In This Issue
Biotechnology Revolution
September 1, 2004 Volume 34 Issue 3

Impact Biomechanics

Wednesday, September 1, 2004

Author: Albert I. King

The goal of impact biomechanics is to protect vehicle occupants from serious injury.

Vehicle safety is on the minds of many would-be car buyers these days, and most of us have either been injured in a crash or know someone who has been injured or even killed in a collision. Impact biomechanics, the science of injury control, is research dedicated to injury prevention through environmental control. The goal of impact biomechanics is to protect vehicle occupants from serious injury at a price the general population can afford. The research is based on the principles of mechanics and an understanding of the physiology and pathophysiology of the human system. Research ranging from the testing of whole-body cadavers to the study of flux of ions across a cell membrane has been going on for at least 65 years.

The study of accidents began in the 1920s when Hugh DeHaven of Cornell set out to understand how people survive falls from great heights. De Haven, a combat pilot in World War I, had survived a free fall from a great height (Andreasson and Backstrom, 2000). Laboratory research began in 1939 at Wayne State University when Steve Gurdjian, a neurosurgeon, and Herbert Lissner, a professor of engineering mechanics, initiated a study on head injuries and skull fractures using anesthetized dogs and cadaveric skulls (Gurdjian and Lissner, 1944). Research on impact biomechanics is now being done in many countries around the world. Centers of biomechanics research include the Bioengineering Center at Wayne State, one of the most versatile research centers in the world, and research laboratories at the University of Michigan, the University of Virginia, the Medical College of Wisconsin, Duke University, and the University of Pennsylvania; new laboratories are being established at the University of California San Francisco, Virginia Polytechnic Institute and State University, and Kettering University.

Because research is still mostly on the macroscopic scale, impact equipment must be able to generate sufficient force or acceleration to cause injury. Wayne State has two pneumatically driven, full-scale impact sleds. The HYGE sled is accelerated from a standstill, and the impact occurs as soon as the sled starts to move. To simulate a frontal impact, the subject faces rearward. The second sled is a deceleration sled that is accelerated up to speed and stopped by a hydraulic snubber to produce the impact. With either sled, the impact pulse can be designed to simulate a certain vehicle or type of crash. Other impact equipment can include linear impactors, which accelerate a moving mass into a body region, or minisleds, which can be transported to a remote facility where a high-speed x-ray unit can be used to measure the motion of internal organs during impact.

The most common parameter measured is body-fixed acceleration of the regions of interest. Other parameters include the force of impact, pressure, strain, displacement, and stretch. In addition, angular acceleration, a parameter thought to be a major cause of brain injury, can now be accurately measured. Before the days of 3-D angular accelerometers, angular acceleration was measured using uniaxial linear accelerometers. For a rigid body, a minimum of six accelerometers are necessary to measure the three linear and three angular components of accelerations. However, the equations used to calculate angular acceleration were in the form of nonlinear differential equations that require a numerical solution. In theory, this method worked fine with six accelerometers, but errors in the measured accelerations tended to magnify errors in the computed acceleration. Thus, if the integration process went on for any period of time, the solution could be far off the mark, especially if low-sensitivity accelerometers were used to measure small accelerations.

In 1975, Padgaonkar et al. developed a new method using nine accelerometers placed on a mount, which has the form of a rectangular Cartesian coordinate system. There were two accelerometers at the end of each axis, normal to that axis, and three at the origin. With this 3-2-2-2 arrangement, the equations become algebraic, thus eliminating the need for integration. Even with the advent of the angular accelerometer, the 3-2-2-2 method is still considered the gold standard against which transducers are compared. This method is just one contribution of biomechanics to applied mechanics. Perhaps in the future, we will find an elegant method of using six accelerometers to measure angular acceleration accurately.

Basic Principles
There are four basic areas of research in impact biomechanics: injury mechanisms; mechanical response; injury tolerance; and simulation of human impact. Cadavers and animals are used to obtain data in the first three areas, but these surrogates are not perfect. For one thing, cadavers lack muscular response, although this does not usually cause serious problems. However, scaling data from animals to humans is prone to error. Tests with volunteers are generally not severe enough to study injury.

