Short stroking caused by BCG

[Hohn]

If the design is properly toleranced on the drawings (and the M16 is), and the parts made to the tolerances specified, then stacking of tolerance will have no effect on operation. The only time "tolerance stacking" causes problems is if the tolerance are not designed correctly, or the parts are not made correctly.

That's the whole point of interchangeability. A rifle made from parts all at one end of the tolerance band should work just as well as ones made at nominal, or a random mix.

Thank you for posting this, I was about to post similar statements. At work (engine engineering) we spend a lot of effort to ensure that the tolerances are specified such that the chances of all "in spec" parts creating an issue are a statistical impossibility. We hold our suppliers to >1.33cPk and the idea that that you just have to hope the AR parts you choose all happen to work out is kind of absurd. If your parts are made to spec by a capable supplier, THEY WILL WORK.

"Tolerance stacking" as it pertains to AR parts isn't a thing. If you have a gun that doesn't work when it comes together (say, feed ramps in barrel extension overhang the ramps in the upper), it's not due to a stacking of tolerances-- it's due to a single part being off. The probability of having all "in spec" parts not work is vanishingly tiny such that the armorer for the biggest organization you know of still hasn't seen it.

And if that last part is not true, then something in the tolerancing scheme of the parts is not properly performed and the DVA (dimensional variation analysis) need to be repeated and corrected. Most commonly IME the problem is that the supplier is measuring the parts incorrectly or interpreting the drawing incorrectly.

[Averageman]

If the design is properly toleranced on the drawings (and the M16 is), and the parts made to the tolerances specified, then stacking of tolerance will have no effect on operation. The only time "tolerance stacking" causes problems is if the tolerance are not designed correctly, or the parts are not made correctly.

That's the whole point of interchangeability. A rifle made from parts all at one end of the tolerance band should work just as well as ones made at nominal, or a random mix.

So, if my cam pin is within tolerances, but just a smidge on the edge of tolerances and it goes in to my bolt, which again, also is within tolerances, but it passes all QC tests but again is just a hair off, then I put in a firing pin that,....You see where I am going, right?

[lysander]

So, if my cam pin is within tolerances, but just a smidge on the edge of tolerances and it goes in to my bolt, which again, also is within tolerances, but it passes all QC tests but again is just a hair off, then I put in a firing pin that,....You see where I am going, right?

No, honestly, I do not.

If your cam pin diameter is 0.3100" to 0.3105", and the hole in the bolt is 0.3120" to 0.3135", they will work correctly. You will have a 0.0015 to 0.0035" clearance.
If the hole in your cam pin is 0.164 to 0.167", and your firing pin diameter is 0.153" to 0.154", they will work correctly. You will have a 0.006" to 0.010" clearance.
I am not going to go through the three thousand or so interfaces in an AR just to show you that all the tolerances work out properly, but rest assured, someone, or a group of someones, has, and they do.

If you are a smidge off the tolerances, you are not within the design requirements, and all bets are off.

So, if you are buying parts that "passes all QC tests but again is just a hair off", you need to reassess who you are buying parts from. The whole purpose of Quality Control (QC) inspections are to ensure the parts are made to within the specified tolerance zone. If they "pass QC" but are a "smidge off", somebody is failing at their job.

[Hohn]

So, if my cam pin is within tolerances, but just a smidge on the edge of tolerances and it goes in to my bolt, which again, also is within tolerances, but it passes all QC tests but again is just a hair off, then I put in a firing pin that,....You see where I am going, right?

QC is not just to ensure everything is “in spec” but that the average of the parts is centered within the spec. And it is assessing the variation within spec to make statistical calculations about the probability of something falling outside.

So you see, tolerances just don’t “stack” all at the maximum amount and all in the same direction. A design engineer sets the tolerances based on how capable a supplier as as well as what is required to meet design intent. For example, if you have a bolt pattern, you have to account not just for the variation in the true position of each threaded hole, but the size and true position of each hole in the through flange (what you are bolting to the hole.) Depending in the number of steps in the stack up, the analysis of the tolerances can become very sophisticated, quickly exceeding what you can do in excel. You need more statistical power from something like CETOL.

Read the Wikipedia article on six sigma and see if it makes sense why you don’t have all the tolerances stack up in the same direction and at maximum value.

[MistWolf]

No, honestly, I do not.

If your cam pin diameter is 0.3100" to 0.3105", and the hole in the bolt is 0.3120" to 0.3135", they will work correctly. You will have a 0.0015 to 0.0035" clearance.
If the hole in your cam pin is 0.164 to 0.167", and your firing pin diameter is 0.153" to 0.154", they will work correctly. You will have a 0.006" to 0.010" clearance.
I am not going to go through the three thousand or so interfaces in an AR just to show you that all the tolerances work out properly, but rest assured, someone, or a group of someones, has, and they do.

If you are a smidge off the tolerances, you are not within the design requirements, and all bets are off.

So, if you are buying parts that "passes all QC tests but again is just a hair off", you need to reassess who you are buying parts from. The whole purpose of Quality Control (QC) inspections are to ensure the parts are made to within the specified tolerance zone. If they "pass QC" but are a "smidge off", somebody is failing at their job.

Lysander is spot on.

[Todd.K]

Lysander is spot on.

Yes, and saved me from going off on a rant.

[Dr. Bullseye]

We are finally getting to my pet gripe and a recurring theme on this board: "Why won't all my Gucci parts work together?" I am really enjoying this.

[Clint]

Lysander and Hohn have nailed the engineering aspect.


In this case of a BCG causing short cycling, the main factors at play are gas circuit leakage vs gas port size.

There are six critical interfaces between twelve part features in the AR gas circuit.

Excess gas leakage comes from one or more of these part features being out of spec and failing to create a "good enough" gas seal.


A gas port can be enlarged over "spec" to compensate for excess gas leakage and that particular combo can actually work.

However, in reality, this is making a second part out of spec to compensate for the first part being out of spec and breaks part interchangeability and can cause future diagnostic issues.

Essentially, this is a "hack" and NOT the correct way to assemble a weapon you can trust.


On the cheap end of the industry, there is a bit of an unholy trifecta that tends to feed on itself involving under powered ammo, leaky gas blocks/gas tubes/bolt carrier groups and over sized gas ports.

Companies interested in doing it right, expect all critical parts to be in spec and the system as a whole to perform at a high level.

[Todd.K]

We are finally getting to my pet gripe and a recurring theme on this board: "Why won't all my Gucci parts work together?" I am really enjoying this.

Because some don’t understand the engineering, some don’t know the tolerances of the part it connects to, and some are manufactured to a “better fit" rather than held to correct tolerance.

[T2C]

I know many do not care for oversize gas ports. I have oversize gas ports on 3 carbines and they run flawlessly on any and all ammunition including steel case Wolf ammunition. I've replaced bolts, bolt carrier groups, gas tubes, etc. purchased from several vendors and I have had zero issues.

Before the ammunition supply started to dry up the carbines were run hard weekly, they could have been maintained better and they continued to run.