David Camm lost 13 years of his life to eight drops of blood.
In 2000, the Indiana state trooper was charged with murder after finding his wife and two children shot to death in their home. During the three trials that followed, the prosecution brought in bloodstain experts who argued that the flecks of blood on the t-shirt Camm had worn the night of the murder were “high-velocity impact spatter”—proof, they said, that he was the shooter. Analysts called by the defense, on the other hand, testified that the blood was actually a transfer stain that had smeared onto Camm’s clothing after he’d tried to help his children.
Testifying for the defense, forensic scientist Robert Shaler disputed the claims of the bloodstain experts on both sides, insisting that the blood’s scant patterning was “too little information from which to draw any meaningful conclusion,” Pamela Coloff at ProPublica reported last year.
Nevertheless, spatter evidence was used against Camm, who was sentenced to 195 years in prison. He remained behind bars until his acquittal in 2013, after the conviction of another man whose DNA had been found at the crime scene several years before.
Camm’s case is only one of many that have come under fire for their heavy reliance on bloodstain pattern analysis, a subfield of forensic science criticized in a report released by the National Academy of Sciences in 2009 for being “more subjective than scientific” and rife with “enormous” uncertainties.
But that might be in the process of changing—thanks in part, perhaps, to the work of a team of physicists that’s marrying the study of blood spatter to fluid dynamics. Their latest model, published today in the journal Physics of Fluids, can now accurately match a bullet of a certain size and shape to the size, number, and speed of blood droplets produced by a gunshot wound.
The approach is all about bringing bloodstain pattern analysis up to speed with modern science, says Patrick Comiskey, who conducted the research as part of his graduate work at the University of Illinois, Chicago. Unlike an individual investigator’s interpretation, he says, physics is something “you can’t change.”
No one scientific finding or model will solve all the problems associated with bloodstain pattern analysis, says Suzanne Bell, a forensics expert at West Virginia University who was not involved in the research. Still, she says, “any time you bring in fundamental science to forensic applications…[it] will help build up the base knowledge we didn’t have necessarily before. People have been working with bloodstains for hundreds of years, but we haven’t seen the depth of investigation that studies like this bring.”
Bridging the existing gaps between forensics and science isn’t trivial. There’s an inherent challenge to analyzing crime scenes, which are, in a sense, the world’s worst time capsules: At best, they partially preserve the remains of a complex, messy problem that can only be solved in reverse.
In many cases, the pattern left by a bloodstain, including the number, size, location, and overall distribution of droplets, can reveal a lot about the shooter, victim, and weapon involved. But some of the methods that blood spatter analysts have used in the past to reverse engineer the trajectory and point of origin of a stain may have been too simple, leaving out some of the intricacies of how blood moves and deforms after leaving the body, says Daniel Attinger, a mechanical engineer at Iowa State University.
Part of the problem is that blood isn’t an easy fluid to work with. Liquids like water flow at the same rate no matter how hard they’re shaken or stirred, but the same isn’t true of certain mixtures like ketchup or blood, which actually get runnier if more force is applied to them. As a result, blood drips slow, but sprays fast.
Blood is also incredibly dynamic: It teems with living cells and active enzymes, and its properties shift under the influence of minute fluctuations in temperature, or the presence of drugs. It’s the “most complex fluid” he’s ever studied, Attinger says.
As soon as it leaves a human body, blood begins to cool and coagulate, effectively thickening, while also warping under the combined influence of gravity and drag from the air it’s moving through.
That’s all before blood meets the surface it ultimately lands on—the properties of which can seriously affect the impressions left behind. If blood’s landing pad has some bounce, it will affect the way the fluid splashes; if it’s rough or full of holes, it’ll change how the blood distributes itself. And while some stains burst on impact, producing a spray of disconnected droplets, others sprout spines that snake away from the bloody epicenter like skeletal fingers.
Like several other disciplines of forensics based in pattern analysis, the study of blood spatter arose and matured largely outside of the scientific community, Bell says. Unlike DNA analysis, which was honed in research laboratories for the purposes of basic biology before being applied to forensics, blood spatter has a far more subjective heritage, based on experiences, anecdotes, and empirical observations. “To get their data, analysts would walk into a room, fill a sponge with blood, and hit it,” she says.
There was nothing intrinsically wrong with this practice, and this kind of experimentation remains necessary, she says. But the practice lacked strong scientific foundations, and for years, little was done to explore the fundamental dynamics behind the paths droplets of blood took after impact.
Thicker than water, and far more complex
This rift in knowledge is where fluid dynamics comes in: Many of the subtleties of blood in motion can be captured with equations, Comiskey says. He, Attinger, and their colleagues have spent years crunching the numbers, coming up with increasingly complex theoretical models that account for more and more of the blood spattering process. The team did their fair share of hands-on experiments, too, firing bullets at everything from sponges soaked in pig’s blood to suspended sacks filled with the real stuff taken from human bodies. After collecting the spatter on clean, smooth surfaces like butcher paper, they’d see how the results of their computer simulations stacked up, and adjust accordingly.
The researchers’ models can now demonstrate a mathematical relationship between the shape and speed of a bullet and the distribution of the resulting blood spatter. Some things are fairly intuitive—like the fact that faster bullets will generate smaller droplets, or that the farther a fleck of blood travels, the more it’s going to be influenced by the pull of gravity. Other factors, though, are more subtle, such as the ways in which droplets are affected by the gas that escapes the muzzle of a gun after it’s fired.