Injury Mechanisms
We must understand how an injury occurs before we can find a way to prevent it. A case in point is the so-called whiplash injury, which can cause intractable neck pain in some people, even after a very minor impact. (Whiplash is discussed in more detail below.)

Mechanical Response
To design and build realistic surrogates, we must first quantify the mechanical responses of various body regions. The crash dummies seen on TV are finely tuned measuring instruments that can indicate to a safety engineer if a body region is likely to be injured in a crash. The response data are also used in the development of computer models that simulate an occupant involved in a crash.

Human Tolerance
Understanding the tolerance level of every body region is crucial for the design engineer. However, human tolerance varies greatly with age and gender. To narrow the range, the tolerance is usually defined for a 50th-percentile, middle-aged male. This means that elderly people, women, and children may be less well protected, and efforts are under way to change government standards to consider tolerance levels of women and children. Information on the responses and tolerances of children is sparse because child cadavers are not readily available.

Simulation of Human Impact
Automobile manufacturers use crash dummies as human surrogates for evaluating safety systems in their vehicles. The Hybrid III family of dummies is figure 2 (see PDF version). The 50th-percentile male dummy, which was designed using frontal impact cadaveric data, has relatively good human-like responses for frontal impact simulations. For side impact simulations, we use side impact dummies, but none of them is really biofidelic at present.

Auto manufacturers do not use cadavers for two reasons. First, there are obvious ethical problems and availability problems associated with their use by a commercial enterprise. Second, cadaveric responses and tolerances are too variable for an efficient evaluation of vehicle safety systems. Thus, human models that simulate human occupants are becoming more popular. The current slogan is “Make the car safe for people, not for dummies.” Rapid increases in computing power and advances in finite element analysis have made it possible to simulate the entire human occupant. The whole body THUMS model developed by Toyota a few years ago is being evaluated by safety professionals around the world (Iwamoto, et al, 2002). Yang (2001) has reviewed all finite element models of the human in the literature.

Before the advent of air bags, 50 percent of automotive fatalities resulted from head injuries. Since then, the fatality rate has declined, but minor and moderate head injuries continue to be a problem. Even mild head injuries can have devastating consequences, such as memory loss, inability to concentrate and process information, increased irritability, and clinical depression. The current Federal Motor Vehicle Safety Standards use a head injury criterion (HIC) limit of 1,000. According to best estimates, the probability of sustaining a life-threatening injury at a HIC of 1,000 is 15 percent (Prasad and Mertz, 1985). We do not have a widely accepted tolerance for a mild traumatic brain injury (MTBI). The HIC has been controversial ever since it was introduced in the early 1970s, because it does not account for the effects of angular acceleration, which is believed by many to be the principal cause of brain injury (Gennarelli et al., 1972; Ommaya and Hirsch, 1971). In fact, Ommaya used animal data to deduce the level of human tolerance for severe brain injury at 1,800 rad/s2.

Recently, researchers at Wayne State University put together the results of several studies and came up with predictors of brain injury (King et al., 2003). First, with the help of a high-speed biplanar x-ray device, the motion of the brain was quantified for mild to moderate impacts, involving a peak linear acceleration of 100 g and a peak angular acceleration of several thousand rad/s2 (Hardy et al., 2001). Figure 3 (see PDF version) shows figure-eight motion patterns of several radio-opaque targets implanted in a cadaveric brain. Brain displacement was ?1 mm due to linear acceleration and as much as ?5 mm due to angular acceleration. However, regardless of the magnitude of the angular acceleration (in excess of 10,000 rad/s2), the displacement did not exceed ?5 mm. Also, the motion was larger near the center of the brain than near the periphery. Another experiment showed that wearing a helmet decreases linear acceleration by more than 20 percent but does not appreciably decrease angular acceleration. Thus, if angular acceleration is the cause of brain injury, how does the helmet protect the brain? We still do not have a satisfactory answer to this question.

