HPT and MPI: still viable and necessary or outdated bureaucracy?

[orionz06]

Simple question for any mfg's that may be on here..when your parts are produced, are they not CMM'ed, or are they just batch tested? Tolerance stacking in certain military aircraft is no longer accepted as they were in the past--I would think in small weapon parts such as the bolt it would be fairly simple and inexpensive to verify the dimensions are in spec?

A few sets of gauges could work for the bolt in particular.

[rob_s]

Dimension
Materials
Treatments & finishes
Assembly

Those are the parts of any specification, military or otherwise. I am unclear on why there seems to be a discussion wherein material testing and dimensional verification are being looked at as either/or. One does not preclude the other.

[Ryan S.]

Its been a coupe years now, but I was told by an manufactor/contractor employee before that pretty much nothing ever fails MPI testing, as in they had worked for several years and a bolt had never failed MPI testing. I'm running MPI tested FN Bolts in my 4 ARs, but they were cheaper than most commercial untested bolts to purchase.

[dpaqu]

I'll preface my post by saying I have little AR exsperience but spent a summer working in a machine shop and we magnafluxed (gearhead speak for MPI) parts if the the customer requested it. I think it's important to point out that the MPI is only as good as the effort put into it by the guy doing it. Doing a bolt must be a bitch becouse it's so small with all those little machining cuts.

With the money guys spend on there guns I would think a "premium" bolt with fancy hard to machine steel thats been cryo treated and MPI would sell well.

[Todd.K]

It was mentioned that H&K does not do this, but aren't 416's noted for breaking bolts? I know there are a ton of broken Del Ton bolts floating around the net these days.

You are missing a HUGE, very important piece of info when you make a leap like above.

How do you know if HPT/MPI would have failed those bolts? You assume the material and manufacture to be equal so that you can pin lack of HPT/MPI on the failures.

The truth is that the low cost manufacturers may use cheaper materials, may not heat treat as precisely, and do poor shot peening (if at all). I would take the non HPT/MPI C158 bolt (of know quality) over a 8620 bolt that passed HPT/MPI.

[rob_s]

8620 bolt that passed HPT/MPI.

As it has been explained to me, the failure rate of bolts and barrels not made from their representative correct materials and test properly in accordance with the military specification(s) would be so high as to be actually cheaper to simply start with the correct material anyway.

This is why it is so important for those who do perform HPT and MPI to be clear about their methodologies, proof load used, rejection criteria, etc.

Some of this (in addition to shot peen method and other expanded datapoints) is addressed in the new Chart, but not all.

[orionz06]

You are missing a HUGE, very important piece of info when you make a leap like above.

How do you know if HPT/MPI would have failed those bolts? You assume the material and manufacture to be equal so that you can pin lack of HPT/MPI on the failures.

The truth is that the low cost manufacturers may use cheaper materials, may not heat treat as precisely, and do poor shot peening (if at all). I would take the non HPT/MPI C158 bolt (of know quality) over a 8620 bolt that passed HPT/MPI.

In the instance of the H&K there are other things going on as well, but I would consider it safe to assume they are using a good material of known quality. Disagree? This is also just running off of a comment made by Rsilvers. If the reports of 416 bolt failures with low round counts are factual, what would be the first step to fixing it? Finding where the problem is and that would involve some testing as well. (I would personally pin the failures on a few other things, but HPT/MPI goes an awful long way to show bolt issues)

The Del-Ton failures are more than likely material alone.

And I would trust the same non HPT/MPI C158 bolt (of known quality) over a 8620 bolt that passed HPT/MPI any day, but does not mean the C158 bolt is good. Using the right materials within the dimensional tolerances of the design and proof tested tells you everything you should know. Removing any one of those opens you up for trouble.

Now what about an alternate means to do so? Batch testing goes further than most, any other ways?

[rsilvers]

Thanks for posting this. I was just about to ask specifically who this "we" is that rsilvers keeps referring to.

Yes, "we" is AAC. I do not have it in my sig as I did not want to use my sig area as a banner ad for AAC.

[rsilvers]

That is likely because they have not been told that matters. Off hand, how many bolts have you seen that have passed everything else make it onto the shelves for sale? I am not suggesting that we step back on the dimensions, but I am curious if cases of good bolts not being correct even exist.

We know that untested bolts will break, just google it, but I am unsure of any dimensional issues.

I can't follow what you are asking. What do you mean about "passing everything else?"

What do you mean about "good bolts not being correct?" If you are asking if there are bolts which have passed MPI but are dimensionally incorrect, of course they exist and are probably not rare.

Untested and tested bolts will break, after 5,000 or 10,000 rounds. You seem to be saying that only untested bolts break. All bolts break.

[rsilvers]

In my mind dimensional tolerances are more important then MPI, but I would have considered it a given that manufactures tested dimensions anyway. MPI, if done, should be on top of that.

Does anyone have any information that would suggest the "left side" of the chart (those who MPI test bolts) also skip dimensional testing? Wouldn't this be relatively simple to accomplish with laser gauges in the large scale and custom gauges/jigs in the small scale?

ETA: If QC budget is the restriction, maybe MPI testing can be considered a marketing expense and taken from that pot instead?

Gauging is the best way. Colt has a govt inspector on site who gauges certain things. Other companies? Don't assume they do. Maybe they do, but if you are going to demand MPI, then they have that many less dollars to do gauging and test firing. Everyone has a fixed QC budget and best not to waste it on tests which do not find as many bad parts.

But to answer directly, there is NO company which checks every part for all dimensions. That would take forever. The best companies do ISO-9000 style statistical sampling where they test so many parts and the number they test is based on a number of factors - such as batch size, and how likely the part is to be bad. Some companies test more than others. Expensive MPI tests take away from how much other testing one can do. I really am convinced that it is a marketing expense.