- Email: [email protected]
Influences of the physical parameters on the risk to neck injuries in low impact speed rear-end collisions
Accid. Anal. and Prev., Vol. 28, No. 4, pp. 493-499, 1996 Copyright 0 1996 Elsevier Science Ltd
Pergamon
Printed in Great Britain. All rights reserved OOOl-4575/96 $15.00 + 0.00
INFLUENCES OF THE PHYSICAL PARAMETERS ON THE RISK TO NECK INJURIES IN LOW IMPACT SPEED REAR-END COLLISIONS * KOSHIRO
ONO’
and
MIJNEKAZU KANNO~
‘Japan Automobile Research Institute, Karima, Tsukuba, Ibaraki 305, Japan; 2College of Engineering of Nihon University, Tokusada, Tamura, Koriyama, Fukushima 963, Japan (Accepted 28 January 1996)
Abstract-The current state of neck injuries sustained in car-to-car rear-end collisions were investigated according to recent automobile accident statistical data in Japan. To clarify the neck injury mechanisms for low impact speed car collisions, the newly developed impact sled experiment which simulates actual car impact acceleration was performed using human subjects.? In order to measure and analyze the physical parameters such as human head rotational acceleration, neck bending moment, shearing and axial forces, the component measurement method with six-degrees of freedom was applied and demonstrated. Furthermore, relationships among the physical parameters, impact speeds, sitting positions, headrest heights and neck muscle tones applied on the subject’s head and neck system were analyzed. These analyses would enable us to comprehend the conditions of the neck muscle tone and the effects of the sitting postures including headrest height, factors which are of vital importance to the understanding of neck injury mechanisms. Copyright 0 1996 Elsevier Science Ltd Keywords-Injury mechanism, Cervical injury, Whiplash injury, Bending moment, Shearing force, Axial force, Sitting position, Muscle tone out for relatively high impact speed ranges of 20 km/h or higher (Desantis 1991; Warner et al. 1991; ForetBruno et al. 1991; Viano 1992; Ewing et al. 1969; Mertz 1984; Mertz et al. 1989). It is reported that minor neck injuries at low impact speeds often involve symptoms of soft tissue injuries called cervical contusions, sprains or distortions, which lack clear objective signs but tend to have prolonged healing times Olsson et al. 1990; Handbook of Whiplash Injuries 1987). The number of such neck injuries is markedly greater than the number of other types of injuries (Handbook of Whiplash Injuries 1987). Only a small number of studies (Mertz and Patrick 1971; Svensson and Lovsund 1992; Foust et al. 1973; Melvin et al. 1972; Ono et al. 1992; Nosaka et al. 1987, 1988; Ono 1992) are found on fundamental aspects of the mechanism of those injuries, due to the complexity of symptoms and the lack of sufficient objective symptoms. As a first step towards the clarification of neck injury mechanisms in such accidents, a sled device, capable of simulating rear-end car collision tests at low impact speeds, was specially developed wherein test subjects could experience actual collisions without any risks. Each human subject was seated on a sled in order to analyze the impact response acting on the area from the head to the neck. This is done by
INTRODUCTION The so-called “whiplash” neck injuries sustained in car rear-end collisions were gradually reduced owing to the effectiveness of the headrest, etc. introduced since the latter half of the 1960s. According to recent insurance statistical data on automobile accidents in Japan, however, approximately 50% of car-to-car traffic accidents have resulted in neck injuries, particularly at low impact speeds. The number of such accidents has increased recently, and neck injuries sustained in rear-end collisions in particular, have shown a major increase again. Studies on neck injuries using human/animal cadavers or volunteers have been carried out, mainly in the USA, since the 1960s (Mertz and Patrick 1967, 1971). More recently, studies employing HY-III dummies have been carried out partly as a result of neck injury criteria being reintroduced as one item to be evaluated to ensure automobile safety in collisions. All of these experimental studies have been carried
*Presented at the 1993 IRCOBI Eindhoven, Netherlands tin this paper, the term “human
Conference, subjects”
September
S-10,
refers to volunteers. 493
K. ONO and M.
494
Rtnnberof 0
100.000
neck
injuries
zoo. 000
300.000
400.000
KANNO
Table 2. Distribution of injured body regions of occupants involved in rear-end collisions
Year 85
(43.9%)
’ 86
(44.8%)
.a7
(47.3%)
. aa
(4-I.‘I%)
Neck
.a9
(47.9%)
Chest/abdomen
90
(48. ID
Extremities
‘91
(50.8%)
Others
Numberof neck injuries Numberof 111other injuries
Head/face
Total
Rear-end collisions (car-to-car)
All other collisions
40,127 (13.5%) 230,594 (77.5%) 11,290 (3.8%) 11,660 (3.9%) 3,746 (1.3%) 297,417 (100.0%)
235,513 (25.4%) 355,703 (38.3%) 97,127 (10.5%) 221,734 (25.2%) 11,453 (1.2%) 927,530 (100.0%)
Fig. 1. Number of neck injuries of occupants involved in all types of car-to-car collisons.
applying a method that measures components using six-degrees of freedom in measuring head accelerations. Moreover, various physical parameters (impact speed, sitting posture, headrest height and degree of occupant’s tension) that influence the impact on the subject’s head to neck area were analyzed. CONDITIONS OF NECK INJURIES SUSTAINED IN REAR-END CAR COLLISIONS IN JAPAN
distortions account for approximately 95% of such injuries (Table 3). It is reported that 1520% of the casualties above suffer from prolonged symptoms (Statistical Analysis Results on the Compulsory Automobile Liability Insurance 1992). It may be said that the trend showing the increase of automobile accidents involving such neck injuries is inevitably due to the increased number of vehicles in use. In Japan, however, the use of a headrest has been mandatory since 1969, and the use of seat belts obligatory since 1986. Considering the fact that vehicle safety has been remarkably enhanced compared to 20 years ago, the obvious question arises as to why such minor neck injuries, especially those having prolonged symptoms, cannot be reduced. In this regard, a clarification of the mechanisms of neck injuries in such accidents is urgently required.
