July 30, 2002
Holden demonstrates first computer brain validation crash test
Sydney, Australia – General Motors Australian division, Holden, has conducted the first in a series of world-leading crash tests to validate a computer brain model that will lead to the ‘virtual’ design of advanced car safety features.
Holden and its partners in this long-term exercise, the Monash University Accident Research Centre (MUARC) and Wayne State University in Detroit, USA, ran the first of five tests based on real life crashes at the Lang Lang proving ground last week.
The loads measured by the crash test dummies during these side impact crash re-enactments will be used as input to a computer brain model, and the injury prediction will be validated against the injuries that car occupants experienced in real world crashes. Researchers will examine the linear and rotational kinematics of the head in selected crashes where the drivers involved suffered varying levels of brain injury.
Holden’s Chief Engineer of Advanced Engineering, Dr Laurie Sparke, said the validation tests were a world first.
“They bring together three centres of expertise – Holden, Wayne State University and MUARC – each specialists in their respective fields. The information each of us is contributing will combine to create the world standard on the effects of car crashes on the human brain.”
King Yang, Professor of Mechanical and Biomedical Engineering at Michigan’s Wayne State University, agrees.
“We at Wayne State started developing a computer brain model in 1990, but computing power has only just reached the stage where we can create sophisticated models.
“We now need to validate our model against real world scenarios. With this series of crash tests, we are identifying five different severities of brain injury.
“We access the patients’ data in the hospital using the extensive collated MUARC crash data. Then we recreate the exact crash circumstances using the side impact expertise and advanced crash testing facilities at Holden. After that, we input and validate the data on the dummies to calibrate the computer models,” Professor Yang said.
“In the collision we have just re-enacted, the severity of the crash forces on the driver were quite high. However, there was no brain injury. We are therefore validating this as the baseline model. To recreate the impact and conditions of the real world crash exactly, we have copied the size and mass of the occupants, the mechanics and structure of the vehicles involved, the weight and speed at which they were travelling and the angle at which they hit,” he explained.
The computer brain model developed at Wayne State University has 300,000 pieces or elements that divide the entire human brain into 5 mm cubes. Different properties can be applied to each individual cube, depending upon whether they comprise grey or white brain matter, or lie in the right or left hemisphere, or in the brain stem.
Researchers can then determine the injuries likely to occur under any given real-world scenario, depending on the forces that are applied to various parts of the brain at various times during the crash. This model will then be validated against real world patient data provided by MUARC.
The brain model should prove to be a far more sophisticated tool than a test dummy, which is limited to measuring acceleration, usually in one direction. The computer brain model will be used for measuring omnidirectional loads and responses, simulating what happens in the real world.
“For instance, we can measure linear acceleration combined with rotational acceleration, which causes the types of combined brain injuries we commonly see. A person’s head doesn’t just move in one direction in a car crash; there are a number of forces that make it move in a variety of ways at different times,” Professor Yang added.
“Australia is the only place in the world where we can utilise accident data (MUARC), biomechanics (Wayne State University) and advanced crash test facilities and side impact expertise (Holden) together in the one place,” he said.
“One day we will be able to utilise a full human body model to measure the risk of injury to an individual, accounting for gender, size and age. In the future, safety engineers will use this computer model to design protection for the whole community of motor vehicle occupants,” Dr Sparke concluded.