Taylor Grieve, graduate research assistant in the MFRC,
prepares a tool mark sample for scanning by profilometer.


In the justice system, a material witness possesses knowledge or expertise so vital to the case that their testimony is critical to the outcome of the trial.

The field of science itself has become a material witness and just as vital to the justice system and homeland security, as demand grows for an ever higher degree of scientific accuracy and reliability in the investigation of forensic evidence.
The Ames Laboratory has built an international reputation in materials science, including the development of advanced tools and techniques for materials characterization of surfaces, structure, chemistry, and physical behaviors.  While the principal mission of the laboratory is finding solutions to energy challenges, these new and increasingly accurate and sophisticated characterization techniques have key benefits in other fields, such as homeland security and forensic science.

Ames Laboratory’s excellence in materials characterization makes the Midwest Forensics Resource Center a good fit for the Lab, which oversees the program in partnership with Iowa State University’s Institute of Physical Research and Technology. The MFRC brings together the Ames Laboratory’s unique research capabilities to support the needs of crime laboratories in 16 states in areas as diverse as molecular and elemental spectroscopy of inks and toners, fracture analysis, toolmark identification, and bloodstain analysis.

Finding print media’s true signature: ink analysis

In the discipline of questioned documents investigation, one of the key clues to verifying the authenticity of a signature or the document itself is the ink on the page. The MFRC is researching ways to bring a higher degree of speed and accuracy to ink analysis.

Ames Laboratory scientist John McClelland and his research team are working with the U.S. Department of Homeland Security and the National Institute of Justice to develop a comprehensive data bank of printing inks and toners as well as computer search software for sample identification.

It’s a painstaking job, building the database

“Our researchers are in the process of collecting samples from manufacturers, and it’s been a slow process, since inks and toners are highly proprietary. They understand the importance of what we’re doing, but we’ve been signing a lot of non-disclosure agreements,” says McClelland about the project under way.

The researchers are using multiple spectroscopic analytical techniques to characterize samples: mass spectrometry, infrared spectroscopy , and x-ray fluorescence.

Ames Laboratory scientist Tonu Reinot has developed computer search algorithms and software that combines and interprets the data from all three of the techniques.

“We find that this approach provides search capability that is more robust, and less subject to false alarms,” says McClelland. “This is particularly important as more and more ink data is put in the library and increasingly only small differences exist between various inks in the library.”

A side-by-side stacked comparison of the infrared spectra
of 450 different inks and toners. The Midwest Forensics
Resource Center is assisting the Department of Homeland
Security and the National Institute of Justice to establish
a comprehensive databank of printing inks and toners.

The MFRC also helped the U.S. Secret Service, one of the foremost questioned document authorities, in establishing a similar database for handwriting inks. For decades, the Secret Service maintained an extensive library of physical ink samples, along with thin-layer chromatography data that identified the unique characteristics of each ink.

“The problem with thin-layer chromatography is it’s fairly time consuming, and not a really easy thing to implement when you have many samples to test and identify,” says McClelland of the process, which uses an adsorbent material on a plate to separate out various analytes of a substance.

McClelland’s research group tested a variety of techniques, including infrared spectroscopy, Raman spectroscopy and also a relatively new mass spectrometry (MS) technique called Direct Analysis in Real Time-MS (DART-MS), which allows a substance to be analyzed without sample preparation and in the open air.

“It turned out that mass spectra are the most definitive for searching and identifying an unknown sample against a database of known samples,” says McClelland. “Thin layer chromatography is a respected analytic technique; but mass spectrometry is more easily applied, and you can get more detail in a mass spectra analysis than in a thin layer chromatogram. It’s a better fingerprint, you could say.”

Break in the case: fracture analysis

Often a crime or accident scene is littered with evidence that comes in pieces: shattered plastic, glass or metal, torn fabric or tape. It’s up to forensic investigators to determine how the pieces of the evidence, and therefore the story, fit together.

A team of materials scientists is applying fracture analysis methods to the study of broken knife blade tips to determine if there is a scientific methodology that can be applied to bolster visual inspection of evidence.

“Much of the forensic examiner’s work is by eye, to see if they can match one piece of a broken object to another,” says Barbara Lograsso, associate scientist with the Ames Laboratory.  “Instead of only trying to match the surfaces visually, we want to be able to show that a technique can significantly and measurably discriminate between matches and nonmatches.”

The scientists are analyzing the fractured surface areas using a 3-D interferometer, which generates a surface height topographical map of the surfaces. A MATLAB software program was developed to “read” the data and perform match analyses.  At this level of analysis, the scientists can provide a lot more detail about the fractured surfaces.

“Some of those features are quite unique to the event,” says Ashraf Bastawros, an Ames Lab associate and an associate professor of aerospace engineering at Iowa State University who specializes in fracture mechanics. “If I break two surfaces I will find some commonalities because they are from the same material, but the additional information is how it was broken. Was it bent? Or bent and twisted? All of those leave unique characteristics on the surface of the fracture.”

A sample broken knife blade. MFRC scientists are researching
methods to describe the unique characteristics of fractured
surfaces to aid in forensic

That precise level of detail may be vital in cases where visual inspection of evidence isn’t sufficient to prove a definite match.

“Investigators may have pieces that broke a couple of times and you cannot put it together like a jigsaw puzzle,” says Bastawros.  “There may be pieces missing altogether. So when you cannot do that, this method can tell more from the unique fractured surface characteristics, something less likely to be disputed in court.”

“What forensic examiners don’t have right now is a technique or process that has some determination of uncertainty,


a statistical error rate, to bolster their judgment,” says Lograsso.  “We hope to provide that with a method and citable literature.”

