More Ammo Testing…On The Way!

Big Brass Ones

After looking back at my ballistics testing, I find myself very troubled by the results of the .38 Spl ammo tests.

In addition to testing more 9mm in the short-barrel Kahr, and more .40S&W in the Glock and the carbine, I procured a batch of different .38 ammo to test. Oddly, not a single retailer stocked the old standby “FBI Load” (.38SPL +P with a 158gr lead semi-wadcutter hollowpoint)

I will be re-testing the Speer Gold Dot short-barrel .38 ammo.  I will be using two .38 snubnose revolvers to verify my findings, and a 4″ gun to give some perspective from a “typical” .38 revolver.

(More)

In addition, I blackened the slugs of the Winchester Silvertip and SXT ammo with a Sharpie, and will be re-testing them as well:

“Mister, we deal in lead”

I wanted to make something more clear as well.  I do not believe that velocity or muzzle energy by themselves are the sole determinant of an ammunition’s effectiveness.  Expanding bullets are designed to expand properly within a range of velocities, with intervening materials like wallboard, glass or clothing negatively affecting their performance. If they are moving too slow, they don’t expand.  If they are moving too fast, they fragment and reduce the depth of penetration (this latter problem isn’t going to affect the .38 ammo).  The best bullet design in the world won’t perform as promised if it isn’t within its performance envelope.

I am trying to see how ammunition performs in my own guns, both from an average velocity perspective, and from a consistency perspective.  If the results I get show that a type of ammo performs close to the manufacturer’s stated velocity, then I feel confident that the bullet will perform as promised.  But if I see a large standard deviation, then some of the rounds will fall considerably short of the average, which could effect performance.

After I find the ammo types with the highest and most consistent velocities, I will make my choice based upon the performance of the bullet designs.

Share and Enjoy:
  • StumbleUpon
  • del.icio.us
  • Facebook
  • Twitter
  • Google Bookmarks
  • Reddit
  • email
  • RSS

Related posts:

4 Responses to “More Ammo Testing…On The Way!”

  1. Ed Harris says:

    In .38 Special some +P loads use slower powders which are optimized for the 4 inch vented test barrel, but perform poorly in revolvers with barrels shorter than 3 inches. Cylinder gap also plays a role here, because a tightly built two-inch gun with minimum cylinder gap of 0.003″ may produce higher velocities with some ammunitions than a four-inch gun having a maximum cylinder gap of 0.008″ As a general rule in .38 Special standard-pressure, lead-bullet ammunnition you can expect a delta-V of 10 f.p.s. for each 0.001″ change in cylinder gap.

  2. Skinnedknuckles says:

    I have 35 rounds of Federal 38 Special (+P) High Velocity 158 Grain (lead) Semi-Wadcutter Hollow Point ammo left from a box purchased last year from Hoffmann’s. I found I couldn’t use it in my revolver because of unacceptable leading (oversized throats). If you want to test it send me a PM. I’m in the Winsted area and I’d be curious to see the results.

  3. Ed Harris says:

    More data from my old test logs:

    Velocity vs. Barrel Length In .38 Special:
    Firearm:———————– Ruger SP101——Ruger Speed Six——–Ruger Service Six—–Marlin1894CB
    ————————————.38 Spl. 2”———.357 Mag. 2-3/4”——-.38 Spl. 4”—————.38 Spl. 18”
    Ammunition: (cylinder gap at 0.06” Mean Assembly Tolerance Throughout)——Solid Barrel Carbine
    Winchester 110-gr. STHP—–851, 19Sd———-886,33Sd——————896, 35Sd————1037, 57Sd
    Winchester 125-gr. STHP+P–737, 14Sd———-not fired——————956, 39Sd————1159, 25Sd
    Federal 147-gr. JHP+P+-LE—-922, 15Sd———-952, 33Sd—————-1065, 40Sd———–1153,44Sd
    Win. X38SMRT 148 HBWC—-682, 25 Sd———734, 11Sd—————-783, 7 Sd————–879,13Sd
    Remington 148HBWC———703, 5Sd————741, 25Sd—————-752, 14Sd————–850, 14Sd
    Norma 158LRN——————721, 14Sd———-743, 14Sd—————-796, 14Sd————–985, 7Sd
    Win. X38SPD 158LHP +P——815, 16Sd———-831, 36Sd—————-879, 21Sd————–1093, 19Sd

    • Ed Harris says:

      CORRECTION – forgive old fumble fingers….

      Cylinder gap was 0.006 inch throughout

      six thousanths of an inch, is “mean assembly” tolerance.

      Minimum gap before proofing front gage must pass 0.003 without resistance with rear gage in place.

      Maximum after proofing must feel resistance on 0.008 front gage with rear gage in place.