“We now have a pretty generalized model,” Comiskey says. “I can say, more or less, that if I have this number of droplets, and they’re this size, and so on, then it was this bullet, this velocity, and [the shooter or victim was at] this location.”
One of the most important factors, Comiskey says, is being able to predict how blood breaks up into smaller drops, a process called atomization. Blood spatter progresses in two directions—backward, toward the shooter, and forward, parallel to the path of the bullet—and neither is very straightforward. The backward case is actually the simpler one, because it doesn’t really matter what the bullet is hitting: Either way, it’ll produce a kind of recoil. Forward spatter, however, produces more droplets that can travel up to twice as fast. It also involves messier physics, because the blood is pushed so quickly ahead of the bullet that it shatters like glass instead of deforming like a typical fluid. The term for this phenomenon is an apt one: “chaotic disintegration.”
Comiskey compares the two physical processes to running a faucet into a cup already filled with liquid. Backward spatter is the upward splash that occurs when the water hits what’s in the cup. Forward spatter is what’s produced when the jet of water is so fast and strong that it blasts through the bottom of the cup. In this case, the cup’s composition and contents can have huge effects on the resulting spray.
Attinger acknowledges that these models and scenarios are still oversimplifications. Sponges and sacks of blood aren’t human bodies, and the walls and floors of crime scenes usually aren’t lined with paper. The different types of fabric used in clothing also present a complication that hasn’t yet been accounted for in crime scene reconstruction. Still, experiments like these, in which the researchers can control and keep track of all the main variables at play, are necessary starting points, he says.
“None of this fundamentally changes what happens if you hit a sponge full of blood,” Bell says. “But it helps you understand why it produces the pattern it produces.”
Blood will tell—but just how much?
The researchers hope that their models will someday become sophisticated and realistic enough to be translated into a chart that spatter analysts can use in the field. An even loftier goal, Comiskey says, would be a cell phone app or handheld device that could scan photos of stains and spatter, crunch the numbers, and spit out a likely rundown of what transpired: the type of gun, the bullets used, the locations of the people involved, each with a score of the algorithm’s confidence in its results.
That possibility is a long way off, though, says Marilyn Miller, a forensics expert at Virginia Commonwealth University who was not involved in the research. Perfect experimental conditions manufactured in a research lab are a far cry from the “highly textured, highly interrupted world of a crime scene,” she says.
All the same, the team’s results offer clarity and coherence to the study of blood spatter, Miller says. “This is the kind of fundamental research that should be the basis for all forensic sciences.”
Attinger, Comiskey, and their colleagues also aren’t alone in their work: Several other research groups have previously investigated the fluid dynamics of blood spatter, including teams at Georgia Tech, Boston University, and Washington State University.
Practically speaking, the mathematical models that come out of these labs could be incredibly powerful tools to test hypotheses generated in the field, Bell says. “The evidence might suggest that a blow was struck here, using this,” she says. “The model can tell you if that’s within the realm of possibility, or unlikely. It’s not going to be, ‘This is what happened,’ but it can assign a probability [to a scenario].”
And that kind of uncertainty—or lack thereof—can make a big difference as evidence makes its way into a courtroom, Bell says. “This is all about making something that wasn’t quantitative, more quantitative.”
But this knowledge won’t do any good if it’s not made available to blood spatter analysts in an understandable, approachable way, Miller says. The fact remains that most experts in this discipline haven’t been trained to interpret research papers grounded in fluid dynamics.
Bell agrees, stressing that much of the scientific gap in these disciplines is about communication, rather than a lack of effort or any kind of willful ignorance. “The problem isn’t that labs don’t want to do better,” she says. “It’s a resource issue: So much is demanded of [the personnel in forensics laboratories], who are busy doing cases...every minute they give to training, learning, and digesting this research takes away from casework.”
Tweaking training regimens for blood spatter analysts or encouraging academic researchers to join forensics labs might make a dent in these issues, Bell says. But more sweeping reforms will be needed to actually put research like this in a position where it can have a real impact.
Part of this will require bringing other relevant voices, like those of judges and attorneys, into the bigger conversation, says Sarah Chu, the Innocence Project’s Senior Advisor on Forensic Science Policy, who was not involved in the research. Accuracy and scientific rigor are critical, but research also needs to be evaluated for its impact on the criminal justice system itself, as well as its ability to be ethically applied in the field, she says.
But the forecast isn’t necessarily gloomy, Chu says. In the decade since the 2009 National Academy of Sciences report, a lot has changed. In 2013, the National Institute for Standards and Technology and the Department of Justice established the Organization of Scientific Area Committees (OSAC) for Forensic Science, which was explicitly tasked with shoring up the scientific foundations of forensics.
OSAC’s subcommittees, which vet methodology for 25 forensic science disciplines, including bloodstain pattern analysis, are some of the few venues in which judges, attorneys, scientists, and other stakeholders are able to come together, Chu says.
Maintaining the momentum gained in the past 10 years, she says, will mean staying engaged and highlighting the goals that traditionally disparate communities have in common, rather than the qualities that might set them apart.
“There hasn’t been a lot of overlap between the world of these scientists and [bloodstain pattern analysts],” she says. “Hopefully publications like these signal the need for that to change.”