The following is a test conducted over the last week to observe the deflection of free-float rails with force applied to them. It is meant to provide information about how the POA of a sighting system (typically laser) may be affected by shooting from a barricade or bipod.

There seems to be an increase in discussion and speculation about this topic, likely due to an increase in night vision capabilities amongst civilians. While it’s likely various government entities and manufacturers have plenty of data about rail deflection of this type, public data is scarce. This is an attempt at a relatively simple, repeatable test.  

This is not a destructive test meant to glean any information about resistance to impacts and permanent deformation or shifting of the rail on the barrel nut. The POA was expected to return to its original position after the force was removed.

Test subjects (samples of 1 each):

BCM MCMR (13.5")
SLR ION Mid (11.7,” previous version)
Forward Controls Design RHF (11.5,” manufactured by CMT Tactical)

METHOD

Each upper was held with a vise 50 yards away from a wall (measured via SIG Kilo1600 laser rangefinder). The vise clamped to an ADM micro mount attached to the top of the upper receiver. The uppers were positioned upside down so weight would be applied to the underside of each rail, as it would when using a barricade or bipod. The laser (Holosun LE221G) was attached to the top side of each rail. Each rail was subject to two iterations of the test, one with the laser at 10” from the upper receiver (or the next aft 1913 slot), and one with the laser as close to the upper receiver as possible without touching (“Laser at Nut,” in the data chart).

The laser was projected onto a piece of white paper and its center point was marked with a blue marker. Force was applied by hanging weight on the rail 10” from the upper receiver. Each weight load was applied and removed 3 times for every rail/weight/laser placement combination. The center of the laser with force applied to the rail was marked with a red marker. The weights used were 2.5lbs, 5lbs, 7.5lbs, and 10lbs. Each rail/weight/laser combination had its own paper.

The distance between blue and red dots on each piece of paper were measured in inches with a digital caliper. Those measurements were recorded and values converted to MOA ([Value]2/1.047), see below.

CONTEXT OF FORCE

To provide context for the weight applied and how that amount of force correlates with real-world use, I used a bathroom scale zeroed with wood supports and a bag rest that provided a base for an AR rail. I then placed a fully assembled AR’s rail on the bag rest and tried to mimic the feel of using a barricade or loading a bipod as closely as possible. Since my testing partner was not available for this portion, I did my best to observe the general range of weight represented on the digital screen.
I witnessed over 18 pounds, although most values were between 12 and 17 pounds.
With the AR’s handguard simply resting on the bag with the rear of the AR raised off the ground, 4.2 pounds was applied.
Therefore, one could expect significantly more deflection during real-world use than what is shown with the maximum weight in the data below.

RESULTS

Attached File


Attached File


Attached File


Attached File


ANOMALIES

It wasn’t until the third rail test that we realized that we hadn’t been using the truck’s outriggers. An outrigger would have mitigated the chances of the rear of the vehicle sagging due to the uppers and different weight loads. We completed all iterations of the test without the outrigger, and then one additional test with the outrigger to observe any difference. The difference was 0.09” at 50 yards, or 0.18 MOA. We did not feel this was a meaningful difference.

One test showed a shift in zero/neutral after the first repetition, the BCM MCMR 10-lb/10”. When the shift was noticed, the new laser position was marked and the weight was applied and removed 4 times to demonstrate consistency. We believe this was caused by slippage of the ADM mount in the vise. Carboard was used between the teeth of the vise jaws and mount, and we tried not to deform the ADM mount when tightening the vise. (The ADM mount seems to have survived the entire test unscathed).

DISCUSSION

I do not have an education in engineering so my predictions about how these rails would compare is based on intuition. We predicted the BCM would exhibit the most deflection. Its design seems to prioritize ergonomics/thinness and lightweight above all else, with very little extraneous material and thin walls. The mounting system is unique and advertised for its reliability but features a comparatively very short barrel nut. The FCD and SLR models feature relatively long barrel nuts and similar mounting systems, with two separate “wedges” that “cradle” the barrel nut and fit into a slot machined into the circumference of the barrel nut. Two bolts clamp the wedges into the nut, and the slot fit should prevent any longitudinal movement. I predicted the FCD would perform slightly better than the SLR, as it purposefully lacks machined cuts along the sides of the 1913 rail protrusion.