According to the automobile accident insurance statistical data in Japan (Statistical Analysis Results on the Compulsory Automobile Liability Insurance 1992), minor neck injuries account for 50% or so of car-to-car accidents, a fact which tends to increase gradually each year (Fig. 1). Looking at the statistical data of 1991 alone, approximately 95% of injuries sustained in car-to-car rear-end collisions were minor injuries of AIS 1 (Table l), and about 80% of the above are concentrated on the neck of the occupant (Table 2). The so-called “whiplash” minor neck injuries involving cervical contusions, sprains and/or
The six-degrees of freedom component measuring method (Ono et al. 1978; Ono and Kanno 1993) per eqn (1) is applied to the measurement of head
Table 1. Distribution of injuries to occupants involved in rear-end collisions (1991)
Table 3. Distribution of types of injuries of occupants involved in rear-end collisions (1991)
AIS 1 AIS 2 AIS 3-6 Unknown
Rear-end collisions (car-to-car)
All other collisions
279,688 (94.0%) 10,264 (3.5%) 3,086 (1.0%) 4,383
719,091 (77.5%) 141,774 (15.3%) 55,562 (6.0%) 11,103 ( 1.2%) 927,530 (100.0%)
(1.5%) Total
297,417
(100.0%)
HEAD ACCELERATION MEASUREMENT AND ANALYSIS OF NECK RESPONSES
Contusion/Sprain Laceration/Avulsion Fracture/Dislocation Others Total
Rear-end collisions (car-to-car)
All other collisions
381,125 (94.5%) 5,493 (1.8%) 6,143 (2.1%) 4,636 (1.6%) 297,417 (100.0%)
666,051 (71.8%) 103,315 (11.1%) 144,523 (15.6%) 13,641 (1.5%) 927,530 (100.0%)
Risk of neck injuries in low impact speed rear-end collisions
acceleration.
where CQis No.k accelerometer, pro is the translational acceleration, 6, is the rotational acceleration at the center of gravity, 0, is the rotational velocity, fk is the location of the No.k attached accelerometer and & is the direction of the No.k attached accelerometer. The head movement in this experiment can be assumed as the two-dimensional movement of the X-Z plane, and the 4-channel measuring method is applied to measure the head acceleration. Components of forces and moments acting on the upper portion (Cl) of the neck can be determined by the application of the component measurement with six-degrees of freedom. Forces and moments at the head’s center of gravity per application of the method to this experiment can be expressed as follows. fi = m&o-fo
(2)
T,=IO,+~ox(Ioo)-T,
(3)
where f, and T, represent the headrest reaction force and the moment, respectively. According to eqns (2) and (3), the forces and the bending moments around the upper portion of the neck can be calculated as follows. J‘N= RJ; = Rf T,=RR=RR+R~&
where R and I represent the coordinate matrix.
(4) (5)
convention
EXPERIMENTS Car collision simulation sled experiment
Based on the experimental data of the two cars, a sled capable of simulating the impact acceleration profile of the struck car was produced (Ono and Kanno 1993). As for the crush characteristic that dominates the sled impact acceleration, coil springs capable of maximum approximation of the impact acceleration-time history waveform of the struck car were used. The seat system used in the experiment was the same as those used in the experimental cars. Conditions and objectives of experiments
Four kinds of physical parameters--(l) impact speed, (2) seatback angle of the occupant’s sitting posture, (3) headrest height and (4) degree of occupant’s tensed condition (neck muscle tonekwere
495
used as parameters that would affect the human headneck responses. Four series of experiments were set up, using a combination of above individual experimental conditions (Ono and Kanno 1993). In the series of experiments, three male human subjects (volunteers) with ages ranging from 22 to 43 wore seat belts and sat in the normal sitting posture for the measurement of the head accelerations, sled impact speeds and occupant motion, etc. Impact speeds were set at 2, 3 and 4 km/h, respectively, for frontal and rearward collisions. For the degree of the occupant’s neck tension electrodes, attached to the skin surface of the muscles of the trapezius and the sternocleidomastoideus on both sides of the neck, were connected to the electromyogram (EMG), which was monitored.
RESULTS
OF EXPERIMENTS DISCUSSION
AND
Head acceleration measurement and neck impact responses
Head dimensions and locations of accelerometers installed on each subject were determined by means of a dimensional measurement using X-ray films. The center of gravity, mass and moment of inertia of the head required for the determination of head-neck responses were obtained from the dimensions determined by the X-ray films and the values reported by Walker (1973). The same values were applied equally to the three human subjects. The head rotational angle was obtained by integrating the rotational accelerations, and the maximum rotational angle and its duration were measured by a high speed video camera in order to minimize any integration error. An analytical comparison was made between the two values. Figure 2 shows an example of an analysis done on rearward collisions at three levels of sled impact speeds. The head slightly rotates clockwise around 50 ms where the acceleration of the sled reaches its maximum value-that is, where “whiplash” starts. The head starts rotating rearward (counterclockwise) around 100 ms where the sled acceleration disappears, but the bending moment and shear force of the neck reach maximum around 110 ms which is slightly after the point of looms. The neck response approaches zero around 200 ms where the head rotational angle reaches maximum. The head-neck motion patterns show similar tendencies in other cases as well. In the case where the impact speed is 2 km/h, however, the head-neck movement delays slightly as the subject’s riding posture is inclined forward slightly.
