I will give this a try when the weather permits, testing it against a known accurate load found the old fashion way, starting at a published low start and working up in .3 grains, 5 rounds each.
A synopsis from the vid:
1. Reference a reloading handbook to find a safe starting load
2. Load 10 cartridges for ladder test, 1st cartridge is starting load, move up in .2 or .3 grain increments
3. Shoot and chronograph each test load
4. Analyze chronograph results, identify loads with little change in velocity
5. Load more cartridges at accuracy nodes
6. Fire, chronograph and analyze accuracy node test loads
7. Load and conduct field test, load final test cartridges and test for accuracy/consistency at distance. (5 shots. If not grouping well change seating depth)
Klem has a great post about this that points out many of the things that concern me as well.
Why don't you send me the data and I will graph it like in the other thread. Lets say, three rounds at each powder increment over ten increments - 30 shots. Shoot at ten targets at the same time you record their velocity. Obviously you will need to be careful when shooting to not confound the results with an external variable like 'shooter'.
That way you can compare Satterlee's velocity method against how accurate they actually are. This is what you want to do, yes?
If you don't have an app like OnTarget that calculates MOA just send photos of the 10 targets with the velocities.
01-16-2020, 05:29 PM (This post was last modified: 01-16-2020, 05:33 PM by imaguy3.)
I do it.. it works.
First time I did it was with my 280ai bolt gun. First bullet/powder combo at .020 off the lands, loading in the middle of the best node had me a 1/2 moa load. I played around with seating depth after that but found no improvement on that gun.
Used it on my grendel as well. In the middle of my node was a 1moa load. Played with seating depth and got it down a little better.
It wasted tons less ammo than when I tried developing a load the old fashioned way for my 30-06 bolt gun.
Why don't you send me the data and I will graph it like in the other thread. Lets say, three rounds at each powder increment over ten increments - 30 shots. Shoot at ten targets at the same time you record their velocity. Obviously you will need to be careful when shooting to not confound the results with an external variable like 'shooter'.
That way you can compare Satterlee's velocity method against how accurate they actually are. This is what you want to do, yes?
If you don't have an app like OnTarget that calculates MOA just send photos of the 10 targets with the velocities.
Klem that sounds like and excellent day at the range! Might be a month or two until weather/temp is fit for the test.
I am more inclined to do four rounds per increment as an accuracy indicator and mostly do five when I test.
What would you say is a validation or a dis-qualifier of each method?
Jimla said:
10-02-2019 06:27 PM
I got a new lot of CFE223 powder, I tried a 10 shot Satterlee's load development and loaded the previous accurate load.
I was surprised how well 10 shot Satterlee's worked, it found my previous 30.8gr and 31.2gr nodes, see graph. At 30.8 it is at a bottom out in a flat area and at 31.2 it is a flat area with a sight drop of Velocity increasing .1 grains.
I think the hand-loader/reloader needs to approach how they will do load development relative to their practical goals with the rifle and application. Some examples:
1. Medium game hunter looking for fastest kills within 300yds
2. Recreational target shooter who likes to shoot 1-2 MOA plates from 200-1000yds
3. Competitor who is shooting for X ring count
4. Starship navigator stranded on earth, equipped with lab-grade scales, electromagnetic or radar velocity-measuring systems layered with each other, barometer, latest ballistics program with every variable you could imagine, and a penchant for demanding extreme precision.
I see myself as a hybrid of all of these guys, minus the X ring paper type shooting.
Depending on whether we're shooting an AR or a bolt gun, we might be wasting our time if we ream primer pockets, anneal necks every x number of firings, weigh and sort brass, etc.
It's up to each hand-loader to determine where they are and what realistic goals they have.
For me, I don't have Sartorius scales yet, and I like to find a load that will shoot as quickly as possible without getting too lost in the weeds for maybe a 1/8 to 1/4 MOA improvement, but I still double-seat my tips to reduce concentricity runout, for example.
So for a more accurate TGT load, I use a combination of an initial 100yd ladder test, then measuring my groups if I can at 300yds, with colored projectile tips assigned to each load. The main reason I like to do the initial 10rd ladder with .3gr increments has more to do with safety for me. I want to establish a trend line of what I should expect to see regarding charge weight increase related to velocity increase. Once I see the increases depart that trend line, then I know I've hit a place with case capacity, charge density, and the specific projectile that I don't really want to be in, establishing a top end limit with that component combo.
Then I can go back in and look at nodes as well as what actually happened on paper and fine-tune that area.
I have noticed that sometimes the flat nodes will have larger dispersion on paper, so this is where Klem's point of focusing on accuracy I think is very valid to consider and not just trusting the tight velocity nodes we might see initially.
Where the tight velocity nodes come into play is with vertical separation at long range, which will depend on what size of a target you're shooting if you need to dial that in.
If you can find a tight velocity node where 2-3 charge weights don't seem to deviate much in speed, and they group well, then you're onto a good load for long range.
