thoughts on 3 vs 5 shot groups

[taliv]

i was doing some chronograph work this weekend, measuring the differences in velocity with cartridge temperature (using an ice chest) for different powders shooting 5 round groups over the chrono. while evaluating the data this morning I noticed something interesting that has me wondering if there may be some other explanations for why 5 shot groups are notoriously more difficult than 3 shot groups.

what i noticed was that the first 3 rounds for some powders seem to be much more consistent than the last 2 rounds out of 5. for other powders, they seem to be a little more random.

example:

2972
2974
2972
2985
2981
------
using 5 shots:
2,977 AVG
6 SD
13 ES

using 3 shots:
2973 AVG
1 SD
2 ES

another example:

2963
2962
2965
2958
2956
------
using 5 shots:
2,961 AVG
4 SD
9 ES

using 3 shots:
2963 AVG
2 SD
3 ES

another example:
2913
2918
2913
2895
2900
------
using 5 shots:
2,908 AVG
10 SD
23 ES

using 3 shots:
2915 AVG
3 SD
5 ES

so, out of 4 groups using this powder, my extreme spread went from
13 --> 2
9 --> 3
23 --> 5
5 --> 5


group sizes with all 5 shots vs first 3 shots were
0.638 0.212
0.427 0.279
0.859 0.406
0.602 0.422

averaging 0.632 vs 0.33 or almost half.

i guess what i'm saying is that while there's no way for me to isolate my mental and physical component and just evaluate the rifle, I am fairly certain flinching or aiming errors or slapping the trigger on my part would not affect the chrono results. something seems to be going on with the gun after 3 shots. however, i'll say again the other powder i tried didn't behave the same, so that also may account for different experiences people report.

what's odd is that i don't believe it is related to the gun "warming up" per se because of the 4 groups, the last 2 rounds were SLOWER than the first 3 in 2 of the groups and unchanged in another. The last 2 rounds were only higher velocity in 1 of the 4 groups. at least, if it was some function of the gun "warming up", it wasn't related to the temperature of the propellant.

anyone performed similar tests?

 

[MrBorland]

Just curious - did you only analyze groups +/- the last 2 shots? You'll need to look at all possible 3-shot groups (shots 1,4&5, shots 2,3,&4, etc)? If there's something special about the first 3, you'll need to compare that group to every other possibility. Otherwise, you're just showing that 3 is smaller than 5, methinks.

 

[taliv]

right, only the first 3 shots vs a total of 5. what's special about the first three is that more often than not, people stop after the first three and don't continue on for the next two.

edit:

you're just showing that 3 is smaller than 5, methinks.

ha! well at least then i wouldn't be wrong

 

[MrBorland]

I just dumped your chrono data into excel an looked at SDs for all 7 3-shot combinations, and it does look like the 1st 3 (1st 2, actually) are the tightest, though you can pretty much see that by simply looking at the data.

It's been my understanding that SD & ES isn't a very good proxy for group size, though, so it'd still be interesting to do a similar analysis on 3-shot group size for all 7 combinations.

 

[taliv]

i don't know if it's a reliable proxy for group size at 100. but it gets very important for the 500+ yard targets i spend most of my time shooting. especially if you want 1st round hits

 

[Andrew Leigh]

I have almost identical experience. My SD's are not as tight as yours though, those in my book are excellent.

Given the ES and SD's you quote there should be little no vertical displacement as a result and those loads are capable of one holing as illustrated by your 3 shot groups.

Although there may be some element of barrel temperature the velocity swing is still not enough to be material at 100yds. I have agonised over this for some time now and have come to the conclusion that I do not have the capacity, from a concentration point of view, to string 5 consecutive shot together. After 3 great shots I get cocky and drop concentration levels.

The other point is that statistically for every increase in shot count there is an increase in risk of throwing one away. On local forums I am known for my stance on flyers ...... I don't believe in them, only bad shots. This is my personal stance as accepting flyers to my mind accepts the possibility that you can discount poor shooting blaming it on equipment or loads.

