This one might be a little involved.  I'm considering putting a 9mm AR together, but far too much of what I read about them indicates accuracy is all over the place.  Everything from great to groups worse than rack grade AKs.  Keep in mind, my accuracy expectations of a 9mm are considerably different than what I'd expect out of a 5.56 or 308.  If it will shoot 2 inches at 50 yards, it's good enough, and since I consider 9mm to be a 50 yard cartridge max, I don't even really care how it performs past 50 yards because I don't see >50 yards as a valid use case of a 9mm carbine.  if it will shoot 1 to 1.5 inches at 50 yards, that'd be awesome but I understand that's getting into cream of the crop territory, and I don't consider that to be especially realistic.

With 5.56 ARs, the single biggest factor in determining accuracy seems to be barrel quality.  This is pretty normal for most rifles.  With 9mm ARs, my observation is that there is more at work than just that, as evidenced by the fact that some RRA LAR-9s shoot well and some of them...not as well.

I want to understand what factors affect 9mm carbines in the accuracy department.  I am not going to spend a bunch of money building something that throws shotgun patterns, so I need to understand what factors I'm trying to control to achieve the desired result.

The obvious thing to look at is the operating mechanism.  Blowback is a very different system from gas operation.  In a gas gun, your bolt doesn't unlock and open the breach until after the projectile has left the barrel.  With blowback, your bolt isn't locked at all to begin with. It makes sense that the breach would start opening before the bullet left the barrel in such a system.  I can see how that might disrupt things a bit.

The prescribed fixes are heavier bolts, heavier buffers, and heavier recoil springs.  9mm bolts are heavier, that's not changing.  I can change my buffer and recoil springs, however.  So I am asking this question for those that have tried it, would running a very heavy buffer (the KVP 10oz tungsten buffer comes to mind) combined with a strong recoil spring, such as a 308 carbine spring, have positive effects on accuracy?  The idea being that these factors keep the bolt in battery longer.  If the answer is yes, then which of the two (heavier recoil springs vs heavier buffer) has the most effect?  Would I be better served with a lighter buffer and an extra power recoil spring?  Certainly, this seems to be what a lot of 9mm carbine manufacturers are doing.  I've noted that the charging handle on the Beretta CX-4 can be challenging to pull compared to an AR because it's recoil spring is especially heavy (incidentally, the CX-4 is probably the gun I'm going to buy if I can't satisfy myself that I can build a 9mm AR that will perform to my liking).

In case it matters, in my case, if I build a 9mm AR I'm looking to build it as a pistol with a 5 to 7.5 inch barrel, to be converted to an SBR once the paperwork comes back.  If what I've said above has merit, the short barrels ought to do a little better.  With the bolt exposed to high pressure gas for a shorter period of time, it ought to not come out of battery quite as far by the time the bullet reaches the muzzle.

This got to be a bit of a rambling post.  Other than what I've already mentioned, assuming what I've mentioned has any merit whatsoever, what other factors affect the accuracy performance of 9mm blowback ARs, and can I control those factors to be more in my favor by selecting certain parts for my build?

James...I've always perceived that 9mm accuracy is dependent on length of bearing surface, fit to bore diameter, and using a cool burning propellant that that does not erode the bullet's base ... I had a P35 High Power that slugged .357" and could shoot 8" plates at 50yd easily with 38super bullets sized .357"...not with 3555"

if you hand load I would suggest a ball process propellant from Winchester.. I use Super Field with a long bearing surface polymer coated  135grain Bayou projectile...it easily shoots 2" at 50yd from my 16"

I shoot a gamer load of a 146 power factor, but tests with factory loads seem to favor a 147 bullet... I don't use a heavy buffer... 5.2oz with a rifle length spring ..bolt and buffer about 22.5oz

Barrel quality matters, as does bullet quality.  Powder has less effect on accuracy, but can have a significant impact on trajectory.  