You won't see much practical value in the tight ES at closer ranges, although there is a consistency in impact speed that some hunters might be looking for if they're frequently on the edge of expansion range characteristics of the hunting bullet they're using.
NRA Basic, Pistol, Rifle, Shotgun, RSO
CCW, CQM, DM, Long Range Rifle Instructor
6.5 Grendel Reloading Handbooks & chamber brushes can be found here:
Lemonaid Wrote:Thanks Jimla, exactly the data I'm looking for!
Glad to help.
One thing you need is a very accurate chronograph for a single shot readings. My old one had error of 1%. The more accurate barrel mounted chronograph made it possible.
I switched to using the satterlee method a few years ago, and it has allowed me to find some of the most accurate and consistent loads you can imagine. 1/3moa and single digit sd. So much so that it doesn?t bug me if my charge master over or under throws by 1-2 tenths. I know they?re all going to be the same speed and go in the same hole.
That being said this is for precision bolt guns and gas guns have a lot more going on. I don?t load for gas guns because I mostly use them for shooting pigs and factory ammo gets me close enough without spending a bunch of time. All that to say I don?t know if the velocity node/accuracy correlation I?ve seen will necessarily be the case when you introduce all of the variables an AR has to offer.
03-08-2020, 12:23 AM (This post was last modified: 03-08-2020, 10:59 PM by Klem.)
Lemonaid,
That's great, and there's plenty of data here to work with.
Here are your results graphed and you can see the relationship of velocity and group size. As I said, choosing optimum charge weight based on velocity (only) is a clumsy proxy for what you really want - which is the smallest group size. Bullets can come out the barrel with good standard deviations of velocity but still be pointing in different directions due to barrel whip, or simply because the bullets are poorly made and not concentric in shape or weight. Suggesting you can find the optimum charge weight by shooting only 10 rounds and looking at velocity is almost negligent. Given shooting is a game of statistical averages imagine a Satterlee outcome which is more towards the edges of the Normal curve than the middle. There's more chance of this happening with only 10 rds. You then spend the rest of that barrel's life shooting a load you are convinced is the best it will ever produce, and you will be wrong.
[ATTACH=CONFIG]15628[/ATTACH]
Unfortunately I am not seeing a useful trend here other than your groups were getting better as the load got hotter. If you started the test by shooting lightest loads first then maybe you got better at shooting on the day or maybe the gun was clean and it took a while to foul and settle. The up-down group size between .2 grain loads by such large group sizes on the target suggest something is, with respect, not right. You cannot say with confidence that a particular load is best for your barrel.
If the barrel follows the usual harmonics there should be a sinusoidal wave of group sizes to indicate the load area that you should hone in on. The velocity line should be as straight as possible to indicate consistent loading, neck tension and the measuring device is fine. If you are loading in the dark, especially compressed loads then a steep curve up in velocity is a danger sign.
I have not used A2200 but QL is telling us that the max load is 28gns, which from that program is always conservative (but it is safe). At your OAL it is 95% full so not a compressed load. That sharp curve upwards approaching 29gns may just be a Magneto precision error and not a danger sign, and I say this because the rest of the velocity line is not exactly straight.
You note two group sizes where one group has an outlier. In your shoes I would not discount any shots unless at the time you are convinced the fault was yours, and not the gun or ammo (e.g. you flinched or slipped).
My thoughts are that you need to re-do this test. Now that you know the load range is safe I would start 5% low and work up instead of 10%. If you want to keep 29gns as the max (absolute value) then shaving 1% off to 1 decimal place is;
29
28.7
28.4
28.1
27.8
I have highlighted two cells in yellow in your data; the '2818' is a standout and you can see by the standard deviation (107) that something is not right here. Having a round 200fps out from the rest is obviously a mistake in the machine, the recording, or loading. The other reading; 2681 in your post is different from your hand written velocities on the target (2691). Only one of them is the correct velocity.
Here is my OCW for 4rd groups of 130AR Hybrids using 8208 out of an 18" barrel. I am looking for a tightest groups closest to the max safe load, which looks to be around the 26.7 mark. So, looking at the red line I am interested in the low points or troughs of the curve. Plus a sinusoidal MOA curve which suggests the barrel has predictable harmonics. Point to note: the max load is a stout load which is a little compressed so it's a good thing the tightest group is a full grain back from the max.
[ATTACH=CONFIG]15629[/ATTACH]
Here's another OCW for the new Nosler RDF (Reduced Drag Factor) line of bullets. Their BC is currently the highest in the 6.5mm 130gn range, and looks like they have achieved this by making it longer and pointier than all others. So they expect users to have plenty of magazine length to accommodate the longer tip; which unfortunately is not the case in an AR.
