M249 SAW: U.S. Ordnance Full Mil-Spec Test

M249 SAW: U.S. Ordnance Full Mil-Spec Test


By Dan Shea

For many years I’ve been involved in performing and chronicling MIL-SPEC testing for firearms. I’ve also been involved in many M249 programs, supplying DoE, DoS, foreign governments and OGAs with the M249. When Steve Helzer told me that US Ord had the production line ready to go and was about to conduct the full MIL-DTL test, I had to be there and get the story….

Manufacturing the M249 machine gun requires a number of different skill sets, and there have been many attempts by companies to go into production. Obviously, the original manufacturer has been successful; FN Herstal as the Minimi, then FNMI for the U.S. government contracts. Besides U.S. Ordnance, there is one other who has had some success, but most crash and burn, especially on the receiver. Fixturing and properly welding the sheet metal receiver has been the bane of many a dream, but U.S. Ordnance has all the needed skill sets: Manufacturing and machining capability, Engineering design capability, and perhaps the most important of all- being a firearms manufacturer with vast and varied experience. If you’re not a “gun maker” you’ll be missing a lot of detail on any firearms project, and fighting through TDPs (Technical Data Packages) that never include all the details needed to make a firearm.

The receiver components here are being prepped to go to the various welding fixtures. Due to the special innovations, these fixtures are proprietary to U.S. Ordnance and not shown. The receiver channel is now considered to be a Title I firearm, a semi-automatic. Once the components are installed, it becomes Title II, a machine gun.

It’s always a great pleasure to visit U.S. Ordnance, they know what they’re doing, and make excellent machine guns. The M249 test visit to the plant in July 2023, was business as usual. We were set up like clockwork to go through the MIL-SPEC, so this article will follow it, as well.

MIL-DTL-70446C is a very boring document. AMENDMENT 1 adds to the “thrilling” reading. The latest document we had available was the 27 March 2009 version, which superseded all others. There are several round counts involved in the test, the longest of which is barrel wear.



This purpose of this MIL-SPEC test is getting the U.S. Ordnance M249 to the contract “First Article.” First Article doesn’t mean the first part or assembly off a production line- it means that a responsible person from the buyer takes an agreed number of the items from the production run, randomly, and puts them through a specified series of tests and measurements. Most contracts allow 270 days to reach First Article, as is the case with the other firearm tested during our program, the US Ord MK19 Mod 3 Grenade Machine Gun, which U.S. Ordnance has been the U.S. government supplier since 2016 and is now the NSN contractor. In this M249 test, it’s exactly what we’re doing on this MIL-SPEC test. In the case of the M249, the First Article language is as follows and, after meeting the full tests done here to prove it out, the customer will repeat the examination as follows, if required. But the contracting officer can forgo further testing if satisfied:

4.2.1 First article quantity. The first article sample shall be representative of the manufacturing methods and processes to be used for quantity production. The first articles shall consist of the quantities specified in Table IV unless otherwise specified.

TABLE IV. First article quantity:

Machine Gun, 5.56mm, M249 (9348199): 10

All components (except unmodified commercial parts): 5

All subassemblies: 5

All assemblies: 5

The testing starts out very simply; does the safety work properly. Note that the tech spec is very specific on this, and yes, red showed when in fire, clicks were audible, and we performed extra tests regarding the springs for proper pressure. The sear and disconnector are visually available in the picture. It’s difficult to quantify 3.4’s sear engagement test in a photo unless there was a cutaway receiver (which defeats the point of testing a real one) but we were able to visually inspect full sear contact, and conducted some buttstock kinetic tests while the sear was engaged to ensure it was in solid contact.

3.3 Small arms safety. The small arms safety, drawing 9348364, shall move manually without binding between the “safe” and “fire” positions and shall remain in the set position under spring pressure until reset. The trigger mechanism shall not function when the safety pin is set in the “safe” position (to the right). The trigger mechanism shall function when the safety is set in the “fire” position (to the left). When moving the safety between the two positions, there shall be an audible and tactical click. When in the “fire” position, the red warning ring shall be displayed.

3.4 Sear. When assembled into the weapon the trigger assembly, drawing 9348354, shall be capable of full engagement with the sear engagement notches of the operating rod assembly, drawing 9348408, and of holding the piston assembly, drawing 9348405, in the rearward, cocked, position.

