Ive been looking for the best combo for my new rifle for quite a while now.

Rifle specs:
BCM upper
AR Performance .223 Wylde SOCOM 16" barrel 3r rifling
Colt/Milspec BCG+bolt
Larue 11" Freefloat quad rail
Larue SPR-S mount
Nightforce 2.5-10X24 MOAR reticle
Larue MBT trigger
Home-brewed 80% lower

Build goal: Build the carbine that I wish I had had in the sandbox. Capable of engaging targets from 0-800 meters but still handy clearing a room and in a vehicle. I understand that 0-600 is far more feasible but hey, shoot for the stars right?


I wanted to make ammo that matched that stated goal. Shooting for average velocities in the 2700-2750 range with MOA or better precision.
As a starting point I ran a 5 shot group of LC M855 through it, and averaged a velocity of 2939, with SD of 21.6, ES 59. A bit faster than whats normally reported from a 16 inch barrel.
Naturally 77gr projectiles, both SMK and Nosler CC. The CC were more consistent in regard to weight so I have been using them over the SMK.

Powders for my round had to be VERY temp stable. Where I live lows in winter can be in the teens, and summer highs in the 115 range.
Powders tested were RL-15, Varget, AR-Comp, IMR 8208 XBR, H4895
Settled on H4895 because it both gave the greatest velocity and was one of the most temp stable. 8208 XBR was a very close second but it turns on you in the blink of an eye, not good. 4895 was much more stable in that regard.

So Ran tests with H4895, LC and Wolf Gold once fired brass, CCI #41 primers, 77Gr CC @ 2.250 +/- .001 COL

Brass deprimed and cleaned via wet tumbler, dried
Brass sorted by weight
Lubed with lanolin/Alcohol mix
Prepped on my MEC rifleman press, Forster FL sizing die, checked in Wilson case gauge
Primer pockets prepped on Lyman case prep center
Trimmed with WFT @ 1.750 +/- .001
Charges metered by RCBS Chargemaster combo to the tenth of a grain.
Bullets sorted and grouped by weight and length
Seated by Forster micrometer seating die to +/- .001 of 2.250 COL

Whew! Ok with all that said my SD SUCK!!!!!Best SD was 12.0, average was in the mid 20's
All chronoed/data collected with Magnetospeed V3

So heres the raw data:
Case/charge/velocity/SD/ES?Group size center to center

WG/23.1/2683/12/26/1.092
WG/23.3/2693/29.7/67/.733
WG/23.5/2725/56/140/1.307 (Flyer, charge off?)
WG/23.7/2715/22.7/56/1.08 (started getting very very light ejector marks)
WG/23.9/2724/12.4/33/.892
WG/24.2/2784/35.2/86/2.30-Obvious overpressure signs, poped primer and HEAVY ejector swipes.

Best groups were at 23.3 (WG .733), 23.9 (LC .841/WG .892), 24.0 (LC .929), 23.6 (WG .929)

LC/23.1/2626/41.7/113
LC/23.3/2646/34.1
LC/23.5/2695/23.1
LC/23.7/2681/23.3 (started getting very very light ejector marks)
LC/23.9/2740/22.5/50/.841
LC/24.2/2775/18/46/shotgun pattern-Obvious over-pressure signs, popped primer and HEAVY ejector swipes.
Missing info for LC tests due to operator error

I actually tested every tenth of a grain between 23.0 and 24.2 for both cases, I just shortened it here.

So, like, where do I go from here? I'm not even sure my data is consistent enough to choose a node to load to.
Looking at it, to my very novice eye, it looks like 23.9 grains is a go to in both WG and LC with (barely) sub MOA groups.
Is the data above good enough to go with 23.9 and start tailoring things like COL to shrink groups or do I simply not have a good data set?

Any and all recommendations welcomed, Ive never done this sort of thing before.

You may very well be right about the mechanical accuracy/shooter limitation. I am no sniper, this is my first scopes rifle more than a 4x acog on my issues M4.

Shooting from the probe with bipod and rear bag. It’s how I’d fight with the rifle, so it’s how I’ll train right? Oh and there is no secure bench surface where I shoot.

Will try that combo out, thanks!

Flash holes deburred? Without pocket uniforming, my SD is about 15fps which is good enough.

With moly'd SMKs, 23.8gr is my max, flying at 2600fps from my 16" carbine. Seems you have a fast barrel, but 2700+fps is still 20" territory.

Agreed that MOA is good enough as well. I consider my Nat'l Match a 0.75 MOA rifle with a 0.25 MOA (at best) shooter.

No doubt about that.
I know that it takes me a few strings of fire to get really comfortable, but after an hour or so I start to get tired and impatient. Need more training/trigger time.

Biggest thing I’m having trouble with is my heartbeat. I can see it bump my reticle and trying to time it is hard, but trigger time is probably the best remedy.