Out of a heavy 26" barrel 6.4*47 bolt gun everything is more precise than an AR; the groups are tighter and more consistent. Although there is a predictable sinusoidal wave of precision and it goes a further 50-metres, the average group size of the Nosler RDF is not as good as the Berger 130AR Hybrids. Whether this is because my barrel 'likes' Bergers better or that the Berger is a more consistently-made/better quality bullet is debatable but the RDF's are going back on the shelf.
[ATTACH=CONFIG]15630[/ATTACH]
This is a pity because the graph is suggesting the most accurate load is also the max safe load...the best of both worlds. But, as I say, the RDF's are not as precise as the 130AR Hybrids, at least in my gun.
Arrrrrrrgh! I flunked the test! You spotted my transcription error, the correct reading was 2681 at 27.6 grains.
I believe the 2818 reading from the 27.0 grains to be correct, a results of the case neck being too thick thus gripping the bullet much harder on chambering and spiking the velocity.
This condition happened somewhat randomly through out the test and I suspected it would throw a monkey wrench into the results.
I used the Speer's reloading data from their site for a max charge of 29.2, max velocity of 2992 from a 24" bbl.
Klem, many thanks for your help. I appreciate the time and effort you put in.
The good thing about flunking the test is that I get a do over!
If you think the neck does not have enough clearance to release the bullet then this is a potentially dangerous situation. The only thing that will solve that is neck turning the case or reaming out the chamber. Not using Lapua cases will also help but that's like the tail wagging the dog, you want to be able to use any brand of case so looking at chamber size is a better solution. What is the diameter of your spent cases at the neck?
It could also be from cases that are too long and need to be trimmed. The mouth of the case gets crimped into the bullet when it goes into battery and cannot release the bullet at normal pressures. Case creep in this calibre is not as severe as other calibres but I take it you are checking the case length of every fired round before reloading it?
Or it could be a mistake in recording the velocity; either by the machine or you. No offence, but I do this sometimes and it can send you down the rabbit hole.
These Lapua cases are some once fired I purchased from PF when they were on sale. I didn't expect any problems after sizing and trimming. I do check case length after sizing and trim between the max and min (1.510).
Measuring Hornday cases I reloaded the case neck is .289, the one round I couldn't chamber (Lapua) neck is .290.
Fired cases Hornady neck = .299 Lapua = .299
I think I was jumping the gun to blame the hard chambering on neck thickness. It looks like it is insufficient sizing in the neck shoulder doughnut area. Sometimes screwing in the sizing die can fix this, others I had to neck turn the doughnut off.
Being that this Lapua brass likely was fired from many different firearms I think most are fine but about 1 in 5 will need more work.
Next test I will ace it!
03-08-2020, 03:37 AM (This post was last modified: 03-08-2020, 04:03 AM by Klem.)
Your neck thickness sounds fine. Mine are .289 also. Allow .001" clearance for safe release of the bullet and it is within the .300 SAAMI chamber blueprint for Grendel.
Your case length is also fine at 1.51", within spec.
Your expanded (fired) neck diameter is also fine at .299". Allowing for .001" springback it sounds like your neck has been reamed to Grendel SAAMI spec.
So, it is not the case size, length or chamber that caused the 200fps spike.
It is too large to be within the error parameters of a Magnetospeed. The Magnetospeed doesn't suffer the vagaries of clouds like the Chrony. Being the only one that was out 200fps means it's not the fault of the chronograph.
If it was a doughnut at the base of the neck it would either not go into battery or fire normally, but certainly not spike 200fps.
This leaves human error I am afraid; an overloaded case, a slip of the pen when transcribing. I've done this before more times than I care to admit...it happens.
I recommend full-length sizing ALL those second hand cases back to SAAMI, regardless if you are using a bolt-gun or auto-loader. Preferably all at the same time. Screw the die right in for a one-time only sizing back to Square One. Then, once they are fired and fire-formed to your chamber unscrew the die slightly if it is an AR and bump .003-.004" to prolong case life. If it is a bolt gun, swap to neck-sizing and bump .001" for accuracy and case life.
A quick glance at your results I'd have guessed the gigantic flat spot between 27.8-28.1 would be the best results. Looked at the pics and it seemed to be, then realized you also posted group size as well, in my defense I was 4 glasses of Wild Turkey deep at the time.
There is a correlation between velocity and group size with the best load at 28.7. This is muddied however by the huge difference at max load 29.1. It is just as likely the wide group at 29 was the shooter (no offence). You need more data. I would focus on 28.6 upwards and shoot a few more groups to get the average.
One thing you need to do is keep all the group sizes the same. No 3rd groups mixed with 5rd groups. This is not a statistically large sample size so your ability to shoot small 3rd groups and large 5rd groups is going to mess up the trends.
Thanks Klem, I concur with your findings. My shooting skills did atrophy during the range closure, the 5 round tests were to try to verify previous accuracy. Now on to the Satterlee 10 round test using Nosler 90 Varmegeddon and H322. It looks to me to do some five round tests at 28.1 and 29.2. Your input on this will again be appreciated!
