How to interpret FBI data?

[Zak Smith]

I have virtually the "same" handgun in three calibers: a Glock 19 (9x19), 32 (.357SIG), and "virtual" 23 (.40SW). I am intrigued by the difference in ballistics - terminal and external - of these rounds.

The FirearmsTactical web site has a large library of terminal ballistics data from the FBI tests. As far as I've been able to figure out, the FBI tests are done as follows:

1. Fire test rounds into both bare and "clothed" gelatin blocks. The gel blocks have been calibrated so that they have a standard consistency.

2. Measure the maximum penetration, and

3. Retrieve the expanded round and measure its expansion.

So, the data ends up looking like this:

9mm 124 grain CCI/Speer Gold Dot JHP +P, 4/17/97,
Test gun SIG P226, Barrel length 4.25", Velocity 1223 fps,
Bare gelatin penetration 13.4", expansion 0.68",
Clothed gelatin 20.25", expansion 0.58".

So, my question is, what is the best way to compare the results? Obviously you want attain a certain penetration depth (what the minimum is is debatable), and you want the round to expand sufficiently.

We can compare max. penetration, or expansion, or penetration * expansion, which should approximate the total volume of destroyed, err, "gel." Or we could use some other comparision.

It would seem more useful to compare the "clothed" gelatin results, since people usually wear clothes.

So, just what is the best way to compare them?


For those who are wondering where I'm going.. We always have discussions like: W has more energy! X has more momentum! Y has more penetration! Z has more expansion!

In my opinion, the gelatin tests would seem to translate much more directly to real-world results than a more abstract measure such as energy, since they have to do with real physical effects. Once I have a way to compare the results of the FBI tests, I can see which particular computed metrics correlate with those results.

Or I can see which rounds have better physical results, penetration, or "wound area" as a function of momentum (related to recoil), or energy (related to case volume).

-z

[James K]

Hi, Smithz,

Is that correct that the clothed gelatin was penetrated deeper (by 7 inches) than the bare gelatin? Did clothing, by reducing velocity and expansion, add that much to penetration?

Jim

[lonegunman]

why does jello need to wear clothes?

sorry, I realize that is not pertinent to the question at hand.

[Apple a Day]

Man, did you just open a can of ugly worms!
Cloth versus naked A plug of cloth could get stuck in the hollow point, filling it, reducing the tendency to expand/resistance so that it goes deeper. You can see from the data that the bullet penetrating the clothed block did not expand as much. All sorts of naked/penetration jokes come to mind but I will refrain.
I am suspicious of claims people make about using hollow points in apartments to 'reduce the chance that they will penetrate walls' since, I think, the drywall will produce the same effect and turn a JHP into a FMJ. I'd love to see someone do some actual tests on drywall and post the results. I remember Oleg doing a similar test on a door (?) and even a .22 went through it.
I would love to be able to take some construction materials- sandwichboard, particleboard, pine, hardwood, sheetrock, glass- and test penetration of different ammo and calibers.

[J. Parker]

You probably know more about gun performance and effectiveness than some geek in a FBI laboratory some where. The FBI is no different than any other LE agency. A different calibre and pistol every other month. You're ingesting WAY to much information.
Light to medium weight good quality hollowpoint in any calibre and you'll do just fine. See, I didn't need a Masters in handgun effectiveness to tell you that. J. Parker

[Archie]

please enlighten the rest of us.

What constitutes "stopping power" has been argued and discussed since Roger Bacon blew the pestle out of his mortar. (He used a heavy pestle, his brother monk used a light pestle.)

Smitty, old boy.... The good thing about shooting gelatin blocks is that they are more or less reproducible. Most of the time. The bad news is they don't alwasy correspond to real life shootings. Just close enough to be "promising".

Keep reading and studying and thinking. Read all tests and statistics with an open yet critical mind. (What is really being tested? What does it show?) Take all conclusions with healthy skepticism.

Then, carry what gives you confidence. Practise to make solid hits. And watch what you are doing.

