The outboard belt configuration with the D-ring in the forward position produced slightly less abdominal deflection but more head excursion as the torso slipped out of the shoulder belt. Again, tests with shoulder or thorax plates limited head excursion but increased rib deflections.
In 60 degree tests, the shoulder belt did not slip away from the dummy shoulder complex. This can be seen in the shoulder belt loads that are in the range 2300 to 2700 N. For the inboard belt test with a high position, the neck shear forces were again much higher than for low position tests, but because of the oblique nature the Y-load shear was lower than for the equivalent 90-degree tests.
DISCUSSION
Far side impacts have received little attention in the literature despite contributing to a significant portion of the injuries and Harm in vehicle crashes. It was noted that 43 % of AIS=3+ injured persons, and 30 % of overall Harm were to occupants on the non-struck side of the side impact collision [Gabler, et al, 2005]. Head injuries remain the predominant injury for far side crash occupants followed closely by torso injuries [Digges, et al, 2005]. The goal in the current series of experiments was to investigate efficacy and trade-offs of generic countermeasures in a far side impact collision. Because previous studies implicated the far side door as a causative agent for head injuries, one of the main purposes of countermeasure design should be to reduce head excursion and avoid contact with the opposite side door. This was accomplished in the generic countermeasures using belt system geometry and placement, as well as thorax or shoulder support in the form of padded plates. An example of the implementation of a thorax or shoulder support has been accomplished using an airbag system in some preliminary attempts at these types of countermeasures [Bostrom and Haland, 2003].
The existing outboard belt restraint systems may have a D-ring located in a position that allows the torso of the occupant to slip out of the shoulder belt in a far side impact collision. In a previous examination of real-world crashes Mackay determined that the torso can slip out of the shoulder belt approximately 35% of the time [Mackay, et al, 1991]. Considering how high the incidence rate is for head injuries in far side crashes, more recent real-world crash data may indicate that belt slip might occur more frequently [Gabler, et al, 2005; Digges, et al, 2005]. All of the tested countermeasures in the present study reduced the head excursions (Table 2) compared to the forward belt condition in the 90-degree test (FSWS133) wherein the dummy torso slipped out of the shoulder belt.
Countermeasures that reduce head excursions must be designed carefully to not increase the likelihood of injury to other body regions. The thorax restraint used in the present study increased chest displacements to likely injurious levels (64 mm). Reverse belt geometry (inboard shoulder belt) demonstrated promise but only when placement was horizontal over the shoulder. This specific placement is easier to control in a dummy occupant than it would be in a real human occupant. A mispositioned shoulder belt, such as the high position for the inboard belt (tests FSWS118 and FSWS126) may cause high loads to the neck structures, placing the internal structures such as the vascular system and spinal column at risk for trauma [Sinson, et al, 2003].
The WorldSID production version dummy provides extensive instrumentation to evaluate restraint system design. It has a self-contained data acquisition system that allows for complete internal wiring of accelerometers, load cells, and deflection sensors. The WorldSID has a unique design of the lumbar spine that looks like an inverted “U” which allows for lateral motion of the torso relative to the pelvis. This lateral torso motion has been shown to be unique in PMHS testing and may be the reason the head can contact the opposite side door in far side crashes [Fildes, et al, 2002]. There are some limitations of this dummy for use in far side impact crashes, a mode that it was not originally designed for. Each of the ribs has an internally mounted IR-TRACC [Rouhana, et al, 1998] that measures deflection best when impacted directly lateral. Because of the interaction of the shoulder belt with the oblique portion of the lower rib cage for the outboard belt configuration, the abdomen deflections recorded in the present test series may be lower than actual oblique deflections at the location of belt interaction. Also, the design of the external portion of the shoulder region that interacts with the belt does not follow a human-like contour. For certain countermeasure belt designs where belt effectiveness depends on proper interaction with the shoulder, belt engagement may not be realistic in the current WorldSID design.
The limitations of this study are linked to the biofidelity of the WorldSID dummy in this crash mode. The WorldSID was chosen because it demonstrated the greatest potential for use as a far side impact crash dummy compared to the BioSID, NHTSA-SID, or EuroSID dummies [Fildes, et al, 2002]. The WorldSID however, does not yet have demonstrated biofidelity in far side impact. A series of PMHS tests in the same far side impact sled buck are planned to examine in more detail, the biofidelity of the WorldSID in this crash mode.