I haven’t deburred flash holes, I’ve thought about it but haven’t bought the kit yet. Might as well add that to my midway USA cart.

2700 is my obsession lol. That’s why I chose an ARP barrel, theyvslways report shooting fast and relatively precisely.

I have only shot 4895 behind lighter bullets, 55 and less so I can't offer much there.

OCW works if you can control all of your conditions and have a good understanding of that.  Don't forget to keep track of what the wind is doing and other factors such as barrel temperature.

I have not found 8208XBR to be a spooky powder and find that it is pretty versatile in 223.  My experience is that it likes to be run on the warmer side and I have had good success with it behind th 75 Amax in my bolt gun and the 77 Nosler CC in my ARs.

It ran good for me too-but kinda jumped the shark at 23.5
Velocities were low-but I got 2690 at 23.2. At 23.5 it jumped to 2745, es of 105 and as of 53. 2786, 2797, 2707, 2692, and an error. Hard pressure signs too. It just kinda flipped out on me around 2700 where 4895 is rock steady.

It also didn’t group as well. My average group size today with 4895 was 1.008 or so, 8208 was 1.5 average over 10 groups, no sub MOA groups.

And to boot 4895 beats it in temp stability.

The Nosler book shows 23.5 grs. H4895 with the 77 gr. bullet as a maximum load. I would shoot ample groups using a slightly lesser charge, say 23 grs. at 100 yards and measure group sizes. Measuring one or two five-shot groups only tells you that you need to fire more groups for verification. I've never seen a need for trying tenth of a grain increments with any rifle cartridge, but I suppose it hurts nothing. Standard deviation often doesn't matter except at very long range, but you may need an incredibly accurate rifle to determine this. H4895 should be an excellent powder for your use. You may not find a better one. Good luck-

Missing info for LC tests due to operator error  

The info you are missing (group sizes) is the only truly important information.  Without it, it really isn’t possible to comment on the LC data.  Based on velocities, brass does not look to be a large factor.

I actually tested every tenth of a grain between 23.0 and 24.2 for both cases, I just shortened it here.

This is too much.  Use 0.2 gr steps at the smallest.  Otherwise, you dilute your sample size and interfere with getting good statistically significant test data.  


So, like, where do I go from here? Is the data above good enough to go with 23.9 and start tailoring things like COL to shrink groups or do I simply not have a good data set?

If 23.7 is already showing pressure signs and 24.2 has obvious pressure signs, I’d say the 23.9 gr load is too hot.  This is true for both types of brass.  

Temperature stability is nice but other factors come into play in making a load unsafe.  I’d say back down.  

Also note, the mean velocity does not change very much as you go down.  

23.3 gr is your best group, gave good velocity and leaves you some margin for error.  23.3 gr yielded 2693 fps which meets, for all intents and purposes, your 2700 fps goal.  

Proper reloading technique following a so-called "ladder" approach is to begin with the Starting load and then increase the charge by a fixed amount for each "rung" on the "ladder" and stop as soon as pressure signs are encountered.

With both sets of cases, you started to get pressure signs at the same place, 23.7 grains.  That should have been your stopping point.  Below 23.7 grains, your most accurate load was at 23.3 grains for the WG brass (and as you noted, we don't know what the most accurate one for the LC brass was because the data is incomplete).  So, if this is going to be your powder choice for this load, you want to focus your next step in load development at 23.3 grains and also try 23.2 and 23.4 and see what kinds of results you get.



Nothing wrong with that except you tie up a lot of bullets, powder and cases in your test placing the "rungs" of the "ladder" so close together.  Personally, with a spread in excess of 1 grain, I would have started out with 0.2 or 0.3 grain increments and then when I found the best one of those; run a second "ladder" at 0.1 grain increments either side of it to see if I did indeed have a "node".



First, recognize that since you were reporting pressure signs at 23.7 grains (i.e. ejector marks), then a 23.9 grain load is NOT a "GO".

Second, run another "ladder" around your most accurate safe load at 23.2, 23.3 and 23.4 grains and see what you get in terms of accuracy.

Third, recognize that your data strongly suggests that the case headstamp doesn't have a meaningful effect on performance, so try a different powder using a single headstamp to cut down on the time and expense of running future ladders.

Forth, be more fastidious on your next range trip to avoid missing the collection of relevant data because of "operator error".

Thank you, seems like sound advice.

I’m going to use just the Wolf Gold since I have so much more of it. I’ll run a ladder test at the mentioned loads this weekend.

I feel the need to clarify that the ‘operator error’ was leaving my notes out around my 4 year old-who proceeded to use them to paint on.

It was one of those too-damn-cute-to-get-mad scenarios lol.

As to pressure signs, I understand that popped primers and hard ejector swipes are a sure sign to stop, but I’ve been told, and it could very well be bullshit, that light ejector swipes aren’t “that big a deal”. It was described to me as akin to a yellow light opposed to the hard swipes/flattened primers/popped primers that are a red light.
Bullshit?