[juliet charley]

The FBI data is good data. It is reproducible with no deep, dark secrets as to its origin (the lab or fertile imagination or a couple of ammunition companies marketing schemes).

The figures after cloth are more meaningful. Generally speaking, you are looking for a round that expands wells (.60+) and penetrates 12 to 16 inches (after cloth) regardless of weight or the velocity. Any penetration less than 9 to 10 inches has the potential to be dangerous to you (may not work). Any penetration over 16 to 18 inches has the potential to be danerous to others than the target (the overpenetration bugaboo). It is generally accepted that more LEOs have been killed when bullets have underpenetrated (and failed to stop the target) than when they have overpenetrated.

I prefer the heavier bullets weights (so long as they meet the above criteria. Some prefer lighter. The emphasis should be on the prefer.

Remember, the lighter/faster rounds tend to penetrate more after cloth (or sheetrock) than the heavier bullets which may be a factor if you live in an apartment building or have several people living in your house. Generally speaking, there will be more muzzle blast and flash associated with the lighter/faster rounds than the heavier rounds (and depending on your firearm, they may be harder on it accelerating wear, etc.).

[Zak Smith]

I entered all the data from the FBI tests that I could find, computed a few things, and set up a web page that allows you to see it sorted in different order.

http://www.demigod.org/~zak/firearms/fbi-pistol.php

It is interesting to compare the kinetic energy order, to the total wound volume, for bare and clothed gelatin.

The results aren't what you might expect. (And, for criminey sakes, avoid CorBon 9x19+P/115! - its result is horrid).

-z

[juliet charley]

Enjoyed it. Noticed a couple of things:

(1) a wide variation between the same round when it was tested more than once (e.g., Remington GS 180);

(2) the .45/.40 advocates seem to be right.

[WESHOOT2]

".........through rose-colored glasses........"




------------------------------

"all my bliss is factory bliss"

[BigG]

Call the 1-900 Psychic Hotline. You will get a straight answer there before you will get one from the feds.

[Zak Smith]

WESHOOT2 & BigG - constructive comments are welcome.

Is there something wrong with the FBI methodology? Gelatin (clothed or not) sure isn't the same as a person, but the advantage of this data is: it's repeatable, controlled, and not inherently biased.

Juliet Charley - I assumed the variation between the "same" round was due to: different muzzle velocity (as shot from a different gun), or a different batch of ammo, since some testing dates differed by years.

-z

[juliet charley]

Smithz -

You are correct about the variances. That was point (evidently, poorly made) There can be a tremendous variation between different lots of various ammunition (food for thought for those who swear by a particular brand/style). Actually, the variances caused by lot generally exceed the variances caused by barrel lengths (at least at service lengths of 3.5 to 5.0 inches--.45 ACP being an exception).

[WESHOOT2]

Testing with controlled media is excellent for comparing ammo performance in controlled media, but individual physiology and specific launch platform performance vary per each instance.

I also suggest that L.E. requirements are different than civilian requirements.

Conclusion: no current test criteria will satisfy the most important question, "How will MY ammo perform on MY specific target?"

I tend to rely on anecdotal testing, based on observed actual 'street' shootings.
I prefer very fast bullets of 115-125g in the .355"-bore guns, and the 40 S&W seems to give good results with weights from 135-180g (go figure; I like the 135's best).
The 45ACP is pefect using correct 230g JHP's.

This is why I often comment on shot placement, as it seems to really be the most important factor in ALL cases.

At this point in time we can't predict bullet performance in humans. We have opinions.

[Zak Smith]

WESHOOT2:

Testing with controlled media is excellent for comparing ammo performance in controlled media, but individual physiologyand specific launch platform performance vary per each instance.

Yes, it is inherently limited in that we are shooting gelatin, not people. I agree. Barriers, clothing, and various alive bits, bones, etc, are all going to act differently than gel. And then there are all the physiological and psychological factors, i.e., if a certain wound is going to "stop" the assailent.

Conclusion: no current test criteria will satisfy the most important question, "How will MY ammo perform on MY specific target?"