Trigger pull is tested in a standard manner. One thing is for sure, a machine gun should never have a “match” trigger, and long experience has taught procurement people and trainers (as well as end users,) that you want a good, solid pull with enough weight required so that the flowing adrenaline prevalent in combat ensures the operator needs to positively engage the trigger. We used a Lyman Professional Electronic Trigger Pull Gauge, and tested in two manners: first, with the trigger group in a vise. This showed a pull of 7 pounds, 2 ounces, and then the more important pull measurement, when assembled and the bolt carrier is engaged in the sear notch and ready to fire to see what the trigger break was. This was a 9-pound, 1.2-ounce pull.

Trigger pull gauges measure an actual weight; gravity related. We’re not talking about foot-pounds, a unit of energy, we’re testing actual pounds as if you were pickup up X pounds with your finger, on the earth, against gravity. It’s “pound-force” not “foot-pounds.” A foot-pound is applying the force of one pound-force linearly, a distance of one foot. An energy measurement. We use this in muzzle energy calculations, and in torque in mechanics. Foot-pounds correspond to joules, not to newtons.

The specification in 3.5 is in “newtons,” a system that is not normally used in the U.S. firearms community but is standard in physics. Generally called a “newton-meter.” One newton is the force necessary to give a mass of 1 kilogram an acceleration of 1-meter-per-second, per second. Yes, it’s a unit of energy but doesn’t correspond to foot-pounds. In contrast, our test instruments are in pound-force as noted above. 1 newton = 0.22481 of pound-force, so in the specification below, 35 N = 7.86835 lbs on the low end, and 70 N = 15.7367 lbs on the upper end. Our test guns were well within the ranges under load.

3.5 Trigger pull. The force on the trigger required to release the operating group from the sear with the bolt in the open position shall not be less than 35.0 Newtons and shall not be greater than 70.0 Newtons.


The spec calls for the completed firearm to be tested with an M197 High Pressure Test round. The M197 uses a 56-grain (3.63 gram) projectile, with SR7641 powder, which is an extruded/flake powder. This powder produces a chamber pressure of 70,000 PSI (pounds per square inch) as opposed to the M193 Ball with 52,000 PSI, or the M855 with 55,000 PSI. You can recognize the M197, it has a silvered (stannic-stained) or nickel-plated cartridge case. This cartridge is used to test M16s also, and should not be fired in normal weapons use.

3.6 Proof firing. Each main and assigned barrel (see 6.7) and bolt assembly shall be capable of withstanding the firing of a Government standard 5.56mm M197 high pressure test cartridge in accordance with MIL-C-46936 or approved equivalent. After firing, each barrel and bolt assembly shall be subjected to visual and magnetic particle inspections to determine that these components are free from cracks, seams and/or other defects.

After the proof testing called for in 3.6, Magnetic Particle Inspection (MPI) is performed. Generally, this is called “magnafluxing” after the manufacturer usually associated with the process. It’s a non-destructive process, and in the case of U.S. Ordnance, they use a “wet” system which is ideal for production needs. In this wet system, a petroleum based “suspension vehicle,” basically a light oil, has a specified quantity of fluorescent magnetic particles added to it (suspended in it). The formula is pretty standard in this use; Magnaglo 14A particles in Carrier II liquid.

This shows the M249 barrel to be tested, supported by a copper coated rod from end to end, and on each end of the machine there is a woven copper screen (copper is non-magnetic). The system is magnetized while the operator applies the wet liquid to it and continues until the magnetization is 100%.

He then uses an ultra-violet light source wand like this Magnaflux EV6000 in a darkened area, to make the magnetic particles visible.

As seen in the MPI enclosed area, the fluorescent magnetic particles on the surface present a unified appearance. If there is a structural break in the metal (this only works on ferromagnetic materials like iron, nickel, cobalt and alloys that feature them prominently, they must be capable of holding a magnetic field), it will appear as a darker area, an anomaly in the color. Thus, fracture lines, breaks, stress areas, will be indicated to the operator’s visual inspection. These are recorded.

The Carrier II is tested for viscosity and color, and in the left jar, the solution in use is tested for color, suspension, and the particle level in the bottom. This must be done periodically to ensure a quality test. When finished, the workpiece has to be demagnetized.