K. ONO and M. KANNO
496
------
100 -
Sled
-
- -
- -
2.2 km/h 3.1 km/h 4.0 km/h
Extension
Sled speeds
-Standard
G? 0_-_\_ 2
0/*
I
L
-.-.l_&lw Headrest 3 ------Hlthout Sled speed 4 krmh
I w
-
B z
z 8 2
-100 50 -
Neck bending
I
+
:
1
moment
0
I
-5o500 -
Neck shear force-FX
-10
Neck axial force-FZ
0
20
Head
Fig. 3. Head-neck
0
100
200
300
Time (ms) Fig. 2. Head-neck
reponses of 2, 3,4 km/h (without headrest).
sled runs for extension
Eflect of diflerences in headrest height on neck impact responses The relationship between the neck bending moment and the head rotational angle is shown in Fig. 3, using different types of headrest-i.e. “standard”, “low” and “without headrest”. Although the bending moment level in the figure is markedly diverse in the three different conditions, it also shows the neck bending moment of a cadaver (without headrest), which constitutes the basic corridor for the neck injury threshold currently proposed (Mertz and Patrick 1971). In the experiment, a sharp rise of bending moment is found in the range where the rotational angle is small, and the so-called “whiplash” movement is observed but is reduced as the rotational angle is increased after the bending moment has reached its maximum level. This is presumably because the cervical muscular force could not resist the impact load. In cases where a cadaver is used, such a sharp rise is absent, the rise being gentle instead. This would be due to the fact that the cadaver’s cervical muscle had hardly any resistance against the impact, while the cervical muscle of the
40
rotational
60
angle
e0
188
(deg.)
responses by the different types of headrest compared to tests using cadavers.
as
human subject resisted the impact. Such a resistive force of a cervical muscle may place an excessive burden on supportive soft tissues (ligament, muscle, intervertebral joint envelop, etc.) of the cervical spine. The head rotational angle is minimized where the “standard” headrest is used, while it becomes larger where the “low” headrest is used, as the bending moment becomes greater. This shows that an appropriate adjustment of headrest height is very important for the prevention of excessive rearward bending of the neck and to reduce the impact load. Parameters that injluence neck impact responses Neck injuries in general and those sustained by the hyper-extension or hyperflexion of the cervical spine in particular have been discussed in terms of neck bending moment and head rotational angle. The responses acting on the neck, however, are not affected by the bending moment alone, but also by the shear force and axial force (compressive force and tension). In this study, the headrest height and seatback inclination angle (sitting posture) are analyzed in terms of three parameters-the neck bending moment, shear force and axial force. Injluence of headrest height. The highest impact response is found with the low headrest, resulting in an excessive load on the neck. In the case where no headrest is installed, the load on the neck is lowest, but the head rotational angle is largest, resulting in cervical hyper-extension. In the case where the “standard” headrest height is used, both the bending moment and the axial force are lowest, but the shear
Risk of neck injuries in low impact speed rear-end collisions
force may increase in proportion to the intensity of impact. This is presumably because the head impact is effectively suppressed by the headrest itself, and the reaction force acts directly on the neck and becomes greater. Therefore, it is reasonable to say that not only the neck bending moment and the head rotational angle but also the neck shear force and the axial force should be analyzed in relation to the neck responses. Influence of sitting posture (seatback angle reclined backward). Although no significant difference
is found in neck bending moment or shear force between the standard and reclined postures, a great difference is found in the axial force. This is presumably because the impact force in the reclined sitting posture is borne in the vertical direction of the neck. Therefore, it will be necessary to analyze not only the neck bending moment and the head rotational angle but also the shear force and the axial force in relation to neck loading conditions, as in the case of headrest height.
491
( km/h)
10 9 8 7 6 5,
4 3, 2 l-
Struck car speed after
1’
/ 01 ~“1’
2:
5-
1
10 15-
3’
4:
5
2 +
6
7 8
3
4
impact
9
10 ( kl,h)
Sled
5 impactspeed
10 'bl e I
20-
Sled lapact Tests
Wm)
Degree of influences of parameters on the risk to neck injuries
It was found that the impact response on the head-neck area was markedly affected by differences in the impact speed, sitting posture and headrest height. Hence, the relative degree of their influence was studied against the impact speeds in actual car experiments. In the study, the neck bending moment alone was used as the indication of the neck impact response. Figure 4 shows the relationship between the impact speed of the striking car and that of the struck car in actual low impact speed car experiments. The sled impact speed, corresponding to the struck car during the impact, is also plotted in the figure. The bending moment acting on the neck is also shown against each parameter. The bending moment acting on the neck is roughly proportional to the sled impact speed. However, the neck bending moment with the “low” headrest is about 50% higher than that of the “standard” headrest, in the case where the sitting posture is “standard”. In the case where the neck is in hypertension, the bending moment is about 40% lower than that of the “standard” headrest. The quantitative analysis of the degree of neck muscle tone is being done at present in relation to trapezius, muscle electromyogram sternocleidomastoideus (EMG) and neck impact response. The comparison between the standard sitting posture and reclined posture shows that the degree of neck responses
Fig. 4. Comparison between actual car impact tests and sled tests together with a comparison between neck bending moments and speeds of sleds under various conditions.
of the latter is about 30% higher than that of the former. Considering such findings, it is reasonable to say that the influence of differences not only in impact speed but also in sitting posture such as headrest height, backward inclination, and occupant’s tension (neck muscle tone) on the neck impact responses cannot be ignored in rear-end collisions at low impact speeds.
CONCLUSION 1. According to automobile accident insurance statistical data in Japan, approximately 50% of the car-to-car accidents involve minor neck injuries. Looking at the car-to-car rear-end collisions alone, neck injuries of AIS 1 account for 80%, and cervical contusions, sprains or distortions account for 95% or so in terms of type of injury.
K.