There is an article by Chris Long on Group Size Diagnosis which I found most interesting. http://www.the-long-family.com/group_size_analysis.htm

 

[taliv]

After 3 great shots I get cocky and drop concentration levels.

that's kind of the notion i've held for years but am challenging now. i usually open groups up a bit in the 4th or 5th shot of a group. always blamed myself, but looking at the chrono, i'm wondering if it's maybe the velocity change

 

[MrBorland]

i don't know if it's a reliable proxy for group size at 100. but it gets very important for the 500+ yard targets i spend most of my time shooting.

but looking at the chrono, i'm wondering if it's maybe the velocity change

If SD/ES becomes a reasonable proxy for groups at long range, and it, not the shooter, was responsible for groups opening up, a good ballistic calculator ought to make a decent prediction about your groups when using your actual chrono data, no? Just thinkin'.

 

[Andrew Leigh]

I'll tell you my reasoning.

I am not a bench shooter but take my hunting loads rather seriously .

I was referring to my 30-06 hunting rifle rather than any specialist piece of kit. When defining the trajectory path in the Sierra Program them a 180gr. bullet at 2 600fps and a 180gr bullet at 2 620fps would show a vertical change of 0.05" which relates to 16.2% of the bullet diameter.

Now that is with a slow and heavy bullet and yet 20fps results in fractional vertical displacement.

I will generally throw shot 4 and then refocus.

Another point I have noticed is that in my case as I am a hunter so I elect to hold the stock forearm off the bench as I would like to simulate my hunting grip as much as possible. I have noted that variable forearm grip is an enemy of good groups as the firmness of the grip will alter muzzle lift. You are probably not holding the forearm. I learnt the lesson with my .375, that rifle taught me to hold on a little tighter .

 

[taliv]

a good ballistic calculator ought to make a decent prediction about your groups when using your actual chrono data, no?

it most certainly does. at distance, you can see it on the target and chrono at same time. they match up well. (at least as far as vertical dispersion is concerned. horizontally, wind comes into play and makes it much more difficult to tell whether it was the shooter, bullet or wind)

When defining the trajectory path in the Sierra Program them a 180gr. bullet at 2 600fps and a 180gr bullet at 2 620fps would show a vertical change of 0.05" which relates to 16.2% of the bullet diameter.

that's just the exterior ballistics of the bullet flight though without taking into account the rifle/barrel. where velocity makes a bigger difference in group size is barrel harmonics. that's why people do ladder tests, and why sometimes you add powder and velocity goes up but your group prints lower on the paper than the last one. because it's about where the barrel is vibrating when the bullet exits.

 

[Andrew Leigh]

that's just the exterior ballistics of the bullet flight though without taking into account the rifle/barrel. where velocity makes a bigger difference in group size is barrel harmonics. that's why people do ladder tests, and why sometimes you add powder and velocity goes up but your group prints lower on the paper than the last one. because it's about where the barrel is vibrating when the bullet exits.

I have always meant to actually shoot paper out to 300m to actually prove what my rifle is actually doing. I don't hunt past 300m and am at better than 0.7MOA so from a hunting point of view I am OK.

You boys are different and us hunters follow the target shooters lead when it comes to accuracy improvements. Think I will have a go at 300m next time out.

Good luck in solving your matter.

 

[tuj]

The only thing I could think of that would explain the drop in velocity or a variation, and believe me, this is reaching, is the barrel temperature. If you were shooting those shots 5 at a time and then letting the gun cool down while you reload or whatever, I could see where there might be an issue of heat around the chamber that is perhaps dropping the combustion process completeness slightly.

You did say you were putting the rounds in an ice chest before shooting? Can you tell us more about your test procedures? How much data did you gather and in general, were the last two rounds slow or faster and what percentage of the last 2 rounds in 5 shot groups were both slower than the average of the previous 3?

 

[taliv]

i put 5 rounds of each powder in a ziploc bag with my kestrel (thermometer) in a cooler with an ice substitute that had been in the freezer to drop the temperature of the ammo.

i removed both and set them near the gun, and fired fairly quickly. my goal was not not let a round sit in the chamber for more than 2-3 seconds because i didn't want the chamber temp to undo the freezer work. I probably averaged 10-12 seconds or so between rounds for 10 rounds, and i alternated between the two groups (one round from one powder then one from the other) so they would experience similar rise in barrel temp.

then i waited a long time for the rifle to cool and repeated that with ammo at ambient temp. after that, i went home, loaded more and returned, so rifle was cool again, and repeated with a different charge. for a total of 8 groups of 5.

tuj, the last 2 rounds were slower velocity than the first three rounds in 2 of the 4 groups and unchanged in another. The last 2 rounds were only higher velocity in 1 of the 4 groups.

