Can top-grade, critical components make a Frankengun a winner?

[sevenhelmet]

Interested to hear what makes one assembly good and another bad. All of the pieces pretty much either snap together, have a simple to tap in roll pin, or have a nut that you just torque to spec with a torque wrench. An AR is unbelievably simple to put together correctly. What are people doing to have a "poorly assembled" AR?

Parts selection is where the battle is won or lost most of the time, IME. Tolerance stacking can be an issue with sub-par parts, so ensuring parts are in spec is important (just knowing what the spec is and how to check can be a research project in itself for the uninitiated.) Barrel to upper receiver fit is very important for maintaining accuracy as the gun warms up. Too loose/tight, or uneven coatings will cause the POI to drift unpredictably. Gas system length, port sizing, and buffer weight all need to be balanced or the firearm may not function reliably or at all. Poor tooling can lead to gritty or loose movement of the trigger, safety, bolt release, CH, etc. All things that can contribute to various firearm failures or less-than-optimal accuracy.

I'm sure there is more, but those are just a few things I picked up along the way.

[WillBrink]

Or does it have to be a through-and-through Colt/BCM/DD/Knight's/etc.?

To me the most critical components are the barrel, BCG, and FCG. Talking about the basic weapon itself, not add-ons like red dots or other glass.

Case in point: perhaps my favorite carbine is built on an 80% lower, engraved as I wanted it to be. The bare upper is a Del-Ton. The barrel is a Colt SOCOM, the BCG is an LMT Enhanced one, the handguard is a KMR, and the FCG is a Geissele SD-C. I also put Knight's anti-rotation pins on it to prevent the ovalling of those holes over time.

My close second is built on a complete BCM ELW upper with KMR. It too has an LMT Enhanced BCG and Geissele SD-C. The lower is a no-name brand I can't even recall off the top of my head.

Now, I do have through-and-through name brand/top tier guns. I am of the opinion, however, that with top-grade critical components you can build a Frankengun that is perfectly suitable as a go-to piece. This assumes that the lower and upper are within specs, and so are the little pins and springs (other than the FCG itself) in the LPK.

Fire away (pun intended).........

Is there a cost saving to the approach or is it done for chits and giggles to more or less build to your liking? Problem for me mostly is I don't feel knowledgeable to decide what's critical and what's not per se, and hence would just default to brands known to be GTG. When I was dedicated 1911 shooter, I did take what's considered base quality guns and make some mods to get an improved gun, but don't think it really saved me any $ in the end from just starting with a better 1911.

[ABNAK]

Is there a cost saving to the approach or is it done for chits and giggles to more or less build to your liking?

Oh I wouldn't say it's a cost-saving thing, as the cost of the individual components can at times exceed just buying a complete weapon. But yeah, it is basically building it to your liking. For instance with the one that has the Colt SOCOM barrel on it......I had the barrel sitting in my parts cupboard new in wrap from Grant. Had it well over a year. Decided WTH, I'll put one together based off of this. Had my buddy cut down the FSB so I could put a KMR on it. Had an 80% lower already so bought a Geissele and LMT Enhanced BCG (damn are those things expensive!). I have two containers of parts so dug out a buffer tube, H2 buffer with spring, stock, and an Aimpoint T-2 that had been on another gun I sold previously. I picked up the Del-Ton upper from Midsouth Shooters Supply (about 15 minutes from my house) and viola! Oh, almost forgot the set of Troy BUIS I ordered. All in all (not counting the T-2) I have about $1500 in it. The T-2 pushes it up over $2K. So you see it is right up there with, or maybe a little more than, a DD.

[Lefty223]

Anywho, he told me one time in a discussion about modifications, a 1911 is a machine. If it is built to spec, it will run. It has to. It doesn't know any other way.

To take that argument further, I would think that if specs are, well, within spec, then, yes.

Not so, as besides being in spec, the greatest bearing on reliability and longevity is WHERE the individual parts are related to those specs. Case in point learned when I was in GE Aircraft Engines's machinist/engineering program.

Ford Motors made and sold a car under the Ford badge while Mazda sold the same car under their badge – both made to the same prints and same specs. The failure rate on the Ford-made transmissions, even though each part was made in spec, greatly exceeded the failure rate for the Mazda product, by a factor or 10 or more, IIRC. And again - all parts were within ‘spec’.

But upon dissection, each individual part within the Ford parts was all over the place within the design tolerance envelope, while for the Mazda parts, each and every part A was at X size, +/- the machine capability production tolerances, while Part B was at Y size, +/- , and so on. It was in effect, the meshing or stack-up of tolerances between the mating parts that had the largest bearing on the usefulness and ‘Quality’ if you will, of the final product.

So while I agree 100% that parts need to be made in spec, there is actually more to it than that on the design side. If the tolerances do not correctly identify which parts can be (or should be) on the high side or which can be on low side for example, is when variation can occur. Such variation can lead to factors, e.g., stress, wear, fatigue, etc., not correctly calculated in the overall design envelope (whether correctly captured in the tolerance scheme or not) causing failures to occur at an alarming rate.

Specs are one thing and are indicative of but one glaring example of ‘why?’ reverse engineering efforts can fail; like years ago when the Russians would steal our military products and reverse engineer them, only to have them fail miserably. Measuring a sample of one or just a sampling of parts gives no direction to the overall dimensional needs of the final design, in terms of what it will eventually mean to functionality, reliability or longevity.

