Surprising ES/SD with Leverevolution

[Harpoon1]

Have had a keg of Lever laying around for a couple of years now and finally got around to trying some. Wow. Can’t believe the ES/SD numbers I’m seeing with it, especially being a spherical powder. Maybe Im late to the party and shouldn’t be surprised.

Anyway...........

Howa HB 20”
Lever/Hornady Brass/ CCI 400
123 ELD-M
2.366 COL
1.770 BTO
-.020 Jump

50F/ sea level

Labradar

3 shot groups/.4 increments

30.1 2433/25/14

30.5. 2477/62/31

30.9. 2500/6/3

31.3. 2533/10/7 (only 2 shots recorded)

31.7. 2562/4/2.6

32.1. 2603/16/8

32.5 2618/3/1.8 (max per “Grendel Reloading Handbook” Vol. 2 and NOT pressure tested!)

Shot 4 more single rounds @ .2 increments looking for pressure

32.7 2634

32.9 2630

33.1 2645

33.3 2658

No signs of pressure noted and all 4 of the last single shots went into a 1 MOA group!


Are those normal numbers for Leverevolution?

[mel]

i think i need t get my hands on some of that powder , 32.7 is prolly as high as id go seeing that speed/dip no increase at 32.9

[Harpoon1]

i think i need t get my hands on some of that powder , 32.7 is prolly as high as id go seeing that speed/dip no increase at 32.9

I was thinking, right in the middle at 32.8!

I’m gonna load 5 of each, 32.7-32.9 and see what the ES/SD looks like right there first, before I do anything brash.

[tdbru]

Harpoon1,
remember, the SAAMI pressure spec for the 6.5 Grendel is pretty low. if you see "typical" pressure signs, i.e. flattened primers, you're likely way over SAAMI pressure limit. no issue with the Howa, but not so good if you happen to be driving an AR. Good luck with the load development though. and good that you have enough Leverevolution for a lot of shooting, as reloading components are pretty scarce up this direction right now.
-tdbru

[Yondering]

i think i need t get my hands on some of that powder , 32.7 is prolly as high as id go seeing that speed/dip no increase at 32.9

You can’t accurately judge something like that based on a single shot, especially with such a small difference in powder charge. It’s almost meaningless in regard to pressure or accuracy potential of those loads.

Be careful not to infer too much from limited sample sizes. People seem to jump to the conclusion that written data is good data, but that’s obviously not true if we stop to think about it, and can be dangerous.

[grayfox]

Your 30.9/ 2500 load is smack dab on the spot where several 20” hb’s have been shooting, mine (now sold) included.
I am back however with a 22” barrel and so far it’s a happy one.

[Yondering]

Are those normal numbers for Leverevolution?

Yes, those seem about right. It likes to be run hot, and generally gives very consistent results while delivering higher velocity than you can get with anything else in this case size.

A number of us have been recommending Lever for exactly those reasons for a handful of different AR15 cartridges.

[Harpoon1]

It’s not the velocity I find all that surprising, it’s the number of single digit ES/SD spreads in the group. I’m gonna try it next in the CZ with a 24” barrel and see what happens. I will be restricted to a much shorter COL, so, should interesting to compare the differences.

[Harpoon1]

You can’t accurately judge something like that based on a single shot, especially with such a small difference in powder charge. It’s almost meaningless in regard to pressure or accuracy potential of those loads.

Respectfully, many think you can, and have been successful. That’s exactly how you perform the “Satterlee” method of load development. Single shot per charge, in small increments, to locate stable combustion nodes.

Here is but one example

https://youtu.be/paUMACbJxTs

[Harpoon1]

Harpoon1,
remember, the SAAMI pressure spec for the 6.5 Grendel is pretty low. if you see "typical" pressure signs, i.e. flattened primers, you're likely way over SAAMI pressure limit. no issue with the Howa, but not so good if you happen to be driving an AR. Good luck with the load development though. and good that you have enough Leverevolution for a lot of shooting, as reloading components are pretty scarce up this direction right now.
-tdbru

Yep, I’m familiar but thanks for the thought. Safety first.