498
Studies on those minor neck injuries with substantial engineering approaches were few, due to the lack of sufficient objective symptoms associated with such injuries. In this regard, experimental studies have been done by simulating car rear-end collisions at low impact speeds using human subjects, for the determination of influences of headrest height, sitting posture, etc. Although the rigidity of the seatback and headrest have significant influences on the neck responses, they are not considered in this study. The following findings have been obtained: The neck bending moment shows a sharp rise where the head rotational angle of the human subject is small. This is due to the resistance force of the cervical muscular force of the subject, which may cause an excessive burden on the soft tissues of the cervical spine. 6) Not only the neck bending moment and head rotational angle but also the neck shear force and axial force affect the neck responses, according to differences in headrest height and sitting posture. (ii) The bending moment that acts on the neck is roughly proportional to the sled impact speed. However, the neck bending moment may differ by 50% or so due to differences in headrest height and sitting posture. (iii) In a case where the neck is in hypertension, the neck bending moment reduces by 40%, while it increases by 30% where the seatback is inclined backwards. The foregoing findings suggest that it is (iv) difficult to evaluate the occurrence of minor neck injuries using car impact speed only. Only a few fundamental studies are available on the mechanism of soft cervical tissue injuries sustained in low impact speed collisions. In these circumstances, the issues which require further study may be summarized as follows: More detailed investigations and analyses are necessary regarding questions of “types of occupants (age and sex) who have sustained neck and “types of injuries”, their “sitting postures” neck injuries” sustained in low impact speed collisions. (i) For the clarification of injury mechanism of cervical soft tissues, it would be necessary to develop appropriate methods for analysis of the injury mechanism, taking into account the influences of occupant sitting posture, cervical muscular forces, and especially the reaction of the cervical spine. (i)
2.
3.
4.
ONO and M. KANNO
(ii)
Experimental studies using human subjects (which allow safe experience of collisions) are clearly limited. Considering the present state of dummy development, it will be necessary to study and develop dummies capable of simulating human head-neck joint characteristics more closely, in order to analyze low impact speed collisions more closely.
REFERENCES Ewing, C.; Thomas, D. et al. Living human dynamic response to -Gx impact acceleration-II. Accelerations measured on the head and neck, Proceedings of 12th Stapp Car Crash Conference. SAE Paper No.690817; 1969. Desantis, D. Development of an improved multi-directional neck prototype. SAE Paper No.912918; 1991. Foret-Bruno, J.; Dauvilliers, F.; Tarriere, C. Influence of the seat and headrest stiffness on the risk of cervical injuries in rear impact. 13th ESV Conference; 1991. Foust, R. et al. Cervical range of motion and dynamic response and strength of cervical muscles. Proceedings of 17th Stapp Car Crash Conference. SAE Paper No.730975; 1973: 285-308. Handbook of whiplash injuries-Toward a comprehensive understanding of whiplash injuries. Japan: Marine and Fire Insurance Association of Japan; March 1987 (Japanese). Melvin, J. et al. Improved neck simulation for anthropometric dummies. SAE Paper No.720958; 1972: 45-60. Mertz, H. Injury assessment values used to evaluate Hybrid III response measurement. NHTSA Docket 74-14, Notice 32, Enclosure 2 of Attachment I of Part III of General Motors Submission USG 2284; March 1984. Mertz, H.; Patrick, L. Investigation of the kinematics and kinetics of whiplash. Proceedings of 11th Stapp Car Crash Conference. SAE Paper No.670919; 1967: 267-317. Mertz, H.; Patrick, L. Strength and response of the human neck. Proceedings of 15th Stapp Car Crash Conference. SAE Paper No.710855; 1971: 207-255. Mertz, H.; Irwin, A.; Melvin, J.; Stalnaker, R.; Beebe, M., Size, weight and biomechanical impact response requirements for adult size small female and large male dummies. SAE Paper No. 890756; 1989. Nosaka, Y.; Ono, K.; Kanno, M. Human cervical bending characteristics-a preliminary report on the measurement methods of dynamic bending moment acting on the cervical spine and responses of the electromyogram on the cervical tissues. The 30th Academic Research Forum of the College of Engineering, Nihon University; December 1987 (Japanese). Nosaka, Y.; Ono, K.; Kanno, M. Human cervical bending characteristics due to external dynamic forces. Kanto District Conference of Human Engineering Society of Japan; December 1988 (Japanese). Olsson, I.; Bunketorp, 0.; Carlsson, G.; Gustafson. C.; Planath, I.; Norin, H.; Ysander, L. An in-depth study of neck injuries in rear end collisions. 1990 IRCOBI Conference; 1990. Ono, K. The analysis of physical parameters involving the
Risk of neck injuries
in low impact
human and neck system and the future theme. Satellite Symposium of the 21th Annual Meeting of the Japan Spine Research Society; 4-8 June 1992 (Japanese). Ono, K.; Kanno, M. Influences of the pysical parameters on the risk to neck injuries in low impact speed rear-end collisions. 1993 IRCOBI Conference; 1993: 201-212. Ono, K. et al. Measurement method using 6-degree of freedom on the dummy head and neck system. Proceeding of The Japan Society of Automobile Engineering, No.14; 1978. Ono, K. et al. The analysis and measurement of physical parameters involving the human head and neck systemfor low speed rear-end car collision. The 4th Bioengineering Symposium, The Japan Society of Mechanical Engineers; July 1992 (Japanese). Statistical Analysis Results on the Compulsory Automobile
speed rear-end
collisions
499
Liability Insurance. Japan: Automobile Insurance Rating Association of Japan; 1992. Svensson, M.; Lovsund, P. A dummy for rear-end collisions-development and validation of a new dummyneck. IRCOBI Conference; 1992: 299-310. Viano, D. Influence of seatback angle on occupant dynamics in simulated rear-end impact. SAE Paper No.922521; 1992. Warner, C.; Stother, C.; James, M.; Decker, R. Occupant protection in rear-end collisions: II. The role of seat back deformation in injury reduction. SAE Paper No.912914; 1991. Walker, L. et al. Mass, volume, center of mass, and mass moment of inertia of head, and head and neck of human body. Proceedings of 17th Stapp Car Crash Conference. SAE Paper No.730985; 1973: 525-537.