 

[tuj]

Well the old sniper's trick goes something like, if you leave cartridges in the sun, the hotter powder will shoot faster. I have no idea if that's true or not. But the one thing I could think of is that your 4th and 5th shot spent the most time out of the cooler, sitting on the table.

I think it would be interesting to repeat the test, but only remove 1 round from the cooler at a time, thus the shot to shot consistency should improve.

 

[taliv]

but as i said, of the 4 groups, the last two rounds were actually SLOWER, more often than they were faster

i would have, but i can't reach into the cooler without losing my cheek weld haha. i'll need an assistant next time.

 

[Owen]

There is a really great thread on ARFCOM about this topic.

http://www.ar15.com/forums/t_3_118/279218_The_Trouble_With_3_Shot_Groups.html

 

[tuj]

Get one of those 'lovely assistants' the magicians use.

But in all seriousness, you've come across something interesting. I think it probably is a test that bears repeating with maybe a slightly different procedure. Or even the same procedure and see if you get the same strange results. Sometimes when we experiment, we come across aberrations that cannot be reproduced. But if it is a reproducible phenomena, then we must all put our thinking caps back on.

Owen's link makes a great point, but it doesn't address what you saw on the chrony exactly.

 

[Comrade Mike]

I shoot 5 shot groups personally. Easier on the wallet

 

[DRYHUMOR]

Interesting.

Curious what sort of temps the bolt face picks up in the firing process. Those temps may affect the stability of the primer and adjacent powder...?

Me thinks I'll take an IR temp gun to the range next time.

 

[Steel Horse Rider]

I think the 4th and 5th shot variations are more related to barrel temperature than to round temperature. I am always surprised at how fast my bolt guns heat up if I fire a magazine in rapid order. Still, for a hobby shooter like myself I am impressed at the uniformity of your loads even if they are not up to your expectations.

 

[jerkface11]

How long are you waiting between shots? When shooting for group size I always set a timer so that the shots are evenly spaced out.

 

[blarby]

Interesting.

On the warming up though- when metal heats, it expands.

This would change pressure, and thus velocity.

It would happen not just in the BBL, but in the chamber as well.



Easiest way to rule that out would be to take 5 shots over the chrony < 3 minutes apart... or whatever it takes for your rifle to fall to the temp of your first shot.


I'm not sure how practical that would be for competition shooting, but from a ballistics standpoint it would give you your answer.

 

[denton]

It is not the temperature of the powder, or the temperature of the air that matters. It is the temperature of the steel in the barrel, especially near the chamber, and it is the temperature of the brass and lead in the cartridge.

If you want to eliminate one variable, you can just leave the round in the chamber for a couple of minutes before firing. The barrel and cartridge will then be at practically the same temperature.

The mechanism for MV loss with colder barrels and ammunition is the robbing of heat energy from the propulsion gas. Colder metals rob more energy than warmer metals. The quantity of heat transferred between two bodies is directly proportional to the difference in their temperatures.

The variables have to be well controlled before you can make a judgment about whether something has affected standard deviation of MV. Think thermocouples attached to barrels. And the stats to decide whether two SDs are really different from each other require fairly large sample sizes.

 

[taliv]

shot one last group tonight (working up a load)



3001
2999
3003
2999
3001

avg 3001
SD 1.7
ES 4

so this one didn't have much difference for the last two rounds... i don't know why. however, it was a different load and was almost exactly 100% case capacity. who knows?


i wouldn't expect the bolt face to get hot at all. at least, not like the chamber.

you're probably right about the barrel temp, but i wonder why that isn't more consistent.

i'm waiting maybe 10-12 seconds between and trying to shoot within 2-3 seconds of closing the bolt to minimize the time the cartridge is exposed to the chamber temp.

while the temp of the barrel may matter, the temp of the powder and air definitely matter. most good ballistics calculators will allow you to input the temp of both the air and the powder. i haven't seen one that allows you to input the temp of the bbl.

i do need larger sample sizes (but they take time and money)

 

[jerkface11]

Try waiting two minutes between shots with the muzzle pointed up.