On loose carrier key screws

**lysander** · 04-03-17, 12:02

Originally Posted by Ned Christiansen

Gunz, there should be a sample on your desk somewhere.

Might be a while before these are on my site but they are available.

These have a 9/64" socket. They will be available with the mil-spec 1/8" socket if there's a call for it. I kinda hope there's not as there is nothing to gain by it, but one mfgr testing it may be required to use it that way. It's my belief that the splines would not or might not technically violate "mispec". I guess that'll be up to the people who deal with that kind of thing on a weekly basis.

I also believe that the milspec screw calls out a 1/8" socket only because at the time of the system's design, the 1936 standards were still in use and the fastener industry had not yet updated cap screw standards to what would become known as "1960 Series", still the industry standard. One notable change in this revamping and improvement to fasters was a change in the 8-32 SHCS from a 1/8" socket to a 9/64 socket.

There is no such thing as a "Military Spec" 8-32-UNC-3A, alloy steel cap screw, undrilled head, with phosphate coating.

The closest thing to a military specification for a "8-32-UNC-3A, alloy steel cap screw, undrilled head, with phosphate coating" was MS35459, but back in 1962 the #6, #8, #10 and 1/4 inch sizes were deleted from the specifications, then in 1964 the whole specification was declared "inactive". The military currently uses standard cap screws that are made to Aerospace Industry Standard NAS1351 or NAS1352, a commercial standard. According to ASME B18.3, and NAS1351/NAS1352, 9/64 is the standard key size, as are the few remaining active military standard cap screws (cadmium plated or drilled CRES screws). The obsolete MS35259, MS35456 and MS35455 did call out a 1/8 inch socket.

That is why the military carrier key screws do not have an "MS" number, but an M16/M4 specific part number: 8448508.

So, while everyone assumes they are standard ASTM A574, 2520 pound tensile proof load, 8-32-UNC-3A, cap screws, or NAS1352-08-4 screws they aren't. Until you can see the requirements set out on DWG #8448508, you just don't know the actual proof load is supposed to be. They might be softer so as not to be so susceptible to cracking.....

Your splines might not cause the screw heads to exceed the dimensional tolerances of ASME B18.3, or NAS1351 or NAS1352, you might check, however, those castellations probably will.

**Ned Christiansen** · 04-03-17, 18:17

My screws definitely and intentionally violate the high limit on the head diameter. For the better of course, but yeah, that would maybe be an issue. I could make it go away if they wanted enough of them, and I'd put them right at the edge of the high limit. My design intent is that the diameter over the splines be no smaller than the smallest counterbore diameter. It will universally be larger and size itself to a perfect fit upon installation.

I have Dwg. No. 8448508. No direct ref there to strengths other than referring to some mil and ASTM standards which I suppose carry that info. I've been over them all in he past but not to the point where I have it all committed to memory. This Army drawing calls for the 1/8" socket so I take that to be a "mil spec". If this is no longer in force I'm unaware of it, but then I would not be the first to know.

I am unfamiliar with the term undrilled head, I take it to mean a socket that is completely cold formed as opposed to drilled and broached...?

**Ned Christiansen** · 04-03-17, 18:21

Originally Posted by mtdawg169

Very interesting idea! Ned, is there any way to accurately test the removal torque vs a standard milspec screw?

With a torque wrench. Not sure how a lab would do it.... I think with a torque wench, probably a really good one :-)

**tom12.7** · 04-03-17, 18:37

Just a normal type dial indicating wrench for this.

**mtdawg169** · 04-03-17, 19:30

Originally Posted by Ned Christiansen

With a torque wrench. Not sure how a lab would do it.... I think with a torque wench, probably a really good one :-)

Well, yeah, of course!

I just wasn't sure if there was a more consistent way to test it. And my torque wrenches only allow you to turn in one direction.

So, do you have any data on the difference in torque required to break the stake vs a regular milspec screw?

**Ned Christiansen** · 04-03-17, 23:41

Data, yes, keep in mind that everything depends upon everything else.

I've done just few tests on the "mil spec" screws for a few reasons, one, I figured they were good and they are, but they are seldom seen except on the brands that do staking right. The other reason is, what I'm telling my screw customers:

"Buying American-made socket head cap screws is a reasonable assurance that the required standards are being adhered to. With some off-shore brands this has proved to be hit-or-miss and in some cases, actual counterfeit fasteners with falsified traceability and quality documentation have been discovered in customs, or when they failed in their application. Buying screws that have the appearance / outward features of the mil-spec screw may carry no assurance of standards compliance unless the actual source of the screws is known and trusted."

What I'm saying is, if we don't get our mil-spec screws from a known source, it's not really possible to say they are for sure the real thing. When you get commercial screws with a brand name on them, you know who made them and they have a vested interest in making them right. Otherwise all we would generally have to go on would be "haz 1/8" socket" but that's about it. I'm not saying I know there are fakes, just saying that, well, I don't know, and I don't think anyone does who is not getting them from a source that can provide documentation.

I will say that in 6-plus months of talking to not just screw companies but also sales reps who rep for cold-heading manufacturers, when I got to the description of what I wanted, a common reply was, "Oh we can get that done for you at one of our off-shore houses, you will be very pleased with the pricing!" That's when I knew I had to deal direct with an existing screw mfgr. of long standing good repute. It goes without saying that they would be a US company!

So Mil-specish screws, the few tests I've done on them, with good stakes, they come in at avg. 110 in pounds. Again the install spec is 50-58 inch pounds and breaking torque spec is 50-100 inch-pounds. So that's pretty good.

Everything depends on everything else part deux: Carrier key tolerances on the rebated part where the staking goes, are wide open. The width range is .022 and the height range is .012 without even considering the tolerance on the carrier diameter and the height of the surface the key seals on. So if you get a narrow, low one, the stake will be less of a stake, pure and simple. Hardness plays a role too-- I've seen some that were file hard for some reason. I suppose there are those that are too soft, but the three or four I've Rockwell tested happened to all be within the 26-32 range.

The ubiquitous YFS screws have not fared well in my testing, as expected. They averaged 76-- in spec, but these are by far the ones I most see loose. That is also partly because there is some correlation between the mfgrs that use YFS and staking thoroughness too. But the YFS has bogus knurling and a rounded head. Maybe an OK screw but not as a carrier key screw. I thought it was interesting that a completely un-knurled 8-32X 1/4 cap screw, with splines, soundly beat the YFS screws at 90 and 96 inch-pounds breaking torque!

Data on the OCKS was a very consistent 132 inch pounds breaking torque, that is with the screws installed at the very bottom of the 50-58 inch-pound installation torque range. I will be conducting about 160 more tests in the coming weeks.

The bottom line though is the breaking torque spec. As long as my carrier keys are not coming loose, I don't care if they key is wide, narrow, low, high, or a few points too hard or soft. Splining the screw heads reduces the criticality of the other factors.

**ScottsBad** · 04-04-17, 00:14

I just want to say thanks for coming up with a great design. I would be much more likely to choose a BCG that uses these screws and your staking system over the ones I see on most BCGs. Every high end BCG should have its gas key fastened with your screws and staking system. It proves that big quality improvements don't have to cost a lot.

BTW - They look cool too.