This weekend I took the AUG out to the 200 yard line, set up sandbags, and fired 5 shot groups of 10 different ammo's.  The goal of the test was to determine point of impact shift with different ammunition.  



The rifle is an 18" AUG with integrated 3X optic.  The test was conducted from sandbags with the for-grip folded and resting on the sandbags.  This was done, with some thought.  The BBL of an AUG is not floated, and by doing this, pressure is put on the barrel.  This could potentially change point of impact, and more represents typical usage of the rifle.  This is not necessarily the most inherently accurate configuration (I could have rested it on the stock, behind the grip), but was chosen because this was felt to more aptly represent common usage.  

Barrel temperature was noted, though no extra effort was made to allow it to cool between strings.  The AUG BBL sheds heat remarkably fast, and does not warm up that fast. It never exceeded 120 F, and fell quickly.  (In comparison, I repeated this test with an M4, and that BBL got up to >140 F very quickly)

Results are summarized graphically below, on a target grid using 1" increments per grid.  This is for 200 yards. MOA deviation will be around half the number of grids markings.  To interpret the graph, Wolf Gold was used as the reference zero, and all other groups are represented by their deviation from that zero.  It should be noted that a repeat of the Wolf Gold group was fired at the end of the test, and the graph shows that repeat almost 1.5 MOA off from the starting group.  Representing the degree of error in this test, so while at least this is an indicator -  the quality of this data is far from supreme.



A few trends and conclusions to make.  

-The commercial 55 gr ball loading's (Wolf, IMI, CCI) all had centers of groups that shot within around 1.5 MOA of each other (obviously some individual impacts where much farther then that).  Note, CCI Lawman 55 gr ball is loaded on LC brass and is very similar to Federal XM193, so without further data available, is assumed a close analog for performance as well.  (Looking anew at the data, one could say the reference zero for commercial 55 ball should probably come up about 1/2 an MOA - tough alas, the AUG scope is 1 MOA per click).
-The 55 gr ball handloads of mine (223-363) are somewhat mild, and are the only 55 ball loading to impact higher than the rest, at about 1.5 MOA high.  Not my first guess for a 200 yard impact, but there it is.
-The M855 62 gr loads (IMI and RORG) both high very close to each other, and about 1-2 MOA higher than the 55 ball.  Slightly to the left, but only in comparison to Wolf Gold.
- The 50 gr AE was the most accurate ammo group of the day, and grouped the most left of all the ammo tested, and higher than most of the 55 gr ball.  Curiously, similar POI as the M855 ammo.
-The Federal Law Enforcement Tactical Bonded 62 gr hit higher still, at about 3 MOA high.  That's enough deviation to start to matter, and a non-perfectly centered shot on a target zero'd for 55 ball might miss a vitals shot, if not zeroed for it.
-The 75 gr match performance is what surprised me most.  Prior testing showed it and the 55 ball to have almost the exact same point of impact at 200 yards as 223-363 (55 ball handloads) were impacting.  Yet today, it was shooting close to 4.5 MOA high.  That's 9" at 200 yards, and that's definitely enough to notice.  I also noted considerable vertical stringing what that group.

I also am beginning to suspect that the AUG point of impact is affect by hold, with more sensitivity than one might first have guessed.  My data is spotty on that, and further testing is needed.

For reference, here is a table of Average To Center (or Mean Radius) of the 5 shot groups.  Note, these are single strings of 5, so not the worlds most statistically definitive.  Update to also include extreme spread

Wolf Gold: 0.708 MOA  (1.8 MOA extreme spread)
Wolf Gold repeat:1.3 MOA (gah)  and (1.24; 0.74 (2); 1.06 (2.7 MOA extreme spread); 1.62 (4 MOA extreme spread) in repeats during sight adjustments - data not included above)
CCI M193: 1.47 (3.98 MOA extreme spread)
IMI M193: 0.986 (2 MOA extreme spread) and (1 MOA in a repeat during sight adjustments - data not included above)
223-363 (55 ball reloads): 1.27 (bad day for this load, normally much better)  (4.2 MOA extreme spread - had a flyer)

