Petey308

Well-Known Member
Joined
Aug 23, 2018
Messages
334
Location
Missouri
So, I plan on discussing this in a video at some point, but figured if you had the time/motivation to do some reading, I'd share a topic/discussion point regarding SD/ES and my thoughts on the matter, specifically how to reduce it. This is the more advanced version of my basic load development as well. Anyways, here goes:

Low SD/ES has much more to do with brass prep and loading quality. It's not so much an indicator of a node as some think it is. You can shoot some great groups with a high SD/ES and you can shoot some terrible groups with a low SD/ES. Your accuracy/group size is influenced more on the consistent release/exit of the bullet from the muzzle and where that muzzle is pointing each time that occurs. That is influenced on the whip/harmonics of the barrel and that in turn is influenced by the pressure curve of the powder charge and everything else occurring when the round is fired. Getting your harmonics consistent (getting into an accuracy node) typically goes hand in hand with lowering SD/ES because it requires getting things consistent with your load to get yourself into a node, and that will typically also affect the SD/ES in a positive way, but not always as much as necessary. If you place yourself in the middle of a great node, you can have a lot of wiggle room on velocity spread and still be in your node. You will, however, still see the affect of high SD/ES at longer ranges, like 1000 yards or more especially. So, with all that said, we can discuss further how you achieve both, together, and with a quality chronograph.

So first off, you start with your powder charge weight node. You look for that accuracy node by finding flat spots in velocity rise as you increase the powder charge during load development.

Once you do find your node, THAT'S when you pay special care to your brass prep methods, although it's best to do it right from the start. It's imperative to use brass of the same headstamp and lot number. You can reduce your SD/ES further by sorting those by internal volume. Some guys simply weigh the cases, others use water to find the exact internal volume measurements. Using top tier brass, like Lapua, Peterson, Alpha, ADG, etc can help eliminate these steps, however.

One very crucial factor in lowering SD/ES is consistent neck tension. That requires different methods depending on overall brass qualities. Some brass is lower quality and will have large inconsistencies in the thickness of the neck. That's when neck turning becomes beneficial- to make the neck thickness consistent and uniform. You're ultimately after a consistent release of the bullet when fired. Consistent/uniform grip on the bullet is a huge factor in achieving that. Neck tension, and neck hardness are the biggest factors here.

You can also use a two-step method to size your brass, using a mandrel to set your final internal neck diameter. Using a mandrel, rather than the expander in most full length sizing dies, will push the inconsistencies in thickness to the outside diameter of the brass, where they don't matter as much. This of course is in comparison to using a bushing die to size your necks. Bushings can work well, but work best when your necks are of a consistent thickness to start. A mandrel, unlike an expander, will also provide more contact surface on the neck, size it in a downward motion as it enters the neck rather than as it exits the neck, and it wont stretch out the shoulders nor induce runout in the neck. Expanders have a tendency to do both, as they're not typically concentric in the die and fully supported to begin with.

The last way to improve neck tension consistency is by annealing. Annealing undoes any work hardening of the brass caused from firing and resizing. As the metal work hardens, it will develop increased spring back, in a sense, due to the metal hardening and losing malleability/ductility. That will also typically be non-uniform throughout the length of the neck and thus create differences in neck tension not just from case to case, but also throughout the neck and shoulder of each case itself. The spring back will be uneven with necks that are uneven in thickness. This can be remedied by turning the necks. Proper annealing also softens the shoulders as well and helps produce consistent sizing, specifically shoulder bump. That will produce a more consistent internal case volume, and thus a more consistent amount of pressure when each round is fired. Consistent pressure goes towards producing consistent velocity.

To get absolute consistency from your brass, and to make them all as close to each other as possible, it's best to go the whole mile. I'll give you an example, by describing my method.