[Fess]

Does anyone know exactly what Hodgdon did to give LVR such unusual properties, especially with the 30-30? 200fps over traditional loads in nothing to sneeze at. My best guess is careful blending of powders with different burn rates, kind of like what they do with Superformance.

[Nordy55]

30.7 under a 120 speer gold dot, 120 rdf shoots both sub moa and 24xx out of 18in tube using magnettospeed chrono

[Yondering]

Respectfully, many think you can, and have been successful. That’s exactly how you perform the “Satterlee” method of load development. Single shot per charge, in small increments, to locate stable combustion nodes.

Here is but one example

https://youtu.be/paUMACbJxTs

Sorry, but “many people” are foolish when it comes to thinking about data variation and sample size. Most results you get from a single sample test will be different if you repeat the same test again. The method above is a particular example of that. Just because people use it and draw conclusions from it does not mean they drew the right conclusions or that it’s a good method.

Single sample results, especially small powder charge increments like you used, place each sample inside the variance range of the samples around it, and to make it worse, with a single sample test you have NO idea what that variance (ES,SD, etc) is. If you shot 5 shots of each instead, you’d see the average velocity be something different than your single samples, and that supposed lack of velocity increase might actually be a normal linear increase and the single samples were just within the ES at each charge.

[Yondering]

It’s not the velocity I find all that surprising, it’s the number of single digit ES/SD spreads in the group. I’m gonna try it next in the CZ with a 24” barrel and see what happens. I will be restricted to a much shorter COL, so, should interesting to compare the differences.

That’s what I said too. The consistency you’re seeing is part of why some of us have been recommending Lever. It’s only just recently caught on in popularity though and people are acting surprised to discover what we’ve been saying all along.

[jasper2408]

Well thanks Harpoon1! I wasn't going to test LVR but now you made me go and order 3 pounds to see what all the commotion is about.

I noticed you used a CCI400 primer. Looks like it performed alright for you.

Seems like everybody that has tried LVR has had success with it.

Thanks for the informative post.

[Harpoon1]

Well thanks Harpoon1! I wasn't going to test LVR but now you made me go and order 3 pounds to see what all the commotion is about.

I noticed you used a CCI400 primer. Looks like it performed alright for you.

Seems like everybody that has tried LVR has had success with it.

Thanks for the informative post.

I have a bunch of 400’s, so, trying to save the 450’s and 41’s specifically for the AR’s in these austere times!

[Harpoon1]

Sorry, but “many people” are foolish when it comes to thinking about data variation and sample size. Most results you get from a single sample test will be different if you repeat the same test again. The method above is a particular example of that. Just because people use it and draw conclusions from it does not mean they drew the right conclusions or that it’s a good method.

Single sample results, especially small powder charge increments like you used, place each sample inside the variance range of the samples around it, and to make it worse, with a single sample test you have NO idea what that variance (ES,SD, etc) is. If you shot 5 shots of each instead, you’d see the average velocity be something different than your single samples, and that supposed lack of velocity increase might actually be a normal linear increase and the single samples were just within the ES at each charge.

I would tend to agree with your observations, however, how do reconcile those who have great success with the “single shot, small increment” method to find a stable node?

For example, here’s someone who is a member of the USA F-Class team. It’s the method he promotes and here is a sample of his success using that method of load development.

https://youtu.be/f5j7VzBX7GU

[grayfox]

400's will do fine for a bolt gun.

On a separate note, I think I will weigh in on the discussion concerning es/sd of load sets developed by handloaders.
The typical sampling and standard deviation, etc, activities concern some existing universe of components, and taking some random sample of those to look for data such as mean, median, std deviation etc. The larger the sample, the more accurate your findings, in general. This would also apply to factory ammo bought as those boxes of rounds are indeed samples of some larger universe -- which presumably one could continue to buy and shoot -- a competitor who finds one such lot for his rifle, usually goes back and buys a lot of that lot#. And this type of random testing is/was also used appropriately in several of my multi-million $$ nuclear regulatory projects, where we were engaged in verifying adequate nuclear safety/licensing of large scale plants, for protecting the health and safety of the public.