3.6 M249 Proof Marking

As called for in the MIL-SPEC, when certain parts are required to be proofed, or magnetic particle inspected, the markings “PM” are applied in a non-destructive manner. This bolt has been proofed with M197 firing and then magnafluxed.

For headspace testing, it’s important to gauge the firearm with the actual 5.56x45mm NATO gauges, not .223. “Headspace” is the distance from the face of the bolt to the point where the cartridge is seated in the chamber- in discussing bottleneck cartridges, it’s from the face of the bolt, to the datum line on the shoulder that is considered the point the cartridge is seated. For rimmed cartridges, it’s the distance from the face of the bolt to the front face of the chamber wall (barrel) and the gauge will be very short with a thick rim. The point is, headspace means different things in different calibers and styles of cases. A chamber for a machine gun will usually be a bit looser in tolerance than a rifle, and as close as .223 Remington and 5.56x45mm NATO are, the NATO chamber is a bit longer out of necessity, the throat is longer thus, the leade would be 0.0125 inches longer. (It’s safe to shoot .223 in the NATO gun, but not always the opposite.) Without getting too far out, there are other differences. Just be sure to use the M249 5.56x45mm NATO gauges on the M249, and not the ones for the British SA80 or the Swedish AK5, or the French FAMAS, all of which are slightly different even though they are all “NATO.” Example in just the U.S.: the civilian .223 “Go” gauge is 1.4640-inch while the M16 “Go” gauge is 1.4646-8-inch and the M249 military “Go” gauge is 1.4940-9-inch. For the “No-Go,” civilian .223 is 1.4670-inch, the M16 is 1.4704-06-inch and the M249 is, well, 1.4980-2-inch.

Just make sure you have the correct gauges.

3.7 Headspace. Headspace in the assembled weapon shall be in accordance with drawing 9348200.

Testing firing pin protrusion is critical to performance as is the next test for indentation. Protrusion is a different subject. Generally, for the size primer in 5.56x45mm NATO ammunition (small rifle) the specification is .030-inch, minimum protrusion, and .055-inch maximum protrusion. U.S. Ordnance has a very clever method of testing this using a milled down barrel extension section, early M249 carry handle, and a special gauge holder that threads into the assembly they created.

The firing pin protrusion test fixture in place in the threaded adapter.

The fixture as prepared for firing pin protrusion testing.

The fixture and gauge in position, with firing pin forward and the gauge measuring. The FP was well within the spec on each M249 we tested.

3.8 Firing pin protrusion. The firing pin protrusion, in the assembled weapon, shall be in accordance with drawing 9348200.

Firing pin indentation is a different type of measure, not just protrusion, but indicates there being enough inertia in the firing pin release to crush the anvil and activate the primer. There are several other indicators in this test, but the specification calls for measuring the original surface of the copper crush cylinder before striking, then the depth of the firing pin indentation compared to that, as well as compared to the potential “pillowing” of the copper around the indentation. U.S. Ordnance made their own gauges as shown in the photos of the testing.

Copper crush cylinders are specified to be 99.9% copper; if silver is added it is counted as copper percentage. They are made from an electrolytic copper rod, and yes, electrical resistance counts in this testing, as an indicator of the rod’s purity. The crush cylinder is specified as 0.399-to-0.4010 inches long and a diameter of 0.2245-to-0.2265 inches; the well in the test fixture must match tightly.

The crush cylinders were made in huge quantities in the WWI and WWII period, and government testing has always drawn from this supply. In 2018, the supplies were dwindling, and new criteria for the cylinders was being worked in – and there are civilian sources for the proper cylinders, such as SAAMI.

U.S. Ordnance has made an excellent fixture for holding the crush cylinders properly. Starting with an M249 5.56x45mm “Go” gauge, the toolroom milled out a spec dimension well for the cylinder. This cylinder has been tested, note the indentation.

Firing pin indent gauge inserted in the barrel chamber with fresh crush cylinder.

After testing, the crush cylinder showing the firing pin indent.

The fixture and crush cylinder are returned to the Mititoyo dial indicator depth gauge in a custom fixture, and measurements are made. Section 3.9 calls for an indent of not less than 0.51mm, which is 0.020-inch. Each test of the U.S. Ordnance M249 produced an indent of 0.32-inch, a solid, proper strike.