So, then, we just throw the baby out with the bathwater and say we have no hope of scientifically characterizing performance ? If you have thrown out media tests, what basis do you have for knowing terminal performance? It seems like saying, there's no need to learn to ride a motorcycle at 5mph, because it's nothing like riding at 60mph..

I tend to rely on anecdotal testing, based on observed actual 'street' shootings.

This strikes me as haphazard. The way we gain reliable knowledge is to use the scientific method.. which that isn't.

If we want to study how "stopping" or incapacitating takes places, we need to know the physical mechanisms that occur that effect the stop. Furthermore, we should be able to measure those mechanisms. Once we understand the mechanisms of stopping or incapacitating, we can create a test which can measure one or more of those effects in a test material. If done carefully, we can be assured that a result of X in the test media will map to result X in an assailent. Repeat for all the known mechanisms.

The problem with many "stopping power" metrics is that they are either too abstract, biased, or not connected with any physical effect related to "stopping".

By "abstract", I mean things like kinetic energy or momentum. We have arguments like "light and fast" vs. "big and slow." Without any more data, these particular arguments are not based on physical effects.

By "biased", I mean the metric is based on a data collection methodology that is inherently flawed. The recent thread on the M&S OSS numbers brought to light some serious problems with that metric.

Finally, by "not connected with any physical effect related to 'stopping'", I mean things we talk about for which there is no known mechanism linking them to incapacitation. It is possible that "energy dump", to the extent that it does not induce tissue damage, is in this category.

Terminal performance is very linked to bullet design, also. A good test must take this into account.

I think the FBI data may be valuable because:

1. it's "scientific" in that it's published, reliable, repeatable, and not biased by sampling,

2. it measures several things very likely directly linked to one of the incapacitating mechanisms (penetration and expansion / wound volume). I have high confidence that this is one such mechanism because more damaged tissue means more blood loss and more mechanical incapacitation in the case of muscle damage.

Problems with the FBI data:

a. it only has one point per load. It would be more interesting if we had a statistically-valid sample from each load / gun, so we knew standard deviation.

b. it would be more interesting to test the same bullet's terminal performance at a range of velocities.

c. we do not know the expansion profile, as a function of penetration depth. If we knew this, we could compute actual wound volume.


I agree that the old, sage advice of "Pick the most powerful caliber you can shoot well in your gun, and use premium bullets" is excellent when considering personal caliber and firearm choice. I feel adequately armed with a good load in 9mm, .40, 357SIG, or .45, but that doesn't mean we should not understand how incapacitation works, which loads perform, and why.


regards
Zak

[Zak Smith]

I just added two more columns, for both of "bare" and "clothed" gelatin. The fields are:

cu/mv - wound volume (inches^3) per momentum (lb*fps)

cu/e - wound volume (inches^3) per energy (ft*lbs)

-z

[WESHOOT2]

See, we agree.

[CaesarI]

Edit:
Nevermind, you mentioned the problems with calculating wound volume.

Despite the flaws I admire the effort, and agree with the condemnation of 115gr Corbon. Sierra bullets in general are not well made or designed. They can't get them to stay together at higher speeds. Hornady XTP's, and Gold Dots are better in this regard.

-Morgan

[Zak Smith]

Staring at the different views of the data - primarily sorted by bare and clothed wound volume - made me notice a few things, some obvious, some not:

1. terminal performance can be a lot different for a round in bare vs. clothed gelatin,

2. small, fast rounds that depend on lots of expansion to get wound volume (or width) can often be foiled by clothing (e.g.: .40 CCI/Speer GD 155gr: bare, it expanded by 110%, to 0.84, but in clothed gelatin, it only expanded by 54%),

3. bullets that start out smaller must expand to a larger percentage of their original size than a larger bullet to attain the same wound width,

4. less massy bullets have less "material" to work with in terms of expansion - they are stretched "more thin",

5. heavier, slower rounds in .40 (or 10mm) and .45 did extremely well in bare gelatin wound volume, but they ruled clothed gelatin wound volume. Their slow velocity did not prevent expansion. e.g.: the 230gr Rem Golden Saber .45 ACP at 871 fps expanded 58% in bare, and 62% in clothed gelatin!