A third experiment by Biokinetics, Inc. of Ottawa, Canada, involves the re-creation of head impacts sustained by NFL players as recorded on video game films. Stereophotogrammetric techniques were used to estimate the velocity impact between helmeted heads and between the helmeted head and the ground. Newman et al. (1999) duplicated the impacts using crash dummies in the laboratory to measure the linear and angular accelerations of both players (one of whom usually suffered a concussion) or of a single player for ground impact. The acceleration data were then fed into a finite element model of the brain to compute brain response, including strain (e), strain rate (de/dt) and the product of strain rate (e•de/dt) (Zhang et al., 2001). These and other response variables, as well as input variables, were subjected to a Logist analysis to determine the strongest predictor of concussion (MTBI), which was found to be e•de/dt, followed by de/dt. Surprisingly, HIC was third, and in the most recent analysis of all available cases, angular acceleration was ninth. These studies show that brain response is a better predictor of injury than input and that a computer model of the brain is essential for understanding brain injury. We also found that the value of HIC to produce a 50-percent probability of an MTBI was 235 and that the average value for angular acceleration to cause an MTBI was 6,400 rad/s2. Thus, the dangers of scaling animal data are obvious.

Another unsolved problem in impact biomechanics is whiplash. Many hypotheses have been proposed to explain neck pain following an accident, the most logical of which appears to be shear in a rear-end impact. The seat back pushes the torso forward leaving the head behind unless a shear force is generated at every level of the cervical spine to bring it along. Experiments carried out by Deng et al. (2000) using the same high-speed x-ray system showed that there is substantial relative motion between the adjacent cervical vertebra, both in translation and rotation, and that the estimated stretch in the facet capsules, particularly in the lower cervical spine, were extremely high, reaching a peak of almost 100 percent. Although it has long been established clinically that facet capsules are a source of neck pain (Wallis et al., 1997), the majority of physicians do not believe in facet pain. Therefore, the controversy continues.

Despite the uncertainty about the cause of whiplash injuries, several attempts have been and are being made to redesign car seats and head rests to prevent or minimize whiplash injuries. Trying to prevent an injury without knowing its cause could potentially put the user at risk. Although engineers do not take the Hippocratic oath, they should obey the maxim—above all else, do no harm.

Importance of Biomechanics
Knowing that the results of my research may save lives and reduce injury is very rewarding. An estimated 12 lives have already been saved as a result of one cadaver test, and the results of that research will continue to save lives (King et al., 1995). Of course, cadaver testing is not the most glamorous job in the world, but young engineers can now use computer models of human responses to study impact. The spine model by Prasad and King (1974) predicted large compressive loads in the thoracolumbar spine when a belted individual was subjected to a frontal impact. At first, this was thought to be an anomaly of the model, but Begeman et al. (1973) showed that the model was correct when they fractured the spine of shoulder-belted cadavers in a series of horizontal sled tests. The explanation after the fact was simple. The thoracic spine is convex rearward, and when the restrained torso is loaded inertially in a frontal impact, it tends to straighten out the curved thoracic spine, causing it to push down on the lumbar spine.

Experimental attempts by several laboratories to cause aortic ruptures in cadavers have been unsuccessful, so Shah et al. 2001 developed a computer model to find the most likely scenario. It is hoped that an improved version of the model can be used as a guide to create this injury and determine the mechanism of injury.

The most obvious thing one can do to protect oneself against injury is to wear a seat belt. If the driver has to sit close to the steering wheel to reach the pedals, the seat should be leaned back as far as is comfortable so the driver is not too close to the air bag. All children should be in the back seat and should use child seats appropriate for their age and weight.
In addition, a word to a senator or congressman can sometimes do wonders to support research. Despite the common occurrence of automotive injuries, no large constituency is clamoring for support on Capitol Hill, and funding for injury research is a miniscule fraction of the federal research budget—one reason improvements in automotive safety have been slow and irregular. Even with air bags, more than 40,000 people are killed every year, and injury is the leading cause of loss of productive years of life, exceeding cancer and heart disease combined. Finally, the cost to society from injury is extremely high, not only in pain and suffering, but also in high taxes and insurance premiums.