I start by decapping (if already fired). Then, I wet tumble all the cases. After they're all clean and dry, I anneal them all to get back to a baseline without any work hardening. Then, I full-length size using just a normal FL sizing die, but without an expander ball/button installed (I sometimes use Redding competition shell holders to set my shoulder bump during this step). That will get all the cases sized uniformly to one another, and leave the necks a bit smaller than final size. I then run all the brass back through a mandrel setup to set my final neck size/tension. The size of the mandrel will determine the amount of tension on the bullet. You can get them custom made to provide your desired amount of tension. Next, I trim/chamfer/deburr all the cases. Then, I uniform the primer pockets and deburrflash holes (this is a one-time process). The last step is to sort all the cases by weight (if not using top tier brass). Uniforming primer pockets and deburring flash holes isn't something you need to do, and I won't always do it. I will also sometimes turn my necks. If you turn necks, do it before final sizing, that way you get all the benefits of turning and produce a very consistent and uniform amount of neck tension. I'd also recommend a fresh anneal before your final sizing if you do turn the necks.

I get SDs around 2fps and ESs around 4-5fps using this method, sometimes lower.

All that said, you don't have to do every step every time, nor do you necessarily have to do every one of those steps at all- ever. Some brass, like Lapua, Peterson, Alpha, ADG, etc are made so consistently that you don't necessarily need to turn the necks, and you can even get by without annealing for several firings.

You can run your own tests to determine what works for you and what doesn't, which I highly encourage. In the end, you want to only do what is truly producing results. No one wants to spend hours upon hours at the reloading bench, especially doing steps that really aren't doing anything significant or ultimately being fruitful.

To add to this:

Here are the absolute best ways I've found to lower SD/ES if you're struggling:

•Use top tier brass, and of the same lot.

•Increase your neck tension. A good amount I've found is .003" And yes, a Factory Crimp Die (FCD) can do this for you, but it's far inferior to just making the whole neck uniform.

•Use a powder scale capable of measuring to an accuracygreater than .1gr (.05gr to even .02gr accuracy is even better/best).

•Use a comparator to measure your OAL from the base of the case to the ogive of the bullet. That will ensure the most consistent seating depth from round to round.

•Use match grade primers

•If using lesser quality brass, batch sort your brass by internal volume (either by weight or using water to measure the actual capacity). Absolutely don't mix headstamps or lot numbers. Sorting by water is even better than just weighing. Using higher quality brass can eliminate this step, however.

Other things can help, but are typically just ways to split hairs, ultimately. If your ES is 20fps or above, you're missing one of the above steps. Fixing it with a crimp isn't the way to go about it, in my opinion and experience. When you get your ES to 10fps and desire lower, then you can add more tedious steps.

Also, regarding crimping, and/or using a Lee Factory Crimp Die (FCD):

Collet crimping, or a Factory Crimp Die (FCD), and/or increased neck tension TYPICALLY reduces velocity due to more energy being spent releasing the bullet from the case. A collet crimp, or increased neck tension, will also reduce your SD/ES because it makes a more consistent contact with the bullet, resulting in a more consistent release.

I used to use just a Lee Full Length (FL) sizing die for several cartridges, and honestly never had any issues with them. I remove the expander ball from it, using it to only FL size the outside dimensions of the case. Then, I use a mandrel in a separate step to set final neck/bullet tension. My tension on most loads is right at .002" and is extremely consistent throughout the length of the neck thanks to the mandrel.
I also turn my necks too, but that's not really necessary when using a mandrel. It's just something I do out of anal retentiveness.

You can get a mandrel setup from somewhere like K&M, Sinclair, 21st​ Century, etc. You can get one machined from some of those places to the exact size you want to give you your desired amount of tension. A mandrel and the press adapter are $30 or less. I'm currently using TiN coated mandrels from Sinclair, which work very smooth and produce great results.

Use that along with your FL sizing die and a quality seatingdie, and you can make very consistent and accurate ammo for pretty cheap.

Also, at .003" of tension, you don't need a crimp. Even at the very common .002" of tension you don't need a crimp either. I don't crimp any of my bottleneck cases and I run semi-autos as well bolt action. They run very consistent, cycle smooth, and I get no bullet setback. My SDs areunder 5fps and my ES is below 10fps.