3.9 Firing pin indent. The firing pin indent, when utilizing copper pressure cylinders in accordance with MIL-C-20109, shall not be less than 0.51mm.

Dispersion and rate-of-fire is part and parcel to a number of related tests. On the table here, we’ve prepped for live fire dispersion, magazine test, and belt pull test, with our two barrels – the main and the assigned (spare) to the gun. These barrels are part of the long-term barrel life test later down, and we kept them in the process.

Firing for record. Here, we fired 18 rounds to get a longer string (the spec calls for 10 rounds). We did that several times and were well within spec and wanted a longer burst. The rate of fire is recorded as 702 rpm, which in section 3.13 is right in the target range – low end (preferred by most SAW gunners). We had one “flyer” and it was still within 20cm, and the group was a 2.09-inch mean radius; which is a 7.3cm square, far exceeding the 20cm square in section 3.10.

Both the main, and assigned barrel were tested with a government supplied 30 round M16 magazine, and we witnessed no malfunctions. As any M249 or Minimi variant gunner knows, the magazine feed can be finicky or unreliable in some manufacturer’s products. This is basically due to the dual feeding lobes needed on the M249 bolt, to feed from either above in the feed tray, or, push through a magazine feed lip set on the lower left. If this is not absolutely perfect in presentation (this is a welding issue for the fixturing of the magazine well) the magazine feed is usually what suffers. U.S. Ordnance was cognizant of this, and we fired many more magazines than required, with no malfunctions.

Magazine test target, no malfunctions, all rounds in the square.

3.10 Dispersion and targeting. When fired at a target located 50 meters from the muzzle, the machine gun with its main and assigned barrel shall meet the following criteria. The weapon must be placed in a government approved mount. Nine out of ten rounds fired in a single burst shall realize a figure of merit H+L (height + length) not exceeding 33cm. No keyholing (defined in 6.7.5) shall be permitted. The mean point of impact of 9 rounds of a 10-round burst shall be within a 20cm by 20cm square. The center of this square shall be 5cm above the point of aim.

3.12 Thirty round magazine. The machine gun shall be capable of functioning without failure (see 6.4) or unserviceable parts (see 6.7.4) with its main and assigned barrel using a government furnished 30 round magazine (see 6.8).

3.13 Cyclic rate of fire. The cyclic rate of fire for each M249 machine gun with its main and assigned barrel shall be between 700 and 850 rounds per minute. This requirement shall be met without failure (see 6.4) or unserviceable parts (see 6.7.4).

In section 3.11, belt pull is specified as pulling 2.86kg (6.305 pounds) while feeding with no malfunction. This test used to be done by mounting the gun very high, and hanging the belt freely at full length. After many studies, it was determined that a more effective pull test to ensure the machine gun transfers the proper amount of energy from moving operating group, to the feed shuttle mechanism, is to use a flat shelf, free roller, X amount of linked rounds and a hanging weight as shown in the picture. The weight is slightly higher than 2.86kg, on purpose, to show the machine gun exceeds the spec.

3.11 Belt pull. The machine gun shall be capable of functioning without failure (see 6.4) or unserviceable parts (see 6.7.4) while pulling 2.86 kilograms (the equivalent of a 200 round free hanging belt).

Section 3.14 of the MIL-SPEC calls for an interchangeability of parts test. The testing quantities can get complex based on monthly productions, IDIQ numbers, first article requests, etc. – that’s a book in itself. In this case, we chose a basic method from the specifications for 10 guns taken randomly from the production line. These were disassembled into components; first as basic assemblies and mixed up, then assembled with random assemblies and function tested and gauged. All passed. Second, the parts were broken down to a lower level and reassembled, then the firearms assembled, function tested, and gauged. All passed. This is a prove-out of the requirement in section 3.14.

Next, some parts from other manufacturers, from older models, common parts with the MK46 machine gun, etc., were used in assembling U.S. Ordnance M249s. All exchanges were successful, and in 3.14.2, it all must be done on the main assemblies, with no tools. A small hammer is required to persuade some new pins, but that is acceptable.

3.14 Interchangeability. Unless otherwise specified on the drawings, all component parts or inseparable subassemblies shall be interchangeable.

