Can crash test dummies really simulate human injuries?

How much do crash test dummies really contribute to crash research?
How much do crash test dummies really contribute to crash research?
©iStockphoto/Jester Caps Photography

We seldom think about how fragile life is until we experience a traumatic event, such as a car accident. Your life could very well hang in the balance of what happens in just a few milliseconds, and the outcome may depend on the construction of your car. While no car is perfectly safe, engineers work to improve designs to minimize the potential of serious injuries.

To do this, auto manufacturers pour millions of dollars into test crashes to study what exactly happens in a crash with different car models. The point is to find out how dangerous a collision would be for a car's driver and passengers. But, of course, who would volunteer to be a human guinea pig in that kind of situation? Even in a controlled environment, it's far too dangerous to test crash a vehicle with human occupants. So, the important task falls to anthropomorphic test devices (ATDs), also known as crash test dummies. These are made in different sizes to imitate the range of a human family -- from infants to adults.

But engineers can't just settle for a simple human-shaped stuffed doll and call it a day. That's because a simple doll wouldn't be able to tell a researcher whether a crash resulted in a broken bone, a cracked rib cage or skin abrasions. Crash test dummies are becoming sophisticated enough to simulate such injuries. A modern ATD has such an advanced, detailed construction that one costs more than $100,000 -- though it lasts dozens of crashes.

Also, in a crash, the car may stop, but your body keeps moving. A person's injuries largely depend on how your body is thrown in the accident. Because of that, the dummy must not only have a realistic human weight in relation to its size, but the weight must be distributed just like a human's. This way, researchers can watch just how hard and quickly a 10-pound head hits an inflating airbag.

The difficulty in creating a sophisticated crash test dummy reminds us of the sheer complexity of the human body. Next, we'll explore the anatomy of a crash test dummy in more depth.

Modern Crash Test Dummies

The evolution of the crash test dummy dates back at least to 1949, when the U.S. Air Force used "Sierra Sam," a dummy developed by Sierra Engineering, to test ejection seats. In the 1970s, General Motors came out with the "Hybrid" dummy, which made several improvements on Sierra Sam. The Hybrid I came first in 1971, followed by Hybrid II in 1972; finally, the ATD still used today, the Hybrid III, appeared in 1976.

Hybrid III ATDs have skeletons of aluminum and steel, including six steel ribs with polymer-based material to imitate a real human chest, encased by vinyl imitation skin. Realistic joints as well as a neck, spine and pelvis made of rubber- or foam-encased metal constructions give a dummy lifelike posture and flexibility -- both of which play a large part in collision injuries.

Beyond its humanlike construction, Hybrid III dummies have extra features that range from simple to sophisticated. Merely smearing the dummies with grease paint allows researchers to see exactly where the dummy hits the car in the crash. Also, sensors inside the dummies measure forces of impact at different points.

The standard Hybrid III represents the 50th percentile male -- the average driver at 5-feet, 10-inches tall and weighing 168 pounds. Federal regulations stipulate the specifications for this ATD as well as the "family" of Hybrid III dummies. Among other things, having dummies of different sizes helps researchers determine the effectiveness of standard seat belts on various body types. In addition to the different Hybrid III dummies, there are also different types of ATDs for different crash tests. Hybrid III dummies are used primarily for frontal impact test crashes. But others include the side impact dummy (SID) and the biofidelic rear impact dummy (BioRID).

The next generation of ATDs is THOR, which has made many improvements on Hybrid III. In particular, THOR can more accurately predict facial injuries because the head is equipped with unidirectional load cells [source: Schmitt]. Other improvements include a new neck and flexible spine design and an advanced rib cage with elliptical ribs.

In recreating a controlled crash, researchers also film it with as many as 20 specialized cameras, which can film at high speeds (about 1,000 frames per second) at different angles [source: Weber]. This way they can watch the crash in clear slow motion to observe every detail.

Simulating Internal Human Injuries

Researchers will even dress dummies in clothes to make the situation as close to reality as possible.
Researchers will even dress dummies in clothes to make the situation as close to reality as possible.

It makes sense that watching a slow-motion video of a properly weighted, humanlike dummy could help researchers determine if a crash will result in external injuries. Perhaps less believable is that a dummy could tell us whether the crash will result in an internal injury. The human body and its sensitive internal organs are so complex that a mere polymer and metal construction seem an inadequate representative.