IMI M855: 0.956
RORG M855: 0.924
Federal 62 gr Bonded: 1.97 (4.12 MOA extreme spread)

75 gr Match: .695 ( 2.1 MOA extreme spread) and would have been much tighter then that, except for a line of vertical stringing over most of that)

American Eagle 50 gr Ballistic tip Varmint: 0.47 MOA (wow; 1.1 MOA extreme spread)  Link to that one:  http://i62.tinypic.com/25jzjvb.jpg



Conclusion.  For most practical purposes, the same zero for M193 and near-M193 spec ball-  as one would expect.   62 gr M855 hits a few MOA higher, but not by a lot, and for most "kill-zone" applications, probably good enough for 200 yard work to keep the same zero - also not that surprising.  The 50 gr varmint loads also grouped high, which isn't quite my expectation, I would have guessed they would impact lower (lighter bullets recoil less causing less muzzle climb).  The 75 gr match loads really surprised me, with group destroying vertical stringing, and just really high point of impact (the horizontal spread was only 1/2 MOA!  Gah!).

Thanks lazyengineer.

I love your presentations because they are, well, presentations/productions, but heck, you're an engineer!

This also explains why my mild hand loads were higher than my pmc strikes. Which I'd been wondering about.

I see you are still getting good groups even with the grip rested on the bags.
I tested mine this weekend with the grip on bags, and it vertically strung the groups.
My gun is on its way to Steyr so they can check out some looseness in the barrel, and its inconsistant accurracy.

1. Groups will open up as the barrel gets hotter





2.  Vertical stringing with an AUG is almost always due to the shooter slightly moving the gun in the vertical axis via the main grip or at the shoulder.  A small sandbag under the rear of the stock helps here.





3.  Heavier rounds do not impact higher because of increased recoil.  By the time the recoil impulse is felt, the bullet is long gone.  Velocity differences and higher ballistic coefficients are in play here (this is why even the 50gr tipped varmint went higher--a higher BC allows the round to carry further/flatter).  The 75gr Match round hit high because it's a very efficient flyer due to its good BC.





4.  Those were damn good groups almost across the board.  M193 type ball ammo at or under 1" MOA mean radius is right at spec.  





5.  Most people will tell you zero your optic using whatever your main accuracy/defense round is (ADI 69gr SMK in my case).  Then the less expensive ball rounds shot for fun or practice just hit where they may around your point of zero.  Bullets do not travel in a perfect arching spiral though the air as show in this graphic.  This is where 'fliers' tend to come from...





 
 

Gentlemen, thank you for your comments.  I have learned in my career that that data is not really analyzed and understood until one composes a report on it (That learning experience was not painless by the way).  Seemingly obvious things are easily missed by simply looking at the data, but not composing a report.  For example, the main purpose of this test was my frustration that point of impact when firing CCI M193 was much lower than for 55 ball handloads, which I thought would be the same.  Yet, only now and during composition do I see that yes indeed, M193 does impact about 3 MOA lower.  And if the zero was 1 click off (easy to do), then the 4 MOA difference I was seeing, makes a lot of sense.  This is encouraging, because I thought my rifle was just randomly loosing zero - (which was pissing me off, a 4 MOA shift in zero hold would have been a no-go, and off the gun would go back to Steyr!)  Now it makes more sense.  

Thanks mcanut for the yaw graphic on M855.  That is quite interesting, I had not seen that before.

To clarify, the posted group performance values are Mean Radius, not to be confused with the more common extreme spread.  This was done because I find mean radius a more valid statisical comparison of performance.  Appologies for the confusion, I will update the OP with the extreme spread results too.  Which will show that while the AUG accuracy is very good with some ammo, by and large it's only pretty good with most.
[edited]

i think most people that have spent time in the arfcom tech forums will agree that mean radius is a better gauge of rifle/bullet accuracy than extreme spread.  the spec for M193 calls for a mean radius of under 1"@100m with 10 round groups from a fixed test barrel



 

I guess the reason I think that recoil from heavier bullets causes impact elevation, is because I saw this published by Julian Hatcher.  I've always considered him the grandfather of modern rifle technical wizardry (secret engineering technical term, we have hats).  He discusses his test results and engineering calculations on this on page 297 of Hatcher's Notebook: (Click on it, and you can read it online)  https://books.google.com/books?id=yESNUKSg5aMC&printsec
He makes a convincing case that heavier bullets recoil more, causing muzzle climb, elevating point of impact.  I've found this consistent with my own rifle range tests, even at short range where the improved ballistic coefficient is irrelevant.