Now, while increasing UNIFORM neck tension can increase your accuracy and decrease your SD/ES, so can utilizing a FCD. However, it's not as superior of a method. Collet crimping puts the peak amount of case neck contact with the bullet right at, and only on, the point of the actual crimp- that being the mouth of the case. If the bullet gets bumped, the runout/concentricity is compromised. It may no longer be consistent with your other rounds. Mishandling of the rounds, rough cycling/feeding, etc is all it takes. Having a bullet out of round is not conducive to accuracy. It will affect the bullet to bore alignment when the round is fired and result in fliers on target.

By setting your entire neck to a consistent amount of tension, you don't need to crimp and it makes the point of contact with the bullet uniform throughout the length of the neck and the bullet's bearing surface. That means it's not easily bumped out of round.

You can find your amount of tension by prepping the brass to the point it's ready for a bullet, and then measure the outside diameter of the case neck. Then, seat a bullet into that case and measure it again. The difference is your amount of tension.

Regarding using mandrels versus other sizing methods:

To elaborate a bit further, the reason the mandrel is superior, is because it sizes the neck in a downward motion and doesn't stretch the neck and shoulder back out as it exits like an expander ball can. It also does a superior job at making the tension very consistent throughout the length if the neck by ironing out any inconsistencies much better than the much shorter surface on most expander ball/buttons.

It's actually better and cheaper than a bushing die setup. Yes, to reiterate, mandrels are actually superior. They'll produce much better SD/ES consistency. To get the same type of results with a bushing, you really need to turn the necks, as previously mentioned.
Brass is made by a series of draws. As in the brass starts out as just a cup and is then gradually drawn out to its cylindrical shape. That process makes, or can make, thicknesses uneven. It's just the nature of the beast. Because of this, if you use a bushing die without turning the necks, you're pushing those inconsistent thicknesses to the inside diameter of the neck. That results in an inconsistent amount of tension on the bullet, where the high spots produce more tension, and low spots produce less, which equals inconsistent bullet release and thus velocity and potentially bullet to bore alignment.

A mandrel pushes those inconsistencies to the outside diameter and leaves a uniform amount of tension on the bullet. No neck turning really necessary.

The neck opens up to the chamber walls when firing, so it no longer affects the bullet. You want the neck to release the bullet evenly and consistently from round to round. That's why it matters having the inside wall uniform.

Once the round fires, the pressure pushes the case into the chamber walls. At that point though it really doesn't matter because you'll have to resize everything again anyways. If you turned the necks, you don't need to do it again. If you didn't, the chamber wall will have pushed those uneven thicknesses back to the inside of the neck. That doesn't matter though, because FL sizing will do the same thing. The mandrel will fix that problem though, pushing them back to the outside diameter once again.

It is good practice to anneal though. A mandrel can't defeat spring-back (loss of malleability/ductility) of the metal as it work hardens. Once work hardened, those inconsistent thicknesses will essentially spring back and tension will once again be uneven. Annealing will undo the work hardening and allow the mandrel to set uniform tension once again.

If you turn your necks, or are using very uniform brass to begin with, a bushing can indeed do the same as a mandrel, just to clarify. Neck turning is another expense, and makes for a longer brass prep session.

I will note too, that neck tension of .002-.003" isn't always desirable, or necessary either. If necks are consistent in regards to thickness and hardness, you can get by just fine with minimal tension. A lot of competitive shooters prefer minimal tension.

I've found that increasing neck tension with lesser quality brass (Federal, Hornady, Winchester, Remington, etc, etc) can help produce better uniform tension due to any inconsistencies in the thickness or hardness of the necks. I also prefer around .002'-.003" personally because I magfeed as well as shoot ARs and prefer the solid amount of tension for overall reliability and to never have issues with bullet setback.

So, the bottom line is that consistency is key. I've found my SD/ES numbers are influenced most by my brass prep and cartridge assembly methods and uniformity. You need to first ensure your brass is of the highest quality, is all sized consistently (annealed if required to achieve that consistency), trimmed, chamfered, and deburred, and then sorted after that if necessary. Then your primers need to be consistent and seated to a consistent depth to ensure consistent ignition. After that, you need to ensure your powder charges are consistent and your bullets are seated to a very consistent depth and with minimal runout. The tension on your bullets should also be very consistent. That's again where annealing can help, and possibly turning necks if thickness isn't uniform.