In conclusion, reverse shoulder belt geometry (inboard) restraints that are positioned directly over the shoulder may hold potential for far side impact countermeasure design. Alternate load carrying paths such as shoulder or thorax restraints may also be effective in reducing head excursion and containing the occupant but the specific design characteristics must be such that local chest deflections are not injurious. The outboard belt configuration in the current test series may cause high oblique lower rib cage deflections when the occupant slips out of the belt at the shoulder.
Figure 2b.
Illustrations of test configurations demonstrating shoulder belt orientation and presence of a shoulder or thorax load plate.
Table 2a.
Inboard belt tests at 90 degrees
| Test ID | FSWS108 | FSWS113 | FSWS118 | FSWS121 |
|---|---|---|---|---|
| Configuration | Low belt + Shoulder | Low belt + Thorax | High belt | Low belt + tension |
| Pelvis-Y (N)
|
4815 | 4164 | 5620 | 4971 |
| Lap Belt (N)
|
180 | 598 | 1847 | 1502 |
| Shoulder Belt (N)
|
320 | 1235 | 2795 | 2244 |
| Shoulder (mm)
|
42.8 | 0.6 | 1.9 | 0.6 |
| Thorax-2 (mm)
|
1.0 | 63.6 | 0.8 | 3 |
| Abdomen-2 (mm)
|
18.0 | 5.1 | 27.4 | 24.7 |
| Lower Neck-Y (N)
|
−1046 | −942 | 2431 | −907 |
| Lower Neck-X (Nm)
|
−145 | −133 | −121 | −183 |
| Head Y-Excursion (mm) | 355 | 428 | 463 | 466 |
Table 2b.
Outboard belt tests at 90 degrees
| Test ID | FSWS109 | FSWS114 | FSWS119 | FSWS133 |
|---|---|---|---|---|
| Configuration | Mid belt + Shoulder | Mid belt + Thorax | Mid belt | Forward Belt |
| Pelvis-Y (N)
|
4575 | 4003 | 5001 | 4942 |
| Lap Belt (N)
|
265 | 515 | 889 | 715 |
| Shoulder Belt (N)
|
695 | 1743 | 2800 | 1693 |
| Shoulder (mm)
|
29.7 | 2 | 2.5 | 2.6 |
| Thorax-2 (mm)
|
0.8 | 58.1 | 4.2 | 8.1 |
| Abdomen-2 (mm)
|
15.2 | 4.5 | 22.4 | 17.1 |
| Lower Neck-Y (N)
|
−858 | −1014 | −1101 | −829 |
| Lower Neck-X (Nm)
|
−117 | −132 | −154 | −137 |
| Head Y-Excursion (mm) | 354 | 400 | 512 | 496 |
Table 2c.
Tests at 60 degrees
| Test ID | FSWS127 | FSWS123 | FSWS126 | FSWS130 |
|---|---|---|---|---|
| Configuration | 60-degree Low belt + tension | 60-degree Low belt + tens + Thorax | 60-degree High belt | 60 Mid belt |
| Pelvis-Y (N)
|
4043 | 2734 | 4043 | 4334 |
| Lap Belt (N)
|
1212 | 949 | 1845 | 1169 |
| Shoulder Belt (N)
|
2331 | 1569 | 2746 | 3014 |
| Shoulder (mm)
|
1.8 | 0.5 | 0.9 | 2.4 |
| Thorax-2 (mm)
|
3.9 | 39 | 1 | 5.8 |
| Abdomen-2 (mm)
|
17.1 | 2.1 | 20.2 | 12.1 |
| Lower Neck-Y (N)
|
−1051 | −700 | 1556 | −925 |
| Lower Neck-X (Nm)
|
−162 | −105 | −148 | −125 |
| Head Y-Excursion (mm) | 384 | 288 | 357 | 427 |
ACKNOWLEDGEMENT
The funding for this study was provided in part by the Australian Research Counsel with cost sharing and support from other sponsors including GM Holden Innovation, and Autoliv Research. Additional funding for this research has been provided by private parties, who have selected Dr. Kennerly Digges [and the FHWA/NHTSA National Crash Analysis Center at the George Washington University] to be an independent solicitor of and funder for research in motor vehicle safety, and to be one of the peer reviewers for the research projects and reports. Neither of the private parties have determined the allocation of funds or had any influence on the content. This study was also partially supported by VA Medical Center Research. The authors would like to thank John Humm, Paul Gromowski, and Mark Meyer of the Neuroscience Research Laboratories at the VA Medical Center in Milwaukee, Wisconsin.
