My quest for the "perfect" cast .223 load for my AR-15

Final update and summary, as of 3/10/19: While there are still plenty of things I think could improve my results (better bullet design, using an alloy with higher tin/antimony content than pure COWW, experimenting with more powders, and using a nicer barrel), I have concluded my .223 through AR-15 tests for the foreseeable future. I achieved just north of 3 MOA with heat-treated 28-30 BHN COWW bullets, from the Lee C225-55-RF mold. This was at around 2600 FPS, using 19.6 grains of H335. The bullets are powder coated using Smoke's Clear PC, and copper Gator Checks brand gas checks. They are sized to .225" after powder coating. For plinking on steel out to 100 yards, these bullets are great. If you want to win competitions, you're going to have to put in more work.

For anyone wanting to shoot cast .223 through an AR-15 rifle, I can suggest ~19.0 grains of H335, using very high BHN 55gr bullets, as an excellent starting point. I have zero cycling issues with this load in my standard 16" barrel and carbine-length gas system. With my less-than-stellar barrel, I'm getting about 3 MOA.

For those who want to do better than 3 MOA, I can suggest the following measures to improve on my results: try mixing in some Linotype alloy with your COWW alloy. If it was me, I'd start by experimenting with a 1:3 ratio of Linotype to COWW, and playing with heat treating to find the optimal BHN for accuracy. Others who are having more success here than I have recommend a BHN in the 18-22 range with this alloy, and IMR 4895 powder to make near-full velocity loads that shoot with excellent accuracy (2700+ FPS).

One last word of caution; it's a really good idea to establish the accuracy potential of your rifle by working up a jacketed load first. I could have saved literally months of testing had I done this first. It turns out my muzzle break was either faulty or incorrectly installed, and was limiting my barrel's accuracy to around 3 MOA regardless of projectile type. After removing the muzzle break, accuracy improved.

Thanks again to everyone who helped me out as I spent the last year or so really learning a ton about how to make cast bullets work in difficult situations. If you have stumbled upon this thread and are trying to do cast .223 in an AR-15, I assure you it's possible, and I'm happy to answer any questions you may have. Feel free to send me a PM.

-Andrew

Original Post:
Hey everyone--

I've spent lots of time reading here and gotten lots of great info. Now that I've joined the ranks of the AR-15 crowd, I'm tackling my most ambitious casting goal yet: an inexpensive cast .223 load that cycles my "M4 clone" reliably, and is accurate out to at least 100 yards, preferably 200.

I'm doing a short video series to carefully document my "somewhat-scientific" approach to chasing down this unicorn load. The documentation is more for myself than anyone else, but if it helps a brother out who is as overwhelmed with information as I was when researching for this, all the better.

This first video is mostly just an introduction and the culmination of my research into the different factors that go into a successful cast load in an AR-15. About 10 minutes of talking and reading off of a paper so I could remember everything I wanted to cover... haha.

https://www.youtube.com/watch?v=EriLIM-_e-A

But this weekend or next, barring a life-changing event (my wife is due to give birth to our first child any day now :D), I'll be testing out my first set of loads to see where they get me. I will be carefully recording my tests, and documenting my findings for the next video.

All testing will be done on my rifle with the following specs, so anyone looking to develop a load for their own rifle will have mine as a starting point:

Anderson Lower
Bear Creek Arsenal Upper
16" chrome-moly barrel with 1:9 twist
Carbine-length gas system
Standard carbine buffer and buffer spring
Standard 8620 steel BCG

Test #1 - Results and Comments in Blue

Video of results:
https://www.youtube.com/watch?v=0tgMzft8gY4&feature=youtu.be

Goals:

• Find out if IMR-4227 will cycle reliably or if I need to find a new powder. Nope, need a new powder.
• Get a rough idea of POI differences between cast loads in the ~2000FPS range and regular factory j-word rounds Seems just a tad lower, but it's not as dramatic of a difference as I was expecting.
• See if any of these charges is grouping obviously better or worse than the others The 12.9 grain load (estimated velocity of 2035 FPS) grouped significantly better than any other load, so I'll see if I can refine that "ideal" velocity in future tests with different powders.