[Falar]

Oh I wouldn't say it's a cost-saving thing, as the cost of the individual components can at times exceed just buying a complete weapon. But yeah, it is basically building it to your liking. For instance with the one that has the Colt SOCOM barrel on it......I had the barrel sitting in my parts cupboard new in wrap from Grant. Had it well over a year. Decided WTH, I'll put one together based off of this. Had my buddy cut down the FSB so I could put a KMR on it. Had an 80% lower already so bought a Geissele and LMT Enhanced BCG (damn are those things expensive!). I have two containers of parts so dug out a buffer tube, H2 buffer with spring, stock, and an Aimpoint T-2 that had been on another gun I sold previously. I picked up the Del-Ton upper from Midsouth Shooters Supply (about 15 minutes from my house) and viola! Oh, almost forgot the set of Troy BUIS I ordered. All in all (not counting the T-2) I have about $1500 in it. The T-2 pushes it up over $2K. So you see it is right up there with, or maybe a little more than, a DD.

This is why I'm so glad for the 6920OEM2.

There isn't a factory made gun that comes with my favorite forend, grip AND stock.

All my future rifles will be 6920OEMs where I finish the build or factory KAC guns.

[ABNAK]

This is why I'm so glad for the 6920OEM2.

There isn't a factory made gun that comes with my favorite forend, grip AND stock.

All my future rifles will be 6920OEMs where I finish the build or factory KAC guns.

You nailed it. I might even add barrel to the list (profile, gas system length), but yeah I customize everything.

[MistWolf]

Lefty, what you're talking about is control. To produce a good product, it must be within tolerance AND in control. If the parts are within tolerance but some on the high side, some on the low side and some in the middle of the tolerance range, your process isn't in control.

If a product is engineered correctly and the manufacturing process is right, there will be no problems with tolerance stack up so long as the parts are all within tolerance, even if they are out of control.

For example, the print calls out for 11 holes to be drilled 1 inch apart +/- .010". 11 holes 1 inch apart means the last hole will be 10 inches +/- .010" from the first. In order to hold that tolerance, the distance of each hole must be measured from the first hole and only the first hole and held to a tolerance of +/- .010" without tolerance stack up. If you were to measure the distance from the first hole to the second, then the distance from the second to the third and so on, the tolerance from the second hole is added to the third and the third to the fourth and so on. While the distance between each hole is within tolerance, by the time you get to the last hole, its location could be off by .10", ten times the allowable tolerance.

That's what "tolerance stack up" really means

[Stickman]

There is a difference between a frankengun, and a custom build.

[lysander]

Lefty, what you're talking about is control. To produce a good product, it must be within tolerance AND in control. If the parts are within tolerance but some on the high side, some on the low side and some in the middle of the tolerance range, your process isn't in control.

If a product is engineered correctly and the manufacturing process is right, there will be no problems with tolerance stack up so long as the parts are all within tolerance, even if they are out of control.

For example, the print calls out for 11 holes to be drilled 1 inch apart +/- .010". 11 holes 1 inch apart means the last hole will be 10 inches +/- .010" from the first. In order to hold that tolerance, the distance of each hole must be measured from the first hole and only the first hole and held to a tolerance of +/- .010" without tolerance stack up. If you were to measure the distance from the first hole to the second, then the distance from the second to the third and so on, the tolerance from the second hole is added to the third and the third to the fourth and so on. While the distance between each hole is within tolerance, by the time you get to the last hole, its location could be off by .10", ten times the allowable tolerance.

That's what "tolerance stack up" really means

BOLD - No, it does not mean that. (And unfortunately, many an engineer doesn't know this either, so you're not alone.)

If the drawing states that '11 holes to be drilled 1 inch apart +/- .010"' (assuming in a straight line), then the allowable tolerance on the distance from the first hole to the last hole is +/-.100" by definition. This is because you have defined the distance between the holes can be up to 1.010" for each.

If 11 holes have to be drilled and the distance from the first hole to the last hole cannot be more that .010" then what you wrote as a drawing is incorrect, and the drawing needs to changed to address the problem. That's a design problem, not a tolerance problem. If the distance from the first to last hole needs to be 10 inches +/-.010", then the drawing should state something like:

"distance from first to last hole 10 +/-.010", 9 holes between equally spaced to within +/-.010", or call out the distance from the first hole to each subsequent hole individually, 1.000 +/-.010", 2.000 +/- .010", 3.000 +/-.010", etc....

If your control is loose, and things wander around inside the tolerance, and this causes problems, then the design is bad. As stated, if everything is within tolerance, things should work correctly, If -

. . . the tolerances do not correctly identify which parts can be (or should be) on the high side or which can be on low side for example, is when variation can occur. Such variation can lead to factors, e.g., stress, wear, fatigue, etc., not correctly calculated in the overall design envelope (whether correctly captured in the tolerance scheme or not) causing failures to occur at an alarming rate.

Then the design is bad and should be changed.

On that note, what is the difference between these four length call-outs:

1) 1.634 +/-.006"
2) 1.628 +.012/-0"
3) 1.640 +0/-.012"
4) 1.628" - 1.640"


Hint, there is a big difference.

[MistWolf]

The part you bolded is the short version. The way most aviation prints are spec'd, the standard practice is that the tolerance of each hole is the distance from the first hole +/- and the distance from the first hole to the last hole is usually the most critical in the run. (Actually, the location of the first and last hole of the run are the most critical.) Many hole patterns on McDonnell-Douglas prints would spec the example I used as "fastener holes are 10 equal spaces" with a dimension line showing 10.00" from the first hole to the eleventh. There would be a note defining what the tolerance will be for a dimension spec'd to two places which, if I recall, is +/- .005"

How to avoid tolerance stacking is one of the first and most important things they taught us. As Lefty pointed out, there's more to manufacturing than just holding a part to spec. The process must be in control and good standard practices and training employees on what the standard practices are, is essential to controlling the process