The graphed data shows linear and consistent deflection, with the exception of the BCM test with the laser at the barrel nut from 7.5 to 10lbs, where there is a more exponential increase.
An entire rail’s data was removed from the final presentation here because we felt the data was too suspect and indicative of experimental error. Not only did it show a pronounced exponential increase in deflection with the laser at the barrel nut, but the amount of deflection exceeded the same rail’s deflection with the laser mounted 10” from the receiver. This simply didn’t make sense to us, and we decided it should be mentioned but wouldn’t be fair to include.

We expected significant deflection from all rails having watched Dugan Ashley produce a sizeable shift at 100 yards from a very stout H&K quad rail: Handguards & Lasers: A Story of Deflection. It was unlikely that any of these rails would perform better.

The choice to include testing with the laser placed at the barrel nut was mostly inspired by this discussion amongst a night fighting SME and other shooting professionals: Aiming Lasers at the Nut. Chuck Pressburg describes knowledgeable professionals mounting their laser modules to the rear of their free-float rails despite the compromises in ergonomics it introduces. We are also seeing recent developments of upper receiver-mounted solutions such as the GBRS Group Hydra and HRF Concepts’ SKIFF prototype. The data from this portion of the test could also be relevant to those who utilize an offset dot clamped to their rail.

There is room for improvement in our methods. We did not measure the projected laser dot at 50 yards. My friend did his best to mark the center each time. We estimate the dot to be 0.75” in diameter. Also, all accessories were removed from the rails except the SLR. Could accessories mounted in the right places have made the difference? Different upper receivers were involved as well. The FCD is mounted on an Innovative Arms WAR upper receiver, the BCM is mounted on an Aero Precision M4E1 receiver, and the SLR is mounted on a Triarc mil-spec receiver. The Innovate Arms receiver has material removed in an area near the front to accommodate the adjustable gas switch. More robust upper receivers than used here do exist from VLTOR, FCD, BCM and others, but it isn’t clear how much those would sway this data.

Despite any experimental/human error introduced, we believe this data still has value in estimating how much deflection one can expect when using their aluminum free-float rail of similar design. We hope that it inspires others to add to the knowledge base and/or inspires further refinement and development of repeatable and accessible testing. I am lucky to have a friend with access to the truck-mounted vise and space that made this possible.

There is more to explore, such as how much upper receiver and/or barrel flex is produced concurrently from these forces, and how M4A1/clones and monolithic designs compare.

Thank you to those that provided suggestions and feedback along the way!

A few photos that couldn't be added to the OP:

This Kodiak C5500 mobile workshop provided a moveable, sturdy vise:
Attached File


Attached File


Attached File


Attached File

Pretty interesting test.
I wonder what the average amount of pressure applied to rails when you shooting normally?
I'd be interested in the Geissele MK4 and the KAC URX 3.1 rail tests.
Thanks for the post OP!

It will absolutely affect point of impact.  I have not done a test with parameters such as this, but my informal shooting has shown significant POI shifts with irons on a free float as well as with optics on a rifle with typical mil-spec handguards and FSB's, depending on how and where the gun was resting on bags or a bipod.

As more folks get into NV and IR lasers get more popular, we will see a migration away from the last decade trend of light weight thin rails to more robust systems. It’s already happening.

I would like to how the MI NightFighter and Ripcord LDR1 would compare in these tests as rigidity for lasers is their claim.

There is no point to shooting to establish rail deflection. This was one of the biggest flaws with that garbage nocturnality test, among other reasons.

Nice, but flawed test...

Those lever, quick detach mounts will move with force applied, and you have cloth between vice jaws and the QD mount.
Both of witch will allow movement.

Glad to, I enjoyed this.


You mean offhand? It would be difficult to measure that in actual use but the bathroom scale showed 4.2lbs for me when resting an AR's fore-end on the bag. This will vary a little depending on the particular gun and what accessories are attached (suppressor, etc.). I use a front rest when zeroing which should neutralize any difference. Shooting prone resting on the magazine could make a difference.


Those things could potentially cause issues, sure. One of the reasons we repeated each weight loading/unloading at least 3 times was to ensure consistency. On the 2 occasions we did see a perceptible change, I
made sure everything was tight and we repeated the loading 4 times with no inconsistency between reps.

If there was a perceptible amount of movement affecting our results, wouldn't we have seen a point-of-aim change between loading/unloading reps?

We used thin cardboard between the vise jaws and QD mount. I probably wouldn't if we repeated this, I feel I got my money's worth out of that mount already.

Thank you for the time and effort you put into this!

I didn’t catch that you weren’t clamping the receiver my first go round and grabbed on to the optic mount instead.

