A word about CNC machines and quality

[garrettwc]

I see this come up all the time in threads - "With all the CNC machines, why can't we match the hand fitted perfection of the old 1911s?"

I recently spent a few days cross training in my company's Quality Assurance dept. I learned there that the function of QA/QC is not to inspect for defects, but to examine materials and processes in relation to stated goals and provide feedback for improvement. In other words, they don't catch defects before they go out the door. They identify broken processes and poor raw materials in an effort to keep the defects from happening in the first place.

What does that have to do with CNC machines and gunsmithing? Read on.

A CNC machine does not insure quality. It insures consistency. Good or bad, every part is just like the one that came before it. This is the stated goals portion of the paragraph above. Most 1911 manufacturers today are just parts assemblers. They don't build them from scratch the way Colt did way back when. So the stated goal is to be able to get parts from various subcontractors that can be drop-in assembled by semi-skilled and often times temporary labor.

As Tuner mentioned in the barrel fitting clinic. The engineers come up with a spec for the barrel lugs that's "close enough" that will allow for a drop-in fit. The stated goal isn't hand fitted perfection, but to build a bunch of guns quickly and at the lowest cost.

And this is why you will not get a custom gunsmith gun out of a CNC machine.

[emiddio]

so let me ask -- how tight can CNC machines build parts to spec ?

if the specs were tight enough -- and we were dealing with a "fresh" design that did not have to be "drop in" for a wide range of historical "specs" -- how good is CNC milling for precision fitting to one ( not a history of) specifications.

also -- regarding barrel making -- we hear about ballistics testing -- shows like CIS -- can barrels and breech faces be made close enough together that you can not identify a barrel a bullet came from -- or case ?

and also -- how many rounds have to be fired in a gun/barrel before the "matching" ability will be blurred enough (CNC made) to make identification non-precise ?

[Tom in Ohio]

I'm an evidence tech. for my department. I asked the same questions at the firearms lab during training. A cartridge can be matched to a particular weapon in several ways, including primer marks, extractor marks, ejector marks, and breech face marks. These marks are microscopic and are basically the grain or texture of the metal. I think that it would be impossible to create a smooth enough surface on the metals use in gun-making that would not leave a unique microscopic signature. I was told that a bullet can still be matched to a particular barrel after 20,000 rounds. The biggest challenge in matching bullets, though, is that corrosion of the fired bullet can happen quickly enough to make matching impossible after a certain amount of time.

[emiddio]

I'm an evidence tech. for my department. I asked the same questions at the firearms lab during training. A cartridge can be matched to a particular weapon in several ways, including primer marks, extractor marks, ejector marks, and breech face marks. These marks are microscopic and are basically the grain or texture of the metal. I think that it would be impossible to create a smooth enough surface on the metals use in gun-making that would not leave a unique microscopic signature. I was told that a bullet can still be matched to a particular barrel after 20,000 rounds. The biggest challenge in matching bullets, though, is that corrosion of the fired bullet can happen quickly enough to make matching impossible after a certain amount of time.

extractors, ejectors can easially be replaced. in 1911 guns barrels can be easially replaced. only the breech face is not replaceable -- until a different slide is used.

editied :

extractors, ejectors belong to the slide in 1911 guns -- and refer to case matching -- not bullet matching.

20000 rnds for barrel matching --ok

barrels can be easially exchanged/replaced.

extractors, ejectors -- are pretty easy to replace also -- how many rounds does it take for these to loose their idenity ?

breech face wear seems like it would take quite a while -- is primer/breech face only matching used ?

[Tom in Ohio]

I'm not trained to do the actual comparisons, so what I can tell you is what the techs have told me. What they've said is that because most cartridge cases (and projectiles) are made of relatively soft metals (brass, copper, lead, etc.) and the firearms are very hard metals, it takes thousands and thousands of rounds to change anything enough to prevent matching. Obviously parts can be switched out, but those original parts can still be later matched to a case or bullet. For prosecution purposes, it wouldn't really matter if a suspect was found to have a complete weapon that matched a particular casing or just a single part that matched. It might even help to show culpable mental state - tampering with evidence to conceal a crime. The only way to really alter the parts enough would be to destroy them or wear them enough either through firing thousands and thousands of rounds or, as most of us can attest to, with a dremel.

[garrettwc]

if the specs were tight enough -- and we were dealing with a "fresh" design that did not have to be "drop in" for a wide range of historical "specs" -- how good is CNC milling for precision fitting to one ( not a history of) specifications.