Pergamon
Printed in Great Britain. All rights reserved OOOl-4575/96 $15.00 + 0.00
INFLUENCES OF THE PHYSICAL PARAMETERS ON THE RISK TO NECK INJURIES IN LOW IMPACT SPEED REAR-END COLLISIONS * KOSHIRO
ONO’
and
MIJNEKAZU KANNO~
‘Japan Automobile Research Institute, Karima, Tsukuba, Ibaraki 305, Japan; 2College of Engineering of Nihon University, Tokusada, Tamura, Koriyama, Fukushima 963, Japan (Accepted 28 January 1996)
Abstract-The current state of neck injuries sustained in car-to-car rear-end collisions were investigated according to recent automobile accident statistical data in Japan. To clarify the neck injury mechanisms for low impact speed car collisions, the newly developed impact sled experiment which simulates actual car impact acceleration was performed using human subjects.? In order to measure and analyze the physical parameters such as human head rotational acceleration, neck bending moment, shearing and axial forces, the component measurement method with six-degrees of freedom was applied and demonstrated. Furthermore, relationships among the physical parameters, impact speeds, sitting positions, headrest heights and neck muscle tones applied on the subject’s head and neck system were analyzed. These analyses would enable us to comprehend the conditions of the neck muscle tone and the effects of the sitting postures including headrest height, factors which are of vital importance to the understanding of neck injury mechanisms. Copyright 0 1996 Elsevier Science Ltd Keywords-Injury mechanism, Cervical injury, Whiplash injury, Bending moment, Shearing force, Axial force, Sitting position, Muscle tone out for relatively high impact speed ranges of 20 km/h or higher (Desantis 1991; Warner et al. 1991; ForetBruno et al. 1991; Viano 1992; Ewing et al. 1969; Mertz 1984; Mertz et al. 1989). It is reported that minor neck injuries at low impact speeds often involve symptoms of soft tissue injuries called cervical contusions, sprains or distortions, which lack clear objective signs but tend to have prolonged healing times Olsson et al. 1990; Handbook of Whiplash Injuries 1987). The number of such neck injuries is markedly greater than the number of other types of injuries (Handbook of Whiplash Injuries 1987). Only a small number of studies (Mertz and Patrick 1971; Svensson and Lovsund 1992; Foust et al. 1973; Melvin et al. 1972; Ono et al. 1992; Nosaka et al. 1987, 1988; Ono 1992) are found on fundamental aspects of the mechanism of those injuries, due to the complexity of symptoms and the lack of sufficient objective symptoms. As a first step towards the clarification of neck injury mechanisms in such accidents, a sled device, capable of simulating rear-end car collision tests at low impact speeds, was specially developed wherein test subjects could experience actual collisions without any risks. Each human subject was seated on a sled in order to analyze the impact response acting on the area from the head to the neck. This is done by
INTRODUCTION The so-called “whiplash” neck injuries sustained in car rear-end collisions were gradually reduced owing to the effectiveness of the headrest, etc. introduced since the latter half of the 1960s. According to recent insurance statistical data on automobile accidents in Japan, however, approximately 50% of car-to-car traffic accidents have resulted in neck injuries, particularly at low impact speeds. The number of such accidents has increased recently, and neck injuries sustained in rear-end collisions in particular, have shown a major increase again. Studies on neck injuries using human/animal cadavers or volunteers have been carried out, mainly in the USA, since the 1960s (Mertz and Patrick 1967, 1971). More recently, studies employing HY-III dummies have been carried out partly as a result of neck injury criteria being reintroduced as one item to be evaluated to ensure automobile safety in collisions. All of these experimental studies have been carried
*Presented at the 1993 IRCOBI Eindhoven, Netherlands tin this paper, the term “human
Conference, subjects”
September
S-10,
refers to volunteers. 493
K. ONO and M.
494
Rtnnberof 0
100.000
neck
injuries
zoo. 000
300.000
400.000
KANNO
Table 2. Distribution of injured body regions of occupants involved in rear-end collisions
Year 85
(43.9%)
’ 86
(44.8%)
.a7
(47.3%)
. aa
(4-I.‘I%)
Neck
.a9
(47.9%)
Chest/abdomen
90
(48. ID
Extremities
‘91
(50.8%)
Others
Numberof neck injuries Numberof 111other injuries
Head/face
Total
Rear-end collisions (car-to-car)
All other collisions
40,127 (13.5%) 230,594 (77.5%) 11,290 (3.8%) 11,660 (3.9%) 3,746 (1.3%) 297,417 (100.0%)
235,513 (25.4%) 355,703 (38.3%) 97,127 (10.5%) 221,734 (25.2%) 11,453 (1.2%) 927,530 (100.0%)
Fig. 1. Number of neck injuries of occupants involved in all types of car-to-car collisons.
applying a method that measures components using six-degrees of freedom in measuring head accelerations. Moreover, various physical parameters (impact speed, sitting posture, headrest height and degree of occupant’s tension) that influence the impact on the subject’s head to neck area were analyzed. CONDITIONS OF NECK INJURIES SUSTAINED IN REAR-END CAR COLLISIONS IN JAPAN
distortions account for approximately 95% of such injuries (Table 3). It is reported that 1520% of the casualties above suffer from prolonged symptoms (Statistical Analysis Results on the Compulsory Automobile Liability Insurance 1992). It may be said that the trend showing the increase of automobile accidents involving such neck injuries is inevitably due to the increased number of vehicles in use. In Japan, however, the use of a headrest has been mandatory since 1969, and the use of seat belts obligatory since 1986. Considering the fact that vehicle safety has been remarkably enhanced compared to 20 years ago, the obvious question arises as to why such minor neck injuries, especially those having prolonged symptoms, cannot be reduced. In this regard, a clarification of the mechanisms of neck injuries in such accidents is urgently required.