While I’m in no way saying the data is irrelevant, it does present an additional deflection point that concerns me with how concrete the results are. While it may or may not change the spreads, it very well could increase the overall totals.

I’d say you need to stabilize the receiver itself to be 100% sure.

My pleasure. Thank you for your online presence.



I see your point. We didn't have a way to quantify the QD to receiver force besides a feel of "really damn tight" at the lever for each upper (and same with the laser). The vise to QD mount seems a little more suspect but as long as there was a clear return to zero and consistency in deflection at the laser how could there be a shift at the vise-to-QD mount interface? I'm trying to imagine how there could be a rotation of the mount in the vise forwards, for example, and have an equal shift backwards when the weights were removed.

I can't prove it didn't have an affect on our data, though, and I have no way to explain the omitted rail's wonky data.

I didn't see an existing upper receiver block that inspired more confidence than this, but those options or purpose-made solutions would be worth a revisit for future testing.

A gentleman in the P&S community is currently developing a more extensive deflection test than I have done here, with many more rails. He gave me permission to post some of the results he's shared so far, which I appreciate.

It won't be an apples-to-apples comparison to my data. He's using a different method involving a MI Upper Receiver Rod, a boresight laser in addition to the rail-mounted laser, and a distance of 6.3 yards (at least for these results in this post). The rail-mounted laser is placed third slot from the end, with the load at the last slot.

This is with a 25-pound load, which is significantly more than my 10-pound max.

ETA: NOTE: THE FOLLOWING MEASUREMENTS REPRESENT THE DIFFERENCE BETWEEN HIS BORESIGHT LASER AND HIS RAIL-MOUNTED LASER, NOT THE ABSOLUTE MOVEMENT OF THE RAIL-MOUNTED LASER FROM THE "ZERO" POSITION.

SOLGW M76 15" : 0.542" at 6.3 yards = 8.21 MOA

Midwest Industries Night Fighter 13.5" : 0.546" at 6.3 yards = 8.27 MOA, with a 1.28 MOA shift/displacement of the rail.

Please correct my math if you see any mistakes.

ETA: I made a mistake. The above numbers are from his "vert downwards" data. His "vert upwards" data more closely resembles my test. Here are those numbers:

SOLGW M76 15" : 10.22 MOA
MI NF 13.5" : 6.02 MOA

That's a much bigger difference. It makes sense that the NF would hold up a little better since it is a shorter rail than the M76, assuming they are roughly equal in rigidity by design, but that's drastic. He did test lateral deflection as well, but so far those are fairly inconsistent (one rail flexed more with e-port down, the other vice versa, etc.). In both, the lowest lateral deflection was greater than the highest vertical deflection however. I think he's still developing and refining things, so don't consider these the absolute final answer. I look forward to his future tests and I think he will provide some great public info where it is sorely lacking.

Awesome; thanks for posting.
We’ll eventually, as a community, have to come up with some sort of standard for how this gets done and build a database.

What if you used a clam shell style upper block that puts the stress on the lugs where the takedown pins go through?  This would better mimic the way the upper is attached to the lower plus would be more repeatable of a test as well.

Absolutely. I really like P&S guy's way of using the MI rod, given that it doesn't have any slop, which I'm sure he would have mentioned if it did because he seems to be pretty meticulous about the whole thing. But, going forward, I would keep my way of placing both the load and rail-mounted laser at 10" every time and doing repeated load/unloads for each iteration.

The biggest challenge for most I think is having the ability to do this at more than a garage's distance. I got lucky and had a friend with an extremely convenient way to set this up at 50 yards but otherwise it could be a big hassle.


Unless I'm not understanding something, I wouldn't be able to test the same dimension as I did because it would require the block to be placed upside-down in the vise, and I didn't have access to a vise that can accomodate that. I don't see how this would be more repeatable, can you please explain that?

Thanks for posting OP.  I believe for the most part,  the small nuances of potential deflection discussed above do not matter so much when looking at the differences of deflection amongst the hand guards,  given they were all tested in exactly the same manner.

Often tho,  hand guards are are loaded 180 degrees opposite,  loading a bipod or resting on a barricade for example.  I would bet that all the tested hand guards would show less deflection if loaded as such.  Why?  As tested,  everything above the receiver threads is in tension….  Loaded 180 degrees opposite would place it compression,  so it would have the added rigidity of the pic rail faces pressing against each other vs freely pulling apart under tension.