REFERENCES
- Augenstein J, Perdeck E, Martin P, et al. Injuries to restrained occupants in far-side crashes. 44th Annual Proceedings, Association for the Advancement of Automotive Medicine; October 2–4, 2000; pp. 57–66. [PMC free article] [PubMed] [Google Scholar]
- Banglmaier RF, Rouhana SW, Beillas P, et al. Lower extremity injuries in lateral impact: A retrospective study. 47th Annual Proceedings, Association for the Advancement of Automotive Medicine; September 22–24, 2003; pp. 425–444. [PMC free article] [PubMed] [Google Scholar]
- Bostrom O, Haland Y. Benefits of a 3+2 point belt system and an inboard torso side support in frontal, far-side, and rollover crashes. 18th ESV conference proceedings; Nagoya, Japan. 2003. [Google Scholar]
- Digges K, Gabler H, Mohan P, et al. Characteristics of the injury environment in far-side crashes. 49th Annual Proceedings, Association for the Advancement of Automotive Medicine; September 12–14, 2005; pp. 185–197. [PMC free article] [PubMed] [Google Scholar]
- Digges KH, Gabler HC. Opportunities for reducing casualties in far-side crashes. SAE World Congress (2006-01-0450); April 3–6, 2006; SP-1997. [Google Scholar]
- Fildes BN, Sparke LJ, Bostrom O, et al. Suitability of current side-impact test dummies in far-side impacts. Proceedings IRCOBI, 19th International Conference on the Biomechanics of Impact; September 18–20, 2002; pp. 43–56. [Google Scholar]
- Frampton RJ, Brown R, Thomas P, et al. Importance of non-struck side occupants in side collisions. 42nd Annual Proceedings, Association for the Advancement of Automotive Medicine; October 5–7, 1998; pp. 303–320. [Google Scholar]
- Franklyn M, Fitzharris M, Yang K, et al. Aortic injuries in side impacts: A preliminary analysis. 46th Annual Proceedings, Association for the Advancement of Automotive Medicine; September 30–October 2, 2002; pp. 1–12. [PubMed] [Google Scholar]
- Gabler HC, Digges K, Fildes BN, Sparke L. Side impact injury risk for belted far side passenger vehicle occupants. SAE World Congress; April 11–14, 2005; paper # 2005-01-0287, Society of Automotive Engineers. [Google Scholar]
- Haland Y, Lovsund P, Nygren A. Estimation of fatalities and disabilities in car-to-car side impacts: An evaluation of different risk factors. 34th Annual Proceedings, Association for the Advancement of Automotive Medicine; October 1–3, 1990; pp. 275–287. [Google Scholar]
- Mackay GM, Parkin S, Hill J, et al. Restrained occupants on the non-struck side in lateral collisions. 35th Annual Proceedings, Association for the Advancement of Automotive Medicine; October 7–9, 1991; pp. 119–132. [Google Scholar]
- Parenteau CS, Viano DC, Lovsund P, et al. Foot-ankle injuries: Influence of crash location, seating position and age. 39th Annual Proceedings, Association for the Advancement of Automotive Medicine; October 16–18, 1995; pp. 177–192. [Google Scholar]
- Rouhana SW, Elhagediab AM, Chapp JJ. A High-Speed Sensor for Measuring Chest Deflection in Crash Test Dummies. Proceedings of 16th International Technical Conference on Enhanced Safety of Vehicles, Technical Paper No. 98-S9-O-15; Windsor, Canada. 1998. [Google Scholar]
- Sinson GP, Yoganandan N, Pintar FA, et al. Carotid artery trauma in motor vehicle crashes: Investigation of the local tensile loading mechanism. Proceedings IRCOBI, 20th International Conference on the Biomechanics of Impact; September 24–25, 2003; pp. 207–216. [Google Scholar]
- Thomas C, Faverjon G, Henry C, et al. Problem of compatibility in car-to-car collisions. 34th Annual Proceedings, Association for the Advancement of Automotive Medicine; October 1–3, 1990; pp. 253–267. [Google Scholar]
- Yoganandan N, Pintar FA, Gennarelli TA, et al. Patterns of abdominal injuries in frontal and side impacts. 44th Annual Proceedings, Association for the Advancement of Automotive Medicine; October 2–4, 2000; pp. 17–36. [PMC free article] [PubMed] [Google Scholar]