For handloaders, however, the issue is a bit different; not totally but somewhat, let me explain what I mean.
I can't speak for every loader but for, say, a relatively conscientious one that is attempting to find accurate loads for his/her rifle. In this case, es/sd plays a part in the load's performance but also has partly to do with the process and various error factors of the loader's routine.

Example: take uniformly sized and re-sized cases, uniform primers (same lot), a single bullet of some percentage consistency (0.3 gr/100 or so), a single lot of powder, an overall length (+/-0.003 for example) or OAL, all within some manner of tolerances, and powder that is weighed and held to, say, +/- 0.1 gr accuracy (out of say 30 grs) of the powder scale. All of those variables have some small % of variability, so if a load, say, exits the muzzle at a consistent node/MV/angle/etc, then one can say such a load could be accurate in one's rifle. As long as you can continue to produce this load with that process and those tolerances, then its performance will be in accordance with the % deviations of the variables listed above... some of which won't have a discernable amount of MV delta, but some will. In those cases, comparing one set of cartridge load combo "A" against a differing set of combo "B" (ie, increasing powder charges for example), one can use es/sd of those load sets as an indicator of potential accuracy. [Certainly more load sets of equal combo are better than a single one.]

In my case I would estimate that, in somewhere up to about 95% of my "good-node-accurate-loads" the resulting MV data remains tightly within an MV range and I would say, I have a good load and the es/sd's also are very good. I can verify that load over several sets of it, over different days at fairly consistent environmental conditions (I shoot at a temp-controlled indoor range).

I'm not a competition shooter so I can't take it to that level, but in my shooting and load development, I am confident that when I find such a load for one of my rifles, every time I have loaded back up that combo, it shoots just like the initial sets I had, with not equal but definitely closely agreeing, es/sd/accuracy/poi etc results.

So I think as loaders we can make a case for using our es/sd values as guides to finding that accurate load, BUT it begins with a boringly consistent and accurate reloading technique.

Therefore, I do think there is a useful case for these in our load development efforts.

[Lemonaid]

Very good 3 part series on precision rifle blog. Check it out! https://precisionrifleblog.com/2020/12/0...-shooters/

[Yondering]

Example: take uniformly sized and re-sized cases, uniform primers (same lot), a single bullet of some percentage consistency (0.3 gr/100 or so), a single lot of powder, an overall length (+/-0.003 for example) or OAL, all within some manner of tolerances, and powder that is weighed and held to, say, +/- 0.1 gr accuracy (out of say 30 grs) of the powder scale. All of those variables have some small % of variability, so if a load, say, exits the muzzle at a consistent node/MV/angle/etc, then one can say such a load could be accurate in one's rifle. As long as you can continue to produce this load with that process and those tolerances, then its performance will be in accordance with the % deviations of the variables listed above... some of which won't have a discernable amount of MV delta, but some will. In those cases, comparing one set of cartridge load combo "A" against a differing set of combo "B" (ie, increasing powder charges for example), one can use es/sd of those load sets as an indicator of potential accuracy. [Certainly more load sets of equal combo are better than a single one.]

While that sounds great in theory, velocity variation doesn’t work that way. If it did, you’d see the same ES/SD for each powder charge for a given load workup, but we all know that’s not true.

On top of that, lowest ES does not repeatably correlate to best accuracy. Yes, it’s necessary at extended range, but inside a few hundred yards if you actually measure both accuracy and velocity for each group in load development you will see what I’m saying. Best accuracy will happen at a certain powder charge or seating depth, and lowest ES will happen at a different one.

As to Harpoon’s example above, being a great shooter doesn’t mean you have a great understanding of statistics and interpreting data. Sorry but I don’t care to spend much time watching YouTube videos, so if there’s a particular point you’re trying to make in that video just say it here.