Once your ammo is assembled to that level, then your SD/ES is more influenced by the pressure curve produced by the particular powder charge, the harmonics/vibrations of the barrel from that pressure curve, and the timing of the powder burn before the bullet exits the barrel. That can be influenced by seating depth/bullet jump as well.

When doing the Satterlee method (or any variation of it), and working in .1gr increments, it's imperative that your loads are precise and as consistent as possible. Most powder scales are only accurate to .1gr so any other difference in the load/round can throw off your results and could either produce false nodes or hide an actual node. I prefer to do .2gr increments and I also do three shots at each charge weight so I can get an average velocity. I get a much more accurate reflection of the velocity for that particular charge weight that way.

Hopefully you were able to make it through all that. I don't consider myself a complete expert, but I do consider myself pretty experienced. I'm always wanting to learn more though and I know I still have things to learn.
 

BFD Guns

Well-Known Member
Joined
Mar 2, 2020
Messages
1,535
Location
Illinois
I'ma gonna look at this in the morning with my coffee. LOL I did get some good advice from Erik Cortina about ES, SD, harmonics, and combustion relationship. You might have covered it, but if not check out his vid. It really explained why some of my most accurate loads weren't with the lowest SD load...
 

Petey308

Well-Known Member
Joined
Aug 23, 2018
Messages
334
Location
Missouri
I’m definitely way behind on watching his videos lol. He’s been pumping them out faster than I can keep up lol.

Definitely let me know what you think of this though. I debated for quite a while on posting it here.
 

BFD Guns

Well-Known Member
Joined
Mar 2, 2020
Messages
1,535
Location
Illinois
I’m definitely way behind on watching his videos lol. He’s been pumping them out faster than I can keep up lol.

Definitely let me know what you think of this though. I debated for quite a while on posting it here.
Roger that. Check out the video I PM'd you! :cool:
 

QuietTexan

Well-Known Member
Joined
Nov 16, 2020
Messages
1,028
Location
Fort Worth, Texas
I get SDs around 2fps and ESs around 4-5fps using this method, sometimes lower. /// If your ES is 20fps or above, you're missing one of the above steps. /// When you get your ES to 10fps and desire lower, then you can add more tedious steps.
Sample size is critically important for anyone making statistical claims.

For SD, any sample of less than 5 shots come with a margin of error so large that in order to be credible is should be presented as a range. Claiming a 2 FPS SD without stating the number of shots taken is very questionable, because the true population SD is almost certainly no where near that number. For example, a 3-shot SD of 2FPS should be shown as 2 +10 FPS SD because the likely SD of 100 identically loaded rounds will be between 2FPS and 12FPS 90% of the time when doing the math on such a small sample.

ES will ultimately approach and eventually slightly exceed 6x of SD, so if your ES is less than 30 fps (translating to a population SD of 5 fps) you almost certainly aren't capturing enough data points to produce valid results capable of providing inferential conclusions, based on the premise that a true population SD of 5 fps is exceptional even in the upper tiers of precision shooting. Extreme spread is the weakest of all statistical solutions as it is calculated from the absolute minimum amount of data of 2 regardless of the size of the sample, while ignoring the highest percentage of the rest of the sample. ES has a place in the descriptive statistics of both muzzle velocity and group size, but in a practical sense only gives negative feedback and does not provide any inferential value.

When based on a meaningful sample size, any SD under 10/ ES under 60 FPS falls squarely into Bryan Litz's category of "exceptional factory ammunition or average hand-loads", and any SD udner 5/ ES under 30 FPS "represents the best hand-loads."

I think this is important to mention because spelling out very precise yet unreliable statistics only sets the readers up for failure by giving them impossible goals to try to achieve. There is no determinative difference between a 2798, 2800, 2802 three shot group and a 2795, 2800, 2805 three shot group, despite the fact that ES and SD more than double between the two samples.
 
Last edited:

Petey308

Well-Known Member
Joined
Aug 23, 2018
Messages
334
Location
Missouri
Sample size is critically important for anyone making statistical claims.