6. the best 9mm rounds tested, according to wound volume in both cases, are the: CCI/Speer Gold Dot 124 +P, and the 147gr Black Talon. This made me re-evaluate my "carry" choice in 9mm, since it was a Proload/GD 115+P. I switched to 124+P.

7. the best .45ACP loads and the best .40 loads were pretty close. "heavy" 9mm's were noticably behind. "light" 9mm's were last,

8. While some 9mm 115's did well relative to the 124gr's and 147gr's in bare gelatin, they scored the worst in clothed gelatin,

9. 357SIG performs similarly to CCI/Speer GD 124 +P. The extra 100 fps didn't help much, either in penetration or wound volume.

10. pick a caliber you like, and then research the best bullet for your platform (as WESHOOT2 said), since the particular load makes a much bigger difference than caliber in some cases (e.g.: a 9mm beating a .45),

10a. stated another way, there are some general trends in terms of wound volume by caliber, but there is significant overlap based on particular load performance,

11. If a round fails to expand "normally" from clothing, there is a good chance it will over-penetrate.. sort by "(clothed gelatin) penetration depth" and notice how many are over 20" penetration,

The big caveat in these conclusions is that they are just one data point per load, and most of the data is pretty old. I wonder if some new loads would perform differently - e.g. would a 357SIG at 1450 FPS perform much better than this one at 1372fps?

Morgan - yeah, the computed wound volume is a bit of a stretch, since we don't know the profiles. These numbers should be about proportional to the actual wound volume, anyway.. the assumption being they expand with a relatively consistent profile, since this is a "smooth" medium, after all.

[BigG]

Maybe your analysis supports the old contention to use the biggest heaviest bullet you can find in a given caliber?

Of course, the ammo makers don't want to tell you that hardball or lead will do as well or better than their dollar apiece premium ammunition. That would be bad for bidness. JMHO

[Zak Smith]

Well, judging the FBI data based on wound volume for clothed and bare gelating does support the conventional hierarchy: .45/10mm - .40 - 9mm.

The heaviest bullets in each caliber generally did do the best in the clothed gelatin tests, but were pretty close in some cases. In bare gelatin, this was not the case, as the 9mm 124+P beat the 147 Black Talon, for example.

Of course, the ammo makers don't want to tell you that hardball or lead will do as well or better than their dollar apiece premium ammunition.

No! This dogma is not reflective of reality.

It is totally wrong and not supported by the data. If we are judging based on wound volume, expansion is critical. Look at the best performers for that metric - they all expanded 50-60% in the clothed gelatin.

For example, take the best performer for wound volume in clothed gelatin, (45ACP Rem G.S. | 230 gr @ 871), it expanded to 0.73 and penetrated 18.85 inches, giving an approximate (as discussed before) wound volume of 7.89 cubic inches. If this round did not expand, and we assume it would not penetrate more (false - it would), the wound volume would be only about 3 cubic inches. Expansion is critical.

-z

[BrokenArrow]

I think ya goofed a little? The 124 +P GD is 20.25/.53 not .58?

The 147 BT did 16.35/.61. Looks like that "beats" the above to me. As far as wound volume anyway, especially if you only measure/score to 18 inches (forget that?).

Many loads that penetrate over 18 inches in gel are _not_ overpenetrationg in actual shootings; can take some oomph to stretch and punch through skin/clothes on the far side, not to mention ribs, shoulder blades, sternum. Looking at some of the guys in my gym and the local biker bars, 18 isn't "too much".

Study was done with San Diego PD looking at performance in real people in real shootings (over 150) with the 147 9mm JHP. Results were very close to gel: avg was about 13/.60.

Stuff can vary from lot to lot, tester to tester. FBI got 17/.58 w older 155 GD, CPRC (Canadian Police Research Centre) got 13.66 shooting through their heavy insulated storm coat and shirts. In bare gel the FBI got 10.7/.84, the CPRC got 13.2/.72. Who knows what your round will do? Might do this when you want that?