I have been following your case with interest. If you could share some of your tabulated performance data, that would be helpful.  My AUG can, and usually does shoot well, However, I can't say I never experience what you complain about.  While trying to zero my scope, I had one group just do random things on me that drove me nuts - consistent with your own complaints.  I wouldn't call it typical, but it did happen.  Here's the target in particular I'm thinking on.  First 2 shots are good (this is at 200 yards, those two are less than 1 MOA from each other)!  Then the next 3 are a major WTF?!?  Something changed, my suspicion is it has more to do with the shooter's hold, than the gun itself, since often the gun would shoot great.  Then again, maybe the gas port got dirty and that affects things?  No idea.  Steyr's response to your service request will provide additional data for me to judge.   I should also probably note, this is with Wolf Gold, and while that's the most accurate bulk ball ammo out there - it does sometimes throw a shot.



But in comparison, when you get groups like this, you relax a bit.  The American Eagle 50 gr Varmint that PSA was selling for $5.79/box was probably the best deal of the year.  I'm glad I bought a bunch.  




The AUG is a good gun, but in my opinion, lacks the inherent accuracy advantages of the AR system.  The barrel is not floated, and has many heavy things with moving parts hanging off of it.  It is thick, and a better quality barrel than most M4 barrels, which helps some, but that only goes so far.  Torque or inconsistencies of hold on the fore-grip, will be torquing on the barrel.  And yes, because the barrel is removable, there is going to be some inevitable play - though quite little from what I can see.  With the pistol grip hanging down, there is more mechanical leverage to amplify that torque.  Upon firing, tens of thousands of PSI of gas shoot into a gas cylinder hanging off to the side, which moves a piece of metal in a tube, and ejects gasses (with some vigor) through a port off to the side.   The short stroke piston flies back (scraping along the gas cylinder attached to the BBL), compressing a spring, also hanging on the side of the BBL.  This metal thing then whacks the shit out of another piece of metal (the piston attached to the bolt carrier group).  It works, but it's a similar system in concept to the M1 Carbine, which is a notoriously inaccurate little POS (yet fun little bastard too, but I digress).   All of this is in comparison to the AR which has a little gas tube that diverts gas pressure back to the receiver.  To me, it's a wonder that AUG's shoot as well as M4's do (usually, close enough), and a testament to the quality of construction.  I shudder to think what a half-baked rendition from an "I can make that too!" country must be like.

I was going to say that it would be an interesting experiment to shoot the same test with an 18 inch same twist AR.

I also am thinking about the statement that heavier bullets don't print high due to recoil. Force is applied equally on the bullet and everywhere else. I am not an engineer but I know that so I would believe recoil is simultaneous unless you're talking the bottoming out of the bolt in the recoiling  cycle in the rearmost  travel of the semi bolt.  FWIW I see the same in bolt rifles. My first deer rifle, a Ruger 77 ,shot 180's higher than150's. It made me thin a of it way back in the eighties.

Also the m1 carbines I have are better than I read about. I don't think they suffer until worn out and rebuilt as a parts gun. The book at the rear of the action and how it mates with the stock block is critical. Both my inland and Winchester CMP service grades shoot fine about 2.5 months.  Which is all they were meant to do. The AUG I am sure was similarly spec'd as a combat rifle.

I can see the troubles one would have trying to bench an AUG.

Stay tuned, because I did.     (Well, not quite: 16" - same twist).  

As to the bolt - interesting thoughts, thanks for sharing on that.  I suppose in some ways, the short stroke piston is better for accuracy than long stroke?  I don't know the timing, but if the short stroke only hits the long stroke after the bullet is gone, then that's a lot less mass moving around the BBL, than a dedicated long stroke (like an AK) would have.  I wonder, has anyone published the timing on that?  Maybe I'm wrong on that, and short stroke is more inherently accurate than Long Stroke?