Test Groups (10 rounds each):

• 12.0 grains of IMR-4227. Estimated pressure is 19733 CUP, and estimated velocity is 1906 FPS. Not accurate, short stroked.
• 12.3 grains of IMR-4227. Estimated pressure is 20253 CUP, and estimated velocity is 1949 FPS. Not accurate, short stroked.
• 12.6 grains of IMR-4227. Estimated pressure is 20773 CUP, and estimated velocity is 1992 FPS. Not accurate, short stroked.
• 12.9 grains of IMR-4227. Estimated pressure is 21293 CUP, and estimated velocity is 2035 FPS. Reasonably accurate, short stroked.
• 13.2 grains of IMR-4227. Estimated pressure is 21813 CUP, and estimated velocity is 2078 FPS. Not accurate, short stroked.

Striving for consistency on as many variables as I can control:

• Projectiles: Boolits are 100% wheel weight alloy, water quenched, using the Lee .225 55 gr flat nose mold.
• Gas check: Aluminum gas checks are installed after casting using a Lee .225" push-through sizing die. My push-through sizing die appears to have a tiny amount of clearance between the ram and the inner walls of the sizing portion of the die. Gas checks are not turning into nice flat bases, but getting thin "fins" of aluminum protruding around the circumference of the base downwards. Any tips for dealing with this would be welcome; it has never happened for my .358" or .314" sizing dies (which I use for 9mm and 7.62x54R respectively).
• Coating: Checked boolits are powder coated using Elvis's "heat to 140 first" shake 'n bake method, then baked at 400F for 20 minutes. Powder coat is Smoke's Clear. Coated boolits are then sized again back down to .225". This powder coat is magic, and puts the HF Red to shame. My bullets look like lead instead of lipstick now and I love it. Mad props to Smoke.
• Hardness: After PCing and sizing, boolits are allowed a minimum of 1 week to age harden before loading. Although I don't have a hardness test, I estimate the process of water quenching WW alloy, then semi-annealing during the PC process, then allowing to age-harden again is leaving me with a hardness somewhere in the range of 15-16 BHN. I was too impatient to wait for them to harden a full week. But no leading, and by the time I get around to shooting any more, those bullets should be well over a week past their PC date.
• Brass: All I have is range pickup, mixed head stamps. All brass is decapped, lubed with Lanolin + 99% isopropyl alcohol, full-length sized using the Lee .223 pacesetter dies, and then wet tumbled with Dawn and Lemishine for two hours before being allowed to thoroughly air dry. I'm going to need to find a solution for reaming or swaging primer pockets. Had trouble getting primers to seat consistently or without excessive force, I assume because a bunch of my range pickup is military brass.
• Primers: All primers are CCI 400 small rifle primers (non-magnum).
• Loading Procedure: Necks are expanded prior to seating boolits using the Lee Universal Neck Expander die. Boolits are seated per reloading manual to a COAL of 2.060". After seating, a light factory crimp is applied to remove belled case mouths and allow the cartridges to chamber, but not enough to swage projectiles down smaller than .225". I'm torn between trying a longer COAL for the sake of accuracy, and leaving it as is for the sake of having somewhat trustworthy estimates on velocity and pressure. I think for now I'll leave them short, and once I have a load that cycles reliably I'll start trying for longer.
• Shooting Procedure: For the sake of consistency, the same 20-round pmag is used for all testing. As my sights are currently zeroed for factory j-word ammo, I'm going to shoot at 25 yards. I am shooting off a bench using a Harris-style bipod. The bipod is rested on a sandbag rather than the hard bench top, and I am leaning into the bipod to load it for each shot. The short range should allow me to get on paper for this test even if POI is dramatically different with these lighter loads, which it probably will be. I realize shooting at 25 yards doesn't give me the best data for measuring accuracy, but gross trends in 10-round groups should be observable, and this test is really more interested in seeing if I'm cycling the action consistently with this powder, and if I'm getting any leading (I shouldn't be). Once I get a reliably cycling load, I'll re-zero my sights and really dial in the load for accuracy at a longer range.

Looking forward to testing these rounds and reporting back.

I look forward to seeing what you come up with.

I would advise you to google the "Houston Warehouse Experiment" and pay close attention to brass prep. Also get some copper checks to compare groups with.

First off, early congratulations on the upcoming life changing event, and unless you have cloned yourself you will find that you will be far too busy for the next month or more to be making any videos. But you will enjoy it, remember every minute of it because they grow up way too fast and post some pics because as you know without pics it never happened...... LOL

Now about the loading, I was working with an NOE mold, if I remember it was 70 grain GC design. I did not do enough work with it to make it worth writing about so I am much interested in seeing what you come up with. Good luck with both endeavors.

