Reconstructing real-world accidents

  • 19-Aug-2011 10:15 EDT
Figure 1.jpg

Vehicle interior surfaces were scanned using a FARO Arm to obtain precise vehicle dimensions and geometry.

The accident began when an SUV traveling on an interstate highway lost control and spun off the road and struck a tree. A young man in the front seat was ejected from his seat and ended up in the rear of the vehicle. Tragically, the injuries he suffered rendered him a paraplegic. Vector Scientific Inc. (VSI) was hired to reconstruct the accident to determine how and why the passenger suffered his injuries. During the course of the investigation, it was determined that the young man was not wearing the available lap and shoulder seatbelt. VSI was then asked to determine the extent to which the passenger’s injuries would have changed if he had been wearing the seatbelt.

Cases such as the one described are all too familiar to VSI, a research and forensic consulting firm specializing in injury biomechanics and accident reconstruction. The company’s expertise in the complementary disciplines of occupant injury analyses and vehicle accident reconstruction includes application of occupant and vehicle computer simulations. “Our assigned tasks start with reconstructing the events that led to an individual’s injuries. In many instances, the person was not wearing a seatbelt and we are asked to compare and contrast what actually happened to what would have happened if the person had been wearing a seatbelt,” said Dr. David Raymond, Senior Biomechanical Engineer for VSI. “This is a very unique and very challenging application of crash simulation tools that have revolutionized the field of forensic biomechanics and accident reconstruction.”

VSI began its forensic reconstruction by determining the vehicle dynamics of the crash. It performed detailed vehicle and scene inspections. It performed laser surveys of the roadway and land surrounding the roadway for the purpose of generating a 3-D map of the accident scene. Positions of physical evidence, such as tire marks and gouges in the dirt and on trees, were meticulously measured and plotted. This information served as input to software used to determine the vehicle crash dynamics. Once vehicle dynamics were comprehended, they served as input to occupant simulation to determine kinematics and injury mechanisms.

VSI used Madymo software to analyze the movement and injury response of the human body to impact. Using the vehicle dynamics as input, Madymo offers the ability to study occupant kinematics under various impact and restraint conditions.

Before simulating the real-world accident, Raymond ran a validation simulation to evaluate the model. In the case of this rear-impact crash, a Federal Motor Vehicle Safety Standard (FMVSS) 301 test was simulated to compare the model’s response of the front seat occupant and seatback to the physical test. FMVSS 301 testing is performed on every vehicle model to determine the integrity of its fuel tank. This test involves a moving barrier striking the rear of the vehicle that is comparable to what happened in this accident. The test also is typically equipped with front seat occupants with instrumentation to measure head, chest, and pelvis acceleration. To comply with the FMVSS 301 requirements, vehicles must withstand certain specified impact tests ranging from 20 to 30 mph, without leaking fuel in excess of 1 oz/min following the tests. These provided VSI data for validation of their rear-impact simulation.

Raymond subjected the Madymo model to the same impact defined in the FMVSS 301 test. He compared the Madymo output to the data and high-speed film taken from the physical test. The simulation results matched up closely, providing a strong indication that the model could be trusted to provide accurate results in this case.

“Validation of a computer simulation is a step performed by every analyst working in a design capacity within the automotive industry,” Raymond said. “Forensic simulation of real-world crashes is especially difficult, however, because many real-world collisions are so unique, there is no data to compare your model to like design engineers have when simulating standardized FMVSS tests. In those instances, VSI will evaluate trends and run design of experiments to gain insight into occupant kinematics and injury mechanisms. In one study where we wanted to understand occupant head motion under various restraint conditions, we ran over 4000 simulations, varying parameters related to the seatbelt position and occupant size and weight in order to understand how these parameters would affect how far the head would have moved if an occupant were restrained.”

With the Madymo model validated, Raymond proceeded to simulate the crash in question. To establish the Madymo model, the vehicle interior was scanned using a FARO Arm to obtain precise vehicle dimensions and geometry. Testing was carried out to obtain input to the model as well. In the case of this high-speed rear impact, the passenger’s seat cushion and seatback were tested to obtain material and structural properties. Raymond then positioned anthropomorphic models in the vehicle with the same size and weight of the actual occupants of the vehicle during the crash in question. He used the vehicle dynamics data to drive the simulation.

VSI uses ESI Group’s Visual-Safe MAD pre- and post-processor for processing and analyzing simulation output. “An important advantage of Visual-Safe MAD is that it allows for analysis and visualization of the simulation’s animation as well as engineering data in a time-synced manner,” Raymond said. “I view the engineering data while watching the simulation to make sure that the results make sense. Visual-Safe MAD and Madymo work well together even when you are dealing with unknown variables associated with a real-world accident. For example, in a situation where you don’t know exactly how the seat was positioned, you can run a design of experiment with ranges to see how it affects the outcome. In many cases, you will discover that the unknown variables either have little effect on the outcome or that you can make a good estimate for the unknown variable by matching the simulation results to physical evidence.”

Raymond first ran a simulation of the accident with the passenger in the front seat being unbelted. The simulation showed the man being thrown through the vehicle and hitting another seat. The simulation matched physical evidence such as contact marks and the occupant’s final point of rest. Next, Raymond ran a new simulation with the front passenger’s seatbelt fastened. “We found that in this case the seatbelt reduced the passenger’s velocity to the point that he was contained mostly in his own seat.”

Another key aspect of VSI’s analysis is assessing occupant injury risk. “We use injury risk assessment tools developed through the automotive industry during our investigations,” Raymond said. “That includes injury criteria developed over the past 50 years. Visual-Safe MAD has many of these criteria functions built into the post-processing software, making injury risk assessment of the simulated crash seamless and straightforward. Using established injury criteria, VSI is able to assess the probability of sustaining a head injury, chest injury, or other injury under both belted and unbelted crash scenarios. We looked at the injury potential in this scenario and concluded that the seatbelt would have mitigated his spine injuries and that he would have sustained only minor injuries.”

Olivier Morisot, Marketing Manager, ESI North America, wrote this article for AEI.

HTML for Linking to Page
Page URL
Rate It
5.00 Avg. Rating

Read More Articles On

Low temperature combustion (LTC) remains an area of high interest with extensive research, even though many challenges must still be overcome before LTC sees commercial acceptance.
Ford and Virginia Tech Transportation Institute are testing a Ford-designed method for self-driving vehicles to signal their "intent" to pedestrians; the research is hoped to help develop common standards for such communication.
Desire for quicker turnaround and more in-depth analysis point to increased automation of ECU testing.
The new vehicle test center in Auburn Hills, MI, will feature a chassis dyno for emissions testing, powertrain development programs and benchmarking.

Related Items

Technical Paper / Journal Article