My RRA 10.5" will shoot 2.5" 5 shot groups at 100 yards with Hornady XTPs, and it'll shoot 4" groups at 100 yards with Xtreme 147 gr heavy plated bullets.

My PSA 16" carbine would do it as well but only from a cool barrel (first 5 rounds or so).  However once I switched to a Ballistic Advantage barrel, I got the same performance I got from my RRA.  I'm not saying that PSA barrels are crap, just that my PSA barrel had some issues, probably related to improper stress relief.

----

If you hand load, you can maximize the velocity of a 9mm round in a carbine.    For example, selecting a slower powder can mean the difference between 1250 fps and 1600 fps with 124 gr bullets in a 16" carbine.

As for the comment above about colloidal ball powders - it's a mixed blessing.  To the extent they are slower burning and can make better use of a longer barrel they can make sense.  However the heavier charge of these fairly abrasive powders does increase wear in the throat.  It's not nearly as bad in 9mm as it is in .357 Mag, but it's something to consider.  The major advantage is that they meter very accurately and you can mass produce accurate ammo.

On the other hand, a slow flake powder like Blue Dot is still very hard to beat for maximum velocity in a 16" 9mm carbine, or a 8"-10" AR9 pistol.    

----

Provided you are using a well designed buffer system, that will consistently operate without excessive bolt rebound or excessive bolt velocity there isn't any significant benefit to a heavier bolt, etc.

----

Given 2 to 2.5 MOA accuracy at 100 yards with high quality bullets and 3.5-4 MOA accuracy with inexpensive projectiles, I regard the 9mm AR carbine as being effective out to 100 yards.   Even with a subsonic 147 gr bullet, you'll find that with a 100 yard zero you are still only 3" high at 50 yards.

If you are shooting a factory loaded 124 gr bullet at around 1250 fps in a carbine, then you cut that mid range trajectory to just 2" above point of aim at 50 yards.  

If you hand load with a slow burning powder to launch a 124 gr bullet to around 1600 fps, then the mid range trajectory is only 1" at 60 yards.  

This last load is very capable - but I've found plated bullets tend to fly apart so it limits your options a bit if you're cost conscious.  I have a 1000 lower priced FMJs on order to see how well they work in this higher velocity application.

Good info.  I do, in fact, reload.  It's harder to justify for 9mm when I can buy Wolf for the cost of handloading it...assuming I am purchasing commercial cast lead bullets.  Fortunately I also cast my own, and if I pursue that, I can load 9mm for the cost of 22lr.

1600 fps would be nice, but that gets dicey with cast.  Gas checks and powder coating or alternative lubes might help here, but it starts getting to be more trouble than it's worth.  Especially since the gas checks will push cpr back up close to Wolf factory ammo.

I'm basically looking at building AR pistols to be SBR'd later, and the calibers I've looked at are 9mm, 22lr, 7.62x39, and 6.5 Grendel.  At best, only two of those are going to happen.  I'm going to have to stop over thinking this and just choose.

By experience I can tell you that a 1-16 twist will give you better accuracy than a 1-10 twist, by far.  I can't explain it, I've owned both and can vouch for that reality!

Mine got way better when I free floated the barrel. It was already in the 2" range at 50 yds with quality commerical ammo in the carbine configuration with a FSB. I haven't had a day to do an accuracy test yet, other than a ragged sub 2" group at 40 yards off a the top of my range bag.

My barrel is a 16" Wilson, 1-10 twist.

Both my RRA 10.5" and my Ballistic Advantage 16" barrels have 1-10" twist and both will shoot 2 MOA at 100 yards with 124 gr and 147 gr Hornady bullets HAPs and XTPs).   They'll do a bit better than that with 90 gr XTPs ripping along at about 1575 fps.

Would 1-16" be better?  I doubt it with good quality bulelts - given the current level of accuracy, even with short, light, fast 90 gr bullets that are probably spinning a whole lot faster than they need too spin.  

However, with low quality bullets where the center of form and center of gravity might not consistently lie on the same axis, slower might be better, particularly with shorter 9mm bullets.