According to the automobile accident insurance statistical data in Japan (Statistical Analysis Results on the Compulsory Automobile Liability Insurance 1992), minor neck injuries account for 50% or so of car-to-car accidents, a fact which tends to increase gradually each year (Fig. 1). Looking at the statistical data of 1991 alone, approximately 95% of injuries sustained in car-to-car rear-end collisions were minor injuries of AIS 1 (Table l), and about 80% of the above are concentrated on the neck of the occupant (Table 2). The so-called “whiplash” minor neck injuries involving cervical contusions, sprains and/or
The six-degrees of freedom component measuring method (Ono et al. 1978; Ono and Kanno 1993) per eqn (1) is applied to the measurement of head
Table 1. Distribution of injuries to occupants involved in rear-end collisions (1991)
Table 3. Distribution of types of injuries of occupants involved in rear-end collisions (1991)
AIS 1 AIS 2 AIS 3-6 Unknown
Rear-end collisions (car-to-car)
All other collisions
279,688 (94.0%) 10,264 (3.5%) 3,086 (1.0%) 4,383
719,091 (77.5%) 141,774 (15.3%) 55,562 (6.0%) 11,103 ( 1.2%) 927,530 (100.0%)
(1.5%) Total
297,417
(100.0%)
HEAD ACCELERATION MEASUREMENT AND ANALYSIS OF NECK RESPONSES
Contusion/Sprain Laceration/Avulsion Fracture/Dislocation Others Total
Rear-end collisions (car-to-car)
All other collisions
381,125 (94.5%) 5,493 (1.8%) 6,143 (2.1%) 4,636 (1.6%) 297,417 (100.0%)
666,051 (71.8%) 103,315 (11.1%) 144,523 (15.6%) 13,641 (1.5%) 927,530 (100.0%)
Risk of neck injuries in low impact speed rear-end collisions
acceleration.
where CQis No.k accelerometer, pro is the translational acceleration, 6, is the rotational acceleration at the center of gravity, 0, is the rotational velocity, fk is the location of the No.k attached accelerometer and & is the direction of the No.k attached accelerometer. The head movement in this experiment can be assumed as the two-dimensional movement of the X-Z plane, and the 4-channel measuring method is applied to measure the head acceleration. Components of forces and moments acting on the upper portion (Cl) of the neck can be determined by the application of the component measurement with six-degrees of freedom. Forces and moments at the head’s center of gravity per application of the method to this experiment can be expressed as follows. fi = m&o-fo
(2)
T,=IO,+~ox(Ioo)-T,
(3)
where f, and T, represent the headrest reaction force and the moment, respectively. According to eqns (2) and (3), the forces and the bending moments around the upper portion of the neck can be calculated as follows. J‘N= RJ; = Rf T,=RR=RR+R~&
where R and I represent the coordinate matrix.
(4) (5)
convention
EXPERIMENTS Car collision simulation sled experiment
Based on the experimental data of the two cars, a sled capable of simulating the impact acceleration profile of the struck car was produced (Ono and Kanno 1993). As for the crush characteristic that dominates the sled impact acceleration, coil springs capable of maximum approximation of the impact acceleration-time history waveform of the struck car were used. The seat system used in the experiment was the same as those used in the experimental cars. Conditions and objectives of experiments
Four kinds of physical parameters--(l) impact speed, (2) seatback angle of the occupant’s sitting posture, (3) headrest height and (4) degree of occupant’s tensed condition (neck muscle tonekwere
495
used as parameters that would affect the human headneck responses. Four series of experiments were set up, using a combination of above individual experimental conditions (Ono and Kanno 1993). In the series of experiments, three male human subjects (volunteers) with ages ranging from 22 to 43 wore seat belts and sat in the normal sitting posture for the measurement of the head accelerations, sled impact speeds and occupant motion, etc. Impact speeds were set at 2, 3 and 4 km/h, respectively, for frontal and rearward collisions. For the degree of the occupant’s neck tension electrodes, attached to the skin surface of the muscles of the trapezius and the sternocleidomastoideus on both sides of the neck, were connected to the electromyogram (EMG), which was monitored.
RESULTS
OF EXPERIMENTS DISCUSSION
AND
Head acceleration measurement and neck impact responses
Head dimensions and locations of accelerometers installed on each subject were determined by means of a dimensional measurement using X-ray films. The center of gravity, mass and moment of inertia of the head required for the determination of head-neck responses were obtained from the dimensions determined by the X-ray films and the values reported by Walker (1973). The same values were applied equally to the three human subjects. The head rotational angle was obtained by integrating the rotational accelerations, and the maximum rotational angle and its duration were measured by a high speed video camera in order to minimize any integration error. An analytical comparison was made between the two values. Figure 2 shows an example of an analysis done on rearward collisions at three levels of sled impact speeds. The head slightly rotates clockwise around 50 ms where the acceleration of the sled reaches its maximum value-that is, where “whiplash” starts. The head starts rotating rearward (counterclockwise) around 100 ms where the sled acceleration disappears, but the bending moment and shear force of the neck reach maximum around 110 ms which is slightly after the point of looms. The neck response approaches zero around 200 ms where the head rotational angle reaches maximum. The head-neck motion patterns show similar tendencies in other cases as well. In the case where the impact speed is 2 km/h, however, the head-neck movement delays slightly as the subject’s riding posture is inclined forward slightly.