Sometimes bolt faces have a burr, but usually if it's more than faint, it's too much.

I'll see if I can get a pic.

This is 23.8gr H4895, moly'd 77gr, I think CCI 400 (450 recommended), with ~0.006" headspace from an older bolt. Can see, but rarely feel the swipes. I'll be backing off to 23.5gr, partly to match TSRA spec velocity.

Attached File

That’s exactly what I started to see at 23.7gr.

OP:
FWIW...........I shoot 23.3gr of H4895 with a 77SMK.
I feel as though I found the best accuracy when I worked up.

I would NOT shoot R15 if you are concerned about Temp variations....... in comparing my various loads,
I feel as though there is a lot of Temp sensitivity with R15.
But is can be accurate.

Yes, total BS!

Popped (failed) primers and is way too late.

Extractor and ejector marks is way too late.  

Flattened primers might qualify as a "yellow light" but are a guarantee you are shooting a high pressure load.  

Heavily cratered primers should be a "red light".


Every reloader needs a margin of safety for their reloads.  You need it for the safety of your rifle.  You need it for the safety of your person.  You need it for the safety of the other shooters on the range.  You need it to protect the reputation of the reloading and the good name of the shooting sports.

You will have to decide what that means in terms of powder charge for the load you are developing and the reloading process you plan to use.  

Remember this - everything is subject to variation when you get into bulk reloading.  You need margin to ensure tolerance stack up a and lot-to-lot variation does not cause an unsafe condition.  



The above statements assume properly prepped brass - headspace, primer decrimp, primer pocket depth, etc.

There's also the fact that with 0.1 grain loading increments, you're going to flood yourself with data that will be very hard to sift through.  And while your 0.1 grain increments may have very consistently measured amounts of powder, it takes an incredible amount of attention to detail - and sophisticated bench rest equipment and techniques - to be able to isolate differences in how different loads pattern when the powder charges are that close together.

A better strategy is to start with the 0.3 grain increments AeroE suggests (0.5 with .308 cases is a good rule). Once you get one or two loads that are consistent and give you fairly small groups, then you can start looking at smaller increments.  This lets you concentrate on real differences, while saving time, effort and sanity.

There is certainly nothing wrong with evaluating loads using powder charges that vary in tenth-grain increments, but I've developed successful loads with half-grain increments, even in small cartridges like the .223. This has also worked well for handgun cartridges, with the sometimes exception of small capacity cases and fast-burning powders.

The OP started his load testing at what I consider a maximum powder charge of H4895 when loading 75/77 grain bullets.

You can get excellent accuracy and good velocity starting with only 22.0 grains of H4895 when loading these bullets.

I use .3 grain increments. I would try 22.1, 22.4, 22.7 and 23.0.

If you are looking to load volume, take a gander at Ramshot TAC.  H4895 is not 223 heavy bullet friendly IMHO.
I have had very good results with TAC in 69 and 77 gr OTM bullet.  Although Western now uses WSR primer in their
data, the previous data used CCI 450.  And, I use them.

Thanks again guys.

I’ve got a pound of TAC but it’s so volatile in terms of temp stability that it’s not good for my needs, I.E. find a good recipe and make a whole lot of it for storage. Otherwise I would have gone with TAC.

I ladder tested today from 23.3 to 23.7 grains of 4895. I wanted to run 23.7 again to make sure I was seeing those swipes and it wasn’t something errant in last weeks loading.

It wasn’t, ejector swipes at 23.7.

Otherwise I think the ladder test showed promise.

25 rounds fired with wolf gold brass. 18 rounds in a just less than1 inch ragged hole, 7 flyers (shooter induced, I tried using a new target and it was really difficult for me to get a good sight picture. My eyes didn’t like the orange color) but all within 2 inches.

Attached File


Velocities were much the same as last week, ranging from 2695 at 23.3 to 2722 at 23.7.

My SD actually got worse though. On the advice of a local I didn’t crimp these loads. My lowest SD was 18.9 at 23.6, all others were in the 25-30 range. I will crimp from here on out.

I also ran the same test with LC 09 brass, just because I have a bunch and I like reloading and shooting. As before LC shot slower-2684 at 23.3 to 2710 at 23.6. No data for 23.7 because the first round showed very heavy swipes so I did not shoot them all.
Lake City also grouped slightly worse.

I’m goong to load mine to 23.4 grains from here on out, that gives me just over 2700 with the Wolf, and just under 2700 with lake city. I’ll play with seating depths to see if there is anything to be gained there. I have the Hornady comparator tool so that’s this weeks project.

And my neutered CA compliant rifle. What we go through to keep a detachable mag

Attached File

2.250 +/- .001

That is just not possible with hollow point designs. You will normally have around .010 COAL variance. Some times up to 0.20 variance exists in a box.

This does not effect on target performance though.