The Demon Chronograph

For my birthday, my girlfriend got me a chronograph.  It pays to be specific about birthday wishes.  My first little test run was a real eye opener.  First, my bullets were quite a bit slower than I expected.  They were going about 1150 feet/second.  I have been progressively downloading to my current 62 grains of blackpowder to gain accuracy.  I’m guessing that it is no accident that my muzzle velocity is right at the speed of sound (1126 ft/s).  The second shocker was the amount of variation.  In 9 rounds I had speeds from 1105 Ft/s to 1193 Ft/s. 

It was time to do a little testing.  I decided to make a batch of 20 rounds in my usual manner, and then make a batch with careful control of bullet weight, case weight, and powder load.  The goal is to see if the careful assembly method yields any real benefits.

Because of limitations on components I have available, the loads aren’t 100% the same as what I compete with, but they are a good standard load.

My base line load:

Creedmoor swaged bullet from Buffalo arms.  510 grain advertised wt. at .459diameter. 20:1 lead:tin.
62.5 grains FFG Swiss blackpowder.
Cases are FC (Federal) nickel plated.
Primers are Winchester Large Rifle Magnum (WLRM).
.030 fiber wad between the bullet and powder. 
Lube SPG as delivered from Buffalo Arms.  No lube cookie.

Weighing the cases (cleaned and decapped), they weighed between 180.2 and 191.4 grains.  For the “precision” loads I gathered 14 that were between 190.0 and 190.5 grains.

Bullets ranged from 512.9 to 514.5 in lubed condition.  For precision loads I used a range from 513.0 to 513.1.

For the “precise” powder load I carefully weighed and “trickled” the charge at 62.5 grains.  My scale reads to the tenth, and is pretty responsive, so I’d guess a range of  62.45 to 62.55 is not unreasonable.  For the “random“  load, I set the barrel measure to 62.5 and just threw them as they came out.  This is pretty much how I make my regular ammo, except I measure about 1 in 10 to be sure I’m not wavering into a dangerous loading.  The results were measured and shown on this graph:

Powder weights as thrown sequentially from an Ideal "55" barrel measure.

One interesting find is the progressive increase in the load and the progressive increase in variation.  It certainly looks like a regular check is in order when I’m doing my typical quantity of 60 or so for a match.

The data I expected to see was a “precise” load that had a slightly narrower band of variation than the “random” load.  I also expected that the averages of the 2 loads would be roughly the same, as they are both the same basic load.  I could not have been more wrong.

The average velocity of the random group was 60 ft /s slower than the precise loads.  Also, the standard deviation of the random loads was 6.7, which is better than the 7.74 of the precise loads.

 Muzzle velocity for "random" and "precise" ammo groups.

Another interesting note is the progressive increase in velocity of the random group, but not the precise group.  I can think of two possible reasons for this velocity climb.  First is the idea that fouling increases through the string of shots.  One would expect a lowering velocity with increased fouling, but there is a slight potential that as fouling holds back the bullet, the increased burn temperature leads to more efficient combustion, so the velocity increases through the last portion of the barrel.  Unlikely, I know.  The second, and more likely possibility is that as my barrel warms, the velocity goes up.  This could either be due to a warmer powder charge burning more efficiently, or due to an expanded barrel having less drag on the bullet.  I will say it was a chilly day, and my barrel warms quite noticeably during a string of shots.  This explanation might also explain the higher velocities of the precise loads.  I shot that string directly after the random string, with only a cleaning in between.  The barrel would have been significantly warmer.

For future tests, I may alternate shots between groups.  The more mathematically correct method would be to randomize the shooting sequence.  I also may wire up the barrel with some thermocouples and do some testing at different temperatures.

One bit of information I was interested in was how much effect the velocity veriation would have on elevation placement of the shots.  I used a simple on-line ballistic calculator to give me the drop at 200 yardsm, then corrected the numbers to zero my lowest shot.  The graph looks the same as the velocity graph, but the scale of the Y axis is an eye opener.

Variation in elevation solely due to velocity variation.

It appears a full 2 inches of group height can be attributed to velocity variation.  Also, the differences in elevation of the two groups is approximately 4.5".  It is starting to look like this is a reasonable place to start chasing accuracy.

I’d like to show you my targets and try to correlate the velocities to accuracy, but the weather was a slight fog that improved throughout the session. The change in conditions was great enough that I wouldn’t really trust any conclusions.

All this points to a lot more loads and a lot more testing.  All blame is to be placed on the Demon Chronograph.