For SD, any sample of less than 5 shots come with a margin of error so large that in order to be credible is should be presented as a range. Claiming a 2 FPS SD without stating the number of shots taken is very questionable, because the true population SD is almost certainly no where near that number. For example, a 3-shot SD of 2FPS should be shown as 2 +10 FPS SD because the likely SD of 100 identically loaded rounds will be between 2FPS and 12FPS 90% of the time when doing the math on such a small sample.

ES will ultimately approach and eventually slightly exceed 6x of SD, so if your ES is less than 30 fps (translating to a population SD of 5 fps) you almost certainly aren't capturing enough data points to produce valid results capable of providing inferential conclusions, based on the premise that a true population SD of 5 fps is exceptional even in the upper tiers of precision shooting. Extreme spread is the weakest of all statistical solutions as it is calculated from the absolute minimum amount of data of 2 regardless of the size of the sample, while ignoring the highest percentage of the rest of the sample. ES has a place in the descriptive statistics of both muzzle velocity and group size, but in a practical sense only gives negative feedback and does not provide any inferential value.

When based on a meaningful sample size, any SD under 10/ ES under 60 FPS falls squarely into Bryan Litz's category of "exceptional factory ammunition or average hand-loads", and any SD udner 5/ ES under 30 FPS "represents the best hand-loads."

I think this is important to mention because spelling out very precise yet unreliable statistics only sets the readers up for failure by giving them impossible goals to try to achieve. There is no determinative difference between a 2798, 2800, 2802 three shot group and a 2795, 2800, 2805 three shot group, despite the fact that ES and SD more than double between the two samples.
I get what you’re saying and your point. I agree as well. When I do my confirmation of a load, I do at least 10 shots. That’s what I base my typical SD/ES off of, and even that is an average, not an exact, because it’s based off multiple loads abd multiple cartridges. It’s meant to give the reader an idea of what I’m getting and perhaps what they could expect. I’m giving specific examples in this write-up.

I will also say that I’m not Bryan Litz and I’m not writing one of his books here lol. I don’t shoot competitively either, as I don’t have time.

All that said, I do really appreciate you adding this point. It’s definitely a valid thing to consider. If I ever do decide to publishing anything I’ve written in a more official place, getting good feedback like this is welcomed. It’s something I’d definitely consider covering better.
 

Dr. Vette

Well-Known Member
Joined
Dec 30, 2009
Messages
2,668
Location
Holland, MI
Nice writeup.

I'll throw in one more observation and see if you agree - the quality of the barrel can affect SD/ES as well. I see this routinely when loading for factory barreled rifles vs the several McWhorter rifles owned by a colleague. I can consistently get much better SD/ES from his rifles than from factory barrels even with the same equipment and loading practices. It's not that a factory rifle can't do it, but the custom surely will.

Thoughts?
 

FEENIX

Well-Known Member
LRH Team Member
Joined
Dec 20, 2008
Messages
18,311
Location
Great Falls, MT
I get what you’re saying and your point. I agree as well. When I do my confirmation of a load, I do at least 10 shots. That’s what I base my typical SD/ES off of, and even that is an average, not an exact, because it’s based off multiple loads abd multiple cartridges. It’s meant to give the reader an idea of what I’m getting and perhaps what they could expect. I’m giving specific examples in this write-up.

I will also say that I’m not Bryan Litz and I’m not writing one of his books here lol. I don’t shoot competitively either, as I don’t have time.

All that said, I do really appreciate you adding this point. It’s definitely a valid thing to consider. If I ever do decide to publishing anything I’ve written in a more official place, getting good feedback like this is welcomed. It’s something I’d definitely consider covering better.
Strength in numbers (sample size, population, etc.) is the nature of the quantitative analysis/approach. I also used to do 10 shots; I reduced it to 5 and eventually 3 for hunting loads. I am sure this will stir the pot/controversy, and I am not looking for any debate either, but it is "my" personal choice and has worked for me for many years now. "I" am by no means trying to convince anyone of my method of madness. I do 3 shot cold bore on different days multiple times. As a pointy-head and nearly 5 decades of hunting, I have yet been afforded the opportunity more than 2 shots at a wild game. :cool:

We all have varying personal experiences and personal choices/preferences.
 