They also intend to produce an open-source software program that allows an image of the material sample to be analyzed, with statistical tools to quantify the quality of a match.

“It’s not possible to measure the total population of all possible specimens, but we’ve got a start with knife tips,” Bastawros says. “We’ll be measuring some materials and once we finish the basic platform, it will have a lot of other applications. We are now working in metal, but it could be applied to ceramics, plastic or glass.”

The tools to build a case: impression evidence analysis

Criminals often leave their mark in the most literal sense—marks left by scrapes and impacts of common household objects like tools. Forensic examiners are tasked with matching this impression evidence to a tool, under the assumption that individual tools and their marks are unique.

“That’s been the problem,” says Scott Chumbley, an Ames Laboratory materials scientist. “I’ve worked with many examiners and they have a high level of experience and integrity. When they say there’s a match, I’ve agreed with them. But an attorney can always argue ‘this is just your opinion.’ So that became the question. If you have a tool, can you show quantitatively that the marks left by it are unique to only that tool?”

ImageChumbley joined with Max Morris, an associate professor in the departments of statistics, and industrial and manufacturing systems engineering at Iowa State University, to find out if they could develop a way to computationally compare toolmarks to the tool that made it, resulting in statistically valid matches.
They began with a common toolbox item, a flat head screwdriver.

“That was the starting point,” says Morris. “Practically speaking, it was not going to get them where they need to be in the crime lab, with a database of hundreds of types of tools, but you have to start with a version of the problem you can get your hands around.”

The research team obtained 50 screwdrivers that were sequentially made on the same manufacturing line. Striations were made by dragging screwdriver tips across sheets of lead at set angles, and traced using both stylus and optical profilometers. Morris developed a statistical algorithm to analyze the toolmarks and compare them with a tool, with the aim of determining whether the tool and toolmarks match.

“We’ve shown that you can have 50 screw drivers that are nearly as identical as possible coming off the manufacturing line, and we are still able to tell one from the other with fairly high degree of certainty,” says Chumbley.

The research team is now studying another common tool, slip joint pliers, and the marks they leave when cutting wire.

“It’s a more complicated version of the original research,” says Chumbley. “In this situation we have two separate cut marks, different from each other but related to the same tool.”

Song  Zhang, an associate professor of mechanical engineering at Iowa State University and a specialist in facial recognition technology, is developing  an open source computer program to generate sample toolmarks for comparison, using Chumbley and Morris’ research as a foundation.

“The idea is to scan a tool tip, get a 3-D representation, and then write a software program that can generate any possible mark from that tool in the virtual world, in any way that we want, with any resolution we want, at any angle we want,” says Zhang. 

Zhang envisions a system, possibly portable, that relies on an optical profilometer to scan marks and tool tips. “Optical profilometers present some design challenges because of data quality, but they are preferred by forensic examiners because they don’t touch the sample and compromise evidence,” he says.

The program could simplify much of the painstaking sample comparisons that forensic examiners do to substantiate a match.

“This software tool will eliminate a lot of the possibilities they might otherwise have to consider before making a determination. We will still need forensic examiners to make the final judgment.”

A pattern of evidence: blood impact spatter analysis

It’s often the most disturbing evidence of a crime: spatters of blood on floors, walls, and other objects. But the patterns formed by these droplets can give investigators a better picture of how a criminal event unfolded.
David Baldwin, director of the MFRC, says what’s been missing from blood spatter study is basic research.

“There was a lot of forensic training that went on where they said ‘this is what happened and this is what you see at the end.’  But they didn’t really know how it got there. People were trained based on experience at crime scenes, not as physicists or as fluid dynamicists. Their explanations may not have made any realistic sense to a scientist.”

To fill the need for research, the MFRC has amassed a library of 500 video clips of blood spatter experiments showing how blood as a fluid behaves when it falls or spatters in relation to objects like bullets, cloth, wood and tools.

“The library is one of our best products, and it’s used constantly. It has very quickly become invaluable to crime scene analysts,” says Baldwin.

The MFRC is continuing that research by characterizing blood impact spatter, or what happens when an object strikes an open source of blood.

“This is basic science. This is physics,” says Baldwin. “I want to understand the parameters that are involved in determining the appearance of bloodstain patterns. No one’s really bothered to study the physics of the distribution, the size of the drops, their locations, and how they are affected by physical parameters that might matter, like velocity and mass. If we don’t do these types of physics experiments, we won’t know that you can or can’t say anything definitive about those end products.”

A series of still photos from a high speed video of a blood impact spatter experiment. Researchers at the MFRC are studying the physics of how blood spatter is formed under different variables such as the quantity of blood and the velocity of impact.

Baldwin’s team of scientists is conducting experiments by constructing a rig or apparatus in a controlled lab environment that eliminates extraneous crime scene details.

“This is a very physics-oriented experiment where we control the surface area, the depth of the blood, the diameter of the pool, the velocity of the impact, the diameter of interaction. Our goal is to determine parameters for the distribution of stains that are relevant and tell us something about what happened,” says Baldwin.

Though the focus of this research is understanding the physics of blood spatter, Baldwin says he could foresee technology that takes an image of blood spatter at a crime scene and analyzes the possible causes, “but that’s in a farther-off future.”
Blood impact spatter research and the more than 60 other research projects supported by the MFRC over the past decade, says Baldwin, are part of the ongoing effort to bring more exacting science to evidence investigation.

“The entire discipline of forensics is about trying to say something based on the end observables, without any control over or knowledge about what happened on the front end,” explains Baldwin. “Scientists would call that inverse problem engineering. Because of the growing demand in the justice system for a citable scientific justification for admitting physical evidence, we’re here to help forensic investigators do just that.”

~ by Laura Millsaps