However, we have a good idea of what the human body can sustain and how much force will result in a critical internal injury. The trick is to find if a crash results in those particular forces.

if you remember from the last page, we mentioned that ATDs are equipped with sensors in various points of the body. These electronic sensors can actually measure the force of impact, accelerations and deformations that are sustained in the collision. In the Hybrid III, these are mounted in the head, neck, chest, pelvis, thighs, legs and ankles. In all, these sensors can record 37,200 different pieces of data in one ATD [source: Mello]. Knowing this information helps researchers to determine what forces will happen to a body and where, which can help them predict all different kinds of injuries. They can use this data to determine if something will likely be a flesh wound or an internal injury.

But another technology has also developed that can perhaps more accurately predict internal injuries. By simulating a crash and a dummy completely in a computer model, researchers believe they can determine whether an internal injury will be sustained. In particular, researchers at the Virginia-based National Crash and Analysis Center have been working on computerized crash tests. They say they can look at details of brain injuries, for instance, that are seen easily in computerized models [source: Science Daily].

Using Cadavers in Crash Testing

Although it may be strange to think about, using cadavers in crash tests gives researchers useful data that might save lives.
Although it may be strange to think about, using cadavers in crash tests gives researchers useful data that might save lives.

We mentioned earlier that the U.S. Air Force used the original crash test dummy in the 1940s. But, in a sense, the history of crash testing subjects date back even further. As early as the 1930s, Wayne State University researcher Lawrence Patrick wanted to test the limits of the human body. Instead of manufacturing a humanlike dummy, he actually used a human -- himself. He took on a 22-pound metal pendulum to the chest among other blows.

But when Patrick needed to test what happens when a body is thrown, he wanted to send a person down an elevator shaft. This time, he again used a human. But rather than himself, he used a dead person.

As gruesome, morbid or unethical as it may sound, using cadavers as crash test subjects makes a lot of sense from a practical standpoint. There is no danger to human life, and researchers can reap some of the most useful, life-saving information from cadaver tests. Patrick has claimed that despite the public outcry that this violated the dignity of the human body, he treated the dead bodies with respect [source: Roach].

The U.S. National Highway Traffic Safety Administration (NHTSA) funds crash tests with cadavers. And, although they don't like to admit it, auto manufacturers are believed to still use cadavers in tests [source: Hyde].

Cadaver testing isn't perfect, though. The available bodies are usually elderly, meaning bones break more easily than the average driver. Also, there is no pressure in the lungs and blood vessels of cadavers, among other problems. Living biological tissue is dramatically different from dead tissue just as it's different from synthetic ATD material. Even using cadavers, it's been difficult for researchers to simulate a pedestrian accident [source: LASTPEP].

In his cadaver crash tests, Patrick had to manipulate the bodies' joints to loosen them. He also took out the brain and put gelatin in its place, in addition to screw-mounting accelerometers to the head [source: Roach]. It's a controversial question as to whether such adjustments, in addition to smashing the bodies full force in a crash test, is inherently disrespectful to a body. It's no wonder auto manufacturers like to keep quiet about cadaver testing.

Related Articles


  • Foreman, Stephen M., Arthur C. Croft. "Whiplash Injuries." Lippincott Williams & Wilkins, 2001. (Nov. 28, 2010)
  • Haywood, D., M. Stonefrost. "The Influence of EuroNCAP on Frontal Impact Safety." "Proceedings of the 1st IMechE Automobile Division Southern Centre Conference on Total Vehicle Technology." John Wiley and Sons, 2001. (Nov. 28, 2010)
  • Hinckley, Jim, Jon G. Robinson. "The Big Book of Car Culture." MotorBooks International, 2005. (Nov. 28, 2010)
  • Hyde, Justin. "How a Cadaver Made Your Car Safer." Jalopnik. Updated Aug. 26, 2010. (Nov. 28, 2010)
  • LASTPEP. "Vehicle Safety, 2002." John Wiley and Sons, 2002. (Nov. 28, 2010)
  • Mello, Tara Baukus. "Meet the Crash Test Dummies." (Nov. 28, 2010)
  • Roach, Mary. "I Was a Human Crash-test Dummy." Nov. 19, 1999. (Nov. 28, 2010)
  • Roach, Mary. "Stiff: The Curious Lives of Human Cadavers." W.W. Norton & Company, 2004. (Nov. 28, 2010)
  • Schmitt, Kai-Uwe, et al. "Trauma Biomechanics." Springr, 2009. (Nov. 28, 2010)
  • Science Daily. "New Generation of Crash Test Dummies." Science Daily. May 1, 2008. (Nov. 28, 2010)
  • Weber, Julian. "Automotive Development Processes." Springer, 2009. (Nov. 29, 2010)