Theoretically yes. The likelihood of success would be greater if they could be built completely in house and the designer could personally check the end product against the original design specs. I'm not sure but I don't think that's happened since JMB walked into Colt.

The reason for the wider tolerances is to accomodate the variances in setup/equipment from subcontractor A to subcontractor B. Even with a detailed spec there will be minor variances from one machine to another, due to the human element of operator setup.

[Mtrclass]

Assuming equal quality of a manual mill and a CNC mill, both would be able to maintain the same tolerances. And depending on the quality, that tolerance can be held very close, in the order of .00010".
However there are a lot of other considerations involved in holding tolerances including but not limited to fixturing, tool design and wear, operator skills.
What a CNC does is make it possible to eliminate a lot of the mistakes or misques input into the process by the operator, and, usually, it allows you to increase production speed while maintaining your tolerance levels.
Assuming a good quality CNC fixturing becomes the most important part of the process and the operators ability to place the part in the fixture the same way each time. A good CNC controller will allow for cutter wear thru cutter comp, and a good program will handle cutter flex. But for a truely nice weapon hand fitting is still needed. Even, like in our case where we make all of our major parts, you still have some tolerance stacking which can only be properly dealt with by hand fitting.

[1911Tuner]

Posts concerning CNC machining are spot on. It doesn't guarantee close tolerances, nor is it meant to. Turning out thousands of identical parts...often with close tolerances...is the object of the exercise. Even with careful control and skilled, dedicated operators, there is still no such thing as a perfect dimension. There is still an allowable tolerance, and the CNC machining or turning center will never be able to match a skilled man taking his time with a file and a stone.

The 1911 pistol was redesigned at one point so that it could be assembled from selective-fit, drop in parts and groups of parts...with the aid of standard gauges in order to eliminate a lot of the hand-fitting that the old ones required. The reason was simple. Interchangeability of parts that come from several different contractors...and it's been like that ever since.

[warmrain]

... you still have some tolerance stacking which can only be properly dealt with by hand fitting.
Now this helps it start to make sense!

[emiddio]

... But for a truely nice weapon hand fitting is still needed. Even, like in our case where we make all of our major parts, you still have some tolerance stacking which can only be properly dealt with by hand fitting.

i think i understand tolerance stacking -- the same thing occurs in computer mathematical calculations -- its has to do with how errors are propogated and accumulate.

in any design -- and in the 1911 in particular -- there is a maximum level of "tolerance stacking" that can occur. Any body know how many levels is the MAX in the building of a 1911 ? It seems it should be related to the maximum level of "part inserted into part" or "part dependant upon part" that occurs. There is a Maximum componding -- what is it ?

i understand Tuner's discussion of "standard guages". any place on this site that discusses what is a complete set of "standard guages" and maybe their proper use ?

If we drew the analogy to the manufacturing of a modern day assembly line production of a gasoline or diesel engine intended for motor vehichles -- is Tuner's concept of "standard guages" used in that industry -- or are all parts considered to be "within spec" and "good enough" if they conform to spec ?

Most parts i presume are CNC machined.

[1911Tuner]

Emiddio's questions:

any place on this site that discusses what is a complete set of "standard guages" and maybe their proper use ?

Not that I know of. Aside from industry standard headpsace gauges and possibly aftermarket slide gauge sets, the only other people who might still have complete gauge sets are the government, and maybe Colt. Goin' on what I've seen in the past 25 years, I wouldn't even bet on Colt. If you can obtain the original blueprints, you can have a toolmaker hand-make'em...but you're probably talkin' about 10,000 bucks for a complete set.
**************************

If we drew the analogy to the manufacturing of a modern day assembly line production of a gasoline or diesel engine intended for motor vehichles -- is Tuner's concept of "standard guages" used in that industry -- or are all parts considered to be "within spec" and "good enough" if they conform to spec.

As long as the gauge and part match, it's supposed to be within spec, and therefore serviceable. Like a headspace set consisting of two gauges.
One is set to minimum static dimension, (.898 inch) and the other to maximum
(.920 inch) as set by the SAAMI. Referred to as simply "GO" and "NO-GO" In practice, "All pistols must go into battery on the GO gauge without resistance, and none may go to battery on the NO-GO. If a pistol should go into full battery on the NO-GO, even with some resistance, the weapon is to be rejected as unfit for service and returned to the unit armorer for repair."