You're probably right on that.  I'm probably not being fair blaming the M1 Carbines accuracy woes on its gas system, entirely.  I know M1 carbine competitors will very carefully use the longest casings they can find to reload, and will pay a lot of attention to the recoil plate.  When they do so, they can get darned near 1 MOA out of a carbine.  And neither of those improvements are related to the gas system.  That said, my God man have you looked at that thing??  The long stroke slide in an M1 Carbine is literally running in rail grooves directly cut into the barrel itself!  That just freaks me out - how can the thing hanging off and literally directly sliding along the barrel in rough-hewn grooves like that be a good idea for accuracy!!?  That said, there's just something about an M1 Carbine, you just grow to love the little POS.  It amuses me that the M1 Carbine shoots a 30 cal bullet using a piston system - designed in the 1940's.  And all the cool kids today are doing everything they can to convert their M4 carbines into piston systems shooting 30 cal bullets...

I'm really looking forward to the AUG vs AR comparison mentioned above! I've been wanting to do that for years. Were they similarly configured?

Somewhat surprising results

Repeat of this test with a 16" AR Middy.

wow, that's pretty frelling badass comparison work.  looks likes accuracy is pretty much a draw between the AUG and AR.



 

Nice work.

So from an accuracy standpoint, the M4 and the AUG are pretty much same ballpark.   Slight advantage to the M4 maybe (maybe) - data quality not good enough to really call that for sure.  The more noteworthy difference is that the M4 seemed to have less sensativity to POI shift due to ammunition type.  The AUG has something like... well, now that I look at it, there is a cluster of 55 ball impacts, and a cluster of 62 gr impacts.  And the difference between those clusters is about 2-2.5 MOA.  Whereas the M4 55/62 ball clusters pretty much overlap, but... OK, let's call it a 1 MOA POI shift down, when going up in weight.  Not a big deal, but a slight advantage to the M4 from a practicality standpoint.  

The 75 gr ball is the biggest mystery.  Not only is this round hitting way the hell off from everyone else, but on the M4 it hits way the Hell low, and on the AUG, it hits way the Hell high.  It's like 10 MOA difference between the two rifles, in shift direction!  I can't figure that one out.  The M4 shoots heavier stuff low, and the AUG shoots heavier stuff high.  

Artifact of these individual guns, or if this is a platform phenomenae, I have no idea.  It would sure be a whole lot cleaner if they both went in the same direction at least!

As a side project, I'm going to start tweaking my 75 gr load, and see if I can find one that has a similar POI to the 55 ball loads, as I would find that a convenience.   Slightly frustrating - prior testing suggested that it already was the same...

Great threads.  

I think the difference is attributable to the barrel/trunion design.  The AR has a huge advantage here.  The fact that the AUG's barrel is QD opens up tolerance in this area where the AR's is essentially zero since it's torqued down.

i remember hearing, going back to the 90's, that AUG barrels produced higher velocities than ARs.  could this velocity increase, along with the increase from 2" more barrel, be the reason for the rounds hitting so much higher than on the AR test?

I've noticed a couple of things in comparing the two sets of results:



1. With ball rounds both the AUG and AR perform about the same



2. With 'match' type rounds, the AUG groups are noticeably smaller



3. The 'horrible' bullpup trigger must not be all that bad if points 1 and 2 are true

Returned to the range today, and confirmed zero's.   Indeed, no changes.  For reference:

A3M1 18" 3X scope settings for 200 yards:

M193 55 ball

i.e. +2 ticks from my reference mark

Mild 55 gr handloads:

i.e. -1 tick from my reference mark

75 gr handloads (~2700 FPS - semi mild)

i.e. -3 ticks from my reference mark (some variability, above testing showed it probably should closer to "-5 ticks" from reference mark, but general direct is about the same.  Only had chance to fire a couple rounds to check zero today.

Good report and photos.  Thanks for your hard work and sharing your information!