New versions of the Gold Dot after cloth:

155/40 GD 18.1/.57
165/40 GD 15.8/.60
180/40 GD 17.5/.60
200/45 GD +P 18.8/.55
230/45 GD 18.9/.59
124/9 GD +P 20/.53
147/9 GD 18/.55
125/357 GD 19/.54

Might as well flip a coin?

This stuff is more useful in determining which ammo to buy than which ammo will "stop" better. Decide what you want it to do, then shoot, measure, score, write the check.

Hit em first, hit em best, hit em often, don't give up!

[Zak Smith]

BrokenArrow - Yes, I have two mistakes. I misquoted the expansion in my first message, and the 147gr BT did have more wound volume.

It's hard to read the table with so much data, sometimes. I plan to build a real database with a web front-end for the data. That should make it easier to analyze.

As far as over penetration, there are several that penetrate 22" or more after clothing. I don't know if that is "over penetration", but it should be noted anyway.

regards
-z

[Zak Smith]

BrokenArrow - what is the source of the "new GD" data?

thanks
Zak

[adad]

Great job Simthz!

I've been wanting to do something like this for quite a while, but never got around to it.

Someone as analytical as you would really appreciate the book "Bullet Penetration" by Duncan MacPherson. You can get it here: http://www.iwba.com . I highly recommend it.

BTW: I wouldn't bother including sort options for momentum and kinetic energy -- there's no correlation between penetration depth and either of these two factors. Here's my favorite quote on why kinetic energy is pointless when it comes to the terminal performance of bullets:

"He compares the kinetic energy of the M16A1 55 grain bullet to that of the 405 grain lead 45-70 bullet. They have the same kinetic energy and he claims, therefore, they are 'capable of producing the same sized temporary cavities' (which, according to Dr. DiMaio, equates to their being of equal "effectiveness"). Actually, the 45-70 produces a smaller temporary cavity than the M16: a greater percentage of its kinetic energy goes into producing its large permanent cavity -- a far more reliable tissue disruption mechanism than the temporary cavity."

"The claim that the M16 has the same killing power as the 45-70 is simply absurd. The absurdity follows from the fallacy that kinetic energy is the sole measure of bullet effectiveness (tissue disruption), or that the size of the temporary cavity determines the effectiveness of rifle bullets. The deep penetration and the 45 or larger caliber hole (depending on degree of bullet expansion) made by the 405 grain 45-70 bullet is the factor that gives the 45-70 bullet the capacity to bring down the largest game on this continent: yet this permanent tissue disruption is completely ignored by Dr. DiMaio's temporary cavity theory."

This quote is from a book review of "Gunshot Wounds -- Practical Aspects of Firearms, Ballistics, and Forensic Techniques." The book review can be found in the Fall 1999 issue of Wound Ballistics Review which is published by the IWBA (www.iwba.com ). The review was written by Martin L. Fackler, MD, FACS and Richard T. Mason, MD, Medical Examiner of Santa Cruz County, CA.

One other thing: you might be surprised how fast expanding ammo expands. When the news about the new expanding FMJ rounds from Federal was just hitting the gun mags and no one but LEOs could get the ammo, the designer (Tom Burczynski
) who came up with the EFMJ posted some pics on TFL of an EFMJ bullet mid-flight, right after going through just two inches of gelatin -- and it looked fully expanded.

Here's a thread with some penetration/expansion data on the EFMJ from (who else?) BrokenArrow: http://www.thefiringline.com/forums/...t=efmj+federal

If you do go ahead with your database, you ought to consider making requests of those who do this sort of testing semi-regularly (BrokenArrow comes to mind, of course, and I've seen someone's data on the 32ACP on the web somewhere). But you'd have to insist on good quality control: bb velocity/ penetration and correction is a must.

I think you really might inspire many folks to finally get started doing some gelatin tests of their own. That would be great! The more (good) data points the better!

Take care and good luck.