Understand that .920 as a maximum is absolute, and many guns that almost accept the NO-GO are still subject to failure and to closer inspection. These are often retested with a field gauge, which...without going into small detail... essentially tells the armorer roughly how close the headspace is to excessive and unserviceable. More field-spec gauges are used on rifles than on pistols. For sidearms, the standard GO/NO-GO test is usually all that's done. I have a .910 and a .912 gauge set in addition to the standard GO and NO-GO set, as well as two that are below minimum, for setting chamber depth
with a semi-finished chamber. I made the gauge set for my own use.


So...most standard gauge sets have a GO and a NO-GO matched set for each check, with which the assembler determines the selective fit. If the part fails, another part is selected until a suitable part is found, or until no parts can be found to work...and the gun or part is rejected for repair or the scrap pile. Again, I have my doubts as to whether the present manufacturing and assembly lines even use such gauges, or if the ones that are in use are correctly dimensioned. If they were, we wouldn't see the frequency of functional issues that we see today.

[warmrain]

It seem though as if some manufacturers (e.g. Kimber?) CNC machine to the tight end of the tolerances so that the "fitting" is really left up to the purchaser and their expended ammo during "break-in"?

[1911Tuner]

It seem though as if some manufacturers (e.g. Kimber?) CNC machine to the tight end of the tolerances so that the "fitting" is really left up to the purchaser and their expended ammo during "break-in"?

Well...Not really. At least not where it really matters. If you're fortunate enough to be able to handle an old, pre-WW2 commercial pistol that's also in close to pristine condition, you'll be surprised to find that it's pretty tight.
In fact, if you use just a little oil in the rails, you'll have to pull pretty hard in all directions to be able to feel any play. Same for the barrel fit...and those pistols ran like a raped ape from the git-go. The early military pistols were built the same way, to the same specs. Just tough to find one these days that old that hasn't been through the wringer.

[SatCong]

I see this come up all the time in threads - "With all the CNC machines, why can't we match the hand fitted perfection of the old 1911s?"

I recently spent a few days cross training in my company's Quality Assurance dept. I learned there that the function of QA/QC is not to inspect for defects, but to examine materials and processes in relation to stated goals and provide feedback for improvement. In other words, they don't catch defects before they go out the door. They identify broken processes and poor raw materials in an effort to keep the defects from happening in the first place.

What does that have to do with CNC machines and gunsmithing? Read on.

A CNC machine does not insure quality. It insures consistency. Good or bad, every part is just like the one that came before it. This is the stated goals portion of the paragraph above. Most 1911 manufacturers today are just parts assemblers. They don't build them from scratch the way Colt did way back when. So the stated goal is to be able to get parts from various subcontractors that can be drop-in assembled by semi-skilled and often times temporary labor.

As Tuner mentioned in the barrel fitting clinic. The engineers come up with a spec for the barrel lugs that's "close enough" that will allow for a drop-in fit. The stated goal isn't hand fitted perfection, but to build a bunch of guns quickly and at the lowest cost.

And this is why you will not get a custom gunsmith gun out of a CNC machine.The are only good as the Company,Programers and Operater's who control the machine. I know I ran them for Cat, 30 years. Some of my jobs I held +or- 2 tens, that with a good machine.

[1911Tuner]

The role of the CNC turning or machining center is, by design and intent, mass production and consistency from first part to last....not precision fitting.

[bearandoldman]

The role of the CNC turning or machining center is, by design and intent, mass production and consistency from first part to last....not precision fitting.

The words of your faithful dog are very true, every morning Bear stick his nose in my ear, tell me to get a move on old man and says the same thing.

[bearandoldman]

The role of the CNC turning or machining center is, by design and intent, mass production and consistency from first part to last....not precision fitting.

The words of your faithful dog are very true, every morning Bear sticks his nose in my ear, tells me "get a move on old man" and says the same thing to me.
Johnny the more we have to do with dogs the less we need people. You cna not have a friend more true through good and bad days as a loving dog. ;) ;)

[brickeyee]

There are a number of places in a modern engine that are still fitted up by hand at assembly. Probably the most common are main bearing shells. An assortment of shells of different thicknesses are used to ensure correct assembly. Many engine blocks are marked with punches near each bearing to indicate the size used for the shells at initial assembly.
The gauge sets mentioned are go and no go for almost every part of the 1911. They are a large set and seldom seen outside a manufacturers QC department. Many were owned by the government and used only by government inspectors.
A 1911 with all parts within spec still requires hand fitting to ensure safe and reliable operation. The chief item is still the thumb safety engagement with the sear. All the hole and parts tolerances combine and Mr. Murphy declares they are invariably in the wrong way.
The design is still very good since fitting of a very few critical items results in a functional and safe weapon.