K. ONO and M. KANNO
496
------
100 -
Sled
-
- -
- -
2.2 km/h 3.1 km/h 4.0 km/h
Extension
Sled speeds
-Standard
G? 0_-_\_ 2
0/*
I
L
-.-.l_&lw Headrest 3 ------Hlthout Sled speed 4 krmh
I w
-
B z
z 8 2
-100 50 -
Neck bending
I
+
:
1
moment
0
I
-5o500 -
Neck shear force-FX
-10
Neck axial force-FZ
0
20
Head
Fig. 3. Head-neck
0
100
200
300
Time (ms) Fig. 2. Head-neck
reponses of 2, 3,4 km/h (without headrest).
sled runs for extension
Eflect of diflerences in headrest height on neck impact responses The relationship between the neck bending moment and the head rotational angle is shown in Fig. 3, using different types of headrest-i.e. “standard”, “low” and “without headrest”. Although the bending moment level in the figure is markedly diverse in the three different conditions, it also shows the neck bending moment of a cadaver (without headrest), which constitutes the basic corridor for the neck injury threshold currently proposed (Mertz and Patrick 1971). In the experiment, a sharp rise of bending moment is found in the range where the rotational angle is small, and the so-called “whiplash” movement is observed but is reduced as the rotational angle is increased after the bending moment has reached its maximum level. This is presumably because the cervical muscular force could not resist the impact load. In cases where a cadaver is used, such a sharp rise is absent, the rise being gentle instead. This would be due to the fact that the cadaver’s cervical muscle had hardly any resistance against the impact, while the cervical muscle of the
40
rotational
60
angle
e0
188
(deg.)
responses by the different types of headrest compared to tests using cadavers.
as
human subject resisted the impact. Such a resistive force of a cervical muscle may place an excessive burden on supportive soft tissues (ligament, muscle, intervertebral joint envelop, etc.) of the cervical spine. The head rotational angle is minimized where the “standard” headrest is used, while it becomes larger where the “low” headrest is used, as the bending moment becomes greater. This shows that an appropriate adjustment of headrest height is very important for the prevention of excessive rearward bending of the neck and to reduce the impact load. Parameters that injluence neck impact responses Neck injuries in general and those sustained by the hyper-extension or hyperflexion of the cervical spine in particular have been discussed in terms of neck bending moment and head rotational angle. The responses acting on the neck, however, are not affected by the bending moment alone, but also by the shear force and axial force (compressive force and tension). In this study, the headrest height and seatback inclination angle (sitting posture) are analyzed in terms of three parameters-the neck bending moment, shear force and axial force. Injluence of headrest height. The highest impact response is found with the low headrest, resulting in an excessive load on the neck. In the case where no headrest is installed, the load on the neck is lowest, but the head rotational angle is largest, resulting in cervical hyper-extension. In the case where the “standard” headrest height is used, both the bending moment and the axial force are lowest, but the shear
Risk of neck injuries in low impact speed rear-end collisions
force may increase in proportion to the intensity of impact. This is presumably because the head impact is effectively suppressed by the headrest itself, and the reaction force acts directly on the neck and becomes greater. Therefore, it is reasonable to say that not only the neck bending moment and the head rotational angle but also the neck shear force and the axial force should be analyzed in relation to the neck responses. Influence of sitting posture (seatback angle reclined backward). Although no significant difference
is found in neck bending moment or shear force between the standard and reclined postures, a great difference is found in the axial force. This is presumably because the impact force in the reclined sitting posture is borne in the vertical direction of the neck. Therefore, it will be necessary to analyze not only the neck bending moment and the head rotational angle but also the shear force and the axial force in relation to neck loading conditions, as in the case of headrest height.
491
( km/h)
10 9 8 7 6 5,
4 3, 2 l-
Struck car speed after
1’
/ 01 ~“1’
2:
5-
1
10 15-
3’
4:
5
2 +
6
7 8
3
4
impact
9
10 ( kl,h)
Sled
5 impactspeed
10 'bl e I
20-
Sled lapact Tests
Wm)
Degree of influences of parameters on the risk to neck injuries
It was found that the impact response on the head-neck area was markedly affected by differences in the impact speed, sitting posture and headrest height. Hence, the relative degree of their influence was studied against the impact speeds in actual car experiments. In the study, the neck bending moment alone was used as the indication of the neck impact response. Figure 4 shows the relationship between the impact speed of the striking car and that of the struck car in actual low impact speed car experiments. The sled impact speed, corresponding to the struck car during the impact, is also plotted in the figure. The bending moment acting on the neck is also shown against each parameter. The bending moment acting on the neck is roughly proportional to the sled impact speed. However, the neck bending moment with the “low” headrest is about 50% higher than that of the “standard” headrest, in the case where the sitting posture is “standard”. In the case where the neck is in hypertension, the bending moment is about 40% lower than that of the “standard” headrest. The quantitative analysis of the degree of neck muscle tone is being done at present in relation to trapezius, muscle electromyogram sternocleidomastoideus (EMG) and neck impact response. The comparison between the standard sitting posture and reclined posture shows that the degree of neck responses
Fig. 4. Comparison between actual car impact tests and sled tests together with a comparison between neck bending moments and speeds of sleds under various conditions.
of the latter is about 30% higher than that of the former. Considering such findings, it is reasonable to say that the influence of differences not only in impact speed but also in sitting posture such as headrest height, backward inclination, and occupant’s tension (neck muscle tone) on the neck impact responses cannot be ignored in rear-end collisions at low impact speeds.
CONCLUSION 1. According to automobile accident insurance statistical data in Japan, approximately 50% of the car-to-car accidents involve minor neck injuries. Looking at the car-to-car rear-end collisions alone, neck injuries of AIS 1 account for 80%, and cervical contusions, sprains or distortions account for 95% or so in terms of type of injury.
K.
