Glass is very hard on bullets, and it’s one of the more expensive portions of the test—a protocol that costs several thousand dollars.

The question often comes up, "You talk about the FBI tests, but exactly what are they?" As I mentioned in passing in my previous column, until the mid-1980s we did not have a scientifically repeatable tissue stimulant in which to test bullet performance. Then Dr. Martin Fackler--working at the U.S. Army Wound Ballistics Research Laboratory, Presidio of San Francisco--developed ballistic gelatin, which permitted researchers to test and compare bullet performance.

After the FBI Miami shootout--in which eight agents fought with two suspects, the latter continuing the battle even after being hit more than once--the Federal Bureau of Investigation set about creating a scientific method to measure bullet performance. I don't think the bureau intended to make the test an industry-wide standard, but that's the way things have turned out.

The process starts with gelatin, specifically Kind & Knox 250, in a 10 percent solution--one part gelatin by weight to nine parts water by weight. The process is strictly defined: The temperature during mixing has to remain within a limited range. If the mixing water is too hot or cool, then the strength of the resulting gel mix changes.

Once mixed, it is poured into molds and cooled. The mold size is defined for either handgun or rifle cartridge use. Handgun tests employ blocks that are 6x6x16 inches, and the blocks are cooled to 39.2 degrees and stored until needed.

The blocks must be shot within 20 minutes of being removed from the refrigerator. The gel blocks each have a thermometer stuck in them to check their temps. If they get too cool or warm, they are either adjusted or discarded.

Once the temperature is checked, then the block is further tested for consistency by having a steel BB fired into it. The BB is chronographed as it is fired, and if it is too slow or fast, the test has to be repeated. The pellet, once fired within an accepted range, has to penetrate a certain amount, plus or minus a small margin for error. Again, if it does not pass this test, the block is pitched.

FBI Ammo Test Video Watch a live FBI ammo test. Click Here

Finally, it has a bullet fired into it. One block, one bullet, 10 feet from muzzle to gel block. In some labs, the block can and will be melted, filtered, cast and re-chilled. In others, one use and it is gone.

Now we get to the interesting part: Repeat as necessary. You see, the FBI, in proper scientific fashion, shoots a number of bullets and then takes the averages of their performance: depth of penetration, expansion, retained weight, etc.

So five shots, each with its own block. Penetration is measured to the nearest quarter-inch, and expansion is measured as well. Expansion is the smallest diameter averaged with the largest diameter on each bullet.

But, wait, it gets better. The FBI recognizes that armed encounters rarely involve naked perpetrators. Bad guys do wear clothes, and they do hide behind things, so the bureau tests for those parameters, too. A full test will involve bare gelatin, as well as intermediate barriers of heavy clothing, plywood, auto glass, sheet metal and wallboard. And just to be thorough, the heavy clothing and auto glass tests are repeated at 20 yards in addition to the 10-foot distance.

The heavy clothing test uses four layers: T-shirt, dress shirt, synthetic insulation and heavy denim are laid against the face of the gel block before firing. The sheet metal test uses two six-inch square pieces of 20-gauge hot-rolled galvanized steel, set three inches apart and 18 inches from the gel block impact face. The gel also has a T-shirt layer on it.

The plywood is a six-inch square piece of AA fir plywood, but just one. Again, the distance is 18 inches from plywood to gel, with a single layer of T-shirt material on the gelatin.

The wallboard is two pieces of standard half-inch gypsum, spaced 3.5 inches apart. Why 3.5 inches, you ask? The actual width of a 2x4, this simulates an interior wall. And again, there are 18 inches of air space between "wall" and gel, with a single layer of T-shirt material.

Last up is the auto glass, and here I have a minor quibble with the FBI. The glass is laminated safety glass: a 15x18-inch piece of quarter-inch windshield glass. The glass is 18 inches from the gel, and yet another long-suffering piece of T-shirt material is on the gel. The glass is tipped back at a 45-degree angle, which is good, simulating an auto windshield to someone on foot.

However, the FBI also insists that the glass be angled 15 degrees to the side, to create a "compound" angle. Guys, study your Euclidean geometry: The intersection of a line and a plane has a simple angle of incidence. No matter how much you tilt that glass, there is no compound angle. But, as I said, it's a minor quibble.

The data gathered by all this are rigorously correct and accurate. Is it, however, a realistic expectation of what bullets do to people? The jury is still out, but in that regard ballistic gelatin is a whole lot closer than earlier test media.

Just what is the desired outcome of all this? What do we learn? Well, the FBI expects a bullet fired in all these tests to expand and not fragment. It must penetrate 12 inches of gelatin or the bureau considers it a failure.

As I've written before, if a bullet traverses a felon's sternum and then expands to the size of a manhole cover, but stops at 11 inches, the FBI considers it a failure. I would not.

The bureau wants full expansion, at least 12 inches of penetration, and up to 18 would be peachy-keen. That's a lot to ask of a handgun bullet, but the manufacturers have stepped up and provided bullets that meet the test. These cost more than bullets developed before the protocols and bullets not meant to meet the standards.

Conducting a full cartridge test is expensive, and the manufacturer has to do it before it hands ammo over to the FBI since there's no point in shipping the bureau ammo that won't pass muster. It takes no fewer than 48 gelatin blocks to do a test, and depending on how parsimonious the testers are, a single cartridge test can use up a good bit of heavy clothing, as well as a number of T-shirts.

Ten pieces of steel, 10 pieces of wallboard, five pieces of plywood and five pieces of auto glass get used up for each cartridge. At an optimistic estimate of costs, each cartridge test runs about $3,000 in test materials. Add in ammunition costs, the overhead for the range, the kitchen and coolers to chill the gel, and the salaries of the agents involved, and I can't see the cost of FBI testing being less than $10,000.

So now when next you see a list of FBI-tested calibers and bullet weights, you have an idea of the work involved as well as the cost. And you have a good idea of why ammo can be expensive.