Crimp or Neck tension?

[JASmith]

Im not that curious, but shouldnt group size n=30 to be significant?

Thanks
SG4247

Many people, including me, are satisfied with 5-shots IF the difference between the means exceeds 2-3 standard deviations.

The SD tends to get smaller as the sample size grows.

Klem's 80 and 108 data points is unusually large making his data more meaningful than most of our tests.

As he indicates, one is free to draw his own inferences from the data. Eighty data points for one set and 108 for the other seriously ups the temptation to declare the difference in average meaningful.

On the other hand, some hard line statisticians might say that the less than one standard deviation between the two means we cannot tell the difference.

What the data does say, however, is that we have a lot of confidence that the less troublesome neck tension only method gives at least as good accuracy as does crimping.

[Boostmeister]

Prior to retirement, I managed our Quality and Reliability group. I had a sign hanging over the doorway to the lab the said, "Without data, you are just another opinion.". Nice job Klem! Everyone can sift through the data to their hearts content, but the data is there for all to see. Refreshing!

[SG4247]

Okay, lets say the data is solid as is.

My interpretation of Klems excellent work, (please take with a grain of sugar) is that there is no significant difference between crimp and .004" neck tension in this case, under these conditions.

Many other test conditions both controlled and uncontrolled, may have a larger impact on the results of crimp vs no crimp than we see here.

I am suggesting this since the difference in average group size is only .080" (a little more than a 1/16"), makes me wonder if the experiment was repeated with a single different parameter - like using a different primer, would the results then be significantly different?

I guess I would consider the results of this test to be basically one null data point - in the endless world of reloading possibilities.

We know by emperical evidence over time, that some cartridges prefer a crimp over light neck tension, to print tight groups. I dont think we reloaders - will ever have the time or resources to understand why. It is fun to experiment tho!

I hope all that makes sense, this is purely my perspective.

[Texas]

Klem,

Thanks you for your work and the data.

[Klem]

Okay, lets say the data is solid as is.

My interpretation of Klems excellent work, (please take with a grain of sugar) is that there is no significant difference between crimp and .004" neck tension in this case, under these conditions.

Many other test conditions both controlled and uncontrolled, may have a larger impact on the results of crimp vs no crimp than we see here.

I am suggesting this since the difference in average group size is only .080" (a little more than a 1/16"), makes me wonder if the experiment was repeated with a single different parameter - like using a different primer, would the results then be significantly different?

I guess I would consider the results of this test to be basically one null data point - in the endless world of reloading possibilities.

We know by emperical evidence over time, that some cartridges prefer a crimp over light neck tension, to print tight groups. I dont think we reloaders - will ever have the time or resources to understand why. It is fun to experiment tho!

I hope all that makes sense, this is purely my perspective.

I understand what you mean.

For the avoidance of doubt I am not trying to generalize to the world population of guns, or even to this calibre. I am providing a data set and saying what I am going to do about it.

I agree with you that the difference in the groups size means is not statistically significant. If we go ahead and do the t-test the null hypothesis (that there is no difference between crimping and neck tension) is not rejected. That is if we use a statistical significance .05 (meaning 95% of all samples will show no statistical difference between crimping and neck tension) in a two-tailed test.

H1. There is no difference between crimping and neck tension (the null hypothesis).
H2. Crimping is more accurate than neck tension.
H3. Neck tension is more accurate than crimping.
(Hence the two-tailed test because at the two outer edges of the Normal curve it is H2 on one side and H3 on the other).



All this would be important if you wanted to generalize to the population of guns and hand loading. In this case however I don't care that the two methods are not significantly different. I have the luxury of treating the group sizes as if they are absolute values. I can do this because I have nothing to lose by choosing one method over the other. I have already paid for the dies and it is the same time and effort to either crimp or neck size. I am comfortable believing the group averages are a fair reflection of what is likely to happen. Actually, using neck tension is one less step in the loading process so even if you believe the group sizes are not statistically different that would be reason enough to choose that neck tension over crimping.

[rabiddawg]

Let me give you guys something simple to ponder.

I measured my expander at .2596

.264-.2596=.0044

Can I say that I am getting the desired neck tension?

[Klem]

Unfortunately those numbers will only give you a clumsy picture of reality.

I measure my Forster expander ball at .2555" which gives .008" of grip when compared against the calibre - That's not right.

The calibre .264 is only a standard by which bullet manufacturers then make bullets slightly wider so they will engrave into the rifling of a barrel. Manufacturers are free to make whatever widths of bullets work for them so there will be variations in bullet diameter. You want to include the width of the bullets you are using.

Different brass has different elasticity so after the case is shaped by your expander it might constrict a little in diameter, or not. Measuring the expander and using that as your metric is also imprecise.

The best (most realistic) way to measure how much a bullet is being gripped by the case is to measure the outside neck of a sized case, and then measure the outside neck again after a bullet has been seated in it.

[rabiddawg]

Ok, thanks. I will do that.

Fwiw, I did measure my Nosler 100 gr bt at exactly .264

[SG4247]

A couple things related to the subject:

I've been my sorting my match ammo by only bullet seating force.

I have a calibrated digital load cell (my video):

https://m.youtube.com/watch?v=6mA10IVnG1Y

I also have a 21st Century Shooters hydro press:

https://m.youtube.com/watch?v=uZ2QFq6Dfx4

Both devices correlate to each other very well.

What I have learned after a years worth of testing seating pressure is:

Consistent seating pressure, with the lowest bullet installation press force SD, within a given population, provides the lowest vertical spread on groups fired at 300-600 yards.

Ive learned that it is extremely difficult to get brass prepped such that bullet seating force is reasonably consistent (within +/- 10lbs).Once you have brass in this range, vertical fliers are no longer a problem.

[Klem]

Guys,

I spoke with the quant guru's at work today (University). Their advice is that while we can use this data to deduce a Z-Score it will be more credible if the samples are both over 30. It then qualifies for a t-test which has a lot more statistical credibility.

It will take a few sessions and I will post in a few weeks when we have 50 groups each.

Off to the range...

[Rickc]

For me it is a mute point. Only thing.I crimp is pistol rounds.

I wonder how many of the world's best long range shooters crimp?

I wonder what your controls are on the neck tension only brass. Freshly annealed? Neck size with bushing or.expander ball? Bullet seating pressure measured?

Lots of variables that are not being.taken into account.

[Klem]

For me it is a mute point. Only thing.I crimp is pistol rounds.

I wonder how many of the world's best long range shooters crimp?

I wonder what your controls are on the neck tension only brass. Freshly annealed? Neck size with bushing or.expander ball? Bullet seating pressure measured?

Lots of variables that are not being.taken into account.

Thanks for that opinion Rick, I'll bear that in mind.

(I think you mean moot, although 'mute' works for me too).

[dammitman]

damm, just read thru this whole thread and all i can say is dang. now i feel like i gotta go try some of this myself. everytime i feel like i am a decent handloader i learn something really interesting and i feel like everything i have done up till then was just guessing. thanks klem. i also like what SG4247 said,,,Consistent seating pressure, with the lowest bullet installation press force SD, within a given population, provides the lowest vertical spread on groups fired at 300-600 yards,,,now thats another nice piece of tech.