H4895 was designed to produce large gas volume to cycle gas operated rifles. I've had good luck getting a 5.56 AR upper to cycle using 16gr of H4895 and 55-62gr boolits. The upper was not mine so I was unable to do much in the way of accuracy testing with it however others have reported decent accuracy using this powder for reduced loads. Some reported needing to use as much as 18gr for proper cycling.

18.5 of 4895 was needed for full functioning on ours.

Quote:

---

Originally Posted by MT Gianni

I would advise you to google the "Houston Warehouse Experiment" and pay close attention to brass prep. Also get some copper checks to compare groups with.

---

I had never heard of this study and it was an extremely interesting read.

Quote:

---

Originally Posted by MT Gianni

I would advise you to google the "Houston Warehouse Experiment" and pay close attention to brass prep. Also get some copper checks to compare groups with.

---

^^^ Yeppers! ^^^

Quote:

---

Originally Posted by BHuij

Hey everyone--

I've spent lots of time reading here and gotten lots of great info. Now that I've joined the ranks of the AR-15 crowd, I'm tackling my most ambitious casting goal yet: an inexpensive cast .223 load that cycles my "M4 clone" reliably, and is accurate out to at least 100 yards, preferably 200.

I'm doing a short video series to carefully document my "somewhat-scientific" approach to chasing down this unicorn load. The documentation is more for myself than anyone else, but if it helps a brother out who is as overwhelmed with information as I was when researching for this, all the better.

This first video is mostly just an introduction and the culmination of my research into the different factors that go into a successful cast load in an AR-15. About 10 minutes of talking and reading off of a paper so I could remember everything I wanted to cover... haha.

https://www.youtube.com/watch?v=EriLIM-_e-A

But this weekend or next, barring a life-changing event (my wife is due to give birth to our first child any day now :D), I'll be testing out my first set of loads to see where they get me. I will be carefully recording my tests, and documenting my findings for the next video.

All testing will be done on my rifle with the following specs, so anyone looking to develop a load for their own rifle will have mine as a starting point:

Anderson Lower
Bear Creek Arsenal Upper
16" chrome-moly barrel with 1:9 twist
Carbine-length gas system
Standard carbine buffer and buffer spring
Standard 8620 steel BCG

---

You can get them a lot harder than 15-16 BHN after PC. I bake mine at 450 for 20 minutes but thats with bigger 308 bullets. Try starting with 425F you'll be surprised how hard they can get

Sent from my SM-G950U1 using Tapatalk

Quote:

---

Originally Posted by dimaprok

You can get them a lot harder than 15-16 BHN after PC. I bake mine at 450 for 20 minutes but thats with bigger 308 bullets. Try starting with 425F you'll be surprised how hard they can get

Sent from my SM-G950U1 using Tapatalk

---

Are you implying that baking hotter during the PC process (without quenching after) actually makes for a harder final boolit? That goes against my theoretical knowledge of how annealing works, so if I'm not misunderstanding here, I'd love for you to expound on this.

Quote:

---

Originally Posted by MT Gianni

I would advise you to google the "Houston Warehouse Experiment" and pay close attention to brass prep. Also get some copper checks to compare groups with.

---

Read it this morning and it was incredibly fascinating. Thanks for the pointer.

To me one of the most astonishing things, aside from the brass prep and skewed results with mixed brass, is the variation of several grains of powder in the same batch and still shooting in the tenths. I really miss Felix when it comes to questions like this.

Quote:

---

Originally Posted by BHuij

Are you implying that baking hotter during the PC process (without quenching after) actually makes for a harder final boolit? That goes against my theoretical knowledge of how annealing works, so if I'm not misunderstanding here, I'd love for you to expound on this.

---

Definately water quench after PC bake without delay.

Sent from my SM-G950U1 using Tapatalk

Well, my wife was most resoundingly NOT in labor today, and the gun range is very close by. So I went for it.

I have amended the “Test #1” post to include results as well as a video going in depth. That is post #2 in this thread.

Not so much luck today, it’s looking like I’ll need to try another powder. So far I have seen recommended:

• CFE223 - Specifically designed for this cartridge (though I assume with j-word bullets in mind)
• H4895 - Specifically designed for large gas volume for use in semiauto gas guns like ARs and AKs
• H335 - Apparently a very accurate powder for .223 projectiles, at least according to two separate people I talked to on Reddit.
• Ramshot Tac - My cousin uses this with great success (again, not with cast).