Petey308

Well-Known Member
Joined
Aug 23, 2018
Messages
334
Location
Missouri
Nice writeup.

I'll throw in one more observation and see if you agree - the quality of the barrel can affect SD/ES as well. I see this routinely when loading for factory barreled rifles vs the several McWhorter rifles owned by a colleague. I can consistently get much better SD/ES from his rifles than from factory barrels even with the same equipment and loading practices. It's not that a factory rifle can't do it, but the custom surely will.

Thoughts?
A good lapped barrel will definitely increase consistency in pressure and this velocity by not producing inconsistencies to the bullet and pressure as it travels through the bore. A lot of factory barrels simply lack the overall quality and consistency and can definitely induce variables to increase SD/ES.
 

Petey308

Well-Known Member
Joined
Aug 23, 2018
Messages
334
Location
Missouri
Strength in numbers (sample size, population, etc.) is the nature of the quantitative analysis/approach. I also used to do 10 shots; I reduced it to 5 and eventually 3 for hunting loads. I am sure this will stir the pot/controversy, and I am not looking for any debate either, but it is "my" personal choice and has worked for me for many years now. "I" am by no means trying to convince anyone of my method of madness. I do 3 shot cold bore on different days multiple times. As a pointy-head and nearly 5 decades of hunting, I have yet been afforded the opportunity more than 2 shots at a wild game. :cool:

We all have varying personal experiences and personal choices/preferences.
The true Satterlee load method uses just one shot at .1gr increments. I already know that statistically that method is flawed due to all the variables involved. That’s why I prefer to do at least 3-5 shots at a given charge and to also go up in .2gr increments rather than just .1gr increments. I take the average MV of those to plot for looking for a node. Once I find what looks to be a node, I confirm with 5-10 shots.

I also mentioned in the write-up that SD isn’t always an indicator for accuracy. It is an indicator for consistency though.

And yes, statistics are always better with a larger sample size. If components weren’t so scare enough and expensive these days, I’d shoot more at a time than I do. I’ve chosen my shot sizes though as a balance between getting useable data and not consuming too many components.
 

BFD Guns

Well-Known Member
Joined
Mar 2, 2020
Messages
1,535
Location
Illinois
What sealed into my brain was the statement by Cortna in regards to handload accuracy issues. "Its either a combustion problem, or its a harmonics problem". When I got that handload to under 5 SD but found it didn't group good, I'd scuttle that recipe and run with the higher SD load that did group good. Although I'm sure the inconsistencies showed up downrange. With my new tuner brake arriving shortly I can return to those low SD recipes and hopefully tune those harmonics.
 

Petey308

Well-Known Member
Joined
Aug 23, 2018
Messages
334
Location
Missouri
What sealed into my brain was the statement by Cortna in regards to handload accuracy issues. "Its either a combustion problem, or its a harmonics problem". When I got that handload to under 5 SD but found it didn't group good, I'd scuttle that recipe and run with the higher SD load that did group good. Although I'm sure the inconsistencies showed up downrange. With my new tuner brake arriving shortly I can return to those low SD recipes and hopefully tune those harmonics.
I actually use QuickLoad to pretty much do all my load development now. I e had a couple guys that are really good with it teach me the ways and now I just use it a long with a few shots to validate and confirm things, tweak as necessary, and I’m golden. A barrel tuner can make those needed tweaks raven easier 🤙🏻
 

asd9055

Well-Known Member
Joined
Nov 15, 2013
Messages
1,695
Location
Texas
I actually use QuickLoad to pretty much do all my load development now. I e had a couple guys that are really good with it teach me the ways and now I just use it a long with a few shots to validate and confirm things, tweak as necessary, and I’m golden. A barrel tuner can make those needed tweaks raven easier 🤙🏻
Have you given Gordon's Relaoding Tool a look? I too use QL, but GRT looks promising as well
 
Top