3.14.1 Interplant. Unless otherwise specified on the drawing, all component parts or inseparable subassemblies shall be interchangeable with weapons representing production from each of the previous manufacturer(s).

3.14.2 Tooling. No tools shall be required to assemble/disassemble the following assemblies from the M249 receiver: Bipod, drawing 13022945, Trigger Assembly, drawing 9348354, Buttstock, drawing 12556935, Gas Cylinder, drawing 9348345, Barrel Assembly, drawing 12011986, and Ejector, drawing 9348223. The Bolt Assembly, drawing 9348412, Slide Assembly, drawing 9348391, and Piston Assembly, drawing 9348405, shall not require any tools to assemble/disassemble from each other. The Heat Shield, drawing 12540405, shall not require any tools to be assembled/disassembled from the Barrel Assembly, drawing 12011986.


U.S. Ordnance has developed a water firing system that is environmentally protective. This is a 155mm howitzer barrel that has been altered so that water runs through it under pressure, and the machine gun fires into the water. At the other end, the water cycles the spent bullets into a container with a conveyer system. The water recycles through the system many times, and is properly filtered.

The conveyor brings the spent projectiles, many of which have the jackets stripped in the water-firing process and dumps them into 55 gallon drums.

In section 3.15, Endurance, the mean rounds between failures, other failures are all defined. During the testing, the records showed well within the spec. I personally didn’t observe any failure that required more than charging the gun or sweeping the feed tray. Fast, normal machine gunner procedures, well within spec.

3.15 Endurance. When subjected to 10,000 rounds of firing using the 200-round magazine, each endurance weapon shall exhibit no more than 4 failures attributable to the machine gun. Of the 4 failures a maximum of 2 failures are allowed which take more than 10 seconds but less than 10 minutes to clear. All remaining failures must be immediately clearable within 10 seconds. No failures which require more than 10 minutes to clear and no instances of uncontrolled firing are allowed. When firing with the government furnished 30 round magazine (see 6.8), 3 failures (attributable to the machine gun), which take less than 10 seconds to clear are allowed in each endurance weapon. Only l failure (attributable to the machine gun) is allowed which takes more than 10 seconds but less than 10 minutes to clear. All incidents shall be recorded. Any incidents not chargeable to the weapon shall be substantiated and reported. No unserviceable parts are allowed during the endurance test. See Table I for endurance summary.

Long barrel life is a priority for procurement people – this is not so much about combat use, as it’s a concern for how much the training barrels are used in training. In order to budget properly, planners need to know how often they need to trade out barrels and other components. If you read the spec, you’ll see that the normal MIL-SPEC test for M249 is a 10,000-round test, which we did. Barrels must go to 15,000 rounds, so another 5,000 for the main barrel and assigned (spare) barrel are required. Top in the photo, is a new barrel, all barrels gauged identically when new. If you look at the second barrel, that is the main barrel in the test and it is at 9900 rounds. The bottom barrel is the assigned barrel at 9700 rounds. M249MG Barrel Erosion Gauge 935009. The pocket ruler is there just to add visual context, as you can see, these barrels have a long way to go before they No-Go.

4.18 Barrel life. The 15,000 rounds barrel life test shall be conducted simultaneously with the 10,000 round endurance test (and 50,000 reliability test when applicable). If only the endurance test is being performed, an additional 5000 rounds shall be fired on the barrel for a total of 15,000. The ammunition, firing schedule, and data recording for the additional 5000 rounds shall be identical to those specified for the endurance test except that use of the government furnished 30 round magazines (see 6.8) is excluded. Likewise, the endurance maintenance schedule shall be continued. Acceptability inspection of the barrel life test shall be determined by magnetic particle inspection in accordance with ASTM E 1444 of the barrel for cracks, projectile velocity, and yaw measurement at 15,000 rounds. Magnetic particle inspection and velocity and yaw measurements (using 20 rds of M855 ammunition) shall be taken in accordance with the procedures specified for the endurance test.

U.S. Ordnance is in full production now on the M249 machine gun. They’ve passed a full MIL-SPEC test regimen, they are in process of delivering hundreds of M249 machine guns to government customers in various countries. The M249 lives on, and many countries (as the United Kingdom just announced) are sticking with the 5.56x45mm cartridge until the mid-2030s.

Contacts for U.S. Ordnance:




training: training@usord.com