Andreasson, R., and C.-G. Backstrom. 2000. The Seat Belt. Stockholm: Vattenfall, AB.
Begeman, P.C., A.I. King, and P. Prasad. 1973. Spinal Loads Resulting From -GDx Acceleration. SAE Technical Paper No. 730977. Warrendale, Pa.: SAE International.
Deng, B., P.C. Begeman, K.H. Yang, S. Tashman, and A.I. King. 2000. Kinematics of human cadaver cervical spine during low speed rear-end impacts. Stapp Car Crash Journal 44: 171–188.
Gennarelli, T.A., L.E. Thibault, and A.K. Ommaya. 1972. Pathophysiologic Responses to Rotational and Translational Accelerations of the Head. SAE Technical Paper No. 720970. Warrendale, Pa.: SAE International.
Gurdjian, E.S., and H.R. Lissner. 1944. Mechanism of head injury as studied by the cathode ray oscilloscope: preliminary report. Journal of Neurosurgery 1: 393–399.
Hardy, W.N., C.D. Foster, M.J. Mason, K.H. Yang, and A.I. King. 2001. Investigation of head injury mechanisms using neutral density technology and high-speed biplanar X-ray. Stapp Car Crash Journal 45: 337–368.
Iwamoto, M., Y. Kisanuki, I. Watanabe, K. Furusu, K. Miki, and J. Hasegawa. 2002. Development of a Finite Element Model of the Total Human Model for Safety (THUMS) and Application to Injury Reconstruction. Pp. 31–42 in Proceedings of the 2002 IRCOBI International Conference on the Biomechanics of Impact. Bron, France: IRCOBI Secretariat.
King, A.I., D.C. Viano, N. Mizeres, and J.S. States. 1995. Humanitarian benefits of cadaveric research on injury prevention. Journal of Trauma 38: 564–569.
King, A.I., K.H. Yang, L. Zhang, and D.C. Viano. 2003. Is head injury caused by linear or angular acceleration? Keynote Lecture. Pp. 1–12 in Proceedings of the 2003 IRCOBI International Conference on the Biomechanics of Impact. Bron, France: IRCOBI Secretariat.
Newman, J., M. Beusenberg, E. Fournier, N. Shewchenko, A.I. King, K.H. Yang, L. Zhang, J. McElhaney, L. Thibault, and G. McGinnis. 1999. A New Biomechanical Assessment of Mild Traumatic Brain Injury: Part I, Methodology. Pp. 17–36 in Proceedings of the 1999 IRCOBI International Conference on the Biomechanics of Impact. Bron, France: IRCOBI Secretariat.
Ommaya, A.K., and A.E. Hirsch. 1971. Tolerances for cerebral concussion from head impact and whiplash in primates. Journal of Biomechanics 4: 13–31.
Padgaonkar, A.J., K.W. Krieger, and A.I. King. 1975. Measurement of angular acceleration of a rigid body using linear accelerometers. Journal of Applied Mechanics 42: 552–556.
Prasad, P., and A.I. King. 1974. An experimentally validated dynamic model of the spine. Journal of Applied Mechanics 41: 546–550.
Prasad, P., and H.J. Mertz. 1985. The Position of the United States Delegation to the ISO Working Group 6 on the Use of the HIC in the Automotive Environment. SAE Technical Paper No. 851246. Warrendale, Pa.: SAE International.
Shah, C.S., K.H. Yang, W.N. Hardy, H.K. Wang, and A.I. King. 2001. Development of a computer model to predict aortic rupture due to impact loading. Stapp Car Crash Journal 45: 161–182.
Wallis, B.J., S.M. Lord, and N. Bogduk. 1997. Resolution of psychological distress of whiplash patients following treatment by radiofrequency neurotomy: a randomized, double-blind, placebo controlled trial. Pain 73: 15–22.
Yang, K.H. 2001. Review of mathematical human models for incorporation into vehicle safety design. International Journal of Vehicle Design 26: 430–441.
Zhang, L., K.H. Yang, R. Dwarampudi, K. Omori, T. Li, K. Chang, W.N. Hardy, T.B. Khalil, and A.I. King. 2001. Recent advances in brain injury research: a new human head model development and validation. Stapp Car Crash Journal 45: 369–394.

About the Author:Albert I. King is Distinguished Professor and Chair of the Department of Biomedical Engineering at Wayne State University.