If anyone has other suggestions or can vouch for one of those powders (particularly if you can vouch for one using cast boolits in .223), I’d love to hear. They all seem to fall right around the same zip code for burn rates, and all of them are significantly slower than 4227.

So barring any really compelling reasons to choose one over the other, I’m inclined to just see what’s cheapest at Sportsman’s Warehouse :D

Oh, and on that off chance that anyone is in Utah and wants to trade me a small amount of one of those powders in exchange for Bullseye, AA#2, or 4227, that would be great. I’d love to avoid a scenario where I spent $30 on a pound of powder that doesn’t work for me. I guess worst case scenario I could always work up a new load for my Mosin with the slower powder and see if I can beat the accuracy I’m getting from 4227 in 7.62x54R.

Currently using 19.5 h4895 for the same boolit from a 16" carbine. Reasonable accuracy at 100 yards. Hasn't been mentioned yet but for best accuracy in these tiny pills you gotta weight sort them ideally within tenth of a grain. Found that my rifle will function at 18grains if I use a filler (dacron). Also was not happy with the aluminum checks and switched to copper for better results. I have h335 & varget on hand as well but haven't done much with it yet.

Still a work in progress for me as well, was considering trying adding copper into my lead, using a heavier/longer boolit, and making a sacrifice to the boolit Gods. Either way it's too cold for going to the outdoor range so this has been on hold for a couple months. Am interested in what you come up with.

From what I'm reading, H4895 is best for function, H335 is best for accuracy, and CFE 223 is the cleanest running. Since I haven't yet gotten mine to cycle, I think H4895 is my next one to try. Maximum port pressure with minimum velocity seems to be the recipe here for these light 55gr cast. If I just absolutely can't get reliable cycling without pushing beyond 2000 FPS, I will consider a lighter buffer spring or a heavier boolit. Thanks for the info! Especially the tip about the Dacron filler allowing for lighter loads, that's good to know.

If you have not already done so, check your accuracy with jacketed bullets. It gives you a baseline of what you and the gun are capable of.

The AR is not an easy platform to work with. If you want blasting ammunition, that is an easier goal than accurate ammunition. You will not achieve accuracy unless you visually inspect and weigh every bullet before lubing/coating and GCing it. If you cannot seat the GC squarely, you are wasting time and resources.

I bought the NATO (I believe about 70 gr) mold to undertake a similar project for my AR’s and decided to abandon it. BTW, A heavier bullet should help. My initial thoughts were to use 4895 and if that did not work, Varget. I think 4895 is a good choice. My alloy was going to be Linotype to get the hardness and if that worked, I would try 92-2-6 next. Both water quenched. My lube was going to be 2500+ from Lars. Lastly, copper GC’s.

My objectives were to be self-sufficient in case I needed to be, and save some money. I do not reload or cast for “fun” like many here do. Facing the prospect of doing all that work casting, bullets sorting, lubing, sizing, load development etc and likely not matching the accuracy or performance of a cheap Jacketed bullet made me question why bother.

One of the vendors here had Hornady 55 gr SP on sale for $7/100 in bulkpacks of 6000. It was an easy decision for me. If the supply of inexpensive jacketed bullets ever becomes an issue, I still have the mold and can address it then. So I follow these kind of threads with interest.

I wish you luck....And yes it can be done. But it will not be easy.

19.5 grains of h4895 was @2100fps from my 16" barrel that was lubed with carnuba blue, copper gc, wd coww, no filler. Would like to try the cfe223 bay see if I can pick some up for the heavier boolits.

Thanks everyone for the info. I've done plenty of reading (here and elsewhere) over the past few days. My wife and I welcomed our first child, a little boy, into the world on Sunday afternoon, and we came home from the hospital only yesterday. Mom and baby both doing great. While we were staying in the hospital, there wasn't a whole lot to do in between feeding the baby and changing diapers, so I spent a good amount of time researching and putting together my next round of tests.

Now that we're home and I'm taking the majority of this week off from work, I'm hoping to be able to put a video together soon describing the next test and what I've decided on. We're kind of in the same boat here as we were in the hospital, time wise. Baby alternates between eating, sleeping, and pooping, mom is napping whenever possible, and I have time to kill, but want to stick around the apartment for obvious reasons. So although I've decided on what I'll be trying out next, and will probably be able to put the test rounds all together by the end of the week, we're going to have to play it by ear to figure out when I'll next be able to spend a few hours at the range. It may be a few weeks.

Anyway, stay tuned.