This brings up another point,  someone previously mentioned URX 3/3.1 rails.  If they do not make tight contact with the receiver,  you basically forfeit a lot of rigidity from the get go.  You essentially have a tube that is the diameter of the receiver bbl extension threads. I bring this up because I literally just sent out (2hrs ago)  a URX 3.1 and stripped receiver to be timed on a lathe.

I was thinking it might have been a better way to secure the upper in the vice and distribute the load a little bit more in line with how the rifle would function.  But then I realized it would be pulling down on the rail instead of pushing up like most normal use cases would (bipods, grips, resting it on things for support, etc) so I don't think that would be a test that would show much other than how much the rail moves.

I'm assuming FCD Testing would also apply to Expo Arms Combat Rails? If so those are a hell of a nice deal for the money. (OEM by CMT, same barrel nut system, etc)

I wouldn't assume it to be 100% the same. The location of the cooling holes vs added MLOK slots (and how they line up around each other) is all going to play a role in how something flexes.

So much of this testing is tracking variations in flex using the upper receiver as the point of reference.

A rail mount IR laser is used to aim rounds fired, so if the rail flexes but the barrel tracks with it, it's not as big of a deal as if the barrel doesn't shift with the rail.  If the flex changes the way the barrel shoots and groups, that's relevant, but the barrel coming out of alignment with upper receiver mounted optics isn't relevant is the rail mounted laser what is being used to aim the rounds.

I tested some of this stuff myself earlier this year.

MI URR in a vise bolted down to a work bench, boresight laser in the chamber, rail mounted laser on the handguard, and I used a hanging scale (like you weigh fish with) to pull the handguard up/down/left/right.

What I noticed as that different combinations of upper receivers (I tested the VLTOR MUR, FCD URF-B, Anderson forged, and SOLGW forged uppers) and different combinations of rails (SOLGW M76, SOLGW EXO2, MI Nightfighter, FCD RHF) all exhibited different degrees of flex (IE: 10lbs of pull to the right might have moved the lasers more if the rail had an aluminum barrel nut vs a steel one) but what I watched was that the boresight laser tracked directly underneath the rail mounted laser, so the two stayed in unison nearly perfectly regardless of the rail/upper tested.

I tested my stuff at a distance of roughly 17' and despite the URR being cranked down in the vise, everything did "move" when it was pulled.  What I paid attention to was where the lasers were at peak force, and where they moved back to once the force was released.  I aimed the lasers on a grid of 1cm squares to track it.  Everything moved, but the barrel tracked evenly with the rail as it deflected.

Needless to say, I'm curious to see more testing on the subject.

That is interesting. I suppose the ultimate test would include actual shots fired with quantified loads applied to the rails.

You're saying that the lasers did NOT return to their original position after they were pulled? How much of an affect did the upper receiver play on the amount of flex/movement of the lasers?

No, so what I was saying was that even on a 1.5" thick plywood work bench, with a URR and all... if you turn your lasers on and start pulling on the handguards, you're going to move things.  Flex in the wood, slop of the upper on the URR... for example if I pulled on it and it tracked 4cm to the right from where it started, once the force was relaxed it would return back 3.5cm back towards its original position, just not precisely where it started before it was pulled on.  You just have to pay attention to where it relaxed back to, not how far it traveled from where you started, if that makes sense.

That does make sense.

I repeated each iteration at least 3 times to try and ensure nothing was shifting to any measurable amount. If the laser did not return to the "zero" or original unloaded position, I made sure everything was tight and restarted the test. When the laser would move back and forth between the "zero" and "loaded" marks 3 consecutive times at 150 feet, we were satisfied.

Good stuff here, with some results that may surprise...

Rigidity Matters: Testing Popular Rails to Find Zero Retention for IR Lasers

Nice.

So in summary:

Most Skinny rails suck: Aero Midwest Geisselle etc. Etc.
If you want the best rail systems for your IR laser or otherwise you will buy a monolithic or integrated upper such as LMT MRP, Seekins Precision, or Aero M4E1.

Got it.

No shooting though, and they were only looking at upper receiver to rail deflection... not bore to rail deflection, which is what matters for the purpose of maintaining the zero on a rail mounted laser.

If the barrel deflects equally with the handguard, it's not relevant where the upper receivers optic is pointing, provided you're aiming with the rail mounted device.

I agree that a "complete" test also includes bore position/deflection. I've crunched some numbers from "the P&S guy's" contributions and it's very illuminating. Those numbers use a laser, not shooting, but we can still see the relationship between the bore and the rail which is what really matters. I'll find the time to put it together into a coherent post if I get permission.