Estimates vary, but a ruptured aorta is a common cause of death in an automobile accident, killing 5,000 to 10,000 people a year in the United States, said John Blebea, chairman of the surgery department at the University of Oklahoma College of Medicine.
Seat belts help, but air bags have not made much difference in preventing this type of injury, Darvish said.
His research goal is twofold. He wants to use his findings to help develop better restraint systems. And for people who survive a crash with a damaged aorta, Darvish wants to help doctors identify which injuries will heal on their own and which require surgery.
"We need to predict what happens before it gets too late," Darvish said.
The key lies in the data from the crash-simulation track, assembled from industrial-grade components made by Precision Technologies USA of Roanoke, Va. The belt-driven system was originally designed for automated manufacturing processes.
It is no surprise what happens when an aluminum "carriage" zooms down the track, hitting another piece of metal that is tethered to one end of a Band-Aid-size piece of pig aorta. The tissue is torn in two, the severed edges curling up a bit, like when you pull apart a piece of saltwater taffy.
To determine precisely what happens as the tissue starts to fail, Darvish needs to measure how it deforms on a scale of millionths of a meter. Each segment of aorta is covered with flecks of paint before being torn apart on the track, so that Darvish and his doctoral students - Mobin Rastgar Agah and Soroush Assaris - can measure how far each fleck moves and how fast. This allows them to calculate the strain on the material.
So far, they have found that when a section of aorta is pulled apart slowly, the tissue will stretch to 150 percent of its original length before separating.
But when it is stretched rapidly, as would happen in a car accident, the tissue fails much sooner, at just 120 percent of its original length.
These measurements and many others go into the creation of a mathematical model, so the engineers can run computer simulations and determine just how much of a tear in the tissue will be problematic. The engineers have conducted other tests on a pig aorta filled with saline solution to simulate blood.
Other researchers have studied such injuries in cadavers, but Darvish said they were an imperfect model of live humans, due to the lack of muscle tension and the position of the organs. A section of pig aorta, on the other hand, allows for controlled, precise measurements, he said.
Aortic tears tend to start on the innermost layer of the vessel, said trauma surgeon Karen Brasel, a professor of surgery at the Medical College of Wisconsin. A big enough tear will lead to delamination - a separation of layers - and ultimately to an aneurysm. These bulging, rupture-prone spots can be repaired by surgeons with a stent graft.
Brasel, who helps investigate accidents for the federal government's Crash Injury Research and Engineering Network, said the thoracic aorta - the upper part - is a vulnerable piece of tissue because it has some give in it.
"The aorta keeps moving forward" in a car crash, she said. "Then it snaps back."
Darvish thinks one answer might be seat belts or other restraints that can be customized to the driver's size. Another issue is that current air bags cushion the body symmetrically, which works well in preventing rib fractures. But inside the body is another story, he said.
"The inner organs are not symmetrical," Darvish said. "Somehow, we need to include that."
He hopes the model also will help identify which kinds of tears will progress to a life-threatening stage, so physicians have a better idea what to look for when they read the results of CT scans and other imaging.
It is noisy work, with loud bang after bang as the results are stored on a computer. If the model does its job, then this type of loud bang will not be so deadly when it happens on the open road.
Watch Temple engineers simulate the impact of a car crash on a piece of aortic tissue at www.inquirer.com/aorta