Tool Mark Identification

[Suddenly]

I am examining the following statement:

"A tool leaves damage on another object that is unique and distinct to that particular tool."

This is in the context of one piece of metal being used to work a second piece of metal. It is claimed that by examining the damage, or tool marks on the second piece of metal, it can be detrmined that the marks came from the first piece of metal to the exclusion of all others.

This is done by taking the first piece of metal, striking it against a third piece of metal similar to the second piece, and comparing the marks with a microscope and optical bridge.

What I am curious about is what kinds of evidence and studies and so forth would you need before you believed that such an identification was "grounded in the scientific method?"

I've run across this topic and I'm curious as to the opinions of others on the issue.

 

[Torlack]

This would be the same thing as firearm identification. I wonder if there are any studies on the reliability of that.

 

[RichardR]

I've seen it on "CSI" so it must be true.

Seriously, I've seen this on CSI and I've wondered if it would really stand up in court.

 

[Beanbag]

It comes down to comparing imprints or traces specific to an individual tool. Suppose I have two sets of pliers, bought at the same time, made by the same machines on the same day. Out of the package, they are identical (pretty much). But maybe I slip while using one set of pliers, and damage the serrations on the jaws somewhere. Now, the imprint of that tool is unique when compared to the other. Damage and alterations occurr over the life of a tool, making it more unique and trackable over time.

Some problems exist. For instance, I use my pocket knife to cut an opening in a piece of plastic, which leaves a set of markings that correspond to the edge of the knife. I then thoroughly resharpen the knife. The markings no loger match.

Most cases, you might use tool marks to exclude what tools might have been used, reducing the pool of "suspects". Paring down where you have to look allows you to concentrate your resources where they are most likely to succeed.

Regards;
Beanbag

 

[Suddenly]

It comes down to comparing imprints or traces specific to an individual tool. Suppose I have two sets of pliers, bought at the same time, made by the same machines on the same day. Out of the package, they are identical (pretty much). But maybe I slip while using one set of pliers, and damage the serrations on the jaws somewhere. Now, the imprint of that tool is unique when compared to the other. Damage and alterations occurr over the life of a tool, making it more unique and trackable over time.

Some problems exist. For instance, I use my pocket knife to cut an opening in a piece of plastic, which leaves a set of markings that correspond to the edge of the knife. I then thoroughly resharpen the knife. The markings no loger match.

Most cases, you might use tool marks to exclude what tools might have been used, reducing the pool of "suspects". Paring down where you have to look allows you to concentrate your resources where they are most likely to succeed.

Regards;
Beanbag

Sure. That is the theory. What I am trying to figure out is if there is any real evidence that "A tool leaves damage on another object that is unique and distinct to that particular tool."

There are guys running around claiming that by the testing I described in the first post that they can say with scientific certainty that they can posititively identify a tool as making a particular mark. This isn't a "one out of a pool" sitiuation, rather it is compared against every other tool in the world that could have made that mark.

I haven't seen a formal testing of this hypothesis. I'm wondering if it is really testable at all. There are those that claim they can make this identification with low error rate, but there are no set objective criteria for what constitutes a "match." There is some talk of "points of comparison" but no real standards. I've seen guys that identify 20 points of comparison testify to the same cerainty of someone that found 2 points of comparison.

This doesn't even get into the fact that "points of comparison" only show that the mark is consistent with a particular tool. It doesn't in itself speak to the uniqueness of the tool mark.

 

[ceptimus]

The same arguments apply to human fingerprint identification. If there are sufficient independent criteria that can be measured, then you can show statistically that the possibility of a false identification is vanishingly small.

Now I am skeptical about this tool identification business. But if no one had ever thought of fingerprint identification before and it was a new idea, I would be skeptical about that too.

Come to think of it, I am still skeptical about how fingerprint identification is so reliable, given the wear and tear our fingertips suffer, the aging process and so on.

 

[Beanbag]

The same arguments apply to human fingerprint identification. If there are sufficient independent criteria that can be measured, then you can show statistically that the possibility of a false identification is vanishingly small.

Now I am skeptical about this tool identification business. But if no one had ever thought of fingerprint identification before and it was a new idea, I would be skeptical about that too.

Come to think of it, I am still skeptical about how fingerprint identification is so reliable, given the wear and tear our fingertips suffer, the aging process and so on.

I would place not too much faith in an identification based solely on a tool mark identification. There are just too many uncertainties in the process, and I doubt a prosecuter would be happy with it as well. A tool mark would be a single component in a body of evidence. A good investigator would use the information provided by the tool mark to look for further evidence to support their case. I would have problems with a case based on a single fingerprint -- think of how many things you unconsciously touch or handle during the day. It could be used along with other evidence to build a case, such as blood evidence, eyewitness testimony that placed a person at the scene at a given time, etc., but just a single fingerprint means that a person touched that object at some time.

You only have to hope that a jury would be as discerning about the evidence.

Regards;
Beanbag

 

[Bikewer]

The technique of toolmark ID is essentially the same as that for firearms, and has the same limitations.
Ideally, the evidence tool would not have been used again after making the original cut or mark, since that would degrade the striations caused by irregularities on the tool's cutting surface.

A test cut would ideally be made on the same sort of material found at the crime scene. (say, a padlock shackle of the same make and model.)

Advanced techniques such as spectrographic analysis can confirm that tiny metal particles adhering to the tool's edges came from the evidence at a crime scene.

If the tool is matched to the crime scene evidence, you must also link the tool successfully to the criminal, of course.

 

[Dymanic]

There are a couple of things you can try that may help
clarify this.

Tear a piece of paper roughly in two, and examine the edges closely under magnification. Tear another piece, and try to match it to the first, so it fits without any gaps. Keep trying until you get an exact match.

It's about informational complexity.

With a piece of paper, orientation on the Z axis is a given, though. Break a stick in two and drag it across a piece of paper, with a piece of carbon paper underneath (if you can still find carbon paper that is). The pattern created will be similar to a bar-code. But significantly changing the stick's angle of presention to the paper will produce a different pattern, so establishing a match may involve some knowledge of what that angle would have been. (If a particular object is known to be the one involved, this process can also be run backwards to establish what the angle was).

 

[Bikewer]

Fortunately, with most of the tools for which identification is attempted, orientation is no great problem. IE-a bolt cutter.

Many folks who would like law enforcement to keep "sample" bullets from each firearm sold are unaware of the difficulties.

The bore, like any other surface that's "worked", is subject to change. The rate at which this happens is conditioned by the type of ammunition used; high-velocity, metal-jacketed bullets will wear the bore rapidly.

Additionally, different bullets may have slightly different signatures. With handguns, the shooter often has the choice of relatively hard jacketed ammo, relatively soft swaged-lead bullets, or somewhat-harder cast bullets.
Handloading increases one's choices dramatically.

Very high velocity ammunition will deform violently on impact, or fragment, making identification difficult or impossible.

Shotguns leave no marks whatever on the projectiles, though the shell case can be identified if recovered. (extractor markings and firing-pin indentations.)