498
Studies on those minor neck injuries with substantial engineering approaches were few, due to the lack of sufficient objective symptoms associated with such injuries. In this regard, experimental studies have been done by simulating car rear-end collisions at low impact speeds using human subjects, for the determination of influences of headrest height, sitting posture, etc. Although the rigidity of the seatback and headrest have significant influences on the neck responses, they are not considered in this study. The following findings have been obtained: The neck bending moment shows a sharp rise where the head rotational angle of the human subject is small. This is due to the resistance force of the cervical muscular force of the subject, which may cause an excessive burden on the soft tissues of the cervical spine. 6) Not only the neck bending moment and head rotational angle but also the neck shear force and axial force affect the neck responses, according to differences in headrest height and sitting posture. (ii) The bending moment that acts on the neck is roughly proportional to the sled impact speed. However, the neck bending moment may differ by 50% or so due to differences in headrest height and sitting posture. (iii) In a case where the neck is in hypertension, the neck bending moment reduces by 40%, while it increases by 30% where the seatback is inclined backwards. The foregoing findings suggest that it is (iv) difficult to evaluate the occurrence of minor neck injuries using car impact speed only. Only a few fundamental studies are available on the mechanism of soft cervical tissue injuries sustained in low impact speed collisions. In these circumstances, the issues which require further study may be summarized as follows: More detailed investigations and analyses are necessary regarding questions of “types of occupants (age and sex) who have sustained neck and “types of injuries”, their “sitting postures” neck injuries” sustained in low impact speed collisions. (i) For the clarification of injury mechanism of cervical soft tissues, it would be necessary to develop appropriate methods for analysis of the injury mechanism, taking into account the influences of occupant sitting posture, cervical muscular forces, and especially the reaction of the cervical spine. (i)
2.
3.
4.
ONO and M. KANNO
(ii)
Experimental studies using human subjects (which allow safe experience of collisions) are clearly limited. Considering the present state of dummy development, it will be necessary to study and develop dummies capable of simulating human head-neck joint characteristics more closely, in order to analyze low impact speed collisions more closely.
REFERENCES Ewing, C.; Thomas, D. et al. Living human dynamic response to -Gx impact acceleration-II. Accelerations measured on the head and neck, Proceedings of 12th Stapp Car Crash Conference. SAE Paper No.690817; 1969. Desantis, D. Development of an improved multi-directional neck prototype. SAE Paper No.912918; 1991. Foret-Bruno, J.; Dauvilliers, F.; Tarriere, C. Influence of the seat and headrest stiffness on the risk of cervical injuries in rear impact. 13th ESV Conference; 1991. Foust, R. et al. Cervical range of motion and dynamic response and strength of cervical muscles. Proceedings of 17th Stapp Car Crash Conference. SAE Paper No.730975; 1973: 285-308. Handbook of whiplash injuries-Toward a comprehensive understanding of whiplash injuries. Japan: Marine and Fire Insurance Association of Japan; March 1987 (Japanese). Melvin, J. et al. Improved neck simulation for anthropometric dummies. SAE Paper No.720958; 1972: 45-60. Mertz, H. Injury assessment values used to evaluate Hybrid III response measurement. NHTSA Docket 74-14, Notice 32, Enclosure 2 of Attachment I of Part III of General Motors Submission USG 2284; March 1984. Mertz, H.; Patrick, L. Investigation of the kinematics and kinetics of whiplash. Proceedings of 11th Stapp Car Crash Conference. SAE Paper No.670919; 1967: 267-317. Mertz, H.; Patrick, L. Strength and response of the human neck. Proceedings of 15th Stapp Car Crash Conference. SAE Paper No.710855; 1971: 207-255. Mertz, H.; Irwin, A.; Melvin, J.; Stalnaker, R.; Beebe, M., Size, weight and biomechanical impact response requirements for adult size small female and large male dummies. SAE Paper No. 890756; 1989. Nosaka, Y.; Ono, K.; Kanno, M. Human cervical bending characteristics-a preliminary report on the measurement methods of dynamic bending moment acting on the cervical spine and responses of the electromyogram on the cervical tissues. The 30th Academic Research Forum of the College of Engineering, Nihon University; December 1987 (Japanese). Nosaka, Y.; Ono, K.; Kanno, M. Human cervical bending characteristics due to external dynamic forces. Kanto District Conference of Human Engineering Society of Japan; December 1988 (Japanese). Olsson, I.; Bunketorp, 0.; Carlsson, G.; Gustafson. C.; Planath, I.; Norin, H.; Ysander, L. An in-depth study of neck injuries in rear end collisions. 1990 IRCOBI Conference; 1990. Ono, K. The analysis of physical parameters involving the
Risk of neck injuries
in low impact
human and neck system and the future theme. Satellite Symposium of the 21th Annual Meeting of the Japan Spine Research Society; 4-8 June 1992 (Japanese). Ono, K.; Kanno, M. Influences of the pysical parameters on the risk to neck injuries in low impact speed rear-end collisions. 1993 IRCOBI Conference; 1993: 201-212. Ono, K. et al. Measurement method using 6-degree of freedom on the dummy head and neck system. Proceeding of The Japan Society of Automobile Engineering, No.14; 1978. Ono, K. et al. The analysis and measurement of physical parameters involving the human head and neck systemfor low speed rear-end car collision. The 4th Bioengineering Symposium, The Japan Society of Mechanical Engineers; July 1992 (Japanese). Statistical Analysis Results on the Compulsory Automobile
speed rear-end
collisions
499
Liability Insurance. Japan: Automobile Insurance Rating Association of Japan; 1992. Svensson, M.; Lovsund, P. A dummy for rear-end collisions-development and validation of a new dummyneck. IRCOBI Conference; 1992: 299-310. Viano, D. Influence of seatback angle on occupant dynamics in simulated rear-end impact. SAE Paper No.922521; 1992. Warner, C.; Stother, C.; James, M.; Decker, R. Occupant protection in rear-end collisions: II. The role of seat back deformation in injury reduction. SAE Paper No.912914; 1991. Walker, L. et al. Mass, volume, center of mass, and mass moment of inertia of head, and head and neck of human body. Proceedings of 17th Stapp Car Crash Conference. SAE Paper No.730985; 1973: 525-537.